swe-latn-10mb-after-ppt-shuff-dyck-100mb-ckpt500_seed3407/checkpoint-21000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.764293215710985,
  "eval_steps": 3000,
  "global_step": 21000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 4.897539186477661,
      "epoch": 0.0004200798151648813,
      "grad_norm": 14.0,
      "learning_rate": 2e-06,
      "loss": 14.0815,
      "mean_token_accuracy": 0.0,
      "num_tokens": 8348.0,
      "step": 5
    },
    {
      "entropy": 4.873638677597046,
      "epoch": 0.0008401596303297626,
      "grad_norm": 13.625,
      "learning_rate": 4.5e-06,
      "loss": 14.0885,
      "mean_token_accuracy": 0.0,
      "num_tokens": 17465.0,
      "step": 10
    },
    {
      "entropy": 4.912403154373169,
      "epoch": 0.001260239445494644,
      "grad_norm": 15.4375,
      "learning_rate": 7e-06,
      "loss": 14.0641,
      "mean_token_accuracy": 0.0,
      "num_tokens": 26627.0,
      "step": 15
    },
    {
      "entropy": 4.9623668670654295,
      "epoch": 0.0016803192606595252,
      "grad_norm": 17.875,
      "learning_rate": 9.5e-06,
      "loss": 13.8986,
      "mean_token_accuracy": 0.0,
      "num_tokens": 36069.0,
      "step": 20
    },
    {
      "entropy": 5.122641944885254,
      "epoch": 0.002100399075824407,
      "grad_norm": 21.0,
      "learning_rate": 1.2e-05,
      "loss": 13.4962,
      "mean_token_accuracy": 0.00036849312018603084,
      "num_tokens": 44967.0,
      "step": 25
    },
    {
      "entropy": 5.440350961685181,
      "epoch": 0.002520478890989288,
      "grad_norm": 25.875,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 13.0078,
      "mean_token_accuracy": 0.00010752688394859433,
      "num_tokens": 55132.0,
      "step": 30
    },
    {
      "entropy": 7.052726125717163,
      "epoch": 0.0029405587061541692,
      "grad_norm": 21.25,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 11.9064,
      "mean_token_accuracy": 0.0,
      "num_tokens": 65141.0,
      "step": 35
    },
    {
      "entropy": 10.193820476531982,
      "epoch": 0.0033606385213190504,
      "grad_norm": 3.171875,
      "learning_rate": 1.95e-05,
      "loss": 10.8374,
      "mean_token_accuracy": 0.0,
      "num_tokens": 74007.0,
      "step": 40
    },
    {
      "entropy": 10.727454471588135,
      "epoch": 0.003780718336483932,
      "grad_norm": 2.984375,
      "learning_rate": 2.2e-05,
      "loss": 10.6658,
      "mean_token_accuracy": 0.004875225311843678,
      "num_tokens": 83736.0,
      "step": 45
    },
    {
      "entropy": 10.743746089935303,
      "epoch": 0.004200798151648814,
      "grad_norm": 2.5625,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 10.4916,
      "mean_token_accuracy": 0.01765382578596473,
      "num_tokens": 92525.0,
      "step": 50
    },
    {
      "entropy": 10.714706707000733,
      "epoch": 0.004620877966813695,
      "grad_norm": 2.234375,
      "learning_rate": 2.7e-05,
      "loss": 10.3094,
      "mean_token_accuracy": 0.01746684079989791,
      "num_tokens": 102015.0,
      "step": 55
    },
    {
      "entropy": 10.654456615447998,
      "epoch": 0.005040957781978576,
      "grad_norm": 1.7578125,
      "learning_rate": 2.95e-05,
      "loss": 10.1455,
      "mean_token_accuracy": 0.019217501021921636,
      "num_tokens": 110887.0,
      "step": 60
    },
    {
      "entropy": 10.696210289001465,
      "epoch": 0.005461037597143457,
      "grad_norm": 1.9140625,
      "learning_rate": 3.2e-05,
      "loss": 10.0256,
      "mean_token_accuracy": 0.018034272640943528,
      "num_tokens": 120442.0,
      "step": 65
    },
    {
      "entropy": 10.72099323272705,
      "epoch": 0.0058811174123083385,
      "grad_norm": 1.90625,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.8848,
      "mean_token_accuracy": 0.029999475181102752,
      "num_tokens": 129297.0,
      "step": 70
    },
    {
      "entropy": 10.713169956207276,
      "epoch": 0.00630119722747322,
      "grad_norm": 1.7890625,
      "learning_rate": 3.7e-05,
      "loss": 9.8186,
      "mean_token_accuracy": 0.03141374848783016,
      "num_tokens": 138305.0,
      "step": 75
    },
    {
      "entropy": 10.702960300445557,
      "epoch": 0.006721277042638101,
      "grad_norm": 1.9140625,
      "learning_rate": 3.95e-05,
      "loss": 9.7553,
      "mean_token_accuracy": 0.023225654661655427,
      "num_tokens": 147640.0,
      "step": 80
    },
    {
      "entropy": 10.704542255401611,
      "epoch": 0.007141356857802983,
      "grad_norm": 1.78125,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.6696,
      "mean_token_accuracy": 0.02769038388505578,
      "num_tokens": 157633.0,
      "step": 85
    },
    {
      "entropy": 10.70653486251831,
      "epoch": 0.007561436672967864,
      "grad_norm": 1.59375,
      "learning_rate": 4.45e-05,
      "loss": 9.607,
      "mean_token_accuracy": 0.04190691113471985,
      "num_tokens": 167984.0,
      "step": 90
    },
    {
      "entropy": 10.694128608703613,
      "epoch": 0.007981516488132745,
      "grad_norm": 1.65625,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.4789,
      "mean_token_accuracy": 0.05244061499834061,
      "num_tokens": 176984.0,
      "step": 95
    },
    {
      "entropy": 10.663659858703614,
      "epoch": 0.008401596303297627,
      "grad_norm": 1.7890625,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.3453,
      "mean_token_accuracy": 0.043341046944260594,
      "num_tokens": 185931.0,
      "step": 100
    },
    {
      "entropy": 10.635926151275635,
      "epoch": 0.008821676118462508,
      "grad_norm": 1.7734375,
      "learning_rate": 5.2e-05,
      "loss": 9.2908,
      "mean_token_accuracy": 0.04105062447488308,
      "num_tokens": 195065.0,
      "step": 105
    },
    {
      "entropy": 10.617516040802002,
      "epoch": 0.00924175593362739,
      "grad_norm": 1.5859375,
      "learning_rate": 5.45e-05,
      "loss": 9.1462,
      "mean_token_accuracy": 0.05056820251047611,
      "num_tokens": 203687.0,
      "step": 110
    },
    {
      "entropy": 10.578640842437744,
      "epoch": 0.00966183574879227,
      "grad_norm": 1.5390625,
      "learning_rate": 5.7e-05,
      "loss": 9.0627,
      "mean_token_accuracy": 0.05324659980833531,
      "num_tokens": 212847.0,
      "step": 115
    },
    {
      "entropy": 10.534226608276366,
      "epoch": 0.010081915563957152,
      "grad_norm": 1.7265625,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.9326,
      "mean_token_accuracy": 0.05340988412499428,
      "num_tokens": 222593.0,
      "step": 120
    },
    {
      "entropy": 10.47627763748169,
      "epoch": 0.010501995379122032,
      "grad_norm": 1.59375,
      "learning_rate": 6.2e-05,
      "loss": 8.7909,
      "mean_token_accuracy": 0.05306763276457786,
      "num_tokens": 231174.0,
      "step": 125
    },
    {
      "entropy": 10.402797031402589,
      "epoch": 0.010922075194286915,
      "grad_norm": 1.515625,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.6666,
      "mean_token_accuracy": 0.0565425343811512,
      "num_tokens": 239833.0,
      "step": 130
    },
    {
      "entropy": 10.321593379974365,
      "epoch": 0.011342155009451797,
      "grad_norm": 1.5078125,
      "learning_rate": 6.7e-05,
      "loss": 8.6377,
      "mean_token_accuracy": 0.05593148916959763,
      "num_tokens": 248794.0,
      "step": 135
    },
    {
      "entropy": 10.22030849456787,
      "epoch": 0.011762234824616677,
      "grad_norm": 1.3515625,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.5488,
      "mean_token_accuracy": 0.05581129565834999,
      "num_tokens": 257123.0,
      "step": 140
    },
    {
      "entropy": 10.1541109085083,
      "epoch": 0.012182314639781559,
      "grad_norm": 1.359375,
      "learning_rate": 7.2e-05,
      "loss": 8.309,
      "mean_token_accuracy": 0.05627111345529556,
      "num_tokens": 266088.0,
      "step": 145
    },
    {
      "entropy": 10.05057725906372,
      "epoch": 0.01260239445494644,
      "grad_norm": 1.15625,
      "learning_rate": 7.45e-05,
      "loss": 8.3916,
      "mean_token_accuracy": 0.05131003558635712,
      "num_tokens": 276074.0,
      "step": 150
    },
    {
      "entropy": 9.931949615478516,
      "epoch": 0.013022474270111321,
      "grad_norm": 1.1484375,
      "learning_rate": 7.7e-05,
      "loss": 8.2337,
      "mean_token_accuracy": 0.05355307124555111,
      "num_tokens": 285280.0,
      "step": 155
    },
    {
      "entropy": 9.78201379776001,
      "epoch": 0.013442554085276202,
      "grad_norm": 1.125,
      "learning_rate": 7.950000000000001e-05,
      "loss": 8.1889,
      "mean_token_accuracy": 0.05086107552051544,
      "num_tokens": 296115.0,
      "step": 160
    },
    {
      "entropy": 9.592757034301759,
      "epoch": 0.013862633900441084,
      "grad_norm": 0.9296875,
      "learning_rate": 8.2e-05,
      "loss": 8.0092,
      "mean_token_accuracy": 0.05353938601911068,
      "num_tokens": 305483.0,
      "step": 165
    },
    {
      "entropy": 9.374695110321046,
      "epoch": 0.014282713715605966,
      "grad_norm": 0.875,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.9652,
      "mean_token_accuracy": 0.05161282308399677,
      "num_tokens": 314000.0,
      "step": 170
    },
    {
      "entropy": 9.172936820983887,
      "epoch": 0.014702793530770846,
      "grad_norm": 0.93359375,
      "learning_rate": 8.7e-05,
      "loss": 7.9024,
      "mean_token_accuracy": 0.0552643246948719,
      "num_tokens": 323667.0,
      "step": 175
    },
    {
      "entropy": 8.918990325927734,
      "epoch": 0.015122873345935728,
      "grad_norm": 0.76953125,
      "learning_rate": 8.95e-05,
      "loss": 7.858,
      "mean_token_accuracy": 0.052525343000888826,
      "num_tokens": 332695.0,
      "step": 180
    },
    {
      "entropy": 8.721237659454346,
      "epoch": 0.015542953161100609,
      "grad_norm": 0.8046875,
      "learning_rate": 9.2e-05,
      "loss": 7.7042,
      "mean_token_accuracy": 0.050673872232437134,
      "num_tokens": 342428.0,
      "step": 185
    },
    {
      "entropy": 8.52247200012207,
      "epoch": 0.01596303297626549,
      "grad_norm": 0.61328125,
      "learning_rate": 9.45e-05,
      "loss": 7.7799,
      "mean_token_accuracy": 0.051399941369891164,
      "num_tokens": 353587.0,
      "step": 190
    },
    {
      "entropy": 8.398729610443116,
      "epoch": 0.01638311279143037,
      "grad_norm": 0.77734375,
      "learning_rate": 9.7e-05,
      "loss": 7.7173,
      "mean_token_accuracy": 0.05443600341677666,
      "num_tokens": 362997.0,
      "step": 195
    },
    {
      "entropy": 8.274804973602295,
      "epoch": 0.016803192606595255,
      "grad_norm": 0.71484375,
      "learning_rate": 9.95e-05,
      "loss": 7.7333,
      "mean_token_accuracy": 0.06009506583213806,
      "num_tokens": 372346.0,
      "step": 200
    },
    {
      "entropy": 8.316196537017822,
      "epoch": 0.017223272421760135,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000102,
      "loss": 7.621,
      "mean_token_accuracy": 0.054996570199728013,
      "num_tokens": 381575.0,
      "step": 205
    },
    {
      "entropy": 8.224405288696289,
      "epoch": 0.017643352236925015,
      "grad_norm": 0.71484375,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.6806,
      "mean_token_accuracy": 0.05832334831357002,
      "num_tokens": 390706.0,
      "step": 210
    },
    {
      "entropy": 8.178961277008057,
      "epoch": 0.018063432052089896,
      "grad_norm": 0.77734375,
      "learning_rate": 0.000107,
      "loss": 7.6844,
      "mean_token_accuracy": 0.056866029649972914,
      "num_tokens": 400000.0,
      "step": 215
    },
    {
      "entropy": 8.118294048309327,
      "epoch": 0.01848351186725478,
      "grad_norm": 0.765625,
      "learning_rate": 0.0001095,
      "loss": 7.6249,
      "mean_token_accuracy": 0.06287443153560161,
      "num_tokens": 409447.0,
      "step": 220
    },
    {
      "entropy": 8.128210067749023,
      "epoch": 0.01890359168241966,
      "grad_norm": 0.84375,
      "learning_rate": 0.000112,
      "loss": 7.6104,
      "mean_token_accuracy": 0.06484438478946686,
      "num_tokens": 418417.0,
      "step": 225
    },
    {
      "entropy": 8.050536060333252,
      "epoch": 0.01932367149758454,
      "grad_norm": 0.90625,
      "learning_rate": 0.0001145,
      "loss": 7.5732,
      "mean_token_accuracy": 0.06463442519307136,
      "num_tokens": 427619.0,
      "step": 230
    },
    {
      "entropy": 8.079150295257568,
      "epoch": 0.019743751312749424,
      "grad_norm": 0.72265625,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.616,
      "mean_token_accuracy": 0.061473137512803075,
      "num_tokens": 437931.0,
      "step": 235
    },
    {
      "entropy": 8.002902173995972,
      "epoch": 0.020163831127914304,
      "grad_norm": 0.73828125,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.6568,
      "mean_token_accuracy": 0.06506373398005963,
      "num_tokens": 447595.0,
      "step": 240
    },
    {
      "entropy": 8.058529567718505,
      "epoch": 0.020583910943079185,
      "grad_norm": 0.796875,
      "learning_rate": 0.000122,
      "loss": 7.5257,
      "mean_token_accuracy": 0.058850279077887535,
      "num_tokens": 457062.0,
      "step": 245
    },
    {
      "entropy": 8.02257719039917,
      "epoch": 0.021003990758244065,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0001245,
      "loss": 7.624,
      "mean_token_accuracy": 0.06412565596401691,
      "num_tokens": 466191.0,
      "step": 250
    },
    {
      "entropy": 8.05047812461853,
      "epoch": 0.02142407057340895,
      "grad_norm": 1.0625,
      "learning_rate": 0.000127,
      "loss": 7.5647,
      "mean_token_accuracy": 0.06097288727760315,
      "num_tokens": 475693.0,
      "step": 255
    },
    {
      "entropy": 8.019621276855469,
      "epoch": 0.02184415038857383,
      "grad_norm": 0.7109375,
      "learning_rate": 0.0001295,
      "loss": 7.589,
      "mean_token_accuracy": 0.06407038681209087,
      "num_tokens": 485173.0,
      "step": 260
    },
    {
      "entropy": 7.99587049484253,
      "epoch": 0.02226423020373871,
      "grad_norm": 0.68359375,
      "learning_rate": 0.000132,
      "loss": 7.5069,
      "mean_token_accuracy": 0.06533381156623363,
      "num_tokens": 493985.0,
      "step": 265
    },
    {
      "entropy": 7.972505903244018,
      "epoch": 0.022684310018903593,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.519,
      "mean_token_accuracy": 0.0665605653077364,
      "num_tokens": 502837.0,
      "step": 270
    },
    {
      "entropy": 7.982478952407837,
      "epoch": 0.023104389834068473,
      "grad_norm": 0.78515625,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.4828,
      "mean_token_accuracy": 0.06852602958679199,
      "num_tokens": 511503.0,
      "step": 275
    },
    {
      "entropy": 7.961443471908569,
      "epoch": 0.023524469649233354,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0001395,
      "loss": 7.6587,
      "mean_token_accuracy": 0.0616368904709816,
      "num_tokens": 521499.0,
      "step": 280
    },
    {
      "entropy": 7.966269254684448,
      "epoch": 0.023944549464398234,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.426,
      "mean_token_accuracy": 0.06728804372251034,
      "num_tokens": 530067.0,
      "step": 285
    },
    {
      "entropy": 7.91850700378418,
      "epoch": 0.024364629279563118,
      "grad_norm": 0.796875,
      "learning_rate": 0.0001445,
      "loss": 7.4148,
      "mean_token_accuracy": 0.07034274153411388,
      "num_tokens": 538559.0,
      "step": 290
    },
    {
      "entropy": 7.979486131668091,
      "epoch": 0.024784709094728,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000147,
      "loss": 7.6056,
      "mean_token_accuracy": 0.06503808572888374,
      "num_tokens": 547288.0,
      "step": 295
    },
    {
      "entropy": 7.885135698318481,
      "epoch": 0.02520478890989288,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0001495,
      "loss": 7.4924,
      "mean_token_accuracy": 0.06286945752799511,
      "num_tokens": 557269.0,
      "step": 300
    },
    {
      "entropy": 8.007584619522095,
      "epoch": 0.025624868725057762,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000152,
      "loss": 7.5193,
      "mean_token_accuracy": 0.06448328271508216,
      "num_tokens": 567280.0,
      "step": 305
    },
    {
      "entropy": 7.853135585784912,
      "epoch": 0.026044948540222643,
      "grad_norm": 0.71875,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.382,
      "mean_token_accuracy": 0.06435618139803409,
      "num_tokens": 576609.0,
      "step": 310
    },
    {
      "entropy": 7.831946516036988,
      "epoch": 0.026465028355387523,
      "grad_norm": 0.79296875,
      "learning_rate": 0.000157,
      "loss": 7.267,
      "mean_token_accuracy": 0.07741407230496407,
      "num_tokens": 586053.0,
      "step": 315
    },
    {
      "entropy": 7.793692779541016,
      "epoch": 0.026885108170552403,
      "grad_norm": 0.921875,
      "learning_rate": 0.0001595,
      "loss": 7.4583,
      "mean_token_accuracy": 0.0726725548505783,
      "num_tokens": 594649.0,
      "step": 320
    },
    {
      "entropy": 7.9273755073547365,
      "epoch": 0.027305187985717287,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000162,
      "loss": 7.4408,
      "mean_token_accuracy": 0.06625295169651509,
      "num_tokens": 603445.0,
      "step": 325
    },
    {
      "entropy": 7.88362865447998,
      "epoch": 0.027725267800882167,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.4937,
      "mean_token_accuracy": 0.06676559969782829,
      "num_tokens": 613611.0,
      "step": 330
    },
    {
      "entropy": 7.917990875244141,
      "epoch": 0.028145347616047048,
      "grad_norm": 1.1875,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.6535,
      "mean_token_accuracy": 0.06991421952843666,
      "num_tokens": 623024.0,
      "step": 335
    },
    {
      "entropy": 7.84550929069519,
      "epoch": 0.02856542743121193,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.37,
      "mean_token_accuracy": 0.07066045552492142,
      "num_tokens": 631624.0,
      "step": 340
    },
    {
      "entropy": 7.743204069137573,
      "epoch": 0.028985507246376812,
      "grad_norm": 1.0,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.3544,
      "mean_token_accuracy": 0.07467443272471427,
      "num_tokens": 640473.0,
      "step": 345
    },
    {
      "entropy": 7.89313235282898,
      "epoch": 0.029405587061541692,
      "grad_norm": 1.546875,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.483,
      "mean_token_accuracy": 0.07056561335921288,
      "num_tokens": 649692.0,
      "step": 350
    },
    {
      "entropy": 7.855760765075684,
      "epoch": 0.029825666876706573,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000177,
      "loss": 7.4521,
      "mean_token_accuracy": 0.07088366113603115,
      "num_tokens": 658236.0,
      "step": 355
    },
    {
      "entropy": 7.798476648330689,
      "epoch": 0.030245746691871456,
      "grad_norm": 1.265625,
      "learning_rate": 0.0001795,
      "loss": 7.2377,
      "mean_token_accuracy": 0.07707448154687882,
      "num_tokens": 667175.0,
      "step": 360
    },
    {
      "entropy": 7.8225892066955565,
      "epoch": 0.030665826507036337,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000182,
      "loss": 7.5145,
      "mean_token_accuracy": 0.07331682369112968,
      "num_tokens": 676456.0,
      "step": 365
    },
    {
      "entropy": 7.92411093711853,
      "epoch": 0.031085906322201217,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0001845,
      "loss": 7.4959,
      "mean_token_accuracy": 0.07029573023319244,
      "num_tokens": 686881.0,
      "step": 370
    },
    {
      "entropy": 7.741179084777832,
      "epoch": 0.0315059861373661,
      "grad_norm": 1.046875,
      "learning_rate": 0.000187,
      "loss": 7.2754,
      "mean_token_accuracy": 0.0755814753472805,
      "num_tokens": 696045.0,
      "step": 375
    },
    {
      "entropy": 7.724820518493653,
      "epoch": 0.03192606595253098,
      "grad_norm": 1.15625,
      "learning_rate": 0.0001895,
      "loss": 7.2387,
      "mean_token_accuracy": 0.08382209464907646,
      "num_tokens": 704729.0,
      "step": 380
    },
    {
      "entropy": 7.6020232200622555,
      "epoch": 0.032346145767695865,
      "grad_norm": 0.73046875,
      "learning_rate": 0.000192,
      "loss": 7.2979,
      "mean_token_accuracy": 0.07353318221867085,
      "num_tokens": 714331.0,
      "step": 385
    },
    {
      "entropy": 7.746062755584717,
      "epoch": 0.03276622558286074,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0001945,
      "loss": 7.26,
      "mean_token_accuracy": 0.07669057212769985,
      "num_tokens": 722788.0,
      "step": 390
    },
    {
      "entropy": 7.785710859298706,
      "epoch": 0.033186305398025626,
      "grad_norm": 1.25,
      "learning_rate": 0.00019700000000000002,
      "loss": 7.3239,
      "mean_token_accuracy": 0.07846143767237664,
      "num_tokens": 731417.0,
      "step": 395
    },
    {
      "entropy": 7.662443017959594,
      "epoch": 0.03360638521319051,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00019950000000000002,
      "loss": 7.2961,
      "mean_token_accuracy": 0.07676250785589218,
      "num_tokens": 741034.0,
      "step": 400
    },
    {
      "entropy": 7.704835367202759,
      "epoch": 0.034026465028355386,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000202,
      "loss": 7.2982,
      "mean_token_accuracy": 0.07422473579645157,
      "num_tokens": 749596.0,
      "step": 405
    },
    {
      "entropy": 7.664615345001221,
      "epoch": 0.03444654484352027,
      "grad_norm": 0.953125,
      "learning_rate": 0.00020449999999999998,
      "loss": 7.2586,
      "mean_token_accuracy": 0.07927079945802688,
      "num_tokens": 758931.0,
      "step": 410
    },
    {
      "entropy": 7.600234937667847,
      "epoch": 0.03486662465868515,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000207,
      "loss": 7.1556,
      "mean_token_accuracy": 0.08096722960472107,
      "num_tokens": 767534.0,
      "step": 415
    },
    {
      "entropy": 7.68730993270874,
      "epoch": 0.03528670447385003,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002095,
      "loss": 7.2278,
      "mean_token_accuracy": 0.07768219709396362,
      "num_tokens": 776456.0,
      "step": 420
    },
    {
      "entropy": 7.634504795074463,
      "epoch": 0.035706784289014915,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000212,
      "loss": 7.2673,
      "mean_token_accuracy": 0.07659058347344398,
      "num_tokens": 786172.0,
      "step": 425
    },
    {
      "entropy": 7.6300970077514645,
      "epoch": 0.03612686410417979,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0002145,
      "loss": 7.1534,
      "mean_token_accuracy": 0.08167968951165676,
      "num_tokens": 795081.0,
      "step": 430
    },
    {
      "entropy": 7.709903621673584,
      "epoch": 0.036546943919344675,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00021700000000000002,
      "loss": 7.2139,
      "mean_token_accuracy": 0.07590967863798141,
      "num_tokens": 804259.0,
      "step": 435
    },
    {
      "entropy": 7.642217683792114,
      "epoch": 0.03696702373450956,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002195,
      "loss": 7.2158,
      "mean_token_accuracy": 0.07940593808889389,
      "num_tokens": 813463.0,
      "step": 440
    },
    {
      "entropy": 7.5950026512146,
      "epoch": 0.037387103549674436,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000222,
      "loss": 7.1477,
      "mean_token_accuracy": 0.08674231953918934,
      "num_tokens": 823029.0,
      "step": 445
    },
    {
      "entropy": 7.576770305633545,
      "epoch": 0.03780718336483932,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002245,
      "loss": 7.2023,
      "mean_token_accuracy": 0.07635540775954723,
      "num_tokens": 832902.0,
      "step": 450
    },
    {
      "entropy": 7.584611654281616,
      "epoch": 0.0382272631800042,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00022700000000000002,
      "loss": 7.1523,
      "mean_token_accuracy": 0.07883834913372993,
      "num_tokens": 842162.0,
      "step": 455
    },
    {
      "entropy": 7.626716089248657,
      "epoch": 0.03864734299516908,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00022950000000000002,
      "loss": 7.1831,
      "mean_token_accuracy": 0.08037797883152961,
      "num_tokens": 852328.0,
      "step": 460
    },
    {
      "entropy": 7.5449056148529055,
      "epoch": 0.039067422810333964,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00023200000000000003,
      "loss": 7.139,
      "mean_token_accuracy": 0.08323637843132019,
      "num_tokens": 860929.0,
      "step": 465
    },
    {
      "entropy": 7.549172639846802,
      "epoch": 0.03948750262549885,
      "grad_norm": 1.15625,
      "learning_rate": 0.00023449999999999998,
      "loss": 7.222,
      "mean_token_accuracy": 0.07719254568219185,
      "num_tokens": 869144.0,
      "step": 470
    },
    {
      "entropy": 7.640847682952881,
      "epoch": 0.039907582440663725,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000237,
      "loss": 7.1591,
      "mean_token_accuracy": 0.08463784530758858,
      "num_tokens": 877447.0,
      "step": 475
    },
    {
      "entropy": 7.579086351394653,
      "epoch": 0.04032766225582861,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002395,
      "loss": 7.1205,
      "mean_token_accuracy": 0.08219385743141175,
      "num_tokens": 887020.0,
      "step": 480
    },
    {
      "entropy": 7.4508629322052,
      "epoch": 0.040747742070993485,
      "grad_norm": 1.140625,
      "learning_rate": 0.000242,
      "loss": 7.1602,
      "mean_token_accuracy": 0.08283634409308434,
      "num_tokens": 895937.0,
      "step": 485
    },
    {
      "entropy": 7.572605180740356,
      "epoch": 0.04116782188615837,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002445,
      "loss": 7.1651,
      "mean_token_accuracy": 0.07650997936725616,
      "num_tokens": 905446.0,
      "step": 490
    },
    {
      "entropy": 7.551622104644776,
      "epoch": 0.04158790170132325,
      "grad_norm": 1.15625,
      "learning_rate": 0.000247,
      "loss": 7.0659,
      "mean_token_accuracy": 0.0878485307097435,
      "num_tokens": 914547.0,
      "step": 495
    },
    {
      "entropy": 7.495913028717041,
      "epoch": 0.04200798151648813,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002495,
      "loss": 7.0595,
      "mean_token_accuracy": 0.08502571135759354,
      "num_tokens": 922900.0,
      "step": 500
    },
    {
      "entropy": 7.497592973709106,
      "epoch": 0.042428061331653014,
      "grad_norm": 1.15625,
      "learning_rate": 0.000252,
      "loss": 7.0901,
      "mean_token_accuracy": 0.08809811025857925,
      "num_tokens": 930876.0,
      "step": 505
    },
    {
      "entropy": 7.464612245559692,
      "epoch": 0.0428481411468179,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002545,
      "loss": 7.1198,
      "mean_token_accuracy": 0.08516570031642914,
      "num_tokens": 939871.0,
      "step": 510
    },
    {
      "entropy": 7.487181520462036,
      "epoch": 0.043268220961982774,
      "grad_norm": 1.1875,
      "learning_rate": 0.000257,
      "loss": 7.1032,
      "mean_token_accuracy": 0.0852330669760704,
      "num_tokens": 948673.0,
      "step": 515
    },
    {
      "entropy": 7.419178056716919,
      "epoch": 0.04368830077714766,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002595,
      "loss": 7.0589,
      "mean_token_accuracy": 0.08271693587303161,
      "num_tokens": 957603.0,
      "step": 520
    },
    {
      "entropy": 7.483588600158692,
      "epoch": 0.04410838059231254,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000262,
      "loss": 7.1144,
      "mean_token_accuracy": 0.08177880719304084,
      "num_tokens": 967731.0,
      "step": 525
    },
    {
      "entropy": 7.489073705673218,
      "epoch": 0.04452846040747742,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00026450000000000003,
      "loss": 7.1329,
      "mean_token_accuracy": 0.08377009257674217,
      "num_tokens": 977427.0,
      "step": 530
    },
    {
      "entropy": 7.537723684310913,
      "epoch": 0.0449485402226423,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00026700000000000004,
      "loss": 7.1087,
      "mean_token_accuracy": 0.08038406819105148,
      "num_tokens": 986758.0,
      "step": 535
    },
    {
      "entropy": 7.401149797439575,
      "epoch": 0.045368620037807186,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00026950000000000005,
      "loss": 7.1028,
      "mean_token_accuracy": 0.0900042973458767,
      "num_tokens": 996377.0,
      "step": 540
    },
    {
      "entropy": 7.488600254058838,
      "epoch": 0.04578869985297206,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00027200000000000005,
      "loss": 7.1394,
      "mean_token_accuracy": 0.07542204968631268,
      "num_tokens": 1006483.0,
      "step": 545
    },
    {
      "entropy": 7.352685880661011,
      "epoch": 0.04620877966813695,
      "grad_norm": 0.875,
      "learning_rate": 0.0002745,
      "loss": 7.0338,
      "mean_token_accuracy": 0.08245834931731225,
      "num_tokens": 1016132.0,
      "step": 550
    },
    {
      "entropy": 7.383127737045288,
      "epoch": 0.04662885948330183,
      "grad_norm": 1.25,
      "learning_rate": 0.000277,
      "loss": 6.9969,
      "mean_token_accuracy": 0.08636636063456535,
      "num_tokens": 1024970.0,
      "step": 555
    },
    {
      "entropy": 7.470243072509765,
      "epoch": 0.04704893929846671,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0002795,
      "loss": 7.0674,
      "mean_token_accuracy": 0.08481250628829003,
      "num_tokens": 1034335.0,
      "step": 560
    },
    {
      "entropy": 7.4651130676269535,
      "epoch": 0.04746901911363159,
      "grad_norm": 1.15625,
      "learning_rate": 0.00028199999999999997,
      "loss": 7.1153,
      "mean_token_accuracy": 0.08971841484308243,
      "num_tokens": 1043954.0,
      "step": 565
    },
    {
      "entropy": 7.4326441287994385,
      "epoch": 0.04788909892879647,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002845,
      "loss": 7.0593,
      "mean_token_accuracy": 0.08415772840380668,
      "num_tokens": 1053554.0,
      "step": 570
    },
    {
      "entropy": 7.493321180343628,
      "epoch": 0.04830917874396135,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000287,
      "loss": 7.0111,
      "mean_token_accuracy": 0.08667988702654839,
      "num_tokens": 1062008.0,
      "step": 575
    },
    {
      "entropy": 7.334539842605591,
      "epoch": 0.048729258559126236,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002895,
      "loss": 7.0738,
      "mean_token_accuracy": 0.08698670640587806,
      "num_tokens": 1070740.0,
      "step": 580
    },
    {
      "entropy": 7.407914113998413,
      "epoch": 0.04914933837429111,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000292,
      "loss": 7.0599,
      "mean_token_accuracy": 0.0839300949126482,
      "num_tokens": 1079681.0,
      "step": 585
    },
    {
      "entropy": 7.3771099090576175,
      "epoch": 0.049569418189456,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002945,
      "loss": 6.9475,
      "mean_token_accuracy": 0.09122345820069314,
      "num_tokens": 1088979.0,
      "step": 590
    },
    {
      "entropy": 7.326050329208374,
      "epoch": 0.04998949800462088,
      "grad_norm": 1.25,
      "learning_rate": 0.000297,
      "loss": 6.9449,
      "mean_token_accuracy": 0.08771323934197425,
      "num_tokens": 1097870.0,
      "step": 595
    },
    {
      "entropy": 7.320249271392822,
      "epoch": 0.05040957781978576,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002995,
      "loss": 7.0421,
      "mean_token_accuracy": 0.08429603464901447,
      "num_tokens": 1107948.0,
      "step": 600
    },
    {
      "entropy": 7.399919700622559,
      "epoch": 0.05082965763495064,
      "grad_norm": 1.078125,
      "learning_rate": 0.000302,
      "loss": 6.9732,
      "mean_token_accuracy": 0.09062434881925582,
      "num_tokens": 1117032.0,
      "step": 605
    },
    {
      "entropy": 7.228488159179688,
      "epoch": 0.051249737450115525,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003045,
      "loss": 6.9617,
      "mean_token_accuracy": 0.08655689135193825,
      "num_tokens": 1127834.0,
      "step": 610
    },
    {
      "entropy": 7.388462066650391,
      "epoch": 0.0516698172652804,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000307,
      "loss": 7.0412,
      "mean_token_accuracy": 0.09550752639770507,
      "num_tokens": 1137382.0,
      "step": 615
    },
    {
      "entropy": 7.257171297073365,
      "epoch": 0.052089897080445285,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003095,
      "loss": 6.8769,
      "mean_token_accuracy": 0.09444506764411927,
      "num_tokens": 1146095.0,
      "step": 620
    },
    {
      "entropy": 7.268283557891846,
      "epoch": 0.05250997689561017,
      "grad_norm": 1.125,
      "learning_rate": 0.000312,
      "loss": 6.9168,
      "mean_token_accuracy": 0.09104742333292962,
      "num_tokens": 1154981.0,
      "step": 625
    },
    {
      "entropy": 7.231864309310913,
      "epoch": 0.052930056710775046,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003145,
      "loss": 6.9162,
      "mean_token_accuracy": 0.09456853494048119,
      "num_tokens": 1164939.0,
      "step": 630
    },
    {
      "entropy": 7.3547971725463865,
      "epoch": 0.05335013652593993,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000317,
      "loss": 7.0466,
      "mean_token_accuracy": 0.08202817216515541,
      "num_tokens": 1174991.0,
      "step": 635
    },
    {
      "entropy": 7.341550350189209,
      "epoch": 0.05377021634110481,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003195,
      "loss": 7.0772,
      "mean_token_accuracy": 0.08350706696510315,
      "num_tokens": 1184885.0,
      "step": 640
    },
    {
      "entropy": 7.257875394821167,
      "epoch": 0.05419029615626969,
      "grad_norm": 1.09375,
      "learning_rate": 0.000322,
      "loss": 6.9751,
      "mean_token_accuracy": 0.08547877669334411,
      "num_tokens": 1193637.0,
      "step": 645
    },
    {
      "entropy": 7.223678970336914,
      "epoch": 0.054610375971434574,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.778,
      "mean_token_accuracy": 0.09493321552872658,
      "num_tokens": 1202188.0,
      "step": 650
    },
    {
      "entropy": 7.3337372779846195,
      "epoch": 0.05503045578659945,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.8974,
      "mean_token_accuracy": 0.0872666135430336,
      "num_tokens": 1210768.0,
      "step": 655
    },
    {
      "entropy": 7.218266105651855,
      "epoch": 0.055450535601764335,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.903,
      "mean_token_accuracy": 0.09044937118887901,
      "num_tokens": 1219819.0,
      "step": 660
    },
    {
      "entropy": 7.269051265716553,
      "epoch": 0.05587061541692922,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.9561,
      "mean_token_accuracy": 0.08509430065751075,
      "num_tokens": 1229703.0,
      "step": 665
    },
    {
      "entropy": 7.323557806015015,
      "epoch": 0.056290695232094096,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.9868,
      "mean_token_accuracy": 0.08851987943053245,
      "num_tokens": 1238942.0,
      "step": 670
    },
    {
      "entropy": 7.31135516166687,
      "epoch": 0.05671077504725898,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000337,
      "loss": 7.0196,
      "mean_token_accuracy": 0.090158711373806,
      "num_tokens": 1248943.0,
      "step": 675
    },
    {
      "entropy": 7.238338661193848,
      "epoch": 0.05713085486242386,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003395,
      "loss": 6.9121,
      "mean_token_accuracy": 0.09068251103162765,
      "num_tokens": 1257761.0,
      "step": 680
    },
    {
      "entropy": 7.152073860168457,
      "epoch": 0.05755093467758874,
      "grad_norm": 1.140625,
      "learning_rate": 0.000342,
      "loss": 6.8534,
      "mean_token_accuracy": 0.0885312557220459,
      "num_tokens": 1267216.0,
      "step": 685
    },
    {
      "entropy": 7.265845727920532,
      "epoch": 0.057971014492753624,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.952,
      "mean_token_accuracy": 0.08522271737456322,
      "num_tokens": 1277210.0,
      "step": 690
    },
    {
      "entropy": 7.267448425292969,
      "epoch": 0.05839109430791851,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000347,
      "loss": 6.8775,
      "mean_token_accuracy": 0.08550659716129302,
      "num_tokens": 1285310.0,
      "step": 695
    },
    {
      "entropy": 7.189469480514527,
      "epoch": 0.058811174123083385,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003495,
      "loss": 6.8822,
      "mean_token_accuracy": 0.09172830283641815,
      "num_tokens": 1294421.0,
      "step": 700
    },
    {
      "entropy": 7.136290740966797,
      "epoch": 0.05923125393824827,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000352,
      "loss": 6.7116,
      "mean_token_accuracy": 0.10014490857720375,
      "num_tokens": 1303281.0,
      "step": 705
    },
    {
      "entropy": 7.132218694686889,
      "epoch": 0.059651333753413145,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003545,
      "loss": 6.8789,
      "mean_token_accuracy": 0.09660332053899764,
      "num_tokens": 1312280.0,
      "step": 710
    },
    {
      "entropy": 7.105095958709716,
      "epoch": 0.06007141356857803,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000357,
      "loss": 6.8389,
      "mean_token_accuracy": 0.08850333318114281,
      "num_tokens": 1321243.0,
      "step": 715
    },
    {
      "entropy": 7.180677127838135,
      "epoch": 0.06049149338374291,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003595,
      "loss": 6.9284,
      "mean_token_accuracy": 0.09514749199151992,
      "num_tokens": 1330324.0,
      "step": 720
    },
    {
      "entropy": 7.1717592716217045,
      "epoch": 0.06091157319890779,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000362,
      "loss": 6.7842,
      "mean_token_accuracy": 0.09439062774181366,
      "num_tokens": 1339485.0,
      "step": 725
    },
    {
      "entropy": 7.2169701099395756,
      "epoch": 0.06133165301407267,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003645,
      "loss": 6.915,
      "mean_token_accuracy": 0.08555234894156456,
      "num_tokens": 1348640.0,
      "step": 730
    },
    {
      "entropy": 7.130293321609497,
      "epoch": 0.06175173282923756,
      "grad_norm": 1.171875,
      "learning_rate": 0.000367,
      "loss": 6.8099,
      "mean_token_accuracy": 0.09767363145947457,
      "num_tokens": 1357581.0,
      "step": 735
    },
    {
      "entropy": 7.150859022140503,
      "epoch": 0.062171812644402434,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003695,
      "loss": 6.8829,
      "mean_token_accuracy": 0.0921759732067585,
      "num_tokens": 1367883.0,
      "step": 740
    },
    {
      "entropy": 7.128716611862183,
      "epoch": 0.06259189245956731,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000372,
      "loss": 6.8294,
      "mean_token_accuracy": 0.08952599465847015,
      "num_tokens": 1376936.0,
      "step": 745
    },
    {
      "entropy": 7.0383734703063965,
      "epoch": 0.0630119722747322,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003745,
      "loss": 6.7534,
      "mean_token_accuracy": 0.09515918046236038,
      "num_tokens": 1386359.0,
      "step": 750
    },
    {
      "entropy": 7.0808555603027346,
      "epoch": 0.06343205208989708,
      "grad_norm": 1.1875,
      "learning_rate": 0.000377,
      "loss": 6.8181,
      "mean_token_accuracy": 0.09296313449740409,
      "num_tokens": 1395223.0,
      "step": 755
    },
    {
      "entropy": 7.2832927227020265,
      "epoch": 0.06385213190506196,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003795,
      "loss": 6.9661,
      "mean_token_accuracy": 0.08833524622023106,
      "num_tokens": 1404917.0,
      "step": 760
    },
    {
      "entropy": 7.180255651473999,
      "epoch": 0.06427221172022685,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000382,
      "loss": 6.857,
      "mean_token_accuracy": 0.10177302286028862,
      "num_tokens": 1413348.0,
      "step": 765
    },
    {
      "entropy": 7.025369262695312,
      "epoch": 0.06469229153539173,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003845,
      "loss": 6.8209,
      "mean_token_accuracy": 0.09210879877209663,
      "num_tokens": 1421726.0,
      "step": 770
    },
    {
      "entropy": 7.136773920059204,
      "epoch": 0.0651123713505566,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.8481,
      "mean_token_accuracy": 0.09178127273917198,
      "num_tokens": 1430686.0,
      "step": 775
    },
    {
      "entropy": 7.201752138137818,
      "epoch": 0.06553245116572148,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.8283,
      "mean_token_accuracy": 0.0948453463613987,
      "num_tokens": 1439499.0,
      "step": 780
    },
    {
      "entropy": 7.1434326171875,
      "epoch": 0.06595253098088637,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.8035,
      "mean_token_accuracy": 0.09644952267408372,
      "num_tokens": 1448220.0,
      "step": 785
    },
    {
      "entropy": 6.982330989837647,
      "epoch": 0.06637261079605125,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.8485,
      "mean_token_accuracy": 0.09032012075185776,
      "num_tokens": 1458217.0,
      "step": 790
    },
    {
      "entropy": 7.011542701721192,
      "epoch": 0.06679269061121614,
      "grad_norm": 1.125,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.7437,
      "mean_token_accuracy": 0.09031623676419258,
      "num_tokens": 1467422.0,
      "step": 795
    },
    {
      "entropy": 7.083204460144043,
      "epoch": 0.06721277042638102,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003995,
      "loss": 6.7143,
      "mean_token_accuracy": 0.09781065881252289,
      "num_tokens": 1476152.0,
      "step": 800
    },
    {
      "entropy": 7.017691135406494,
      "epoch": 0.06763285024154589,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000402,
      "loss": 6.8197,
      "mean_token_accuracy": 0.09335827901959419,
      "num_tokens": 1485248.0,
      "step": 805
    },
    {
      "entropy": 7.028779029846191,
      "epoch": 0.06805293005671077,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004045,
      "loss": 6.7816,
      "mean_token_accuracy": 0.0929147683084011,
      "num_tokens": 1494248.0,
      "step": 810
    },
    {
      "entropy": 7.107398557662964,
      "epoch": 0.06847300987187566,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.8809,
      "mean_token_accuracy": 0.09237429052591324,
      "num_tokens": 1503565.0,
      "step": 815
    },
    {
      "entropy": 7.178503799438476,
      "epoch": 0.06889308968704054,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004095,
      "loss": 6.9674,
      "mean_token_accuracy": 0.08681917861104012,
      "num_tokens": 1513227.0,
      "step": 820
    },
    {
      "entropy": 7.083085680007935,
      "epoch": 0.06931316950220542,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000412,
      "loss": 6.7482,
      "mean_token_accuracy": 0.09896269217133521,
      "num_tokens": 1522312.0,
      "step": 825
    },
    {
      "entropy": 7.000125169754028,
      "epoch": 0.0697332493173703,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004145,
      "loss": 6.7224,
      "mean_token_accuracy": 0.09859632030129432,
      "num_tokens": 1531720.0,
      "step": 830
    },
    {
      "entropy": 7.015534973144531,
      "epoch": 0.07015332913253518,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000417,
      "loss": 6.7888,
      "mean_token_accuracy": 0.09380378648638725,
      "num_tokens": 1541238.0,
      "step": 835
    },
    {
      "entropy": 7.136422204971313,
      "epoch": 0.07057340894770006,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004195,
      "loss": 6.8785,
      "mean_token_accuracy": 0.09427597969770432,
      "num_tokens": 1550875.0,
      "step": 840
    },
    {
      "entropy": 6.93135290145874,
      "epoch": 0.07099348876286495,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000422,
      "loss": 6.8248,
      "mean_token_accuracy": 0.09056826308369637,
      "num_tokens": 1560287.0,
      "step": 845
    },
    {
      "entropy": 7.053813314437866,
      "epoch": 0.07141356857802983,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004245,
      "loss": 6.7055,
      "mean_token_accuracy": 0.10149886757135392,
      "num_tokens": 1569043.0,
      "step": 850
    },
    {
      "entropy": 6.978960704803467,
      "epoch": 0.07183364839319471,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000427,
      "loss": 6.726,
      "mean_token_accuracy": 0.09670688211917877,
      "num_tokens": 1578112.0,
      "step": 855
    },
    {
      "entropy": 6.866673278808594,
      "epoch": 0.07225372820835958,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004295,
      "loss": 6.6514,
      "mean_token_accuracy": 0.10521448627114297,
      "num_tokens": 1586587.0,
      "step": 860
    },
    {
      "entropy": 7.09016604423523,
      "epoch": 0.07267380802352447,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000432,
      "loss": 6.8069,
      "mean_token_accuracy": 0.09529538303613663,
      "num_tokens": 1595585.0,
      "step": 865
    },
    {
      "entropy": 6.990141916275024,
      "epoch": 0.07309388783868935,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004345,
      "loss": 6.7599,
      "mean_token_accuracy": 0.09652992263436318,
      "num_tokens": 1605355.0,
      "step": 870
    },
    {
      "entropy": 7.018218660354615,
      "epoch": 0.07351396765385423,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000437,
      "loss": 6.7838,
      "mean_token_accuracy": 0.0944277286529541,
      "num_tokens": 1613637.0,
      "step": 875
    },
    {
      "entropy": 6.990546417236328,
      "epoch": 0.07393404746901912,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004395,
      "loss": 6.7417,
      "mean_token_accuracy": 0.09869495183229446,
      "num_tokens": 1622731.0,
      "step": 880
    },
    {
      "entropy": 7.067089796066284,
      "epoch": 0.074354127284184,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000442,
      "loss": 6.7144,
      "mean_token_accuracy": 0.09230813458561897,
      "num_tokens": 1632098.0,
      "step": 885
    },
    {
      "entropy": 6.908722448348999,
      "epoch": 0.07477420709934887,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004445,
      "loss": 6.7079,
      "mean_token_accuracy": 0.09467918276786805,
      "num_tokens": 1641259.0,
      "step": 890
    },
    {
      "entropy": 6.984257411956787,
      "epoch": 0.07519428691451376,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000447,
      "loss": 6.7835,
      "mean_token_accuracy": 0.09350861608982086,
      "num_tokens": 1651362.0,
      "step": 895
    },
    {
      "entropy": 6.942133188247681,
      "epoch": 0.07561436672967864,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.698,
      "mean_token_accuracy": 0.09561712741851806,
      "num_tokens": 1660190.0,
      "step": 900
    },
    {
      "entropy": 6.949711561203003,
      "epoch": 0.07603444654484352,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.7202,
      "mean_token_accuracy": 0.09667671620845794,
      "num_tokens": 1669020.0,
      "step": 905
    },
    {
      "entropy": 7.059488010406494,
      "epoch": 0.0764545263600084,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.7692,
      "mean_token_accuracy": 0.10004925951361657,
      "num_tokens": 1678158.0,
      "step": 910
    },
    {
      "entropy": 6.941928768157959,
      "epoch": 0.07687460617517328,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.7751,
      "mean_token_accuracy": 0.08887208923697472,
      "num_tokens": 1687481.0,
      "step": 915
    },
    {
      "entropy": 6.903632497787475,
      "epoch": 0.07729468599033816,
      "grad_norm": 1.0,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.7248,
      "mean_token_accuracy": 0.10260720252990722,
      "num_tokens": 1696782.0,
      "step": 920
    },
    {
      "entropy": 6.950085973739624,
      "epoch": 0.07771476580550304,
      "grad_norm": 1.0,
      "learning_rate": 0.000462,
      "loss": 6.7327,
      "mean_token_accuracy": 0.10223789662122726,
      "num_tokens": 1706153.0,
      "step": 925
    },
    {
      "entropy": 6.981480550765991,
      "epoch": 0.07813484562066793,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004645,
      "loss": 6.7583,
      "mean_token_accuracy": 0.09635938182473183,
      "num_tokens": 1715585.0,
      "step": 930
    },
    {
      "entropy": 7.110754156112671,
      "epoch": 0.07855492543583281,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000467,
      "loss": 6.8789,
      "mean_token_accuracy": 0.09635257050395012,
      "num_tokens": 1724857.0,
      "step": 935
    },
    {
      "entropy": 6.890129041671753,
      "epoch": 0.0789750052509977,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004695,
      "loss": 6.6811,
      "mean_token_accuracy": 0.10242518037557602,
      "num_tokens": 1733528.0,
      "step": 940
    },
    {
      "entropy": 7.000388526916504,
      "epoch": 0.07939508506616257,
      "grad_norm": 0.875,
      "learning_rate": 0.000472,
      "loss": 6.8096,
      "mean_token_accuracy": 0.09821859821677208,
      "num_tokens": 1742953.0,
      "step": 945
    },
    {
      "entropy": 6.962129068374634,
      "epoch": 0.07981516488132745,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004745,
      "loss": 6.757,
      "mean_token_accuracy": 0.10171304419636726,
      "num_tokens": 1752155.0,
      "step": 950
    },
    {
      "entropy": 6.876726579666138,
      "epoch": 0.08023524469649233,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000477,
      "loss": 6.6364,
      "mean_token_accuracy": 0.09934233203530311,
      "num_tokens": 1760562.0,
      "step": 955
    },
    {
      "entropy": 6.884459018707275,
      "epoch": 0.08065532451165722,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004795,
      "loss": 6.6462,
      "mean_token_accuracy": 0.1006347469985485,
      "num_tokens": 1769631.0,
      "step": 960
    },
    {
      "entropy": 6.902465105056763,
      "epoch": 0.0810754043268221,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000482,
      "loss": 6.7043,
      "mean_token_accuracy": 0.10013237074017525,
      "num_tokens": 1779080.0,
      "step": 965
    },
    {
      "entropy": 6.916742420196533,
      "epoch": 0.08149548414198697,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004845,
      "loss": 6.6819,
      "mean_token_accuracy": 0.10164752677083015,
      "num_tokens": 1787830.0,
      "step": 970
    },
    {
      "entropy": 6.823171949386596,
      "epoch": 0.08191556395715185,
      "grad_norm": 0.9375,
      "learning_rate": 0.000487,
      "loss": 6.6048,
      "mean_token_accuracy": 0.10189055874943734,
      "num_tokens": 1796998.0,
      "step": 975
    },
    {
      "entropy": 6.977118110656738,
      "epoch": 0.08233564377231674,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004895,
      "loss": 6.6814,
      "mean_token_accuracy": 0.10083219707012177,
      "num_tokens": 1806194.0,
      "step": 980
    },
    {
      "entropy": 6.672053623199463,
      "epoch": 0.08275572358748162,
      "grad_norm": 1.046875,
      "learning_rate": 0.000492,
      "loss": 6.5647,
      "mean_token_accuracy": 0.10488267168402672,
      "num_tokens": 1815751.0,
      "step": 985
    },
    {
      "entropy": 6.7590797424316404,
      "epoch": 0.0831758034026465,
      "grad_norm": 1.0,
      "learning_rate": 0.0004945,
      "loss": 6.6428,
      "mean_token_accuracy": 0.10192798376083374,
      "num_tokens": 1825379.0,
      "step": 990
    },
    {
      "entropy": 6.884732866287232,
      "epoch": 0.08359588321781139,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000497,
      "loss": 6.6025,
      "mean_token_accuracy": 0.10135274231433869,
      "num_tokens": 1834158.0,
      "step": 995
    },
    {
      "entropy": 6.860666513442993,
      "epoch": 0.08401596303297626,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995,
      "loss": 6.5689,
      "mean_token_accuracy": 0.10223893150687217,
      "num_tokens": 1842724.0,
      "step": 1000
    },
    {
      "entropy": 6.812324571609497,
      "epoch": 0.08443604284814114,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499999998724557,
      "loss": 6.5774,
      "mean_token_accuracy": 0.10202668011188507,
      "num_tokens": 1852485.0,
      "step": 1005
    },
    {
      "entropy": 6.779099035263061,
      "epoch": 0.08485612266330603,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999999935430703,
      "loss": 6.6115,
      "mean_token_accuracy": 0.10559484660625458,
      "num_tokens": 1861303.0,
      "step": 1010
    },
    {
      "entropy": 6.770784664154053,
      "epoch": 0.08527620247847091,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999999843758243,
      "loss": 6.5927,
      "mean_token_accuracy": 0.1131505697965622,
      "num_tokens": 1870859.0,
      "step": 1015
    },
    {
      "entropy": 6.940572690963745,
      "epoch": 0.0856962822936358,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999999712228196,
      "loss": 6.7566,
      "mean_token_accuracy": 0.09317019283771515,
      "num_tokens": 1880295.0,
      "step": 1020
    },
    {
      "entropy": 6.93710470199585,
      "epoch": 0.08611636210880068,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999999540840562,
      "loss": 6.6611,
      "mean_token_accuracy": 0.10422083139419555,
      "num_tokens": 1889193.0,
      "step": 1025
    },
    {
      "entropy": 6.811184549331665,
      "epoch": 0.08653644192396555,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999999329595345,
      "loss": 6.7584,
      "mean_token_accuracy": 0.09407720491290092,
      "num_tokens": 1899437.0,
      "step": 1030
    },
    {
      "entropy": 6.87418532371521,
      "epoch": 0.08695652173913043,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999999078492548,
      "loss": 6.6473,
      "mean_token_accuracy": 0.1043108843266964,
      "num_tokens": 1907882.0,
      "step": 1035
    },
    {
      "entropy": 6.7778688907623295,
      "epoch": 0.08737660155429532,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999998787532176,
      "loss": 6.5571,
      "mean_token_accuracy": 0.10416635349392891,
      "num_tokens": 1916872.0,
      "step": 1040
    },
    {
      "entropy": 6.838136434555054,
      "epoch": 0.0877966813694602,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999998456714234,
      "loss": 6.7226,
      "mean_token_accuracy": 0.10190053880214692,
      "num_tokens": 1926636.0,
      "step": 1045
    },
    {
      "entropy": 6.815088891983033,
      "epoch": 0.08821676118462508,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999998086038729,
      "loss": 6.6294,
      "mean_token_accuracy": 0.1084713064134121,
      "num_tokens": 1935962.0,
      "step": 1050
    },
    {
      "entropy": 6.798814964294434,
      "epoch": 0.08863684099978995,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999997675505665,
      "loss": 6.5955,
      "mean_token_accuracy": 0.10102920681238174,
      "num_tokens": 1944600.0,
      "step": 1055
    },
    {
      "entropy": 6.8712811946868895,
      "epoch": 0.08905692081495484,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999997225115052,
      "loss": 6.7682,
      "mean_token_accuracy": 0.09951933771371842,
      "num_tokens": 1954234.0,
      "step": 1060
    },
    {
      "entropy": 6.951000690460205,
      "epoch": 0.08947700063011972,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999996734866896,
      "loss": 6.7171,
      "mean_token_accuracy": 0.09834984764456749,
      "num_tokens": 1964499.0,
      "step": 1065
    },
    {
      "entropy": 6.700433588027954,
      "epoch": 0.0898970804452846,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999996204761206,
      "loss": 6.4396,
      "mean_token_accuracy": 0.11416423171758652,
      "num_tokens": 1973635.0,
      "step": 1070
    },
    {
      "entropy": 6.754154396057129,
      "epoch": 0.09031716026044949,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999995634797993,
      "loss": 6.5931,
      "mean_token_accuracy": 0.10663349330425262,
      "num_tokens": 1983509.0,
      "step": 1075
    },
    {
      "entropy": 6.815823554992676,
      "epoch": 0.09073724007561437,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999995024977265,
      "loss": 6.5793,
      "mean_token_accuracy": 0.10800760760903358,
      "num_tokens": 1992336.0,
      "step": 1080
    },
    {
      "entropy": 6.820423316955567,
      "epoch": 0.09115731989077924,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999994375299034,
      "loss": 6.5959,
      "mean_token_accuracy": 0.10795153826475143,
      "num_tokens": 2001931.0,
      "step": 1085
    },
    {
      "entropy": 6.72901930809021,
      "epoch": 0.09157739970594413,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499999368576331,
      "loss": 6.4682,
      "mean_token_accuracy": 0.11168377995491027,
      "num_tokens": 2010935.0,
      "step": 1090
    },
    {
      "entropy": 6.699783134460449,
      "epoch": 0.09199747952110901,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999992956370109,
      "loss": 6.5304,
      "mean_token_accuracy": 0.10683032646775245,
      "num_tokens": 2020587.0,
      "step": 1095
    },
    {
      "entropy": 6.677279663085938,
      "epoch": 0.0924175593362739,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499999218711944,
      "loss": 6.5573,
      "mean_token_accuracy": 0.1078680157661438,
      "num_tokens": 2029743.0,
      "step": 1100
    },
    {
      "entropy": 6.796489238739014,
      "epoch": 0.09283763915143878,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999991378011317,
      "loss": 6.5775,
      "mean_token_accuracy": 0.1102212406694889,
      "num_tokens": 2038468.0,
      "step": 1105
    },
    {
      "entropy": 6.65820722579956,
      "epoch": 0.09325771896660366,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999990529045757,
      "loss": 6.4917,
      "mean_token_accuracy": 0.10695874094963073,
      "num_tokens": 2047456.0,
      "step": 1110
    },
    {
      "entropy": 6.895982646942139,
      "epoch": 0.09367779878176853,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999989640222771,
      "loss": 6.8041,
      "mean_token_accuracy": 0.09525922834873199,
      "num_tokens": 2056691.0,
      "step": 1115
    },
    {
      "entropy": 6.853526449203491,
      "epoch": 0.09409787859693342,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499998871154238,
      "loss": 6.5875,
      "mean_token_accuracy": 0.10508395135402679,
      "num_tokens": 2066068.0,
      "step": 1120
    },
    {
      "entropy": 6.754317712783814,
      "epoch": 0.0945179584120983,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999987743004597,
      "loss": 6.5313,
      "mean_token_accuracy": 0.11206126511096955,
      "num_tokens": 2075113.0,
      "step": 1125
    },
    {
      "entropy": 6.816254997253418,
      "epoch": 0.09493803822726318,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999986734609438,
      "loss": 6.6612,
      "mean_token_accuracy": 0.1047824963927269,
      "num_tokens": 2084557.0,
      "step": 1130
    },
    {
      "entropy": 6.785951089859009,
      "epoch": 0.09535811804242807,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999985686356923,
      "loss": 6.5497,
      "mean_token_accuracy": 0.10589791536331176,
      "num_tokens": 2093424.0,
      "step": 1135
    },
    {
      "entropy": 6.7812634944915775,
      "epoch": 0.09577819785759294,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499998459824707,
      "loss": 6.6748,
      "mean_token_accuracy": 0.10217821151018143,
      "num_tokens": 2103066.0,
      "step": 1140
    },
    {
      "entropy": 6.756265497207641,
      "epoch": 0.09619827767275782,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00049999834702799,
      "loss": 6.5477,
      "mean_token_accuracy": 0.11097522377967835,
      "num_tokens": 2112447.0,
      "step": 1145
    },
    {
      "entropy": 6.820600318908691,
      "epoch": 0.0966183574879227,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999982302455431,
      "loss": 6.5692,
      "mean_token_accuracy": 0.11053408607840538,
      "num_tokens": 2121949.0,
      "step": 1150
    },
    {
      "entropy": 6.731539011001587,
      "epoch": 0.09703843730308759,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999981094773683,
      "loss": 6.4735,
      "mean_token_accuracy": 0.11045165956020356,
      "num_tokens": 2130464.0,
      "step": 1155
    },
    {
      "entropy": 6.791986989974975,
      "epoch": 0.09745851711825247,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499997984723468,
      "loss": 6.627,
      "mean_token_accuracy": 0.10332545191049576,
      "num_tokens": 2139577.0,
      "step": 1160
    },
    {
      "entropy": 6.555002975463867,
      "epoch": 0.09787859693341736,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999978559838441,
      "loss": 6.3652,
      "mean_token_accuracy": 0.10894591435790062,
      "num_tokens": 2147919.0,
      "step": 1165
    },
    {
      "entropy": 6.623918771743774,
      "epoch": 0.09829867674858223,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999977232584991,
      "loss": 6.5152,
      "mean_token_accuracy": 0.1095739871263504,
      "num_tokens": 2156936.0,
      "step": 1170
    },
    {
      "entropy": 6.789136934280395,
      "epoch": 0.09871875656374711,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999975865474354,
      "loss": 6.5864,
      "mean_token_accuracy": 0.10703087151050568,
      "num_tokens": 2165362.0,
      "step": 1175
    },
    {
      "entropy": 6.65335168838501,
      "epoch": 0.099138836378912,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999974458506551,
      "loss": 6.5154,
      "mean_token_accuracy": 0.10435130968689918,
      "num_tokens": 2173665.0,
      "step": 1180
    },
    {
      "entropy": 6.8575843334197994,
      "epoch": 0.09955891619407688,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499997301168161,
      "loss": 6.4988,
      "mean_token_accuracy": 0.11348986551165581,
      "num_tokens": 2182222.0,
      "step": 1185
    },
    {
      "entropy": 6.672175312042237,
      "epoch": 0.09997899600924176,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999971524999556,
      "loss": 6.5809,
      "mean_token_accuracy": 0.11042713969945908,
      "num_tokens": 2192358.0,
      "step": 1190
    },
    {
      "entropy": 6.7970047950744625,
      "epoch": 0.10039907582440663,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999969998460414,
      "loss": 6.5411,
      "mean_token_accuracy": 0.10771783664822579,
      "num_tokens": 2201889.0,
      "step": 1195
    },
    {
      "entropy": 6.672988796234131,
      "epoch": 0.10081915563957151,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999968432064213,
      "loss": 6.54,
      "mean_token_accuracy": 0.12040728628635407,
      "num_tokens": 2211810.0,
      "step": 1200
    },
    {
      "entropy": 6.662283229827881,
      "epoch": 0.1012392354547364,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999966825810979,
      "loss": 6.4935,
      "mean_token_accuracy": 0.10729209631681443,
      "num_tokens": 2221123.0,
      "step": 1205
    },
    {
      "entropy": 6.682981157302857,
      "epoch": 0.10165931526990128,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999965179700742,
      "loss": 6.4347,
      "mean_token_accuracy": 0.11437768638134002,
      "num_tokens": 2230129.0,
      "step": 1210
    },
    {
      "entropy": 6.584838199615478,
      "epoch": 0.10207939508506617,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499996349373353,
      "loss": 6.5049,
      "mean_token_accuracy": 0.11030597984790802,
      "num_tokens": 2239929.0,
      "step": 1215
    },
    {
      "entropy": 6.768577909469604,
      "epoch": 0.10249947490023105,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999961767909374,
      "loss": 6.478,
      "mean_token_accuracy": 0.11449568048119545,
      "num_tokens": 2248078.0,
      "step": 1220
    },
    {
      "entropy": 6.615856742858886,
      "epoch": 0.10291955471539592,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999960002228303,
      "loss": 6.5666,
      "mean_token_accuracy": 0.11093122437596321,
      "num_tokens": 2256975.0,
      "step": 1225
    },
    {
      "entropy": 6.733388185501099,
      "epoch": 0.1033396345305608,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999958196690349,
      "loss": 6.4212,
      "mean_token_accuracy": 0.11245058998465537,
      "num_tokens": 2265797.0,
      "step": 1230
    },
    {
      "entropy": 6.668054151535034,
      "epoch": 0.10375971434572569,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999956351295545,
      "loss": 6.5197,
      "mean_token_accuracy": 0.1140691690146923,
      "num_tokens": 2274099.0,
      "step": 1235
    },
    {
      "entropy": 6.614612483978272,
      "epoch": 0.10417979416089057,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999954466043922,
      "loss": 6.4309,
      "mean_token_accuracy": 0.11669434607028961,
      "num_tokens": 2282360.0,
      "step": 1240
    },
    {
      "entropy": 6.6785484790802006,
      "epoch": 0.10459987397605545,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999952540935514,
      "loss": 6.5283,
      "mean_token_accuracy": 0.10534609332680703,
      "num_tokens": 2292714.0,
      "step": 1245
    },
    {
      "entropy": 6.669230937957764,
      "epoch": 0.10501995379122034,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999950575970356,
      "loss": 6.4759,
      "mean_token_accuracy": 0.11022288277745247,
      "num_tokens": 2301633.0,
      "step": 1250
    },
    {
      "entropy": 6.690946578979492,
      "epoch": 0.10544003360638521,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999948571148482,
      "loss": 6.4607,
      "mean_token_accuracy": 0.11288312897086143,
      "num_tokens": 2310067.0,
      "step": 1255
    },
    {
      "entropy": 6.663266944885254,
      "epoch": 0.10586011342155009,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999946526469927,
      "loss": 6.5329,
      "mean_token_accuracy": 0.11119069010019303,
      "num_tokens": 2320090.0,
      "step": 1260
    },
    {
      "entropy": 6.619077062606811,
      "epoch": 0.10628019323671498,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999944441934728,
      "loss": 6.4811,
      "mean_token_accuracy": 0.11513725593686104,
      "num_tokens": 2329255.0,
      "step": 1265
    },
    {
      "entropy": 6.743935060501099,
      "epoch": 0.10670027305187986,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999942317542922,
      "loss": 6.5708,
      "mean_token_accuracy": 0.10842081904411316,
      "num_tokens": 2339535.0,
      "step": 1270
    },
    {
      "entropy": 6.647988319396973,
      "epoch": 0.10712035286704474,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999940153294546,
      "loss": 6.4584,
      "mean_token_accuracy": 0.11369285136461257,
      "num_tokens": 2348948.0,
      "step": 1275
    },
    {
      "entropy": 6.573079681396484,
      "epoch": 0.10754043268220961,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499993794918964,
      "loss": 6.4903,
      "mean_token_accuracy": 0.11098795756697655,
      "num_tokens": 2359141.0,
      "step": 1280
    },
    {
      "entropy": 6.623832654953003,
      "epoch": 0.1079605124973745,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999935705228241,
      "loss": 6.5296,
      "mean_token_accuracy": 0.10813465043902397,
      "num_tokens": 2368906.0,
      "step": 1285
    },
    {
      "entropy": 6.781174278259277,
      "epoch": 0.10838059231253938,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999933421410389,
      "loss": 6.5156,
      "mean_token_accuracy": 0.11340161636471749,
      "num_tokens": 2377029.0,
      "step": 1290
    },
    {
      "entropy": 6.664435195922851,
      "epoch": 0.10880067212770426,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004999931097736125,
      "loss": 6.5678,
      "mean_token_accuracy": 0.10811809226870536,
      "num_tokens": 2387088.0,
      "step": 1295
    },
    {
      "entropy": 6.711423254013061,
      "epoch": 0.10922075194286915,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999928734205492,
      "loss": 6.4748,
      "mean_token_accuracy": 0.11095682382583619,
      "num_tokens": 2395596.0,
      "step": 1300
    },
    {
      "entropy": 6.609859323501587,
      "epoch": 0.10964083175803403,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999926330818528,
      "loss": 6.4489,
      "mean_token_accuracy": 0.11987485438585281,
      "num_tokens": 2404506.0,
      "step": 1305
    },
    {
      "entropy": 6.6314843654632565,
      "epoch": 0.1100609115731989,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999923887575278,
      "loss": 6.4954,
      "mean_token_accuracy": 0.11150062307715417,
      "num_tokens": 2414342.0,
      "step": 1310
    },
    {
      "entropy": 6.703009176254272,
      "epoch": 0.11048099138836379,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999921404475785,
      "loss": 6.4686,
      "mean_token_accuracy": 0.11657750681042671,
      "num_tokens": 2423076.0,
      "step": 1315
    },
    {
      "entropy": 6.580958127975464,
      "epoch": 0.11090107120352867,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004999918881520093,
      "loss": 6.4108,
      "mean_token_accuracy": 0.11714263036847114,
      "num_tokens": 2432492.0,
      "step": 1320
    },
    {
      "entropy": 6.586594104766846,
      "epoch": 0.11132115101869355,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999916318708246,
      "loss": 6.377,
      "mean_token_accuracy": 0.11821310296654701,
      "num_tokens": 2441916.0,
      "step": 1325
    },
    {
      "entropy": 6.610158824920655,
      "epoch": 0.11174123083385844,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999913716040291,
      "loss": 6.4429,
      "mean_token_accuracy": 0.11613110527396202,
      "num_tokens": 2450932.0,
      "step": 1330
    },
    {
      "entropy": 6.58526463508606,
      "epoch": 0.11216131064902331,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999911073516272,
      "loss": 6.4417,
      "mean_token_accuracy": 0.11660609394311905,
      "num_tokens": 2460058.0,
      "step": 1335
    },
    {
      "entropy": 6.543748760223389,
      "epoch": 0.11258139046418819,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999908391136237,
      "loss": 6.3882,
      "mean_token_accuracy": 0.12162717878818512,
      "num_tokens": 2469607.0,
      "step": 1340
    },
    {
      "entropy": 6.594651317596435,
      "epoch": 0.11300147027935308,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999905668900234,
      "loss": 6.4404,
      "mean_token_accuracy": 0.11361290663480758,
      "num_tokens": 2478345.0,
      "step": 1345
    },
    {
      "entropy": 6.574893379211426,
      "epoch": 0.11342155009451796,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499990290680831,
      "loss": 6.3587,
      "mean_token_accuracy": 0.11955868676304818,
      "num_tokens": 2486662.0,
      "step": 1350
    },
    {
      "entropy": 6.634180545806885,
      "epoch": 0.11384162990968284,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999900104860516,
      "loss": 6.4884,
      "mean_token_accuracy": 0.11496279463171959,
      "num_tokens": 2495392.0,
      "step": 1355
    },
    {
      "entropy": 6.6078431606292725,
      "epoch": 0.11426170972484773,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999897263056898,
      "loss": 6.5187,
      "mean_token_accuracy": 0.11110892221331596,
      "num_tokens": 2505254.0,
      "step": 1360
    },
    {
      "entropy": 6.676256227493286,
      "epoch": 0.1146817895400126,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499989438139751,
      "loss": 6.3391,
      "mean_token_accuracy": 0.11575900986790658,
      "num_tokens": 2514096.0,
      "step": 1365
    },
    {
      "entropy": 6.550286197662354,
      "epoch": 0.11510186935517748,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999891459882401,
      "loss": 6.3458,
      "mean_token_accuracy": 0.11746028289198876,
      "num_tokens": 2523635.0,
      "step": 1370
    },
    {
      "entropy": 6.538182258605957,
      "epoch": 0.11552194917034236,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999888498511624,
      "loss": 6.4382,
      "mean_token_accuracy": 0.1144244447350502,
      "num_tokens": 2532528.0,
      "step": 1375
    },
    {
      "entropy": 6.527873754501343,
      "epoch": 0.11594202898550725,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999885497285229,
      "loss": 6.3289,
      "mean_token_accuracy": 0.11667455583810807,
      "num_tokens": 2541893.0,
      "step": 1380
    },
    {
      "entropy": 6.541602611541748,
      "epoch": 0.11636210880067213,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999882456203273,
      "loss": 6.3923,
      "mean_token_accuracy": 0.1184983029961586,
      "num_tokens": 2551551.0,
      "step": 1385
    },
    {
      "entropy": 6.588447904586792,
      "epoch": 0.11678218861583702,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999879375265806,
      "loss": 6.347,
      "mean_token_accuracy": 0.11739155948162079,
      "num_tokens": 2560183.0,
      "step": 1390
    },
    {
      "entropy": 6.454002952575683,
      "epoch": 0.11720226843100189,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999876254472886,
      "loss": 6.2601,
      "mean_token_accuracy": 0.12417450621724128,
      "num_tokens": 2568697.0,
      "step": 1395
    },
    {
      "entropy": 6.591037464141846,
      "epoch": 0.11762234824616677,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999873093824565,
      "loss": 6.4467,
      "mean_token_accuracy": 0.11451651230454445,
      "num_tokens": 2578151.0,
      "step": 1400
    },
    {
      "entropy": 6.704318237304688,
      "epoch": 0.11804242806133165,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999869893320902,
      "loss": 6.5716,
      "mean_token_accuracy": 0.11275385171175004,
      "num_tokens": 2585901.0,
      "step": 1405
    },
    {
      "entropy": 6.5783727169036865,
      "epoch": 0.11846250787649654,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999866652961952,
      "loss": 6.3901,
      "mean_token_accuracy": 0.11096679419279099,
      "num_tokens": 2595655.0,
      "step": 1410
    },
    {
      "entropy": 6.604079818725586,
      "epoch": 0.11888258769166142,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999863372747773,
      "loss": 6.3632,
      "mean_token_accuracy": 0.11101334393024445,
      "num_tokens": 2604949.0,
      "step": 1415
    },
    {
      "entropy": 6.592713022232056,
      "epoch": 0.11930266750682629,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999860052678423,
      "loss": 6.4367,
      "mean_token_accuracy": 0.11833727061748504,
      "num_tokens": 2614260.0,
      "step": 1420
    },
    {
      "entropy": 6.529619216918945,
      "epoch": 0.11972274732199117,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999856692753959,
      "loss": 6.4205,
      "mean_token_accuracy": 0.11050314903259277,
      "num_tokens": 2623740.0,
      "step": 1425
    },
    {
      "entropy": 6.583321142196655,
      "epoch": 0.12014282713715606,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999853292974444,
      "loss": 6.3153,
      "mean_token_accuracy": 0.12103430330753326,
      "num_tokens": 2631998.0,
      "step": 1430
    },
    {
      "entropy": 6.522999906539917,
      "epoch": 0.12056290695232094,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999849853339936,
      "loss": 6.4682,
      "mean_token_accuracy": 0.12008080482482911,
      "num_tokens": 2641169.0,
      "step": 1435
    },
    {
      "entropy": 6.666995334625244,
      "epoch": 0.12098298676748583,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004999846373850497,
      "loss": 6.3232,
      "mean_token_accuracy": 0.11853480413556099,
      "num_tokens": 2650576.0,
      "step": 1440
    },
    {
      "entropy": 6.456968832015991,
      "epoch": 0.12140306658265071,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999842854506186,
      "loss": 6.4062,
      "mean_token_accuracy": 0.11563461422920226,
      "num_tokens": 2660817.0,
      "step": 1445
    },
    {
      "entropy": 6.559107828140259,
      "epoch": 0.12182314639781558,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999839295307069,
      "loss": 6.3481,
      "mean_token_accuracy": 0.11764338314533233,
      "num_tokens": 2669338.0,
      "step": 1450
    },
    {
      "entropy": 6.621726083755493,
      "epoch": 0.12224322621298046,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999835696253206,
      "loss": 6.4021,
      "mean_token_accuracy": 0.11631894037127495,
      "num_tokens": 2679108.0,
      "step": 1455
    },
    {
      "entropy": 6.533853530883789,
      "epoch": 0.12266330602814535,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999832057344664,
      "loss": 6.37,
      "mean_token_accuracy": 0.11607012003660203,
      "num_tokens": 2688126.0,
      "step": 1460
    },
    {
      "entropy": 6.420909214019775,
      "epoch": 0.12308338584331023,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999828378581504,
      "loss": 6.3282,
      "mean_token_accuracy": 0.12565367370843888,
      "num_tokens": 2697245.0,
      "step": 1465
    },
    {
      "entropy": 6.610960960388184,
      "epoch": 0.12350346565847511,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999824659963793,
      "loss": 6.3873,
      "mean_token_accuracy": 0.11763433516025543,
      "num_tokens": 2705934.0,
      "step": 1470
    },
    {
      "entropy": 6.490250873565674,
      "epoch": 0.12392354547364,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999820901491598,
      "loss": 6.314,
      "mean_token_accuracy": 0.12268458679318428,
      "num_tokens": 2714367.0,
      "step": 1475
    },
    {
      "entropy": 6.484623718261719,
      "epoch": 0.12434362528880487,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999817103164983,
      "loss": 6.3522,
      "mean_token_accuracy": 0.12056686282157898,
      "num_tokens": 2724366.0,
      "step": 1480
    },
    {
      "entropy": 6.5408772945404055,
      "epoch": 0.12476370510396975,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999813264984017,
      "loss": 6.3547,
      "mean_token_accuracy": 0.11844174638390541,
      "num_tokens": 2733980.0,
      "step": 1485
    },
    {
      "entropy": 6.583751821517945,
      "epoch": 0.12518378491913462,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999809386948767,
      "loss": 6.3494,
      "mean_token_accuracy": 0.12221027910709381,
      "num_tokens": 2744013.0,
      "step": 1490
    },
    {
      "entropy": 6.452259206771851,
      "epoch": 0.12560386473429952,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999805469059302,
      "loss": 6.4272,
      "mean_token_accuracy": 0.11441577672958374,
      "num_tokens": 2753385.0,
      "step": 1495
    },
    {
      "entropy": 6.51189775466919,
      "epoch": 0.1260239445494644,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999801511315693,
      "loss": 6.2651,
      "mean_token_accuracy": 0.12032239362597466,
      "num_tokens": 2762875.0,
      "step": 1500
    },
    {
      "entropy": 6.593140554428101,
      "epoch": 0.1264440243646293,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999797513718007,
      "loss": 6.345,
      "mean_token_accuracy": 0.12291451320052146,
      "num_tokens": 2772182.0,
      "step": 1505
    },
    {
      "entropy": 6.4095139503479,
      "epoch": 0.12686410417979416,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999793476266317,
      "loss": 6.2994,
      "mean_token_accuracy": 0.12340676560997962,
      "num_tokens": 2780814.0,
      "step": 1510
    },
    {
      "entropy": 6.818328523635865,
      "epoch": 0.12728418399495905,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999789398960695,
      "loss": 6.5676,
      "mean_token_accuracy": 0.1171888455748558,
      "num_tokens": 2791104.0,
      "step": 1515
    },
    {
      "entropy": 6.400639343261719,
      "epoch": 0.12770426381012392,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999785281801212,
      "loss": 6.2622,
      "mean_token_accuracy": 0.12264100611209869,
      "num_tokens": 2800081.0,
      "step": 1520
    },
    {
      "entropy": 6.553307294845581,
      "epoch": 0.1281243436252888,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499978112478794,
      "loss": 6.3896,
      "mean_token_accuracy": 0.11860985979437828,
      "num_tokens": 2809096.0,
      "step": 1525
    },
    {
      "entropy": 6.57843770980835,
      "epoch": 0.1285444234404537,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999776927920955,
      "loss": 6.365,
      "mean_token_accuracy": 0.12388472035527229,
      "num_tokens": 2818857.0,
      "step": 1530
    },
    {
      "entropy": 6.497845268249511,
      "epoch": 0.12896450325561856,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499977269120033,
      "loss": 6.4425,
      "mean_token_accuracy": 0.11523956879973411,
      "num_tokens": 2829332.0,
      "step": 1535
    },
    {
      "entropy": 6.516994428634644,
      "epoch": 0.12938458307078346,
      "grad_norm": 0.875,
      "learning_rate": 0.000499976841462614,
      "loss": 6.3434,
      "mean_token_accuracy": 0.11322477459907532,
      "num_tokens": 2839193.0,
      "step": 1540
    },
    {
      "entropy": 6.506987857818603,
      "epoch": 0.12980466288594833,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499976409819846,
      "loss": 6.352,
      "mean_token_accuracy": 0.11825493946671486,
      "num_tokens": 2848535.0,
      "step": 1545
    },
    {
      "entropy": 6.395366477966308,
      "epoch": 0.1302247427011132,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999759741917369,
      "loss": 6.2444,
      "mean_token_accuracy": 0.12733460888266562,
      "num_tokens": 2858090.0,
      "step": 1550
    },
    {
      "entropy": 6.548688650131226,
      "epoch": 0.1306448225162781,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999755345782941,
      "loss": 6.3894,
      "mean_token_accuracy": 0.11661416217684746,
      "num_tokens": 2866984.0,
      "step": 1555
    },
    {
      "entropy": 6.389467811584472,
      "epoch": 0.13106490233144297,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999750909795256,
      "loss": 6.2012,
      "mean_token_accuracy": 0.12793270647525787,
      "num_tokens": 2876550.0,
      "step": 1560
    },
    {
      "entropy": 6.517248106002808,
      "epoch": 0.13148498214660786,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999746433954394,
      "loss": 6.3019,
      "mean_token_accuracy": 0.12286014333367348,
      "num_tokens": 2885782.0,
      "step": 1565
    },
    {
      "entropy": 6.436316442489624,
      "epoch": 0.13190506196177273,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499974191826043,
      "loss": 6.2806,
      "mean_token_accuracy": 0.13175936564803123,
      "num_tokens": 2894807.0,
      "step": 1570
    },
    {
      "entropy": 6.556414270401001,
      "epoch": 0.1323251417769376,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999737362713448,
      "loss": 6.3319,
      "mean_token_accuracy": 0.12141326144337654,
      "num_tokens": 2904076.0,
      "step": 1575
    },
    {
      "entropy": 6.438077783584594,
      "epoch": 0.1327452215921025,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999732767313527,
      "loss": 6.2301,
      "mean_token_accuracy": 0.12145909070968627,
      "num_tokens": 2913761.0,
      "step": 1580
    },
    {
      "entropy": 6.605003738403321,
      "epoch": 0.13316530140726737,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999728132060746,
      "loss": 6.4615,
      "mean_token_accuracy": 0.11997142881155014,
      "num_tokens": 2922848.0,
      "step": 1585
    },
    {
      "entropy": 6.545976686477661,
      "epoch": 0.13358538122243227,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999723456955192,
      "loss": 6.3361,
      "mean_token_accuracy": 0.11663061752915382,
      "num_tokens": 2932718.0,
      "step": 1590
    },
    {
      "entropy": 6.370933151245117,
      "epoch": 0.13400546103759714,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999718741996945,
      "loss": 6.2878,
      "mean_token_accuracy": 0.12183845415711403,
      "num_tokens": 2942686.0,
      "step": 1595
    },
    {
      "entropy": 6.422312498092651,
      "epoch": 0.13442554085276204,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499971398718609,
      "loss": 6.2711,
      "mean_token_accuracy": 0.11760244071483612,
      "num_tokens": 2952096.0,
      "step": 1600
    },
    {
      "entropy": 6.4970005512237545,
      "epoch": 0.1348456206679269,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999709192522708,
      "loss": 6.3449,
      "mean_token_accuracy": 0.12237487882375717,
      "num_tokens": 2960660.0,
      "step": 1605
    },
    {
      "entropy": 6.604963493347168,
      "epoch": 0.13526570048309178,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999704358006887,
      "loss": 6.3497,
      "mean_token_accuracy": 0.11823863610625267,
      "num_tokens": 2969834.0,
      "step": 1610
    },
    {
      "entropy": 6.454269361495972,
      "epoch": 0.13568578029825668,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999699483638712,
      "loss": 6.3256,
      "mean_token_accuracy": 0.12492895498871803,
      "num_tokens": 2979023.0,
      "step": 1615
    },
    {
      "entropy": 6.496698570251465,
      "epoch": 0.13610586011342155,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999694569418269,
      "loss": 6.3222,
      "mean_token_accuracy": 0.12362956255674362,
      "num_tokens": 2988083.0,
      "step": 1620
    },
    {
      "entropy": 6.424844694137573,
      "epoch": 0.13652593992858644,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999689615345645,
      "loss": 6.2399,
      "mean_token_accuracy": 0.12576239481568335,
      "num_tokens": 2997240.0,
      "step": 1625
    },
    {
      "entropy": 6.554689359664917,
      "epoch": 0.1369460197437513,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999684621420928,
      "loss": 6.3191,
      "mean_token_accuracy": 0.12176118865609169,
      "num_tokens": 3007077.0,
      "step": 1630
    },
    {
      "entropy": 6.500646114349365,
      "epoch": 0.13736609955891618,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999679587644205,
      "loss": 6.3628,
      "mean_token_accuracy": 0.11905800476670265,
      "num_tokens": 3015821.0,
      "step": 1635
    },
    {
      "entropy": 6.445851802825928,
      "epoch": 0.13778617937408108,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999674514015568,
      "loss": 6.2803,
      "mean_token_accuracy": 0.12496163100004196,
      "num_tokens": 3025858.0,
      "step": 1640
    },
    {
      "entropy": 6.520554733276367,
      "epoch": 0.13820625918924595,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999669400535105,
      "loss": 6.2466,
      "mean_token_accuracy": 0.12024016380310058,
      "num_tokens": 3035537.0,
      "step": 1645
    },
    {
      "entropy": 6.2957807064056395,
      "epoch": 0.13862633900441085,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999664247202907,
      "loss": 6.1784,
      "mean_token_accuracy": 0.12314105778932571,
      "num_tokens": 3044204.0,
      "step": 1650
    },
    {
      "entropy": 6.51614408493042,
      "epoch": 0.13904641881957572,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999659054019066,
      "loss": 6.3301,
      "mean_token_accuracy": 0.1245350755751133,
      "num_tokens": 3053111.0,
      "step": 1655
    },
    {
      "entropy": 6.3797486305236815,
      "epoch": 0.1394664986347406,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999653820983673,
      "loss": 6.2542,
      "mean_token_accuracy": 0.12539379447698593,
      "num_tokens": 3062456.0,
      "step": 1660
    },
    {
      "entropy": 6.448072814941407,
      "epoch": 0.13988657844990549,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499964854809682,
      "loss": 6.2791,
      "mean_token_accuracy": 0.1257368341088295,
      "num_tokens": 3071132.0,
      "step": 1665
    },
    {
      "entropy": 6.44848141670227,
      "epoch": 0.14030665826507036,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999643235358602,
      "loss": 6.2275,
      "mean_token_accuracy": 0.13099895641207696,
      "num_tokens": 3080892.0,
      "step": 1670
    },
    {
      "entropy": 6.366451406478882,
      "epoch": 0.14072673808023525,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999637882769112,
      "loss": 6.1829,
      "mean_token_accuracy": 0.12473593726754188,
      "num_tokens": 3089874.0,
      "step": 1675
    },
    {
      "entropy": 6.438631200790406,
      "epoch": 0.14114681789540012,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999632490328447,
      "loss": 6.3097,
      "mean_token_accuracy": 0.12419775351881981,
      "num_tokens": 3099535.0,
      "step": 1680
    },
    {
      "entropy": 6.457270622253418,
      "epoch": 0.14156689771056502,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999627058036699,
      "loss": 6.2762,
      "mean_token_accuracy": 0.12024275660514831,
      "num_tokens": 3108772.0,
      "step": 1685
    },
    {
      "entropy": 6.486367511749267,
      "epoch": 0.1419869775257299,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999621585893966,
      "loss": 6.2888,
      "mean_token_accuracy": 0.11725226789712906,
      "num_tokens": 3118333.0,
      "step": 1690
    },
    {
      "entropy": 6.429450368881225,
      "epoch": 0.14240705734089476,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999616073900346,
      "loss": 6.3284,
      "mean_token_accuracy": 0.12171539664268494,
      "num_tokens": 3127356.0,
      "step": 1695
    },
    {
      "entropy": 6.498106622695923,
      "epoch": 0.14282713715605966,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999610522055935,
      "loss": 6.2912,
      "mean_token_accuracy": 0.11646916791796684,
      "num_tokens": 3136859.0,
      "step": 1700
    },
    {
      "entropy": 6.471587181091309,
      "epoch": 0.14324721697122453,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999604930360832,
      "loss": 6.33,
      "mean_token_accuracy": 0.11686689779162407,
      "num_tokens": 3146607.0,
      "step": 1705
    },
    {
      "entropy": 6.408087730407715,
      "epoch": 0.14366729678638943,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999599298815136,
      "loss": 6.2621,
      "mean_token_accuracy": 0.1279582552611828,
      "num_tokens": 3156327.0,
      "step": 1710
    },
    {
      "entropy": 6.365849542617798,
      "epoch": 0.1440873766015543,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004999593627418947,
      "loss": 6.1983,
      "mean_token_accuracy": 0.13225716799497605,
      "num_tokens": 3165559.0,
      "step": 1715
    },
    {
      "entropy": 6.51579737663269,
      "epoch": 0.14450745641671917,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999587916172365,
      "loss": 6.2988,
      "mean_token_accuracy": 0.11835923567414283,
      "num_tokens": 3173850.0,
      "step": 1720
    },
    {
      "entropy": 6.430272150039673,
      "epoch": 0.14492753623188406,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999582165075492,
      "loss": 6.2413,
      "mean_token_accuracy": 0.12055112421512604,
      "num_tokens": 3182838.0,
      "step": 1725
    },
    {
      "entropy": 6.354161643981934,
      "epoch": 0.14534761604704893,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999576374128429,
      "loss": 6.2389,
      "mean_token_accuracy": 0.1231518879532814,
      "num_tokens": 3191692.0,
      "step": 1730
    },
    {
      "entropy": 6.491216516494751,
      "epoch": 0.14576769586221383,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999570543331279,
      "loss": 6.2699,
      "mean_token_accuracy": 0.1253180719912052,
      "num_tokens": 3200069.0,
      "step": 1735
    },
    {
      "entropy": 6.4589770317077635,
      "epoch": 0.1461877756773787,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999564672684145,
      "loss": 6.3418,
      "mean_token_accuracy": 0.11855114102363587,
      "num_tokens": 3209653.0,
      "step": 1740
    },
    {
      "entropy": 6.512481784820556,
      "epoch": 0.14660785549254357,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999558762187131,
      "loss": 6.2232,
      "mean_token_accuracy": 0.13238132372498512,
      "num_tokens": 3218313.0,
      "step": 1745
    },
    {
      "entropy": 6.309611749649048,
      "epoch": 0.14702793530770847,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999552811840342,
      "loss": 6.1676,
      "mean_token_accuracy": 0.12678530514240266,
      "num_tokens": 3227525.0,
      "step": 1750
    },
    {
      "entropy": 6.419345808029175,
      "epoch": 0.14744801512287334,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999546821643884,
      "loss": 6.2824,
      "mean_token_accuracy": 0.12407411858439446,
      "num_tokens": 3237022.0,
      "step": 1755
    },
    {
      "entropy": 6.376602506637573,
      "epoch": 0.14786809493803824,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999540791597861,
      "loss": 6.1608,
      "mean_token_accuracy": 0.1266619324684143,
      "num_tokens": 3246605.0,
      "step": 1760
    },
    {
      "entropy": 6.292014646530151,
      "epoch": 0.1482881747532031,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999534721702383,
      "loss": 6.139,
      "mean_token_accuracy": 0.13051704913377762,
      "num_tokens": 3255587.0,
      "step": 1765
    },
    {
      "entropy": 6.414491844177246,
      "epoch": 0.148708254568368,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999528611957553,
      "loss": 6.2196,
      "mean_token_accuracy": 0.1258571393787861,
      "num_tokens": 3265669.0,
      "step": 1770
    },
    {
      "entropy": 6.458701848983765,
      "epoch": 0.14912833438353287,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999522462363485,
      "loss": 6.2058,
      "mean_token_accuracy": 0.12836851552128792,
      "num_tokens": 3275013.0,
      "step": 1775
    },
    {
      "entropy": 6.390926885604858,
      "epoch": 0.14954841419869774,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999516272920283,
      "loss": 6.3077,
      "mean_token_accuracy": 0.12579302862286568,
      "num_tokens": 3284723.0,
      "step": 1780
    },
    {
      "entropy": 6.3027690887451175,
      "epoch": 0.14996849401386264,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499951004362806,
      "loss": 6.1375,
      "mean_token_accuracy": 0.13238742649555207,
      "num_tokens": 3293860.0,
      "step": 1785
    },
    {
      "entropy": 6.337384557723999,
      "epoch": 0.1503885738290275,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999503774486924,
      "loss": 6.1876,
      "mean_token_accuracy": 0.1245276965200901,
      "num_tokens": 3303158.0,
      "step": 1790
    },
    {
      "entropy": 6.296294498443603,
      "epoch": 0.1508086536441924,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999497465496987,
      "loss": 6.1254,
      "mean_token_accuracy": 0.12188991010189057,
      "num_tokens": 3313068.0,
      "step": 1795
    },
    {
      "entropy": 6.346179628372193,
      "epoch": 0.15122873345935728,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499949111665836,
      "loss": 6.21,
      "mean_token_accuracy": 0.12482204511761666,
      "num_tokens": 3321885.0,
      "step": 1800
    },
    {
      "entropy": 6.417145109176635,
      "epoch": 0.15164881327452215,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999484727971158,
      "loss": 6.2049,
      "mean_token_accuracy": 0.12663445696234704,
      "num_tokens": 3330924.0,
      "step": 1805
    },
    {
      "entropy": 6.3640861988067625,
      "epoch": 0.15206889308968705,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499947829943549,
      "loss": 6.2421,
      "mean_token_accuracy": 0.12180857658386231,
      "num_tokens": 3340070.0,
      "step": 1810
    },
    {
      "entropy": 6.42901873588562,
      "epoch": 0.15248897290485192,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999471831051474,
      "loss": 6.2161,
      "mean_token_accuracy": 0.13666835352778434,
      "num_tokens": 3349870.0,
      "step": 1815
    },
    {
      "entropy": 6.384167289733886,
      "epoch": 0.1529090527200168,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999465322819222,
      "loss": 6.2561,
      "mean_token_accuracy": 0.1196154922246933,
      "num_tokens": 3359573.0,
      "step": 1820
    },
    {
      "entropy": 6.420805263519287,
      "epoch": 0.15332913253518168,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999458774738851,
      "loss": 6.2022,
      "mean_token_accuracy": 0.1329035244882107,
      "num_tokens": 3368577.0,
      "step": 1825
    },
    {
      "entropy": 6.3645374298095705,
      "epoch": 0.15374921235034655,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999452186810476,
      "loss": 6.1723,
      "mean_token_accuracy": 0.13094222992658616,
      "num_tokens": 3377801.0,
      "step": 1830
    },
    {
      "entropy": 6.414077711105347,
      "epoch": 0.15416929216551145,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999445559034214,
      "loss": 6.2325,
      "mean_token_accuracy": 0.12632885575294495,
      "num_tokens": 3386666.0,
      "step": 1835
    },
    {
      "entropy": 6.492283153533935,
      "epoch": 0.15458937198067632,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999438891410181,
      "loss": 6.3646,
      "mean_token_accuracy": 0.12365243062376977,
      "num_tokens": 3396086.0,
      "step": 1840
    },
    {
      "entropy": 6.3361475467681885,
      "epoch": 0.15500945179584122,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999432183938496,
      "loss": 6.2714,
      "mean_token_accuracy": 0.12596381455659866,
      "num_tokens": 3404894.0,
      "step": 1845
    },
    {
      "entropy": 6.407006740570068,
      "epoch": 0.1554295316110061,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999425436619279,
      "loss": 6.2544,
      "mean_token_accuracy": 0.12557967007160187,
      "num_tokens": 3414172.0,
      "step": 1850
    },
    {
      "entropy": 6.476001453399658,
      "epoch": 0.15584961142617096,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499941864945265,
      "loss": 6.2327,
      "mean_token_accuracy": 0.12348362207412719,
      "num_tokens": 3423409.0,
      "step": 1855
    },
    {
      "entropy": 6.296181011199951,
      "epoch": 0.15626969124133586,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999411822438726,
      "loss": 6.1767,
      "mean_token_accuracy": 0.12874680012464523,
      "num_tokens": 3433047.0,
      "step": 1860
    },
    {
      "entropy": 6.446085119247437,
      "epoch": 0.15668977105650073,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499940495557763,
      "loss": 6.1762,
      "mean_token_accuracy": 0.12680488154292108,
      "num_tokens": 3442490.0,
      "step": 1865
    },
    {
      "entropy": 6.367494440078735,
      "epoch": 0.15710985087166562,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999398048869485,
      "loss": 6.2463,
      "mean_token_accuracy": 0.1263427644968033,
      "num_tokens": 3451804.0,
      "step": 1870
    },
    {
      "entropy": 6.448274326324463,
      "epoch": 0.1575299306868305,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499939110231441,
      "loss": 6.225,
      "mean_token_accuracy": 0.12612806633114815,
      "num_tokens": 3461481.0,
      "step": 1875
    },
    {
      "entropy": 6.3886480808258055,
      "epoch": 0.1579500105019954,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999384115912531,
      "loss": 6.2713,
      "mean_token_accuracy": 0.12860373705625533,
      "num_tokens": 3471798.0,
      "step": 1880
    },
    {
      "entropy": 6.294258403778076,
      "epoch": 0.15837009031716026,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499937708966397,
      "loss": 6.1694,
      "mean_token_accuracy": 0.1269647814333439,
      "num_tokens": 3481386.0,
      "step": 1885
    },
    {
      "entropy": 6.380875062942505,
      "epoch": 0.15879017013232513,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999370023568853,
      "loss": 6.1648,
      "mean_token_accuracy": 0.1258099302649498,
      "num_tokens": 3489981.0,
      "step": 1890
    },
    {
      "entropy": 6.298674488067627,
      "epoch": 0.15921024994749003,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999362917627304,
      "loss": 6.1603,
      "mean_token_accuracy": 0.1279548816382885,
      "num_tokens": 3498551.0,
      "step": 1895
    },
    {
      "entropy": 6.364571666717529,
      "epoch": 0.1596303297626549,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999355771839448,
      "loss": 6.1198,
      "mean_token_accuracy": 0.12712288051843643,
      "num_tokens": 3507921.0,
      "step": 1900
    },
    {
      "entropy": 6.472431182861328,
      "epoch": 0.1600504095778198,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999348586205414,
      "loss": 6.3119,
      "mean_token_accuracy": 0.12473546266555786,
      "num_tokens": 3517570.0,
      "step": 1905
    },
    {
      "entropy": 6.44771819114685,
      "epoch": 0.16047048939298467,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999341360725327,
      "loss": 6.2778,
      "mean_token_accuracy": 0.12144239619374275,
      "num_tokens": 3526774.0,
      "step": 1910
    },
    {
      "entropy": 6.367665719985962,
      "epoch": 0.16089056920814954,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999334095399317,
      "loss": 6.2164,
      "mean_token_accuracy": 0.13476391285657882,
      "num_tokens": 3535319.0,
      "step": 1915
    },
    {
      "entropy": 6.29778151512146,
      "epoch": 0.16131064902331443,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999326790227512,
      "loss": 6.1759,
      "mean_token_accuracy": 0.12915017306804658,
      "num_tokens": 3544468.0,
      "step": 1920
    },
    {
      "entropy": 6.263013315200806,
      "epoch": 0.1617307288384793,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999319445210041,
      "loss": 6.074,
      "mean_token_accuracy": 0.13183464705944062,
      "num_tokens": 3553529.0,
      "step": 1925
    },
    {
      "entropy": 6.341449880599976,
      "epoch": 0.1621508086536442,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999312060347034,
      "loss": 6.1214,
      "mean_token_accuracy": 0.1265263617038727,
      "num_tokens": 3563053.0,
      "step": 1930
    },
    {
      "entropy": 6.272969150543213,
      "epoch": 0.16257088846880907,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999304635638621,
      "loss": 6.0616,
      "mean_token_accuracy": 0.13436255902051925,
      "num_tokens": 3571877.0,
      "step": 1935
    },
    {
      "entropy": 6.305766344070435,
      "epoch": 0.16299096828397394,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999297171084935,
      "loss": 6.1234,
      "mean_token_accuracy": 0.1326524868607521,
      "num_tokens": 3581496.0,
      "step": 1940
    },
    {
      "entropy": 6.313105726242066,
      "epoch": 0.16341104809913884,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999289666686109,
      "loss": 6.1344,
      "mean_token_accuracy": 0.1272105999290943,
      "num_tokens": 3590752.0,
      "step": 1945
    },
    {
      "entropy": 6.1983654499053955,
      "epoch": 0.1638311279143037,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999282122442274,
      "loss": 6.1416,
      "mean_token_accuracy": 0.13213918060064317,
      "num_tokens": 3599885.0,
      "step": 1950
    },
    {
      "entropy": 6.440747499465942,
      "epoch": 0.1642512077294686,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999274538353564,
      "loss": 6.2106,
      "mean_token_accuracy": 0.12646706998348237,
      "num_tokens": 3610039.0,
      "step": 1955
    },
    {
      "entropy": 6.272910118103027,
      "epoch": 0.16467128754463348,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999266914420114,
      "loss": 6.1332,
      "mean_token_accuracy": 0.12781383618712425,
      "num_tokens": 3619954.0,
      "step": 1960
    },
    {
      "entropy": 6.3279284000396725,
      "epoch": 0.16509136735979837,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499925925064206,
      "loss": 6.1033,
      "mean_token_accuracy": 0.13286505788564681,
      "num_tokens": 3628164.0,
      "step": 1965
    },
    {
      "entropy": 6.360192775726318,
      "epoch": 0.16551144717496324,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999251547019535,
      "loss": 6.2577,
      "mean_token_accuracy": 0.12934610545635222,
      "num_tokens": 3636778.0,
      "step": 1970
    },
    {
      "entropy": 6.414323568344116,
      "epoch": 0.16593152699012811,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999243803552678,
      "loss": 6.197,
      "mean_token_accuracy": 0.1297714054584503,
      "num_tokens": 3647046.0,
      "step": 1975
    },
    {
      "entropy": 6.239211511611939,
      "epoch": 0.166351606805293,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999236020241625,
      "loss": 6.1263,
      "mean_token_accuracy": 0.12911003157496453,
      "num_tokens": 3656130.0,
      "step": 1980
    },
    {
      "entropy": 6.400930309295655,
      "epoch": 0.16677168662045788,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999228197086514,
      "loss": 6.2075,
      "mean_token_accuracy": 0.12356689870357514,
      "num_tokens": 3666145.0,
      "step": 1985
    },
    {
      "entropy": 6.2993443489074705,
      "epoch": 0.16719176643562278,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999220334087484,
      "loss": 6.2466,
      "mean_token_accuracy": 0.12334830835461616,
      "num_tokens": 3676722.0,
      "step": 1990
    },
    {
      "entropy": 6.383305788040161,
      "epoch": 0.16761184625078765,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999212431244673,
      "loss": 6.2353,
      "mean_token_accuracy": 0.12262691259384155,
      "num_tokens": 3685880.0,
      "step": 1995
    },
    {
      "entropy": 6.260931539535522,
      "epoch": 0.16803192606595252,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999204488558222,
      "loss": 6.064,
      "mean_token_accuracy": 0.13560437187552452,
      "num_tokens": 3695167.0,
      "step": 2000
    },
    {
      "entropy": 6.341617202758789,
      "epoch": 0.16845200588111742,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999196506028273,
      "loss": 6.1714,
      "mean_token_accuracy": 0.13033221215009688,
      "num_tokens": 3703700.0,
      "step": 2005
    },
    {
      "entropy": 6.330134630203247,
      "epoch": 0.1688720856962823,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999188483654965,
      "loss": 6.1229,
      "mean_token_accuracy": 0.1314982570707798,
      "num_tokens": 3712825.0,
      "step": 2010
    },
    {
      "entropy": 6.2663736820220945,
      "epoch": 0.16929216551144718,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999180421438442,
      "loss": 6.0855,
      "mean_token_accuracy": 0.13337414115667343,
      "num_tokens": 3721807.0,
      "step": 2015
    },
    {
      "entropy": 6.426154756546021,
      "epoch": 0.16971224532661205,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999172319378846,
      "loss": 6.2558,
      "mean_token_accuracy": 0.12170653715729714,
      "num_tokens": 3730502.0,
      "step": 2020
    },
    {
      "entropy": 6.357564783096313,
      "epoch": 0.17013232514177692,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999164177476319,
      "loss": 6.1304,
      "mean_token_accuracy": 0.13129208087921143,
      "num_tokens": 3739696.0,
      "step": 2025
    },
    {
      "entropy": 6.183953857421875,
      "epoch": 0.17055240495694182,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999155995731009,
      "loss": 6.1493,
      "mean_token_accuracy": 0.1318361982703209,
      "num_tokens": 3748675.0,
      "step": 2030
    },
    {
      "entropy": 6.50171127319336,
      "epoch": 0.1709724847721067,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999147774143057,
      "loss": 6.2267,
      "mean_token_accuracy": 0.12189607992768288,
      "num_tokens": 3757714.0,
      "step": 2035
    },
    {
      "entropy": 6.228915214538574,
      "epoch": 0.1713925645872716,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499913951271261,
      "loss": 6.0545,
      "mean_token_accuracy": 0.13067708387970925,
      "num_tokens": 3767589.0,
      "step": 2040
    },
    {
      "entropy": 6.328529357910156,
      "epoch": 0.17181264440243646,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999131211439816,
      "loss": 6.1523,
      "mean_token_accuracy": 0.13088449314236641,
      "num_tokens": 3777261.0,
      "step": 2045
    },
    {
      "entropy": 6.358458805084228,
      "epoch": 0.17223272421760136,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499912287032482,
      "loss": 6.105,
      "mean_token_accuracy": 0.13954608961939813,
      "num_tokens": 3786658.0,
      "step": 2050
    },
    {
      "entropy": 6.194888353347778,
      "epoch": 0.17265280403276623,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499911448936777,
      "loss": 6.1021,
      "mean_token_accuracy": 0.1387820027768612,
      "num_tokens": 3794977.0,
      "step": 2055
    },
    {
      "entropy": 6.274205017089844,
      "epoch": 0.1730728838479311,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999106068568816,
      "loss": 6.1767,
      "mean_token_accuracy": 0.1296847067773342,
      "num_tokens": 3805138.0,
      "step": 2060
    },
    {
      "entropy": 6.353550148010254,
      "epoch": 0.173492963663096,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999097607928106,
      "loss": 6.1193,
      "mean_token_accuracy": 0.13772593289613724,
      "num_tokens": 3814444.0,
      "step": 2065
    },
    {
      "entropy": 6.28287091255188,
      "epoch": 0.17391304347826086,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999089107445788,
      "loss": 6.081,
      "mean_token_accuracy": 0.1331343598663807,
      "num_tokens": 3822859.0,
      "step": 2070
    },
    {
      "entropy": 6.187959289550781,
      "epoch": 0.17433312329342576,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999080567122016,
      "loss": 6.0928,
      "mean_token_accuracy": 0.13325781002640724,
      "num_tokens": 3833159.0,
      "step": 2075
    },
    {
      "entropy": 6.2891417980194095,
      "epoch": 0.17475320310859063,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999071986956941,
      "loss": 6.1081,
      "mean_token_accuracy": 0.13443495854735374,
      "num_tokens": 3842136.0,
      "step": 2080
    },
    {
      "entropy": 6.291293048858643,
      "epoch": 0.1751732829237555,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999063366950713,
      "loss": 6.1855,
      "mean_token_accuracy": 0.1292753279209137,
      "num_tokens": 3851406.0,
      "step": 2085
    },
    {
      "entropy": 6.308997011184692,
      "epoch": 0.1755933627389204,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999054707103486,
      "loss": 6.0941,
      "mean_token_accuracy": 0.1310033306479454,
      "num_tokens": 3861061.0,
      "step": 2090
    },
    {
      "entropy": 6.280231618881226,
      "epoch": 0.17601344255408527,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999046007415412,
      "loss": 6.0761,
      "mean_token_accuracy": 0.1290605030953884,
      "num_tokens": 3870357.0,
      "step": 2095
    },
    {
      "entropy": 6.3239013671875,
      "epoch": 0.17643352236925017,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999037267886646,
      "loss": 6.1119,
      "mean_token_accuracy": 0.1301138088107109,
      "num_tokens": 3879393.0,
      "step": 2100
    },
    {
      "entropy": 6.217884635925293,
      "epoch": 0.17685360218441504,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999028488517343,
      "loss": 6.1106,
      "mean_token_accuracy": 0.13041476532816887,
      "num_tokens": 3888030.0,
      "step": 2105
    },
    {
      "entropy": 6.330727005004883,
      "epoch": 0.1772736819995799,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999019669307659,
      "loss": 6.1128,
      "mean_token_accuracy": 0.14087287336587906,
      "num_tokens": 3897430.0,
      "step": 2110
    },
    {
      "entropy": 6.268360662460327,
      "epoch": 0.1776937618147448,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999010810257749,
      "loss": 6.1294,
      "mean_token_accuracy": 0.12782176434993744,
      "num_tokens": 3907711.0,
      "step": 2115
    },
    {
      "entropy": 6.222604417800904,
      "epoch": 0.17811384162990967,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999001911367771,
      "loss": 6.0703,
      "mean_token_accuracy": 0.13519175425171853,
      "num_tokens": 3915816.0,
      "step": 2120
    },
    {
      "entropy": 6.279330682754517,
      "epoch": 0.17853392144507457,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998992972637883,
      "loss": 6.1784,
      "mean_token_accuracy": 0.12573839649558066,
      "num_tokens": 3925162.0,
      "step": 2125
    },
    {
      "entropy": 6.312918043136596,
      "epoch": 0.17895400126023944,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998983994068242,
      "loss": 6.0804,
      "mean_token_accuracy": 0.12800241634249687,
      "num_tokens": 3934476.0,
      "step": 2130
    },
    {
      "entropy": 6.257489347457886,
      "epoch": 0.17937408107540434,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998974975659006,
      "loss": 6.117,
      "mean_token_accuracy": 0.13179219737648964,
      "num_tokens": 3943501.0,
      "step": 2135
    },
    {
      "entropy": 6.269851350784302,
      "epoch": 0.1797941608905692,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998965917410338,
      "loss": 6.105,
      "mean_token_accuracy": 0.12811433225870134,
      "num_tokens": 3953663.0,
      "step": 2140
    },
    {
      "entropy": 6.263305425643921,
      "epoch": 0.18021424070573408,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998956819322397,
      "loss": 6.0649,
      "mean_token_accuracy": 0.1344923309981823,
      "num_tokens": 3962634.0,
      "step": 2145
    },
    {
      "entropy": 6.2131294250488285,
      "epoch": 0.18063432052089898,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998947681395343,
      "loss": 6.0787,
      "mean_token_accuracy": 0.1363943800330162,
      "num_tokens": 3972496.0,
      "step": 2150
    },
    {
      "entropy": 6.464788103103638,
      "epoch": 0.18105440033606385,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499893850362934,
      "loss": 6.3219,
      "mean_token_accuracy": 0.12087765410542488,
      "num_tokens": 3980724.0,
      "step": 2155
    },
    {
      "entropy": 6.291199588775635,
      "epoch": 0.18147448015122875,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998929286024548,
      "loss": 6.1445,
      "mean_token_accuracy": 0.12979334518313407,
      "num_tokens": 3989842.0,
      "step": 2160
    },
    {
      "entropy": 6.278742074966431,
      "epoch": 0.18189455996639362,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998920028581133,
      "loss": 6.0687,
      "mean_token_accuracy": 0.14321179836988449,
      "num_tokens": 3998534.0,
      "step": 2165
    },
    {
      "entropy": 6.2926966667175295,
      "epoch": 0.18231463978155849,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998910731299258,
      "loss": 6.0834,
      "mean_token_accuracy": 0.12944981455802917,
      "num_tokens": 4007677.0,
      "step": 2170
    },
    {
      "entropy": 6.248470973968506,
      "epoch": 0.18273471959672338,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998901394179085,
      "loss": 6.1367,
      "mean_token_accuracy": 0.12592731565237045,
      "num_tokens": 4016347.0,
      "step": 2175
    },
    {
      "entropy": 6.264119005203247,
      "epoch": 0.18315479941188825,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998892017220784,
      "loss": 6.0249,
      "mean_token_accuracy": 0.13742422610521315,
      "num_tokens": 4025199.0,
      "step": 2180
    },
    {
      "entropy": 6.263695335388183,
      "epoch": 0.18357487922705315,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998882600424519,
      "loss": 6.0944,
      "mean_token_accuracy": 0.12306004837155342,
      "num_tokens": 4033933.0,
      "step": 2185
    },
    {
      "entropy": 6.230836820602417,
      "epoch": 0.18399495904221802,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998873143790455,
      "loss": 6.0121,
      "mean_token_accuracy": 0.13827981501817704,
      "num_tokens": 4042891.0,
      "step": 2190
    },
    {
      "entropy": 6.326285457611084,
      "epoch": 0.1844150388573829,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998863647318763,
      "loss": 6.1254,
      "mean_token_accuracy": 0.13076284304261207,
      "num_tokens": 4051123.0,
      "step": 2195
    },
    {
      "entropy": 6.218794679641723,
      "epoch": 0.1848351186725478,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998854111009608,
      "loss": 6.1018,
      "mean_token_accuracy": 0.12811416238546372,
      "num_tokens": 4060025.0,
      "step": 2200
    },
    {
      "entropy": 6.201221132278443,
      "epoch": 0.18525519848771266,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998844534863161,
      "loss": 6.0147,
      "mean_token_accuracy": 0.12775525376200675,
      "num_tokens": 4069363.0,
      "step": 2205
    },
    {
      "entropy": 6.318734121322632,
      "epoch": 0.18567527830287756,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998834918879592,
      "loss": 6.1746,
      "mean_token_accuracy": 0.13352699503302573,
      "num_tokens": 4078855.0,
      "step": 2210
    },
    {
      "entropy": 6.284113597869873,
      "epoch": 0.18609535811804243,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499882526305907,
      "loss": 6.1205,
      "mean_token_accuracy": 0.13283827975392343,
      "num_tokens": 4087801.0,
      "step": 2215
    },
    {
      "entropy": 6.278976345062256,
      "epoch": 0.18651543793320732,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998815567401765,
      "loss": 6.1183,
      "mean_token_accuracy": 0.13487844318151473,
      "num_tokens": 4096949.0,
      "step": 2220
    },
    {
      "entropy": 6.322417879104615,
      "epoch": 0.1869355177483722,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998805831907851,
      "loss": 6.092,
      "mean_token_accuracy": 0.12811465859413146,
      "num_tokens": 4105399.0,
      "step": 2225
    },
    {
      "entropy": 6.261762046813965,
      "epoch": 0.18735559756353706,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998796056577501,
      "loss": 6.0294,
      "mean_token_accuracy": 0.12984329760074614,
      "num_tokens": 4113873.0,
      "step": 2230
    },
    {
      "entropy": 6.17222752571106,
      "epoch": 0.18777567737870196,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998786241410886,
      "loss": 6.0911,
      "mean_token_accuracy": 0.13395953848958014,
      "num_tokens": 4123528.0,
      "step": 2235
    },
    {
      "entropy": 6.337759065628052,
      "epoch": 0.18819575719386683,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499877638640818,
      "loss": 6.1045,
      "mean_token_accuracy": 0.12952830642461777,
      "num_tokens": 4133370.0,
      "step": 2240
    },
    {
      "entropy": 6.218608713150024,
      "epoch": 0.18861583700903173,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499876649156956,
      "loss": 6.027,
      "mean_token_accuracy": 0.13578148186206818,
      "num_tokens": 4142370.0,
      "step": 2245
    },
    {
      "entropy": 6.224829864501953,
      "epoch": 0.1890359168241966,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998756556895196,
      "loss": 6.1118,
      "mean_token_accuracy": 0.13433386385440826,
      "num_tokens": 4152367.0,
      "step": 2250
    },
    {
      "entropy": 6.281910085678101,
      "epoch": 0.18945599663936147,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499874658238527,
      "loss": 6.0938,
      "mean_token_accuracy": 0.134437657892704,
      "num_tokens": 4161126.0,
      "step": 2255
    },
    {
      "entropy": 6.240273094177246,
      "epoch": 0.18987607645452637,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998736568039957,
      "loss": 5.9943,
      "mean_token_accuracy": 0.13461074903607367,
      "num_tokens": 4169910.0,
      "step": 2260
    },
    {
      "entropy": 6.2365155696868895,
      "epoch": 0.19029615626969124,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998726513859432,
      "loss": 6.1408,
      "mean_token_accuracy": 0.1260552354156971,
      "num_tokens": 4179893.0,
      "step": 2265
    },
    {
      "entropy": 6.301045608520508,
      "epoch": 0.19071623608485613,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998716419843875,
      "loss": 6.1429,
      "mean_token_accuracy": 0.13841021955013275,
      "num_tokens": 4190065.0,
      "step": 2270
    },
    {
      "entropy": 6.13691897392273,
      "epoch": 0.191136315900021,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998706285993465,
      "loss": 6.0687,
      "mean_token_accuracy": 0.13612305223941804,
      "num_tokens": 4198395.0,
      "step": 2275
    },
    {
      "entropy": 6.308142518997192,
      "epoch": 0.19155639571518587,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998696112308381,
      "loss": 6.069,
      "mean_token_accuracy": 0.13221055194735526,
      "num_tokens": 4207555.0,
      "step": 2280
    },
    {
      "entropy": 6.1573583602905275,
      "epoch": 0.19197647553035077,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998685898788803,
      "loss": 6.0272,
      "mean_token_accuracy": 0.13469478338956833,
      "num_tokens": 4216533.0,
      "step": 2285
    },
    {
      "entropy": 6.334969472885132,
      "epoch": 0.19239655534551564,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998675645434914,
      "loss": 6.1333,
      "mean_token_accuracy": 0.134019573032856,
      "num_tokens": 4225575.0,
      "step": 2290
    },
    {
      "entropy": 6.167663669586181,
      "epoch": 0.19281663516068054,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998665352246891,
      "loss": 5.9307,
      "mean_token_accuracy": 0.14072284027934073,
      "num_tokens": 4234306.0,
      "step": 2295
    },
    {
      "entropy": 6.1863524436950685,
      "epoch": 0.1932367149758454,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998655019224921,
      "loss": 6.1102,
      "mean_token_accuracy": 0.13416126370429993,
      "num_tokens": 4243998.0,
      "step": 2300
    },
    {
      "entropy": 6.274348163604737,
      "epoch": 0.19365679479101028,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998644646369185,
      "loss": 6.0024,
      "mean_token_accuracy": 0.13099471628665924,
      "num_tokens": 4253653.0,
      "step": 2305
    },
    {
      "entropy": 6.157945919036865,
      "epoch": 0.19407687460617518,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998634233679865,
      "loss": 6.0922,
      "mean_token_accuracy": 0.12541876286268233,
      "num_tokens": 4263305.0,
      "step": 2310
    },
    {
      "entropy": 6.220145130157471,
      "epoch": 0.19449695442134005,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499862378115715,
      "loss": 5.9623,
      "mean_token_accuracy": 0.13993363454937935,
      "num_tokens": 4272212.0,
      "step": 2315
    },
    {
      "entropy": 6.2961366176605225,
      "epoch": 0.19491703423650494,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998613288801221,
      "loss": 6.1759,
      "mean_token_accuracy": 0.12927026972174643,
      "num_tokens": 4281445.0,
      "step": 2320
    },
    {
      "entropy": 6.306110906600952,
      "epoch": 0.1953371140516698,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998602756612267,
      "loss": 6.0867,
      "mean_token_accuracy": 0.13294345438480376,
      "num_tokens": 4290938.0,
      "step": 2325
    },
    {
      "entropy": 6.204567623138428,
      "epoch": 0.1957571938668347,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998592184590471,
      "loss": 6.1218,
      "mean_token_accuracy": 0.13037385493516923,
      "num_tokens": 4300022.0,
      "step": 2330
    },
    {
      "entropy": 6.182180976867675,
      "epoch": 0.19617727368199958,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998581572736024,
      "loss": 6.0055,
      "mean_token_accuracy": 0.13552708327770233,
      "num_tokens": 4308910.0,
      "step": 2335
    },
    {
      "entropy": 6.136313104629517,
      "epoch": 0.19659735349716445,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998570921049112,
      "loss": 5.982,
      "mean_token_accuracy": 0.1380659468472004,
      "num_tokens": 4317136.0,
      "step": 2340
    },
    {
      "entropy": 6.223579549789429,
      "epoch": 0.19701743331232935,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998560229529924,
      "loss": 6.0392,
      "mean_token_accuracy": 0.1369388960301876,
      "num_tokens": 4326163.0,
      "step": 2345
    },
    {
      "entropy": 6.337650966644287,
      "epoch": 0.19743751312749422,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998549498178649,
      "loss": 6.1756,
      "mean_token_accuracy": 0.1331799015402794,
      "num_tokens": 4335837.0,
      "step": 2350
    },
    {
      "entropy": 6.207294178009033,
      "epoch": 0.19785759294265912,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998538726995477,
      "loss": 6.096,
      "mean_token_accuracy": 0.13218674436211586,
      "num_tokens": 4345108.0,
      "step": 2355
    },
    {
      "entropy": 6.251865100860596,
      "epoch": 0.198277672757824,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049985279159806,
      "loss": 6.1203,
      "mean_token_accuracy": 0.13030959963798522,
      "num_tokens": 4353761.0,
      "step": 2360
    },
    {
      "entropy": 6.238997983932495,
      "epoch": 0.19869775257298886,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998517065134208,
      "loss": 6.0679,
      "mean_token_accuracy": 0.1309155412018299,
      "num_tokens": 4363244.0,
      "step": 2365
    },
    {
      "entropy": 6.240653324127197,
      "epoch": 0.19911783238815375,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998506174456494,
      "loss": 6.0563,
      "mean_token_accuracy": 0.13212330862879754,
      "num_tokens": 4373034.0,
      "step": 2370
    },
    {
      "entropy": 6.222096681594849,
      "epoch": 0.19953791220331862,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998495243947653,
      "loss": 6.0012,
      "mean_token_accuracy": 0.12821247354149817,
      "num_tokens": 4382554.0,
      "step": 2375
    },
    {
      "entropy": 6.245941638946533,
      "epoch": 0.19995799201848352,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998484273607875,
      "loss": 6.0093,
      "mean_token_accuracy": 0.1390307828783989,
      "num_tokens": 4391001.0,
      "step": 2380
    },
    {
      "entropy": 6.051425838470459,
      "epoch": 0.2003780718336484,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998473263437356,
      "loss": 5.9437,
      "mean_token_accuracy": 0.13520744293928147,
      "num_tokens": 4400632.0,
      "step": 2385
    },
    {
      "entropy": 6.14297890663147,
      "epoch": 0.20079815164881326,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499846221343629,
      "loss": 6.0297,
      "mean_token_accuracy": 0.13219423592090607,
      "num_tokens": 4409565.0,
      "step": 2390
    },
    {
      "entropy": 6.155060577392578,
      "epoch": 0.20121823146397816,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998451123604875,
      "loss": 5.9866,
      "mean_token_accuracy": 0.14023150131106377,
      "num_tokens": 4418384.0,
      "step": 2395
    },
    {
      "entropy": 6.256305456161499,
      "epoch": 0.20163831127914303,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998439993943306,
      "loss": 6.1021,
      "mean_token_accuracy": 0.1406387820839882,
      "num_tokens": 4427581.0,
      "step": 2400
    },
    {
      "entropy": 6.275782299041748,
      "epoch": 0.20205839109430793,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998428824451779,
      "loss": 6.0884,
      "mean_token_accuracy": 0.12875301465392114,
      "num_tokens": 4436572.0,
      "step": 2405
    },
    {
      "entropy": 6.213010787963867,
      "epoch": 0.2024784709094728,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998417615130495,
      "loss": 6.0989,
      "mean_token_accuracy": 0.12865509316325188,
      "num_tokens": 4445230.0,
      "step": 2410
    },
    {
      "entropy": 6.289054584503174,
      "epoch": 0.2028985507246377,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998406365979649,
      "loss": 6.166,
      "mean_token_accuracy": 0.13029074743390084,
      "num_tokens": 4454251.0,
      "step": 2415
    },
    {
      "entropy": 6.1775250911712645,
      "epoch": 0.20331863053980256,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998395076999443,
      "loss": 5.9922,
      "mean_token_accuracy": 0.13723801597952842,
      "num_tokens": 4463949.0,
      "step": 2420
    },
    {
      "entropy": 6.2511374950408936,
      "epoch": 0.20373871035496743,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998383748190076,
      "loss": 6.2021,
      "mean_token_accuracy": 0.1281643107533455,
      "num_tokens": 4473373.0,
      "step": 2425
    },
    {
      "entropy": 6.318125200271607,
      "epoch": 0.20415879017013233,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998372379551748,
      "loss": 6.0174,
      "mean_token_accuracy": 0.13771012797951698,
      "num_tokens": 4482303.0,
      "step": 2430
    },
    {
      "entropy": 6.15513801574707,
      "epoch": 0.2045788699852972,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998360971084663,
      "loss": 5.9882,
      "mean_token_accuracy": 0.1334059588611126,
      "num_tokens": 4491214.0,
      "step": 2435
    },
    {
      "entropy": 6.118799161911011,
      "epoch": 0.2049989498004621,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998349522789019,
      "loss": 5.9164,
      "mean_token_accuracy": 0.14181319028139114,
      "num_tokens": 4500099.0,
      "step": 2440
    },
    {
      "entropy": 6.1538893699646,
      "epoch": 0.20541902961562697,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998338034665021,
      "loss": 5.9967,
      "mean_token_accuracy": 0.13947194814682007,
      "num_tokens": 4509893.0,
      "step": 2445
    },
    {
      "entropy": 6.1470013618469235,
      "epoch": 0.20583910943079184,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998326506712872,
      "loss": 5.9665,
      "mean_token_accuracy": 0.1369723781943321,
      "num_tokens": 4518606.0,
      "step": 2450
    },
    {
      "entropy": 6.206533908843994,
      "epoch": 0.20625918924595674,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998314938932778,
      "loss": 6.0509,
      "mean_token_accuracy": 0.1360008053481579,
      "num_tokens": 4528392.0,
      "step": 2455
    },
    {
      "entropy": 6.256601238250733,
      "epoch": 0.2066792690611216,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998303331324943,
      "loss": 6.0235,
      "mean_token_accuracy": 0.13803394809365271,
      "num_tokens": 4536983.0,
      "step": 2460
    },
    {
      "entropy": 6.104164505004883,
      "epoch": 0.2070993488762865,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998291683889571,
      "loss": 5.9373,
      "mean_token_accuracy": 0.14096372500061988,
      "num_tokens": 4544967.0,
      "step": 2465
    },
    {
      "entropy": 6.204864883422852,
      "epoch": 0.20751942869145137,
      "grad_norm": 1.0,
      "learning_rate": 0.000499827999662687,
      "loss": 6.001,
      "mean_token_accuracy": 0.13221748918294907,
      "num_tokens": 4554646.0,
      "step": 2470
    },
    {
      "entropy": 6.243742561340332,
      "epoch": 0.20793950850661624,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998268269537046,
      "loss": 6.0222,
      "mean_token_accuracy": 0.13807824179530143,
      "num_tokens": 4564040.0,
      "step": 2475
    },
    {
      "entropy": 6.120432043075562,
      "epoch": 0.20835958832178114,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998256502620308,
      "loss": 6.0382,
      "mean_token_accuracy": 0.14363139793276786,
      "num_tokens": 4573758.0,
      "step": 2480
    },
    {
      "entropy": 6.265066432952881,
      "epoch": 0.208779668136946,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998244695876864,
      "loss": 6.0777,
      "mean_token_accuracy": 0.1315403826534748,
      "num_tokens": 4582097.0,
      "step": 2485
    },
    {
      "entropy": 6.097950839996338,
      "epoch": 0.2091997479521109,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998232849306921,
      "loss": 6.0435,
      "mean_token_accuracy": 0.1351850524544716,
      "num_tokens": 4590687.0,
      "step": 2490
    },
    {
      "entropy": 6.293486022949219,
      "epoch": 0.20961982776727578,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998220962910693,
      "loss": 6.0235,
      "mean_token_accuracy": 0.13292664289474487,
      "num_tokens": 4599497.0,
      "step": 2495
    },
    {
      "entropy": 6.173567104339599,
      "epoch": 0.21003990758244068,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998209036688386,
      "loss": 5.9939,
      "mean_token_accuracy": 0.13765107467770576,
      "num_tokens": 4607958.0,
      "step": 2500
    },
    {
      "entropy": 6.268425703048706,
      "epoch": 0.21045998739760555,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998197070640216,
      "loss": 6.1098,
      "mean_token_accuracy": 0.129291183501482,
      "num_tokens": 4617515.0,
      "step": 2505
    },
    {
      "entropy": 6.28843879699707,
      "epoch": 0.21088006721277042,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998185064766391,
      "loss": 6.0059,
      "mean_token_accuracy": 0.1318512476980686,
      "num_tokens": 4627037.0,
      "step": 2510
    },
    {
      "entropy": 6.095378351211548,
      "epoch": 0.21130014702793531,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998173019067127,
      "loss": 6.0089,
      "mean_token_accuracy": 0.1362810969352722,
      "num_tokens": 4637393.0,
      "step": 2515
    },
    {
      "entropy": 6.151039409637451,
      "epoch": 0.21172022684310018,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998160933542633,
      "loss": 6.0518,
      "mean_token_accuracy": 0.12635309770703315,
      "num_tokens": 4646832.0,
      "step": 2520
    },
    {
      "entropy": 6.288768529891968,
      "epoch": 0.21214030665826508,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998148808193128,
      "loss": 6.0821,
      "mean_token_accuracy": 0.13684087693691255,
      "num_tokens": 4655719.0,
      "step": 2525
    },
    {
      "entropy": 6.185772228240967,
      "epoch": 0.21256038647342995,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998136643018823,
      "loss": 6.0231,
      "mean_token_accuracy": 0.13462624847888946,
      "num_tokens": 4665364.0,
      "step": 2530
    },
    {
      "entropy": 6.2163975715637205,
      "epoch": 0.21298046628859482,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998124438019935,
      "loss": 6.0079,
      "mean_token_accuracy": 0.13166589960455893,
      "num_tokens": 4674760.0,
      "step": 2535
    },
    {
      "entropy": 6.0905904293060305,
      "epoch": 0.21340054610375972,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998112193196681,
      "loss": 5.9166,
      "mean_token_accuracy": 0.13693561255931855,
      "num_tokens": 4683900.0,
      "step": 2540
    },
    {
      "entropy": 6.067295026779175,
      "epoch": 0.2138206259189246,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998099908549277,
      "loss": 5.9706,
      "mean_token_accuracy": 0.13006317764520645,
      "num_tokens": 4693915.0,
      "step": 2545
    },
    {
      "entropy": 6.123124837875366,
      "epoch": 0.2142407057340895,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499808758407794,
      "loss": 5.8413,
      "mean_token_accuracy": 0.1404764860868454,
      "num_tokens": 4703102.0,
      "step": 2550
    },
    {
      "entropy": 6.155497789382935,
      "epoch": 0.21466078554925436,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998075219782889,
      "loss": 6.0483,
      "mean_token_accuracy": 0.13892892748117447,
      "num_tokens": 4712925.0,
      "step": 2555
    },
    {
      "entropy": 6.1989946365356445,
      "epoch": 0.21508086536441923,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998062815664344,
      "loss": 5.9819,
      "mean_token_accuracy": 0.13216783180832864,
      "num_tokens": 4722641.0,
      "step": 2560
    },
    {
      "entropy": 6.098575496673584,
      "epoch": 0.21550094517958412,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998050371722524,
      "loss": 6.0475,
      "mean_token_accuracy": 0.13535310626029967,
      "num_tokens": 4732603.0,
      "step": 2565
    },
    {
      "entropy": 6.0733263969421385,
      "epoch": 0.215921024994749,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004998037887957649,
      "loss": 5.8768,
      "mean_token_accuracy": 0.1405107393860817,
      "num_tokens": 4742644.0,
      "step": 2570
    },
    {
      "entropy": 6.274937391281128,
      "epoch": 0.2163411048099139,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998025364369939,
      "loss": 6.2137,
      "mean_token_accuracy": 0.13099505975842476,
      "num_tokens": 4751482.0,
      "step": 2575
    },
    {
      "entropy": 6.293144750595093,
      "epoch": 0.21676118462507876,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998012800959619,
      "loss": 6.0839,
      "mean_token_accuracy": 0.13207904621958733,
      "num_tokens": 4760593.0,
      "step": 2580
    },
    {
      "entropy": 6.261121511459351,
      "epoch": 0.21718126444024366,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998000197726909,
      "loss": 6.0643,
      "mean_token_accuracy": 0.13794807344675064,
      "num_tokens": 4769294.0,
      "step": 2585
    },
    {
      "entropy": 6.1975377082824705,
      "epoch": 0.21760134425540853,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004997987554672033,
      "loss": 5.9825,
      "mean_token_accuracy": 0.13557377234101295,
      "num_tokens": 4779239.0,
      "step": 2590
    },
    {
      "entropy": 6.179704809188843,
      "epoch": 0.2180214240705734,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997974871795215,
      "loss": 6.0438,
      "mean_token_accuracy": 0.13279528692364692,
      "num_tokens": 4788211.0,
      "step": 2595
    },
    {
      "entropy": 6.169263744354248,
      "epoch": 0.2184415038857383,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499796214909668,
      "loss": 6.0046,
      "mean_token_accuracy": 0.14151790514588355,
      "num_tokens": 4797921.0,
      "step": 2600
    },
    {
      "entropy": 6.179717254638672,
      "epoch": 0.21886158370090317,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997949386576653,
      "loss": 6.021,
      "mean_token_accuracy": 0.13185311183333398,
      "num_tokens": 4807772.0,
      "step": 2605
    },
    {
      "entropy": 6.09913649559021,
      "epoch": 0.21928166351606806,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499793658423536,
      "loss": 6.0302,
      "mean_token_accuracy": 0.1336025021970272,
      "num_tokens": 4817999.0,
      "step": 2610
    },
    {
      "entropy": 6.236963701248169,
      "epoch": 0.21970174333123293,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997923742073028,
      "loss": 5.9808,
      "mean_token_accuracy": 0.14284878820180893,
      "num_tokens": 4826679.0,
      "step": 2615
    },
    {
      "entropy": 6.087517356872558,
      "epoch": 0.2201218231463978,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997910860089884,
      "loss": 5.9814,
      "mean_token_accuracy": 0.13721209689974784,
      "num_tokens": 4834998.0,
      "step": 2620
    },
    {
      "entropy": 6.181451654434204,
      "epoch": 0.2205419029615627,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997897938286156,
      "loss": 5.9418,
      "mean_token_accuracy": 0.140488101541996,
      "num_tokens": 4843635.0,
      "step": 2625
    },
    {
      "entropy": 6.164580821990967,
      "epoch": 0.22096198277672757,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997884976662075,
      "loss": 6.0616,
      "mean_token_accuracy": 0.13446614518761635,
      "num_tokens": 4852027.0,
      "step": 2630
    },
    {
      "entropy": 6.231269407272339,
      "epoch": 0.22138206259189247,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997871975217868,
      "loss": 5.9769,
      "mean_token_accuracy": 0.14198934510350228,
      "num_tokens": 4861244.0,
      "step": 2635
    },
    {
      "entropy": 6.045923233032227,
      "epoch": 0.22180214240705734,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997858933953768,
      "loss": 5.8995,
      "mean_token_accuracy": 0.14127301722764968,
      "num_tokens": 4869902.0,
      "step": 2640
    },
    {
      "entropy": 6.087070560455322,
      "epoch": 0.2222222222222222,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997845852870004,
      "loss": 5.879,
      "mean_token_accuracy": 0.14420521035790443,
      "num_tokens": 4878502.0,
      "step": 2645
    },
    {
      "entropy": 6.088281726837158,
      "epoch": 0.2226423020373871,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997832731966806,
      "loss": 5.9405,
      "mean_token_accuracy": 0.14226024821400643,
      "num_tokens": 4888348.0,
      "step": 2650
    },
    {
      "entropy": 6.156372451782227,
      "epoch": 0.22306238185255198,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997819571244411,
      "loss": 5.9976,
      "mean_token_accuracy": 0.14085667207837105,
      "num_tokens": 4897302.0,
      "step": 2655
    },
    {
      "entropy": 6.102771759033203,
      "epoch": 0.22348246166771688,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997806370703049,
      "loss": 6.0167,
      "mean_token_accuracy": 0.1412481315433979,
      "num_tokens": 4907078.0,
      "step": 2660
    },
    {
      "entropy": 6.086454391479492,
      "epoch": 0.22390254148288175,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004997793130342954,
      "loss": 5.8462,
      "mean_token_accuracy": 0.14060726314783095,
      "num_tokens": 4917489.0,
      "step": 2665
    },
    {
      "entropy": 6.059194183349609,
      "epoch": 0.22432262129804661,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997779850164363,
      "loss": 5.9408,
      "mean_token_accuracy": 0.13854166269302368,
      "num_tokens": 4927073.0,
      "step": 2670
    },
    {
      "entropy": 6.217087936401367,
      "epoch": 0.2247427011132115,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997766530167508,
      "loss": 6.0634,
      "mean_token_accuracy": 0.13044301494956018,
      "num_tokens": 4935464.0,
      "step": 2675
    },
    {
      "entropy": 6.28878345489502,
      "epoch": 0.22516278092837638,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997753170352627,
      "loss": 6.1414,
      "mean_token_accuracy": 0.13435352668166162,
      "num_tokens": 4944718.0,
      "step": 2680
    },
    {
      "entropy": 6.156159353256226,
      "epoch": 0.22558286074354128,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997739770719955,
      "loss": 6.0181,
      "mean_token_accuracy": 0.13306965455412864,
      "num_tokens": 4954223.0,
      "step": 2685
    },
    {
      "entropy": 6.152837419509888,
      "epoch": 0.22600294055870615,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499772633126973,
      "loss": 6.0448,
      "mean_token_accuracy": 0.13488989993929862,
      "num_tokens": 4963371.0,
      "step": 2690
    },
    {
      "entropy": 6.077383518218994,
      "epoch": 0.22642302037387105,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997712852002192,
      "loss": 5.8942,
      "mean_token_accuracy": 0.14509259015321732,
      "num_tokens": 4972973.0,
      "step": 2695
    },
    {
      "entropy": 6.145831155776977,
      "epoch": 0.22684310018903592,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997699332917578,
      "loss": 6.1551,
      "mean_token_accuracy": 0.12881803661584854,
      "num_tokens": 4982808.0,
      "step": 2700
    },
    {
      "entropy": 6.264477157592774,
      "epoch": 0.2272631800042008,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997685774016127,
      "loss": 6.0218,
      "mean_token_accuracy": 0.13483983278274536,
      "num_tokens": 4992427.0,
      "step": 2705
    },
    {
      "entropy": 6.240380764007568,
      "epoch": 0.22768325981936569,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499767217529808,
      "loss": 6.1836,
      "mean_token_accuracy": 0.1287289910018444,
      "num_tokens": 5003562.0,
      "step": 2710
    },
    {
      "entropy": 6.103500318527222,
      "epoch": 0.22810333963453056,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004997658536763678,
      "loss": 5.9025,
      "mean_token_accuracy": 0.13835748061537742,
      "num_tokens": 5013429.0,
      "step": 2715
    },
    {
      "entropy": 6.204726886749268,
      "epoch": 0.22852341944969545,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997644858413163,
      "loss": 6.0337,
      "mean_token_accuracy": 0.13989441394805907,
      "num_tokens": 5022045.0,
      "step": 2720
    },
    {
      "entropy": 6.083020782470703,
      "epoch": 0.22894349926486032,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997631140246775,
      "loss": 5.8567,
      "mean_token_accuracy": 0.14418582767248153,
      "num_tokens": 5032260.0,
      "step": 2725
    },
    {
      "entropy": 6.0916870594024655,
      "epoch": 0.2293635790800252,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499761738226476,
      "loss": 5.9035,
      "mean_token_accuracy": 0.1432834282517433,
      "num_tokens": 5041688.0,
      "step": 2730
    },
    {
      "entropy": 6.093435859680175,
      "epoch": 0.2297836588951901,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499760358446736,
      "loss": 6.0111,
      "mean_token_accuracy": 0.13241263329982758,
      "num_tokens": 5051005.0,
      "step": 2735
    },
    {
      "entropy": 6.205419874191284,
      "epoch": 0.23020373871035496,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499758974685482,
      "loss": 5.9348,
      "mean_token_accuracy": 0.13710809499025345,
      "num_tokens": 5060084.0,
      "step": 2740
    },
    {
      "entropy": 6.123983669281006,
      "epoch": 0.23062381852551986,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997575869427385,
      "loss": 5.948,
      "mean_token_accuracy": 0.14109890162944794,
      "num_tokens": 5069081.0,
      "step": 2745
    },
    {
      "entropy": 6.1220703125,
      "epoch": 0.23104389834068473,
      "grad_norm": 0.921875,
      "learning_rate": 0.00049975619521853,
      "loss": 5.9356,
      "mean_token_accuracy": 0.13786115869879723,
      "num_tokens": 5078597.0,
      "step": 2750
    },
    {
      "entropy": 6.122048711776733,
      "epoch": 0.2314639781558496,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997547995128814,
      "loss": 5.9968,
      "mean_token_accuracy": 0.14226307570934296,
      "num_tokens": 5087607.0,
      "step": 2755
    },
    {
      "entropy": 6.144302225112915,
      "epoch": 0.2318840579710145,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997533998258171,
      "loss": 5.9821,
      "mean_token_accuracy": 0.14107478111982347,
      "num_tokens": 5097412.0,
      "step": 2760
    },
    {
      "entropy": 6.224723386764526,
      "epoch": 0.23230413778617937,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997519961573622,
      "loss": 6.053,
      "mean_token_accuracy": 0.13255186080932618,
      "num_tokens": 5105817.0,
      "step": 2765
    },
    {
      "entropy": 6.279865074157715,
      "epoch": 0.23272421760134426,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997505885075414,
      "loss": 6.0941,
      "mean_token_accuracy": 0.13345747739076613,
      "num_tokens": 5114958.0,
      "step": 2770
    },
    {
      "entropy": 6.129254007339478,
      "epoch": 0.23314429741650913,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997491768763795,
      "loss": 6.0302,
      "mean_token_accuracy": 0.13575156703591346,
      "num_tokens": 5123728.0,
      "step": 2775
    },
    {
      "entropy": 6.134473133087158,
      "epoch": 0.23356437723167403,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997477612639018,
      "loss": 6.0582,
      "mean_token_accuracy": 0.13159941956400872,
      "num_tokens": 5134099.0,
      "step": 2780
    },
    {
      "entropy": 6.231441497802734,
      "epoch": 0.2339844570468389,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997463416701332,
      "loss": 6.0689,
      "mean_token_accuracy": 0.1301489420235157,
      "num_tokens": 5142934.0,
      "step": 2785
    },
    {
      "entropy": 6.0879510879516605,
      "epoch": 0.23440453686200377,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997449180950989,
      "loss": 5.8957,
      "mean_token_accuracy": 0.15564696341753007,
      "num_tokens": 5151835.0,
      "step": 2790
    },
    {
      "entropy": 6.114243698120117,
      "epoch": 0.23482461667716867,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997434905388241,
      "loss": 5.9528,
      "mean_token_accuracy": 0.14443679004907609,
      "num_tokens": 5161136.0,
      "step": 2795
    },
    {
      "entropy": 6.104035425186157,
      "epoch": 0.23524469649233354,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499742059001334,
      "loss": 5.8911,
      "mean_token_accuracy": 0.14185608699917793,
      "num_tokens": 5170741.0,
      "step": 2800
    },
    {
      "entropy": 6.075469923019409,
      "epoch": 0.23566477630749844,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997406234826541,
      "loss": 5.9262,
      "mean_token_accuracy": 0.14401806518435478,
      "num_tokens": 5180549.0,
      "step": 2805
    },
    {
      "entropy": 6.064437532424927,
      "epoch": 0.2360848561226633,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004997391839828098,
      "loss": 5.8936,
      "mean_token_accuracy": 0.14641853421926498,
      "num_tokens": 5189486.0,
      "step": 2810
    },
    {
      "entropy": 6.1467596054077145,
      "epoch": 0.23650493593782818,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997377405018266,
      "loss": 5.9515,
      "mean_token_accuracy": 0.1352544255554676,
      "num_tokens": 5198525.0,
      "step": 2815
    },
    {
      "entropy": 6.114045858383179,
      "epoch": 0.23692501575299307,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00049973629303973,
      "loss": 6.0104,
      "mean_token_accuracy": 0.13799095004796982,
      "num_tokens": 5207124.0,
      "step": 2820
    },
    {
      "entropy": 6.020995473861694,
      "epoch": 0.23734509556815794,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004997348415965457,
      "loss": 5.8289,
      "mean_token_accuracy": 0.14437564313411713,
      "num_tokens": 5216529.0,
      "step": 2825
    },
    {
      "entropy": 6.182903099060058,
      "epoch": 0.23776517538332284,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997333861722995,
      "loss": 5.9792,
      "mean_token_accuracy": 0.14223513826727868,
      "num_tokens": 5225796.0,
      "step": 2830
    },
    {
      "entropy": 6.194612598419189,
      "epoch": 0.2381852551984877,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499731926767017,
      "loss": 5.9941,
      "mean_token_accuracy": 0.14215042740106582,
      "num_tokens": 5233876.0,
      "step": 2835
    },
    {
      "entropy": 6.092140197753906,
      "epoch": 0.23860533501365258,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004997304633807242,
      "loss": 5.9958,
      "mean_token_accuracy": 0.13835923075675965,
      "num_tokens": 5244782.0,
      "step": 2840
    },
    {
      "entropy": 6.103197765350342,
      "epoch": 0.23902541482881748,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997289960134468,
      "loss": 5.9434,
      "mean_token_accuracy": 0.14274498224258422,
      "num_tokens": 5253453.0,
      "step": 2845
    },
    {
      "entropy": 6.109802532196045,
      "epoch": 0.23944549464398235,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997275246652111,
      "loss": 5.9697,
      "mean_token_accuracy": 0.1414198510348797,
      "num_tokens": 5262355.0,
      "step": 2850
    },
    {
      "entropy": 6.074161100387573,
      "epoch": 0.23986557445914725,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499726049336043,
      "loss": 5.884,
      "mean_token_accuracy": 0.14384180679917336,
      "num_tokens": 5271959.0,
      "step": 2855
    },
    {
      "entropy": 6.123191499710083,
      "epoch": 0.24028565427431212,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997245700259686,
      "loss": 5.9216,
      "mean_token_accuracy": 0.14565154165029526,
      "num_tokens": 5281393.0,
      "step": 2860
    },
    {
      "entropy": 6.1445637226104735,
      "epoch": 0.240705734089477,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997230867350141,
      "loss": 6.0674,
      "mean_token_accuracy": 0.13226635977625847,
      "num_tokens": 5290979.0,
      "step": 2865
    },
    {
      "entropy": 6.2341917037963865,
      "epoch": 0.24112581390464188,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997215994632059,
      "loss": 6.0149,
      "mean_token_accuracy": 0.14298992082476616,
      "num_tokens": 5300263.0,
      "step": 2870
    },
    {
      "entropy": 6.114161205291748,
      "epoch": 0.24154589371980675,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997201082105704,
      "loss": 6.0298,
      "mean_token_accuracy": 0.13639863729476928,
      "num_tokens": 5309522.0,
      "step": 2875
    },
    {
      "entropy": 6.144821310043335,
      "epoch": 0.24196597353497165,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997186129771338,
      "loss": 6.0256,
      "mean_token_accuracy": 0.1396084174513817,
      "num_tokens": 5319770.0,
      "step": 2880
    },
    {
      "entropy": 6.21076397895813,
      "epoch": 0.24238605335013652,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997171137629226,
      "loss": 6.0217,
      "mean_token_accuracy": 0.1428602673113346,
      "num_tokens": 5328400.0,
      "step": 2885
    },
    {
      "entropy": 6.059879302978516,
      "epoch": 0.24280613316530142,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997156105679636,
      "loss": 5.8456,
      "mean_token_accuracy": 0.1514780417084694,
      "num_tokens": 5336338.0,
      "step": 2890
    },
    {
      "entropy": 6.016794109344483,
      "epoch": 0.2432262129804663,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997141033922832,
      "loss": 5.927,
      "mean_token_accuracy": 0.13949262946844102,
      "num_tokens": 5345391.0,
      "step": 2895
    },
    {
      "entropy": 6.142647981643677,
      "epoch": 0.24364629279563116,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997125922359081,
      "loss": 5.9624,
      "mean_token_accuracy": 0.13398301377892494,
      "num_tokens": 5354709.0,
      "step": 2900
    },
    {
      "entropy": 6.111950016021728,
      "epoch": 0.24406637261079606,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997110770988652,
      "loss": 5.8818,
      "mean_token_accuracy": 0.14174696654081345,
      "num_tokens": 5363738.0,
      "step": 2905
    },
    {
      "entropy": 6.079838371276855,
      "epoch": 0.24448645242596093,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997095579811813,
      "loss": 6.0081,
      "mean_token_accuracy": 0.13893039748072625,
      "num_tokens": 5373583.0,
      "step": 2910
    },
    {
      "entropy": 6.187237548828125,
      "epoch": 0.24490653224112582,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004997080348828833,
      "loss": 6.0554,
      "mean_token_accuracy": 0.13840119615197183,
      "num_tokens": 5383486.0,
      "step": 2915
    },
    {
      "entropy": 6.126826524734497,
      "epoch": 0.2453266120562907,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997065078039981,
      "loss": 5.9624,
      "mean_token_accuracy": 0.1391638047993183,
      "num_tokens": 5391974.0,
      "step": 2920
    },
    {
      "entropy": 6.164314794540405,
      "epoch": 0.24574669187145556,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997049767445529,
      "loss": 5.999,
      "mean_token_accuracy": 0.1359800137579441,
      "num_tokens": 5400882.0,
      "step": 2925
    },
    {
      "entropy": 6.199448156356811,
      "epoch": 0.24616677168662046,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997034417045746,
      "loss": 5.9647,
      "mean_token_accuracy": 0.13757251799106598,
      "num_tokens": 5410538.0,
      "step": 2930
    },
    {
      "entropy": 6.0790434837341305,
      "epoch": 0.24658685150178533,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997019026840907,
      "loss": 5.846,
      "mean_token_accuracy": 0.14040058851242065,
      "num_tokens": 5419406.0,
      "step": 2935
    },
    {
      "entropy": 5.975088357925415,
      "epoch": 0.24700693131695023,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997003596831282,
      "loss": 5.9572,
      "mean_token_accuracy": 0.1371507503092289,
      "num_tokens": 5428817.0,
      "step": 2940
    },
    {
      "entropy": 6.196776390075684,
      "epoch": 0.2474270111321151,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996988127017145,
      "loss": 5.9872,
      "mean_token_accuracy": 0.13953725397586822,
      "num_tokens": 5438277.0,
      "step": 2945
    },
    {
      "entropy": 6.091228246688843,
      "epoch": 0.24784709094728,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996972617398772,
      "loss": 6.0034,
      "mean_token_accuracy": 0.13882646411657334,
      "num_tokens": 5447440.0,
      "step": 2950
    },
    {
      "entropy": 6.129848289489746,
      "epoch": 0.24826717076244487,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996957067976435,
      "loss": 5.9254,
      "mean_token_accuracy": 0.13761287704110145,
      "num_tokens": 5455988.0,
      "step": 2955
    },
    {
      "entropy": 6.139088869094849,
      "epoch": 0.24868725057760974,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996941478750411,
      "loss": 5.9357,
      "mean_token_accuracy": 0.14046952053904532,
      "num_tokens": 5464996.0,
      "step": 2960
    },
    {
      "entropy": 6.173188400268555,
      "epoch": 0.24910733039277463,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996925849720975,
      "loss": 6.0717,
      "mean_token_accuracy": 0.13539182469248773,
      "num_tokens": 5474174.0,
      "step": 2965
    },
    {
      "entropy": 6.202445459365845,
      "epoch": 0.2495274102079395,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996910180888405,
      "loss": 5.9624,
      "mean_token_accuracy": 0.14169209375977515,
      "num_tokens": 5482838.0,
      "step": 2970
    },
    {
      "entropy": 6.113552713394165,
      "epoch": 0.2499474900231044,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996894472252977,
      "loss": 5.9734,
      "mean_token_accuracy": 0.13861241191625595,
      "num_tokens": 5491616.0,
      "step": 2975
    },
    {
      "entropy": 6.091303777694702,
      "epoch": 0.25036756983826924,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996878723814973,
      "loss": 5.9555,
      "mean_token_accuracy": 0.13805645182728768,
      "num_tokens": 5500942.0,
      "step": 2980
    },
    {
      "entropy": 6.159912872314453,
      "epoch": 0.25078764965343414,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996862935574667,
      "loss": 5.8989,
      "mean_token_accuracy": 0.13720250651240348,
      "num_tokens": 5510078.0,
      "step": 2985
    },
    {
      "entropy": 6.035575675964355,
      "epoch": 0.25120772946859904,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996847107532342,
      "loss": 5.953,
      "mean_token_accuracy": 0.13990190252661705,
      "num_tokens": 5518924.0,
      "step": 2990
    },
    {
      "entropy": 6.1541543960571286,
      "epoch": 0.25162780928376394,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996831239688277,
      "loss": 5.9479,
      "mean_token_accuracy": 0.135651333630085,
      "num_tokens": 5527385.0,
      "step": 2995
    },
    {
      "entropy": 5.980669641494751,
      "epoch": 0.2520478890989288,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996815332042754,
      "loss": 5.8093,
      "mean_token_accuracy": 0.14832203537225724,
      "num_tokens": 5536781.0,
      "step": 3000
    },
    {
      "epoch": 0.2520478890989288,
      "eval_entropy": 5.811351248549843,
      "eval_loss": 5.969449043273926,
      "eval_mean_token_accuracy": 0.14523225678583587,
      "eval_num_tokens": 5536781.0,
      "eval_runtime": 20.9301,
      "eval_samples_per_second": 1785.275,
      "eval_steps_per_second": 223.171,
      "step": 3000
    },
    {
      "entropy": 6.094974756240845,
      "epoch": 0.2524679689140937,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996799384596054,
      "loss": 5.9892,
      "mean_token_accuracy": 0.13971033170819283,
      "num_tokens": 5545893.0,
      "step": 3005
    },
    {
      "entropy": 6.159230375289917,
      "epoch": 0.2528880487292586,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996783397348461,
      "loss": 5.9342,
      "mean_token_accuracy": 0.1357440948486328,
      "num_tokens": 5555818.0,
      "step": 3010
    },
    {
      "entropy": 6.10536036491394,
      "epoch": 0.2533081285444234,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004996767370300256,
      "loss": 5.9092,
      "mean_token_accuracy": 0.14052527099847795,
      "num_tokens": 5565331.0,
      "step": 3015
    },
    {
      "entropy": 6.084013175964356,
      "epoch": 0.2537282083595883,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996751303451724,
      "loss": 5.9063,
      "mean_token_accuracy": 0.1431153953075409,
      "num_tokens": 5574003.0,
      "step": 3020
    },
    {
      "entropy": 6.077633428573608,
      "epoch": 0.2541482881747532,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996735196803149,
      "loss": 5.8124,
      "mean_token_accuracy": 0.1457200452685356,
      "num_tokens": 5582517.0,
      "step": 3025
    },
    {
      "entropy": 6.136652421951294,
      "epoch": 0.2545683679899181,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004996719050354818,
      "loss": 6.0331,
      "mean_token_accuracy": 0.13732432797551156,
      "num_tokens": 5591952.0,
      "step": 3030
    },
    {
      "entropy": 6.048459148406982,
      "epoch": 0.25498844780508295,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996702864107015,
      "loss": 5.9271,
      "mean_token_accuracy": 0.1418526716530323,
      "num_tokens": 5601460.0,
      "step": 3035
    },
    {
      "entropy": 6.284002351760864,
      "epoch": 0.25540852762024785,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996686638060028,
      "loss": 6.0376,
      "mean_token_accuracy": 0.13265891820192338,
      "num_tokens": 5610776.0,
      "step": 3040
    },
    {
      "entropy": 6.0990218162536625,
      "epoch": 0.25582860743541275,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996670372214144,
      "loss": 5.9517,
      "mean_token_accuracy": 0.14587848633527756,
      "num_tokens": 5619627.0,
      "step": 3045
    },
    {
      "entropy": 5.947749042510987,
      "epoch": 0.2562486872505776,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996654066569651,
      "loss": 5.8302,
      "mean_token_accuracy": 0.14839987307786942,
      "num_tokens": 5628969.0,
      "step": 3050
    },
    {
      "entropy": 6.065955305099488,
      "epoch": 0.2566687670657425,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996637721126839,
      "loss": 5.8796,
      "mean_token_accuracy": 0.1424746796488762,
      "num_tokens": 5638629.0,
      "step": 3055
    },
    {
      "entropy": 6.107337474822998,
      "epoch": 0.2570888468809074,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996621335885996,
      "loss": 5.96,
      "mean_token_accuracy": 0.14037469550967216,
      "num_tokens": 5647571.0,
      "step": 3060
    },
    {
      "entropy": 6.1402346134185795,
      "epoch": 0.2575089266960722,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004996604910847413,
      "loss": 5.8758,
      "mean_token_accuracy": 0.1501998595893383,
      "num_tokens": 5656709.0,
      "step": 3065
    },
    {
      "entropy": 6.089657640457153,
      "epoch": 0.2579290065112371,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499658844601138,
      "loss": 6.0454,
      "mean_token_accuracy": 0.1388375774025917,
      "num_tokens": 5665714.0,
      "step": 3070
    },
    {
      "entropy": 6.122017049789429,
      "epoch": 0.258349086326402,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499657194137819,
      "loss": 6.0085,
      "mean_token_accuracy": 0.14038632139563562,
      "num_tokens": 5675854.0,
      "step": 3075
    },
    {
      "entropy": 6.120520496368409,
      "epoch": 0.2587691661415669,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996555396948136,
      "loss": 5.8388,
      "mean_token_accuracy": 0.13863911256194114,
      "num_tokens": 5685690.0,
      "step": 3080
    },
    {
      "entropy": 6.04867787361145,
      "epoch": 0.25918924595673176,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996538812721509,
      "loss": 5.8979,
      "mean_token_accuracy": 0.14566717743873597,
      "num_tokens": 5695766.0,
      "step": 3085
    },
    {
      "entropy": 6.1084822654724125,
      "epoch": 0.25960932577189666,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996522188698603,
      "loss": 5.9477,
      "mean_token_accuracy": 0.14305482804775238,
      "num_tokens": 5704365.0,
      "step": 3090
    },
    {
      "entropy": 6.1373619556427,
      "epoch": 0.26002940558706156,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996505524879714,
      "loss": 6.0346,
      "mean_token_accuracy": 0.1344648614525795,
      "num_tokens": 5713345.0,
      "step": 3095
    },
    {
      "entropy": 6.0922935962677,
      "epoch": 0.2604494854022264,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996488821265137,
      "loss": 5.8491,
      "mean_token_accuracy": 0.14658251106739045,
      "num_tokens": 5722907.0,
      "step": 3100
    },
    {
      "entropy": 6.012635898590088,
      "epoch": 0.2608695652173913,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996472077855166,
      "loss": 5.8975,
      "mean_token_accuracy": 0.1446256823837757,
      "num_tokens": 5731589.0,
      "step": 3105
    },
    {
      "entropy": 6.104847145080567,
      "epoch": 0.2612896450325562,
      "grad_norm": 0.953125,
      "learning_rate": 0.00049964552946501,
      "loss": 5.8693,
      "mean_token_accuracy": 0.14166852980852127,
      "num_tokens": 5739922.0,
      "step": 3110
    },
    {
      "entropy": 5.955222845077515,
      "epoch": 0.2617097248477211,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996438471650235,
      "loss": 5.8075,
      "mean_token_accuracy": 0.14593275487422944,
      "num_tokens": 5749206.0,
      "step": 3115
    },
    {
      "entropy": 6.058114147186279,
      "epoch": 0.26212980466288593,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996421608855869,
      "loss": 5.8672,
      "mean_token_accuracy": 0.14658411368727683,
      "num_tokens": 5758803.0,
      "step": 3120
    },
    {
      "entropy": 6.108597993850708,
      "epoch": 0.26254988447805083,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996404706267301,
      "loss": 5.9424,
      "mean_token_accuracy": 0.1372783973813057,
      "num_tokens": 5768368.0,
      "step": 3125
    },
    {
      "entropy": 6.006940793991089,
      "epoch": 0.26296996429321573,
      "grad_norm": 1.0,
      "learning_rate": 0.000499638776388483,
      "loss": 5.7986,
      "mean_token_accuracy": 0.15118449479341506,
      "num_tokens": 5776707.0,
      "step": 3130
    },
    {
      "entropy": 6.002774715423584,
      "epoch": 0.26339004410838057,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996370781708757,
      "loss": 5.9905,
      "mean_token_accuracy": 0.1297268383204937,
      "num_tokens": 5787037.0,
      "step": 3135
    },
    {
      "entropy": 6.140283203125,
      "epoch": 0.26381012392354547,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996353759739382,
      "loss": 5.9369,
      "mean_token_accuracy": 0.14406516328454017,
      "num_tokens": 5796630.0,
      "step": 3140
    },
    {
      "entropy": 6.036018514633179,
      "epoch": 0.26423020373871037,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996336697977007,
      "loss": 5.9374,
      "mean_token_accuracy": 0.14153894111514093,
      "num_tokens": 5806402.0,
      "step": 3145
    },
    {
      "entropy": 6.037613487243652,
      "epoch": 0.2646502835538752,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996319596421933,
      "loss": 5.8841,
      "mean_token_accuracy": 0.14679263159632683,
      "num_tokens": 5815742.0,
      "step": 3150
    },
    {
      "entropy": 6.028949451446533,
      "epoch": 0.2650703633690401,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996302455074466,
      "loss": 5.9032,
      "mean_token_accuracy": 0.14230084121227266,
      "num_tokens": 5824915.0,
      "step": 3155
    },
    {
      "entropy": 6.096432209014893,
      "epoch": 0.265490443184205,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996285273934906,
      "loss": 5.9252,
      "mean_token_accuracy": 0.1397813640534878,
      "num_tokens": 5834978.0,
      "step": 3160
    },
    {
      "entropy": 6.111389875411987,
      "epoch": 0.2659105229993699,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499626805300356,
      "loss": 6.0775,
      "mean_token_accuracy": 0.14250660091638565,
      "num_tokens": 5845684.0,
      "step": 3165
    },
    {
      "entropy": 6.151995420455933,
      "epoch": 0.26633060281453474,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996250792280732,
      "loss": 5.9483,
      "mean_token_accuracy": 0.13529476448893546,
      "num_tokens": 5854905.0,
      "step": 3170
    },
    {
      "entropy": 6.133824729919434,
      "epoch": 0.26675068262969964,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996233491766727,
      "loss": 5.9717,
      "mean_token_accuracy": 0.13773170113563538,
      "num_tokens": 5863654.0,
      "step": 3175
    },
    {
      "entropy": 6.108005428314209,
      "epoch": 0.26717076244486454,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996216151461854,
      "loss": 5.9717,
      "mean_token_accuracy": 0.14263901710510254,
      "num_tokens": 5872442.0,
      "step": 3180
    },
    {
      "entropy": 6.124168252944946,
      "epoch": 0.2675908422600294,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996198771366417,
      "loss": 5.8956,
      "mean_token_accuracy": 0.1422991782426834,
      "num_tokens": 5882372.0,
      "step": 3185
    },
    {
      "entropy": 5.892281866073608,
      "epoch": 0.2680109220751943,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996181351480726,
      "loss": 5.7197,
      "mean_token_accuracy": 0.14947764426469803,
      "num_tokens": 5891113.0,
      "step": 3190
    },
    {
      "entropy": 5.995995616912841,
      "epoch": 0.2684310018903592,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996163891805089,
      "loss": 5.9396,
      "mean_token_accuracy": 0.14690565466880798,
      "num_tokens": 5899582.0,
      "step": 3195
    },
    {
      "entropy": 6.141474914550781,
      "epoch": 0.2688510817055241,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996146392339815,
      "loss": 5.9025,
      "mean_token_accuracy": 0.13870225176215173,
      "num_tokens": 5908938.0,
      "step": 3200
    },
    {
      "entropy": 6.058559274673462,
      "epoch": 0.2692711615206889,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996128853085215,
      "loss": 5.8555,
      "mean_token_accuracy": 0.14446185976266862,
      "num_tokens": 5918055.0,
      "step": 3205
    },
    {
      "entropy": 6.094514513015747,
      "epoch": 0.2696912413358538,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996111274041598,
      "loss": 5.8536,
      "mean_token_accuracy": 0.14117878526449204,
      "num_tokens": 5926744.0,
      "step": 3210
    },
    {
      "entropy": 6.07371563911438,
      "epoch": 0.2701113211510187,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004996093655209277,
      "loss": 5.9284,
      "mean_token_accuracy": 0.1441202089190483,
      "num_tokens": 5936521.0,
      "step": 3215
    },
    {
      "entropy": 6.157954692840576,
      "epoch": 0.27053140096618356,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996075996588563,
      "loss": 6.0011,
      "mean_token_accuracy": 0.1351262703537941,
      "num_tokens": 5945010.0,
      "step": 3220
    },
    {
      "entropy": 6.07440071105957,
      "epoch": 0.27095148078134845,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499605829817977,
      "loss": 5.8945,
      "mean_token_accuracy": 0.14755850434303283,
      "num_tokens": 5953766.0,
      "step": 3225
    },
    {
      "entropy": 6.051190042495728,
      "epoch": 0.27137156059651335,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499604055998321,
      "loss": 5.8222,
      "mean_token_accuracy": 0.14777183532714844,
      "num_tokens": 5962168.0,
      "step": 3230
    },
    {
      "entropy": 5.972162342071533,
      "epoch": 0.2717916404116782,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004996022781999198,
      "loss": 5.8634,
      "mean_token_accuracy": 0.14441936314105988,
      "num_tokens": 5971627.0,
      "step": 3235
    },
    {
      "entropy": 6.052579498291015,
      "epoch": 0.2722117202268431,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499600496422805,
      "loss": 5.9405,
      "mean_token_accuracy": 0.14112165570259094,
      "num_tokens": 5981775.0,
      "step": 3240
    },
    {
      "entropy": 6.032028532028198,
      "epoch": 0.272631800042008,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499598710667008,
      "loss": 5.8729,
      "mean_token_accuracy": 0.14314236044883727,
      "num_tokens": 5991097.0,
      "step": 3245
    },
    {
      "entropy": 6.038925361633301,
      "epoch": 0.2730518798571729,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995969209325604,
      "loss": 5.9177,
      "mean_token_accuracy": 0.13474914729595183,
      "num_tokens": 5999517.0,
      "step": 3250
    },
    {
      "entropy": 6.042639112472534,
      "epoch": 0.2734719596723377,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995951272194941,
      "loss": 5.9046,
      "mean_token_accuracy": 0.13345034569501876,
      "num_tokens": 6008545.0,
      "step": 3255
    },
    {
      "entropy": 6.164342308044434,
      "epoch": 0.2738920394875026,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995933295278407,
      "loss": 5.8797,
      "mean_token_accuracy": 0.13910974860191344,
      "num_tokens": 6017366.0,
      "step": 3260
    },
    {
      "entropy": 6.04072585105896,
      "epoch": 0.2743121193026675,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995915278576321,
      "loss": 5.8405,
      "mean_token_accuracy": 0.14348409026861192,
      "num_tokens": 6025597.0,
      "step": 3265
    },
    {
      "entropy": 6.078993320465088,
      "epoch": 0.27473219911783237,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995897222089004,
      "loss": 5.9484,
      "mean_token_accuracy": 0.14113520383834838,
      "num_tokens": 6034239.0,
      "step": 3270
    },
    {
      "entropy": 6.241492223739624,
      "epoch": 0.27515227893299726,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995879125816772,
      "loss": 5.9783,
      "mean_token_accuracy": 0.14251597672700883,
      "num_tokens": 6043837.0,
      "step": 3275
    },
    {
      "entropy": 6.017462110519409,
      "epoch": 0.27557235874816216,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004995860989759949,
      "loss": 5.9027,
      "mean_token_accuracy": 0.1499466359615326,
      "num_tokens": 6053217.0,
      "step": 3280
    },
    {
      "entropy": 6.126773357391357,
      "epoch": 0.27599243856332706,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995842813918855,
      "loss": 5.9321,
      "mean_token_accuracy": 0.14304194226861,
      "num_tokens": 6061553.0,
      "step": 3285
    },
    {
      "entropy": 6.039501285552978,
      "epoch": 0.2764125183784919,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995824598293812,
      "loss": 5.8019,
      "mean_token_accuracy": 0.14597292095422745,
      "num_tokens": 6070080.0,
      "step": 3290
    },
    {
      "entropy": 6.095096731185913,
      "epoch": 0.2768325981936568,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995806342885142,
      "loss": 5.9473,
      "mean_token_accuracy": 0.1479337990283966,
      "num_tokens": 6078438.0,
      "step": 3295
    },
    {
      "entropy": 6.091133499145508,
      "epoch": 0.2772526780088217,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499578804769317,
      "loss": 5.9386,
      "mean_token_accuracy": 0.1408208929002285,
      "num_tokens": 6087794.0,
      "step": 3300
    },
    {
      "entropy": 6.1423423290252686,
      "epoch": 0.27767275782398654,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004995769712718218,
      "loss": 5.9397,
      "mean_token_accuracy": 0.14154492914676667,
      "num_tokens": 6096709.0,
      "step": 3305
    },
    {
      "entropy": 6.03666353225708,
      "epoch": 0.27809283763915144,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995751337960613,
      "loss": 5.8786,
      "mean_token_accuracy": 0.13932624831795692,
      "num_tokens": 6105866.0,
      "step": 3310
    },
    {
      "entropy": 6.069159746170044,
      "epoch": 0.27851291745431633,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995732923420679,
      "loss": 5.8264,
      "mean_token_accuracy": 0.14724719822406768,
      "num_tokens": 6114882.0,
      "step": 3315
    },
    {
      "entropy": 5.9742859363555905,
      "epoch": 0.2789329972694812,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995714469098743,
      "loss": 5.8038,
      "mean_token_accuracy": 0.14107296913862227,
      "num_tokens": 6123978.0,
      "step": 3320
    },
    {
      "entropy": 6.065083360671997,
      "epoch": 0.2793530770846461,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499569597499513,
      "loss": 5.9488,
      "mean_token_accuracy": 0.14292607307434083,
      "num_tokens": 6133246.0,
      "step": 3325
    },
    {
      "entropy": 6.032466030120849,
      "epoch": 0.27977315689981097,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004995677441110172,
      "loss": 5.7988,
      "mean_token_accuracy": 0.14654416292905809,
      "num_tokens": 6142865.0,
      "step": 3330
    },
    {
      "entropy": 6.065610027313232,
      "epoch": 0.28019323671497587,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995658867444192,
      "loss": 5.9007,
      "mean_token_accuracy": 0.1368338242173195,
      "num_tokens": 6152492.0,
      "step": 3335
    },
    {
      "entropy": 6.032893276214599,
      "epoch": 0.2806133165301407,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995640253997523,
      "loss": 5.9101,
      "mean_token_accuracy": 0.13812195584177972,
      "num_tokens": 6161953.0,
      "step": 3340
    },
    {
      "entropy": 5.920506429672241,
      "epoch": 0.2810333963453056,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004995621600770492,
      "loss": 5.7589,
      "mean_token_accuracy": 0.15040778815746308,
      "num_tokens": 6171467.0,
      "step": 3345
    },
    {
      "entropy": 6.0409363269805905,
      "epoch": 0.2814534761604705,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004995602907763431,
      "loss": 5.84,
      "mean_token_accuracy": 0.14080849960446357,
      "num_tokens": 6180646.0,
      "step": 3350
    },
    {
      "entropy": 6.033467626571655,
      "epoch": 0.28187355597563535,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995584174976672,
      "loss": 5.8301,
      "mean_token_accuracy": 0.13724913001060485,
      "num_tokens": 6189832.0,
      "step": 3355
    },
    {
      "entropy": 5.980023050308228,
      "epoch": 0.28229363579080025,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995565402410544,
      "loss": 5.7669,
      "mean_token_accuracy": 0.15104614943265915,
      "num_tokens": 6198339.0,
      "step": 3360
    },
    {
      "entropy": 6.0385888576507565,
      "epoch": 0.28271371560596514,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995546590065383,
      "loss": 5.8364,
      "mean_token_accuracy": 0.14058903455734253,
      "num_tokens": 6207564.0,
      "step": 3365
    },
    {
      "entropy": 6.024218225479126,
      "epoch": 0.28313379542113004,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995527737941518,
      "loss": 5.8995,
      "mean_token_accuracy": 0.14139530062675476,
      "num_tokens": 6216056.0,
      "step": 3370
    },
    {
      "entropy": 5.981421661376953,
      "epoch": 0.2835538752362949,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995508846039287,
      "loss": 5.8633,
      "mean_token_accuracy": 0.14600031226873397,
      "num_tokens": 6225573.0,
      "step": 3375
    },
    {
      "entropy": 6.090993881225586,
      "epoch": 0.2839739550514598,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995489914359023,
      "loss": 5.9921,
      "mean_token_accuracy": 0.13607787638902663,
      "num_tokens": 6235057.0,
      "step": 3380
    },
    {
      "entropy": 6.113264799118042,
      "epoch": 0.2843940348666247,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995470942901061,
      "loss": 5.8982,
      "mean_token_accuracy": 0.14372147843241692,
      "num_tokens": 6244164.0,
      "step": 3385
    },
    {
      "entropy": 6.123813199996948,
      "epoch": 0.2848141146817895,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995451931665738,
      "loss": 5.9112,
      "mean_token_accuracy": 0.14370959922671317,
      "num_tokens": 6253095.0,
      "step": 3390
    },
    {
      "entropy": 6.027105283737183,
      "epoch": 0.2852341944969544,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499543288065339,
      "loss": 5.8576,
      "mean_token_accuracy": 0.1425433225929737,
      "num_tokens": 6261134.0,
      "step": 3395
    },
    {
      "entropy": 5.999712657928467,
      "epoch": 0.2856542743121193,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995413789864354,
      "loss": 5.8402,
      "mean_token_accuracy": 0.14678002744913102,
      "num_tokens": 6270384.0,
      "step": 3400
    },
    {
      "entropy": 5.960616159439087,
      "epoch": 0.28607435412728416,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995394659298971,
      "loss": 5.7915,
      "mean_token_accuracy": 0.15043452978134156,
      "num_tokens": 6279702.0,
      "step": 3405
    },
    {
      "entropy": 6.005735635757446,
      "epoch": 0.28649443394244906,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995375488957576,
      "loss": 5.8309,
      "mean_token_accuracy": 0.13979510068893433,
      "num_tokens": 6288297.0,
      "step": 3410
    },
    {
      "entropy": 6.013660907745361,
      "epoch": 0.28691451375761395,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499535627884051,
      "loss": 5.9396,
      "mean_token_accuracy": 0.13572793826460838,
      "num_tokens": 6297288.0,
      "step": 3415
    },
    {
      "entropy": 6.181674480438232,
      "epoch": 0.28733459357277885,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004995337028948115,
      "loss": 5.9473,
      "mean_token_accuracy": 0.13616690784692764,
      "num_tokens": 6306719.0,
      "step": 3420
    },
    {
      "entropy": 6.01987738609314,
      "epoch": 0.2877546733879437,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995317739280731,
      "loss": 5.7786,
      "mean_token_accuracy": 0.15149419456720353,
      "num_tokens": 6316639.0,
      "step": 3425
    },
    {
      "entropy": 6.0435620784759525,
      "epoch": 0.2881747532031086,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995298409838699,
      "loss": 5.8956,
      "mean_token_accuracy": 0.14308716356754303,
      "num_tokens": 6326879.0,
      "step": 3430
    },
    {
      "entropy": 5.951386880874634,
      "epoch": 0.2885948330182735,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499527904062236,
      "loss": 5.7971,
      "mean_token_accuracy": 0.147957743704319,
      "num_tokens": 6335729.0,
      "step": 3435
    },
    {
      "entropy": 6.049512481689453,
      "epoch": 0.28901491283343833,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995259631632061,
      "loss": 5.8867,
      "mean_token_accuracy": 0.13639168068766594,
      "num_tokens": 6345154.0,
      "step": 3440
    },
    {
      "entropy": 6.043827152252197,
      "epoch": 0.28943499264860323,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004995240182868143,
      "loss": 5.8349,
      "mean_token_accuracy": 0.14297354817390442,
      "num_tokens": 6354309.0,
      "step": 3445
    },
    {
      "entropy": 5.9356239318847654,
      "epoch": 0.2898550724637681,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004995220694330951,
      "loss": 5.7975,
      "mean_token_accuracy": 0.14842799976468085,
      "num_tokens": 6363389.0,
      "step": 3450
    },
    {
      "entropy": 6.005477046966552,
      "epoch": 0.290275152278933,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995201166020832,
      "loss": 5.8706,
      "mean_token_accuracy": 0.14157460033893585,
      "num_tokens": 6372475.0,
      "step": 3455
    },
    {
      "entropy": 6.111435508728027,
      "epoch": 0.29069523209409787,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499518159793813,
      "loss": 5.8116,
      "mean_token_accuracy": 0.15400618463754653,
      "num_tokens": 6380906.0,
      "step": 3460
    },
    {
      "entropy": 5.945709609985352,
      "epoch": 0.29111531190926276,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499516199008319,
      "loss": 5.833,
      "mean_token_accuracy": 0.1436382308602333,
      "num_tokens": 6390085.0,
      "step": 3465
    },
    {
      "entropy": 6.070226430892944,
      "epoch": 0.29153539172442766,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995142342456364,
      "loss": 5.887,
      "mean_token_accuracy": 0.14058329910039902,
      "num_tokens": 6399441.0,
      "step": 3470
    },
    {
      "entropy": 6.092825603485108,
      "epoch": 0.2919554715395925,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995122655057997,
      "loss": 5.9566,
      "mean_token_accuracy": 0.14255323261022568,
      "num_tokens": 6408995.0,
      "step": 3475
    },
    {
      "entropy": 5.959508466720581,
      "epoch": 0.2923755513547574,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995102927888437,
      "loss": 5.7104,
      "mean_token_accuracy": 0.15224269554018974,
      "num_tokens": 6418080.0,
      "step": 3480
    },
    {
      "entropy": 6.040865230560303,
      "epoch": 0.2927956311699223,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995083160948036,
      "loss": 5.8828,
      "mean_token_accuracy": 0.13958164304494858,
      "num_tokens": 6426732.0,
      "step": 3485
    },
    {
      "entropy": 5.988566875457764,
      "epoch": 0.29321571098508714,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995063354237141,
      "loss": 5.8849,
      "mean_token_accuracy": 0.14826264083385468,
      "num_tokens": 6435957.0,
      "step": 3490
    },
    {
      "entropy": 6.052132320404053,
      "epoch": 0.29363579080025204,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995043507756107,
      "loss": 5.8284,
      "mean_token_accuracy": 0.14379840195178986,
      "num_tokens": 6445642.0,
      "step": 3495
    },
    {
      "entropy": 6.018087530136109,
      "epoch": 0.29405587061541694,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995023621505282,
      "loss": 5.8853,
      "mean_token_accuracy": 0.1442800186574459,
      "num_tokens": 6454664.0,
      "step": 3500
    },
    {
      "entropy": 5.923487854003906,
      "epoch": 0.29447595043058183,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499500369548502,
      "loss": 5.8031,
      "mean_token_accuracy": 0.14537445902824403,
      "num_tokens": 6463224.0,
      "step": 3505
    },
    {
      "entropy": 6.171301317214966,
      "epoch": 0.2948960302457467,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994983729695674,
      "loss": 6.0242,
      "mean_token_accuracy": 0.13893685415387153,
      "num_tokens": 6473112.0,
      "step": 3510
    },
    {
      "entropy": 6.045923900604248,
      "epoch": 0.2953161100609116,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994963724137595,
      "loss": 5.8734,
      "mean_token_accuracy": 0.1453737996518612,
      "num_tokens": 6482062.0,
      "step": 3515
    },
    {
      "entropy": 6.005832099914551,
      "epoch": 0.29573618987607647,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994943678811142,
      "loss": 5.8543,
      "mean_token_accuracy": 0.1401112914085388,
      "num_tokens": 6490568.0,
      "step": 3520
    },
    {
      "entropy": 6.070190906524658,
      "epoch": 0.2961562696912413,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004994923593716667,
      "loss": 5.9152,
      "mean_token_accuracy": 0.14401544779539108,
      "num_tokens": 6500815.0,
      "step": 3525
    },
    {
      "entropy": 5.988713598251342,
      "epoch": 0.2965763495064062,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994903468854527,
      "loss": 5.7801,
      "mean_token_accuracy": 0.1514952600002289,
      "num_tokens": 6509529.0,
      "step": 3530
    },
    {
      "entropy": 5.968388557434082,
      "epoch": 0.2969964293215711,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994883304225077,
      "loss": 5.8498,
      "mean_token_accuracy": 0.1457517549395561,
      "num_tokens": 6517934.0,
      "step": 3535
    },
    {
      "entropy": 6.0856983184814455,
      "epoch": 0.297416509136736,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004994863099828675,
      "loss": 5.824,
      "mean_token_accuracy": 0.14552584439516067,
      "num_tokens": 6526098.0,
      "step": 3540
    },
    {
      "entropy": 5.9941321849823,
      "epoch": 0.29783658895190085,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499484285566568,
      "loss": 5.8578,
      "mean_token_accuracy": 0.1432398647069931,
      "num_tokens": 6535831.0,
      "step": 3545
    },
    {
      "entropy": 5.94168758392334,
      "epoch": 0.29825666876706575,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004994822571736449,
      "loss": 5.7625,
      "mean_token_accuracy": 0.14396894872188568,
      "num_tokens": 6545704.0,
      "step": 3550
    },
    {
      "entropy": 6.0284003734588625,
      "epoch": 0.29867674858223064,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994802248041342,
      "loss": 5.7903,
      "mean_token_accuracy": 0.14962279424071312,
      "num_tokens": 6554423.0,
      "step": 3555
    },
    {
      "entropy": 6.024318981170654,
      "epoch": 0.2990968283973955,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499478188458072,
      "loss": 5.8398,
      "mean_token_accuracy": 0.1454595111310482,
      "num_tokens": 6563989.0,
      "step": 3560
    },
    {
      "entropy": 6.043286228179932,
      "epoch": 0.2995169082125604,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994761481354943,
      "loss": 5.9785,
      "mean_token_accuracy": 0.14439435824751853,
      "num_tokens": 6572745.0,
      "step": 3565
    },
    {
      "entropy": 6.224496841430664,
      "epoch": 0.2999369880277253,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004994741038364371,
      "loss": 5.9792,
      "mean_token_accuracy": 0.1366976760327816,
      "num_tokens": 6581723.0,
      "step": 3570
    },
    {
      "entropy": 5.943874788284302,
      "epoch": 0.3003570678428901,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994720555609369,
      "loss": 5.7171,
      "mean_token_accuracy": 0.14883559197187424,
      "num_tokens": 6590342.0,
      "step": 3575
    },
    {
      "entropy": 5.911714887619018,
      "epoch": 0.300777147658055,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994700033090297,
      "loss": 5.7724,
      "mean_token_accuracy": 0.15467880815267562,
      "num_tokens": 6599206.0,
      "step": 3580
    },
    {
      "entropy": 6.1032740592956545,
      "epoch": 0.3011972274732199,
      "grad_norm": 1.0,
      "learning_rate": 0.000499467947080752,
      "loss": 6.0443,
      "mean_token_accuracy": 0.1365446515381336,
      "num_tokens": 6608947.0,
      "step": 3585
    },
    {
      "entropy": 6.099864816665649,
      "epoch": 0.3016173072883848,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994658868761402,
      "loss": 5.8715,
      "mean_token_accuracy": 0.14974160194396974,
      "num_tokens": 6618378.0,
      "step": 3590
    },
    {
      "entropy": 5.99675874710083,
      "epoch": 0.30203738710354966,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994638226952307,
      "loss": 5.9162,
      "mean_token_accuracy": 0.1387033462524414,
      "num_tokens": 6627527.0,
      "step": 3595
    },
    {
      "entropy": 6.10715618133545,
      "epoch": 0.30245746691871456,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994617545380604,
      "loss": 5.8484,
      "mean_token_accuracy": 0.14276407659053802,
      "num_tokens": 6636964.0,
      "step": 3600
    },
    {
      "entropy": 5.94628529548645,
      "epoch": 0.30287754673387945,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994596824046656,
      "loss": 5.8094,
      "mean_token_accuracy": 0.14509298652410507,
      "num_tokens": 6646074.0,
      "step": 3605
    },
    {
      "entropy": 6.038099336624145,
      "epoch": 0.3032976265490443,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499457606295083,
      "loss": 5.8782,
      "mean_token_accuracy": 0.14421509206295013,
      "num_tokens": 6655027.0,
      "step": 3610
    },
    {
      "entropy": 5.900383329391479,
      "epoch": 0.3037177063642092,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994555262093495,
      "loss": 5.6651,
      "mean_token_accuracy": 0.1515209808945656,
      "num_tokens": 6663747.0,
      "step": 3615
    },
    {
      "entropy": 6.1385828971862795,
      "epoch": 0.3041377861793741,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499453442147502,
      "loss": 5.9912,
      "mean_token_accuracy": 0.13430312201380729,
      "num_tokens": 6672922.0,
      "step": 3620
    },
    {
      "entropy": 6.011907768249512,
      "epoch": 0.304557865994539,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994513541095773,
      "loss": 5.7983,
      "mean_token_accuracy": 0.15209768787026406,
      "num_tokens": 6682233.0,
      "step": 3625
    },
    {
      "entropy": 5.979143190383911,
      "epoch": 0.30497794580970383,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994492620956126,
      "loss": 5.858,
      "mean_token_accuracy": 0.1495204821228981,
      "num_tokens": 6691593.0,
      "step": 3630
    },
    {
      "entropy": 6.0055419921875,
      "epoch": 0.30539802562486873,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004994471661056445,
      "loss": 5.8508,
      "mean_token_accuracy": 0.15055460929870607,
      "num_tokens": 6701318.0,
      "step": 3635
    },
    {
      "entropy": 6.076199960708618,
      "epoch": 0.3058181054400336,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994450661397106,
      "loss": 5.8559,
      "mean_token_accuracy": 0.14739347100257874,
      "num_tokens": 6710059.0,
      "step": 3640
    },
    {
      "entropy": 6.129534769058227,
      "epoch": 0.30623818525519847,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499442962197848,
      "loss": 5.9629,
      "mean_token_accuracy": 0.13774683177471161,
      "num_tokens": 6719811.0,
      "step": 3645
    },
    {
      "entropy": 5.9699358463287355,
      "epoch": 0.30665826507036337,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994408542800937,
      "loss": 5.8199,
      "mean_token_accuracy": 0.1469706729054451,
      "num_tokens": 6728789.0,
      "step": 3650
    },
    {
      "entropy": 5.977092027664185,
      "epoch": 0.30707834488552826,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994387423864855,
      "loss": 5.8145,
      "mean_token_accuracy": 0.14784644544124603,
      "num_tokens": 6737706.0,
      "step": 3655
    },
    {
      "entropy": 5.975707149505615,
      "epoch": 0.3074984247006931,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994366265170603,
      "loss": 5.7648,
      "mean_token_accuracy": 0.16093844026327134,
      "num_tokens": 6746861.0,
      "step": 3660
    },
    {
      "entropy": 6.098724269866944,
      "epoch": 0.307918504515858,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994345066718558,
      "loss": 5.9542,
      "mean_token_accuracy": 0.13871566131711005,
      "num_tokens": 6755242.0,
      "step": 3665
    },
    {
      "entropy": 6.0416632175445555,
      "epoch": 0.3083385843310229,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994323828509098,
      "loss": 5.8976,
      "mean_token_accuracy": 0.14007208868861198,
      "num_tokens": 6764549.0,
      "step": 3670
    },
    {
      "entropy": 5.993009233474732,
      "epoch": 0.3087586641461878,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994302550542596,
      "loss": 5.877,
      "mean_token_accuracy": 0.15204267650842668,
      "num_tokens": 6774123.0,
      "step": 3675
    },
    {
      "entropy": 5.879176998138428,
      "epoch": 0.30917874396135264,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499428123281943,
      "loss": 5.6764,
      "mean_token_accuracy": 0.15400367826223374,
      "num_tokens": 6782922.0,
      "step": 3680
    },
    {
      "entropy": 5.992125606536865,
      "epoch": 0.30959882377651754,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994259875339978,
      "loss": 5.9092,
      "mean_token_accuracy": 0.14591203182935714,
      "num_tokens": 6792042.0,
      "step": 3685
    },
    {
      "entropy": 6.150802707672119,
      "epoch": 0.31001890359168244,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994238478104617,
      "loss": 5.9127,
      "mean_token_accuracy": 0.14117644280195235,
      "num_tokens": 6800994.0,
      "step": 3690
    },
    {
      "entropy": 5.969630241394043,
      "epoch": 0.3104389834068473,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994217041113727,
      "loss": 5.8293,
      "mean_token_accuracy": 0.14889054223895073,
      "num_tokens": 6809938.0,
      "step": 3695
    },
    {
      "entropy": 6.048539876937866,
      "epoch": 0.3108590632220122,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004994195564367688,
      "loss": 5.9521,
      "mean_token_accuracy": 0.14025204703211785,
      "num_tokens": 6820289.0,
      "step": 3700
    },
    {
      "entropy": 6.0517502784729,
      "epoch": 0.3112791430371771,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994174047866882,
      "loss": 5.7924,
      "mean_token_accuracy": 0.14613612070679666,
      "num_tokens": 6830068.0,
      "step": 3705
    },
    {
      "entropy": 5.912678909301758,
      "epoch": 0.3116992228523419,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994152491611686,
      "loss": 5.8273,
      "mean_token_accuracy": 0.14581120014190674,
      "num_tokens": 6838591.0,
      "step": 3710
    },
    {
      "entropy": 5.982998657226562,
      "epoch": 0.3121193026675068,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994130895602485,
      "loss": 5.7956,
      "mean_token_accuracy": 0.14477218091487884,
      "num_tokens": 6847796.0,
      "step": 3715
    },
    {
      "entropy": 6.063366651535034,
      "epoch": 0.3125393824826717,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499410925983966,
      "loss": 5.8849,
      "mean_token_accuracy": 0.1452285960316658,
      "num_tokens": 6856585.0,
      "step": 3720
    },
    {
      "entropy": 5.898049783706665,
      "epoch": 0.3129594622978366,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994087584323596,
      "loss": 5.8038,
      "mean_token_accuracy": 0.15066958963871002,
      "num_tokens": 6865757.0,
      "step": 3725
    },
    {
      "entropy": 5.990089702606201,
      "epoch": 0.31337954211300145,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994065869054676,
      "loss": 5.8226,
      "mean_token_accuracy": 0.14529856890439988,
      "num_tokens": 6875371.0,
      "step": 3730
    },
    {
      "entropy": 6.091077661514282,
      "epoch": 0.31379962192816635,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004994044114033283,
      "loss": 5.8973,
      "mean_token_accuracy": 0.14080470502376558,
      "num_tokens": 6884050.0,
      "step": 3735
    },
    {
      "entropy": 6.091219329833985,
      "epoch": 0.31421970174333125,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994022319259806,
      "loss": 5.8638,
      "mean_token_accuracy": 0.14875171110033988,
      "num_tokens": 6893079.0,
      "step": 3740
    },
    {
      "entropy": 5.994270372390747,
      "epoch": 0.3146397815584961,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004994000484734629,
      "loss": 5.9464,
      "mean_token_accuracy": 0.14748625531792642,
      "num_tokens": 6903100.0,
      "step": 3745
    },
    {
      "entropy": 6.005830955505371,
      "epoch": 0.315059861373661,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004993978610458137,
      "loss": 5.7979,
      "mean_token_accuracy": 0.14891340732574462,
      "num_tokens": 6912164.0,
      "step": 3750
    },
    {
      "entropy": 5.970021390914917,
      "epoch": 0.3154799411888259,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993956696430721,
      "loss": 5.8128,
      "mean_token_accuracy": 0.14630623608827592,
      "num_tokens": 6921183.0,
      "step": 3755
    },
    {
      "entropy": 6.014796590805053,
      "epoch": 0.3159000210039908,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004993934742652768,
      "loss": 5.8857,
      "mean_token_accuracy": 0.1476880669593811,
      "num_tokens": 6931325.0,
      "step": 3760
    },
    {
      "entropy": 6.025224351882935,
      "epoch": 0.3163201008191556,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004993912749124665,
      "loss": 5.7949,
      "mean_token_accuracy": 0.15172581374645233,
      "num_tokens": 6940234.0,
      "step": 3765
    },
    {
      "entropy": 5.975033617019653,
      "epoch": 0.3167401806343205,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993890715846804,
      "loss": 5.875,
      "mean_token_accuracy": 0.14686811119318008,
      "num_tokens": 6949067.0,
      "step": 3770
    },
    {
      "entropy": 6.006247806549072,
      "epoch": 0.3171602604494854,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993868642819574,
      "loss": 5.8475,
      "mean_token_accuracy": 0.14386743307113647,
      "num_tokens": 6959085.0,
      "step": 3775
    },
    {
      "entropy": 5.984570598602295,
      "epoch": 0.31758034026465026,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993846530043367,
      "loss": 5.8889,
      "mean_token_accuracy": 0.14338204190135,
      "num_tokens": 6967392.0,
      "step": 3780
    },
    {
      "entropy": 5.983299684524536,
      "epoch": 0.31800042007981516,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993824377518574,
      "loss": 5.8149,
      "mean_token_accuracy": 0.1506180763244629,
      "num_tokens": 6976369.0,
      "step": 3785
    },
    {
      "entropy": 6.051072645187378,
      "epoch": 0.31842049989498006,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993802185245587,
      "loss": 5.8256,
      "mean_token_accuracy": 0.1495169520378113,
      "num_tokens": 6985889.0,
      "step": 3790
    },
    {
      "entropy": 5.973819303512573,
      "epoch": 0.3188405797101449,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00049937799532248,
      "loss": 5.8527,
      "mean_token_accuracy": 0.13678822666406631,
      "num_tokens": 6995396.0,
      "step": 3795
    },
    {
      "entropy": 6.154268550872803,
      "epoch": 0.3192606595253098,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993757681456607,
      "loss": 5.9123,
      "mean_token_accuracy": 0.14265349358320237,
      "num_tokens": 7004666.0,
      "step": 3800
    },
    {
      "entropy": 6.029079341888428,
      "epoch": 0.3196807393404747,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004993735369941401,
      "loss": 5.9402,
      "mean_token_accuracy": 0.13829794898629189,
      "num_tokens": 7014608.0,
      "step": 3805
    },
    {
      "entropy": 6.041004323959351,
      "epoch": 0.3201008191556396,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993713018679579,
      "loss": 5.8043,
      "mean_token_accuracy": 0.14656506404280661,
      "num_tokens": 7023671.0,
      "step": 3810
    },
    {
      "entropy": 6.019459915161133,
      "epoch": 0.32052089897080444,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993690627671536,
      "loss": 5.8571,
      "mean_token_accuracy": 0.13951837345957757,
      "num_tokens": 7033786.0,
      "step": 3815
    },
    {
      "entropy": 5.97732892036438,
      "epoch": 0.32094097878596933,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993668196917669,
      "loss": 5.7672,
      "mean_token_accuracy": 0.15004572570323943,
      "num_tokens": 7042162.0,
      "step": 3820
    },
    {
      "entropy": 6.027564477920532,
      "epoch": 0.32136105860113423,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004993645726418375,
      "loss": 5.8878,
      "mean_token_accuracy": 0.14644107297062875,
      "num_tokens": 7051903.0,
      "step": 3825
    },
    {
      "entropy": 5.969810867309571,
      "epoch": 0.3217811384162991,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993623216174053,
      "loss": 5.7484,
      "mean_token_accuracy": 0.1605222076177597,
      "num_tokens": 7060229.0,
      "step": 3830
    },
    {
      "entropy": 5.917182302474975,
      "epoch": 0.32220121823146397,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00049936006661851,
      "loss": 5.8379,
      "mean_token_accuracy": 0.14710989743471145,
      "num_tokens": 7069040.0,
      "step": 3835
    },
    {
      "entropy": 5.992334508895874,
      "epoch": 0.32262129804662887,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993578076451917,
      "loss": 5.7103,
      "mean_token_accuracy": 0.1498522937297821,
      "num_tokens": 7078409.0,
      "step": 3840
    },
    {
      "entropy": 5.879006052017212,
      "epoch": 0.32304137786179377,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004993555446974903,
      "loss": 5.7865,
      "mean_token_accuracy": 0.14535705298185347,
      "num_tokens": 7087983.0,
      "step": 3845
    },
    {
      "entropy": 5.901397848129273,
      "epoch": 0.3234614576769586,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499353277775446,
      "loss": 5.7473,
      "mean_token_accuracy": 0.15219927430152894,
      "num_tokens": 7097277.0,
      "step": 3850
    },
    {
      "entropy": 5.960507011413574,
      "epoch": 0.3238815374921235,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993510068790989,
      "loss": 5.6508,
      "mean_token_accuracy": 0.16467221826314926,
      "num_tokens": 7105918.0,
      "step": 3855
    },
    {
      "entropy": 5.879998397827149,
      "epoch": 0.3243016173072884,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993487320084892,
      "loss": 5.7106,
      "mean_token_accuracy": 0.15136535316705704,
      "num_tokens": 7115049.0,
      "step": 3860
    },
    {
      "entropy": 5.982763433456421,
      "epoch": 0.32472169712245325,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004993464531636573,
      "loss": 5.8137,
      "mean_token_accuracy": 0.14752050638198852,
      "num_tokens": 7124862.0,
      "step": 3865
    },
    {
      "entropy": 5.8926033020019535,
      "epoch": 0.32514177693761814,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004993441703446435,
      "loss": 5.7049,
      "mean_token_accuracy": 0.15468737334012986,
      "num_tokens": 7133280.0,
      "step": 3870
    },
    {
      "entropy": 6.015898132324219,
      "epoch": 0.32556185675278304,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004993418835514882,
      "loss": 5.8984,
      "mean_token_accuracy": 0.14468531757593156,
      "num_tokens": 7142446.0,
      "step": 3875
    },
    {
      "entropy": 5.991433334350586,
      "epoch": 0.3259819365679479,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004993395927842321,
      "loss": 5.8105,
      "mean_token_accuracy": 0.14096677601337432,
      "num_tokens": 7152143.0,
      "step": 3880
    },
    {
      "entropy": 6.059468841552734,
      "epoch": 0.3264020163831128,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993372980429155,
      "loss": 5.888,
      "mean_token_accuracy": 0.1420806512236595,
      "num_tokens": 7162046.0,
      "step": 3885
    },
    {
      "entropy": 5.982543849945069,
      "epoch": 0.3268220961982777,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004993349993275792,
      "loss": 5.774,
      "mean_token_accuracy": 0.1508569300174713,
      "num_tokens": 7171557.0,
      "step": 3890
    },
    {
      "entropy": 5.789723920822143,
      "epoch": 0.3272421760134426,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004993326966382639,
      "loss": 5.692,
      "mean_token_accuracy": 0.15062906444072724,
      "num_tokens": 7180927.0,
      "step": 3895
    },
    {
      "entropy": 5.92172474861145,
      "epoch": 0.3276622558286074,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993303899750104,
      "loss": 5.7555,
      "mean_token_accuracy": 0.14762243777513503,
      "num_tokens": 7189552.0,
      "step": 3900
    },
    {
      "entropy": 6.041365337371826,
      "epoch": 0.3280823356437723,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993280793378595,
      "loss": 5.7899,
      "mean_token_accuracy": 0.14609607309103012,
      "num_tokens": 7197857.0,
      "step": 3905
    },
    {
      "entropy": 5.951012849807739,
      "epoch": 0.3285024154589372,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993257647268522,
      "loss": 5.7448,
      "mean_token_accuracy": 0.15203747153282166,
      "num_tokens": 7206785.0,
      "step": 3910
    },
    {
      "entropy": 5.97521710395813,
      "epoch": 0.32892249527410206,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004993234461420295,
      "loss": 5.8492,
      "mean_token_accuracy": 0.14864396750926973,
      "num_tokens": 7216360.0,
      "step": 3915
    },
    {
      "entropy": 5.943229389190674,
      "epoch": 0.32934257508926695,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993211235834326,
      "loss": 5.6438,
      "mean_token_accuracy": 0.16730980798602105,
      "num_tokens": 7224890.0,
      "step": 3920
    },
    {
      "entropy": 5.844398403167725,
      "epoch": 0.32976265490443185,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993187970511023,
      "loss": 5.6993,
      "mean_token_accuracy": 0.1706005111336708,
      "num_tokens": 7234442.0,
      "step": 3925
    },
    {
      "entropy": 5.963161754608154,
      "epoch": 0.33018273471959675,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993164665450801,
      "loss": 5.8445,
      "mean_token_accuracy": 0.14895002990961076,
      "num_tokens": 7244023.0,
      "step": 3930
    },
    {
      "entropy": 5.904555988311768,
      "epoch": 0.3306028145347616,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004993141320654072,
      "loss": 5.6987,
      "mean_token_accuracy": 0.15213721096515656,
      "num_tokens": 7253548.0,
      "step": 3935
    },
    {
      "entropy": 5.913322687149048,
      "epoch": 0.3310228943499265,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499311793612125,
      "loss": 5.7659,
      "mean_token_accuracy": 0.15264057517051696,
      "num_tokens": 7262962.0,
      "step": 3940
    },
    {
      "entropy": 6.002100086212158,
      "epoch": 0.3314429741650914,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004993094511852748,
      "loss": 5.7999,
      "mean_token_accuracy": 0.14923355728387833,
      "num_tokens": 7272234.0,
      "step": 3945
    },
    {
      "entropy": 5.9879292964935305,
      "epoch": 0.33186305398025623,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993071047848983,
      "loss": 5.7677,
      "mean_token_accuracy": 0.14813442677259445,
      "num_tokens": 7281524.0,
      "step": 3950
    },
    {
      "entropy": 5.884419012069702,
      "epoch": 0.3322831337954211,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993047544110368,
      "loss": 5.6773,
      "mean_token_accuracy": 0.15349545404314996,
      "num_tokens": 7289601.0,
      "step": 3955
    },
    {
      "entropy": 5.788786172866821,
      "epoch": 0.332703213610586,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993024000637321,
      "loss": 5.6501,
      "mean_token_accuracy": 0.1583729490637779,
      "num_tokens": 7298508.0,
      "step": 3960
    },
    {
      "entropy": 5.918163013458252,
      "epoch": 0.33312329342575087,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993000417430259,
      "loss": 5.8565,
      "mean_token_accuracy": 0.1440684713423252,
      "num_tokens": 7309065.0,
      "step": 3965
    },
    {
      "entropy": 6.068691110610962,
      "epoch": 0.33354337324091576,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00049929767944896,
      "loss": 5.8795,
      "mean_token_accuracy": 0.14953403770923615,
      "num_tokens": 7319669.0,
      "step": 3970
    },
    {
      "entropy": 6.0281017303466795,
      "epoch": 0.33396345305608066,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992953131815761,
      "loss": 5.8292,
      "mean_token_accuracy": 0.1508055806159973,
      "num_tokens": 7328425.0,
      "step": 3975
    },
    {
      "entropy": 5.922671604156494,
      "epoch": 0.33438353287124556,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004992929429409164,
      "loss": 5.7009,
      "mean_token_accuracy": 0.15846544951200486,
      "num_tokens": 7337369.0,
      "step": 3980
    },
    {
      "entropy": 5.8830712795257565,
      "epoch": 0.3348036126864104,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992905687270225,
      "loss": 5.7797,
      "mean_token_accuracy": 0.1513069674372673,
      "num_tokens": 7346829.0,
      "step": 3985
    },
    {
      "entropy": 5.999729156494141,
      "epoch": 0.3352236925015753,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992881905399368,
      "loss": 5.8263,
      "mean_token_accuracy": 0.1449328303337097,
      "num_tokens": 7355976.0,
      "step": 3990
    },
    {
      "entropy": 5.988975906372071,
      "epoch": 0.3356437723167402,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992858083797013,
      "loss": 5.7855,
      "mean_token_accuracy": 0.14504586905241013,
      "num_tokens": 7365210.0,
      "step": 3995
    },
    {
      "entropy": 5.991175603866577,
      "epoch": 0.33606385213190504,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992834222463581,
      "loss": 5.8537,
      "mean_token_accuracy": 0.13908673748373984,
      "num_tokens": 7374175.0,
      "step": 4000
    },
    {
      "entropy": 5.972196435928344,
      "epoch": 0.33648393194706994,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004992810321399496,
      "loss": 5.8636,
      "mean_token_accuracy": 0.14633251801133157,
      "num_tokens": 7383302.0,
      "step": 4005
    },
    {
      "entropy": 6.040848684310913,
      "epoch": 0.33690401176223483,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992786380605182,
      "loss": 5.8386,
      "mean_token_accuracy": 0.14676400497555733,
      "num_tokens": 7392746.0,
      "step": 4010
    },
    {
      "entropy": 5.938001918792724,
      "epoch": 0.33732409157739973,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992762400081062,
      "loss": 5.6882,
      "mean_token_accuracy": 0.15358891040086747,
      "num_tokens": 7401604.0,
      "step": 4015
    },
    {
      "entropy": 5.918675374984741,
      "epoch": 0.3377441713925646,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992738379827559,
      "loss": 5.789,
      "mean_token_accuracy": 0.14934294447302818,
      "num_tokens": 7410594.0,
      "step": 4020
    },
    {
      "entropy": 6.007134819030762,
      "epoch": 0.33816425120772947,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992714319845101,
      "loss": 5.6879,
      "mean_token_accuracy": 0.15554973632097244,
      "num_tokens": 7418831.0,
      "step": 4025
    },
    {
      "entropy": 5.873322439193726,
      "epoch": 0.33858433102289437,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992690220134116,
      "loss": 5.7446,
      "mean_token_accuracy": 0.15233736038208007,
      "num_tokens": 7427731.0,
      "step": 4030
    },
    {
      "entropy": 6.057994699478149,
      "epoch": 0.3390044108380592,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004992666080695027,
      "loss": 5.8518,
      "mean_token_accuracy": 0.1464371532201767,
      "num_tokens": 7436447.0,
      "step": 4035
    },
    {
      "entropy": 5.948624610900879,
      "epoch": 0.3394244906532241,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004992641901528262,
      "loss": 5.7627,
      "mean_token_accuracy": 0.15126800090074538,
      "num_tokens": 7445352.0,
      "step": 4040
    },
    {
      "entropy": 5.9620507717132565,
      "epoch": 0.339844570468389,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992617682634252,
      "loss": 5.8182,
      "mean_token_accuracy": 0.15263011008501054,
      "num_tokens": 7454298.0,
      "step": 4045
    },
    {
      "entropy": 5.97614221572876,
      "epoch": 0.34026465028355385,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004992593424013424,
      "loss": 5.8337,
      "mean_token_accuracy": 0.14937392994761467,
      "num_tokens": 7463543.0,
      "step": 4050
    },
    {
      "entropy": 6.002068901062012,
      "epoch": 0.34068473009871875,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004992569125666209,
      "loss": 5.8464,
      "mean_token_accuracy": 0.14506482928991318,
      "num_tokens": 7472701.0,
      "step": 4055
    },
    {
      "entropy": 6.06530499458313,
      "epoch": 0.34110480991388364,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004992544787593037,
      "loss": 5.8289,
      "mean_token_accuracy": 0.14331392347812652,
      "num_tokens": 7481123.0,
      "step": 4060
    },
    {
      "entropy": 5.987830924987793,
      "epoch": 0.34152488972904854,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992520409794338,
      "loss": 5.8815,
      "mean_token_accuracy": 0.14786751866340636,
      "num_tokens": 7490439.0,
      "step": 4065
    },
    {
      "entropy": 5.97078309059143,
      "epoch": 0.3419449695442134,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992495992270544,
      "loss": 5.7612,
      "mean_token_accuracy": 0.15085898339748383,
      "num_tokens": 7499326.0,
      "step": 4070
    },
    {
      "entropy": 5.969767665863037,
      "epoch": 0.3423650493593783,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004992471535022089,
      "loss": 5.8137,
      "mean_token_accuracy": 0.14518282786011696,
      "num_tokens": 7509407.0,
      "step": 4075
    },
    {
      "entropy": 5.964113235473633,
      "epoch": 0.3427851291745432,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992447038049405,
      "loss": 5.8624,
      "mean_token_accuracy": 0.14494732469320298,
      "num_tokens": 7518443.0,
      "step": 4080
    },
    {
      "entropy": 5.919957160949707,
      "epoch": 0.343205208989708,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004992422501352927,
      "loss": 5.7254,
      "mean_token_accuracy": 0.15868185311555863,
      "num_tokens": 7527609.0,
      "step": 4085
    },
    {
      "entropy": 5.9979315280914305,
      "epoch": 0.3436252888048729,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004992397924933089,
      "loss": 5.8053,
      "mean_token_accuracy": 0.14575055539608,
      "num_tokens": 7536890.0,
      "step": 4090
    },
    {
      "entropy": 5.9897191524505615,
      "epoch": 0.3440453686200378,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992373308790325,
      "loss": 5.7589,
      "mean_token_accuracy": 0.15681823194026948,
      "num_tokens": 7546509.0,
      "step": 4095
    },
    {
      "entropy": 5.846250724792481,
      "epoch": 0.3444654484352027,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004992348652925074,
      "loss": 5.8077,
      "mean_token_accuracy": 0.14950673207640647,
      "num_tokens": 7555336.0,
      "step": 4100
    },
    {
      "entropy": 5.961744737625122,
      "epoch": 0.34488552825036756,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992323957337771,
      "loss": 5.7555,
      "mean_token_accuracy": 0.14984272718429564,
      "num_tokens": 7565210.0,
      "step": 4105
    },
    {
      "entropy": 6.007063913345337,
      "epoch": 0.34530560806553245,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004992299222028855,
      "loss": 5.833,
      "mean_token_accuracy": 0.15445573925971984,
      "num_tokens": 7574516.0,
      "step": 4110
    },
    {
      "entropy": 5.898674392700196,
      "epoch": 0.34572568788069735,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004992274446998761,
      "loss": 5.6943,
      "mean_token_accuracy": 0.15527215451002122,
      "num_tokens": 7583219.0,
      "step": 4115
    },
    {
      "entropy": 5.95861120223999,
      "epoch": 0.3461457676958622,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992249632247929,
      "loss": 5.9258,
      "mean_token_accuracy": 0.1425651490688324,
      "num_tokens": 7592050.0,
      "step": 4120
    },
    {
      "entropy": 6.043584823608398,
      "epoch": 0.3465658475110271,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992224777776802,
      "loss": 5.7612,
      "mean_token_accuracy": 0.14657285809516907,
      "num_tokens": 7600718.0,
      "step": 4125
    },
    {
      "entropy": 5.957071924209595,
      "epoch": 0.346985927326192,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992199883585816,
      "loss": 5.7941,
      "mean_token_accuracy": 0.14970231726765632,
      "num_tokens": 7609191.0,
      "step": 4130
    },
    {
      "entropy": 5.986074066162109,
      "epoch": 0.34740600714135683,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004992174949675413,
      "loss": 5.8307,
      "mean_token_accuracy": 0.14882247895002365,
      "num_tokens": 7618509.0,
      "step": 4135
    },
    {
      "entropy": 5.975029325485229,
      "epoch": 0.34782608695652173,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992149976046037,
      "loss": 5.7472,
      "mean_token_accuracy": 0.150260391831398,
      "num_tokens": 7627851.0,
      "step": 4140
    },
    {
      "entropy": 5.925207090377808,
      "epoch": 0.3482461667716866,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004992124962698128,
      "loss": 5.8042,
      "mean_token_accuracy": 0.14737849459052085,
      "num_tokens": 7636748.0,
      "step": 4145
    },
    {
      "entropy": 5.972150611877441,
      "epoch": 0.3486662465868515,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499209990963213,
      "loss": 5.7307,
      "mean_token_accuracy": 0.1522408828139305,
      "num_tokens": 7645436.0,
      "step": 4150
    },
    {
      "entropy": 6.006196355819702,
      "epoch": 0.34908632640201637,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004992074816848487,
      "loss": 5.8544,
      "mean_token_accuracy": 0.14600287303328513,
      "num_tokens": 7655414.0,
      "step": 4155
    },
    {
      "entropy": 5.8425483226776125,
      "epoch": 0.34950640621718126,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004992049684347642,
      "loss": 5.6383,
      "mean_token_accuracy": 0.15359618812799453,
      "num_tokens": 7664295.0,
      "step": 4160
    },
    {
      "entropy": 5.967918491363525,
      "epoch": 0.34992648603234616,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992024512130042,
      "loss": 5.7694,
      "mean_token_accuracy": 0.1513090416789055,
      "num_tokens": 7673295.0,
      "step": 4165
    },
    {
      "entropy": 5.854374551773072,
      "epoch": 0.350346565847511,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004991999300196132,
      "loss": 5.7737,
      "mean_token_accuracy": 0.1478977158665657,
      "num_tokens": 7682932.0,
      "step": 4170
    },
    {
      "entropy": 6.003021907806397,
      "epoch": 0.3507666456626759,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991974048546359,
      "loss": 5.7846,
      "mean_token_accuracy": 0.14676762744784355,
      "num_tokens": 7692105.0,
      "step": 4175
    },
    {
      "entropy": 5.946457290649414,
      "epoch": 0.3511867254778408,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499194875718117,
      "loss": 5.7816,
      "mean_token_accuracy": 0.14983211755752562,
      "num_tokens": 7701294.0,
      "step": 4180
    },
    {
      "entropy": 5.9310308456420895,
      "epoch": 0.3516068052930057,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991923426101013,
      "loss": 5.7713,
      "mean_token_accuracy": 0.14655268639326097,
      "num_tokens": 7710964.0,
      "step": 4185
    },
    {
      "entropy": 6.007804489135742,
      "epoch": 0.35202688510817054,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991898055306337,
      "loss": 5.8946,
      "mean_token_accuracy": 0.14021267369389534,
      "num_tokens": 7719938.0,
      "step": 4190
    },
    {
      "entropy": 6.032896995544434,
      "epoch": 0.35244696492333544,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004991872644797591,
      "loss": 5.8087,
      "mean_token_accuracy": 0.14767822921276091,
      "num_tokens": 7729129.0,
      "step": 4195
    },
    {
      "entropy": 5.918514633178711,
      "epoch": 0.35286704473850034,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991847194575226,
      "loss": 5.8201,
      "mean_token_accuracy": 0.14519443809986116,
      "num_tokens": 7738506.0,
      "step": 4200
    },
    {
      "entropy": 6.051172924041748,
      "epoch": 0.3532871245536652,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991821704639693,
      "loss": 5.9203,
      "mean_token_accuracy": 0.14418965578079224,
      "num_tokens": 7749320.0,
      "step": 4205
    },
    {
      "entropy": 6.062065458297729,
      "epoch": 0.3537072043688301,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004991796174991443,
      "loss": 5.7829,
      "mean_token_accuracy": 0.14587843269109727,
      "num_tokens": 7758735.0,
      "step": 4210
    },
    {
      "entropy": 5.849421215057373,
      "epoch": 0.354127284183995,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991770605630927,
      "loss": 5.751,
      "mean_token_accuracy": 0.15034677758812903,
      "num_tokens": 7767556.0,
      "step": 4215
    },
    {
      "entropy": 5.936790323257446,
      "epoch": 0.3545473639991598,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004991744996558599,
      "loss": 5.7616,
      "mean_token_accuracy": 0.15403031408786774,
      "num_tokens": 7776615.0,
      "step": 4220
    },
    {
      "entropy": 5.955025720596313,
      "epoch": 0.3549674438143247,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991719347774913,
      "loss": 5.7946,
      "mean_token_accuracy": 0.1537439852952957,
      "num_tokens": 7785288.0,
      "step": 4225
    },
    {
      "entropy": 5.960596990585327,
      "epoch": 0.3553875236294896,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004991693659280324,
      "loss": 5.715,
      "mean_token_accuracy": 0.1543234631419182,
      "num_tokens": 7794381.0,
      "step": 4230
    },
    {
      "entropy": 5.904143905639648,
      "epoch": 0.3558076034446545,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004991667931075284,
      "loss": 5.6835,
      "mean_token_accuracy": 0.14821650087833405,
      "num_tokens": 7803265.0,
      "step": 4235
    },
    {
      "entropy": 5.919279670715332,
      "epoch": 0.35622768325981935,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991642163160252,
      "loss": 5.7894,
      "mean_token_accuracy": 0.15157712996006012,
      "num_tokens": 7812445.0,
      "step": 4240
    },
    {
      "entropy": 6.0007213115692135,
      "epoch": 0.35664776307498425,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004991616355535684,
      "loss": 5.7926,
      "mean_token_accuracy": 0.15207020044326783,
      "num_tokens": 7822073.0,
      "step": 4245
    },
    {
      "entropy": 5.957857275009156,
      "epoch": 0.35706784289014915,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991590508202036,
      "loss": 5.7608,
      "mean_token_accuracy": 0.15158121064305305,
      "num_tokens": 7831193.0,
      "step": 4250
    },
    {
      "entropy": 5.963643503189087,
      "epoch": 0.357487922705314,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004991564621159766,
      "loss": 5.8016,
      "mean_token_accuracy": 0.14848409742116928,
      "num_tokens": 7840311.0,
      "step": 4255
    },
    {
      "entropy": 5.973238897323609,
      "epoch": 0.3579080025204789,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991538694409334,
      "loss": 5.8361,
      "mean_token_accuracy": 0.14516036435961724,
      "num_tokens": 7849622.0,
      "step": 4260
    },
    {
      "entropy": 5.913772773742676,
      "epoch": 0.3583280823356438,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004991512727951198,
      "loss": 5.7903,
      "mean_token_accuracy": 0.1450696036219597,
      "num_tokens": 7859494.0,
      "step": 4265
    },
    {
      "entropy": 6.065109920501709,
      "epoch": 0.3587481621508087,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991486721785818,
      "loss": 5.8808,
      "mean_token_accuracy": 0.1437905862927437,
      "num_tokens": 7868526.0,
      "step": 4270
    },
    {
      "entropy": 5.933383369445801,
      "epoch": 0.3591682419659735,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004991460675913655,
      "loss": 5.7167,
      "mean_token_accuracy": 0.14970996454358101,
      "num_tokens": 7877631.0,
      "step": 4275
    },
    {
      "entropy": 5.946811819076538,
      "epoch": 0.3595883217811384,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499143459033517,
      "loss": 5.7765,
      "mean_token_accuracy": 0.15149033516645433,
      "num_tokens": 7886814.0,
      "step": 4280
    },
    {
      "entropy": 5.832566404342652,
      "epoch": 0.3600084015963033,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991408465050825,
      "loss": 5.6169,
      "mean_token_accuracy": 0.15575773864984513,
      "num_tokens": 7896337.0,
      "step": 4285
    },
    {
      "entropy": 5.902747249603271,
      "epoch": 0.36042848141146816,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004991382300061084,
      "loss": 5.8474,
      "mean_token_accuracy": 0.14084347635507583,
      "num_tokens": 7906071.0,
      "step": 4290
    },
    {
      "entropy": 5.995443725585938,
      "epoch": 0.36084856122663306,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004991356095366409,
      "loss": 5.8339,
      "mean_token_accuracy": 0.15023995712399482,
      "num_tokens": 7915003.0,
      "step": 4295
    },
    {
      "entropy": 5.958809423446655,
      "epoch": 0.36126864104179796,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991329850967266,
      "loss": 5.724,
      "mean_token_accuracy": 0.15122520327568054,
      "num_tokens": 7924408.0,
      "step": 4300
    },
    {
      "entropy": 5.900264692306519,
      "epoch": 0.3616887208569628,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991303566864118,
      "loss": 5.6744,
      "mean_token_accuracy": 0.1535492181777954,
      "num_tokens": 7934717.0,
      "step": 4305
    },
    {
      "entropy": 5.847307777404785,
      "epoch": 0.3621088006721277,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991277243057431,
      "loss": 5.7314,
      "mean_token_accuracy": 0.14550352916121484,
      "num_tokens": 7944278.0,
      "step": 4310
    },
    {
      "entropy": 5.89778242111206,
      "epoch": 0.3625288804872926,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004991250879547673,
      "loss": 5.7601,
      "mean_token_accuracy": 0.15223362892866135,
      "num_tokens": 7953344.0,
      "step": 4315
    },
    {
      "entropy": 5.903073406219482,
      "epoch": 0.3629489603024575,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004991224476335309,
      "loss": 5.7762,
      "mean_token_accuracy": 0.14970767721533776,
      "num_tokens": 7962869.0,
      "step": 4320
    },
    {
      "entropy": 6.026950740814209,
      "epoch": 0.36336904011762233,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991198033420807,
      "loss": 5.7736,
      "mean_token_accuracy": 0.14771935492753982,
      "num_tokens": 7971981.0,
      "step": 4325
    },
    {
      "entropy": 5.912839651107788,
      "epoch": 0.36378911993278723,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004991171550804636,
      "loss": 5.7281,
      "mean_token_accuracy": 0.14824195951223373,
      "num_tokens": 7980979.0,
      "step": 4330
    },
    {
      "entropy": 5.94692907333374,
      "epoch": 0.36420919974795213,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991145028487266,
      "loss": 5.8003,
      "mean_token_accuracy": 0.1460677921772003,
      "num_tokens": 7989607.0,
      "step": 4335
    },
    {
      "entropy": 5.898593473434448,
      "epoch": 0.36462927956311697,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991118466469165,
      "loss": 5.6285,
      "mean_token_accuracy": 0.15354615002870559,
      "num_tokens": 7998356.0,
      "step": 4340
    },
    {
      "entropy": 5.863179397583008,
      "epoch": 0.36504935937828187,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004991091864750805,
      "loss": 5.726,
      "mean_token_accuracy": 0.15553184300661088,
      "num_tokens": 8007596.0,
      "step": 4345
    },
    {
      "entropy": 5.962289476394654,
      "epoch": 0.36546943919344677,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004991065223332655,
      "loss": 5.7878,
      "mean_token_accuracy": 0.14825710654258728,
      "num_tokens": 8016493.0,
      "step": 4350
    },
    {
      "entropy": 5.949162864685059,
      "epoch": 0.36588951900861166,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991038542215191,
      "loss": 5.7608,
      "mean_token_accuracy": 0.14747046008706094,
      "num_tokens": 8025867.0,
      "step": 4355
    },
    {
      "entropy": 5.890843343734741,
      "epoch": 0.3663095988237765,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004991011821398882,
      "loss": 5.7889,
      "mean_token_accuracy": 0.15656512528657912,
      "num_tokens": 8036251.0,
      "step": 4360
    },
    {
      "entropy": 5.949850082397461,
      "epoch": 0.3667296786389414,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990985060884202,
      "loss": 5.7395,
      "mean_token_accuracy": 0.1522827535867691,
      "num_tokens": 8045647.0,
      "step": 4365
    },
    {
      "entropy": 5.987407255172729,
      "epoch": 0.3671497584541063,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004990958260671627,
      "loss": 5.8127,
      "mean_token_accuracy": 0.1425602063536644,
      "num_tokens": 8056025.0,
      "step": 4370
    },
    {
      "entropy": 5.904713535308838,
      "epoch": 0.36756983826927114,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990931420761629,
      "loss": 5.7529,
      "mean_token_accuracy": 0.15707799047231674,
      "num_tokens": 8065029.0,
      "step": 4375
    },
    {
      "entropy": 5.966267681121826,
      "epoch": 0.36798991808443604,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004990904541154685,
      "loss": 5.7108,
      "mean_token_accuracy": 0.16287750899791717,
      "num_tokens": 8073249.0,
      "step": 4380
    },
    {
      "entropy": 5.987148952484131,
      "epoch": 0.36840999789960094,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990877621851271,
      "loss": 5.8252,
      "mean_token_accuracy": 0.1494595602154732,
      "num_tokens": 8082039.0,
      "step": 4385
    },
    {
      "entropy": 5.861302328109741,
      "epoch": 0.3688300777147658,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004990850662851863,
      "loss": 5.6716,
      "mean_token_accuracy": 0.15636952966451645,
      "num_tokens": 8090011.0,
      "step": 4390
    },
    {
      "entropy": 5.960203027725219,
      "epoch": 0.3692501575299307,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004990823664156941,
      "loss": 5.7847,
      "mean_token_accuracy": 0.1592693328857422,
      "num_tokens": 8099934.0,
      "step": 4395
    },
    {
      "entropy": 5.991479682922363,
      "epoch": 0.3696702373450956,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004990796625766981,
      "loss": 5.7909,
      "mean_token_accuracy": 0.1480449065566063,
      "num_tokens": 8108969.0,
      "step": 4400
    },
    {
      "entropy": 5.9183997631073,
      "epoch": 0.3700903171602605,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004990769547682462,
      "loss": 5.7222,
      "mean_token_accuracy": 0.15052928328514098,
      "num_tokens": 8117372.0,
      "step": 4405
    },
    {
      "entropy": 5.994330787658692,
      "epoch": 0.3705103969754253,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004990742429903866,
      "loss": 5.8896,
      "mean_token_accuracy": 0.14699996411800384,
      "num_tokens": 8127108.0,
      "step": 4410
    },
    {
      "entropy": 6.001091718673706,
      "epoch": 0.3709304767905902,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499071527243167,
      "loss": 5.8728,
      "mean_token_accuracy": 0.14656655341386796,
      "num_tokens": 8137392.0,
      "step": 4415
    },
    {
      "entropy": 5.959682178497315,
      "epoch": 0.3713505566057551,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004990688075266357,
      "loss": 5.7523,
      "mean_token_accuracy": 0.15625606924295427,
      "num_tokens": 8146257.0,
      "step": 4420
    },
    {
      "entropy": 5.867766714096069,
      "epoch": 0.37177063642091995,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004990660838408409,
      "loss": 5.7074,
      "mean_token_accuracy": 0.1495484985411167,
      "num_tokens": 8154952.0,
      "step": 4425
    },
    {
      "entropy": 5.991319561004639,
      "epoch": 0.37219071623608485,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004990633561858308,
      "loss": 5.7495,
      "mean_token_accuracy": 0.1485943764448166,
      "num_tokens": 8164365.0,
      "step": 4430
    },
    {
      "entropy": 5.964152002334595,
      "epoch": 0.37261079605124975,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990606245616537,
      "loss": 5.7532,
      "mean_token_accuracy": 0.1534093663096428,
      "num_tokens": 8172614.0,
      "step": 4435
    },
    {
      "entropy": 5.937890291213989,
      "epoch": 0.37303087586641465,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990578889683579,
      "loss": 5.8143,
      "mean_token_accuracy": 0.14677512943744658,
      "num_tokens": 8182445.0,
      "step": 4440
    },
    {
      "entropy": 5.963037729263306,
      "epoch": 0.3734509556815795,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004990551494059921,
      "loss": 5.6942,
      "mean_token_accuracy": 0.1528533488512039,
      "num_tokens": 8191871.0,
      "step": 4445
    },
    {
      "entropy": 5.989839935302735,
      "epoch": 0.3738710354967444,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004990524058746047,
      "loss": 5.8572,
      "mean_token_accuracy": 0.15349839329719545,
      "num_tokens": 8200658.0,
      "step": 4450
    },
    {
      "entropy": 5.928804111480713,
      "epoch": 0.3742911153119093,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990496583742443,
      "loss": 5.768,
      "mean_token_accuracy": 0.15008518993854522,
      "num_tokens": 8209776.0,
      "step": 4455
    },
    {
      "entropy": 5.890834903717041,
      "epoch": 0.3747111951270741,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004990469069049596,
      "loss": 5.703,
      "mean_token_accuracy": 0.15505637228488922,
      "num_tokens": 8219401.0,
      "step": 4460
    },
    {
      "entropy": 5.889174842834473,
      "epoch": 0.375131274942239,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004990441514667993,
      "loss": 5.7637,
      "mean_token_accuracy": 0.15405186712741853,
      "num_tokens": 8228762.0,
      "step": 4465
    },
    {
      "entropy": 5.968860626220703,
      "epoch": 0.3755513547574039,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990413920598121,
      "loss": 5.7636,
      "mean_token_accuracy": 0.15373183339834212,
      "num_tokens": 8236612.0,
      "step": 4470
    },
    {
      "entropy": 5.943127727508545,
      "epoch": 0.37597143457256876,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004990386286840471,
      "loss": 5.7608,
      "mean_token_accuracy": 0.14784819036722183,
      "num_tokens": 8245043.0,
      "step": 4475
    },
    {
      "entropy": 6.054802799224854,
      "epoch": 0.37639151438773366,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990358613395532,
      "loss": 5.8574,
      "mean_token_accuracy": 0.1462250158190727,
      "num_tokens": 8255270.0,
      "step": 4480
    },
    {
      "entropy": 5.998329496383667,
      "epoch": 0.37681159420289856,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004990330900263792,
      "loss": 5.8289,
      "mean_token_accuracy": 0.14600027352571487,
      "num_tokens": 8264761.0,
      "step": 4485
    },
    {
      "entropy": 5.965425777435303,
      "epoch": 0.37723167401806346,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004990303147445745,
      "loss": 5.7775,
      "mean_token_accuracy": 0.15009373426437378,
      "num_tokens": 8274308.0,
      "step": 4490
    },
    {
      "entropy": 5.8652503967285154,
      "epoch": 0.3776517538332283,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990275354941881,
      "loss": 5.6572,
      "mean_token_accuracy": 0.16007554829120635,
      "num_tokens": 8283323.0,
      "step": 4495
    },
    {
      "entropy": 5.926560640335083,
      "epoch": 0.3780718336483932,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004990247522752694,
      "loss": 5.9835,
      "mean_token_accuracy": 0.14134320095181466,
      "num_tokens": 8293452.0,
      "step": 4500
    },
    {
      "entropy": 5.975225973129272,
      "epoch": 0.3784919134635581,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004990219650878674,
      "loss": 5.6744,
      "mean_token_accuracy": 0.15657551288604737,
      "num_tokens": 8302941.0,
      "step": 4505
    },
    {
      "entropy": 5.864366960525513,
      "epoch": 0.37891199327872294,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004990191739320318,
      "loss": 5.6893,
      "mean_token_accuracy": 0.1629151314496994,
      "num_tokens": 8311811.0,
      "step": 4510
    },
    {
      "entropy": 5.829625272750855,
      "epoch": 0.37933207309388783,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990163788078117,
      "loss": 5.6083,
      "mean_token_accuracy": 0.16099839359521867,
      "num_tokens": 8321130.0,
      "step": 4515
    },
    {
      "entropy": 5.864146566390991,
      "epoch": 0.37975215290905273,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004990135797152569,
      "loss": 5.7098,
      "mean_token_accuracy": 0.15075407326221466,
      "num_tokens": 8330233.0,
      "step": 4520
    },
    {
      "entropy": 5.879518222808838,
      "epoch": 0.3801722327242176,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990107766544169,
      "loss": 5.6939,
      "mean_token_accuracy": 0.15840827226638793,
      "num_tokens": 8338585.0,
      "step": 4525
    },
    {
      "entropy": 5.919932222366333,
      "epoch": 0.38059231253938247,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004990079696253413,
      "loss": 5.7301,
      "mean_token_accuracy": 0.15791035294532776,
      "num_tokens": 8346618.0,
      "step": 4530
    },
    {
      "entropy": 5.939203500747681,
      "epoch": 0.38101239235454737,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990051586280799,
      "loss": 5.7227,
      "mean_token_accuracy": 0.15326298028230667,
      "num_tokens": 8356273.0,
      "step": 4535
    },
    {
      "entropy": 5.922527170181274,
      "epoch": 0.38143247216971227,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004990023436626824,
      "loss": 5.7148,
      "mean_token_accuracy": 0.15422383397817613,
      "num_tokens": 8366668.0,
      "step": 4540
    },
    {
      "entropy": 6.001941728591919,
      "epoch": 0.3818525519848771,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004989995247291988,
      "loss": 5.8211,
      "mean_token_accuracy": 0.15321157723665238,
      "num_tokens": 8375610.0,
      "step": 4545
    },
    {
      "entropy": 5.933878087997437,
      "epoch": 0.382272631800042,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004989967018276789,
      "loss": 5.7133,
      "mean_token_accuracy": 0.1544783115386963,
      "num_tokens": 8384455.0,
      "step": 4550
    },
    {
      "entropy": 5.826741123199463,
      "epoch": 0.3826927116152069,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004989938749581727,
      "loss": 5.7382,
      "mean_token_accuracy": 0.15217722058296204,
      "num_tokens": 8393868.0,
      "step": 4555
    },
    {
      "entropy": 5.93110761642456,
      "epoch": 0.38311279143037175,
      "grad_norm": 0.875,
      "learning_rate": 0.0004989910441207305,
      "loss": 5.7615,
      "mean_token_accuracy": 0.14792898297309875,
      "num_tokens": 8402916.0,
      "step": 4560
    },
    {
      "entropy": 5.909101343154907,
      "epoch": 0.38353287124553664,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004989882093154023,
      "loss": 5.6917,
      "mean_token_accuracy": 0.15420530289411544,
      "num_tokens": 8411649.0,
      "step": 4565
    },
    {
      "entropy": 5.909681701660157,
      "epoch": 0.38395295106070154,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004989853705422381,
      "loss": 5.7919,
      "mean_token_accuracy": 0.14452365189790725,
      "num_tokens": 8420393.0,
      "step": 4570
    },
    {
      "entropy": 5.895758485794067,
      "epoch": 0.38437303087586644,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004989825278012886,
      "loss": 5.6875,
      "mean_token_accuracy": 0.15371473133563995,
      "num_tokens": 8429404.0,
      "step": 4575
    },
    {
      "entropy": 5.917196273803711,
      "epoch": 0.3847931106910313,
      "grad_norm": 1.078125,
      "learning_rate": 0.000498979681092604,
      "loss": 5.7344,
      "mean_token_accuracy": 0.1492347776889801,
      "num_tokens": 8438299.0,
      "step": 4580
    },
    {
      "entropy": 5.835894346237183,
      "epoch": 0.3852131905061962,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004989768304162345,
      "loss": 5.6859,
      "mean_token_accuracy": 0.15611978322267533,
      "num_tokens": 8447392.0,
      "step": 4585
    },
    {
      "entropy": 5.929430866241455,
      "epoch": 0.3856332703213611,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004989739757722308,
      "loss": 5.7782,
      "mean_token_accuracy": 0.1487022116780281,
      "num_tokens": 8456361.0,
      "step": 4590
    },
    {
      "entropy": 5.930190563201904,
      "epoch": 0.3860533501365259,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004989711171606436,
      "loss": 5.7047,
      "mean_token_accuracy": 0.1564156025648117,
      "num_tokens": 8465548.0,
      "step": 4595
    },
    {
      "entropy": 5.932565641403198,
      "epoch": 0.3864734299516908,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989682545815232,
      "loss": 5.7218,
      "mean_token_accuracy": 0.15004338324069977,
      "num_tokens": 8474454.0,
      "step": 4600
    },
    {
      "entropy": 5.86793704032898,
      "epoch": 0.3868935097668557,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004989653880349207,
      "loss": 5.6391,
      "mean_token_accuracy": 0.15498715043067932,
      "num_tokens": 8482694.0,
      "step": 4605
    },
    {
      "entropy": 5.911267232894898,
      "epoch": 0.38731358958202056,
      "grad_norm": 1.0,
      "learning_rate": 0.0004989625175208864,
      "loss": 5.7515,
      "mean_token_accuracy": 0.14617059528827667,
      "num_tokens": 8491162.0,
      "step": 4610
    },
    {
      "entropy": 5.838245964050293,
      "epoch": 0.38773366939718545,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004989596430394717,
      "loss": 5.6052,
      "mean_token_accuracy": 0.16923634558916092,
      "num_tokens": 8500716.0,
      "step": 4615
    },
    {
      "entropy": 5.836051511764526,
      "epoch": 0.38815374921235035,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498956764590727,
      "loss": 5.6697,
      "mean_token_accuracy": 0.15317125618457794,
      "num_tokens": 8508871.0,
      "step": 4620
    },
    {
      "entropy": 5.983802938461304,
      "epoch": 0.38857382902751525,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004989538821747037,
      "loss": 5.8615,
      "mean_token_accuracy": 0.1454290270805359,
      "num_tokens": 8518450.0,
      "step": 4625
    },
    {
      "entropy": 5.942924976348877,
      "epoch": 0.3889939088426801,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004989509957914527,
      "loss": 5.7594,
      "mean_token_accuracy": 0.14524921625852585,
      "num_tokens": 8528238.0,
      "step": 4630
    },
    {
      "entropy": 5.879286861419677,
      "epoch": 0.389413988657845,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004989481054410251,
      "loss": 5.6593,
      "mean_token_accuracy": 0.1511693000793457,
      "num_tokens": 8537587.0,
      "step": 4635
    },
    {
      "entropy": 5.879880046844482,
      "epoch": 0.3898340684730099,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004989452111234721,
      "loss": 5.7671,
      "mean_token_accuracy": 0.1519084557890892,
      "num_tokens": 8547703.0,
      "step": 4640
    },
    {
      "entropy": 5.887255001068115,
      "epoch": 0.39025414828817473,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000498942312838845,
      "loss": 5.719,
      "mean_token_accuracy": 0.15363839119672776,
      "num_tokens": 8557001.0,
      "step": 4645
    },
    {
      "entropy": 5.827616882324219,
      "epoch": 0.3906742281033396,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989394105871952,
      "loss": 5.5931,
      "mean_token_accuracy": 0.1651986852288246,
      "num_tokens": 8565638.0,
      "step": 4650
    },
    {
      "entropy": 5.949290990829468,
      "epoch": 0.3910943079185045,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000498936504368574,
      "loss": 5.7626,
      "mean_token_accuracy": 0.15302542299032212,
      "num_tokens": 8574428.0,
      "step": 4655
    },
    {
      "entropy": 5.850303411483765,
      "epoch": 0.3915143877336694,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004989335941830329,
      "loss": 5.7144,
      "mean_token_accuracy": 0.15317893475294114,
      "num_tokens": 8583157.0,
      "step": 4660
    },
    {
      "entropy": 5.870885944366455,
      "epoch": 0.39193446754883426,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004989306800306236,
      "loss": 5.6849,
      "mean_token_accuracy": 0.14882708638906478,
      "num_tokens": 8592382.0,
      "step": 4665
    },
    {
      "entropy": 5.876427268981933,
      "epoch": 0.39235454736399916,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004989277619113975,
      "loss": 5.6754,
      "mean_token_accuracy": 0.15828215777873994,
      "num_tokens": 8601058.0,
      "step": 4670
    },
    {
      "entropy": 5.938532400131225,
      "epoch": 0.39277462717916406,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989248398254065,
      "loss": 5.7703,
      "mean_token_accuracy": 0.14732256829738616,
      "num_tokens": 8609479.0,
      "step": 4675
    },
    {
      "entropy": 5.900654602050781,
      "epoch": 0.3931947069943289,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004989219137727021,
      "loss": 5.7294,
      "mean_token_accuracy": 0.1546520099043846,
      "num_tokens": 8618860.0,
      "step": 4680
    },
    {
      "entropy": 5.897889614105225,
      "epoch": 0.3936147868094938,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989189837533365,
      "loss": 5.6571,
      "mean_token_accuracy": 0.15630604475736617,
      "num_tokens": 8627462.0,
      "step": 4685
    },
    {
      "entropy": 5.992354822158814,
      "epoch": 0.3940348666246587,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004989160497673613,
      "loss": 5.8569,
      "mean_token_accuracy": 0.146843983232975,
      "num_tokens": 8637569.0,
      "step": 4690
    },
    {
      "entropy": 5.945879602432251,
      "epoch": 0.39445494643982354,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004989131118148286,
      "loss": 5.6573,
      "mean_token_accuracy": 0.14983191341161728,
      "num_tokens": 8645440.0,
      "step": 4695
    },
    {
      "entropy": 5.849710083007812,
      "epoch": 0.39487502625498844,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989101698957904,
      "loss": 5.7926,
      "mean_token_accuracy": 0.1551379680633545,
      "num_tokens": 8655077.0,
      "step": 4700
    },
    {
      "entropy": 5.917178297042847,
      "epoch": 0.39529510607015333,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989072240102988,
      "loss": 5.7388,
      "mean_token_accuracy": 0.15438668057322502,
      "num_tokens": 8663126.0,
      "step": 4705
    },
    {
      "entropy": 5.974384880065918,
      "epoch": 0.39571518588531823,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004989042741584061,
      "loss": 5.7074,
      "mean_token_accuracy": 0.15212598592042922,
      "num_tokens": 8672386.0,
      "step": 4710
    },
    {
      "entropy": 5.768937015533448,
      "epoch": 0.3961352657004831,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989013203401645,
      "loss": 5.6548,
      "mean_token_accuracy": 0.1561252072453499,
      "num_tokens": 8681930.0,
      "step": 4715
    },
    {
      "entropy": 5.873412132263184,
      "epoch": 0.396555345515648,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004988983625556264,
      "loss": 5.7083,
      "mean_token_accuracy": 0.15193997621536254,
      "num_tokens": 8690993.0,
      "step": 4720
    },
    {
      "entropy": 5.8479430198669435,
      "epoch": 0.39697542533081287,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004988954008048438,
      "loss": 5.7116,
      "mean_token_accuracy": 0.15675423592329024,
      "num_tokens": 8699497.0,
      "step": 4725
    },
    {
      "entropy": 6.041143608093262,
      "epoch": 0.3973955051459777,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004988924350878697,
      "loss": 5.8809,
      "mean_token_accuracy": 0.14092395827174187,
      "num_tokens": 8709274.0,
      "step": 4730
    },
    {
      "entropy": 5.993929195404053,
      "epoch": 0.3978155849611426,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004988894654047563,
      "loss": 5.7498,
      "mean_token_accuracy": 0.1469295635819435,
      "num_tokens": 8718158.0,
      "step": 4735
    },
    {
      "entropy": 5.827867841720581,
      "epoch": 0.3982356647763075,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004988864917555562,
      "loss": 5.6291,
      "mean_token_accuracy": 0.1548566371202469,
      "num_tokens": 8727459.0,
      "step": 4740
    },
    {
      "entropy": 5.879483222961426,
      "epoch": 0.3986557445914724,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004988835141403224,
      "loss": 5.7594,
      "mean_token_accuracy": 0.15677372515201568,
      "num_tokens": 8737614.0,
      "step": 4745
    },
    {
      "entropy": 5.8809814453125,
      "epoch": 0.39907582440663725,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004988805325591073,
      "loss": 5.5943,
      "mean_token_accuracy": 0.15506676733493804,
      "num_tokens": 8746799.0,
      "step": 4750
    },
    {
      "entropy": 5.892252826690674,
      "epoch": 0.39949590422180214,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004988775470119639,
      "loss": 5.7641,
      "mean_token_accuracy": 0.1494814820587635,
      "num_tokens": 8756555.0,
      "step": 4755
    },
    {
      "entropy": 5.835883665084839,
      "epoch": 0.39991598403696704,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004988745574989451,
      "loss": 5.7763,
      "mean_token_accuracy": 0.1567599043250084,
      "num_tokens": 8765849.0,
      "step": 4760
    },
    {
      "entropy": 6.027160596847534,
      "epoch": 0.4003360638521319,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004988715640201036,
      "loss": 5.8543,
      "mean_token_accuracy": 0.14386038482189178,
      "num_tokens": 8775713.0,
      "step": 4765
    },
    {
      "entropy": 5.901659107208252,
      "epoch": 0.4007561436672968,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004988685665754928,
      "loss": 5.6779,
      "mean_token_accuracy": 0.15384853333234788,
      "num_tokens": 8784717.0,
      "step": 4770
    },
    {
      "entropy": 5.849430847167969,
      "epoch": 0.4011762234824617,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004988655651651656,
      "loss": 5.6953,
      "mean_token_accuracy": 0.15288221091032028,
      "num_tokens": 8794388.0,
      "step": 4775
    },
    {
      "entropy": 5.8477579116821286,
      "epoch": 0.4015963032976265,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004988625597891751,
      "loss": 5.7118,
      "mean_token_accuracy": 0.15216539055109024,
      "num_tokens": 8802436.0,
      "step": 4780
    },
    {
      "entropy": 5.939417314529419,
      "epoch": 0.4020163831127914,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988595504475746,
      "loss": 5.6698,
      "mean_token_accuracy": 0.15575191825628282,
      "num_tokens": 8811184.0,
      "step": 4785
    },
    {
      "entropy": 5.918405437469483,
      "epoch": 0.4024364629279563,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988565371404175,
      "loss": 5.7523,
      "mean_token_accuracy": 0.1554136499762535,
      "num_tokens": 8820525.0,
      "step": 4790
    },
    {
      "entropy": 5.843054962158203,
      "epoch": 0.4028565427431212,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004988535198677571,
      "loss": 5.6217,
      "mean_token_accuracy": 0.15804577022790908,
      "num_tokens": 8828928.0,
      "step": 4795
    },
    {
      "entropy": 5.970369100570679,
      "epoch": 0.40327662255828606,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988504986296469,
      "loss": 5.8132,
      "mean_token_accuracy": 0.14200191348791122,
      "num_tokens": 8838615.0,
      "step": 4800
    },
    {
      "entropy": 5.933490657806397,
      "epoch": 0.40369670237345096,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004988474734261404,
      "loss": 5.7889,
      "mean_token_accuracy": 0.14515041634440423,
      "num_tokens": 8848709.0,
      "step": 4805
    },
    {
      "entropy": 5.920221948623658,
      "epoch": 0.40411678218861585,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004988444442572911,
      "loss": 5.7527,
      "mean_token_accuracy": 0.14759574979543685,
      "num_tokens": 8858277.0,
      "step": 4810
    },
    {
      "entropy": 5.874279880523682,
      "epoch": 0.4045368620037807,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004988414111231528,
      "loss": 5.707,
      "mean_token_accuracy": 0.15657222270965576,
      "num_tokens": 8868436.0,
      "step": 4815
    },
    {
      "entropy": 5.909940195083618,
      "epoch": 0.4049569418189456,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000498838374023779,
      "loss": 5.7047,
      "mean_token_accuracy": 0.14395527988672258,
      "num_tokens": 8877740.0,
      "step": 4820
    },
    {
      "entropy": 5.964693403244018,
      "epoch": 0.4053770216341105,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004988353329592239,
      "loss": 5.6772,
      "mean_token_accuracy": 0.1548558935523033,
      "num_tokens": 8887408.0,
      "step": 4825
    },
    {
      "entropy": 5.922537040710449,
      "epoch": 0.4057971014492754,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004988322879295409,
      "loss": 5.8427,
      "mean_token_accuracy": 0.14574310183525085,
      "num_tokens": 8897141.0,
      "step": 4830
    },
    {
      "entropy": 5.805268859863281,
      "epoch": 0.40621718126444023,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004988292389347844,
      "loss": 5.6104,
      "mean_token_accuracy": 0.16397565603256226,
      "num_tokens": 8905747.0,
      "step": 4835
    },
    {
      "entropy": 5.965035152435303,
      "epoch": 0.40663726107960513,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000498826185975008,
      "loss": 5.7752,
      "mean_token_accuracy": 0.15198375433683395,
      "num_tokens": 8914926.0,
      "step": 4840
    },
    {
      "entropy": 5.8813001155853275,
      "epoch": 0.40705734089477,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004988231290502662,
      "loss": 5.7777,
      "mean_token_accuracy": 0.15258675962686538,
      "num_tokens": 8923956.0,
      "step": 4845
    },
    {
      "entropy": 5.9237809658050535,
      "epoch": 0.40747742070993487,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004988200681606127,
      "loss": 5.6491,
      "mean_token_accuracy": 0.14926957190036774,
      "num_tokens": 8932654.0,
      "step": 4850
    },
    {
      "entropy": 5.894003057479859,
      "epoch": 0.40789750052509977,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000498817003306102,
      "loss": 5.6451,
      "mean_token_accuracy": 0.1597268909215927,
      "num_tokens": 8941716.0,
      "step": 4855
    },
    {
      "entropy": 5.817833948135376,
      "epoch": 0.40831758034026466,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004988139344867884,
      "loss": 5.724,
      "mean_token_accuracy": 0.1522026114165783,
      "num_tokens": 8950377.0,
      "step": 4860
    },
    {
      "entropy": 5.875721025466919,
      "epoch": 0.4087376601554295,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004988108617027261,
      "loss": 5.6699,
      "mean_token_accuracy": 0.15545763820409775,
      "num_tokens": 8959857.0,
      "step": 4865
    },
    {
      "entropy": 5.833681583404541,
      "epoch": 0.4091577399705944,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004988077849539698,
      "loss": 5.6224,
      "mean_token_accuracy": 0.15625977516174316,
      "num_tokens": 8968272.0,
      "step": 4870
    },
    {
      "entropy": 5.867157077789306,
      "epoch": 0.4095778197857593,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988047042405736,
      "loss": 5.6935,
      "mean_token_accuracy": 0.16176266372203826,
      "num_tokens": 8977445.0,
      "step": 4875
    },
    {
      "entropy": 5.968780374526977,
      "epoch": 0.4099978996009242,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004988016195625924,
      "loss": 5.76,
      "mean_token_accuracy": 0.14859384894371033,
      "num_tokens": 8987315.0,
      "step": 4880
    },
    {
      "entropy": 5.913091421127319,
      "epoch": 0.41041797941608904,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004987985309200807,
      "loss": 5.7589,
      "mean_token_accuracy": 0.1500951334834099,
      "num_tokens": 8998119.0,
      "step": 4885
    },
    {
      "entropy": 5.845153284072876,
      "epoch": 0.41083805923125394,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987954383130934,
      "loss": 5.6392,
      "mean_token_accuracy": 0.1648591250181198,
      "num_tokens": 9007167.0,
      "step": 4890
    },
    {
      "entropy": 5.868191337585449,
      "epoch": 0.41125813904641884,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000498792341741685,
      "loss": 5.6988,
      "mean_token_accuracy": 0.1461891144514084,
      "num_tokens": 9016690.0,
      "step": 4895
    },
    {
      "entropy": 5.933396625518799,
      "epoch": 0.4116782188615837,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004987892412059106,
      "loss": 5.7776,
      "mean_token_accuracy": 0.1512982115149498,
      "num_tokens": 9026117.0,
      "step": 4900
    },
    {
      "entropy": 5.841189861297607,
      "epoch": 0.4120982986767486,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004987861367058251,
      "loss": 5.6685,
      "mean_token_accuracy": 0.1527989387512207,
      "num_tokens": 9035754.0,
      "step": 4905
    },
    {
      "entropy": 5.89400463104248,
      "epoch": 0.4125183784919135,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004987830282414833,
      "loss": 5.6681,
      "mean_token_accuracy": 0.15465910285711287,
      "num_tokens": 9045453.0,
      "step": 4910
    },
    {
      "entropy": 5.913017749786377,
      "epoch": 0.41293845830707837,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004987799158129404,
      "loss": 5.7865,
      "mean_token_accuracy": 0.15383710861206054,
      "num_tokens": 9056045.0,
      "step": 4915
    },
    {
      "entropy": 5.873713541030884,
      "epoch": 0.4133585381222432,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987767994202516,
      "loss": 5.6591,
      "mean_token_accuracy": 0.149123851954937,
      "num_tokens": 9065728.0,
      "step": 4920
    },
    {
      "entropy": 5.842416429519654,
      "epoch": 0.4137786179374081,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004987736790634719,
      "loss": 5.6964,
      "mean_token_accuracy": 0.1515074759721756,
      "num_tokens": 9075522.0,
      "step": 4925
    },
    {
      "entropy": 5.839908504486084,
      "epoch": 0.414198697752573,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004987705547426568,
      "loss": 5.6657,
      "mean_token_accuracy": 0.14872333407402039,
      "num_tokens": 9084412.0,
      "step": 4930
    },
    {
      "entropy": 5.913540363311768,
      "epoch": 0.41461877756773785,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004987674264578615,
      "loss": 5.7332,
      "mean_token_accuracy": 0.14811609983444213,
      "num_tokens": 9094289.0,
      "step": 4935
    },
    {
      "entropy": 5.8725332736969,
      "epoch": 0.41503885738290275,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004987642942091414,
      "loss": 5.6496,
      "mean_token_accuracy": 0.15519289672374725,
      "num_tokens": 9103124.0,
      "step": 4940
    },
    {
      "entropy": 5.8630242347717285,
      "epoch": 0.41545893719806765,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004987611579965523,
      "loss": 5.5939,
      "mean_token_accuracy": 0.1533403366804123,
      "num_tokens": 9112794.0,
      "step": 4945
    },
    {
      "entropy": 5.913699436187744,
      "epoch": 0.4158790170132325,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004987580178201492,
      "loss": 5.7582,
      "mean_token_accuracy": 0.1595202386379242,
      "num_tokens": 9122718.0,
      "step": 4950
    },
    {
      "entropy": 5.891729164123535,
      "epoch": 0.4162990968283974,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004987548736799882,
      "loss": 5.7729,
      "mean_token_accuracy": 0.1482948213815689,
      "num_tokens": 9131855.0,
      "step": 4955
    },
    {
      "entropy": 5.892034864425659,
      "epoch": 0.4167191766435623,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004987517255761248,
      "loss": 5.6301,
      "mean_token_accuracy": 0.15863611847162246,
      "num_tokens": 9141102.0,
      "step": 4960
    },
    {
      "entropy": 5.8408544063568115,
      "epoch": 0.4171392564587272,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987485735086148,
      "loss": 5.7023,
      "mean_token_accuracy": 0.15845317095518113,
      "num_tokens": 9150552.0,
      "step": 4965
    },
    {
      "entropy": 5.915563631057739,
      "epoch": 0.417559336273892,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000498745417477514,
      "loss": 5.6944,
      "mean_token_accuracy": 0.15274149924516678,
      "num_tokens": 9160105.0,
      "step": 4970
    },
    {
      "entropy": 5.851050758361817,
      "epoch": 0.4179794160890569,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004987422574828784,
      "loss": 5.6882,
      "mean_token_accuracy": 0.15457934886217117,
      "num_tokens": 9169367.0,
      "step": 4975
    },
    {
      "entropy": 5.845413064956665,
      "epoch": 0.4183994959042218,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004987390935247639,
      "loss": 5.5584,
      "mean_token_accuracy": 0.16090269088745118,
      "num_tokens": 9177872.0,
      "step": 4980
    },
    {
      "entropy": 5.878557348251343,
      "epoch": 0.41881957571938666,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004987359256032265,
      "loss": 5.7719,
      "mean_token_accuracy": 0.1481206126511097,
      "num_tokens": 9187879.0,
      "step": 4985
    },
    {
      "entropy": 5.854932594299316,
      "epoch": 0.41923965553455156,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004987327537183225,
      "loss": 5.6934,
      "mean_token_accuracy": 0.15052096396684647,
      "num_tokens": 9198281.0,
      "step": 4990
    },
    {
      "entropy": 5.8744182109832765,
      "epoch": 0.41965973534971646,
      "grad_norm": 0.875,
      "learning_rate": 0.0004987295778701078,
      "loss": 5.6809,
      "mean_token_accuracy": 0.1535363718867302,
      "num_tokens": 9207670.0,
      "step": 4995
    },
    {
      "entropy": 5.941969966888427,
      "epoch": 0.42007981516488135,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000498726398058639,
      "loss": 5.6845,
      "mean_token_accuracy": 0.15731745064258576,
      "num_tokens": 9216995.0,
      "step": 5000
    },
    {
      "entropy": 5.869586181640625,
      "epoch": 0.4204998949800462,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004987232142839723,
      "loss": 5.7815,
      "mean_token_accuracy": 0.14722964018583298,
      "num_tokens": 9227330.0,
      "step": 5005
    },
    {
      "entropy": 5.857760953903198,
      "epoch": 0.4209199747952111,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004987200265461638,
      "loss": 5.6921,
      "mean_token_accuracy": 0.160021048784256,
      "num_tokens": 9236666.0,
      "step": 5010
    },
    {
      "entropy": 5.903905725479126,
      "epoch": 0.421340054610376,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004987168348452705,
      "loss": 5.6911,
      "mean_token_accuracy": 0.15569491684436798,
      "num_tokens": 9246388.0,
      "step": 5015
    },
    {
      "entropy": 5.847051525115967,
      "epoch": 0.42176013442554083,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004987136391813485,
      "loss": 5.6417,
      "mean_token_accuracy": 0.16386419981718064,
      "num_tokens": 9255239.0,
      "step": 5020
    },
    {
      "entropy": 5.802522277832031,
      "epoch": 0.42218021424070573,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004987104395544547,
      "loss": 5.6203,
      "mean_token_accuracy": 0.15661859810352324,
      "num_tokens": 9264468.0,
      "step": 5025
    },
    {
      "entropy": 5.871787643432617,
      "epoch": 0.42260029405587063,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987072359646455,
      "loss": 5.6899,
      "mean_token_accuracy": 0.15836804807186128,
      "num_tokens": 9274140.0,
      "step": 5030
    },
    {
      "entropy": 5.8818848609924315,
      "epoch": 0.42302037387103547,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004987040284119778,
      "loss": 5.6734,
      "mean_token_accuracy": 0.15363640189170838,
      "num_tokens": 9283539.0,
      "step": 5035
    },
    {
      "entropy": 5.812870168685913,
      "epoch": 0.42344045368620037,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004987008168965087,
      "loss": 5.6668,
      "mean_token_accuracy": 0.15175789594650269,
      "num_tokens": 9292664.0,
      "step": 5040
    },
    {
      "entropy": 5.904106283187867,
      "epoch": 0.42386053350136527,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004986976014182946,
      "loss": 5.7649,
      "mean_token_accuracy": 0.15362361818552017,
      "num_tokens": 9302814.0,
      "step": 5045
    },
    {
      "entropy": 5.973753833770752,
      "epoch": 0.42428061331653016,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004986943819773927,
      "loss": 5.7642,
      "mean_token_accuracy": 0.15095529705286026,
      "num_tokens": 9312654.0,
      "step": 5050
    },
    {
      "entropy": 5.961247158050537,
      "epoch": 0.424700693131695,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00049869115857386,
      "loss": 5.7709,
      "mean_token_accuracy": 0.15128740966320037,
      "num_tokens": 9322271.0,
      "step": 5055
    },
    {
      "entropy": 5.890190887451172,
      "epoch": 0.4251207729468599,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004986879312077536,
      "loss": 5.7245,
      "mean_token_accuracy": 0.15066271498799325,
      "num_tokens": 9331341.0,
      "step": 5060
    },
    {
      "entropy": 5.8575903415679935,
      "epoch": 0.4255408527620248,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986846998791308,
      "loss": 5.6574,
      "mean_token_accuracy": 0.15662601292133332,
      "num_tokens": 9339863.0,
      "step": 5065
    },
    {
      "entropy": 5.815864372253418,
      "epoch": 0.42596093257718964,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004986814645880485,
      "loss": 5.6238,
      "mean_token_accuracy": 0.15884864330291748,
      "num_tokens": 9349488.0,
      "step": 5070
    },
    {
      "entropy": 5.82549033164978,
      "epoch": 0.42638101239235454,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004986782253345645,
      "loss": 5.6362,
      "mean_token_accuracy": 0.15622432827949523,
      "num_tokens": 9357977.0,
      "step": 5075
    },
    {
      "entropy": 5.88819260597229,
      "epoch": 0.42680109220751944,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004986749821187358,
      "loss": 5.7393,
      "mean_token_accuracy": 0.15062530636787413,
      "num_tokens": 9367449.0,
      "step": 5080
    },
    {
      "entropy": 5.933618497848511,
      "epoch": 0.42722117202268434,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00049867173494062,
      "loss": 5.7602,
      "mean_token_accuracy": 0.15464712232351302,
      "num_tokens": 9377070.0,
      "step": 5085
    },
    {
      "entropy": 5.820448541641236,
      "epoch": 0.4276412518378492,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004986684838002744,
      "loss": 5.5628,
      "mean_token_accuracy": 0.1470932886004448,
      "num_tokens": 9385881.0,
      "step": 5090
    },
    {
      "entropy": 5.8892427444458,
      "epoch": 0.4280613316530141,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004986652286977569,
      "loss": 5.6922,
      "mean_token_accuracy": 0.14868162721395492,
      "num_tokens": 9395159.0,
      "step": 5095
    },
    {
      "entropy": 5.860781717300415,
      "epoch": 0.428481411468179,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004986619696331252,
      "loss": 5.6359,
      "mean_token_accuracy": 0.1528971217572689,
      "num_tokens": 9404590.0,
      "step": 5100
    },
    {
      "entropy": 5.8955230712890625,
      "epoch": 0.4289014912833438,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004986587066064367,
      "loss": 5.6716,
      "mean_token_accuracy": 0.15910475552082062,
      "num_tokens": 9414452.0,
      "step": 5105
    },
    {
      "entropy": 5.946217107772827,
      "epoch": 0.4293215710985087,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004986554396177494,
      "loss": 5.7956,
      "mean_token_accuracy": 0.14509906619787216,
      "num_tokens": 9424004.0,
      "step": 5110
    },
    {
      "entropy": 5.867961645126343,
      "epoch": 0.4297416509136736,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004986521686671212,
      "loss": 5.6669,
      "mean_token_accuracy": 0.16597846299409866,
      "num_tokens": 9433487.0,
      "step": 5115
    },
    {
      "entropy": 5.847635078430176,
      "epoch": 0.43016173072883845,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00049864889375461,
      "loss": 5.7274,
      "mean_token_accuracy": 0.15033599734306335,
      "num_tokens": 9442742.0,
      "step": 5120
    },
    {
      "entropy": 5.905732250213623,
      "epoch": 0.43058181054400335,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004986456148802738,
      "loss": 5.7928,
      "mean_token_accuracy": 0.15171840339899062,
      "num_tokens": 9452550.0,
      "step": 5125
    },
    {
      "entropy": 5.9744799613952635,
      "epoch": 0.43100189035916825,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986423320441707,
      "loss": 5.7355,
      "mean_token_accuracy": 0.14951937571167945,
      "num_tokens": 9461920.0,
      "step": 5130
    },
    {
      "entropy": 5.8919281482696535,
      "epoch": 0.43142197017433315,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004986390452463588,
      "loss": 5.6598,
      "mean_token_accuracy": 0.14998603016138076,
      "num_tokens": 9470817.0,
      "step": 5135
    },
    {
      "entropy": 5.724545240402222,
      "epoch": 0.431842049989498,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004986357544868964,
      "loss": 5.6199,
      "mean_token_accuracy": 0.15861552208662033,
      "num_tokens": 9479936.0,
      "step": 5140
    },
    {
      "entropy": 5.891957855224609,
      "epoch": 0.4322621298046629,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004986324597658418,
      "loss": 5.647,
      "mean_token_accuracy": 0.15907271206378937,
      "num_tokens": 9489818.0,
      "step": 5145
    },
    {
      "entropy": 5.789335918426514,
      "epoch": 0.4326822096198278,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004986291610832533,
      "loss": 5.65,
      "mean_token_accuracy": 0.15387662947177888,
      "num_tokens": 9499688.0,
      "step": 5150
    },
    {
      "entropy": 5.953478908538818,
      "epoch": 0.4331022894349926,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004986258584391892,
      "loss": 5.7152,
      "mean_token_accuracy": 0.15160334408283233,
      "num_tokens": 9509581.0,
      "step": 5155
    },
    {
      "entropy": 5.943544149398804,
      "epoch": 0.4335223692501575,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004986225518337084,
      "loss": 5.7801,
      "mean_token_accuracy": 0.15453668758273126,
      "num_tokens": 9518556.0,
      "step": 5160
    },
    {
      "entropy": 5.779713249206543,
      "epoch": 0.4339424490653224,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004986192412668692,
      "loss": 5.6852,
      "mean_token_accuracy": 0.15138726085424423,
      "num_tokens": 9527612.0,
      "step": 5165
    },
    {
      "entropy": 5.834102582931519,
      "epoch": 0.4343625288804873,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004986159267387302,
      "loss": 5.5783,
      "mean_token_accuracy": 0.16340289860963822,
      "num_tokens": 9535882.0,
      "step": 5170
    },
    {
      "entropy": 5.829035234451294,
      "epoch": 0.43478260869565216,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004986126082493502,
      "loss": 5.6932,
      "mean_token_accuracy": 0.15901082158088684,
      "num_tokens": 9544799.0,
      "step": 5175
    },
    {
      "entropy": 5.856446266174316,
      "epoch": 0.43520268851081706,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004986092857987881,
      "loss": 5.5866,
      "mean_token_accuracy": 0.15803252607584,
      "num_tokens": 9553805.0,
      "step": 5180
    },
    {
      "entropy": 5.8282328128814695,
      "epoch": 0.43562276832598196,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004986059593871026,
      "loss": 5.6406,
      "mean_token_accuracy": 0.15629224628210067,
      "num_tokens": 9563493.0,
      "step": 5185
    },
    {
      "entropy": 5.88116488456726,
      "epoch": 0.4360428481411468,
      "grad_norm": 0.875,
      "learning_rate": 0.0004986026290143527,
      "loss": 5.7235,
      "mean_token_accuracy": 0.15083191096782683,
      "num_tokens": 9572297.0,
      "step": 5190
    },
    {
      "entropy": 6.01215090751648,
      "epoch": 0.4364629279563117,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004985992946805973,
      "loss": 5.8338,
      "mean_token_accuracy": 0.14679813981056214,
      "num_tokens": 9581967.0,
      "step": 5195
    },
    {
      "entropy": 5.860457944869995,
      "epoch": 0.4368830077714766,
      "grad_norm": 0.875,
      "learning_rate": 0.0004985959563858955,
      "loss": 5.7621,
      "mean_token_accuracy": 0.15618911683559417,
      "num_tokens": 9590885.0,
      "step": 5200
    },
    {
      "entropy": 5.920520925521851,
      "epoch": 0.43730308758664144,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004985926141303066,
      "loss": 5.6794,
      "mean_token_accuracy": 0.15353824943304062,
      "num_tokens": 9599247.0,
      "step": 5205
    },
    {
      "entropy": 5.904012393951416,
      "epoch": 0.43772316740180633,
      "grad_norm": 1.0,
      "learning_rate": 0.0004985892679138896,
      "loss": 5.6055,
      "mean_token_accuracy": 0.1609176740050316,
      "num_tokens": 9608296.0,
      "step": 5210
    },
    {
      "entropy": 5.878478050231934,
      "epoch": 0.43814324721697123,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004985859177367038,
      "loss": 5.6557,
      "mean_token_accuracy": 0.15502529889345168,
      "num_tokens": 9616734.0,
      "step": 5215
    },
    {
      "entropy": 5.882960367202759,
      "epoch": 0.43856332703213613,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985825635988087,
      "loss": 5.7492,
      "mean_token_accuracy": 0.15253912508487702,
      "num_tokens": 9626246.0,
      "step": 5220
    },
    {
      "entropy": 5.820113277435302,
      "epoch": 0.43898340684730097,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004985792055002635,
      "loss": 5.6064,
      "mean_token_accuracy": 0.1569197416305542,
      "num_tokens": 9634963.0,
      "step": 5225
    },
    {
      "entropy": 5.904363775253296,
      "epoch": 0.43940348666246587,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004985758434411278,
      "loss": 5.693,
      "mean_token_accuracy": 0.15850198715925218,
      "num_tokens": 9643615.0,
      "step": 5230
    },
    {
      "entropy": 5.883202981948853,
      "epoch": 0.43982356647763077,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004985724774214613,
      "loss": 5.6582,
      "mean_token_accuracy": 0.15674181655049324,
      "num_tokens": 9653306.0,
      "step": 5235
    },
    {
      "entropy": 5.843471193313599,
      "epoch": 0.4402436462927956,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004985691074413233,
      "loss": 5.6905,
      "mean_token_accuracy": 0.15597492307424546,
      "num_tokens": 9662389.0,
      "step": 5240
    },
    {
      "entropy": 5.821241474151611,
      "epoch": 0.4406637261079605,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985657335007739,
      "loss": 5.6905,
      "mean_token_accuracy": 0.1512298621237278,
      "num_tokens": 9671183.0,
      "step": 5245
    },
    {
      "entropy": 5.886975622177124,
      "epoch": 0.4410838059231254,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004985623555998725,
      "loss": 5.6624,
      "mean_token_accuracy": 0.16423219591379165,
      "num_tokens": 9680544.0,
      "step": 5250
    },
    {
      "entropy": 5.857367658615113,
      "epoch": 0.4415038857382903,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004985589737386791,
      "loss": 5.6992,
      "mean_token_accuracy": 0.1553041160106659,
      "num_tokens": 9690137.0,
      "step": 5255
    },
    {
      "entropy": 5.809094667434692,
      "epoch": 0.44192396555345514,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985555879172535,
      "loss": 5.6499,
      "mean_token_accuracy": 0.15782568752765655,
      "num_tokens": 9699149.0,
      "step": 5260
    },
    {
      "entropy": 5.934548234939575,
      "epoch": 0.44234404536862004,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000498552198135656,
      "loss": 5.7042,
      "mean_token_accuracy": 0.16048139780759813,
      "num_tokens": 9709308.0,
      "step": 5265
    },
    {
      "entropy": 5.857599496841431,
      "epoch": 0.44276412518378494,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004985488043939462,
      "loss": 5.6462,
      "mean_token_accuracy": 0.15583646297454834,
      "num_tokens": 9718462.0,
      "step": 5270
    },
    {
      "entropy": 5.84532356262207,
      "epoch": 0.4431842049989498,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985454066921846,
      "loss": 5.5768,
      "mean_token_accuracy": 0.1588665559887886,
      "num_tokens": 9727626.0,
      "step": 5275
    },
    {
      "entropy": 5.721759366989136,
      "epoch": 0.4436042848141147,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985420050304312,
      "loss": 5.6106,
      "mean_token_accuracy": 0.1551545485854149,
      "num_tokens": 9737091.0,
      "step": 5280
    },
    {
      "entropy": 5.822771167755127,
      "epoch": 0.4440243646292796,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004985385994087462,
      "loss": 5.668,
      "mean_token_accuracy": 0.15709339901804925,
      "num_tokens": 9746135.0,
      "step": 5285
    },
    {
      "entropy": 5.899721670150757,
      "epoch": 0.4444444444444444,
      "grad_norm": 1.0,
      "learning_rate": 0.0004985351898271901,
      "loss": 5.6148,
      "mean_token_accuracy": 0.15749678909778594,
      "num_tokens": 9754549.0,
      "step": 5290
    },
    {
      "entropy": 5.891132783889771,
      "epoch": 0.4448645242596093,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004985317762858231,
      "loss": 5.7448,
      "mean_token_accuracy": 0.14913515299558638,
      "num_tokens": 9764219.0,
      "step": 5295
    },
    {
      "entropy": 5.863654899597168,
      "epoch": 0.4452846040747742,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498528358784706,
      "loss": 5.5821,
      "mean_token_accuracy": 0.1614499032497406,
      "num_tokens": 9772234.0,
      "step": 5300
    },
    {
      "entropy": 5.825886964797974,
      "epoch": 0.4457046838899391,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000498524937323899,
      "loss": 5.6503,
      "mean_token_accuracy": 0.16151614040136336,
      "num_tokens": 9781417.0,
      "step": 5305
    },
    {
      "entropy": 5.958418083190918,
      "epoch": 0.44612476370510395,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004985215119034628,
      "loss": 5.7848,
      "mean_token_accuracy": 0.14706624820828437,
      "num_tokens": 9791286.0,
      "step": 5310
    },
    {
      "entropy": 5.83896541595459,
      "epoch": 0.44654484352026885,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004985180825234582,
      "loss": 5.7521,
      "mean_token_accuracy": 0.15418604761362076,
      "num_tokens": 9802157.0,
      "step": 5315
    },
    {
      "entropy": 5.961184072494507,
      "epoch": 0.44696492333543375,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004985146491839459,
      "loss": 5.7422,
      "mean_token_accuracy": 0.1454138860106468,
      "num_tokens": 9812646.0,
      "step": 5320
    },
    {
      "entropy": 5.93065824508667,
      "epoch": 0.4473850031505986,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004985112118849865,
      "loss": 5.7505,
      "mean_token_accuracy": 0.1481009006500244,
      "num_tokens": 9822274.0,
      "step": 5325
    },
    {
      "entropy": 5.776948642730713,
      "epoch": 0.4478050829657635,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004985077706266412,
      "loss": 5.564,
      "mean_token_accuracy": 0.15706712752580643,
      "num_tokens": 9831337.0,
      "step": 5330
    },
    {
      "entropy": 5.817965841293335,
      "epoch": 0.4482251627809284,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004985043254089708,
      "loss": 5.6766,
      "mean_token_accuracy": 0.14964559227228164,
      "num_tokens": 9840798.0,
      "step": 5335
    },
    {
      "entropy": 5.79916353225708,
      "epoch": 0.44864524259609323,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985008762320364,
      "loss": 5.6555,
      "mean_token_accuracy": 0.15808326303958892,
      "num_tokens": 9850117.0,
      "step": 5340
    },
    {
      "entropy": 5.808034658432007,
      "epoch": 0.4490653224112581,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000498497423095899,
      "loss": 5.6089,
      "mean_token_accuracy": 0.1625373423099518,
      "num_tokens": 9858227.0,
      "step": 5345
    },
    {
      "entropy": 5.813161182403564,
      "epoch": 0.449485402226423,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004984939660006199,
      "loss": 5.7006,
      "mean_token_accuracy": 0.15393247902393342,
      "num_tokens": 9867157.0,
      "step": 5350
    },
    {
      "entropy": 5.847505283355713,
      "epoch": 0.4499054820415879,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004984905049462602,
      "loss": 5.6205,
      "mean_token_accuracy": 0.15501179546117783,
      "num_tokens": 9877045.0,
      "step": 5355
    },
    {
      "entropy": 5.961196565628052,
      "epoch": 0.45032556185675277,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004984870399328814,
      "loss": 5.7394,
      "mean_token_accuracy": 0.1546921819448471,
      "num_tokens": 9886637.0,
      "step": 5360
    },
    {
      "entropy": 5.801573133468628,
      "epoch": 0.45074564167191766,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004984835709605446,
      "loss": 5.617,
      "mean_token_accuracy": 0.16214425712823868,
      "num_tokens": 9895601.0,
      "step": 5365
    },
    {
      "entropy": 5.878483295440674,
      "epoch": 0.45116572148708256,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004984800980293116,
      "loss": 5.7673,
      "mean_token_accuracy": 0.15574561953544616,
      "num_tokens": 9904775.0,
      "step": 5370
    },
    {
      "entropy": 5.883770704269409,
      "epoch": 0.4515858013022474,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004984766211392435,
      "loss": 5.7029,
      "mean_token_accuracy": 0.1528709352016449,
      "num_tokens": 9913795.0,
      "step": 5375
    },
    {
      "entropy": 5.857630348205566,
      "epoch": 0.4520058811174123,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004984731402904024,
      "loss": 5.5597,
      "mean_token_accuracy": 0.15977992564439775,
      "num_tokens": 9922576.0,
      "step": 5380
    },
    {
      "entropy": 5.829555082321167,
      "epoch": 0.4524259609325772,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004984696554828496,
      "loss": 5.5229,
      "mean_token_accuracy": 0.1608197048306465,
      "num_tokens": 9930971.0,
      "step": 5385
    },
    {
      "entropy": 5.845896244049072,
      "epoch": 0.4528460407477421,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004984661667166468,
      "loss": 5.6432,
      "mean_token_accuracy": 0.15930829644203187,
      "num_tokens": 9939628.0,
      "step": 5390
    },
    {
      "entropy": 5.826462650299073,
      "epoch": 0.45326612056290694,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004984626739918561,
      "loss": 5.6198,
      "mean_token_accuracy": 0.16187043488025665,
      "num_tokens": 9948397.0,
      "step": 5395
    },
    {
      "entropy": 5.897841453552246,
      "epoch": 0.45368620037807184,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004984591773085391,
      "loss": 5.7046,
      "mean_token_accuracy": 0.1579720750451088,
      "num_tokens": 9957683.0,
      "step": 5400
    },
    {
      "entropy": 5.910743761062622,
      "epoch": 0.45410628019323673,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004984556766667578,
      "loss": 5.6996,
      "mean_token_accuracy": 0.15666956454515457,
      "num_tokens": 9966756.0,
      "step": 5405
    },
    {
      "entropy": 5.825371026992798,
      "epoch": 0.4545263600084016,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004984521720665743,
      "loss": 5.684,
      "mean_token_accuracy": 0.15819198042154312,
      "num_tokens": 9976000.0,
      "step": 5410
    },
    {
      "entropy": 5.866469955444336,
      "epoch": 0.4549464398235665,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004984486635080507,
      "loss": 5.6874,
      "mean_token_accuracy": 0.16008011251688004,
      "num_tokens": 9985509.0,
      "step": 5415
    },
    {
      "entropy": 5.797957992553711,
      "epoch": 0.45536651963873137,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004984451509912489,
      "loss": 5.6301,
      "mean_token_accuracy": 0.15894405841827391,
      "num_tokens": 9994342.0,
      "step": 5420
    },
    {
      "entropy": 5.786969327926636,
      "epoch": 0.4557865994538962,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004984416345162315,
      "loss": 5.6643,
      "mean_token_accuracy": 0.15164350271224974,
      "num_tokens": 10004249.0,
      "step": 5425
    },
    {
      "entropy": 5.855116987228394,
      "epoch": 0.4562066792690611,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004984381140830605,
      "loss": 5.6561,
      "mean_token_accuracy": 0.15875716507434845,
      "num_tokens": 10012430.0,
      "step": 5430
    },
    {
      "entropy": 5.869373321533203,
      "epoch": 0.456626759084226,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004984345896917984,
      "loss": 5.6467,
      "mean_token_accuracy": 0.15343497991561889,
      "num_tokens": 10021434.0,
      "step": 5435
    },
    {
      "entropy": 5.851734781265259,
      "epoch": 0.4570468388993909,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004984310613425076,
      "loss": 5.6384,
      "mean_token_accuracy": 0.159539994597435,
      "num_tokens": 10030473.0,
      "step": 5440
    },
    {
      "entropy": 5.888415145874023,
      "epoch": 0.45746691871455575,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004984275290352506,
      "loss": 5.6459,
      "mean_token_accuracy": 0.15975773483514785,
      "num_tokens": 10039057.0,
      "step": 5445
    },
    {
      "entropy": 5.906418371200561,
      "epoch": 0.45788699852972065,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004984239927700899,
      "loss": 5.7257,
      "mean_token_accuracy": 0.15392611026763917,
      "num_tokens": 10047998.0,
      "step": 5450
    },
    {
      "entropy": 5.940126276016235,
      "epoch": 0.45830707834488554,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004984204525470883,
      "loss": 5.655,
      "mean_token_accuracy": 0.14976596385240554,
      "num_tokens": 10057479.0,
      "step": 5455
    },
    {
      "entropy": 5.804240036010742,
      "epoch": 0.4587271581600504,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004984169083663084,
      "loss": 5.6319,
      "mean_token_accuracy": 0.15475388765335082,
      "num_tokens": 10067754.0,
      "step": 5460
    },
    {
      "entropy": 5.835357999801635,
      "epoch": 0.4591472379752153,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004984133602278129,
      "loss": 5.7106,
      "mean_token_accuracy": 0.1532885193824768,
      "num_tokens": 10076815.0,
      "step": 5465
    },
    {
      "entropy": 5.961859178543091,
      "epoch": 0.4595673177903802,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000498409808131665,
      "loss": 5.7057,
      "mean_token_accuracy": 0.15364886671304703,
      "num_tokens": 10086300.0,
      "step": 5470
    },
    {
      "entropy": 5.828978395462036,
      "epoch": 0.4599873976055451,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004984062520779272,
      "loss": 5.6238,
      "mean_token_accuracy": 0.15949156731367112,
      "num_tokens": 10095383.0,
      "step": 5475
    },
    {
      "entropy": 5.7631326675415036,
      "epoch": 0.4604074774207099,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004984026920666628,
      "loss": 5.5992,
      "mean_token_accuracy": 0.16033411771059036,
      "num_tokens": 10103971.0,
      "step": 5480
    },
    {
      "entropy": 5.845707273483276,
      "epoch": 0.4608275572358748,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004983991280979347,
      "loss": 5.5979,
      "mean_token_accuracy": 0.16333702802658082,
      "num_tokens": 10113028.0,
      "step": 5485
    },
    {
      "entropy": 5.881896495819092,
      "epoch": 0.4612476370510397,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004983955601718061,
      "loss": 5.5699,
      "mean_token_accuracy": 0.15939705371856688,
      "num_tokens": 10121890.0,
      "step": 5490
    },
    {
      "entropy": 5.889582633972168,
      "epoch": 0.46166771686620456,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004983919882883401,
      "loss": 5.6941,
      "mean_token_accuracy": 0.1580561488866806,
      "num_tokens": 10131655.0,
      "step": 5495
    },
    {
      "entropy": 5.924855709075928,
      "epoch": 0.46208779668136946,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004983884124476,
      "loss": 5.6927,
      "mean_token_accuracy": 0.15681440383195877,
      "num_tokens": 10140778.0,
      "step": 5500
    },
    {
      "entropy": 5.877159309387207,
      "epoch": 0.46250787649653435,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004983848326496494,
      "loss": 5.7491,
      "mean_token_accuracy": 0.15238432586193085,
      "num_tokens": 10150229.0,
      "step": 5505
    },
    {
      "entropy": 5.874773073196411,
      "epoch": 0.4629279563116992,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004983812488945513,
      "loss": 5.6384,
      "mean_token_accuracy": 0.155783973634243,
      "num_tokens": 10158939.0,
      "step": 5510
    },
    {
      "entropy": 5.835632228851319,
      "epoch": 0.4633480361268641,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004983776611823696,
      "loss": 5.6313,
      "mean_token_accuracy": 0.15373607724905014,
      "num_tokens": 10168383.0,
      "step": 5515
    },
    {
      "entropy": 5.781601142883301,
      "epoch": 0.463768115942029,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004983740695131676,
      "loss": 5.6409,
      "mean_token_accuracy": 0.16217319369316102,
      "num_tokens": 10178678.0,
      "step": 5520
    },
    {
      "entropy": 5.814817571640015,
      "epoch": 0.4641881957571939,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000498370473887009,
      "loss": 5.6264,
      "mean_token_accuracy": 0.15971693769097328,
      "num_tokens": 10188964.0,
      "step": 5525
    },
    {
      "entropy": 5.860387706756592,
      "epoch": 0.46460827557235873,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004983668743039573,
      "loss": 5.6544,
      "mean_token_accuracy": 0.16176948100328445,
      "num_tokens": 10198333.0,
      "step": 5530
    },
    {
      "entropy": 5.823429822921753,
      "epoch": 0.46502835538752363,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004983632707640766,
      "loss": 5.6641,
      "mean_token_accuracy": 0.15528512597084046,
      "num_tokens": 10207876.0,
      "step": 5535
    },
    {
      "entropy": 5.837881183624267,
      "epoch": 0.4654484352026885,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004983596632674306,
      "loss": 5.6144,
      "mean_token_accuracy": 0.16058907508850098,
      "num_tokens": 10216822.0,
      "step": 5540
    },
    {
      "entropy": 5.89295072555542,
      "epoch": 0.46586851501785337,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004983560518140831,
      "loss": 5.7334,
      "mean_token_accuracy": 0.1498328372836113,
      "num_tokens": 10226887.0,
      "step": 5545
    },
    {
      "entropy": 5.870006895065307,
      "epoch": 0.46628859483301827,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004983524364040982,
      "loss": 5.5705,
      "mean_token_accuracy": 0.16448622196912766,
      "num_tokens": 10235935.0,
      "step": 5550
    },
    {
      "entropy": 5.8303587436676025,
      "epoch": 0.46670867464818316,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004983488170375399,
      "loss": 5.5323,
      "mean_token_accuracy": 0.16216835826635362,
      "num_tokens": 10245590.0,
      "step": 5555
    },
    {
      "entropy": 5.7766870021820065,
      "epoch": 0.46712875446334806,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004983451937144723,
      "loss": 5.6226,
      "mean_token_accuracy": 0.15672850459814072,
      "num_tokens": 10255104.0,
      "step": 5560
    },
    {
      "entropy": 5.665926456451416,
      "epoch": 0.4675488342785129,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004983415664349595,
      "loss": 5.4807,
      "mean_token_accuracy": 0.17375965118408204,
      "num_tokens": 10264236.0,
      "step": 5565
    },
    {
      "entropy": 5.877076864242554,
      "epoch": 0.4679689140936778,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004983379351990659,
      "loss": 5.599,
      "mean_token_accuracy": 0.1592295229434967,
      "num_tokens": 10273335.0,
      "step": 5570
    },
    {
      "entropy": 5.77634220123291,
      "epoch": 0.4683889939088427,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004983343000068559,
      "loss": 5.5563,
      "mean_token_accuracy": 0.16145434379577636,
      "num_tokens": 10282206.0,
      "step": 5575
    },
    {
      "entropy": 5.734559774398804,
      "epoch": 0.46880907372400754,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004983306608583937,
      "loss": 5.5184,
      "mean_token_accuracy": 0.17214130163192748,
      "num_tokens": 10290056.0,
      "step": 5580
    },
    {
      "entropy": 5.843502569198608,
      "epoch": 0.46922915353917244,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004983270177537438,
      "loss": 5.5842,
      "mean_token_accuracy": 0.15954413563013076,
      "num_tokens": 10299726.0,
      "step": 5585
    },
    {
      "entropy": 5.810264730453492,
      "epoch": 0.46964923335433734,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004983233706929708,
      "loss": 5.6605,
      "mean_token_accuracy": 0.15892852544784547,
      "num_tokens": 10308696.0,
      "step": 5590
    },
    {
      "entropy": 5.862246799468994,
      "epoch": 0.4700693131695022,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004983197196761392,
      "loss": 5.7304,
      "mean_token_accuracy": 0.14815479144454002,
      "num_tokens": 10317845.0,
      "step": 5595
    },
    {
      "entropy": 5.828830194473267,
      "epoch": 0.4704893929846671,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004983160647033139,
      "loss": 5.6313,
      "mean_token_accuracy": 0.16088498681783675,
      "num_tokens": 10326563.0,
      "step": 5600
    },
    {
      "entropy": 5.822518301010132,
      "epoch": 0.470909472799832,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004983124057745595,
      "loss": 5.6024,
      "mean_token_accuracy": 0.15862844586372377,
      "num_tokens": 10335931.0,
      "step": 5605
    },
    {
      "entropy": 5.762676143646241,
      "epoch": 0.47132955261499687,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004983087428899408,
      "loss": 5.5949,
      "mean_token_accuracy": 0.15312879234552385,
      "num_tokens": 10344984.0,
      "step": 5610
    },
    {
      "entropy": 5.84120922088623,
      "epoch": 0.4717496324301617,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004983050760495227,
      "loss": 5.6335,
      "mean_token_accuracy": 0.15357797592878342,
      "num_tokens": 10353522.0,
      "step": 5615
    },
    {
      "entropy": 5.845833110809326,
      "epoch": 0.4721697122453266,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004983014052533702,
      "loss": 5.6424,
      "mean_token_accuracy": 0.15998404920101167,
      "num_tokens": 10363527.0,
      "step": 5620
    },
    {
      "entropy": 5.777768898010254,
      "epoch": 0.4725897920604915,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004982977305015481,
      "loss": 5.5728,
      "mean_token_accuracy": 0.15610021501779556,
      "num_tokens": 10372040.0,
      "step": 5625
    },
    {
      "entropy": 5.841713190078735,
      "epoch": 0.47300987187565635,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004982940517941219,
      "loss": 5.5468,
      "mean_token_accuracy": 0.15774558037519454,
      "num_tokens": 10381279.0,
      "step": 5630
    },
    {
      "entropy": 5.853977680206299,
      "epoch": 0.47342995169082125,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004982903691311564,
      "loss": 5.7304,
      "mean_token_accuracy": 0.15029727220535277,
      "num_tokens": 10390608.0,
      "step": 5635
    },
    {
      "entropy": 5.7988402366638185,
      "epoch": 0.47385003150598615,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004982866825127172,
      "loss": 5.5192,
      "mean_token_accuracy": 0.16449746191501619,
      "num_tokens": 10399851.0,
      "step": 5640
    },
    {
      "entropy": 5.867706441879273,
      "epoch": 0.47427011132115104,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004982829919388692,
      "loss": 5.7887,
      "mean_token_accuracy": 0.14914158508181571,
      "num_tokens": 10410425.0,
      "step": 5645
    },
    {
      "entropy": 5.776743412017822,
      "epoch": 0.4746901911363159,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004982792974096781,
      "loss": 5.5749,
      "mean_token_accuracy": 0.16128528863191605,
      "num_tokens": 10418783.0,
      "step": 5650
    },
    {
      "entropy": 5.915902757644654,
      "epoch": 0.4751102709514808,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000498275598925209,
      "loss": 5.7303,
      "mean_token_accuracy": 0.15162914544343947,
      "num_tokens": 10427360.0,
      "step": 5655
    },
    {
      "entropy": 5.907740354537964,
      "epoch": 0.4755303507666457,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004982718964855277,
      "loss": 5.6872,
      "mean_token_accuracy": 0.15605778843164445,
      "num_tokens": 10436613.0,
      "step": 5660
    },
    {
      "entropy": 5.824285364151001,
      "epoch": 0.4759504305818105,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004982681900907,
      "loss": 5.7422,
      "mean_token_accuracy": 0.1537246122956276,
      "num_tokens": 10445055.0,
      "step": 5665
    },
    {
      "entropy": 5.7850565910339355,
      "epoch": 0.4763705103969754,
      "grad_norm": 0.875,
      "learning_rate": 0.000498264479740791,
      "loss": 5.5536,
      "mean_token_accuracy": 0.16351223587989808,
      "num_tokens": 10454516.0,
      "step": 5670
    },
    {
      "entropy": 5.895203351974487,
      "epoch": 0.4767905902121403,
      "grad_norm": 0.875,
      "learning_rate": 0.0004982607654358668,
      "loss": 5.6981,
      "mean_token_accuracy": 0.15540930181741713,
      "num_tokens": 10463771.0,
      "step": 5675
    },
    {
      "entropy": 5.829155921936035,
      "epoch": 0.47721067002730516,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000498257047175993,
      "loss": 5.6192,
      "mean_token_accuracy": 0.15894964784383775,
      "num_tokens": 10473783.0,
      "step": 5680
    },
    {
      "entropy": 5.822129917144776,
      "epoch": 0.47763074984247006,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004982533249612357,
      "loss": 5.6042,
      "mean_token_accuracy": 0.15907904505729675,
      "num_tokens": 10483424.0,
      "step": 5685
    },
    {
      "entropy": 5.735984182357788,
      "epoch": 0.47805082965763496,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004982495987916607,
      "loss": 5.5272,
      "mean_token_accuracy": 0.16477657556533815,
      "num_tokens": 10492536.0,
      "step": 5690
    },
    {
      "entropy": 5.813406753540039,
      "epoch": 0.47847090947279985,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004982458686673339,
      "loss": 5.641,
      "mean_token_accuracy": 0.1593659669160843,
      "num_tokens": 10501616.0,
      "step": 5695
    },
    {
      "entropy": 5.93604531288147,
      "epoch": 0.4788909892879647,
      "grad_norm": 1.0,
      "learning_rate": 0.0004982421345883217,
      "loss": 5.6726,
      "mean_token_accuracy": 0.15468072444200515,
      "num_tokens": 10511190.0,
      "step": 5700
    },
    {
      "entropy": 5.774876070022583,
      "epoch": 0.4793110691031296,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004982383965546898,
      "loss": 5.6325,
      "mean_token_accuracy": 0.15240147411823274,
      "num_tokens": 10520310.0,
      "step": 5705
    },
    {
      "entropy": 5.855574083328247,
      "epoch": 0.4797311489182945,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004982346545665048,
      "loss": 5.6007,
      "mean_token_accuracy": 0.1570914402604103,
      "num_tokens": 10528711.0,
      "step": 5710
    },
    {
      "entropy": 5.828868722915649,
      "epoch": 0.48015122873345933,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004982309086238328,
      "loss": 5.6853,
      "mean_token_accuracy": 0.15123532563447953,
      "num_tokens": 10538484.0,
      "step": 5715
    },
    {
      "entropy": 5.844063568115234,
      "epoch": 0.48057130854862423,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004982271587267403,
      "loss": 5.6234,
      "mean_token_accuracy": 0.1559097334742546,
      "num_tokens": 10547623.0,
      "step": 5720
    },
    {
      "entropy": 5.850908041000366,
      "epoch": 0.48099138836378913,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004982234048752935,
      "loss": 5.5951,
      "mean_token_accuracy": 0.1569741114974022,
      "num_tokens": 10556234.0,
      "step": 5725
    },
    {
      "entropy": 5.9303936004638675,
      "epoch": 0.481411468178954,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498219647069559,
      "loss": 5.8031,
      "mean_token_accuracy": 0.149222881346941,
      "num_tokens": 10566308.0,
      "step": 5730
    },
    {
      "entropy": 5.862217617034912,
      "epoch": 0.48183154799411887,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004982158853096035,
      "loss": 5.7366,
      "mean_token_accuracy": 0.1512727364897728,
      "num_tokens": 10575212.0,
      "step": 5735
    },
    {
      "entropy": 5.835469961166382,
      "epoch": 0.48225162780928377,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004982121195954935,
      "loss": 5.5055,
      "mean_token_accuracy": 0.1651215985417366,
      "num_tokens": 10584590.0,
      "step": 5740
    },
    {
      "entropy": 5.801496887207032,
      "epoch": 0.48267170762444866,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004982083499272957,
      "loss": 5.563,
      "mean_token_accuracy": 0.16047275960445403,
      "num_tokens": 10593997.0,
      "step": 5745
    },
    {
      "entropy": 5.821460342407226,
      "epoch": 0.4830917874396135,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004982045763050768,
      "loss": 5.7078,
      "mean_token_accuracy": 0.15620884150266648,
      "num_tokens": 10603299.0,
      "step": 5750
    },
    {
      "entropy": 5.863379764556885,
      "epoch": 0.4835118672547784,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004982007987289041,
      "loss": 5.6286,
      "mean_token_accuracy": 0.15687449425458908,
      "num_tokens": 10613546.0,
      "step": 5755
    },
    {
      "entropy": 5.822661638259888,
      "epoch": 0.4839319470699433,
      "grad_norm": 1.0,
      "learning_rate": 0.0004981970171988439,
      "loss": 5.6026,
      "mean_token_accuracy": 0.16541317775845527,
      "num_tokens": 10622966.0,
      "step": 5760
    },
    {
      "entropy": 5.828573179244995,
      "epoch": 0.48435202688510814,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004981932317149636,
      "loss": 5.6795,
      "mean_token_accuracy": 0.15333941504359244,
      "num_tokens": 10633441.0,
      "step": 5765
    },
    {
      "entropy": 5.888944625854492,
      "epoch": 0.48477210670027304,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00049818944227733,
      "loss": 5.673,
      "mean_token_accuracy": 0.15536948442459106,
      "num_tokens": 10643124.0,
      "step": 5770
    },
    {
      "entropy": 5.874996185302734,
      "epoch": 0.48519218651543794,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004981856488860105,
      "loss": 5.645,
      "mean_token_accuracy": 0.14975375980138778,
      "num_tokens": 10652517.0,
      "step": 5775
    },
    {
      "entropy": 5.824694728851318,
      "epoch": 0.48561226633060284,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004981818515410721,
      "loss": 5.6734,
      "mean_token_accuracy": 0.1480187088251114,
      "num_tokens": 10663352.0,
      "step": 5780
    },
    {
      "entropy": 5.810465860366821,
      "epoch": 0.4860323461457677,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004981780502425821,
      "loss": 5.7037,
      "mean_token_accuracy": 0.15300920158624648,
      "num_tokens": 10672430.0,
      "step": 5785
    },
    {
      "entropy": 5.845775556564331,
      "epoch": 0.4864524259609326,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004981742449906079,
      "loss": 5.63,
      "mean_token_accuracy": 0.1613398462533951,
      "num_tokens": 10681908.0,
      "step": 5790
    },
    {
      "entropy": 5.840546178817749,
      "epoch": 0.4868725057760975,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004981704357852168,
      "loss": 5.6314,
      "mean_token_accuracy": 0.15710638910532,
      "num_tokens": 10691259.0,
      "step": 5795
    },
    {
      "entropy": 5.816694116592407,
      "epoch": 0.4872925855912623,
      "grad_norm": 1.0,
      "learning_rate": 0.0004981666226264764,
      "loss": 5.5358,
      "mean_token_accuracy": 0.1578466773033142,
      "num_tokens": 10699668.0,
      "step": 5800
    },
    {
      "entropy": 5.808956050872803,
      "epoch": 0.4877126654064272,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004981628055144542,
      "loss": 5.5867,
      "mean_token_accuracy": 0.16105588674545288,
      "num_tokens": 10709146.0,
      "step": 5805
    },
    {
      "entropy": 5.8652105808258055,
      "epoch": 0.4881327452215921,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004981589844492177,
      "loss": 5.6745,
      "mean_token_accuracy": 0.1511348471045494,
      "num_tokens": 10718724.0,
      "step": 5810
    },
    {
      "entropy": 5.8189291000366214,
      "epoch": 0.488552825036757,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004981551594308349,
      "loss": 5.6182,
      "mean_token_accuracy": 0.15741762444376944,
      "num_tokens": 10728101.0,
      "step": 5815
    },
    {
      "entropy": 5.91320481300354,
      "epoch": 0.48897290485192185,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004981513304593733,
      "loss": 5.6369,
      "mean_token_accuracy": 0.16215446591377258,
      "num_tokens": 10736750.0,
      "step": 5820
    },
    {
      "entropy": 5.888871526718139,
      "epoch": 0.48939298466708675,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004981474975349006,
      "loss": 5.8152,
      "mean_token_accuracy": 0.15442975759506225,
      "num_tokens": 10746914.0,
      "step": 5825
    },
    {
      "entropy": 5.866163349151611,
      "epoch": 0.48981306448225165,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000498143660657485,
      "loss": 5.6537,
      "mean_token_accuracy": 0.15946730375289916,
      "num_tokens": 10755786.0,
      "step": 5830
    },
    {
      "entropy": 5.732654571533203,
      "epoch": 0.4902331442974165,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004981398198271944,
      "loss": 5.5308,
      "mean_token_accuracy": 0.16208673119544983,
      "num_tokens": 10764821.0,
      "step": 5835
    },
    {
      "entropy": 5.8535669326782225,
      "epoch": 0.4906532241125814,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004981359750440968,
      "loss": 5.6202,
      "mean_token_accuracy": 0.15754470974206924,
      "num_tokens": 10773569.0,
      "step": 5840
    },
    {
      "entropy": 5.786517763137818,
      "epoch": 0.4910733039277463,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004981321263082603,
      "loss": 5.5918,
      "mean_token_accuracy": 0.1573310151696205,
      "num_tokens": 10782298.0,
      "step": 5845
    },
    {
      "entropy": 5.761690998077393,
      "epoch": 0.4914933837429111,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000498128273619753,
      "loss": 5.5705,
      "mean_token_accuracy": 0.16008659899234773,
      "num_tokens": 10792087.0,
      "step": 5850
    },
    {
      "entropy": 5.80288519859314,
      "epoch": 0.491913463558076,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004981244169786433,
      "loss": 5.6796,
      "mean_token_accuracy": 0.15478396266698838,
      "num_tokens": 10801641.0,
      "step": 5855
    },
    {
      "entropy": 5.914400911331176,
      "epoch": 0.4923335433732409,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004981205563849994,
      "loss": 5.733,
      "mean_token_accuracy": 0.15336815193295478,
      "num_tokens": 10811612.0,
      "step": 5860
    },
    {
      "entropy": 5.819022130966187,
      "epoch": 0.4927536231884058,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004981166918388897,
      "loss": 5.5649,
      "mean_token_accuracy": 0.1595151573419571,
      "num_tokens": 10821608.0,
      "step": 5865
    },
    {
      "entropy": 5.741395044326782,
      "epoch": 0.49317370300357066,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004981128233403828,
      "loss": 5.5233,
      "mean_token_accuracy": 0.16548489332199096,
      "num_tokens": 10830679.0,
      "step": 5870
    },
    {
      "entropy": 5.786447334289551,
      "epoch": 0.49359378281873556,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000498108950889547,
      "loss": 5.5906,
      "mean_token_accuracy": 0.15901205837726592,
      "num_tokens": 10839669.0,
      "step": 5875
    },
    {
      "entropy": 5.854657602310181,
      "epoch": 0.49401386263390046,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004981050744864512,
      "loss": 5.5673,
      "mean_token_accuracy": 0.15820209830999374,
      "num_tokens": 10849666.0,
      "step": 5880
    },
    {
      "entropy": 5.776970052719117,
      "epoch": 0.4944339424490653,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004981011941311638,
      "loss": 5.4867,
      "mean_token_accuracy": 0.16672125905752183,
      "num_tokens": 10858225.0,
      "step": 5885
    },
    {
      "entropy": 5.745939016342163,
      "epoch": 0.4948540222642302,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004980973098237535,
      "loss": 5.5962,
      "mean_token_accuracy": 0.15611702501773833,
      "num_tokens": 10867466.0,
      "step": 5890
    },
    {
      "entropy": 5.851855373382568,
      "epoch": 0.4952741020793951,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004980934215642894,
      "loss": 5.6284,
      "mean_token_accuracy": 0.16179797202348709,
      "num_tokens": 10875850.0,
      "step": 5895
    },
    {
      "entropy": 5.7541014671325685,
      "epoch": 0.49569418189456,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00049808952935284,
      "loss": 5.5593,
      "mean_token_accuracy": 0.16514879912137986,
      "num_tokens": 10885154.0,
      "step": 5900
    },
    {
      "entropy": 5.800168514251709,
      "epoch": 0.49611426170972484,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004980856331894747,
      "loss": 5.6605,
      "mean_token_accuracy": 0.15418792963027955,
      "num_tokens": 10894080.0,
      "step": 5905
    },
    {
      "entropy": 5.813483572006225,
      "epoch": 0.49653434152488973,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004980817330742621,
      "loss": 5.6576,
      "mean_token_accuracy": 0.15561991333961486,
      "num_tokens": 10903248.0,
      "step": 5910
    },
    {
      "entropy": 5.844412422180175,
      "epoch": 0.49695442134005463,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004980778290072716,
      "loss": 5.6122,
      "mean_token_accuracy": 0.16167805790901185,
      "num_tokens": 10912939.0,
      "step": 5915
    },
    {
      "entropy": 5.842522192001343,
      "epoch": 0.4973745011552195,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004980739209885722,
      "loss": 5.6308,
      "mean_token_accuracy": 0.15903968065977098,
      "num_tokens": 10921505.0,
      "step": 5920
    },
    {
      "entropy": 5.840193128585815,
      "epoch": 0.49779458097038437,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004980700090182331,
      "loss": 5.7062,
      "mean_token_accuracy": 0.15969687849283218,
      "num_tokens": 10931861.0,
      "step": 5925
    },
    {
      "entropy": 5.907305431365967,
      "epoch": 0.49821466078554927,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004980660930963238,
      "loss": 5.6246,
      "mean_token_accuracy": 0.15740073174238206,
      "num_tokens": 10940810.0,
      "step": 5930
    },
    {
      "entropy": 5.757447147369385,
      "epoch": 0.4986347406007141,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004980621732229133,
      "loss": 5.5067,
      "mean_token_accuracy": 0.15881143212318422,
      "num_tokens": 10949514.0,
      "step": 5935
    },
    {
      "entropy": 5.834470319747925,
      "epoch": 0.499054820415879,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004980582493980714,
      "loss": 5.7198,
      "mean_token_accuracy": 0.14656566232442855,
      "num_tokens": 10959161.0,
      "step": 5940
    },
    {
      "entropy": 5.808456516265869,
      "epoch": 0.4994749002310439,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004980543216218674,
      "loss": 5.5881,
      "mean_token_accuracy": 0.17089201658964157,
      "num_tokens": 10968983.0,
      "step": 5945
    },
    {
      "entropy": 5.847827053070068,
      "epoch": 0.4998949800462088,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004980503898943711,
      "loss": 5.7092,
      "mean_token_accuracy": 0.15717339664697647,
      "num_tokens": 10978044.0,
      "step": 5950
    },
    {
      "entropy": 5.825947999954224,
      "epoch": 0.5003150598613737,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004980464542156519,
      "loss": 5.6317,
      "mean_token_accuracy": 0.16252256333827972,
      "num_tokens": 10986980.0,
      "step": 5955
    },
    {
      "entropy": 5.819803857803345,
      "epoch": 0.5007351396765385,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004980425145857796,
      "loss": 5.5665,
      "mean_token_accuracy": 0.1698974922299385,
      "num_tokens": 10995163.0,
      "step": 5960
    },
    {
      "entropy": 5.767242860794068,
      "epoch": 0.5011552194917034,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000498038571004824,
      "loss": 5.4995,
      "mean_token_accuracy": 0.1690700516104698,
      "num_tokens": 11003722.0,
      "step": 5965
    },
    {
      "entropy": 5.714049768447876,
      "epoch": 0.5015752993068683,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004980346234728549,
      "loss": 5.564,
      "mean_token_accuracy": 0.16594353169202805,
      "num_tokens": 11013176.0,
      "step": 5970
    },
    {
      "entropy": 5.8587408542633055,
      "epoch": 0.5019953791220332,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004980306719899424,
      "loss": 5.6238,
      "mean_token_accuracy": 0.155168816447258,
      "num_tokens": 11022636.0,
      "step": 5975
    },
    {
      "entropy": 5.788705062866211,
      "epoch": 0.5024154589371981,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004980267165561564,
      "loss": 5.5592,
      "mean_token_accuracy": 0.16169780045747756,
      "num_tokens": 11031896.0,
      "step": 5980
    },
    {
      "entropy": 5.778947973251343,
      "epoch": 0.502835538752363,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004980227571715669,
      "loss": 5.6229,
      "mean_token_accuracy": 0.15644263625144958,
      "num_tokens": 11040802.0,
      "step": 5985
    },
    {
      "entropy": 5.795914840698242,
      "epoch": 0.5032556185675279,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004980187938362441,
      "loss": 5.5497,
      "mean_token_accuracy": 0.1557212933897972,
      "num_tokens": 11049701.0,
      "step": 5990
    },
    {
      "entropy": 5.845397663116455,
      "epoch": 0.5036756983826927,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004980148265502581,
      "loss": 5.7159,
      "mean_token_accuracy": 0.14954705834388732,
      "num_tokens": 11059555.0,
      "step": 5995
    },
    {
      "entropy": 5.861802339553833,
      "epoch": 0.5040957781978576,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004980108553136795,
      "loss": 5.6437,
      "mean_token_accuracy": 0.16013213694095613,
      "num_tokens": 11068940.0,
      "step": 6000
    },
    {
      "epoch": 0.5040957781978576,
      "eval_entropy": 5.670633745714113,
      "eval_loss": 5.616561412811279,
      "eval_mean_token_accuracy": 0.16696001174898584,
      "eval_num_tokens": 11068940.0,
      "eval_runtime": 20.941,
      "eval_samples_per_second": 1784.347,
      "eval_steps_per_second": 223.055,
      "step": 6000
    },
    {
      "entropy": 5.892019891738892,
      "epoch": 0.5045158580130225,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004980068801265783,
      "loss": 5.6127,
      "mean_token_accuracy": 0.16208406910300255,
      "num_tokens": 11079014.0,
      "step": 6005
    },
    {
      "entropy": 5.872786378860473,
      "epoch": 0.5049359378281874,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004980029009890251,
      "loss": 5.6911,
      "mean_token_accuracy": 0.1607103869318962,
      "num_tokens": 11089526.0,
      "step": 6010
    },
    {
      "entropy": 5.848162651062012,
      "epoch": 0.5053560176433523,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004979989179010904,
      "loss": 5.5885,
      "mean_token_accuracy": 0.1622884288430214,
      "num_tokens": 11099156.0,
      "step": 6015
    },
    {
      "entropy": 5.774669027328491,
      "epoch": 0.5057760974585171,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004979949308628445,
      "loss": 5.5678,
      "mean_token_accuracy": 0.16341736614704133,
      "num_tokens": 11108242.0,
      "step": 6020
    },
    {
      "entropy": 5.7315993309021,
      "epoch": 0.506196177273682,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004979909398743584,
      "loss": 5.5785,
      "mean_token_accuracy": 0.16548175141215324,
      "num_tokens": 11118076.0,
      "step": 6025
    },
    {
      "entropy": 5.869733619689941,
      "epoch": 0.5066162570888468,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004979869449357026,
      "loss": 5.6594,
      "mean_token_accuracy": 0.16638825982809066,
      "num_tokens": 11127265.0,
      "step": 6030
    },
    {
      "entropy": 5.842595624923706,
      "epoch": 0.5070363369040117,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004979829460469478,
      "loss": 5.5583,
      "mean_token_accuracy": 0.16190421134233474,
      "num_tokens": 11136429.0,
      "step": 6035
    },
    {
      "entropy": 5.795467519760132,
      "epoch": 0.5074564167191766,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004979789432081649,
      "loss": 5.5752,
      "mean_token_accuracy": 0.158992238342762,
      "num_tokens": 11146201.0,
      "step": 6040
    },
    {
      "entropy": 5.842192792892456,
      "epoch": 0.5078764965343415,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000497974936419425,
      "loss": 5.602,
      "mean_token_accuracy": 0.16134665459394454,
      "num_tokens": 11154867.0,
      "step": 6045
    },
    {
      "entropy": 5.719891548156738,
      "epoch": 0.5082965763495064,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004979709256807989,
      "loss": 5.6234,
      "mean_token_accuracy": 0.159073668718338,
      "num_tokens": 11164092.0,
      "step": 6050
    },
    {
      "entropy": 5.8586814403533936,
      "epoch": 0.5087166561646713,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004979669109923575,
      "loss": 5.6608,
      "mean_token_accuracy": 0.15583281964063644,
      "num_tokens": 11173176.0,
      "step": 6055
    },
    {
      "entropy": 5.8806725025177,
      "epoch": 0.5091367359798362,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004979628923541721,
      "loss": 5.6322,
      "mean_token_accuracy": 0.1580136314034462,
      "num_tokens": 11182397.0,
      "step": 6060
    },
    {
      "entropy": 5.8351644515991214,
      "epoch": 0.509556815795001,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000497958869766314,
      "loss": 5.6564,
      "mean_token_accuracy": 0.15806965231895448,
      "num_tokens": 11191790.0,
      "step": 6065
    },
    {
      "entropy": 5.795283222198487,
      "epoch": 0.5099768956101659,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004979548432288543,
      "loss": 5.5694,
      "mean_token_accuracy": 0.16827381253242493,
      "num_tokens": 11201104.0,
      "step": 6070
    },
    {
      "entropy": 5.825335931777954,
      "epoch": 0.5103969754253308,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004979508127418643,
      "loss": 5.5689,
      "mean_token_accuracy": 0.1612475499510765,
      "num_tokens": 11209578.0,
      "step": 6075
    },
    {
      "entropy": 5.79924750328064,
      "epoch": 0.5108170552404957,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004979467783054155,
      "loss": 5.5304,
      "mean_token_accuracy": 0.16983622163534165,
      "num_tokens": 11218380.0,
      "step": 6080
    },
    {
      "entropy": 5.728046512603759,
      "epoch": 0.5112371350556606,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004979427399195793,
      "loss": 5.5533,
      "mean_token_accuracy": 0.15939420759677886,
      "num_tokens": 11227810.0,
      "step": 6085
    },
    {
      "entropy": 5.797479343414307,
      "epoch": 0.5116572148708255,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004979386975844274,
      "loss": 5.5678,
      "mean_token_accuracy": 0.1583065688610077,
      "num_tokens": 11236631.0,
      "step": 6090
    },
    {
      "entropy": 5.7882331848144535,
      "epoch": 0.5120772946859904,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004979346513000311,
      "loss": 5.6299,
      "mean_token_accuracy": 0.15695726424455642,
      "num_tokens": 11247418.0,
      "step": 6095
    },
    {
      "entropy": 5.794447517395019,
      "epoch": 0.5124973745011552,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004979306010664623,
      "loss": 5.5274,
      "mean_token_accuracy": 0.16943604946136476,
      "num_tokens": 11256246.0,
      "step": 6100
    },
    {
      "entropy": 5.6955108642578125,
      "epoch": 0.5129174543163201,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004979265468837927,
      "loss": 5.522,
      "mean_token_accuracy": 0.1628118321299553,
      "num_tokens": 11265980.0,
      "step": 6105
    },
    {
      "entropy": 5.830253028869629,
      "epoch": 0.513337534131485,
      "grad_norm": 0.90625,
      "learning_rate": 0.000497922488752094,
      "loss": 5.5666,
      "mean_token_accuracy": 0.16191249042749406,
      "num_tokens": 11276158.0,
      "step": 6110
    },
    {
      "entropy": 5.754393672943115,
      "epoch": 0.5137576139466499,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004979184266714383,
      "loss": 5.4801,
      "mean_token_accuracy": 0.16416939049959184,
      "num_tokens": 11284957.0,
      "step": 6115
    },
    {
      "entropy": 5.68437647819519,
      "epoch": 0.5141776937618148,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004979143606418974,
      "loss": 5.4989,
      "mean_token_accuracy": 0.1635582208633423,
      "num_tokens": 11294340.0,
      "step": 6120
    },
    {
      "entropy": 5.882603931427002,
      "epoch": 0.5145977735769797,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004979102906635435,
      "loss": 5.7626,
      "mean_token_accuracy": 0.15395218133926392,
      "num_tokens": 11303344.0,
      "step": 6125
    },
    {
      "entropy": 5.8543013572692875,
      "epoch": 0.5150178533921445,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004979062167364486,
      "loss": 5.6329,
      "mean_token_accuracy": 0.1685298874974251,
      "num_tokens": 11311338.0,
      "step": 6130
    },
    {
      "entropy": 5.742776393890381,
      "epoch": 0.5154379332073094,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004979021388606847,
      "loss": 5.4557,
      "mean_token_accuracy": 0.17043802589178086,
      "num_tokens": 11320194.0,
      "step": 6135
    },
    {
      "entropy": 5.752661800384521,
      "epoch": 0.5158580130224742,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004978980570363243,
      "loss": 5.6292,
      "mean_token_accuracy": 0.16262891739606858,
      "num_tokens": 11329952.0,
      "step": 6140
    },
    {
      "entropy": 5.749009799957276,
      "epoch": 0.5162780928376391,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004978939712634396,
      "loss": 5.5872,
      "mean_token_accuracy": 0.15799860954284667,
      "num_tokens": 11339384.0,
      "step": 6145
    },
    {
      "entropy": 5.864343166351318,
      "epoch": 0.516698172652804,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004978898815421029,
      "loss": 5.7398,
      "mean_token_accuracy": 0.1542922481894493,
      "num_tokens": 11348409.0,
      "step": 6150
    },
    {
      "entropy": 5.910318660736084,
      "epoch": 0.5171182524679689,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004978857878723867,
      "loss": 5.6355,
      "mean_token_accuracy": 0.16374070346355438,
      "num_tokens": 11357478.0,
      "step": 6155
    },
    {
      "entropy": 5.852247953414917,
      "epoch": 0.5175383322831338,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004978816902543636,
      "loss": 5.6758,
      "mean_token_accuracy": 0.15980569571256636,
      "num_tokens": 11366379.0,
      "step": 6160
    },
    {
      "entropy": 5.8074695587158205,
      "epoch": 0.5179584120982986,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004978775886881062,
      "loss": 5.6683,
      "mean_token_accuracy": 0.1595798946917057,
      "num_tokens": 11376357.0,
      "step": 6165
    },
    {
      "entropy": 5.806489706039429,
      "epoch": 0.5183784919134635,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000497873483173687,
      "loss": 5.5574,
      "mean_token_accuracy": 0.1633630320429802,
      "num_tokens": 11384995.0,
      "step": 6170
    },
    {
      "entropy": 5.765994882583618,
      "epoch": 0.5187985717286284,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004978693737111787,
      "loss": 5.5498,
      "mean_token_accuracy": 0.16460832059383393,
      "num_tokens": 11395363.0,
      "step": 6175
    },
    {
      "entropy": 5.784793043136597,
      "epoch": 0.5192186515437933,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004978652603006543,
      "loss": 5.547,
      "mean_token_accuracy": 0.16279931366443634,
      "num_tokens": 11404511.0,
      "step": 6180
    },
    {
      "entropy": 5.849311590194702,
      "epoch": 0.5196387313589582,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004978611429421866,
      "loss": 5.5901,
      "mean_token_accuracy": 0.1584021270275116,
      "num_tokens": 11413400.0,
      "step": 6185
    },
    {
      "entropy": 5.8417051315307615,
      "epoch": 0.5200588111741231,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004978570216358485,
      "loss": 5.6448,
      "mean_token_accuracy": 0.14908464699983598,
      "num_tokens": 11423693.0,
      "step": 6190
    },
    {
      "entropy": 5.817714643478394,
      "epoch": 0.520478890989288,
      "grad_norm": 0.875,
      "learning_rate": 0.000497852896381713,
      "loss": 5.6142,
      "mean_token_accuracy": 0.1532459408044815,
      "num_tokens": 11433195.0,
      "step": 6195
    },
    {
      "entropy": 5.828872537612915,
      "epoch": 0.5208989708044528,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004978487671798531,
      "loss": 5.7567,
      "mean_token_accuracy": 0.1504699096083641,
      "num_tokens": 11443416.0,
      "step": 6200
    },
    {
      "entropy": 5.889235591888427,
      "epoch": 0.5213190506196177,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004978446340303422,
      "loss": 5.6046,
      "mean_token_accuracy": 0.15886004865169526,
      "num_tokens": 11452487.0,
      "step": 6205
    },
    {
      "entropy": 5.838255834579468,
      "epoch": 0.5217391304347826,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004978404969332533,
      "loss": 5.612,
      "mean_token_accuracy": 0.16928871273994445,
      "num_tokens": 11461893.0,
      "step": 6210
    },
    {
      "entropy": 5.759885358810425,
      "epoch": 0.5221592102499475,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004978363558886597,
      "loss": 5.5448,
      "mean_token_accuracy": 0.15130849480628966,
      "num_tokens": 11471238.0,
      "step": 6215
    },
    {
      "entropy": 5.782651662826538,
      "epoch": 0.5225792900651124,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004978322108966348,
      "loss": 5.6447,
      "mean_token_accuracy": 0.15431271195411683,
      "num_tokens": 11480571.0,
      "step": 6220
    },
    {
      "entropy": 5.777035188674927,
      "epoch": 0.5229993698802773,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004978280619572521,
      "loss": 5.6347,
      "mean_token_accuracy": 0.15825602263212205,
      "num_tokens": 11489552.0,
      "step": 6225
    },
    {
      "entropy": 5.847810840606689,
      "epoch": 0.5234194496954422,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000497823909070585,
      "loss": 5.6792,
      "mean_token_accuracy": 0.15690901577472688,
      "num_tokens": 11498715.0,
      "step": 6230
    },
    {
      "entropy": 5.850524520874023,
      "epoch": 0.523839529510607,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004978197522367071,
      "loss": 5.623,
      "mean_token_accuracy": 0.152306155115366,
      "num_tokens": 11508472.0,
      "step": 6235
    },
    {
      "entropy": 5.872546720504761,
      "epoch": 0.5242596093257719,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004978155914556919,
      "loss": 5.5565,
      "mean_token_accuracy": 0.16584993600845338,
      "num_tokens": 11517620.0,
      "step": 6240
    },
    {
      "entropy": 5.77497296333313,
      "epoch": 0.5246796891409368,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004978114267276134,
      "loss": 5.6051,
      "mean_token_accuracy": 0.16037460714578627,
      "num_tokens": 11526106.0,
      "step": 6245
    },
    {
      "entropy": 5.8083820819854735,
      "epoch": 0.5250997689561017,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004978072580525451,
      "loss": 5.6483,
      "mean_token_accuracy": 0.16743095070123673,
      "num_tokens": 11535840.0,
      "step": 6250
    },
    {
      "entropy": 5.864418029785156,
      "epoch": 0.5255198487712666,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000497803085430561,
      "loss": 5.6222,
      "mean_token_accuracy": 0.16470124199986458,
      "num_tokens": 11545110.0,
      "step": 6255
    },
    {
      "entropy": 5.888049030303955,
      "epoch": 0.5259399285864315,
      "grad_norm": 1.0,
      "learning_rate": 0.0004977989088617349,
      "loss": 5.6516,
      "mean_token_accuracy": 0.1539329208433628,
      "num_tokens": 11554382.0,
      "step": 6260
    },
    {
      "entropy": 5.767773246765136,
      "epoch": 0.5263600084015964,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000497794728346141,
      "loss": 5.5138,
      "mean_token_accuracy": 0.16381160616874696,
      "num_tokens": 11562821.0,
      "step": 6265
    },
    {
      "entropy": 5.893789291381836,
      "epoch": 0.5267800882167611,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004977905438838531,
      "loss": 5.7273,
      "mean_token_accuracy": 0.15106379389762878,
      "num_tokens": 11571705.0,
      "step": 6270
    },
    {
      "entropy": 5.702169561386109,
      "epoch": 0.527200168031926,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004977863554749453,
      "loss": 5.5516,
      "mean_token_accuracy": 0.16104080229997636,
      "num_tokens": 11580692.0,
      "step": 6275
    },
    {
      "entropy": 5.785028123855591,
      "epoch": 0.5276202478470909,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004977821631194922,
      "loss": 5.5577,
      "mean_token_accuracy": 0.15807193517684937,
      "num_tokens": 11589966.0,
      "step": 6280
    },
    {
      "entropy": 5.850618124008179,
      "epoch": 0.5280403276622558,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004977779668175677,
      "loss": 5.6314,
      "mean_token_accuracy": 0.15651484876871108,
      "num_tokens": 11599627.0,
      "step": 6285
    },
    {
      "entropy": 5.793644523620605,
      "epoch": 0.5284604074774207,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004977737665692461,
      "loss": 5.6144,
      "mean_token_accuracy": 0.16737879067659378,
      "num_tokens": 11608431.0,
      "step": 6290
    },
    {
      "entropy": 5.801764106750488,
      "epoch": 0.5288804872925856,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004977695623746021,
      "loss": 5.4887,
      "mean_token_accuracy": 0.15994460731744767,
      "num_tokens": 11617552.0,
      "step": 6295
    },
    {
      "entropy": 5.761782693862915,
      "epoch": 0.5293005671077504,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004977653542337099,
      "loss": 5.5293,
      "mean_token_accuracy": 0.16859177947044374,
      "num_tokens": 11626828.0,
      "step": 6300
    },
    {
      "entropy": 5.7847129821777346,
      "epoch": 0.5297206469229153,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004977611421466443,
      "loss": 5.6242,
      "mean_token_accuracy": 0.1562230333685875,
      "num_tokens": 11635867.0,
      "step": 6305
    },
    {
      "entropy": 5.839254665374756,
      "epoch": 0.5301407267380802,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004977569261134797,
      "loss": 5.5369,
      "mean_token_accuracy": 0.16260915696620942,
      "num_tokens": 11644711.0,
      "step": 6310
    },
    {
      "entropy": 5.811285734176636,
      "epoch": 0.5305608065532451,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004977527061342908,
      "loss": 5.6101,
      "mean_token_accuracy": 0.1652897745370865,
      "num_tokens": 11653320.0,
      "step": 6315
    },
    {
      "entropy": 5.794872760772705,
      "epoch": 0.53098088636841,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004977484822091524,
      "loss": 5.5717,
      "mean_token_accuracy": 0.16471439003944396,
      "num_tokens": 11662753.0,
      "step": 6320
    },
    {
      "entropy": 5.816572046279907,
      "epoch": 0.5314009661835749,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004977442543381394,
      "loss": 5.6096,
      "mean_token_accuracy": 0.1561722218990326,
      "num_tokens": 11671622.0,
      "step": 6325
    },
    {
      "entropy": 5.834009218215942,
      "epoch": 0.5318210459987398,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004977400225213266,
      "loss": 5.5972,
      "mean_token_accuracy": 0.15700093060731887,
      "num_tokens": 11679964.0,
      "step": 6330
    },
    {
      "entropy": 5.764689970016479,
      "epoch": 0.5322411258139046,
      "grad_norm": 0.921875,
      "learning_rate": 0.000497735786758789,
      "loss": 5.5543,
      "mean_token_accuracy": 0.16563338190317153,
      "num_tokens": 11688700.0,
      "step": 6335
    },
    {
      "entropy": 5.801178979873657,
      "epoch": 0.5326612056290695,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004977315470506016,
      "loss": 5.6646,
      "mean_token_accuracy": 0.1598043382167816,
      "num_tokens": 11698425.0,
      "step": 6340
    },
    {
      "entropy": 5.897024726867675,
      "epoch": 0.5330812854442344,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004977273033968397,
      "loss": 5.6493,
      "mean_token_accuracy": 0.15482330769300462,
      "num_tokens": 11707705.0,
      "step": 6345
    },
    {
      "entropy": 5.781996250152588,
      "epoch": 0.5335013652593993,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004977230557975782,
      "loss": 5.5507,
      "mean_token_accuracy": 0.16277959197759628,
      "num_tokens": 11717079.0,
      "step": 6350
    },
    {
      "entropy": 5.79433765411377,
      "epoch": 0.5339214450745642,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004977188042528923,
      "loss": 5.5477,
      "mean_token_accuracy": 0.15767836570739746,
      "num_tokens": 11725504.0,
      "step": 6355
    },
    {
      "entropy": 5.8038781642913815,
      "epoch": 0.5343415248897291,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004977145487628576,
      "loss": 5.6133,
      "mean_token_accuracy": 0.16047950983047485,
      "num_tokens": 11735282.0,
      "step": 6360
    },
    {
      "entropy": 5.802271318435669,
      "epoch": 0.534761604704894,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004977102893275494,
      "loss": 5.5949,
      "mean_token_accuracy": 0.15906881988048555,
      "num_tokens": 11744827.0,
      "step": 6365
    },
    {
      "entropy": 5.830772161483765,
      "epoch": 0.5351816845200588,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000497706025947043,
      "loss": 5.5545,
      "mean_token_accuracy": 0.16256064921617508,
      "num_tokens": 11753066.0,
      "step": 6370
    },
    {
      "entropy": 5.8030619621276855,
      "epoch": 0.5356017643352237,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004977017586214142,
      "loss": 5.5912,
      "mean_token_accuracy": 0.1619131088256836,
      "num_tokens": 11761190.0,
      "step": 6375
    },
    {
      "entropy": 5.814495420455932,
      "epoch": 0.5360218441503886,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004976974873507382,
      "loss": 5.5577,
      "mean_token_accuracy": 0.16317787915468215,
      "num_tokens": 11770321.0,
      "step": 6380
    },
    {
      "entropy": 5.764397287368775,
      "epoch": 0.5364419239655535,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000497693212135091,
      "loss": 5.6192,
      "mean_token_accuracy": 0.15596141070127487,
      "num_tokens": 11778388.0,
      "step": 6385
    },
    {
      "entropy": 5.84173469543457,
      "epoch": 0.5368620037807184,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004976889329745482,
      "loss": 5.4878,
      "mean_token_accuracy": 0.1627799853682518,
      "num_tokens": 11786250.0,
      "step": 6390
    },
    {
      "entropy": 5.719529485702514,
      "epoch": 0.5372820835958833,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004976846498691857,
      "loss": 5.451,
      "mean_token_accuracy": 0.16578584611415864,
      "num_tokens": 11794831.0,
      "step": 6395
    },
    {
      "entropy": 5.760023069381714,
      "epoch": 0.5377021634110482,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004976803628190792,
      "loss": 5.5108,
      "mean_token_accuracy": 0.17098050117492675,
      "num_tokens": 11803550.0,
      "step": 6400
    },
    {
      "entropy": 5.79886531829834,
      "epoch": 0.5381222432262129,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004976760718243047,
      "loss": 5.5775,
      "mean_token_accuracy": 0.16045399457216264,
      "num_tokens": 11812478.0,
      "step": 6405
    },
    {
      "entropy": 5.7917827606201175,
      "epoch": 0.5385423230413778,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004976717768849383,
      "loss": 5.5449,
      "mean_token_accuracy": 0.1533448353409767,
      "num_tokens": 11822463.0,
      "step": 6410
    },
    {
      "entropy": 5.791978216171264,
      "epoch": 0.5389624028565427,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004976674780010561,
      "loss": 5.5904,
      "mean_token_accuracy": 0.1517584055662155,
      "num_tokens": 11831853.0,
      "step": 6415
    },
    {
      "entropy": 5.8261794090271,
      "epoch": 0.5393824826717076,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000497663175172734,
      "loss": 5.6033,
      "mean_token_accuracy": 0.157386177778244,
      "num_tokens": 11841574.0,
      "step": 6420
    },
    {
      "entropy": 5.813488960266113,
      "epoch": 0.5398025624868725,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004976588684000486,
      "loss": 5.6851,
      "mean_token_accuracy": 0.14666915237903594,
      "num_tokens": 11852489.0,
      "step": 6425
    },
    {
      "entropy": 5.8010648727417,
      "epoch": 0.5402226423020374,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004976545576830759,
      "loss": 5.5748,
      "mean_token_accuracy": 0.1548174649477005,
      "num_tokens": 11861499.0,
      "step": 6430
    },
    {
      "entropy": 5.796115446090698,
      "epoch": 0.5406427221172023,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004976502430218924,
      "loss": 5.646,
      "mean_token_accuracy": 0.15776485875248908,
      "num_tokens": 11871685.0,
      "step": 6435
    },
    {
      "entropy": 5.803308534622192,
      "epoch": 0.5410628019323671,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004976459244165744,
      "loss": 5.5567,
      "mean_token_accuracy": 0.16087824404239653,
      "num_tokens": 11881340.0,
      "step": 6440
    },
    {
      "entropy": 5.778725004196167,
      "epoch": 0.541482881747532,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004976416018671986,
      "loss": 5.5702,
      "mean_token_accuracy": 0.16107980906963348,
      "num_tokens": 11890700.0,
      "step": 6445
    },
    {
      "entropy": 5.78373007774353,
      "epoch": 0.5419029615626969,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004976372753738415,
      "loss": 5.5726,
      "mean_token_accuracy": 0.1566349372267723,
      "num_tokens": 11900329.0,
      "step": 6450
    },
    {
      "entropy": 5.905847835540771,
      "epoch": 0.5423230413778618,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004976329449365795,
      "loss": 5.6072,
      "mean_token_accuracy": 0.15164827257394792,
      "num_tokens": 11909915.0,
      "step": 6455
    },
    {
      "entropy": 5.798289060592651,
      "epoch": 0.5427431211930267,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004976286105554897,
      "loss": 5.6337,
      "mean_token_accuracy": 0.15849536657333374,
      "num_tokens": 11918302.0,
      "step": 6460
    },
    {
      "entropy": 5.803639078140259,
      "epoch": 0.5431632010081916,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004976242722306487,
      "loss": 5.5735,
      "mean_token_accuracy": 0.16022982448339462,
      "num_tokens": 11927794.0,
      "step": 6465
    },
    {
      "entropy": 5.831341409683228,
      "epoch": 0.5435832808233564,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004976199299621333,
      "loss": 5.6019,
      "mean_token_accuracy": 0.16211526244878768,
      "num_tokens": 11937701.0,
      "step": 6470
    },
    {
      "entropy": 5.733603811264038,
      "epoch": 0.5440033606385213,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004976155837500205,
      "loss": 5.517,
      "mean_token_accuracy": 0.1622432976961136,
      "num_tokens": 11946106.0,
      "step": 6475
    },
    {
      "entropy": 5.745950984954834,
      "epoch": 0.5444234404536862,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004976112335943872,
      "loss": 5.4524,
      "mean_token_accuracy": 0.16185834854841233,
      "num_tokens": 11954604.0,
      "step": 6480
    },
    {
      "entropy": 5.6879777908325195,
      "epoch": 0.5448435202688511,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004976068794953106,
      "loss": 5.5075,
      "mean_token_accuracy": 0.16735649704933167,
      "num_tokens": 11963664.0,
      "step": 6485
    },
    {
      "entropy": 5.789367198944092,
      "epoch": 0.545263600084016,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004976025214528677,
      "loss": 5.4925,
      "mean_token_accuracy": 0.1637357786297798,
      "num_tokens": 11973426.0,
      "step": 6490
    },
    {
      "entropy": 5.750045394897461,
      "epoch": 0.5456836798991809,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004975981594671359,
      "loss": 5.5561,
      "mean_token_accuracy": 0.1580976128578186,
      "num_tokens": 11982339.0,
      "step": 6495
    },
    {
      "entropy": 5.818325185775757,
      "epoch": 0.5461037597143458,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004975937935381921,
      "loss": 5.5905,
      "mean_token_accuracy": 0.16282343864440918,
      "num_tokens": 11992016.0,
      "step": 6500
    },
    {
      "entropy": 5.715883493423462,
      "epoch": 0.5465238395295106,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000497589423666114,
      "loss": 5.6008,
      "mean_token_accuracy": 0.15767328143119813,
      "num_tokens": 12000616.0,
      "step": 6505
    },
    {
      "entropy": 5.636272621154785,
      "epoch": 0.5469439193446755,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004975850498509789,
      "loss": 5.5032,
      "mean_token_accuracy": 0.16611733734607698,
      "num_tokens": 12009717.0,
      "step": 6510
    },
    {
      "entropy": 5.756056118011474,
      "epoch": 0.5473639991598404,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004975806720928642,
      "loss": 5.5833,
      "mean_token_accuracy": 0.16398886144161223,
      "num_tokens": 12018020.0,
      "step": 6515
    },
    {
      "entropy": 5.786145830154419,
      "epoch": 0.5477840789750053,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004975762903918475,
      "loss": 5.5716,
      "mean_token_accuracy": 0.15816354602575303,
      "num_tokens": 12027119.0,
      "step": 6520
    },
    {
      "entropy": 5.822166204452515,
      "epoch": 0.5482041587901701,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004975719047480064,
      "loss": 5.5654,
      "mean_token_accuracy": 0.1643454521894455,
      "num_tokens": 12035566.0,
      "step": 6525
    },
    {
      "entropy": 5.764070987701416,
      "epoch": 0.548624238605335,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004975675151614187,
      "loss": 5.4705,
      "mean_token_accuracy": 0.16931594610214235,
      "num_tokens": 12044505.0,
      "step": 6530
    },
    {
      "entropy": 5.648342418670654,
      "epoch": 0.5490443184204999,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000497563121632162,
      "loss": 5.5423,
      "mean_token_accuracy": 0.16287525296211242,
      "num_tokens": 12053338.0,
      "step": 6535
    },
    {
      "entropy": 5.7813294410705565,
      "epoch": 0.5494643982356647,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004975587241603142,
      "loss": 5.5433,
      "mean_token_accuracy": 0.15935882478952407,
      "num_tokens": 12063235.0,
      "step": 6540
    },
    {
      "entropy": 5.866084289550781,
      "epoch": 0.5498844780508296,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004975543227459533,
      "loss": 5.6222,
      "mean_token_accuracy": 0.15872533470392228,
      "num_tokens": 12072490.0,
      "step": 6545
    },
    {
      "entropy": 5.813476037979126,
      "epoch": 0.5503045578659945,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004975499173891571,
      "loss": 5.6719,
      "mean_token_accuracy": 0.15454915761947632,
      "num_tokens": 12081474.0,
      "step": 6550
    },
    {
      "entropy": 5.751346492767334,
      "epoch": 0.5507246376811594,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004975455080900037,
      "loss": 5.5374,
      "mean_token_accuracy": 0.1650713264942169,
      "num_tokens": 12090963.0,
      "step": 6555
    },
    {
      "entropy": 5.78758487701416,
      "epoch": 0.5511447174963243,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004975410948485713,
      "loss": 5.5407,
      "mean_token_accuracy": 0.16168846040964127,
      "num_tokens": 12099786.0,
      "step": 6560
    },
    {
      "entropy": 5.749557781219482,
      "epoch": 0.5515647973114892,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004975366776649379,
      "loss": 5.5506,
      "mean_token_accuracy": 0.16442051529884338,
      "num_tokens": 12108469.0,
      "step": 6565
    },
    {
      "entropy": 5.772415590286255,
      "epoch": 0.5519848771266541,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004975322565391818,
      "loss": 5.5196,
      "mean_token_accuracy": 0.16210478246212007,
      "num_tokens": 12118287.0,
      "step": 6570
    },
    {
      "entropy": 5.87158613204956,
      "epoch": 0.5524049569418189,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004975278314713814,
      "loss": 5.691,
      "mean_token_accuracy": 0.1591292105615139,
      "num_tokens": 12127122.0,
      "step": 6575
    },
    {
      "entropy": 5.80229229927063,
      "epoch": 0.5528250367569838,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004975234024616152,
      "loss": 5.5818,
      "mean_token_accuracy": 0.1666329622268677,
      "num_tokens": 12136395.0,
      "step": 6580
    },
    {
      "entropy": 5.67426233291626,
      "epoch": 0.5532451165721487,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004975189695099613,
      "loss": 5.5549,
      "mean_token_accuracy": 0.1633012667298317,
      "num_tokens": 12145025.0,
      "step": 6585
    },
    {
      "entropy": 5.831414222717285,
      "epoch": 0.5536651963873136,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004975145326164985,
      "loss": 5.6074,
      "mean_token_accuracy": 0.15783063024282457,
      "num_tokens": 12154352.0,
      "step": 6590
    },
    {
      "entropy": 5.772958660125733,
      "epoch": 0.5540852762024785,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004975100917813055,
      "loss": 5.5004,
      "mean_token_accuracy": 0.16332688182592392,
      "num_tokens": 12163802.0,
      "step": 6595
    },
    {
      "entropy": 5.773453712463379,
      "epoch": 0.5545053560176434,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004975056470044606,
      "loss": 5.5248,
      "mean_token_accuracy": 0.16305141448974608,
      "num_tokens": 12173111.0,
      "step": 6600
    },
    {
      "entropy": 5.801900053024292,
      "epoch": 0.5549254358328082,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004975011982860428,
      "loss": 5.5841,
      "mean_token_accuracy": 0.162354776263237,
      "num_tokens": 12182048.0,
      "step": 6605
    },
    {
      "entropy": 5.766907835006714,
      "epoch": 0.5553455156479731,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004974967456261309,
      "loss": 5.5636,
      "mean_token_accuracy": 0.16079073697328566,
      "num_tokens": 12191501.0,
      "step": 6610
    },
    {
      "entropy": 5.792457342147827,
      "epoch": 0.555765595463138,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004974922890248036,
      "loss": 5.5858,
      "mean_token_accuracy": 0.16431280374526977,
      "num_tokens": 12201132.0,
      "step": 6615
    },
    {
      "entropy": 5.886992406845093,
      "epoch": 0.5561856752783029,
      "grad_norm": 0.984375,
      "learning_rate": 0.00049748782848214,
      "loss": 5.7236,
      "mean_token_accuracy": 0.15739185065031053,
      "num_tokens": 12211082.0,
      "step": 6620
    },
    {
      "entropy": 5.804878759384155,
      "epoch": 0.5566057550934678,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004974833639982192,
      "loss": 5.5381,
      "mean_token_accuracy": 0.16676106601953505,
      "num_tokens": 12219946.0,
      "step": 6625
    },
    {
      "entropy": 5.824381923675537,
      "epoch": 0.5570258349086327,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00049747889557312,
      "loss": 5.6496,
      "mean_token_accuracy": 0.15940177738666533,
      "num_tokens": 12229668.0,
      "step": 6630
    },
    {
      "entropy": 5.82171139717102,
      "epoch": 0.5574459147237976,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004974744232069219,
      "loss": 5.6258,
      "mean_token_accuracy": 0.16216015815734863,
      "num_tokens": 12238750.0,
      "step": 6635
    },
    {
      "entropy": 5.768255758285522,
      "epoch": 0.5578659945389624,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004974699468997038,
      "loss": 5.5745,
      "mean_token_accuracy": 0.15991929471492766,
      "num_tokens": 12246825.0,
      "step": 6640
    },
    {
      "entropy": 5.724027681350708,
      "epoch": 0.5582860743541272,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004974654666515452,
      "loss": 5.5063,
      "mean_token_accuracy": 0.16335234940052032,
      "num_tokens": 12256413.0,
      "step": 6645
    },
    {
      "entropy": 5.777611970901489,
      "epoch": 0.5587061541692921,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004974609824625254,
      "loss": 5.5374,
      "mean_token_accuracy": 0.1716291382908821,
      "num_tokens": 12265458.0,
      "step": 6650
    },
    {
      "entropy": 5.680338525772095,
      "epoch": 0.559126233984457,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004974564943327239,
      "loss": 5.4761,
      "mean_token_accuracy": 0.16565128564834594,
      "num_tokens": 12274124.0,
      "step": 6655
    },
    {
      "entropy": 5.616445541381836,
      "epoch": 0.5595463137996219,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00049745200226222,
      "loss": 5.4574,
      "mean_token_accuracy": 0.1757717177271843,
      "num_tokens": 12283513.0,
      "step": 6660
    },
    {
      "entropy": 5.72118616104126,
      "epoch": 0.5599663936147868,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004974475062510936,
      "loss": 5.578,
      "mean_token_accuracy": 0.16630735546350478,
      "num_tokens": 12292396.0,
      "step": 6665
    },
    {
      "entropy": 5.790207576751709,
      "epoch": 0.5603864734299517,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004974430062994242,
      "loss": 5.6008,
      "mean_token_accuracy": 0.1657990723848343,
      "num_tokens": 12301604.0,
      "step": 6670
    },
    {
      "entropy": 5.872246789932251,
      "epoch": 0.5608065532451165,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004974385024072912,
      "loss": 5.6379,
      "mean_token_accuracy": 0.1586868643760681,
      "num_tokens": 12310458.0,
      "step": 6675
    },
    {
      "entropy": 5.845408535003662,
      "epoch": 0.5612266330602814,
      "grad_norm": 0.96875,
      "learning_rate": 0.000497433994574775,
      "loss": 5.6392,
      "mean_token_accuracy": 0.15612249374389647,
      "num_tokens": 12319620.0,
      "step": 6680
    },
    {
      "entropy": 5.836484575271607,
      "epoch": 0.5616467128754463,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000497429482801955,
      "loss": 5.6913,
      "mean_token_accuracy": 0.15891646072268487,
      "num_tokens": 12329518.0,
      "step": 6685
    },
    {
      "entropy": 5.771274280548096,
      "epoch": 0.5620667926906112,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004974249670889111,
      "loss": 5.5072,
      "mean_token_accuracy": 0.16867185533046722,
      "num_tokens": 12338244.0,
      "step": 6690
    },
    {
      "entropy": 5.84527816772461,
      "epoch": 0.5624868725057761,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004974204474357237,
      "loss": 5.6853,
      "mean_token_accuracy": 0.1586710885167122,
      "num_tokens": 12347962.0,
      "step": 6695
    },
    {
      "entropy": 5.82287187576294,
      "epoch": 0.562906952320941,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004974159238424723,
      "loss": 5.5952,
      "mean_token_accuracy": 0.15904147922992706,
      "num_tokens": 12357020.0,
      "step": 6700
    },
    {
      "entropy": 5.77509069442749,
      "epoch": 0.5633270321361059,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004974113963092376,
      "loss": 5.5756,
      "mean_token_accuracy": 0.16422966718673707,
      "num_tokens": 12366108.0,
      "step": 6705
    },
    {
      "entropy": 5.840948581695557,
      "epoch": 0.5637471119512707,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004974068648360995,
      "loss": 5.4841,
      "mean_token_accuracy": 0.1757053405046463,
      "num_tokens": 12374508.0,
      "step": 6710
    },
    {
      "entropy": 5.755139255523682,
      "epoch": 0.5641671917664356,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004974023294231383,
      "loss": 5.5189,
      "mean_token_accuracy": 0.16722812950611116,
      "num_tokens": 12383555.0,
      "step": 6715
    },
    {
      "entropy": 5.765905952453613,
      "epoch": 0.5645872715816005,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004973977900704342,
      "loss": 5.6112,
      "mean_token_accuracy": 0.15942420959472656,
      "num_tokens": 12392680.0,
      "step": 6720
    },
    {
      "entropy": 5.813912057876587,
      "epoch": 0.5650073513967654,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004973932467780679,
      "loss": 5.6505,
      "mean_token_accuracy": 0.16003197729587554,
      "num_tokens": 12401881.0,
      "step": 6725
    },
    {
      "entropy": 5.828183031082153,
      "epoch": 0.5654274312119303,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004973886995461197,
      "loss": 5.6395,
      "mean_token_accuracy": 0.15504083037376404,
      "num_tokens": 12411487.0,
      "step": 6730
    },
    {
      "entropy": 5.759853315353394,
      "epoch": 0.5658475110270952,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004973841483746703,
      "loss": 5.4293,
      "mean_token_accuracy": 0.17268578708171844,
      "num_tokens": 12420376.0,
      "step": 6735
    },
    {
      "entropy": 5.671838998794556,
      "epoch": 0.5662675908422601,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004973795932638001,
      "loss": 5.4884,
      "mean_token_accuracy": 0.17086241841316224,
      "num_tokens": 12429518.0,
      "step": 6740
    },
    {
      "entropy": 5.758874130249024,
      "epoch": 0.5666876706574249,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00049737503421359,
      "loss": 5.4741,
      "mean_token_accuracy": 0.16709426939487457,
      "num_tokens": 12438952.0,
      "step": 6745
    },
    {
      "entropy": 5.718228816986084,
      "epoch": 0.5671077504725898,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004973704712241206,
      "loss": 5.4791,
      "mean_token_accuracy": 0.16316720843315125,
      "num_tokens": 12448576.0,
      "step": 6750
    },
    {
      "entropy": 5.711690616607666,
      "epoch": 0.5675278302877547,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004973659042954729,
      "loss": 5.5202,
      "mean_token_accuracy": 0.16409691125154496,
      "num_tokens": 12458166.0,
      "step": 6755
    },
    {
      "entropy": 5.662349796295166,
      "epoch": 0.5679479101029196,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004973613334277277,
      "loss": 5.4523,
      "mean_token_accuracy": 0.17179197520017625,
      "num_tokens": 12467271.0,
      "step": 6760
    },
    {
      "entropy": 5.774337673187256,
      "epoch": 0.5683679899180845,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004973567586209658,
      "loss": 5.616,
      "mean_token_accuracy": 0.15810115188360213,
      "num_tokens": 12476255.0,
      "step": 6765
    },
    {
      "entropy": 5.776514148712158,
      "epoch": 0.5687880697332494,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004973521798752686,
      "loss": 5.582,
      "mean_token_accuracy": 0.16264407485723495,
      "num_tokens": 12485096.0,
      "step": 6770
    },
    {
      "entropy": 5.905115127563477,
      "epoch": 0.5692081495484141,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000497347597190717,
      "loss": 5.5998,
      "mean_token_accuracy": 0.16689000129699708,
      "num_tokens": 12494405.0,
      "step": 6775
    },
    {
      "entropy": 5.762172269821167,
      "epoch": 0.569628229363579,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004973430105673921,
      "loss": 5.5338,
      "mean_token_accuracy": 0.16077927649021148,
      "num_tokens": 12503349.0,
      "step": 6780
    },
    {
      "entropy": 5.765121984481811,
      "epoch": 0.5700483091787439,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004973384200053754,
      "loss": 5.6144,
      "mean_token_accuracy": 0.16267796456813813,
      "num_tokens": 12513122.0,
      "step": 6785
    },
    {
      "entropy": 5.78830852508545,
      "epoch": 0.5704683889939088,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000497333825504748,
      "loss": 5.5488,
      "mean_token_accuracy": 0.16491103917360306,
      "num_tokens": 12523614.0,
      "step": 6790
    },
    {
      "entropy": 5.805237054824829,
      "epoch": 0.5708884688090737,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004973292270655914,
      "loss": 5.5971,
      "mean_token_accuracy": 0.1578866049647331,
      "num_tokens": 12532031.0,
      "step": 6795
    },
    {
      "entropy": 5.813123607635498,
      "epoch": 0.5713085486242386,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000497324624687987,
      "loss": 5.677,
      "mean_token_accuracy": 0.1546449840068817,
      "num_tokens": 12542239.0,
      "step": 6800
    },
    {
      "entropy": 5.870275640487671,
      "epoch": 0.5717286284394035,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004973200183720164,
      "loss": 5.5922,
      "mean_token_accuracy": 0.15739912986755372,
      "num_tokens": 12552608.0,
      "step": 6805
    },
    {
      "entropy": 5.71568284034729,
      "epoch": 0.5721487082545683,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004973154081177611,
      "loss": 5.4416,
      "mean_token_accuracy": 0.16019771695137025,
      "num_tokens": 12562020.0,
      "step": 6810
    },
    {
      "entropy": 5.750686836242676,
      "epoch": 0.5725687880697332,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004973107939253027,
      "loss": 5.5208,
      "mean_token_accuracy": 0.17395547032356262,
      "num_tokens": 12570519.0,
      "step": 6815
    },
    {
      "entropy": 5.701910352706909,
      "epoch": 0.5729888678848981,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004973061757947233,
      "loss": 5.5173,
      "mean_token_accuracy": 0.16468419283628463,
      "num_tokens": 12579324.0,
      "step": 6820
    },
    {
      "entropy": 5.725820779800415,
      "epoch": 0.573408947700063,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004973015537261043,
      "loss": 5.5783,
      "mean_token_accuracy": 0.16338938027620314,
      "num_tokens": 12588014.0,
      "step": 6825
    },
    {
      "entropy": 5.828612613677978,
      "epoch": 0.5738290275152279,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004972969277195279,
      "loss": 5.5811,
      "mean_token_accuracy": 0.16146877110004426,
      "num_tokens": 12596882.0,
      "step": 6830
    },
    {
      "entropy": 5.755074739456177,
      "epoch": 0.5742491073303928,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004972922977750757,
      "loss": 5.4987,
      "mean_token_accuracy": 0.16406602859497071,
      "num_tokens": 12606069.0,
      "step": 6835
    },
    {
      "entropy": 5.772974300384521,
      "epoch": 0.5746691871455577,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00049728766389283,
      "loss": 5.5212,
      "mean_token_accuracy": 0.15817257314920424,
      "num_tokens": 12615167.0,
      "step": 6840
    },
    {
      "entropy": 5.709498500823974,
      "epoch": 0.5750892669607225,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004972830260728729,
      "loss": 5.5682,
      "mean_token_accuracy": 0.1608594685792923,
      "num_tokens": 12624230.0,
      "step": 6845
    },
    {
      "entropy": 5.799942398071289,
      "epoch": 0.5755093467758874,
      "grad_norm": 1.0,
      "learning_rate": 0.0004972783843152863,
      "loss": 5.5534,
      "mean_token_accuracy": 0.16636108458042145,
      "num_tokens": 12633158.0,
      "step": 6850
    },
    {
      "entropy": 5.7593261241912845,
      "epoch": 0.5759294265910523,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004972737386201527,
      "loss": 5.5088,
      "mean_token_accuracy": 0.15915744751691818,
      "num_tokens": 12641465.0,
      "step": 6855
    },
    {
      "entropy": 5.718925714492798,
      "epoch": 0.5763495064062172,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004972690889875541,
      "loss": 5.463,
      "mean_token_accuracy": 0.16929690539836884,
      "num_tokens": 12650437.0,
      "step": 6860
    },
    {
      "entropy": 5.901365756988525,
      "epoch": 0.5767695862213821,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004972644354175732,
      "loss": 5.677,
      "mean_token_accuracy": 0.1593965582549572,
      "num_tokens": 12660072.0,
      "step": 6865
    },
    {
      "entropy": 5.84463062286377,
      "epoch": 0.577189666036547,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004972597779102922,
      "loss": 5.6865,
      "mean_token_accuracy": 0.15710159689188002,
      "num_tokens": 12670405.0,
      "step": 6870
    },
    {
      "entropy": 5.71471757888794,
      "epoch": 0.5776097458517119,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004972551164657937,
      "loss": 5.5645,
      "mean_token_accuracy": 0.16354330629110336,
      "num_tokens": 12679992.0,
      "step": 6875
    },
    {
      "entropy": 5.821878576278687,
      "epoch": 0.5780298256668767,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004972504510841602,
      "loss": 5.6182,
      "mean_token_accuracy": 0.1593255043029785,
      "num_tokens": 12690289.0,
      "step": 6880
    },
    {
      "entropy": 5.8895659923553465,
      "epoch": 0.5784499054820416,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004972457817654745,
      "loss": 5.6131,
      "mean_token_accuracy": 0.15650860518217086,
      "num_tokens": 12700518.0,
      "step": 6885
    },
    {
      "entropy": 5.845453262329102,
      "epoch": 0.5788699852972065,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004972411085098191,
      "loss": 5.6617,
      "mean_token_accuracy": 0.1576239213347435,
      "num_tokens": 12710603.0,
      "step": 6890
    },
    {
      "entropy": 5.801397609710693,
      "epoch": 0.5792900651123714,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000497236431317277,
      "loss": 5.5627,
      "mean_token_accuracy": 0.16420939862728118,
      "num_tokens": 12719298.0,
      "step": 6895
    },
    {
      "entropy": 5.783726787567138,
      "epoch": 0.5797101449275363,
      "grad_norm": 1.0625,
      "learning_rate": 0.000497231750187931,
      "loss": 5.539,
      "mean_token_accuracy": 0.16443853229284286,
      "num_tokens": 12728368.0,
      "step": 6900
    },
    {
      "entropy": 5.848576545715332,
      "epoch": 0.5801302247427012,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004972270651218638,
      "loss": 5.6102,
      "mean_token_accuracy": 0.1655212238430977,
      "num_tokens": 12737898.0,
      "step": 6905
    },
    {
      "entropy": 5.8078995704650875,
      "epoch": 0.580550304557866,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004972223761191587,
      "loss": 5.5571,
      "mean_token_accuracy": 0.16010729670524598,
      "num_tokens": 12746761.0,
      "step": 6910
    },
    {
      "entropy": 5.714229488372803,
      "epoch": 0.5809703843730308,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004972176831798986,
      "loss": 5.4897,
      "mean_token_accuracy": 0.17168698459863663,
      "num_tokens": 12755128.0,
      "step": 6915
    },
    {
      "entropy": 5.799271202087402,
      "epoch": 0.5813904641881957,
      "grad_norm": 0.875,
      "learning_rate": 0.0004972129863041667,
      "loss": 5.6659,
      "mean_token_accuracy": 0.157497800886631,
      "num_tokens": 12764727.0,
      "step": 6920
    },
    {
      "entropy": 5.808268928527832,
      "epoch": 0.5818105440033606,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004972082854920462,
      "loss": 5.5235,
      "mean_token_accuracy": 0.16768158674240113,
      "num_tokens": 12773557.0,
      "step": 6925
    },
    {
      "entropy": 5.7645995140075685,
      "epoch": 0.5822306238185255,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004972035807436203,
      "loss": 5.5202,
      "mean_token_accuracy": 0.16713482439517974,
      "num_tokens": 12782525.0,
      "step": 6930
    },
    {
      "entropy": 5.844976472854614,
      "epoch": 0.5826507036336904,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004971988720589723,
      "loss": 5.62,
      "mean_token_accuracy": 0.16590498089790345,
      "num_tokens": 12791534.0,
      "step": 6935
    },
    {
      "entropy": 5.817744731903076,
      "epoch": 0.5830707834488553,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004971941594381858,
      "loss": 5.4978,
      "mean_token_accuracy": 0.16802848428487777,
      "num_tokens": 12800662.0,
      "step": 6940
    },
    {
      "entropy": 5.835689687728882,
      "epoch": 0.5834908632640201,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004971894428813441,
      "loss": 5.5611,
      "mean_token_accuracy": 0.16315991580486297,
      "num_tokens": 12809440.0,
      "step": 6945
    },
    {
      "entropy": 5.788361358642578,
      "epoch": 0.583910943079185,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000497184722388531,
      "loss": 5.6248,
      "mean_token_accuracy": 0.15915456414222717,
      "num_tokens": 12818560.0,
      "step": 6950
    },
    {
      "entropy": 5.8209641456604,
      "epoch": 0.5843310228943499,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004971799979598297,
      "loss": 5.5612,
      "mean_token_accuracy": 0.16092651188373566,
      "num_tokens": 12827898.0,
      "step": 6955
    },
    {
      "entropy": 5.7494245052337645,
      "epoch": 0.5847511027095148,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004971752695953243,
      "loss": 5.5193,
      "mean_token_accuracy": 0.16733047217130662,
      "num_tokens": 12837199.0,
      "step": 6960
    },
    {
      "entropy": 5.748213577270508,
      "epoch": 0.5851711825246797,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004971705372950984,
      "loss": 5.5239,
      "mean_token_accuracy": 0.16277478337287904,
      "num_tokens": 12846493.0,
      "step": 6965
    },
    {
      "entropy": 5.800312566757202,
      "epoch": 0.5855912623398446,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004971658010592358,
      "loss": 5.5579,
      "mean_token_accuracy": 0.1590779483318329,
      "num_tokens": 12855026.0,
      "step": 6970
    },
    {
      "entropy": 5.753565406799316,
      "epoch": 0.5860113421550095,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004971610608878205,
      "loss": 5.6207,
      "mean_token_accuracy": 0.16079339683055877,
      "num_tokens": 12864563.0,
      "step": 6975
    },
    {
      "entropy": 5.871353530883789,
      "epoch": 0.5864314219701743,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004971563167809363,
      "loss": 5.5577,
      "mean_token_accuracy": 0.16129229664802552,
      "num_tokens": 12874358.0,
      "step": 6980
    },
    {
      "entropy": 5.771938753128052,
      "epoch": 0.5868515017853392,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004971515687386674,
      "loss": 5.5778,
      "mean_token_accuracy": 0.1612066924571991,
      "num_tokens": 12883110.0,
      "step": 6985
    },
    {
      "entropy": 5.791098928451538,
      "epoch": 0.5872715816005041,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004971468167610978,
      "loss": 5.621,
      "mean_token_accuracy": 0.16476011723279954,
      "num_tokens": 12892977.0,
      "step": 6990
    },
    {
      "entropy": 5.769950342178345,
      "epoch": 0.587691661415669,
      "grad_norm": 0.875,
      "learning_rate": 0.0004971420608483117,
      "loss": 5.4432,
      "mean_token_accuracy": 0.17349021434783934,
      "num_tokens": 12902327.0,
      "step": 6995
    },
    {
      "entropy": 5.676387453079224,
      "epoch": 0.5881117412308339,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004971373010003936,
      "loss": 5.4521,
      "mean_token_accuracy": 0.17477503120899202,
      "num_tokens": 12911957.0,
      "step": 7000
    },
    {
      "entropy": 5.7588342189788815,
      "epoch": 0.5885318210459988,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004971325372174274,
      "loss": 5.5521,
      "mean_token_accuracy": 0.15923598557710647,
      "num_tokens": 12920380.0,
      "step": 7005
    },
    {
      "entropy": 5.770251941680908,
      "epoch": 0.5889519008611637,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004971277694994976,
      "loss": 5.6139,
      "mean_token_accuracy": 0.16216087639331817,
      "num_tokens": 12929670.0,
      "step": 7010
    },
    {
      "entropy": 5.7871832847595215,
      "epoch": 0.5893719806763285,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000497122997846689,
      "loss": 5.5289,
      "mean_token_accuracy": 0.16881115585565568,
      "num_tokens": 12938185.0,
      "step": 7015
    },
    {
      "entropy": 5.7700306415557865,
      "epoch": 0.5897920604914934,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004971182222590857,
      "loss": 5.5391,
      "mean_token_accuracy": 0.1693275436758995,
      "num_tokens": 12947706.0,
      "step": 7020
    },
    {
      "entropy": 5.708819961547851,
      "epoch": 0.5902121403066583,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004971134427367725,
      "loss": 5.5322,
      "mean_token_accuracy": 0.1619589224457741,
      "num_tokens": 12957393.0,
      "step": 7025
    },
    {
      "entropy": 5.755391454696655,
      "epoch": 0.5906322201218231,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000497108659279834,
      "loss": 5.4255,
      "mean_token_accuracy": 0.16974055171012878,
      "num_tokens": 12967165.0,
      "step": 7030
    },
    {
      "entropy": 5.824974870681762,
      "epoch": 0.591052299936988,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004971038718883551,
      "loss": 5.5698,
      "mean_token_accuracy": 0.15800926685333253,
      "num_tokens": 12976490.0,
      "step": 7035
    },
    {
      "entropy": 5.798360872268677,
      "epoch": 0.5914723797521529,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004970990805624203,
      "loss": 5.5616,
      "mean_token_accuracy": 0.1576763302087784,
      "num_tokens": 12985423.0,
      "step": 7040
    },
    {
      "entropy": 5.68902554512024,
      "epoch": 0.5918924595673178,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004970942853021147,
      "loss": 5.4549,
      "mean_token_accuracy": 0.17013241201639176,
      "num_tokens": 12994510.0,
      "step": 7045
    },
    {
      "entropy": 5.772824716567993,
      "epoch": 0.5923125393824826,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004970894861075232,
      "loss": 5.5688,
      "mean_token_accuracy": 0.16687131375074388,
      "num_tokens": 13003383.0,
      "step": 7050
    },
    {
      "entropy": 5.77887659072876,
      "epoch": 0.5927326191976475,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004970846829787309,
      "loss": 5.5206,
      "mean_token_accuracy": 0.16513603180646896,
      "num_tokens": 13012550.0,
      "step": 7055
    },
    {
      "entropy": 5.800196886062622,
      "epoch": 0.5931526990128124,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004970798759158227,
      "loss": 5.5808,
      "mean_token_accuracy": 0.1571718215942383,
      "num_tokens": 13022066.0,
      "step": 7060
    },
    {
      "entropy": 5.748064422607422,
      "epoch": 0.5935727788279773,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004970750649188839,
      "loss": 5.5557,
      "mean_token_accuracy": 0.1705468252301216,
      "num_tokens": 13031008.0,
      "step": 7065
    },
    {
      "entropy": 5.7538933753967285,
      "epoch": 0.5939928586431422,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004970702499879998,
      "loss": 5.5271,
      "mean_token_accuracy": 0.1642756849527359,
      "num_tokens": 13040366.0,
      "step": 7070
    },
    {
      "entropy": 5.732847690582275,
      "epoch": 0.5944129384583071,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004970654311232554,
      "loss": 5.553,
      "mean_token_accuracy": 0.16143419444561005,
      "num_tokens": 13051140.0,
      "step": 7075
    },
    {
      "entropy": 5.755773305892944,
      "epoch": 0.594833018273472,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004970606083247362,
      "loss": 5.4933,
      "mean_token_accuracy": 0.16373573541641234,
      "num_tokens": 13059835.0,
      "step": 7080
    },
    {
      "entropy": 5.705280590057373,
      "epoch": 0.5952530980886368,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004970557815925278,
      "loss": 5.4334,
      "mean_token_accuracy": 0.16496694684028626,
      "num_tokens": 13068909.0,
      "step": 7085
    },
    {
      "entropy": 5.699723863601685,
      "epoch": 0.5956731779038017,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004970509509267155,
      "loss": 5.5233,
      "mean_token_accuracy": 0.16164302676916123,
      "num_tokens": 13078380.0,
      "step": 7090
    },
    {
      "entropy": 5.80332989692688,
      "epoch": 0.5960932577189666,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004970461163273849,
      "loss": 5.5633,
      "mean_token_accuracy": 0.15941598862409592,
      "num_tokens": 13087774.0,
      "step": 7095
    },
    {
      "entropy": 5.708866786956787,
      "epoch": 0.5965133375341315,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004970412777946219,
      "loss": 5.3747,
      "mean_token_accuracy": 0.17025242894887924,
      "num_tokens": 13095938.0,
      "step": 7100
    },
    {
      "entropy": 5.662532758712769,
      "epoch": 0.5969334173492964,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004970364353285117,
      "loss": 5.5225,
      "mean_token_accuracy": 0.16549988687038422,
      "num_tokens": 13104661.0,
      "step": 7105
    },
    {
      "entropy": 5.797566175460815,
      "epoch": 0.5973534971644613,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004970315889291405,
      "loss": 5.5197,
      "mean_token_accuracy": 0.1630711078643799,
      "num_tokens": 13114505.0,
      "step": 7110
    },
    {
      "entropy": 5.672900009155273,
      "epoch": 0.5977735769796261,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004970267385965941,
      "loss": 5.4601,
      "mean_token_accuracy": 0.1651099517941475,
      "num_tokens": 13124590.0,
      "step": 7115
    },
    {
      "entropy": 5.7219311714172365,
      "epoch": 0.598193656794791,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004970218843309583,
      "loss": 5.4543,
      "mean_token_accuracy": 0.16958350390195848,
      "num_tokens": 13134026.0,
      "step": 7120
    },
    {
      "entropy": 5.831587266921997,
      "epoch": 0.5986137366099559,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004970170261323192,
      "loss": 5.6046,
      "mean_token_accuracy": 0.16518358737230301,
      "num_tokens": 13142654.0,
      "step": 7125
    },
    {
      "entropy": 5.683307886123657,
      "epoch": 0.5990338164251208,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004970121640007627,
      "loss": 5.5414,
      "mean_token_accuracy": 0.16407201141119004,
      "num_tokens": 13151177.0,
      "step": 7130
    },
    {
      "entropy": 5.782469606399536,
      "epoch": 0.5994538962402857,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004970072979363751,
      "loss": 5.516,
      "mean_token_accuracy": 0.1628323569893837,
      "num_tokens": 13159689.0,
      "step": 7135
    },
    {
      "entropy": 5.748354053497314,
      "epoch": 0.5998739760554506,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004970024279392425,
      "loss": 5.5585,
      "mean_token_accuracy": 0.15765938758850098,
      "num_tokens": 13168601.0,
      "step": 7140
    },
    {
      "entropy": 5.752061033248902,
      "epoch": 0.6002940558706155,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004969975540094513,
      "loss": 5.5431,
      "mean_token_accuracy": 0.16534297466278075,
      "num_tokens": 13177035.0,
      "step": 7145
    },
    {
      "entropy": 5.790233945846557,
      "epoch": 0.6007141356857802,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004969926761470876,
      "loss": 5.5017,
      "mean_token_accuracy": 0.17208170592784883,
      "num_tokens": 13185444.0,
      "step": 7150
    },
    {
      "entropy": 5.72742395401001,
      "epoch": 0.6011342155009451,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000496987794352238,
      "loss": 5.4877,
      "mean_token_accuracy": 0.17015279084444046,
      "num_tokens": 13194987.0,
      "step": 7155
    },
    {
      "entropy": 5.682320976257325,
      "epoch": 0.60155429531611,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004969829086249889,
      "loss": 5.5274,
      "mean_token_accuracy": 0.16477164924144744,
      "num_tokens": 13203807.0,
      "step": 7160
    },
    {
      "entropy": 5.81383113861084,
      "epoch": 0.6019743751312749,
      "grad_norm": 1.0,
      "learning_rate": 0.000496978018965427,
      "loss": 5.6255,
      "mean_token_accuracy": 0.1601822167634964,
      "num_tokens": 13214362.0,
      "step": 7165
    },
    {
      "entropy": 5.842028665542602,
      "epoch": 0.6023944549464398,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004969731253736387,
      "loss": 5.6288,
      "mean_token_accuracy": 0.15746766030788423,
      "num_tokens": 13224192.0,
      "step": 7170
    },
    {
      "entropy": 5.685931777954101,
      "epoch": 0.6028145347616047,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004969682278497109,
      "loss": 5.5813,
      "mean_token_accuracy": 0.16571438312530518,
      "num_tokens": 13234430.0,
      "step": 7175
    },
    {
      "entropy": 5.704985523223877,
      "epoch": 0.6032346145767696,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004969633263937301,
      "loss": 5.484,
      "mean_token_accuracy": 0.16353561729192734,
      "num_tokens": 13243681.0,
      "step": 7180
    },
    {
      "entropy": 5.922461128234863,
      "epoch": 0.6036546943919344,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004969584210057832,
      "loss": 5.7514,
      "mean_token_accuracy": 0.15289916098117828,
      "num_tokens": 13254334.0,
      "step": 7185
    },
    {
      "entropy": 5.8753338813781735,
      "epoch": 0.6040747742070993,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004969535116859573,
      "loss": 5.5647,
      "mean_token_accuracy": 0.16856275945901872,
      "num_tokens": 13263781.0,
      "step": 7190
    },
    {
      "entropy": 5.686918592453003,
      "epoch": 0.6044948540222642,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004969485984343392,
      "loss": 5.4807,
      "mean_token_accuracy": 0.1674156203866005,
      "num_tokens": 13272831.0,
      "step": 7195
    },
    {
      "entropy": 5.815481042861938,
      "epoch": 0.6049149338374291,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000496943681251016,
      "loss": 5.5295,
      "mean_token_accuracy": 0.16091856956481934,
      "num_tokens": 13281621.0,
      "step": 7200
    },
    {
      "entropy": 5.706494903564453,
      "epoch": 0.605335013652594,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004969387601360747,
      "loss": 5.5144,
      "mean_token_accuracy": 0.1580105721950531,
      "num_tokens": 13291021.0,
      "step": 7205
    },
    {
      "entropy": 5.756741666793824,
      "epoch": 0.6057550934677589,
      "grad_norm": 1.0,
      "learning_rate": 0.0004969338350896026,
      "loss": 5.5374,
      "mean_token_accuracy": 0.1676796019077301,
      "num_tokens": 13299752.0,
      "step": 7210
    },
    {
      "entropy": 5.803439140319824,
      "epoch": 0.6061751732829238,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004969289061116869,
      "loss": 5.558,
      "mean_token_accuracy": 0.1542640320956707,
      "num_tokens": 13309112.0,
      "step": 7215
    },
    {
      "entropy": 5.79882845878601,
      "epoch": 0.6065952530980886,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004969239732024148,
      "loss": 5.5461,
      "mean_token_accuracy": 0.16634639352560043,
      "num_tokens": 13318328.0,
      "step": 7220
    },
    {
      "entropy": 5.639520168304443,
      "epoch": 0.6070153329132535,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004969190363618739,
      "loss": 5.4522,
      "mean_token_accuracy": 0.16655466556549073,
      "num_tokens": 13328940.0,
      "step": 7225
    },
    {
      "entropy": 5.732606840133667,
      "epoch": 0.6074354127284184,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004969140955901516,
      "loss": 5.4752,
      "mean_token_accuracy": 0.16986453533172607,
      "num_tokens": 13337829.0,
      "step": 7230
    },
    {
      "entropy": 5.858270359039307,
      "epoch": 0.6078554925435833,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004969091508873352,
      "loss": 5.6344,
      "mean_token_accuracy": 0.1569068744778633,
      "num_tokens": 13348289.0,
      "step": 7235
    },
    {
      "entropy": 5.757472801208496,
      "epoch": 0.6082755723587482,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004969042022535126,
      "loss": 5.56,
      "mean_token_accuracy": 0.16294955611228942,
      "num_tokens": 13357292.0,
      "step": 7240
    },
    {
      "entropy": 5.799214792251587,
      "epoch": 0.6086956521739131,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004968992496887713,
      "loss": 5.5965,
      "mean_token_accuracy": 0.16106789112091063,
      "num_tokens": 13366640.0,
      "step": 7245
    },
    {
      "entropy": 5.769832897186279,
      "epoch": 0.609115731989078,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004968942931931989,
      "loss": 5.5126,
      "mean_token_accuracy": 0.1701967105269432,
      "num_tokens": 13377509.0,
      "step": 7250
    },
    {
      "entropy": 5.774058771133423,
      "epoch": 0.6095358118042428,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004968893327668835,
      "loss": 5.6077,
      "mean_token_accuracy": 0.15826145708560943,
      "num_tokens": 13386573.0,
      "step": 7255
    },
    {
      "entropy": 5.701208162307739,
      "epoch": 0.6099558916194077,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004968843684099128,
      "loss": 5.4581,
      "mean_token_accuracy": 0.17081232666969298,
      "num_tokens": 13395790.0,
      "step": 7260
    },
    {
      "entropy": 5.741069650650024,
      "epoch": 0.6103759714345726,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004968794001223747,
      "loss": 5.5052,
      "mean_token_accuracy": 0.16164683550596237,
      "num_tokens": 13405265.0,
      "step": 7265
    },
    {
      "entropy": 5.760489416122437,
      "epoch": 0.6107960512497375,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004968744279043574,
      "loss": 5.4925,
      "mean_token_accuracy": 0.17164491713047028,
      "num_tokens": 13413796.0,
      "step": 7270
    },
    {
      "entropy": 5.7925914287567135,
      "epoch": 0.6112161310649024,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004968694517559488,
      "loss": 5.559,
      "mean_token_accuracy": 0.16426575034856797,
      "num_tokens": 13423299.0,
      "step": 7275
    },
    {
      "entropy": 5.750715351104736,
      "epoch": 0.6116362108800673,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004968644716772371,
      "loss": 5.4781,
      "mean_token_accuracy": 0.16807011514902115,
      "num_tokens": 13432267.0,
      "step": 7280
    },
    {
      "entropy": 5.704412889480591,
      "epoch": 0.612056290695232,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004968594876683105,
      "loss": 5.568,
      "mean_token_accuracy": 0.1630061075091362,
      "num_tokens": 13442332.0,
      "step": 7285
    },
    {
      "entropy": 5.732030010223388,
      "epoch": 0.6124763705103969,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004968544997292572,
      "loss": 5.5136,
      "mean_token_accuracy": 0.16694193333387375,
      "num_tokens": 13451700.0,
      "step": 7290
    },
    {
      "entropy": 5.783506631851196,
      "epoch": 0.6128964503255618,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004968495078601659,
      "loss": 5.6165,
      "mean_token_accuracy": 0.15932738035917282,
      "num_tokens": 13461009.0,
      "step": 7295
    },
    {
      "entropy": 5.788569831848145,
      "epoch": 0.6133165301407267,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004968445120611247,
      "loss": 5.6097,
      "mean_token_accuracy": 0.16252483278512955,
      "num_tokens": 13470341.0,
      "step": 7300
    },
    {
      "entropy": 5.810060262680054,
      "epoch": 0.6137366099558916,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004968395123322223,
      "loss": 5.52,
      "mean_token_accuracy": 0.16331708431243896,
      "num_tokens": 13479898.0,
      "step": 7305
    },
    {
      "entropy": 5.741236209869385,
      "epoch": 0.6141566897710565,
      "grad_norm": 0.953125,
      "learning_rate": 0.000496834508673547,
      "loss": 5.4506,
      "mean_token_accuracy": 0.16923084259033203,
      "num_tokens": 13488116.0,
      "step": 7310
    },
    {
      "entropy": 5.722527027130127,
      "epoch": 0.6145767695862214,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004968295010851877,
      "loss": 5.4824,
      "mean_token_accuracy": 0.17146271616220474,
      "num_tokens": 13497814.0,
      "step": 7315
    },
    {
      "entropy": 5.742617845535278,
      "epoch": 0.6149968494013862,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004968244895672331,
      "loss": 5.485,
      "mean_token_accuracy": 0.1658101573586464,
      "num_tokens": 13506617.0,
      "step": 7320
    },
    {
      "entropy": 5.739045333862305,
      "epoch": 0.6154169292165511,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004968194741197718,
      "loss": 5.6458,
      "mean_token_accuracy": 0.1604766994714737,
      "num_tokens": 13516632.0,
      "step": 7325
    },
    {
      "entropy": 5.85946102142334,
      "epoch": 0.615837009031716,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004968144547428927,
      "loss": 5.5773,
      "mean_token_accuracy": 0.16585769951343537,
      "num_tokens": 13526452.0,
      "step": 7330
    },
    {
      "entropy": 5.794593858718872,
      "epoch": 0.6162570888468809,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004968094314366848,
      "loss": 5.4818,
      "mean_token_accuracy": 0.1677970305085182,
      "num_tokens": 13535663.0,
      "step": 7335
    },
    {
      "entropy": 5.676696920394898,
      "epoch": 0.6166771686620458,
      "grad_norm": 0.9375,
      "learning_rate": 0.000496804404201237,
      "loss": 5.406,
      "mean_token_accuracy": 0.17962536066770554,
      "num_tokens": 13544574.0,
      "step": 7340
    },
    {
      "entropy": 5.855955123901367,
      "epoch": 0.6170972484772107,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004967993730366385,
      "loss": 5.5846,
      "mean_token_accuracy": 0.1650419920682907,
      "num_tokens": 13553041.0,
      "step": 7345
    },
    {
      "entropy": 5.75057430267334,
      "epoch": 0.6175173282923756,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004967943379429781,
      "loss": 5.5358,
      "mean_token_accuracy": 0.16116965860128402,
      "num_tokens": 13562108.0,
      "step": 7350
    },
    {
      "entropy": 5.887640190124512,
      "epoch": 0.6179374081075404,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004967892989203454,
      "loss": 5.6873,
      "mean_token_accuracy": 0.15722633749246598,
      "num_tokens": 13571500.0,
      "step": 7355
    },
    {
      "entropy": 5.831484985351563,
      "epoch": 0.6183574879227053,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004967842559688295,
      "loss": 5.595,
      "mean_token_accuracy": 0.1585787534713745,
      "num_tokens": 13581304.0,
      "step": 7360
    },
    {
      "entropy": 5.750160121917725,
      "epoch": 0.6187775677378702,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004967792090885195,
      "loss": 5.459,
      "mean_token_accuracy": 0.16869706958532332,
      "num_tokens": 13590734.0,
      "step": 7365
    },
    {
      "entropy": 5.681668090820312,
      "epoch": 0.6191976475530351,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004967741582795052,
      "loss": 5.5397,
      "mean_token_accuracy": 0.16230125427246095,
      "num_tokens": 13600486.0,
      "step": 7370
    },
    {
      "entropy": 5.827136898040772,
      "epoch": 0.6196177273682,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004967691035418758,
      "loss": 5.561,
      "mean_token_accuracy": 0.15649450570344925,
      "num_tokens": 13610542.0,
      "step": 7375
    },
    {
      "entropy": 5.7486302852630615,
      "epoch": 0.6200378071833649,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000496764044875721,
      "loss": 5.5092,
      "mean_token_accuracy": 0.1684138521552086,
      "num_tokens": 13619431.0,
      "step": 7380
    },
    {
      "entropy": 5.700806188583374,
      "epoch": 0.6204578869985298,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004967589822811303,
      "loss": 5.5359,
      "mean_token_accuracy": 0.1648167222738266,
      "num_tokens": 13629930.0,
      "step": 7385
    },
    {
      "entropy": 5.896169185638428,
      "epoch": 0.6208779668136946,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004967539157581934,
      "loss": 5.6599,
      "mean_token_accuracy": 0.15638630613684654,
      "num_tokens": 13639439.0,
      "step": 7390
    },
    {
      "entropy": 5.835882091522217,
      "epoch": 0.6212980466288595,
      "grad_norm": 0.921875,
      "learning_rate": 0.000496748845307,
      "loss": 5.5854,
      "mean_token_accuracy": 0.16385838612914086,
      "num_tokens": 13648548.0,
      "step": 7395
    },
    {
      "entropy": 5.834676933288574,
      "epoch": 0.6217181264440244,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004967437709276401,
      "loss": 5.6395,
      "mean_token_accuracy": 0.1629529520869255,
      "num_tokens": 13657658.0,
      "step": 7400
    },
    {
      "entropy": 5.69852991104126,
      "epoch": 0.6221382062591893,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004967386926202034,
      "loss": 5.4126,
      "mean_token_accuracy": 0.17070243656635284,
      "num_tokens": 13666763.0,
      "step": 7405
    },
    {
      "entropy": 5.810741710662842,
      "epoch": 0.6225582860743542,
      "grad_norm": 0.953125,
      "learning_rate": 0.00049673361038478,
      "loss": 5.6322,
      "mean_token_accuracy": 0.15381688475608826,
      "num_tokens": 13676527.0,
      "step": 7410
    },
    {
      "entropy": 5.784683704376221,
      "epoch": 0.622978365889519,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004967285242214599,
      "loss": 5.5995,
      "mean_token_accuracy": 0.16754818111658096,
      "num_tokens": 13685404.0,
      "step": 7415
    },
    {
      "entropy": 5.741398477554322,
      "epoch": 0.6233984457046838,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000496723434130333,
      "loss": 5.4257,
      "mean_token_accuracy": 0.16745139211416243,
      "num_tokens": 13693118.0,
      "step": 7420
    },
    {
      "entropy": 5.722853708267212,
      "epoch": 0.6238185255198487,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004967183401114898,
      "loss": 5.4955,
      "mean_token_accuracy": 0.1658550038933754,
      "num_tokens": 13702015.0,
      "step": 7425
    },
    {
      "entropy": 5.740091228485108,
      "epoch": 0.6242386053350136,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004967132421650203,
      "loss": 5.5105,
      "mean_token_accuracy": 0.16067489981651306,
      "num_tokens": 13711658.0,
      "step": 7430
    },
    {
      "entropy": 5.706097173690796,
      "epoch": 0.6246586851501785,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004967081402910149,
      "loss": 5.5426,
      "mean_token_accuracy": 0.16329398304224013,
      "num_tokens": 13720718.0,
      "step": 7435
    },
    {
      "entropy": 5.775422716140747,
      "epoch": 0.6250787649653434,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000496703034489564,
      "loss": 5.3974,
      "mean_token_accuracy": 0.16986681073904036,
      "num_tokens": 13729364.0,
      "step": 7440
    },
    {
      "entropy": 5.832866621017456,
      "epoch": 0.6254988447805083,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004966979247607579,
      "loss": 5.7002,
      "mean_token_accuracy": 0.15824775993824006,
      "num_tokens": 13739436.0,
      "step": 7445
    },
    {
      "entropy": 5.813727521896363,
      "epoch": 0.6259189245956732,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004966928111046873,
      "loss": 5.6138,
      "mean_token_accuracy": 0.16856314688920976,
      "num_tokens": 13749196.0,
      "step": 7450
    },
    {
      "entropy": 5.757702589035034,
      "epoch": 0.626339004410838,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004966876935214426,
      "loss": 5.4585,
      "mean_token_accuracy": 0.16779254078865052,
      "num_tokens": 13758414.0,
      "step": 7455
    },
    {
      "entropy": 5.726636123657227,
      "epoch": 0.6267590842260029,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004966825720111147,
      "loss": 5.5029,
      "mean_token_accuracy": 0.1628955826163292,
      "num_tokens": 13767496.0,
      "step": 7460
    },
    {
      "entropy": 5.810227489471435,
      "epoch": 0.6271791640411678,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004966774465737942,
      "loss": 5.6482,
      "mean_token_accuracy": 0.16314300298690795,
      "num_tokens": 13777033.0,
      "step": 7465
    },
    {
      "entropy": 5.857483291625977,
      "epoch": 0.6275992438563327,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004966723172095717,
      "loss": 5.5925,
      "mean_token_accuracy": 0.16579401046037673,
      "num_tokens": 13786313.0,
      "step": 7470
    },
    {
      "entropy": 5.770004558563232,
      "epoch": 0.6280193236714976,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004966671839185384,
      "loss": 5.5256,
      "mean_token_accuracy": 0.17164823561906814,
      "num_tokens": 13795257.0,
      "step": 7475
    },
    {
      "entropy": 5.668675422668457,
      "epoch": 0.6284394034866625,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004966620467007851,
      "loss": 5.4329,
      "mean_token_accuracy": 0.17163733541965484,
      "num_tokens": 13804582.0,
      "step": 7480
    },
    {
      "entropy": 5.726530838012695,
      "epoch": 0.6288594833018274,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004966569055564027,
      "loss": 5.4239,
      "mean_token_accuracy": 0.16691428869962693,
      "num_tokens": 13813248.0,
      "step": 7485
    },
    {
      "entropy": 5.820185995101928,
      "epoch": 0.6292795631169922,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004966517604854823,
      "loss": 5.6962,
      "mean_token_accuracy": 0.1543672889471054,
      "num_tokens": 13823301.0,
      "step": 7490
    },
    {
      "entropy": 5.736224126815796,
      "epoch": 0.6296996429321571,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004966466114881152,
      "loss": 5.4233,
      "mean_token_accuracy": 0.17319636940956115,
      "num_tokens": 13832040.0,
      "step": 7495
    },
    {
      "entropy": 5.762654495239258,
      "epoch": 0.630119722747322,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004966414585643925,
      "loss": 5.6244,
      "mean_token_accuracy": 0.1584588885307312,
      "num_tokens": 13841874.0,
      "step": 7500
    },
    {
      "entropy": 5.746025705337525,
      "epoch": 0.6305398025624869,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004966363017144055,
      "loss": 5.438,
      "mean_token_accuracy": 0.17738869339227675,
      "num_tokens": 13850755.0,
      "step": 7505
    },
    {
      "entropy": 5.689277791976929,
      "epoch": 0.6309598823776518,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004966311409382455,
      "loss": 5.5106,
      "mean_token_accuracy": 0.16660621762275696,
      "num_tokens": 13860009.0,
      "step": 7510
    },
    {
      "entropy": 5.6767165184021,
      "epoch": 0.6313799621928167,
      "grad_norm": 1.0,
      "learning_rate": 0.0004966259762360039,
      "loss": 5.4394,
      "mean_token_accuracy": 0.1712989777326584,
      "num_tokens": 13868476.0,
      "step": 7515
    },
    {
      "entropy": 5.661019659042358,
      "epoch": 0.6318000420079816,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004966208076077723,
      "loss": 5.4443,
      "mean_token_accuracy": 0.17133119106292724,
      "num_tokens": 13877367.0,
      "step": 7520
    },
    {
      "entropy": 5.765606021881103,
      "epoch": 0.6322201218231464,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004966156350536422,
      "loss": 5.5305,
      "mean_token_accuracy": 0.15655245184898375,
      "num_tokens": 13885985.0,
      "step": 7525
    },
    {
      "entropy": 5.723706483840942,
      "epoch": 0.6326402016383113,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004966104585737054,
      "loss": 5.4402,
      "mean_token_accuracy": 0.16649931967258452,
      "num_tokens": 13895059.0,
      "step": 7530
    },
    {
      "entropy": 5.725945901870728,
      "epoch": 0.6330602814534761,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004966052781680534,
      "loss": 5.506,
      "mean_token_accuracy": 0.16364648938179016,
      "num_tokens": 13903789.0,
      "step": 7535
    },
    {
      "entropy": 5.799894857406616,
      "epoch": 0.633480361268641,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004966000938367778,
      "loss": 5.4927,
      "mean_token_accuracy": 0.1635666087269783,
      "num_tokens": 13913377.0,
      "step": 7540
    },
    {
      "entropy": 5.639234495162964,
      "epoch": 0.6339004410838059,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004965949055799708,
      "loss": 5.4375,
      "mean_token_accuracy": 0.17239714413881302,
      "num_tokens": 13922141.0,
      "step": 7545
    },
    {
      "entropy": 5.756970500946045,
      "epoch": 0.6343205208989708,
      "grad_norm": 1.0,
      "learning_rate": 0.0004965897133977241,
      "loss": 5.5016,
      "mean_token_accuracy": 0.16107288450002671,
      "num_tokens": 13930717.0,
      "step": 7550
    },
    {
      "entropy": 5.829464769363403,
      "epoch": 0.6347406007141357,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004965845172901298,
      "loss": 5.5649,
      "mean_token_accuracy": 0.16199764609336853,
      "num_tokens": 13940344.0,
      "step": 7555
    },
    {
      "entropy": 5.733773422241211,
      "epoch": 0.6351606805293005,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004965793172572798,
      "loss": 5.4274,
      "mean_token_accuracy": 0.17065520584583282,
      "num_tokens": 13948400.0,
      "step": 7560
    },
    {
      "entropy": 5.689055871963501,
      "epoch": 0.6355807603444654,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004965741132992663,
      "loss": 5.5288,
      "mean_token_accuracy": 0.16113974153995514,
      "num_tokens": 13957939.0,
      "step": 7565
    },
    {
      "entropy": 5.806535863876343,
      "epoch": 0.6360008401596303,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004965689054161814,
      "loss": 5.5104,
      "mean_token_accuracy": 0.16322817504405976,
      "num_tokens": 13966943.0,
      "step": 7570
    },
    {
      "entropy": 5.712642097473145,
      "epoch": 0.6364209199747952,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004965636936081176,
      "loss": 5.4257,
      "mean_token_accuracy": 0.16673171520233154,
      "num_tokens": 13975850.0,
      "step": 7575
    },
    {
      "entropy": 5.79060549736023,
      "epoch": 0.6368409997899601,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000496558477875167,
      "loss": 5.5007,
      "mean_token_accuracy": 0.1744964301586151,
      "num_tokens": 13985059.0,
      "step": 7580
    },
    {
      "entropy": 5.809859275817871,
      "epoch": 0.637261079605125,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000496553258217422,
      "loss": 5.5634,
      "mean_token_accuracy": 0.1576758250594139,
      "num_tokens": 13993571.0,
      "step": 7585
    },
    {
      "entropy": 5.810078382492065,
      "epoch": 0.6376811594202898,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004965480346349751,
      "loss": 5.5531,
      "mean_token_accuracy": 0.16398203521966934,
      "num_tokens": 14002326.0,
      "step": 7590
    },
    {
      "entropy": 5.877778625488281,
      "epoch": 0.6381012392354547,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000496542807127919,
      "loss": 5.685,
      "mean_token_accuracy": 0.16269021555781366,
      "num_tokens": 14012002.0,
      "step": 7595
    },
    {
      "entropy": 5.7747259616851805,
      "epoch": 0.6385213190506196,
      "grad_norm": 0.921875,
      "learning_rate": 0.000496537575696346,
      "loss": 5.5489,
      "mean_token_accuracy": 0.16207897216081618,
      "num_tokens": 14022085.0,
      "step": 7600
    },
    {
      "entropy": 5.7344865798950195,
      "epoch": 0.6389413988657845,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004965323403403488,
      "loss": 5.4429,
      "mean_token_accuracy": 0.16591739356517793,
      "num_tokens": 14030706.0,
      "step": 7605
    },
    {
      "entropy": 5.687538003921508,
      "epoch": 0.6393614786809494,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004965271010600205,
      "loss": 5.4678,
      "mean_token_accuracy": 0.1682354286313057,
      "num_tokens": 14039520.0,
      "step": 7610
    },
    {
      "entropy": 5.758507633209229,
      "epoch": 0.6397815584961143,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004965218578554535,
      "loss": 5.5768,
      "mean_token_accuracy": 0.16726803928613662,
      "num_tokens": 14048407.0,
      "step": 7615
    },
    {
      "entropy": 5.69767746925354,
      "epoch": 0.6402016383112792,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000496516610726741,
      "loss": 5.4949,
      "mean_token_accuracy": 0.1704343616962433,
      "num_tokens": 14057534.0,
      "step": 7620
    },
    {
      "entropy": 5.718931722640991,
      "epoch": 0.640621718126444,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004965113596739759,
      "loss": 5.4342,
      "mean_token_accuracy": 0.17019388526678086,
      "num_tokens": 14065992.0,
      "step": 7625
    },
    {
      "entropy": 5.666730785369873,
      "epoch": 0.6410417979416089,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004965061046972508,
      "loss": 5.4219,
      "mean_token_accuracy": 0.1642868459224701,
      "num_tokens": 14074806.0,
      "step": 7630
    },
    {
      "entropy": 5.70978798866272,
      "epoch": 0.6414618777567738,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004965008457966594,
      "loss": 5.4974,
      "mean_token_accuracy": 0.16396250426769257,
      "num_tokens": 14083813.0,
      "step": 7635
    },
    {
      "entropy": 5.73560585975647,
      "epoch": 0.6418819575719387,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004964955829722945,
      "loss": 5.4397,
      "mean_token_accuracy": 0.16967316418886186,
      "num_tokens": 14092193.0,
      "step": 7640
    },
    {
      "entropy": 5.836223840713501,
      "epoch": 0.6423020373871036,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004964903162242493,
      "loss": 5.6404,
      "mean_token_accuracy": 0.15832567512989043,
      "num_tokens": 14102797.0,
      "step": 7645
    },
    {
      "entropy": 5.709490633010864,
      "epoch": 0.6427221172022685,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004964850455526173,
      "loss": 5.5084,
      "mean_token_accuracy": 0.16814197450876237,
      "num_tokens": 14112226.0,
      "step": 7650
    },
    {
      "entropy": 5.640767669677734,
      "epoch": 0.6431421970174334,
      "grad_norm": 1.0,
      "learning_rate": 0.0004964797709574917,
      "loss": 5.4332,
      "mean_token_accuracy": 0.16021016240119934,
      "num_tokens": 14121775.0,
      "step": 7655
    },
    {
      "entropy": 5.720970726013183,
      "epoch": 0.6435622768325981,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000496474492438966,
      "loss": 5.4235,
      "mean_token_accuracy": 0.16737012714147567,
      "num_tokens": 14130415.0,
      "step": 7660
    },
    {
      "entropy": 5.7541563510894775,
      "epoch": 0.643982356647763,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004964692099971338,
      "loss": 5.4533,
      "mean_token_accuracy": 0.16918079406023026,
      "num_tokens": 14140204.0,
      "step": 7665
    },
    {
      "entropy": 5.718717336654663,
      "epoch": 0.6444024364629279,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004964639236320885,
      "loss": 5.3952,
      "mean_token_accuracy": 0.16594948023557662,
      "num_tokens": 14149595.0,
      "step": 7670
    },
    {
      "entropy": 5.639620018005371,
      "epoch": 0.6448225162780928,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004964586333439239,
      "loss": 5.4751,
      "mean_token_accuracy": 0.16469965428113936,
      "num_tokens": 14158865.0,
      "step": 7675
    },
    {
      "entropy": 5.6987968444824215,
      "epoch": 0.6452425960932577,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004964533391327335,
      "loss": 5.4479,
      "mean_token_accuracy": 0.17683874666690827,
      "num_tokens": 14167962.0,
      "step": 7680
    },
    {
      "entropy": 5.761344957351684,
      "epoch": 0.6456626759084226,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004964480409986113,
      "loss": 5.4858,
      "mean_token_accuracy": 0.1746539369225502,
      "num_tokens": 14176479.0,
      "step": 7685
    },
    {
      "entropy": 5.813274955749511,
      "epoch": 0.6460827557235875,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004964427389416512,
      "loss": 5.4973,
      "mean_token_accuracy": 0.1644265055656433,
      "num_tokens": 14185408.0,
      "step": 7690
    },
    {
      "entropy": 5.661887788772583,
      "epoch": 0.6465028355387523,
      "grad_norm": 1.046875,
      "learning_rate": 0.000496437432961947,
      "loss": 5.5021,
      "mean_token_accuracy": 0.1675272688269615,
      "num_tokens": 14194155.0,
      "step": 7695
    },
    {
      "entropy": 5.6674519062042235,
      "epoch": 0.6469229153539172,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004964321230595925,
      "loss": 5.5358,
      "mean_token_accuracy": 0.1637965828180313,
      "num_tokens": 14202779.0,
      "step": 7700
    },
    {
      "entropy": 5.854550075531006,
      "epoch": 0.6473429951690821,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004964268092346821,
      "loss": 5.7104,
      "mean_token_accuracy": 0.15778978168964386,
      "num_tokens": 14212552.0,
      "step": 7705
    },
    {
      "entropy": 5.839205121994018,
      "epoch": 0.647763074984247,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004964214914873098,
      "loss": 5.4966,
      "mean_token_accuracy": 0.16174962222576142,
      "num_tokens": 14222783.0,
      "step": 7710
    },
    {
      "entropy": 5.649949645996093,
      "epoch": 0.6481831547994119,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004964161698175697,
      "loss": 5.3966,
      "mean_token_accuracy": 0.16576698720455169,
      "num_tokens": 14232085.0,
      "step": 7715
    },
    {
      "entropy": 5.705242538452149,
      "epoch": 0.6486032346145768,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004964108442255562,
      "loss": 5.5497,
      "mean_token_accuracy": 0.16135221123695373,
      "num_tokens": 14241969.0,
      "step": 7720
    },
    {
      "entropy": 5.675537252426148,
      "epoch": 0.6490233144297417,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004964055147113637,
      "loss": 5.4516,
      "mean_token_accuracy": 0.174803626537323,
      "num_tokens": 14251012.0,
      "step": 7725
    },
    {
      "entropy": 5.804049396514893,
      "epoch": 0.6494433942449065,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004964001812750864,
      "loss": 5.5535,
      "mean_token_accuracy": 0.16239049136638642,
      "num_tokens": 14261110.0,
      "step": 7730
    },
    {
      "entropy": 5.785045480728149,
      "epoch": 0.6498634740600714,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000496394843916819,
      "loss": 5.5415,
      "mean_token_accuracy": 0.1608967199921608,
      "num_tokens": 14270869.0,
      "step": 7735
    },
    {
      "entropy": 5.727395629882812,
      "epoch": 0.6502835538752363,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004963895026366558,
      "loss": 5.5201,
      "mean_token_accuracy": 0.16277528554201126,
      "num_tokens": 14279607.0,
      "step": 7740
    },
    {
      "entropy": 5.684513425827026,
      "epoch": 0.6507036336904012,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004963841574346917,
      "loss": 5.483,
      "mean_token_accuracy": 0.163617080450058,
      "num_tokens": 14289282.0,
      "step": 7745
    },
    {
      "entropy": 5.676046323776245,
      "epoch": 0.6511237135055661,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004963788083110212,
      "loss": 5.4234,
      "mean_token_accuracy": 0.1703727886080742,
      "num_tokens": 14298658.0,
      "step": 7750
    },
    {
      "entropy": 5.807625865936279,
      "epoch": 0.651543793320731,
      "grad_norm": 0.90625,
      "learning_rate": 0.000496373455265739,
      "loss": 5.5023,
      "mean_token_accuracy": 0.16605711132287979,
      "num_tokens": 14307832.0,
      "step": 7755
    },
    {
      "entropy": 5.710317754745484,
      "epoch": 0.6519638731358958,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004963680982989402,
      "loss": 5.4191,
      "mean_token_accuracy": 0.17270244359970094,
      "num_tokens": 14317122.0,
      "step": 7760
    },
    {
      "entropy": 5.68309645652771,
      "epoch": 0.6523839529510607,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004963627374107195,
      "loss": 5.4602,
      "mean_token_accuracy": 0.17099864333868026,
      "num_tokens": 14326069.0,
      "step": 7765
    },
    {
      "entropy": 5.686089515686035,
      "epoch": 0.6528040327662256,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004963573726011717,
      "loss": 5.451,
      "mean_token_accuracy": 0.16986725032329558,
      "num_tokens": 14335260.0,
      "step": 7770
    },
    {
      "entropy": 5.811730098724365,
      "epoch": 0.6532241125813905,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004963520038703922,
      "loss": 5.5649,
      "mean_token_accuracy": 0.1554375059902668,
      "num_tokens": 14345823.0,
      "step": 7775
    },
    {
      "entropy": 5.721672868728637,
      "epoch": 0.6536441923965554,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000496346631218476,
      "loss": 5.4314,
      "mean_token_accuracy": 0.17085641324520112,
      "num_tokens": 14354316.0,
      "step": 7780
    },
    {
      "entropy": 5.678657960891724,
      "epoch": 0.6540642722117203,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000496341254645518,
      "loss": 5.4635,
      "mean_token_accuracy": 0.17192131727933885,
      "num_tokens": 14364539.0,
      "step": 7785
    },
    {
      "entropy": 5.742990207672119,
      "epoch": 0.6544843520268852,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004963358741516138,
      "loss": 5.5754,
      "mean_token_accuracy": 0.15854422003030777,
      "num_tokens": 14374081.0,
      "step": 7790
    },
    {
      "entropy": 5.760342597961426,
      "epoch": 0.6549044318420499,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004963304897368585,
      "loss": 5.4639,
      "mean_token_accuracy": 0.16030970215797424,
      "num_tokens": 14383255.0,
      "step": 7795
    },
    {
      "entropy": 5.8018804550170895,
      "epoch": 0.6553245116572148,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004963251014013475,
      "loss": 5.6113,
      "mean_token_accuracy": 0.15945181399583816,
      "num_tokens": 14392417.0,
      "step": 7800
    },
    {
      "entropy": 5.86241455078125,
      "epoch": 0.6557445914723797,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004963197091451763,
      "loss": 5.6384,
      "mean_token_accuracy": 0.15713931620121002,
      "num_tokens": 14401899.0,
      "step": 7805
    },
    {
      "entropy": 5.831144714355469,
      "epoch": 0.6561646712875446,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004963143129684405,
      "loss": 5.5971,
      "mean_token_accuracy": 0.15792884901165963,
      "num_tokens": 14411245.0,
      "step": 7810
    },
    {
      "entropy": 5.676599359512329,
      "epoch": 0.6565847511027095,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004963089128712355,
      "loss": 5.4808,
      "mean_token_accuracy": 0.16693222820758818,
      "num_tokens": 14419710.0,
      "step": 7815
    },
    {
      "entropy": 5.69451642036438,
      "epoch": 0.6570048309178744,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004963035088536571,
      "loss": 5.436,
      "mean_token_accuracy": 0.17514338344335556,
      "num_tokens": 14430266.0,
      "step": 7820
    },
    {
      "entropy": 5.740038776397705,
      "epoch": 0.6574249107330393,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004962981009158012,
      "loss": 5.4254,
      "mean_token_accuracy": 0.16121056228876113,
      "num_tokens": 14439515.0,
      "step": 7825
    },
    {
      "entropy": 5.725498914718628,
      "epoch": 0.6578449905482041,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004962926890577632,
      "loss": 5.4854,
      "mean_token_accuracy": 0.1698240667581558,
      "num_tokens": 14448091.0,
      "step": 7830
    },
    {
      "entropy": 5.7471672058105465,
      "epoch": 0.658265070363369,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000496287273279639,
      "loss": 5.5145,
      "mean_token_accuracy": 0.16248589754104614,
      "num_tokens": 14457744.0,
      "step": 7835
    },
    {
      "entropy": 5.785703563690186,
      "epoch": 0.6586851501785339,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000496281853581525,
      "loss": 5.5009,
      "mean_token_accuracy": 0.16848925054073333,
      "num_tokens": 14467597.0,
      "step": 7840
    },
    {
      "entropy": 5.740370082855224,
      "epoch": 0.6591052299936988,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004962764299635168,
      "loss": 5.4873,
      "mean_token_accuracy": 0.1677161082625389,
      "num_tokens": 14476662.0,
      "step": 7845
    },
    {
      "entropy": 5.792033338546753,
      "epoch": 0.6595253098088637,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004962710024257105,
      "loss": 5.5578,
      "mean_token_accuracy": 0.1600222632288933,
      "num_tokens": 14486583.0,
      "step": 7850
    },
    {
      "entropy": 5.7833571434021,
      "epoch": 0.6599453896240286,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004962655709682025,
      "loss": 5.5613,
      "mean_token_accuracy": 0.16222136318683625,
      "num_tokens": 14496528.0,
      "step": 7855
    },
    {
      "entropy": 5.736913347244263,
      "epoch": 0.6603654694391935,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004962601355910887,
      "loss": 5.556,
      "mean_token_accuracy": 0.1588379979133606,
      "num_tokens": 14507026.0,
      "step": 7860
    },
    {
      "entropy": 5.64929895401001,
      "epoch": 0.6607855492543583,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004962546962944656,
      "loss": 5.426,
      "mean_token_accuracy": 0.16645535379648208,
      "num_tokens": 14516480.0,
      "step": 7865
    },
    {
      "entropy": 5.687178373336792,
      "epoch": 0.6612056290695232,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004962492530784295,
      "loss": 5.3726,
      "mean_token_accuracy": 0.1767757385969162,
      "num_tokens": 14525068.0,
      "step": 7870
    },
    {
      "entropy": 5.677102279663086,
      "epoch": 0.6616257088846881,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004962438059430768,
      "loss": 5.4875,
      "mean_token_accuracy": 0.17044864147901534,
      "num_tokens": 14534441.0,
      "step": 7875
    },
    {
      "entropy": 5.80464277267456,
      "epoch": 0.662045788699853,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004962383548885039,
      "loss": 5.5817,
      "mean_token_accuracy": 0.16445917934179305,
      "num_tokens": 14543026.0,
      "step": 7880
    },
    {
      "entropy": 5.721199655532837,
      "epoch": 0.6624658685150179,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004962328999148075,
      "loss": 5.4407,
      "mean_token_accuracy": 0.17542991489171983,
      "num_tokens": 14552068.0,
      "step": 7885
    },
    {
      "entropy": 5.754334926605225,
      "epoch": 0.6628859483301828,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004962274410220842,
      "loss": 5.5836,
      "mean_token_accuracy": 0.16184947192668914,
      "num_tokens": 14561587.0,
      "step": 7890
    },
    {
      "entropy": 5.801696395874023,
      "epoch": 0.6633060281453477,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004962219782104308,
      "loss": 5.5872,
      "mean_token_accuracy": 0.16888613402843475,
      "num_tokens": 14571020.0,
      "step": 7895
    },
    {
      "entropy": 5.800302648544312,
      "epoch": 0.6637261079605125,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004962165114799439,
      "loss": 5.5389,
      "mean_token_accuracy": 0.1558610826730728,
      "num_tokens": 14580638.0,
      "step": 7900
    },
    {
      "entropy": 5.715132999420166,
      "epoch": 0.6641461877756774,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004962110408307204,
      "loss": 5.4571,
      "mean_token_accuracy": 0.16448279917240144,
      "num_tokens": 14590173.0,
      "step": 7905
    },
    {
      "entropy": 5.654195117950439,
      "epoch": 0.6645662675908423,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004962055662628571,
      "loss": 5.4531,
      "mean_token_accuracy": 0.16645844876766205,
      "num_tokens": 14598635.0,
      "step": 7910
    },
    {
      "entropy": 5.763539838790893,
      "epoch": 0.6649863474060071,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004962000877764513,
      "loss": 5.4877,
      "mean_token_accuracy": 0.17148942798376082,
      "num_tokens": 14607233.0,
      "step": 7915
    },
    {
      "entropy": 5.882718420028686,
      "epoch": 0.665406427221172,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004961946053715998,
      "loss": 5.6366,
      "mean_token_accuracy": 0.15461080521345139,
      "num_tokens": 14617483.0,
      "step": 7920
    },
    {
      "entropy": 5.721769142150879,
      "epoch": 0.665826507036337,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004961891190483997,
      "loss": 5.4446,
      "mean_token_accuracy": 0.16694557815790176,
      "num_tokens": 14625805.0,
      "step": 7925
    },
    {
      "entropy": 5.659672784805298,
      "epoch": 0.6662465868515017,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004961836288069483,
      "loss": 5.3918,
      "mean_token_accuracy": 0.1696493521332741,
      "num_tokens": 14634605.0,
      "step": 7930
    },
    {
      "entropy": 5.79138560295105,
      "epoch": 0.6666666666666666,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004961781346473428,
      "loss": 5.5955,
      "mean_token_accuracy": 0.15645502209663392,
      "num_tokens": 14644970.0,
      "step": 7935
    },
    {
      "entropy": 5.763145494461059,
      "epoch": 0.6670867464818315,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004961726365696805,
      "loss": 5.4754,
      "mean_token_accuracy": 0.16234245151281357,
      "num_tokens": 14655043.0,
      "step": 7940
    },
    {
      "entropy": 5.747127676010132,
      "epoch": 0.6675068262969964,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004961671345740589,
      "loss": 5.4556,
      "mean_token_accuracy": 0.16612063944339753,
      "num_tokens": 14663994.0,
      "step": 7945
    },
    {
      "entropy": 5.676699924468994,
      "epoch": 0.6679269061121613,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004961616286605753,
      "loss": 5.4514,
      "mean_token_accuracy": 0.16224169731140137,
      "num_tokens": 14674101.0,
      "step": 7950
    },
    {
      "entropy": 5.73637170791626,
      "epoch": 0.6683469859273262,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004961561188293273,
      "loss": 5.5209,
      "mean_token_accuracy": 0.16538771167397498,
      "num_tokens": 14684156.0,
      "step": 7955
    },
    {
      "entropy": 5.655863046646118,
      "epoch": 0.6687670657424911,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004961506050804126,
      "loss": 5.4529,
      "mean_token_accuracy": 0.17343426644802093,
      "num_tokens": 14693223.0,
      "step": 7960
    },
    {
      "entropy": 5.757249593734741,
      "epoch": 0.6691871455576559,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000496145087413929,
      "loss": 5.4526,
      "mean_token_accuracy": 0.1625904157757759,
      "num_tokens": 14702959.0,
      "step": 7965
    },
    {
      "entropy": 5.854363918304443,
      "epoch": 0.6696072253728208,
      "grad_norm": 1.0,
      "learning_rate": 0.0004961395658299737,
      "loss": 5.5793,
      "mean_token_accuracy": 0.1612630233168602,
      "num_tokens": 14712146.0,
      "step": 7970
    },
    {
      "entropy": 5.723653364181518,
      "epoch": 0.6700273051879857,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004961340403286451,
      "loss": 5.4817,
      "mean_token_accuracy": 0.15962589234113694,
      "num_tokens": 14721932.0,
      "step": 7975
    },
    {
      "entropy": 5.698150777816773,
      "epoch": 0.6704473850031506,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004961285109100408,
      "loss": 5.4143,
      "mean_token_accuracy": 0.17404233068227767,
      "num_tokens": 14731080.0,
      "step": 7980
    },
    {
      "entropy": 5.610926008224487,
      "epoch": 0.6708674648183155,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004961229775742587,
      "loss": 5.4347,
      "mean_token_accuracy": 0.17332813590765,
      "num_tokens": 14740057.0,
      "step": 7985
    },
    {
      "entropy": 5.813832473754883,
      "epoch": 0.6712875446334804,
      "grad_norm": 1.046875,
      "learning_rate": 0.000496117440321397,
      "loss": 5.5084,
      "mean_token_accuracy": 0.17006851732730865,
      "num_tokens": 14748399.0,
      "step": 7990
    },
    {
      "entropy": 5.791306734085083,
      "epoch": 0.6717076244486453,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004961118991515537,
      "loss": 5.5058,
      "mean_token_accuracy": 0.1638423651456833,
      "num_tokens": 14757215.0,
      "step": 7995
    },
    {
      "entropy": 5.686156272888184,
      "epoch": 0.6721277042638101,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000496106354064827,
      "loss": 5.5156,
      "mean_token_accuracy": 0.16654904037714005,
      "num_tokens": 14766191.0,
      "step": 8000
    },
    {
      "entropy": 5.8185874938964846,
      "epoch": 0.672547784078975,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004961008050613149,
      "loss": 5.5678,
      "mean_token_accuracy": 0.16079447120428086,
      "num_tokens": 14775220.0,
      "step": 8005
    },
    {
      "entropy": 5.785628366470337,
      "epoch": 0.6729678638941399,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004960952521411161,
      "loss": 5.5348,
      "mean_token_accuracy": 0.16216191947460173,
      "num_tokens": 14784287.0,
      "step": 8010
    },
    {
      "entropy": 5.833697271347046,
      "epoch": 0.6733879437093048,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004960896953043287,
      "loss": 5.5933,
      "mean_token_accuracy": 0.16402601152658464,
      "num_tokens": 14794219.0,
      "step": 8015
    },
    {
      "entropy": 5.7764135837554935,
      "epoch": 0.6738080235244697,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004960841345510511,
      "loss": 5.5046,
      "mean_token_accuracy": 0.16632075160741805,
      "num_tokens": 14803324.0,
      "step": 8020
    },
    {
      "entropy": 5.732311344146728,
      "epoch": 0.6742281033396346,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000496078569881382,
      "loss": 5.528,
      "mean_token_accuracy": 0.1683580070734024,
      "num_tokens": 14811963.0,
      "step": 8025
    },
    {
      "entropy": 5.725998973846435,
      "epoch": 0.6746481831547995,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004960730012954198,
      "loss": 5.4591,
      "mean_token_accuracy": 0.16402439773082733,
      "num_tokens": 14821903.0,
      "step": 8030
    },
    {
      "entropy": 5.676053524017334,
      "epoch": 0.6750682629699643,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004960674287932634,
      "loss": 5.473,
      "mean_token_accuracy": 0.16029030084609985,
      "num_tokens": 14831215.0,
      "step": 8035
    },
    {
      "entropy": 5.750786447525025,
      "epoch": 0.6754883427851291,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004960618523750111,
      "loss": 5.3928,
      "mean_token_accuracy": 0.17306858897209168,
      "num_tokens": 14840354.0,
      "step": 8040
    },
    {
      "entropy": 5.741588592529297,
      "epoch": 0.675908422600294,
      "grad_norm": 0.96875,
      "learning_rate": 0.000496056272040762,
      "loss": 5.558,
      "mean_token_accuracy": 0.16385315358638763,
      "num_tokens": 14849660.0,
      "step": 8045
    },
    {
      "entropy": 5.753055763244629,
      "epoch": 0.6763285024154589,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004960506877906149,
      "loss": 5.4605,
      "mean_token_accuracy": 0.16291872560977935,
      "num_tokens": 14859819.0,
      "step": 8050
    },
    {
      "entropy": 5.771967840194702,
      "epoch": 0.6767485822306238,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004960450996246686,
      "loss": 5.4832,
      "mean_token_accuracy": 0.16809024140238762,
      "num_tokens": 14869260.0,
      "step": 8055
    },
    {
      "entropy": 5.732260227203369,
      "epoch": 0.6771686620457887,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004960395075430222,
      "loss": 5.4522,
      "mean_token_accuracy": 0.165982024371624,
      "num_tokens": 14878685.0,
      "step": 8060
    },
    {
      "entropy": 5.693307256698608,
      "epoch": 0.6775887418609536,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004960339115457748,
      "loss": 5.4529,
      "mean_token_accuracy": 0.1676463320851326,
      "num_tokens": 14888456.0,
      "step": 8065
    },
    {
      "entropy": 5.718993425369263,
      "epoch": 0.6780088216761184,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004960283116330255,
      "loss": 5.5505,
      "mean_token_accuracy": 0.16544234454631807,
      "num_tokens": 14897401.0,
      "step": 8070
    },
    {
      "entropy": 5.767450761795044,
      "epoch": 0.6784289014912833,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004960227078048735,
      "loss": 5.4803,
      "mean_token_accuracy": 0.1697261393070221,
      "num_tokens": 14906741.0,
      "step": 8075
    },
    {
      "entropy": 5.745333003997803,
      "epoch": 0.6788489813064482,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004960171000614179,
      "loss": 5.365,
      "mean_token_accuracy": 0.17516455948352813,
      "num_tokens": 14916002.0,
      "step": 8080
    },
    {
      "entropy": 5.6103489875793455,
      "epoch": 0.6792690611216131,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004960114884027583,
      "loss": 5.3111,
      "mean_token_accuracy": 0.17905542850494385,
      "num_tokens": 14925247.0,
      "step": 8085
    },
    {
      "entropy": 5.647380781173706,
      "epoch": 0.679689140936778,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004960058728289939,
      "loss": 5.4204,
      "mean_token_accuracy": 0.16654787361621856,
      "num_tokens": 14933925.0,
      "step": 8090
    },
    {
      "entropy": 5.8051458358764645,
      "epoch": 0.6801092207519429,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004960002533402243,
      "loss": 5.4976,
      "mean_token_accuracy": 0.17039394080638887,
      "num_tokens": 14943368.0,
      "step": 8095
    },
    {
      "entropy": 5.749414873123169,
      "epoch": 0.6805293005671077,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004959946299365491,
      "loss": 5.5039,
      "mean_token_accuracy": 0.16465162187814714,
      "num_tokens": 14953710.0,
      "step": 8100
    },
    {
      "entropy": 5.780334091186523,
      "epoch": 0.6809493803822726,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004959890026180677,
      "loss": 5.5232,
      "mean_token_accuracy": 0.16230493485927583,
      "num_tokens": 14962814.0,
      "step": 8105
    },
    {
      "entropy": 5.6274302959442135,
      "epoch": 0.6813694601974375,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049598337138488,
      "loss": 5.4317,
      "mean_token_accuracy": 0.17269947230815888,
      "num_tokens": 14971631.0,
      "step": 8110
    },
    {
      "entropy": 5.702581739425659,
      "epoch": 0.6817895400126024,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004959777362370855,
      "loss": 5.4131,
      "mean_token_accuracy": 0.17140980064868927,
      "num_tokens": 14980528.0,
      "step": 8115
    },
    {
      "entropy": 5.709146404266358,
      "epoch": 0.6822096198277673,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004959720971747843,
      "loss": 5.4482,
      "mean_token_accuracy": 0.16945860683918,
      "num_tokens": 14989331.0,
      "step": 8120
    },
    {
      "entropy": 5.675239133834839,
      "epoch": 0.6826296996429322,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004959664541980762,
      "loss": 5.4418,
      "mean_token_accuracy": 0.17257071286439896,
      "num_tokens": 14999403.0,
      "step": 8125
    },
    {
      "entropy": 5.721805572509766,
      "epoch": 0.6830497794580971,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004959608073070612,
      "loss": 5.5259,
      "mean_token_accuracy": 0.16475262492895126,
      "num_tokens": 15009388.0,
      "step": 8130
    },
    {
      "entropy": 5.782740736007691,
      "epoch": 0.6834698592732619,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004959551565018392,
      "loss": 5.4411,
      "mean_token_accuracy": 0.17187750041484834,
      "num_tokens": 15018586.0,
      "step": 8135
    },
    {
      "entropy": 5.691683387756347,
      "epoch": 0.6838899390884268,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004959495017825104,
      "loss": 5.4529,
      "mean_token_accuracy": 0.17066957354545592,
      "num_tokens": 15027982.0,
      "step": 8140
    },
    {
      "entropy": 5.6663110733032225,
      "epoch": 0.6843100189035917,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004959438431491749,
      "loss": 5.4368,
      "mean_token_accuracy": 0.17165957242250443,
      "num_tokens": 15037103.0,
      "step": 8145
    },
    {
      "entropy": 5.687498998641968,
      "epoch": 0.6847300987187566,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000495938180601933,
      "loss": 5.5516,
      "mean_token_accuracy": 0.1639342337846756,
      "num_tokens": 15046739.0,
      "step": 8150
    },
    {
      "entropy": 5.783039665222168,
      "epoch": 0.6851501785339215,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004959325141408851,
      "loss": 5.4828,
      "mean_token_accuracy": 0.16961948871612548,
      "num_tokens": 15056586.0,
      "step": 8155
    },
    {
      "entropy": 5.715682792663574,
      "epoch": 0.6855702583490864,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004959268437661313,
      "loss": 5.4587,
      "mean_token_accuracy": 0.16696280092000962,
      "num_tokens": 15066622.0,
      "step": 8160
    },
    {
      "entropy": 5.737662839889526,
      "epoch": 0.6859903381642513,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004959211694777724,
      "loss": 5.4505,
      "mean_token_accuracy": 0.1721181496977806,
      "num_tokens": 15075415.0,
      "step": 8165
    },
    {
      "entropy": 5.663425445556641,
      "epoch": 0.686410417979416,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004959154912759086,
      "loss": 5.4307,
      "mean_token_accuracy": 0.16645318865776063,
      "num_tokens": 15085087.0,
      "step": 8170
    },
    {
      "entropy": 5.727186918258667,
      "epoch": 0.6868304977945809,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004959098091606406,
      "loss": 5.4409,
      "mean_token_accuracy": 0.17099099531769751,
      "num_tokens": 15093580.0,
      "step": 8175
    },
    {
      "entropy": 5.624398899078369,
      "epoch": 0.6872505776097458,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004959041231320692,
      "loss": 5.4356,
      "mean_token_accuracy": 0.1736039474606514,
      "num_tokens": 15104033.0,
      "step": 8180
    },
    {
      "entropy": 5.660421991348267,
      "epoch": 0.6876706574249107,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004958984331902951,
      "loss": 5.52,
      "mean_token_accuracy": 0.16186250150203704,
      "num_tokens": 15113164.0,
      "step": 8185
    },
    {
      "entropy": 5.733459997177124,
      "epoch": 0.6880907372400756,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004958927393354188,
      "loss": 5.4547,
      "mean_token_accuracy": 0.1694747194647789,
      "num_tokens": 15122215.0,
      "step": 8190
    },
    {
      "entropy": 5.758696889877319,
      "epoch": 0.6885108170552405,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004958870415675415,
      "loss": 5.4395,
      "mean_token_accuracy": 0.16364239901304245,
      "num_tokens": 15130877.0,
      "step": 8195
    },
    {
      "entropy": 5.675120258331299,
      "epoch": 0.6889308968704054,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004958813398867639,
      "loss": 5.4189,
      "mean_token_accuracy": 0.1747715577483177,
      "num_tokens": 15140227.0,
      "step": 8200
    },
    {
      "entropy": 5.826716184616089,
      "epoch": 0.6893509766855702,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004958756342931872,
      "loss": 5.5949,
      "mean_token_accuracy": 0.15757112503051757,
      "num_tokens": 15150006.0,
      "step": 8205
    },
    {
      "entropy": 5.753350257873535,
      "epoch": 0.6897710565007351,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004958699247869122,
      "loss": 5.4716,
      "mean_token_accuracy": 0.16771091967821122,
      "num_tokens": 15160032.0,
      "step": 8210
    },
    {
      "entropy": 5.696915531158448,
      "epoch": 0.6901911363159,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004958642113680404,
      "loss": 5.4281,
      "mean_token_accuracy": 0.17119741290807725,
      "num_tokens": 15168966.0,
      "step": 8215
    },
    {
      "entropy": 5.8364509582519535,
      "epoch": 0.6906112161310649,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004958584940366727,
      "loss": 5.5947,
      "mean_token_accuracy": 0.16665195673704147,
      "num_tokens": 15179337.0,
      "step": 8220
    },
    {
      "entropy": 5.796460723876953,
      "epoch": 0.6910312959462298,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004958527727929106,
      "loss": 5.5162,
      "mean_token_accuracy": 0.16378459185361863,
      "num_tokens": 15188395.0,
      "step": 8225
    },
    {
      "entropy": 5.732974672317505,
      "epoch": 0.6914513757613947,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004958470476368552,
      "loss": 5.4369,
      "mean_token_accuracy": 0.17375798597931863,
      "num_tokens": 15198669.0,
      "step": 8230
    },
    {
      "entropy": 5.716420412063599,
      "epoch": 0.6918714555765595,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004958413185686082,
      "loss": 5.4581,
      "mean_token_accuracy": 0.16865925788879393,
      "num_tokens": 15207371.0,
      "step": 8235
    },
    {
      "entropy": 5.751028299331665,
      "epoch": 0.6922915353917244,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004958355855882709,
      "loss": 5.4921,
      "mean_token_accuracy": 0.16884441524744034,
      "num_tokens": 15215694.0,
      "step": 8240
    },
    {
      "entropy": 5.747837352752685,
      "epoch": 0.6927116152068893,
      "grad_norm": 0.921875,
      "learning_rate": 0.000495829848695945,
      "loss": 5.4679,
      "mean_token_accuracy": 0.1624392107129097,
      "num_tokens": 15224963.0,
      "step": 8245
    },
    {
      "entropy": 5.62935700416565,
      "epoch": 0.6931316950220542,
      "grad_norm": 0.953125,
      "learning_rate": 0.000495824107891732,
      "loss": 5.2962,
      "mean_token_accuracy": 0.17529756128787993,
      "num_tokens": 15233569.0,
      "step": 8250
    },
    {
      "entropy": 5.710500335693359,
      "epoch": 0.6935517748372191,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004958183631757336,
      "loss": 5.4676,
      "mean_token_accuracy": 0.16601490527391433,
      "num_tokens": 15242671.0,
      "step": 8255
    },
    {
      "entropy": 5.703408718109131,
      "epoch": 0.693971854652384,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004958126145480517,
      "loss": 5.4113,
      "mean_token_accuracy": 0.17612217962741852,
      "num_tokens": 15251698.0,
      "step": 8260
    },
    {
      "entropy": 5.7834920406341555,
      "epoch": 0.6943919344675489,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004958068620087879,
      "loss": 5.5402,
      "mean_token_accuracy": 0.17006468921899795,
      "num_tokens": 15260608.0,
      "step": 8265
    },
    {
      "entropy": 5.6925029277801515,
      "epoch": 0.6948120142827137,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004958011055580443,
      "loss": 5.4184,
      "mean_token_accuracy": 0.17269738465547563,
      "num_tokens": 15268866.0,
      "step": 8270
    },
    {
      "entropy": 5.629737043380738,
      "epoch": 0.6952320940978786,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004957953451959229,
      "loss": 5.3744,
      "mean_token_accuracy": 0.1811777964234352,
      "num_tokens": 15277600.0,
      "step": 8275
    },
    {
      "entropy": 5.682136154174804,
      "epoch": 0.6956521739130435,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004957895809225254,
      "loss": 5.3968,
      "mean_token_accuracy": 0.17215245068073273,
      "num_tokens": 15286016.0,
      "step": 8280
    },
    {
      "entropy": 5.705394601821899,
      "epoch": 0.6960722537282084,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004957838127379544,
      "loss": 5.4481,
      "mean_token_accuracy": 0.17101877182722092,
      "num_tokens": 15294676.0,
      "step": 8285
    },
    {
      "entropy": 5.725075101852417,
      "epoch": 0.6964923335433733,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004957780406423118,
      "loss": 5.4408,
      "mean_token_accuracy": 0.1685130462050438,
      "num_tokens": 15304084.0,
      "step": 8290
    },
    {
      "entropy": 5.7422380447387695,
      "epoch": 0.6969124133585382,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004957722646356999,
      "loss": 5.4413,
      "mean_token_accuracy": 0.16818782836198806,
      "num_tokens": 15314182.0,
      "step": 8295
    },
    {
      "entropy": 5.712948179244995,
      "epoch": 0.697332493173703,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004957664847182209,
      "loss": 5.5483,
      "mean_token_accuracy": 0.16249415427446365,
      "num_tokens": 15324213.0,
      "step": 8300
    },
    {
      "entropy": 5.754847049713135,
      "epoch": 0.6977525729888678,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004957607008899774,
      "loss": 5.4966,
      "mean_token_accuracy": 0.1627067044377327,
      "num_tokens": 15333122.0,
      "step": 8305
    },
    {
      "entropy": 5.795898056030273,
      "epoch": 0.6981726528040327,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004957549131510717,
      "loss": 5.5902,
      "mean_token_accuracy": 0.1592757970094681,
      "num_tokens": 15342199.0,
      "step": 8310
    },
    {
      "entropy": 5.825842666625976,
      "epoch": 0.6985927326191976,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004957491215016065,
      "loss": 5.5456,
      "mean_token_accuracy": 0.16149616092443467,
      "num_tokens": 15352463.0,
      "step": 8315
    },
    {
      "entropy": 5.6511616706848145,
      "epoch": 0.6990128124343625,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004957433259416841,
      "loss": 5.3651,
      "mean_token_accuracy": 0.1684968426823616,
      "num_tokens": 15361815.0,
      "step": 8320
    },
    {
      "entropy": 5.731157350540161,
      "epoch": 0.6994328922495274,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004957375264714075,
      "loss": 5.4793,
      "mean_token_accuracy": 0.16277330964803696,
      "num_tokens": 15371773.0,
      "step": 8325
    },
    {
      "entropy": 5.667555856704712,
      "epoch": 0.6998529720646923,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004957317230908792,
      "loss": 5.4412,
      "mean_token_accuracy": 0.16636938005685806,
      "num_tokens": 15380881.0,
      "step": 8330
    },
    {
      "entropy": 5.65048303604126,
      "epoch": 0.7002730518798572,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004957259158002022,
      "loss": 5.3176,
      "mean_token_accuracy": 0.1781625434756279,
      "num_tokens": 15389310.0,
      "step": 8335
    },
    {
      "entropy": 5.609999656677246,
      "epoch": 0.700693131695022,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004957201045994791,
      "loss": 5.4033,
      "mean_token_accuracy": 0.16689541339874267,
      "num_tokens": 15398584.0,
      "step": 8340
    },
    {
      "entropy": 5.720047378540039,
      "epoch": 0.7011132115101869,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004957142894888131,
      "loss": 5.4414,
      "mean_token_accuracy": 0.17400545477867127,
      "num_tokens": 15407208.0,
      "step": 8345
    },
    {
      "entropy": 5.7197003841400145,
      "epoch": 0.7015332913253518,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004957084704683071,
      "loss": 5.4709,
      "mean_token_accuracy": 0.16283994913101196,
      "num_tokens": 15416474.0,
      "step": 8350
    },
    {
      "entropy": 5.724147510528565,
      "epoch": 0.7019533711405167,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004957026475380642,
      "loss": 5.4757,
      "mean_token_accuracy": 0.17140310853719712,
      "num_tokens": 15426101.0,
      "step": 8355
    },
    {
      "entropy": 5.765893840789795,
      "epoch": 0.7023734509556816,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004956968206981875,
      "loss": 5.5211,
      "mean_token_accuracy": 0.1647307574748993,
      "num_tokens": 15435910.0,
      "step": 8360
    },
    {
      "entropy": 5.774050664901734,
      "epoch": 0.7027935307708465,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004956909899487803,
      "loss": 5.5322,
      "mean_token_accuracy": 0.16614156365394592,
      "num_tokens": 15445494.0,
      "step": 8365
    },
    {
      "entropy": 5.682026767730713,
      "epoch": 0.7032136105860114,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004956851552899459,
      "loss": 5.4382,
      "mean_token_accuracy": 0.17470966875553132,
      "num_tokens": 15455332.0,
      "step": 8370
    },
    {
      "entropy": 5.714264822006226,
      "epoch": 0.7036336904011762,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004956793167217874,
      "loss": 5.5105,
      "mean_token_accuracy": 0.16363029927015305,
      "num_tokens": 15464241.0,
      "step": 8375
    },
    {
      "entropy": 5.780676937103271,
      "epoch": 0.7040537702163411,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004956734742444087,
      "loss": 5.5017,
      "mean_token_accuracy": 0.17035040408372878,
      "num_tokens": 15473473.0,
      "step": 8380
    },
    {
      "entropy": 5.716037273406982,
      "epoch": 0.704473850031506,
      "grad_norm": 1.0,
      "learning_rate": 0.0004956676278579129,
      "loss": 5.3806,
      "mean_token_accuracy": 0.17329385429620742,
      "num_tokens": 15482494.0,
      "step": 8385
    },
    {
      "entropy": 5.651085138320923,
      "epoch": 0.7048939298466709,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004956617775624037,
      "loss": 5.4101,
      "mean_token_accuracy": 0.16964631527662277,
      "num_tokens": 15491180.0,
      "step": 8390
    },
    {
      "entropy": 5.6696333408355715,
      "epoch": 0.7053140096618358,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004956559233579848,
      "loss": 5.4555,
      "mean_token_accuracy": 0.16501674056053162,
      "num_tokens": 15501035.0,
      "step": 8395
    },
    {
      "entropy": 5.720309400558472,
      "epoch": 0.7057340894770007,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004956500652447598,
      "loss": 5.4416,
      "mean_token_accuracy": 0.17136951833963393,
      "num_tokens": 15510191.0,
      "step": 8400
    },
    {
      "entropy": 5.675808477401733,
      "epoch": 0.7061541692921655,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004956442032228324,
      "loss": 5.5188,
      "mean_token_accuracy": 0.16620295494794846,
      "num_tokens": 15519253.0,
      "step": 8405
    },
    {
      "entropy": 5.686394500732422,
      "epoch": 0.7065742491073304,
      "grad_norm": 1.0,
      "learning_rate": 0.0004956383372923067,
      "loss": 5.4754,
      "mean_token_accuracy": 0.16609471291303635,
      "num_tokens": 15528348.0,
      "step": 8410
    },
    {
      "entropy": 5.836600017547608,
      "epoch": 0.7069943289224953,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004956324674532864,
      "loss": 5.5603,
      "mean_token_accuracy": 0.16521759927272797,
      "num_tokens": 15537557.0,
      "step": 8415
    },
    {
      "entropy": 5.787467193603516,
      "epoch": 0.7074144087376601,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004956265937058757,
      "loss": 5.4804,
      "mean_token_accuracy": 0.1682910680770874,
      "num_tokens": 15546745.0,
      "step": 8420
    },
    {
      "entropy": 5.707254266738891,
      "epoch": 0.707834488552825,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004956207160501784,
      "loss": 5.3952,
      "mean_token_accuracy": 0.17201364487409593,
      "num_tokens": 15555532.0,
      "step": 8425
    },
    {
      "entropy": 5.690491533279419,
      "epoch": 0.70825456836799,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004956148344862987,
      "loss": 5.4374,
      "mean_token_accuracy": 0.17301695346832274,
      "num_tokens": 15564189.0,
      "step": 8430
    },
    {
      "entropy": 5.630844163894653,
      "epoch": 0.7086746481831548,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004956089490143408,
      "loss": 5.4712,
      "mean_token_accuracy": 0.16781792789697647,
      "num_tokens": 15574116.0,
      "step": 8435
    },
    {
      "entropy": 5.815407657623291,
      "epoch": 0.7090947279983196,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004956030596344089,
      "loss": 5.4479,
      "mean_token_accuracy": 0.16956783086061478,
      "num_tokens": 15583031.0,
      "step": 8440
    },
    {
      "entropy": 5.764626502990723,
      "epoch": 0.7095148078134845,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004955971663466075,
      "loss": 5.5794,
      "mean_token_accuracy": 0.16632352769374847,
      "num_tokens": 15592576.0,
      "step": 8445
    },
    {
      "entropy": 5.791698408126831,
      "epoch": 0.7099348876286494,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004955912691510407,
      "loss": 5.5076,
      "mean_token_accuracy": 0.16888453364372252,
      "num_tokens": 15601065.0,
      "step": 8450
    },
    {
      "entropy": 5.7341821670532225,
      "epoch": 0.7103549674438143,
      "grad_norm": 1.0,
      "learning_rate": 0.0004955853680478134,
      "loss": 5.462,
      "mean_token_accuracy": 0.15883099138736725,
      "num_tokens": 15610112.0,
      "step": 8455
    },
    {
      "entropy": 5.723237371444702,
      "epoch": 0.7107750472589792,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004955794630370297,
      "loss": 5.4256,
      "mean_token_accuracy": 0.16989553719758987,
      "num_tokens": 15618890.0,
      "step": 8460
    },
    {
      "entropy": 5.7183808326721195,
      "epoch": 0.7111951270741441,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004955735541187945,
      "loss": 5.4667,
      "mean_token_accuracy": 0.16563509553670883,
      "num_tokens": 15627678.0,
      "step": 8465
    },
    {
      "entropy": 5.794507265090942,
      "epoch": 0.711615206889309,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004955676412932124,
      "loss": 5.4421,
      "mean_token_accuracy": 0.173024220764637,
      "num_tokens": 15636833.0,
      "step": 8470
    },
    {
      "entropy": 5.714361238479614,
      "epoch": 0.7120352867044738,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004955617245603881,
      "loss": 5.4575,
      "mean_token_accuracy": 0.1654538705945015,
      "num_tokens": 15646571.0,
      "step": 8475
    },
    {
      "entropy": 5.6584114074707035,
      "epoch": 0.7124553665196387,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004955558039204263,
      "loss": 5.5003,
      "mean_token_accuracy": 0.17047171145677567,
      "num_tokens": 15654907.0,
      "step": 8480
    },
    {
      "entropy": 5.769586610794067,
      "epoch": 0.7128754463348036,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004955498793734321,
      "loss": 5.4361,
      "mean_token_accuracy": 0.16981849670410157,
      "num_tokens": 15664336.0,
      "step": 8485
    },
    {
      "entropy": 5.760800886154175,
      "epoch": 0.7132955261499685,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004955439509195103,
      "loss": 5.5124,
      "mean_token_accuracy": 0.1660913497209549,
      "num_tokens": 15674000.0,
      "step": 8490
    },
    {
      "entropy": 5.781082010269165,
      "epoch": 0.7137156059651334,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004955380185587661,
      "loss": 5.5052,
      "mean_token_accuracy": 0.1704909771680832,
      "num_tokens": 15684214.0,
      "step": 8495
    },
    {
      "entropy": 5.754578447341919,
      "epoch": 0.7141356857802983,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004955320822913043,
      "loss": 5.5162,
      "mean_token_accuracy": 0.1640373393893242,
      "num_tokens": 15693546.0,
      "step": 8500
    },
    {
      "entropy": 5.726579427719116,
      "epoch": 0.7145557655954632,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004955261421172302,
      "loss": 5.4062,
      "mean_token_accuracy": 0.167661589384079,
      "num_tokens": 15702310.0,
      "step": 8505
    },
    {
      "entropy": 5.678763675689697,
      "epoch": 0.714975845410628,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004955201980366493,
      "loss": 5.4737,
      "mean_token_accuracy": 0.16903873085975646,
      "num_tokens": 15711544.0,
      "step": 8510
    },
    {
      "entropy": 5.594863843917847,
      "epoch": 0.7153959252257929,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004955142500496665,
      "loss": 5.3656,
      "mean_token_accuracy": 0.17074762731790544,
      "num_tokens": 15720914.0,
      "step": 8515
    },
    {
      "entropy": 5.749608850479126,
      "epoch": 0.7158160050409578,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004955082981563872,
      "loss": 5.4642,
      "mean_token_accuracy": 0.16272155642509462,
      "num_tokens": 15729825.0,
      "step": 8520
    },
    {
      "entropy": 5.720403575897217,
      "epoch": 0.7162360848561227,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000495502342356917,
      "loss": 5.4694,
      "mean_token_accuracy": 0.17173092514276506,
      "num_tokens": 15739649.0,
      "step": 8525
    },
    {
      "entropy": 5.714644861221314,
      "epoch": 0.7166561646712876,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004954963826513614,
      "loss": 5.3531,
      "mean_token_accuracy": 0.17039962708950043,
      "num_tokens": 15747805.0,
      "step": 8530
    },
    {
      "entropy": 5.757282876968384,
      "epoch": 0.7170762444864525,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000495490419039826,
      "loss": 5.4986,
      "mean_token_accuracy": 0.16411223262548447,
      "num_tokens": 15757267.0,
      "step": 8535
    },
    {
      "entropy": 5.730868721008301,
      "epoch": 0.7174963243016174,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004954844515224162,
      "loss": 5.4571,
      "mean_token_accuracy": 0.16875616163015367,
      "num_tokens": 15767412.0,
      "step": 8540
    },
    {
      "entropy": 5.691779232025146,
      "epoch": 0.7179164041167821,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004954784800992379,
      "loss": 5.4588,
      "mean_token_accuracy": 0.16602378189563752,
      "num_tokens": 15776813.0,
      "step": 8545
    },
    {
      "entropy": 5.759685468673706,
      "epoch": 0.718336483931947,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004954725047703969,
      "loss": 5.4929,
      "mean_token_accuracy": 0.16820817142724992,
      "num_tokens": 15786258.0,
      "step": 8550
    },
    {
      "entropy": 5.724852085113525,
      "epoch": 0.7187565637471119,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000495466525535999,
      "loss": 5.4597,
      "mean_token_accuracy": 0.1727804183959961,
      "num_tokens": 15795673.0,
      "step": 8555
    },
    {
      "entropy": 5.736737442016602,
      "epoch": 0.7191766435622768,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004954605423961501,
      "loss": 5.4573,
      "mean_token_accuracy": 0.16743139922618866,
      "num_tokens": 15805050.0,
      "step": 8560
    },
    {
      "entropy": 5.666402006149292,
      "epoch": 0.7195967233774417,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004954545553509562,
      "loss": 5.4085,
      "mean_token_accuracy": 0.17864352017641066,
      "num_tokens": 15813347.0,
      "step": 8565
    },
    {
      "entropy": 5.774226522445678,
      "epoch": 0.7200168031926066,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004954485644005235,
      "loss": 5.528,
      "mean_token_accuracy": 0.1669788345694542,
      "num_tokens": 15823528.0,
      "step": 8570
    },
    {
      "entropy": 5.746664953231812,
      "epoch": 0.7204368830077714,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004954425695449578,
      "loss": 5.4508,
      "mean_token_accuracy": 0.16504298746585847,
      "num_tokens": 15832727.0,
      "step": 8575
    },
    {
      "entropy": 5.7676012992858885,
      "epoch": 0.7208569628229363,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004954365707843657,
      "loss": 5.5178,
      "mean_token_accuracy": 0.16330185979604722,
      "num_tokens": 15842402.0,
      "step": 8580
    },
    {
      "entropy": 5.679731559753418,
      "epoch": 0.7212770426381012,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004954305681188531,
      "loss": 5.3759,
      "mean_token_accuracy": 0.1738160789012909,
      "num_tokens": 15850886.0,
      "step": 8585
    },
    {
      "entropy": 5.858513593673706,
      "epoch": 0.7216971224532661,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004954245615485265,
      "loss": 5.6717,
      "mean_token_accuracy": 0.16439133360981942,
      "num_tokens": 15860093.0,
      "step": 8590
    },
    {
      "entropy": 5.713961744308472,
      "epoch": 0.722117202268431,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004954185510734924,
      "loss": 5.3886,
      "mean_token_accuracy": 0.17300495952367784,
      "num_tokens": 15868681.0,
      "step": 8595
    },
    {
      "entropy": 5.688973331451416,
      "epoch": 0.7225372820835959,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004954125366938571,
      "loss": 5.4504,
      "mean_token_accuracy": 0.1770208552479744,
      "num_tokens": 15878041.0,
      "step": 8600
    },
    {
      "entropy": 5.701451826095581,
      "epoch": 0.7229573618987608,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004954065184097271,
      "loss": 5.4422,
      "mean_token_accuracy": 0.1692481279373169,
      "num_tokens": 15887562.0,
      "step": 8605
    },
    {
      "entropy": 5.695767259597778,
      "epoch": 0.7233774417139256,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004954004962212092,
      "loss": 5.3745,
      "mean_token_accuracy": 0.17987406849861146,
      "num_tokens": 15896480.0,
      "step": 8610
    },
    {
      "entropy": 5.813690280914306,
      "epoch": 0.7237975215290905,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004953944701284101,
      "loss": 5.6009,
      "mean_token_accuracy": 0.1619385376572609,
      "num_tokens": 15906743.0,
      "step": 8615
    },
    {
      "entropy": 5.743529319763184,
      "epoch": 0.7242176013442554,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004953884401314363,
      "loss": 5.5401,
      "mean_token_accuracy": 0.15503934770822525,
      "num_tokens": 15915981.0,
      "step": 8620
    },
    {
      "entropy": 5.726900053024292,
      "epoch": 0.7246376811594203,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004953824062303949,
      "loss": 5.3852,
      "mean_token_accuracy": 0.16863519996404647,
      "num_tokens": 15924117.0,
      "step": 8625
    },
    {
      "entropy": 5.671165132522583,
      "epoch": 0.7250577609745852,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004953763684253926,
      "loss": 5.4228,
      "mean_token_accuracy": 0.1744863882660866,
      "num_tokens": 15933124.0,
      "step": 8630
    },
    {
      "entropy": 5.676993417739868,
      "epoch": 0.7254778407897501,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004953703267165364,
      "loss": 5.3196,
      "mean_token_accuracy": 0.1732983782887459,
      "num_tokens": 15942422.0,
      "step": 8635
    },
    {
      "entropy": 5.728943920135498,
      "epoch": 0.725897920604915,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004953642811039332,
      "loss": 5.5281,
      "mean_token_accuracy": 0.16218795329332353,
      "num_tokens": 15950989.0,
      "step": 8640
    },
    {
      "entropy": 5.7657547950744625,
      "epoch": 0.7263180004200798,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004953582315876904,
      "loss": 5.53,
      "mean_token_accuracy": 0.16450190246105195,
      "num_tokens": 15959659.0,
      "step": 8645
    },
    {
      "entropy": 5.7511814594268795,
      "epoch": 0.7267380802352447,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000495352178167915,
      "loss": 5.4005,
      "mean_token_accuracy": 0.18288260698318481,
      "num_tokens": 15968102.0,
      "step": 8650
    },
    {
      "entropy": 5.785438919067383,
      "epoch": 0.7271581600504096,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004953461208447143,
      "loss": 5.5326,
      "mean_token_accuracy": 0.16101690530776977,
      "num_tokens": 15977705.0,
      "step": 8655
    },
    {
      "entropy": 5.739281797409058,
      "epoch": 0.7275782398655745,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004953400596181953,
      "loss": 5.522,
      "mean_token_accuracy": 0.16212962567806244,
      "num_tokens": 15986703.0,
      "step": 8660
    },
    {
      "entropy": 5.755725765228272,
      "epoch": 0.7279983196807394,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004953339944884657,
      "loss": 5.4248,
      "mean_token_accuracy": 0.17208922058343887,
      "num_tokens": 15995672.0,
      "step": 8665
    },
    {
      "entropy": 5.587175512313843,
      "epoch": 0.7284183994959043,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004953279254556329,
      "loss": 5.3964,
      "mean_token_accuracy": 0.17940126806497575,
      "num_tokens": 16004437.0,
      "step": 8670
    },
    {
      "entropy": 5.701867580413818,
      "epoch": 0.7288384793110692,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004953218525198043,
      "loss": 5.4279,
      "mean_token_accuracy": 0.17154791951179504,
      "num_tokens": 16012847.0,
      "step": 8675
    },
    {
      "entropy": 5.788314533233643,
      "epoch": 0.7292585591262339,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004953157756810876,
      "loss": 5.4375,
      "mean_token_accuracy": 0.17364862710237502,
      "num_tokens": 16022213.0,
      "step": 8680
    },
    {
      "entropy": 5.689052963256836,
      "epoch": 0.7296786389413988,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004953096949395902,
      "loss": 5.5076,
      "mean_token_accuracy": 0.17346307635307312,
      "num_tokens": 16031411.0,
      "step": 8685
    },
    {
      "entropy": 5.719053745269775,
      "epoch": 0.7300987187565637,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004953036102954202,
      "loss": 5.5491,
      "mean_token_accuracy": 0.16453250944614412,
      "num_tokens": 16041227.0,
      "step": 8690
    },
    {
      "entropy": 5.6874908924102785,
      "epoch": 0.7305187985717286,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004952975217486852,
      "loss": 5.3687,
      "mean_token_accuracy": 0.17681472301483153,
      "num_tokens": 16049777.0,
      "step": 8695
    },
    {
      "entropy": 5.735844707489013,
      "epoch": 0.7309388783868935,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004952914292994928,
      "loss": 5.4537,
      "mean_token_accuracy": 0.17558913975954055,
      "num_tokens": 16059093.0,
      "step": 8700
    },
    {
      "entropy": 5.7709338665008545,
      "epoch": 0.7313589582020584,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004952853329479514,
      "loss": 5.4978,
      "mean_token_accuracy": 0.1736828625202179,
      "num_tokens": 16068550.0,
      "step": 8705
    },
    {
      "entropy": 5.749593019485474,
      "epoch": 0.7317790380172233,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004952792326941686,
      "loss": 5.5216,
      "mean_token_accuracy": 0.16563831269741058,
      "num_tokens": 16078286.0,
      "step": 8710
    },
    {
      "entropy": 5.751245927810669,
      "epoch": 0.7321991178323881,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004952731285382527,
      "loss": 5.4629,
      "mean_token_accuracy": 0.17020100355148315,
      "num_tokens": 16087560.0,
      "step": 8715
    },
    {
      "entropy": 5.657774114608765,
      "epoch": 0.732619197647553,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004952670204803118,
      "loss": 5.4111,
      "mean_token_accuracy": 0.17724975645542146,
      "num_tokens": 16097478.0,
      "step": 8720
    },
    {
      "entropy": 5.730371761322021,
      "epoch": 0.7330392774627179,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004952609085204539,
      "loss": 5.5315,
      "mean_token_accuracy": 0.17183186411857604,
      "num_tokens": 16106884.0,
      "step": 8725
    },
    {
      "entropy": 5.733107042312622,
      "epoch": 0.7334593572778828,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004952547926587876,
      "loss": 5.4493,
      "mean_token_accuracy": 0.16233453452587127,
      "num_tokens": 16115689.0,
      "step": 8730
    },
    {
      "entropy": 5.706781721115112,
      "epoch": 0.7338794370930477,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004952486728954209,
      "loss": 5.3842,
      "mean_token_accuracy": 0.17700668871402742,
      "num_tokens": 16125237.0,
      "step": 8735
    },
    {
      "entropy": 5.676010227203369,
      "epoch": 0.7342995169082126,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004952425492304624,
      "loss": 5.4004,
      "mean_token_accuracy": 0.17442144006490706,
      "num_tokens": 16133940.0,
      "step": 8740
    },
    {
      "entropy": 5.725145864486694,
      "epoch": 0.7347195967233774,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004952364216640207,
      "loss": 5.4792,
      "mean_token_accuracy": 0.16769580841064452,
      "num_tokens": 16143256.0,
      "step": 8745
    },
    {
      "entropy": 5.746494483947754,
      "epoch": 0.7351396765385423,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000495230290196204,
      "loss": 5.3788,
      "mean_token_accuracy": 0.17094354331493378,
      "num_tokens": 16153259.0,
      "step": 8750
    },
    {
      "entropy": 5.758786964416504,
      "epoch": 0.7355597563537072,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004952241548271212,
      "loss": 5.618,
      "mean_token_accuracy": 0.15930501893162727,
      "num_tokens": 16162125.0,
      "step": 8755
    },
    {
      "entropy": 5.791958570480347,
      "epoch": 0.7359798361688721,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004952180155568809,
      "loss": 5.5292,
      "mean_token_accuracy": 0.16540495604276656,
      "num_tokens": 16171680.0,
      "step": 8760
    },
    {
      "entropy": 5.773538208007812,
      "epoch": 0.736399915984037,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004952118723855919,
      "loss": 5.5271,
      "mean_token_accuracy": 0.1690017431974411,
      "num_tokens": 16181559.0,
      "step": 8765
    },
    {
      "entropy": 5.749522876739502,
      "epoch": 0.7368199957992019,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004952057253133628,
      "loss": 5.478,
      "mean_token_accuracy": 0.1682865560054779,
      "num_tokens": 16190611.0,
      "step": 8770
    },
    {
      "entropy": 5.7609809875488285,
      "epoch": 0.7372400756143668,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004951995743403028,
      "loss": 5.4943,
      "mean_token_accuracy": 0.1672732874751091,
      "num_tokens": 16200156.0,
      "step": 8775
    },
    {
      "entropy": 5.677986335754395,
      "epoch": 0.7376601554295316,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004951934194665208,
      "loss": 5.4703,
      "mean_token_accuracy": 0.16116737425327302,
      "num_tokens": 16209808.0,
      "step": 8780
    },
    {
      "entropy": 5.662946367263794,
      "epoch": 0.7380802352446965,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004951872606921257,
      "loss": 5.4139,
      "mean_token_accuracy": 0.17300440967082978,
      "num_tokens": 16219243.0,
      "step": 8785
    },
    {
      "entropy": 5.688099145889282,
      "epoch": 0.7385003150598614,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004951810980172265,
      "loss": 5.4322,
      "mean_token_accuracy": 0.18030907660722734,
      "num_tokens": 16228180.0,
      "step": 8790
    },
    {
      "entropy": 5.713538980484008,
      "epoch": 0.7389203948750263,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004951749314419327,
      "loss": 5.4606,
      "mean_token_accuracy": 0.1713638111948967,
      "num_tokens": 16237045.0,
      "step": 8795
    },
    {
      "entropy": 5.729268360137939,
      "epoch": 0.7393404746901912,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004951687609663533,
      "loss": 5.3593,
      "mean_token_accuracy": 0.17546552717685698,
      "num_tokens": 16245307.0,
      "step": 8800
    },
    {
      "entropy": 5.705272102355957,
      "epoch": 0.739760554505356,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004951625865905977,
      "loss": 5.4152,
      "mean_token_accuracy": 0.1655416816473007,
      "num_tokens": 16255047.0,
      "step": 8805
    },
    {
      "entropy": 5.683449697494507,
      "epoch": 0.740180634320521,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004951564083147753,
      "loss": 5.4428,
      "mean_token_accuracy": 0.1721823990345001,
      "num_tokens": 16264969.0,
      "step": 8810
    },
    {
      "entropy": 5.735998058319092,
      "epoch": 0.7406007141356857,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004951502261389953,
      "loss": 5.5395,
      "mean_token_accuracy": 0.1642596483230591,
      "num_tokens": 16274757.0,
      "step": 8815
    },
    {
      "entropy": 5.696336126327514,
      "epoch": 0.7410207939508506,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004951440400633677,
      "loss": 5.4385,
      "mean_token_accuracy": 0.18188624531030656,
      "num_tokens": 16283409.0,
      "step": 8820
    },
    {
      "entropy": 5.680422639846801,
      "epoch": 0.7414408737660155,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004951378500880015,
      "loss": 5.4064,
      "mean_token_accuracy": 0.17475765198469162,
      "num_tokens": 16293206.0,
      "step": 8825
    },
    {
      "entropy": 5.750751066207886,
      "epoch": 0.7418609535811804,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004951316562130067,
      "loss": 5.4371,
      "mean_token_accuracy": 0.16920247077941894,
      "num_tokens": 16303121.0,
      "step": 8830
    },
    {
      "entropy": 5.6800635814666744,
      "epoch": 0.7422810333963453,
      "grad_norm": 0.875,
      "learning_rate": 0.000495125458438493,
      "loss": 5.3993,
      "mean_token_accuracy": 0.17428747564554214,
      "num_tokens": 16312710.0,
      "step": 8835
    },
    {
      "entropy": 5.816542863845825,
      "epoch": 0.7427011132115102,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004951192567645702,
      "loss": 5.5965,
      "mean_token_accuracy": 0.16230317950248718,
      "num_tokens": 16322280.0,
      "step": 8840
    },
    {
      "entropy": 5.66908860206604,
      "epoch": 0.7431211930266751,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004951130511913481,
      "loss": 5.4274,
      "mean_token_accuracy": 0.1729036808013916,
      "num_tokens": 16331656.0,
      "step": 8845
    },
    {
      "entropy": 5.699582195281982,
      "epoch": 0.7435412728418399,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004951068417189366,
      "loss": 5.4602,
      "mean_token_accuracy": 0.1716711178421974,
      "num_tokens": 16341074.0,
      "step": 8850
    },
    {
      "entropy": 5.752342224121094,
      "epoch": 0.7439613526570048,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004951006283474457,
      "loss": 5.4418,
      "mean_token_accuracy": 0.1679142266511917,
      "num_tokens": 16350097.0,
      "step": 8855
    },
    {
      "entropy": 5.576859951019287,
      "epoch": 0.7443814324721697,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004950944110769856,
      "loss": 5.3652,
      "mean_token_accuracy": 0.17501350343227387,
      "num_tokens": 16359274.0,
      "step": 8860
    },
    {
      "entropy": 5.6189018249511715,
      "epoch": 0.7448015122873346,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004950881899076663,
      "loss": 5.3279,
      "mean_token_accuracy": 0.1871775731444359,
      "num_tokens": 16368445.0,
      "step": 8865
    },
    {
      "entropy": 5.748156213760376,
      "epoch": 0.7452215921024995,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004950819648395979,
      "loss": 5.4319,
      "mean_token_accuracy": 0.1757849007844925,
      "num_tokens": 16377689.0,
      "step": 8870
    },
    {
      "entropy": 5.702897644042968,
      "epoch": 0.7456416719176644,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000495075735872891,
      "loss": 5.4008,
      "mean_token_accuracy": 0.16995237171649932,
      "num_tokens": 16386713.0,
      "step": 8875
    },
    {
      "entropy": 5.722459173202514,
      "epoch": 0.7460617517328293,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004950695030076557,
      "loss": 5.4167,
      "mean_token_accuracy": 0.17003950476646423,
      "num_tokens": 16395390.0,
      "step": 8880
    },
    {
      "entropy": 5.709754419326782,
      "epoch": 0.7464818315479941,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004950632662440027,
      "loss": 5.4968,
      "mean_token_accuracy": 0.17370348274707795,
      "num_tokens": 16404531.0,
      "step": 8885
    },
    {
      "entropy": 5.6589555740356445,
      "epoch": 0.746901911363159,
      "grad_norm": 1.0,
      "learning_rate": 0.0004950570255820419,
      "loss": 5.3987,
      "mean_token_accuracy": 0.17150552719831466,
      "num_tokens": 16413649.0,
      "step": 8890
    },
    {
      "entropy": 5.680497932434082,
      "epoch": 0.7473219911783239,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004950507810218843,
      "loss": 5.511,
      "mean_token_accuracy": 0.16552116945385933,
      "num_tokens": 16423247.0,
      "step": 8895
    },
    {
      "entropy": 5.797210931777954,
      "epoch": 0.7477420709934888,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004950445325636405,
      "loss": 5.4271,
      "mean_token_accuracy": 0.16588872224092482,
      "num_tokens": 16432190.0,
      "step": 8900
    },
    {
      "entropy": 5.725117492675781,
      "epoch": 0.7481621508086537,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004950382802074211,
      "loss": 5.3915,
      "mean_token_accuracy": 0.1822303220629692,
      "num_tokens": 16443091.0,
      "step": 8905
    },
    {
      "entropy": 5.6469872951507565,
      "epoch": 0.7485822306238186,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004950320239533369,
      "loss": 5.4387,
      "mean_token_accuracy": 0.16911691129207612,
      "num_tokens": 16452077.0,
      "step": 8910
    },
    {
      "entropy": 5.7999378681182865,
      "epoch": 0.7490023104389834,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004950257638014986,
      "loss": 5.5597,
      "mean_token_accuracy": 0.1630978539586067,
      "num_tokens": 16461893.0,
      "step": 8915
    },
    {
      "entropy": 5.756207513809204,
      "epoch": 0.7494223902541483,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004950194997520172,
      "loss": 5.405,
      "mean_token_accuracy": 0.1679363951086998,
      "num_tokens": 16470904.0,
      "step": 8920
    },
    {
      "entropy": 5.68512544631958,
      "epoch": 0.7498424700693131,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004950132318050037,
      "loss": 5.4649,
      "mean_token_accuracy": 0.1691165268421173,
      "num_tokens": 16480130.0,
      "step": 8925
    },
    {
      "entropy": 5.703501272201538,
      "epoch": 0.750262549884478,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004950069599605691,
      "loss": 5.5055,
      "mean_token_accuracy": 0.1704162359237671,
      "num_tokens": 16489485.0,
      "step": 8930
    },
    {
      "entropy": 5.765615510940552,
      "epoch": 0.750682629699643,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004950006842188245,
      "loss": 5.4669,
      "mean_token_accuracy": 0.1749041572213173,
      "num_tokens": 16498529.0,
      "step": 8935
    },
    {
      "entropy": 5.703298091888428,
      "epoch": 0.7511027095148078,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000494994404579881,
      "loss": 5.3843,
      "mean_token_accuracy": 0.17271067798137665,
      "num_tokens": 16508094.0,
      "step": 8940
    },
    {
      "entropy": 5.711791753768921,
      "epoch": 0.7515227893299727,
      "grad_norm": 1.015625,
      "learning_rate": 0.00049498812104385,
      "loss": 5.4846,
      "mean_token_accuracy": 0.1700764775276184,
      "num_tokens": 16517620.0,
      "step": 8945
    },
    {
      "entropy": 5.717977094650268,
      "epoch": 0.7519428691451375,
      "grad_norm": 1.0,
      "learning_rate": 0.0004949818336108425,
      "loss": 5.4782,
      "mean_token_accuracy": 0.16688166409730912,
      "num_tokens": 16526720.0,
      "step": 8950
    },
    {
      "entropy": 5.681493616104126,
      "epoch": 0.7523629489603024,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004949755422809703,
      "loss": 5.4475,
      "mean_token_accuracy": 0.1643982008099556,
      "num_tokens": 16535979.0,
      "step": 8955
    },
    {
      "entropy": 5.698634004592895,
      "epoch": 0.7527830287754673,
      "grad_norm": 1.0,
      "learning_rate": 0.0004949692470543446,
      "loss": 5.3374,
      "mean_token_accuracy": 0.18242864608764647,
      "num_tokens": 16544538.0,
      "step": 8960
    },
    {
      "entropy": 5.673680353164673,
      "epoch": 0.7532031085906322,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004949629479310769,
      "loss": 5.4116,
      "mean_token_accuracy": 0.17440479397773742,
      "num_tokens": 16553962.0,
      "step": 8965
    },
    {
      "entropy": 5.689835166931152,
      "epoch": 0.7536231884057971,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004949566449112788,
      "loss": 5.3409,
      "mean_token_accuracy": 0.1779365137219429,
      "num_tokens": 16562652.0,
      "step": 8970
    },
    {
      "entropy": 5.750735950469971,
      "epoch": 0.754043268220962,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004949503379950621,
      "loss": 5.4492,
      "mean_token_accuracy": 0.1657186269760132,
      "num_tokens": 16570887.0,
      "step": 8975
    },
    {
      "entropy": 5.755009412765503,
      "epoch": 0.7544633480361269,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004949440271825385,
      "loss": 5.551,
      "mean_token_accuracy": 0.16502224653959274,
      "num_tokens": 16581469.0,
      "step": 8980
    },
    {
      "entropy": 5.750725841522216,
      "epoch": 0.7548834278512917,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004949377124738196,
      "loss": 5.4568,
      "mean_token_accuracy": 0.16357554495334625,
      "num_tokens": 16590213.0,
      "step": 8985
    },
    {
      "entropy": 5.720483827590942,
      "epoch": 0.7553035076664566,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004949313938690174,
      "loss": 5.4353,
      "mean_token_accuracy": 0.16786181479692458,
      "num_tokens": 16598384.0,
      "step": 8990
    },
    {
      "entropy": 5.654509162902832,
      "epoch": 0.7557235874816215,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004949250713682438,
      "loss": 5.415,
      "mean_token_accuracy": 0.17460085153579713,
      "num_tokens": 16607670.0,
      "step": 8995
    },
    {
      "entropy": 5.74861364364624,
      "epoch": 0.7561436672967864,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004949187449716107,
      "loss": 5.5033,
      "mean_token_accuracy": 0.16701496690511702,
      "num_tokens": 16617560.0,
      "step": 9000
    },
    {
      "epoch": 0.7561436672967864,
      "eval_entropy": 5.503128333307134,
      "eval_loss": 5.453708171844482,
      "eval_mean_token_accuracy": 0.17778601125042207,
      "eval_num_tokens": 16617560.0,
      "eval_runtime": 21.0065,
      "eval_samples_per_second": 1778.783,
      "eval_steps_per_second": 222.36,
      "step": 9000
    },
    {
      "entropy": 5.709756469726562,
      "epoch": 0.7565637471119513,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004949124146792304,
      "loss": 5.4137,
      "mean_token_accuracy": 0.17349707037210466,
      "num_tokens": 16626038.0,
      "step": 9005
    },
    {
      "entropy": 5.6661951541900635,
      "epoch": 0.7569838269271162,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004949060804912149,
      "loss": 5.4201,
      "mean_token_accuracy": 0.17093457877635956,
      "num_tokens": 16636490.0,
      "step": 9010
    },
    {
      "entropy": 5.738677263259888,
      "epoch": 0.7574039067422811,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004948997424076764,
      "loss": 5.4192,
      "mean_token_accuracy": 0.17112809866666795,
      "num_tokens": 16645369.0,
      "step": 9015
    },
    {
      "entropy": 5.774097013473511,
      "epoch": 0.7578239865574459,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004948934004287272,
      "loss": 5.5116,
      "mean_token_accuracy": 0.16631682962179184,
      "num_tokens": 16654348.0,
      "step": 9020
    },
    {
      "entropy": 5.787421560287475,
      "epoch": 0.7582440663726108,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004948870545544796,
      "loss": 5.5089,
      "mean_token_accuracy": 0.16316519677639008,
      "num_tokens": 16664009.0,
      "step": 9025
    },
    {
      "entropy": 5.711210584640503,
      "epoch": 0.7586641461877757,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000494880704785046,
      "loss": 5.5115,
      "mean_token_accuracy": 0.1665276974439621,
      "num_tokens": 16674079.0,
      "step": 9030
    },
    {
      "entropy": 5.765767765045166,
      "epoch": 0.7590842260029406,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004948743511205392,
      "loss": 5.4453,
      "mean_token_accuracy": 0.16655828878283502,
      "num_tokens": 16683687.0,
      "step": 9035
    },
    {
      "entropy": 5.731180429458618,
      "epoch": 0.7595043058181055,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004948679935610712,
      "loss": 5.3593,
      "mean_token_accuracy": 0.18364747762680053,
      "num_tokens": 16693311.0,
      "step": 9040
    },
    {
      "entropy": 5.6440253257751465,
      "epoch": 0.7599243856332704,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000494861632106755,
      "loss": 5.4036,
      "mean_token_accuracy": 0.1687808722257614,
      "num_tokens": 16702121.0,
      "step": 9045
    },
    {
      "entropy": 5.7021595478057865,
      "epoch": 0.7603444654484351,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004948552667577033,
      "loss": 5.4279,
      "mean_token_accuracy": 0.17165983766317366,
      "num_tokens": 16711883.0,
      "step": 9050
    },
    {
      "entropy": 5.736509943008423,
      "epoch": 0.7607645452636,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004948488975140286,
      "loss": 5.4872,
      "mean_token_accuracy": 0.1662601962685585,
      "num_tokens": 16721449.0,
      "step": 9055
    },
    {
      "entropy": 5.719401693344116,
      "epoch": 0.7611846250787649,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000494842524375844,
      "loss": 5.4228,
      "mean_token_accuracy": 0.17060478180646896,
      "num_tokens": 16730068.0,
      "step": 9060
    },
    {
      "entropy": 5.7237855911254885,
      "epoch": 0.7616047048939298,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004948361473432623,
      "loss": 5.442,
      "mean_token_accuracy": 0.17317203432321548,
      "num_tokens": 16739970.0,
      "step": 9065
    },
    {
      "entropy": 5.78205361366272,
      "epoch": 0.7620247847090947,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004948297664163964,
      "loss": 5.5112,
      "mean_token_accuracy": 0.16682129949331284,
      "num_tokens": 16749461.0,
      "step": 9070
    },
    {
      "entropy": 5.745754098892212,
      "epoch": 0.7624448645242596,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004948233815953593,
      "loss": 5.5791,
      "mean_token_accuracy": 0.1631207138299942,
      "num_tokens": 16758747.0,
      "step": 9075
    },
    {
      "entropy": 5.68776683807373,
      "epoch": 0.7628649443394245,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004948169928802643,
      "loss": 5.2934,
      "mean_token_accuracy": 0.18020772486925124,
      "num_tokens": 16767212.0,
      "step": 9080
    },
    {
      "entropy": 5.7201744556427006,
      "epoch": 0.7632850241545893,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004948106002712245,
      "loss": 5.4386,
      "mean_token_accuracy": 0.17203060090541838,
      "num_tokens": 16776514.0,
      "step": 9085
    },
    {
      "entropy": 5.7265181064605715,
      "epoch": 0.7637051039697542,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004948042037683529,
      "loss": 5.4369,
      "mean_token_accuracy": 0.16726903468370438,
      "num_tokens": 16786310.0,
      "step": 9090
    },
    {
      "entropy": 5.734868001937866,
      "epoch": 0.7641251837849191,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004947978033717632,
      "loss": 5.4533,
      "mean_token_accuracy": 0.17024074047803878,
      "num_tokens": 16795551.0,
      "step": 9095
    },
    {
      "entropy": 5.738364267349243,
      "epoch": 0.764545263600084,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004947913990815684,
      "loss": 5.4356,
      "mean_token_accuracy": 0.17073904126882553,
      "num_tokens": 16805099.0,
      "step": 9100
    },
    {
      "entropy": 5.721946477890015,
      "epoch": 0.7649653434152489,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004947849908978824,
      "loss": 5.4867,
      "mean_token_accuracy": 0.1674138903617859,
      "num_tokens": 16813963.0,
      "step": 9105
    },
    {
      "entropy": 5.792682552337647,
      "epoch": 0.7653854232304138,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004947785788208182,
      "loss": 5.5074,
      "mean_token_accuracy": 0.16473163217306136,
      "num_tokens": 16822814.0,
      "step": 9110
    },
    {
      "entropy": 5.7823121547698975,
      "epoch": 0.7658055030455787,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004947721628504898,
      "loss": 5.5312,
      "mean_token_accuracy": 0.16430546790361406,
      "num_tokens": 16831906.0,
      "step": 9115
    },
    {
      "entropy": 5.678847742080689,
      "epoch": 0.7662255828607435,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004947657429870108,
      "loss": 5.3638,
      "mean_token_accuracy": 0.17285381406545638,
      "num_tokens": 16840050.0,
      "step": 9120
    },
    {
      "entropy": 5.671515274047851,
      "epoch": 0.7666456626759084,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004947593192304946,
      "loss": 5.3857,
      "mean_token_accuracy": 0.17043459713459014,
      "num_tokens": 16848404.0,
      "step": 9125
    },
    {
      "entropy": 5.674127531051636,
      "epoch": 0.7670657424910733,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004947528915810554,
      "loss": 5.3789,
      "mean_token_accuracy": 0.16918594390153885,
      "num_tokens": 16856568.0,
      "step": 9130
    },
    {
      "entropy": 5.684866571426392,
      "epoch": 0.7674858223062382,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004947464600388066,
      "loss": 5.4084,
      "mean_token_accuracy": 0.17281172573566436,
      "num_tokens": 16864936.0,
      "step": 9135
    },
    {
      "entropy": 5.887958574295044,
      "epoch": 0.7679059021214031,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004947400246038627,
      "loss": 5.5398,
      "mean_token_accuracy": 0.16796707659959792,
      "num_tokens": 16874504.0,
      "step": 9140
    },
    {
      "entropy": 5.648350620269776,
      "epoch": 0.768325981936568,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004947335852763374,
      "loss": 5.3033,
      "mean_token_accuracy": 0.17359408885240554,
      "num_tokens": 16883365.0,
      "step": 9145
    },
    {
      "entropy": 5.68917932510376,
      "epoch": 0.7687460617517329,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004947271420563447,
      "loss": 5.5471,
      "mean_token_accuracy": 0.15678050816059114,
      "num_tokens": 16892701.0,
      "step": 9150
    },
    {
      "entropy": 5.722358512878418,
      "epoch": 0.7691661415668977,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004947206949439989,
      "loss": 5.3812,
      "mean_token_accuracy": 0.16777743250131608,
      "num_tokens": 16901864.0,
      "step": 9155
    },
    {
      "entropy": 5.7149741649627686,
      "epoch": 0.7695862213820626,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000494714243939414,
      "loss": 5.4032,
      "mean_token_accuracy": 0.1772664561867714,
      "num_tokens": 16910908.0,
      "step": 9160
    },
    {
      "entropy": 5.702767944335937,
      "epoch": 0.7700063011972275,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004947077890427045,
      "loss": 5.4322,
      "mean_token_accuracy": 0.17422608882188798,
      "num_tokens": 16920299.0,
      "step": 9165
    },
    {
      "entropy": 5.803005790710449,
      "epoch": 0.7704263810123924,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004947013302539846,
      "loss": 5.5643,
      "mean_token_accuracy": 0.15801274180412292,
      "num_tokens": 16930027.0,
      "step": 9170
    },
    {
      "entropy": 5.81136531829834,
      "epoch": 0.7708464608275573,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004946948675733688,
      "loss": 5.4455,
      "mean_token_accuracy": 0.16715119332075118,
      "num_tokens": 16939387.0,
      "step": 9175
    },
    {
      "entropy": 5.704103660583496,
      "epoch": 0.7712665406427222,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004946884010009714,
      "loss": 5.4232,
      "mean_token_accuracy": 0.1728737622499466,
      "num_tokens": 16950024.0,
      "step": 9180
    },
    {
      "entropy": 5.644105005264282,
      "epoch": 0.771686620457887,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004946819305369073,
      "loss": 5.3459,
      "mean_token_accuracy": 0.1768750235438347,
      "num_tokens": 16958219.0,
      "step": 9185
    },
    {
      "entropy": 5.664517211914062,
      "epoch": 0.7721067002730518,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004946754561812909,
      "loss": 5.3094,
      "mean_token_accuracy": 0.17455064058303832,
      "num_tokens": 16966829.0,
      "step": 9190
    },
    {
      "entropy": 5.656622743606567,
      "epoch": 0.7725267800882167,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004946689779342367,
      "loss": 5.409,
      "mean_token_accuracy": 0.17283372431993485,
      "num_tokens": 16975585.0,
      "step": 9195
    },
    {
      "entropy": 5.707811164855957,
      "epoch": 0.7729468599033816,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004946624957958599,
      "loss": 5.393,
      "mean_token_accuracy": 0.1730600118637085,
      "num_tokens": 16984848.0,
      "step": 9200
    },
    {
      "entropy": 5.709238576889038,
      "epoch": 0.7733669397185465,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000494656009766275,
      "loss": 5.4003,
      "mean_token_accuracy": 0.17551617324352264,
      "num_tokens": 16993179.0,
      "step": 9205
    },
    {
      "entropy": 5.632018613815307,
      "epoch": 0.7737870195337114,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000494649519845597,
      "loss": 5.4216,
      "mean_token_accuracy": 0.16661988645792009,
      "num_tokens": 17002563.0,
      "step": 9210
    },
    {
      "entropy": 5.700537967681885,
      "epoch": 0.7742070993488763,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004946430260339409,
      "loss": 5.4583,
      "mean_token_accuracy": 0.1632947728037834,
      "num_tokens": 17011805.0,
      "step": 9215
    },
    {
      "entropy": 5.723238897323609,
      "epoch": 0.7746271791640411,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004946365283314216,
      "loss": 5.3939,
      "mean_token_accuracy": 0.17818556129932403,
      "num_tokens": 17020398.0,
      "step": 9220
    },
    {
      "entropy": 5.651641130447388,
      "epoch": 0.775047258979206,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004946300267381545,
      "loss": 5.3806,
      "mean_token_accuracy": 0.17204211801290512,
      "num_tokens": 17030805.0,
      "step": 9225
    },
    {
      "entropy": 5.737525987625122,
      "epoch": 0.7754673387943709,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004946235212542544,
      "loss": 5.4191,
      "mean_token_accuracy": 0.17021374851465226,
      "num_tokens": 17040164.0,
      "step": 9230
    },
    {
      "entropy": 5.719857692718506,
      "epoch": 0.7758874186095358,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004946170118798367,
      "loss": 5.4889,
      "mean_token_accuracy": 0.167935411632061,
      "num_tokens": 17049519.0,
      "step": 9235
    },
    {
      "entropy": 5.692773962020874,
      "epoch": 0.7763074984247007,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004946104986150167,
      "loss": 5.4307,
      "mean_token_accuracy": 0.1688208520412445,
      "num_tokens": 17058042.0,
      "step": 9240
    },
    {
      "entropy": 5.705029535293579,
      "epoch": 0.7767275782398656,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004946039814599099,
      "loss": 5.4297,
      "mean_token_accuracy": 0.1713069662451744,
      "num_tokens": 17067107.0,
      "step": 9245
    },
    {
      "entropy": 5.69770622253418,
      "epoch": 0.7771476580550305,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004945974604146316,
      "loss": 5.522,
      "mean_token_accuracy": 0.17614929676055907,
      "num_tokens": 17076975.0,
      "step": 9250
    },
    {
      "entropy": 5.649022912979126,
      "epoch": 0.7775677378701953,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004945909354792974,
      "loss": 5.3712,
      "mean_token_accuracy": 0.17921458929777145,
      "num_tokens": 17086405.0,
      "step": 9255
    },
    {
      "entropy": 5.71134581565857,
      "epoch": 0.7779878176853602,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004945844066540229,
      "loss": 5.4518,
      "mean_token_accuracy": 0.16803888380527496,
      "num_tokens": 17095333.0,
      "step": 9260
    },
    {
      "entropy": 5.768996047973633,
      "epoch": 0.7784078975005251,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004945778739389236,
      "loss": 5.4946,
      "mean_token_accuracy": 0.16734545230865477,
      "num_tokens": 17103631.0,
      "step": 9265
    },
    {
      "entropy": 5.740355157852173,
      "epoch": 0.77882797731569,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004945713373341152,
      "loss": 5.404,
      "mean_token_accuracy": 0.16756715774536132,
      "num_tokens": 17112612.0,
      "step": 9270
    },
    {
      "entropy": 5.7405647277832035,
      "epoch": 0.7792480571308549,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004945647968397139,
      "loss": 5.4419,
      "mean_token_accuracy": 0.172185018658638,
      "num_tokens": 17121592.0,
      "step": 9275
    },
    {
      "entropy": 5.67391414642334,
      "epoch": 0.7796681369460198,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004945582524558352,
      "loss": 5.4437,
      "mean_token_accuracy": 0.17422326803207397,
      "num_tokens": 17131003.0,
      "step": 9280
    },
    {
      "entropy": 5.7484899997711185,
      "epoch": 0.7800882167611847,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000494551704182595,
      "loss": 5.4631,
      "mean_token_accuracy": 0.16607047468423844,
      "num_tokens": 17140013.0,
      "step": 9285
    },
    {
      "entropy": 5.83637113571167,
      "epoch": 0.7805082965763495,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004945451520201095,
      "loss": 5.577,
      "mean_token_accuracy": 0.15845307111740112,
      "num_tokens": 17150406.0,
      "step": 9290
    },
    {
      "entropy": 5.772165727615357,
      "epoch": 0.7809283763915144,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004945385959684947,
      "loss": 5.4444,
      "mean_token_accuracy": 0.17058080285787583,
      "num_tokens": 17159757.0,
      "step": 9295
    },
    {
      "entropy": 5.694557762145996,
      "epoch": 0.7813484562066793,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004945320360278667,
      "loss": 5.4657,
      "mean_token_accuracy": 0.175720676779747,
      "num_tokens": 17169317.0,
      "step": 9300
    },
    {
      "entropy": 5.752211475372315,
      "epoch": 0.7817685360218442,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004945254721983416,
      "loss": 5.4479,
      "mean_token_accuracy": 0.17670038044452668,
      "num_tokens": 17178410.0,
      "step": 9305
    },
    {
      "entropy": 5.742282390594482,
      "epoch": 0.782188615837009,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000494518904480036,
      "loss": 5.3887,
      "mean_token_accuracy": 0.17418605387210845,
      "num_tokens": 17186922.0,
      "step": 9310
    },
    {
      "entropy": 5.725687694549561,
      "epoch": 0.782608695652174,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004945123328730659,
      "loss": 5.482,
      "mean_token_accuracy": 0.16463595926761626,
      "num_tokens": 17197125.0,
      "step": 9315
    },
    {
      "entropy": 5.682478141784668,
      "epoch": 0.7830287754673388,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000494505757377548,
      "loss": 5.3982,
      "mean_token_accuracy": 0.167322738468647,
      "num_tokens": 17206169.0,
      "step": 9320
    },
    {
      "entropy": 5.6084565162658695,
      "epoch": 0.7834488552825036,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004944991779935985,
      "loss": 5.3459,
      "mean_token_accuracy": 0.17241707891225816,
      "num_tokens": 17214607.0,
      "step": 9325
    },
    {
      "entropy": 5.628190898895264,
      "epoch": 0.7838689350976685,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000494492594721334,
      "loss": 5.3265,
      "mean_token_accuracy": 0.177450430393219,
      "num_tokens": 17223616.0,
      "step": 9330
    },
    {
      "entropy": 5.755640316009521,
      "epoch": 0.7842890149128334,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004944860075608715,
      "loss": 5.4181,
      "mean_token_accuracy": 0.1696034088730812,
      "num_tokens": 17232729.0,
      "step": 9335
    },
    {
      "entropy": 5.686833667755127,
      "epoch": 0.7847090947279983,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004944794165123272,
      "loss": 5.4719,
      "mean_token_accuracy": 0.17232580333948136,
      "num_tokens": 17242128.0,
      "step": 9340
    },
    {
      "entropy": 5.6870192527771,
      "epoch": 0.7851291745431632,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000494472821575818,
      "loss": 5.368,
      "mean_token_accuracy": 0.17366537749767302,
      "num_tokens": 17250806.0,
      "step": 9345
    },
    {
      "entropy": 5.846062517166137,
      "epoch": 0.7855492543583281,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004944662227514609,
      "loss": 5.6121,
      "mean_token_accuracy": 0.15955153554677964,
      "num_tokens": 17260888.0,
      "step": 9350
    },
    {
      "entropy": 5.662187814712524,
      "epoch": 0.785969334173493,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004944596200393726,
      "loss": 5.357,
      "mean_token_accuracy": 0.17413095086812974,
      "num_tokens": 17270387.0,
      "step": 9355
    },
    {
      "entropy": 5.685818815231324,
      "epoch": 0.7863894139886578,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004944530134396702,
      "loss": 5.4039,
      "mean_token_accuracy": 0.1744739294052124,
      "num_tokens": 17279866.0,
      "step": 9360
    },
    {
      "entropy": 5.63991346359253,
      "epoch": 0.7868094938038227,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004944464029524707,
      "loss": 5.3988,
      "mean_token_accuracy": 0.17386460453271865,
      "num_tokens": 17289233.0,
      "step": 9365
    },
    {
      "entropy": 5.722629499435425,
      "epoch": 0.7872295736189876,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000494439788577891,
      "loss": 5.4679,
      "mean_token_accuracy": 0.16804894655942917,
      "num_tokens": 17298705.0,
      "step": 9370
    },
    {
      "entropy": 5.72796049118042,
      "epoch": 0.7876496534341525,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004944331703160486,
      "loss": 5.4517,
      "mean_token_accuracy": 0.16966647952795028,
      "num_tokens": 17307793.0,
      "step": 9375
    },
    {
      "entropy": 5.674992895126342,
      "epoch": 0.7880697332493174,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004944265481670605,
      "loss": 5.4987,
      "mean_token_accuracy": 0.16292951852083207,
      "num_tokens": 17318248.0,
      "step": 9380
    },
    {
      "entropy": 5.752004241943359,
      "epoch": 0.7884898130644823,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004944199221310441,
      "loss": 5.4187,
      "mean_token_accuracy": 0.164162315428257,
      "num_tokens": 17327281.0,
      "step": 9385
    },
    {
      "entropy": 5.708621597290039,
      "epoch": 0.7889098928796471,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004944132922081168,
      "loss": 5.416,
      "mean_token_accuracy": 0.17573025971651077,
      "num_tokens": 17336805.0,
      "step": 9390
    },
    {
      "entropy": 5.664907693862915,
      "epoch": 0.789329972694812,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004944066583983961,
      "loss": 5.3865,
      "mean_token_accuracy": 0.16414855122566224,
      "num_tokens": 17346024.0,
      "step": 9395
    },
    {
      "entropy": 5.728114223480224,
      "epoch": 0.7897500525099769,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004944000207019992,
      "loss": 5.4732,
      "mean_token_accuracy": 0.16647982150316237,
      "num_tokens": 17355100.0,
      "step": 9400
    },
    {
      "entropy": 5.788735866546631,
      "epoch": 0.7901701323251418,
      "grad_norm": 0.875,
      "learning_rate": 0.0004943933791190441,
      "loss": 5.4903,
      "mean_token_accuracy": 0.16659328043460847,
      "num_tokens": 17364769.0,
      "step": 9405
    },
    {
      "entropy": 5.696931028366089,
      "epoch": 0.7905902121403067,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004943867336496482,
      "loss": 5.3552,
      "mean_token_accuracy": 0.17439349442720414,
      "num_tokens": 17374082.0,
      "step": 9410
    },
    {
      "entropy": 5.62298002243042,
      "epoch": 0.7910102919554716,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004943800842939293,
      "loss": 5.407,
      "mean_token_accuracy": 0.1706312343478203,
      "num_tokens": 17383570.0,
      "step": 9415
    },
    {
      "entropy": 5.728837776184082,
      "epoch": 0.7914303717706365,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000494373431052005,
      "loss": 5.4098,
      "mean_token_accuracy": 0.1693436995148659,
      "num_tokens": 17392105.0,
      "step": 9420
    },
    {
      "entropy": 5.637105083465576,
      "epoch": 0.7918504515858013,
      "grad_norm": 1.0,
      "learning_rate": 0.0004943667739239935,
      "loss": 5.3681,
      "mean_token_accuracy": 0.17188186794519425,
      "num_tokens": 17401363.0,
      "step": 9425
    },
    {
      "entropy": 5.735306644439698,
      "epoch": 0.7922705314009661,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004943601129100125,
      "loss": 5.3833,
      "mean_token_accuracy": 0.17236204594373702,
      "num_tokens": 17411333.0,
      "step": 9430
    },
    {
      "entropy": 5.745853757858276,
      "epoch": 0.792690611216131,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004943534480101801,
      "loss": 5.4395,
      "mean_token_accuracy": 0.17388227730989456,
      "num_tokens": 17421162.0,
      "step": 9435
    },
    {
      "entropy": 5.685050010681152,
      "epoch": 0.793110691031296,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004943467792246142,
      "loss": 5.3703,
      "mean_token_accuracy": 0.1735716700553894,
      "num_tokens": 17430119.0,
      "step": 9440
    },
    {
      "entropy": 5.67172155380249,
      "epoch": 0.7935307708464608,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004943401065534332,
      "loss": 5.3936,
      "mean_token_accuracy": 0.1731569990515709,
      "num_tokens": 17439617.0,
      "step": 9445
    },
    {
      "entropy": 5.664157104492188,
      "epoch": 0.7939508506616257,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004943334299967551,
      "loss": 5.5383,
      "mean_token_accuracy": 0.16610010340809822,
      "num_tokens": 17448720.0,
      "step": 9450
    },
    {
      "entropy": 5.696771287918091,
      "epoch": 0.7943709304767906,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004943267495546982,
      "loss": 5.3921,
      "mean_token_accuracy": 0.17423522472381592,
      "num_tokens": 17457458.0,
      "step": 9455
    },
    {
      "entropy": 5.779187822341919,
      "epoch": 0.7947910102919554,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004943200652273809,
      "loss": 5.4217,
      "mean_token_accuracy": 0.17195487916469573,
      "num_tokens": 17467095.0,
      "step": 9460
    },
    {
      "entropy": 5.665433406829834,
      "epoch": 0.7952110901071203,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004943133770149216,
      "loss": 5.4473,
      "mean_token_accuracy": 0.16560591757297516,
      "num_tokens": 17476247.0,
      "step": 9465
    },
    {
      "entropy": 5.74268946647644,
      "epoch": 0.7956311699222852,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004943066849174386,
      "loss": 5.4408,
      "mean_token_accuracy": 0.178628671169281,
      "num_tokens": 17486352.0,
      "step": 9470
    },
    {
      "entropy": 5.727765035629273,
      "epoch": 0.7960512497374501,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004942999889350508,
      "loss": 5.4197,
      "mean_token_accuracy": 0.1730722039937973,
      "num_tokens": 17495633.0,
      "step": 9475
    },
    {
      "entropy": 5.76165132522583,
      "epoch": 0.796471329552615,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004942932890678765,
      "loss": 5.474,
      "mean_token_accuracy": 0.16914799362421035,
      "num_tokens": 17504325.0,
      "step": 9480
    },
    {
      "entropy": 5.749494743347168,
      "epoch": 0.7968914093677799,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004942865853160346,
      "loss": 5.469,
      "mean_token_accuracy": 0.16953212916851043,
      "num_tokens": 17513265.0,
      "step": 9485
    },
    {
      "entropy": 5.761033821105957,
      "epoch": 0.7973114891829448,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004942798776796436,
      "loss": 5.4983,
      "mean_token_accuracy": 0.16615088433027267,
      "num_tokens": 17522939.0,
      "step": 9490
    },
    {
      "entropy": 5.79576940536499,
      "epoch": 0.7977315689981096,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004942731661588226,
      "loss": 5.507,
      "mean_token_accuracy": 0.1620532378554344,
      "num_tokens": 17532250.0,
      "step": 9495
    },
    {
      "entropy": 5.834904146194458,
      "epoch": 0.7981516488132745,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004942664507536904,
      "loss": 5.5139,
      "mean_token_accuracy": 0.1728232830762863,
      "num_tokens": 17541368.0,
      "step": 9500
    },
    {
      "entropy": 5.6685301780700685,
      "epoch": 0.7985717286284394,
      "grad_norm": 1.0,
      "learning_rate": 0.0004942597314643659,
      "loss": 5.4465,
      "mean_token_accuracy": 0.16970769464969634,
      "num_tokens": 17550871.0,
      "step": 9505
    },
    {
      "entropy": 5.684246969223023,
      "epoch": 0.7989918084436043,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004942530082909681,
      "loss": 5.3733,
      "mean_token_accuracy": 0.17839082032442094,
      "num_tokens": 17559683.0,
      "step": 9510
    },
    {
      "entropy": 5.702172517776489,
      "epoch": 0.7994118882587692,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004942462812336163,
      "loss": 5.387,
      "mean_token_accuracy": 0.1743641749024391,
      "num_tokens": 17568877.0,
      "step": 9515
    },
    {
      "entropy": 5.807447338104248,
      "epoch": 0.7998319680739341,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004942395502924293,
      "loss": 5.537,
      "mean_token_accuracy": 0.16214424520730972,
      "num_tokens": 17578202.0,
      "step": 9520
    },
    {
      "entropy": 5.702708673477173,
      "epoch": 0.800252047889099,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004942328154675268,
      "loss": 5.3712,
      "mean_token_accuracy": 0.17558752745389938,
      "num_tokens": 17587342.0,
      "step": 9525
    },
    {
      "entropy": 5.675125074386597,
      "epoch": 0.8006721277042638,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004942260767590277,
      "loss": 5.2503,
      "mean_token_accuracy": 0.17630501389503478,
      "num_tokens": 17595671.0,
      "step": 9530
    },
    {
      "entropy": 5.694116020202637,
      "epoch": 0.8010922075194287,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004942193341670516,
      "loss": 5.5507,
      "mean_token_accuracy": 0.1648133561015129,
      "num_tokens": 17605649.0,
      "step": 9535
    },
    {
      "entropy": 5.7488264560699465,
      "epoch": 0.8015122873345936,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004942125876917178,
      "loss": 5.4473,
      "mean_token_accuracy": 0.16340883672237397,
      "num_tokens": 17615286.0,
      "step": 9540
    },
    {
      "entropy": 5.65223069190979,
      "epoch": 0.8019323671497585,
      "grad_norm": 0.953125,
      "learning_rate": 0.000494205837333146,
      "loss": 5.4183,
      "mean_token_accuracy": 0.17017140835523606,
      "num_tokens": 17624583.0,
      "step": 9545
    },
    {
      "entropy": 5.68720703125,
      "epoch": 0.8023524469649234,
      "grad_norm": 1.0,
      "learning_rate": 0.0004941990830914557,
      "loss": 5.4056,
      "mean_token_accuracy": 0.17712142914533616,
      "num_tokens": 17633894.0,
      "step": 9550
    },
    {
      "entropy": 5.781897449493409,
      "epoch": 0.8027725267800883,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004941923249667663,
      "loss": 5.5048,
      "mean_token_accuracy": 0.16426765769720078,
      "num_tokens": 17643172.0,
      "step": 9555
    },
    {
      "entropy": 5.735560035705566,
      "epoch": 0.803192606595253,
      "grad_norm": 1.0,
      "learning_rate": 0.0004941855629591979,
      "loss": 5.3957,
      "mean_token_accuracy": 0.17035981267690659,
      "num_tokens": 17651901.0,
      "step": 9560
    },
    {
      "entropy": 5.6513436794281,
      "epoch": 0.8036126864104179,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004941787970688701,
      "loss": 5.3832,
      "mean_token_accuracy": 0.17904664129018782,
      "num_tokens": 17660806.0,
      "step": 9565
    },
    {
      "entropy": 5.756799602508545,
      "epoch": 0.8040327662255828,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004941720272959027,
      "loss": 5.4506,
      "mean_token_accuracy": 0.17744832932949067,
      "num_tokens": 17669157.0,
      "step": 9570
    },
    {
      "entropy": 5.659553527832031,
      "epoch": 0.8044528460407477,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004941652536404157,
      "loss": 5.3538,
      "mean_token_accuracy": 0.174832521378994,
      "num_tokens": 17678664.0,
      "step": 9575
    },
    {
      "entropy": 5.6976377964019775,
      "epoch": 0.8048729258559126,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004941584761025291,
      "loss": 5.4054,
      "mean_token_accuracy": 0.16897910833358765,
      "num_tokens": 17688252.0,
      "step": 9580
    },
    {
      "entropy": 5.699000835418701,
      "epoch": 0.8052930056710775,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000494151694682363,
      "loss": 5.4238,
      "mean_token_accuracy": 0.17228419706225395,
      "num_tokens": 17696473.0,
      "step": 9585
    },
    {
      "entropy": 5.674478816986084,
      "epoch": 0.8057130854862424,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004941449093800374,
      "loss": 5.4406,
      "mean_token_accuracy": 0.1735398665070534,
      "num_tokens": 17706177.0,
      "step": 9590
    },
    {
      "entropy": 5.6636717319488525,
      "epoch": 0.8061331653014072,
      "grad_norm": 1.0,
      "learning_rate": 0.0004941381201956726,
      "loss": 5.2956,
      "mean_token_accuracy": 0.18299725353717805,
      "num_tokens": 17715355.0,
      "step": 9595
    },
    {
      "entropy": 5.655506324768067,
      "epoch": 0.8065532451165721,
      "grad_norm": 0.875,
      "learning_rate": 0.0004941313271293889,
      "loss": 5.3927,
      "mean_token_accuracy": 0.1772886872291565,
      "num_tokens": 17724345.0,
      "step": 9600
    },
    {
      "entropy": 5.658014869689941,
      "epoch": 0.806973324931737,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004941245301813065,
      "loss": 5.3252,
      "mean_token_accuracy": 0.18592872619628906,
      "num_tokens": 17732805.0,
      "step": 9605
    },
    {
      "entropy": 5.628007125854492,
      "epoch": 0.8073934047469019,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004941177293515459,
      "loss": 5.3607,
      "mean_token_accuracy": 0.1731885313987732,
      "num_tokens": 17741963.0,
      "step": 9610
    },
    {
      "entropy": 5.645999813079834,
      "epoch": 0.8078134845620668,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004941109246402275,
      "loss": 5.3907,
      "mean_token_accuracy": 0.17028539776802062,
      "num_tokens": 17751858.0,
      "step": 9615
    },
    {
      "entropy": 5.786959218978882,
      "epoch": 0.8082335643772317,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004941041160474721,
      "loss": 5.4966,
      "mean_token_accuracy": 0.16842888593673705,
      "num_tokens": 17761152.0,
      "step": 9620
    },
    {
      "entropy": 5.724298334121704,
      "epoch": 0.8086536441923966,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004940973035733999,
      "loss": 5.4355,
      "mean_token_accuracy": 0.1722586676478386,
      "num_tokens": 17770493.0,
      "step": 9625
    },
    {
      "entropy": 5.7816143989562985,
      "epoch": 0.8090737240075614,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004940904872181318,
      "loss": 5.421,
      "mean_token_accuracy": 0.16961349546909332,
      "num_tokens": 17779871.0,
      "step": 9630
    },
    {
      "entropy": 5.784200763702392,
      "epoch": 0.8094938038227263,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004940836669817887,
      "loss": 5.466,
      "mean_token_accuracy": 0.166531477868557,
      "num_tokens": 17788606.0,
      "step": 9635
    },
    {
      "entropy": 5.627583169937134,
      "epoch": 0.8099138836378912,
      "grad_norm": 1.0,
      "learning_rate": 0.0004940768428644911,
      "loss": 5.3947,
      "mean_token_accuracy": 0.17269753366708757,
      "num_tokens": 17797458.0,
      "step": 9640
    },
    {
      "entropy": 5.615399503707886,
      "epoch": 0.8103339634530561,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004940700148663601,
      "loss": 5.3243,
      "mean_token_accuracy": 0.17325042486190795,
      "num_tokens": 17806902.0,
      "step": 9645
    },
    {
      "entropy": 5.782258558273315,
      "epoch": 0.810754043268221,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004940631829875165,
      "loss": 5.4567,
      "mean_token_accuracy": 0.17239793986082078,
      "num_tokens": 17816374.0,
      "step": 9650
    },
    {
      "entropy": 5.732754325866699,
      "epoch": 0.8111741230833859,
      "grad_norm": 1.0,
      "learning_rate": 0.0004940563472280815,
      "loss": 5.4485,
      "mean_token_accuracy": 0.17549648135900497,
      "num_tokens": 17825267.0,
      "step": 9655
    },
    {
      "entropy": 5.706888723373413,
      "epoch": 0.8115942028985508,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004940495075881761,
      "loss": 5.3625,
      "mean_token_accuracy": 0.18070753812789916,
      "num_tokens": 17834027.0,
      "step": 9660
    },
    {
      "entropy": 5.641294002532959,
      "epoch": 0.8120142827137156,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004940426640679214,
      "loss": 5.3727,
      "mean_token_accuracy": 0.16692933738231658,
      "num_tokens": 17843587.0,
      "step": 9665
    },
    {
      "entropy": 5.730035972595215,
      "epoch": 0.8124343625288805,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004940358166674388,
      "loss": 5.3968,
      "mean_token_accuracy": 0.17473787367343901,
      "num_tokens": 17852284.0,
      "step": 9670
    },
    {
      "entropy": 5.752172183990479,
      "epoch": 0.8128544423440454,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004940289653868494,
      "loss": 5.4001,
      "mean_token_accuracy": 0.17687258422374724,
      "num_tokens": 17860896.0,
      "step": 9675
    },
    {
      "entropy": 5.6480035305023195,
      "epoch": 0.8132745221592103,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004940221102262747,
      "loss": 5.3986,
      "mean_token_accuracy": 0.16510132998228072,
      "num_tokens": 17870796.0,
      "step": 9680
    },
    {
      "entropy": 5.706752681732178,
      "epoch": 0.8136946019743752,
      "grad_norm": 0.875,
      "learning_rate": 0.0004940152511858361,
      "loss": 5.4655,
      "mean_token_accuracy": 0.1687447413802147,
      "num_tokens": 17880016.0,
      "step": 9685
    },
    {
      "entropy": 5.7532110691070555,
      "epoch": 0.81411468178954,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004940083882656551,
      "loss": 5.5079,
      "mean_token_accuracy": 0.16772048622369767,
      "num_tokens": 17889348.0,
      "step": 9690
    },
    {
      "entropy": 5.694590091705322,
      "epoch": 0.814534761604705,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004940015214658532,
      "loss": 5.3839,
      "mean_token_accuracy": 0.17515652626752853,
      "num_tokens": 17898392.0,
      "step": 9695
    },
    {
      "entropy": 5.745860624313354,
      "epoch": 0.8149548414198697,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004939946507865522,
      "loss": 5.4682,
      "mean_token_accuracy": 0.17255045771598815,
      "num_tokens": 17907141.0,
      "step": 9700
    },
    {
      "entropy": 5.718906402587891,
      "epoch": 0.8153749212350346,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004939877762278737,
      "loss": 5.3413,
      "mean_token_accuracy": 0.1744229093194008,
      "num_tokens": 17915792.0,
      "step": 9705
    },
    {
      "entropy": 5.721134757995605,
      "epoch": 0.8157950010501995,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004939808977899396,
      "loss": 5.4995,
      "mean_token_accuracy": 0.16537131369113922,
      "num_tokens": 17925603.0,
      "step": 9710
    },
    {
      "entropy": 5.722070932388306,
      "epoch": 0.8162150808653644,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004939740154728716,
      "loss": 5.4544,
      "mean_token_accuracy": 0.17068614512681962,
      "num_tokens": 17934436.0,
      "step": 9715
    },
    {
      "entropy": 5.724423408508301,
      "epoch": 0.8166351606805293,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004939671292767915,
      "loss": 5.396,
      "mean_token_accuracy": 0.18011061400175093,
      "num_tokens": 17942969.0,
      "step": 9720
    },
    {
      "entropy": 5.755416202545166,
      "epoch": 0.8170552404956942,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004939602392018216,
      "loss": 5.485,
      "mean_token_accuracy": 0.16732489466667175,
      "num_tokens": 17952053.0,
      "step": 9725
    },
    {
      "entropy": 5.703722524642944,
      "epoch": 0.817475320310859,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004939533452480839,
      "loss": 5.4575,
      "mean_token_accuracy": 0.17452422529459,
      "num_tokens": 17960707.0,
      "step": 9730
    },
    {
      "entropy": 5.769480848312378,
      "epoch": 0.8178954001260239,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004939464474157003,
      "loss": 5.5168,
      "mean_token_accuracy": 0.16033145487308503,
      "num_tokens": 17971035.0,
      "step": 9735
    },
    {
      "entropy": 5.719961500167846,
      "epoch": 0.8183154799411888,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004939395457047932,
      "loss": 5.3982,
      "mean_token_accuracy": 0.1691040962934494,
      "num_tokens": 17980656.0,
      "step": 9740
    },
    {
      "entropy": 5.768946886062622,
      "epoch": 0.8187355597563537,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004939326401154847,
      "loss": 5.4355,
      "mean_token_accuracy": 0.16254517287015915,
      "num_tokens": 17990977.0,
      "step": 9745
    },
    {
      "entropy": 5.655762434005737,
      "epoch": 0.8191556395715186,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004939257306478973,
      "loss": 5.4312,
      "mean_token_accuracy": 0.17181824296712875,
      "num_tokens": 18000186.0,
      "step": 9750
    },
    {
      "entropy": 5.67215929031372,
      "epoch": 0.8195757193866835,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004939188173021532,
      "loss": 5.3988,
      "mean_token_accuracy": 0.16897007077932358,
      "num_tokens": 18010269.0,
      "step": 9755
    },
    {
      "entropy": 5.696268558502197,
      "epoch": 0.8199957992018484,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004939119000783751,
      "loss": 5.3569,
      "mean_token_accuracy": 0.1821137249469757,
      "num_tokens": 18018461.0,
      "step": 9760
    },
    {
      "entropy": 5.670961475372314,
      "epoch": 0.8204158790170132,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004939049789766855,
      "loss": 5.3623,
      "mean_token_accuracy": 0.16979527473449707,
      "num_tokens": 18027173.0,
      "step": 9765
    },
    {
      "entropy": 5.681652688980103,
      "epoch": 0.8208359588321781,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004938980539972068,
      "loss": 5.4673,
      "mean_token_accuracy": 0.1694917768239975,
      "num_tokens": 18036791.0,
      "step": 9770
    },
    {
      "entropy": 5.661679315567016,
      "epoch": 0.821256038647343,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004938911251400617,
      "loss": 5.4079,
      "mean_token_accuracy": 0.1746845379471779,
      "num_tokens": 18046908.0,
      "step": 9775
    },
    {
      "entropy": 5.641901636123658,
      "epoch": 0.8216761184625079,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004938841924053731,
      "loss": 5.3413,
      "mean_token_accuracy": 0.18411260843276978,
      "num_tokens": 18055825.0,
      "step": 9780
    },
    {
      "entropy": 5.784076595306397,
      "epoch": 0.8220961982776728,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004938772557932637,
      "loss": 5.4615,
      "mean_token_accuracy": 0.16495970487594605,
      "num_tokens": 18065334.0,
      "step": 9785
    },
    {
      "entropy": 5.7153181552886965,
      "epoch": 0.8225162780928377,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004938703153038565,
      "loss": 5.3533,
      "mean_token_accuracy": 0.17671273052692413,
      "num_tokens": 18073999.0,
      "step": 9790
    },
    {
      "entropy": 5.61739821434021,
      "epoch": 0.8229363579080026,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004938633709372744,
      "loss": 5.3993,
      "mean_token_accuracy": 0.16987203061580658,
      "num_tokens": 18083665.0,
      "step": 9795
    },
    {
      "entropy": 5.6752242088317875,
      "epoch": 0.8233564377231674,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004938564226936403,
      "loss": 5.4013,
      "mean_token_accuracy": 0.17371127158403396,
      "num_tokens": 18092501.0,
      "step": 9800
    },
    {
      "entropy": 5.692128038406372,
      "epoch": 0.8237765175383323,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004938494705730773,
      "loss": 5.3686,
      "mean_token_accuracy": 0.174516025185585,
      "num_tokens": 18101320.0,
      "step": 9805
    },
    {
      "entropy": 5.686312437057495,
      "epoch": 0.8241965973534972,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004938425145757087,
      "loss": 5.4227,
      "mean_token_accuracy": 0.166843281686306,
      "num_tokens": 18110190.0,
      "step": 9810
    },
    {
      "entropy": 5.692755556106567,
      "epoch": 0.824616677168662,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004938355547016577,
      "loss": 5.3989,
      "mean_token_accuracy": 0.17334626764059066,
      "num_tokens": 18119301.0,
      "step": 9815
    },
    {
      "entropy": 5.767399263381958,
      "epoch": 0.825036756983827,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004938285909510474,
      "loss": 5.4607,
      "mean_token_accuracy": 0.1654415801167488,
      "num_tokens": 18128959.0,
      "step": 9820
    },
    {
      "entropy": 5.711994314193726,
      "epoch": 0.8254568367989918,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004938216233240014,
      "loss": 5.4367,
      "mean_token_accuracy": 0.17324853539466858,
      "num_tokens": 18138156.0,
      "step": 9825
    },
    {
      "entropy": 5.7775660991668705,
      "epoch": 0.8258769166141567,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000493814651820643,
      "loss": 5.4647,
      "mean_token_accuracy": 0.16314569264650344,
      "num_tokens": 18147244.0,
      "step": 9830
    },
    {
      "entropy": 5.759236145019531,
      "epoch": 0.8262969964293215,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004938076764410956,
      "loss": 5.4541,
      "mean_token_accuracy": 0.17017538100481033,
      "num_tokens": 18156040.0,
      "step": 9835
    },
    {
      "entropy": 5.712372207641602,
      "epoch": 0.8267170762444864,
      "grad_norm": 1.046875,
      "learning_rate": 0.000493800697185483,
      "loss": 5.3891,
      "mean_token_accuracy": 0.16826028227806092,
      "num_tokens": 18165210.0,
      "step": 9840
    },
    {
      "entropy": 5.745231628417969,
      "epoch": 0.8271371560596513,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004937937140539288,
      "loss": 5.4594,
      "mean_token_accuracy": 0.1697965607047081,
      "num_tokens": 18174841.0,
      "step": 9845
    },
    {
      "entropy": 5.626189851760865,
      "epoch": 0.8275572358748162,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004937867270465564,
      "loss": 5.3318,
      "mean_token_accuracy": 0.1725539356470108,
      "num_tokens": 18184112.0,
      "step": 9850
    },
    {
      "entropy": 5.779374408721924,
      "epoch": 0.8279773156899811,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004937797361634899,
      "loss": 5.5208,
      "mean_token_accuracy": 0.16566380262374877,
      "num_tokens": 18193564.0,
      "step": 9855
    },
    {
      "entropy": 5.639378690719605,
      "epoch": 0.828397395505146,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000493772741404853,
      "loss": 5.2572,
      "mean_token_accuracy": 0.18007946759462357,
      "num_tokens": 18202836.0,
      "step": 9860
    },
    {
      "entropy": 5.678477144241333,
      "epoch": 0.8288174753203108,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004937657427707698,
      "loss": 5.3962,
      "mean_token_accuracy": 0.17871991395950318,
      "num_tokens": 18212098.0,
      "step": 9865
    },
    {
      "entropy": 5.70985894203186,
      "epoch": 0.8292375551354757,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004937587402613639,
      "loss": 5.4172,
      "mean_token_accuracy": 0.1705652341246605,
      "num_tokens": 18221541.0,
      "step": 9870
    },
    {
      "entropy": 5.68948221206665,
      "epoch": 0.8296576349506406,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004937517338767597,
      "loss": 5.4225,
      "mean_token_accuracy": 0.16736140549182893,
      "num_tokens": 18231015.0,
      "step": 9875
    },
    {
      "entropy": 5.737211036682129,
      "epoch": 0.8300777147658055,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004937447236170811,
      "loss": 5.4395,
      "mean_token_accuracy": 0.17107924073934555,
      "num_tokens": 18239729.0,
      "step": 9880
    },
    {
      "entropy": 5.748801898956299,
      "epoch": 0.8304977945809704,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004937377094824523,
      "loss": 5.4811,
      "mean_token_accuracy": 0.17103148773312568,
      "num_tokens": 18249773.0,
      "step": 9885
    },
    {
      "entropy": 5.725286960601807,
      "epoch": 0.8309178743961353,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004937306914729977,
      "loss": 5.4373,
      "mean_token_accuracy": 0.169718599319458,
      "num_tokens": 18259179.0,
      "step": 9890
    },
    {
      "entropy": 5.6264832496643065,
      "epoch": 0.8313379542113002,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004937236695888416,
      "loss": 5.3355,
      "mean_token_accuracy": 0.1819911703467369,
      "num_tokens": 18268164.0,
      "step": 9895
    },
    {
      "entropy": 5.717627286911011,
      "epoch": 0.831758034026465,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004937166438301082,
      "loss": 5.5021,
      "mean_token_accuracy": 0.16980272978544236,
      "num_tokens": 18276259.0,
      "step": 9900
    },
    {
      "entropy": 5.691333293914795,
      "epoch": 0.8321781138416299,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004937096141969221,
      "loss": 5.4657,
      "mean_token_accuracy": 0.1689844086766243,
      "num_tokens": 18285729.0,
      "step": 9905
    },
    {
      "entropy": 5.81794753074646,
      "epoch": 0.8325981936567948,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004937025806894077,
      "loss": 5.62,
      "mean_token_accuracy": 0.15900368988513947,
      "num_tokens": 18295873.0,
      "step": 9910
    },
    {
      "entropy": 5.767088890075684,
      "epoch": 0.8330182734719597,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004936955433076899,
      "loss": 5.4494,
      "mean_token_accuracy": 0.17725514471530915,
      "num_tokens": 18305135.0,
      "step": 9915
    },
    {
      "entropy": 5.783304357528687,
      "epoch": 0.8334383532871246,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000493688502051893,
      "loss": 5.4901,
      "mean_token_accuracy": 0.1678687147796154,
      "num_tokens": 18314251.0,
      "step": 9920
    },
    {
      "entropy": 5.6946399211883545,
      "epoch": 0.8338584331022895,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004936814569221421,
      "loss": 5.331,
      "mean_token_accuracy": 0.18811306059360505,
      "num_tokens": 18322863.0,
      "step": 9925
    },
    {
      "entropy": 5.615930509567261,
      "epoch": 0.8342785129174544,
      "grad_norm": 1.0,
      "learning_rate": 0.0004936744079185616,
      "loss": 5.3413,
      "mean_token_accuracy": 0.17328823506832122,
      "num_tokens": 18332129.0,
      "step": 9930
    },
    {
      "entropy": 5.724304056167602,
      "epoch": 0.8346985927326191,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004936673550412767,
      "loss": 5.443,
      "mean_token_accuracy": 0.16912543028593063,
      "num_tokens": 18341457.0,
      "step": 9935
    },
    {
      "entropy": 5.72011365890503,
      "epoch": 0.835118672547784,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000493660298290412,
      "loss": 5.4188,
      "mean_token_accuracy": 0.16655368953943253,
      "num_tokens": 18351397.0,
      "step": 9940
    },
    {
      "entropy": 5.688337135314941,
      "epoch": 0.8355387523629489,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004936532376660929,
      "loss": 5.3668,
      "mean_token_accuracy": 0.1733811840415001,
      "num_tokens": 18360005.0,
      "step": 9945
    },
    {
      "entropy": 5.740257215499878,
      "epoch": 0.8359588321781138,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004936461731684442,
      "loss": 5.449,
      "mean_token_accuracy": 0.17170872688293456,
      "num_tokens": 18369707.0,
      "step": 9950
    },
    {
      "entropy": 5.800839757919311,
      "epoch": 0.8363789119932787,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004936391047975912,
      "loss": 5.5688,
      "mean_token_accuracy": 0.16948766261339188,
      "num_tokens": 18379514.0,
      "step": 9955
    },
    {
      "entropy": 5.70206618309021,
      "epoch": 0.8367989918084436,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004936320325536589,
      "loss": 5.2949,
      "mean_token_accuracy": 0.17217475324869155,
      "num_tokens": 18388854.0,
      "step": 9960
    },
    {
      "entropy": 5.736205625534057,
      "epoch": 0.8372190716236085,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004936249564367729,
      "loss": 5.4516,
      "mean_token_accuracy": 0.17096713334321975,
      "num_tokens": 18397806.0,
      "step": 9965
    },
    {
      "entropy": 5.628139543533325,
      "epoch": 0.8376391514387733,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004936178764470583,
      "loss": 5.3109,
      "mean_token_accuracy": 0.1726173624396324,
      "num_tokens": 18406645.0,
      "step": 9970
    },
    {
      "entropy": 5.614950513839721,
      "epoch": 0.8380592312539382,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004936107925846405,
      "loss": 5.3389,
      "mean_token_accuracy": 0.17455190271139145,
      "num_tokens": 18415730.0,
      "step": 9975
    },
    {
      "entropy": 5.713989686965943,
      "epoch": 0.8384793110691031,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004936037048496452,
      "loss": 5.4588,
      "mean_token_accuracy": 0.17137794941663742,
      "num_tokens": 18424638.0,
      "step": 9980
    },
    {
      "entropy": 5.690193700790405,
      "epoch": 0.838899390884268,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004935966132421977,
      "loss": 5.4871,
      "mean_token_accuracy": 0.16778844147920607,
      "num_tokens": 18434090.0,
      "step": 9985
    },
    {
      "entropy": 5.652533102035522,
      "epoch": 0.8393194706994329,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004935895177624239,
      "loss": 5.3172,
      "mean_token_accuracy": 0.1746213749051094,
      "num_tokens": 18442965.0,
      "step": 9990
    },
    {
      "entropy": 5.743184900283813,
      "epoch": 0.8397395505145978,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004935824184104493,
      "loss": 5.3479,
      "mean_token_accuracy": 0.1808331713080406,
      "num_tokens": 18451553.0,
      "step": 9995
    },
    {
      "entropy": 5.66016788482666,
      "epoch": 0.8401596303297627,
      "grad_norm": 1.0,
      "learning_rate": 0.0004935753151863997,
      "loss": 5.4178,
      "mean_token_accuracy": 0.16960439532995225,
      "num_tokens": 18461325.0,
      "step": 10000
    },
    {
      "entropy": 5.70008807182312,
      "epoch": 0.8405797101449275,
      "grad_norm": 1.0,
      "learning_rate": 0.0004935682080904009,
      "loss": 5.4087,
      "mean_token_accuracy": 0.1788218528032303,
      "num_tokens": 18469977.0,
      "step": 10005
    },
    {
      "entropy": 5.745610094070434,
      "epoch": 0.8409997899600924,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004935610971225789,
      "loss": 5.379,
      "mean_token_accuracy": 0.1749153032898903,
      "num_tokens": 18479534.0,
      "step": 10010
    },
    {
      "entropy": 5.6720983505249025,
      "epoch": 0.8414198697752573,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004935539822830597,
      "loss": 5.483,
      "mean_token_accuracy": 0.16612620204687117,
      "num_tokens": 18488800.0,
      "step": 10015
    },
    {
      "entropy": 5.728128814697266,
      "epoch": 0.8418399495904222,
      "grad_norm": 1.078125,
      "learning_rate": 0.000493546863571969,
      "loss": 5.4392,
      "mean_token_accuracy": 0.17366082668304444,
      "num_tokens": 18498083.0,
      "step": 10020
    },
    {
      "entropy": 5.703526878356934,
      "epoch": 0.8422600294055871,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004935397409894333,
      "loss": 5.4009,
      "mean_token_accuracy": 0.16526015102863312,
      "num_tokens": 18508265.0,
      "step": 10025
    },
    {
      "entropy": 5.7603973865509035,
      "epoch": 0.842680109220752,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004935326145355787,
      "loss": 5.4195,
      "mean_token_accuracy": 0.1707303047180176,
      "num_tokens": 18517283.0,
      "step": 10030
    },
    {
      "entropy": 5.740194511413574,
      "epoch": 0.8431001890359168,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004935254842105311,
      "loss": 5.4489,
      "mean_token_accuracy": 0.17193007469177246,
      "num_tokens": 18526482.0,
      "step": 10035
    },
    {
      "entropy": 5.616510200500488,
      "epoch": 0.8435202688510817,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004935183500144173,
      "loss": 5.271,
      "mean_token_accuracy": 0.189638751745224,
      "num_tokens": 18536150.0,
      "step": 10040
    },
    {
      "entropy": 5.7846485614776615,
      "epoch": 0.8439403486662466,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004935112119473634,
      "loss": 5.4776,
      "mean_token_accuracy": 0.16828864365816115,
      "num_tokens": 18545168.0,
      "step": 10045
    },
    {
      "entropy": 5.724568033218384,
      "epoch": 0.8443604284814115,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004935040700094959,
      "loss": 5.4124,
      "mean_token_accuracy": 0.17468920797109605,
      "num_tokens": 18553363.0,
      "step": 10050
    },
    {
      "entropy": 5.631190776824951,
      "epoch": 0.8447805082965764,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004934969242009412,
      "loss": 5.3439,
      "mean_token_accuracy": 0.17899968475103378,
      "num_tokens": 18562546.0,
      "step": 10055
    },
    {
      "entropy": 5.665628671646118,
      "epoch": 0.8452005881117413,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004934897745218262,
      "loss": 5.4272,
      "mean_token_accuracy": 0.17086587101221085,
      "num_tokens": 18572149.0,
      "step": 10060
    },
    {
      "entropy": 5.713841772079467,
      "epoch": 0.8456206679269062,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004934826209722772,
      "loss": 5.3109,
      "mean_token_accuracy": 0.1781184881925583,
      "num_tokens": 18580842.0,
      "step": 10065
    },
    {
      "entropy": 5.743235778808594,
      "epoch": 0.8460407477420709,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004934754635524211,
      "loss": 5.4287,
      "mean_token_accuracy": 0.1735612466931343,
      "num_tokens": 18589765.0,
      "step": 10070
    },
    {
      "entropy": 5.703540706634522,
      "epoch": 0.8464608275572358,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004934683022623847,
      "loss": 5.4011,
      "mean_token_accuracy": 0.16834982633590698,
      "num_tokens": 18599532.0,
      "step": 10075
    },
    {
      "entropy": 5.630594253540039,
      "epoch": 0.8468809073724007,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004934611371022947,
      "loss": 5.324,
      "mean_token_accuracy": 0.1759598508477211,
      "num_tokens": 18608438.0,
      "step": 10080
    },
    {
      "entropy": 5.744466638565063,
      "epoch": 0.8473009871875656,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004934539680722783,
      "loss": 5.4663,
      "mean_token_accuracy": 0.16937509179115295,
      "num_tokens": 18617313.0,
      "step": 10085
    },
    {
      "entropy": 5.679561805725098,
      "epoch": 0.8477210670027305,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004934467951724622,
      "loss": 5.3002,
      "mean_token_accuracy": 0.17695471048355102,
      "num_tokens": 18625880.0,
      "step": 10090
    },
    {
      "entropy": 5.6911262512207035,
      "epoch": 0.8481411468178954,
      "grad_norm": 0.875,
      "learning_rate": 0.0004934396184029737,
      "loss": 5.3777,
      "mean_token_accuracy": 0.17113883197307586,
      "num_tokens": 18635727.0,
      "step": 10095
    },
    {
      "entropy": 5.7253765106201175,
      "epoch": 0.8485612266330603,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004934324377639398,
      "loss": 5.4268,
      "mean_token_accuracy": 0.17339333593845369,
      "num_tokens": 18645619.0,
      "step": 10100
    },
    {
      "entropy": 5.651005363464355,
      "epoch": 0.8489813064482251,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004934252532554878,
      "loss": 5.3499,
      "mean_token_accuracy": 0.17403531968593597,
      "num_tokens": 18654901.0,
      "step": 10105
    },
    {
      "entropy": 5.824486207962036,
      "epoch": 0.84940138626339,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004934180648777449,
      "loss": 5.5813,
      "mean_token_accuracy": 0.1697644844651222,
      "num_tokens": 18664523.0,
      "step": 10110
    },
    {
      "entropy": 5.764934015274048,
      "epoch": 0.8498214660785549,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004934108726308384,
      "loss": 5.4394,
      "mean_token_accuracy": 0.16460866928100587,
      "num_tokens": 18673685.0,
      "step": 10115
    },
    {
      "entropy": 5.695920705795288,
      "epoch": 0.8502415458937198,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004934036765148958,
      "loss": 5.3921,
      "mean_token_accuracy": 0.17187559455633164,
      "num_tokens": 18682889.0,
      "step": 10120
    },
    {
      "entropy": 5.723449754714966,
      "epoch": 0.8506616257088847,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004933964765300446,
      "loss": 5.4381,
      "mean_token_accuracy": 0.16791050732135773,
      "num_tokens": 18692978.0,
      "step": 10125
    },
    {
      "entropy": 5.692808341979981,
      "epoch": 0.8510817055240496,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000493389272676412,
      "loss": 5.3531,
      "mean_token_accuracy": 0.17797447592020035,
      "num_tokens": 18701846.0,
      "step": 10130
    },
    {
      "entropy": 5.675617122650147,
      "epoch": 0.8515017853392145,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004933820649541262,
      "loss": 5.3984,
      "mean_token_accuracy": 0.1820225700736046,
      "num_tokens": 18711492.0,
      "step": 10135
    },
    {
      "entropy": 5.657948780059814,
      "epoch": 0.8519218651543793,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004933748533633145,
      "loss": 5.3219,
      "mean_token_accuracy": 0.18281747102737428,
      "num_tokens": 18720407.0,
      "step": 10140
    },
    {
      "entropy": 5.682277488708496,
      "epoch": 0.8523419449695442,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004933676379041045,
      "loss": 5.3795,
      "mean_token_accuracy": 0.1766313672065735,
      "num_tokens": 18729968.0,
      "step": 10145
    },
    {
      "entropy": 5.737059164047241,
      "epoch": 0.8527620247847091,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004933604185766245,
      "loss": 5.4905,
      "mean_token_accuracy": 0.1653348386287689,
      "num_tokens": 18739525.0,
      "step": 10150
    },
    {
      "entropy": 5.687533521652222,
      "epoch": 0.853182104599874,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004933531953810019,
      "loss": 5.3674,
      "mean_token_accuracy": 0.18290079087018968,
      "num_tokens": 18749087.0,
      "step": 10155
    },
    {
      "entropy": 5.753673076629639,
      "epoch": 0.8536021844150389,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004933459683173652,
      "loss": 5.4062,
      "mean_token_accuracy": 0.17374185025691985,
      "num_tokens": 18758174.0,
      "step": 10160
    },
    {
      "entropy": 5.7686608791351315,
      "epoch": 0.8540222642302038,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004933387373858418,
      "loss": 5.4349,
      "mean_token_accuracy": 0.17194124907255173,
      "num_tokens": 18767679.0,
      "step": 10165
    },
    {
      "entropy": 5.679949474334717,
      "epoch": 0.8544423440453687,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004933315025865602,
      "loss": 5.3765,
      "mean_token_accuracy": 0.16869062632322313,
      "num_tokens": 18776749.0,
      "step": 10170
    },
    {
      "entropy": 5.745608186721801,
      "epoch": 0.8548624238605335,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004933242639196485,
      "loss": 5.5565,
      "mean_token_accuracy": 0.1623409777879715,
      "num_tokens": 18786313.0,
      "step": 10175
    },
    {
      "entropy": 5.816283702850342,
      "epoch": 0.8552825036756984,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004933170213852348,
      "loss": 5.4356,
      "mean_token_accuracy": 0.1684453934431076,
      "num_tokens": 18795340.0,
      "step": 10180
    },
    {
      "entropy": 5.6982838153839115,
      "epoch": 0.8557025834908633,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004933097749834476,
      "loss": 5.3559,
      "mean_token_accuracy": 0.16831685453653336,
      "num_tokens": 18804114.0,
      "step": 10185
    },
    {
      "entropy": 5.660799646377564,
      "epoch": 0.8561226633060282,
      "grad_norm": 0.921875,
      "learning_rate": 0.000493302524714415,
      "loss": 5.3623,
      "mean_token_accuracy": 0.16940130889415742,
      "num_tokens": 18813797.0,
      "step": 10190
    },
    {
      "entropy": 5.701170539855957,
      "epoch": 0.856542743121193,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004932952705782657,
      "loss": 5.4174,
      "mean_token_accuracy": 0.16815407276153566,
      "num_tokens": 18822410.0,
      "step": 10195
    },
    {
      "entropy": 5.650213670730591,
      "epoch": 0.856962822936358,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000493288012575128,
      "loss": 5.3544,
      "mean_token_accuracy": 0.17864344418048858,
      "num_tokens": 18832091.0,
      "step": 10200
    },
    {
      "entropy": 5.691934537887573,
      "epoch": 0.8573829027515227,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004932807507051307,
      "loss": 5.3797,
      "mean_token_accuracy": 0.16768632382154464,
      "num_tokens": 18841298.0,
      "step": 10205
    },
    {
      "entropy": 5.661048746109008,
      "epoch": 0.8578029825666876,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004932734849684022,
      "loss": 5.3532,
      "mean_token_accuracy": 0.1751982182264328,
      "num_tokens": 18849683.0,
      "step": 10210
    },
    {
      "entropy": 5.695506620407104,
      "epoch": 0.8582230623818525,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004932662153650712,
      "loss": 5.2941,
      "mean_token_accuracy": 0.1761276513338089,
      "num_tokens": 18858832.0,
      "step": 10215
    },
    {
      "entropy": 5.620655202865601,
      "epoch": 0.8586431421970174,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004932589418952668,
      "loss": 5.3416,
      "mean_token_accuracy": 0.17267124354839325,
      "num_tokens": 18867652.0,
      "step": 10220
    },
    {
      "entropy": 5.711537647247314,
      "epoch": 0.8590632220121823,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004932516645591175,
      "loss": 5.4125,
      "mean_token_accuracy": 0.17468912303447723,
      "num_tokens": 18877282.0,
      "step": 10225
    },
    {
      "entropy": 5.7473063468933105,
      "epoch": 0.8594833018273472,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004932443833567524,
      "loss": 5.5401,
      "mean_token_accuracy": 0.16876454651355743,
      "num_tokens": 18886565.0,
      "step": 10230
    },
    {
      "entropy": 5.756016969680786,
      "epoch": 0.8599033816425121,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004932370982883003,
      "loss": 5.4478,
      "mean_token_accuracy": 0.17323363274335862,
      "num_tokens": 18896440.0,
      "step": 10235
    },
    {
      "entropy": 5.771557998657227,
      "epoch": 0.8603234614576769,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004932298093538905,
      "loss": 5.4944,
      "mean_token_accuracy": 0.16832654625177385,
      "num_tokens": 18906246.0,
      "step": 10240
    },
    {
      "entropy": 5.659944868087768,
      "epoch": 0.8607435412728418,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000493222516553652,
      "loss": 5.3917,
      "mean_token_accuracy": 0.16731180995702744,
      "num_tokens": 18915108.0,
      "step": 10245
    },
    {
      "entropy": 5.711893224716187,
      "epoch": 0.8611636210880067,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004932152198877139,
      "loss": 5.3829,
      "mean_token_accuracy": 0.1741176500916481,
      "num_tokens": 18923664.0,
      "step": 10250
    },
    {
      "entropy": 5.729861831665039,
      "epoch": 0.8615837009031716,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004932079193562057,
      "loss": 5.4773,
      "mean_token_accuracy": 0.1685989648103714,
      "num_tokens": 18933496.0,
      "step": 10255
    },
    {
      "entropy": 5.679082345962525,
      "epoch": 0.8620037807183365,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004932006149592564,
      "loss": 5.3506,
      "mean_token_accuracy": 0.1752501279115677,
      "num_tokens": 18942222.0,
      "step": 10260
    },
    {
      "entropy": 5.763818883895874,
      "epoch": 0.8624238605335014,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004931933066969957,
      "loss": 5.3724,
      "mean_token_accuracy": 0.17565049976110458,
      "num_tokens": 18952057.0,
      "step": 10265
    },
    {
      "entropy": 5.715308570861817,
      "epoch": 0.8628439403486663,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004931859945695528,
      "loss": 5.4405,
      "mean_token_accuracy": 0.16643152832984925,
      "num_tokens": 18961664.0,
      "step": 10270
    },
    {
      "entropy": 5.591107273101807,
      "epoch": 0.8632640201638311,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004931786785770575,
      "loss": 5.2206,
      "mean_token_accuracy": 0.187975712120533,
      "num_tokens": 18969900.0,
      "step": 10275
    },
    {
      "entropy": 5.726562786102295,
      "epoch": 0.863684099978996,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004931713587196392,
      "loss": 5.5124,
      "mean_token_accuracy": 0.16591396033763886,
      "num_tokens": 18979286.0,
      "step": 10280
    },
    {
      "entropy": 5.792297124862671,
      "epoch": 0.8641041797941609,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004931640349974275,
      "loss": 5.4034,
      "mean_token_accuracy": 0.1727977752685547,
      "num_tokens": 18987553.0,
      "step": 10285
    },
    {
      "entropy": 5.744881200790405,
      "epoch": 0.8645242596093258,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004931567074105524,
      "loss": 5.4668,
      "mean_token_accuracy": 0.17214933782815933,
      "num_tokens": 18996354.0,
      "step": 10290
    },
    {
      "entropy": 5.652358627319336,
      "epoch": 0.8649443394244907,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004931493759591435,
      "loss": 5.355,
      "mean_token_accuracy": 0.17563411891460418,
      "num_tokens": 19005150.0,
      "step": 10295
    },
    {
      "entropy": 5.733958196640015,
      "epoch": 0.8653644192396556,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004931420406433308,
      "loss": 5.3935,
      "mean_token_accuracy": 0.16878109872341157,
      "num_tokens": 19014572.0,
      "step": 10300
    },
    {
      "entropy": 5.678887748718262,
      "epoch": 0.8657844990548205,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000493134701463244,
      "loss": 5.232,
      "mean_token_accuracy": 0.18580519855022432,
      "num_tokens": 19023462.0,
      "step": 10305
    },
    {
      "entropy": 5.612130975723266,
      "epoch": 0.8662045788699853,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004931273584190135,
      "loss": 5.3436,
      "mean_token_accuracy": 0.1780797243118286,
      "num_tokens": 19032460.0,
      "step": 10310
    },
    {
      "entropy": 5.6639501571655275,
      "epoch": 0.8666246586851502,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004931200115107691,
      "loss": 5.3708,
      "mean_token_accuracy": 0.17969484329223634,
      "num_tokens": 19041734.0,
      "step": 10315
    },
    {
      "entropy": 5.681323003768921,
      "epoch": 0.867044738500315,
      "grad_norm": 1.03125,
      "learning_rate": 0.000493112660738641,
      "loss": 5.3458,
      "mean_token_accuracy": 0.16926186978816987,
      "num_tokens": 19050867.0,
      "step": 10320
    },
    {
      "entropy": 5.6649250984191895,
      "epoch": 0.86746481831548,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004931053061027594,
      "loss": 5.3459,
      "mean_token_accuracy": 0.17134073823690416,
      "num_tokens": 19060518.0,
      "step": 10325
    },
    {
      "entropy": 5.67543215751648,
      "epoch": 0.8678848981306448,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004930979476032546,
      "loss": 5.344,
      "mean_token_accuracy": 0.17601591795682908,
      "num_tokens": 19069588.0,
      "step": 10330
    },
    {
      "entropy": 5.676546955108643,
      "epoch": 0.8683049779458097,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000493090585240257,
      "loss": 5.3927,
      "mean_token_accuracy": 0.16369630247354508,
      "num_tokens": 19079060.0,
      "step": 10335
    },
    {
      "entropy": 5.676814413070678,
      "epoch": 0.8687250577609746,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004930832190138969,
      "loss": 5.3202,
      "mean_token_accuracy": 0.16793182939291001,
      "num_tokens": 19087721.0,
      "step": 10340
    },
    {
      "entropy": 5.669615268707275,
      "epoch": 0.8691451375761394,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000493075848924305,
      "loss": 5.3669,
      "mean_token_accuracy": 0.1751513212919235,
      "num_tokens": 19096800.0,
      "step": 10345
    },
    {
      "entropy": 5.647520875930786,
      "epoch": 0.8695652173913043,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004930684749716117,
      "loss": 5.4278,
      "mean_token_accuracy": 0.16903601884841918,
      "num_tokens": 19106774.0,
      "step": 10350
    },
    {
      "entropy": 5.700551700592041,
      "epoch": 0.8699852972064692,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004930610971559476,
      "loss": 5.3463,
      "mean_token_accuracy": 0.17723934203386307,
      "num_tokens": 19116413.0,
      "step": 10355
    },
    {
      "entropy": 5.721123695373535,
      "epoch": 0.8704053770216341,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004930537154774436,
      "loss": 5.3839,
      "mean_token_accuracy": 0.17043269276618958,
      "num_tokens": 19125363.0,
      "step": 10360
    },
    {
      "entropy": 5.732427453994751,
      "epoch": 0.870825456836799,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004930463299362302,
      "loss": 5.4615,
      "mean_token_accuracy": 0.1629556119441986,
      "num_tokens": 19135461.0,
      "step": 10365
    },
    {
      "entropy": 5.7652373790740965,
      "epoch": 0.8712455366519639,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004930389405324383,
      "loss": 5.3359,
      "mean_token_accuracy": 0.1800934851169586,
      "num_tokens": 19144085.0,
      "step": 10370
    },
    {
      "entropy": 5.7150733947753904,
      "epoch": 0.8716656164671287,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004930315472661987,
      "loss": 5.3468,
      "mean_token_accuracy": 0.17935262322425843,
      "num_tokens": 19153291.0,
      "step": 10375
    },
    {
      "entropy": 5.635470867156982,
      "epoch": 0.8720856962822936,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004930241501376428,
      "loss": 5.365,
      "mean_token_accuracy": 0.17207663059234618,
      "num_tokens": 19163514.0,
      "step": 10380
    },
    {
      "entropy": 5.588260221481323,
      "epoch": 0.8725057760974585,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004930167491469013,
      "loss": 5.2557,
      "mean_token_accuracy": 0.178853639960289,
      "num_tokens": 19172103.0,
      "step": 10385
    },
    {
      "entropy": 5.714194631576538,
      "epoch": 0.8729258559126234,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004930093442941053,
      "loss": 5.3405,
      "mean_token_accuracy": 0.1713276445865631,
      "num_tokens": 19180893.0,
      "step": 10390
    },
    {
      "entropy": 5.641024541854859,
      "epoch": 0.8733459357277883,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004930019355793858,
      "loss": 5.2755,
      "mean_token_accuracy": 0.17605988532304764,
      "num_tokens": 19190495.0,
      "step": 10395
    },
    {
      "entropy": 5.648091554641724,
      "epoch": 0.8737660155429532,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004929945230028746,
      "loss": 5.3618,
      "mean_token_accuracy": 0.18059074580669404,
      "num_tokens": 19198988.0,
      "step": 10400
    },
    {
      "entropy": 5.6576454639434814,
      "epoch": 0.8741860953581181,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004929871065647024,
      "loss": 5.2684,
      "mean_token_accuracy": 0.17943289875984192,
      "num_tokens": 19208014.0,
      "step": 10405
    },
    {
      "entropy": 5.729404544830322,
      "epoch": 0.8746061751732829,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004929796862650011,
      "loss": 5.4613,
      "mean_token_accuracy": 0.1700380802154541,
      "num_tokens": 19218220.0,
      "step": 10410
    },
    {
      "entropy": 5.650667810440064,
      "epoch": 0.8750262549884478,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004929722621039018,
      "loss": 5.3576,
      "mean_token_accuracy": 0.17190681844949723,
      "num_tokens": 19227176.0,
      "step": 10415
    },
    {
      "entropy": 5.633760738372803,
      "epoch": 0.8754463348036127,
      "grad_norm": 1.0,
      "learning_rate": 0.0004929648340815362,
      "loss": 5.3915,
      "mean_token_accuracy": 0.16977670639753342,
      "num_tokens": 19236085.0,
      "step": 10420
    },
    {
      "entropy": 5.72749137878418,
      "epoch": 0.8758664146187776,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004929574021980355,
      "loss": 5.4259,
      "mean_token_accuracy": 0.16803160458803176,
      "num_tokens": 19246671.0,
      "step": 10425
    },
    {
      "entropy": 5.692175722122192,
      "epoch": 0.8762864944339425,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004929499664535319,
      "loss": 5.3476,
      "mean_token_accuracy": 0.17221028804779054,
      "num_tokens": 19256321.0,
      "step": 10430
    },
    {
      "entropy": 5.651491928100586,
      "epoch": 0.8767065742491074,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004929425268481569,
      "loss": 5.3056,
      "mean_token_accuracy": 0.17936076670885087,
      "num_tokens": 19265518.0,
      "step": 10435
    },
    {
      "entropy": 5.648326396942139,
      "epoch": 0.8771266540642723,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004929350833820422,
      "loss": 5.3195,
      "mean_token_accuracy": 0.17637630552053452,
      "num_tokens": 19274120.0,
      "step": 10440
    },
    {
      "entropy": 5.665770435333252,
      "epoch": 0.877546733879437,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004929276360553197,
      "loss": 5.3851,
      "mean_token_accuracy": 0.17357644885778428,
      "num_tokens": 19284377.0,
      "step": 10445
    },
    {
      "entropy": 5.623538446426392,
      "epoch": 0.8779668136946019,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004929201848681213,
      "loss": 5.2579,
      "mean_token_accuracy": 0.17334801256656646,
      "num_tokens": 19293326.0,
      "step": 10450
    },
    {
      "entropy": 5.660878467559814,
      "epoch": 0.8783868935097668,
      "grad_norm": 1.0,
      "learning_rate": 0.0004929127298205792,
      "loss": 5.3088,
      "mean_token_accuracy": 0.1824418544769287,
      "num_tokens": 19302086.0,
      "step": 10455
    },
    {
      "entropy": 5.717710876464844,
      "epoch": 0.8788069733249317,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004929052709128251,
      "loss": 5.3253,
      "mean_token_accuracy": 0.18175015598535538,
      "num_tokens": 19310124.0,
      "step": 10460
    },
    {
      "entropy": 5.619649648666382,
      "epoch": 0.8792270531400966,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004928978081449914,
      "loss": 5.3434,
      "mean_token_accuracy": 0.17558258026838303,
      "num_tokens": 19321269.0,
      "step": 10465
    },
    {
      "entropy": 5.656749677658081,
      "epoch": 0.8796471329552615,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004928903415172103,
      "loss": 5.3625,
      "mean_token_accuracy": 0.18061470985412598,
      "num_tokens": 19330390.0,
      "step": 10470
    },
    {
      "entropy": 5.702829742431641,
      "epoch": 0.8800672127704264,
      "grad_norm": 0.96875,
      "learning_rate": 0.000492882871029614,
      "loss": 5.3673,
      "mean_token_accuracy": 0.17414373457431792,
      "num_tokens": 19339457.0,
      "step": 10475
    },
    {
      "entropy": 5.713857698440552,
      "epoch": 0.8804872925855912,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004928753966823348,
      "loss": 5.4266,
      "mean_token_accuracy": 0.1727289468050003,
      "num_tokens": 19348710.0,
      "step": 10480
    },
    {
      "entropy": 5.752920246124267,
      "epoch": 0.8809073724007561,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004928679184755051,
      "loss": 5.4713,
      "mean_token_accuracy": 0.1758141413331032,
      "num_tokens": 19357215.0,
      "step": 10485
    },
    {
      "entropy": 5.7117071628570555,
      "epoch": 0.881327452215921,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004928604364092574,
      "loss": 5.3907,
      "mean_token_accuracy": 0.17333034425973892,
      "num_tokens": 19366043.0,
      "step": 10490
    },
    {
      "entropy": 5.753298950195313,
      "epoch": 0.8817475320310859,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004928529504837243,
      "loss": 5.4486,
      "mean_token_accuracy": 0.1724349707365036,
      "num_tokens": 19375468.0,
      "step": 10495
    },
    {
      "entropy": 5.752193260192871,
      "epoch": 0.8821676118462508,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004928454606990383,
      "loss": 5.3143,
      "mean_token_accuracy": 0.17740447521209718,
      "num_tokens": 19384467.0,
      "step": 10500
    },
    {
      "entropy": 5.6545494556427,
      "epoch": 0.8825876916614157,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004928379670553322,
      "loss": 5.3822,
      "mean_token_accuracy": 0.1732860341668129,
      "num_tokens": 19393618.0,
      "step": 10505
    },
    {
      "entropy": 5.761135721206665,
      "epoch": 0.8830077714765806,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004928304695527387,
      "loss": 5.4411,
      "mean_token_accuracy": 0.17005382031202315,
      "num_tokens": 19402921.0,
      "step": 10510
    },
    {
      "entropy": 5.791484260559082,
      "epoch": 0.8834278512917454,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004928229681913905,
      "loss": 5.4076,
      "mean_token_accuracy": 0.1751767173409462,
      "num_tokens": 19412048.0,
      "step": 10515
    },
    {
      "entropy": 5.7464762210845945,
      "epoch": 0.8838479311069103,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004928154629714207,
      "loss": 5.4029,
      "mean_token_accuracy": 0.17286426573991776,
      "num_tokens": 19420993.0,
      "step": 10520
    },
    {
      "entropy": 5.647195482254029,
      "epoch": 0.8842680109220752,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000492807953892962,
      "loss": 5.3633,
      "mean_token_accuracy": 0.17356964200735092,
      "num_tokens": 19430145.0,
      "step": 10525
    },
    {
      "entropy": 5.628022575378418,
      "epoch": 0.8846880907372401,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004928004409561476,
      "loss": 5.2849,
      "mean_token_accuracy": 0.17803787291049958,
      "num_tokens": 19438918.0,
      "step": 10530
    },
    {
      "entropy": 5.642681217193603,
      "epoch": 0.885108170552405,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004927929241611106,
      "loss": 5.3361,
      "mean_token_accuracy": 0.1790485993027687,
      "num_tokens": 19448490.0,
      "step": 10535
    },
    {
      "entropy": 5.690312719345092,
      "epoch": 0.8855282503675699,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000492785403507984,
      "loss": 5.3891,
      "mean_token_accuracy": 0.1705471768975258,
      "num_tokens": 19457098.0,
      "step": 10540
    },
    {
      "entropy": 5.692985725402832,
      "epoch": 0.8859483301827347,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004927778789969012,
      "loss": 5.3928,
      "mean_token_accuracy": 0.1723649874329567,
      "num_tokens": 19466419.0,
      "step": 10545
    },
    {
      "entropy": 5.668448448181152,
      "epoch": 0.8863684099978996,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004927703506279955,
      "loss": 5.434,
      "mean_token_accuracy": 0.16586271971464156,
      "num_tokens": 19475882.0,
      "step": 10550
    },
    {
      "entropy": 5.764997482299805,
      "epoch": 0.8867884898130645,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004927628184014,
      "loss": 5.4744,
      "mean_token_accuracy": 0.1650938495993614,
      "num_tokens": 19485917.0,
      "step": 10555
    },
    {
      "entropy": 5.7592438697814945,
      "epoch": 0.8872085696282294,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004927552823172483,
      "loss": 5.4107,
      "mean_token_accuracy": 0.16981758028268815,
      "num_tokens": 19494984.0,
      "step": 10560
    },
    {
      "entropy": 5.702795553207397,
      "epoch": 0.8876286494433943,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000492747742375674,
      "loss": 5.3557,
      "mean_token_accuracy": 0.17847150862216948,
      "num_tokens": 19504087.0,
      "step": 10565
    },
    {
      "entropy": 5.72349328994751,
      "epoch": 0.8880487292585592,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004927401985768106,
      "loss": 5.3958,
      "mean_token_accuracy": 0.17886022478342056,
      "num_tokens": 19512880.0,
      "step": 10570
    },
    {
      "entropy": 5.6543807029724125,
      "epoch": 0.888468809073724,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004927326509207915,
      "loss": 5.3588,
      "mean_token_accuracy": 0.1801150843501091,
      "num_tokens": 19521723.0,
      "step": 10575
    },
    {
      "entropy": 5.738810873031616,
      "epoch": 0.8888888888888888,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004927250994077508,
      "loss": 5.4448,
      "mean_token_accuracy": 0.17018781006336212,
      "num_tokens": 19531352.0,
      "step": 10580
    },
    {
      "entropy": 5.756013679504394,
      "epoch": 0.8893089687040537,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000492717544037822,
      "loss": 5.5023,
      "mean_token_accuracy": 0.17919861972332002,
      "num_tokens": 19540943.0,
      "step": 10585
    },
    {
      "entropy": 5.617977857589722,
      "epoch": 0.8897290485192186,
      "grad_norm": 0.921875,
      "learning_rate": 0.000492709984811139,
      "loss": 5.2981,
      "mean_token_accuracy": 0.18022771179676056,
      "num_tokens": 19550527.0,
      "step": 10590
    },
    {
      "entropy": 5.635468912124634,
      "epoch": 0.8901491283343835,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004927024217278358,
      "loss": 5.3009,
      "mean_token_accuracy": 0.18382543027400972,
      "num_tokens": 19559746.0,
      "step": 10595
    },
    {
      "entropy": 5.731223821640015,
      "epoch": 0.8905692081495484,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004926948547880462,
      "loss": 5.4681,
      "mean_token_accuracy": 0.16403882205486298,
      "num_tokens": 19569286.0,
      "step": 10600
    },
    {
      "entropy": 5.647329664230346,
      "epoch": 0.8909892879647133,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004926872839919044,
      "loss": 5.3337,
      "mean_token_accuracy": 0.1724674642086029,
      "num_tokens": 19578245.0,
      "step": 10605
    },
    {
      "entropy": 5.6380500316619875,
      "epoch": 0.8914093677798782,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004926797093395446,
      "loss": 5.3183,
      "mean_token_accuracy": 0.17766014933586122,
      "num_tokens": 19587244.0,
      "step": 10610
    },
    {
      "entropy": 5.684698104858398,
      "epoch": 0.891829447595043,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004926721308311006,
      "loss": 5.3797,
      "mean_token_accuracy": 0.1778306022286415,
      "num_tokens": 19596932.0,
      "step": 10615
    },
    {
      "entropy": 5.759928178787232,
      "epoch": 0.8922495274102079,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004926645484667069,
      "loss": 5.488,
      "mean_token_accuracy": 0.17161238938570023,
      "num_tokens": 19606256.0,
      "step": 10620
    },
    {
      "entropy": 5.810886716842651,
      "epoch": 0.8926696072253728,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004926569622464979,
      "loss": 5.4937,
      "mean_token_accuracy": 0.16746273785829544,
      "num_tokens": 19615726.0,
      "step": 10625
    },
    {
      "entropy": 5.687451601028442,
      "epoch": 0.8930896870405377,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004926493721706079,
      "loss": 5.3537,
      "mean_token_accuracy": 0.17123373448848725,
      "num_tokens": 19624037.0,
      "step": 10630
    },
    {
      "entropy": 5.652608346939087,
      "epoch": 0.8935097668557026,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004926417782391713,
      "loss": 5.3595,
      "mean_token_accuracy": 0.1797306016087532,
      "num_tokens": 19632882.0,
      "step": 10635
    },
    {
      "entropy": 5.710896730422974,
      "epoch": 0.8939298466708675,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004926341804523227,
      "loss": 5.474,
      "mean_token_accuracy": 0.17128193974494935,
      "num_tokens": 19642686.0,
      "step": 10640
    },
    {
      "entropy": 5.685780954360962,
      "epoch": 0.8943499264860324,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004926265788101966,
      "loss": 5.3605,
      "mean_token_accuracy": 0.17778971791267395,
      "num_tokens": 19651380.0,
      "step": 10645
    },
    {
      "entropy": 5.654682922363281,
      "epoch": 0.8947700063011972,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004926189733129278,
      "loss": 5.3067,
      "mean_token_accuracy": 0.17653356194496156,
      "num_tokens": 19660136.0,
      "step": 10650
    },
    {
      "entropy": 5.6517448902130125,
      "epoch": 0.8951900861163621,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004926113639606509,
      "loss": 5.3566,
      "mean_token_accuracy": 0.18426522314548494,
      "num_tokens": 19669146.0,
      "step": 10655
    },
    {
      "entropy": 5.727730321884155,
      "epoch": 0.895610165931527,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004926037507535008,
      "loss": 5.4712,
      "mean_token_accuracy": 0.17460092306137084,
      "num_tokens": 19678627.0,
      "step": 10660
    },
    {
      "entropy": 5.6961911678314205,
      "epoch": 0.8960302457466919,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004925961336916122,
      "loss": 5.4074,
      "mean_token_accuracy": 0.17401352077722548,
      "num_tokens": 19688033.0,
      "step": 10665
    },
    {
      "entropy": 5.742551946640015,
      "epoch": 0.8964503255618568,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004925885127751202,
      "loss": 5.4069,
      "mean_token_accuracy": 0.17422727793455123,
      "num_tokens": 19696523.0,
      "step": 10670
    },
    {
      "entropy": 5.757106876373291,
      "epoch": 0.8968704053770217,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004925808880041596,
      "loss": 5.3255,
      "mean_token_accuracy": 0.17283156812191008,
      "num_tokens": 19706339.0,
      "step": 10675
    },
    {
      "entropy": 5.691051340103149,
      "epoch": 0.8972904851921865,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004925732593788658,
      "loss": 5.3726,
      "mean_token_accuracy": 0.17111489921808243,
      "num_tokens": 19714779.0,
      "step": 10680
    },
    {
      "entropy": 5.6845439910888675,
      "epoch": 0.8977105650073514,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004925656268993737,
      "loss": 5.4493,
      "mean_token_accuracy": 0.1686468631029129,
      "num_tokens": 19723727.0,
      "step": 10685
    },
    {
      "entropy": 5.659609317779541,
      "epoch": 0.8981306448225163,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004925579905658185,
      "loss": 5.4164,
      "mean_token_accuracy": 0.1788722828030586,
      "num_tokens": 19732783.0,
      "step": 10690
    },
    {
      "entropy": 5.728063011169434,
      "epoch": 0.8985507246376812,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004925503503783355,
      "loss": 5.3679,
      "mean_token_accuracy": 0.16937527656555176,
      "num_tokens": 19741268.0,
      "step": 10695
    },
    {
      "entropy": 5.7297547340393065,
      "epoch": 0.898970804452846,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004925427063370601,
      "loss": 5.3017,
      "mean_token_accuracy": 0.17625733464956284,
      "num_tokens": 19751490.0,
      "step": 10700
    },
    {
      "entropy": 5.696488666534424,
      "epoch": 0.899390884268011,
      "grad_norm": 1.0,
      "learning_rate": 0.0004925350584421278,
      "loss": 5.3409,
      "mean_token_accuracy": 0.17375748455524445,
      "num_tokens": 19760487.0,
      "step": 10705
    },
    {
      "entropy": 5.748647928237915,
      "epoch": 0.8998109640831758,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004925274066936738,
      "loss": 5.3112,
      "mean_token_accuracy": 0.1847700893878937,
      "num_tokens": 19768984.0,
      "step": 10710
    },
    {
      "entropy": 5.646275901794434,
      "epoch": 0.9002310438983406,
      "grad_norm": 0.875,
      "learning_rate": 0.0004925197510918339,
      "loss": 5.2945,
      "mean_token_accuracy": 0.18154353648424149,
      "num_tokens": 19778335.0,
      "step": 10715
    },
    {
      "entropy": 5.688979053497315,
      "epoch": 0.9006511237135055,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004925120916367435,
      "loss": 5.4349,
      "mean_token_accuracy": 0.16783437132835388,
      "num_tokens": 19789082.0,
      "step": 10720
    },
    {
      "entropy": 5.601105880737305,
      "epoch": 0.9010712035286704,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004925044283285384,
      "loss": 5.1853,
      "mean_token_accuracy": 0.19120420813560485,
      "num_tokens": 19797902.0,
      "step": 10725
    },
    {
      "entropy": 5.571681499481201,
      "epoch": 0.9014912833438353,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004924967611673544,
      "loss": 5.3554,
      "mean_token_accuracy": 0.17599836885929107,
      "num_tokens": 19806481.0,
      "step": 10730
    },
    {
      "entropy": 5.5609723091125485,
      "epoch": 0.9019113631590002,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004924890901533273,
      "loss": 5.2547,
      "mean_token_accuracy": 0.18766673505306244,
      "num_tokens": 19815226.0,
      "step": 10735
    },
    {
      "entropy": 5.803448152542114,
      "epoch": 0.9023314429741651,
      "grad_norm": 1.0,
      "learning_rate": 0.0004924814152865929,
      "loss": 5.456,
      "mean_token_accuracy": 0.1663283735513687,
      "num_tokens": 19824577.0,
      "step": 10740
    },
    {
      "entropy": 5.7356963634490965,
      "epoch": 0.90275152278933,
      "grad_norm": 1.0,
      "learning_rate": 0.0004924737365672873,
      "loss": 5.3701,
      "mean_token_accuracy": 0.17286850064992904,
      "num_tokens": 19832936.0,
      "step": 10745
    },
    {
      "entropy": 5.781580877304077,
      "epoch": 0.9031716026044948,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004924660539955463,
      "loss": 5.4904,
      "mean_token_accuracy": 0.17319288551807405,
      "num_tokens": 19841946.0,
      "step": 10750
    },
    {
      "entropy": 5.7288895606994625,
      "epoch": 0.9035916824196597,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004924583675715063,
      "loss": 5.3975,
      "mean_token_accuracy": 0.17567060738801957,
      "num_tokens": 19851469.0,
      "step": 10755
    },
    {
      "entropy": 5.7112936019897464,
      "epoch": 0.9040117622348246,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004924506772953031,
      "loss": 5.4491,
      "mean_token_accuracy": 0.17362731099128723,
      "num_tokens": 19860731.0,
      "step": 10760
    },
    {
      "entropy": 5.694790554046631,
      "epoch": 0.9044318420499895,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004924429831670733,
      "loss": 5.4771,
      "mean_token_accuracy": 0.16899607628583907,
      "num_tokens": 19869717.0,
      "step": 10765
    },
    {
      "entropy": 5.74413104057312,
      "epoch": 0.9048519218651544,
      "grad_norm": 0.9375,
      "learning_rate": 0.000492435285186953,
      "loss": 5.3997,
      "mean_token_accuracy": 0.17111445665359498,
      "num_tokens": 19879229.0,
      "step": 10770
    },
    {
      "entropy": 5.790788555145264,
      "epoch": 0.9052720016803193,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004924275833550785,
      "loss": 5.4112,
      "mean_token_accuracy": 0.17046762108802796,
      "num_tokens": 19888260.0,
      "step": 10775
    },
    {
      "entropy": 5.74414381980896,
      "epoch": 0.9056920814954842,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004924198776715865,
      "loss": 5.4313,
      "mean_token_accuracy": 0.17878833264112473,
      "num_tokens": 19897070.0,
      "step": 10780
    },
    {
      "entropy": 5.744823598861695,
      "epoch": 0.906112161310649,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004924121681366132,
      "loss": 5.4088,
      "mean_token_accuracy": 0.16657472401857376,
      "num_tokens": 19907170.0,
      "step": 10785
    },
    {
      "entropy": 5.704545688629151,
      "epoch": 0.9065322411258139,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004924044547502951,
      "loss": 5.353,
      "mean_token_accuracy": 0.1740161508321762,
      "num_tokens": 19917220.0,
      "step": 10790
    },
    {
      "entropy": 5.66229019165039,
      "epoch": 0.9069523209409788,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004923967375127692,
      "loss": 5.4258,
      "mean_token_accuracy": 0.1747304081916809,
      "num_tokens": 19926724.0,
      "step": 10795
    },
    {
      "entropy": 5.79721360206604,
      "epoch": 0.9073724007561437,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000492389016424172,
      "loss": 5.5227,
      "mean_token_accuracy": 0.16669196933507918,
      "num_tokens": 19936429.0,
      "step": 10800
    },
    {
      "entropy": 5.716966581344605,
      "epoch": 0.9077924805713086,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004923812914846404,
      "loss": 5.2966,
      "mean_token_accuracy": 0.18898213356733323,
      "num_tokens": 19945096.0,
      "step": 10805
    },
    {
      "entropy": 5.688310003280639,
      "epoch": 0.9082125603864735,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004923735626943111,
      "loss": 5.3743,
      "mean_token_accuracy": 0.1788196086883545,
      "num_tokens": 19953560.0,
      "step": 10810
    },
    {
      "entropy": 5.711865472793579,
      "epoch": 0.9086326402016384,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004923658300533211,
      "loss": 5.3591,
      "mean_token_accuracy": 0.17323427945375441,
      "num_tokens": 19962669.0,
      "step": 10815
    },
    {
      "entropy": 5.762352132797242,
      "epoch": 0.9090527200168032,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004923580935618073,
      "loss": 5.4075,
      "mean_token_accuracy": 0.1719552233815193,
      "num_tokens": 19971990.0,
      "step": 10820
    },
    {
      "entropy": 5.674012804031372,
      "epoch": 0.909472799831968,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004923503532199069,
      "loss": 5.377,
      "mean_token_accuracy": 0.17736288160085678,
      "num_tokens": 19981850.0,
      "step": 10825
    },
    {
      "entropy": 5.705214977264404,
      "epoch": 0.909892879647133,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004923426090277567,
      "loss": 5.4275,
      "mean_token_accuracy": 0.16890623271465302,
      "num_tokens": 19991574.0,
      "step": 10830
    },
    {
      "entropy": 5.729412794113159,
      "epoch": 0.9103129594622978,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004923348609854943,
      "loss": 5.3803,
      "mean_token_accuracy": 0.17469983398914338,
      "num_tokens": 20001392.0,
      "step": 10835
    },
    {
      "entropy": 5.7025830268859865,
      "epoch": 0.9107330392774627,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004923271090932566,
      "loss": 5.424,
      "mean_token_accuracy": 0.17207745164632798,
      "num_tokens": 20011277.0,
      "step": 10840
    },
    {
      "entropy": 5.663294696807862,
      "epoch": 0.9111531190926276,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004923193533511812,
      "loss": 5.3381,
      "mean_token_accuracy": 0.17635092735290528,
      "num_tokens": 20021171.0,
      "step": 10845
    },
    {
      "entropy": 5.805743455886841,
      "epoch": 0.9115731989077924,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004923115937594053,
      "loss": 5.4156,
      "mean_token_accuracy": 0.17307167649269103,
      "num_tokens": 20030189.0,
      "step": 10850
    },
    {
      "entropy": 5.761628484725952,
      "epoch": 0.9119932787229573,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004923038303180664,
      "loss": 5.3854,
      "mean_token_accuracy": 0.18065591603517533,
      "num_tokens": 20038287.0,
      "step": 10855
    },
    {
      "entropy": 5.694932651519776,
      "epoch": 0.9124133585381222,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000492296063027302,
      "loss": 5.4209,
      "mean_token_accuracy": 0.16729898005723953,
      "num_tokens": 20047653.0,
      "step": 10860
    },
    {
      "entropy": 5.6990941047668455,
      "epoch": 0.9128334383532871,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004922882918872498,
      "loss": 5.4297,
      "mean_token_accuracy": 0.16623241305351258,
      "num_tokens": 20057415.0,
      "step": 10865
    },
    {
      "entropy": 5.8091846942901615,
      "epoch": 0.913253518168452,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004922805168980475,
      "loss": 5.4335,
      "mean_token_accuracy": 0.17249839901924133,
      "num_tokens": 20065996.0,
      "step": 10870
    },
    {
      "entropy": 5.684816312789917,
      "epoch": 0.9136735979836169,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004922727380598326,
      "loss": 5.3657,
      "mean_token_accuracy": 0.17473243474960326,
      "num_tokens": 20075376.0,
      "step": 10875
    },
    {
      "entropy": 5.641756153106689,
      "epoch": 0.9140936777987818,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000492264955372743,
      "loss": 5.401,
      "mean_token_accuracy": 0.16607374250888823,
      "num_tokens": 20084950.0,
      "step": 10880
    },
    {
      "entropy": 5.751581430435181,
      "epoch": 0.9145137576139466,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004922571688369165,
      "loss": 5.3881,
      "mean_token_accuracy": 0.17759778648614882,
      "num_tokens": 20094011.0,
      "step": 10885
    },
    {
      "entropy": 5.7211121082305905,
      "epoch": 0.9149338374291115,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004922493784524914,
      "loss": 5.3524,
      "mean_token_accuracy": 0.1753992959856987,
      "num_tokens": 20103037.0,
      "step": 10890
    },
    {
      "entropy": 5.722981595993042,
      "epoch": 0.9153539172442764,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004922415842196052,
      "loss": 5.4954,
      "mean_token_accuracy": 0.16313080489635468,
      "num_tokens": 20112727.0,
      "step": 10895
    },
    {
      "entropy": 5.591765022277832,
      "epoch": 0.9157739970594413,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004922337861383963,
      "loss": 5.3086,
      "mean_token_accuracy": 0.18054919838905334,
      "num_tokens": 20122341.0,
      "step": 10900
    },
    {
      "entropy": 5.657147073745728,
      "epoch": 0.9161940768746062,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004922259842090027,
      "loss": 5.2986,
      "mean_token_accuracy": 0.1779008224606514,
      "num_tokens": 20131354.0,
      "step": 10905
    },
    {
      "entropy": 5.711278343200684,
      "epoch": 0.9166141566897711,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004922181784315627,
      "loss": 5.3422,
      "mean_token_accuracy": 0.17677853852510453,
      "num_tokens": 20140440.0,
      "step": 10910
    },
    {
      "entropy": 5.676589775085449,
      "epoch": 0.917034236504936,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004922103688062145,
      "loss": 5.349,
      "mean_token_accuracy": 0.17499181777238845,
      "num_tokens": 20149331.0,
      "step": 10915
    },
    {
      "entropy": 5.637252330780029,
      "epoch": 0.9174543163201008,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004922025553330964,
      "loss": 5.3109,
      "mean_token_accuracy": 0.18377403616905214,
      "num_tokens": 20158566.0,
      "step": 10920
    },
    {
      "entropy": 5.703567123413086,
      "epoch": 0.9178743961352657,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000492194738012347,
      "loss": 5.3992,
      "mean_token_accuracy": 0.17752688080072404,
      "num_tokens": 20168339.0,
      "step": 10925
    },
    {
      "entropy": 5.774370861053467,
      "epoch": 0.9182944759504306,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004921869168441045,
      "loss": 5.4187,
      "mean_token_accuracy": 0.16602367162704468,
      "num_tokens": 20177967.0,
      "step": 10930
    },
    {
      "entropy": 5.7096333503723145,
      "epoch": 0.9187145557655955,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004921790918285077,
      "loss": 5.4187,
      "mean_token_accuracy": 0.17528679966926575,
      "num_tokens": 20187279.0,
      "step": 10935
    },
    {
      "entropy": 5.691339206695557,
      "epoch": 0.9191346355807604,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004921712629656951,
      "loss": 5.4991,
      "mean_token_accuracy": 0.18109170645475386,
      "num_tokens": 20195324.0,
      "step": 10940
    },
    {
      "entropy": 5.718715572357178,
      "epoch": 0.9195547153959253,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004921634302558054,
      "loss": 5.4287,
      "mean_token_accuracy": 0.17527840733528138,
      "num_tokens": 20204985.0,
      "step": 10945
    },
    {
      "entropy": 5.721715116500855,
      "epoch": 0.9199747952110902,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004921555936989773,
      "loss": 5.4447,
      "mean_token_accuracy": 0.1718514993786812,
      "num_tokens": 20214553.0,
      "step": 10950
    },
    {
      "entropy": 5.774837923049927,
      "epoch": 0.9203948750262549,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004921477532953497,
      "loss": 5.3547,
      "mean_token_accuracy": 0.17544877976179124,
      "num_tokens": 20224118.0,
      "step": 10955
    },
    {
      "entropy": 5.647385692596435,
      "epoch": 0.9208149548414198,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004921399090450616,
      "loss": 5.3188,
      "mean_token_accuracy": 0.171922767162323,
      "num_tokens": 20233719.0,
      "step": 10960
    },
    {
      "entropy": 5.67383074760437,
      "epoch": 0.9212350346565847,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004921320609482517,
      "loss": 5.4157,
      "mean_token_accuracy": 0.17376127541065217,
      "num_tokens": 20242311.0,
      "step": 10965
    },
    {
      "entropy": 5.75290060043335,
      "epoch": 0.9216551144717496,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004921242090050591,
      "loss": 5.4505,
      "mean_token_accuracy": 0.17223982661962509,
      "num_tokens": 20252998.0,
      "step": 10970
    },
    {
      "entropy": 5.7609776020050045,
      "epoch": 0.9220751942869145,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000492116353215623,
      "loss": 5.4945,
      "mean_token_accuracy": 0.17329947650432587,
      "num_tokens": 20262456.0,
      "step": 10975
    },
    {
      "entropy": 5.628494024276733,
      "epoch": 0.9224952741020794,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004921084935800825,
      "loss": 5.2587,
      "mean_token_accuracy": 0.1845813661813736,
      "num_tokens": 20271516.0,
      "step": 10980
    },
    {
      "entropy": 5.622925424575806,
      "epoch": 0.9229153539172443,
      "grad_norm": 1.0,
      "learning_rate": 0.0004921006300985768,
      "loss": 5.3017,
      "mean_token_accuracy": 0.17941147834062576,
      "num_tokens": 20280373.0,
      "step": 10985
    },
    {
      "entropy": 5.672286891937256,
      "epoch": 0.9233354337324091,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004920927627712453,
      "loss": 5.2981,
      "mean_token_accuracy": 0.1797472581267357,
      "num_tokens": 20289426.0,
      "step": 10990
    },
    {
      "entropy": 5.752888011932373,
      "epoch": 0.923755513547574,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004920848915982273,
      "loss": 5.4457,
      "mean_token_accuracy": 0.17115329205989838,
      "num_tokens": 20298045.0,
      "step": 10995
    },
    {
      "entropy": 5.656572151184082,
      "epoch": 0.9241755933627389,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004920770165796622,
      "loss": 5.3458,
      "mean_token_accuracy": 0.17374630123376847,
      "num_tokens": 20307352.0,
      "step": 11000
    },
    {
      "entropy": 5.679479598999023,
      "epoch": 0.9245956731779038,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004920691377156895,
      "loss": 5.3666,
      "mean_token_accuracy": 0.1757354497909546,
      "num_tokens": 20316448.0,
      "step": 11005
    },
    {
      "entropy": 5.784104871749878,
      "epoch": 0.9250157529930687,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004920612550064488,
      "loss": 5.5019,
      "mean_token_accuracy": 0.1707981675863266,
      "num_tokens": 20326440.0,
      "step": 11010
    },
    {
      "entropy": 5.6779622554779055,
      "epoch": 0.9254358328082336,
      "grad_norm": 1.0,
      "learning_rate": 0.0004920533684520797,
      "loss": 5.2938,
      "mean_token_accuracy": 0.18185334652662277,
      "num_tokens": 20335447.0,
      "step": 11015
    },
    {
      "entropy": 5.694982051849365,
      "epoch": 0.9258559126233984,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000492045478052722,
      "loss": 5.448,
      "mean_token_accuracy": 0.17321226745843887,
      "num_tokens": 20344523.0,
      "step": 11020
    },
    {
      "entropy": 5.691788530349731,
      "epoch": 0.9262759924385633,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004920375838085154,
      "loss": 5.394,
      "mean_token_accuracy": 0.17579833120107652,
      "num_tokens": 20354267.0,
      "step": 11025
    },
    {
      "entropy": 5.712007617950439,
      "epoch": 0.9266960722537282,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004920296857195998,
      "loss": 5.4686,
      "mean_token_accuracy": 0.17234223932027817,
      "num_tokens": 20364137.0,
      "step": 11030
    },
    {
      "entropy": 5.695209932327271,
      "epoch": 0.9271161520688931,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000492021783786115,
      "loss": 5.3605,
      "mean_token_accuracy": 0.17994005978107452,
      "num_tokens": 20372583.0,
      "step": 11035
    },
    {
      "entropy": 5.647004222869873,
      "epoch": 0.927536231884058,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004920138780082011,
      "loss": 5.3118,
      "mean_token_accuracy": 0.17195972055196762,
      "num_tokens": 20382050.0,
      "step": 11040
    },
    {
      "entropy": 5.6490295886993405,
      "epoch": 0.9279563116992229,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004920059683859981,
      "loss": 5.2645,
      "mean_token_accuracy": 0.18437047600746154,
      "num_tokens": 20391425.0,
      "step": 11045
    },
    {
      "entropy": 5.746783065795898,
      "epoch": 0.9283763915143878,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004919980549196461,
      "loss": 5.4396,
      "mean_token_accuracy": 0.17227613627910615,
      "num_tokens": 20400559.0,
      "step": 11050
    },
    {
      "entropy": 5.691994285583496,
      "epoch": 0.9287964713295526,
      "grad_norm": 1.0,
      "learning_rate": 0.0004919901376092853,
      "loss": 5.3477,
      "mean_token_accuracy": 0.17798608243465425,
      "num_tokens": 20408985.0,
      "step": 11055
    },
    {
      "entropy": 5.6829784393310545,
      "epoch": 0.9292165511447175,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004919822164550559,
      "loss": 5.4641,
      "mean_token_accuracy": 0.16510661095380783,
      "num_tokens": 20417855.0,
      "step": 11060
    },
    {
      "entropy": 5.732713508605957,
      "epoch": 0.9296366309598824,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004919742914570983,
      "loss": 5.4242,
      "mean_token_accuracy": 0.1783280700445175,
      "num_tokens": 20426191.0,
      "step": 11065
    },
    {
      "entropy": 5.698213052749634,
      "epoch": 0.9300567107750473,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000491966362615553,
      "loss": 5.3835,
      "mean_token_accuracy": 0.16715701371431352,
      "num_tokens": 20435592.0,
      "step": 11070
    },
    {
      "entropy": 5.749117851257324,
      "epoch": 0.9304767905902122,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00049195842993056,
      "loss": 5.4086,
      "mean_token_accuracy": 0.17477771788835525,
      "num_tokens": 20445504.0,
      "step": 11075
    },
    {
      "entropy": 5.705859041213989,
      "epoch": 0.930896870405377,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004919504934022604,
      "loss": 5.3759,
      "mean_token_accuracy": 0.16921173632144929,
      "num_tokens": 20455153.0,
      "step": 11080
    },
    {
      "entropy": 5.682570886611939,
      "epoch": 0.931316950220542,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004919425530307943,
      "loss": 5.3456,
      "mean_token_accuracy": 0.17448251694440842,
      "num_tokens": 20465101.0,
      "step": 11085
    },
    {
      "entropy": 5.69701361656189,
      "epoch": 0.9317370300357067,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004919346088163028,
      "loss": 5.397,
      "mean_token_accuracy": 0.1750429481267929,
      "num_tokens": 20474700.0,
      "step": 11090
    },
    {
      "entropy": 5.727405309677124,
      "epoch": 0.9321571098508716,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004919266607589263,
      "loss": 5.4386,
      "mean_token_accuracy": 0.16773997992277145,
      "num_tokens": 20483945.0,
      "step": 11095
    },
    {
      "entropy": 5.691340780258178,
      "epoch": 0.9325771896660365,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004919187088588057,
      "loss": 5.403,
      "mean_token_accuracy": 0.17400172501802444,
      "num_tokens": 20493307.0,
      "step": 11100
    },
    {
      "entropy": 5.6985382556915285,
      "epoch": 0.9329972694812014,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004919107531160819,
      "loss": 5.3344,
      "mean_token_accuracy": 0.18474431931972504,
      "num_tokens": 20501889.0,
      "step": 11105
    },
    {
      "entropy": 5.721513652801514,
      "epoch": 0.9334173492963663,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004919027935308957,
      "loss": 5.3689,
      "mean_token_accuracy": 0.17995511889457702,
      "num_tokens": 20510577.0,
      "step": 11110
    },
    {
      "entropy": 5.632858562469482,
      "epoch": 0.9338374291115312,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004918948301033884,
      "loss": 5.3165,
      "mean_token_accuracy": 0.1768263667821884,
      "num_tokens": 20520025.0,
      "step": 11115
    },
    {
      "entropy": 5.719315910339356,
      "epoch": 0.9342575089266961,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004918868628337007,
      "loss": 5.3865,
      "mean_token_accuracy": 0.17152649015188218,
      "num_tokens": 20528989.0,
      "step": 11120
    },
    {
      "entropy": 5.727646160125732,
      "epoch": 0.9346775887418609,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004918788917219739,
      "loss": 5.3322,
      "mean_token_accuracy": 0.17481573671102524,
      "num_tokens": 20538328.0,
      "step": 11125
    },
    {
      "entropy": 5.708231258392334,
      "epoch": 0.9350976685570258,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004918709167683493,
      "loss": 5.49,
      "mean_token_accuracy": 0.17153393775224685,
      "num_tokens": 20548069.0,
      "step": 11130
    },
    {
      "entropy": 5.653415298461914,
      "epoch": 0.9355177483721907,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004918629379729681,
      "loss": 5.2243,
      "mean_token_accuracy": 0.18639769405126572,
      "num_tokens": 20557128.0,
      "step": 11135
    },
    {
      "entropy": 5.631262969970703,
      "epoch": 0.9359378281873556,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004918549553359715,
      "loss": 5.3566,
      "mean_token_accuracy": 0.17522759288549422,
      "num_tokens": 20566352.0,
      "step": 11140
    },
    {
      "entropy": 5.690695142745971,
      "epoch": 0.9363579080025205,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004918469688575012,
      "loss": 5.3845,
      "mean_token_accuracy": 0.17432625889778136,
      "num_tokens": 20575814.0,
      "step": 11145
    },
    {
      "entropy": 5.7054613590240475,
      "epoch": 0.9367779878176854,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004918389785376983,
      "loss": 5.255,
      "mean_token_accuracy": 0.17823997735977173,
      "num_tokens": 20584715.0,
      "step": 11150
    },
    {
      "entropy": 5.639838075637817,
      "epoch": 0.9371980676328503,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004918309843767047,
      "loss": 5.333,
      "mean_token_accuracy": 0.17500682324171066,
      "num_tokens": 20594630.0,
      "step": 11155
    },
    {
      "entropy": 5.638676309585572,
      "epoch": 0.9376181474480151,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004918229863746618,
      "loss": 5.3184,
      "mean_token_accuracy": 0.1756111577153206,
      "num_tokens": 20603653.0,
      "step": 11160
    },
    {
      "entropy": 5.7117600440979,
      "epoch": 0.93803822726318,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004918149845317114,
      "loss": 5.3999,
      "mean_token_accuracy": 0.17521323263645172,
      "num_tokens": 20612188.0,
      "step": 11165
    },
    {
      "entropy": 5.688275718688965,
      "epoch": 0.9384583070783449,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004918069788479952,
      "loss": 5.325,
      "mean_token_accuracy": 0.17852505743503572,
      "num_tokens": 20620933.0,
      "step": 11170
    },
    {
      "entropy": 5.686825752258301,
      "epoch": 0.9388783868935098,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004917989693236549,
      "loss": 5.3624,
      "mean_token_accuracy": 0.18025224506855012,
      "num_tokens": 20629919.0,
      "step": 11175
    },
    {
      "entropy": 5.686520147323608,
      "epoch": 0.9392984667086747,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004917909559588326,
      "loss": 5.3398,
      "mean_token_accuracy": 0.17267686873674393,
      "num_tokens": 20638475.0,
      "step": 11180
    },
    {
      "entropy": 5.794269943237305,
      "epoch": 0.9397185465238396,
      "grad_norm": 1.078125,
      "learning_rate": 0.00049178293875367,
      "loss": 5.4462,
      "mean_token_accuracy": 0.1658658653497696,
      "num_tokens": 20648105.0,
      "step": 11185
    },
    {
      "entropy": 5.709870529174805,
      "epoch": 0.9401386263390044,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004917749177083094,
      "loss": 5.3611,
      "mean_token_accuracy": 0.17066793292760848,
      "num_tokens": 20657527.0,
      "step": 11190
    },
    {
      "entropy": 5.649870777130127,
      "epoch": 0.9405587061541693,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004917668928228927,
      "loss": 5.3518,
      "mean_token_accuracy": 0.17344795316457748,
      "num_tokens": 20666375.0,
      "step": 11195
    },
    {
      "entropy": 5.7032557964324955,
      "epoch": 0.9409787859693342,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004917588640975622,
      "loss": 5.302,
      "mean_token_accuracy": 0.18351777046918868,
      "num_tokens": 20675350.0,
      "step": 11200
    },
    {
      "entropy": 5.634568071365356,
      "epoch": 0.941398865784499,
      "grad_norm": 1.078125,
      "learning_rate": 0.00049175083153246,
      "loss": 5.2435,
      "mean_token_accuracy": 0.17860905081033707,
      "num_tokens": 20684072.0,
      "step": 11205
    },
    {
      "entropy": 5.612369060516357,
      "epoch": 0.941818945599664,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004917427951277284,
      "loss": 5.3442,
      "mean_token_accuracy": 0.17951606214046478,
      "num_tokens": 20692989.0,
      "step": 11210
    },
    {
      "entropy": 5.724744606018066,
      "epoch": 0.9422390254148288,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004917347548835097,
      "loss": 5.3072,
      "mean_token_accuracy": 0.17899496406316756,
      "num_tokens": 20701269.0,
      "step": 11215
    },
    {
      "entropy": 5.738250255584717,
      "epoch": 0.9426591052299937,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004917267107999466,
      "loss": 5.4186,
      "mean_token_accuracy": 0.1694338619709015,
      "num_tokens": 20709739.0,
      "step": 11220
    },
    {
      "entropy": 5.670594787597656,
      "epoch": 0.9430791850451585,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004917186628771812,
      "loss": 5.3511,
      "mean_token_accuracy": 0.17989387214183808,
      "num_tokens": 20718950.0,
      "step": 11225
    },
    {
      "entropy": 5.6569007396697994,
      "epoch": 0.9434992648603234,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004917106111153565,
      "loss": 5.3268,
      "mean_token_accuracy": 0.17909297347068787,
      "num_tokens": 20729469.0,
      "step": 11230
    },
    {
      "entropy": 5.711237812042237,
      "epoch": 0.9439193446754883,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004917025555146148,
      "loss": 5.3636,
      "mean_token_accuracy": 0.18594632893800736,
      "num_tokens": 20738231.0,
      "step": 11235
    },
    {
      "entropy": 5.721692657470703,
      "epoch": 0.9443394244906532,
      "grad_norm": 0.890625,
      "learning_rate": 0.000491694496075099,
      "loss": 5.5361,
      "mean_token_accuracy": 0.16001801192760468,
      "num_tokens": 20748578.0,
      "step": 11240
    },
    {
      "entropy": 5.792121410369873,
      "epoch": 0.9447595043058181,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004916864327969517,
      "loss": 5.4773,
      "mean_token_accuracy": 0.16235322058200835,
      "num_tokens": 20759284.0,
      "step": 11245
    },
    {
      "entropy": 5.751742935180664,
      "epoch": 0.945179584120983,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004916783656803158,
      "loss": 5.4322,
      "mean_token_accuracy": 0.17575230002403258,
      "num_tokens": 20768186.0,
      "step": 11250
    },
    {
      "entropy": 5.657699394226074,
      "epoch": 0.9455996639361479,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004916702947253342,
      "loss": 5.2735,
      "mean_token_accuracy": 0.18520920723676682,
      "num_tokens": 20776711.0,
      "step": 11255
    },
    {
      "entropy": 5.703401327133179,
      "epoch": 0.9460197437513127,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004916622199321501,
      "loss": 5.3587,
      "mean_token_accuracy": 0.17892625778913498,
      "num_tokens": 20785154.0,
      "step": 11260
    },
    {
      "entropy": 5.6899854183197025,
      "epoch": 0.9464398235664776,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004916541413009062,
      "loss": 5.2996,
      "mean_token_accuracy": 0.18226084262132644,
      "num_tokens": 20794114.0,
      "step": 11265
    },
    {
      "entropy": 5.725108480453491,
      "epoch": 0.9468599033816425,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004916460588317458,
      "loss": 5.3781,
      "mean_token_accuracy": 0.18178620487451552,
      "num_tokens": 20803892.0,
      "step": 11270
    },
    {
      "entropy": 5.644835662841797,
      "epoch": 0.9472799831968074,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004916379725248118,
      "loss": 5.2925,
      "mean_token_accuracy": 0.17907506674528123,
      "num_tokens": 20812892.0,
      "step": 11275
    },
    {
      "entropy": 5.695964097976685,
      "epoch": 0.9477000630119723,
      "grad_norm": 1.0,
      "learning_rate": 0.0004916298823802479,
      "loss": 5.3626,
      "mean_token_accuracy": 0.1705339878797531,
      "num_tokens": 20821934.0,
      "step": 11280
    },
    {
      "entropy": 5.651394414901733,
      "epoch": 0.9481201428271372,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004916217883981971,
      "loss": 5.272,
      "mean_token_accuracy": 0.18131529539823532,
      "num_tokens": 20830100.0,
      "step": 11285
    },
    {
      "entropy": 5.664659643173218,
      "epoch": 0.9485402226423021,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004916136905788029,
      "loss": 5.3611,
      "mean_token_accuracy": 0.17631006389856338,
      "num_tokens": 20839890.0,
      "step": 11290
    },
    {
      "entropy": 5.748161506652832,
      "epoch": 0.9489603024574669,
      "grad_norm": 1.125,
      "learning_rate": 0.0004916055889222087,
      "loss": 5.4677,
      "mean_token_accuracy": 0.1618989586830139,
      "num_tokens": 20848670.0,
      "step": 11295
    },
    {
      "entropy": 5.717919921875,
      "epoch": 0.9493803822726318,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000491597483428558,
      "loss": 5.3185,
      "mean_token_accuracy": 0.18559250235557556,
      "num_tokens": 20857291.0,
      "step": 11300
    },
    {
      "entropy": 5.580420303344726,
      "epoch": 0.9498004620877967,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004915893740979944,
      "loss": 5.3046,
      "mean_token_accuracy": 0.18592865020036697,
      "num_tokens": 20865341.0,
      "step": 11305
    },
    {
      "entropy": 5.6954912662506105,
      "epoch": 0.9502205419029616,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004915812609306617,
      "loss": 5.4033,
      "mean_token_accuracy": 0.1736812561750412,
      "num_tokens": 20875194.0,
      "step": 11310
    },
    {
      "entropy": 5.702722787857056,
      "epoch": 0.9506406217181265,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004915731439267034,
      "loss": 5.3373,
      "mean_token_accuracy": 0.1742067500948906,
      "num_tokens": 20884831.0,
      "step": 11315
    },
    {
      "entropy": 5.604478073120117,
      "epoch": 0.9510607015332914,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004915650230862634,
      "loss": 5.221,
      "mean_token_accuracy": 0.1844581037759781,
      "num_tokens": 20893790.0,
      "step": 11320
    },
    {
      "entropy": 5.660559034347534,
      "epoch": 0.9514807813484563,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004915568984094854,
      "loss": 5.3391,
      "mean_token_accuracy": 0.17990740090608598,
      "num_tokens": 20902175.0,
      "step": 11325
    },
    {
      "entropy": 5.8010705471038815,
      "epoch": 0.951900861163621,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004915487698965136,
      "loss": 5.4611,
      "mean_token_accuracy": 0.16570978313684465,
      "num_tokens": 20911484.0,
      "step": 11330
    },
    {
      "entropy": 5.771519184112549,
      "epoch": 0.952320940978786,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004915406375474917,
      "loss": 5.4216,
      "mean_token_accuracy": 0.1652315691113472,
      "num_tokens": 20920916.0,
      "step": 11335
    },
    {
      "entropy": 5.701288890838623,
      "epoch": 0.9527410207939508,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000491532501362564,
      "loss": 5.4262,
      "mean_token_accuracy": 0.17851725369691848,
      "num_tokens": 20930219.0,
      "step": 11340
    },
    {
      "entropy": 5.651402187347412,
      "epoch": 0.9531611006091157,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004915243613418745,
      "loss": 5.2713,
      "mean_token_accuracy": 0.17963684797286988,
      "num_tokens": 20939591.0,
      "step": 11345
    },
    {
      "entropy": 5.710078477859497,
      "epoch": 0.9535811804242806,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004915162174855675,
      "loss": 5.4263,
      "mean_token_accuracy": 0.168277545273304,
      "num_tokens": 20950035.0,
      "step": 11350
    },
    {
      "entropy": 5.7000363826751705,
      "epoch": 0.9540012602394455,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004915080697937872,
      "loss": 5.3375,
      "mean_token_accuracy": 0.17775082290172578,
      "num_tokens": 20959168.0,
      "step": 11355
    },
    {
      "entropy": 5.66070671081543,
      "epoch": 0.9544213400546103,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004914999182666779,
      "loss": 5.2575,
      "mean_token_accuracy": 0.18143768310546876,
      "num_tokens": 20967887.0,
      "step": 11360
    },
    {
      "entropy": 5.759396409988403,
      "epoch": 0.9548414198697752,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004914917629043839,
      "loss": 5.3747,
      "mean_token_accuracy": 0.16917349696159362,
      "num_tokens": 20977558.0,
      "step": 11365
    },
    {
      "entropy": 5.622768831253052,
      "epoch": 0.9552614996849401,
      "grad_norm": 1.015625,
      "learning_rate": 0.00049148360370705,
      "loss": 5.3613,
      "mean_token_accuracy": 0.17598101645708084,
      "num_tokens": 20986118.0,
      "step": 11370
    },
    {
      "entropy": 5.691233921051025,
      "epoch": 0.955681579500105,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004914754406748204,
      "loss": 5.2673,
      "mean_token_accuracy": 0.17957736402750016,
      "num_tokens": 20994623.0,
      "step": 11375
    },
    {
      "entropy": 5.741635704040528,
      "epoch": 0.9561016593152699,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00049146727380784,
      "loss": 5.4492,
      "mean_token_accuracy": 0.16830119043588637,
      "num_tokens": 21004193.0,
      "step": 11380
    },
    {
      "entropy": 5.656800889968872,
      "epoch": 0.9565217391304348,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004914591031062531,
      "loss": 5.2915,
      "mean_token_accuracy": 0.18386186808347701,
      "num_tokens": 21013125.0,
      "step": 11385
    },
    {
      "entropy": 5.55534782409668,
      "epoch": 0.9569418189455997,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004914509285702048,
      "loss": 5.1773,
      "mean_token_accuracy": 0.1943739041686058,
      "num_tokens": 21021402.0,
      "step": 11390
    },
    {
      "entropy": 5.670525217056275,
      "epoch": 0.9573618987607645,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004914427501998397,
      "loss": 5.3079,
      "mean_token_accuracy": 0.17549702823162078,
      "num_tokens": 21029639.0,
      "step": 11395
    },
    {
      "entropy": 5.69216251373291,
      "epoch": 0.9577819785759294,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004914345679953027,
      "loss": 5.3231,
      "mean_token_accuracy": 0.1795396089553833,
      "num_tokens": 21037525.0,
      "step": 11400
    },
    {
      "entropy": 5.725239086151123,
      "epoch": 0.9582020583910943,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004914263819567388,
      "loss": 5.432,
      "mean_token_accuracy": 0.16999363154172897,
      "num_tokens": 21047702.0,
      "step": 11405
    },
    {
      "entropy": 5.766044092178345,
      "epoch": 0.9586221382062592,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000491418192084293,
      "loss": 5.3185,
      "mean_token_accuracy": 0.18020146936178208,
      "num_tokens": 21056379.0,
      "step": 11410
    },
    {
      "entropy": 5.666383695602417,
      "epoch": 0.9590422180214241,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004914099983781104,
      "loss": 5.3222,
      "mean_token_accuracy": 0.18027336299419403,
      "num_tokens": 21065283.0,
      "step": 11415
    },
    {
      "entropy": 5.650647783279419,
      "epoch": 0.959462297836589,
      "grad_norm": 1.03125,
      "learning_rate": 0.000491401800838336,
      "loss": 5.4529,
      "mean_token_accuracy": 0.17243248224258423,
      "num_tokens": 21074938.0,
      "step": 11420
    },
    {
      "entropy": 5.674296140670776,
      "epoch": 0.9598823776517539,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004913935994651153,
      "loss": 5.2953,
      "mean_token_accuracy": 0.18081222623586654,
      "num_tokens": 21084729.0,
      "step": 11425
    },
    {
      "entropy": 5.6087418556213375,
      "epoch": 0.9603024574669187,
      "grad_norm": 1.0,
      "learning_rate": 0.0004913853942585932,
      "loss": 5.1992,
      "mean_token_accuracy": 0.18500538915395737,
      "num_tokens": 21093456.0,
      "step": 11430
    },
    {
      "entropy": 5.621715021133423,
      "epoch": 0.9607225372820836,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004913771852189155,
      "loss": 5.3215,
      "mean_token_accuracy": 0.17773586362600327,
      "num_tokens": 21102980.0,
      "step": 11435
    },
    {
      "entropy": 5.797303867340088,
      "epoch": 0.9611426170972485,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004913689723462271,
      "loss": 5.5249,
      "mean_token_accuracy": 0.18506134003400804,
      "num_tokens": 21112777.0,
      "step": 11440
    },
    {
      "entropy": 5.709644746780396,
      "epoch": 0.9615626969124134,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000491360755640674,
      "loss": 5.4379,
      "mean_token_accuracy": 0.1728661462664604,
      "num_tokens": 21122139.0,
      "step": 11445
    },
    {
      "entropy": 5.641683149337768,
      "epoch": 0.9619827767275783,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004913525351024014,
      "loss": 5.3162,
      "mean_token_accuracy": 0.17775568664073943,
      "num_tokens": 21131425.0,
      "step": 11450
    },
    {
      "entropy": 5.662838792800903,
      "epoch": 0.9624028565427432,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004913443107315552,
      "loss": 5.3044,
      "mean_token_accuracy": 0.172211217880249,
      "num_tokens": 21140784.0,
      "step": 11455
    },
    {
      "entropy": 5.693063879013062,
      "epoch": 0.962822936357908,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004913360825282807,
      "loss": 5.3104,
      "mean_token_accuracy": 0.17996297627687455,
      "num_tokens": 21150408.0,
      "step": 11460
    },
    {
      "entropy": 5.652466821670532,
      "epoch": 0.9632430161730728,
      "grad_norm": 1.0,
      "learning_rate": 0.000491327850492724,
      "loss": 5.4142,
      "mean_token_accuracy": 0.17556226253509521,
      "num_tokens": 21158915.0,
      "step": 11465
    },
    {
      "entropy": 5.558532238006592,
      "epoch": 0.9636630959882377,
      "grad_norm": 1.0,
      "learning_rate": 0.0004913196146250309,
      "loss": 5.2093,
      "mean_token_accuracy": 0.18567204624414443,
      "num_tokens": 21167336.0,
      "step": 11470
    },
    {
      "entropy": 5.727208852767944,
      "epoch": 0.9640831758034026,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004913113749253472,
      "loss": 5.4852,
      "mean_token_accuracy": 0.17642497569322585,
      "num_tokens": 21177499.0,
      "step": 11475
    },
    {
      "entropy": 5.850815153121948,
      "epoch": 0.9645032556185675,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004913031313938188,
      "loss": 5.4275,
      "mean_token_accuracy": 0.16683491468429565,
      "num_tokens": 21186961.0,
      "step": 11480
    },
    {
      "entropy": 5.7186026096344,
      "epoch": 0.9649233354337324,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004912948840305919,
      "loss": 5.3142,
      "mean_token_accuracy": 0.18248960077762605,
      "num_tokens": 21196364.0,
      "step": 11485
    },
    {
      "entropy": 5.668226718902588,
      "epoch": 0.9653434152488973,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004912866328358125,
      "loss": 5.3697,
      "mean_token_accuracy": 0.1763843223452568,
      "num_tokens": 21206376.0,
      "step": 11490
    },
    {
      "entropy": 5.674428987503052,
      "epoch": 0.9657634950640621,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004912783778096266,
      "loss": 5.3492,
      "mean_token_accuracy": 0.18101761788129805,
      "num_tokens": 21215889.0,
      "step": 11495
    },
    {
      "entropy": 5.709354543685913,
      "epoch": 0.966183574879227,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004912701189521808,
      "loss": 5.3652,
      "mean_token_accuracy": 0.17971180081367494,
      "num_tokens": 21224959.0,
      "step": 11500
    },
    {
      "entropy": 5.736634159088135,
      "epoch": 0.9666036546943919,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004912618562636211,
      "loss": 5.4772,
      "mean_token_accuracy": 0.16933317929506303,
      "num_tokens": 21234495.0,
      "step": 11505
    },
    {
      "entropy": 5.653110933303833,
      "epoch": 0.9670237345095568,
      "grad_norm": 0.921875,
      "learning_rate": 0.000491253589744094,
      "loss": 5.3419,
      "mean_token_accuracy": 0.17508088797330856,
      "num_tokens": 21244555.0,
      "step": 11510
    },
    {
      "entropy": 5.7864092826843265,
      "epoch": 0.9674438143247217,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004912453193937459,
      "loss": 5.4699,
      "mean_token_accuracy": 0.17372365146875382,
      "num_tokens": 21254199.0,
      "step": 11515
    },
    {
      "entropy": 5.694921398162842,
      "epoch": 0.9678638941398866,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004912370452127234,
      "loss": 5.3654,
      "mean_token_accuracy": 0.1714448183774948,
      "num_tokens": 21262723.0,
      "step": 11520
    },
    {
      "entropy": 5.690516710281372,
      "epoch": 0.9682839739550515,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004912287672011728,
      "loss": 5.2819,
      "mean_token_accuracy": 0.18121316581964492,
      "num_tokens": 21271283.0,
      "step": 11525
    },
    {
      "entropy": 5.656310987472534,
      "epoch": 0.9687040537702163,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004912204853592411,
      "loss": 5.3391,
      "mean_token_accuracy": 0.18518260568380357,
      "num_tokens": 21279542.0,
      "step": 11530
    },
    {
      "entropy": 5.6403100967407225,
      "epoch": 0.9691241335853812,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004912121996870748,
      "loss": 5.3093,
      "mean_token_accuracy": 0.17596265971660613,
      "num_tokens": 21288678.0,
      "step": 11535
    },
    {
      "entropy": 5.713146638870239,
      "epoch": 0.9695442134005461,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004912039101848207,
      "loss": 5.4671,
      "mean_token_accuracy": 0.17476562708616256,
      "num_tokens": 21298982.0,
      "step": 11540
    },
    {
      "entropy": 5.712697839736938,
      "epoch": 0.969964293215711,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004911956168526257,
      "loss": 5.4,
      "mean_token_accuracy": 0.17398983091115952,
      "num_tokens": 21307663.0,
      "step": 11545
    },
    {
      "entropy": 5.746055030822754,
      "epoch": 0.9703843730308759,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004911873196906366,
      "loss": 5.4004,
      "mean_token_accuracy": 0.1680943712592125,
      "num_tokens": 21318004.0,
      "step": 11550
    },
    {
      "entropy": 5.617781686782837,
      "epoch": 0.9708044528460408,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004911790186990005,
      "loss": 5.22,
      "mean_token_accuracy": 0.19249431192874908,
      "num_tokens": 21327373.0,
      "step": 11555
    },
    {
      "entropy": 5.632816791534424,
      "epoch": 0.9712245326612057,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004911707138778643,
      "loss": 5.3331,
      "mean_token_accuracy": 0.18021902143955232,
      "num_tokens": 21335654.0,
      "step": 11560
    },
    {
      "entropy": 5.706795358657837,
      "epoch": 0.9716446124763705,
      "grad_norm": 1.0,
      "learning_rate": 0.0004911624052273754,
      "loss": 5.3962,
      "mean_token_accuracy": 0.17298138290643691,
      "num_tokens": 21344464.0,
      "step": 11565
    },
    {
      "entropy": 5.779402256011963,
      "epoch": 0.9720646922915354,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004911540927476807,
      "loss": 5.4707,
      "mean_token_accuracy": 0.17176631540060044,
      "num_tokens": 21354121.0,
      "step": 11570
    },
    {
      "entropy": 5.708774375915527,
      "epoch": 0.9724847721067003,
      "grad_norm": 1.0,
      "learning_rate": 0.0004911457764389275,
      "loss": 5.4073,
      "mean_token_accuracy": 0.1765870526432991,
      "num_tokens": 21363395.0,
      "step": 11575
    },
    {
      "entropy": 5.693147850036621,
      "epoch": 0.9729048519218652,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004911374563012633,
      "loss": 5.3705,
      "mean_token_accuracy": 0.17218163460493088,
      "num_tokens": 21372126.0,
      "step": 11580
    },
    {
      "entropy": 5.771610403060913,
      "epoch": 0.97332493173703,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004911291323348352,
      "loss": 5.4433,
      "mean_token_accuracy": 0.17207276523113252,
      "num_tokens": 21380554.0,
      "step": 11585
    },
    {
      "entropy": 5.688725519180298,
      "epoch": 0.973745011552195,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004911208045397909,
      "loss": 5.3368,
      "mean_token_accuracy": 0.17727535218000412,
      "num_tokens": 21389317.0,
      "step": 11590
    },
    {
      "entropy": 5.731136083602905,
      "epoch": 0.9741650913673598,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004911124729162778,
      "loss": 5.4392,
      "mean_token_accuracy": 0.17163546830415727,
      "num_tokens": 21398926.0,
      "step": 11595
    },
    {
      "entropy": 5.710000658035279,
      "epoch": 0.9745851711825246,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004911041374644435,
      "loss": 5.2571,
      "mean_token_accuracy": 0.18613769114017487,
      "num_tokens": 21406962.0,
      "step": 11600
    },
    {
      "entropy": 5.7084471702575685,
      "epoch": 0.9750052509976895,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004910957981844357,
      "loss": 5.3592,
      "mean_token_accuracy": 0.17707916647195815,
      "num_tokens": 21415868.0,
      "step": 11605
    },
    {
      "entropy": 5.7771618366241455,
      "epoch": 0.9754253308128544,
      "grad_norm": 1.0,
      "learning_rate": 0.0004910874550764022,
      "loss": 5.4719,
      "mean_token_accuracy": 0.17531527429819108,
      "num_tokens": 21424544.0,
      "step": 11610
    },
    {
      "entropy": 5.656285572052002,
      "epoch": 0.9758454106280193,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004910791081404907,
      "loss": 5.3397,
      "mean_token_accuracy": 0.17917796969413757,
      "num_tokens": 21433589.0,
      "step": 11615
    },
    {
      "entropy": 5.712646007537842,
      "epoch": 0.9762654904431842,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004910707573768489,
      "loss": 5.4094,
      "mean_token_accuracy": 0.16997252255678177,
      "num_tokens": 21442084.0,
      "step": 11620
    },
    {
      "entropy": 5.663222455978394,
      "epoch": 0.9766855702583491,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004910624027856251,
      "loss": 5.314,
      "mean_token_accuracy": 0.1778140112757683,
      "num_tokens": 21450962.0,
      "step": 11625
    },
    {
      "entropy": 5.719225645065308,
      "epoch": 0.977105650073514,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004910540443669669,
      "loss": 5.3955,
      "mean_token_accuracy": 0.17166452258825302,
      "num_tokens": 21461322.0,
      "step": 11630
    },
    {
      "entropy": 5.685926818847657,
      "epoch": 0.9775257298886788,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004910456821210227,
      "loss": 5.3881,
      "mean_token_accuracy": 0.17719679772853852,
      "num_tokens": 21470800.0,
      "step": 11635
    },
    {
      "entropy": 5.668769216537475,
      "epoch": 0.9779458097038437,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004910373160479404,
      "loss": 5.2373,
      "mean_token_accuracy": 0.1857740804553032,
      "num_tokens": 21479707.0,
      "step": 11640
    },
    {
      "entropy": 5.681134605407715,
      "epoch": 0.9783658895190086,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004910289461478683,
      "loss": 5.4103,
      "mean_token_accuracy": 0.17056867629289627,
      "num_tokens": 21489469.0,
      "step": 11645
    },
    {
      "entropy": 5.721190071105957,
      "epoch": 0.9787859693341735,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004910205724209547,
      "loss": 5.3686,
      "mean_token_accuracy": 0.17375414818525314,
      "num_tokens": 21499226.0,
      "step": 11650
    },
    {
      "entropy": 5.635466384887695,
      "epoch": 0.9792060491493384,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004910121948673478,
      "loss": 5.2604,
      "mean_token_accuracy": 0.18351707756519317,
      "num_tokens": 21508129.0,
      "step": 11655
    },
    {
      "entropy": 5.65350227355957,
      "epoch": 0.9796261289645033,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004910038134871962,
      "loss": 5.2906,
      "mean_token_accuracy": 0.18378447443246843,
      "num_tokens": 21516293.0,
      "step": 11660
    },
    {
      "entropy": 5.763593530654907,
      "epoch": 0.9800462087796681,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004909954282806482,
      "loss": 5.4536,
      "mean_token_accuracy": 0.174176824092865,
      "num_tokens": 21525393.0,
      "step": 11665
    },
    {
      "entropy": 5.664162254333496,
      "epoch": 0.980466288594833,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004909870392478524,
      "loss": 5.3104,
      "mean_token_accuracy": 0.17742753475904466,
      "num_tokens": 21534585.0,
      "step": 11670
    },
    {
      "entropy": 5.621233224868774,
      "epoch": 0.9808863684099979,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004909786463889575,
      "loss": 5.2376,
      "mean_token_accuracy": 0.18252046257257462,
      "num_tokens": 21542947.0,
      "step": 11675
    },
    {
      "entropy": 5.683218097686767,
      "epoch": 0.9813064482251628,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004909702497041121,
      "loss": 5.3445,
      "mean_token_accuracy": 0.1765931636095047,
      "num_tokens": 21552168.0,
      "step": 11680
    },
    {
      "entropy": 5.708033514022827,
      "epoch": 0.9817265280403277,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004909618491934648,
      "loss": 5.3741,
      "mean_token_accuracy": 0.17811677902936934,
      "num_tokens": 21562131.0,
      "step": 11685
    },
    {
      "entropy": 5.656235837936402,
      "epoch": 0.9821466078554926,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004909534448571647,
      "loss": 5.311,
      "mean_token_accuracy": 0.18598451614379882,
      "num_tokens": 21571363.0,
      "step": 11690
    },
    {
      "entropy": 5.662360334396363,
      "epoch": 0.9825666876706575,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004909450366953604,
      "loss": 5.2964,
      "mean_token_accuracy": 0.18597533851861953,
      "num_tokens": 21580754.0,
      "step": 11695
    },
    {
      "entropy": 5.6616133689880375,
      "epoch": 0.9829867674858223,
      "grad_norm": 0.96875,
      "learning_rate": 0.000490936624708201,
      "loss": 5.408,
      "mean_token_accuracy": 0.17438692450523377,
      "num_tokens": 21590053.0,
      "step": 11700
    },
    {
      "entropy": 5.700369262695313,
      "epoch": 0.9834068473009872,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004909282088958356,
      "loss": 5.338,
      "mean_token_accuracy": 0.17859577238559723,
      "num_tokens": 21598681.0,
      "step": 11705
    },
    {
      "entropy": 5.755769157409668,
      "epoch": 0.983826927116152,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000490919789258413,
      "loss": 5.3609,
      "mean_token_accuracy": 0.18359744995832444,
      "num_tokens": 21607465.0,
      "step": 11710
    },
    {
      "entropy": 5.727117586135864,
      "epoch": 0.984247006931317,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004909113657960826,
      "loss": 5.4323,
      "mean_token_accuracy": 0.1684650734066963,
      "num_tokens": 21617480.0,
      "step": 11715
    },
    {
      "entropy": 5.641330623626709,
      "epoch": 0.9846670867464818,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004909029385089935,
      "loss": 5.3549,
      "mean_token_accuracy": 0.17785272151231765,
      "num_tokens": 21626434.0,
      "step": 11720
    },
    {
      "entropy": 5.672800159454345,
      "epoch": 0.9850871665616467,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000490894507397295,
      "loss": 5.3474,
      "mean_token_accuracy": 0.18236180692911147,
      "num_tokens": 21635627.0,
      "step": 11725
    },
    {
      "entropy": 5.739930486679077,
      "epoch": 0.9855072463768116,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004908860724611365,
      "loss": 5.3472,
      "mean_token_accuracy": 0.17631836086511612,
      "num_tokens": 21644789.0,
      "step": 11730
    },
    {
      "entropy": 5.644730091094971,
      "epoch": 0.9859273261919764,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004908776337006675,
      "loss": 5.349,
      "mean_token_accuracy": 0.1759209454059601,
      "num_tokens": 21653696.0,
      "step": 11735
    },
    {
      "entropy": 5.687325096130371,
      "epoch": 0.9863474060071413,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004908691911160373,
      "loss": 5.3355,
      "mean_token_accuracy": 0.16616807430982589,
      "num_tokens": 21664420.0,
      "step": 11740
    },
    {
      "entropy": 5.675135755538941,
      "epoch": 0.9867674858223062,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004908607447073954,
      "loss": 5.325,
      "mean_token_accuracy": 0.18100758790969848,
      "num_tokens": 21673716.0,
      "step": 11745
    },
    {
      "entropy": 5.691566371917725,
      "epoch": 0.9871875656374711,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004908522944748917,
      "loss": 5.3231,
      "mean_token_accuracy": 0.1852756157517433,
      "num_tokens": 21682860.0,
      "step": 11750
    },
    {
      "entropy": 5.591435766220092,
      "epoch": 0.987607645452636,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004908438404186758,
      "loss": 5.3241,
      "mean_token_accuracy": 0.18076315075159072,
      "num_tokens": 21691915.0,
      "step": 11755
    },
    {
      "entropy": 5.729312562942505,
      "epoch": 0.9880277252678009,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004908353825388973,
      "loss": 5.4449,
      "mean_token_accuracy": 0.1711159124970436,
      "num_tokens": 21701666.0,
      "step": 11760
    },
    {
      "entropy": 5.779921340942383,
      "epoch": 0.9884478050829658,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004908269208357062,
      "loss": 5.3752,
      "mean_token_accuracy": 0.18159790635108947,
      "num_tokens": 21709267.0,
      "step": 11765
    },
    {
      "entropy": 5.637835550308227,
      "epoch": 0.9888678848981306,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004908184553092523,
      "loss": 5.2625,
      "mean_token_accuracy": 0.17955298721790314,
      "num_tokens": 21718117.0,
      "step": 11770
    },
    {
      "entropy": 5.689300107955932,
      "epoch": 0.9892879647132955,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004908099859596856,
      "loss": 5.4132,
      "mean_token_accuracy": 0.1810736581683159,
      "num_tokens": 21727952.0,
      "step": 11775
    },
    {
      "entropy": 5.712216997146607,
      "epoch": 0.9897080445284604,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004908015127871561,
      "loss": 5.2623,
      "mean_token_accuracy": 0.17740319967269896,
      "num_tokens": 21737878.0,
      "step": 11780
    },
    {
      "entropy": 5.628698682785034,
      "epoch": 0.9901281243436253,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000490793035791814,
      "loss": 5.2306,
      "mean_token_accuracy": 0.1800789549946785,
      "num_tokens": 21747391.0,
      "step": 11785
    },
    {
      "entropy": 5.638668870925903,
      "epoch": 0.9905482041587902,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004907845549738093,
      "loss": 5.2565,
      "mean_token_accuracy": 0.18442873507738114,
      "num_tokens": 21756791.0,
      "step": 11790
    },
    {
      "entropy": 5.600989484786988,
      "epoch": 0.9909682839739551,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004907760703332923,
      "loss": 5.3115,
      "mean_token_accuracy": 0.17620126605033876,
      "num_tokens": 21766020.0,
      "step": 11795
    },
    {
      "entropy": 5.739794158935547,
      "epoch": 0.99138836378912,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004907675818704134,
      "loss": 5.4094,
      "mean_token_accuracy": 0.16773414760828018,
      "num_tokens": 21775895.0,
      "step": 11800
    },
    {
      "entropy": 5.661771154403686,
      "epoch": 0.9918084436042848,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004907590895853228,
      "loss": 5.3327,
      "mean_token_accuracy": 0.17859974652528762,
      "num_tokens": 21784543.0,
      "step": 11805
    },
    {
      "entropy": 5.653049087524414,
      "epoch": 0.9922285234194497,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004907505934781712,
      "loss": 5.3859,
      "mean_token_accuracy": 0.1735259398818016,
      "num_tokens": 21793938.0,
      "step": 11810
    },
    {
      "entropy": 5.729407882690429,
      "epoch": 0.9926486032346146,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004907420935491087,
      "loss": 5.3505,
      "mean_token_accuracy": 0.1748668819665909,
      "num_tokens": 21803641.0,
      "step": 11815
    },
    {
      "entropy": 5.696050357818604,
      "epoch": 0.9930686830497795,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004907335897982862,
      "loss": 5.2707,
      "mean_token_accuracy": 0.18189048767089844,
      "num_tokens": 21812542.0,
      "step": 11820
    },
    {
      "entropy": 5.617125511169434,
      "epoch": 0.9934887628649444,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004907250822258543,
      "loss": 5.3619,
      "mean_token_accuracy": 0.17457778453826905,
      "num_tokens": 21821847.0,
      "step": 11825
    },
    {
      "entropy": 5.7730554103851315,
      "epoch": 0.9939088426801093,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004907165708319637,
      "loss": 5.4144,
      "mean_token_accuracy": 0.17439746856689453,
      "num_tokens": 21830799.0,
      "step": 11830
    },
    {
      "entropy": 5.680804634094239,
      "epoch": 0.994328922495274,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004907080556167651,
      "loss": 5.3257,
      "mean_token_accuracy": 0.17789210379123688,
      "num_tokens": 21840202.0,
      "step": 11835
    },
    {
      "entropy": 5.715157318115234,
      "epoch": 0.994749002310439,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004906995365804093,
      "loss": 5.4193,
      "mean_token_accuracy": 0.17181181609630586,
      "num_tokens": 21849701.0,
      "step": 11840
    },
    {
      "entropy": 5.706198883056641,
      "epoch": 0.9951690821256038,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004906910137230472,
      "loss": 5.3188,
      "mean_token_accuracy": 0.18163561522960664,
      "num_tokens": 21859191.0,
      "step": 11845
    },
    {
      "entropy": 5.682278156280518,
      "epoch": 0.9955891619407687,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00049068248704483,
      "loss": 5.3096,
      "mean_token_accuracy": 0.17822230458259583,
      "num_tokens": 21867944.0,
      "step": 11850
    },
    {
      "entropy": 5.619589805603027,
      "epoch": 0.9960092417559336,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004906739565459085,
      "loss": 5.3422,
      "mean_token_accuracy": 0.17374778985977174,
      "num_tokens": 21876368.0,
      "step": 11855
    },
    {
      "entropy": 5.767853784561157,
      "epoch": 0.9964293215710985,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000490665422226434,
      "loss": 5.4246,
      "mean_token_accuracy": 0.16895324289798735,
      "num_tokens": 21885634.0,
      "step": 11860
    },
    {
      "entropy": 5.605723142623901,
      "epoch": 0.9968494013862634,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004906568840865576,
      "loss": 5.2408,
      "mean_token_accuracy": 0.18185186982154847,
      "num_tokens": 21894315.0,
      "step": 11865
    },
    {
      "entropy": 5.6185742855072025,
      "epoch": 0.9972694812014282,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004906483421264305,
      "loss": 5.3504,
      "mean_token_accuracy": 0.18069129288196564,
      "num_tokens": 21903342.0,
      "step": 11870
    },
    {
      "entropy": 5.72865481376648,
      "epoch": 0.9976895610165931,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000490639796346204,
      "loss": 5.4659,
      "mean_token_accuracy": 0.16910600364208223,
      "num_tokens": 21914158.0,
      "step": 11875
    },
    {
      "entropy": 5.76618127822876,
      "epoch": 0.998109640831758,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004906312467460297,
      "loss": 5.3706,
      "mean_token_accuracy": 0.17504575103521347,
      "num_tokens": 21922639.0,
      "step": 11880
    },
    {
      "entropy": 5.658193683624267,
      "epoch": 0.9985297206469229,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004906226933260588,
      "loss": 5.3224,
      "mean_token_accuracy": 0.17745455950498581,
      "num_tokens": 21931385.0,
      "step": 11885
    },
    {
      "entropy": 5.714499616622925,
      "epoch": 0.9989498004620878,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004906141360864429,
      "loss": 5.3509,
      "mean_token_accuracy": 0.17270935773849488,
      "num_tokens": 21940788.0,
      "step": 11890
    },
    {
      "entropy": 5.725809097290039,
      "epoch": 0.9993698802772527,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004906055750273336,
      "loss": 5.3635,
      "mean_token_accuracy": 0.18086509853601457,
      "num_tokens": 21950309.0,
      "step": 11895
    },
    {
      "entropy": 5.65466947555542,
      "epoch": 0.9997899600924176,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004905970101488826,
      "loss": 5.37,
      "mean_token_accuracy": 0.18021659702062606,
      "num_tokens": 21959141.0,
      "step": 11900
    },
    {
      "entropy": 5.745198143853082,
      "epoch": 1.000168031926066,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004905884414512416,
      "loss": 5.3987,
      "mean_token_accuracy": 0.18021918170981938,
      "num_tokens": 21966665.0,
      "step": 11905
    },
    {
      "entropy": 5.720191812515258,
      "epoch": 1.0005881117412307,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004905798689345623,
      "loss": 5.3374,
      "mean_token_accuracy": 0.1821998745203018,
      "num_tokens": 21976728.0,
      "step": 11910
    },
    {
      "entropy": 5.656696033477783,
      "epoch": 1.0010081915563958,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004905712925989968,
      "loss": 5.199,
      "mean_token_accuracy": 0.17787499874830245,
      "num_tokens": 21985915.0,
      "step": 11915
    },
    {
      "entropy": 5.6643424987792965,
      "epoch": 1.0014282713715605,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004905627124446967,
      "loss": 5.2414,
      "mean_token_accuracy": 0.17900198549032212,
      "num_tokens": 21995826.0,
      "step": 11920
    },
    {
      "entropy": 5.644786405563354,
      "epoch": 1.0018483511867255,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004905541284718142,
      "loss": 5.2058,
      "mean_token_accuracy": 0.18394928723573684,
      "num_tokens": 22005299.0,
      "step": 11925
    },
    {
      "entropy": 5.66067156791687,
      "epoch": 1.0022684310018903,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004905455406805011,
      "loss": 5.2416,
      "mean_token_accuracy": 0.1824041485786438,
      "num_tokens": 22014499.0,
      "step": 11930
    },
    {
      "entropy": 5.73095474243164,
      "epoch": 1.0026885108170553,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00049053694907091,
      "loss": 5.3911,
      "mean_token_accuracy": 0.17511750459671022,
      "num_tokens": 22024531.0,
      "step": 11935
    },
    {
      "entropy": 5.708370590209961,
      "epoch": 1.0031085906322201,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004905283536431928,
      "loss": 5.3003,
      "mean_token_accuracy": 0.17841449081897737,
      "num_tokens": 22034036.0,
      "step": 11940
    },
    {
      "entropy": 5.661829805374145,
      "epoch": 1.003528670447385,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004905197543975017,
      "loss": 5.2089,
      "mean_token_accuracy": 0.18137877881526948,
      "num_tokens": 22042910.0,
      "step": 11945
    },
    {
      "entropy": 5.711597871780396,
      "epoch": 1.00394875026255,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004905111513339892,
      "loss": 5.2958,
      "mean_token_accuracy": 0.18205167055130006,
      "num_tokens": 22052242.0,
      "step": 11950
    },
    {
      "entropy": 5.6694591522216795,
      "epoch": 1.0043688300777147,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004905025444528076,
      "loss": 5.2721,
      "mean_token_accuracy": 0.17857326120138167,
      "num_tokens": 22061467.0,
      "step": 11955
    },
    {
      "entropy": 5.5813757419586185,
      "epoch": 1.0047889098928797,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004904939337541093,
      "loss": 5.1285,
      "mean_token_accuracy": 0.18635020554065704,
      "num_tokens": 22070300.0,
      "step": 11960
    },
    {
      "entropy": 5.714624357223511,
      "epoch": 1.0052089897080445,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004904853192380472,
      "loss": 5.2974,
      "mean_token_accuracy": 0.1818109780550003,
      "num_tokens": 22078960.0,
      "step": 11965
    },
    {
      "entropy": 5.67142391204834,
      "epoch": 1.0056290695232095,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004904767009047733,
      "loss": 5.2415,
      "mean_token_accuracy": 0.18054623007774354,
      "num_tokens": 22088135.0,
      "step": 11970
    },
    {
      "entropy": 5.702590990066528,
      "epoch": 1.0060491493383743,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004904680787544408,
      "loss": 5.3527,
      "mean_token_accuracy": 0.17716111987829208,
      "num_tokens": 22098004.0,
      "step": 11975
    },
    {
      "entropy": 5.754019546508789,
      "epoch": 1.006469229153539,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004904594527872022,
      "loss": 5.3232,
      "mean_token_accuracy": 0.1762009248137474,
      "num_tokens": 22107680.0,
      "step": 11980
    },
    {
      "entropy": 5.712886142730713,
      "epoch": 1.006889308968704,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004904508230032103,
      "loss": 5.2959,
      "mean_token_accuracy": 0.1786690518260002,
      "num_tokens": 22118004.0,
      "step": 11985
    },
    {
      "entropy": 5.686515188217163,
      "epoch": 1.0073093887838689,
      "grad_norm": 1.1875,
      "learning_rate": 0.000490442189402618,
      "loss": 5.2622,
      "mean_token_accuracy": 0.18934257626533507,
      "num_tokens": 22127825.0,
      "step": 11990
    },
    {
      "entropy": 5.665019083023071,
      "epoch": 1.007729468599034,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004904335519855783,
      "loss": 5.192,
      "mean_token_accuracy": 0.18694633692502977,
      "num_tokens": 22136448.0,
      "step": 11995
    },
    {
      "entropy": 5.659635210037232,
      "epoch": 1.0081495484141987,
      "grad_norm": 1.0,
      "learning_rate": 0.0004904249107522442,
      "loss": 5.2987,
      "mean_token_accuracy": 0.18026018887758255,
      "num_tokens": 22146415.0,
      "step": 12000
    },
    {
      "epoch": 1.0081495484141987,
      "eval_entropy": 5.424837325068975,
      "eval_loss": 5.348167419433594,
      "eval_mean_token_accuracy": 0.1852694718273657,
      "eval_num_tokens": 22146415.0,
      "eval_runtime": 20.9126,
      "eval_samples_per_second": 1786.766,
      "eval_steps_per_second": 223.358,
      "step": 12000
    },
    {
      "entropy": 5.729191446304322,
      "epoch": 1.0085696282293637,
      "grad_norm": 1.125,
      "learning_rate": 0.0004904162657027685,
      "loss": 5.4359,
      "mean_token_accuracy": 0.1786647319793701,
      "num_tokens": 22156327.0,
      "step": 12005
    },
    {
      "entropy": 5.648891830444336,
      "epoch": 1.0089897080445285,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004904076168373049,
      "loss": 5.2254,
      "mean_token_accuracy": 0.1822052225470543,
      "num_tokens": 22165677.0,
      "step": 12010
    },
    {
      "entropy": 5.696646451950073,
      "epoch": 1.0094097878596933,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004903989641560061,
      "loss": 5.3605,
      "mean_token_accuracy": 0.1811189815402031,
      "num_tokens": 22175232.0,
      "step": 12015
    },
    {
      "entropy": 5.7059225082397464,
      "epoch": 1.0098298676748583,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004903903076590256,
      "loss": 5.2484,
      "mean_token_accuracy": 0.18019332736730576,
      "num_tokens": 22184026.0,
      "step": 12020
    },
    {
      "entropy": 5.604170560836792,
      "epoch": 1.010249947490023,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004903816473465167,
      "loss": 5.1484,
      "mean_token_accuracy": 0.19093165397644044,
      "num_tokens": 22192020.0,
      "step": 12025
    },
    {
      "entropy": 5.581792259216309,
      "epoch": 1.010670027305188,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004903729832186328,
      "loss": 5.1306,
      "mean_token_accuracy": 0.19227272570133208,
      "num_tokens": 22200060.0,
      "step": 12030
    },
    {
      "entropy": 5.639505529403687,
      "epoch": 1.0110901071203529,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004903643152755274,
      "loss": 5.1986,
      "mean_token_accuracy": 0.1794219434261322,
      "num_tokens": 22208625.0,
      "step": 12035
    },
    {
      "entropy": 5.621885538101196,
      "epoch": 1.0115101869355176,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004903556435173541,
      "loss": 5.1696,
      "mean_token_accuracy": 0.1890147164463997,
      "num_tokens": 22217781.0,
      "step": 12040
    },
    {
      "entropy": 5.655472707748413,
      "epoch": 1.0119302667506826,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004903469679442665,
      "loss": 5.3033,
      "mean_token_accuracy": 0.17978367507457732,
      "num_tokens": 22226432.0,
      "step": 12045
    },
    {
      "entropy": 5.643258810043335,
      "epoch": 1.0123503465658474,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004903382885564181,
      "loss": 5.3029,
      "mean_token_accuracy": 0.18215753883123398,
      "num_tokens": 22234811.0,
      "step": 12050
    },
    {
      "entropy": 5.606886100769043,
      "epoch": 1.0127704263810124,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000490329605353963,
      "loss": 5.1914,
      "mean_token_accuracy": 0.18768127709627153,
      "num_tokens": 22242808.0,
      "step": 12055
    },
    {
      "entropy": 5.751883697509766,
      "epoch": 1.0131905061961772,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004903209183370547,
      "loss": 5.2486,
      "mean_token_accuracy": 0.1865461513400078,
      "num_tokens": 22251371.0,
      "step": 12060
    },
    {
      "entropy": 5.789570379257202,
      "epoch": 1.0136105860113422,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004903122275058472,
      "loss": 5.33,
      "mean_token_accuracy": 0.18375689536333084,
      "num_tokens": 22260868.0,
      "step": 12065
    },
    {
      "entropy": 5.5965673446655275,
      "epoch": 1.014030665826507,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004903035328604944,
      "loss": 5.2269,
      "mean_token_accuracy": 0.1812163144350052,
      "num_tokens": 22270554.0,
      "step": 12070
    },
    {
      "entropy": 5.596884918212891,
      "epoch": 1.0144507456416718,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004902948344011506,
      "loss": 5.2352,
      "mean_token_accuracy": 0.17911261767148973,
      "num_tokens": 22279170.0,
      "step": 12075
    },
    {
      "entropy": 5.700472497940064,
      "epoch": 1.0148708254568368,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004902861321279694,
      "loss": 5.3621,
      "mean_token_accuracy": 0.1765124335885048,
      "num_tokens": 22288788.0,
      "step": 12080
    },
    {
      "entropy": 5.638234233856201,
      "epoch": 1.0152909052720016,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004902774260411055,
      "loss": 5.1702,
      "mean_token_accuracy": 0.18393722921609879,
      "num_tokens": 22297501.0,
      "step": 12085
    },
    {
      "entropy": 5.619339513778686,
      "epoch": 1.0157109850871666,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004902687161407126,
      "loss": 5.1258,
      "mean_token_accuracy": 0.1942564532160759,
      "num_tokens": 22306181.0,
      "step": 12090
    },
    {
      "entropy": 5.600225639343262,
      "epoch": 1.0161310649023314,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004902600024269454,
      "loss": 5.2654,
      "mean_token_accuracy": 0.1864081636071205,
      "num_tokens": 22315762.0,
      "step": 12095
    },
    {
      "entropy": 5.5886448383331295,
      "epoch": 1.0165511447174964,
      "grad_norm": 1.0625,
      "learning_rate": 0.000490251284899958,
      "loss": 5.2191,
      "mean_token_accuracy": 0.1840157151222229,
      "num_tokens": 22325127.0,
      "step": 12100
    },
    {
      "entropy": 5.6344461917877195,
      "epoch": 1.0169712245326612,
      "grad_norm": 1.046875,
      "learning_rate": 0.000490242563559905,
      "loss": 5.3013,
      "mean_token_accuracy": 0.17882361710071565,
      "num_tokens": 22334038.0,
      "step": 12105
    },
    {
      "entropy": 5.644755268096924,
      "epoch": 1.017391304347826,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004902338384069408,
      "loss": 5.1722,
      "mean_token_accuracy": 0.18300676792860032,
      "num_tokens": 22342658.0,
      "step": 12110
    },
    {
      "entropy": 5.709315013885498,
      "epoch": 1.017811384162991,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049022510944122,
      "loss": 5.3214,
      "mean_token_accuracy": 0.17637363970279693,
      "num_tokens": 22352559.0,
      "step": 12115
    },
    {
      "entropy": 5.723194980621338,
      "epoch": 1.0182314639781558,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004902163766628972,
      "loss": 5.2503,
      "mean_token_accuracy": 0.1825407102704048,
      "num_tokens": 22361455.0,
      "step": 12120
    },
    {
      "entropy": 5.728067684173584,
      "epoch": 1.0186515437933208,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004902076400721271,
      "loss": 5.2558,
      "mean_token_accuracy": 0.1816166654229164,
      "num_tokens": 22371163.0,
      "step": 12125
    },
    {
      "entropy": 5.73334002494812,
      "epoch": 1.0190716236084856,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004901988996690645,
      "loss": 5.2816,
      "mean_token_accuracy": 0.18191347718238832,
      "num_tokens": 22379975.0,
      "step": 12130
    },
    {
      "entropy": 5.731075668334961,
      "epoch": 1.0194917034236506,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004901901554538641,
      "loss": 5.3004,
      "mean_token_accuracy": 0.1808861345052719,
      "num_tokens": 22389657.0,
      "step": 12135
    },
    {
      "entropy": 5.615172624588013,
      "epoch": 1.0199117832388154,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000490181407426681,
      "loss": 5.1486,
      "mean_token_accuracy": 0.18932715505361558,
      "num_tokens": 22398320.0,
      "step": 12140
    },
    {
      "entropy": 5.69921350479126,
      "epoch": 1.0203318630539802,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004901726555876701,
      "loss": 5.3364,
      "mean_token_accuracy": 0.17337931394577027,
      "num_tokens": 22406634.0,
      "step": 12145
    },
    {
      "entropy": 5.718554162979126,
      "epoch": 1.0207519428691452,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004901638999369862,
      "loss": 5.3819,
      "mean_token_accuracy": 0.17468169182538987,
      "num_tokens": 22415939.0,
      "step": 12150
    },
    {
      "entropy": 5.657704448699951,
      "epoch": 1.02117202268431,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004901551404747847,
      "loss": 5.3277,
      "mean_token_accuracy": 0.17230201959609986,
      "num_tokens": 22425256.0,
      "step": 12155
    },
    {
      "entropy": 5.689808464050293,
      "epoch": 1.021592102499475,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004901463772012209,
      "loss": 5.3621,
      "mean_token_accuracy": 0.17679853439331056,
      "num_tokens": 22434750.0,
      "step": 12160
    },
    {
      "entropy": 5.699192380905151,
      "epoch": 1.0220121823146397,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004901376101164495,
      "loss": 5.269,
      "mean_token_accuracy": 0.1764159694314003,
      "num_tokens": 22443426.0,
      "step": 12165
    },
    {
      "entropy": 5.67895998954773,
      "epoch": 1.0224322621298048,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004901288392206263,
      "loss": 5.2548,
      "mean_token_accuracy": 0.1735469728708267,
      "num_tokens": 22452778.0,
      "step": 12170
    },
    {
      "entropy": 5.630224895477295,
      "epoch": 1.0228523419449695,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004901200645139064,
      "loss": 5.217,
      "mean_token_accuracy": 0.18821492344141005,
      "num_tokens": 22462864.0,
      "step": 12175
    },
    {
      "entropy": 5.627038049697876,
      "epoch": 1.0232724217601343,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004901112859964454,
      "loss": 5.2763,
      "mean_token_accuracy": 0.17913248836994172,
      "num_tokens": 22472849.0,
      "step": 12180
    },
    {
      "entropy": 5.654596328735352,
      "epoch": 1.0236925015752993,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004901025036683987,
      "loss": 5.2004,
      "mean_token_accuracy": 0.17802743762731552,
      "num_tokens": 22481693.0,
      "step": 12185
    },
    {
      "entropy": 5.646336936950684,
      "epoch": 1.0241125813904641,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004900937175299219,
      "loss": 5.2324,
      "mean_token_accuracy": 0.18271609246730805,
      "num_tokens": 22490934.0,
      "step": 12190
    },
    {
      "entropy": 5.5874090671539305,
      "epoch": 1.0245326612056291,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004900849275811707,
      "loss": 5.2652,
      "mean_token_accuracy": 0.17981857657432557,
      "num_tokens": 22500457.0,
      "step": 12195
    },
    {
      "entropy": 5.645065021514893,
      "epoch": 1.024952741020794,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004900761338223007,
      "loss": 5.2031,
      "mean_token_accuracy": 0.18214412033557892,
      "num_tokens": 22509641.0,
      "step": 12200
    },
    {
      "entropy": 5.652515459060669,
      "epoch": 1.025372820835959,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004900673362534677,
      "loss": 5.1541,
      "mean_token_accuracy": 0.1891545057296753,
      "num_tokens": 22518616.0,
      "step": 12205
    },
    {
      "entropy": 5.641676378250122,
      "epoch": 1.0257929006511237,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004900585348748277,
      "loss": 5.2923,
      "mean_token_accuracy": 0.18081348687410354,
      "num_tokens": 22527599.0,
      "step": 12210
    },
    {
      "entropy": 5.623777151107788,
      "epoch": 1.0262129804662885,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004900497296865365,
      "loss": 5.2989,
      "mean_token_accuracy": 0.17388005405664445,
      "num_tokens": 22537399.0,
      "step": 12215
    },
    {
      "entropy": 5.841449975967407,
      "epoch": 1.0266330602814535,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004900409206887499,
      "loss": 5.4693,
      "mean_token_accuracy": 0.17230376452207566,
      "num_tokens": 22546746.0,
      "step": 12220
    },
    {
      "entropy": 5.74645504951477,
      "epoch": 1.0270531400966183,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004900321078816243,
      "loss": 5.2814,
      "mean_token_accuracy": 0.1865550920367241,
      "num_tokens": 22555735.0,
      "step": 12225
    },
    {
      "entropy": 5.705350399017334,
      "epoch": 1.0274732199117833,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004900232912653156,
      "loss": 5.2748,
      "mean_token_accuracy": 0.17938363701105117,
      "num_tokens": 22565010.0,
      "step": 12230
    },
    {
      "entropy": 5.651552391052246,
      "epoch": 1.027893299726948,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00049001447083998,
      "loss": 5.2571,
      "mean_token_accuracy": 0.1776842936873436,
      "num_tokens": 22573565.0,
      "step": 12235
    },
    {
      "entropy": 5.658456611633301,
      "epoch": 1.028313379542113,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004900056466057737,
      "loss": 5.2634,
      "mean_token_accuracy": 0.1766846224665642,
      "num_tokens": 22582549.0,
      "step": 12240
    },
    {
      "entropy": 5.650886583328247,
      "epoch": 1.028733459357278,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004899968185628531,
      "loss": 5.2903,
      "mean_token_accuracy": 0.17502767890691756,
      "num_tokens": 22592112.0,
      "step": 12245
    },
    {
      "entropy": 5.61941876411438,
      "epoch": 1.0291535391724427,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004899879867113746,
      "loss": 5.1609,
      "mean_token_accuracy": 0.18664622008800508,
      "num_tokens": 22600581.0,
      "step": 12250
    },
    {
      "entropy": 5.728427410125732,
      "epoch": 1.0295736189876077,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004899791510514945,
      "loss": 5.3422,
      "mean_token_accuracy": 0.1744697704911232,
      "num_tokens": 22610822.0,
      "step": 12255
    },
    {
      "entropy": 5.699223279953003,
      "epoch": 1.0299936988027725,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004899703115833696,
      "loss": 5.3666,
      "mean_token_accuracy": 0.17521671056747437,
      "num_tokens": 22619484.0,
      "step": 12260
    },
    {
      "entropy": 5.643547582626343,
      "epoch": 1.0304137786179375,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004899614683071563,
      "loss": 5.1902,
      "mean_token_accuracy": 0.18214986920356752,
      "num_tokens": 22629038.0,
      "step": 12265
    },
    {
      "entropy": 5.699086380004883,
      "epoch": 1.0308338584331023,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004899526212230112,
      "loss": 5.296,
      "mean_token_accuracy": 0.17284860014915465,
      "num_tokens": 22638619.0,
      "step": 12270
    },
    {
      "entropy": 5.637109327316284,
      "epoch": 1.0312539382482673,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004899437703310912,
      "loss": 5.2669,
      "mean_token_accuracy": 0.17988306730985643,
      "num_tokens": 22648065.0,
      "step": 12275
    },
    {
      "entropy": 5.690742111206054,
      "epoch": 1.031674018063432,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004899349156315529,
      "loss": 5.3259,
      "mean_token_accuracy": 0.17899182438850403,
      "num_tokens": 22658107.0,
      "step": 12280
    },
    {
      "entropy": 5.66560926437378,
      "epoch": 1.0320940978785969,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004899260571245533,
      "loss": 5.2239,
      "mean_token_accuracy": 0.18416753262281418,
      "num_tokens": 22667103.0,
      "step": 12285
    },
    {
      "entropy": 5.64478554725647,
      "epoch": 1.0325141776937619,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004899171948102492,
      "loss": 5.1982,
      "mean_token_accuracy": 0.18252110481262207,
      "num_tokens": 22676792.0,
      "step": 12290
    },
    {
      "entropy": 5.6361888408660885,
      "epoch": 1.0329342575089266,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004899083286887977,
      "loss": 5.2241,
      "mean_token_accuracy": 0.18449742943048478,
      "num_tokens": 22685344.0,
      "step": 12295
    },
    {
      "entropy": 5.721011543273926,
      "epoch": 1.0333543373240917,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004898994587603559,
      "loss": 5.2853,
      "mean_token_accuracy": 0.17731328159570695,
      "num_tokens": 22694387.0,
      "step": 12300
    },
    {
      "entropy": 5.6244641780853275,
      "epoch": 1.0337744171392564,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004898905850250807,
      "loss": 5.3191,
      "mean_token_accuracy": 0.1827871397137642,
      "num_tokens": 22704203.0,
      "step": 12305
    },
    {
      "entropy": 5.699512958526611,
      "epoch": 1.0341944969544214,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004898817074831295,
      "loss": 5.3669,
      "mean_token_accuracy": 0.17662356346845626,
      "num_tokens": 22713518.0,
      "step": 12310
    },
    {
      "entropy": 5.738121795654297,
      "epoch": 1.0346145767695862,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004898728261346595,
      "loss": 5.3604,
      "mean_token_accuracy": 0.17533956468105316,
      "num_tokens": 22722997.0,
      "step": 12315
    },
    {
      "entropy": 5.726660442352295,
      "epoch": 1.035034656584751,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000489863940979828,
      "loss": 5.2939,
      "mean_token_accuracy": 0.176972895860672,
      "num_tokens": 22732385.0,
      "step": 12320
    },
    {
      "entropy": 5.6302389144897464,
      "epoch": 1.035454736399916,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004898550520187925,
      "loss": 5.1951,
      "mean_token_accuracy": 0.18711263686418533,
      "num_tokens": 22741148.0,
      "step": 12325
    },
    {
      "entropy": 5.628175973892212,
      "epoch": 1.0358748162150808,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004898461592517103,
      "loss": 5.1782,
      "mean_token_accuracy": 0.18838218003511428,
      "num_tokens": 22750239.0,
      "step": 12330
    },
    {
      "entropy": 5.7032843112945555,
      "epoch": 1.0362948960302458,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004898372626787391,
      "loss": 5.3331,
      "mean_token_accuracy": 0.17208956331014633,
      "num_tokens": 22759290.0,
      "step": 12335
    },
    {
      "entropy": 5.741462564468383,
      "epoch": 1.0367149758454106,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004898283623000364,
      "loss": 5.3542,
      "mean_token_accuracy": 0.17169689387083054,
      "num_tokens": 22768450.0,
      "step": 12340
    },
    {
      "entropy": 5.670726251602173,
      "epoch": 1.0371350556605754,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004898194581157598,
      "loss": 5.2246,
      "mean_token_accuracy": 0.17666156888008117,
      "num_tokens": 22777711.0,
      "step": 12345
    },
    {
      "entropy": 5.6758898258209225,
      "epoch": 1.0375551354757404,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004898105501260671,
      "loss": 5.3054,
      "mean_token_accuracy": 0.1748965635895729,
      "num_tokens": 22787153.0,
      "step": 12350
    },
    {
      "entropy": 5.730099678039551,
      "epoch": 1.0379752152909052,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004898016383311163,
      "loss": 5.3086,
      "mean_token_accuracy": 0.1841848835349083,
      "num_tokens": 22797125.0,
      "step": 12355
    },
    {
      "entropy": 5.7143354415893555,
      "epoch": 1.0383952951060702,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000489792722731065,
      "loss": 5.3071,
      "mean_token_accuracy": 0.1762430787086487,
      "num_tokens": 22806478.0,
      "step": 12360
    },
    {
      "entropy": 5.6345418930053714,
      "epoch": 1.038815374921235,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004897838033260712,
      "loss": 5.3206,
      "mean_token_accuracy": 0.1706371322274208,
      "num_tokens": 22815375.0,
      "step": 12365
    },
    {
      "entropy": 5.683670473098755,
      "epoch": 1.0392354547364,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004897748801162929,
      "loss": 5.2667,
      "mean_token_accuracy": 0.18158774971961975,
      "num_tokens": 22824401.0,
      "step": 12370
    },
    {
      "entropy": 5.725633573532105,
      "epoch": 1.0396555345515648,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004897659531018882,
      "loss": 5.3862,
      "mean_token_accuracy": 0.18009344339370728,
      "num_tokens": 22833933.0,
      "step": 12375
    },
    {
      "entropy": 5.644766807556152,
      "epoch": 1.0400756143667296,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004897570222830152,
      "loss": 5.2465,
      "mean_token_accuracy": 0.1782798796892166,
      "num_tokens": 22843779.0,
      "step": 12380
    },
    {
      "entropy": 5.772410154342651,
      "epoch": 1.0404956941818946,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004897480876598322,
      "loss": 5.3514,
      "mean_token_accuracy": 0.17785338312387466,
      "num_tokens": 22852951.0,
      "step": 12385
    },
    {
      "entropy": 5.742561531066895,
      "epoch": 1.0409157739970594,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004897391492324974,
      "loss": 5.3542,
      "mean_token_accuracy": 0.17231353521347045,
      "num_tokens": 22861398.0,
      "step": 12390
    },
    {
      "entropy": 5.687098503112793,
      "epoch": 1.0413358538122244,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004897302070011691,
      "loss": 5.2351,
      "mean_token_accuracy": 0.18651540130376815,
      "num_tokens": 22870518.0,
      "step": 12395
    },
    {
      "entropy": 5.631335163116455,
      "epoch": 1.0417559336273892,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004897212609660058,
      "loss": 5.3124,
      "mean_token_accuracy": 0.17701956480741501,
      "num_tokens": 22879389.0,
      "step": 12400
    },
    {
      "entropy": 5.651618003845215,
      "epoch": 1.0421760134425542,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004897123111271659,
      "loss": 5.3021,
      "mean_token_accuracy": 0.18335544168949128,
      "num_tokens": 22888977.0,
      "step": 12405
    },
    {
      "entropy": 5.7171238422393795,
      "epoch": 1.042596093257719,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004897033574848079,
      "loss": 5.3017,
      "mean_token_accuracy": 0.18249097913503648,
      "num_tokens": 22898446.0,
      "step": 12410
    },
    {
      "entropy": 5.682340431213379,
      "epoch": 1.0430161730728837,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004896944000390907,
      "loss": 5.3105,
      "mean_token_accuracy": 0.18068617582321167,
      "num_tokens": 22908044.0,
      "step": 12415
    },
    {
      "entropy": 5.740937519073486,
      "epoch": 1.0434362528880488,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004896854387901725,
      "loss": 5.3636,
      "mean_token_accuracy": 0.17493544071912764,
      "num_tokens": 22917330.0,
      "step": 12420
    },
    {
      "entropy": 5.750636625289917,
      "epoch": 1.0438563327032135,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004896764737382124,
      "loss": 5.3135,
      "mean_token_accuracy": 0.18605137169361113,
      "num_tokens": 22927160.0,
      "step": 12425
    },
    {
      "entropy": 5.748177099227905,
      "epoch": 1.0442764125183785,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004896675048833691,
      "loss": 5.2871,
      "mean_token_accuracy": 0.17707371860742568,
      "num_tokens": 22936755.0,
      "step": 12430
    },
    {
      "entropy": 5.695187520980835,
      "epoch": 1.0446964923335433,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004896585322258014,
      "loss": 5.2813,
      "mean_token_accuracy": 0.18082156777381897,
      "num_tokens": 22945699.0,
      "step": 12435
    },
    {
      "entropy": 5.657007694244385,
      "epoch": 1.0451165721487083,
      "grad_norm": 1.125,
      "learning_rate": 0.0004896495557656685,
      "loss": 5.2461,
      "mean_token_accuracy": 0.1876405581831932,
      "num_tokens": 22954001.0,
      "step": 12440
    },
    {
      "entropy": 5.751598787307739,
      "epoch": 1.0455366519638731,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004896405755031293,
      "loss": 5.3503,
      "mean_token_accuracy": 0.1742752104997635,
      "num_tokens": 22963805.0,
      "step": 12445
    },
    {
      "entropy": 5.659220790863037,
      "epoch": 1.045956731779038,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004896315914383427,
      "loss": 5.2768,
      "mean_token_accuracy": 0.16998640447854996,
      "num_tokens": 22973542.0,
      "step": 12450
    },
    {
      "entropy": 5.588300800323486,
      "epoch": 1.046376811594203,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004896226035714679,
      "loss": 5.1708,
      "mean_token_accuracy": 0.18437206745147705,
      "num_tokens": 22982417.0,
      "step": 12455
    },
    {
      "entropy": 5.655200386047364,
      "epoch": 1.0467968914093677,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004896136119026642,
      "loss": 5.2768,
      "mean_token_accuracy": 0.17781902849674225,
      "num_tokens": 22992879.0,
      "step": 12460
    },
    {
      "entropy": 5.639383172988891,
      "epoch": 1.0472169712245327,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004896046164320911,
      "loss": 5.1594,
      "mean_token_accuracy": 0.18606820255517958,
      "num_tokens": 23001344.0,
      "step": 12465
    },
    {
      "entropy": 5.604008531570434,
      "epoch": 1.0476370510396975,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004895956171599075,
      "loss": 5.2242,
      "mean_token_accuracy": 0.18745152354240419,
      "num_tokens": 23010007.0,
      "step": 12470
    },
    {
      "entropy": 5.6880372047424315,
      "epoch": 1.0480571308548625,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004895866140862731,
      "loss": 5.3404,
      "mean_token_accuracy": 0.17760687470436096,
      "num_tokens": 23019120.0,
      "step": 12475
    },
    {
      "entropy": 5.686089754104614,
      "epoch": 1.0484772106700273,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004895776072113473,
      "loss": 5.3182,
      "mean_token_accuracy": 0.18031392693519593,
      "num_tokens": 23028562.0,
      "step": 12480
    },
    {
      "entropy": 5.663020181655884,
      "epoch": 1.048897290485192,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004895685965352898,
      "loss": 5.2566,
      "mean_token_accuracy": 0.17938933223485948,
      "num_tokens": 23037687.0,
      "step": 12485
    },
    {
      "entropy": 5.6676054954528805,
      "epoch": 1.049317370300357,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004895595820582601,
      "loss": 5.2603,
      "mean_token_accuracy": 0.18234266489744186,
      "num_tokens": 23047475.0,
      "step": 12490
    },
    {
      "entropy": 5.648822069168091,
      "epoch": 1.0497374501155219,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004895505637804177,
      "loss": 5.2863,
      "mean_token_accuracy": 0.17406089454889298,
      "num_tokens": 23057475.0,
      "step": 12495
    },
    {
      "entropy": 5.593143701553345,
      "epoch": 1.050157529930687,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004895415417019227,
      "loss": 5.2527,
      "mean_token_accuracy": 0.17980754971504212,
      "num_tokens": 23066419.0,
      "step": 12500
    },
    {
      "entropy": 5.690524435043335,
      "epoch": 1.0505776097458517,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004895325158229346,
      "loss": 5.3296,
      "mean_token_accuracy": 0.17947423756122588,
      "num_tokens": 23075516.0,
      "step": 12505
    },
    {
      "entropy": 5.651531553268432,
      "epoch": 1.0509976895610167,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004895234861436136,
      "loss": 5.2045,
      "mean_token_accuracy": 0.18796317279338837,
      "num_tokens": 23084132.0,
      "step": 12510
    },
    {
      "entropy": 5.699585199356079,
      "epoch": 1.0514177693761815,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004895144526641194,
      "loss": 5.254,
      "mean_token_accuracy": 0.18298893719911574,
      "num_tokens": 23093958.0,
      "step": 12515
    },
    {
      "entropy": 5.7669291496276855,
      "epoch": 1.0518378491913463,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004895054153846123,
      "loss": 5.3084,
      "mean_token_accuracy": 0.1774764746427536,
      "num_tokens": 23103524.0,
      "step": 12520
    },
    {
      "entropy": 5.672692918777466,
      "epoch": 1.0522579290065113,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004894963743052521,
      "loss": 5.2507,
      "mean_token_accuracy": 0.17901405841112136,
      "num_tokens": 23112445.0,
      "step": 12525
    },
    {
      "entropy": 5.687171268463135,
      "epoch": 1.052678008821676,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004894873294261991,
      "loss": 5.3053,
      "mean_token_accuracy": 0.18307436406612396,
      "num_tokens": 23121299.0,
      "step": 12530
    },
    {
      "entropy": 5.736151885986328,
      "epoch": 1.053098088636841,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004894782807476134,
      "loss": 5.306,
      "mean_token_accuracy": 0.1737381860613823,
      "num_tokens": 23130260.0,
      "step": 12535
    },
    {
      "entropy": 5.696302127838135,
      "epoch": 1.0535181684520059,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004894692282696555,
      "loss": 5.2344,
      "mean_token_accuracy": 0.1846111685037613,
      "num_tokens": 23139335.0,
      "step": 12540
    },
    {
      "entropy": 5.624364614486694,
      "epoch": 1.0539382482671709,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004894601719924857,
      "loss": 5.2377,
      "mean_token_accuracy": 0.1824282705783844,
      "num_tokens": 23149299.0,
      "step": 12545
    },
    {
      "entropy": 5.625687932968139,
      "epoch": 1.0543583280823356,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004894511119162644,
      "loss": 5.2061,
      "mean_token_accuracy": 0.18733200579881668,
      "num_tokens": 23158651.0,
      "step": 12550
    },
    {
      "entropy": 5.713938856124878,
      "epoch": 1.0547784078975004,
      "grad_norm": 0.984375,
      "learning_rate": 0.000489442048041152,
      "loss": 5.2849,
      "mean_token_accuracy": 0.17450997680425645,
      "num_tokens": 23167629.0,
      "step": 12555
    },
    {
      "entropy": 5.619251251220703,
      "epoch": 1.0551984877126654,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004894329803673092,
      "loss": 5.2901,
      "mean_token_accuracy": 0.17884936928749084,
      "num_tokens": 23177026.0,
      "step": 12560
    },
    {
      "entropy": 5.637700033187866,
      "epoch": 1.0556185675278302,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004894239088948964,
      "loss": 5.2408,
      "mean_token_accuracy": 0.18238873332738875,
      "num_tokens": 23185297.0,
      "step": 12565
    },
    {
      "entropy": 5.639225006103516,
      "epoch": 1.0560386473429952,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004894148336240747,
      "loss": 5.2433,
      "mean_token_accuracy": 0.18484221398830414,
      "num_tokens": 23194804.0,
      "step": 12570
    },
    {
      "entropy": 5.683013772964477,
      "epoch": 1.05645872715816,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004894057545550045,
      "loss": 5.2924,
      "mean_token_accuracy": 0.18169104009866716,
      "num_tokens": 23205063.0,
      "step": 12575
    },
    {
      "entropy": 5.629480934143066,
      "epoch": 1.056878806973325,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004893966716878467,
      "loss": 5.2323,
      "mean_token_accuracy": 0.1764781579375267,
      "num_tokens": 23215038.0,
      "step": 12580
    },
    {
      "entropy": 5.721001672744751,
      "epoch": 1.0572988867884898,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004893875850227624,
      "loss": 5.3772,
      "mean_token_accuracy": 0.17522427290678025,
      "num_tokens": 23223530.0,
      "step": 12585
    },
    {
      "entropy": 5.699247169494629,
      "epoch": 1.0577189666036546,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004893784945599124,
      "loss": 5.3245,
      "mean_token_accuracy": 0.18004236370325089,
      "num_tokens": 23232547.0,
      "step": 12590
    },
    {
      "entropy": 5.681890249252319,
      "epoch": 1.0581390464188196,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004893694002994577,
      "loss": 5.3502,
      "mean_token_accuracy": 0.1771372765302658,
      "num_tokens": 23241305.0,
      "step": 12595
    },
    {
      "entropy": 5.823739385604858,
      "epoch": 1.0585591262339844,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004893603022415595,
      "loss": 5.3857,
      "mean_token_accuracy": 0.17562371790409087,
      "num_tokens": 23250708.0,
      "step": 12600
    },
    {
      "entropy": 5.742749118804932,
      "epoch": 1.0589792060491494,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004893512003863788,
      "loss": 5.2809,
      "mean_token_accuracy": 0.17282908111810685,
      "num_tokens": 23260161.0,
      "step": 12605
    },
    {
      "entropy": 5.602674341201782,
      "epoch": 1.0593992858643142,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004893420947340771,
      "loss": 5.1944,
      "mean_token_accuracy": 0.18067559748888015,
      "num_tokens": 23268932.0,
      "step": 12610
    },
    {
      "entropy": 5.661489677429199,
      "epoch": 1.0598193656794792,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004893329852848155,
      "loss": 5.2948,
      "mean_token_accuracy": 0.17977765798568726,
      "num_tokens": 23277741.0,
      "step": 12615
    },
    {
      "entropy": 5.6706501007080075,
      "epoch": 1.060239445494644,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004893238720387555,
      "loss": 5.2974,
      "mean_token_accuracy": 0.17933399379253387,
      "num_tokens": 23286982.0,
      "step": 12620
    },
    {
      "entropy": 5.612814140319824,
      "epoch": 1.0606595253098088,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004893147549960584,
      "loss": 5.2274,
      "mean_token_accuracy": 0.18449618071317672,
      "num_tokens": 23296902.0,
      "step": 12625
    },
    {
      "entropy": 5.619822931289673,
      "epoch": 1.0610796051249738,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004893056341568857,
      "loss": 5.2219,
      "mean_token_accuracy": 0.1864966332912445,
      "num_tokens": 23305443.0,
      "step": 12630
    },
    {
      "entropy": 5.611188983917236,
      "epoch": 1.0614996849401386,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004892965095213992,
      "loss": 5.1918,
      "mean_token_accuracy": 0.1887941837310791,
      "num_tokens": 23315420.0,
      "step": 12635
    },
    {
      "entropy": 5.709328317642212,
      "epoch": 1.0619197647553036,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004892873810897604,
      "loss": 5.29,
      "mean_token_accuracy": 0.176090507209301,
      "num_tokens": 23324540.0,
      "step": 12640
    },
    {
      "entropy": 5.69493522644043,
      "epoch": 1.0623398445704684,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004892782488621308,
      "loss": 5.2348,
      "mean_token_accuracy": 0.18402790427207946,
      "num_tokens": 23334282.0,
      "step": 12645
    },
    {
      "entropy": 5.662745952606201,
      "epoch": 1.0627599243856332,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004892691128386725,
      "loss": 5.2276,
      "mean_token_accuracy": 0.18393418043851853,
      "num_tokens": 23342836.0,
      "step": 12650
    },
    {
      "entropy": 5.671319150924683,
      "epoch": 1.0631800042007982,
      "grad_norm": 1.0,
      "learning_rate": 0.0004892599730195471,
      "loss": 5.2269,
      "mean_token_accuracy": 0.1864285260438919,
      "num_tokens": 23351863.0,
      "step": 12655
    },
    {
      "entropy": 5.741396522521972,
      "epoch": 1.063600084015963,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004892508294049167,
      "loss": 5.3542,
      "mean_token_accuracy": 0.18320610225200654,
      "num_tokens": 23361788.0,
      "step": 12660
    },
    {
      "entropy": 5.6810461521148685,
      "epoch": 1.064020163831128,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004892416819949431,
      "loss": 5.2175,
      "mean_token_accuracy": 0.1792144164443016,
      "num_tokens": 23370175.0,
      "step": 12665
    },
    {
      "entropy": 5.618555974960327,
      "epoch": 1.0644402436462927,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004892325307897886,
      "loss": 5.253,
      "mean_token_accuracy": 0.17969499081373214,
      "num_tokens": 23378835.0,
      "step": 12670
    },
    {
      "entropy": 5.6547294616699215,
      "epoch": 1.0648603234614578,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004892233757896149,
      "loss": 5.251,
      "mean_token_accuracy": 0.17920616120100022,
      "num_tokens": 23389390.0,
      "step": 12675
    },
    {
      "entropy": 5.71104793548584,
      "epoch": 1.0652804032766225,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004892142169945845,
      "loss": 5.2793,
      "mean_token_accuracy": 0.17964175790548326,
      "num_tokens": 23398802.0,
      "step": 12680
    },
    {
      "entropy": 5.648165655136109,
      "epoch": 1.0657004830917876,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004892050544048596,
      "loss": 5.2497,
      "mean_token_accuracy": 0.17950946241617202,
      "num_tokens": 23407731.0,
      "step": 12685
    },
    {
      "entropy": 5.6626380443572994,
      "epoch": 1.0661205629069523,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004891958880206024,
      "loss": 5.2774,
      "mean_token_accuracy": 0.17747126519680023,
      "num_tokens": 23417046.0,
      "step": 12690
    },
    {
      "entropy": 5.690803146362304,
      "epoch": 1.0665406427221171,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004891867178419753,
      "loss": 5.2651,
      "mean_token_accuracy": 0.18215234577655792,
      "num_tokens": 23426107.0,
      "step": 12695
    },
    {
      "entropy": 5.699212074279785,
      "epoch": 1.0669607225372821,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004891775438691408,
      "loss": 5.3394,
      "mean_token_accuracy": 0.18157806098461152,
      "num_tokens": 23435523.0,
      "step": 12700
    },
    {
      "entropy": 5.694417333602905,
      "epoch": 1.067380802352447,
      "grad_norm": 1.25,
      "learning_rate": 0.0004891683661022615,
      "loss": 5.2634,
      "mean_token_accuracy": 0.18492066264152526,
      "num_tokens": 23444185.0,
      "step": 12705
    },
    {
      "entropy": 5.765972566604614,
      "epoch": 1.067800882167612,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004891591845414997,
      "loss": 5.4318,
      "mean_token_accuracy": 0.1651720330119133,
      "num_tokens": 23454100.0,
      "step": 12710
    },
    {
      "entropy": 5.7497951030731205,
      "epoch": 1.0682209619827767,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004891499991870184,
      "loss": 5.328,
      "mean_token_accuracy": 0.17413219511508943,
      "num_tokens": 23463415.0,
      "step": 12715
    },
    {
      "entropy": 5.682176828384399,
      "epoch": 1.0686410417979415,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048914081003898,
      "loss": 5.2518,
      "mean_token_accuracy": 0.17763317376375198,
      "num_tokens": 23471515.0,
      "step": 12720
    },
    {
      "entropy": 5.715403127670288,
      "epoch": 1.0690611216131065,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004891316170975475,
      "loss": 5.3001,
      "mean_token_accuracy": 0.1811108559370041,
      "num_tokens": 23481696.0,
      "step": 12725
    },
    {
      "entropy": 5.730748319625855,
      "epoch": 1.0694812014282713,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004891224203628836,
      "loss": 5.2737,
      "mean_token_accuracy": 0.18413740545511245,
      "num_tokens": 23490714.0,
      "step": 12730
    },
    {
      "entropy": 5.5975260734558105,
      "epoch": 1.0699012812434363,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004891132198351514,
      "loss": 5.2228,
      "mean_token_accuracy": 0.18698084354400635,
      "num_tokens": 23500368.0,
      "step": 12735
    },
    {
      "entropy": 5.556104278564453,
      "epoch": 1.070321361058601,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004891040155145137,
      "loss": 5.1886,
      "mean_token_accuracy": 0.18293554037809373,
      "num_tokens": 23508857.0,
      "step": 12740
    },
    {
      "entropy": 5.612934827804565,
      "epoch": 1.070741440873766,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004890948074011335,
      "loss": 5.1752,
      "mean_token_accuracy": 0.18947744816541673,
      "num_tokens": 23518128.0,
      "step": 12745
    },
    {
      "entropy": 5.718452072143554,
      "epoch": 1.071161520688931,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004890855954951741,
      "loss": 5.2784,
      "mean_token_accuracy": 0.1832232728600502,
      "num_tokens": 23527292.0,
      "step": 12750
    },
    {
      "entropy": 5.722832536697387,
      "epoch": 1.0715816005040957,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004890763797967987,
      "loss": 5.2736,
      "mean_token_accuracy": 0.17744670808315277,
      "num_tokens": 23535694.0,
      "step": 12755
    },
    {
      "entropy": 5.645816326141357,
      "epoch": 1.0720016803192607,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004890671603061704,
      "loss": 5.2573,
      "mean_token_accuracy": 0.18388551473617554,
      "num_tokens": 23544766.0,
      "step": 12760
    },
    {
      "entropy": 5.623814296722412,
      "epoch": 1.0724217601344255,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004890579370234526,
      "loss": 5.2368,
      "mean_token_accuracy": 0.18621060848236085,
      "num_tokens": 23554037.0,
      "step": 12765
    },
    {
      "entropy": 5.726473999023438,
      "epoch": 1.0728418399495905,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004890487099488086,
      "loss": 5.3198,
      "mean_token_accuracy": 0.17537978589534758,
      "num_tokens": 23562282.0,
      "step": 12770
    },
    {
      "entropy": 5.770499324798584,
      "epoch": 1.0732619197647553,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000489039479082402,
      "loss": 5.3562,
      "mean_token_accuracy": 0.17187156826257705,
      "num_tokens": 23571955.0,
      "step": 12775
    },
    {
      "entropy": 5.675796270370483,
      "epoch": 1.0736819995799203,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004890302444243962,
      "loss": 5.2564,
      "mean_token_accuracy": 0.17287126332521438,
      "num_tokens": 23580996.0,
      "step": 12780
    },
    {
      "entropy": 5.684293460845947,
      "epoch": 1.074102079395085,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004890210059749549,
      "loss": 5.3542,
      "mean_token_accuracy": 0.1722228541970253,
      "num_tokens": 23589618.0,
      "step": 12785
    },
    {
      "entropy": 5.656562805175781,
      "epoch": 1.0745221592102498,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004890117637342416,
      "loss": 5.1881,
      "mean_token_accuracy": 0.17950229346752167,
      "num_tokens": 23599574.0,
      "step": 12790
    },
    {
      "entropy": 5.654933118820191,
      "epoch": 1.0749422390254149,
      "grad_norm": 1.0,
      "learning_rate": 0.0004890025177024202,
      "loss": 5.2486,
      "mean_token_accuracy": 0.18002490401268006,
      "num_tokens": 23609205.0,
      "step": 12795
    },
    {
      "entropy": 5.631871080398559,
      "epoch": 1.0753623188405796,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004889932678796543,
      "loss": 5.2813,
      "mean_token_accuracy": 0.16844987273216247,
      "num_tokens": 23617554.0,
      "step": 12800
    },
    {
      "entropy": 5.6981847286224365,
      "epoch": 1.0757823986557447,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004889840142661078,
      "loss": 5.3514,
      "mean_token_accuracy": 0.17695426493883132,
      "num_tokens": 23626757.0,
      "step": 12805
    },
    {
      "entropy": 5.736290121078492,
      "epoch": 1.0762024784709094,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004889747568619447,
      "loss": 5.2972,
      "mean_token_accuracy": 0.1785106763243675,
      "num_tokens": 23636111.0,
      "step": 12810
    },
    {
      "entropy": 5.685405969619751,
      "epoch": 1.0766225582860744,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004889654956673291,
      "loss": 5.2806,
      "mean_token_accuracy": 0.18219675868749619,
      "num_tokens": 23644579.0,
      "step": 12815
    },
    {
      "entropy": 5.677073287963867,
      "epoch": 1.0770426381012392,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004889562306824248,
      "loss": 5.2028,
      "mean_token_accuracy": 0.18450048565864563,
      "num_tokens": 23653263.0,
      "step": 12820
    },
    {
      "entropy": 5.556289720535278,
      "epoch": 1.077462717916404,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000488946961907396,
      "loss": 5.1521,
      "mean_token_accuracy": 0.1935470461845398,
      "num_tokens": 23662529.0,
      "step": 12825
    },
    {
      "entropy": 5.583633852005005,
      "epoch": 1.077882797731569,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004889376893424071,
      "loss": 5.1823,
      "mean_token_accuracy": 0.18970176428556443,
      "num_tokens": 23671491.0,
      "step": 12830
    },
    {
      "entropy": 5.640367937088013,
      "epoch": 1.0783028775467338,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004889284129876221,
      "loss": 5.1889,
      "mean_token_accuracy": 0.1866241529583931,
      "num_tokens": 23680121.0,
      "step": 12835
    },
    {
      "entropy": 5.629704380035401,
      "epoch": 1.0787229573618988,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004889191328432054,
      "loss": 5.2547,
      "mean_token_accuracy": 0.17927492409944534,
      "num_tokens": 23689008.0,
      "step": 12840
    },
    {
      "entropy": 5.669335079193115,
      "epoch": 1.0791430371770636,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004889098489093215,
      "loss": 5.2684,
      "mean_token_accuracy": 0.17769426107406616,
      "num_tokens": 23698551.0,
      "step": 12845
    },
    {
      "entropy": 5.756178951263427,
      "epoch": 1.0795631169922286,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004889005611861347,
      "loss": 5.4286,
      "mean_token_accuracy": 0.1771997630596161,
      "num_tokens": 23707438.0,
      "step": 12850
    },
    {
      "entropy": 5.667813444137574,
      "epoch": 1.0799831968073934,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004888912696738096,
      "loss": 5.2823,
      "mean_token_accuracy": 0.18185589611530303,
      "num_tokens": 23715822.0,
      "step": 12855
    },
    {
      "entropy": 5.704867792129517,
      "epoch": 1.0804032766225582,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004888819743725108,
      "loss": 5.3024,
      "mean_token_accuracy": 0.18461446315050126,
      "num_tokens": 23725426.0,
      "step": 12860
    },
    {
      "entropy": 5.696760892868042,
      "epoch": 1.0808233564377232,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000488872675282403,
      "loss": 5.2826,
      "mean_token_accuracy": 0.18007925599813462,
      "num_tokens": 23735092.0,
      "step": 12865
    },
    {
      "entropy": 5.662566995620727,
      "epoch": 1.081243436252888,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004888633724036509,
      "loss": 5.2828,
      "mean_token_accuracy": 0.17780134975910186,
      "num_tokens": 23744255.0,
      "step": 12870
    },
    {
      "entropy": 5.642053270339966,
      "epoch": 1.081663516068053,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004888540657364192,
      "loss": 5.1421,
      "mean_token_accuracy": 0.18846668004989625,
      "num_tokens": 23752978.0,
      "step": 12875
    },
    {
      "entropy": 5.683878660202026,
      "epoch": 1.0820835958832178,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004888447552808729,
      "loss": 5.2175,
      "mean_token_accuracy": 0.18439697921276094,
      "num_tokens": 23761051.0,
      "step": 12880
    },
    {
      "entropy": 5.722139167785644,
      "epoch": 1.0825036756983828,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004888354410371768,
      "loss": 5.341,
      "mean_token_accuracy": 0.17446444779634476,
      "num_tokens": 23770818.0,
      "step": 12885
    },
    {
      "entropy": 5.776684045791626,
      "epoch": 1.0829237555135476,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000488826123005496,
      "loss": 5.327,
      "mean_token_accuracy": 0.1792567104101181,
      "num_tokens": 23780597.0,
      "step": 12890
    },
    {
      "entropy": 5.637493038177491,
      "epoch": 1.0833438353287124,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004888168011859957,
      "loss": 5.1968,
      "mean_token_accuracy": 0.18231092095375062,
      "num_tokens": 23790119.0,
      "step": 12895
    },
    {
      "entropy": 5.652245378494262,
      "epoch": 1.0837639151438774,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004888074755788407,
      "loss": 5.2525,
      "mean_token_accuracy": 0.18587874472141266,
      "num_tokens": 23798972.0,
      "step": 12900
    },
    {
      "entropy": 5.644486951828003,
      "epoch": 1.0841839949590422,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004887981461841963,
      "loss": 5.2305,
      "mean_token_accuracy": 0.1889592945575714,
      "num_tokens": 23808685.0,
      "step": 12905
    },
    {
      "entropy": 5.711899375915527,
      "epoch": 1.0846040747742072,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004887888130022279,
      "loss": 5.2357,
      "mean_token_accuracy": 0.17830779552459716,
      "num_tokens": 23817721.0,
      "step": 12910
    },
    {
      "entropy": 5.580431652069092,
      "epoch": 1.085024154589372,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004887794760331008,
      "loss": 5.2072,
      "mean_token_accuracy": 0.18942340910434724,
      "num_tokens": 23826892.0,
      "step": 12915
    },
    {
      "entropy": 5.658565664291382,
      "epoch": 1.085444234404537,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004887701352769804,
      "loss": 5.1555,
      "mean_token_accuracy": 0.18969567865133286,
      "num_tokens": 23835717.0,
      "step": 12920
    },
    {
      "entropy": 5.683448934555054,
      "epoch": 1.0858643142197018,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000488760790734032,
      "loss": 5.2371,
      "mean_token_accuracy": 0.18935709744691848,
      "num_tokens": 23845814.0,
      "step": 12925
    },
    {
      "entropy": 5.706423282623291,
      "epoch": 1.0862843940348665,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004887514424044214,
      "loss": 5.2308,
      "mean_token_accuracy": 0.17566871494054795,
      "num_tokens": 23854779.0,
      "step": 12930
    },
    {
      "entropy": 5.63693151473999,
      "epoch": 1.0867044738500315,
      "grad_norm": 1.09375,
      "learning_rate": 0.000488742090288314,
      "loss": 5.2895,
      "mean_token_accuracy": 0.18086664080619813,
      "num_tokens": 23863533.0,
      "step": 12935
    },
    {
      "entropy": 5.67240047454834,
      "epoch": 1.0871245536651963,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004887327343858755,
      "loss": 5.3083,
      "mean_token_accuracy": 0.1804274633526802,
      "num_tokens": 23872725.0,
      "step": 12940
    },
    {
      "entropy": 5.683504199981689,
      "epoch": 1.0875446334803613,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004887233746972717,
      "loss": 5.3077,
      "mean_token_accuracy": 0.17329630702733995,
      "num_tokens": 23881799.0,
      "step": 12945
    },
    {
      "entropy": 5.658789873123169,
      "epoch": 1.0879647132955261,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004887140112226684,
      "loss": 5.3132,
      "mean_token_accuracy": 0.1773510679602623,
      "num_tokens": 23890628.0,
      "step": 12950
    },
    {
      "entropy": 5.639373111724853,
      "epoch": 1.088384793110691,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004887046439622314,
      "loss": 5.2924,
      "mean_token_accuracy": 0.1874363124370575,
      "num_tokens": 23899968.0,
      "step": 12955
    },
    {
      "entropy": 5.6947096824646,
      "epoch": 1.088804872925856,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004886952729161267,
      "loss": 5.1782,
      "mean_token_accuracy": 0.19132802188396453,
      "num_tokens": 23908634.0,
      "step": 12960
    },
    {
      "entropy": 5.703873920440674,
      "epoch": 1.0892249527410207,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004886858980845202,
      "loss": 5.3077,
      "mean_token_accuracy": 0.18086057305335998,
      "num_tokens": 23917925.0,
      "step": 12965
    },
    {
      "entropy": 5.6058183193206785,
      "epoch": 1.0896450325561857,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004886765194675782,
      "loss": 5.2239,
      "mean_token_accuracy": 0.1841082513332367,
      "num_tokens": 23927173.0,
      "step": 12970
    },
    {
      "entropy": 5.6475348472595215,
      "epoch": 1.0900651123713505,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004886671370654665,
      "loss": 5.1693,
      "mean_token_accuracy": 0.18463612645864486,
      "num_tokens": 23936258.0,
      "step": 12975
    },
    {
      "entropy": 5.659107637405396,
      "epoch": 1.0904851921865155,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004886577508783516,
      "loss": 5.1573,
      "mean_token_accuracy": 0.19075287878513336,
      "num_tokens": 23944215.0,
      "step": 12980
    },
    {
      "entropy": 5.691824865341187,
      "epoch": 1.0909052720016803,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004886483609063997,
      "loss": 5.2362,
      "mean_token_accuracy": 0.18063196539878845,
      "num_tokens": 23953151.0,
      "step": 12985
    },
    {
      "entropy": 5.580561590194702,
      "epoch": 1.0913253518168453,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004886389671497769,
      "loss": 5.2266,
      "mean_token_accuracy": 0.19029317498207093,
      "num_tokens": 23962919.0,
      "step": 12990
    },
    {
      "entropy": 5.691284465789795,
      "epoch": 1.09174543163201,
      "grad_norm": 1.0,
      "learning_rate": 0.00048862956960865,
      "loss": 5.2651,
      "mean_token_accuracy": 0.18001859486103058,
      "num_tokens": 23971900.0,
      "step": 12995
    },
    {
      "entropy": 5.679598665237426,
      "epoch": 1.0921655114471749,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004886201682831852,
      "loss": 5.2142,
      "mean_token_accuracy": 0.18552780002355576,
      "num_tokens": 23980945.0,
      "step": 13000
    },
    {
      "entropy": 5.615207624435425,
      "epoch": 1.09258559126234,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004886107631735491,
      "loss": 5.1905,
      "mean_token_accuracy": 0.18361067324876784,
      "num_tokens": 23990460.0,
      "step": 13005
    },
    {
      "entropy": 5.721375131607056,
      "epoch": 1.0930056710775047,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004886013542799083,
      "loss": 5.3516,
      "mean_token_accuracy": 0.1659088134765625,
      "num_tokens": 23999925.0,
      "step": 13010
    },
    {
      "entropy": 5.649724245071411,
      "epoch": 1.0934257508926697,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004885919416024296,
      "loss": 5.2089,
      "mean_token_accuracy": 0.18353860527276994,
      "num_tokens": 24009039.0,
      "step": 13015
    },
    {
      "entropy": 5.7273773670196535,
      "epoch": 1.0938458307078345,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004885825251412796,
      "loss": 5.2565,
      "mean_token_accuracy": 0.18467819541692734,
      "num_tokens": 24017725.0,
      "step": 13020
    },
    {
      "entropy": 5.701650142669678,
      "epoch": 1.0942659105229993,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004885731048966252,
      "loss": 5.2517,
      "mean_token_accuracy": 0.1813005894422531,
      "num_tokens": 24027158.0,
      "step": 13025
    },
    {
      "entropy": 5.647268342971802,
      "epoch": 1.0946859903381643,
      "grad_norm": 1.125,
      "learning_rate": 0.0004885636808686331,
      "loss": 5.3101,
      "mean_token_accuracy": 0.18107652068138122,
      "num_tokens": 24037224.0,
      "step": 13030
    },
    {
      "entropy": 5.655750751495361,
      "epoch": 1.095106070153329,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004885542530574705,
      "loss": 5.2889,
      "mean_token_accuracy": 0.1789359360933304,
      "num_tokens": 24046097.0,
      "step": 13035
    },
    {
      "entropy": 5.62082986831665,
      "epoch": 1.095526149968494,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004885448214633042,
      "loss": 5.1641,
      "mean_token_accuracy": 0.18469917625188828,
      "num_tokens": 24055270.0,
      "step": 13040
    },
    {
      "entropy": 5.734753561019898,
      "epoch": 1.0959462297836589,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004885353860863013,
      "loss": 5.3452,
      "mean_token_accuracy": 0.17018480151891707,
      "num_tokens": 24064995.0,
      "step": 13045
    },
    {
      "entropy": 5.791645336151123,
      "epoch": 1.0963663095988239,
      "grad_norm": 1.09375,
      "learning_rate": 0.000488525946926629,
      "loss": 5.3757,
      "mean_token_accuracy": 0.17358792424201966,
      "num_tokens": 24075523.0,
      "step": 13050
    },
    {
      "entropy": 5.696471214294434,
      "epoch": 1.0967863894139886,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004885165039844545,
      "loss": 5.2579,
      "mean_token_accuracy": 0.18176131397485734,
      "num_tokens": 24084933.0,
      "step": 13055
    },
    {
      "entropy": 5.696083831787109,
      "epoch": 1.0972064692291534,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004885070572599452,
      "loss": 5.2802,
      "mean_token_accuracy": 0.17447678446769715,
      "num_tokens": 24093964.0,
      "step": 13060
    },
    {
      "entropy": 5.664876413345337,
      "epoch": 1.0976265490443184,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004884976067532681,
      "loss": 5.2343,
      "mean_token_accuracy": 0.16958827525377274,
      "num_tokens": 24103951.0,
      "step": 13065
    },
    {
      "entropy": 5.646041393280029,
      "epoch": 1.0980466288594832,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000488488152464591,
      "loss": 5.3402,
      "mean_token_accuracy": 0.17385809421539306,
      "num_tokens": 24113392.0,
      "step": 13070
    },
    {
      "entropy": 5.633843469619751,
      "epoch": 1.0984667086746482,
      "grad_norm": 1.0,
      "learning_rate": 0.0004884786943940812,
      "loss": 5.2016,
      "mean_token_accuracy": 0.18126096129417418,
      "num_tokens": 24123165.0,
      "step": 13075
    },
    {
      "entropy": 5.633621644973755,
      "epoch": 1.098886788489813,
      "grad_norm": 1.0,
      "learning_rate": 0.0004884692325419063,
      "loss": 5.2558,
      "mean_token_accuracy": 0.18120505809783935,
      "num_tokens": 24132176.0,
      "step": 13080
    },
    {
      "entropy": 5.683441257476806,
      "epoch": 1.099306868304978,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004884597669082336,
      "loss": 5.3084,
      "mean_token_accuracy": 0.17819964289665222,
      "num_tokens": 24141737.0,
      "step": 13085
    },
    {
      "entropy": 5.690486097335816,
      "epoch": 1.0997269481201428,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004884502974932313,
      "loss": 5.2463,
      "mean_token_accuracy": 0.18067810237407683,
      "num_tokens": 24150477.0,
      "step": 13090
    },
    {
      "entropy": 5.779782342910766,
      "epoch": 1.1001470279353076,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004884408242970668,
      "loss": 5.3347,
      "mean_token_accuracy": 0.17849907279014587,
      "num_tokens": 24158739.0,
      "step": 13095
    },
    {
      "entropy": 5.630479192733764,
      "epoch": 1.1005671077504726,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004884313473199081,
      "loss": 5.1897,
      "mean_token_accuracy": 0.18633594810962678,
      "num_tokens": 24167511.0,
      "step": 13100
    },
    {
      "entropy": 5.599885320663452,
      "epoch": 1.1009871875656374,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004884218665619229,
      "loss": 5.1882,
      "mean_token_accuracy": 0.18263051509857178,
      "num_tokens": 24176413.0,
      "step": 13105
    },
    {
      "entropy": 5.6354567527771,
      "epoch": 1.1014072673808024,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004884123820232792,
      "loss": 5.1429,
      "mean_token_accuracy": 0.19051122665405273,
      "num_tokens": 24185135.0,
      "step": 13110
    },
    {
      "entropy": 5.665712022781372,
      "epoch": 1.1018273471959672,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004884028937041451,
      "loss": 5.2255,
      "mean_token_accuracy": 0.18888204544782639,
      "num_tokens": 24193273.0,
      "step": 13115
    },
    {
      "entropy": 5.711168479919434,
      "epoch": 1.1022474270111322,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004883934016046886,
      "loss": 5.2832,
      "mean_token_accuracy": 0.17425669282674788,
      "num_tokens": 24202509.0,
      "step": 13120
    },
    {
      "entropy": 5.701364803314209,
      "epoch": 1.102667506826297,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000488383905725078,
      "loss": 5.2883,
      "mean_token_accuracy": 0.17556007504463195,
      "num_tokens": 24212644.0,
      "step": 13125
    },
    {
      "entropy": 5.644434022903442,
      "epoch": 1.1030875866414618,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004883744060654811,
      "loss": 5.1741,
      "mean_token_accuracy": 0.18276613354682922,
      "num_tokens": 24221838.0,
      "step": 13130
    },
    {
      "entropy": 5.612702989578247,
      "epoch": 1.1035076664566268,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004883649026260667,
      "loss": 5.2595,
      "mean_token_accuracy": 0.18412424474954606,
      "num_tokens": 24230987.0,
      "step": 13135
    },
    {
      "entropy": 5.668402528762817,
      "epoch": 1.1039277462717916,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004883553954070028,
      "loss": 5.2193,
      "mean_token_accuracy": 0.18840075433254241,
      "num_tokens": 24240523.0,
      "step": 13140
    },
    {
      "entropy": 5.706751918792724,
      "epoch": 1.1043478260869566,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000488345884408458,
      "loss": 5.3062,
      "mean_token_accuracy": 0.18301960825920105,
      "num_tokens": 24249799.0,
      "step": 13145
    },
    {
      "entropy": 5.640453672409057,
      "epoch": 1.1047679059021214,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004883363696306007,
      "loss": 5.2583,
      "mean_token_accuracy": 0.18457481265068054,
      "num_tokens": 24259361.0,
      "step": 13150
    },
    {
      "entropy": 5.671831560134888,
      "epoch": 1.1051879857172864,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004883268510735995,
      "loss": 5.2215,
      "mean_token_accuracy": 0.17966383546590806,
      "num_tokens": 24268010.0,
      "step": 13155
    },
    {
      "entropy": 5.585932111740112,
      "epoch": 1.1056080655324512,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004883173287376229,
      "loss": 5.2742,
      "mean_token_accuracy": 0.18010523021221161,
      "num_tokens": 24277416.0,
      "step": 13160
    },
    {
      "entropy": 5.731878232955933,
      "epoch": 1.106028145347616,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004883078026228397,
      "loss": 5.3388,
      "mean_token_accuracy": 0.17820267677307128,
      "num_tokens": 24286185.0,
      "step": 13165
    },
    {
      "entropy": 5.691918611526489,
      "epoch": 1.106448225162781,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004882982727294187,
      "loss": 5.1976,
      "mean_token_accuracy": 0.17962379902601242,
      "num_tokens": 24295382.0,
      "step": 13170
    },
    {
      "entropy": 5.630468797683716,
      "epoch": 1.1068683049779457,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004882887390575284,
      "loss": 5.2223,
      "mean_token_accuracy": 0.1822338104248047,
      "num_tokens": 24305197.0,
      "step": 13175
    },
    {
      "entropy": 5.706013345718384,
      "epoch": 1.1072883847931108,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004882792016073381,
      "loss": 5.3088,
      "mean_token_accuracy": 0.17204824686050416,
      "num_tokens": 24314149.0,
      "step": 13180
    },
    {
      "entropy": 5.675956964492798,
      "epoch": 1.1077084646082755,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048826966037901655,
      "loss": 5.2266,
      "mean_token_accuracy": 0.1876238390803337,
      "num_tokens": 24323737.0,
      "step": 13185
    },
    {
      "entropy": 5.645000410079956,
      "epoch": 1.1081285444234406,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048826011537273276,
      "loss": 5.2121,
      "mean_token_accuracy": 0.1822698086500168,
      "num_tokens": 24332853.0,
      "step": 13190
    },
    {
      "entropy": 5.670597171783447,
      "epoch": 1.1085486242386053,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004882505665886558,
      "loss": 5.3415,
      "mean_token_accuracy": 0.1786191612482071,
      "num_tokens": 24342632.0,
      "step": 13195
    },
    {
      "entropy": 5.63273253440857,
      "epoch": 1.1089687040537701,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048824101402695493,
      "loss": 5.1951,
      "mean_token_accuracy": 0.18174874931573867,
      "num_tokens": 24351659.0,
      "step": 13200
    },
    {
      "entropy": 5.625222539901733,
      "epoch": 1.1093887838689351,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004882314576877993,
      "loss": 5.2126,
      "mean_token_accuracy": 0.17604063898324968,
      "num_tokens": 24360938.0,
      "step": 13205
    },
    {
      "entropy": 5.627703523635864,
      "epoch": 1.1098088636841,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004882218975713581,
      "loss": 5.3039,
      "mean_token_accuracy": 0.18238244354724883,
      "num_tokens": 24369603.0,
      "step": 13210
    },
    {
      "entropy": 5.593412446975708,
      "epoch": 1.110228943499265,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004882123336778009,
      "loss": 5.2412,
      "mean_token_accuracy": 0.18330066204071044,
      "num_tokens": 24377605.0,
      "step": 13215
    },
    {
      "entropy": 5.672424554824829,
      "epoch": 1.1106490233144297,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004882027660072969,
      "loss": 5.2832,
      "mean_token_accuracy": 0.1787025362253189,
      "num_tokens": 24386930.0,
      "step": 13220
    },
    {
      "entropy": 5.662126779556274,
      "epoch": 1.1110691031295947,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004881931945600157,
      "loss": 5.2267,
      "mean_token_accuracy": 0.19292469918727875,
      "num_tokens": 24396473.0,
      "step": 13225
    },
    {
      "entropy": 5.698059463500977,
      "epoch": 1.1114891829447595,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004881836193361269,
      "loss": 5.3193,
      "mean_token_accuracy": 0.1764673709869385,
      "num_tokens": 24405461.0,
      "step": 13230
    },
    {
      "entropy": 5.691852617263794,
      "epoch": 1.1119092627599243,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004881740403358,
      "loss": 5.2576,
      "mean_token_accuracy": 0.1866246536374092,
      "num_tokens": 24414138.0,
      "step": 13235
    },
    {
      "entropy": 5.691903877258301,
      "epoch": 1.1123293425750893,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048816445755920474,
      "loss": 5.2726,
      "mean_token_accuracy": 0.17981644123792648,
      "num_tokens": 24423386.0,
      "step": 13240
    },
    {
      "entropy": 5.692438411712646,
      "epoch": 1.112749422390254,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004881548710065109,
      "loss": 5.2582,
      "mean_token_accuracy": 0.18377391248941422,
      "num_tokens": 24433637.0,
      "step": 13245
    },
    {
      "entropy": 5.670820093154907,
      "epoch": 1.113169502205419,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004881452806778883,
      "loss": 5.3144,
      "mean_token_accuracy": 0.18136941492557526,
      "num_tokens": 24443677.0,
      "step": 13250
    },
    {
      "entropy": 5.6963848114013675,
      "epoch": 1.113589582020584,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048813568657350676,
      "loss": 5.2055,
      "mean_token_accuracy": 0.1869214192032814,
      "num_tokens": 24452317.0,
      "step": 13255
    },
    {
      "entropy": 5.688091087341308,
      "epoch": 1.1140096618357487,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004881260886935363,
      "loss": 5.239,
      "mean_token_accuracy": 0.18475331664085387,
      "num_tokens": 24460626.0,
      "step": 13260
    },
    {
      "entropy": 5.7285982131958,
      "epoch": 1.1144297416509137,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048811648703814693,
      "loss": 5.3207,
      "mean_token_accuracy": 0.1685321196913719,
      "num_tokens": 24469583.0,
      "step": 13265
    },
    {
      "entropy": 5.642823314666748,
      "epoch": 1.1148498214660785,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004881068816075087,
      "loss": 5.2451,
      "mean_token_accuracy": 0.17854614108800887,
      "num_tokens": 24478811.0,
      "step": 13270
    },
    {
      "entropy": 5.658578634262085,
      "epoch": 1.1152699012812435,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048809727240179193,
      "loss": 5.287,
      "mean_token_accuracy": 0.18240704089403154,
      "num_tokens": 24487818.0,
      "step": 13275
    },
    {
      "entropy": 5.681069564819336,
      "epoch": 1.1156899810964083,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004880876594211665,
      "loss": 5.2707,
      "mean_token_accuracy": 0.17547988444566726,
      "num_tokens": 24497087.0,
      "step": 13280
    },
    {
      "entropy": 5.774662446975708,
      "epoch": 1.1161100609115733,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048807804266580304,
      "loss": 5.2245,
      "mean_token_accuracy": 0.1806109294295311,
      "num_tokens": 24505347.0,
      "step": 13285
    },
    {
      "entropy": 5.719064426422119,
      "epoch": 1.116530140726738,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004880684221358717,
      "loss": 5.243,
      "mean_token_accuracy": 0.17971936911344527,
      "num_tokens": 24514732.0,
      "step": 13290
    },
    {
      "entropy": 5.706452226638794,
      "epoch": 1.116950220541903,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048805879783154305,
      "loss": 5.2804,
      "mean_token_accuracy": 0.17748649418354034,
      "num_tokens": 24523295.0,
      "step": 13295
    },
    {
      "entropy": 5.625413322448731,
      "epoch": 1.1173703003570679,
      "grad_norm": 1.0,
      "learning_rate": 0.00048804916975298744,
      "loss": 5.1882,
      "mean_token_accuracy": 0.18580531775951387,
      "num_tokens": 24532415.0,
      "step": 13300
    },
    {
      "entropy": 5.70411057472229,
      "epoch": 1.1177903801722326,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004880395379003755,
      "loss": 5.3263,
      "mean_token_accuracy": 0.18090976625680924,
      "num_tokens": 24541856.0,
      "step": 13305
    },
    {
      "entropy": 5.6655196189880375,
      "epoch": 1.1182104599873977,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048802990227387797,
      "loss": 5.3044,
      "mean_token_accuracy": 0.1737400248646736,
      "num_tokens": 24550982.0,
      "step": 13310
    },
    {
      "entropy": 5.694133138656616,
      "epoch": 1.1186305398025624,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048802026287366525,
      "loss": 5.3648,
      "mean_token_accuracy": 0.1724078133702278,
      "num_tokens": 24561176.0,
      "step": 13315
    },
    {
      "entropy": 5.7212522506713865,
      "epoch": 1.1190506196177274,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048801061969990834,
      "loss": 5.2453,
      "mean_token_accuracy": 0.182855486869812,
      "num_tokens": 24570741.0,
      "step": 13320
    },
    {
      "entropy": 5.6823338031768795,
      "epoch": 1.1194706994328922,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048800097275277795,
      "loss": 5.2568,
      "mean_token_accuracy": 0.1868749588727951,
      "num_tokens": 24580175.0,
      "step": 13325
    },
    {
      "entropy": 5.679512405395508,
      "epoch": 1.119890779248057,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000487991322032445,
      "loss": 5.2472,
      "mean_token_accuracy": 0.18417306393384933,
      "num_tokens": 24588754.0,
      "step": 13330
    },
    {
      "entropy": 5.78442416191101,
      "epoch": 1.120310859063222,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004879816675390805,
      "loss": 5.4221,
      "mean_token_accuracy": 0.175226566195488,
      "num_tokens": 24599429.0,
      "step": 13335
    },
    {
      "entropy": 5.638189506530762,
      "epoch": 1.1207309388783868,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048797200927285547,
      "loss": 5.1643,
      "mean_token_accuracy": 0.18462345004081726,
      "num_tokens": 24608767.0,
      "step": 13340
    },
    {
      "entropy": 5.669074964523316,
      "epoch": 1.1211510186935518,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004879623472339409,
      "loss": 5.3211,
      "mean_token_accuracy": 0.17880960255861283,
      "num_tokens": 24618232.0,
      "step": 13345
    },
    {
      "entropy": 5.712857675552368,
      "epoch": 1.1215710985087166,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000487952681422508,
      "loss": 5.213,
      "mean_token_accuracy": 0.1808924123644829,
      "num_tokens": 24626986.0,
      "step": 13350
    },
    {
      "entropy": 5.61821494102478,
      "epoch": 1.1219911783238816,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000487943011838728,
      "loss": 5.1038,
      "mean_token_accuracy": 0.19026215523481368,
      "num_tokens": 24635283.0,
      "step": 13355
    },
    {
      "entropy": 5.576014995574951,
      "epoch": 1.1224112581390464,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004879333384827722,
      "loss": 5.233,
      "mean_token_accuracy": 0.178211909532547,
      "num_tokens": 24644451.0,
      "step": 13360
    },
    {
      "entropy": 5.72415566444397,
      "epoch": 1.1228313379542114,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004879236613548119,
      "loss": 5.3318,
      "mean_token_accuracy": 0.17672962993383406,
      "num_tokens": 24654811.0,
      "step": 13365
    },
    {
      "entropy": 5.703073263168335,
      "epoch": 1.1232514177693762,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004879139804550187,
      "loss": 5.2767,
      "mean_token_accuracy": 0.18055036514997483,
      "num_tokens": 24663712.0,
      "step": 13370
    },
    {
      "entropy": 5.7139424800872805,
      "epoch": 1.123671497584541,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048790429578356387,
      "loss": 5.3587,
      "mean_token_accuracy": 0.18081534206867217,
      "num_tokens": 24672518.0,
      "step": 13375
    },
    {
      "entropy": 5.692781448364258,
      "epoch": 1.124091577399706,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048789460734061915,
      "loss": 5.2381,
      "mean_token_accuracy": 0.1923122689127922,
      "num_tokens": 24681900.0,
      "step": 13380
    },
    {
      "entropy": 5.6932816982269285,
      "epoch": 1.1245116572148708,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004878849151263561,
      "loss": 5.2575,
      "mean_token_accuracy": 0.18621968924999238,
      "num_tokens": 24691760.0,
      "step": 13385
    },
    {
      "entropy": 5.677702617645264,
      "epoch": 1.1249317370300358,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004878752191409463,
      "loss": 5.2108,
      "mean_token_accuracy": 0.1840396046638489,
      "num_tokens": 24700742.0,
      "step": 13390
    },
    {
      "entropy": 5.6617100715637205,
      "epoch": 1.1253518168452006,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004878655193845616,
      "loss": 5.2765,
      "mean_token_accuracy": 0.18067796230316163,
      "num_tokens": 24709329.0,
      "step": 13395
    },
    {
      "entropy": 5.7200017929077145,
      "epoch": 1.1257718966603654,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048785581585737394,
      "loss": 5.4228,
      "mean_token_accuracy": 0.17013319581747055,
      "num_tokens": 24718475.0,
      "step": 13400
    },
    {
      "entropy": 5.728095197677613,
      "epoch": 1.1261919764755304,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000487846108559555,
      "loss": 5.2747,
      "mean_token_accuracy": 0.18746441453695298,
      "num_tokens": 24727817.0,
      "step": 13405
    },
    {
      "entropy": 5.681646919250488,
      "epoch": 1.1266120562906952,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048783639749127694,
      "loss": 5.2609,
      "mean_token_accuracy": 0.17848434150218964,
      "num_tokens": 24737057.0,
      "step": 13410
    },
    {
      "entropy": 5.764699554443359,
      "epoch": 1.1270321361058602,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004878266826527116,
      "loss": 5.2948,
      "mean_token_accuracy": 0.17870959341526033,
      "num_tokens": 24746016.0,
      "step": 13415
    },
    {
      "entropy": 5.768991279602051,
      "epoch": 1.127452215921025,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048781696404403126,
      "loss": 5.2986,
      "mean_token_accuracy": 0.178691628575325,
      "num_tokens": 24755978.0,
      "step": 13420
    },
    {
      "entropy": 5.629824829101563,
      "epoch": 1.12787229573619,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048780724166540794,
      "loss": 5.2052,
      "mean_token_accuracy": 0.1819114401936531,
      "num_tokens": 24765255.0,
      "step": 13425
    },
    {
      "entropy": 5.632645750045777,
      "epoch": 1.1282923755513548,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004877975155170139,
      "loss": 5.2678,
      "mean_token_accuracy": 0.17932751923799514,
      "num_tokens": 24774339.0,
      "step": 13430
    },
    {
      "entropy": 5.687859344482422,
      "epoch": 1.1287124553665198,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004877877855990215,
      "loss": 5.2772,
      "mean_token_accuracy": 0.17869479209184647,
      "num_tokens": 24783236.0,
      "step": 13435
    },
    {
      "entropy": 5.623878526687622,
      "epoch": 1.1291325351816845,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000487778051911603,
      "loss": 5.2072,
      "mean_token_accuracy": 0.18106629252433776,
      "num_tokens": 24792168.0,
      "step": 13440
    },
    {
      "entropy": 5.736230516433716,
      "epoch": 1.1295526149968493,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004877683144549308,
      "loss": 5.3259,
      "mean_token_accuracy": 0.17919528484344482,
      "num_tokens": 24800843.0,
      "step": 13445
    },
    {
      "entropy": 5.688276720046997,
      "epoch": 1.1299726948120143,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048775857322917753,
      "loss": 5.2071,
      "mean_token_accuracy": 0.17959577739238738,
      "num_tokens": 24810475.0,
      "step": 13450
    },
    {
      "entropy": 5.642524385452271,
      "epoch": 1.1303927746271791,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004877488282345158,
      "loss": 5.2815,
      "mean_token_accuracy": 0.17896010279655455,
      "num_tokens": 24820486.0,
      "step": 13455
    },
    {
      "entropy": 5.757071781158447,
      "epoch": 1.1308128544423441,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000487739079471118,
      "loss": 5.3465,
      "mean_token_accuracy": 0.17830272614955903,
      "num_tokens": 24830243.0,
      "step": 13460
    },
    {
      "entropy": 5.715437364578247,
      "epoch": 1.131232934257509,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000487729326939157,
      "loss": 5.2671,
      "mean_token_accuracy": 0.17592093497514724,
      "num_tokens": 24839090.0,
      "step": 13465
    },
    {
      "entropy": 5.6381776332855225,
      "epoch": 1.1316530140726737,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048771957063880553,
      "loss": 5.2606,
      "mean_token_accuracy": 0.18127904534339906,
      "num_tokens": 24847933.0,
      "step": 13470
    },
    {
      "entropy": 5.685086727142334,
      "epoch": 1.1320730938878387,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004877098105702363,
      "loss": 5.2466,
      "mean_token_accuracy": 0.18696131259202958,
      "num_tokens": 24857037.0,
      "step": 13475
    },
    {
      "entropy": 5.555217838287353,
      "epoch": 1.1324931737030035,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048770004673362243,
      "loss": 5.0951,
      "mean_token_accuracy": 0.19122041165828704,
      "num_tokens": 24866042.0,
      "step": 13480
    },
    {
      "entropy": 5.533856153488159,
      "epoch": 1.1329132535181685,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048769027912913673,
      "loss": 5.061,
      "mean_token_accuracy": 0.19720048308372498,
      "num_tokens": 24873735.0,
      "step": 13485
    },
    {
      "entropy": 5.573809671401977,
      "epoch": 1.1333333333333333,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004876805077569522,
      "loss": 5.1478,
      "mean_token_accuracy": 0.18459829986095427,
      "num_tokens": 24882277.0,
      "step": 13490
    },
    {
      "entropy": 5.60994668006897,
      "epoch": 1.133753413148498,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048767073261724204,
      "loss": 5.2675,
      "mean_token_accuracy": 0.17937127500772476,
      "num_tokens": 24891354.0,
      "step": 13495
    },
    {
      "entropy": 5.667100048065185,
      "epoch": 1.134173492963663,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004876609537101793,
      "loss": 5.2394,
      "mean_token_accuracy": 0.17811793833971024,
      "num_tokens": 24899887.0,
      "step": 13500
    },
    {
      "entropy": 5.774137115478515,
      "epoch": 1.1345935727788279,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004876511710359374,
      "loss": 5.2631,
      "mean_token_accuracy": 0.18105413019657135,
      "num_tokens": 24908616.0,
      "step": 13505
    },
    {
      "entropy": 5.681682968139649,
      "epoch": 1.135013652593993,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048764138459468935,
      "loss": 5.2954,
      "mean_token_accuracy": 0.18110547959804535,
      "num_tokens": 24917864.0,
      "step": 13510
    },
    {
      "entropy": 5.774509143829346,
      "epoch": 1.1354337324091577,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048763159438660876,
      "loss": 5.3245,
      "mean_token_accuracy": 0.18085842430591584,
      "num_tokens": 24927864.0,
      "step": 13515
    },
    {
      "entropy": 5.6376869678497314,
      "epoch": 1.1358538122243227,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048762180041186893,
      "loss": 5.2173,
      "mean_token_accuracy": 0.18946526795625687,
      "num_tokens": 24937146.0,
      "step": 13520
    },
    {
      "entropy": 5.7010866641998295,
      "epoch": 1.1362738920394875,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004876120026706434,
      "loss": 5.2933,
      "mean_token_accuracy": 0.18329816162586213,
      "num_tokens": 24945694.0,
      "step": 13525
    },
    {
      "entropy": 5.653949785232544,
      "epoch": 1.1366939718546525,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004876022011631057,
      "loss": 5.2072,
      "mean_token_accuracy": 0.18480013459920883,
      "num_tokens": 24955325.0,
      "step": 13530
    },
    {
      "entropy": 5.604835748672485,
      "epoch": 1.1371140516698173,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004875923958894295,
      "loss": 5.0938,
      "mean_token_accuracy": 0.18716135025024414,
      "num_tokens": 24964028.0,
      "step": 13535
    },
    {
      "entropy": 5.669315528869629,
      "epoch": 1.137534131484982,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048758258684978846,
      "loss": 5.2682,
      "mean_token_accuracy": 0.1837463393807411,
      "num_tokens": 24972923.0,
      "step": 13540
    },
    {
      "entropy": 5.629849243164062,
      "epoch": 1.137954211300147,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048757277404435636,
      "loss": 5.1736,
      "mean_token_accuracy": 0.1847257986664772,
      "num_tokens": 24982156.0,
      "step": 13545
    },
    {
      "entropy": 5.637995862960816,
      "epoch": 1.1383742911153119,
      "grad_norm": 1.015625,
      "learning_rate": 0.000487562957473307,
      "loss": 5.2268,
      "mean_token_accuracy": 0.18306522220373153,
      "num_tokens": 24991616.0,
      "step": 13550
    },
    {
      "entropy": 5.619869184494019,
      "epoch": 1.1387943709304769,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004875531371368144,
      "loss": 5.2695,
      "mean_token_accuracy": 0.18245483785867692,
      "num_tokens": 25001140.0,
      "step": 13555
    },
    {
      "entropy": 5.617695569992065,
      "epoch": 1.1392144507456416,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048754331303505236,
      "loss": 5.1852,
      "mean_token_accuracy": 0.1882428079843521,
      "num_tokens": 25010863.0,
      "step": 13560
    },
    {
      "entropy": 5.698365259170532,
      "epoch": 1.1396345305608064,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048753348516819496,
      "loss": 5.2926,
      "mean_token_accuracy": 0.181953302025795,
      "num_tokens": 25019770.0,
      "step": 13565
    },
    {
      "entropy": 5.778852033615112,
      "epoch": 1.1400546103759714,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004875236535364163,
      "loss": 5.3286,
      "mean_token_accuracy": 0.17123429775238036,
      "num_tokens": 25029900.0,
      "step": 13570
    },
    {
      "entropy": 5.721889066696167,
      "epoch": 1.1404746901911362,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004875138181398906,
      "loss": 5.294,
      "mean_token_accuracy": 0.18345095813274384,
      "num_tokens": 25039428.0,
      "step": 13575
    },
    {
      "entropy": 5.671444892883301,
      "epoch": 1.1408947700063012,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000487503978978792,
      "loss": 5.2694,
      "mean_token_accuracy": 0.17903570234775543,
      "num_tokens": 25049145.0,
      "step": 13580
    },
    {
      "entropy": 5.683608484268189,
      "epoch": 1.141314849821466,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048749413605329487,
      "loss": 5.286,
      "mean_token_accuracy": 0.1833672672510147,
      "num_tokens": 25058772.0,
      "step": 13585
    },
    {
      "entropy": 5.676011800765991,
      "epoch": 1.141734929636631,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048748428936357346,
      "loss": 5.2148,
      "mean_token_accuracy": 0.18349185436964036,
      "num_tokens": 25067249.0,
      "step": 13590
    },
    {
      "entropy": 5.624266576766968,
      "epoch": 1.1421550094517958,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004874744389098024,
      "loss": 5.1761,
      "mean_token_accuracy": 0.18314225375652313,
      "num_tokens": 25076893.0,
      "step": 13595
    },
    {
      "entropy": 5.591076374053955,
      "epoch": 1.1425750892669608,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004874645846921559,
      "loss": 5.1817,
      "mean_token_accuracy": 0.19296690821647644,
      "num_tokens": 25086238.0,
      "step": 13600
    },
    {
      "entropy": 5.660066604614258,
      "epoch": 1.1429951690821256,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048745472671080884,
      "loss": 5.2123,
      "mean_token_accuracy": 0.1812112271785736,
      "num_tokens": 25095334.0,
      "step": 13605
    },
    {
      "entropy": 5.677392959594727,
      "epoch": 1.1434152488972904,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048744486496593565,
      "loss": 5.2131,
      "mean_token_accuracy": 0.18197848349809648,
      "num_tokens": 25104136.0,
      "step": 13610
    },
    {
      "entropy": 5.652319240570068,
      "epoch": 1.1438353287124554,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000487434999457711,
      "loss": 5.1979,
      "mean_token_accuracy": 0.19154253453016282,
      "num_tokens": 25112629.0,
      "step": 13615
    },
    {
      "entropy": 5.659061670303345,
      "epoch": 1.1442554085276202,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004874251301863098,
      "loss": 5.2357,
      "mean_token_accuracy": 0.18086742907762526,
      "num_tokens": 25121014.0,
      "step": 13620
    },
    {
      "entropy": 5.636931753158569,
      "epoch": 1.1446754883427852,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048741525715190675,
      "loss": 5.2837,
      "mean_token_accuracy": 0.17920210659503938,
      "num_tokens": 25130097.0,
      "step": 13625
    },
    {
      "entropy": 5.723202323913574,
      "epoch": 1.14509556815795,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004874053803546769,
      "loss": 5.274,
      "mean_token_accuracy": 0.18140329122543336,
      "num_tokens": 25139065.0,
      "step": 13630
    },
    {
      "entropy": 5.647499036788941,
      "epoch": 1.1455156479731148,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000487395499794795,
      "loss": 5.2491,
      "mean_token_accuracy": 0.1823761895298958,
      "num_tokens": 25148852.0,
      "step": 13635
    },
    {
      "entropy": 5.604208755493164,
      "epoch": 1.1459357277882798,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004873856154724362,
      "loss": 5.1505,
      "mean_token_accuracy": 0.19094152450561525,
      "num_tokens": 25157580.0,
      "step": 13640
    },
    {
      "entropy": 5.694274425506592,
      "epoch": 1.1463558076034446,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004873757273877756,
      "loss": 5.2531,
      "mean_token_accuracy": 0.17768404334783555,
      "num_tokens": 25166243.0,
      "step": 13645
    },
    {
      "entropy": 5.6587567806243895,
      "epoch": 1.1467758874186096,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048736583554098836,
      "loss": 5.2732,
      "mean_token_accuracy": 0.17753455638885499,
      "num_tokens": 25174674.0,
      "step": 13650
    },
    {
      "entropy": 5.589343690872193,
      "epoch": 1.1471959672337744,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048735593993224973,
      "loss": 5.172,
      "mean_token_accuracy": 0.19389842450618744,
      "num_tokens": 25183892.0,
      "step": 13655
    },
    {
      "entropy": 5.603427839279175,
      "epoch": 1.1476160470489394,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048734604056173495,
      "loss": 5.223,
      "mean_token_accuracy": 0.182113941013813,
      "num_tokens": 25192731.0,
      "step": 13660
    },
    {
      "entropy": 5.684765005111695,
      "epoch": 1.1480361268641042,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048733613742961933,
      "loss": 5.3253,
      "mean_token_accuracy": 0.17760632783174515,
      "num_tokens": 25201280.0,
      "step": 13665
    },
    {
      "entropy": 5.66482458114624,
      "epoch": 1.1484562066792692,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048732623053607846,
      "loss": 5.1874,
      "mean_token_accuracy": 0.18178779482841492,
      "num_tokens": 25209929.0,
      "step": 13670
    },
    {
      "entropy": 5.660930252075195,
      "epoch": 1.148876286494434,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004873163198812877,
      "loss": 5.1339,
      "mean_token_accuracy": 0.18998833298683165,
      "num_tokens": 25218583.0,
      "step": 13675
    },
    {
      "entropy": 5.684929895401001,
      "epoch": 1.1492963663095987,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004873064054654227,
      "loss": 5.2983,
      "mean_token_accuracy": 0.17942339330911636,
      "num_tokens": 25228949.0,
      "step": 13680
    },
    {
      "entropy": 5.658124494552612,
      "epoch": 1.1497164461247638,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048729648728865904,
      "loss": 5.1853,
      "mean_token_accuracy": 0.1956553876399994,
      "num_tokens": 25238603.0,
      "step": 13685
    },
    {
      "entropy": 5.664039134979248,
      "epoch": 1.1501365259399285,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048728656535117237,
      "loss": 5.2929,
      "mean_token_accuracy": 0.17787734419107437,
      "num_tokens": 25248265.0,
      "step": 13690
    },
    {
      "entropy": 5.657366037368774,
      "epoch": 1.1505566057550936,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004872766396531386,
      "loss": 5.259,
      "mean_token_accuracy": 0.18284955322742463,
      "num_tokens": 25258195.0,
      "step": 13695
    },
    {
      "entropy": 5.754784488677979,
      "epoch": 1.1509766855702583,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048726671019473335,
      "loss": 5.222,
      "mean_token_accuracy": 0.1874355599284172,
      "num_tokens": 25267886.0,
      "step": 13700
    },
    {
      "entropy": 5.6908159255981445,
      "epoch": 1.1513967653854231,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048725677697613267,
      "loss": 5.2582,
      "mean_token_accuracy": 0.18304793238639833,
      "num_tokens": 25277304.0,
      "step": 13705
    },
    {
      "entropy": 5.654964303970337,
      "epoch": 1.1518168452005881,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004872468399975125,
      "loss": 5.257,
      "mean_token_accuracy": 0.18120323717594147,
      "num_tokens": 25286771.0,
      "step": 13710
    },
    {
      "entropy": 5.75738935470581,
      "epoch": 1.152236925015753,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048723689925904884,
      "loss": 5.3184,
      "mean_token_accuracy": 0.1802498534321785,
      "num_tokens": 25296018.0,
      "step": 13715
    },
    {
      "entropy": 5.682253980636597,
      "epoch": 1.152657004830918,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004872269547609179,
      "loss": 5.283,
      "mean_token_accuracy": 0.18282630890607834,
      "num_tokens": 25305737.0,
      "step": 13720
    },
    {
      "entropy": 5.5883081436157225,
      "epoch": 1.1530770846460827,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004872170065032956,
      "loss": 5.1116,
      "mean_token_accuracy": 0.19130570739507674,
      "num_tokens": 25314625.0,
      "step": 13725
    },
    {
      "entropy": 5.6583571434021,
      "epoch": 1.1534971644612477,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004872070544863584,
      "loss": 5.2393,
      "mean_token_accuracy": 0.18421585559844972,
      "num_tokens": 25323453.0,
      "step": 13730
    },
    {
      "entropy": 5.680066013336182,
      "epoch": 1.1539172442764125,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004871970987102824,
      "loss": 5.2696,
      "mean_token_accuracy": 0.18223375380039214,
      "num_tokens": 25333236.0,
      "step": 13735
    },
    {
      "entropy": 5.648608779907226,
      "epoch": 1.1543373240915775,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004871871391752442,
      "loss": 5.1669,
      "mean_token_accuracy": 0.1813027337193489,
      "num_tokens": 25341993.0,
      "step": 13740
    },
    {
      "entropy": 5.699672603607178,
      "epoch": 1.1547574039067423,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048717717588141993,
      "loss": 5.1986,
      "mean_token_accuracy": 0.18249054849147797,
      "num_tokens": 25350695.0,
      "step": 13745
    },
    {
      "entropy": 5.670484828948974,
      "epoch": 1.155177483721907,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004871672088289863,
      "loss": 5.2416,
      "mean_token_accuracy": 0.1824687734246254,
      "num_tokens": 25359044.0,
      "step": 13750
    },
    {
      "entropy": 5.603364896774292,
      "epoch": 1.155597563537072,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048715723801811986,
      "loss": 5.2625,
      "mean_token_accuracy": 0.18367746025323867,
      "num_tokens": 25367959.0,
      "step": 13755
    },
    {
      "entropy": 5.66744213104248,
      "epoch": 1.156017643352237,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048714726344899716,
      "loss": 5.2789,
      "mean_token_accuracy": 0.18486496806144714,
      "num_tokens": 25376968.0,
      "step": 13760
    },
    {
      "entropy": 5.6204994201660154,
      "epoch": 1.156437723167402,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004871372851217949,
      "loss": 5.1663,
      "mean_token_accuracy": 0.19086449295282365,
      "num_tokens": 25385381.0,
      "step": 13765
    },
    {
      "entropy": 5.668493366241455,
      "epoch": 1.1568578029825667,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004871273030366899,
      "loss": 5.2751,
      "mean_token_accuracy": 0.17997864931821822,
      "num_tokens": 25394647.0,
      "step": 13770
    },
    {
      "entropy": 5.643715238571167,
      "epoch": 1.1572778827977315,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004871173171938589,
      "loss": 5.2152,
      "mean_token_accuracy": 0.1937444716691971,
      "num_tokens": 25403973.0,
      "step": 13775
    },
    {
      "entropy": 5.693225955963134,
      "epoch": 1.1576979626128965,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004871073275934789,
      "loss": 5.2222,
      "mean_token_accuracy": 0.18283227682113648,
      "num_tokens": 25412319.0,
      "step": 13780
    },
    {
      "entropy": 5.624289894104004,
      "epoch": 1.1581180424280613,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048709733423572685,
      "loss": 5.2434,
      "mean_token_accuracy": 0.18010500073432922,
      "num_tokens": 25420558.0,
      "step": 13785
    },
    {
      "entropy": 5.607655429840088,
      "epoch": 1.1585381222432263,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048708733712077973,
      "loss": 5.1815,
      "mean_token_accuracy": 0.18967724293470384,
      "num_tokens": 25429258.0,
      "step": 13790
    },
    {
      "entropy": 5.664348602294922,
      "epoch": 1.158958202058391,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004870773362488146,
      "loss": 5.1633,
      "mean_token_accuracy": 0.18552689999341965,
      "num_tokens": 25438005.0,
      "step": 13795
    },
    {
      "entropy": 5.628162240982055,
      "epoch": 1.159378281873556,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004870673316200087,
      "loss": 5.1847,
      "mean_token_accuracy": 0.18308848291635513,
      "num_tokens": 25447120.0,
      "step": 13800
    },
    {
      "entropy": 5.6105506896972654,
      "epoch": 1.1597983616887209,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004870573232345392,
      "loss": 5.1728,
      "mean_token_accuracy": 0.17910649478435517,
      "num_tokens": 25456216.0,
      "step": 13805
    },
    {
      "entropy": 5.827288389205933,
      "epoch": 1.1602184415038856,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004870473110925834,
      "loss": 5.4148,
      "mean_token_accuracy": 0.17368083745241164,
      "num_tokens": 25466456.0,
      "step": 13810
    },
    {
      "entropy": 5.615953493118286,
      "epoch": 1.1606385213190507,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004870372951943187,
      "loss": 5.0827,
      "mean_token_accuracy": 0.19566079080104828,
      "num_tokens": 25475217.0,
      "step": 13815
    },
    {
      "entropy": 5.7068931579589846,
      "epoch": 1.1610586011342154,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048702727553992243,
      "loss": 5.3724,
      "mean_token_accuracy": 0.16979683935642242,
      "num_tokens": 25484617.0,
      "step": 13820
    },
    {
      "entropy": 5.664984226226807,
      "epoch": 1.1614786809493804,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048701725212957223,
      "loss": 5.1921,
      "mean_token_accuracy": 0.1883465677499771,
      "num_tokens": 25493936.0,
      "step": 13825
    },
    {
      "entropy": 5.640343379974365,
      "epoch": 1.1618987607645452,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004870072249634455,
      "loss": 5.1655,
      "mean_token_accuracy": 0.1913938343524933,
      "num_tokens": 25502306.0,
      "step": 13830
    },
    {
      "entropy": 5.576923274993897,
      "epoch": 1.1623188405797102,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048699719404172006,
      "loss": 5.2228,
      "mean_token_accuracy": 0.1856873705983162,
      "num_tokens": 25511247.0,
      "step": 13835
    },
    {
      "entropy": 5.638578224182129,
      "epoch": 1.162738920394875,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048698715936457344,
      "loss": 5.2619,
      "mean_token_accuracy": 0.18379607796669006,
      "num_tokens": 25520482.0,
      "step": 13840
    },
    {
      "entropy": 5.670048999786377,
      "epoch": 1.1631590002100398,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048697712093218336,
      "loss": 5.16,
      "mean_token_accuracy": 0.18658994138240814,
      "num_tokens": 25529854.0,
      "step": 13845
    },
    {
      "entropy": 5.6327488899230955,
      "epoch": 1.1635790800252048,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004869670787447279,
      "loss": 5.1141,
      "mean_token_accuracy": 0.18678332418203353,
      "num_tokens": 25538251.0,
      "step": 13850
    },
    {
      "entropy": 5.594510698318482,
      "epoch": 1.1639991598403696,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004869570328023846,
      "loss": 5.217,
      "mean_token_accuracy": 0.18038427829742432,
      "num_tokens": 25546889.0,
      "step": 13855
    },
    {
      "entropy": 5.600675392150879,
      "epoch": 1.1644192396555346,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048694698310533177,
      "loss": 5.2327,
      "mean_token_accuracy": 0.18146373629570006,
      "num_tokens": 25557040.0,
      "step": 13860
    },
    {
      "entropy": 5.740075254440308,
      "epoch": 1.1648393194706994,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004869369296537472,
      "loss": 5.4084,
      "mean_token_accuracy": 0.1690896049141884,
      "num_tokens": 25565798.0,
      "step": 13865
    },
    {
      "entropy": 5.757442855834961,
      "epoch": 1.1652593992858642,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004869268724478091,
      "loss": 5.2488,
      "mean_token_accuracy": 0.18638191968202591,
      "num_tokens": 25575039.0,
      "step": 13870
    },
    {
      "entropy": 5.738666296005249,
      "epoch": 1.1656794791010292,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048691681148769545,
      "loss": 5.2525,
      "mean_token_accuracy": 0.1840719014406204,
      "num_tokens": 25584635.0,
      "step": 13875
    },
    {
      "entropy": 5.595553588867188,
      "epoch": 1.166099558916194,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004869067467735847,
      "loss": 5.192,
      "mean_token_accuracy": 0.18165479153394698,
      "num_tokens": 25593736.0,
      "step": 13880
    },
    {
      "entropy": 5.650736093521118,
      "epoch": 1.166519638731359,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004868966783056551,
      "loss": 5.1512,
      "mean_token_accuracy": 0.1938465118408203,
      "num_tokens": 25602685.0,
      "step": 13885
    },
    {
      "entropy": 5.654877853393555,
      "epoch": 1.1669397185465238,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048688660608408484,
      "loss": 5.226,
      "mean_token_accuracy": 0.18053028732538223,
      "num_tokens": 25610690.0,
      "step": 13890
    },
    {
      "entropy": 5.591117095947266,
      "epoch": 1.1673597983616888,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048687653010905254,
      "loss": 5.1054,
      "mean_token_accuracy": 0.18998621702194213,
      "num_tokens": 25619805.0,
      "step": 13895
    },
    {
      "entropy": 5.69170298576355,
      "epoch": 1.1677798781768536,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048686645038073664,
      "loss": 5.3059,
      "mean_token_accuracy": 0.1793782263994217,
      "num_tokens": 25629447.0,
      "step": 13900
    },
    {
      "entropy": 5.642636632919311,
      "epoch": 1.1681999579920186,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048685636689931554,
      "loss": 5.1872,
      "mean_token_accuracy": 0.185796856880188,
      "num_tokens": 25638619.0,
      "step": 13905
    },
    {
      "entropy": 5.688897180557251,
      "epoch": 1.1686200378071834,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048684627966496803,
      "loss": 5.2528,
      "mean_token_accuracy": 0.18300721347332,
      "num_tokens": 25648255.0,
      "step": 13910
    },
    {
      "entropy": 5.650645112991333,
      "epoch": 1.1690401176223482,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048683618867787284,
      "loss": 5.2748,
      "mean_token_accuracy": 0.18363728523254394,
      "num_tokens": 25657881.0,
      "step": 13915
    },
    {
      "entropy": 5.7307051658630375,
      "epoch": 1.1694601974375132,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004868260939382086,
      "loss": 5.3241,
      "mean_token_accuracy": 0.17719403952360152,
      "num_tokens": 25666773.0,
      "step": 13920
    },
    {
      "entropy": 5.666933012008667,
      "epoch": 1.169880277252678,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004868159954461542,
      "loss": 5.213,
      "mean_token_accuracy": 0.18616429418325425,
      "num_tokens": 25675152.0,
      "step": 13925
    },
    {
      "entropy": 5.742690801620483,
      "epoch": 1.170300357067843,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048680589320188847,
      "loss": 5.3113,
      "mean_token_accuracy": 0.17788237035274507,
      "num_tokens": 25684962.0,
      "step": 13930
    },
    {
      "entropy": 5.622620582580566,
      "epoch": 1.1707204368830078,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004867957872055904,
      "loss": 5.2336,
      "mean_token_accuracy": 0.18070897608995437,
      "num_tokens": 25693782.0,
      "step": 13935
    },
    {
      "entropy": 5.605896043777466,
      "epoch": 1.1711405166981725,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048678567745743905,
      "loss": 5.1853,
      "mean_token_accuracy": 0.1891879841685295,
      "num_tokens": 25703081.0,
      "step": 13940
    },
    {
      "entropy": 5.620411539077759,
      "epoch": 1.1715605965133375,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004867755639576135,
      "loss": 5.2145,
      "mean_token_accuracy": 0.18457203954458237,
      "num_tokens": 25711628.0,
      "step": 13945
    },
    {
      "entropy": 5.59106855392456,
      "epoch": 1.1719806763285023,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004867654467062928,
      "loss": 5.228,
      "mean_token_accuracy": 0.185416641831398,
      "num_tokens": 25720676.0,
      "step": 13950
    },
    {
      "entropy": 5.611890506744385,
      "epoch": 1.1724007561436673,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048675532570365633,
      "loss": 5.1939,
      "mean_token_accuracy": 0.1874692142009735,
      "num_tokens": 25729920.0,
      "step": 13955
    },
    {
      "entropy": 5.640635251998901,
      "epoch": 1.1728208359588321,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048674520094988327,
      "loss": 5.1834,
      "mean_token_accuracy": 0.19211795628070832,
      "num_tokens": 25739745.0,
      "step": 13960
    },
    {
      "entropy": 5.6493833541870115,
      "epoch": 1.1732409157739971,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048673507244515303,
      "loss": 5.182,
      "mean_token_accuracy": 0.18443135023117066,
      "num_tokens": 25748636.0,
      "step": 13965
    },
    {
      "entropy": 5.689094686508179,
      "epoch": 1.173660995589162,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000486724940189645,
      "loss": 5.3272,
      "mean_token_accuracy": 0.1801094800233841,
      "num_tokens": 25758393.0,
      "step": 13970
    },
    {
      "entropy": 5.690440559387207,
      "epoch": 1.174081075404327,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004867148041835386,
      "loss": 5.3102,
      "mean_token_accuracy": 0.17558927536010743,
      "num_tokens": 25768520.0,
      "step": 13975
    },
    {
      "entropy": 5.62732081413269,
      "epoch": 1.1745011552194917,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004867046644270136,
      "loss": 5.1134,
      "mean_token_accuracy": 0.190086205303669,
      "num_tokens": 25777168.0,
      "step": 13980
    },
    {
      "entropy": 5.776018524169922,
      "epoch": 1.1749212350346565,
      "grad_norm": 1.0,
      "learning_rate": 0.0004866945209202494,
      "loss": 5.4284,
      "mean_token_accuracy": 0.16918590068817138,
      "num_tokens": 25787042.0,
      "step": 13985
    },
    {
      "entropy": 5.722411108016968,
      "epoch": 1.1753413148498215,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004866843736634258,
      "loss": 5.2971,
      "mean_token_accuracy": 0.1831905409693718,
      "num_tokens": 25796784.0,
      "step": 13990
    },
    {
      "entropy": 5.743758678436279,
      "epoch": 1.1757613946649863,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004866742226567225,
      "loss": 5.3391,
      "mean_token_accuracy": 0.17153186202049256,
      "num_tokens": 25806285.0,
      "step": 13995
    },
    {
      "entropy": 5.6624620914459225,
      "epoch": 1.1761814744801513,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048666406790031936,
      "loss": 5.1839,
      "mean_token_accuracy": 0.18119580298662186,
      "num_tokens": 25814889.0,
      "step": 14000
    },
    {
      "entropy": 5.623716068267822,
      "epoch": 1.176601554295316,
      "grad_norm": 1.0,
      "learning_rate": 0.0004866539093943962,
      "loss": 5.2325,
      "mean_token_accuracy": 0.1807695984840393,
      "num_tokens": 25824551.0,
      "step": 14005
    },
    {
      "entropy": 5.692598628997803,
      "epoch": 1.1770216341104809,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048664374713913304,
      "loss": 5.2484,
      "mean_token_accuracy": 0.18016877323389052,
      "num_tokens": 25834482.0,
      "step": 14010
    },
    {
      "entropy": 5.730862617492676,
      "epoch": 1.177441713925646,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004866335811347099,
      "loss": 5.2912,
      "mean_token_accuracy": 0.17793221324682235,
      "num_tokens": 25843274.0,
      "step": 14015
    },
    {
      "entropy": 5.750838327407837,
      "epoch": 1.1778617937408107,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048662341138130683,
      "loss": 5.288,
      "mean_token_accuracy": 0.17654060572385788,
      "num_tokens": 25852482.0,
      "step": 14020
    },
    {
      "entropy": 5.70770754814148,
      "epoch": 1.1782818735559757,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048661323787910405,
      "loss": 5.2781,
      "mean_token_accuracy": 0.1758618801832199,
      "num_tokens": 25862657.0,
      "step": 14025
    },
    {
      "entropy": 5.63136510848999,
      "epoch": 1.1787019533711405,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004866030606282817,
      "loss": 5.2191,
      "mean_token_accuracy": 0.18614309281110764,
      "num_tokens": 25871492.0,
      "step": 14030
    },
    {
      "entropy": 5.6397780418396,
      "epoch": 1.1791220331863055,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048659287962902006,
      "loss": 5.2286,
      "mean_token_accuracy": 0.1863328218460083,
      "num_tokens": 25880979.0,
      "step": 14035
    },
    {
      "entropy": 5.673973655700683,
      "epoch": 1.1795421130014703,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048658269488149945,
      "loss": 5.2037,
      "mean_token_accuracy": 0.1820206716656685,
      "num_tokens": 25891060.0,
      "step": 14040
    },
    {
      "entropy": 5.773657894134521,
      "epoch": 1.1799621928166353,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004865725063859005,
      "loss": 5.3368,
      "mean_token_accuracy": 0.17824782133102418,
      "num_tokens": 25900421.0,
      "step": 14045
    },
    {
      "entropy": 5.616058111190796,
      "epoch": 1.1803822726318,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048656231414240345,
      "loss": 5.2466,
      "mean_token_accuracy": 0.17775113135576248,
      "num_tokens": 25909614.0,
      "step": 14050
    },
    {
      "entropy": 5.590089321136475,
      "epoch": 1.1808023524469649,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000486552118151189,
      "loss": 5.2766,
      "mean_token_accuracy": 0.17353931814432144,
      "num_tokens": 25919324.0,
      "step": 14055
    },
    {
      "entropy": 5.676435899734497,
      "epoch": 1.1812224322621299,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048654191841243763,
      "loss": 5.265,
      "mean_token_accuracy": 0.18936726897954942,
      "num_tokens": 25928818.0,
      "step": 14060
    },
    {
      "entropy": 5.716177654266358,
      "epoch": 1.1816425120772946,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004865317149263301,
      "loss": 5.3236,
      "mean_token_accuracy": 0.18030305653810502,
      "num_tokens": 25938148.0,
      "step": 14065
    },
    {
      "entropy": 5.687571096420288,
      "epoch": 1.1820625918924597,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004865215076930473,
      "loss": 5.233,
      "mean_token_accuracy": 0.1857810452580452,
      "num_tokens": 25947210.0,
      "step": 14070
    },
    {
      "entropy": 5.645349073410034,
      "epoch": 1.1824826717076244,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004865112967127697,
      "loss": 5.2207,
      "mean_token_accuracy": 0.1853203907608986,
      "num_tokens": 25955949.0,
      "step": 14075
    },
    {
      "entropy": 5.577921962738037,
      "epoch": 1.1829027515227892,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004865010819856786,
      "loss": 5.1554,
      "mean_token_accuracy": 0.1812880367040634,
      "num_tokens": 25964193.0,
      "step": 14080
    },
    {
      "entropy": 5.636929321289062,
      "epoch": 1.1833228313379542,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004864908635119546,
      "loss": 5.2528,
      "mean_token_accuracy": 0.18184894621372222,
      "num_tokens": 25973141.0,
      "step": 14085
    },
    {
      "entropy": 5.726170635223388,
      "epoch": 1.183742911153119,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004864806412917788,
      "loss": 5.3003,
      "mean_token_accuracy": 0.18087310045957566,
      "num_tokens": 25982650.0,
      "step": 14090
    },
    {
      "entropy": 5.7547783851623535,
      "epoch": 1.184162990968284,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004864704153253325,
      "loss": 5.3007,
      "mean_token_accuracy": 0.17744034379720688,
      "num_tokens": 25992096.0,
      "step": 14095
    },
    {
      "entropy": 5.737809467315674,
      "epoch": 1.1845830707834488,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048646018561279665,
      "loss": 5.2845,
      "mean_token_accuracy": 0.17958581447601318,
      "num_tokens": 26002063.0,
      "step": 14100
    },
    {
      "entropy": 5.5846648693084715,
      "epoch": 1.1850031505986138,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048644995215435245,
      "loss": 5.1398,
      "mean_token_accuracy": 0.18903977274894715,
      "num_tokens": 26010716.0,
      "step": 14105
    },
    {
      "entropy": 5.651952314376831,
      "epoch": 1.1854232304137786,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004864397149501812,
      "loss": 5.203,
      "mean_token_accuracy": 0.18372938483953477,
      "num_tokens": 26019136.0,
      "step": 14110
    },
    {
      "entropy": 5.649115896224975,
      "epoch": 1.1858433102289434,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048642947400046434,
      "loss": 5.2385,
      "mean_token_accuracy": 0.18921892046928407,
      "num_tokens": 26028029.0,
      "step": 14115
    },
    {
      "entropy": 5.753901767730713,
      "epoch": 1.1862633900441084,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048641922930538325,
      "loss": 5.3637,
      "mean_token_accuracy": 0.17756056487560273,
      "num_tokens": 26038025.0,
      "step": 14120
    },
    {
      "entropy": 5.711358213424683,
      "epoch": 1.1866834698592732,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004864089808651193,
      "loss": 5.3398,
      "mean_token_accuracy": 0.1663362666964531,
      "num_tokens": 26048427.0,
      "step": 14125
    },
    {
      "entropy": 5.647575187683105,
      "epoch": 1.1871035496744382,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004863987286798541,
      "loss": 5.1447,
      "mean_token_accuracy": 0.1833113580942154,
      "num_tokens": 26057682.0,
      "step": 14130
    },
    {
      "entropy": 5.6246720314025875,
      "epoch": 1.187523629489603,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004863884727497693,
      "loss": 5.2222,
      "mean_token_accuracy": 0.1814057245850563,
      "num_tokens": 26066562.0,
      "step": 14135
    },
    {
      "entropy": 5.597769260406494,
      "epoch": 1.187943709304768,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004863782130750466,
      "loss": 5.1707,
      "mean_token_accuracy": 0.18318833112716676,
      "num_tokens": 26075633.0,
      "step": 14140
    },
    {
      "entropy": 5.678184938430786,
      "epoch": 1.1883637891199328,
      "grad_norm": 1.0,
      "learning_rate": 0.00048636794965586764,
      "loss": 5.3058,
      "mean_token_accuracy": 0.17958724945783616,
      "num_tokens": 26085160.0,
      "step": 14145
    },
    {
      "entropy": 5.683256673812866,
      "epoch": 1.1887838689350976,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048635768249241434,
      "loss": 5.2069,
      "mean_token_accuracy": 0.18133674263954164,
      "num_tokens": 26094157.0,
      "step": 14150
    },
    {
      "entropy": 5.7356878280639645,
      "epoch": 1.1892039487502626,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004863474115848685,
      "loss": 5.31,
      "mean_token_accuracy": 0.1802584409713745,
      "num_tokens": 26104459.0,
      "step": 14155
    },
    {
      "entropy": 5.596679544448852,
      "epoch": 1.1896240285654274,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048633713693341214,
      "loss": 5.2555,
      "mean_token_accuracy": 0.18036611825227739,
      "num_tokens": 26114468.0,
      "step": 14160
    },
    {
      "entropy": 5.630685472488404,
      "epoch": 1.1900441083805924,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048632685853822714,
      "loss": 5.2322,
      "mean_token_accuracy": 0.17721411138772963,
      "num_tokens": 26123408.0,
      "step": 14165
    },
    {
      "entropy": 5.67579083442688,
      "epoch": 1.1904641881957572,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004863165763994957,
      "loss": 5.2505,
      "mean_token_accuracy": 0.1760255441069603,
      "num_tokens": 26132692.0,
      "step": 14170
    },
    {
      "entropy": 5.775873327255249,
      "epoch": 1.190884268010922,
      "grad_norm": 1.125,
      "learning_rate": 0.0004863062905173999,
      "loss": 5.3931,
      "mean_token_accuracy": 0.17803334295749665,
      "num_tokens": 26142259.0,
      "step": 14175
    },
    {
      "entropy": 5.69611177444458,
      "epoch": 1.191304347826087,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000486296000892122,
      "loss": 5.2437,
      "mean_token_accuracy": 0.18153987526893617,
      "num_tokens": 26151782.0,
      "step": 14180
    },
    {
      "entropy": 5.609896326065064,
      "epoch": 1.1917244276412517,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048628570752384424,
      "loss": 5.1012,
      "mean_token_accuracy": 0.19075109362602233,
      "num_tokens": 26160449.0,
      "step": 14185
    },
    {
      "entropy": 5.726301431655884,
      "epoch": 1.1921445074564168,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048627541041274897,
      "loss": 5.3464,
      "mean_token_accuracy": 0.17585661858320237,
      "num_tokens": 26169764.0,
      "step": 14190
    },
    {
      "entropy": 5.679019927978516,
      "epoch": 1.1925645872715815,
      "grad_norm": 1.0,
      "learning_rate": 0.00048626510955901854,
      "loss": 5.1934,
      "mean_token_accuracy": 0.1849956139922142,
      "num_tokens": 26178759.0,
      "step": 14195
    },
    {
      "entropy": 5.724093294143676,
      "epoch": 1.1929846670867466,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004862548049628356,
      "loss": 5.3041,
      "mean_token_accuracy": 0.18281467258930206,
      "num_tokens": 26187904.0,
      "step": 14200
    },
    {
      "entropy": 5.719123506546021,
      "epoch": 1.1934047469019113,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004862444966243824,
      "loss": 5.245,
      "mean_token_accuracy": 0.1862848535180092,
      "num_tokens": 26196563.0,
      "step": 14205
    },
    {
      "entropy": 5.773376846313477,
      "epoch": 1.1938248267170763,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004862341845438419,
      "loss": 5.2477,
      "mean_token_accuracy": 0.17970482259988785,
      "num_tokens": 26206573.0,
      "step": 14210
    },
    {
      "entropy": 5.635081720352173,
      "epoch": 1.1942449065322411,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048622386872139645,
      "loss": 5.1803,
      "mean_token_accuracy": 0.1860246941447258,
      "num_tokens": 26215308.0,
      "step": 14215
    },
    {
      "entropy": 5.605758142471314,
      "epoch": 1.194664986347406,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000486213549157229,
      "loss": 5.2474,
      "mean_token_accuracy": 0.18344548344612122,
      "num_tokens": 26224379.0,
      "step": 14220
    },
    {
      "entropy": 5.638043308258057,
      "epoch": 1.195085066162571,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004862032258515222,
      "loss": 5.2098,
      "mean_token_accuracy": 0.18943456560373306,
      "num_tokens": 26233620.0,
      "step": 14225
    },
    {
      "entropy": 5.718277788162231,
      "epoch": 1.1955051459777357,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004861928988044592,
      "loss": 5.2956,
      "mean_token_accuracy": 0.17665328234434127,
      "num_tokens": 26242556.0,
      "step": 14230
    },
    {
      "entropy": 5.721251630783081,
      "epoch": 1.1959252257929007,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004861825680162226,
      "loss": 5.2652,
      "mean_token_accuracy": 0.18234376907348632,
      "num_tokens": 26251561.0,
      "step": 14235
    },
    {
      "entropy": 5.630056858062744,
      "epoch": 1.1963453056080655,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048617223348699546,
      "loss": 5.1909,
      "mean_token_accuracy": 0.18268014043569564,
      "num_tokens": 26261115.0,
      "step": 14240
    },
    {
      "entropy": 5.701509952545166,
      "epoch": 1.1967653854232303,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004861618952169611,
      "loss": 5.3258,
      "mean_token_accuracy": 0.18601730167865754,
      "num_tokens": 26271165.0,
      "step": 14245
    },
    {
      "entropy": 5.66995096206665,
      "epoch": 1.1971854652383953,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004861515532063025,
      "loss": 5.3119,
      "mean_token_accuracy": 0.17776518911123276,
      "num_tokens": 26280822.0,
      "step": 14250
    },
    {
      "entropy": 5.681661701202392,
      "epoch": 1.19760554505356,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048614120745520275,
      "loss": 5.2114,
      "mean_token_accuracy": 0.18382159918546676,
      "num_tokens": 26288747.0,
      "step": 14255
    },
    {
      "entropy": 5.710450839996338,
      "epoch": 1.198025624868725,
      "grad_norm": 1.0,
      "learning_rate": 0.00048613085796384524,
      "loss": 5.2568,
      "mean_token_accuracy": 0.17893887907266617,
      "num_tokens": 26298387.0,
      "step": 14260
    },
    {
      "entropy": 5.608849430084229,
      "epoch": 1.19844570468389,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048612050473241335,
      "loss": 5.1552,
      "mean_token_accuracy": 0.18917208015918732,
      "num_tokens": 26307016.0,
      "step": 14265
    },
    {
      "entropy": 5.667242765426636,
      "epoch": 1.198865784499055,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004861101477610905,
      "loss": 5.2737,
      "mean_token_accuracy": 0.1792106047272682,
      "num_tokens": 26316296.0,
      "step": 14270
    },
    {
      "entropy": 5.667968463897705,
      "epoch": 1.1992858643142197,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048609978705006,
      "loss": 5.274,
      "mean_token_accuracy": 0.180621962249279,
      "num_tokens": 26325525.0,
      "step": 14275
    },
    {
      "entropy": 5.633566522598267,
      "epoch": 1.1997059441293847,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004860894225995055,
      "loss": 5.1516,
      "mean_token_accuracy": 0.19144971072673797,
      "num_tokens": 26334195.0,
      "step": 14280
    },
    {
      "entropy": 5.64380464553833,
      "epoch": 1.2001260239445495,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048607905440961054,
      "loss": 5.2741,
      "mean_token_accuracy": 0.17881516963243485,
      "num_tokens": 26343933.0,
      "step": 14285
    },
    {
      "entropy": 5.677962398529052,
      "epoch": 1.2005461037597143,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048606868248055887,
      "loss": 5.2191,
      "mean_token_accuracy": 0.18560283333063127,
      "num_tokens": 26353455.0,
      "step": 14290
    },
    {
      "entropy": 5.750855302810669,
      "epoch": 1.2009661835748793,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004860583068125341,
      "loss": 5.2236,
      "mean_token_accuracy": 0.18569577038288115,
      "num_tokens": 26362662.0,
      "step": 14295
    },
    {
      "entropy": 5.67245717048645,
      "epoch": 1.201386263390044,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004860479274057202,
      "loss": 5.2395,
      "mean_token_accuracy": 0.17771499902009963,
      "num_tokens": 26371536.0,
      "step": 14300
    },
    {
      "entropy": 5.697594881057739,
      "epoch": 1.201806343205209,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048603754426030087,
      "loss": 5.3058,
      "mean_token_accuracy": 0.18444887101650237,
      "num_tokens": 26381925.0,
      "step": 14305
    },
    {
      "entropy": 5.673886919021607,
      "epoch": 1.2022264230203739,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048602715737646016,
      "loss": 5.1891,
      "mean_token_accuracy": 0.18497975915670395,
      "num_tokens": 26391111.0,
      "step": 14310
    },
    {
      "entropy": 5.74705924987793,
      "epoch": 1.2026465028355386,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048601676675438197,
      "loss": 5.326,
      "mean_token_accuracy": 0.17124558985233307,
      "num_tokens": 26401667.0,
      "step": 14315
    },
    {
      "entropy": 5.652735424041748,
      "epoch": 1.2030665826507037,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048600637239425045,
      "loss": 5.1608,
      "mean_token_accuracy": 0.1903965011239052,
      "num_tokens": 26411261.0,
      "step": 14320
    },
    {
      "entropy": 5.688894748687744,
      "epoch": 1.2034866624658684,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048599597429624966,
      "loss": 5.32,
      "mean_token_accuracy": 0.17646912783384322,
      "num_tokens": 26419808.0,
      "step": 14325
    },
    {
      "entropy": 5.65993013381958,
      "epoch": 1.2039067422810334,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048598557246056385,
      "loss": 5.2346,
      "mean_token_accuracy": 0.1829381287097931,
      "num_tokens": 26429160.0,
      "step": 14330
    },
    {
      "entropy": 5.667569780349732,
      "epoch": 1.2043268220961982,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048597516688737727,
      "loss": 5.193,
      "mean_token_accuracy": 0.18618611842393876,
      "num_tokens": 26437675.0,
      "step": 14335
    },
    {
      "entropy": 5.698027420043945,
      "epoch": 1.2047469019113632,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048596475757687425,
      "loss": 5.2555,
      "mean_token_accuracy": 0.18135085701942444,
      "num_tokens": 26446317.0,
      "step": 14340
    },
    {
      "entropy": 5.687911653518677,
      "epoch": 1.205166981726528,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048595434452923915,
      "loss": 5.28,
      "mean_token_accuracy": 0.1906139075756073,
      "num_tokens": 26456183.0,
      "step": 14345
    },
    {
      "entropy": 5.669294261932373,
      "epoch": 1.205587061541693,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048594392774465656,
      "loss": 5.2462,
      "mean_token_accuracy": 0.17616432160139084,
      "num_tokens": 26466324.0,
      "step": 14350
    },
    {
      "entropy": 5.678545331954956,
      "epoch": 1.2060071413568578,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048593350722331074,
      "loss": 5.2414,
      "mean_token_accuracy": 0.18592250496149063,
      "num_tokens": 26475560.0,
      "step": 14355
    },
    {
      "entropy": 5.625226831436157,
      "epoch": 1.2064272211720226,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048592308296538654,
      "loss": 5.2077,
      "mean_token_accuracy": 0.18210819065570832,
      "num_tokens": 26484955.0,
      "step": 14360
    },
    {
      "entropy": 5.676719903945923,
      "epoch": 1.2068473009871876,
      "grad_norm": 1.0,
      "learning_rate": 0.0004859126549710686,
      "loss": 5.2,
      "mean_token_accuracy": 0.18677111119031906,
      "num_tokens": 26494306.0,
      "step": 14365
    },
    {
      "entropy": 5.667259693145752,
      "epoch": 1.2072673808023524,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048590222324054153,
      "loss": 5.1841,
      "mean_token_accuracy": 0.1903291955590248,
      "num_tokens": 26503871.0,
      "step": 14370
    },
    {
      "entropy": 5.750825977325439,
      "epoch": 1.2076874606175174,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004858917877739901,
      "loss": 5.3067,
      "mean_token_accuracy": 0.17933351993560792,
      "num_tokens": 26511929.0,
      "step": 14375
    },
    {
      "entropy": 5.676896047592163,
      "epoch": 1.2081075404326822,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004858813485715994,
      "loss": 5.3025,
      "mean_token_accuracy": 0.1749362140893936,
      "num_tokens": 26520469.0,
      "step": 14380
    },
    {
      "entropy": 5.679920434951782,
      "epoch": 1.208527620247847,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004858709056335541,
      "loss": 5.2505,
      "mean_token_accuracy": 0.18464099019765853,
      "num_tokens": 26530102.0,
      "step": 14385
    },
    {
      "entropy": 5.675498914718628,
      "epoch": 1.208947700063012,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048586045896003926,
      "loss": 5.2535,
      "mean_token_accuracy": 0.18120122104883193,
      "num_tokens": 26538705.0,
      "step": 14390
    },
    {
      "entropy": 5.712258815765381,
      "epoch": 1.2093677798781768,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004858500085512401,
      "loss": 5.3282,
      "mean_token_accuracy": 0.1784636303782463,
      "num_tokens": 26548315.0,
      "step": 14395
    },
    {
      "entropy": 5.6935735702514645,
      "epoch": 1.2097878596933418,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048583955440734144,
      "loss": 5.1845,
      "mean_token_accuracy": 0.18441999405622483,
      "num_tokens": 26556412.0,
      "step": 14400
    },
    {
      "entropy": 5.709979104995727,
      "epoch": 1.2102079395085066,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048582909652852873,
      "loss": 5.3304,
      "mean_token_accuracy": 0.17204978168010712,
      "num_tokens": 26566146.0,
      "step": 14405
    },
    {
      "entropy": 5.633830451965332,
      "epoch": 1.2106280193236716,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004858186349149871,
      "loss": 5.2319,
      "mean_token_accuracy": 0.18733145594596862,
      "num_tokens": 26576019.0,
      "step": 14410
    },
    {
      "entropy": 5.593112564086914,
      "epoch": 1.2110480991388364,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000485808169566902,
      "loss": 5.1044,
      "mean_token_accuracy": 0.1886659324169159,
      "num_tokens": 26585461.0,
      "step": 14415
    },
    {
      "entropy": 5.605729818344116,
      "epoch": 1.2114681789540014,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048579770048445863,
      "loss": 5.1434,
      "mean_token_accuracy": 0.20051885545253753,
      "num_tokens": 26594021.0,
      "step": 14420
    },
    {
      "entropy": 5.739030361175537,
      "epoch": 1.2118882587691662,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048578722766784253,
      "loss": 5.2896,
      "mean_token_accuracy": 0.1818574398756027,
      "num_tokens": 26602712.0,
      "step": 14425
    },
    {
      "entropy": 5.63052077293396,
      "epoch": 1.212308338584331,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048577675111723925,
      "loss": 5.0019,
      "mean_token_accuracy": 0.19522817879915239,
      "num_tokens": 26610970.0,
      "step": 14430
    },
    {
      "entropy": 5.665298795700073,
      "epoch": 1.212728418399496,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048576627083283435,
      "loss": 5.2619,
      "mean_token_accuracy": 0.18590751439332961,
      "num_tokens": 26619840.0,
      "step": 14435
    },
    {
      "entropy": 5.6351042747497555,
      "epoch": 1.2131484982146608,
      "grad_norm": 1.0,
      "learning_rate": 0.0004857557868148136,
      "loss": 5.1563,
      "mean_token_accuracy": 0.1908823013305664,
      "num_tokens": 26629271.0,
      "step": 14440
    },
    {
      "entropy": 5.631720352172851,
      "epoch": 1.2135685780298258,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004857452990633625,
      "loss": 5.2113,
      "mean_token_accuracy": 0.18432663530111312,
      "num_tokens": 26638610.0,
      "step": 14445
    },
    {
      "entropy": 5.769435119628906,
      "epoch": 1.2139886578449905,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048573480757866695,
      "loss": 5.336,
      "mean_token_accuracy": 0.18053234964609147,
      "num_tokens": 26648504.0,
      "step": 14450
    },
    {
      "entropy": 5.691094207763672,
      "epoch": 1.2144087376601553,
      "grad_norm": 1.0,
      "learning_rate": 0.00048572431236091284,
      "loss": 5.2447,
      "mean_token_accuracy": 0.18206262737512588,
      "num_tokens": 26658084.0,
      "step": 14455
    },
    {
      "entropy": 5.657041692733765,
      "epoch": 1.2148288174753203,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048571381341028604,
      "loss": 5.3184,
      "mean_token_accuracy": 0.17985238134860992,
      "num_tokens": 26666933.0,
      "step": 14460
    },
    {
      "entropy": 5.713371515274048,
      "epoch": 1.2152488972904851,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004857033107269725,
      "loss": 5.2123,
      "mean_token_accuracy": 0.18746238201856613,
      "num_tokens": 26675049.0,
      "step": 14465
    },
    {
      "entropy": 5.623864555358887,
      "epoch": 1.2156689771056501,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048569280431115823,
      "loss": 5.2565,
      "mean_token_accuracy": 0.1853037014603615,
      "num_tokens": 26684223.0,
      "step": 14470
    },
    {
      "entropy": 5.723515033721924,
      "epoch": 1.216089056920815,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004856822941630296,
      "loss": 5.222,
      "mean_token_accuracy": 0.18003584444522858,
      "num_tokens": 26693605.0,
      "step": 14475
    },
    {
      "entropy": 5.728838014602661,
      "epoch": 1.2165091367359797,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048567178028277255,
      "loss": 5.285,
      "mean_token_accuracy": 0.18274309933185579,
      "num_tokens": 26702829.0,
      "step": 14480
    },
    {
      "entropy": 5.719204998016357,
      "epoch": 1.2169292165511447,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004856612626705733,
      "loss": 5.3283,
      "mean_token_accuracy": 0.17774795442819596,
      "num_tokens": 26712466.0,
      "step": 14485
    },
    {
      "entropy": 5.670398378372193,
      "epoch": 1.2173492963663095,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004856507413266183,
      "loss": 5.1988,
      "mean_token_accuracy": 0.19019820243120195,
      "num_tokens": 26721730.0,
      "step": 14490
    },
    {
      "entropy": 5.653546905517578,
      "epoch": 1.2177693761814745,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000485640216251094,
      "loss": 5.2999,
      "mean_token_accuracy": 0.17735347151756287,
      "num_tokens": 26731017.0,
      "step": 14495
    },
    {
      "entropy": 5.680216884613037,
      "epoch": 1.2181894559966393,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048562968744418665,
      "loss": 5.2568,
      "mean_token_accuracy": 0.18216293007135392,
      "num_tokens": 26739588.0,
      "step": 14500
    },
    {
      "entropy": 5.748578119277954,
      "epoch": 1.2186095358118043,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004856191549060828,
      "loss": 5.3407,
      "mean_token_accuracy": 0.1774814710021019,
      "num_tokens": 26748889.0,
      "step": 14505
    },
    {
      "entropy": 5.756321287155151,
      "epoch": 1.219029615626969,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048560861863696913,
      "loss": 5.3011,
      "mean_token_accuracy": 0.18165266662836074,
      "num_tokens": 26757979.0,
      "step": 14510
    },
    {
      "entropy": 5.679617214202881,
      "epoch": 1.219449695442134,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004855980786370322,
      "loss": 5.2356,
      "mean_token_accuracy": 0.1798142358660698,
      "num_tokens": 26767225.0,
      "step": 14515
    },
    {
      "entropy": 5.641861009597778,
      "epoch": 1.219869775257299,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004855875349064588,
      "loss": 5.166,
      "mean_token_accuracy": 0.18415713161230088,
      "num_tokens": 26776289.0,
      "step": 14520
    },
    {
      "entropy": 5.73698468208313,
      "epoch": 1.2202898550724637,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004855769874454356,
      "loss": 5.288,
      "mean_token_accuracy": 0.18133005648851394,
      "num_tokens": 26785631.0,
      "step": 14525
    },
    {
      "entropy": 5.7060293674469,
      "epoch": 1.2207099348876287,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004855664362541495,
      "loss": 5.2811,
      "mean_token_accuracy": 0.17822755873203278,
      "num_tokens": 26795285.0,
      "step": 14530
    },
    {
      "entropy": 5.627279233932495,
      "epoch": 1.2211300147027935,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048555588133278744,
      "loss": 5.2236,
      "mean_token_accuracy": 0.17522367238998413,
      "num_tokens": 26804584.0,
      "step": 14535
    },
    {
      "entropy": 5.585305547714233,
      "epoch": 1.2215500945179585,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004855453226815363,
      "loss": 5.1125,
      "mean_token_accuracy": 0.1881616547703743,
      "num_tokens": 26814354.0,
      "step": 14540
    },
    {
      "entropy": 5.573113489151001,
      "epoch": 1.2219701743331233,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048553476030058326,
      "loss": 5.1315,
      "mean_token_accuracy": 0.18794713616371156,
      "num_tokens": 26824274.0,
      "step": 14545
    },
    {
      "entropy": 5.596048069000244,
      "epoch": 1.222390254148288,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048552419419011536,
      "loss": 5.2589,
      "mean_token_accuracy": 0.18113338947296143,
      "num_tokens": 26833155.0,
      "step": 14550
    },
    {
      "entropy": 5.65484652519226,
      "epoch": 1.222810333963453,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004855136243503196,
      "loss": 5.2046,
      "mean_token_accuracy": 0.18324665278196334,
      "num_tokens": 26842545.0,
      "step": 14555
    },
    {
      "entropy": 5.723557090759277,
      "epoch": 1.2232304137786179,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048550305078138363,
      "loss": 5.2716,
      "mean_token_accuracy": 0.18153285086154938,
      "num_tokens": 26851772.0,
      "step": 14560
    },
    {
      "entropy": 5.633531188964843,
      "epoch": 1.2236504935937829,
      "grad_norm": 1.0,
      "learning_rate": 0.00048549247348349435,
      "loss": 5.1593,
      "mean_token_accuracy": 0.19334136247634887,
      "num_tokens": 26860884.0,
      "step": 14565
    },
    {
      "entropy": 5.6667492389678955,
      "epoch": 1.2240705734089476,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048548189245683934,
      "loss": 5.2812,
      "mean_token_accuracy": 0.18448901772499085,
      "num_tokens": 26869435.0,
      "step": 14570
    },
    {
      "entropy": 5.648065423965454,
      "epoch": 1.2244906532241127,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048547130770160596,
      "loss": 5.1796,
      "mean_token_accuracy": 0.18526097238063813,
      "num_tokens": 26878852.0,
      "step": 14575
    },
    {
      "entropy": 5.6628560543060305,
      "epoch": 1.2249107330392774,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004854607192179817,
      "loss": 5.16,
      "mean_token_accuracy": 0.18675457537174225,
      "num_tokens": 26887532.0,
      "step": 14580
    },
    {
      "entropy": 5.795173597335816,
      "epoch": 1.2253308128544425,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004854501270061543,
      "loss": 5.383,
      "mean_token_accuracy": 0.17961090952157974,
      "num_tokens": 26897459.0,
      "step": 14585
    },
    {
      "entropy": 5.634214258193969,
      "epoch": 1.2257508926696072,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048543953106631115,
      "loss": 5.1442,
      "mean_token_accuracy": 0.1891936853528023,
      "num_tokens": 26907156.0,
      "step": 14590
    },
    {
      "entropy": 5.716067981719971,
      "epoch": 1.226170972484772,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004854289313986401,
      "loss": 5.2553,
      "mean_token_accuracy": 0.17971518486738206,
      "num_tokens": 26915764.0,
      "step": 14595
    },
    {
      "entropy": 5.579642915725708,
      "epoch": 1.226591052299937,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004854183280033289,
      "loss": 5.1239,
      "mean_token_accuracy": 0.18705009371042253,
      "num_tokens": 26924166.0,
      "step": 14600
    },
    {
      "entropy": 5.67133526802063,
      "epoch": 1.2270111321151018,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004854077208805654,
      "loss": 5.3355,
      "mean_token_accuracy": 0.1774318367242813,
      "num_tokens": 26933546.0,
      "step": 14605
    },
    {
      "entropy": 5.702149438858032,
      "epoch": 1.2274312119302668,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004853971100305374,
      "loss": 5.2767,
      "mean_token_accuracy": 0.18572707921266557,
      "num_tokens": 26943213.0,
      "step": 14610
    },
    {
      "entropy": 5.715191268920899,
      "epoch": 1.2278512917454316,
      "grad_norm": 0.9375,
      "learning_rate": 0.000485386495453433,
      "loss": 5.2423,
      "mean_token_accuracy": 0.1881014436483383,
      "num_tokens": 26952968.0,
      "step": 14615
    },
    {
      "entropy": 5.651744318008423,
      "epoch": 1.2282713715605964,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048537587714944007,
      "loss": 5.2077,
      "mean_token_accuracy": 0.18450172245502472,
      "num_tokens": 26962230.0,
      "step": 14620
    },
    {
      "entropy": 5.61191873550415,
      "epoch": 1.2286914513757614,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004853652551187469,
      "loss": 5.2808,
      "mean_token_accuracy": 0.1884658619761467,
      "num_tokens": 26970985.0,
      "step": 14625
    },
    {
      "entropy": 5.675854349136353,
      "epoch": 1.2291115311909262,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048535462936154147,
      "loss": 5.289,
      "mean_token_accuracy": 0.18243741989135742,
      "num_tokens": 26981138.0,
      "step": 14630
    },
    {
      "entropy": 5.64106855392456,
      "epoch": 1.2295316110060912,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004853439998780122,
      "loss": 5.1701,
      "mean_token_accuracy": 0.1906170889735222,
      "num_tokens": 26990158.0,
      "step": 14635
    },
    {
      "entropy": 5.6716265201568605,
      "epoch": 1.229951690821256,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004853333666683472,
      "loss": 5.3004,
      "mean_token_accuracy": 0.17534388601779938,
      "num_tokens": 26998889.0,
      "step": 14640
    },
    {
      "entropy": 5.671046924591065,
      "epoch": 1.230371770636421,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048532272973273496,
      "loss": 5.2284,
      "mean_token_accuracy": 0.18123542964458467,
      "num_tokens": 27008912.0,
      "step": 14645
    },
    {
      "entropy": 5.64111819267273,
      "epoch": 1.2307918504515858,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048531208907136384,
      "loss": 5.1293,
      "mean_token_accuracy": 0.19974811375141144,
      "num_tokens": 27017573.0,
      "step": 14650
    },
    {
      "entropy": 5.629122018814087,
      "epoch": 1.2312119302667508,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048530144468442236,
      "loss": 5.2132,
      "mean_token_accuracy": 0.18198008239269256,
      "num_tokens": 27027205.0,
      "step": 14655
    },
    {
      "entropy": 5.633598756790161,
      "epoch": 1.2316320100819156,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048529079657209906,
      "loss": 5.1625,
      "mean_token_accuracy": 0.18555975556373597,
      "num_tokens": 27035882.0,
      "step": 14660
    },
    {
      "entropy": 5.607516860961914,
      "epoch": 1.2320520898970804,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004852801447345826,
      "loss": 5.2481,
      "mean_token_accuracy": 0.184847691655159,
      "num_tokens": 27044761.0,
      "step": 14665
    },
    {
      "entropy": 5.6852864742279055,
      "epoch": 1.2324721697122454,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004852694891720617,
      "loss": 5.2644,
      "mean_token_accuracy": 0.178885018825531,
      "num_tokens": 27054149.0,
      "step": 14670
    },
    {
      "entropy": 5.7014556407928465,
      "epoch": 1.2328922495274102,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000485258829884725,
      "loss": 5.297,
      "mean_token_accuracy": 0.18527219146490098,
      "num_tokens": 27063145.0,
      "step": 14675
    },
    {
      "entropy": 5.735172414779663,
      "epoch": 1.2333123293425752,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004852481668727614,
      "loss": 5.2677,
      "mean_token_accuracy": 0.1818074628710747,
      "num_tokens": 27072378.0,
      "step": 14680
    },
    {
      "entropy": 5.571155595779419,
      "epoch": 1.23373240915774,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048523750013635986,
      "loss": 5.1374,
      "mean_token_accuracy": 0.1853396102786064,
      "num_tokens": 27082241.0,
      "step": 14685
    },
    {
      "entropy": 5.576324701309204,
      "epoch": 1.2341524889729047,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004852268296757092,
      "loss": 5.1547,
      "mean_token_accuracy": 0.1842060700058937,
      "num_tokens": 27091488.0,
      "step": 14690
    },
    {
      "entropy": 5.721409463882447,
      "epoch": 1.2345725687880698,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004852161554909985,
      "loss": 5.2053,
      "mean_token_accuracy": 0.1878787100315094,
      "num_tokens": 27100378.0,
      "step": 14695
    },
    {
      "entropy": 5.69066104888916,
      "epoch": 1.2349926486032345,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048520547758241686,
      "loss": 5.2299,
      "mean_token_accuracy": 0.1835736706852913,
      "num_tokens": 27110341.0,
      "step": 14700
    },
    {
      "entropy": 5.674981307983399,
      "epoch": 1.2354127284183996,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048519479595015343,
      "loss": 5.1898,
      "mean_token_accuracy": 0.18298552185297012,
      "num_tokens": 27119381.0,
      "step": 14705
    },
    {
      "entropy": 5.631501388549805,
      "epoch": 1.2358328082335643,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048518411059439746,
      "loss": 5.2638,
      "mean_token_accuracy": 0.17899573892354964,
      "num_tokens": 27129167.0,
      "step": 14710
    },
    {
      "entropy": 5.69050555229187,
      "epoch": 1.2362528880487293,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048517342151533813,
      "loss": 5.2858,
      "mean_token_accuracy": 0.17756334096193313,
      "num_tokens": 27138479.0,
      "step": 14715
    },
    {
      "entropy": 5.680474185943604,
      "epoch": 1.2366729678638941,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004851627287131649,
      "loss": 5.1534,
      "mean_token_accuracy": 0.18793304413557052,
      "num_tokens": 27147197.0,
      "step": 14720
    },
    {
      "entropy": 5.628408765792846,
      "epoch": 1.2370930476790591,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004851520321880672,
      "loss": 5.1889,
      "mean_token_accuracy": 0.1886487066745758,
      "num_tokens": 27155854.0,
      "step": 14725
    },
    {
      "entropy": 5.643164825439453,
      "epoch": 1.237513127494224,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004851413319402344,
      "loss": 5.1559,
      "mean_token_accuracy": 0.18393366634845734,
      "num_tokens": 27165069.0,
      "step": 14730
    },
    {
      "entropy": 5.647834968566895,
      "epoch": 1.2379332073093887,
      "grad_norm": 1.125,
      "learning_rate": 0.0004851306279698561,
      "loss": 5.2078,
      "mean_token_accuracy": 0.1795213133096695,
      "num_tokens": 27174070.0,
      "step": 14735
    },
    {
      "entropy": 5.716778659820557,
      "epoch": 1.2383532871245537,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004851199202771219,
      "loss": 5.2757,
      "mean_token_accuracy": 0.1870725393295288,
      "num_tokens": 27182903.0,
      "step": 14740
    },
    {
      "entropy": 5.6445080757141115,
      "epoch": 1.2387733669397185,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004851092088622216,
      "loss": 5.1791,
      "mean_token_accuracy": 0.18956800401210785,
      "num_tokens": 27192747.0,
      "step": 14745
    },
    {
      "entropy": 5.697391223907471,
      "epoch": 1.2391934467548835,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004850984937253448,
      "loss": 5.2202,
      "mean_token_accuracy": 0.18787728548049926,
      "num_tokens": 27201657.0,
      "step": 14750
    },
    {
      "entropy": 5.6604596138000485,
      "epoch": 1.2396135265700483,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004850877748666814,
      "loss": 5.2313,
      "mean_token_accuracy": 0.18318491727113723,
      "num_tokens": 27211794.0,
      "step": 14755
    },
    {
      "entropy": 5.620717144012451,
      "epoch": 1.240033606385213,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048507705228642117,
      "loss": 5.1895,
      "mean_token_accuracy": 0.17854712903499603,
      "num_tokens": 27221852.0,
      "step": 14760
    },
    {
      "entropy": 5.638713407516479,
      "epoch": 1.240453686200378,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004850663259847542,
      "loss": 5.2474,
      "mean_token_accuracy": 0.18233464211225509,
      "num_tokens": 27231558.0,
      "step": 14765
    },
    {
      "entropy": 5.666490268707276,
      "epoch": 1.240873766015543,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048505559596187037,
      "loss": 5.2328,
      "mean_token_accuracy": 0.18314465880393982,
      "num_tokens": 27241053.0,
      "step": 14770
    },
    {
      "entropy": 5.623038387298584,
      "epoch": 1.241293845830708,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004850448622179599,
      "loss": 5.1102,
      "mean_token_accuracy": 0.18834349811077117,
      "num_tokens": 27249770.0,
      "step": 14775
    },
    {
      "entropy": 5.7206199169158936,
      "epoch": 1.2417139256458727,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004850341247532128,
      "loss": 5.3603,
      "mean_token_accuracy": 0.18124401420354844,
      "num_tokens": 27258883.0,
      "step": 14780
    },
    {
      "entropy": 5.712884950637817,
      "epoch": 1.2421340054610377,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004850233835678194,
      "loss": 5.2552,
      "mean_token_accuracy": 0.18546027988195418,
      "num_tokens": 27268056.0,
      "step": 14785
    },
    {
      "entropy": 5.67972412109375,
      "epoch": 1.2425540852762025,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004850126386619699,
      "loss": 5.1487,
      "mean_token_accuracy": 0.1934316039085388,
      "num_tokens": 27276965.0,
      "step": 14790
    },
    {
      "entropy": 5.646904802322387,
      "epoch": 1.2429741650913673,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004850018900358545,
      "loss": 5.2001,
      "mean_token_accuracy": 0.19088197350502015,
      "num_tokens": 27286173.0,
      "step": 14795
    },
    {
      "entropy": 5.655182933807373,
      "epoch": 1.2433942449065323,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048499113768966386,
      "loss": 5.1939,
      "mean_token_accuracy": 0.1882569596171379,
      "num_tokens": 27294863.0,
      "step": 14800
    },
    {
      "entropy": 5.703532981872558,
      "epoch": 1.243814324721697,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004849803816235884,
      "loss": 5.2495,
      "mean_token_accuracy": 0.18228195905685424,
      "num_tokens": 27304427.0,
      "step": 14805
    },
    {
      "entropy": 5.740639162063599,
      "epoch": 1.244234404536862,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004849696218378185,
      "loss": 5.2918,
      "mean_token_accuracy": 0.1824310004711151,
      "num_tokens": 27313716.0,
      "step": 14810
    },
    {
      "entropy": 5.721193599700928,
      "epoch": 1.2446544843520269,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004849588583325449,
      "loss": 5.1978,
      "mean_token_accuracy": 0.19424594044685364,
      "num_tokens": 27322342.0,
      "step": 14815
    },
    {
      "entropy": 5.7138207912445065,
      "epoch": 1.2450745641671919,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004849480911079583,
      "loss": 5.2847,
      "mean_token_accuracy": 0.16878207325935363,
      "num_tokens": 27331892.0,
      "step": 14820
    },
    {
      "entropy": 5.687543296813965,
      "epoch": 1.2454946439823567,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004849373201642493,
      "loss": 5.2535,
      "mean_token_accuracy": 0.17534445226192474,
      "num_tokens": 27340428.0,
      "step": 14825
    },
    {
      "entropy": 5.64434232711792,
      "epoch": 1.2459147237975214,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004849265455016088,
      "loss": 5.2422,
      "mean_token_accuracy": 0.182173652946949,
      "num_tokens": 27349224.0,
      "step": 14830
    },
    {
      "entropy": 5.700295782089233,
      "epoch": 1.2463348036126864,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004849157671202277,
      "loss": 5.2357,
      "mean_token_accuracy": 0.18692852407693863,
      "num_tokens": 27357480.0,
      "step": 14835
    },
    {
      "entropy": 5.6686258792877195,
      "epoch": 1.2467548834278512,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004849049850202968,
      "loss": 5.1439,
      "mean_token_accuracy": 0.18896788209676743,
      "num_tokens": 27366732.0,
      "step": 14840
    },
    {
      "entropy": 5.665991258621216,
      "epoch": 1.2471749632430162,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004848941992020072,
      "loss": 5.2564,
      "mean_token_accuracy": 0.18047528117895126,
      "num_tokens": 27375834.0,
      "step": 14845
    },
    {
      "entropy": 5.722081804275513,
      "epoch": 1.247595043058181,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004848834096655499,
      "loss": 5.2476,
      "mean_token_accuracy": 0.18529427349567412,
      "num_tokens": 27385311.0,
      "step": 14850
    },
    {
      "entropy": 5.678635931015014,
      "epoch": 1.2480151228733458,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048487261641111607,
      "loss": 5.2876,
      "mean_token_accuracy": 0.1805171936750412,
      "num_tokens": 27394587.0,
      "step": 14855
    },
    {
      "entropy": 5.558681201934815,
      "epoch": 1.2484352026885108,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000484861819438897,
      "loss": 5.1557,
      "mean_token_accuracy": 0.18218595534563065,
      "num_tokens": 27403316.0,
      "step": 14860
    },
    {
      "entropy": 5.6447343826293945,
      "epoch": 1.2488552825036756,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004848510187490838,
      "loss": 5.2123,
      "mean_token_accuracy": 0.19038472026586534,
      "num_tokens": 27412709.0,
      "step": 14865
    },
    {
      "entropy": 5.709834432601928,
      "epoch": 1.2492753623188406,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004848402143418679,
      "loss": 5.2666,
      "mean_token_accuracy": 0.17966692745685578,
      "num_tokens": 27422004.0,
      "step": 14870
    },
    {
      "entropy": 5.663084840774536,
      "epoch": 1.2496954421340054,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048482940621744053,
      "loss": 5.2805,
      "mean_token_accuracy": 0.1832091048359871,
      "num_tokens": 27431931.0,
      "step": 14875
    },
    {
      "entropy": 5.617640638351441,
      "epoch": 1.2501155219491704,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004848185943759934,
      "loss": 5.1284,
      "mean_token_accuracy": 0.19388483166694642,
      "num_tokens": 27441527.0,
      "step": 14880
    },
    {
      "entropy": 5.730137777328491,
      "epoch": 1.2505356017643352,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048480777881771786,
      "loss": 5.27,
      "mean_token_accuracy": 0.19130437523126603,
      "num_tokens": 27449964.0,
      "step": 14885
    },
    {
      "entropy": 5.6717531204223635,
      "epoch": 1.2509556815795002,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004847969595428056,
      "loss": 5.2659,
      "mean_token_accuracy": 0.17987626791000366,
      "num_tokens": 27459044.0,
      "step": 14890
    },
    {
      "entropy": 5.582613611221314,
      "epoch": 1.251375761394665,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048478613655144817,
      "loss": 5.2613,
      "mean_token_accuracy": 0.18128221184015275,
      "num_tokens": 27467644.0,
      "step": 14895
    },
    {
      "entropy": 5.706760835647583,
      "epoch": 1.2517958412098298,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004847753098438374,
      "loss": 5.2834,
      "mean_token_accuracy": 0.1772915691137314,
      "num_tokens": 27476899.0,
      "step": 14900
    },
    {
      "entropy": 5.700392007827759,
      "epoch": 1.2522159210249948,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000484764479420165,
      "loss": 5.1836,
      "mean_token_accuracy": 0.18831791877746581,
      "num_tokens": 27485167.0,
      "step": 14905
    },
    {
      "entropy": 5.708838176727295,
      "epoch": 1.2526360008401596,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048475364528062287,
      "loss": 5.2114,
      "mean_token_accuracy": 0.17930906116962433,
      "num_tokens": 27493986.0,
      "step": 14910
    },
    {
      "entropy": 5.704959869384766,
      "epoch": 1.2530560806553246,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004847428074254029,
      "loss": 5.252,
      "mean_token_accuracy": 0.19165757447481155,
      "num_tokens": 27503896.0,
      "step": 14915
    },
    {
      "entropy": 5.65868444442749,
      "epoch": 1.2534761604704894,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048473196585469713,
      "loss": 5.2191,
      "mean_token_accuracy": 0.18710336536169053,
      "num_tokens": 27513485.0,
      "step": 14920
    },
    {
      "entropy": 5.713653230667115,
      "epoch": 1.2538962402856542,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048472112056869763,
      "loss": 5.2735,
      "mean_token_accuracy": 0.1760294795036316,
      "num_tokens": 27523164.0,
      "step": 14925
    },
    {
      "entropy": 5.726631689071655,
      "epoch": 1.2543163201008192,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004847102715675964,
      "loss": 5.2274,
      "mean_token_accuracy": 0.1807127609848976,
      "num_tokens": 27531387.0,
      "step": 14930
    },
    {
      "entropy": 5.624379014968872,
      "epoch": 1.254736399915984,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004846994188515857,
      "loss": 5.2168,
      "mean_token_accuracy": 0.1880618840456009,
      "num_tokens": 27541754.0,
      "step": 14935
    },
    {
      "entropy": 5.735077381134033,
      "epoch": 1.255156479731149,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004846885624208578,
      "loss": 5.3011,
      "mean_token_accuracy": 0.17986671179533004,
      "num_tokens": 27551458.0,
      "step": 14940
    },
    {
      "entropy": 5.680077791213989,
      "epoch": 1.2555765595463138,
      "grad_norm": 1.015625,
      "learning_rate": 0.000484677702275605,
      "loss": 5.21,
      "mean_token_accuracy": 0.18616114258766175,
      "num_tokens": 27560797.0,
      "step": 14945
    },
    {
      "entropy": 5.712385368347168,
      "epoch": 1.2559966393614788,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048466683841601963,
      "loss": 5.2121,
      "mean_token_accuracy": 0.1836182802915573,
      "num_tokens": 27570166.0,
      "step": 14950
    },
    {
      "entropy": 5.601553106307984,
      "epoch": 1.2564167191766435,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048465597084229416,
      "loss": 5.1128,
      "mean_token_accuracy": 0.18998402655124663,
      "num_tokens": 27579411.0,
      "step": 14955
    },
    {
      "entropy": 5.702762079238892,
      "epoch": 1.2568367989918086,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004846450995546212,
      "loss": 5.341,
      "mean_token_accuracy": 0.1797543078660965,
      "num_tokens": 27589124.0,
      "step": 14960
    },
    {
      "entropy": 5.711671638488769,
      "epoch": 1.2572568788069733,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004846342245531932,
      "loss": 5.324,
      "mean_token_accuracy": 0.17847648411989211,
      "num_tokens": 27598664.0,
      "step": 14965
    },
    {
      "entropy": 5.750071287155151,
      "epoch": 1.2576769586221381,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004846233458382029,
      "loss": 5.2455,
      "mean_token_accuracy": 0.18522500693798066,
      "num_tokens": 27607189.0,
      "step": 14970
    },
    {
      "entropy": 5.722580146789551,
      "epoch": 1.2580970384373031,
      "grad_norm": 1.0,
      "learning_rate": 0.00048461246340984293,
      "loss": 5.2799,
      "mean_token_accuracy": 0.18317657709121704,
      "num_tokens": 27616415.0,
      "step": 14975
    },
    {
      "entropy": 5.687801313400269,
      "epoch": 1.258517118252468,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004846015772683061,
      "loss": 5.2541,
      "mean_token_accuracy": 0.1864374503493309,
      "num_tokens": 27624492.0,
      "step": 14980
    },
    {
      "entropy": 5.575812482833863,
      "epoch": 1.258937198067633,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048459068741378526,
      "loss": 5.161,
      "mean_token_accuracy": 0.18146611750125885,
      "num_tokens": 27634243.0,
      "step": 14985
    },
    {
      "entropy": 5.680055236816406,
      "epoch": 1.2593572778827977,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004845797938464734,
      "loss": 5.2519,
      "mean_token_accuracy": 0.1848156377673149,
      "num_tokens": 27642887.0,
      "step": 14990
    },
    {
      "entropy": 5.742097806930542,
      "epoch": 1.2597773576979625,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004845688965665633,
      "loss": 5.2625,
      "mean_token_accuracy": 0.18163831233978273,
      "num_tokens": 27652524.0,
      "step": 14995
    },
    {
      "entropy": 5.660221672058105,
      "epoch": 1.2601974375131275,
      "grad_norm": 1.125,
      "learning_rate": 0.00048455799557424814,
      "loss": 5.1232,
      "mean_token_accuracy": 0.19624098092317582,
      "num_tokens": 27661306.0,
      "step": 15000
    },
    {
      "epoch": 1.2601974375131275,
      "eval_entropy": 5.480510376219932,
      "eval_loss": 5.287484169006348,
      "eval_mean_token_accuracy": 0.19071965733213259,
      "eval_num_tokens": 27661306.0,
      "eval_runtime": 20.8899,
      "eval_samples_per_second": 1788.707,
      "eval_steps_per_second": 223.6,
      "step": 15000
    },
    {
      "entropy": 5.682873106002807,
      "epoch": 1.2606175173282923,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004845470908697209,
      "loss": 5.3226,
      "mean_token_accuracy": 0.18663743138313293,
      "num_tokens": 27671728.0,
      "step": 15005
    },
    {
      "entropy": 5.6125036716461185,
      "epoch": 1.2610375971434573,
      "grad_norm": 0.921875,
      "learning_rate": 0.000484536182453175,
      "loss": 5.1145,
      "mean_token_accuracy": 0.19209557622671128,
      "num_tokens": 27680740.0,
      "step": 15010
    },
    {
      "entropy": 5.689800357818603,
      "epoch": 1.261457676958622,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004845252703248035,
      "loss": 5.203,
      "mean_token_accuracy": 0.18818854093551635,
      "num_tokens": 27689865.0,
      "step": 15015
    },
    {
      "entropy": 5.671578598022461,
      "epoch": 1.2618777567737869,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004845143544847997,
      "loss": 5.2302,
      "mean_token_accuracy": 0.18501729518175125,
      "num_tokens": 27700366.0,
      "step": 15020
    },
    {
      "entropy": 5.664003229141235,
      "epoch": 1.262297836588952,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048450343493335697,
      "loss": 5.1327,
      "mean_token_accuracy": 0.1915966585278511,
      "num_tokens": 27708893.0,
      "step": 15025
    },
    {
      "entropy": 5.62802267074585,
      "epoch": 1.262717916404117,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004844925116706688,
      "loss": 5.1721,
      "mean_token_accuracy": 0.18221084475517274,
      "num_tokens": 27717494.0,
      "step": 15030
    },
    {
      "entropy": 5.563887357711792,
      "epoch": 1.2631379962192817,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048448158469692866,
      "loss": 5.0965,
      "mean_token_accuracy": 0.19990942925214766,
      "num_tokens": 27726487.0,
      "step": 15035
    },
    {
      "entropy": 5.742855215072632,
      "epoch": 1.2635580760344465,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004844706540123301,
      "loss": 5.3116,
      "mean_token_accuracy": 0.175772725045681,
      "num_tokens": 27736602.0,
      "step": 15040
    },
    {
      "entropy": 5.757776308059692,
      "epoch": 1.2639781558496115,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048445971961706675,
      "loss": 5.3028,
      "mean_token_accuracy": 0.18110672384500504,
      "num_tokens": 27746322.0,
      "step": 15045
    },
    {
      "entropy": 5.637805652618408,
      "epoch": 1.2643982356647763,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004844487815113323,
      "loss": 5.1841,
      "mean_token_accuracy": 0.1905662328004837,
      "num_tokens": 27754941.0,
      "step": 15050
    },
    {
      "entropy": 5.644527959823608,
      "epoch": 1.2648183154799413,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004844378396953206,
      "loss": 5.2305,
      "mean_token_accuracy": 0.18545543849468232,
      "num_tokens": 27763941.0,
      "step": 15055
    },
    {
      "entropy": 5.703114748001099,
      "epoch": 1.265238395295106,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048442689416922536,
      "loss": 5.2638,
      "mean_token_accuracy": 0.18711112588644027,
      "num_tokens": 27773087.0,
      "step": 15060
    },
    {
      "entropy": 5.588543272018432,
      "epoch": 1.2656584751102709,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048441594493324057,
      "loss": 5.0897,
      "mean_token_accuracy": 0.19230256080627442,
      "num_tokens": 27782648.0,
      "step": 15065
    },
    {
      "entropy": 5.651916742324829,
      "epoch": 1.2660785549254359,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048440499198756015,
      "loss": 5.2961,
      "mean_token_accuracy": 0.17800186276435853,
      "num_tokens": 27791567.0,
      "step": 15070
    },
    {
      "entropy": 5.674059867858887,
      "epoch": 1.2664986347406006,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048439403533237816,
      "loss": 5.2848,
      "mean_token_accuracy": 0.17919526845216752,
      "num_tokens": 27801397.0,
      "step": 15075
    },
    {
      "entropy": 5.756672143936157,
      "epoch": 1.2669187145557657,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004843830749678886,
      "loss": 5.2758,
      "mean_token_accuracy": 0.1832936331629753,
      "num_tokens": 27810831.0,
      "step": 15080
    },
    {
      "entropy": 5.6129858016967775,
      "epoch": 1.2673387943709304,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004843721108942856,
      "loss": 5.1911,
      "mean_token_accuracy": 0.1895334839820862,
      "num_tokens": 27819591.0,
      "step": 15085
    },
    {
      "entropy": 5.606224489212036,
      "epoch": 1.2677588741860952,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004843611431117636,
      "loss": 5.2026,
      "mean_token_accuracy": 0.19222567677497865,
      "num_tokens": 27828614.0,
      "step": 15090
    },
    {
      "entropy": 5.67724928855896,
      "epoch": 1.2681789540012602,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004843501716205167,
      "loss": 5.2364,
      "mean_token_accuracy": 0.1809759482741356,
      "num_tokens": 27837549.0,
      "step": 15095
    },
    {
      "entropy": 5.699390983581543,
      "epoch": 1.2685990338164252,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004843391964207393,
      "loss": 5.2315,
      "mean_token_accuracy": 0.18013533651828767,
      "num_tokens": 27846678.0,
      "step": 15100
    },
    {
      "entropy": 5.741962528228759,
      "epoch": 1.26901911363159,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004843282175126258,
      "loss": 5.2545,
      "mean_token_accuracy": 0.18150699585676194,
      "num_tokens": 27855734.0,
      "step": 15105
    },
    {
      "entropy": 5.665146589279175,
      "epoch": 1.2694391934467548,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048431723489637086,
      "loss": 5.1884,
      "mean_token_accuracy": 0.18633033633232116,
      "num_tokens": 27865111.0,
      "step": 15110
    },
    {
      "entropy": 5.678395843505859,
      "epoch": 1.2698592732619198,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048430624857216876,
      "loss": 5.2165,
      "mean_token_accuracy": 0.18687239587306975,
      "num_tokens": 27874495.0,
      "step": 15115
    },
    {
      "entropy": 5.639739227294922,
      "epoch": 1.2702793530770846,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004842952585402143,
      "loss": 5.2427,
      "mean_token_accuracy": 0.18856698721647264,
      "num_tokens": 27884531.0,
      "step": 15120
    },
    {
      "entropy": 5.618935203552246,
      "epoch": 1.2706994328922496,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000484284264800702,
      "loss": 5.1293,
      "mean_token_accuracy": 0.1902562662959099,
      "num_tokens": 27893463.0,
      "step": 15125
    },
    {
      "entropy": 5.720178413391113,
      "epoch": 1.2711195127074144,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048427326735382687,
      "loss": 5.2521,
      "mean_token_accuracy": 0.17922103255987168,
      "num_tokens": 27903015.0,
      "step": 15130
    },
    {
      "entropy": 5.684845972061157,
      "epoch": 1.2715395925225792,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004842622661997834,
      "loss": 5.2257,
      "mean_token_accuracy": 0.1872514918446541,
      "num_tokens": 27912207.0,
      "step": 15135
    },
    {
      "entropy": 5.697420120239258,
      "epoch": 1.2719596723377442,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004842512613387668,
      "loss": 5.2307,
      "mean_token_accuracy": 0.17940394431352616,
      "num_tokens": 27921566.0,
      "step": 15140
    },
    {
      "entropy": 5.663734292984008,
      "epoch": 1.272379752152909,
      "grad_norm": 1.0,
      "learning_rate": 0.0004842402527709718,
      "loss": 5.1778,
      "mean_token_accuracy": 0.18691326230764388,
      "num_tokens": 27930633.0,
      "step": 15145
    },
    {
      "entropy": 5.697439527511596,
      "epoch": 1.272799831968074,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004842292404965934,
      "loss": 5.2874,
      "mean_token_accuracy": 0.18131062388420105,
      "num_tokens": 27939887.0,
      "step": 15150
    },
    {
      "entropy": 5.717975711822509,
      "epoch": 1.2732199117832388,
      "grad_norm": 1.125,
      "learning_rate": 0.0004842182245158268,
      "loss": 5.2949,
      "mean_token_accuracy": 0.18785111010074615,
      "num_tokens": 27949090.0,
      "step": 15155
    },
    {
      "entropy": 5.6314410209655765,
      "epoch": 1.2736399915984036,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048420720482886715,
      "loss": 5.1208,
      "mean_token_accuracy": 0.19463839530944824,
      "num_tokens": 27958141.0,
      "step": 15160
    },
    {
      "entropy": 5.6016974449157715,
      "epoch": 1.2740600714135686,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004841961814359095,
      "loss": 5.1747,
      "mean_token_accuracy": 0.18878703713417053,
      "num_tokens": 27967780.0,
      "step": 15165
    },
    {
      "entropy": 5.711613321304322,
      "epoch": 1.2744801512287336,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048418515433714917,
      "loss": 5.2728,
      "mean_token_accuracy": 0.1843690827488899,
      "num_tokens": 27976243.0,
      "step": 15170
    },
    {
      "entropy": 5.649885654449463,
      "epoch": 1.2749002310438984,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004841741235327817,
      "loss": 5.1476,
      "mean_token_accuracy": 0.19189854413270951,
      "num_tokens": 27985874.0,
      "step": 15175
    },
    {
      "entropy": 5.767988395690918,
      "epoch": 1.2753203108590632,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048416308902300215,
      "loss": 5.3622,
      "mean_token_accuracy": 0.17386006265878678,
      "num_tokens": 27995111.0,
      "step": 15180
    },
    {
      "entropy": 5.663547468185425,
      "epoch": 1.2757403906742282,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004841520508080063,
      "loss": 5.1953,
      "mean_token_accuracy": 0.1885661691427231,
      "num_tokens": 28003948.0,
      "step": 15185
    },
    {
      "entropy": 5.655944442749023,
      "epoch": 1.276160470489393,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048414100888798957,
      "loss": 5.2028,
      "mean_token_accuracy": 0.18404938280582428,
      "num_tokens": 28012941.0,
      "step": 15190
    },
    {
      "entropy": 5.610059642791748,
      "epoch": 1.276580550304558,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004841299632631475,
      "loss": 5.1973,
      "mean_token_accuracy": 0.1828886941075325,
      "num_tokens": 28022195.0,
      "step": 15195
    },
    {
      "entropy": 5.65628228187561,
      "epoch": 1.2770006301197228,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004841189139336759,
      "loss": 5.1384,
      "mean_token_accuracy": 0.19217560738325118,
      "num_tokens": 28031446.0,
      "step": 15200
    },
    {
      "entropy": 5.68044228553772,
      "epoch": 1.2774207099348875,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004841078608997703,
      "loss": 5.1516,
      "mean_token_accuracy": 0.1946975737810135,
      "num_tokens": 28040906.0,
      "step": 15205
    },
    {
      "entropy": 5.653388595581054,
      "epoch": 1.2778407897500526,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004840968041616267,
      "loss": 5.1613,
      "mean_token_accuracy": 0.18839018642902375,
      "num_tokens": 28049848.0,
      "step": 15210
    },
    {
      "entropy": 5.659143257141113,
      "epoch": 1.2782608695652173,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048408574371944094,
      "loss": 5.1698,
      "mean_token_accuracy": 0.1873503729701042,
      "num_tokens": 28058276.0,
      "step": 15215
    },
    {
      "entropy": 5.684235143661499,
      "epoch": 1.2786809493803823,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004840746795734088,
      "loss": 5.2603,
      "mean_token_accuracy": 0.185902838408947,
      "num_tokens": 28068185.0,
      "step": 15220
    },
    {
      "entropy": 5.7272453784942625,
      "epoch": 1.2791010291955471,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004840636117237264,
      "loss": 5.3182,
      "mean_token_accuracy": 0.18122888803482057,
      "num_tokens": 28077532.0,
      "step": 15225
    },
    {
      "entropy": 5.609941482543945,
      "epoch": 1.279521109010712,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004840525401705897,
      "loss": 5.1743,
      "mean_token_accuracy": 0.18322077840566636,
      "num_tokens": 28087593.0,
      "step": 15230
    },
    {
      "entropy": 5.646874237060547,
      "epoch": 1.279941188825877,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048404146491419503,
      "loss": 5.1497,
      "mean_token_accuracy": 0.19319449067115785,
      "num_tokens": 28096256.0,
      "step": 15235
    },
    {
      "entropy": 5.705491209030152,
      "epoch": 1.2803612686410417,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048403038595473837,
      "loss": 5.1882,
      "mean_token_accuracy": 0.18828644454479218,
      "num_tokens": 28105048.0,
      "step": 15240
    },
    {
      "entropy": 5.669730281829834,
      "epoch": 1.2807813484562067,
      "grad_norm": 1.015625,
      "learning_rate": 0.000484019303292416,
      "loss": 5.251,
      "mean_token_accuracy": 0.17236898988485336,
      "num_tokens": 28114330.0,
      "step": 15245
    },
    {
      "entropy": 5.639583587646484,
      "epoch": 1.2812014282713715,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048400821692742434,
      "loss": 5.1817,
      "mean_token_accuracy": 0.19096322357654572,
      "num_tokens": 28123147.0,
      "step": 15250
    },
    {
      "entropy": 5.70355453491211,
      "epoch": 1.2816215080865365,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048399712685995983,
      "loss": 5.2985,
      "mean_token_accuracy": 0.18468752205371858,
      "num_tokens": 28132477.0,
      "step": 15255
    },
    {
      "entropy": 5.677632284164429,
      "epoch": 1.2820415879017013,
      "grad_norm": 1.125,
      "learning_rate": 0.00048398603309021877,
      "loss": 5.289,
      "mean_token_accuracy": 0.17882544100284575,
      "num_tokens": 28141350.0,
      "step": 15260
    },
    {
      "entropy": 5.723903512954712,
      "epoch": 1.2824616677168663,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004839749356183978,
      "loss": 5.2398,
      "mean_token_accuracy": 0.18105954974889754,
      "num_tokens": 28149522.0,
      "step": 15265
    },
    {
      "entropy": 5.698184061050415,
      "epoch": 1.282881747532031,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004839638344446933,
      "loss": 5.3148,
      "mean_token_accuracy": 0.1846858873963356,
      "num_tokens": 28159646.0,
      "step": 15270
    },
    {
      "entropy": 5.722110557556152,
      "epoch": 1.283301827347196,
      "grad_norm": 1.0,
      "learning_rate": 0.0004839527295693023,
      "loss": 5.219,
      "mean_token_accuracy": 0.18618374317884445,
      "num_tokens": 28168408.0,
      "step": 15275
    },
    {
      "entropy": 5.692686939239502,
      "epoch": 1.283721907162361,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004839416209924211,
      "loss": 5.2513,
      "mean_token_accuracy": 0.17623980045318605,
      "num_tokens": 28177744.0,
      "step": 15280
    },
    {
      "entropy": 5.725609254837036,
      "epoch": 1.2841419869775257,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048393050871424676,
      "loss": 5.2926,
      "mean_token_accuracy": 0.18197290003299713,
      "num_tokens": 28186811.0,
      "step": 15285
    },
    {
      "entropy": 5.6954731941223145,
      "epoch": 1.2845620667926907,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000483919392734976,
      "loss": 5.2694,
      "mean_token_accuracy": 0.18221310824155806,
      "num_tokens": 28197052.0,
      "step": 15290
    },
    {
      "entropy": 5.649689245223999,
      "epoch": 1.2849821466078555,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004839082730548058,
      "loss": 5.1348,
      "mean_token_accuracy": 0.19537210911512376,
      "num_tokens": 28206000.0,
      "step": 15295
    },
    {
      "entropy": 5.621656656265259,
      "epoch": 1.2854022264230203,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004838971496739331,
      "loss": 5.1124,
      "mean_token_accuracy": 0.18375589400529863,
      "num_tokens": 28214679.0,
      "step": 15300
    },
    {
      "entropy": 5.589047002792358,
      "epoch": 1.2858223062381853,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000483886022592555,
      "loss": 5.2356,
      "mean_token_accuracy": 0.18311792463064194,
      "num_tokens": 28223890.0,
      "step": 15305
    },
    {
      "entropy": 5.629820728302002,
      "epoch": 1.28624238605335,
      "grad_norm": 1.0,
      "learning_rate": 0.0004838748918108685,
      "loss": 5.1778,
      "mean_token_accuracy": 0.18884078711271285,
      "num_tokens": 28232422.0,
      "step": 15310
    },
    {
      "entropy": 5.704855155944824,
      "epoch": 1.286662465868515,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048386375732907083,
      "loss": 5.2105,
      "mean_token_accuracy": 0.18572533130645752,
      "num_tokens": 28242079.0,
      "step": 15315
    },
    {
      "entropy": 5.771668434143066,
      "epoch": 1.2870825456836799,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048385261914735936,
      "loss": 5.3854,
      "mean_token_accuracy": 0.17527088671922683,
      "num_tokens": 28252510.0,
      "step": 15320
    },
    {
      "entropy": 5.740034103393555,
      "epoch": 1.2875026254988446,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048384147726593125,
      "loss": 5.3147,
      "mean_token_accuracy": 0.1816914215683937,
      "num_tokens": 28261348.0,
      "step": 15325
    },
    {
      "entropy": 5.7191802024841305,
      "epoch": 1.2879227053140097,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004838303316849839,
      "loss": 5.2106,
      "mean_token_accuracy": 0.1815898060798645,
      "num_tokens": 28270739.0,
      "step": 15330
    },
    {
      "entropy": 5.701415634155273,
      "epoch": 1.2883427851291747,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048381918240471473,
      "loss": 5.2603,
      "mean_token_accuracy": 0.17692821174860002,
      "num_tokens": 28279370.0,
      "step": 15335
    },
    {
      "entropy": 5.707439947128296,
      "epoch": 1.2887628649443394,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048380802942532124,
      "loss": 5.1914,
      "mean_token_accuracy": 0.18212359696626662,
      "num_tokens": 28287955.0,
      "step": 15340
    },
    {
      "entropy": 5.5799088954925535,
      "epoch": 1.2891829447595042,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048379687274700107,
      "loss": 5.1519,
      "mean_token_accuracy": 0.19197224080562592,
      "num_tokens": 28296832.0,
      "step": 15345
    },
    {
      "entropy": 5.615931415557862,
      "epoch": 1.2896030245746692,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048378571236995185,
      "loss": 5.1754,
      "mean_token_accuracy": 0.1838840901851654,
      "num_tokens": 28305778.0,
      "step": 15350
    },
    {
      "entropy": 5.674046230316162,
      "epoch": 1.290023104389834,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048377454829437124,
      "loss": 5.2378,
      "mean_token_accuracy": 0.18029564321041108,
      "num_tokens": 28314615.0,
      "step": 15355
    },
    {
      "entropy": 5.735815954208374,
      "epoch": 1.290443184204999,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004837633805204569,
      "loss": 5.2955,
      "mean_token_accuracy": 0.1824686661362648,
      "num_tokens": 28324478.0,
      "step": 15360
    },
    {
      "entropy": 5.733227872848511,
      "epoch": 1.2908632640201638,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004837522090484069,
      "loss": 5.2389,
      "mean_token_accuracy": 0.1820377603173256,
      "num_tokens": 28333532.0,
      "step": 15365
    },
    {
      "entropy": 5.679190063476563,
      "epoch": 1.2912833438353286,
      "grad_norm": 1.0,
      "learning_rate": 0.00048374103387841894,
      "loss": 5.2162,
      "mean_token_accuracy": 0.1791183039546013,
      "num_tokens": 28343723.0,
      "step": 15370
    },
    {
      "entropy": 5.6929620742797855,
      "epoch": 1.2917034236504936,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048372985501069106,
      "loss": 5.2103,
      "mean_token_accuracy": 0.18511388450860977,
      "num_tokens": 28351992.0,
      "step": 15375
    },
    {
      "entropy": 5.6732401847839355,
      "epoch": 1.2921235034656584,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004837186724454213,
      "loss": 5.1917,
      "mean_token_accuracy": 0.19064346849918365,
      "num_tokens": 28361141.0,
      "step": 15380
    },
    {
      "entropy": 5.632470369338989,
      "epoch": 1.2925435832808234,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004837074861828077,
      "loss": 5.1824,
      "mean_token_accuracy": 0.1903940409421921,
      "num_tokens": 28370339.0,
      "step": 15385
    },
    {
      "entropy": 5.7076191902160645,
      "epoch": 1.2929636630959882,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004836962962230485,
      "loss": 5.3001,
      "mean_token_accuracy": 0.17998291701078414,
      "num_tokens": 28379242.0,
      "step": 15390
    },
    {
      "entropy": 5.648181629180908,
      "epoch": 1.293383742911153,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004836851025663418,
      "loss": 5.1751,
      "mean_token_accuracy": 0.19270177632570268,
      "num_tokens": 28388864.0,
      "step": 15395
    },
    {
      "entropy": 5.7533660411834715,
      "epoch": 1.293803822726318,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000483673905212886,
      "loss": 5.2742,
      "mean_token_accuracy": 0.1859456941485405,
      "num_tokens": 28398000.0,
      "step": 15400
    },
    {
      "entropy": 5.576316547393799,
      "epoch": 1.294223902541483,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004836627041628794,
      "loss": 5.2216,
      "mean_token_accuracy": 0.18475604355335234,
      "num_tokens": 28407652.0,
      "step": 15405
    },
    {
      "entropy": 5.7184394836425785,
      "epoch": 1.2946439823566478,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004836514994165205,
      "loss": 5.2689,
      "mean_token_accuracy": 0.1789836183190346,
      "num_tokens": 28417694.0,
      "step": 15410
    },
    {
      "entropy": 5.728481292724609,
      "epoch": 1.2950640621718126,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048364029097400777,
      "loss": 5.2168,
      "mean_token_accuracy": 0.18802534639835358,
      "num_tokens": 28426928.0,
      "step": 15415
    },
    {
      "entropy": 5.677914047241211,
      "epoch": 1.2954841419869776,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048362907883553956,
      "loss": 5.2484,
      "mean_token_accuracy": 0.17647462338209152,
      "num_tokens": 28436176.0,
      "step": 15420
    },
    {
      "entropy": 5.710226535797119,
      "epoch": 1.2959042218021424,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048361786300131477,
      "loss": 5.3396,
      "mean_token_accuracy": 0.17654576599597932,
      "num_tokens": 28445277.0,
      "step": 15425
    },
    {
      "entropy": 5.72326512336731,
      "epoch": 1.2963243016173074,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004836066434715319,
      "loss": 5.2087,
      "mean_token_accuracy": 0.18456294983625413,
      "num_tokens": 28453959.0,
      "step": 15430
    },
    {
      "entropy": 5.6885395526885985,
      "epoch": 1.2967443814324722,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004835954202463898,
      "loss": 5.2925,
      "mean_token_accuracy": 0.18077042698860168,
      "num_tokens": 28463780.0,
      "step": 15435
    },
    {
      "entropy": 5.6299995422363285,
      "epoch": 1.297164461247637,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004835841933260872,
      "loss": 5.167,
      "mean_token_accuracy": 0.18045762479305266,
      "num_tokens": 28473299.0,
      "step": 15440
    },
    {
      "entropy": 5.641190481185913,
      "epoch": 1.297584541062802,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048357296271082305,
      "loss": 5.1984,
      "mean_token_accuracy": 0.18333597034215926,
      "num_tokens": 28481859.0,
      "step": 15445
    },
    {
      "entropy": 5.7216263771057125,
      "epoch": 1.2980046208779668,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048356172840079625,
      "loss": 5.2464,
      "mean_token_accuracy": 0.1827081874012947,
      "num_tokens": 28491034.0,
      "step": 15450
    },
    {
      "entropy": 5.7014641761779785,
      "epoch": 1.2984247006931318,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004835504903962058,
      "loss": 5.1857,
      "mean_token_accuracy": 0.18175535798072814,
      "num_tokens": 28499829.0,
      "step": 15455
    },
    {
      "entropy": 5.644228124618531,
      "epoch": 1.2988447805082965,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048353924869725084,
      "loss": 5.1741,
      "mean_token_accuracy": 0.19424838572740555,
      "num_tokens": 28508188.0,
      "step": 15460
    },
    {
      "entropy": 5.604212713241577,
      "epoch": 1.2992648603234613,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004835280033041305,
      "loss": 5.0892,
      "mean_token_accuracy": 0.19110470563173293,
      "num_tokens": 28516509.0,
      "step": 15465
    },
    {
      "entropy": 5.617927169799804,
      "epoch": 1.2996849401386263,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004835167542170439,
      "loss": 5.2981,
      "mean_token_accuracy": 0.18407307267189027,
      "num_tokens": 28526457.0,
      "step": 15470
    },
    {
      "entropy": 5.63437967300415,
      "epoch": 1.3001050199537914,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004835055014361904,
      "loss": 5.2132,
      "mean_token_accuracy": 0.18324276357889174,
      "num_tokens": 28536149.0,
      "step": 15475
    },
    {
      "entropy": 5.743523502349854,
      "epoch": 1.3005250997689561,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048349424496176924,
      "loss": 5.2711,
      "mean_token_accuracy": 0.17861638963222504,
      "num_tokens": 28545486.0,
      "step": 15480
    },
    {
      "entropy": 5.735276889801026,
      "epoch": 1.300945179584121,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048348298479397996,
      "loss": 5.1713,
      "mean_token_accuracy": 0.1864497110247612,
      "num_tokens": 28554555.0,
      "step": 15485
    },
    {
      "entropy": 5.6189697265625,
      "epoch": 1.301365259399286,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048347172093302196,
      "loss": 5.203,
      "mean_token_accuracy": 0.19588252902030945,
      "num_tokens": 28563387.0,
      "step": 15490
    },
    {
      "entropy": 5.650023031234741,
      "epoch": 1.3017853392144507,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048346045337909475,
      "loss": 5.1947,
      "mean_token_accuracy": 0.1863280192017555,
      "num_tokens": 28573437.0,
      "step": 15495
    },
    {
      "entropy": 5.602910995483398,
      "epoch": 1.3022054190296157,
      "grad_norm": 0.984375,
      "learning_rate": 0.000483449182132398,
      "loss": 5.1407,
      "mean_token_accuracy": 0.19586918950080873,
      "num_tokens": 28583362.0,
      "step": 15500
    },
    {
      "entropy": 5.7878899574279785,
      "epoch": 1.3026254988447805,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048343790719313124,
      "loss": 5.3217,
      "mean_token_accuracy": 0.17867469489574433,
      "num_tokens": 28593201.0,
      "step": 15505
    },
    {
      "entropy": 5.678332710266114,
      "epoch": 1.3030455786599453,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048342662856149427,
      "loss": 5.2475,
      "mean_token_accuracy": 0.17691996395587922,
      "num_tokens": 28602486.0,
      "step": 15510
    },
    {
      "entropy": 5.61362476348877,
      "epoch": 1.3034656584751103,
      "grad_norm": 1.046875,
      "learning_rate": 0.000483415346237687,
      "loss": 5.2504,
      "mean_token_accuracy": 0.1779778480529785,
      "num_tokens": 28611643.0,
      "step": 15515
    },
    {
      "entropy": 5.720222854614258,
      "epoch": 1.303885738290275,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004834040602219091,
      "loss": 5.2747,
      "mean_token_accuracy": 0.1848591074347496,
      "num_tokens": 28620545.0,
      "step": 15520
    },
    {
      "entropy": 5.66220235824585,
      "epoch": 1.30430581810544,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048339277051436067,
      "loss": 5.2306,
      "mean_token_accuracy": 0.18619415611028672,
      "num_tokens": 28630024.0,
      "step": 15525
    },
    {
      "entropy": 5.751428318023682,
      "epoch": 1.304725897920605,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004833814771152415,
      "loss": 5.2876,
      "mean_token_accuracy": 0.18809086382389067,
      "num_tokens": 28638995.0,
      "step": 15530
    },
    {
      "entropy": 5.697282791137695,
      "epoch": 1.3051459777357697,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048337018002475184,
      "loss": 5.2442,
      "mean_token_accuracy": 0.18519910275936127,
      "num_tokens": 28647833.0,
      "step": 15535
    },
    {
      "entropy": 5.625005197525025,
      "epoch": 1.3055660575509347,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004833588792430917,
      "loss": 5.1166,
      "mean_token_accuracy": 0.19010624140501023,
      "num_tokens": 28657441.0,
      "step": 15540
    },
    {
      "entropy": 5.658119821548462,
      "epoch": 1.3059861373660997,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004833475747704614,
      "loss": 5.2432,
      "mean_token_accuracy": 0.1856137290596962,
      "num_tokens": 28666666.0,
      "step": 15545
    },
    {
      "entropy": 5.677402925491333,
      "epoch": 1.3064062171812645,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000483336266607061,
      "loss": 5.2401,
      "mean_token_accuracy": 0.18404926657676696,
      "num_tokens": 28676770.0,
      "step": 15550
    },
    {
      "entropy": 5.678611469268799,
      "epoch": 1.3068262969964293,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048332495475309097,
      "loss": 5.1599,
      "mean_token_accuracy": 0.18388599306344985,
      "num_tokens": 28685610.0,
      "step": 15555
    },
    {
      "entropy": 5.6823639392852785,
      "epoch": 1.3072463768115943,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048331363920875155,
      "loss": 5.2706,
      "mean_token_accuracy": 0.18541425764560698,
      "num_tokens": 28695082.0,
      "step": 15560
    },
    {
      "entropy": 5.649327278137207,
      "epoch": 1.307666456626759,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048330231997424335,
      "loss": 5.2106,
      "mean_token_accuracy": 0.1831612080335617,
      "num_tokens": 28704006.0,
      "step": 15565
    },
    {
      "entropy": 5.693876314163208,
      "epoch": 1.308086536441924,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004832909970497668,
      "loss": 5.2094,
      "mean_token_accuracy": 0.18668109327554702,
      "num_tokens": 28713665.0,
      "step": 15570
    },
    {
      "entropy": 5.6811493873596195,
      "epoch": 1.3085066162570889,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048327967043552245,
      "loss": 5.1856,
      "mean_token_accuracy": 0.18585191369056703,
      "num_tokens": 28722920.0,
      "step": 15575
    },
    {
      "entropy": 5.631540441513062,
      "epoch": 1.3089266960722536,
      "grad_norm": 1.0,
      "learning_rate": 0.00048326834013171107,
      "loss": 5.1255,
      "mean_token_accuracy": 0.19284412115812302,
      "num_tokens": 28731689.0,
      "step": 15580
    },
    {
      "entropy": 5.620257806777954,
      "epoch": 1.3093467758874187,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004832570061385332,
      "loss": 5.2243,
      "mean_token_accuracy": 0.19059069603681564,
      "num_tokens": 28741308.0,
      "step": 15585
    },
    {
      "entropy": 5.647295379638672,
      "epoch": 1.3097668557025834,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004832456684561898,
      "loss": 5.2191,
      "mean_token_accuracy": 0.18846675753593445,
      "num_tokens": 28750190.0,
      "step": 15590
    },
    {
      "entropy": 5.6955914974212645,
      "epoch": 1.3101869355177485,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004832343270848815,
      "loss": 5.2846,
      "mean_token_accuracy": 0.1793016329407692,
      "num_tokens": 28759588.0,
      "step": 15595
    },
    {
      "entropy": 5.68325366973877,
      "epoch": 1.3106070153329132,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048322298202480935,
      "loss": 5.2821,
      "mean_token_accuracy": 0.18034479320049285,
      "num_tokens": 28768800.0,
      "step": 15600
    },
    {
      "entropy": 5.70915470123291,
      "epoch": 1.311027095148078,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048321163327617433,
      "loss": 5.2236,
      "mean_token_accuracy": 0.18359237164258957,
      "num_tokens": 28778108.0,
      "step": 15605
    },
    {
      "entropy": 5.714198160171509,
      "epoch": 1.311447174963243,
      "grad_norm": 1.0,
      "learning_rate": 0.0004832002808391775,
      "loss": 5.1904,
      "mean_token_accuracy": 0.18482305705547333,
      "num_tokens": 28787202.0,
      "step": 15610
    },
    {
      "entropy": 5.6883039474487305,
      "epoch": 1.3118672547784078,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004831889247140198,
      "loss": 5.2196,
      "mean_token_accuracy": 0.1867629885673523,
      "num_tokens": 28797482.0,
      "step": 15615
    },
    {
      "entropy": 5.621526765823364,
      "epoch": 1.3122873345935728,
      "grad_norm": 1.0,
      "learning_rate": 0.00048317756490090253,
      "loss": 5.1623,
      "mean_token_accuracy": 0.1883018672466278,
      "num_tokens": 28805872.0,
      "step": 15620
    },
    {
      "entropy": 5.636060905456543,
      "epoch": 1.3127074144087376,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048316620140002685,
      "loss": 5.2712,
      "mean_token_accuracy": 0.17940105348825455,
      "num_tokens": 28814836.0,
      "step": 15625
    },
    {
      "entropy": 5.74907374382019,
      "epoch": 1.3131274942239024,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004831548342115942,
      "loss": 5.2735,
      "mean_token_accuracy": 0.17501763254404068,
      "num_tokens": 28824727.0,
      "step": 15630
    },
    {
      "entropy": 5.772703027725219,
      "epoch": 1.3135475740390674,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048314346333580576,
      "loss": 5.3672,
      "mean_token_accuracy": 0.1810245618224144,
      "num_tokens": 28833848.0,
      "step": 15635
    },
    {
      "entropy": 5.659271287918091,
      "epoch": 1.3139676538542324,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004831320887728631,
      "loss": 5.1325,
      "mean_token_accuracy": 0.19095175564289094,
      "num_tokens": 28842198.0,
      "step": 15640
    },
    {
      "entropy": 5.703675413131714,
      "epoch": 1.3143877336693972,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004831207105229676,
      "loss": 5.2217,
      "mean_token_accuracy": 0.18384626656770706,
      "num_tokens": 28851804.0,
      "step": 15645
    },
    {
      "entropy": 5.605397081375122,
      "epoch": 1.314807813484562,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048310932858632087,
      "loss": 5.1351,
      "mean_token_accuracy": 0.18884730488061904,
      "num_tokens": 28860181.0,
      "step": 15650
    },
    {
      "entropy": 5.6268641471862795,
      "epoch": 1.315227893299727,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048309794296312467,
      "loss": 5.1935,
      "mean_token_accuracy": 0.1949946716427803,
      "num_tokens": 28869945.0,
      "step": 15655
    },
    {
      "entropy": 5.652247285842895,
      "epoch": 1.3156479731148918,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048308655365358053,
      "loss": 5.2552,
      "mean_token_accuracy": 0.1884856104850769,
      "num_tokens": 28880343.0,
      "step": 15660
    },
    {
      "entropy": 5.822339344024658,
      "epoch": 1.3160680529300568,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048307516065789017,
      "loss": 5.3205,
      "mean_token_accuracy": 0.1764918491244316,
      "num_tokens": 28889441.0,
      "step": 15665
    },
    {
      "entropy": 5.750012636184692,
      "epoch": 1.3164881327452216,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048306376397625546,
      "loss": 5.28,
      "mean_token_accuracy": 0.1805360808968544,
      "num_tokens": 28898154.0,
      "step": 15670
    },
    {
      "entropy": 5.706697082519531,
      "epoch": 1.3169082125603864,
      "grad_norm": 1.0,
      "learning_rate": 0.00048305236360887834,
      "loss": 5.2454,
      "mean_token_accuracy": 0.18432848900556564,
      "num_tokens": 28908359.0,
      "step": 15675
    },
    {
      "entropy": 5.636640024185181,
      "epoch": 1.3173282923755514,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048304095955596074,
      "loss": 5.2442,
      "mean_token_accuracy": 0.17749034464359284,
      "num_tokens": 28918416.0,
      "step": 15680
    },
    {
      "entropy": 5.74223518371582,
      "epoch": 1.3177483721907162,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004830295518177047,
      "loss": 5.1691,
      "mean_token_accuracy": 0.19347862005233765,
      "num_tokens": 28927412.0,
      "step": 15685
    },
    {
      "entropy": 5.666182374954223,
      "epoch": 1.3181684520058812,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048301814039431227,
      "loss": 5.2102,
      "mean_token_accuracy": 0.18556985557079314,
      "num_tokens": 28936106.0,
      "step": 15690
    },
    {
      "entropy": 5.658556127548218,
      "epoch": 1.318588531821046,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048300672528598553,
      "loss": 5.2437,
      "mean_token_accuracy": 0.1865576907992363,
      "num_tokens": 28945197.0,
      "step": 15695
    },
    {
      "entropy": 5.754442167282105,
      "epoch": 1.3190086116362107,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004829953064929268,
      "loss": 5.266,
      "mean_token_accuracy": 0.17908058762550355,
      "num_tokens": 28954278.0,
      "step": 15700
    },
    {
      "entropy": 5.794744157791138,
      "epoch": 1.3194286914513758,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004829838840153383,
      "loss": 5.3341,
      "mean_token_accuracy": 0.18118251413106917,
      "num_tokens": 28963101.0,
      "step": 15705
    },
    {
      "entropy": 5.685741233825683,
      "epoch": 1.3198487712665408,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004829724578534224,
      "loss": 5.2195,
      "mean_token_accuracy": 0.18284594565629958,
      "num_tokens": 28972063.0,
      "step": 15710
    },
    {
      "entropy": 5.674992322921753,
      "epoch": 1.3202688510817056,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048296102800738153,
      "loss": 5.1612,
      "mean_token_accuracy": 0.1890750363469124,
      "num_tokens": 28981617.0,
      "step": 15715
    },
    {
      "entropy": 5.700266647338867,
      "epoch": 1.3206889308968703,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048294959447741807,
      "loss": 5.1844,
      "mean_token_accuracy": 0.18180931508541107,
      "num_tokens": 28989442.0,
      "step": 15720
    },
    {
      "entropy": 5.64476842880249,
      "epoch": 1.3211090107120353,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048293815726373467,
      "loss": 5.1616,
      "mean_token_accuracy": 0.18973579704761506,
      "num_tokens": 28999104.0,
      "step": 15725
    },
    {
      "entropy": 5.701443767547607,
      "epoch": 1.3215290905272001,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048292671636653386,
      "loss": 5.2462,
      "mean_token_accuracy": 0.18270690143108367,
      "num_tokens": 29008645.0,
      "step": 15730
    },
    {
      "entropy": 5.678061199188233,
      "epoch": 1.3219491703423651,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004829152717860184,
      "loss": 5.216,
      "mean_token_accuracy": 0.18288564532995225,
      "num_tokens": 29018655.0,
      "step": 15735
    },
    {
      "entropy": 5.706061220169067,
      "epoch": 1.32236925015753,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048290382352239087,
      "loss": 5.231,
      "mean_token_accuracy": 0.1897312343120575,
      "num_tokens": 29027109.0,
      "step": 15740
    },
    {
      "entropy": 5.661923503875732,
      "epoch": 1.3227893299726947,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048289237157585424,
      "loss": 5.0636,
      "mean_token_accuracy": 0.19740405529737473,
      "num_tokens": 29035535.0,
      "step": 15745
    },
    {
      "entropy": 5.63719687461853,
      "epoch": 1.3232094097878597,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004828809159466112,
      "loss": 5.2097,
      "mean_token_accuracy": 0.18439700305461884,
      "num_tokens": 29044723.0,
      "step": 15750
    },
    {
      "entropy": 5.674361085891723,
      "epoch": 1.3236294896030245,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004828694566348648,
      "loss": 5.3683,
      "mean_token_accuracy": 0.1732844293117523,
      "num_tokens": 29053636.0,
      "step": 15755
    },
    {
      "entropy": 5.765580749511718,
      "epoch": 1.3240495694181895,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048285799364081806,
      "loss": 5.26,
      "mean_token_accuracy": 0.18049966394901276,
      "num_tokens": 29062940.0,
      "step": 15760
    },
    {
      "entropy": 5.700114440917969,
      "epoch": 1.3244696492333543,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048284652696467404,
      "loss": 5.1766,
      "mean_token_accuracy": 0.18752314001321793,
      "num_tokens": 29072159.0,
      "step": 15765
    },
    {
      "entropy": 5.732506132125854,
      "epoch": 1.324889729048519,
      "grad_norm": 1.0,
      "learning_rate": 0.00048283505660663575,
      "loss": 5.2537,
      "mean_token_accuracy": 0.1843470439314842,
      "num_tokens": 29081544.0,
      "step": 15770
    },
    {
      "entropy": 5.664053440093994,
      "epoch": 1.325309808863684,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004828235825669064,
      "loss": 5.2148,
      "mean_token_accuracy": 0.18739358931779862,
      "num_tokens": 29090710.0,
      "step": 15775
    },
    {
      "entropy": 5.698979806900025,
      "epoch": 1.325729888678849,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048281210484568937,
      "loss": 5.2393,
      "mean_token_accuracy": 0.18073878884315492,
      "num_tokens": 29098988.0,
      "step": 15780
    },
    {
      "entropy": 5.633983993530274,
      "epoch": 1.326149968494014,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048280062344318794,
      "loss": 5.2537,
      "mean_token_accuracy": 0.1760525569319725,
      "num_tokens": 29108926.0,
      "step": 15785
    },
    {
      "entropy": 5.656332731246948,
      "epoch": 1.3265700483091787,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004827891383596054,
      "loss": 5.1976,
      "mean_token_accuracy": 0.1850075200200081,
      "num_tokens": 29118065.0,
      "step": 15790
    },
    {
      "entropy": 5.643776750564575,
      "epoch": 1.3269901281243437,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048277764959514524,
      "loss": 5.147,
      "mean_token_accuracy": 0.18783850222826004,
      "num_tokens": 29127030.0,
      "step": 15795
    },
    {
      "entropy": 5.733533763885498,
      "epoch": 1.3274102079395085,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004827661571500111,
      "loss": 5.2756,
      "mean_token_accuracy": 0.17884863913059235,
      "num_tokens": 29137200.0,
      "step": 15800
    },
    {
      "entropy": 5.731985092163086,
      "epoch": 1.3278302877546735,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048275466102440644,
      "loss": 5.2403,
      "mean_token_accuracy": 0.1873842164874077,
      "num_tokens": 29147029.0,
      "step": 15805
    },
    {
      "entropy": 5.666528749465942,
      "epoch": 1.3282503675698383,
      "grad_norm": 1.125,
      "learning_rate": 0.00048274316121853494,
      "loss": 5.1622,
      "mean_token_accuracy": 0.18753918558359145,
      "num_tokens": 29155675.0,
      "step": 15810
    },
    {
      "entropy": 5.685599756240845,
      "epoch": 1.328670447385003,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048273165773260023,
      "loss": 5.1922,
      "mean_token_accuracy": 0.18945285826921462,
      "num_tokens": 29164730.0,
      "step": 15815
    },
    {
      "entropy": 5.663177633285523,
      "epoch": 1.329090527200168,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004827201505668063,
      "loss": 5.2113,
      "mean_token_accuracy": 0.18898718804121017,
      "num_tokens": 29173074.0,
      "step": 15820
    },
    {
      "entropy": 5.749929809570313,
      "epoch": 1.3295106070153329,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004827086397213568,
      "loss": 5.3192,
      "mean_token_accuracy": 0.18125005960464477,
      "num_tokens": 29182175.0,
      "step": 15825
    },
    {
      "entropy": 5.841700172424316,
      "epoch": 1.3299306868304979,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004826971251964557,
      "loss": 5.5124,
      "mean_token_accuracy": 0.17815138399600983,
      "num_tokens": 29192910.0,
      "step": 15830
    },
    {
      "entropy": 5.655148458480835,
      "epoch": 1.3303507666456627,
      "grad_norm": 0.953125,
      "learning_rate": 0.000482685606992307,
      "loss": 5.1832,
      "mean_token_accuracy": 0.18534803092479707,
      "num_tokens": 29201969.0,
      "step": 15835
    },
    {
      "entropy": 5.774865055084229,
      "epoch": 1.3307708464608274,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048267408510911463,
      "loss": 5.3062,
      "mean_token_accuracy": 0.17872634381055832,
      "num_tokens": 29210475.0,
      "step": 15840
    },
    {
      "entropy": 5.651780223846435,
      "epoch": 1.3311909262759924,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004826625595470829,
      "loss": 5.1825,
      "mean_token_accuracy": 0.18137686550617219,
      "num_tokens": 29222586.0,
      "step": 15845
    },
    {
      "entropy": 5.6573486328125,
      "epoch": 1.3316110060911575,
      "grad_norm": 1.0,
      "learning_rate": 0.00048265103030641575,
      "loss": 5.2512,
      "mean_token_accuracy": 0.18388060629367828,
      "num_tokens": 29231503.0,
      "step": 15850
    },
    {
      "entropy": 5.629065036773682,
      "epoch": 1.3320310859063222,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004826394973873176,
      "loss": 5.1994,
      "mean_token_accuracy": 0.18393610119819642,
      "num_tokens": 29241534.0,
      "step": 15855
    },
    {
      "entropy": 5.661872720718383,
      "epoch": 1.332451165721487,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048262796078999266,
      "loss": 5.2142,
      "mean_token_accuracy": 0.18339792490005494,
      "num_tokens": 29250381.0,
      "step": 15860
    },
    {
      "entropy": 5.6986337184906,
      "epoch": 1.332871245536652,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004826164205146453,
      "loss": 5.2997,
      "mean_token_accuracy": 0.1787050783634186,
      "num_tokens": 29259205.0,
      "step": 15865
    },
    {
      "entropy": 5.635449981689453,
      "epoch": 1.3332913253518168,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048260487656147995,
      "loss": 5.1754,
      "mean_token_accuracy": 0.18985574692487717,
      "num_tokens": 29267723.0,
      "step": 15870
    },
    {
      "entropy": 5.646294021606446,
      "epoch": 1.3337114051669818,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048259332893070106,
      "loss": 5.1962,
      "mean_token_accuracy": 0.18741147220134735,
      "num_tokens": 29277102.0,
      "step": 15875
    },
    {
      "entropy": 5.689813184738159,
      "epoch": 1.3341314849821466,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004825817776225133,
      "loss": 5.17,
      "mean_token_accuracy": 0.18359510600566864,
      "num_tokens": 29286484.0,
      "step": 15880
    },
    {
      "entropy": 5.676045227050781,
      "epoch": 1.3345515647973114,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048257022263712123,
      "loss": 5.2582,
      "mean_token_accuracy": 0.19178806841373444,
      "num_tokens": 29296528.0,
      "step": 15885
    },
    {
      "entropy": 5.575808811187744,
      "epoch": 1.3349716446124764,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048255866397472954,
      "loss": 5.1322,
      "mean_token_accuracy": 0.1910013422369957,
      "num_tokens": 29305283.0,
      "step": 15890
    },
    {
      "entropy": 5.6476006507873535,
      "epoch": 1.3353917244276412,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000482547101635543,
      "loss": 5.1193,
      "mean_token_accuracy": 0.18840920627117158,
      "num_tokens": 29315088.0,
      "step": 15895
    },
    {
      "entropy": 5.694641256332398,
      "epoch": 1.3358118042428062,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048253553561976645,
      "loss": 5.1503,
      "mean_token_accuracy": 0.19069553762674332,
      "num_tokens": 29323793.0,
      "step": 15900
    },
    {
      "entropy": 5.618024921417236,
      "epoch": 1.336231884057971,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004825239659276047,
      "loss": 5.2077,
      "mean_token_accuracy": 0.18389461636543275,
      "num_tokens": 29334015.0,
      "step": 15905
    },
    {
      "entropy": 5.719866037368774,
      "epoch": 1.3366519638731358,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004825123925592628,
      "loss": 5.3219,
      "mean_token_accuracy": 0.18156332969665528,
      "num_tokens": 29343221.0,
      "step": 15910
    },
    {
      "entropy": 5.681499338150024,
      "epoch": 1.3370720436883008,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048250081551494574,
      "loss": 5.1748,
      "mean_token_accuracy": 0.1876213014125824,
      "num_tokens": 29352261.0,
      "step": 15915
    },
    {
      "entropy": 5.683165836334228,
      "epoch": 1.3374921235034656,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004824892347948586,
      "loss": 5.2784,
      "mean_token_accuracy": 0.17981873899698259,
      "num_tokens": 29362138.0,
      "step": 15920
    },
    {
      "entropy": 5.6785056591033936,
      "epoch": 1.3379122033186306,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004824776503992064,
      "loss": 5.1725,
      "mean_token_accuracy": 0.1873041197657585,
      "num_tokens": 29371234.0,
      "step": 15925
    },
    {
      "entropy": 5.647764873504639,
      "epoch": 1.3383322831337954,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004824660623281945,
      "loss": 5.2192,
      "mean_token_accuracy": 0.19202394634485245,
      "num_tokens": 29380371.0,
      "step": 15930
    },
    {
      "entropy": 5.711520624160767,
      "epoch": 1.3387523629489604,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048245447058202815,
      "loss": 5.3283,
      "mean_token_accuracy": 0.18305360674858093,
      "num_tokens": 29389230.0,
      "step": 15935
    },
    {
      "entropy": 5.707727432250977,
      "epoch": 1.3391724427641252,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004824428751609126,
      "loss": 5.2224,
      "mean_token_accuracy": 0.19087693393230437,
      "num_tokens": 29398753.0,
      "step": 15940
    },
    {
      "entropy": 5.674511051177978,
      "epoch": 1.3395925225792902,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048243127606505343,
      "loss": 5.2033,
      "mean_token_accuracy": 0.1884998083114624,
      "num_tokens": 29407487.0,
      "step": 15945
    },
    {
      "entropy": 5.595352745056152,
      "epoch": 1.340012602394455,
      "grad_norm": 1.0625,
      "learning_rate": 0.000482419673294656,
      "loss": 5.1925,
      "mean_token_accuracy": 0.18538012951612473,
      "num_tokens": 29416140.0,
      "step": 15950
    },
    {
      "entropy": 5.619274044036866,
      "epoch": 1.3404326822096198,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004824080668499259,
      "loss": 5.2138,
      "mean_token_accuracy": 0.18466936200857162,
      "num_tokens": 29424763.0,
      "step": 15955
    },
    {
      "entropy": 5.74043550491333,
      "epoch": 1.3408527620247848,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048239645673106855,
      "loss": 5.2103,
      "mean_token_accuracy": 0.18446938544511796,
      "num_tokens": 29434589.0,
      "step": 15960
    },
    {
      "entropy": 5.7114753246307375,
      "epoch": 1.3412728418399495,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048238484293828995,
      "loss": 5.2319,
      "mean_token_accuracy": 0.1793540120124817,
      "num_tokens": 29443549.0,
      "step": 15965
    },
    {
      "entropy": 5.69881420135498,
      "epoch": 1.3416929216551146,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004823732254717955,
      "loss": 5.2279,
      "mean_token_accuracy": 0.18338126689195633,
      "num_tokens": 29452457.0,
      "step": 15970
    },
    {
      "entropy": 5.649531650543213,
      "epoch": 1.3421130014702793,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004823616043317912,
      "loss": 5.2128,
      "mean_token_accuracy": 0.18218692094087602,
      "num_tokens": 29461238.0,
      "step": 15975
    },
    {
      "entropy": 5.694186639785767,
      "epoch": 1.3425330812854441,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048234997951848284,
      "loss": 5.2732,
      "mean_token_accuracy": 0.18032973855733872,
      "num_tokens": 29471170.0,
      "step": 15980
    },
    {
      "entropy": 5.759691190719605,
      "epoch": 1.3429531611006091,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004823383510320764,
      "loss": 5.2763,
      "mean_token_accuracy": 0.17922571152448655,
      "num_tokens": 29481017.0,
      "step": 15985
    },
    {
      "entropy": 5.7579868793487545,
      "epoch": 1.343373240915774,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048232671887277786,
      "loss": 5.2285,
      "mean_token_accuracy": 0.18784973174333572,
      "num_tokens": 29489809.0,
      "step": 15990
    },
    {
      "entropy": 5.648686456680298,
      "epoch": 1.343793320730939,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048231508304079313,
      "loss": 5.2608,
      "mean_token_accuracy": 0.1844817191362381,
      "num_tokens": 29499499.0,
      "step": 15995
    },
    {
      "entropy": 5.710238218307495,
      "epoch": 1.3442134005461037,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048230344353632855,
      "loss": 5.1959,
      "mean_token_accuracy": 0.18722955137491226,
      "num_tokens": 29508526.0,
      "step": 16000
    },
    {
      "entropy": 5.698666381835937,
      "epoch": 1.3446334803612685,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004822918003595902,
      "loss": 5.1403,
      "mean_token_accuracy": 0.18768172562122346,
      "num_tokens": 29517516.0,
      "step": 16005
    },
    {
      "entropy": 5.713804244995117,
      "epoch": 1.3450535601764335,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004822801535107843,
      "loss": 5.2397,
      "mean_token_accuracy": 0.17905609160661698,
      "num_tokens": 29526949.0,
      "step": 16010
    },
    {
      "entropy": 5.611508989334107,
      "epoch": 1.3454736399915985,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004822685029901173,
      "loss": 5.1542,
      "mean_token_accuracy": 0.1868077903985977,
      "num_tokens": 29536696.0,
      "step": 16015
    },
    {
      "entropy": 5.650392532348633,
      "epoch": 1.3458937198067633,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004822568487977954,
      "loss": 5.2475,
      "mean_token_accuracy": 0.18699756413698196,
      "num_tokens": 29545672.0,
      "step": 16020
    },
    {
      "entropy": 5.709530353546143,
      "epoch": 1.346313799621928,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048224519093402517,
      "loss": 5.2674,
      "mean_token_accuracy": 0.18413971960544587,
      "num_tokens": 29554888.0,
      "step": 16025
    },
    {
      "entropy": 5.718653774261474,
      "epoch": 1.346733879437093,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048223352939901317,
      "loss": 5.2083,
      "mean_token_accuracy": 0.18655561804771423,
      "num_tokens": 29564798.0,
      "step": 16030
    },
    {
      "entropy": 5.67486343383789,
      "epoch": 1.347153959252258,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004822218641929658,
      "loss": 5.2291,
      "mean_token_accuracy": 0.19045621752738953,
      "num_tokens": 29574802.0,
      "step": 16035
    },
    {
      "entropy": 5.730597066879272,
      "epoch": 1.347574039067423,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004822101953160899,
      "loss": 5.2313,
      "mean_token_accuracy": 0.18180814236402512,
      "num_tokens": 29583056.0,
      "step": 16040
    },
    {
      "entropy": 5.65416374206543,
      "epoch": 1.3479941188825877,
      "grad_norm": 1.109375,
      "learning_rate": 0.000482198522768592,
      "loss": 5.2119,
      "mean_token_accuracy": 0.18013266921043397,
      "num_tokens": 29591935.0,
      "step": 16045
    },
    {
      "entropy": 5.60307731628418,
      "epoch": 1.3484141986977525,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048218684655067907,
      "loss": 5.1419,
      "mean_token_accuracy": 0.18614706248044968,
      "num_tokens": 29600812.0,
      "step": 16050
    },
    {
      "entropy": 5.715270471572876,
      "epoch": 1.3488342785129175,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004821751666625577,
      "loss": 5.2374,
      "mean_token_accuracy": 0.18497280180454254,
      "num_tokens": 29610735.0,
      "step": 16055
    },
    {
      "entropy": 5.66562671661377,
      "epoch": 1.3492543583280823,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048216348310443506,
      "loss": 5.1848,
      "mean_token_accuracy": 0.18085191249847413,
      "num_tokens": 29620295.0,
      "step": 16060
    },
    {
      "entropy": 5.644209623336792,
      "epoch": 1.3496744381432473,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048215179587651795,
      "loss": 5.0735,
      "mean_token_accuracy": 0.1948539063334465,
      "num_tokens": 29628214.0,
      "step": 16065
    },
    {
      "entropy": 5.650769567489624,
      "epoch": 1.350094517958412,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004821401049790134,
      "loss": 5.2011,
      "mean_token_accuracy": 0.1889786899089813,
      "num_tokens": 29636598.0,
      "step": 16070
    },
    {
      "entropy": 5.656390953063965,
      "epoch": 1.3505145977735769,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004821284104121286,
      "loss": 5.1698,
      "mean_token_accuracy": 0.1902052253484726,
      "num_tokens": 29646052.0,
      "step": 16075
    },
    {
      "entropy": 5.660060262680053,
      "epoch": 1.3509346775887419,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048211671217607066,
      "loss": 5.2087,
      "mean_token_accuracy": 0.17866480946540833,
      "num_tokens": 29655310.0,
      "step": 16080
    },
    {
      "entropy": 5.710606479644776,
      "epoch": 1.3513547574039069,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004821050102710468,
      "loss": 5.2286,
      "mean_token_accuracy": 0.18574794232845307,
      "num_tokens": 29664020.0,
      "step": 16085
    },
    {
      "entropy": 5.681233167648315,
      "epoch": 1.3517748372190717,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048209330469726433,
      "loss": 5.2677,
      "mean_token_accuracy": 0.17854526042938232,
      "num_tokens": 29672416.0,
      "step": 16090
    },
    {
      "entropy": 5.745069169998169,
      "epoch": 1.3521949170342364,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048208159545493057,
      "loss": 5.1812,
      "mean_token_accuracy": 0.1929494246840477,
      "num_tokens": 29681148.0,
      "step": 16095
    },
    {
      "entropy": 5.6490894794464115,
      "epoch": 1.3526149968494015,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004820698825442531,
      "loss": 5.149,
      "mean_token_accuracy": 0.19246250987052918,
      "num_tokens": 29689089.0,
      "step": 16100
    },
    {
      "entropy": 5.641401147842407,
      "epoch": 1.3530350766645662,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048205816596543914,
      "loss": 5.2844,
      "mean_token_accuracy": 0.1790076032280922,
      "num_tokens": 29697704.0,
      "step": 16105
    },
    {
      "entropy": 5.69409499168396,
      "epoch": 1.3534551564797312,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048204644571869646,
      "loss": 5.2742,
      "mean_token_accuracy": 0.17900375425815582,
      "num_tokens": 29706966.0,
      "step": 16110
    },
    {
      "entropy": 5.677479314804077,
      "epoch": 1.353875236294896,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004820347218042326,
      "loss": 5.171,
      "mean_token_accuracy": 0.18637843281030655,
      "num_tokens": 29715817.0,
      "step": 16115
    },
    {
      "entropy": 5.701815795898438,
      "epoch": 1.3542953161100608,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004820229942222553,
      "loss": 5.2671,
      "mean_token_accuracy": 0.1821471408009529,
      "num_tokens": 29725500.0,
      "step": 16120
    },
    {
      "entropy": 5.6442993640899655,
      "epoch": 1.3547153959252258,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048201126297297214,
      "loss": 5.1935,
      "mean_token_accuracy": 0.18330859690904616,
      "num_tokens": 29734774.0,
      "step": 16125
    },
    {
      "entropy": 5.6478804588317875,
      "epoch": 1.3551354757403906,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004819995280565911,
      "loss": 5.1645,
      "mean_token_accuracy": 0.19132495522499085,
      "num_tokens": 29744667.0,
      "step": 16130
    },
    {
      "entropy": 5.7679986476898195,
      "epoch": 1.3555555555555556,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048198778947332,
      "loss": 5.2574,
      "mean_token_accuracy": 0.18357103914022446,
      "num_tokens": 29753644.0,
      "step": 16135
    },
    {
      "entropy": 5.7797150135040285,
      "epoch": 1.3559756353707204,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004819760472233668,
      "loss": 5.1989,
      "mean_token_accuracy": 0.19173940122127534,
      "num_tokens": 29762977.0,
      "step": 16140
    },
    {
      "entropy": 5.628623199462891,
      "epoch": 1.3563957151858852,
      "grad_norm": 1.0,
      "learning_rate": 0.00048196430130693956,
      "loss": 5.2063,
      "mean_token_accuracy": 0.1935465380549431,
      "num_tokens": 29772221.0,
      "step": 16145
    },
    {
      "entropy": 5.629550123214722,
      "epoch": 1.3568157950010502,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048195255172424627,
      "loss": 5.1813,
      "mean_token_accuracy": 0.19236460477113723,
      "num_tokens": 29781240.0,
      "step": 16150
    },
    {
      "entropy": 5.704911184310913,
      "epoch": 1.3572358748162152,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048194079847549507,
      "loss": 5.172,
      "mean_token_accuracy": 0.18543580919504166,
      "num_tokens": 29790330.0,
      "step": 16155
    },
    {
      "entropy": 5.726726722717285,
      "epoch": 1.35765595463138,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004819290415608942,
      "loss": 5.2761,
      "mean_token_accuracy": 0.18457056879997252,
      "num_tokens": 29800945.0,
      "step": 16160
    },
    {
      "entropy": 5.7215804100036625,
      "epoch": 1.3580760344465448,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004819172809806519,
      "loss": 5.3266,
      "mean_token_accuracy": 0.17797631919384002,
      "num_tokens": 29810391.0,
      "step": 16165
    },
    {
      "entropy": 5.675590658187867,
      "epoch": 1.3584961142617098,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048190551673497645,
      "loss": 5.1965,
      "mean_token_accuracy": 0.18425641804933549,
      "num_tokens": 29819511.0,
      "step": 16170
    },
    {
      "entropy": 5.668870687484741,
      "epoch": 1.3589161940768746,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004818937488240764,
      "loss": 5.2515,
      "mean_token_accuracy": 0.19021306931972504,
      "num_tokens": 29828313.0,
      "step": 16175
    },
    {
      "entropy": 5.6326165199279785,
      "epoch": 1.3593362738920396,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048188197724816014,
      "loss": 5.1529,
      "mean_token_accuracy": 0.19183594435453416,
      "num_tokens": 29837940.0,
      "step": 16180
    },
    {
      "entropy": 5.696715307235718,
      "epoch": 1.3597563537072044,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048187020200743613,
      "loss": 5.123,
      "mean_token_accuracy": 0.18970063030719758,
      "num_tokens": 29846799.0,
      "step": 16185
    },
    {
      "entropy": 5.706109189987183,
      "epoch": 1.3601764335223692,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000481858423102113,
      "loss": 5.2591,
      "mean_token_accuracy": 0.1786084234714508,
      "num_tokens": 29856263.0,
      "step": 16190
    },
    {
      "entropy": 5.629780912399292,
      "epoch": 1.3605965133375342,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004818466405323994,
      "loss": 5.1795,
      "mean_token_accuracy": 0.18698889911174774,
      "num_tokens": 29864335.0,
      "step": 16195
    },
    {
      "entropy": 5.702047729492188,
      "epoch": 1.361016593152699,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048183485429850417,
      "loss": 5.223,
      "mean_token_accuracy": 0.18214884400367737,
      "num_tokens": 29873466.0,
      "step": 16200
    },
    {
      "entropy": 5.610985946655274,
      "epoch": 1.361436672967864,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004818230644006359,
      "loss": 5.2052,
      "mean_token_accuracy": 0.19018648713827133,
      "num_tokens": 29883051.0,
      "step": 16205
    },
    {
      "entropy": 5.657182216644287,
      "epoch": 1.3618567527830288,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004818112708390036,
      "loss": 5.1462,
      "mean_token_accuracy": 0.18624110966920854,
      "num_tokens": 29891823.0,
      "step": 16210
    },
    {
      "entropy": 5.664607667922974,
      "epoch": 1.3622768325981935,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004817994736138162,
      "loss": 5.1552,
      "mean_token_accuracy": 0.1890528604388237,
      "num_tokens": 29900735.0,
      "step": 16215
    },
    {
      "entropy": 5.684064865112305,
      "epoch": 1.3626969124133586,
      "grad_norm": 1.0,
      "learning_rate": 0.0004817876727252824,
      "loss": 5.2565,
      "mean_token_accuracy": 0.19108208119869233,
      "num_tokens": 29910345.0,
      "step": 16220
    },
    {
      "entropy": 5.647084808349609,
      "epoch": 1.3631169922285233,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048177586817361166,
      "loss": 5.199,
      "mean_token_accuracy": 0.18699009865522384,
      "num_tokens": 29919650.0,
      "step": 16225
    },
    {
      "entropy": 5.7393255710601805,
      "epoch": 1.3635370720436883,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004817640599590128,
      "loss": 5.2482,
      "mean_token_accuracy": 0.1830395445227623,
      "num_tokens": 29928851.0,
      "step": 16230
    },
    {
      "entropy": 5.754026794433594,
      "epoch": 1.3639571518588531,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048175224808169506,
      "loss": 5.3171,
      "mean_token_accuracy": 0.17834263294935226,
      "num_tokens": 29939146.0,
      "step": 16235
    },
    {
      "entropy": 5.671836280822754,
      "epoch": 1.3643772316740181,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048174043254186775,
      "loss": 5.1608,
      "mean_token_accuracy": 0.18206970393657684,
      "num_tokens": 29947556.0,
      "step": 16240
    },
    {
      "entropy": 5.649259328842163,
      "epoch": 1.364797311489183,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004817286133397401,
      "loss": 5.2495,
      "mean_token_accuracy": 0.1835629940032959,
      "num_tokens": 29957319.0,
      "step": 16245
    },
    {
      "entropy": 5.6803196430206295,
      "epoch": 1.365217391304348,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004817167904755216,
      "loss": 5.2463,
      "mean_token_accuracy": 0.183173069357872,
      "num_tokens": 29966697.0,
      "step": 16250
    },
    {
      "entropy": 5.7281121730804445,
      "epoch": 1.3656374711195127,
      "grad_norm": 1.0,
      "learning_rate": 0.00048170496394942154,
      "loss": 5.2338,
      "mean_token_accuracy": 0.18870728015899657,
      "num_tokens": 29975103.0,
      "step": 16255
    },
    {
      "entropy": 5.622225761413574,
      "epoch": 1.3660575509346775,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048169313376164943,
      "loss": 5.153,
      "mean_token_accuracy": 0.18643567711114883,
      "num_tokens": 29984865.0,
      "step": 16260
    },
    {
      "entropy": 5.6989726543426515,
      "epoch": 1.3664776307498425,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048168129991241497,
      "loss": 5.182,
      "mean_token_accuracy": 0.18622321784496307,
      "num_tokens": 29994376.0,
      "step": 16265
    },
    {
      "entropy": 5.798887777328491,
      "epoch": 1.3668977105650073,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004816694624019277,
      "loss": 5.3627,
      "mean_token_accuracy": 0.18208667039871215,
      "num_tokens": 30004846.0,
      "step": 16270
    },
    {
      "entropy": 5.683022689819336,
      "epoch": 1.3673177903801723,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048165762123039723,
      "loss": 5.1925,
      "mean_token_accuracy": 0.18435177505016326,
      "num_tokens": 30014083.0,
      "step": 16275
    },
    {
      "entropy": 5.66888689994812,
      "epoch": 1.367737870195337,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048164577639803354,
      "loss": 5.1797,
      "mean_token_accuracy": 0.18978354334831238,
      "num_tokens": 30023606.0,
      "step": 16280
    },
    {
      "entropy": 5.586885023117065,
      "epoch": 1.3681579500105019,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004816339279050463,
      "loss": 5.1706,
      "mean_token_accuracy": 0.18460664749145508,
      "num_tokens": 30033657.0,
      "step": 16285
    },
    {
      "entropy": 5.624463367462158,
      "epoch": 1.368578029825667,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048162207575164537,
      "loss": 5.2221,
      "mean_token_accuracy": 0.18487675935029985,
      "num_tokens": 30043230.0,
      "step": 16290
    },
    {
      "entropy": 5.687584257125854,
      "epoch": 1.3689981096408317,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048161021993804075,
      "loss": 5.2568,
      "mean_token_accuracy": 0.18299194276332856,
      "num_tokens": 30054457.0,
      "step": 16295
    },
    {
      "entropy": 5.629288101196289,
      "epoch": 1.3694181894559967,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048159836046444255,
      "loss": 5.1093,
      "mean_token_accuracy": 0.1924107328057289,
      "num_tokens": 30062912.0,
      "step": 16300
    },
    {
      "entropy": 5.670363473892212,
      "epoch": 1.3698382692711615,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004815864973310607,
      "loss": 5.262,
      "mean_token_accuracy": 0.18307853043079375,
      "num_tokens": 30071340.0,
      "step": 16305
    },
    {
      "entropy": 5.767958879470825,
      "epoch": 1.3702583490863263,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048157463053810553,
      "loss": 5.3162,
      "mean_token_accuracy": 0.17506110966205596,
      "num_tokens": 30080334.0,
      "step": 16310
    },
    {
      "entropy": 5.695189142227173,
      "epoch": 1.3706784289014913,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048156276008578706,
      "loss": 5.1905,
      "mean_token_accuracy": 0.18693476766347886,
      "num_tokens": 30089391.0,
      "step": 16315
    },
    {
      "entropy": 5.656703138351441,
      "epoch": 1.3710985087166563,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004815508859743157,
      "loss": 5.143,
      "mean_token_accuracy": 0.19052834957838058,
      "num_tokens": 30099027.0,
      "step": 16320
    },
    {
      "entropy": 5.648742151260376,
      "epoch": 1.371518588531821,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004815390082039017,
      "loss": 5.1528,
      "mean_token_accuracy": 0.18969078809022905,
      "num_tokens": 30108088.0,
      "step": 16325
    },
    {
      "entropy": 5.614395427703857,
      "epoch": 1.3719386683469859,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048152712677475556,
      "loss": 5.1404,
      "mean_token_accuracy": 0.18849417418241501,
      "num_tokens": 30117768.0,
      "step": 16330
    },
    {
      "entropy": 5.727328443527222,
      "epoch": 1.3723587481621509,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048151524168708773,
      "loss": 5.2744,
      "mean_token_accuracy": 0.18134679645299911,
      "num_tokens": 30126364.0,
      "step": 16335
    },
    {
      "entropy": 5.680858421325683,
      "epoch": 1.3727788279773157,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048150335294110867,
      "loss": 5.2058,
      "mean_token_accuracy": 0.1897210881114006,
      "num_tokens": 30135365.0,
      "step": 16340
    },
    {
      "entropy": 5.719281530380249,
      "epoch": 1.3731989077924807,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048149146053702915,
      "loss": 5.2599,
      "mean_token_accuracy": 0.19516786932945251,
      "num_tokens": 30145542.0,
      "step": 16345
    },
    {
      "entropy": 5.72928147315979,
      "epoch": 1.3736189876076454,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004814795644750597,
      "loss": 5.3022,
      "mean_token_accuracy": 0.18064761012792588,
      "num_tokens": 30154100.0,
      "step": 16350
    },
    {
      "entropy": 5.664049482345581,
      "epoch": 1.3740390674228102,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048146766475541105,
      "loss": 5.1866,
      "mean_token_accuracy": 0.1865238815546036,
      "num_tokens": 30162647.0,
      "step": 16355
    },
    {
      "entropy": 5.828630256652832,
      "epoch": 1.3744591472379752,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048145576137829406,
      "loss": 5.321,
      "mean_token_accuracy": 0.1765571653842926,
      "num_tokens": 30172518.0,
      "step": 16360
    },
    {
      "entropy": 5.710259532928466,
      "epoch": 1.37487922705314,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004814438543439195,
      "loss": 5.2559,
      "mean_token_accuracy": 0.1834559127688408,
      "num_tokens": 30183124.0,
      "step": 16365
    },
    {
      "entropy": 5.691094875335693,
      "epoch": 1.375299306868305,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004814319436524984,
      "loss": 5.1978,
      "mean_token_accuracy": 0.18797856271266938,
      "num_tokens": 30191861.0,
      "step": 16370
    },
    {
      "entropy": 5.591437816619873,
      "epoch": 1.3757193866834698,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048142002930424174,
      "loss": 5.1117,
      "mean_token_accuracy": 0.18482332080602645,
      "num_tokens": 30200308.0,
      "step": 16375
    },
    {
      "entropy": 5.677455806732178,
      "epoch": 1.3761394664986346,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004814081112993605,
      "loss": 5.2173,
      "mean_token_accuracy": 0.18738737106323242,
      "num_tokens": 30209380.0,
      "step": 16380
    },
    {
      "entropy": 5.728288364410401,
      "epoch": 1.3765595463137996,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004813961896380659,
      "loss": 5.2842,
      "mean_token_accuracy": 0.18687527030706405,
      "num_tokens": 30218549.0,
      "step": 16385
    },
    {
      "entropy": 5.696465682983399,
      "epoch": 1.3769796261289646,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004813842643205691,
      "loss": 5.2172,
      "mean_token_accuracy": 0.18507154583930968,
      "num_tokens": 30228119.0,
      "step": 16390
    },
    {
      "entropy": 5.729372644424439,
      "epoch": 1.3773997059441294,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004813723353470813,
      "loss": 5.2119,
      "mean_token_accuracy": 0.18475860953330994,
      "num_tokens": 30236765.0,
      "step": 16395
    },
    {
      "entropy": 5.6812145709991455,
      "epoch": 1.3778197857592942,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004813604027178139,
      "loss": 5.1536,
      "mean_token_accuracy": 0.18649637401103974,
      "num_tokens": 30246089.0,
      "step": 16400
    },
    {
      "entropy": 5.681843519210815,
      "epoch": 1.3782398655744592,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048134846643297817,
      "loss": 5.2687,
      "mean_token_accuracy": 0.18046996742486954,
      "num_tokens": 30255806.0,
      "step": 16405
    },
    {
      "entropy": 5.733788967132568,
      "epoch": 1.378659945389624,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004813365264927856,
      "loss": 5.3098,
      "mean_token_accuracy": 0.17495220452547072,
      "num_tokens": 30267112.0,
      "step": 16410
    },
    {
      "entropy": 5.700532341003418,
      "epoch": 1.379080025204789,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004813245828974477,
      "loss": 5.2088,
      "mean_token_accuracy": 0.18486867100000381,
      "num_tokens": 30276168.0,
      "step": 16415
    },
    {
      "entropy": 5.643577003479004,
      "epoch": 1.3795001050199538,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004813126356471761,
      "loss": 5.2219,
      "mean_token_accuracy": 0.1871683344244957,
      "num_tokens": 30285723.0,
      "step": 16420
    },
    {
      "entropy": 5.766163539886475,
      "epoch": 1.3799201848351186,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004813006847421824,
      "loss": 5.2707,
      "mean_token_accuracy": 0.1844539687037468,
      "num_tokens": 30294790.0,
      "step": 16425
    },
    {
      "entropy": 5.711303186416626,
      "epoch": 1.3803402646502836,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004812887301826783,
      "loss": 5.2091,
      "mean_token_accuracy": 0.188228540122509,
      "num_tokens": 30303439.0,
      "step": 16430
    },
    {
      "entropy": 5.655287599563598,
      "epoch": 1.3807603444654484,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004812767719688755,
      "loss": 5.1915,
      "mean_token_accuracy": 0.18581496328115463,
      "num_tokens": 30312493.0,
      "step": 16435
    },
    {
      "entropy": 5.702940559387207,
      "epoch": 1.3811804242806134,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004812648101009859,
      "loss": 5.2199,
      "mean_token_accuracy": 0.19138378500938416,
      "num_tokens": 30321637.0,
      "step": 16440
    },
    {
      "entropy": 5.757873582839966,
      "epoch": 1.3816005040957782,
      "grad_norm": 1.0,
      "learning_rate": 0.0004812528445792215,
      "loss": 5.3416,
      "mean_token_accuracy": 0.17706894874572754,
      "num_tokens": 30330730.0,
      "step": 16445
    },
    {
      "entropy": 5.659854936599731,
      "epoch": 1.382020583910943,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048124087540379407,
      "loss": 5.1942,
      "mean_token_accuracy": 0.1843529924750328,
      "num_tokens": 30339568.0,
      "step": 16450
    },
    {
      "entropy": 5.675431632995606,
      "epoch": 1.382440663726108,
      "grad_norm": 1.125,
      "learning_rate": 0.00048122890257491573,
      "loss": 5.241,
      "mean_token_accuracy": 0.17641419023275376,
      "num_tokens": 30349225.0,
      "step": 16455
    },
    {
      "entropy": 5.733782482147217,
      "epoch": 1.382860743541273,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048121692609279866,
      "loss": 5.2249,
      "mean_token_accuracy": 0.18905517905950547,
      "num_tokens": 30358804.0,
      "step": 16460
    },
    {
      "entropy": 5.712367248535156,
      "epoch": 1.3832808233564378,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004812049459576549,
      "loss": 5.3053,
      "mean_token_accuracy": 0.18559854477643967,
      "num_tokens": 30368490.0,
      "step": 16465
    },
    {
      "entropy": 5.75775671005249,
      "epoch": 1.3837009031716025,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004811929621696966,
      "loss": 5.2039,
      "mean_token_accuracy": 0.1845212161540985,
      "num_tokens": 30377117.0,
      "step": 16470
    },
    {
      "entropy": 5.612489175796509,
      "epoch": 1.3841209829867676,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048118097472913627,
      "loss": 5.0772,
      "mean_token_accuracy": 0.19641234576702118,
      "num_tokens": 30385151.0,
      "step": 16475
    },
    {
      "entropy": 5.589596366882324,
      "epoch": 1.3845410628019323,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004811689836361861,
      "loss": 5.1127,
      "mean_token_accuracy": 0.18863447904586791,
      "num_tokens": 30394837.0,
      "step": 16480
    },
    {
      "entropy": 5.682020092010498,
      "epoch": 1.3849611426170974,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004811569888910585,
      "loss": 5.1789,
      "mean_token_accuracy": 0.19005471765995025,
      "num_tokens": 30403507.0,
      "step": 16485
    },
    {
      "entropy": 5.664462995529175,
      "epoch": 1.3853812224322621,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004811449904939661,
      "loss": 5.1717,
      "mean_token_accuracy": 0.19051122814416885,
      "num_tokens": 30412941.0,
      "step": 16490
    },
    {
      "entropy": 5.628215360641479,
      "epoch": 1.385801302247427,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048113298844512127,
      "loss": 5.1874,
      "mean_token_accuracy": 0.19550154060125352,
      "num_tokens": 30421823.0,
      "step": 16495
    },
    {
      "entropy": 5.6825000762939455,
      "epoch": 1.386221382062592,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004811209827447367,
      "loss": 5.2658,
      "mean_token_accuracy": 0.17943769246339797,
      "num_tokens": 30431901.0,
      "step": 16500
    },
    {
      "entropy": 5.6902892112731935,
      "epoch": 1.3866414618777567,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048110897339302504,
      "loss": 5.1976,
      "mean_token_accuracy": 0.18454488664865493,
      "num_tokens": 30442037.0,
      "step": 16505
    },
    {
      "entropy": 5.686445093154907,
      "epoch": 1.3870615416929217,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048109696039019915,
      "loss": 5.1736,
      "mean_token_accuracy": 0.19129129648208618,
      "num_tokens": 30451189.0,
      "step": 16510
    },
    {
      "entropy": 5.675273799896241,
      "epoch": 1.3874816215080865,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004810849437364716,
      "loss": 5.2597,
      "mean_token_accuracy": 0.18512654304504395,
      "num_tokens": 30460214.0,
      "step": 16515
    },
    {
      "entropy": 5.697692012786865,
      "epoch": 1.3879017013232513,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048107292343205546,
      "loss": 5.2643,
      "mean_token_accuracy": 0.1834980934858322,
      "num_tokens": 30469936.0,
      "step": 16520
    },
    {
      "entropy": 5.70015811920166,
      "epoch": 1.3883217811384163,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004810608994771636,
      "loss": 5.2123,
      "mean_token_accuracy": 0.18490938246250152,
      "num_tokens": 30479282.0,
      "step": 16525
    },
    {
      "entropy": 5.737960290908814,
      "epoch": 1.388741860953581,
      "grad_norm": 1.0,
      "learning_rate": 0.000481048871872009,
      "loss": 5.2359,
      "mean_token_accuracy": 0.1861589878797531,
      "num_tokens": 30487839.0,
      "step": 16530
    },
    {
      "entropy": 5.755791521072387,
      "epoch": 1.389161940768746,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048103684061680463,
      "loss": 5.2903,
      "mean_token_accuracy": 0.18441792130470275,
      "num_tokens": 30497327.0,
      "step": 16535
    },
    {
      "entropy": 5.683746957778931,
      "epoch": 1.389582020583911,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048102480571176384,
      "loss": 5.1801,
      "mean_token_accuracy": 0.18794931769371032,
      "num_tokens": 30506996.0,
      "step": 16540
    },
    {
      "entropy": 5.667428731918335,
      "epoch": 1.390002100399076,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004810127671570997,
      "loss": 5.1195,
      "mean_token_accuracy": 0.19714997112751007,
      "num_tokens": 30515627.0,
      "step": 16545
    },
    {
      "entropy": 5.691244411468506,
      "epoch": 1.3904221802142407,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048100072495302544,
      "loss": 5.2459,
      "mean_token_accuracy": 0.18254385143518448,
      "num_tokens": 30525858.0,
      "step": 16550
    },
    {
      "entropy": 5.615413522720337,
      "epoch": 1.3908422600294057,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004809886790997544,
      "loss": 5.1485,
      "mean_token_accuracy": 0.19099547117948532,
      "num_tokens": 30536331.0,
      "step": 16555
    },
    {
      "entropy": 5.7060610294342045,
      "epoch": 1.3912623398445705,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004809766295975,
      "loss": 5.1935,
      "mean_token_accuracy": 0.18938792794942855,
      "num_tokens": 30545329.0,
      "step": 16560
    },
    {
      "entropy": 5.658728551864624,
      "epoch": 1.3916824196597353,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004809645764464757,
      "loss": 5.1591,
      "mean_token_accuracy": 0.1850077986717224,
      "num_tokens": 30554357.0,
      "step": 16565
    },
    {
      "entropy": 5.721221685409546,
      "epoch": 1.3921024994749003,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048095251964689494,
      "loss": 5.3188,
      "mean_token_accuracy": 0.1804337978363037,
      "num_tokens": 30563548.0,
      "step": 16570
    },
    {
      "entropy": 5.709998416900635,
      "epoch": 1.392522579290065,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048094045919897134,
      "loss": 5.2098,
      "mean_token_accuracy": 0.1864153116941452,
      "num_tokens": 30572844.0,
      "step": 16575
    },
    {
      "entropy": 5.6381920337677,
      "epoch": 1.39294265910523,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004809283951029185,
      "loss": 5.1356,
      "mean_token_accuracy": 0.19570783972740174,
      "num_tokens": 30580930.0,
      "step": 16580
    },
    {
      "entropy": 5.6930629253387455,
      "epoch": 1.3933627389203949,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004809163273589503,
      "loss": 5.1383,
      "mean_token_accuracy": 0.19307959526777269,
      "num_tokens": 30589917.0,
      "step": 16585
    },
    {
      "entropy": 5.656709337234497,
      "epoch": 1.3937828187355596,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048090425596728035,
      "loss": 5.2374,
      "mean_token_accuracy": 0.18964896351099014,
      "num_tokens": 30599282.0,
      "step": 16590
    },
    {
      "entropy": 5.688169622421265,
      "epoch": 1.3942028985507247,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048089218092812254,
      "loss": 5.2299,
      "mean_token_accuracy": 0.18261457085609437,
      "num_tokens": 30608244.0,
      "step": 16595
    },
    {
      "entropy": 5.745567798614502,
      "epoch": 1.3946229783658894,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048088010224169064,
      "loss": 5.3277,
      "mean_token_accuracy": 0.18508842587471008,
      "num_tokens": 30617340.0,
      "step": 16600
    },
    {
      "entropy": 5.80283489227295,
      "epoch": 1.3950430581810545,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048086801990819886,
      "loss": 5.2794,
      "mean_token_accuracy": 0.17990150898694993,
      "num_tokens": 30626244.0,
      "step": 16605
    },
    {
      "entropy": 5.698116302490234,
      "epoch": 1.3954631379962192,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048085593392786113,
      "loss": 5.2433,
      "mean_token_accuracy": 0.18673592656850815,
      "num_tokens": 30635279.0,
      "step": 16610
    },
    {
      "entropy": 5.705820941925049,
      "epoch": 1.395883217811384,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004808438443008915,
      "loss": 5.383,
      "mean_token_accuracy": 0.17610064297914504,
      "num_tokens": 30645790.0,
      "step": 16615
    },
    {
      "entropy": 5.708877611160278,
      "epoch": 1.396303297626549,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004808317510275041,
      "loss": 5.233,
      "mean_token_accuracy": 0.17889177948236465,
      "num_tokens": 30654497.0,
      "step": 16620
    },
    {
      "entropy": 5.751762914657593,
      "epoch": 1.396723377441714,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004808196541079133,
      "loss": 5.2833,
      "mean_token_accuracy": 0.18053341060876846,
      "num_tokens": 30663760.0,
      "step": 16625
    },
    {
      "entropy": 5.706173324584961,
      "epoch": 1.3971434572568788,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048080755354233326,
      "loss": 5.2814,
      "mean_token_accuracy": 0.1849368005990982,
      "num_tokens": 30674263.0,
      "step": 16630
    },
    {
      "entropy": 5.7207845687866214,
      "epoch": 1.3975635370720436,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004807954493309784,
      "loss": 5.1703,
      "mean_token_accuracy": 0.18548215180635452,
      "num_tokens": 30683501.0,
      "step": 16635
    },
    {
      "entropy": 5.716302537918091,
      "epoch": 1.3979836168872086,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048078334147406314,
      "loss": 5.1688,
      "mean_token_accuracy": 0.20160829424858093,
      "num_tokens": 30691917.0,
      "step": 16640
    },
    {
      "entropy": 5.638992595672607,
      "epoch": 1.3984036967023734,
      "grad_norm": 1.125,
      "learning_rate": 0.00048077122997180197,
      "loss": 5.2454,
      "mean_token_accuracy": 0.1845514416694641,
      "num_tokens": 30701753.0,
      "step": 16645
    },
    {
      "entropy": 5.614064693450928,
      "epoch": 1.3988237765175384,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004807591148244093,
      "loss": 5.2108,
      "mean_token_accuracy": 0.18467018008232117,
      "num_tokens": 30710878.0,
      "step": 16650
    },
    {
      "entropy": 5.66349778175354,
      "epoch": 1.3992438563327032,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004807469960321,
      "loss": 5.1053,
      "mean_token_accuracy": 0.1917671635746956,
      "num_tokens": 30719372.0,
      "step": 16655
    },
    {
      "entropy": 5.743276643753052,
      "epoch": 1.399663936147868,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048073487359508854,
      "loss": 5.2658,
      "mean_token_accuracy": 0.18102390319108963,
      "num_tokens": 30728529.0,
      "step": 16660
    },
    {
      "entropy": 5.723650646209717,
      "epoch": 1.400084015963033,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048072274751358976,
      "loss": 5.1997,
      "mean_token_accuracy": 0.19160695970058442,
      "num_tokens": 30737704.0,
      "step": 16665
    },
    {
      "entropy": 5.658043146133423,
      "epoch": 1.4005040957781978,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048071061778781843,
      "loss": 5.1824,
      "mean_token_accuracy": 0.18578770160675048,
      "num_tokens": 30747836.0,
      "step": 16670
    },
    {
      "entropy": 5.649416732788086,
      "epoch": 1.4009241755933628,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004806984844179894,
      "loss": 5.2438,
      "mean_token_accuracy": 0.18415799140930175,
      "num_tokens": 30757881.0,
      "step": 16675
    },
    {
      "entropy": 5.741478347778321,
      "epoch": 1.4013442554085276,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048068634740431774,
      "loss": 5.2511,
      "mean_token_accuracy": 0.18036163747310638,
      "num_tokens": 30767592.0,
      "step": 16680
    },
    {
      "entropy": 5.706129503250122,
      "epoch": 1.4017643352236924,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004806742067470182,
      "loss": 5.2417,
      "mean_token_accuracy": 0.18404701650142669,
      "num_tokens": 30776633.0,
      "step": 16685
    },
    {
      "entropy": 5.738908576965332,
      "epoch": 1.4021844150388574,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048066206244630613,
      "loss": 5.1852,
      "mean_token_accuracy": 0.18462662696838378,
      "num_tokens": 30785195.0,
      "step": 16690
    },
    {
      "entropy": 5.64884204864502,
      "epoch": 1.4026044948540224,
      "grad_norm": 1.0,
      "learning_rate": 0.00048064991450239643,
      "loss": 5.174,
      "mean_token_accuracy": 0.1860155552625656,
      "num_tokens": 30794397.0,
      "step": 16695
    },
    {
      "entropy": 5.793358564376831,
      "epoch": 1.4030245746691872,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048063776291550444,
      "loss": 5.3409,
      "mean_token_accuracy": 0.17929955422878266,
      "num_tokens": 30803312.0,
      "step": 16700
    },
    {
      "entropy": 5.706845998764038,
      "epoch": 1.403444654484352,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048062560768584537,
      "loss": 5.2262,
      "mean_token_accuracy": 0.19814287573099137,
      "num_tokens": 30812519.0,
      "step": 16705
    },
    {
      "entropy": 5.669205665588379,
      "epoch": 1.403864734299517,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048061344881363444,
      "loss": 5.1921,
      "mean_token_accuracy": 0.190933957695961,
      "num_tokens": 30821558.0,
      "step": 16710
    },
    {
      "entropy": 5.731124401092529,
      "epoch": 1.4042848141146818,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004806012862990873,
      "loss": 5.224,
      "mean_token_accuracy": 0.18388185054063796,
      "num_tokens": 30831521.0,
      "step": 16715
    },
    {
      "entropy": 5.691718339920044,
      "epoch": 1.4047048939298468,
      "grad_norm": 1.0,
      "learning_rate": 0.00048058912014241914,
      "loss": 5.1936,
      "mean_token_accuracy": 0.18455802649259567,
      "num_tokens": 30841191.0,
      "step": 16720
    },
    {
      "entropy": 5.725134992599488,
      "epoch": 1.4051249737450116,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004805769503438456,
      "loss": 5.3034,
      "mean_token_accuracy": 0.1815813049674034,
      "num_tokens": 30850556.0,
      "step": 16725
    },
    {
      "entropy": 5.686429357528686,
      "epoch": 1.4055450535601763,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048056477690358227,
      "loss": 5.2025,
      "mean_token_accuracy": 0.1881520226597786,
      "num_tokens": 30859410.0,
      "step": 16730
    },
    {
      "entropy": 5.757922697067261,
      "epoch": 1.4059651333753413,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004805525998218447,
      "loss": 5.2241,
      "mean_token_accuracy": 0.1792560875415802,
      "num_tokens": 30868048.0,
      "step": 16735
    },
    {
      "entropy": 5.712350368499756,
      "epoch": 1.4063852131905061,
      "grad_norm": 1.125,
      "learning_rate": 0.00048054041909884873,
      "loss": 5.2521,
      "mean_token_accuracy": 0.1838813006877899,
      "num_tokens": 30876785.0,
      "step": 16740
    },
    {
      "entropy": 5.776911640167237,
      "epoch": 1.4068052930056711,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048052823473481007,
      "loss": 5.3113,
      "mean_token_accuracy": 0.17895396649837494,
      "num_tokens": 30886158.0,
      "step": 16745
    },
    {
      "entropy": 5.749477767944336,
      "epoch": 1.407225372820836,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048051604672994446,
      "loss": 5.1804,
      "mean_token_accuracy": 0.18243900686502457,
      "num_tokens": 30895283.0,
      "step": 16750
    },
    {
      "entropy": 5.689336585998535,
      "epoch": 1.4076454526360007,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048050385508446804,
      "loss": 5.1844,
      "mean_token_accuracy": 0.19056810736656188,
      "num_tokens": 30905514.0,
      "step": 16755
    },
    {
      "entropy": 5.688450527191162,
      "epoch": 1.4080655324511657,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048049165979859655,
      "loss": 5.1198,
      "mean_token_accuracy": 0.20527548491954803,
      "num_tokens": 30914794.0,
      "step": 16760
    },
    {
      "entropy": 5.638699722290039,
      "epoch": 1.4084856122663307,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048047946087254615,
      "loss": 5.1545,
      "mean_token_accuracy": 0.18647514283657074,
      "num_tokens": 30923823.0,
      "step": 16765
    },
    {
      "entropy": 5.668554210662842,
      "epoch": 1.4089056920814955,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048046725830653295,
      "loss": 5.2485,
      "mean_token_accuracy": 0.18687618225812913,
      "num_tokens": 30932738.0,
      "step": 16770
    },
    {
      "entropy": 5.70009298324585,
      "epoch": 1.4093257718966603,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048045505210077304,
      "loss": 5.284,
      "mean_token_accuracy": 0.17593639642000197,
      "num_tokens": 30942302.0,
      "step": 16775
    },
    {
      "entropy": 5.689658355712891,
      "epoch": 1.4097458517118253,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004804428422554826,
      "loss": 5.174,
      "mean_token_accuracy": 0.18801504075527192,
      "num_tokens": 30951662.0,
      "step": 16780
    },
    {
      "entropy": 5.683955383300781,
      "epoch": 1.41016593152699,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004804306287708782,
      "loss": 5.2166,
      "mean_token_accuracy": 0.18863149583339692,
      "num_tokens": 30960475.0,
      "step": 16785
    },
    {
      "entropy": 5.639825916290283,
      "epoch": 1.410586011342155,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048041841164717574,
      "loss": 5.044,
      "mean_token_accuracy": 0.19546734541654587,
      "num_tokens": 30969075.0,
      "step": 16790
    },
    {
      "entropy": 5.654620504379272,
      "epoch": 1.41100609115732,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004804061908845921,
      "loss": 5.1254,
      "mean_token_accuracy": 0.19384037107229232,
      "num_tokens": 30978030.0,
      "step": 16795
    },
    {
      "entropy": 5.690433406829834,
      "epoch": 1.4114261709724847,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048039396648334346,
      "loss": 5.1073,
      "mean_token_accuracy": 0.1880299359560013,
      "num_tokens": 30985639.0,
      "step": 16800
    },
    {
      "entropy": 5.681980085372925,
      "epoch": 1.4118462507876497,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004803817384436465,
      "loss": 5.229,
      "mean_token_accuracy": 0.18509876281023024,
      "num_tokens": 30994811.0,
      "step": 16805
    },
    {
      "entropy": 5.68304557800293,
      "epoch": 1.4122663306028145,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004803695067657178,
      "loss": 5.1954,
      "mean_token_accuracy": 0.18453514724969863,
      "num_tokens": 31003813.0,
      "step": 16810
    },
    {
      "entropy": 5.663872814178466,
      "epoch": 1.4126864104179795,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000480357271449774,
      "loss": 5.1689,
      "mean_token_accuracy": 0.1981577157974243,
      "num_tokens": 31012488.0,
      "step": 16815
    },
    {
      "entropy": 5.729624271392822,
      "epoch": 1.4131064902331443,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004803450324960318,
      "loss": 5.1766,
      "mean_token_accuracy": 0.18569174706935881,
      "num_tokens": 31021089.0,
      "step": 16820
    },
    {
      "entropy": 5.685629987716675,
      "epoch": 1.413526570048309,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048033278990470825,
      "loss": 5.2051,
      "mean_token_accuracy": 0.18730328381061553,
      "num_tokens": 31029903.0,
      "step": 16825
    },
    {
      "entropy": 5.655189180374146,
      "epoch": 1.413946649863474,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048032054367601996,
      "loss": 5.1896,
      "mean_token_accuracy": 0.1882103145122528,
      "num_tokens": 31039207.0,
      "step": 16830
    },
    {
      "entropy": 5.659401798248291,
      "epoch": 1.414366729678639,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048030829381018396,
      "loss": 5.2387,
      "mean_token_accuracy": 0.18444830179214478,
      "num_tokens": 31048190.0,
      "step": 16835
    },
    {
      "entropy": 5.723845958709717,
      "epoch": 1.4147868094938039,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004802960403074173,
      "loss": 5.2848,
      "mean_token_accuracy": 0.1912844106554985,
      "num_tokens": 31058769.0,
      "step": 16840
    },
    {
      "entropy": 5.738271188735962,
      "epoch": 1.4152068893089687,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048028378316793705,
      "loss": 5.2727,
      "mean_token_accuracy": 0.18357594907283784,
      "num_tokens": 31066830.0,
      "step": 16845
    },
    {
      "entropy": 5.755136919021607,
      "epoch": 1.4156269691241337,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004802715223919602,
      "loss": 5.2978,
      "mean_token_accuracy": 0.1875277042388916,
      "num_tokens": 31077205.0,
      "step": 16850
    },
    {
      "entropy": 5.7461535930633545,
      "epoch": 1.4160470489392984,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048025925797970403,
      "loss": 5.2105,
      "mean_token_accuracy": 0.19118614345788956,
      "num_tokens": 31087327.0,
      "step": 16855
    },
    {
      "entropy": 5.660273933410645,
      "epoch": 1.4164671287544635,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048024698993138587,
      "loss": 5.157,
      "mean_token_accuracy": 0.19000198394060136,
      "num_tokens": 31096501.0,
      "step": 16860
    },
    {
      "entropy": 5.707158851623535,
      "epoch": 1.4168872085696282,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048023471824722294,
      "loss": 5.3504,
      "mean_token_accuracy": 0.17294915169477462,
      "num_tokens": 31105949.0,
      "step": 16865
    },
    {
      "entropy": 5.718421936035156,
      "epoch": 1.417307288384793,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048022244292743256,
      "loss": 5.2416,
      "mean_token_accuracy": 0.17726877331733704,
      "num_tokens": 31115482.0,
      "step": 16870
    },
    {
      "entropy": 5.724588251113891,
      "epoch": 1.417727368199958,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048021016397223234,
      "loss": 5.2155,
      "mean_token_accuracy": 0.1878615215420723,
      "num_tokens": 31124758.0,
      "step": 16875
    },
    {
      "entropy": 5.60503306388855,
      "epoch": 1.4181474480151228,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048019788138183977,
      "loss": 5.0617,
      "mean_token_accuracy": 0.199758680164814,
      "num_tokens": 31134114.0,
      "step": 16880
    },
    {
      "entropy": 5.598604965209961,
      "epoch": 1.4185675278302878,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048018559515647244,
      "loss": 5.1243,
      "mean_token_accuracy": 0.19241459369659425,
      "num_tokens": 31142667.0,
      "step": 16885
    },
    {
      "entropy": 5.649864339828492,
      "epoch": 1.4189876076454526,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048017330529634785,
      "loss": 5.2327,
      "mean_token_accuracy": 0.17943748831748962,
      "num_tokens": 31152105.0,
      "step": 16890
    },
    {
      "entropy": 5.712868595123291,
      "epoch": 1.4194076874606174,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048016101180168376,
      "loss": 5.2131,
      "mean_token_accuracy": 0.19259319007396697,
      "num_tokens": 31160277.0,
      "step": 16895
    },
    {
      "entropy": 5.820760250091553,
      "epoch": 1.4198277672757824,
      "grad_norm": 1.0,
      "learning_rate": 0.00048014871467269804,
      "loss": 5.4002,
      "mean_token_accuracy": 0.17599798291921614,
      "num_tokens": 31170677.0,
      "step": 16900
    },
    {
      "entropy": 5.696310091018677,
      "epoch": 1.4202478470909472,
      "grad_norm": 1.0,
      "learning_rate": 0.00048013641390960856,
      "loss": 5.1949,
      "mean_token_accuracy": 0.18230001777410507,
      "num_tokens": 31179298.0,
      "step": 16905
    },
    {
      "entropy": 5.676075410842896,
      "epoch": 1.4206679269061122,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004801241095126331,
      "loss": 5.2141,
      "mean_token_accuracy": 0.18268964290618897,
      "num_tokens": 31188547.0,
      "step": 16910
    },
    {
      "entropy": 5.668713903427124,
      "epoch": 1.421088006721277,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004801118014819896,
      "loss": 5.2208,
      "mean_token_accuracy": 0.18514768034219742,
      "num_tokens": 31197680.0,
      "step": 16915
    },
    {
      "entropy": 5.673295354843139,
      "epoch": 1.421508086536442,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004800994898178962,
      "loss": 5.1659,
      "mean_token_accuracy": 0.18991825729608536,
      "num_tokens": 31206351.0,
      "step": 16920
    },
    {
      "entropy": 5.713836240768432,
      "epoch": 1.4219281663516068,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004800871745205708,
      "loss": 5.359,
      "mean_token_accuracy": 0.17899004518985748,
      "num_tokens": 31216478.0,
      "step": 16925
    },
    {
      "entropy": 5.787197780609131,
      "epoch": 1.4223482461667718,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048007485559023195,
      "loss": 5.3016,
      "mean_token_accuracy": 0.1779828578233719,
      "num_tokens": 31225920.0,
      "step": 16930
    },
    {
      "entropy": 5.68327488899231,
      "epoch": 1.4227683259819366,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004800625330270975,
      "loss": 5.1887,
      "mean_token_accuracy": 0.18587670624256133,
      "num_tokens": 31235061.0,
      "step": 16935
    },
    {
      "entropy": 5.6192492008209225,
      "epoch": 1.4231884057971014,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004800502068313859,
      "loss": 5.1753,
      "mean_token_accuracy": 0.1914423793554306,
      "num_tokens": 31243448.0,
      "step": 16940
    },
    {
      "entropy": 5.708370733261108,
      "epoch": 1.4236084856122664,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004800378770033154,
      "loss": 5.2884,
      "mean_token_accuracy": 0.18514792174100875,
      "num_tokens": 31252569.0,
      "step": 16945
    },
    {
      "entropy": 5.70576286315918,
      "epoch": 1.4240285654274312,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004800255435431046,
      "loss": 5.1408,
      "mean_token_accuracy": 0.19442081600427627,
      "num_tokens": 31261905.0,
      "step": 16950
    },
    {
      "entropy": 5.643242311477661,
      "epoch": 1.4244486452425962,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048001320645097177,
      "loss": 5.1298,
      "mean_token_accuracy": 0.19192965626716613,
      "num_tokens": 31271203.0,
      "step": 16955
    },
    {
      "entropy": 5.647221803665161,
      "epoch": 1.424868725057761,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048000086572713566,
      "loss": 5.1638,
      "mean_token_accuracy": 0.19440150260925293,
      "num_tokens": 31279812.0,
      "step": 16960
    },
    {
      "entropy": 5.609462785720825,
      "epoch": 1.4252888048729258,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004799885213718147,
      "loss": 5.1811,
      "mean_token_accuracy": 0.18293211311101915,
      "num_tokens": 31289615.0,
      "step": 16965
    },
    {
      "entropy": 5.625906658172608,
      "epoch": 1.4257088846880908,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047997617338522763,
      "loss": 5.1383,
      "mean_token_accuracy": 0.19404136389493942,
      "num_tokens": 31298947.0,
      "step": 16970
    },
    {
      "entropy": 5.642147064208984,
      "epoch": 1.4261289645032555,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047996382176759324,
      "loss": 5.1378,
      "mean_token_accuracy": 0.18950241059064865,
      "num_tokens": 31307465.0,
      "step": 16975
    },
    {
      "entropy": 5.61986780166626,
      "epoch": 1.4265490443184206,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004799514665191303,
      "loss": 5.2389,
      "mean_token_accuracy": 0.1823584407567978,
      "num_tokens": 31317682.0,
      "step": 16980
    },
    {
      "entropy": 5.674328184127807,
      "epoch": 1.4269691241335853,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004799391076400576,
      "loss": 5.2464,
      "mean_token_accuracy": 0.1848309963941574,
      "num_tokens": 31326113.0,
      "step": 16985
    },
    {
      "entropy": 5.725942850112915,
      "epoch": 1.4273892039487501,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047992674513059415,
      "loss": 5.2738,
      "mean_token_accuracy": 0.18790781795978545,
      "num_tokens": 31335263.0,
      "step": 16990
    },
    {
      "entropy": 5.70635290145874,
      "epoch": 1.4278092837639151,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047991437899095896,
      "loss": 5.2075,
      "mean_token_accuracy": 0.19285413324832917,
      "num_tokens": 31344503.0,
      "step": 16995
    },
    {
      "entropy": 5.693903064727783,
      "epoch": 1.4282293635790801,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047990200922137105,
      "loss": 5.2572,
      "mean_token_accuracy": 0.18187833130359649,
      "num_tokens": 31354530.0,
      "step": 17000
    },
    {
      "entropy": 5.639449596405029,
      "epoch": 1.428649443394245,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004798896358220496,
      "loss": 5.0958,
      "mean_token_accuracy": 0.19277971088886262,
      "num_tokens": 31362761.0,
      "step": 17005
    },
    {
      "entropy": 5.62120246887207,
      "epoch": 1.4290695232094097,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004798772587932137,
      "loss": 5.1143,
      "mean_token_accuracy": 0.18538062274456024,
      "num_tokens": 31372933.0,
      "step": 17010
    },
    {
      "entropy": 5.73318829536438,
      "epoch": 1.4294896030245747,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004798648781350826,
      "loss": 5.289,
      "mean_token_accuracy": 0.1861738756299019,
      "num_tokens": 31382651.0,
      "step": 17015
    },
    {
      "entropy": 5.671647644042968,
      "epoch": 1.4299096828397395,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004798524938478758,
      "loss": 5.2378,
      "mean_token_accuracy": 0.18281141072511672,
      "num_tokens": 31392272.0,
      "step": 17020
    },
    {
      "entropy": 5.704460716247558,
      "epoch": 1.4303297626549045,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004798401059318124,
      "loss": 5.1626,
      "mean_token_accuracy": 0.18633147478103637,
      "num_tokens": 31400684.0,
      "step": 17025
    },
    {
      "entropy": 5.674871110916138,
      "epoch": 1.4307498424700693,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004798277143871122,
      "loss": 5.1585,
      "mean_token_accuracy": 0.19512604176998138,
      "num_tokens": 31409082.0,
      "step": 17030
    },
    {
      "entropy": 5.584090089797973,
      "epoch": 1.431169922285234,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004798153192139944,
      "loss": 5.128,
      "mean_token_accuracy": 0.1931147590279579,
      "num_tokens": 31417415.0,
      "step": 17035
    },
    {
      "entropy": 5.670648908615112,
      "epoch": 1.431590002100399,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004798029204126786,
      "loss": 5.2551,
      "mean_token_accuracy": 0.1869749367237091,
      "num_tokens": 31427510.0,
      "step": 17040
    },
    {
      "entropy": 5.671906518936157,
      "epoch": 1.432010081915564,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004797905179833847,
      "loss": 5.1533,
      "mean_token_accuracy": 0.1875850513577461,
      "num_tokens": 31436187.0,
      "step": 17045
    },
    {
      "entropy": 5.698179054260254,
      "epoch": 1.432430161730729,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004797781119263321,
      "loss": 5.1447,
      "mean_token_accuracy": 0.18765630722045898,
      "num_tokens": 31445179.0,
      "step": 17050
    },
    {
      "entropy": 5.703338861465454,
      "epoch": 1.4328502415458937,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004797657022417408,
      "loss": 5.2187,
      "mean_token_accuracy": 0.18669373095035552,
      "num_tokens": 31454434.0,
      "step": 17055
    },
    {
      "entropy": 5.6224359512329105,
      "epoch": 1.4332703213610585,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047975328892983045,
      "loss": 5.2324,
      "mean_token_accuracy": 0.1895969957113266,
      "num_tokens": 31464202.0,
      "step": 17060
    },
    {
      "entropy": 5.612545967102051,
      "epoch": 1.4336904011762235,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047974087199082095,
      "loss": 5.1099,
      "mean_token_accuracy": 0.19621628969907762,
      "num_tokens": 31473158.0,
      "step": 17065
    },
    {
      "entropy": 5.634983396530151,
      "epoch": 1.4341104809913885,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047972845142493244,
      "loss": 5.1573,
      "mean_token_accuracy": 0.18245962709188462,
      "num_tokens": 31482643.0,
      "step": 17070
    },
    {
      "entropy": 5.666369485855102,
      "epoch": 1.4345305608065533,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004797160272323848,
      "loss": 5.217,
      "mean_token_accuracy": 0.18681353628635405,
      "num_tokens": 31492080.0,
      "step": 17075
    },
    {
      "entropy": 5.658629846572876,
      "epoch": 1.434950640621718,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047970359941339815,
      "loss": 5.1591,
      "mean_token_accuracy": 0.18926746994256974,
      "num_tokens": 31501990.0,
      "step": 17080
    },
    {
      "entropy": 5.678166198730469,
      "epoch": 1.435370720436883,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004796911679681926,
      "loss": 5.2415,
      "mean_token_accuracy": 0.18354712128639222,
      "num_tokens": 31510548.0,
      "step": 17085
    },
    {
      "entropy": 5.698126411437988,
      "epoch": 1.4357908002520479,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047967873289698847,
      "loss": 5.2057,
      "mean_token_accuracy": 0.18615905642509462,
      "num_tokens": 31518695.0,
      "step": 17090
    },
    {
      "entropy": 5.8190758228302,
      "epoch": 1.4362108800672129,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047966629420000595,
      "loss": 5.3189,
      "mean_token_accuracy": 0.18429006934165953,
      "num_tokens": 31528021.0,
      "step": 17095
    },
    {
      "entropy": 5.781438112258911,
      "epoch": 1.4366309598823777,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004796538518774654,
      "loss": 5.2924,
      "mean_token_accuracy": 0.1826125994324684,
      "num_tokens": 31537786.0,
      "step": 17100
    },
    {
      "entropy": 5.655714321136474,
      "epoch": 1.4370510396975424,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047964140592958725,
      "loss": 5.2456,
      "mean_token_accuracy": 0.18805236518383026,
      "num_tokens": 31548006.0,
      "step": 17105
    },
    {
      "entropy": 5.691010284423828,
      "epoch": 1.4374711195127075,
      "grad_norm": 1.09375,
      "learning_rate": 0.000479628956356592,
      "loss": 5.2046,
      "mean_token_accuracy": 0.1842111423611641,
      "num_tokens": 31557042.0,
      "step": 17110
    },
    {
      "entropy": 5.720591259002686,
      "epoch": 1.4378911993278722,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004796165031587001,
      "loss": 5.2152,
      "mean_token_accuracy": 0.18544068634510041,
      "num_tokens": 31566661.0,
      "step": 17115
    },
    {
      "entropy": 5.741116619110107,
      "epoch": 1.4383112791430372,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004796040463361323,
      "loss": 5.1835,
      "mean_token_accuracy": 0.19755398333072663,
      "num_tokens": 31575724.0,
      "step": 17120
    },
    {
      "entropy": 5.715835189819336,
      "epoch": 1.438731358958202,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004795915858891091,
      "loss": 5.2494,
      "mean_token_accuracy": 0.18857370615005492,
      "num_tokens": 31585068.0,
      "step": 17125
    },
    {
      "entropy": 5.704570055007935,
      "epoch": 1.4391514387733668,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004795791218178514,
      "loss": 5.3028,
      "mean_token_accuracy": 0.18356307446956635,
      "num_tokens": 31594629.0,
      "step": 17130
    },
    {
      "entropy": 5.6732875347137455,
      "epoch": 1.4395715185885318,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047956665412257984,
      "loss": 5.179,
      "mean_token_accuracy": 0.1890894055366516,
      "num_tokens": 31603469.0,
      "step": 17135
    },
    {
      "entropy": 5.673287868499756,
      "epoch": 1.4399915984036968,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047955418280351526,
      "loss": 5.1402,
      "mean_token_accuracy": 0.1981142595410347,
      "num_tokens": 31611674.0,
      "step": 17140
    },
    {
      "entropy": 5.784692096710205,
      "epoch": 1.4404116782188616,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004795417078608788,
      "loss": 5.3804,
      "mean_token_accuracy": 0.1801501840353012,
      "num_tokens": 31621863.0,
      "step": 17145
    },
    {
      "entropy": 5.764338588714599,
      "epoch": 1.4408317580340264,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047952922929489126,
      "loss": 5.22,
      "mean_token_accuracy": 0.18685618191957473,
      "num_tokens": 31630968.0,
      "step": 17150
    },
    {
      "entropy": 5.69820351600647,
      "epoch": 1.4412518378491914,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047951674710577366,
      "loss": 5.242,
      "mean_token_accuracy": 0.18607798367738723,
      "num_tokens": 31640643.0,
      "step": 17155
    },
    {
      "entropy": 5.5981019020080565,
      "epoch": 1.4416719176643562,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047950426129374723,
      "loss": 5.1487,
      "mean_token_accuracy": 0.192959064245224,
      "num_tokens": 31648941.0,
      "step": 17160
    },
    {
      "entropy": 5.682616329193115,
      "epoch": 1.4420919974795212,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047949177185903314,
      "loss": 5.2359,
      "mean_token_accuracy": 0.19235930144786834,
      "num_tokens": 31658019.0,
      "step": 17165
    },
    {
      "entropy": 5.722033548355102,
      "epoch": 1.442512077294686,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004794792788018526,
      "loss": 5.2765,
      "mean_token_accuracy": 0.17878874093294145,
      "num_tokens": 31668050.0,
      "step": 17170
    },
    {
      "entropy": 5.667446899414062,
      "epoch": 1.4429321571098508,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000479466782122427,
      "loss": 5.1237,
      "mean_token_accuracy": 0.18951182812452316,
      "num_tokens": 31676727.0,
      "step": 17175
    },
    {
      "entropy": 5.678989410400391,
      "epoch": 1.4433522369250158,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047945428182097756,
      "loss": 5.2296,
      "mean_token_accuracy": 0.17958298623561858,
      "num_tokens": 31686205.0,
      "step": 17180
    },
    {
      "entropy": 5.690536642074585,
      "epoch": 1.4437723167401806,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047944177789772583,
      "loss": 5.2519,
      "mean_token_accuracy": 0.18357335329055785,
      "num_tokens": 31695521.0,
      "step": 17185
    },
    {
      "entropy": 5.799330854415894,
      "epoch": 1.4441923965553456,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004794292703528932,
      "loss": 5.2742,
      "mean_token_accuracy": 0.1733855739235878,
      "num_tokens": 31706606.0,
      "step": 17190
    },
    {
      "entropy": 5.720072078704834,
      "epoch": 1.4446124763705104,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047941675918670133,
      "loss": 5.3628,
      "mean_token_accuracy": 0.1763266794383526,
      "num_tokens": 31716881.0,
      "step": 17195
    },
    {
      "entropy": 5.662317419052124,
      "epoch": 1.4450325561856752,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004794042443993719,
      "loss": 5.1701,
      "mean_token_accuracy": 0.1836187794804573,
      "num_tokens": 31725878.0,
      "step": 17200
    },
    {
      "entropy": 5.6713676929473875,
      "epoch": 1.4454526360008402,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004793917259911265,
      "loss": 5.2433,
      "mean_token_accuracy": 0.18535358905792237,
      "num_tokens": 31735033.0,
      "step": 17205
    },
    {
      "entropy": 5.6287740707397464,
      "epoch": 1.445872715816005,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004793792039621869,
      "loss": 5.1938,
      "mean_token_accuracy": 0.18655352890491486,
      "num_tokens": 31744887.0,
      "step": 17210
    },
    {
      "entropy": 5.731176042556763,
      "epoch": 1.44629279563117,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047936667831277504,
      "loss": 5.2669,
      "mean_token_accuracy": 0.17561186999082565,
      "num_tokens": 31754137.0,
      "step": 17215
    },
    {
      "entropy": 5.730813121795654,
      "epoch": 1.4467128754463348,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004793541490431126,
      "loss": 5.088,
      "mean_token_accuracy": 0.19306415617465972,
      "num_tokens": 31763394.0,
      "step": 17220
    },
    {
      "entropy": 5.729691839218139,
      "epoch": 1.4471329552614998,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004793416161534216,
      "loss": 5.2183,
      "mean_token_accuracy": 0.18121858835220336,
      "num_tokens": 31771905.0,
      "step": 17225
    },
    {
      "entropy": 5.566930294036865,
      "epoch": 1.4475530350766646,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047932907964392423,
      "loss": 5.0495,
      "mean_token_accuracy": 0.20088715404272078,
      "num_tokens": 31780788.0,
      "step": 17230
    },
    {
      "entropy": 5.7301013469696045,
      "epoch": 1.4479731148918296,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047931653951484234,
      "loss": 5.2239,
      "mean_token_accuracy": 0.17787783443927765,
      "num_tokens": 31790198.0,
      "step": 17235
    },
    {
      "entropy": 5.691083860397339,
      "epoch": 1.4483931947069943,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047930399576639815,
      "loss": 5.2099,
      "mean_token_accuracy": 0.19164225906133653,
      "num_tokens": 31799396.0,
      "step": 17240
    },
    {
      "entropy": 5.631149053573608,
      "epoch": 1.4488132745221591,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047929144839881386,
      "loss": 5.0716,
      "mean_token_accuracy": 0.20279514491558076,
      "num_tokens": 31807680.0,
      "step": 17245
    },
    {
      "entropy": 5.7048932075500485,
      "epoch": 1.4492333543373241,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047927889741231186,
      "loss": 5.2163,
      "mean_token_accuracy": 0.18269665986299516,
      "num_tokens": 31817406.0,
      "step": 17250
    },
    {
      "entropy": 5.698895454406738,
      "epoch": 1.449653434152489,
      "grad_norm": 1.125,
      "learning_rate": 0.00047926634280711435,
      "loss": 5.201,
      "mean_token_accuracy": 0.1862025037407875,
      "num_tokens": 31826518.0,
      "step": 17255
    },
    {
      "entropy": 5.688116073608398,
      "epoch": 1.450073513967654,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004792537845834437,
      "loss": 5.2728,
      "mean_token_accuracy": 0.1805024579167366,
      "num_tokens": 31835538.0,
      "step": 17260
    },
    {
      "entropy": 5.65452094078064,
      "epoch": 1.4504935937828187,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004792412227415224,
      "loss": 5.1454,
      "mean_token_accuracy": 0.19079507887363434,
      "num_tokens": 31844899.0,
      "step": 17265
    },
    {
      "entropy": 5.699429893493653,
      "epoch": 1.4509136735979835,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047922865728157314,
      "loss": 5.2038,
      "mean_token_accuracy": 0.1905255511403084,
      "num_tokens": 31854322.0,
      "step": 17270
    },
    {
      "entropy": 5.641614198684692,
      "epoch": 1.4513337534131485,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004792160882038183,
      "loss": 5.1452,
      "mean_token_accuracy": 0.18870719373226166,
      "num_tokens": 31863657.0,
      "step": 17275
    },
    {
      "entropy": 5.653892564773559,
      "epoch": 1.4517538332283133,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004792035155084806,
      "loss": 5.1284,
      "mean_token_accuracy": 0.18968206644058228,
      "num_tokens": 31873468.0,
      "step": 17280
    },
    {
      "entropy": 5.68013710975647,
      "epoch": 1.4521739130434783,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047919093919578283,
      "loss": 5.2465,
      "mean_token_accuracy": 0.1853806793689728,
      "num_tokens": 31882391.0,
      "step": 17285
    },
    {
      "entropy": 5.670394039154052,
      "epoch": 1.452593992858643,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004791783592659476,
      "loss": 5.2674,
      "mean_token_accuracy": 0.18263329416513444,
      "num_tokens": 31891370.0,
      "step": 17290
    },
    {
      "entropy": 5.637481307983398,
      "epoch": 1.4530140726738079,
      "grad_norm": 1.078125,
      "learning_rate": 0.000479165775719198,
      "loss": 5.1834,
      "mean_token_accuracy": 0.18398721814155578,
      "num_tokens": 31900688.0,
      "step": 17295
    },
    {
      "entropy": 5.7036802768707275,
      "epoch": 1.453434152488973,
      "grad_norm": 1.890625,
      "learning_rate": 0.00047915318855575674,
      "loss": 5.204,
      "mean_token_accuracy": 0.19701419323682784,
      "num_tokens": 31909359.0,
      "step": 17300
    },
    {
      "entropy": 5.657992267608643,
      "epoch": 1.453854232304138,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047914059777584686,
      "loss": 5.1888,
      "mean_token_accuracy": 0.19062976390123368,
      "num_tokens": 31918529.0,
      "step": 17305
    },
    {
      "entropy": 5.672594785690308,
      "epoch": 1.4542743121193027,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047912800337969144,
      "loss": 5.2571,
      "mean_token_accuracy": 0.18015563488006592,
      "num_tokens": 31928310.0,
      "step": 17310
    },
    {
      "entropy": 5.707254838943482,
      "epoch": 1.4546943919344675,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047911540536751355,
      "loss": 5.1635,
      "mean_token_accuracy": 0.1872883766889572,
      "num_tokens": 31937077.0,
      "step": 17315
    },
    {
      "entropy": 5.705650424957275,
      "epoch": 1.4551144717496325,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004791028037395363,
      "loss": 5.233,
      "mean_token_accuracy": 0.17971020340919494,
      "num_tokens": 31946023.0,
      "step": 17320
    },
    {
      "entropy": 5.602953910827637,
      "epoch": 1.4555345515647973,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047909019849598305,
      "loss": 5.0776,
      "mean_token_accuracy": 0.19422320723533631,
      "num_tokens": 31954741.0,
      "step": 17325
    },
    {
      "entropy": 5.639042568206787,
      "epoch": 1.4559546313799623,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047907758963707696,
      "loss": 5.1675,
      "mean_token_accuracy": 0.1882924422621727,
      "num_tokens": 31963516.0,
      "step": 17330
    },
    {
      "entropy": 5.685782861709595,
      "epoch": 1.456374711195127,
      "grad_norm": 1.125,
      "learning_rate": 0.00047906497716304153,
      "loss": 5.2097,
      "mean_token_accuracy": 0.18441124260425568,
      "num_tokens": 31971917.0,
      "step": 17335
    },
    {
      "entropy": 5.741246271133423,
      "epoch": 1.4567947910102919,
      "grad_norm": 1.125,
      "learning_rate": 0.0004790523610741001,
      "loss": 5.2652,
      "mean_token_accuracy": 0.18316615074872972,
      "num_tokens": 31980718.0,
      "step": 17340
    },
    {
      "entropy": 5.750768852233887,
      "epoch": 1.4572148708254569,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047903974137047614,
      "loss": 5.2035,
      "mean_token_accuracy": 0.187534798681736,
      "num_tokens": 31988664.0,
      "step": 17345
    },
    {
      "entropy": 5.706573724746704,
      "epoch": 1.4576349506406217,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047902711805239325,
      "loss": 5.267,
      "mean_token_accuracy": 0.18419413417577743,
      "num_tokens": 31998415.0,
      "step": 17350
    },
    {
      "entropy": 5.734539461135864,
      "epoch": 1.4580550304557867,
      "grad_norm": 1.0,
      "learning_rate": 0.00047901449112007494,
      "loss": 5.2751,
      "mean_token_accuracy": 0.1885323092341423,
      "num_tokens": 32007915.0,
      "step": 17355
    },
    {
      "entropy": 5.684205818176269,
      "epoch": 1.4584751102709514,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047900186057374514,
      "loss": 5.203,
      "mean_token_accuracy": 0.18459141850471497,
      "num_tokens": 32016582.0,
      "step": 17360
    },
    {
      "entropy": 5.660382843017578,
      "epoch": 1.4588951900861162,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047898922641362724,
      "loss": 5.1999,
      "mean_token_accuracy": 0.18248110711574556,
      "num_tokens": 32026008.0,
      "step": 17365
    },
    {
      "entropy": 5.718503999710083,
      "epoch": 1.4593152699012812,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004789765886399453,
      "loss": 5.2532,
      "mean_token_accuracy": 0.18320031017065047,
      "num_tokens": 32034554.0,
      "step": 17370
    },
    {
      "entropy": 5.784012985229492,
      "epoch": 1.4597353497164463,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047896394725292313,
      "loss": 5.2361,
      "mean_token_accuracy": 0.19398827999830245,
      "num_tokens": 32044003.0,
      "step": 17375
    },
    {
      "entropy": 5.686049604415894,
      "epoch": 1.460155429531611,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047895130225278473,
      "loss": 5.2156,
      "mean_token_accuracy": 0.18879769146442413,
      "num_tokens": 32053753.0,
      "step": 17380
    },
    {
      "entropy": 5.64119963645935,
      "epoch": 1.4605755093467758,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004789386536397539,
      "loss": 5.217,
      "mean_token_accuracy": 0.18385226726531984,
      "num_tokens": 32062459.0,
      "step": 17385
    },
    {
      "entropy": 5.729079532623291,
      "epoch": 1.4609955891619408,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004789260014140549,
      "loss": 5.2902,
      "mean_token_accuracy": 0.1848796397447586,
      "num_tokens": 32072544.0,
      "step": 17390
    },
    {
      "entropy": 5.722535419464111,
      "epoch": 1.4614156689771056,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047891334557591177,
      "loss": 5.2384,
      "mean_token_accuracy": 0.18093919903039932,
      "num_tokens": 32082015.0,
      "step": 17395
    },
    {
      "entropy": 5.651052761077881,
      "epoch": 1.4618357487922706,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004789006861255488,
      "loss": 5.1622,
      "mean_token_accuracy": 0.19277044087648393,
      "num_tokens": 32091622.0,
      "step": 17400
    },
    {
      "entropy": 5.734331703186035,
      "epoch": 1.4622558286074354,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004788880230631901,
      "loss": 5.3272,
      "mean_token_accuracy": 0.1776832416653633,
      "num_tokens": 32102716.0,
      "step": 17405
    },
    {
      "entropy": 5.743254804611206,
      "epoch": 1.4626759084226002,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047887535638906005,
      "loss": 5.1263,
      "mean_token_accuracy": 0.19768501520156861,
      "num_tokens": 32111051.0,
      "step": 17410
    },
    {
      "entropy": 5.611501169204712,
      "epoch": 1.4630959882377652,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000478862686103383,
      "loss": 5.1338,
      "mean_token_accuracy": 0.19645296037197113,
      "num_tokens": 32119781.0,
      "step": 17415
    },
    {
      "entropy": 5.664831924438476,
      "epoch": 1.46351606805293,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047885001220638354,
      "loss": 5.2283,
      "mean_token_accuracy": 0.18855584859848024,
      "num_tokens": 32128849.0,
      "step": 17420
    },
    {
      "entropy": 5.740883636474609,
      "epoch": 1.463936147868095,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047883733469828604,
      "loss": 5.2627,
      "mean_token_accuracy": 0.18940636962652208,
      "num_tokens": 32138046.0,
      "step": 17425
    },
    {
      "entropy": 5.8068986415863035,
      "epoch": 1.4643562276832598,
      "grad_norm": 1.0,
      "learning_rate": 0.00047882465357931516,
      "loss": 5.3076,
      "mean_token_accuracy": 0.18139575570821762,
      "num_tokens": 32147994.0,
      "step": 17430
    },
    {
      "entropy": 5.755560350418091,
      "epoch": 1.4647763074984246,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004788119688496954,
      "loss": 5.2574,
      "mean_token_accuracy": 0.18773732781410218,
      "num_tokens": 32156835.0,
      "step": 17435
    },
    {
      "entropy": 5.6850920677185055,
      "epoch": 1.4651963873135896,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004787992805096516,
      "loss": 5.2024,
      "mean_token_accuracy": 0.18811201453208923,
      "num_tokens": 32166751.0,
      "step": 17440
    },
    {
      "entropy": 5.69700517654419,
      "epoch": 1.4656164671287546,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047878658855940855,
      "loss": 5.2946,
      "mean_token_accuracy": 0.17590508460998536,
      "num_tokens": 32175705.0,
      "step": 17445
    },
    {
      "entropy": 5.778347682952881,
      "epoch": 1.4660365469439194,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004787738929991909,
      "loss": 5.3266,
      "mean_token_accuracy": 0.1763791933655739,
      "num_tokens": 32185404.0,
      "step": 17450
    },
    {
      "entropy": 5.721336698532104,
      "epoch": 1.4664566267590842,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047876119382922374,
      "loss": 5.2074,
      "mean_token_accuracy": 0.19106389433145524,
      "num_tokens": 32194054.0,
      "step": 17455
    },
    {
      "entropy": 5.666349267959594,
      "epoch": 1.4668767065742492,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047874849104973194,
      "loss": 5.2855,
      "mean_token_accuracy": 0.1720316082239151,
      "num_tokens": 32204080.0,
      "step": 17460
    },
    {
      "entropy": 5.695929527282715,
      "epoch": 1.467296786389414,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047873578466094054,
      "loss": 5.1811,
      "mean_token_accuracy": 0.18279954195022582,
      "num_tokens": 32213279.0,
      "step": 17465
    },
    {
      "entropy": 5.66992712020874,
      "epoch": 1.467716866204579,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004787230746630746,
      "loss": 5.1948,
      "mean_token_accuracy": 0.18772552460432052,
      "num_tokens": 32221668.0,
      "step": 17470
    },
    {
      "entropy": 5.702330017089844,
      "epoch": 1.4681369460197438,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004787103610563593,
      "loss": 5.1378,
      "mean_token_accuracy": 0.1922280713915825,
      "num_tokens": 32229683.0,
      "step": 17475
    },
    {
      "entropy": 5.694551277160644,
      "epoch": 1.4685570258349085,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047869764384101993,
      "loss": 5.2042,
      "mean_token_accuracy": 0.18950177133083343,
      "num_tokens": 32238948.0,
      "step": 17480
    },
    {
      "entropy": 5.730488300323486,
      "epoch": 1.4689771056500736,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047868492301728164,
      "loss": 5.2291,
      "mean_token_accuracy": 0.18351960927248,
      "num_tokens": 32248079.0,
      "step": 17485
    },
    {
      "entropy": 5.670637130737305,
      "epoch": 1.4693971854652383,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047867219858536975,
      "loss": 5.0826,
      "mean_token_accuracy": 0.196705362200737,
      "num_tokens": 32256413.0,
      "step": 17490
    },
    {
      "entropy": 5.683293056488037,
      "epoch": 1.4698172652804034,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004786594705455098,
      "loss": 5.22,
      "mean_token_accuracy": 0.17982027530670167,
      "num_tokens": 32265954.0,
      "step": 17495
    },
    {
      "entropy": 5.677178382873535,
      "epoch": 1.4702373450955681,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004786467388979272,
      "loss": 5.1446,
      "mean_token_accuracy": 0.18972351998090745,
      "num_tokens": 32273817.0,
      "step": 17500
    },
    {
      "entropy": 5.670801830291748,
      "epoch": 1.470657424910733,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047863400364284744,
      "loss": 5.1906,
      "mean_token_accuracy": 0.1893660232424736,
      "num_tokens": 32283025.0,
      "step": 17505
    },
    {
      "entropy": 5.660158634185791,
      "epoch": 1.471077504725898,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047862126478049623,
      "loss": 5.177,
      "mean_token_accuracy": 0.19247282147407532,
      "num_tokens": 32292321.0,
      "step": 17510
    },
    {
      "entropy": 5.754831218719483,
      "epoch": 1.4714975845410627,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047860852231109915,
      "loss": 5.2629,
      "mean_token_accuracy": 0.17347700744867325,
      "num_tokens": 32302203.0,
      "step": 17515
    },
    {
      "entropy": 5.636993265151977,
      "epoch": 1.4719176643562277,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004785957762348819,
      "loss": 5.1243,
      "mean_token_accuracy": 0.19010509550571442,
      "num_tokens": 32310893.0,
      "step": 17520
    },
    {
      "entropy": 5.6034149646759035,
      "epoch": 1.4723377441713925,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004785830265520703,
      "loss": 5.1471,
      "mean_token_accuracy": 0.19206166118383408,
      "num_tokens": 32320320.0,
      "step": 17525
    },
    {
      "entropy": 5.604295492172241,
      "epoch": 1.4727578239865575,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047857027326289023,
      "loss": 5.094,
      "mean_token_accuracy": 0.19663815796375275,
      "num_tokens": 32329196.0,
      "step": 17530
    },
    {
      "entropy": 5.693138313293457,
      "epoch": 1.4731779038017223,
      "grad_norm": 1.0,
      "learning_rate": 0.00047855751636756763,
      "loss": 5.211,
      "mean_token_accuracy": 0.18184043616056442,
      "num_tokens": 32338529.0,
      "step": 17535
    },
    {
      "entropy": 5.698631620407104,
      "epoch": 1.4735979836168873,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004785447558663284,
      "loss": 5.2077,
      "mean_token_accuracy": 0.1898406371474266,
      "num_tokens": 32347114.0,
      "step": 17540
    },
    {
      "entropy": 5.767664861679077,
      "epoch": 1.474018063432052,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047853199175939865,
      "loss": 5.3556,
      "mean_token_accuracy": 0.1817089319229126,
      "num_tokens": 32356765.0,
      "step": 17545
    },
    {
      "entropy": 5.816174793243408,
      "epoch": 1.474438143247217,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004785192240470045,
      "loss": 5.326,
      "mean_token_accuracy": 0.1815305233001709,
      "num_tokens": 32366175.0,
      "step": 17550
    },
    {
      "entropy": 5.679656839370727,
      "epoch": 1.474858223062382,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000478506452729372,
      "loss": 5.1118,
      "mean_token_accuracy": 0.19194453358650207,
      "num_tokens": 32375063.0,
      "step": 17555
    },
    {
      "entropy": 5.638524723052979,
      "epoch": 1.4752783028775467,
      "grad_norm": 1.125,
      "learning_rate": 0.00047849367780672755,
      "loss": 5.1743,
      "mean_token_accuracy": 0.1911807492375374,
      "num_tokens": 32384596.0,
      "step": 17560
    },
    {
      "entropy": 5.632683563232422,
      "epoch": 1.4756983826927117,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004784808992792974,
      "loss": 5.1512,
      "mean_token_accuracy": 0.18825320899486542,
      "num_tokens": 32393489.0,
      "step": 17565
    },
    {
      "entropy": 5.713040018081665,
      "epoch": 1.4761184625078765,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004784681171473079,
      "loss": 5.1423,
      "mean_token_accuracy": 0.19032877385616304,
      "num_tokens": 32402192.0,
      "step": 17570
    },
    {
      "entropy": 5.746954107284546,
      "epoch": 1.4765385423230413,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047845533141098543,
      "loss": 5.2148,
      "mean_token_accuracy": 0.18130073249340056,
      "num_tokens": 32411317.0,
      "step": 17575
    },
    {
      "entropy": 5.75069580078125,
      "epoch": 1.4769586221382063,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004784425420705565,
      "loss": 5.2935,
      "mean_token_accuracy": 0.18136326372623443,
      "num_tokens": 32420308.0,
      "step": 17580
    },
    {
      "entropy": 5.624954795837402,
      "epoch": 1.477378701953371,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004784297491262477,
      "loss": 5.2304,
      "mean_token_accuracy": 0.18571205735206603,
      "num_tokens": 32429532.0,
      "step": 17585
    },
    {
      "entropy": 5.714317464828492,
      "epoch": 1.477798781768536,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004784169525782858,
      "loss": 5.2017,
      "mean_token_accuracy": 0.18757518082857133,
      "num_tokens": 32439382.0,
      "step": 17590
    },
    {
      "entropy": 5.713875198364258,
      "epoch": 1.4782188615837009,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004784041524268971,
      "loss": 5.189,
      "mean_token_accuracy": 0.1898956224322319,
      "num_tokens": 32447893.0,
      "step": 17595
    },
    {
      "entropy": 5.6334796905517575,
      "epoch": 1.4786389413988656,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047839134867230874,
      "loss": 5.1839,
      "mean_token_accuracy": 0.1923775404691696,
      "num_tokens": 32457770.0,
      "step": 17600
    },
    {
      "entropy": 5.719471073150634,
      "epoch": 1.4790590212140307,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047837854131474726,
      "loss": 5.2914,
      "mean_token_accuracy": 0.18435123711824417,
      "num_tokens": 32467247.0,
      "step": 17605
    },
    {
      "entropy": 5.783801555633545,
      "epoch": 1.4794791010291957,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047836573035443976,
      "loss": 5.2677,
      "mean_token_accuracy": 0.18225768357515335,
      "num_tokens": 32477453.0,
      "step": 17610
    },
    {
      "entropy": 5.774486017227173,
      "epoch": 1.4798991808443605,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047835291579161293,
      "loss": 5.2461,
      "mean_token_accuracy": 0.18947134912014008,
      "num_tokens": 32486278.0,
      "step": 17615
    },
    {
      "entropy": 5.64484601020813,
      "epoch": 1.4803192606595252,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004783400976264941,
      "loss": 5.1745,
      "mean_token_accuracy": 0.19212258756160736,
      "num_tokens": 32495523.0,
      "step": 17620
    },
    {
      "entropy": 5.7044206142425535,
      "epoch": 1.4807393404746902,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047832727585930997,
      "loss": 5.208,
      "mean_token_accuracy": 0.18657382279634477,
      "num_tokens": 32504952.0,
      "step": 17625
    },
    {
      "entropy": 5.6819816589355465,
      "epoch": 1.481159420289855,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004783144504902879,
      "loss": 5.1797,
      "mean_token_accuracy": 0.18655804246664048,
      "num_tokens": 32515620.0,
      "step": 17630
    },
    {
      "entropy": 5.624293279647827,
      "epoch": 1.48157950010502,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000478301621519655,
      "loss": 5.1644,
      "mean_token_accuracy": 0.1925906077027321,
      "num_tokens": 32524549.0,
      "step": 17635
    },
    {
      "entropy": 5.683595991134643,
      "epoch": 1.4819995799201848,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004782887889476386,
      "loss": 5.0542,
      "mean_token_accuracy": 0.19920557737350464,
      "num_tokens": 32533043.0,
      "step": 17640
    },
    {
      "entropy": 5.69859824180603,
      "epoch": 1.4824196597353496,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000478275952774466,
      "loss": 5.1597,
      "mean_token_accuracy": 0.18621080964803696,
      "num_tokens": 32541679.0,
      "step": 17645
    },
    {
      "entropy": 5.732716035842896,
      "epoch": 1.4828397395505146,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004782631130003646,
      "loss": 5.2911,
      "mean_token_accuracy": 0.18713412135839463,
      "num_tokens": 32550922.0,
      "step": 17650
    },
    {
      "entropy": 5.700401544570923,
      "epoch": 1.4832598193656794,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004782502696255617,
      "loss": 5.2771,
      "mean_token_accuracy": 0.18194049447774888,
      "num_tokens": 32560063.0,
      "step": 17655
    },
    {
      "entropy": 5.629743432998657,
      "epoch": 1.4836798991808444,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047823742265028495,
      "loss": 5.1583,
      "mean_token_accuracy": 0.19046991616487502,
      "num_tokens": 32569476.0,
      "step": 17660
    },
    {
      "entropy": 5.712680339813232,
      "epoch": 1.4840999789960092,
      "grad_norm": 1.03125,
      "learning_rate": 0.000478224572074762,
      "loss": 5.2063,
      "mean_token_accuracy": 0.1944539800286293,
      "num_tokens": 32578552.0,
      "step": 17665
    },
    {
      "entropy": 5.726285552978515,
      "epoch": 1.484520058811174,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004782117178992203,
      "loss": 5.2036,
      "mean_token_accuracy": 0.19127248525619506,
      "num_tokens": 32589074.0,
      "step": 17670
    },
    {
      "entropy": 5.6908238410949705,
      "epoch": 1.484940138626339,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004781988601238878,
      "loss": 5.2341,
      "mean_token_accuracy": 0.1834852710366249,
      "num_tokens": 32599288.0,
      "step": 17675
    },
    {
      "entropy": 5.7375006675720215,
      "epoch": 1.485360218441504,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000478185998748992,
      "loss": 5.2917,
      "mean_token_accuracy": 0.1794963300228119,
      "num_tokens": 32609430.0,
      "step": 17680
    },
    {
      "entropy": 5.636043214797974,
      "epoch": 1.4857802982566688,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047817313377476083,
      "loss": 5.139,
      "mean_token_accuracy": 0.19061165452003478,
      "num_tokens": 32617763.0,
      "step": 17685
    },
    {
      "entropy": 5.645914888381958,
      "epoch": 1.4862003780718336,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047816026520142234,
      "loss": 5.2095,
      "mean_token_accuracy": 0.1839400574564934,
      "num_tokens": 32627465.0,
      "step": 17690
    },
    {
      "entropy": 5.749820232391357,
      "epoch": 1.4866204578869986,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004781473930292043,
      "loss": 5.1389,
      "mean_token_accuracy": 0.19504829198122026,
      "num_tokens": 32635984.0,
      "step": 17695
    },
    {
      "entropy": 5.615261793136597,
      "epoch": 1.4870405377021634,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004781345172583348,
      "loss": 5.0817,
      "mean_token_accuracy": 0.19573403149843216,
      "num_tokens": 32644346.0,
      "step": 17700
    },
    {
      "entropy": 5.628323221206665,
      "epoch": 1.4874606175173284,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047812163788904196,
      "loss": 5.2001,
      "mean_token_accuracy": 0.18402860462665557,
      "num_tokens": 32654118.0,
      "step": 17705
    },
    {
      "entropy": 5.706335258483887,
      "epoch": 1.4878806973324932,
      "grad_norm": 1.0,
      "learning_rate": 0.00047810875492155386,
      "loss": 5.2045,
      "mean_token_accuracy": 0.19060197174549104,
      "num_tokens": 32664258.0,
      "step": 17710
    },
    {
      "entropy": 5.689349603652954,
      "epoch": 1.488300777147658,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004780958683560987,
      "loss": 5.2788,
      "mean_token_accuracy": 0.17915300279855728,
      "num_tokens": 32673672.0,
      "step": 17715
    },
    {
      "entropy": 5.693923473358154,
      "epoch": 1.488720856962823,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004780829781929049,
      "loss": 5.2551,
      "mean_token_accuracy": 0.176225508749485,
      "num_tokens": 32682901.0,
      "step": 17720
    },
    {
      "entropy": 5.741204261779785,
      "epoch": 1.4891409367779878,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004780700844322007,
      "loss": 5.1719,
      "mean_token_accuracy": 0.1943327158689499,
      "num_tokens": 32691384.0,
      "step": 17725
    },
    {
      "entropy": 5.66629810333252,
      "epoch": 1.4895610165931528,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047805718707421446,
      "loss": 5.2214,
      "mean_token_accuracy": 0.18961918503046035,
      "num_tokens": 32700758.0,
      "step": 17730
    },
    {
      "entropy": 5.757991123199463,
      "epoch": 1.4899810964083176,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047804428611917475,
      "loss": 5.328,
      "mean_token_accuracy": 0.1790391057729721,
      "num_tokens": 32709676.0,
      "step": 17735
    },
    {
      "entropy": 5.728940963745117,
      "epoch": 1.4904011762234823,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047803138156731,
      "loss": 5.2225,
      "mean_token_accuracy": 0.18449746817350388,
      "num_tokens": 32718102.0,
      "step": 17740
    },
    {
      "entropy": 5.7407714366912845,
      "epoch": 1.4908212560386473,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047801847341884897,
      "loss": 5.2126,
      "mean_token_accuracy": 0.18912291377782822,
      "num_tokens": 32727356.0,
      "step": 17745
    },
    {
      "entropy": 5.702575635910034,
      "epoch": 1.4912413358538124,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004780055616740202,
      "loss": 5.201,
      "mean_token_accuracy": 0.18499555438756943,
      "num_tokens": 32736605.0,
      "step": 17750
    },
    {
      "entropy": 5.653643321990967,
      "epoch": 1.4916614156689771,
      "grad_norm": 1.0,
      "learning_rate": 0.0004779926463330524,
      "loss": 5.1481,
      "mean_token_accuracy": 0.1887960359454155,
      "num_tokens": 32745573.0,
      "step": 17755
    },
    {
      "entropy": 5.656874179840088,
      "epoch": 1.492081495484142,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004779797273961744,
      "loss": 5.2077,
      "mean_token_accuracy": 0.1881374716758728,
      "num_tokens": 32755695.0,
      "step": 17760
    },
    {
      "entropy": 5.647112655639648,
      "epoch": 1.492501575299307,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004779668048636151,
      "loss": 5.135,
      "mean_token_accuracy": 0.1951678141951561,
      "num_tokens": 32763570.0,
      "step": 17765
    },
    {
      "entropy": 5.688737249374389,
      "epoch": 1.4929216551144717,
      "grad_norm": 1.125,
      "learning_rate": 0.00047795387873560336,
      "loss": 5.2227,
      "mean_token_accuracy": 0.18130321204662322,
      "num_tokens": 32772006.0,
      "step": 17770
    },
    {
      "entropy": 5.695094203948974,
      "epoch": 1.4933417349296367,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004779409490123681,
      "loss": 5.1809,
      "mean_token_accuracy": 0.18205158710479735,
      "num_tokens": 32781080.0,
      "step": 17775
    },
    {
      "entropy": 5.650719881057739,
      "epoch": 1.4937618147448015,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004779280156941384,
      "loss": 5.1391,
      "mean_token_accuracy": 0.1863284394145012,
      "num_tokens": 32789880.0,
      "step": 17780
    },
    {
      "entropy": 5.728087997436523,
      "epoch": 1.4941818945599663,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047791507878114354,
      "loss": 5.1869,
      "mean_token_accuracy": 0.18853655308485032,
      "num_tokens": 32799222.0,
      "step": 17785
    },
    {
      "entropy": 5.659189653396607,
      "epoch": 1.4946019743751313,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004779021382736124,
      "loss": 5.1757,
      "mean_token_accuracy": 0.18746853321790696,
      "num_tokens": 32808945.0,
      "step": 17790
    },
    {
      "entropy": 5.607034921646118,
      "epoch": 1.495022054190296,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004778891941717745,
      "loss": 5.088,
      "mean_token_accuracy": 0.19595674723386763,
      "num_tokens": 32818386.0,
      "step": 17795
    },
    {
      "entropy": 5.601575899124145,
      "epoch": 1.495442134005461,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004778762464758589,
      "loss": 5.1687,
      "mean_token_accuracy": 0.17886970043182374,
      "num_tokens": 32828364.0,
      "step": 17800
    },
    {
      "entropy": 5.77184567451477,
      "epoch": 1.495862213820626,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047786329518609505,
      "loss": 5.2851,
      "mean_token_accuracy": 0.1866486445069313,
      "num_tokens": 32837399.0,
      "step": 17805
    },
    {
      "entropy": 5.675094032287598,
      "epoch": 1.4962822936357907,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047785034030271243,
      "loss": 5.1162,
      "mean_token_accuracy": 0.19946875870227815,
      "num_tokens": 32846111.0,
      "step": 17810
    },
    {
      "entropy": 5.634919786453247,
      "epoch": 1.4967023734509557,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004778373818259404,
      "loss": 5.0445,
      "mean_token_accuracy": 0.20159802734851837,
      "num_tokens": 32855839.0,
      "step": 17815
    },
    {
      "entropy": 5.758076477050781,
      "epoch": 1.4971224532661207,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047782441975600866,
      "loss": 5.3171,
      "mean_token_accuracy": 0.18615912199020385,
      "num_tokens": 32865946.0,
      "step": 17820
    },
    {
      "entropy": 5.765147733688354,
      "epoch": 1.4975425330812855,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004778114540931468,
      "loss": 5.2771,
      "mean_token_accuracy": 0.18547076880931854,
      "num_tokens": 32875310.0,
      "step": 17825
    },
    {
      "entropy": 5.683182716369629,
      "epoch": 1.4979626128964503,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047779848483758445,
      "loss": 5.2386,
      "mean_token_accuracy": 0.1867612838745117,
      "num_tokens": 32885315.0,
      "step": 17830
    },
    {
      "entropy": 5.688684415817261,
      "epoch": 1.4983826927116153,
      "grad_norm": 1.125,
      "learning_rate": 0.00047778551198955133,
      "loss": 5.2002,
      "mean_token_accuracy": 0.19151496291160583,
      "num_tokens": 32894055.0,
      "step": 17835
    },
    {
      "entropy": 5.682215213775635,
      "epoch": 1.49880277252678,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004777725355492773,
      "loss": 5.216,
      "mean_token_accuracy": 0.19545311331748963,
      "num_tokens": 32903030.0,
      "step": 17840
    },
    {
      "entropy": 5.684957504272461,
      "epoch": 1.499222852341945,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004777595555169922,
      "loss": 5.1535,
      "mean_token_accuracy": 0.19222939610481263,
      "num_tokens": 32911562.0,
      "step": 17845
    },
    {
      "entropy": 5.705116319656372,
      "epoch": 1.4996429321571099,
      "grad_norm": 1.03125,
      "learning_rate": 0.000477746571892926,
      "loss": 5.2636,
      "mean_token_accuracy": 0.18123303055763246,
      "num_tokens": 32920376.0,
      "step": 17850
    },
    {
      "entropy": 5.736515951156616,
      "epoch": 1.5000630119722747,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004777335846773087,
      "loss": 5.2006,
      "mean_token_accuracy": 0.18063124865293503,
      "num_tokens": 32929374.0,
      "step": 17855
    },
    {
      "entropy": 5.624841403961182,
      "epoch": 1.5004830917874397,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047772059387037025,
      "loss": 5.1194,
      "mean_token_accuracy": 0.18409956395626068,
      "num_tokens": 32938695.0,
      "step": 17860
    },
    {
      "entropy": 5.762369823455811,
      "epoch": 1.5009031716026044,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004777075994723409,
      "loss": 5.2048,
      "mean_token_accuracy": 0.19374118000268936,
      "num_tokens": 32947725.0,
      "step": 17865
    },
    {
      "entropy": 5.712083864212036,
      "epoch": 1.5013232514177695,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047769460148345085,
      "loss": 5.2012,
      "mean_token_accuracy": 0.17866237461566925,
      "num_tokens": 32957017.0,
      "step": 17870
    },
    {
      "entropy": 5.65272011756897,
      "epoch": 1.5017433312329342,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004776815999039303,
      "loss": 5.181,
      "mean_token_accuracy": 0.18819639682769776,
      "num_tokens": 32965944.0,
      "step": 17875
    },
    {
      "entropy": 5.6268964290618895,
      "epoch": 1.502163411048099,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004776685947340096,
      "loss": 5.1822,
      "mean_token_accuracy": 0.19175078123807907,
      "num_tokens": 32975368.0,
      "step": 17880
    },
    {
      "entropy": 5.692888259887695,
      "epoch": 1.502583490863264,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004776555859739191,
      "loss": 5.2354,
      "mean_token_accuracy": 0.18508009016513824,
      "num_tokens": 32984603.0,
      "step": 17885
    },
    {
      "entropy": 5.726054859161377,
      "epoch": 1.503003570678429,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047764257362388913,
      "loss": 5.2006,
      "mean_token_accuracy": 0.18924659490585327,
      "num_tokens": 32993621.0,
      "step": 17890
    },
    {
      "entropy": 5.631523466110229,
      "epoch": 1.5034236504935938,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004776295576841504,
      "loss": 5.1931,
      "mean_token_accuracy": 0.19043040424585342,
      "num_tokens": 33002637.0,
      "step": 17895
    },
    {
      "entropy": 5.650270175933838,
      "epoch": 1.5038437303087586,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047761653815493337,
      "loss": 5.1328,
      "mean_token_accuracy": 0.1974516913294792,
      "num_tokens": 33011964.0,
      "step": 17900
    },
    {
      "entropy": 5.664147901535034,
      "epoch": 1.5042638101239234,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047760351503646877,
      "loss": 5.2054,
      "mean_token_accuracy": 0.1873147264122963,
      "num_tokens": 33020626.0,
      "step": 17905
    },
    {
      "entropy": 5.697901010513306,
      "epoch": 1.5046838899390884,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004775904883289871,
      "loss": 5.156,
      "mean_token_accuracy": 0.18911846429109574,
      "num_tokens": 33029212.0,
      "step": 17910
    },
    {
      "entropy": 5.6613422393798825,
      "epoch": 1.5051039697542534,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047757745803271936,
      "loss": 5.223,
      "mean_token_accuracy": 0.18291569352149964,
      "num_tokens": 33038893.0,
      "step": 17915
    },
    {
      "entropy": 5.664108037948608,
      "epoch": 1.5055240495694182,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004775644241478962,
      "loss": 5.2065,
      "mean_token_accuracy": 0.18577029705047607,
      "num_tokens": 33048058.0,
      "step": 17920
    },
    {
      "entropy": 5.6448201656341555,
      "epoch": 1.505944129384583,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047755138667474864,
      "loss": 5.1329,
      "mean_token_accuracy": 0.19166362136602402,
      "num_tokens": 33057106.0,
      "step": 17925
    },
    {
      "entropy": 5.649065685272217,
      "epoch": 1.506364209199748,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004775383456135075,
      "loss": 5.2648,
      "mean_token_accuracy": 0.18357754051685332,
      "num_tokens": 33066400.0,
      "step": 17930
    },
    {
      "entropy": 5.661255550384522,
      "epoch": 1.5067842890149128,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004775253009644038,
      "loss": 5.1247,
      "mean_token_accuracy": 0.19647203981876374,
      "num_tokens": 33075357.0,
      "step": 17935
    },
    {
      "entropy": 5.739010190963745,
      "epoch": 1.5072043688300778,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047751225272766885,
      "loss": 5.2318,
      "mean_token_accuracy": 0.18486902564764024,
      "num_tokens": 33085707.0,
      "step": 17940
    },
    {
      "entropy": 5.815869665145874,
      "epoch": 1.5076244486452426,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004774992009035335,
      "loss": 5.311,
      "mean_token_accuracy": 0.18061485439538955,
      "num_tokens": 33095825.0,
      "step": 17945
    },
    {
      "entropy": 5.661799383163452,
      "epoch": 1.5080445284604074,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004774861454922291,
      "loss": 5.1383,
      "mean_token_accuracy": 0.19088542014360427,
      "num_tokens": 33105130.0,
      "step": 17950
    },
    {
      "entropy": 5.658465957641601,
      "epoch": 1.5084646082755724,
      "grad_norm": 1.125,
      "learning_rate": 0.0004774730864939869,
      "loss": 5.1777,
      "mean_token_accuracy": 0.18474355041980745,
      "num_tokens": 33113226.0,
      "step": 17955
    },
    {
      "entropy": 5.69106822013855,
      "epoch": 1.5088846880907374,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047746002390903824,
      "loss": 5.1605,
      "mean_token_accuracy": 0.1924551621079445,
      "num_tokens": 33120824.0,
      "step": 17960
    },
    {
      "entropy": 5.714562177658081,
      "epoch": 1.5093047679059022,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004774469577376145,
      "loss": 5.1538,
      "mean_token_accuracy": 0.1928742066025734,
      "num_tokens": 33129503.0,
      "step": 17965
    },
    {
      "entropy": 5.573778820037842,
      "epoch": 1.509724847721067,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047743388797994715,
      "loss": 5.0766,
      "mean_token_accuracy": 0.19486554861068725,
      "num_tokens": 33138838.0,
      "step": 17970
    },
    {
      "entropy": 5.647586870193481,
      "epoch": 1.5101449275362318,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047742081463626767,
      "loss": 5.1816,
      "mean_token_accuracy": 0.1911126509308815,
      "num_tokens": 33148142.0,
      "step": 17975
    },
    {
      "entropy": 5.663047027587891,
      "epoch": 1.5105650073513968,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004774077377068078,
      "loss": 5.1906,
      "mean_token_accuracy": 0.1948083758354187,
      "num_tokens": 33156750.0,
      "step": 17980
    },
    {
      "entropy": 5.763590002059937,
      "epoch": 1.5109850871665618,
      "grad_norm": 1.03125,
      "learning_rate": 0.000477394657191799,
      "loss": 5.3168,
      "mean_token_accuracy": 0.1840188294649124,
      "num_tokens": 33166511.0,
      "step": 17985
    },
    {
      "entropy": 5.71966438293457,
      "epoch": 1.5114051669817266,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047738157309147307,
      "loss": 5.2786,
      "mean_token_accuracy": 0.1871291682124138,
      "num_tokens": 33175812.0,
      "step": 17990
    },
    {
      "entropy": 5.653257751464844,
      "epoch": 1.5118252467968913,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047736848540606174,
      "loss": 5.1296,
      "mean_token_accuracy": 0.18715905994176865,
      "num_tokens": 33185201.0,
      "step": 17995
    },
    {
      "entropy": 5.6673171043396,
      "epoch": 1.5122453266120561,
      "grad_norm": 1.078125,
      "learning_rate": 0.000477355394135797,
      "loss": 5.1178,
      "mean_token_accuracy": 0.18796557784080506,
      "num_tokens": 33195151.0,
      "step": 18000
    },
    {
      "epoch": 1.5122453266120561,
      "eval_entropy": 5.422196337610668,
      "eval_loss": 5.247309684753418,
      "eval_mean_token_accuracy": 0.19431857194562485,
      "eval_num_tokens": 33195151.0,
      "eval_runtime": 20.7342,
      "eval_samples_per_second": 1802.139,
      "eval_steps_per_second": 225.279,
      "step": 18000
    },
    {
      "entropy": 5.68146653175354,
      "epoch": 1.5126654064272211,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004773422992809106,
      "loss": 5.1828,
      "mean_token_accuracy": 0.19170676469802855,
      "num_tokens": 33204800.0,
      "step": 18005
    },
    {
      "entropy": 5.665374088287353,
      "epoch": 1.5130854862423861,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004773292008416346,
      "loss": 5.2,
      "mean_token_accuracy": 0.19000470042228698,
      "num_tokens": 33214529.0,
      "step": 18010
    },
    {
      "entropy": 5.711228847503662,
      "epoch": 1.513505566057551,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047731609881820095,
      "loss": 5.2145,
      "mean_token_accuracy": 0.18299998193979264,
      "num_tokens": 33224522.0,
      "step": 18015
    },
    {
      "entropy": 5.759230470657348,
      "epoch": 1.5139256458727157,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047730299321084173,
      "loss": 5.2153,
      "mean_token_accuracy": 0.1894364148378372,
      "num_tokens": 33233220.0,
      "step": 18020
    },
    {
      "entropy": 5.706881809234619,
      "epoch": 1.5143457256878807,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047728988401978916,
      "loss": 5.1622,
      "mean_token_accuracy": 0.18817603290081025,
      "num_tokens": 33242277.0,
      "step": 18025
    },
    {
      "entropy": 5.739811992645263,
      "epoch": 1.5147658055030457,
      "grad_norm": 1.0,
      "learning_rate": 0.0004772767712452756,
      "loss": 5.201,
      "mean_token_accuracy": 0.19743338525295256,
      "num_tokens": 33251113.0,
      "step": 18030
    },
    {
      "entropy": 5.67248649597168,
      "epoch": 1.5151858853182105,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047726365488753305,
      "loss": 5.3346,
      "mean_token_accuracy": 0.1813394695520401,
      "num_tokens": 33261055.0,
      "step": 18035
    },
    {
      "entropy": 5.73041090965271,
      "epoch": 1.5156059651333753,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047725053494679403,
      "loss": 5.3075,
      "mean_token_accuracy": 0.18020932972431183,
      "num_tokens": 33270981.0,
      "step": 18040
    },
    {
      "entropy": 5.8066658020019535,
      "epoch": 1.51602604494854,
      "grad_norm": 1.125,
      "learning_rate": 0.00047723741142329104,
      "loss": 5.2469,
      "mean_token_accuracy": 0.18234001845121384,
      "num_tokens": 33279516.0,
      "step": 18045
    },
    {
      "entropy": 5.679580545425415,
      "epoch": 1.516446124763705,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047722428431725637,
      "loss": 5.1828,
      "mean_token_accuracy": 0.19380246996879577,
      "num_tokens": 33288300.0,
      "step": 18050
    },
    {
      "entropy": 5.615900468826294,
      "epoch": 1.5168662045788701,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004772111536289226,
      "loss": 5.1997,
      "mean_token_accuracy": 0.18502001017332076,
      "num_tokens": 33299059.0,
      "step": 18055
    },
    {
      "entropy": 5.677860164642334,
      "epoch": 1.517286284394035,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047719801935852235,
      "loss": 5.2441,
      "mean_token_accuracy": 0.18329039514064788,
      "num_tokens": 33308879.0,
      "step": 18060
    },
    {
      "entropy": 5.7792737007141115,
      "epoch": 1.5177063642091997,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004771848815062883,
      "loss": 5.3322,
      "mean_token_accuracy": 0.17884444892406465,
      "num_tokens": 33318615.0,
      "step": 18065
    },
    {
      "entropy": 5.755047941207886,
      "epoch": 1.5181264440243645,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004771717400724532,
      "loss": 5.3745,
      "mean_token_accuracy": 0.1764811635017395,
      "num_tokens": 33328748.0,
      "step": 18070
    },
    {
      "entropy": 5.722303867340088,
      "epoch": 1.5185465238395295,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004771585950572499,
      "loss": 5.1912,
      "mean_token_accuracy": 0.18234504908323287,
      "num_tokens": 33338350.0,
      "step": 18075
    },
    {
      "entropy": 5.6933564186096195,
      "epoch": 1.5189666036546945,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004771454464609111,
      "loss": 5.1954,
      "mean_token_accuracy": 0.18834950923919677,
      "num_tokens": 33348202.0,
      "step": 18080
    },
    {
      "entropy": 5.614488649368286,
      "epoch": 1.5193866834698593,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004771322942836699,
      "loss": 5.2062,
      "mean_token_accuracy": 0.18710137605667115,
      "num_tokens": 33356996.0,
      "step": 18085
    },
    {
      "entropy": 5.759280633926392,
      "epoch": 1.519806763285024,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004771191385257592,
      "loss": 5.2863,
      "mean_token_accuracy": 0.1781959354877472,
      "num_tokens": 33366173.0,
      "step": 18090
    },
    {
      "entropy": 5.748492670059204,
      "epoch": 1.520226843100189,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004771059791874119,
      "loss": 5.2413,
      "mean_token_accuracy": 0.17910731881856917,
      "num_tokens": 33375921.0,
      "step": 18095
    },
    {
      "entropy": 5.669404411315918,
      "epoch": 1.520646922915354,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004770928162688613,
      "loss": 5.1842,
      "mean_token_accuracy": 0.1894295647740364,
      "num_tokens": 33385538.0,
      "step": 18100
    },
    {
      "entropy": 5.664581203460694,
      "epoch": 1.5210670027305189,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047707964977034055,
      "loss": 5.1287,
      "mean_token_accuracy": 0.19599489420652388,
      "num_tokens": 33393728.0,
      "step": 18105
    },
    {
      "entropy": 5.744066190719605,
      "epoch": 1.5214870825456837,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004770664796920828,
      "loss": 5.2221,
      "mean_token_accuracy": 0.18306789100170134,
      "num_tokens": 33402540.0,
      "step": 18110
    },
    {
      "entropy": 5.674205780029297,
      "epoch": 1.5219071623608484,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004770533060343215,
      "loss": 5.1943,
      "mean_token_accuracy": 0.1871560588479042,
      "num_tokens": 33411706.0,
      "step": 18115
    },
    {
      "entropy": 5.643727445602417,
      "epoch": 1.5223272421760135,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004770401287972899,
      "loss": 5.1422,
      "mean_token_accuracy": 0.19051759243011473,
      "num_tokens": 33420604.0,
      "step": 18120
    },
    {
      "entropy": 5.636025905609131,
      "epoch": 1.5227473219911785,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047702694798122143,
      "loss": 5.1153,
      "mean_token_accuracy": 0.19872678816318512,
      "num_tokens": 33429558.0,
      "step": 18125
    },
    {
      "entropy": 5.789419507980346,
      "epoch": 1.5231674018063432,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047701376358634957,
      "loss": 5.3244,
      "mean_token_accuracy": 0.17577954828739167,
      "num_tokens": 33439620.0,
      "step": 18130
    },
    {
      "entropy": 5.700473594665527,
      "epoch": 1.523587481621508,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047700057561290797,
      "loss": 5.2718,
      "mean_token_accuracy": 0.18329745680093765,
      "num_tokens": 33449067.0,
      "step": 18135
    },
    {
      "entropy": 5.625346231460571,
      "epoch": 1.5240075614366728,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004769873840611302,
      "loss": 5.1646,
      "mean_token_accuracy": 0.18640013486146928,
      "num_tokens": 33458089.0,
      "step": 18140
    },
    {
      "entropy": 5.6967747688293455,
      "epoch": 1.5244276412518378,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004769741889312499,
      "loss": 5.2974,
      "mean_token_accuracy": 0.18416070938110352,
      "num_tokens": 33466883.0,
      "step": 18145
    },
    {
      "entropy": 5.730316209793091,
      "epoch": 1.5248477210670028,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047696099022350087,
      "loss": 5.3098,
      "mean_token_accuracy": 0.18089912682771683,
      "num_tokens": 33476649.0,
      "step": 18150
    },
    {
      "entropy": 5.8102867603302,
      "epoch": 1.5252678008821676,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047694778793811685,
      "loss": 5.2745,
      "mean_token_accuracy": 0.1810872420668602,
      "num_tokens": 33486274.0,
      "step": 18155
    },
    {
      "entropy": 5.720560503005982,
      "epoch": 1.5256878806973324,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047693458207533177,
      "loss": 5.1629,
      "mean_token_accuracy": 0.18541543930768967,
      "num_tokens": 33494950.0,
      "step": 18160
    },
    {
      "entropy": 5.667433452606201,
      "epoch": 1.5261079605124974,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004769213726353795,
      "loss": 5.2072,
      "mean_token_accuracy": 0.19074438363313675,
      "num_tokens": 33503545.0,
      "step": 18165
    },
    {
      "entropy": 5.6646253108978275,
      "epoch": 1.5265280403276622,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047690815961849416,
      "loss": 5.2381,
      "mean_token_accuracy": 0.18631891906261444,
      "num_tokens": 33512871.0,
      "step": 18170
    },
    {
      "entropy": 5.703803443908692,
      "epoch": 1.5269481201428272,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004768949430249097,
      "loss": 5.1853,
      "mean_token_accuracy": 0.18463631123304367,
      "num_tokens": 33521933.0,
      "step": 18175
    },
    {
      "entropy": 5.671959257125854,
      "epoch": 1.527368199957992,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004768817228548603,
      "loss": 5.1543,
      "mean_token_accuracy": 0.18730782717466354,
      "num_tokens": 33531370.0,
      "step": 18180
    },
    {
      "entropy": 5.749475717544556,
      "epoch": 1.5277882797731568,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004768684991085802,
      "loss": 5.2413,
      "mean_token_accuracy": 0.18674177825450897,
      "num_tokens": 33540310.0,
      "step": 18185
    },
    {
      "entropy": 5.748261022567749,
      "epoch": 1.5282083595883218,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047685527178630347,
      "loss": 5.245,
      "mean_token_accuracy": 0.18378521800041198,
      "num_tokens": 33549943.0,
      "step": 18190
    },
    {
      "entropy": 5.751106071472168,
      "epoch": 1.5286284394034868,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004768420408882646,
      "loss": 5.3106,
      "mean_token_accuracy": 0.18207121789455413,
      "num_tokens": 33560167.0,
      "step": 18195
    },
    {
      "entropy": 5.7595922470092775,
      "epoch": 1.5290485192186516,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047682880641469787,
      "loss": 5.194,
      "mean_token_accuracy": 0.18333557695150376,
      "num_tokens": 33569604.0,
      "step": 18200
    },
    {
      "entropy": 5.7331043720245365,
      "epoch": 1.5294685990338164,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004768155683658378,
      "loss": 5.1974,
      "mean_token_accuracy": 0.1852698877453804,
      "num_tokens": 33578400.0,
      "step": 18205
    },
    {
      "entropy": 5.614318132400513,
      "epoch": 1.5298886788489812,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004768023267419188,
      "loss": 5.1813,
      "mean_token_accuracy": 0.18292288929224015,
      "num_tokens": 33587527.0,
      "step": 18210
    },
    {
      "entropy": 5.6538426876068115,
      "epoch": 1.5303087586641462,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004767890815431756,
      "loss": 5.1082,
      "mean_token_accuracy": 0.18878984898328782,
      "num_tokens": 33596026.0,
      "step": 18215
    },
    {
      "entropy": 5.667317533493042,
      "epoch": 1.5307288384793112,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047677583276984264,
      "loss": 5.1843,
      "mean_token_accuracy": 0.18641254305839539,
      "num_tokens": 33605906.0,
      "step": 18220
    },
    {
      "entropy": 5.653371906280517,
      "epoch": 1.531148918294476,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004767625804221548,
      "loss": 5.1614,
      "mean_token_accuracy": 0.1875321701169014,
      "num_tokens": 33615758.0,
      "step": 18225
    },
    {
      "entropy": 5.678629159927368,
      "epoch": 1.5315689981096408,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004767493245003466,
      "loss": 5.2198,
      "mean_token_accuracy": 0.1989634558558464,
      "num_tokens": 33625486.0,
      "step": 18230
    },
    {
      "entropy": 5.677112483978272,
      "epoch": 1.5319890779248058,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047673606500465315,
      "loss": 5.1588,
      "mean_token_accuracy": 0.19794049561023713,
      "num_tokens": 33633954.0,
      "step": 18235
    },
    {
      "entropy": 5.669197988510132,
      "epoch": 1.5324091577399706,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000476722801935309,
      "loss": 5.2344,
      "mean_token_accuracy": 0.18432164192199707,
      "num_tokens": 33642478.0,
      "step": 18240
    },
    {
      "entropy": 5.627944231033325,
      "epoch": 1.5328292375551356,
      "grad_norm": 1.125,
      "learning_rate": 0.0004767095352925495,
      "loss": 5.1658,
      "mean_token_accuracy": 0.19252669662237168,
      "num_tokens": 33650785.0,
      "step": 18245
    },
    {
      "entropy": 5.653904724121094,
      "epoch": 1.5332493173703003,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004766962650766093,
      "loss": 5.1265,
      "mean_token_accuracy": 0.1938754290342331,
      "num_tokens": 33659677.0,
      "step": 18250
    },
    {
      "entropy": 5.773008537292481,
      "epoch": 1.5336693971854651,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047668299128772365,
      "loss": 5.3028,
      "mean_token_accuracy": 0.17765135765075685,
      "num_tokens": 33669493.0,
      "step": 18255
    },
    {
      "entropy": 5.778806591033936,
      "epoch": 1.5340894770006301,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004766697139261277,
      "loss": 5.2898,
      "mean_token_accuracy": 0.18563032597303392,
      "num_tokens": 33678446.0,
      "step": 18260
    },
    {
      "entropy": 5.6903032779693605,
      "epoch": 1.5345095568157952,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004766564329920566,
      "loss": 5.1452,
      "mean_token_accuracy": 0.19144199341535567,
      "num_tokens": 33687647.0,
      "step": 18265
    },
    {
      "entropy": 5.729109191894532,
      "epoch": 1.53492963663096,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004766431484857456,
      "loss": 5.2128,
      "mean_token_accuracy": 0.19211371690034867,
      "num_tokens": 33697395.0,
      "step": 18270
    },
    {
      "entropy": 5.71789436340332,
      "epoch": 1.5353497164461247,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047662986040743004,
      "loss": 5.2152,
      "mean_token_accuracy": 0.19618483632802963,
      "num_tokens": 33706779.0,
      "step": 18275
    },
    {
      "entropy": 5.663815832138061,
      "epoch": 1.5357697962612895,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004766165687573454,
      "loss": 5.2013,
      "mean_token_accuracy": 0.18823830038309097,
      "num_tokens": 33714828.0,
      "step": 18280
    },
    {
      "entropy": 5.720313262939453,
      "epoch": 1.5361898760764545,
      "grad_norm": 0.96875,
      "learning_rate": 0.000476603273535727,
      "loss": 5.2074,
      "mean_token_accuracy": 0.1864815652370453,
      "num_tokens": 33724730.0,
      "step": 18285
    },
    {
      "entropy": 5.746868419647217,
      "epoch": 1.5366099558916195,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004765899747428104,
      "loss": 5.2587,
      "mean_token_accuracy": 0.18759524822235107,
      "num_tokens": 33734374.0,
      "step": 18290
    },
    {
      "entropy": 5.751424884796142,
      "epoch": 1.5370300357067843,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047657667237883125,
      "loss": 5.2393,
      "mean_token_accuracy": 0.1928840845823288,
      "num_tokens": 33743395.0,
      "step": 18295
    },
    {
      "entropy": 5.757344961166382,
      "epoch": 1.537450115521949,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047656336644402513,
      "loss": 5.2905,
      "mean_token_accuracy": 0.18564702272415162,
      "num_tokens": 33752526.0,
      "step": 18300
    },
    {
      "entropy": 5.756756067276001,
      "epoch": 1.5378701953371139,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004765500569386278,
      "loss": 5.2482,
      "mean_token_accuracy": 0.19399198591709138,
      "num_tokens": 33761310.0,
      "step": 18305
    },
    {
      "entropy": 5.691047954559326,
      "epoch": 1.538290275152279,
      "grad_norm": 0.96875,
      "learning_rate": 0.000476536743862875,
      "loss": 5.149,
      "mean_token_accuracy": 0.1888882428407669,
      "num_tokens": 33770870.0,
      "step": 18310
    },
    {
      "entropy": 5.6254003047943115,
      "epoch": 1.538710354967444,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047652342721700246,
      "loss": 5.1233,
      "mean_token_accuracy": 0.18723627477884291,
      "num_tokens": 33779648.0,
      "step": 18315
    },
    {
      "entropy": 5.709557342529297,
      "epoch": 1.5391304347826087,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004765101070012462,
      "loss": 5.2824,
      "mean_token_accuracy": 0.1799091339111328,
      "num_tokens": 33789172.0,
      "step": 18320
    },
    {
      "entropy": 5.771592998504639,
      "epoch": 1.5395505145977735,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047649678321584214,
      "loss": 5.2694,
      "mean_token_accuracy": 0.17828236222267152,
      "num_tokens": 33798069.0,
      "step": 18325
    },
    {
      "entropy": 5.7512492656707765,
      "epoch": 1.5399705944129385,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047648345586102643,
      "loss": 5.2272,
      "mean_token_accuracy": 0.18744209855794908,
      "num_tokens": 33806214.0,
      "step": 18330
    },
    {
      "entropy": 5.704847288131714,
      "epoch": 1.5403906742281035,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000476470124937035,
      "loss": 5.2256,
      "mean_token_accuracy": 0.18707503229379654,
      "num_tokens": 33815365.0,
      "step": 18335
    },
    {
      "entropy": 5.740296745300293,
      "epoch": 1.5408107540432683,
      "grad_norm": 1.046875,
      "learning_rate": 0.000476456790444104,
      "loss": 5.1359,
      "mean_token_accuracy": 0.19656677395105362,
      "num_tokens": 33825204.0,
      "step": 18340
    },
    {
      "entropy": 5.76303071975708,
      "epoch": 1.541230833858433,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004764434523824697,
      "loss": 5.2589,
      "mean_token_accuracy": 0.18746478855609894,
      "num_tokens": 33834439.0,
      "step": 18345
    },
    {
      "entropy": 5.680177307128906,
      "epoch": 1.5416509136735979,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047643011075236845,
      "loss": 5.2335,
      "mean_token_accuracy": 0.18748681247234344,
      "num_tokens": 33843959.0,
      "step": 18350
    },
    {
      "entropy": 5.757766628265381,
      "epoch": 1.5420709934887629,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047641676555403646,
      "loss": 5.2825,
      "mean_token_accuracy": 0.1790398359298706,
      "num_tokens": 33853234.0,
      "step": 18355
    },
    {
      "entropy": 5.720400714874268,
      "epoch": 1.5424910733039279,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004764034167877102,
      "loss": 5.1856,
      "mean_token_accuracy": 0.1870778262615204,
      "num_tokens": 33861755.0,
      "step": 18360
    },
    {
      "entropy": 5.748347759246826,
      "epoch": 1.5429111531190927,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047639006445362607,
      "loss": 5.2881,
      "mean_token_accuracy": 0.18777570724487305,
      "num_tokens": 33870956.0,
      "step": 18365
    },
    {
      "entropy": 5.602409362792969,
      "epoch": 1.5433312329342574,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004763767085520207,
      "loss": 5.1467,
      "mean_token_accuracy": 0.18963173925876617,
      "num_tokens": 33880568.0,
      "step": 18370
    },
    {
      "entropy": 5.786648082733154,
      "epoch": 1.5437513127494222,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004763633490831306,
      "loss": 5.3324,
      "mean_token_accuracy": 0.17457073628902436,
      "num_tokens": 33890145.0,
      "step": 18375
    },
    {
      "entropy": 5.696844816207886,
      "epoch": 1.5441713925645872,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004763499860471925,
      "loss": 5.1888,
      "mean_token_accuracy": 0.1918278530240059,
      "num_tokens": 33899155.0,
      "step": 18380
    },
    {
      "entropy": 5.667385625839233,
      "epoch": 1.5445914723797523,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000476336619444443,
      "loss": 5.216,
      "mean_token_accuracy": 0.17915960848331453,
      "num_tokens": 33909410.0,
      "step": 18385
    },
    {
      "entropy": 5.663455677032471,
      "epoch": 1.545011552194917,
      "grad_norm": 1.015625,
      "learning_rate": 0.000476323249275119,
      "loss": 5.1119,
      "mean_token_accuracy": 0.1997802197933197,
      "num_tokens": 33918451.0,
      "step": 18390
    },
    {
      "entropy": 5.664191722869873,
      "epoch": 1.5454316320100818,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004763098755394573,
      "loss": 5.1396,
      "mean_token_accuracy": 0.19072637856006622,
      "num_tokens": 33928317.0,
      "step": 18395
    },
    {
      "entropy": 5.705172443389893,
      "epoch": 1.5458517118252468,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004762964982376949,
      "loss": 5.3077,
      "mean_token_accuracy": 0.18540750294923783,
      "num_tokens": 33938010.0,
      "step": 18400
    },
    {
      "entropy": 5.715775442123413,
      "epoch": 1.5462717916404118,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047628311737006856,
      "loss": 5.1279,
      "mean_token_accuracy": 0.1897299110889435,
      "num_tokens": 33946964.0,
      "step": 18405
    },
    {
      "entropy": 5.665321493148804,
      "epoch": 1.5466918714555766,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047626973293681555,
      "loss": 5.15,
      "mean_token_accuracy": 0.18568734228610992,
      "num_tokens": 33956026.0,
      "step": 18410
    },
    {
      "entropy": 5.632046175003052,
      "epoch": 1.5471119512707414,
      "grad_norm": 1.0,
      "learning_rate": 0.0004762563449381728,
      "loss": 5.1997,
      "mean_token_accuracy": 0.18289818465709687,
      "num_tokens": 33965787.0,
      "step": 18415
    },
    {
      "entropy": 5.713032484054565,
      "epoch": 1.5475320310859062,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047624295337437753,
      "loss": 5.2263,
      "mean_token_accuracy": 0.1853696033358574,
      "num_tokens": 33974178.0,
      "step": 18420
    },
    {
      "entropy": 5.646288299560547,
      "epoch": 1.5479521109010712,
      "grad_norm": 1.0,
      "learning_rate": 0.0004762295582456669,
      "loss": 5.105,
      "mean_token_accuracy": 0.19382406920194625,
      "num_tokens": 33983652.0,
      "step": 18425
    },
    {
      "entropy": 5.7271651268005375,
      "epoch": 1.5483721907162362,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047621615955227835,
      "loss": 5.1753,
      "mean_token_accuracy": 0.19231128245592116,
      "num_tokens": 33991938.0,
      "step": 18430
    },
    {
      "entropy": 5.623953628540039,
      "epoch": 1.548792270531401,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004762027572944491,
      "loss": 5.1714,
      "mean_token_accuracy": 0.18704987764358522,
      "num_tokens": 33999918.0,
      "step": 18435
    },
    {
      "entropy": 5.6429015636444095,
      "epoch": 1.5492123503465658,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047618935147241667,
      "loss": 5.17,
      "mean_token_accuracy": 0.19808271080255507,
      "num_tokens": 34008416.0,
      "step": 18440
    },
    {
      "entropy": 5.730793476104736,
      "epoch": 1.5496324301617306,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004761759420864184,
      "loss": 5.2379,
      "mean_token_accuracy": 0.18340663611888885,
      "num_tokens": 34017616.0,
      "step": 18445
    },
    {
      "entropy": 5.7416540622711185,
      "epoch": 1.5500525099768956,
      "grad_norm": 1.0,
      "learning_rate": 0.000476162529136692,
      "loss": 5.1783,
      "mean_token_accuracy": 0.1859962150454521,
      "num_tokens": 34026064.0,
      "step": 18450
    },
    {
      "entropy": 5.67342586517334,
      "epoch": 1.5504725897920606,
      "grad_norm": 1.0,
      "learning_rate": 0.0004761491126234749,
      "loss": 5.0969,
      "mean_token_accuracy": 0.1947190970182419,
      "num_tokens": 34035378.0,
      "step": 18455
    },
    {
      "entropy": 5.675426769256592,
      "epoch": 1.5508926696072254,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004761356925470049,
      "loss": 5.1577,
      "mean_token_accuracy": 0.18379280269145964,
      "num_tokens": 34044600.0,
      "step": 18460
    },
    {
      "entropy": 5.706506872177124,
      "epoch": 1.5513127494223902,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047612226890751956,
      "loss": 5.2194,
      "mean_token_accuracy": 0.18591221421957016,
      "num_tokens": 34054680.0,
      "step": 18465
    },
    {
      "entropy": 5.725556659698486,
      "epoch": 1.5517328292375552,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047610884170525697,
      "loss": 5.16,
      "mean_token_accuracy": 0.19319739192724228,
      "num_tokens": 34063034.0,
      "step": 18470
    },
    {
      "entropy": 5.65951452255249,
      "epoch": 1.55215290905272,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004760954109404547,
      "loss": 5.1633,
      "mean_token_accuracy": 0.1872850701212883,
      "num_tokens": 34072122.0,
      "step": 18475
    },
    {
      "entropy": 5.6789820194244385,
      "epoch": 1.552572988867885,
      "grad_norm": 1.0,
      "learning_rate": 0.0004760819766133508,
      "loss": 5.1672,
      "mean_token_accuracy": 0.18463978320360183,
      "num_tokens": 34081493.0,
      "step": 18480
    },
    {
      "entropy": 5.724473762512207,
      "epoch": 1.5529930686830498,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047606853872418317,
      "loss": 5.2287,
      "mean_token_accuracy": 0.18234089016914368,
      "num_tokens": 34090872.0,
      "step": 18485
    },
    {
      "entropy": 5.613226079940796,
      "epoch": 1.5534131484982145,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004760550972731899,
      "loss": 5.0976,
      "mean_token_accuracy": 0.2005658194422722,
      "num_tokens": 34100729.0,
      "step": 18490
    },
    {
      "entropy": 5.60742883682251,
      "epoch": 1.5538332283133796,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004760416522606092,
      "loss": 5.0904,
      "mean_token_accuracy": 0.1989050254225731,
      "num_tokens": 34109492.0,
      "step": 18495
    },
    {
      "entropy": 5.61601824760437,
      "epoch": 1.5542533081285446,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004760282036866791,
      "loss": 5.1746,
      "mean_token_accuracy": 0.1915330171585083,
      "num_tokens": 34119529.0,
      "step": 18500
    },
    {
      "entropy": 5.748079395294189,
      "epoch": 1.5546733879437094,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004760147515516379,
      "loss": 5.2451,
      "mean_token_accuracy": 0.18769406974315644,
      "num_tokens": 34128261.0,
      "step": 18505
    },
    {
      "entropy": 5.729319143295288,
      "epoch": 1.5550934677588741,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047600129585572386,
      "loss": 5.2326,
      "mean_token_accuracy": 0.18961698412895203,
      "num_tokens": 34136916.0,
      "step": 18510
    },
    {
      "entropy": 5.669271802902221,
      "epoch": 1.555513547574039,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004759878365991754,
      "loss": 5.1318,
      "mean_token_accuracy": 0.19420424550771714,
      "num_tokens": 34146400.0,
      "step": 18515
    },
    {
      "entropy": 5.6788330078125,
      "epoch": 1.555933627389204,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004759743737822309,
      "loss": 5.1849,
      "mean_token_accuracy": 0.19295579195022583,
      "num_tokens": 34155611.0,
      "step": 18520
    },
    {
      "entropy": 5.732957410812378,
      "epoch": 1.556353707204369,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047596090740512884,
      "loss": 5.2204,
      "mean_token_accuracy": 0.18947670310735704,
      "num_tokens": 34165301.0,
      "step": 18525
    },
    {
      "entropy": 5.710750722885132,
      "epoch": 1.5567737870195337,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047594743746810786,
      "loss": 5.1912,
      "mean_token_accuracy": 0.18710286766290665,
      "num_tokens": 34174655.0,
      "step": 18530
    },
    {
      "entropy": 5.762575244903564,
      "epoch": 1.5571938668346985,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047593396397140644,
      "loss": 5.3344,
      "mean_token_accuracy": 0.1839095965027809,
      "num_tokens": 34184293.0,
      "step": 18535
    },
    {
      "entropy": 5.74411153793335,
      "epoch": 1.5576139466498635,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004759204869152632,
      "loss": 5.2298,
      "mean_token_accuracy": 0.18207360953092575,
      "num_tokens": 34193025.0,
      "step": 18540
    },
    {
      "entropy": 5.6920819759368895,
      "epoch": 1.5580340264650283,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004759070062999171,
      "loss": 5.1577,
      "mean_token_accuracy": 0.18634756356477739,
      "num_tokens": 34201082.0,
      "step": 18545
    },
    {
      "entropy": 5.741053104400635,
      "epoch": 1.5584541062801933,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004758935221256069,
      "loss": 5.2834,
      "mean_token_accuracy": 0.18015584498643875,
      "num_tokens": 34211210.0,
      "step": 18550
    },
    {
      "entropy": 5.746921730041504,
      "epoch": 1.558874186095358,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047588003439257134,
      "loss": 5.2245,
      "mean_token_accuracy": 0.18685204535722733,
      "num_tokens": 34220309.0,
      "step": 18555
    },
    {
      "entropy": 5.746979141235352,
      "epoch": 1.559294265910523,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047586654310104946,
      "loss": 5.2307,
      "mean_token_accuracy": 0.17383143901824952,
      "num_tokens": 34229532.0,
      "step": 18560
    },
    {
      "entropy": 5.778536224365235,
      "epoch": 1.559714345725688,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004758530482512801,
      "loss": 5.43,
      "mean_token_accuracy": 0.17496731877326965,
      "num_tokens": 34239543.0,
      "step": 18565
    },
    {
      "entropy": 5.721138191223145,
      "epoch": 1.560134425540853,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004758395498435024,
      "loss": 5.2259,
      "mean_token_accuracy": 0.1903236299753189,
      "num_tokens": 34248654.0,
      "step": 18570
    },
    {
      "entropy": 5.717340564727783,
      "epoch": 1.5605545053560177,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047582604787795555,
      "loss": 5.2422,
      "mean_token_accuracy": 0.1781938150525093,
      "num_tokens": 34258757.0,
      "step": 18575
    },
    {
      "entropy": 5.698209571838379,
      "epoch": 1.5609745851711825,
      "grad_norm": 1.0,
      "learning_rate": 0.0004758125423548787,
      "loss": 5.2215,
      "mean_token_accuracy": 0.1816542387008667,
      "num_tokens": 34268253.0,
      "step": 18580
    },
    {
      "entropy": 5.722633171081543,
      "epoch": 1.5613946649863473,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047579903327451097,
      "loss": 5.291,
      "mean_token_accuracy": 0.18011696338653566,
      "num_tokens": 34277361.0,
      "step": 18585
    },
    {
      "entropy": 5.667633199691773,
      "epoch": 1.5618147448015123,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004757855206370919,
      "loss": 5.1652,
      "mean_token_accuracy": 0.18839871734380723,
      "num_tokens": 34285923.0,
      "step": 18590
    },
    {
      "entropy": 5.634246206283569,
      "epoch": 1.5622348246166773,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047577200444286064,
      "loss": 5.156,
      "mean_token_accuracy": 0.18964882642030717,
      "num_tokens": 34296300.0,
      "step": 18595
    },
    {
      "entropy": 5.758458280563355,
      "epoch": 1.562654904431842,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004757584846920567,
      "loss": 5.1829,
      "mean_token_accuracy": 0.19113355726003647,
      "num_tokens": 34305757.0,
      "step": 18600
    },
    {
      "entropy": 5.736627626419067,
      "epoch": 1.5630749842470069,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004757449613849196,
      "loss": 5.2498,
      "mean_token_accuracy": 0.17898643463850022,
      "num_tokens": 34314714.0,
      "step": 18605
    },
    {
      "entropy": 5.789560461044312,
      "epoch": 1.5634950640621716,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047573143452168883,
      "loss": 5.2527,
      "mean_token_accuracy": 0.18915975689888,
      "num_tokens": 34323501.0,
      "step": 18610
    },
    {
      "entropy": 5.72806944847107,
      "epoch": 1.5639151438773367,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047571790410260405,
      "loss": 5.1847,
      "mean_token_accuracy": 0.19050463736057283,
      "num_tokens": 34331752.0,
      "step": 18615
    },
    {
      "entropy": 5.738928508758545,
      "epoch": 1.5643352236925017,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000475704370127905,
      "loss": 5.2329,
      "mean_token_accuracy": 0.18271356225013732,
      "num_tokens": 34341479.0,
      "step": 18620
    },
    {
      "entropy": 5.707871818542481,
      "epoch": 1.5647553035076665,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004756908325978314,
      "loss": 5.2373,
      "mean_token_accuracy": 0.1824428215622902,
      "num_tokens": 34350991.0,
      "step": 18625
    },
    {
      "entropy": 5.69806227684021,
      "epoch": 1.5651753833228312,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047567729151262305,
      "loss": 5.1723,
      "mean_token_accuracy": 0.18900625854730607,
      "num_tokens": 34360089.0,
      "step": 18630
    },
    {
      "entropy": 5.663226509094239,
      "epoch": 1.5655954631379962,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004756637468725198,
      "loss": 5.1355,
      "mean_token_accuracy": 0.18475776314735412,
      "num_tokens": 34370352.0,
      "step": 18635
    },
    {
      "entropy": 5.687634134292603,
      "epoch": 1.5660155429531613,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004756501986777616,
      "loss": 5.1475,
      "mean_token_accuracy": 0.1822171464562416,
      "num_tokens": 34378958.0,
      "step": 18640
    },
    {
      "entropy": 5.644553804397583,
      "epoch": 1.566435622768326,
      "grad_norm": 1.125,
      "learning_rate": 0.00047563664692858843,
      "loss": 5.0887,
      "mean_token_accuracy": 0.1987448588013649,
      "num_tokens": 34387723.0,
      "step": 18645
    },
    {
      "entropy": 5.707383918762207,
      "epoch": 1.5668557025834908,
      "grad_norm": 1.0,
      "learning_rate": 0.0004756230916252404,
      "loss": 5.2212,
      "mean_token_accuracy": 0.1899727240204811,
      "num_tokens": 34397089.0,
      "step": 18650
    },
    {
      "entropy": 5.822610139846802,
      "epoch": 1.5672757823986556,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047560953276795756,
      "loss": 5.2434,
      "mean_token_accuracy": 0.1865297257900238,
      "num_tokens": 34406278.0,
      "step": 18655
    },
    {
      "entropy": 5.699982118606568,
      "epoch": 1.5676958622138206,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047559597035698014,
      "loss": 5.2086,
      "mean_token_accuracy": 0.18752616494894028,
      "num_tokens": 34415404.0,
      "step": 18660
    },
    {
      "entropy": 5.656954574584961,
      "epoch": 1.5681159420289856,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004755824043925485,
      "loss": 5.2456,
      "mean_token_accuracy": 0.18446249812841414,
      "num_tokens": 34425036.0,
      "step": 18665
    },
    {
      "entropy": 5.711710596084595,
      "epoch": 1.5685360218441504,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004755688348749027,
      "loss": 5.1681,
      "mean_token_accuracy": 0.18997115641832352,
      "num_tokens": 34434246.0,
      "step": 18670
    },
    {
      "entropy": 5.649768590927124,
      "epoch": 1.5689561016593152,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004755552618042834,
      "loss": 5.1566,
      "mean_token_accuracy": 0.1906455397605896,
      "num_tokens": 34444189.0,
      "step": 18675
    },
    {
      "entropy": 5.699088048934937,
      "epoch": 1.56937618147448,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004755416851809308,
      "loss": 5.1704,
      "mean_token_accuracy": 0.18978362828493117,
      "num_tokens": 34453727.0,
      "step": 18680
    },
    {
      "entropy": 5.643474149703979,
      "epoch": 1.569796261289645,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004755281050050856,
      "loss": 5.1661,
      "mean_token_accuracy": 0.18719941824674607,
      "num_tokens": 34462835.0,
      "step": 18685
    },
    {
      "entropy": 5.663916778564453,
      "epoch": 1.57021634110481,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004755145212769882,
      "loss": 5.2049,
      "mean_token_accuracy": 0.19053178429603576,
      "num_tokens": 34471642.0,
      "step": 18690
    },
    {
      "entropy": 5.663637399673462,
      "epoch": 1.5706364209199748,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047550093399687936,
      "loss": 5.2065,
      "mean_token_accuracy": 0.18711315393447875,
      "num_tokens": 34480468.0,
      "step": 18695
    },
    {
      "entropy": 5.7650031566619875,
      "epoch": 1.5710565007351396,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004754873431649997,
      "loss": 5.1819,
      "mean_token_accuracy": 0.18817597329616548,
      "num_tokens": 34490299.0,
      "step": 18700
    },
    {
      "entropy": 5.698501873016357,
      "epoch": 1.5714765805503046,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047547374878159003,
      "loss": 5.2622,
      "mean_token_accuracy": 0.17971642911434174,
      "num_tokens": 34498831.0,
      "step": 18705
    },
    {
      "entropy": 5.691864395141602,
      "epoch": 1.5718966603654696,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004754601508468911,
      "loss": 5.2012,
      "mean_token_accuracy": 0.18786772936582566,
      "num_tokens": 34508048.0,
      "step": 18710
    },
    {
      "entropy": 5.721207094192505,
      "epoch": 1.5723167401806344,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004754465493611438,
      "loss": 5.3109,
      "mean_token_accuracy": 0.17956671267747878,
      "num_tokens": 34517070.0,
      "step": 18715
    },
    {
      "entropy": 5.656190824508667,
      "epoch": 1.5727368199957992,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047543294432458904,
      "loss": 5.1193,
      "mean_token_accuracy": 0.19062748700380325,
      "num_tokens": 34525934.0,
      "step": 18720
    },
    {
      "entropy": 5.760162591934204,
      "epoch": 1.573156899810964,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000475419335737468,
      "loss": 5.3041,
      "mean_token_accuracy": 0.18648409396409987,
      "num_tokens": 34534222.0,
      "step": 18725
    },
    {
      "entropy": 5.782601261138916,
      "epoch": 1.573576979626129,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047540572360002157,
      "loss": 5.2819,
      "mean_token_accuracy": 0.18312918394804,
      "num_tokens": 34543291.0,
      "step": 18730
    },
    {
      "entropy": 5.73994197845459,
      "epoch": 1.573997059441294,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047539210791249095,
      "loss": 5.1378,
      "mean_token_accuracy": 0.19613551050424577,
      "num_tokens": 34552383.0,
      "step": 18735
    },
    {
      "entropy": 5.7127539157867435,
      "epoch": 1.5744171392564588,
      "grad_norm": 1.125,
      "learning_rate": 0.0004753784886751173,
      "loss": 5.1499,
      "mean_token_accuracy": 0.19371048510074615,
      "num_tokens": 34560311.0,
      "step": 18740
    },
    {
      "entropy": 5.6537212371826175,
      "epoch": 1.5748372190716236,
      "grad_norm": 1.0,
      "learning_rate": 0.0004753648658881419,
      "loss": 5.1895,
      "mean_token_accuracy": 0.19350127130746841,
      "num_tokens": 34569903.0,
      "step": 18745
    },
    {
      "entropy": 5.680458450317383,
      "epoch": 1.5752572988867883,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047535123955180607,
      "loss": 5.1938,
      "mean_token_accuracy": 0.1860976740717888,
      "num_tokens": 34579735.0,
      "step": 18750
    },
    {
      "entropy": 5.74421238899231,
      "epoch": 1.5756773787019533,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004753376096663512,
      "loss": 5.2375,
      "mean_token_accuracy": 0.18463196307420732,
      "num_tokens": 34589105.0,
      "step": 18755
    },
    {
      "entropy": 5.636974811553955,
      "epoch": 1.5760974585171184,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047532397623201877,
      "loss": 5.1724,
      "mean_token_accuracy": 0.19587232023477555,
      "num_tokens": 34597883.0,
      "step": 18760
    },
    {
      "entropy": 5.669177389144897,
      "epoch": 1.5765175383322831,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047531033924905024,
      "loss": 5.1472,
      "mean_token_accuracy": 0.19323616176843644,
      "num_tokens": 34606666.0,
      "step": 18765
    },
    {
      "entropy": 5.728579807281494,
      "epoch": 1.576937618147448,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004752966987176873,
      "loss": 5.2429,
      "mean_token_accuracy": 0.18624092936515807,
      "num_tokens": 34616547.0,
      "step": 18770
    },
    {
      "entropy": 5.681903743743897,
      "epoch": 1.577357697962613,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004752830546381713,
      "loss": 5.2294,
      "mean_token_accuracy": 0.19251570254564285,
      "num_tokens": 34625679.0,
      "step": 18775
    },
    {
      "entropy": 5.61891016960144,
      "epoch": 1.5777777777777777,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004752694070107442,
      "loss": 5.196,
      "mean_token_accuracy": 0.18695205301046372,
      "num_tokens": 34635633.0,
      "step": 18780
    },
    {
      "entropy": 5.722997188568115,
      "epoch": 1.5781978575929427,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004752557558356476,
      "loss": 5.2161,
      "mean_token_accuracy": 0.18560741394758223,
      "num_tokens": 34645206.0,
      "step": 18785
    },
    {
      "entropy": 5.702025318145752,
      "epoch": 1.5786179374081075,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004752421011131234,
      "loss": 5.2386,
      "mean_token_accuracy": 0.17885345965623856,
      "num_tokens": 34653884.0,
      "step": 18790
    },
    {
      "entropy": 5.633797597885132,
      "epoch": 1.5790380172232723,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047522844284341364,
      "loss": 5.0837,
      "mean_token_accuracy": 0.19601554721593856,
      "num_tokens": 34662170.0,
      "step": 18795
    },
    {
      "entropy": 5.629519510269165,
      "epoch": 1.5794580970384373,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004752147810267601,
      "loss": 5.2413,
      "mean_token_accuracy": 0.1837887153029442,
      "num_tokens": 34672548.0,
      "step": 18800
    },
    {
      "entropy": 5.7391496181488035,
      "epoch": 1.5798781768536023,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047520111566340465,
      "loss": 5.2408,
      "mean_token_accuracy": 0.18808806240558623,
      "num_tokens": 34680972.0,
      "step": 18805
    },
    {
      "entropy": 5.690294933319092,
      "epoch": 1.580298256668767,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047518744675358965,
      "loss": 5.1154,
      "mean_token_accuracy": 0.19420301169157028,
      "num_tokens": 34689589.0,
      "step": 18810
    },
    {
      "entropy": 5.670215177536011,
      "epoch": 1.580718336483932,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004751737742975571,
      "loss": 5.1475,
      "mean_token_accuracy": 0.18692880868911743,
      "num_tokens": 34698747.0,
      "step": 18815
    },
    {
      "entropy": 5.682676553726196,
      "epoch": 1.5811384162990967,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047516009829554913,
      "loss": 5.2175,
      "mean_token_accuracy": 0.1871643528342247,
      "num_tokens": 34707502.0,
      "step": 18820
    },
    {
      "entropy": 5.669141960144043,
      "epoch": 1.5815584961142617,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047514641874780815,
      "loss": 5.1337,
      "mean_token_accuracy": 0.18977122455835344,
      "num_tokens": 34715879.0,
      "step": 18825
    },
    {
      "entropy": 5.664966678619384,
      "epoch": 1.5819785759294267,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047513273565457644,
      "loss": 5.3155,
      "mean_token_accuracy": 0.18593798726797103,
      "num_tokens": 34726090.0,
      "step": 18830
    },
    {
      "entropy": 5.779621982574463,
      "epoch": 1.5823986557445915,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004751190490160964,
      "loss": 5.2686,
      "mean_token_accuracy": 0.1767435759305954,
      "num_tokens": 34736014.0,
      "step": 18835
    },
    {
      "entropy": 5.755038595199585,
      "epoch": 1.5828187355597563,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047510535883261035,
      "loss": 5.1961,
      "mean_token_accuracy": 0.1897487610578537,
      "num_tokens": 34745648.0,
      "step": 18840
    },
    {
      "entropy": 5.700213479995727,
      "epoch": 1.5832388153749213,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000475091665104361,
      "loss": 5.1693,
      "mean_token_accuracy": 0.1924079403281212,
      "num_tokens": 34753908.0,
      "step": 18845
    },
    {
      "entropy": 5.682164049148559,
      "epoch": 1.583658895190086,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004750779678315908,
      "loss": 5.0461,
      "mean_token_accuracy": 0.20134453624486923,
      "num_tokens": 34762303.0,
      "step": 18850
    },
    {
      "entropy": 5.650493192672729,
      "epoch": 1.584078975005251,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004750642670145424,
      "loss": 5.2265,
      "mean_token_accuracy": 0.18128631561994552,
      "num_tokens": 34771463.0,
      "step": 18855
    },
    {
      "entropy": 5.790189456939697,
      "epoch": 1.5844990548204159,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004750505626534585,
      "loss": 5.296,
      "mean_token_accuracy": 0.18232108578085898,
      "num_tokens": 34780704.0,
      "step": 18860
    },
    {
      "entropy": 5.672183704376221,
      "epoch": 1.5849191346355807,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047503685474858194,
      "loss": 5.1387,
      "mean_token_accuracy": 0.19334539920091628,
      "num_tokens": 34790262.0,
      "step": 18865
    },
    {
      "entropy": 5.705787420272827,
      "epoch": 1.5853392144507457,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004750231433001555,
      "loss": 5.1573,
      "mean_token_accuracy": 0.1922369748353958,
      "num_tokens": 34799450.0,
      "step": 18870
    },
    {
      "entropy": 5.741850662231445,
      "epoch": 1.5857592942659107,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004750094283084221,
      "loss": 5.223,
      "mean_token_accuracy": 0.17889637500047684,
      "num_tokens": 34808220.0,
      "step": 18875
    },
    {
      "entropy": 5.745956707000732,
      "epoch": 1.5861793740810755,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047499570977362467,
      "loss": 5.2206,
      "mean_token_accuracy": 0.1817813202738762,
      "num_tokens": 34817846.0,
      "step": 18880
    },
    {
      "entropy": 5.644904613494873,
      "epoch": 1.5865994538962402,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047498198769600617,
      "loss": 5.2321,
      "mean_token_accuracy": 0.19198177456855775,
      "num_tokens": 34826962.0,
      "step": 18885
    },
    {
      "entropy": 5.635780382156372,
      "epoch": 1.587019533711405,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004749682620758097,
      "loss": 5.1709,
      "mean_token_accuracy": 0.1871061369776726,
      "num_tokens": 34837170.0,
      "step": 18890
    },
    {
      "entropy": 5.69831805229187,
      "epoch": 1.58743961352657,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047495453291327854,
      "loss": 5.1734,
      "mean_token_accuracy": 0.18981715589761733,
      "num_tokens": 34845336.0,
      "step": 18895
    },
    {
      "entropy": 5.688724565505981,
      "epoch": 1.587859693341735,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047494080020865577,
      "loss": 5.1714,
      "mean_token_accuracy": 0.1918397828936577,
      "num_tokens": 34854613.0,
      "step": 18900
    },
    {
      "entropy": 5.7010026454925535,
      "epoch": 1.5882797731568998,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004749270639621846,
      "loss": 5.2378,
      "mean_token_accuracy": 0.1874014601111412,
      "num_tokens": 34864254.0,
      "step": 18905
    },
    {
      "entropy": 5.7555805206298825,
      "epoch": 1.5886998529720646,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004749133241741085,
      "loss": 5.2815,
      "mean_token_accuracy": 0.18555965423583984,
      "num_tokens": 34874380.0,
      "step": 18910
    },
    {
      "entropy": 5.739329147338867,
      "epoch": 1.5891199327872296,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004748995808446708,
      "loss": 5.2398,
      "mean_token_accuracy": 0.18198596984148024,
      "num_tokens": 34883688.0,
      "step": 18915
    },
    {
      "entropy": 5.673891305923462,
      "epoch": 1.5895400126023944,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047488583397411495,
      "loss": 5.1139,
      "mean_token_accuracy": 0.19461620599031448,
      "num_tokens": 34892831.0,
      "step": 18920
    },
    {
      "entropy": 5.709463596343994,
      "epoch": 1.5899600924175594,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047487208356268454,
      "loss": 5.1894,
      "mean_token_accuracy": 0.19562061429023742,
      "num_tokens": 34901517.0,
      "step": 18925
    },
    {
      "entropy": 5.746156930923462,
      "epoch": 1.5903801722327242,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047485832961062296,
      "loss": 5.202,
      "mean_token_accuracy": 0.19207657724618912,
      "num_tokens": 34910765.0,
      "step": 18930
    },
    {
      "entropy": 5.742528820037842,
      "epoch": 1.590800252047889,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047484457211817405,
      "loss": 5.2357,
      "mean_token_accuracy": 0.18313186913728713,
      "num_tokens": 34919799.0,
      "step": 18935
    },
    {
      "entropy": 5.630531692504883,
      "epoch": 1.591220331863054,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047483081108558143,
      "loss": 5.1122,
      "mean_token_accuracy": 0.19282832592725754,
      "num_tokens": 34928199.0,
      "step": 18940
    },
    {
      "entropy": 5.705439710617066,
      "epoch": 1.591640411678219,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000474817046513089,
      "loss": 5.2532,
      "mean_token_accuracy": 0.18635836690664292,
      "num_tokens": 34937751.0,
      "step": 18945
    },
    {
      "entropy": 5.805593776702881,
      "epoch": 1.5920604914933838,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004748032784009403,
      "loss": 5.1665,
      "mean_token_accuracy": 0.19168122112751007,
      "num_tokens": 34946052.0,
      "step": 18950
    },
    {
      "entropy": 5.7219563007354735,
      "epoch": 1.5924805713085486,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004747895067493796,
      "loss": 5.1849,
      "mean_token_accuracy": 0.18367738127708436,
      "num_tokens": 34954932.0,
      "step": 18955
    },
    {
      "entropy": 5.707137775421143,
      "epoch": 1.5929006511237134,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004747757315586505,
      "loss": 5.1644,
      "mean_token_accuracy": 0.19541464149951934,
      "num_tokens": 34963581.0,
      "step": 18960
    },
    {
      "entropy": 5.539660310745239,
      "epoch": 1.5933207309388784,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047476195282899727,
      "loss": 4.987,
      "mean_token_accuracy": 0.19968956410884858,
      "num_tokens": 34972844.0,
      "step": 18965
    },
    {
      "entropy": 5.655507802963257,
      "epoch": 1.5937408107540434,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047474817056066383,
      "loss": 5.2063,
      "mean_token_accuracy": 0.19344731867313386,
      "num_tokens": 34981998.0,
      "step": 18970
    },
    {
      "entropy": 5.646384143829346,
      "epoch": 1.5941608905692082,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047473438475389453,
      "loss": 5.1327,
      "mean_token_accuracy": 0.1968084305524826,
      "num_tokens": 34990552.0,
      "step": 18975
    },
    {
      "entropy": 5.661907052993774,
      "epoch": 1.594580970384373,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004747205954089333,
      "loss": 5.1346,
      "mean_token_accuracy": 0.19390660673379898,
      "num_tokens": 35000259.0,
      "step": 18980
    },
    {
      "entropy": 5.727808237075806,
      "epoch": 1.5950010501995378,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004747068025260247,
      "loss": 5.2256,
      "mean_token_accuracy": 0.1863940566778183,
      "num_tokens": 35009592.0,
      "step": 18985
    },
    {
      "entropy": 5.754229879379272,
      "epoch": 1.5954211300147028,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004746930061054129,
      "loss": 5.2709,
      "mean_token_accuracy": 0.17970754951238632,
      "num_tokens": 35019356.0,
      "step": 18990
    },
    {
      "entropy": 5.673370456695556,
      "epoch": 1.5958412098298678,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047467920614734224,
      "loss": 5.1853,
      "mean_token_accuracy": 0.1934608682990074,
      "num_tokens": 35028764.0,
      "step": 18995
    },
    {
      "entropy": 5.6985283374786375,
      "epoch": 1.5962612896450326,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004746654026520573,
      "loss": 5.2004,
      "mean_token_accuracy": 0.18951717615127564,
      "num_tokens": 35037903.0,
      "step": 19000
    },
    {
      "entropy": 5.686356925964356,
      "epoch": 1.5966813694601973,
      "grad_norm": 1.125,
      "learning_rate": 0.0004746515956198026,
      "loss": 5.1258,
      "mean_token_accuracy": 0.1899218186736107,
      "num_tokens": 35046326.0,
      "step": 19005
    },
    {
      "entropy": 5.783659839630127,
      "epoch": 1.5971014492753624,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047463778505082266,
      "loss": 5.336,
      "mean_token_accuracy": 0.18169979453086854,
      "num_tokens": 35055551.0,
      "step": 19010
    },
    {
      "entropy": 5.65393180847168,
      "epoch": 1.5975215290905274,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004746239709453621,
      "loss": 5.1051,
      "mean_token_accuracy": 0.1973701074719429,
      "num_tokens": 35065595.0,
      "step": 19015
    },
    {
      "entropy": 5.738381719589233,
      "epoch": 1.5979416089056921,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004746101533036658,
      "loss": 5.2121,
      "mean_token_accuracy": 0.1850927397608757,
      "num_tokens": 35075097.0,
      "step": 19020
    },
    {
      "entropy": 5.815371704101563,
      "epoch": 1.598361688720857,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047459633212597834,
      "loss": 5.3053,
      "mean_token_accuracy": 0.17791522145271302,
      "num_tokens": 35084092.0,
      "step": 19025
    },
    {
      "entropy": 5.7111162662506105,
      "epoch": 1.5987817685360217,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004745825074125447,
      "loss": 5.1849,
      "mean_token_accuracy": 0.1901321828365326,
      "num_tokens": 35093007.0,
      "step": 19030
    },
    {
      "entropy": 5.767176389694214,
      "epoch": 1.5992018483511867,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004745686791636097,
      "loss": 5.269,
      "mean_token_accuracy": 0.1814927041530609,
      "num_tokens": 35103094.0,
      "step": 19035
    },
    {
      "entropy": 5.631485414505005,
      "epoch": 1.5996219281663517,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047455484737941823,
      "loss": 5.082,
      "mean_token_accuracy": 0.19262613356113434,
      "num_tokens": 35112561.0,
      "step": 19040
    },
    {
      "entropy": 5.655822038650513,
      "epoch": 1.6000420079815165,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004745410120602155,
      "loss": 5.1766,
      "mean_token_accuracy": 0.18818106055259703,
      "num_tokens": 35121718.0,
      "step": 19045
    },
    {
      "entropy": 5.705010032653808,
      "epoch": 1.6004620877966813,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047452717320624647,
      "loss": 5.1419,
      "mean_token_accuracy": 0.20210612565279007,
      "num_tokens": 35130073.0,
      "step": 19050
    },
    {
      "entropy": 5.714133977890015,
      "epoch": 1.600882167611846,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004745133308177562,
      "loss": 5.1846,
      "mean_token_accuracy": 0.18922341465950013,
      "num_tokens": 35138876.0,
      "step": 19055
    },
    {
      "entropy": 5.716654300689697,
      "epoch": 1.601302247427011,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047449948489499007,
      "loss": 5.1839,
      "mean_token_accuracy": 0.18111678063869477,
      "num_tokens": 35147750.0,
      "step": 19060
    },
    {
      "entropy": 5.7355094909667965,
      "epoch": 1.6017223272421761,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047448563543819335,
      "loss": 5.1978,
      "mean_token_accuracy": 0.19010573476552964,
      "num_tokens": 35156955.0,
      "step": 19065
    },
    {
      "entropy": 5.646643877029419,
      "epoch": 1.602142407057341,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004744717824476112,
      "loss": 5.2188,
      "mean_token_accuracy": 0.18415475338697435,
      "num_tokens": 35166542.0,
      "step": 19070
    },
    {
      "entropy": 5.698222541809082,
      "epoch": 1.6025624868725057,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047445792592348926,
      "loss": 5.1716,
      "mean_token_accuracy": 0.19387024492025376,
      "num_tokens": 35175258.0,
      "step": 19075
    },
    {
      "entropy": 5.7199615955352785,
      "epoch": 1.6029825666876707,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004744440658660729,
      "loss": 5.1885,
      "mean_token_accuracy": 0.18246723860502242,
      "num_tokens": 35184970.0,
      "step": 19080
    },
    {
      "entropy": 5.698621463775635,
      "epoch": 1.6034026465028357,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004744302022756075,
      "loss": 5.1745,
      "mean_token_accuracy": 0.18502653390169144,
      "num_tokens": 35193948.0,
      "step": 19085
    },
    {
      "entropy": 5.5929741859436035,
      "epoch": 1.6038227263180005,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047441633515233874,
      "loss": 5.1299,
      "mean_token_accuracy": 0.19354627132415772,
      "num_tokens": 35203792.0,
      "step": 19090
    },
    {
      "entropy": 5.6906232833862305,
      "epoch": 1.6042428061331653,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004744024644965123,
      "loss": 5.2715,
      "mean_token_accuracy": 0.18539106398820876,
      "num_tokens": 35212684.0,
      "step": 19095
    },
    {
      "entropy": 5.676451396942139,
      "epoch": 1.60466288594833,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047438859030837397,
      "loss": 5.1034,
      "mean_token_accuracy": 0.19492622315883637,
      "num_tokens": 35220830.0,
      "step": 19100
    },
    {
      "entropy": 5.722379541397094,
      "epoch": 1.605082965763495,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047437471258816936,
      "loss": 5.1605,
      "mean_token_accuracy": 0.1878805175423622,
      "num_tokens": 35230171.0,
      "step": 19105
    },
    {
      "entropy": 5.675529813766479,
      "epoch": 1.60550304557866,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047436083133614446,
      "loss": 5.0958,
      "mean_token_accuracy": 0.19459396749734878,
      "num_tokens": 35239022.0,
      "step": 19110
    },
    {
      "entropy": 5.673115158081055,
      "epoch": 1.6059231253938249,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047434694655254495,
      "loss": 5.1241,
      "mean_token_accuracy": 0.18892850428819657,
      "num_tokens": 35247564.0,
      "step": 19115
    },
    {
      "entropy": 5.671248865127564,
      "epoch": 1.6063432052089897,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000474333058237617,
      "loss": 5.1684,
      "mean_token_accuracy": 0.18273677676916122,
      "num_tokens": 35256175.0,
      "step": 19120
    },
    {
      "entropy": 5.844353246688843,
      "epoch": 1.6067632850241544,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047431916639160656,
      "loss": 5.2905,
      "mean_token_accuracy": 0.18649296909570695,
      "num_tokens": 35265278.0,
      "step": 19125
    },
    {
      "entropy": 5.6216577053070065,
      "epoch": 1.6071833648393195,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004743052710147598,
      "loss": 5.0319,
      "mean_token_accuracy": 0.20057274103164674,
      "num_tokens": 35274715.0,
      "step": 19130
    },
    {
      "entropy": 5.544546365737915,
      "epoch": 1.6076034446544845,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047429137210732266,
      "loss": 5.1418,
      "mean_token_accuracy": 0.1894964635372162,
      "num_tokens": 35285450.0,
      "step": 19135
    },
    {
      "entropy": 5.700664901733399,
      "epoch": 1.6080235244696492,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004742774696695415,
      "loss": 5.1774,
      "mean_token_accuracy": 0.18049427270889282,
      "num_tokens": 35294531.0,
      "step": 19140
    },
    {
      "entropy": 5.760405158996582,
      "epoch": 1.608443604284814,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047426356370166266,
      "loss": 5.1977,
      "mean_token_accuracy": 0.18217170536518096,
      "num_tokens": 35303749.0,
      "step": 19145
    },
    {
      "entropy": 5.638873100280762,
      "epoch": 1.608863684099979,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004742496542039324,
      "loss": 5.1779,
      "mean_token_accuracy": 0.18119246363639832,
      "num_tokens": 35312994.0,
      "step": 19150
    },
    {
      "entropy": 5.6744140625,
      "epoch": 1.6092837639151438,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047423574117659703,
      "loss": 5.1442,
      "mean_token_accuracy": 0.19014629125595092,
      "num_tokens": 35322533.0,
      "step": 19155
    },
    {
      "entropy": 5.6861967086792,
      "epoch": 1.6097038437303088,
      "grad_norm": 1.0,
      "learning_rate": 0.00047422182461990316,
      "loss": 5.1619,
      "mean_token_accuracy": 0.19348402172327042,
      "num_tokens": 35331872.0,
      "step": 19160
    },
    {
      "entropy": 5.674013757705689,
      "epoch": 1.6101239235454736,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047420790453409724,
      "loss": 5.2184,
      "mean_token_accuracy": 0.1855485185980797,
      "num_tokens": 35341517.0,
      "step": 19165
    },
    {
      "entropy": 5.608324718475342,
      "epoch": 1.6105440033606384,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004741939809194258,
      "loss": 5.1078,
      "mean_token_accuracy": 0.19698781222105027,
      "num_tokens": 35350291.0,
      "step": 19170
    },
    {
      "entropy": 5.723851728439331,
      "epoch": 1.6109640831758034,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047418005377613566,
      "loss": 5.2849,
      "mean_token_accuracy": 0.18126029819250106,
      "num_tokens": 35360711.0,
      "step": 19175
    },
    {
      "entropy": 5.702187156677246,
      "epoch": 1.6113841629909684,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004741661231044733,
      "loss": 5.2179,
      "mean_token_accuracy": 0.18911915868520737,
      "num_tokens": 35370069.0,
      "step": 19180
    },
    {
      "entropy": 5.735180187225342,
      "epoch": 1.6118042428061332,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047415218890468577,
      "loss": 5.1705,
      "mean_token_accuracy": 0.1961536481976509,
      "num_tokens": 35380389.0,
      "step": 19185
    },
    {
      "entropy": 5.641651582717896,
      "epoch": 1.612224322621298,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004741382511770197,
      "loss": 5.1811,
      "mean_token_accuracy": 0.18602068573236466,
      "num_tokens": 35389420.0,
      "step": 19190
    },
    {
      "entropy": 5.722196435928344,
      "epoch": 1.6126444024364628,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047412430992172205,
      "loss": 5.2858,
      "mean_token_accuracy": 0.18047426789999008,
      "num_tokens": 35399418.0,
      "step": 19195
    },
    {
      "entropy": 5.671954107284546,
      "epoch": 1.6130644822516278,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047411036513903974,
      "loss": 5.1696,
      "mean_token_accuracy": 0.18835400640964509,
      "num_tokens": 35408717.0,
      "step": 19200
    },
    {
      "entropy": 5.685611915588379,
      "epoch": 1.6134845620667928,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047409641682921987,
      "loss": 5.1347,
      "mean_token_accuracy": 0.20273468196392058,
      "num_tokens": 35417118.0,
      "step": 19205
    },
    {
      "entropy": 5.681270742416382,
      "epoch": 1.6139046418819576,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004740824649925096,
      "loss": 5.2066,
      "mean_token_accuracy": 0.18966880589723586,
      "num_tokens": 35425526.0,
      "step": 19210
    },
    {
      "entropy": 5.634865617752075,
      "epoch": 1.6143247216971224,
      "grad_norm": 1.125,
      "learning_rate": 0.0004740685096291559,
      "loss": 5.2169,
      "mean_token_accuracy": 0.18445804417133332,
      "num_tokens": 35434932.0,
      "step": 19215
    },
    {
      "entropy": 5.7745729923248295,
      "epoch": 1.6147448015122874,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047405455073940597,
      "loss": 5.2382,
      "mean_token_accuracy": 0.18467941880226135,
      "num_tokens": 35443909.0,
      "step": 19220
    },
    {
      "entropy": 5.764472389221192,
      "epoch": 1.6151648813274522,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004740405883235072,
      "loss": 5.2029,
      "mean_token_accuracy": 0.19218032360076903,
      "num_tokens": 35454082.0,
      "step": 19225
    },
    {
      "entropy": 5.78025598526001,
      "epoch": 1.6155849611426172,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047402662238170694,
      "loss": 5.2714,
      "mean_token_accuracy": 0.184135365486145,
      "num_tokens": 35464547.0,
      "step": 19230
    },
    {
      "entropy": 5.704647970199585,
      "epoch": 1.616005040957782,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004740126529142526,
      "loss": 5.156,
      "mean_token_accuracy": 0.18805668503046036,
      "num_tokens": 35473310.0,
      "step": 19235
    },
    {
      "entropy": 5.681270551681519,
      "epoch": 1.6164251207729468,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004739986799213915,
      "loss": 5.2542,
      "mean_token_accuracy": 0.1918073281645775,
      "num_tokens": 35483502.0,
      "step": 19240
    },
    {
      "entropy": 5.705972051620483,
      "epoch": 1.6168452005881118,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004739847034033713,
      "loss": 5.2201,
      "mean_token_accuracy": 0.18866525292396547,
      "num_tokens": 35493063.0,
      "step": 19245
    },
    {
      "entropy": 5.677617454528809,
      "epoch": 1.6172652804032768,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047397072336043957,
      "loss": 5.1776,
      "mean_token_accuracy": 0.18951303213834764,
      "num_tokens": 35501829.0,
      "step": 19250
    },
    {
      "entropy": 5.715717840194702,
      "epoch": 1.6176853602184416,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047395673979284383,
      "loss": 5.1882,
      "mean_token_accuracy": 0.18066412955522537,
      "num_tokens": 35510411.0,
      "step": 19255
    },
    {
      "entropy": 5.73548469543457,
      "epoch": 1.6181054400336063,
      "grad_norm": 1.015625,
      "learning_rate": 0.000473942752700832,
      "loss": 5.2031,
      "mean_token_accuracy": 0.18600752353668212,
      "num_tokens": 35519571.0,
      "step": 19260
    },
    {
      "entropy": 5.703898286819458,
      "epoch": 1.6185255198487711,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047392876208465166,
      "loss": 5.1677,
      "mean_token_accuracy": 0.18549326360225676,
      "num_tokens": 35527306.0,
      "step": 19265
    },
    {
      "entropy": 5.740292978286743,
      "epoch": 1.6189455996639361,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004739147679445508,
      "loss": 5.1635,
      "mean_token_accuracy": 0.1828925833106041,
      "num_tokens": 35536126.0,
      "step": 19270
    },
    {
      "entropy": 5.6792552947998045,
      "epoch": 1.6193656794791011,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004739007702807773,
      "loss": 5.2141,
      "mean_token_accuracy": 0.18502748012542725,
      "num_tokens": 35545593.0,
      "step": 19275
    },
    {
      "entropy": 5.610373783111572,
      "epoch": 1.619785759294266,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047388676909357894,
      "loss": 5.1679,
      "mean_token_accuracy": 0.18376809507608413,
      "num_tokens": 35554780.0,
      "step": 19280
    },
    {
      "entropy": 5.720436859130859,
      "epoch": 1.6202058391094307,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047387276438320394,
      "loss": 5.1577,
      "mean_token_accuracy": 0.19272739291191102,
      "num_tokens": 35562982.0,
      "step": 19285
    },
    {
      "entropy": 5.740250492095948,
      "epoch": 1.6206259189245955,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004738587561499003,
      "loss": 5.2774,
      "mean_token_accuracy": 0.18635699301958084,
      "num_tokens": 35571528.0,
      "step": 19290
    },
    {
      "entropy": 5.609196186065674,
      "epoch": 1.6210459987397605,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047384474439391615,
      "loss": 5.1109,
      "mean_token_accuracy": 0.19280455261468887,
      "num_tokens": 35580386.0,
      "step": 19295
    },
    {
      "entropy": 5.678215837478637,
      "epoch": 1.6214660785549255,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004738307291154998,
      "loss": 5.0824,
      "mean_token_accuracy": 0.1869479775428772,
      "num_tokens": 35589456.0,
      "step": 19300
    },
    {
      "entropy": 5.664905643463134,
      "epoch": 1.6218861583700903,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004738167103148995,
      "loss": 5.1794,
      "mean_token_accuracy": 0.1862914189696312,
      "num_tokens": 35598116.0,
      "step": 19305
    },
    {
      "entropy": 5.696725606918335,
      "epoch": 1.622306238185255,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047380268799236355,
      "loss": 5.1738,
      "mean_token_accuracy": 0.184284570813179,
      "num_tokens": 35606481.0,
      "step": 19310
    },
    {
      "entropy": 5.7014477252960205,
      "epoch": 1.62272631800042,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047378866214814024,
      "loss": 5.1518,
      "mean_token_accuracy": 0.18474017679691315,
      "num_tokens": 35615517.0,
      "step": 19315
    },
    {
      "entropy": 5.682359790802002,
      "epoch": 1.6231463978155851,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047377463278247827,
      "loss": 5.1856,
      "mean_token_accuracy": 0.18000807762145996,
      "num_tokens": 35625100.0,
      "step": 19320
    },
    {
      "entropy": 5.663484048843384,
      "epoch": 1.62356647763075,
      "grad_norm": 1.03125,
      "learning_rate": 0.000473760599895626,
      "loss": 5.1267,
      "mean_token_accuracy": 0.1869422346353531,
      "num_tokens": 35634572.0,
      "step": 19325
    },
    {
      "entropy": 5.71120285987854,
      "epoch": 1.6239865574459147,
      "grad_norm": 1.09375,
      "learning_rate": 0.000473746563487832,
      "loss": 5.1522,
      "mean_token_accuracy": 0.18967004418373107,
      "num_tokens": 35643883.0,
      "step": 19330
    },
    {
      "entropy": 5.667012023925781,
      "epoch": 1.6244066372610795,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047373252355934506,
      "loss": 5.2303,
      "mean_token_accuracy": 0.19216018468141555,
      "num_tokens": 35652527.0,
      "step": 19335
    },
    {
      "entropy": 5.7506133079528805,
      "epoch": 1.6248267170762445,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047371848011041375,
      "loss": 5.2255,
      "mean_token_accuracy": 0.19073550403118134,
      "num_tokens": 35662436.0,
      "step": 19340
    },
    {
      "entropy": 5.6976700782775875,
      "epoch": 1.6252467968914095,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047370443314128687,
      "loss": 5.1571,
      "mean_token_accuracy": 0.1936761423945427,
      "num_tokens": 35672302.0,
      "step": 19345
    },
    {
      "entropy": 5.738803386688232,
      "epoch": 1.6256668767065743,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004736903826522132,
      "loss": 5.2096,
      "mean_token_accuracy": 0.18296123892068863,
      "num_tokens": 35680852.0,
      "step": 19350
    },
    {
      "entropy": 5.696914577484131,
      "epoch": 1.626086956521739,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004736763286434419,
      "loss": 5.1826,
      "mean_token_accuracy": 0.18774507343769073,
      "num_tokens": 35690159.0,
      "step": 19355
    },
    {
      "entropy": 5.671940803527832,
      "epoch": 1.6265070363369039,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004736622711152216,
      "loss": 5.123,
      "mean_token_accuracy": 0.1900339186191559,
      "num_tokens": 35699165.0,
      "step": 19360
    },
    {
      "entropy": 5.700385856628418,
      "epoch": 1.6269271161520689,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004736482100678015,
      "loss": 5.1762,
      "mean_token_accuracy": 0.1880303665995598,
      "num_tokens": 35708910.0,
      "step": 19365
    },
    {
      "entropy": 5.694727277755737,
      "epoch": 1.6273471959672339,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047363414550143063,
      "loss": 5.2379,
      "mean_token_accuracy": 0.18436530530452727,
      "num_tokens": 35718218.0,
      "step": 19370
    },
    {
      "entropy": 5.683092975616455,
      "epoch": 1.6277672757823987,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047362007741635816,
      "loss": 5.1666,
      "mean_token_accuracy": 0.18449121564626694,
      "num_tokens": 35727076.0,
      "step": 19375
    },
    {
      "entropy": 5.7194297313690186,
      "epoch": 1.6281873555975634,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004736060058128333,
      "loss": 5.2659,
      "mean_token_accuracy": 0.1832773819565773,
      "num_tokens": 35736316.0,
      "step": 19380
    },
    {
      "entropy": 5.718344974517822,
      "epoch": 1.6286074354127285,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047359193069110533,
      "loss": 5.2167,
      "mean_token_accuracy": 0.18936008512973784,
      "num_tokens": 35745747.0,
      "step": 19385
    },
    {
      "entropy": 5.742183256149292,
      "epoch": 1.6290275152278935,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047357785205142354,
      "loss": 5.203,
      "mean_token_accuracy": 0.19118784070014955,
      "num_tokens": 35754825.0,
      "step": 19390
    },
    {
      "entropy": 5.654837703704834,
      "epoch": 1.6294475950430583,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004735637698940374,
      "loss": 5.1463,
      "mean_token_accuracy": 0.19106266498565674,
      "num_tokens": 35764504.0,
      "step": 19395
    },
    {
      "entropy": 5.783188772201538,
      "epoch": 1.629867674858223,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004735496842191963,
      "loss": 5.2377,
      "mean_token_accuracy": 0.18378744125366211,
      "num_tokens": 35774195.0,
      "step": 19400
    },
    {
      "entropy": 5.646369695663452,
      "epoch": 1.6302877546733878,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047353559502714976,
      "loss": 5.1009,
      "mean_token_accuracy": 0.19447090923786164,
      "num_tokens": 35783721.0,
      "step": 19405
    },
    {
      "entropy": 5.669447326660157,
      "epoch": 1.6307078344885528,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004735215023181474,
      "loss": 5.1954,
      "mean_token_accuracy": 0.18814494758844375,
      "num_tokens": 35792821.0,
      "step": 19410
    },
    {
      "entropy": 5.709923219680786,
      "epoch": 1.6311279143037178,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047350740609243883,
      "loss": 5.2257,
      "mean_token_accuracy": 0.1841375485062599,
      "num_tokens": 35802746.0,
      "step": 19415
    },
    {
      "entropy": 5.775618362426758,
      "epoch": 1.6315479941188826,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004734933063502738,
      "loss": 5.1853,
      "mean_token_accuracy": 0.19306959062814713,
      "num_tokens": 35811196.0,
      "step": 19420
    },
    {
      "entropy": 5.780186700820923,
      "epoch": 1.6319680739340474,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047347920309190203,
      "loss": 5.2681,
      "mean_token_accuracy": 0.18416139334440232,
      "num_tokens": 35820787.0,
      "step": 19425
    },
    {
      "entropy": 5.672123193740845,
      "epoch": 1.6323881537492122,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004734650963175734,
      "loss": 5.2061,
      "mean_token_accuracy": 0.1873942121863365,
      "num_tokens": 35831247.0,
      "step": 19430
    },
    {
      "entropy": 5.710813760757446,
      "epoch": 1.6328082335643772,
      "grad_norm": 1.0,
      "learning_rate": 0.00047345098602753777,
      "loss": 5.2735,
      "mean_token_accuracy": 0.18416493386030197,
      "num_tokens": 35840759.0,
      "step": 19435
    },
    {
      "entropy": 5.692849731445312,
      "epoch": 1.6332283133795422,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004734368722220451,
      "loss": 5.2047,
      "mean_token_accuracy": 0.18460565507411958,
      "num_tokens": 35850137.0,
      "step": 19440
    },
    {
      "entropy": 5.675331068038941,
      "epoch": 1.633648393194707,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004734227549013455,
      "loss": 5.0367,
      "mean_token_accuracy": 0.20037066787481309,
      "num_tokens": 35858412.0,
      "step": 19445
    },
    {
      "entropy": 5.672222089767456,
      "epoch": 1.6340684730098718,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004734086340656889,
      "loss": 5.1432,
      "mean_token_accuracy": 0.1918720558285713,
      "num_tokens": 35868202.0,
      "step": 19450
    },
    {
      "entropy": 5.638163137435913,
      "epoch": 1.6344885528250368,
      "grad_norm": 1.125,
      "learning_rate": 0.0004733945097153255,
      "loss": 5.1892,
      "mean_token_accuracy": 0.19344068169593812,
      "num_tokens": 35877237.0,
      "step": 19455
    },
    {
      "entropy": 5.622659015655517,
      "epoch": 1.6349086326402016,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004733803818505055,
      "loss": 5.0667,
      "mean_token_accuracy": 0.1967902213335037,
      "num_tokens": 35887016.0,
      "step": 19460
    },
    {
      "entropy": 5.729615306854248,
      "epoch": 1.6353287124553666,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047336625047147924,
      "loss": 5.1559,
      "mean_token_accuracy": 0.19411284923553468,
      "num_tokens": 35896393.0,
      "step": 19465
    },
    {
      "entropy": 5.699149560928345,
      "epoch": 1.6357487922705314,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047335211557849693,
      "loss": 5.1943,
      "mean_token_accuracy": 0.18804482072591783,
      "num_tokens": 35905237.0,
      "step": 19470
    },
    {
      "entropy": 5.676076650619507,
      "epoch": 1.6361688720856962,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004733379771718092,
      "loss": 5.1962,
      "mean_token_accuracy": 0.19026329070329667,
      "num_tokens": 35914352.0,
      "step": 19475
    },
    {
      "entropy": 5.7170275211334225,
      "epoch": 1.6365889519008612,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004733238352516661,
      "loss": 5.2747,
      "mean_token_accuracy": 0.18386934995651244,
      "num_tokens": 35923785.0,
      "step": 19480
    },
    {
      "entropy": 5.722577857971191,
      "epoch": 1.6370090317160262,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047330968981831856,
      "loss": 5.1967,
      "mean_token_accuracy": 0.18313570618629454,
      "num_tokens": 35932495.0,
      "step": 19485
    },
    {
      "entropy": 5.720559167861938,
      "epoch": 1.637429111531191,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047329554087201687,
      "loss": 5.1666,
      "mean_token_accuracy": 0.19785182774066926,
      "num_tokens": 35941745.0,
      "step": 19490
    },
    {
      "entropy": 5.79425835609436,
      "epoch": 1.6378491913463558,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047328138841301186,
      "loss": 5.2476,
      "mean_token_accuracy": 0.18498873263597487,
      "num_tokens": 35950281.0,
      "step": 19495
    },
    {
      "entropy": 5.654531812667846,
      "epoch": 1.6382692711615205,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004732672324415541,
      "loss": 5.1739,
      "mean_token_accuracy": 0.19395357370376587,
      "num_tokens": 35959531.0,
      "step": 19500
    },
    {
      "entropy": 5.72698335647583,
      "epoch": 1.6386893509766856,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004732530729578945,
      "loss": 5.2288,
      "mean_token_accuracy": 0.18878337293863295,
      "num_tokens": 35969462.0,
      "step": 19505
    },
    {
      "entropy": 5.719201898574829,
      "epoch": 1.6391094307918506,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004732389099622837,
      "loss": 5.2286,
      "mean_token_accuracy": 0.1823413550853729,
      "num_tokens": 35978022.0,
      "step": 19510
    },
    {
      "entropy": 5.767617416381836,
      "epoch": 1.6395295106070154,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047322474345497267,
      "loss": 5.2021,
      "mean_token_accuracy": 0.185215499997139,
      "num_tokens": 35988193.0,
      "step": 19515
    },
    {
      "entropy": 5.7725255489349365,
      "epoch": 1.6399495904221801,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047321057343621247,
      "loss": 5.23,
      "mean_token_accuracy": 0.1830961987376213,
      "num_tokens": 35997404.0,
      "step": 19520
    },
    {
      "entropy": 5.639719676971436,
      "epoch": 1.6403696702373451,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047319639990625395,
      "loss": 5.1169,
      "mean_token_accuracy": 0.20056988149881363,
      "num_tokens": 36005356.0,
      "step": 19525
    },
    {
      "entropy": 5.784581708908081,
      "epoch": 1.64078975005251,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047318222286534824,
      "loss": 5.3776,
      "mean_token_accuracy": 0.17723228484392167,
      "num_tokens": 36015305.0,
      "step": 19530
    },
    {
      "entropy": 5.769973802566528,
      "epoch": 1.641209829867675,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047316804231374663,
      "loss": 5.2233,
      "mean_token_accuracy": 0.18442749679088594,
      "num_tokens": 36024278.0,
      "step": 19535
    },
    {
      "entropy": 5.676090812683105,
      "epoch": 1.6416299096828397,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004731538582517001,
      "loss": 5.055,
      "mean_token_accuracy": 0.19409469068050383,
      "num_tokens": 36032870.0,
      "step": 19540
    },
    {
      "entropy": 5.646928882598877,
      "epoch": 1.6420499894980045,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047313967067945996,
      "loss": 5.0974,
      "mean_token_accuracy": 0.1983703076839447,
      "num_tokens": 36041725.0,
      "step": 19545
    },
    {
      "entropy": 5.712642002105713,
      "epoch": 1.6424700693131695,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004731254795972777,
      "loss": 5.2152,
      "mean_token_accuracy": 0.18961652219295502,
      "num_tokens": 36050929.0,
      "step": 19550
    },
    {
      "entropy": 5.7799293994903564,
      "epoch": 1.6428901491283345,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004731112850054045,
      "loss": 5.2158,
      "mean_token_accuracy": 0.18771993666887282,
      "num_tokens": 36060059.0,
      "step": 19555
    },
    {
      "entropy": 5.6676586151123045,
      "epoch": 1.6433102289434993,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004730970869040919,
      "loss": 5.1506,
      "mean_token_accuracy": 0.201340888440609,
      "num_tokens": 36069445.0,
      "step": 19560
    },
    {
      "entropy": 5.732652711868286,
      "epoch": 1.643730308758664,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047308288529359147,
      "loss": 5.2685,
      "mean_token_accuracy": 0.1837436407804489,
      "num_tokens": 36079129.0,
      "step": 19565
    },
    {
      "entropy": 5.772090864181519,
      "epoch": 1.644150388573829,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004730686801741547,
      "loss": 5.1699,
      "mean_token_accuracy": 0.19059444814920426,
      "num_tokens": 36088320.0,
      "step": 19570
    },
    {
      "entropy": 5.752588558197021,
      "epoch": 1.644570468388994,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004730544715460332,
      "loss": 5.2317,
      "mean_token_accuracy": 0.1882110208272934,
      "num_tokens": 36097728.0,
      "step": 19575
    },
    {
      "entropy": 5.771630716323853,
      "epoch": 1.644990548204159,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047304025940947875,
      "loss": 5.2422,
      "mean_token_accuracy": 0.18970264941453935,
      "num_tokens": 36106566.0,
      "step": 19580
    },
    {
      "entropy": 5.726252174377441,
      "epoch": 1.6454106280193237,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047302604376474306,
      "loss": 5.1933,
      "mean_token_accuracy": 0.18556971997022628,
      "num_tokens": 36115475.0,
      "step": 19585
    },
    {
      "entropy": 5.68648943901062,
      "epoch": 1.6458307078344885,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047301182461207807,
      "loss": 5.2715,
      "mean_token_accuracy": 0.18692764043807983,
      "num_tokens": 36124404.0,
      "step": 19590
    },
    {
      "entropy": 5.645534706115723,
      "epoch": 1.6462507876496533,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047299760195173554,
      "loss": 5.1242,
      "mean_token_accuracy": 0.1949109047651291,
      "num_tokens": 36132987.0,
      "step": 19595
    },
    {
      "entropy": 5.750514268875122,
      "epoch": 1.6466708674648183,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004729833757839673,
      "loss": 5.2719,
      "mean_token_accuracy": 0.18709372729063034,
      "num_tokens": 36142163.0,
      "step": 19600
    },
    {
      "entropy": 5.713512563705445,
      "epoch": 1.6470909472799833,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047296914610902565,
      "loss": 5.2245,
      "mean_token_accuracy": 0.18468474596738815,
      "num_tokens": 36152561.0,
      "step": 19605
    },
    {
      "entropy": 5.680805683135986,
      "epoch": 1.647511027095148,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047295491292716245,
      "loss": 5.1661,
      "mean_token_accuracy": 0.18546382635831832,
      "num_tokens": 36161877.0,
      "step": 19610
    },
    {
      "entropy": 5.727251625061035,
      "epoch": 1.6479311069103129,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047294067623862996,
      "loss": 5.2122,
      "mean_token_accuracy": 0.18415965139865875,
      "num_tokens": 36171523.0,
      "step": 19615
    },
    {
      "entropy": 5.672529697418213,
      "epoch": 1.6483511867254779,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047292643604368025,
      "loss": 5.134,
      "mean_token_accuracy": 0.19474033266305923,
      "num_tokens": 36180339.0,
      "step": 19620
    },
    {
      "entropy": 5.703769540786743,
      "epoch": 1.6487712665406429,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004729121923425657,
      "loss": 5.2507,
      "mean_token_accuracy": 0.1831184819340706,
      "num_tokens": 36191584.0,
      "step": 19625
    },
    {
      "entropy": 5.789300727844238,
      "epoch": 1.6491913463558077,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004728979451355385,
      "loss": 5.2447,
      "mean_token_accuracy": 0.1936808094382286,
      "num_tokens": 36200738.0,
      "step": 19630
    },
    {
      "entropy": 5.681628704071045,
      "epoch": 1.6496114261709725,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047288369442285115,
      "loss": 5.086,
      "mean_token_accuracy": 0.19748593270778655,
      "num_tokens": 36209394.0,
      "step": 19635
    },
    {
      "entropy": 5.672049856185913,
      "epoch": 1.6500315059861372,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047286944020475606,
      "loss": 5.2032,
      "mean_token_accuracy": 0.1919265180826187,
      "num_tokens": 36218268.0,
      "step": 19640
    },
    {
      "entropy": 5.668643999099731,
      "epoch": 1.6504515858013022,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004728551824815057,
      "loss": 5.1666,
      "mean_token_accuracy": 0.1924498811364174,
      "num_tokens": 36226974.0,
      "step": 19645
    },
    {
      "entropy": 5.639807033538818,
      "epoch": 1.6508716656164673,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047284092125335277,
      "loss": 5.1133,
      "mean_token_accuracy": 0.1968569651246071,
      "num_tokens": 36235892.0,
      "step": 19650
    },
    {
      "entropy": 5.606718492507935,
      "epoch": 1.651291745431632,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004728266565205497,
      "loss": 5.1223,
      "mean_token_accuracy": 0.1933310627937317,
      "num_tokens": 36244750.0,
      "step": 19655
    },
    {
      "entropy": 5.681778240203857,
      "epoch": 1.6517118252467968,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047281238828334924,
      "loss": 5.186,
      "mean_token_accuracy": 0.19129592031240464,
      "num_tokens": 36254902.0,
      "step": 19660
    },
    {
      "entropy": 5.735056066513062,
      "epoch": 1.6521319050619616,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004727981165420042,
      "loss": 5.2067,
      "mean_token_accuracy": 0.19015057533979415,
      "num_tokens": 36265546.0,
      "step": 19665
    },
    {
      "entropy": 5.640113830566406,
      "epoch": 1.6525519848771266,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004727838412967674,
      "loss": 5.1489,
      "mean_token_accuracy": 0.1906261622905731,
      "num_tokens": 36273978.0,
      "step": 19670
    },
    {
      "entropy": 5.708683347702026,
      "epoch": 1.6529720646922916,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004727695625478917,
      "loss": 5.1491,
      "mean_token_accuracy": 0.191907300055027,
      "num_tokens": 36283117.0,
      "step": 19675
    },
    {
      "entropy": 5.75423526763916,
      "epoch": 1.6533921445074564,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047275528029562996,
      "loss": 5.1623,
      "mean_token_accuracy": 0.18650306463241578,
      "num_tokens": 36293031.0,
      "step": 19680
    },
    {
      "entropy": 5.667601537704468,
      "epoch": 1.6538122243226212,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047274099454023535,
      "loss": 5.1852,
      "mean_token_accuracy": 0.19305343329906463,
      "num_tokens": 36302080.0,
      "step": 19685
    },
    {
      "entropy": 5.6549454689025875,
      "epoch": 1.6542323041377862,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047272670528196084,
      "loss": 5.1882,
      "mean_token_accuracy": 0.1860254541039467,
      "num_tokens": 36311077.0,
      "step": 19690
    },
    {
      "entropy": 5.618292713165284,
      "epoch": 1.6546523839529512,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004727124125210595,
      "loss": 5.12,
      "mean_token_accuracy": 0.19889096468687056,
      "num_tokens": 36320300.0,
      "step": 19695
    },
    {
      "entropy": 5.6836010932922365,
      "epoch": 1.655072463768116,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047269811625778456,
      "loss": 5.182,
      "mean_token_accuracy": 0.1898993268609047,
      "num_tokens": 36330184.0,
      "step": 19700
    },
    {
      "entropy": 5.661662673950195,
      "epoch": 1.6554925435832808,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004726838164923893,
      "loss": 5.2077,
      "mean_token_accuracy": 0.18399138748645782,
      "num_tokens": 36339526.0,
      "step": 19705
    },
    {
      "entropy": 5.701456832885742,
      "epoch": 1.6559126233984456,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047266951322512716,
      "loss": 5.1661,
      "mean_token_accuracy": 0.19235066771507264,
      "num_tokens": 36348849.0,
      "step": 19710
    },
    {
      "entropy": 5.701607275009155,
      "epoch": 1.6563327032136106,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047265520645625123,
      "loss": 5.1578,
      "mean_token_accuracy": 0.18250928372144698,
      "num_tokens": 36358924.0,
      "step": 19715
    },
    {
      "entropy": 5.782726573944092,
      "epoch": 1.6567527830287756,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047264089618601513,
      "loss": 5.2389,
      "mean_token_accuracy": 0.1896510824561119,
      "num_tokens": 36367130.0,
      "step": 19720
    },
    {
      "entropy": 5.713455390930176,
      "epoch": 1.6571728628439404,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004726265824146724,
      "loss": 5.1603,
      "mean_token_accuracy": 0.18987968415021897,
      "num_tokens": 36376575.0,
      "step": 19725
    },
    {
      "entropy": 5.641866016387939,
      "epoch": 1.6575929426591052,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004726122651424764,
      "loss": 5.1157,
      "mean_token_accuracy": 0.1913418158888817,
      "num_tokens": 36385010.0,
      "step": 19730
    },
    {
      "entropy": 5.5748037815094,
      "epoch": 1.65801302247427,
      "grad_norm": 1.078125,
      "learning_rate": 0.000472597944369681,
      "loss": 4.9226,
      "mean_token_accuracy": 0.2045608177781105,
      "num_tokens": 36393574.0,
      "step": 19735
    },
    {
      "entropy": 5.645991420745849,
      "epoch": 1.658433102289435,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047258362009653965,
      "loss": 5.14,
      "mean_token_accuracy": 0.1947878286242485,
      "num_tokens": 36401992.0,
      "step": 19740
    },
    {
      "entropy": 5.704039669036865,
      "epoch": 1.6588531821046,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047256929232330624,
      "loss": 5.2654,
      "mean_token_accuracy": 0.18167627304792405,
      "num_tokens": 36411712.0,
      "step": 19745
    },
    {
      "entropy": 5.664850187301636,
      "epoch": 1.6592732619197648,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004725549610502346,
      "loss": 5.0977,
      "mean_token_accuracy": 0.19135782718658448,
      "num_tokens": 36420240.0,
      "step": 19750
    },
    {
      "entropy": 5.706680059432983,
      "epoch": 1.6596933417349296,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047254062627757854,
      "loss": 5.2067,
      "mean_token_accuracy": 0.1940309703350067,
      "num_tokens": 36430068.0,
      "step": 19755
    },
    {
      "entropy": 5.752298212051391,
      "epoch": 1.6601134215500946,
      "grad_norm": 1.0,
      "learning_rate": 0.000472526288005592,
      "loss": 5.2224,
      "mean_token_accuracy": 0.1867004945874214,
      "num_tokens": 36439808.0,
      "step": 19760
    },
    {
      "entropy": 5.685444307327271,
      "epoch": 1.6605335013652593,
      "grad_norm": 1.0625,
      "learning_rate": 0.000472511946234529,
      "loss": 5.1796,
      "mean_token_accuracy": 0.1937822237610817,
      "num_tokens": 36449609.0,
      "step": 19765
    },
    {
      "entropy": 5.764244222640992,
      "epoch": 1.6609535811804244,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004724976009646435,
      "loss": 5.1705,
      "mean_token_accuracy": 0.1914839431643486,
      "num_tokens": 36457700.0,
      "step": 19770
    },
    {
      "entropy": 5.731533432006836,
      "epoch": 1.6613736609955891,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004724832521961897,
      "loss": 5.1988,
      "mean_token_accuracy": 0.1886048898100853,
      "num_tokens": 36466881.0,
      "step": 19775
    },
    {
      "entropy": 5.759159135818481,
      "epoch": 1.661793740810754,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047246889992942187,
      "loss": 5.2956,
      "mean_token_accuracy": 0.17890715450048447,
      "num_tokens": 36475433.0,
      "step": 19780
    },
    {
      "entropy": 5.721084213256836,
      "epoch": 1.662213820625919,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004724545441645941,
      "loss": 5.223,
      "mean_token_accuracy": 0.1863424837589264,
      "num_tokens": 36484232.0,
      "step": 19785
    },
    {
      "entropy": 5.749406147003174,
      "epoch": 1.662633900441084,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004724401849019608,
      "loss": 5.323,
      "mean_token_accuracy": 0.1770969569683075,
      "num_tokens": 36493588.0,
      "step": 19790
    },
    {
      "entropy": 5.6702492237091064,
      "epoch": 1.6630539802562487,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047242582214177616,
      "loss": 5.1182,
      "mean_token_accuracy": 0.18691457360982894,
      "num_tokens": 36502289.0,
      "step": 19795
    },
    {
      "entropy": 5.7376142024993895,
      "epoch": 1.6634740600714135,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047241145588429483,
      "loss": 5.249,
      "mean_token_accuracy": 0.18273103088140488,
      "num_tokens": 36511978.0,
      "step": 19800
    },
    {
      "entropy": 5.728275775909424,
      "epoch": 1.6638941398865783,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004723970861297712,
      "loss": 5.2274,
      "mean_token_accuracy": 0.1878792569041252,
      "num_tokens": 36520378.0,
      "step": 19805
    },
    {
      "entropy": 5.723746490478516,
      "epoch": 1.6643142197017433,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004723827128784599,
      "loss": 5.198,
      "mean_token_accuracy": 0.18488269448280334,
      "num_tokens": 36529965.0,
      "step": 19810
    },
    {
      "entropy": 5.82201132774353,
      "epoch": 1.6647342995169083,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047236833613061534,
      "loss": 5.2208,
      "mean_token_accuracy": 0.1863059416413307,
      "num_tokens": 36539394.0,
      "step": 19815
    },
    {
      "entropy": 5.722127771377563,
      "epoch": 1.665154379332073,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004723539558864925,
      "loss": 5.2718,
      "mean_token_accuracy": 0.1857059568166733,
      "num_tokens": 36548608.0,
      "step": 19820
    },
    {
      "entropy": 5.691886043548584,
      "epoch": 1.665574459147238,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004723395721463459,
      "loss": 5.1665,
      "mean_token_accuracy": 0.18764622509479523,
      "num_tokens": 36557736.0,
      "step": 19825
    },
    {
      "entropy": 5.707265186309814,
      "epoch": 1.665994538962403,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004723251849104303,
      "loss": 5.1747,
      "mean_token_accuracy": 0.18576338291168212,
      "num_tokens": 36566745.0,
      "step": 19830
    },
    {
      "entropy": 5.69480333328247,
      "epoch": 1.6664146187775677,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047231079417900076,
      "loss": 5.1312,
      "mean_token_accuracy": 0.18793563097715377,
      "num_tokens": 36575956.0,
      "step": 19835
    },
    {
      "entropy": 5.700874996185303,
      "epoch": 1.6668346985927327,
      "grad_norm": 1.125,
      "learning_rate": 0.000472296399952312,
      "loss": 5.1608,
      "mean_token_accuracy": 0.192136187851429,
      "num_tokens": 36584673.0,
      "step": 19840
    },
    {
      "entropy": 5.702242660522461,
      "epoch": 1.6672547784078975,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004722820022306192,
      "loss": 5.2176,
      "mean_token_accuracy": 0.18994000256061555,
      "num_tokens": 36593758.0,
      "step": 19845
    },
    {
      "entropy": 5.618290185928345,
      "epoch": 1.6676748582230623,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004722676010141773,
      "loss": 5.0875,
      "mean_token_accuracy": 0.19148144870996475,
      "num_tokens": 36603722.0,
      "step": 19850
    },
    {
      "entropy": 5.728228330612183,
      "epoch": 1.6680949380382273,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047225319630324136,
      "loss": 5.1449,
      "mean_token_accuracy": 0.18877629786729813,
      "num_tokens": 36612478.0,
      "step": 19855
    },
    {
      "entropy": 5.75853796005249,
      "epoch": 1.6685150178533923,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004722387880980667,
      "loss": 5.3343,
      "mean_token_accuracy": 0.17735987454652785,
      "num_tokens": 36622399.0,
      "step": 19860
    },
    {
      "entropy": 5.736996746063232,
      "epoch": 1.668935097668557,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047222437639890844,
      "loss": 5.1661,
      "mean_token_accuracy": 0.18879138976335524,
      "num_tokens": 36631798.0,
      "step": 19865
    },
    {
      "entropy": 5.630416917800903,
      "epoch": 1.6693551774837219,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047220996120602197,
      "loss": 5.1866,
      "mean_token_accuracy": 0.18843229562044145,
      "num_tokens": 36640405.0,
      "step": 19870
    },
    {
      "entropy": 5.808610725402832,
      "epoch": 1.6697752572988867,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047219554251966246,
      "loss": 5.3059,
      "mean_token_accuracy": 0.17979287058115007,
      "num_tokens": 36650209.0,
      "step": 19875
    },
    {
      "entropy": 5.778086757659912,
      "epoch": 1.6701953371140517,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004721811203400855,
      "loss": 5.2753,
      "mean_token_accuracy": 0.1846320554614067,
      "num_tokens": 36660248.0,
      "step": 19880
    },
    {
      "entropy": 5.705180978775024,
      "epoch": 1.6706154169292167,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047216669466754657,
      "loss": 5.1615,
      "mean_token_accuracy": 0.1876298099756241,
      "num_tokens": 36669938.0,
      "step": 19885
    },
    {
      "entropy": 5.641536855697632,
      "epoch": 1.6710354967443815,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004721522655023012,
      "loss": 5.1752,
      "mean_token_accuracy": 0.19321906417608262,
      "num_tokens": 36679903.0,
      "step": 19890
    },
    {
      "entropy": 5.761877107620239,
      "epoch": 1.6714555765595462,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004721378328446049,
      "loss": 5.2525,
      "mean_token_accuracy": 0.19090331792831422,
      "num_tokens": 36688424.0,
      "step": 19895
    },
    {
      "entropy": 5.814555311203003,
      "epoch": 1.6718756563747112,
      "grad_norm": 1.0,
      "learning_rate": 0.0004721233966947134,
      "loss": 5.3189,
      "mean_token_accuracy": 0.1844882920384407,
      "num_tokens": 36698715.0,
      "step": 19900
    },
    {
      "entropy": 5.581263828277588,
      "epoch": 1.672295736189876,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047210895705288237,
      "loss": 5.1843,
      "mean_token_accuracy": 0.20274797677993775,
      "num_tokens": 36708456.0,
      "step": 19905
    },
    {
      "entropy": 5.724348783493042,
      "epoch": 1.672715816005041,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004720945139193678,
      "loss": 5.1961,
      "mean_token_accuracy": 0.1857710972428322,
      "num_tokens": 36717596.0,
      "step": 19910
    },
    {
      "entropy": 5.783181619644165,
      "epoch": 1.6731358958202058,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004720800672944253,
      "loss": 5.2845,
      "mean_token_accuracy": 0.18206883519887923,
      "num_tokens": 36727092.0,
      "step": 19915
    },
    {
      "entropy": 5.709772348403931,
      "epoch": 1.6735559756353706,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004720656171783109,
      "loss": 5.0183,
      "mean_token_accuracy": 0.20082486867904664,
      "num_tokens": 36735910.0,
      "step": 19920
    },
    {
      "entropy": 5.6188493251800535,
      "epoch": 1.6739760554505356,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004720511635712806,
      "loss": 5.1315,
      "mean_token_accuracy": 0.19211622178554535,
      "num_tokens": 36745237.0,
      "step": 19925
    },
    {
      "entropy": 5.722348022460937,
      "epoch": 1.6743961352657006,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047203670647359035,
      "loss": 5.2724,
      "mean_token_accuracy": 0.18657598197460173,
      "num_tokens": 36753603.0,
      "step": 19930
    },
    {
      "entropy": 5.809965753555298,
      "epoch": 1.6748162150808654,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004720222458854964,
      "loss": 5.2676,
      "mean_token_accuracy": 0.1822372257709503,
      "num_tokens": 36763010.0,
      "step": 19935
    },
    {
      "entropy": 5.732275819778442,
      "epoch": 1.6752362948960302,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047200778180725477,
      "loss": 5.1824,
      "mean_token_accuracy": 0.19167154133319855,
      "num_tokens": 36772156.0,
      "step": 19940
    },
    {
      "entropy": 5.685102796554565,
      "epoch": 1.675656374711195,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047199331423912174,
      "loss": 5.1031,
      "mean_token_accuracy": 0.19108653217554092,
      "num_tokens": 36781386.0,
      "step": 19945
    },
    {
      "entropy": 5.758098506927491,
      "epoch": 1.67607645452636,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004719788431813536,
      "loss": 5.2448,
      "mean_token_accuracy": 0.18592525869607926,
      "num_tokens": 36790754.0,
      "step": 19950
    },
    {
      "entropy": 5.790366888046265,
      "epoch": 1.676496534341525,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004719643686342066,
      "loss": 5.2259,
      "mean_token_accuracy": 0.1881207749247551,
      "num_tokens": 36799623.0,
      "step": 19955
    },
    {
      "entropy": 5.585662221908569,
      "epoch": 1.6769166141566898,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004719498905979373,
      "loss": 5.0546,
      "mean_token_accuracy": 0.19857629686594008,
      "num_tokens": 36808662.0,
      "step": 19960
    },
    {
      "entropy": 5.699353122711182,
      "epoch": 1.6773366939718546,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004719354090728021,
      "loss": 5.1552,
      "mean_token_accuracy": 0.1899635300040245,
      "num_tokens": 36817730.0,
      "step": 19965
    },
    {
      "entropy": 5.732377529144287,
      "epoch": 1.6777567737870194,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047192092405905743,
      "loss": 5.1368,
      "mean_token_accuracy": 0.1907976046204567,
      "num_tokens": 36827203.0,
      "step": 19970
    },
    {
      "entropy": 5.78850908279419,
      "epoch": 1.6781768536021844,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004719064355569601,
      "loss": 5.2958,
      "mean_token_accuracy": 0.18399682492017747,
      "num_tokens": 36836145.0,
      "step": 19975
    },
    {
      "entropy": 5.795878601074219,
      "epoch": 1.6785969334173494,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047189194356676666,
      "loss": 5.2861,
      "mean_token_accuracy": 0.18683785200119019,
      "num_tokens": 36845609.0,
      "step": 19980
    },
    {
      "entropy": 5.724456548690796,
      "epoch": 1.6790170132325142,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047187744808873386,
      "loss": 5.309,
      "mean_token_accuracy": 0.18302555680274962,
      "num_tokens": 36855367.0,
      "step": 19985
    },
    {
      "entropy": 5.752686977386475,
      "epoch": 1.679437093047679,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047186294912311835,
      "loss": 5.2601,
      "mean_token_accuracy": 0.18197859674692154,
      "num_tokens": 36864808.0,
      "step": 19990
    },
    {
      "entropy": 5.738765001296997,
      "epoch": 1.679857172862844,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047184844667017705,
      "loss": 5.1344,
      "mean_token_accuracy": 0.19879636019468308,
      "num_tokens": 36873651.0,
      "step": 19995
    },
    {
      "entropy": 5.719749450683594,
      "epoch": 1.680277252678009,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047183394073016695,
      "loss": 5.2716,
      "mean_token_accuracy": 0.18222263753414153,
      "num_tokens": 36883227.0,
      "step": 20000
    },
    {
      "entropy": 5.644468164443969,
      "epoch": 1.6806973324931738,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047181943130334493,
      "loss": 5.0599,
      "mean_token_accuracy": 0.19848745614290236,
      "num_tokens": 36891628.0,
      "step": 20005
    },
    {
      "entropy": 5.678480434417724,
      "epoch": 1.6811174123083386,
      "grad_norm": 1.0625,
      "learning_rate": 0.000471804918389968,
      "loss": 5.1711,
      "mean_token_accuracy": 0.18836675882339476,
      "num_tokens": 36901819.0,
      "step": 20010
    },
    {
      "entropy": 5.666687774658203,
      "epoch": 1.6815374921235033,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004717904019902933,
      "loss": 5.1982,
      "mean_token_accuracy": 0.18926816284656525,
      "num_tokens": 36911206.0,
      "step": 20015
    },
    {
      "entropy": 5.658950519561768,
      "epoch": 1.6819575719386683,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000471775882104578,
      "loss": 5.1503,
      "mean_token_accuracy": 0.1888556957244873,
      "num_tokens": 36920830.0,
      "step": 20020
    },
    {
      "entropy": 5.642233037948609,
      "epoch": 1.6823776517538334,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047176135873307917,
      "loss": 5.0627,
      "mean_token_accuracy": 0.19084397405385972,
      "num_tokens": 36929702.0,
      "step": 20025
    },
    {
      "entropy": 5.7849259853363035,
      "epoch": 1.6827977315689981,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004717468318760543,
      "loss": 5.2783,
      "mean_token_accuracy": 0.18618108928203583,
      "num_tokens": 36938423.0,
      "step": 20030
    },
    {
      "entropy": 5.745562267303467,
      "epoch": 1.683217811384163,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047173230153376057,
      "loss": 5.2008,
      "mean_token_accuracy": 0.18411597162485122,
      "num_tokens": 36947198.0,
      "step": 20035
    },
    {
      "entropy": 5.736875104904175,
      "epoch": 1.6836378911993277,
      "grad_norm": 1.125,
      "learning_rate": 0.0004717177677064554,
      "loss": 5.1757,
      "mean_token_accuracy": 0.19616892337799072,
      "num_tokens": 36955636.0,
      "step": 20040
    },
    {
      "entropy": 5.625310468673706,
      "epoch": 1.6840579710144927,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047170323039439634,
      "loss": 5.1344,
      "mean_token_accuracy": 0.18652003556489943,
      "num_tokens": 36964463.0,
      "step": 20045
    },
    {
      "entropy": 5.769357395172119,
      "epoch": 1.6844780508296577,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004716886895978408,
      "loss": 5.2497,
      "mean_token_accuracy": 0.18749321848154069,
      "num_tokens": 36974043.0,
      "step": 20050
    },
    {
      "entropy": 5.693659067153931,
      "epoch": 1.6848981306448225,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047167414531704637,
      "loss": 5.1326,
      "mean_token_accuracy": 0.18924250602722167,
      "num_tokens": 36983856.0,
      "step": 20055
    },
    {
      "entropy": 5.687331962585449,
      "epoch": 1.6853182104599873,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047165959755227077,
      "loss": 5.171,
      "mean_token_accuracy": 0.1922578826546669,
      "num_tokens": 36992664.0,
      "step": 20060
    },
    {
      "entropy": 5.704150581359864,
      "epoch": 1.6857382902751523,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047164504630377166,
      "loss": 5.208,
      "mean_token_accuracy": 0.19431670606136323,
      "num_tokens": 37001826.0,
      "step": 20065
    },
    {
      "entropy": 5.824293947219848,
      "epoch": 1.6861583700903173,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047163049157180676,
      "loss": 5.2499,
      "mean_token_accuracy": 0.18831366151571274,
      "num_tokens": 37010821.0,
      "step": 20070
    },
    {
      "entropy": 5.733322620391846,
      "epoch": 1.6865784499054821,
      "grad_norm": 1.015625,
      "learning_rate": 0.000471615933356634,
      "loss": 5.2946,
      "mean_token_accuracy": 0.18128153830766677,
      "num_tokens": 37021293.0,
      "step": 20075
    },
    {
      "entropy": 5.671109390258789,
      "epoch": 1.686998529720647,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004716013716585112,
      "loss": 5.1099,
      "mean_token_accuracy": 0.19193612188100814,
      "num_tokens": 37031063.0,
      "step": 20080
    },
    {
      "entropy": 5.6495428562164305,
      "epoch": 1.6874186095358117,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004715868064776964,
      "loss": 5.1703,
      "mean_token_accuracy": 0.19583162665367126,
      "num_tokens": 37040879.0,
      "step": 20085
    },
    {
      "entropy": 5.682411813735962,
      "epoch": 1.6878386893509767,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004715722378144474,
      "loss": 5.0832,
      "mean_token_accuracy": 0.1980195537209511,
      "num_tokens": 37049452.0,
      "step": 20090
    },
    {
      "entropy": 5.649366569519043,
      "epoch": 1.6882587691661417,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004715576656690225,
      "loss": 5.0477,
      "mean_token_accuracy": 0.19456881880760193,
      "num_tokens": 37058010.0,
      "step": 20095
    },
    {
      "entropy": 5.725043106079101,
      "epoch": 1.6886788489813065,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047154309004167984,
      "loss": 5.2418,
      "mean_token_accuracy": 0.17966294437646865,
      "num_tokens": 37067580.0,
      "step": 20100
    },
    {
      "entropy": 5.648288011550903,
      "epoch": 1.6890989287964713,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047152851093267744,
      "loss": 5.1424,
      "mean_token_accuracy": 0.18703756481409073,
      "num_tokens": 37076584.0,
      "step": 20105
    },
    {
      "entropy": 5.685269403457641,
      "epoch": 1.689519008611636,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004715139283422737,
      "loss": 5.1649,
      "mean_token_accuracy": 0.18410371840000153,
      "num_tokens": 37086330.0,
      "step": 20110
    },
    {
      "entropy": 5.782552814483642,
      "epoch": 1.689939088426801,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000471499342270727,
      "loss": 5.2348,
      "mean_token_accuracy": 0.18300855308771133,
      "num_tokens": 37096323.0,
      "step": 20115
    },
    {
      "entropy": 5.660893249511719,
      "epoch": 1.690359168241966,
      "grad_norm": 1.0,
      "learning_rate": 0.00047148475271829556,
      "loss": 5.1532,
      "mean_token_accuracy": 0.18976059556007385,
      "num_tokens": 37106281.0,
      "step": 20120
    },
    {
      "entropy": 5.610343599319458,
      "epoch": 1.6907792480571309,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004714701596852379,
      "loss": 5.1435,
      "mean_token_accuracy": 0.19357367604970932,
      "num_tokens": 37116002.0,
      "step": 20125
    },
    {
      "entropy": 5.693322563171387,
      "epoch": 1.6911993278722957,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004714555631718125,
      "loss": 5.2072,
      "mean_token_accuracy": 0.19255559146404266,
      "num_tokens": 37125125.0,
      "step": 20130
    },
    {
      "entropy": 5.665072870254517,
      "epoch": 1.6916194076874607,
      "grad_norm": 1.109375,
      "learning_rate": 0.000471440963178278,
      "loss": 5.0781,
      "mean_token_accuracy": 0.196279776096344,
      "num_tokens": 37134358.0,
      "step": 20135
    },
    {
      "entropy": 5.7656152725219725,
      "epoch": 1.6920394875026254,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047142635970489293,
      "loss": 5.2338,
      "mean_token_accuracy": 0.18758096545934677,
      "num_tokens": 37143732.0,
      "step": 20140
    },
    {
      "entropy": 5.683068895339966,
      "epoch": 1.6924595673177905,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004714117527519161,
      "loss": 5.1391,
      "mean_token_accuracy": 0.18386450558900833,
      "num_tokens": 37153809.0,
      "step": 20145
    },
    {
      "entropy": 5.6533409595489506,
      "epoch": 1.6928796471329552,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047139714231960616,
      "loss": 5.1612,
      "mean_token_accuracy": 0.18370364010334014,
      "num_tokens": 37163272.0,
      "step": 20150
    },
    {
      "entropy": 5.634261274337769,
      "epoch": 1.69329972694812,
      "grad_norm": 1.15625,
      "learning_rate": 0.000471382528408222,
      "loss": 5.0949,
      "mean_token_accuracy": 0.20036439448595048,
      "num_tokens": 37172323.0,
      "step": 20155
    },
    {
      "entropy": 5.744928550720215,
      "epoch": 1.693719806763285,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004713679110180225,
      "loss": 5.2955,
      "mean_token_accuracy": 0.18750677406787872,
      "num_tokens": 37181262.0,
      "step": 20160
    },
    {
      "entropy": 5.775751829147339,
      "epoch": 1.69413988657845,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004713532901492666,
      "loss": 5.2519,
      "mean_token_accuracy": 0.18976220190525056,
      "num_tokens": 37189576.0,
      "step": 20165
    },
    {
      "entropy": 5.7796690464019775,
      "epoch": 1.6945599663936148,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004713386658022132,
      "loss": 5.2366,
      "mean_token_accuracy": 0.18957336097955704,
      "num_tokens": 37199502.0,
      "step": 20170
    },
    {
      "entropy": 5.711120510101319,
      "epoch": 1.6949800462087796,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004713240379771214,
      "loss": 5.1693,
      "mean_token_accuracy": 0.18549246937036515,
      "num_tokens": 37209028.0,
      "step": 20175
    },
    {
      "entropy": 5.803004407882691,
      "epoch": 1.6954001260239444,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004713094066742505,
      "loss": 5.2938,
      "mean_token_accuracy": 0.18527641147375107,
      "num_tokens": 37218087.0,
      "step": 20180
    },
    {
      "entropy": 5.781187677383423,
      "epoch": 1.6958202058391094,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047129477189385946,
      "loss": 5.2423,
      "mean_token_accuracy": 0.18856725543737413,
      "num_tokens": 37227345.0,
      "step": 20185
    },
    {
      "entropy": 5.770312547683716,
      "epoch": 1.6962402856542744,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004712801336362076,
      "loss": 5.1831,
      "mean_token_accuracy": 0.18440685868263246,
      "num_tokens": 37236011.0,
      "step": 20190
    },
    {
      "entropy": 5.6984014987945555,
      "epoch": 1.6966603654694392,
      "grad_norm": 1.125,
      "learning_rate": 0.0004712654919015543,
      "loss": 5.1648,
      "mean_token_accuracy": 0.19053589999675752,
      "num_tokens": 37244613.0,
      "step": 20195
    },
    {
      "entropy": 5.693202877044678,
      "epoch": 1.697080445284604,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004712508466901588,
      "loss": 5.2042,
      "mean_token_accuracy": 0.18793748766183854,
      "num_tokens": 37253768.0,
      "step": 20200
    },
    {
      "entropy": 5.756471395492554,
      "epoch": 1.697500525099769,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047123619800228057,
      "loss": 5.2512,
      "mean_token_accuracy": 0.18018776327371597,
      "num_tokens": 37263230.0,
      "step": 20205
    },
    {
      "entropy": 5.71506896018982,
      "epoch": 1.6979206049149338,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004712215458381792,
      "loss": 5.1453,
      "mean_token_accuracy": 0.18742875009775162,
      "num_tokens": 37272752.0,
      "step": 20210
    },
    {
      "entropy": 5.772667169570923,
      "epoch": 1.6983406847300988,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004712068901981142,
      "loss": 5.2062,
      "mean_token_accuracy": 0.19145152121782302,
      "num_tokens": 37281465.0,
      "step": 20215
    },
    {
      "entropy": 5.697954273223877,
      "epoch": 1.6987607645452636,
      "grad_norm": 1.25,
      "learning_rate": 0.0004711922310823452,
      "loss": 5.2022,
      "mean_token_accuracy": 0.18724939972162247,
      "num_tokens": 37290408.0,
      "step": 20220
    },
    {
      "entropy": 5.691194200515747,
      "epoch": 1.6991808443604284,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004711775684911318,
      "loss": 5.1508,
      "mean_token_accuracy": 0.19201064854860306,
      "num_tokens": 37298890.0,
      "step": 20225
    },
    {
      "entropy": 5.66837363243103,
      "epoch": 1.6996009241755934,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047116290242473375,
      "loss": 5.172,
      "mean_token_accuracy": 0.18652581572532653,
      "num_tokens": 37307720.0,
      "step": 20230
    },
    {
      "entropy": 5.709434556961059,
      "epoch": 1.7000210039907584,
      "grad_norm": 1.03125,
      "learning_rate": 0.000471148232883411,
      "loss": 5.2119,
      "mean_token_accuracy": 0.18651415705680846,
      "num_tokens": 37317145.0,
      "step": 20235
    },
    {
      "entropy": 5.691120529174805,
      "epoch": 1.7004410838059232,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047113355986742325,
      "loss": 5.1399,
      "mean_token_accuracy": 0.1948271930217743,
      "num_tokens": 37326579.0,
      "step": 20240
    },
    {
      "entropy": 5.719968986511231,
      "epoch": 1.700861163621088,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047111888337703046,
      "loss": 5.2038,
      "mean_token_accuracy": 0.18470802158117294,
      "num_tokens": 37336065.0,
      "step": 20245
    },
    {
      "entropy": 5.6678883075714115,
      "epoch": 1.7012812434362528,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004711042034124926,
      "loss": 5.0988,
      "mean_token_accuracy": 0.19020232558250427,
      "num_tokens": 37345297.0,
      "step": 20250
    },
    {
      "entropy": 5.7430346488952635,
      "epoch": 1.7017013232514178,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004710895199740698,
      "loss": 5.2472,
      "mean_token_accuracy": 0.18694171607494353,
      "num_tokens": 37354942.0,
      "step": 20255
    },
    {
      "entropy": 5.785901594161987,
      "epoch": 1.7021214030665828,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004710748330620222,
      "loss": 5.1476,
      "mean_token_accuracy": 0.19667835533618927,
      "num_tokens": 37364068.0,
      "step": 20260
    },
    {
      "entropy": 5.715058422088623,
      "epoch": 1.7025414828817476,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004710601426766098,
      "loss": 5.2615,
      "mean_token_accuracy": 0.18408912122249604,
      "num_tokens": 37373256.0,
      "step": 20265
    },
    {
      "entropy": 5.635310173034668,
      "epoch": 1.7029615626969123,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047104544881809295,
      "loss": 5.0874,
      "mean_token_accuracy": 0.1965874508023262,
      "num_tokens": 37382098.0,
      "step": 20270
    },
    {
      "entropy": 5.648442173004151,
      "epoch": 1.7033816425120771,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004710307514867319,
      "loss": 5.1053,
      "mean_token_accuracy": 0.19250808656215668,
      "num_tokens": 37390844.0,
      "step": 20275
    },
    {
      "entropy": 5.762359285354615,
      "epoch": 1.7038017223272421,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004710160506827871,
      "loss": 5.1934,
      "mean_token_accuracy": 0.182874895632267,
      "num_tokens": 37399617.0,
      "step": 20280
    },
    {
      "entropy": 5.75444655418396,
      "epoch": 1.7042218021424071,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004710013464065189,
      "loss": 5.2804,
      "mean_token_accuracy": 0.18793714493513108,
      "num_tokens": 37409368.0,
      "step": 20285
    },
    {
      "entropy": 5.658962726593018,
      "epoch": 1.704641881957572,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004709866386581877,
      "loss": 5.1032,
      "mean_token_accuracy": 0.1986285462975502,
      "num_tokens": 37418026.0,
      "step": 20290
    },
    {
      "entropy": 5.677364683151245,
      "epoch": 1.7050619617727367,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047097192743805413,
      "loss": 5.1168,
      "mean_token_accuracy": 0.19037076085805893,
      "num_tokens": 37426850.0,
      "step": 20295
    },
    {
      "entropy": 5.669489479064941,
      "epoch": 1.7054820415879017,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004709572127463788,
      "loss": 5.146,
      "mean_token_accuracy": 0.19233414083719252,
      "num_tokens": 37436631.0,
      "step": 20300
    },
    {
      "entropy": 5.706404829025269,
      "epoch": 1.7059021214030667,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004709424945834223,
      "loss": 5.1693,
      "mean_token_accuracy": 0.18836882710456848,
      "num_tokens": 37445619.0,
      "step": 20305
    },
    {
      "entropy": 5.677819061279297,
      "epoch": 1.7063222012182315,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047092777294944544,
      "loss": 5.1374,
      "mean_token_accuracy": 0.1925772026181221,
      "num_tokens": 37454205.0,
      "step": 20310
    },
    {
      "entropy": 5.713298177719116,
      "epoch": 1.7067422810333963,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000470913047844709,
      "loss": 5.244,
      "mean_token_accuracy": 0.1860815927386284,
      "num_tokens": 37463301.0,
      "step": 20315
    },
    {
      "entropy": 5.7082874298095705,
      "epoch": 1.707162360848561,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047089831926947374,
      "loss": 5.2286,
      "mean_token_accuracy": 0.19539083391427994,
      "num_tokens": 37471937.0,
      "step": 20320
    },
    {
      "entropy": 5.776408863067627,
      "epoch": 1.707582440663726,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004708835872240007,
      "loss": 5.1923,
      "mean_token_accuracy": 0.19652780294418334,
      "num_tokens": 37480779.0,
      "step": 20325
    },
    {
      "entropy": 5.743436098098755,
      "epoch": 1.7080025204788911,
      "grad_norm": 1.0,
      "learning_rate": 0.00047086885170855074,
      "loss": 5.2211,
      "mean_token_accuracy": 0.17878660261631013,
      "num_tokens": 37491053.0,
      "step": 20330
    },
    {
      "entropy": 5.740151739120483,
      "epoch": 1.708422600294056,
      "grad_norm": 1.234375,
      "learning_rate": 0.000470854112723385,
      "loss": 5.1862,
      "mean_token_accuracy": 0.1885814055800438,
      "num_tokens": 37499091.0,
      "step": 20335
    },
    {
      "entropy": 5.7016510486602785,
      "epoch": 1.7088426801092207,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004708393702687644,
      "loss": 5.2487,
      "mean_token_accuracy": 0.18208204954862595,
      "num_tokens": 37507882.0,
      "step": 20340
    },
    {
      "entropy": 5.716127538681031,
      "epoch": 1.7092627599243855,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047082462434495015,
      "loss": 5.1902,
      "mean_token_accuracy": 0.19423203468322753,
      "num_tokens": 37517048.0,
      "step": 20345
    },
    {
      "entropy": 5.787512397766113,
      "epoch": 1.7096828397395505,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004708098749522036,
      "loss": 5.2274,
      "mean_token_accuracy": 0.1848258912563324,
      "num_tokens": 37526355.0,
      "step": 20350
    },
    {
      "entropy": 5.746285820007325,
      "epoch": 1.7101029195547155,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004707951220907859,
      "loss": 5.2875,
      "mean_token_accuracy": 0.18427462875843048,
      "num_tokens": 37535746.0,
      "step": 20355
    },
    {
      "entropy": 5.742266750335693,
      "epoch": 1.7105229993698803,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004707803657609585,
      "loss": 5.2136,
      "mean_token_accuracy": 0.17928823679685593,
      "num_tokens": 37546479.0,
      "step": 20360
    },
    {
      "entropy": 5.799523162841797,
      "epoch": 1.710943079185045,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047076560596298275,
      "loss": 5.257,
      "mean_token_accuracy": 0.18474181443452836,
      "num_tokens": 37556805.0,
      "step": 20365
    },
    {
      "entropy": 5.804301166534424,
      "epoch": 1.71136315900021,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047075084269712,
      "loss": 5.2571,
      "mean_token_accuracy": 0.19106352478265762,
      "num_tokens": 37564748.0,
      "step": 20370
    },
    {
      "entropy": 5.661684322357178,
      "epoch": 1.711783238815375,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004707360759636319,
      "loss": 5.0729,
      "mean_token_accuracy": 0.2015371561050415,
      "num_tokens": 37574674.0,
      "step": 20375
    },
    {
      "entropy": 5.712034416198731,
      "epoch": 1.7122033186305399,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047072130576278,
      "loss": 5.1739,
      "mean_token_accuracy": 0.1889556735754013,
      "num_tokens": 37584459.0,
      "step": 20380
    },
    {
      "entropy": 5.748602867126465,
      "epoch": 1.7126233984457047,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004707065320948259,
      "loss": 5.2189,
      "mean_token_accuracy": 0.19481065571308137,
      "num_tokens": 37593570.0,
      "step": 20385
    },
    {
      "entropy": 5.722158050537109,
      "epoch": 1.7130434782608694,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047069175496003147,
      "loss": 5.2411,
      "mean_token_accuracy": 0.18068318516016008,
      "num_tokens": 37603032.0,
      "step": 20390
    },
    {
      "entropy": 5.692116212844849,
      "epoch": 1.7134635580760345,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004706769743586583,
      "loss": 5.1473,
      "mean_token_accuracy": 0.19191100150346757,
      "num_tokens": 37612404.0,
      "step": 20395
    },
    {
      "entropy": 5.684495544433593,
      "epoch": 1.7138836378911995,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047066219029096837,
      "loss": 5.173,
      "mean_token_accuracy": 0.18579856753349305,
      "num_tokens": 37621933.0,
      "step": 20400
    },
    {
      "entropy": 5.745543146133423,
      "epoch": 1.7143037177063642,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004706474027572234,
      "loss": 5.1865,
      "mean_token_accuracy": 0.19278233945369722,
      "num_tokens": 37632078.0,
      "step": 20405
    },
    {
      "entropy": 5.652585554122925,
      "epoch": 1.714723797521529,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047063261175768543,
      "loss": 5.1247,
      "mean_token_accuracy": 0.1871877908706665,
      "num_tokens": 37641665.0,
      "step": 20410
    },
    {
      "entropy": 5.751262760162353,
      "epoch": 1.7151438773366938,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047061781729261656,
      "loss": 5.1878,
      "mean_token_accuracy": 0.18009450882673264,
      "num_tokens": 37650751.0,
      "step": 20415
    },
    {
      "entropy": 5.685685777664185,
      "epoch": 1.7155639571518588,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047060301936227865,
      "loss": 5.1779,
      "mean_token_accuracy": 0.1908721938729286,
      "num_tokens": 37659165.0,
      "step": 20420
    },
    {
      "entropy": 5.706720590591431,
      "epoch": 1.7159840369670238,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004705882179669341,
      "loss": 5.1877,
      "mean_token_accuracy": 0.18835581690073014,
      "num_tokens": 37668057.0,
      "step": 20425
    },
    {
      "entropy": 5.7053382873535154,
      "epoch": 1.7164041167821886,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004705734131068449,
      "loss": 5.155,
      "mean_token_accuracy": 0.18740259855985641,
      "num_tokens": 37677674.0,
      "step": 20430
    },
    {
      "entropy": 5.720551586151123,
      "epoch": 1.7168241965973534,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004705586047822734,
      "loss": 5.1958,
      "mean_token_accuracy": 0.18874522298574448,
      "num_tokens": 37687009.0,
      "step": 20435
    },
    {
      "entropy": 5.691572523117065,
      "epoch": 1.7172442764125184,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047054379299348194,
      "loss": 5.0611,
      "mean_token_accuracy": 0.1907883405685425,
      "num_tokens": 37696723.0,
      "step": 20440
    },
    {
      "entropy": 5.698027753829956,
      "epoch": 1.7176643562276832,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047052897774073295,
      "loss": 5.1916,
      "mean_token_accuracy": 0.1832045152783394,
      "num_tokens": 37706560.0,
      "step": 20445
    },
    {
      "entropy": 5.703041028976441,
      "epoch": 1.7180844360428482,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047051415902428875,
      "loss": 5.215,
      "mean_token_accuracy": 0.1900903344154358,
      "num_tokens": 37715176.0,
      "step": 20450
    },
    {
      "entropy": 5.696533107757569,
      "epoch": 1.718504515858013,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004704993368444119,
      "loss": 5.1769,
      "mean_token_accuracy": 0.19120315909385682,
      "num_tokens": 37723956.0,
      "step": 20455
    },
    {
      "entropy": 5.707925415039062,
      "epoch": 1.7189245956731778,
      "grad_norm": 1.125,
      "learning_rate": 0.0004704845112013649,
      "loss": 5.1757,
      "mean_token_accuracy": 0.1890694320201874,
      "num_tokens": 37733236.0,
      "step": 20460
    },
    {
      "entropy": 5.73171100616455,
      "epoch": 1.7193446754883428,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004704696820954105,
      "loss": 5.2366,
      "mean_token_accuracy": 0.1865750402212143,
      "num_tokens": 37742626.0,
      "step": 20465
    },
    {
      "entropy": 5.682947778701783,
      "epoch": 1.7197647553035078,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004704548495268113,
      "loss": 5.1199,
      "mean_token_accuracy": 0.19956742227077484,
      "num_tokens": 37751854.0,
      "step": 20470
    },
    {
      "entropy": 5.666686344146728,
      "epoch": 1.7201848351186726,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047044001349583,
      "loss": 5.1518,
      "mean_token_accuracy": 0.19155270904302596,
      "num_tokens": 37760993.0,
      "step": 20475
    },
    {
      "entropy": 5.696970367431641,
      "epoch": 1.7206049149338374,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047042517400272966,
      "loss": 5.2285,
      "mean_token_accuracy": 0.19628766924142838,
      "num_tokens": 37771714.0,
      "step": 20480
    },
    {
      "entropy": 5.740558004379272,
      "epoch": 1.7210249947490022,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004704103310477729,
      "loss": 5.1831,
      "mean_token_accuracy": 0.19742557853460313,
      "num_tokens": 37780653.0,
      "step": 20485
    },
    {
      "entropy": 5.697887134552002,
      "epoch": 1.7214450745641672,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004703954846312228,
      "loss": 5.2148,
      "mean_token_accuracy": 0.1889414057135582,
      "num_tokens": 37790450.0,
      "step": 20490
    },
    {
      "entropy": 5.705566883087158,
      "epoch": 1.7218651543793322,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004703806347533423,
      "loss": 5.2071,
      "mean_token_accuracy": 0.18833397030830384,
      "num_tokens": 37800450.0,
      "step": 20495
    },
    {
      "entropy": 5.755374193191528,
      "epoch": 1.722285234194497,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004703657814143945,
      "loss": 5.2291,
      "mean_token_accuracy": 0.17914355397224427,
      "num_tokens": 37809261.0,
      "step": 20500
    },
    {
      "entropy": 5.713159704208374,
      "epoch": 1.7227053140096618,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004703509246146424,
      "loss": 5.0406,
      "mean_token_accuracy": 0.19280193895101547,
      "num_tokens": 37818244.0,
      "step": 20505
    },
    {
      "entropy": 5.713351678848267,
      "epoch": 1.7231253938248268,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004703360643543493,
      "loss": 5.1405,
      "mean_token_accuracy": 0.19644814878702163,
      "num_tokens": 37828555.0,
      "step": 20510
    },
    {
      "entropy": 5.627626752853393,
      "epoch": 1.7235454736399916,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047032120063377836,
      "loss": 5.1163,
      "mean_token_accuracy": 0.188557967543602,
      "num_tokens": 37837840.0,
      "step": 20515
    },
    {
      "entropy": 5.733564472198486,
      "epoch": 1.7239655534551566,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047030633345319293,
      "loss": 5.165,
      "mean_token_accuracy": 0.18708669394254684,
      "num_tokens": 37846910.0,
      "step": 20520
    },
    {
      "entropy": 5.5913389205932615,
      "epoch": 1.7243856332703213,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047029146281285647,
      "loss": 5.024,
      "mean_token_accuracy": 0.20557920783758163,
      "num_tokens": 37855642.0,
      "step": 20525
    },
    {
      "entropy": 5.700211715698242,
      "epoch": 1.7248057130854861,
      "grad_norm": 1.125,
      "learning_rate": 0.0004702765887130322,
      "loss": 5.1984,
      "mean_token_accuracy": 0.19025878757238388,
      "num_tokens": 37864439.0,
      "step": 20530
    },
    {
      "entropy": 5.7580955028533936,
      "epoch": 1.7252257929006511,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047026171115398377,
      "loss": 5.2585,
      "mean_token_accuracy": 0.18447816520929336,
      "num_tokens": 37873801.0,
      "step": 20535
    },
    {
      "entropy": 5.695023918151856,
      "epoch": 1.7256458727158162,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004702468301359746,
      "loss": 5.1277,
      "mean_token_accuracy": 0.20137574821710585,
      "num_tokens": 37883915.0,
      "step": 20540
    },
    {
      "entropy": 5.727406549453735,
      "epoch": 1.726065952530981,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004702319456592684,
      "loss": 5.2423,
      "mean_token_accuracy": 0.18847788870334625,
      "num_tokens": 37894083.0,
      "step": 20545
    },
    {
      "entropy": 5.741250944137573,
      "epoch": 1.7264860323461457,
      "grad_norm": 1.125,
      "learning_rate": 0.00047021705772412885,
      "loss": 5.251,
      "mean_token_accuracy": 0.18876909762620925,
      "num_tokens": 37902264.0,
      "step": 20550
    },
    {
      "entropy": 5.706295394897461,
      "epoch": 1.7269061121613105,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047020216633081964,
      "loss": 5.135,
      "mean_token_accuracy": 0.192874276638031,
      "num_tokens": 37911071.0,
      "step": 20555
    },
    {
      "entropy": 5.683909320831299,
      "epoch": 1.7273261919764755,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047018727147960453,
      "loss": 5.2357,
      "mean_token_accuracy": 0.18047783225774766,
      "num_tokens": 37920048.0,
      "step": 20560
    },
    {
      "entropy": 5.7170051574707035,
      "epoch": 1.7277462717916405,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047017237317074743,
      "loss": 5.2146,
      "mean_token_accuracy": 0.18825803250074385,
      "num_tokens": 37928877.0,
      "step": 20565
    },
    {
      "entropy": 5.767697334289551,
      "epoch": 1.7281663516068053,
      "grad_norm": 1.0,
      "learning_rate": 0.0004701574714045123,
      "loss": 5.2203,
      "mean_token_accuracy": 0.1853915199637413,
      "num_tokens": 37937860.0,
      "step": 20570
    },
    {
      "entropy": 5.738599586486816,
      "epoch": 1.72858643142197,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047014256618116304,
      "loss": 5.2187,
      "mean_token_accuracy": 0.17895897030830382,
      "num_tokens": 37947588.0,
      "step": 20575
    },
    {
      "entropy": 5.719818592071533,
      "epoch": 1.729006511237135,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047012765750096365,
      "loss": 5.1402,
      "mean_token_accuracy": 0.1950746074318886,
      "num_tokens": 37957598.0,
      "step": 20580
    },
    {
      "entropy": 5.709773397445678,
      "epoch": 1.7294265910523,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047011274536417827,
      "loss": 5.1321,
      "mean_token_accuracy": 0.1952793687582016,
      "num_tokens": 37965294.0,
      "step": 20585
    },
    {
      "entropy": 5.668940782546997,
      "epoch": 1.729846670867465,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047009782977107113,
      "loss": 5.1418,
      "mean_token_accuracy": 0.19681655764579772,
      "num_tokens": 37973977.0,
      "step": 20590
    },
    {
      "entropy": 5.795746183395385,
      "epoch": 1.7302667506826297,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047008291072190634,
      "loss": 5.2803,
      "mean_token_accuracy": 0.18668508976697923,
      "num_tokens": 37984492.0,
      "step": 20595
    },
    {
      "entropy": 5.824600505828857,
      "epoch": 1.7306868304977945,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004700679882169482,
      "loss": 5.2213,
      "mean_token_accuracy": 0.18736924678087236,
      "num_tokens": 37994404.0,
      "step": 20600
    },
    {
      "entropy": 5.676775789260864,
      "epoch": 1.7311069103129595,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004700530622564613,
      "loss": 5.1384,
      "mean_token_accuracy": 0.19940583556890487,
      "num_tokens": 38002659.0,
      "step": 20605
    },
    {
      "entropy": 5.700324249267578,
      "epoch": 1.7315269901281245,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004700381328407096,
      "loss": 5.0991,
      "mean_token_accuracy": 0.1939309373497963,
      "num_tokens": 38012290.0,
      "step": 20610
    },
    {
      "entropy": 5.778468799591065,
      "epoch": 1.7319470699432893,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004700231999699579,
      "loss": 5.224,
      "mean_token_accuracy": 0.1876577228307724,
      "num_tokens": 38022163.0,
      "step": 20615
    },
    {
      "entropy": 5.708511638641357,
      "epoch": 1.732367149758454,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004700082636444706,
      "loss": 5.1677,
      "mean_token_accuracy": 0.1864674434065819,
      "num_tokens": 38031051.0,
      "step": 20620
    },
    {
      "entropy": 5.735758447647095,
      "epoch": 1.7327872295736189,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046999332386451245,
      "loss": 5.1969,
      "mean_token_accuracy": 0.1918060600757599,
      "num_tokens": 38040474.0,
      "step": 20625
    },
    {
      "entropy": 5.715891122817993,
      "epoch": 1.7332073093887839,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046997838063034784,
      "loss": 5.185,
      "mean_token_accuracy": 0.1889556258916855,
      "num_tokens": 38049620.0,
      "step": 20630
    },
    {
      "entropy": 5.672065401077271,
      "epoch": 1.7336273892039489,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046996343394224173,
      "loss": 5.1316,
      "mean_token_accuracy": 0.19490093737840652,
      "num_tokens": 38059866.0,
      "step": 20635
    },
    {
      "entropy": 5.712813663482666,
      "epoch": 1.7340474690191137,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046994848380045866,
      "loss": 5.1272,
      "mean_token_accuracy": 0.19151117354631425,
      "num_tokens": 38068948.0,
      "step": 20640
    },
    {
      "entropy": 5.786874198913575,
      "epoch": 1.7344675488342784,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046993353020526366,
      "loss": 5.2865,
      "mean_token_accuracy": 0.18291630744934081,
      "num_tokens": 38079239.0,
      "step": 20645
    },
    {
      "entropy": 5.732482624053955,
      "epoch": 1.7348876286494432,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004699185731569215,
      "loss": 5.2073,
      "mean_token_accuracy": 0.18588967621326447,
      "num_tokens": 38087999.0,
      "step": 20650
    },
    {
      "entropy": 5.775139999389649,
      "epoch": 1.7353077084646082,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004699036126556972,
      "loss": 5.1922,
      "mean_token_accuracy": 0.18828590959310532,
      "num_tokens": 38096586.0,
      "step": 20655
    },
    {
      "entropy": 5.711859846115113,
      "epoch": 1.7357277882797733,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004698886487018558,
      "loss": 5.166,
      "mean_token_accuracy": 0.1887360394001007,
      "num_tokens": 38104766.0,
      "step": 20660
    },
    {
      "entropy": 5.672710037231445,
      "epoch": 1.736147868094938,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004698736812956623,
      "loss": 5.1808,
      "mean_token_accuracy": 0.1914621040225029,
      "num_tokens": 38113574.0,
      "step": 20665
    },
    {
      "entropy": 5.716278219223023,
      "epoch": 1.7365679479101028,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004698587104373819,
      "loss": 5.1351,
      "mean_token_accuracy": 0.18250806331634523,
      "num_tokens": 38122513.0,
      "step": 20670
    },
    {
      "entropy": 5.66591968536377,
      "epoch": 1.7369880277252678,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046984373612727975,
      "loss": 5.1374,
      "mean_token_accuracy": 0.1822051003575325,
      "num_tokens": 38131105.0,
      "step": 20675
    },
    {
      "entropy": 5.678977632522583,
      "epoch": 1.7374081075404328,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046982875836562116,
      "loss": 5.2074,
      "mean_token_accuracy": 0.18018703907728195,
      "num_tokens": 38140106.0,
      "step": 20680
    },
    {
      "entropy": 5.703153944015503,
      "epoch": 1.7378281873555976,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046981377715267145,
      "loss": 5.1692,
      "mean_token_accuracy": 0.19266608208417893,
      "num_tokens": 38149215.0,
      "step": 20685
    },
    {
      "entropy": 5.74072642326355,
      "epoch": 1.7382482671707624,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000469798792488696,
      "loss": 5.091,
      "mean_token_accuracy": 0.19755076766014099,
      "num_tokens": 38157591.0,
      "step": 20690
    },
    {
      "entropy": 5.643893194198609,
      "epoch": 1.7386683469859272,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004697838043739602,
      "loss": 5.2011,
      "mean_token_accuracy": 0.18713759481906891,
      "num_tokens": 38167673.0,
      "step": 20695
    },
    {
      "entropy": 5.771096134185791,
      "epoch": 1.7390884268010922,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046976881280872974,
      "loss": 5.1706,
      "mean_token_accuracy": 0.19010260105133056,
      "num_tokens": 38177586.0,
      "step": 20700
    },
    {
      "entropy": 5.775089597702026,
      "epoch": 1.7395085066162572,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004697538177932699,
      "loss": 5.182,
      "mean_token_accuracy": 0.19288280308246614,
      "num_tokens": 38187020.0,
      "step": 20705
    },
    {
      "entropy": 5.579402160644531,
      "epoch": 1.739928586431422,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004697388193278465,
      "loss": 4.9598,
      "mean_token_accuracy": 0.21227476745843887,
      "num_tokens": 38195705.0,
      "step": 20710
    },
    {
      "entropy": 5.621748638153076,
      "epoch": 1.7403486662465868,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004697238174127252,
      "loss": 5.0892,
      "mean_token_accuracy": 0.19653374552726746,
      "num_tokens": 38204726.0,
      "step": 20715
    },
    {
      "entropy": 5.683867359161377,
      "epoch": 1.7407687460617516,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004697088120481717,
      "loss": 5.1978,
      "mean_token_accuracy": 0.1896982803940773,
      "num_tokens": 38214376.0,
      "step": 20720
    },
    {
      "entropy": 5.67055082321167,
      "epoch": 1.7411888258769166,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004696938032344519,
      "loss": 5.0946,
      "mean_token_accuracy": 0.19605673849582672,
      "num_tokens": 38223631.0,
      "step": 20725
    },
    {
      "entropy": 5.722799777984619,
      "epoch": 1.7416089056920816,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004696787909718317,
      "loss": 5.1365,
      "mean_token_accuracy": 0.20132384747266768,
      "num_tokens": 38233519.0,
      "step": 20730
    },
    {
      "entropy": 5.71354455947876,
      "epoch": 1.7420289855072464,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046966377526057686,
      "loss": 5.1188,
      "mean_token_accuracy": 0.1973813757300377,
      "num_tokens": 38242340.0,
      "step": 20735
    },
    {
      "entropy": 5.653933048248291,
      "epoch": 1.7424490653224112,
      "grad_norm": 1.125,
      "learning_rate": 0.0004696487561009535,
      "loss": 5.0986,
      "mean_token_accuracy": 0.19205507934093474,
      "num_tokens": 38251194.0,
      "step": 20740
    },
    {
      "entropy": 5.658127164840698,
      "epoch": 1.7428691451375762,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004696337334932277,
      "loss": 5.1791,
      "mean_token_accuracy": 0.1819814473390579,
      "num_tokens": 38259938.0,
      "step": 20745
    },
    {
      "entropy": 5.697735738754273,
      "epoch": 1.743289224952741,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046961870743766546,
      "loss": 5.178,
      "mean_token_accuracy": 0.18942587822675705,
      "num_tokens": 38268073.0,
      "step": 20750
    },
    {
      "entropy": 5.727962875366211,
      "epoch": 1.743709304767906,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046960367793453313,
      "loss": 5.2604,
      "mean_token_accuracy": 0.1859055832028389,
      "num_tokens": 38277667.0,
      "step": 20755
    },
    {
      "entropy": 5.7806943416595455,
      "epoch": 1.7441293845830708,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046958864498409673,
      "loss": 5.2066,
      "mean_token_accuracy": 0.1874053567647934,
      "num_tokens": 38287142.0,
      "step": 20760
    },
    {
      "entropy": 5.767076063156128,
      "epoch": 1.7445494643982355,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046957360858662276,
      "loss": 5.196,
      "mean_token_accuracy": 0.193208509683609,
      "num_tokens": 38296199.0,
      "step": 20765
    },
    {
      "entropy": 5.734447050094604,
      "epoch": 1.7449695442134006,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004695585687423775,
      "loss": 5.2081,
      "mean_token_accuracy": 0.1833066925406456,
      "num_tokens": 38305412.0,
      "step": 20770
    },
    {
      "entropy": 5.630394220352173,
      "epoch": 1.7453896240285656,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004695435254516273,
      "loss": 5.1485,
      "mean_token_accuracy": 0.19791667014360428,
      "num_tokens": 38313898.0,
      "step": 20775
    },
    {
      "entropy": 5.763130521774292,
      "epoch": 1.7458097038437304,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004695284787146388,
      "loss": 5.2432,
      "mean_token_accuracy": 0.1843920901417732,
      "num_tokens": 38322835.0,
      "step": 20780
    },
    {
      "entropy": 5.66911358833313,
      "epoch": 1.7462297836588951,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004695134285316784,
      "loss": 5.0752,
      "mean_token_accuracy": 0.19947534650564194,
      "num_tokens": 38331448.0,
      "step": 20785
    },
    {
      "entropy": 5.725335931777954,
      "epoch": 1.74664986347406,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046949837490301293,
      "loss": 5.1975,
      "mean_token_accuracy": 0.1880056545138359,
      "num_tokens": 38340837.0,
      "step": 20790
    },
    {
      "entropy": 5.714566469192505,
      "epoch": 1.747069943289225,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004694833178289088,
      "loss": 5.1491,
      "mean_token_accuracy": 0.19455148428678512,
      "num_tokens": 38349363.0,
      "step": 20795
    },
    {
      "entropy": 5.682427597045899,
      "epoch": 1.74749002310439,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004694682573096328,
      "loss": 5.1852,
      "mean_token_accuracy": 0.18709126114845276,
      "num_tokens": 38358017.0,
      "step": 20800
    },
    {
      "entropy": 5.6699927806854244,
      "epoch": 1.7479101029195547,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046945319334545184,
      "loss": 5.1889,
      "mean_token_accuracy": 0.18587440401315689,
      "num_tokens": 38367256.0,
      "step": 20805
    },
    {
      "entropy": 5.669847393035889,
      "epoch": 1.7483301827347195,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004694381259366327,
      "loss": 5.175,
      "mean_token_accuracy": 0.18843374103307725,
      "num_tokens": 38376169.0,
      "step": 20810
    },
    {
      "entropy": 5.729050397872925,
      "epoch": 1.7487502625498845,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046942305508344216,
      "loss": 5.1502,
      "mean_token_accuracy": 0.18918030709028244,
      "num_tokens": 38385379.0,
      "step": 20815
    },
    {
      "entropy": 5.758592033386231,
      "epoch": 1.7491703423650493,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004694079807861473,
      "loss": 5.2259,
      "mean_token_accuracy": 0.18782172054052354,
      "num_tokens": 38395217.0,
      "step": 20820
    },
    {
      "entropy": 5.7068071365356445,
      "epoch": 1.7495904221802143,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004693929030450153,
      "loss": 5.1487,
      "mean_token_accuracy": 0.19230168610811232,
      "num_tokens": 38404347.0,
      "step": 20825
    },
    {
      "entropy": 5.748762893676758,
      "epoch": 1.750010501995379,
      "grad_norm": 1.125,
      "learning_rate": 0.00046937782186031303,
      "loss": 5.101,
      "mean_token_accuracy": 0.1895531088113785,
      "num_tokens": 38413394.0,
      "step": 20830
    },
    {
      "entropy": 5.697626066207886,
      "epoch": 1.750430581810544,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004693627372323078,
      "loss": 5.1821,
      "mean_token_accuracy": 0.18667229413986205,
      "num_tokens": 38422043.0,
      "step": 20835
    },
    {
      "entropy": 5.783128452301026,
      "epoch": 1.750850661625709,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004693476491612667,
      "loss": 5.3131,
      "mean_token_accuracy": 0.18373072892427444,
      "num_tokens": 38430792.0,
      "step": 20840
    },
    {
      "entropy": 5.59730076789856,
      "epoch": 1.751270741440874,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004693325576474571,
      "loss": 5.1194,
      "mean_token_accuracy": 0.19102779477834703,
      "num_tokens": 38439105.0,
      "step": 20845
    },
    {
      "entropy": 5.718141889572143,
      "epoch": 1.7516908212560387,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004693174626911463,
      "loss": 5.1533,
      "mean_token_accuracy": 0.19048849493265152,
      "num_tokens": 38447944.0,
      "step": 20850
    },
    {
      "entropy": 5.7344249248504635,
      "epoch": 1.7521109010712035,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046930236429260173,
      "loss": 5.1861,
      "mean_token_accuracy": 0.18425924181938172,
      "num_tokens": 38457206.0,
      "step": 20855
    },
    {
      "entropy": 5.767601680755615,
      "epoch": 1.7525309808863683,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004692872624520908,
      "loss": 5.2675,
      "mean_token_accuracy": 0.1812077760696411,
      "num_tokens": 38467085.0,
      "step": 20860
    },
    {
      "entropy": 5.7734287738800045,
      "epoch": 1.7529510607015333,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000469272157169881,
      "loss": 5.1218,
      "mean_token_accuracy": 0.19055947363376619,
      "num_tokens": 38475970.0,
      "step": 20865
    },
    {
      "entropy": 5.717898511886597,
      "epoch": 1.7533711405166983,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004692570484462401,
      "loss": 5.2303,
      "mean_token_accuracy": 0.1893252745270729,
      "num_tokens": 38484579.0,
      "step": 20870
    },
    {
      "entropy": 5.718835926055908,
      "epoch": 1.753791220331863,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046924193628143554,
      "loss": 5.2551,
      "mean_token_accuracy": 0.1856676310300827,
      "num_tokens": 38495107.0,
      "step": 20875
    },
    {
      "entropy": 5.805339622497558,
      "epoch": 1.7542113001470279,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046922682067573516,
      "loss": 5.2561,
      "mean_token_accuracy": 0.18898842632770538,
      "num_tokens": 38505731.0,
      "step": 20880
    },
    {
      "entropy": 5.684660625457764,
      "epoch": 1.7546313799621929,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046921170162940657,
      "loss": 5.175,
      "mean_token_accuracy": 0.19159777760505675,
      "num_tokens": 38514483.0,
      "step": 20885
    },
    {
      "entropy": 5.70018424987793,
      "epoch": 1.7550514597773577,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046919657914271774,
      "loss": 5.0766,
      "mean_token_accuracy": 0.19790809750556945,
      "num_tokens": 38522953.0,
      "step": 20890
    },
    {
      "entropy": 5.681619691848755,
      "epoch": 1.7554715395925227,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004691814532159365,
      "loss": 5.0747,
      "mean_token_accuracy": 0.20359453111886977,
      "num_tokens": 38531891.0,
      "step": 20895
    },
    {
      "entropy": 5.725100326538086,
      "epoch": 1.7558916194076875,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004691663238493308,
      "loss": 5.2519,
      "mean_token_accuracy": 0.19216530919075012,
      "num_tokens": 38541609.0,
      "step": 20900
    },
    {
      "entropy": 5.729188251495361,
      "epoch": 1.7563116992228522,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004691511910431686,
      "loss": 5.2567,
      "mean_token_accuracy": 0.18739428818225862,
      "num_tokens": 38550348.0,
      "step": 20905
    },
    {
      "entropy": 5.697227478027344,
      "epoch": 1.7567317790380172,
      "grad_norm": 1.125,
      "learning_rate": 0.0004691360547977181,
      "loss": 5.1069,
      "mean_token_accuracy": 0.19705526232719422,
      "num_tokens": 38559493.0,
      "step": 20910
    },
    {
      "entropy": 5.719883012771606,
      "epoch": 1.7571518588531823,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004691209151132474,
      "loss": 5.1402,
      "mean_token_accuracy": 0.17976868748664857,
      "num_tokens": 38567888.0,
      "step": 20915
    },
    {
      "entropy": 5.719273424148559,
      "epoch": 1.757571938668347,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004691057719900246,
      "loss": 5.2091,
      "mean_token_accuracy": 0.18506319224834442,
      "num_tokens": 38577216.0,
      "step": 20920
    },
    {
      "entropy": 5.69208402633667,
      "epoch": 1.7579920184835118,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046909062542831794,
      "loss": 5.1508,
      "mean_token_accuracy": 0.19662615954875945,
      "num_tokens": 38586258.0,
      "step": 20925
    },
    {
      "entropy": 5.707313680648804,
      "epoch": 1.7584120982986766,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004690754754283959,
      "loss": 5.1,
      "mean_token_accuracy": 0.1959293767809868,
      "num_tokens": 38594900.0,
      "step": 20930
    },
    {
      "entropy": 5.664283275604248,
      "epoch": 1.7588321781138416,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004690603219905266,
      "loss": 5.1865,
      "mean_token_accuracy": 0.1892174407839775,
      "num_tokens": 38603980.0,
      "step": 20935
    },
    {
      "entropy": 5.749043989181518,
      "epoch": 1.7592522579290066,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046904516511497873,
      "loss": 5.2609,
      "mean_token_accuracy": 0.18150665163993834,
      "num_tokens": 38613888.0,
      "step": 20940
    },
    {
      "entropy": 5.747987747192383,
      "epoch": 1.7596723377441714,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046903000480202065,
      "loss": 5.206,
      "mean_token_accuracy": 0.18404597640037537,
      "num_tokens": 38623969.0,
      "step": 20945
    },
    {
      "entropy": 5.684590673446655,
      "epoch": 1.7600924175593362,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046901484105192094,
      "loss": 5.1776,
      "mean_token_accuracy": 0.18877159655094147,
      "num_tokens": 38633387.0,
      "step": 20950
    },
    {
      "entropy": 5.743253946304321,
      "epoch": 1.760512497374501,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046899967386494816,
      "loss": 5.2315,
      "mean_token_accuracy": 0.18162560015916823,
      "num_tokens": 38642481.0,
      "step": 20955
    },
    {
      "entropy": 5.748624849319458,
      "epoch": 1.760932577189666,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004689845032413712,
      "loss": 5.1987,
      "mean_token_accuracy": 0.1854836642742157,
      "num_tokens": 38652345.0,
      "step": 20960
    },
    {
      "entropy": 5.792650461196899,
      "epoch": 1.761352657004831,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004689693291814586,
      "loss": 5.2422,
      "mean_token_accuracy": 0.1849406659603119,
      "num_tokens": 38661529.0,
      "step": 20965
    },
    {
      "entropy": 5.707007217407226,
      "epoch": 1.7617727368199958,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004689541516854791,
      "loss": 5.1347,
      "mean_token_accuracy": 0.18912262320518494,
      "num_tokens": 38670191.0,
      "step": 20970
    },
    {
      "entropy": 5.7172935962677,
      "epoch": 1.7621928166351606,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004689389707537018,
      "loss": 5.2329,
      "mean_token_accuracy": 0.18542350232601165,
      "num_tokens": 38679089.0,
      "step": 20975
    },
    {
      "entropy": 5.728964757919312,
      "epoch": 1.7626128964503256,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046892378638639545,
      "loss": 5.1558,
      "mean_token_accuracy": 0.1913009688258171,
      "num_tokens": 38688821.0,
      "step": 20980
    },
    {
      "entropy": 5.7453234672546385,
      "epoch": 1.7630329762654906,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046890859858382913,
      "loss": 5.2437,
      "mean_token_accuracy": 0.17898193448781968,
      "num_tokens": 38698232.0,
      "step": 20985
    },
    {
      "entropy": 5.840619850158691,
      "epoch": 1.7634530560806554,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004688934073462718,
      "loss": 5.3014,
      "mean_token_accuracy": 0.17906804382801056,
      "num_tokens": 38708090.0,
      "step": 20990
    },
    {
      "entropy": 5.756931304931641,
      "epoch": 1.7638731358958202,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046887821267399256,
      "loss": 5.1924,
      "mean_token_accuracy": 0.19723644107580185,
      "num_tokens": 38717370.0,
      "step": 20995
    },
    {
      "entropy": 5.746652746200562,
      "epoch": 1.764293215710985,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004688630145672607,
      "loss": 5.176,
      "mean_token_accuracy": 0.18925876170396805,
      "num_tokens": 38726758.0,
      "step": 21000
    },
    {
      "epoch": 1.764293215710985,
      "eval_entropy": 5.487146126435365,
      "eval_loss": 5.215159893035889,
      "eval_mean_token_accuracy": 0.19715142748520986,
      "eval_num_tokens": 38726758.0,
      "eval_runtime": 20.9191,
      "eval_samples_per_second": 1786.211,
      "eval_steps_per_second": 223.288,
      "step": 21000
    }
  ],
  "logging_steps": 5,
  "max_steps": 119020,
  "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": 8398625696317440.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}