{
  "best_global_step": 2250,
  "best_metric": 1.8998193594895059,
  "best_model_checkpoint": "/content/cot-split2/checkpoint-2000",
  "epoch": 2.0,
  "eval_steps": 250,
  "global_step": 2554,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0,
      "eval_bpd": 5.449506205976702,
      "eval_entropy": 3.8259646681880883,
      "eval_loss": 3.7819511890411377,
      "eval_mean_token_accuracy": 0.3298281561563004,
      "eval_nll": 3.777309862116675,
      "eval_num_tokens": 0.0,
      "eval_ppl": 43.69832894461198,
      "eval_runtime": 318.4675,
      "eval_samples_per_second": 570.259,
      "eval_steps_per_second": 8.911,
      "step": 0
    },
    {
      "entropy": 4.035161377492861,
      "epoch": 0.0007831466833737959,
      "grad_norm": 2.9375,
      "learning_rate": 0.0,
      "loss": 3.6375203132629395,
      "mean_token_accuracy": 0.3242907148120855,
      "num_tokens": 156431.0,
      "step": 1
    },
    {
      "entropy": 4.157279884560115,
      "epoch": 0.019578667084344897,
      "grad_norm": 1.8515625,
      "learning_rate": 9.375e-05,
      "loss": 3.7293694814046225,
      "mean_token_accuracy": 0.3223868495477843,
      "num_tokens": 3882398.0,
      "step": 25
    },
    {
      "entropy": 4.069613211207327,
      "epoch": 0.03915733416868979,
      "grad_norm": 6.65625,
      "learning_rate": 0.00019140625,
      "loss": 3.286188659667969,
      "mean_token_accuracy": 0.3470086962011918,
      "num_tokens": 7771579.0,
      "step": 50
    },
    {
      "entropy": 3.712915897718112,
      "epoch": 0.05873600125303469,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002890625,
      "loss": 2.9255511474609377,
      "mean_token_accuracy": 0.37689020381328076,
      "num_tokens": 11677887.0,
      "step": 75
    },
    {
      "entropy": 3.4658741959475385,
      "epoch": 0.07831466833737959,
      "grad_norm": 1.671875,
      "learning_rate": 0.00038671875,
      "loss": 2.5379095458984375,
      "mean_token_accuracy": 0.41490777290923164,
      "num_tokens": 15584019.0,
      "step": 100
    },
    {
      "entropy": 3.295554994185651,
      "epoch": 0.09789333542172449,
      "grad_norm": 0.59765625,
      "learning_rate": 0.000484375,
      "loss": 2.405628204345703,
      "mean_token_accuracy": 0.4313112262648087,
      "num_tokens": 19484337.0,
      "step": 125
    },
    {
      "entropy": 3.1617948360015533,
      "epoch": 0.11747200250606939,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004999075642455791,
      "loss": 2.3207965087890625,
      "mean_token_accuracy": 0.4436994497174724,
      "num_tokens": 23377294.0,
      "step": 150
    },
    {
      "entropy": 3.0947463825081942,
      "epoch": 0.1370506695904143,
      "grad_norm": 0.6640625,
      "learning_rate": 0.0004995565798713207,
      "loss": 2.2583859252929686,
      "mean_token_accuracy": 0.45017372579314874,
      "num_tokens": 27287783.0,
      "step": 175
    },
    {
      "entropy": 3.045135160417696,
      "epoch": 0.15662933667475917,
      "grad_norm": 0.65625,
      "learning_rate": 0.0004989440608306114,
      "loss": 2.230653076171875,
      "mean_token_accuracy": 0.4551180636124264,
      "num_tokens": 31205249.0,
      "step": 200
    },
    {
      "entropy": 3.0288844077617805,
      "epoch": 0.1762080037591041,
      "grad_norm": 0.625,
      "learning_rate": 0.0004980706490418613,
      "loss": 2.214619140625,
      "mean_token_accuracy": 0.4571791829436394,
      "num_tokens": 35081817.0,
      "step": 225
    },
    {
      "entropy": 2.9857863094922594,
      "epoch": 0.19578667084344897,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0004969372598384225,
      "loss": 2.183477783203125,
      "mean_token_accuracy": 0.4623076980301137,
      "num_tokens": 38961791.0,
      "step": 250
    },
    {
      "epoch": 0.19578667084344897,
      "eval_bpd": 3.0542670212451286,
      "eval_entropy": 2.831413693293467,
      "eval_loss": 2.114036798477173,
      "eval_mean_token_accuracy": 0.471896302096762,
      "eval_nll": 2.1170565744532834,
      "eval_num_tokens": 38961791.0,
      "eval_ppl": 8.306651458853036,
      "eval_runtime": 317.7783,
      "eval_samples_per_second": 571.496,
      "eval_steps_per_second": 8.931,
      "step": 250
    },
    {
      "entropy": 2.960300336004756,
      "epoch": 0.2153653379277939,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0004955450810093227,
      "loss": 2.1587040710449217,
      "mean_token_accuracy": 0.46429396638467557,
      "num_tokens": 42868157.0,
      "step": 275
    },
    {
      "entropy": 2.9496937473585083,
      "epoch": 0.23494400501213877,
      "grad_norm": 0.6953125,
      "learning_rate": 0.0004938955715544644,
      "loss": 2.1702507019042967,
      "mean_token_accuracy": 0.4644734879386833,
      "num_tokens": 46774567.0,
      "step": 300
    },
    {
      "entropy": 2.925756508647974,
      "epoch": 0.25452267209648366,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0004919904601555969,
      "loss": 2.1320841979980467,
      "mean_token_accuracy": 0.4678374269200783,
      "num_tokens": 50651509.0,
      "step": 325
    },
    {
      "entropy": 2.916845726644516,
      "epoch": 0.2741013391808286,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0004898317433646626,
      "loss": 2.149468994140625,
      "mean_token_accuracy": 0.46948018886239345,
      "num_tokens": 54545113.0,
      "step": 350
    },
    {
      "entropy": 2.8672960236041836,
      "epoch": 0.2936800062651735,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00048742168351141525,
      "loss": 2.095373077392578,
      "mean_token_accuracy": 0.47492304637997923,
      "num_tokens": 58450259.0,
      "step": 375
    },
    {
      "entropy": 2.866450867223391,
      "epoch": 0.31325867334951835,
      "grad_norm": 0.609375,
      "learning_rate": 0.00048476280633250404,
      "loss": 2.093819122314453,
      "mean_token_accuracy": 0.4762016071737522,
      "num_tokens": 62334830.0,
      "step": 400
    },
    {
      "entropy": 2.8444264384733913,
      "epoch": 0.33283734043386326,
      "grad_norm": 0.671875,
      "learning_rate": 0.00048185789832450813,
      "loss": 2.07142822265625,
      "mean_token_accuracy": 0.47914267030296726,
      "num_tokens": 66254709.0,
      "step": 425
    },
    {
      "entropy": 2.8643093723566215,
      "epoch": 0.3524160075182082,
      "grad_norm": 0.578125,
      "learning_rate": 0.00047871000382369524,
      "loss": 2.0907044982910157,
      "mean_token_accuracy": 0.475262880935016,
      "num_tokens": 70149264.0,
      "step": 450
    },
    {
      "entropy": 2.8425666356253885,
      "epoch": 0.37199467460255303,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004753224218155648,
      "loss": 2.0714340209960938,
      "mean_token_accuracy": 0.47639239522332366,
      "num_tokens": 74036195.0,
      "step": 475
    },
    {
      "entropy": 2.8495917513303857,
      "epoch": 0.39157334168689795,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0004716987024775197,
      "loss": 2.0957228088378907,
      "mean_token_accuracy": 0.4775797151536097,
      "num_tokens": 77964744.0,
      "step": 500
    },
    {
      "epoch": 0.39157334168689795,
      "eval_bpd": 2.9013035889897933,
      "eval_entropy": 2.6662824092598547,
      "eval_loss": 2.007856607437134,
      "eval_mean_token_accuracy": 0.48950937525505117,
      "eval_nll": 2.0110304026567256,
      "eval_num_tokens": 77964744.0,
      "eval_ppl": 7.471011532715633,
      "eval_runtime": 317.7246,
      "eval_samples_per_second": 571.593,
      "eval_steps_per_second": 8.932,
      "step": 500
    },
    {
      "entropy": 2.8462999362059764,
      "epoch": 0.41115200877124286,
      "grad_norm": 0.59765625,
      "learning_rate": 0.000467842643458289,
      "loss": 2.0840892028808593,
      "mean_token_accuracy": 0.47706405281698494,
      "num_tokens": 81861726.0,
      "step": 525
    },
    {
      "entropy": 2.8082084615239284,
      "epoch": 0.4307306758555878,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0004637582858980017,
      "loss": 2.041808624267578,
      "mean_token_accuracy": 0.4826572669515401,
      "num_tokens": 85760943.0,
      "step": 550
    },
    {
      "entropy": 2.8388380295514866,
      "epoch": 0.45030934293993263,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0004594499101930825,
      "loss": 2.078563385009766,
      "mean_token_accuracy": 0.4769605679393806,
      "num_tokens": 89653726.0,
      "step": 575
    },
    {
      "entropy": 2.8302025972271743,
      "epoch": 0.46988801002427755,
      "grad_norm": 0.578125,
      "learning_rate": 0.00045492203151040655,
      "loss": 2.0540911865234377,
      "mean_token_accuracy": 0.478844059310469,
      "num_tokens": 93570215.0,
      "step": 600
    },
    {
      "entropy": 2.792682902628425,
      "epoch": 0.48946667710862246,
      "grad_norm": 0.69140625,
      "learning_rate": 0.00045017939505541626,
      "loss": 2.0480902099609377,
      "mean_token_accuracy": 0.4848035732412953,
      "num_tokens": 97477693.0,
      "step": 625
    },
    {
      "entropy": 2.7777595483840716,
      "epoch": 0.5090453441929673,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00044522697109915813,
      "loss": 2.0343661499023438,
      "mean_token_accuracy": 0.48608020488521336,
      "num_tokens": 101394752.0,
      "step": 650
    },
    {
      "entropy": 2.742625050496606,
      "epoch": 0.5286240112773123,
      "grad_norm": 0.65625,
      "learning_rate": 0.0004400699497694506,
      "loss": 2.0052694702148437,
      "mean_token_accuracy": 0.4903364669038516,
      "num_tokens": 105270418.0,
      "step": 675
    },
    {
      "entropy": 2.8160673456556586,
      "epoch": 0.5482026783616571,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00043471373561164396,
      "loss": 2.0440855407714844,
      "mean_token_accuracy": 0.4817835272308287,
      "num_tokens": 109173756.0,
      "step": 700
    },
    {
      "entropy": 2.731675431270922,
      "epoch": 0.567781345446002,
      "grad_norm": 0.625,
      "learning_rate": 0.00042916394192466997,
      "loss": 1.9994474792480468,
      "mean_token_accuracy": 0.49231067122440925,
      "num_tokens": 113075567.0,
      "step": 725
    },
    {
      "entropy": 2.761901786283376,
      "epoch": 0.587360012530347,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00042342638487831874,
      "loss": 2.0215072631835938,
      "mean_token_accuracy": 0.48973995306839846,
      "num_tokens": 116976712.0,
      "step": 750
    },
    {
      "epoch": 0.587360012530347,
      "eval_bpd": 2.8346400005221386,
      "eval_entropy": 2.564649322232512,
      "eval_loss": 1.9616204500198364,
      "eval_mean_token_accuracy": 0.4976015657363103,
      "eval_nll": 1.964822724264362,
      "eval_num_tokens": 116976712.0,
      "eval_ppl": 7.133647855387014,
      "eval_runtime": 317.7915,
      "eval_samples_per_second": 571.472,
      "eval_steps_per_second": 8.93,
      "step": 750
    },
    {
      "entropy": 2.772676199080625,
      "epoch": 0.6069386796146918,
      "grad_norm": 0.6484375,
      "learning_rate": 0.0004175070774179078,
      "loss": 2.0193243408203125,
      "mean_token_accuracy": 0.4871398110581337,
      "num_tokens": 120856970.0,
      "step": 775
    },
    {
      "entropy": 2.7522809489804128,
      "epoch": 0.6265173466990367,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0004114122229627302,
      "loss": 2.0201119995117187,
      "mean_token_accuracy": 0.4884969423881558,
      "num_tokens": 124747012.0,
      "step": 800
    },
    {
      "entropy": 2.728323356258852,
      "epoch": 0.6460960137833817,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00040514820890488664,
      "loss": 2.000582427978516,
      "mean_token_accuracy": 0.49145337661979344,
      "num_tokens": 128633902.0,
      "step": 825
    },
    {
      "entropy": 2.7372595225009375,
      "epoch": 0.6656746808677265,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00039872159991531533,
      "loss": 2.00209716796875,
      "mean_token_accuracy": 0.4912540972003919,
      "num_tokens": 132518194.0,
      "step": 850
    },
    {
      "entropy": 2.738841588033162,
      "epoch": 0.6852533479520714,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0003921391310640325,
      "loss": 1.994276580810547,
      "mean_token_accuracy": 0.491252567341933,
      "num_tokens": 136385199.0,
      "step": 875
    },
    {
      "entropy": 2.7295696415657966,
      "epoch": 0.7048320150364163,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0003854077007617969,
      "loss": 1.9865512084960937,
      "mean_token_accuracy": 0.49203635280876573,
      "num_tokens": 140293961.0,
      "step": 900
    },
    {
      "entropy": 2.7351411749253978,
      "epoch": 0.7244106821207612,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0003785343635305919,
      "loss": 1.9895498657226562,
      "mean_token_accuracy": 0.49209919251085,
      "num_tokens": 144186204.0,
      "step": 925
    },
    {
      "entropy": 2.7418933621077652,
      "epoch": 0.7439893492051061,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00037152632261050483,
      "loss": 2.0013661193847656,
      "mean_token_accuracy": 0.49214851600044945,
      "num_tokens": 148108105.0,
      "step": 950
    },
    {
      "entropy": 2.7243121079421706,
      "epoch": 0.763568016289451,
      "grad_norm": 0.640625,
      "learning_rate": 0.0003643909224107492,
      "loss": 1.991259307861328,
      "mean_token_accuracy": 0.4933121377369521,
      "num_tokens": 152015615.0,
      "step": 975
    },
    {
      "entropy": 2.778145840960037,
      "epoch": 0.7831466833737959,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00035713564081274257,
      "loss": 2.0278318786621092,
      "mean_token_accuracy": 0.48574800621666386,
      "num_tokens": 155927990.0,
      "step": 1000
    },
    {
      "epoch": 0.7831466833737959,
      "eval_bpd": 2.789343226908039,
      "eval_entropy": 2.525867688934465,
      "eval_loss": 1.9301713705062866,
      "eval_mean_token_accuracy": 0.502722952650009,
      "eval_nll": 1.9334253933452872,
      "eval_num_tokens": 155927990.0,
      "eval_ppl": 6.913149988716967,
      "eval_runtime": 317.7734,
      "eval_samples_per_second": 571.505,
      "eval_steps_per_second": 8.931,
      "step": 1000
    },
    {
      "entropy": 2.6940799960317015,
      "epoch": 0.8027253504581409,
      "grad_norm": 0.73046875,
      "learning_rate": 0.0003497680813333055,
      "loss": 1.9498268127441407,
      "mean_token_accuracy": 0.4984518737087283,
      "num_tokens": 159830176.0,
      "step": 1025
    },
    {
      "entropy": 2.7343128778140633,
      "epoch": 0.8223040175424857,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0003422959651561952,
      "loss": 1.9870970153808594,
      "mean_token_accuracy": 0.4924939781360015,
      "num_tokens": 163728360.0,
      "step": 1050
    },
    {
      "entropy": 2.7091974552364038,
      "epoch": 0.8418826846268306,
      "grad_norm": 0.546875,
      "learning_rate": 0.0003347271230403245,
      "loss": 1.9727362060546876,
      "mean_token_accuracy": 0.49556485851019944,
      "num_tokens": 167642672.0,
      "step": 1075
    },
    {
      "entropy": 2.7602979798373886,
      "epoch": 0.8614613517111755,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00032706948711314597,
      "loss": 1.992530059814453,
      "mean_token_accuracy": 0.490016912306627,
      "num_tokens": 171525383.0,
      "step": 1100
    },
    {
      "entropy": 2.721992457947341,
      "epoch": 0.8810400187955204,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0003193310825578028,
      "loss": 1.984290771484375,
      "mean_token_accuracy": 0.4936226868736485,
      "num_tokens": 175403307.0,
      "step": 1125
    },
    {
      "entropy": 2.739525592921455,
      "epoch": 0.9006186858798653,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0003115200192027566,
      "loss": 2.005537261962891,
      "mean_token_accuracy": 0.4910348970524133,
      "num_tokens": 179313014.0,
      "step": 1150
    },
    {
      "entropy": 2.7139533455144047,
      "epoch": 0.9201973529642102,
      "grad_norm": 0.7578125,
      "learning_rate": 0.00030364448302270787,
      "loss": 1.9704147338867188,
      "mean_token_accuracy": 0.4949843998242547,
      "num_tokens": 183213621.0,
      "step": 1175
    },
    {
      "entropy": 2.7396354848939612,
      "epoch": 0.9397760200485551,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002957127275597154,
      "loss": 2.010960998535156,
      "mean_token_accuracy": 0.4896595800467586,
      "num_tokens": 187081439.0,
      "step": 1200
    },
    {
      "entropy": 2.768980867176604,
      "epoch": 0.9593546871329,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002877330652735044,
      "loss": 2.016462554931641,
      "mean_token_accuracy": 0.48757137446833304,
      "num_tokens": 190967002.0,
      "step": 1225
    },
    {
      "entropy": 2.704207567026802,
      "epoch": 0.9789333542172449,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0002797138588300303,
      "loss": 1.9740122985839843,
      "mean_token_accuracy": 0.4958090399193563,
      "num_tokens": 194892883.0,
      "step": 1250
    },
    {
      "epoch": 0.9789333542172449,
      "eval_bpd": 2.767466912988666,
      "eval_entropy": 2.5821667777041397,
      "eval_loss": 1.9150066375732422,
      "eval_mean_token_accuracy": 0.5045345765416368,
      "eval_nll": 1.9182618880310294,
      "eval_num_tokens": 194892883.0,
      "eval_ppl": 6.809113176939714,
      "eval_runtime": 317.7422,
      "eval_samples_per_second": 571.561,
      "eval_steps_per_second": 8.932,
      "step": 1250
    },
    {
      "entropy": 2.6901174929114173,
      "epoch": 0.9985120213015898,
      "grad_norm": 0.609375,
      "learning_rate": 0.0002716635123374255,
      "loss": 1.9728475952148437,
      "mean_token_accuracy": 0.49665255697471494,
      "num_tokens": 198787951.0,
      "step": 1275
    },
    {
      "entropy": 2.7251224381449632,
      "epoch": 1.0180123737175972,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00026359046253851537,
      "loss": 1.9876043701171875,
      "mean_token_accuracy": 0.49355816180166034,
      "num_tokens": 202648704.0,
      "step": 1300
    },
    {
      "entropy": 2.7282011647258444,
      "epoch": 1.0375910408019422,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00025550316996913314,
      "loss": 1.9952455139160157,
      "mean_token_accuracy": 0.4923797138191662,
      "num_tokens": 206542240.0,
      "step": 1325
    },
    {
      "entropy": 2.7167434371623265,
      "epoch": 1.0571697078862872,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00024741011009149944,
      "loss": 1.985081329345703,
      "mean_token_accuracy": 0.49359039342358696,
      "num_tokens": 210432789.0,
      "step": 1350
    },
    {
      "entropy": 2.71175808717535,
      "epoch": 1.076748374970632,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002393197644119594,
      "loss": 1.96485595703125,
      "mean_token_accuracy": 0.4930601012845071,
      "num_tokens": 214328522.0,
      "step": 1375
    },
    {
      "entropy": 2.692236351652126,
      "epoch": 1.0963270420549769,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00023124061159238553,
      "loss": 1.9634054565429688,
      "mean_token_accuracy": 0.49771587887169566,
      "num_tokens": 218240312.0,
      "step": 1400
    },
    {
      "entropy": 2.6694742179626036,
      "epoch": 1.1159057091393219,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002231811185645613,
      "loss": 1.92927978515625,
      "mean_token_accuracy": 0.5001408948135679,
      "num_tokens": 222131527.0,
      "step": 1425
    },
    {
      "entropy": 2.698059647261731,
      "epoch": 1.1354843762236666,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00021514973165685783,
      "loss": 1.9672230529785155,
      "mean_token_accuracy": 0.49608284242814266,
      "num_tokens": 226067443.0,
      "step": 1450
    },
    {
      "entropy": 2.6789358364283884,
      "epoch": 1.1550630433080116,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00020715486774250342,
      "loss": 1.956893310546875,
      "mean_token_accuracy": 0.4982129548229775,
      "num_tokens": 229991406.0,
      "step": 1475
    },
    {
      "entropy": 2.7062208250785704,
      "epoch": 1.1746417103923565,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0001992049054187214,
      "loss": 1.9547114562988281,
      "mean_token_accuracy": 0.49589620364152964,
      "num_tokens": 233890134.0,
      "step": 1500
    },
    {
      "epoch": 1.1746417103923565,
      "eval_bpd": 2.7522503005513013,
      "eval_entropy": 2.5481301208510145,
      "eval_loss": 1.9044488668441772,
      "eval_mean_token_accuracy": 0.5064063985939268,
      "eval_nll": 1.9077145360223966,
      "eval_num_tokens": 233890134.0,
      "eval_ppl": 6.737672480774532,
      "eval_runtime": 317.8596,
      "eval_samples_per_second": 571.35,
      "eval_steps_per_second": 8.928,
      "step": 1500
    },
    {
      "entropy": 2.7129361564886767,
      "epoch": 1.1942203774767013,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00019130817622598124,
      "loss": 1.9719258117675782,
      "mean_token_accuracy": 0.49460158206696975,
      "num_tokens": 237758654.0,
      "step": 1525
    },
    {
      "entropy": 2.6747650715948335,
      "epoch": 1.2137990445610463,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0001834729559165655,
      "loss": 1.9511233520507814,
      "mean_token_accuracy": 0.5000184974177605,
      "num_tokens": 241628710.0,
      "step": 1550
    },
    {
      "entropy": 2.71732418091414,
      "epoch": 1.2333777116453912,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00017570745578160198,
      "loss": 1.970428466796875,
      "mean_token_accuracy": 0.4950899709651767,
      "num_tokens": 245505115.0,
      "step": 1575
    },
    {
      "entropy": 2.7108077335151752,
      "epoch": 1.2529563787297362,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0001680198140456516,
      "loss": 1.9686761474609376,
      "mean_token_accuracy": 0.4953383364533806,
      "num_tokens": 249424811.0,
      "step": 1600
    },
    {
      "entropy": 2.703162872704107,
      "epoch": 1.272535045814081,
      "grad_norm": 0.546875,
      "learning_rate": 0.00016041808733786915,
      "loss": 1.964592742919922,
      "mean_token_accuracy": 0.49645553220386396,
      "num_tokens": 253343499.0,
      "step": 1625
    },
    {
      "entropy": 2.66976916182448,
      "epoch": 1.292113712898426,
      "grad_norm": 0.6328125,
      "learning_rate": 0.0001529102422486769,
      "loss": 1.9394204711914063,
      "mean_token_accuracy": 0.5004235959752784,
      "num_tokens": 257234610.0,
      "step": 1650
    },
    {
      "entropy": 2.721756610187284,
      "epoch": 1.3116923799827709,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00014550414698079727,
      "loss": 1.973395538330078,
      "mean_token_accuracy": 0.4934210284984832,
      "num_tokens": 261125376.0,
      "step": 1675
    },
    {
      "entropy": 2.675206924819273,
      "epoch": 1.3312710470671156,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00013820756310339665,
      "loss": 1.9494314575195313,
      "mean_token_accuracy": 0.5002018785718743,
      "num_tokens": 265045653.0,
      "step": 1700
    },
    {
      "entropy": 2.6791174841836285,
      "epoch": 1.3508497141514606,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00013102813741798025,
      "loss": 1.9530244445800782,
      "mean_token_accuracy": 0.4984120513628194,
      "num_tokens": 268938718.0,
      "step": 1725
    },
    {
      "entropy": 2.6993336161214603,
      "epoch": 1.3704283812358056,
      "grad_norm": 0.53125,
      "learning_rate": 0.00012397339394456356,
      "loss": 1.9576156616210938,
      "mean_token_accuracy": 0.4969041736325842,
      "num_tokens": 272852464.0,
      "step": 1750
    },
    {
      "epoch": 1.3704283812358056,
      "eval_bpd": 2.745460742423444,
      "eval_entropy": 2.540863456538099,
      "eval_loss": 1.8997492790222168,
      "eval_mean_token_accuracy": 0.5073028806361339,
      "eval_nll": 1.9030083729488245,
      "eval_num_tokens": 272852464.0,
      "eval_ppl": 6.7060383913214,
      "eval_runtime": 317.8678,
      "eval_samples_per_second": 571.335,
      "eval_steps_per_second": 8.928,
      "step": 1750
    },
    {
      "entropy": 2.6804251901539,
      "epoch": 1.3900070483201503,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00011705072603651829,
      "loss": 1.966283721923828,
      "mean_token_accuracy": 0.499262063001487,
      "num_tokens": 276760333.0,
      "step": 1775
    },
    {
      "entropy": 2.6867454461529983,
      "epoch": 1.4095857154044953,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00011026738863235674,
      "loss": 1.9592491149902345,
      "mean_token_accuracy": 0.49931023020028253,
      "num_tokens": 280651605.0,
      "step": 1800
    },
    {
      "entropy": 2.743908617496538,
      "epoch": 1.4291643824888403,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00010363049065257416,
      "loss": 1.9920062255859374,
      "mean_token_accuracy": 0.49101824791568294,
      "num_tokens": 284514323.0,
      "step": 1825
    },
    {
      "entropy": 2.7106320855852593,
      "epoch": 1.448743049573185,
      "grad_norm": 0.54296875,
      "learning_rate": 9.71469875495186e-05,
      "loss": 1.970909881591797,
      "mean_token_accuracy": 0.49648378670825294,
      "num_tokens": 288427479.0,
      "step": 1850
    },
    {
      "entropy": 2.6800678695504154,
      "epoch": 1.46832171665753,
      "grad_norm": 0.54296875,
      "learning_rate": 9.082367401809396e-05,
      "loss": 1.948299102783203,
      "mean_token_accuracy": 0.4985318521843758,
      "num_tokens": 292338476.0,
      "step": 1875
    },
    {
      "entropy": 2.7168465775671655,
      "epoch": 1.487900383741875,
      "grad_norm": 0.6015625,
      "learning_rate": 8.466717687493703e-05,
      "loss": 1.9598713684082032,
      "mean_token_accuracy": 0.4938074109471161,
      "num_tokens": 296239840.0,
      "step": 1900
    },
    {
      "entropy": 2.7053632534812735,
      "epoch": 1.5074790508262197,
      "grad_norm": 0.54296875,
      "learning_rate": 7.868394811353039e-05,
      "loss": 1.9856521606445312,
      "mean_token_accuracy": 0.4938124810525953,
      "num_tokens": 300130010.0,
      "step": 1925
    },
    {
      "entropy": 2.6828013073823707,
      "epoch": 1.5270577179105647,
      "grad_norm": 0.5390625,
      "learning_rate": 7.288025814252918e-05,
      "loss": 1.9554457092285156,
      "mean_token_accuracy": 0.4990507491186839,
      "num_tokens": 304047616.0,
      "step": 1950
    },
    {
      "entropy": 2.6844564534367192,
      "epoch": 1.5466363849949096,
      "grad_norm": 0.515625,
      "learning_rate": 6.726218921438915e-05,
      "loss": 1.946611328125,
      "mean_token_accuracy": 0.4998197396157143,
      "num_tokens": 307919609.0,
      "step": 1975
    },
    {
      "entropy": 2.677877172118015,
      "epoch": 1.5662150520792544,
      "grad_norm": 0.53125,
      "learning_rate": 6.183562905118107e-05,
      "loss": 1.9413558959960937,
      "mean_token_accuracy": 0.4990722147850662,
      "num_tokens": 311806214.0,
      "step": 2000
    },
    {
      "epoch": 1.5662150520792544,
      "eval_bpd": 2.741979334114817,
      "eval_entropy": 2.526833154664517,
      "eval_loss": 1.8973326683044434,
      "eval_mean_token_accuracy": 0.5078092915779853,
      "eval_nll": 1.9005952445953216,
      "eval_num_tokens": 311806214.0,
      "eval_ppl": 6.689875369506092,
      "eval_runtime": 317.8347,
      "eval_samples_per_second": 571.395,
      "eval_steps_per_second": 8.929,
      "step": 2000
    },
    {
      "entropy": 2.7164615675383526,
      "epoch": 1.5857937191635993,
      "grad_norm": 0.5234375,
      "learning_rate": 5.660626467427307e-05,
      "loss": 1.9794667053222657,
      "mean_token_accuracy": 0.4929969155160395,
      "num_tokens": 315733264.0,
      "step": 2025
    },
    {
      "entropy": 2.7066155855772465,
      "epoch": 1.6053723862479443,
      "grad_norm": 0.48046875,
      "learning_rate": 5.157957644434627e-05,
      "loss": 1.9774436950683594,
      "mean_token_accuracy": 0.4956127546833086,
      "num_tokens": 319661012.0,
      "step": 2050
    },
    {
      "entropy": 2.670972178570003,
      "epoch": 1.624951053332289,
      "grad_norm": 0.54296875,
      "learning_rate": 4.676083231799086e-05,
      "loss": 1.9275872802734375,
      "mean_token_accuracy": 0.5002606406547293,
      "num_tokens": 323574025.0,
      "step": 2075
    },
    {
      "entropy": 2.711799482902222,
      "epoch": 1.644529720416634,
      "grad_norm": 0.5625,
      "learning_rate": 4.2155082326900877e-05,
      "loss": 1.9813533020019531,
      "mean_token_accuracy": 0.49327668520870743,
      "num_tokens": 327471092.0,
      "step": 2100
    },
    {
      "entropy": 2.6970261707972836,
      "epoch": 1.664108387500979,
      "grad_norm": 0.57421875,
      "learning_rate": 3.7767153285453686e-05,
      "loss": 1.9576905822753907,
      "mean_token_accuracy": 0.49707455761602976,
      "num_tokens": 331389183.0,
      "step": 2125
    },
    {
      "entropy": 2.677772254526351,
      "epoch": 1.6836870545853238,
      "grad_norm": 0.59375,
      "learning_rate": 3.360164373222063e-05,
      "loss": 1.957755126953125,
      "mean_token_accuracy": 0.49902625771959186,
      "num_tokens": 335283715.0,
      "step": 2150
    },
    {
      "entropy": 2.6987997353257436,
      "epoch": 1.7032657216696687,
      "grad_norm": 0.7421875,
      "learning_rate": 2.966291911071012e-05,
      "loss": 1.9490525817871094,
      "mean_token_accuracy": 0.4974400475960837,
      "num_tokens": 339192022.0,
      "step": 2175
    },
    {
      "entropy": 2.6854547652224467,
      "epoch": 1.7228443887540137,
      "grad_norm": 0.59765625,
      "learning_rate": 2.5955107194394062e-05,
      "loss": 1.9636650085449219,
      "mean_token_accuracy": 0.49844507096807833,
      "num_tokens": 343093506.0,
      "step": 2200
    },
    {
      "entropy": 2.745404202931747,
      "epoch": 1.7424230558383584,
      "grad_norm": 0.5625,
      "learning_rate": 2.2482093760811617e-05,
      "loss": 1.998858642578125,
      "mean_token_accuracy": 0.49048741045360417,
      "num_tokens": 346973141.0,
      "step": 2225
    },
    {
      "entropy": 2.688500183765981,
      "epoch": 1.7620017229227034,
      "grad_norm": 0.53125,
      "learning_rate": 1.9247518519283968e-05,
      "loss": 1.945882568359375,
      "mean_token_accuracy": 0.49931712291043334,
      "num_tokens": 350842781.0,
      "step": 2250
    },
    {
      "epoch": 1.7620017229227034,
      "eval_bpd": 2.740859968520357,
      "eval_entropy": 2.529529855314099,
      "eval_loss": 1.896554946899414,
      "eval_mean_token_accuracy": 0.5079694414016338,
      "eval_nll": 1.8998193594895059,
      "eval_num_tokens": 350842781.0,
      "eval_ppl": 6.684686807971229,
      "eval_runtime": 317.8554,
      "eval_samples_per_second": 571.357,
      "eval_steps_per_second": 8.929,
      "step": 2250
    },
    {
      "entropy": 2.705022756460653,
      "epoch": 1.7815803900070484,
      "grad_norm": 0.546875,
      "learning_rate": 1.625477129650835e-05,
      "loss": 1.9555226135253907,
      "mean_token_accuracy": 0.4978065436114544,
      "num_tokens": 354757724.0,
      "step": 2275
    },
    {
      "entropy": 2.7167964107785805,
      "epoch": 1.8011590570913931,
      "grad_norm": 0.51953125,
      "learning_rate": 1.3506988484027926e-05,
      "loss": 1.9781422424316406,
      "mean_token_accuracy": 0.4929085231575454,
      "num_tokens": 358654333.0,
      "step": 2300
    },
    {
      "entropy": 2.685449053375581,
      "epoch": 1.820737724175738,
      "grad_norm": 0.5625,
      "learning_rate": 1.1007049751301789e-05,
      "loss": 1.9553031921386719,
      "mean_token_accuracy": 0.498819113219479,
      "num_tokens": 362545095.0,
      "step": 2325
    },
    {
      "entropy": 2.67209738728332,
      "epoch": 1.840316391260083,
      "grad_norm": 0.5234375,
      "learning_rate": 8.757575027818238e-06,
      "loss": 1.9598594665527345,
      "mean_token_accuracy": 0.5002877914117176,
      "num_tokens": 366446317.0,
      "step": 2350
    },
    {
      "entropy": 2.70602050797425,
      "epoch": 1.8598950583444278,
      "grad_norm": 0.62890625,
      "learning_rate": 6.760921757415439e-06,
      "loss": 1.963476104736328,
      "mean_token_accuracy": 0.49664319603072654,
      "num_tokens": 370335504.0,
      "step": 2375
    },
    {
      "entropy": 2.717956616262883,
      "epoch": 1.8794737254287728,
      "grad_norm": 0.55859375,
      "learning_rate": 5.0191824276862585e-06,
      "loss": 1.9688818359375,
      "mean_token_accuracy": 0.49326475180265583,
      "num_tokens": 374199425.0,
      "step": 2400
    },
    {
      "entropy": 2.715174971645218,
      "epoch": 1.8990523925131177,
      "grad_norm": 0.56640625,
      "learning_rate": 3.5341823770564385e-06,
      "loss": 1.9765179443359375,
      "mean_token_accuracy": 0.49364000601175223,
      "num_tokens": 378109465.0,
      "step": 2425
    },
    {
      "entropy": 2.671021122239945,
      "epoch": 1.9186310595974625,
      "grad_norm": 0.5,
      "learning_rate": 2.307477881834663e-06,
      "loss": 1.9456196594238282,
      "mean_token_accuracy": 0.4995694641029147,
      "num_tokens": 382015089.0,
      "step": 2450
    },
    {
      "entropy": 2.6898520242951838,
      "epoch": 1.9382097266818075,
      "grad_norm": 0.64453125,
      "learning_rate": 1.340354525238785e-06,
      "loss": 1.966468048095703,
      "mean_token_accuracy": 0.49751518802294914,
      "num_tokens": 385896919.0,
      "step": 2475
    },
    {
      "entropy": 2.6790956659680316,
      "epoch": 1.9577883937661524,
      "grad_norm": 0.53125,
      "learning_rate": 6.33825850107933e-07,
      "loss": 1.9549522399902344,
      "mean_token_accuracy": 0.500419180984845,
      "num_tokens": 389799725.0,
      "step": 2500
    },
    {
      "epoch": 1.9577883937661524,
      "eval_bpd": 2.7410050480160044,
      "eval_entropy": 2.5233276360670986,
      "eval_loss": 1.8966543674468994,
      "eval_mean_token_accuracy": 0.5080129268654007,
      "eval_nll": 1.8999199209328708,
      "eval_num_tokens": 389799725.0,
      "eval_ppl": 6.6853590635260085,
      "eval_runtime": 317.8734,
      "eval_samples_per_second": 571.325,
      "eval_steps_per_second": 8.928,
      "step": 2500
    },
    {
      "entropy": 2.700464563382669,
      "epoch": 1.9773670608504972,
      "grad_norm": 0.5,
      "learning_rate": 1.88632296712149e-07,
      "loss": 1.967222900390625,
      "mean_token_accuracy": 0.49522956458534473,
      "num_tokens": 393678888.0,
      "step": 2525
    },
    {
      "entropy": 2.6681114449302297,
      "epoch": 1.9969457279348422,
      "grad_norm": 0.77734375,
      "learning_rate": 5.240426772828988e-09,
      "loss": 1.9312892150878906,
      "mean_token_accuracy": 0.5011825757761065,
      "num_tokens": 397533693.0,
      "step": 2550
    },
    {
      "epoch": 2.0,
      "eval_bpd": 2.7410092079469774,
      "eval_entropy": 2.5233090361563453,
      "eval_loss": 1.8966583013534546,
      "eval_mean_token_accuracy": 0.5080199040107065,
      "eval_nll": 1.8999228043772962,
      "eval_num_tokens": 398148860.0,
      "eval_ppl": 6.685378340415125,
      "eval_runtime": 318.7016,
      "eval_samples_per_second": 569.84,
      "eval_steps_per_second": 8.905,
      "step": 2554
    }
  ],
  "logging_steps": 25,
  "max_steps": 2554,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.2938339313930778e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}