{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1072,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0046641791044776115,
      "grad_norm": 1.9764376015368916,
      "learning_rate": 4.6296296296296296e-06,
      "loss": 0.8462,
      "step": 5
    },
    {
      "epoch": 0.009328358208955223,
      "grad_norm": 1.4314413200657392,
      "learning_rate": 9.259259259259259e-06,
      "loss": 0.8236,
      "step": 10
    },
    {
      "epoch": 0.013992537313432836,
      "grad_norm": 1.120748220000551,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.7682,
      "step": 15
    },
    {
      "epoch": 0.018656716417910446,
      "grad_norm": 0.6728096457482404,
      "learning_rate": 1.8518518518518518e-05,
      "loss": 0.7273,
      "step": 20
    },
    {
      "epoch": 0.02332089552238806,
      "grad_norm": 0.5046065652987821,
      "learning_rate": 2.314814814814815e-05,
      "loss": 0.6963,
      "step": 25
    },
    {
      "epoch": 0.027985074626865673,
      "grad_norm": 0.4613424801601521,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.6736,
      "step": 30
    },
    {
      "epoch": 0.03264925373134328,
      "grad_norm": 0.4135275973660254,
      "learning_rate": 3.240740740740741e-05,
      "loss": 0.6448,
      "step": 35
    },
    {
      "epoch": 0.03731343283582089,
      "grad_norm": 0.42041723465008374,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 0.6379,
      "step": 40
    },
    {
      "epoch": 0.04197761194029851,
      "grad_norm": 0.4086127540673641,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.6333,
      "step": 45
    },
    {
      "epoch": 0.04664179104477612,
      "grad_norm": 0.39099983349763345,
      "learning_rate": 4.62962962962963e-05,
      "loss": 0.6148,
      "step": 50
    },
    {
      "epoch": 0.051305970149253734,
      "grad_norm": 0.39954713933549074,
      "learning_rate": 4.999989285883431e-05,
      "loss": 0.6341,
      "step": 55
    },
    {
      "epoch": 0.055970149253731345,
      "grad_norm": 0.4350948496115011,
      "learning_rate": 4.999614302517356e-05,
      "loss": 0.6272,
      "step": 60
    },
    {
      "epoch": 0.06063432835820896,
      "grad_norm": 0.40177852748659404,
      "learning_rate": 4.99870371535606e-05,
      "loss": 0.6059,
      "step": 65
    },
    {
      "epoch": 0.06529850746268656,
      "grad_norm": 0.38264613611154347,
      "learning_rate": 4.997257741198456e-05,
      "loss": 0.635,
      "step": 70
    },
    {
      "epoch": 0.06996268656716417,
      "grad_norm": 0.4686192193288965,
      "learning_rate": 4.9952767243121146e-05,
      "loss": 0.6122,
      "step": 75
    },
    {
      "epoch": 0.07462686567164178,
      "grad_norm": 0.49202199024337895,
      "learning_rate": 4.992761136351291e-05,
      "loss": 0.6076,
      "step": 80
    },
    {
      "epoch": 0.07929104477611941,
      "grad_norm": 0.5069244133631586,
      "learning_rate": 4.989711576244639e-05,
      "loss": 0.5999,
      "step": 85
    },
    {
      "epoch": 0.08395522388059702,
      "grad_norm": 0.4125232537952896,
      "learning_rate": 4.986128770052603e-05,
      "loss": 0.5894,
      "step": 90
    },
    {
      "epoch": 0.08861940298507463,
      "grad_norm": 0.4690088592451492,
      "learning_rate": 4.9820135707945634e-05,
      "loss": 0.5914,
      "step": 95
    },
    {
      "epoch": 0.09328358208955224,
      "grad_norm": 0.4970882976092545,
      "learning_rate": 4.9773669582457364e-05,
      "loss": 0.6045,
      "step": 100
    },
    {
      "epoch": 0.09794776119402986,
      "grad_norm": 0.4461715071663368,
      "learning_rate": 4.972190038703905e-05,
      "loss": 0.6099,
      "step": 105
    },
    {
      "epoch": 0.10261194029850747,
      "grad_norm": 0.4195725756826533,
      "learning_rate": 4.966484044726024e-05,
      "loss": 0.5968,
      "step": 110
    },
    {
      "epoch": 0.10727611940298508,
      "grad_norm": 0.44648158143178057,
      "learning_rate": 4.9602503348347625e-05,
      "loss": 0.5874,
      "step": 115
    },
    {
      "epoch": 0.11194029850746269,
      "grad_norm": 0.47598304401561387,
      "learning_rate": 4.953490393195063e-05,
      "loss": 0.5947,
      "step": 120
    },
    {
      "epoch": 0.1166044776119403,
      "grad_norm": 0.49998566726270804,
      "learning_rate": 4.9462058292607735e-05,
      "loss": 0.5921,
      "step": 125
    },
    {
      "epoch": 0.12126865671641791,
      "grad_norm": 0.4251340155158614,
      "learning_rate": 4.938398377391461e-05,
      "loss": 0.5849,
      "step": 130
    },
    {
      "epoch": 0.1259328358208955,
      "grad_norm": 0.40634416819865915,
      "learning_rate": 4.930069896439485e-05,
      "loss": 0.5847,
      "step": 135
    },
    {
      "epoch": 0.13059701492537312,
      "grad_norm": 0.538942175595294,
      "learning_rate": 4.921222369307427e-05,
      "loss": 0.5868,
      "step": 140
    },
    {
      "epoch": 0.13526119402985073,
      "grad_norm": 0.43479698529404903,
      "learning_rate": 4.9118579024759854e-05,
      "loss": 0.5891,
      "step": 145
    },
    {
      "epoch": 0.13992537313432835,
      "grad_norm": 0.476799188570071,
      "learning_rate": 4.901978725502454e-05,
      "loss": 0.573,
      "step": 150
    },
    {
      "epoch": 0.14458955223880596,
      "grad_norm": 0.43892813052018925,
      "learning_rate": 4.891587190489891e-05,
      "loss": 0.5775,
      "step": 155
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 0.42447943188758713,
      "learning_rate": 4.880685771527114e-05,
      "loss": 0.5852,
      "step": 160
    },
    {
      "epoch": 0.15391791044776118,
      "grad_norm": 0.38843326538800166,
      "learning_rate": 4.869277064099654e-05,
      "loss": 0.581,
      "step": 165
    },
    {
      "epoch": 0.15858208955223882,
      "grad_norm": 0.42022325604145055,
      "learning_rate": 4.8573637844718e-05,
      "loss": 0.5915,
      "step": 170
    },
    {
      "epoch": 0.16324626865671643,
      "grad_norm": 0.4020541945983881,
      "learning_rate": 4.844948769039896e-05,
      "loss": 0.5857,
      "step": 175
    },
    {
      "epoch": 0.16791044776119404,
      "grad_norm": 0.3878492346710094,
      "learning_rate": 4.83203497365703e-05,
      "loss": 0.5812,
      "step": 180
    },
    {
      "epoch": 0.17257462686567165,
      "grad_norm": 0.3959416689503896,
      "learning_rate": 4.818625472929286e-05,
      "loss": 0.5741,
      "step": 185
    },
    {
      "epoch": 0.17723880597014927,
      "grad_norm": 0.3438324253179286,
      "learning_rate": 4.8047234594837143e-05,
      "loss": 0.5677,
      "step": 190
    },
    {
      "epoch": 0.18190298507462688,
      "grad_norm": 0.43286612650725903,
      "learning_rate": 4.7903322432082185e-05,
      "loss": 0.568,
      "step": 195
    },
    {
      "epoch": 0.1865671641791045,
      "grad_norm": 0.3687114375153602,
      "learning_rate": 4.775455250463507e-05,
      "loss": 0.5677,
      "step": 200
    },
    {
      "epoch": 0.1912313432835821,
      "grad_norm": 0.4113086559648166,
      "learning_rate": 4.760096023267322e-05,
      "loss": 0.5815,
      "step": 205
    },
    {
      "epoch": 0.1958955223880597,
      "grad_norm": 0.36461363089638427,
      "learning_rate": 4.744258218451135e-05,
      "loss": 0.5732,
      "step": 210
    },
    {
      "epoch": 0.20055970149253732,
      "grad_norm": 0.5014961882683915,
      "learning_rate": 4.7279456067895e-05,
      "loss": 0.5709,
      "step": 215
    },
    {
      "epoch": 0.20522388059701493,
      "grad_norm": 0.37909263832337364,
      "learning_rate": 4.71116207210228e-05,
      "loss": 0.5666,
      "step": 220
    },
    {
      "epoch": 0.20988805970149255,
      "grad_norm": 0.3659556863078565,
      "learning_rate": 4.6939116103299655e-05,
      "loss": 0.5594,
      "step": 225
    },
    {
      "epoch": 0.21455223880597016,
      "grad_norm": 0.42583667948413667,
      "learning_rate": 4.676198328582288e-05,
      "loss": 0.5684,
      "step": 230
    },
    {
      "epoch": 0.21921641791044777,
      "grad_norm": 0.3859154861405004,
      "learning_rate": 4.6580264441603724e-05,
      "loss": 0.5754,
      "step": 235
    },
    {
      "epoch": 0.22388059701492538,
      "grad_norm": 0.39106576385368313,
      "learning_rate": 4.6394002835526535e-05,
      "loss": 0.5744,
      "step": 240
    },
    {
      "epoch": 0.228544776119403,
      "grad_norm": 0.3780182344000619,
      "learning_rate": 4.6203242814047946e-05,
      "loss": 0.5594,
      "step": 245
    },
    {
      "epoch": 0.2332089552238806,
      "grad_norm": 0.35412847272293113,
      "learning_rate": 4.6008029794638596e-05,
      "loss": 0.5598,
      "step": 250
    },
    {
      "epoch": 0.23787313432835822,
      "grad_norm": 0.3430551499365869,
      "learning_rate": 4.580841025496974e-05,
      "loss": 0.5574,
      "step": 255
    },
    {
      "epoch": 0.24253731343283583,
      "grad_norm": 0.33636534225361475,
      "learning_rate": 4.560443172184763e-05,
      "loss": 0.5666,
      "step": 260
    },
    {
      "epoch": 0.24720149253731344,
      "grad_norm": 0.32163202329313323,
      "learning_rate": 4.539614275989793e-05,
      "loss": 0.5551,
      "step": 265
    },
    {
      "epoch": 0.251865671641791,
      "grad_norm": 0.3362540905976733,
      "learning_rate": 4.5183592960003104e-05,
      "loss": 0.5539,
      "step": 270
    },
    {
      "epoch": 0.25652985074626866,
      "grad_norm": 0.4086822028521931,
      "learning_rate": 4.496683292749555e-05,
      "loss": 0.5583,
      "step": 275
    },
    {
      "epoch": 0.26119402985074625,
      "grad_norm": 0.40171933144466027,
      "learning_rate": 4.4745914270109055e-05,
      "loss": 0.5616,
      "step": 280
    },
    {
      "epoch": 0.2658582089552239,
      "grad_norm": 0.3048217989028901,
      "learning_rate": 4.4520889585691705e-05,
      "loss": 0.5669,
      "step": 285
    },
    {
      "epoch": 0.27052238805970147,
      "grad_norm": 0.3215470193697081,
      "learning_rate": 4.429181244968301e-05,
      "loss": 0.5511,
      "step": 290
    },
    {
      "epoch": 0.2751865671641791,
      "grad_norm": 0.2998971881368716,
      "learning_rate": 4.4058737402358295e-05,
      "loss": 0.5452,
      "step": 295
    },
    {
      "epoch": 0.2798507462686567,
      "grad_norm": 0.39015675590409005,
      "learning_rate": 4.38217199358434e-05,
      "loss": 0.5612,
      "step": 300
    },
    {
      "epoch": 0.28451492537313433,
      "grad_norm": 0.33178439784975494,
      "learning_rate": 4.3580816480902656e-05,
      "loss": 0.5454,
      "step": 305
    },
    {
      "epoch": 0.2891791044776119,
      "grad_norm": 0.3351700685619042,
      "learning_rate": 4.3336084393503545e-05,
      "loss": 0.5537,
      "step": 310
    },
    {
      "epoch": 0.29384328358208955,
      "grad_norm": 0.3562063334394124,
      "learning_rate": 4.308758194116094e-05,
      "loss": 0.5565,
      "step": 315
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 0.40270107421568574,
      "learning_rate": 4.283536828906436e-05,
      "loss": 0.5679,
      "step": 320
    },
    {
      "epoch": 0.3031716417910448,
      "grad_norm": 0.3650604303797052,
      "learning_rate": 4.2579503485991567e-05,
      "loss": 0.5523,
      "step": 325
    },
    {
      "epoch": 0.30783582089552236,
      "grad_norm": 0.38144306509308035,
      "learning_rate": 4.2320048450011684e-05,
      "loss": 0.5591,
      "step": 330
    },
    {
      "epoch": 0.3125,
      "grad_norm": 0.40997811569050646,
      "learning_rate": 4.205706495398143e-05,
      "loss": 0.5547,
      "step": 335
    },
    {
      "epoch": 0.31716417910447764,
      "grad_norm": 0.33609193713035646,
      "learning_rate": 4.179061561083777e-05,
      "loss": 0.5405,
      "step": 340
    },
    {
      "epoch": 0.3218283582089552,
      "grad_norm": 0.3397441908674912,
      "learning_rate": 4.1520763858690644e-05,
      "loss": 0.5552,
      "step": 345
    },
    {
      "epoch": 0.32649253731343286,
      "grad_norm": 0.33843363405675947,
      "learning_rate": 4.124757394571914e-05,
      "loss": 0.5516,
      "step": 350
    },
    {
      "epoch": 0.33115671641791045,
      "grad_norm": 0.32566981148675417,
      "learning_rate": 4.097111091487486e-05,
      "loss": 0.5454,
      "step": 355
    },
    {
      "epoch": 0.3358208955223881,
      "grad_norm": 0.37872534358247,
      "learning_rate": 4.069144058839605e-05,
      "loss": 0.5631,
      "step": 360
    },
    {
      "epoch": 0.34048507462686567,
      "grad_norm": 0.33032062301451925,
      "learning_rate": 4.040862955213615e-05,
      "loss": 0.5571,
      "step": 365
    },
    {
      "epoch": 0.3451492537313433,
      "grad_norm": 0.3201126166701232,
      "learning_rate": 4.012274513971061e-05,
      "loss": 0.5591,
      "step": 370
    },
    {
      "epoch": 0.3498134328358209,
      "grad_norm": 0.3466600975636123,
      "learning_rate": 3.9833855416465624e-05,
      "loss": 0.5468,
      "step": 375
    },
    {
      "epoch": 0.35447761194029853,
      "grad_norm": 0.3573267138974116,
      "learning_rate": 3.954202916327264e-05,
      "loss": 0.5457,
      "step": 380
    },
    {
      "epoch": 0.3591417910447761,
      "grad_norm": 0.3003122339186495,
      "learning_rate": 3.924733586015257e-05,
      "loss": 0.5443,
      "step": 385
    },
    {
      "epoch": 0.36380597014925375,
      "grad_norm": 0.3047012978878461,
      "learning_rate": 3.894984566973346e-05,
      "loss": 0.5478,
      "step": 390
    },
    {
      "epoch": 0.36847014925373134,
      "grad_norm": 0.33242216638118766,
      "learning_rate": 3.864962942054572e-05,
      "loss": 0.5543,
      "step": 395
    },
    {
      "epoch": 0.373134328358209,
      "grad_norm": 0.32390824699094883,
      "learning_rate": 3.834675859015876e-05,
      "loss": 0.546,
      "step": 400
    },
    {
      "epoch": 0.37779850746268656,
      "grad_norm": 0.3368746943555459,
      "learning_rate": 3.804130528816312e-05,
      "loss": 0.5485,
      "step": 405
    },
    {
      "epoch": 0.3824626865671642,
      "grad_norm": 0.3294436304011171,
      "learning_rate": 3.77333422390021e-05,
      "loss": 0.5342,
      "step": 410
    },
    {
      "epoch": 0.3871268656716418,
      "grad_norm": 0.2992567678994227,
      "learning_rate": 3.7422942764657054e-05,
      "loss": 0.5456,
      "step": 415
    },
    {
      "epoch": 0.3917910447761194,
      "grad_norm": 0.34623245219018867,
      "learning_rate": 3.711018076719034e-05,
      "loss": 0.5506,
      "step": 420
    },
    {
      "epoch": 0.396455223880597,
      "grad_norm": 0.3264253891806827,
      "learning_rate": 3.679513071115025e-05,
      "loss": 0.5538,
      "step": 425
    },
    {
      "epoch": 0.40111940298507465,
      "grad_norm": 0.34180367842194637,
      "learning_rate": 3.647786760584194e-05,
      "loss": 0.5508,
      "step": 430
    },
    {
      "epoch": 0.40578358208955223,
      "grad_norm": 0.3329961640111932,
      "learning_rate": 3.615846698746869e-05,
      "loss": 0.5545,
      "step": 435
    },
    {
      "epoch": 0.41044776119402987,
      "grad_norm": 0.3218838080487748,
      "learning_rate": 3.583700490114776e-05,
      "loss": 0.5468,
      "step": 440
    },
    {
      "epoch": 0.41511194029850745,
      "grad_norm": 0.2962813621413357,
      "learning_rate": 3.5513557882805e-05,
      "loss": 0.5451,
      "step": 445
    },
    {
      "epoch": 0.4197761194029851,
      "grad_norm": 0.35180011811550493,
      "learning_rate": 3.518820294095267e-05,
      "loss": 0.5477,
      "step": 450
    },
    {
      "epoch": 0.4244402985074627,
      "grad_norm": 0.324559555224438,
      "learning_rate": 3.486101753835468e-05,
      "loss": 0.5405,
      "step": 455
    },
    {
      "epoch": 0.4291044776119403,
      "grad_norm": 0.31920055255458074,
      "learning_rate": 3.453207957358377e-05,
      "loss": 0.5364,
      "step": 460
    },
    {
      "epoch": 0.4337686567164179,
      "grad_norm": 0.266971924437904,
      "learning_rate": 3.420146736247487e-05,
      "loss": 0.5273,
      "step": 465
    },
    {
      "epoch": 0.43843283582089554,
      "grad_norm": 0.32004814052652386,
      "learning_rate": 3.386925961947906e-05,
      "loss": 0.5513,
      "step": 470
    },
    {
      "epoch": 0.4430970149253731,
      "grad_norm": 0.3395919916093928,
      "learning_rate": 3.353553543892277e-05,
      "loss": 0.5508,
      "step": 475
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 0.3372834952110793,
      "learning_rate": 3.320037427617639e-05,
      "loss": 0.5427,
      "step": 480
    },
    {
      "epoch": 0.45242537313432835,
      "grad_norm": 0.30138144367379316,
      "learning_rate": 3.2863855928737026e-05,
      "loss": 0.5376,
      "step": 485
    },
    {
      "epoch": 0.457089552238806,
      "grad_norm": 0.3042709717398133,
      "learning_rate": 3.252606051722972e-05,
      "loss": 0.5521,
      "step": 490
    },
    {
      "epoch": 0.46175373134328357,
      "grad_norm": 0.3147603049850918,
      "learning_rate": 3.218706846633183e-05,
      "loss": 0.5362,
      "step": 495
    },
    {
      "epoch": 0.4664179104477612,
      "grad_norm": 0.32467383926014504,
      "learning_rate": 3.1846960485624886e-05,
      "loss": 0.5321,
      "step": 500
    },
    {
      "epoch": 0.4710820895522388,
      "grad_norm": 0.31072918948975164,
      "learning_rate": 3.150581755037877e-05,
      "loss": 0.5444,
      "step": 505
    },
    {
      "epoch": 0.47574626865671643,
      "grad_norm": 0.2875028374675,
      "learning_rate": 3.1163720882272516e-05,
      "loss": 0.539,
      "step": 510
    },
    {
      "epoch": 0.480410447761194,
      "grad_norm": 0.27864392303111696,
      "learning_rate": 3.08207519300565e-05,
      "loss": 0.5379,
      "step": 515
    },
    {
      "epoch": 0.48507462686567165,
      "grad_norm": 0.27495835874383767,
      "learning_rate": 3.047699235016056e-05,
      "loss": 0.5294,
      "step": 520
    },
    {
      "epoch": 0.48973880597014924,
      "grad_norm": 0.32292561340550724,
      "learning_rate": 3.0132523987252658e-05,
      "loss": 0.5328,
      "step": 525
    },
    {
      "epoch": 0.4944029850746269,
      "grad_norm": 0.28190497507760415,
      "learning_rate": 2.9787428854752736e-05,
      "loss": 0.5229,
      "step": 530
    },
    {
      "epoch": 0.49906716417910446,
      "grad_norm": 0.29893632016465344,
      "learning_rate": 2.9441789115306402e-05,
      "loss": 0.5444,
      "step": 535
    },
    {
      "epoch": 0.503731343283582,
      "grad_norm": 0.2937647666675581,
      "learning_rate": 2.9095687061223058e-05,
      "loss": 0.5404,
      "step": 540
    },
    {
      "epoch": 0.5083955223880597,
      "grad_norm": 0.32554217421674203,
      "learning_rate": 2.874920509488319e-05,
      "loss": 0.5436,
      "step": 545
    },
    {
      "epoch": 0.5130597014925373,
      "grad_norm": 0.325931285389485,
      "learning_rate": 2.8402425709119435e-05,
      "loss": 0.5386,
      "step": 550
    },
    {
      "epoch": 0.5177238805970149,
      "grad_norm": 0.29051430193129457,
      "learning_rate": 2.8055431467576106e-05,
      "loss": 0.5366,
      "step": 555
    },
    {
      "epoch": 0.5223880597014925,
      "grad_norm": 0.3096308868889565,
      "learning_rate": 2.7708304985051868e-05,
      "loss": 0.54,
      "step": 560
    },
    {
      "epoch": 0.5270522388059702,
      "grad_norm": 0.3160895951421515,
      "learning_rate": 2.7361128907830253e-05,
      "loss": 0.5309,
      "step": 565
    },
    {
      "epoch": 0.5317164179104478,
      "grad_norm": 0.2831546985708715,
      "learning_rate": 2.7013985894002623e-05,
      "loss": 0.5214,
      "step": 570
    },
    {
      "epoch": 0.5363805970149254,
      "grad_norm": 0.28807162338524384,
      "learning_rate": 2.6666958593788405e-05,
      "loss": 0.5364,
      "step": 575
    },
    {
      "epoch": 0.5410447761194029,
      "grad_norm": 0.3026251457722979,
      "learning_rate": 2.6320129629857093e-05,
      "loss": 0.5408,
      "step": 580
    },
    {
      "epoch": 0.5457089552238806,
      "grad_norm": 0.2869243020660439,
      "learning_rate": 2.597358157765692e-05,
      "loss": 0.5287,
      "step": 585
    },
    {
      "epoch": 0.5503731343283582,
      "grad_norm": 0.27339105580071865,
      "learning_rate": 2.56273969457547e-05,
      "loss": 0.5261,
      "step": 590
    },
    {
      "epoch": 0.5550373134328358,
      "grad_norm": 0.30400318199346166,
      "learning_rate": 2.528165815619162e-05,
      "loss": 0.5361,
      "step": 595
    },
    {
      "epoch": 0.5597014925373134,
      "grad_norm": 0.2723235085262856,
      "learning_rate": 2.4936447524859625e-05,
      "loss": 0.5273,
      "step": 600
    },
    {
      "epoch": 0.5643656716417911,
      "grad_norm": 0.2694137789877158,
      "learning_rate": 2.459184724190308e-05,
      "loss": 0.5195,
      "step": 605
    },
    {
      "epoch": 0.5690298507462687,
      "grad_norm": 0.31626957671199124,
      "learning_rate": 2.4247939352150386e-05,
      "loss": 0.5339,
      "step": 610
    },
    {
      "epoch": 0.5736940298507462,
      "grad_norm": 0.29413680883017335,
      "learning_rate": 2.390480573558012e-05,
      "loss": 0.5405,
      "step": 615
    },
    {
      "epoch": 0.5783582089552238,
      "grad_norm": 0.3065351668730826,
      "learning_rate": 2.3562528087826573e-05,
      "loss": 0.5251,
      "step": 620
    },
    {
      "epoch": 0.5830223880597015,
      "grad_norm": 0.26708916261058924,
      "learning_rate": 2.3221187900729003e-05,
      "loss": 0.5289,
      "step": 625
    },
    {
      "epoch": 0.5876865671641791,
      "grad_norm": 0.2836658322888292,
      "learning_rate": 2.2880866442929544e-05,
      "loss": 0.5304,
      "step": 630
    },
    {
      "epoch": 0.5923507462686567,
      "grad_norm": 0.27046810963493595,
      "learning_rate": 2.254164474052416e-05,
      "loss": 0.5336,
      "step": 635
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 0.2801464415588618,
      "learning_rate": 2.2203603557771447e-05,
      "loss": 0.5179,
      "step": 640
    },
    {
      "epoch": 0.601679104477612,
      "grad_norm": 0.2799176199165856,
      "learning_rate": 2.186682337786365e-05,
      "loss": 0.5301,
      "step": 645
    },
    {
      "epoch": 0.6063432835820896,
      "grad_norm": 0.2787107214909981,
      "learning_rate": 2.153138438376473e-05,
      "loss": 0.5177,
      "step": 650
    },
    {
      "epoch": 0.6110074626865671,
      "grad_norm": 0.2888838182238187,
      "learning_rate": 2.119736643911979e-05,
      "loss": 0.5265,
      "step": 655
    },
    {
      "epoch": 0.6156716417910447,
      "grad_norm": 0.2787832383670051,
      "learning_rate": 2.0864849069240645e-05,
      "loss": 0.5302,
      "step": 660
    },
    {
      "epoch": 0.6203358208955224,
      "grad_norm": 0.2627154396461022,
      "learning_rate": 2.0533911442171805e-05,
      "loss": 0.5263,
      "step": 665
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.29534395087775367,
      "learning_rate": 2.0204632349841667e-05,
      "loss": 0.5237,
      "step": 670
    },
    {
      "epoch": 0.6296641791044776,
      "grad_norm": 0.2655096554332185,
      "learning_rate": 1.9877090189303182e-05,
      "loss": 0.5309,
      "step": 675
    },
    {
      "epoch": 0.6343283582089553,
      "grad_norm": 0.2700729312781095,
      "learning_rate": 1.9551362944068462e-05,
      "loss": 0.5237,
      "step": 680
    },
    {
      "epoch": 0.6389925373134329,
      "grad_norm": 0.2611011089050749,
      "learning_rate": 1.922752816554204e-05,
      "loss": 0.5234,
      "step": 685
    },
    {
      "epoch": 0.6436567164179104,
      "grad_norm": 0.25922947776157407,
      "learning_rate": 1.890566295455678e-05,
      "loss": 0.5173,
      "step": 690
    },
    {
      "epoch": 0.648320895522388,
      "grad_norm": 0.24853625361203247,
      "learning_rate": 1.858584394301728e-05,
      "loss": 0.5216,
      "step": 695
    },
    {
      "epoch": 0.6529850746268657,
      "grad_norm": 0.27687546685694786,
      "learning_rate": 1.8268147275654707e-05,
      "loss": 0.5386,
      "step": 700
    },
    {
      "epoch": 0.6576492537313433,
      "grad_norm": 0.24785393612070897,
      "learning_rate": 1.7952648591897858e-05,
      "loss": 0.5211,
      "step": 705
    },
    {
      "epoch": 0.6623134328358209,
      "grad_norm": 0.29293847501280135,
      "learning_rate": 1.7639423007864252e-05,
      "loss": 0.5132,
      "step": 710
    },
    {
      "epoch": 0.6669776119402985,
      "grad_norm": 0.24923844657487307,
      "learning_rate": 1.7328545098476106e-05,
      "loss": 0.5167,
      "step": 715
    },
    {
      "epoch": 0.6716417910447762,
      "grad_norm": 0.26570613648581376,
      "learning_rate": 1.702008887970491e-05,
      "loss": 0.519,
      "step": 720
    },
    {
      "epoch": 0.6763059701492538,
      "grad_norm": 0.26373066926606076,
      "learning_rate": 1.671412779094926e-05,
      "loss": 0.5228,
      "step": 725
    },
    {
      "epoch": 0.6809701492537313,
      "grad_norm": 0.2526155940226532,
      "learning_rate": 1.6410734677549872e-05,
      "loss": 0.519,
      "step": 730
    },
    {
      "epoch": 0.6856343283582089,
      "grad_norm": 0.27439542585060434,
      "learning_rate": 1.6109981773446036e-05,
      "loss": 0.5204,
      "step": 735
    },
    {
      "epoch": 0.6902985074626866,
      "grad_norm": 0.24862245672468689,
      "learning_rate": 1.58119406839777e-05,
      "loss": 0.5189,
      "step": 740
    },
    {
      "epoch": 0.6949626865671642,
      "grad_norm": 0.24323442659034997,
      "learning_rate": 1.5516682368837133e-05,
      "loss": 0.5143,
      "step": 745
    },
    {
      "epoch": 0.6996268656716418,
      "grad_norm": 0.2783897852969829,
      "learning_rate": 1.5224277125174388e-05,
      "loss": 0.5307,
      "step": 750
    },
    {
      "epoch": 0.7042910447761194,
      "grad_norm": 0.2663619652723149,
      "learning_rate": 1.4934794570860416e-05,
      "loss": 0.5314,
      "step": 755
    },
    {
      "epoch": 0.7089552238805971,
      "grad_norm": 0.26323091306556967,
      "learning_rate": 1.464830362791204e-05,
      "loss": 0.5178,
      "step": 760
    },
    {
      "epoch": 0.7136194029850746,
      "grad_norm": 0.2609500205391141,
      "learning_rate": 1.4364872506082425e-05,
      "loss": 0.5255,
      "step": 765
    },
    {
      "epoch": 0.7182835820895522,
      "grad_norm": 0.2710473414376264,
      "learning_rate": 1.4084568686621314e-05,
      "loss": 0.5286,
      "step": 770
    },
    {
      "epoch": 0.7229477611940298,
      "grad_norm": 0.26623093173720286,
      "learning_rate": 1.3807458906208546e-05,
      "loss": 0.5355,
      "step": 775
    },
    {
      "epoch": 0.7276119402985075,
      "grad_norm": 0.2516613759006932,
      "learning_rate": 1.3533609141065008e-05,
      "loss": 0.5148,
      "step": 780
    },
    {
      "epoch": 0.7322761194029851,
      "grad_norm": 0.2552482036562621,
      "learning_rate": 1.326308459124447e-05,
      "loss": 0.5179,
      "step": 785
    },
    {
      "epoch": 0.7369402985074627,
      "grad_norm": 0.24831784749149477,
      "learning_rate": 1.299594966511038e-05,
      "loss": 0.5265,
      "step": 790
    },
    {
      "epoch": 0.7416044776119403,
      "grad_norm": 0.2542556781092361,
      "learning_rate": 1.2732267964001033e-05,
      "loss": 0.5147,
      "step": 795
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 0.26221364798745955,
      "learning_rate": 1.2472102267086904e-05,
      "loss": 0.5212,
      "step": 800
    },
    {
      "epoch": 0.7509328358208955,
      "grad_norm": 0.24312122488283722,
      "learning_rate": 1.2215514516423813e-05,
      "loss": 0.5353,
      "step": 805
    },
    {
      "epoch": 0.7555970149253731,
      "grad_norm": 0.26282735529843265,
      "learning_rate": 1.1962565802205255e-05,
      "loss": 0.5139,
      "step": 810
    },
    {
      "epoch": 0.7602611940298507,
      "grad_norm": 0.2476722606801948,
      "learning_rate": 1.1713316348217673e-05,
      "loss": 0.5187,
      "step": 815
    },
    {
      "epoch": 0.7649253731343284,
      "grad_norm": 0.25096104248356926,
      "learning_rate": 1.1467825497501954e-05,
      "loss": 0.5033,
      "step": 820
    },
    {
      "epoch": 0.769589552238806,
      "grad_norm": 0.24224168196533652,
      "learning_rate": 1.1226151698224597e-05,
      "loss": 0.5243,
      "step": 825
    },
    {
      "epoch": 0.7742537313432836,
      "grad_norm": 0.2522377650971933,
      "learning_rate": 1.0988352489762006e-05,
      "loss": 0.5283,
      "step": 830
    },
    {
      "epoch": 0.7789179104477612,
      "grad_norm": 0.24586034947790467,
      "learning_rate": 1.0754484489001085e-05,
      "loss": 0.5185,
      "step": 835
    },
    {
      "epoch": 0.7835820895522388,
      "grad_norm": 0.2540046185584379,
      "learning_rate": 1.052460337685951e-05,
      "loss": 0.5174,
      "step": 840
    },
    {
      "epoch": 0.7882462686567164,
      "grad_norm": 0.24424035510396644,
      "learning_rate": 1.0298763885028839e-05,
      "loss": 0.5135,
      "step": 845
    },
    {
      "epoch": 0.792910447761194,
      "grad_norm": 0.24097259911522662,
      "learning_rate": 1.0077019782943584e-05,
      "loss": 0.5191,
      "step": 850
    },
    {
      "epoch": 0.7975746268656716,
      "grad_norm": 0.22966397148660536,
      "learning_rate": 9.859423864979441e-06,
      "loss": 0.5075,
      "step": 855
    },
    {
      "epoch": 0.8022388059701493,
      "grad_norm": 0.25454516096421553,
      "learning_rate": 9.646027937883622e-06,
      "loss": 0.5267,
      "step": 860
    },
    {
      "epoch": 0.8069029850746269,
      "grad_norm": 0.25947472933275617,
      "learning_rate": 9.436882808440334e-06,
      "loss": 0.5279,
      "step": 865
    },
    {
      "epoch": 0.8115671641791045,
      "grad_norm": 0.2333429022867138,
      "learning_rate": 9.232038271374377e-06,
      "loss": 0.5099,
      "step": 870
    },
    {
      "epoch": 0.816231343283582,
      "grad_norm": 0.23720076251289113,
      "learning_rate": 9.031543097495638e-06,
      "loss": 0.5145,
      "step": 875
    },
    {
      "epoch": 0.8208955223880597,
      "grad_norm": 0.2608860891878713,
      "learning_rate": 8.835445022087426e-06,
      "loss": 0.5125,
      "step": 880
    },
    {
      "epoch": 0.8255597014925373,
      "grad_norm": 0.248796206189008,
      "learning_rate": 8.6437907335413e-06,
      "loss": 0.5225,
      "step": 885
    },
    {
      "epoch": 0.8302238805970149,
      "grad_norm": 0.2363486956815597,
      "learning_rate": 8.456625862241193e-06,
      "loss": 0.5248,
      "step": 890
    },
    {
      "epoch": 0.8348880597014925,
      "grad_norm": 0.24134673811694132,
      "learning_rate": 8.273994969699394e-06,
      "loss": 0.5189,
      "step": 895
    },
    {
      "epoch": 0.8395522388059702,
      "grad_norm": 0.22438925840490204,
      "learning_rate": 8.095941537947057e-06,
      "loss": 0.5218,
      "step": 900
    },
    {
      "epoch": 0.8442164179104478,
      "grad_norm": 0.24284130133539145,
      "learning_rate": 7.922507959181673e-06,
      "loss": 0.5101,
      "step": 905
    },
    {
      "epoch": 0.8488805970149254,
      "grad_norm": 0.23178936992573468,
      "learning_rate": 7.753735525674059e-06,
      "loss": 0.5163,
      "step": 910
    },
    {
      "epoch": 0.8535447761194029,
      "grad_norm": 0.2531112545315579,
      "learning_rate": 7.58966441993719e-06,
      "loss": 0.513,
      "step": 915
    },
    {
      "epoch": 0.8582089552238806,
      "grad_norm": 0.25236625441627875,
      "learning_rate": 7.430333705159286e-06,
      "loss": 0.5339,
      "step": 920
    },
    {
      "epoch": 0.8628731343283582,
      "grad_norm": 0.2713665169762642,
      "learning_rate": 7.275781315903374e-06,
      "loss": 0.5324,
      "step": 925
    },
    {
      "epoch": 0.8675373134328358,
      "grad_norm": 0.23950738331964438,
      "learning_rate": 7.126044049075548e-06,
      "loss": 0.5237,
      "step": 930
    },
    {
      "epoch": 0.8722014925373134,
      "grad_norm": 0.23323024234821804,
      "learning_rate": 6.9811575551641224e-06,
      "loss": 0.5115,
      "step": 935
    },
    {
      "epoch": 0.8768656716417911,
      "grad_norm": 0.24257474110632124,
      "learning_rate": 6.8411563297516995e-06,
      "loss": 0.5055,
      "step": 940
    },
    {
      "epoch": 0.8815298507462687,
      "grad_norm": 0.245170836351988,
      "learning_rate": 6.706073705302254e-06,
      "loss": 0.5169,
      "step": 945
    },
    {
      "epoch": 0.8861940298507462,
      "grad_norm": 0.23886895938544617,
      "learning_rate": 6.575941843225068e-06,
      "loss": 0.5072,
      "step": 950
    },
    {
      "epoch": 0.8908582089552238,
      "grad_norm": 0.23235561835641955,
      "learning_rate": 6.450791726217538e-06,
      "loss": 0.5148,
      "step": 955
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 0.22966650210630168,
      "learning_rate": 6.330653150888617e-06,
      "loss": 0.5185,
      "step": 960
    },
    {
      "epoch": 0.9001865671641791,
      "grad_norm": 0.2297504409214819,
      "learning_rate": 6.215554720664598e-06,
      "loss": 0.5245,
      "step": 965
    },
    {
      "epoch": 0.9048507462686567,
      "grad_norm": 0.24374759590853393,
      "learning_rate": 6.105523838979022e-06,
      "loss": 0.5244,
      "step": 970
    },
    {
      "epoch": 0.9095149253731343,
      "grad_norm": 0.22861004334133614,
      "learning_rate": 6.000586702748301e-06,
      "loss": 0.5099,
      "step": 975
    },
    {
      "epoch": 0.914179104477612,
      "grad_norm": 0.2245890338301874,
      "learning_rate": 5.900768296134551e-06,
      "loss": 0.5122,
      "step": 980
    },
    {
      "epoch": 0.9188432835820896,
      "grad_norm": 0.23552028614217635,
      "learning_rate": 5.8060923845971825e-06,
      "loss": 0.5201,
      "step": 985
    },
    {
      "epoch": 0.9235074626865671,
      "grad_norm": 0.22537028329185735,
      "learning_rate": 5.7165815092346825e-06,
      "loss": 0.5058,
      "step": 990
    },
    {
      "epoch": 0.9281716417910447,
      "grad_norm": 0.7371806401244925,
      "learning_rate": 5.632256981417845e-06,
      "loss": 0.5141,
      "step": 995
    },
    {
      "epoch": 0.9328358208955224,
      "grad_norm": 0.23518208398659468,
      "learning_rate": 5.553138877715833e-06,
      "loss": 0.5162,
      "step": 1000
    },
    {
      "epoch": 0.9375,
      "grad_norm": 0.22409637826011708,
      "learning_rate": 5.479246035116201e-06,
      "loss": 0.5091,
      "step": 1005
    },
    {
      "epoch": 0.9421641791044776,
      "grad_norm": 0.23088030429859727,
      "learning_rate": 5.410596046540051e-06,
      "loss": 0.5113,
      "step": 1010
    },
    {
      "epoch": 0.9468283582089553,
      "grad_norm": 0.24167691781648407,
      "learning_rate": 5.347205256653387e-06,
      "loss": 0.5141,
      "step": 1015
    },
    {
      "epoch": 0.9514925373134329,
      "grad_norm": 0.2505803611107525,
      "learning_rate": 5.28908875797568e-06,
      "loss": 0.5072,
      "step": 1020
    },
    {
      "epoch": 0.9561567164179104,
      "grad_norm": 0.23002826210892982,
      "learning_rate": 5.236260387286509e-06,
      "loss": 0.5134,
      "step": 1025
    },
    {
      "epoch": 0.960820895522388,
      "grad_norm": 0.2337459592766504,
      "learning_rate": 5.1887327223312296e-06,
      "loss": 0.5101,
      "step": 1030
    },
    {
      "epoch": 0.9654850746268657,
      "grad_norm": 0.2271925278965666,
      "learning_rate": 5.1465170788263595e-06,
      "loss": 0.518,
      "step": 1035
    },
    {
      "epoch": 0.9701492537313433,
      "grad_norm": 0.235300174359742,
      "learning_rate": 5.109623507765466e-06,
      "loss": 0.5137,
      "step": 1040
    },
    {
      "epoch": 0.9748134328358209,
      "grad_norm": 0.233671111506403,
      "learning_rate": 5.07806079302615e-06,
      "loss": 0.5208,
      "step": 1045
    },
    {
      "epoch": 0.9794776119402985,
      "grad_norm": 0.22627242259525954,
      "learning_rate": 5.051836449278715e-06,
      "loss": 0.5154,
      "step": 1050
    },
    {
      "epoch": 0.9841417910447762,
      "grad_norm": 0.23340526855169283,
      "learning_rate": 5.030956720197035e-06,
      "loss": 0.508,
      "step": 1055
    },
    {
      "epoch": 0.9888059701492538,
      "grad_norm": 0.22844780564307632,
      "learning_rate": 5.015426576972003e-06,
      "loss": 0.5257,
      "step": 1060
    },
    {
      "epoch": 0.9934701492537313,
      "grad_norm": 0.24227361546235135,
      "learning_rate": 5.005249717127964e-06,
      "loss": 0.5167,
      "step": 1065
    },
    {
      "epoch": 0.9981343283582089,
      "grad_norm": 0.2376538962488628,
      "learning_rate": 5.000428563642382e-06,
      "loss": 0.5171,
      "step": 1070
    },
    {
      "epoch": 1.0,
      "step": 1072,
      "total_flos": 488621249396736.0,
      "train_loss": 0.0,
      "train_runtime": 1.4378,
      "train_samples_per_second": 23857.288,
      "train_steps_per_second": 745.605
    }
  ],
  "logging_steps": 5,
  "max_steps": 1072,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 488621249396736.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}