{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 536,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.009328358208955223,
      "grad_norm": 1.453897786558338,
      "learning_rate": 9.259259259259259e-06,
      "loss": 0.8491,
      "step": 5
    },
    {
      "epoch": 0.018656716417910446,
      "grad_norm": 1.1644840596539492,
      "learning_rate": 1.8518518518518518e-05,
      "loss": 0.8006,
      "step": 10
    },
    {
      "epoch": 0.027985074626865673,
      "grad_norm": 0.7401001416095307,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.7315,
      "step": 15
    },
    {
      "epoch": 0.03731343283582089,
      "grad_norm": 0.45715435120985837,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 0.6803,
      "step": 20
    },
    {
      "epoch": 0.04664179104477612,
      "grad_norm": 0.452456199763512,
      "learning_rate": 4.62962962962963e-05,
      "loss": 0.6578,
      "step": 25
    },
    {
      "epoch": 0.055970149253731345,
      "grad_norm": 0.3770950269873419,
      "learning_rate": 4.999614302517356e-05,
      "loss": 0.6581,
      "step": 30
    },
    {
      "epoch": 0.06529850746268656,
      "grad_norm": 0.3867121454463547,
      "learning_rate": 4.997257741198456e-05,
      "loss": 0.6419,
      "step": 35
    },
    {
      "epoch": 0.07462686567164178,
      "grad_norm": 0.30411968998458416,
      "learning_rate": 4.992761136351291e-05,
      "loss": 0.627,
      "step": 40
    },
    {
      "epoch": 0.08395522388059702,
      "grad_norm": 0.3178169814537855,
      "learning_rate": 4.986128770052603e-05,
      "loss": 0.6084,
      "step": 45
    },
    {
      "epoch": 0.09328358208955224,
      "grad_norm": 0.2807145763333074,
      "learning_rate": 4.9773669582457364e-05,
      "loss": 0.61,
      "step": 50
    },
    {
      "epoch": 0.10261194029850747,
      "grad_norm": 0.34303198218488506,
      "learning_rate": 4.966484044726024e-05,
      "loss": 0.6137,
      "step": 55
    },
    {
      "epoch": 0.11194029850746269,
      "grad_norm": 0.32762450828392975,
      "learning_rate": 4.953490393195063e-05,
      "loss": 0.6003,
      "step": 60
    },
    {
      "epoch": 0.12126865671641791,
      "grad_norm": 0.38462981393609424,
      "learning_rate": 4.938398377391461e-05,
      "loss": 0.5972,
      "step": 65
    },
    {
      "epoch": 0.13059701492537312,
      "grad_norm": 0.4924998750785629,
      "learning_rate": 4.921222369307427e-05,
      "loss": 0.5939,
      "step": 70
    },
    {
      "epoch": 0.13992537313432835,
      "grad_norm": 0.4428236238520031,
      "learning_rate": 4.901978725502454e-05,
      "loss": 0.5895,
      "step": 75
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 0.39121495707754167,
      "learning_rate": 4.880685771527114e-05,
      "loss": 0.5895,
      "step": 80
    },
    {
      "epoch": 0.15858208955223882,
      "grad_norm": 0.42141569967189374,
      "learning_rate": 4.8573637844718e-05,
      "loss": 0.5945,
      "step": 85
    },
    {
      "epoch": 0.16791044776119404,
      "grad_norm": 0.37097437462801275,
      "learning_rate": 4.83203497365703e-05,
      "loss": 0.591,
      "step": 90
    },
    {
      "epoch": 0.17723880597014927,
      "grad_norm": 0.4736806212925385,
      "learning_rate": 4.8047234594837143e-05,
      "loss": 0.5782,
      "step": 95
    },
    {
      "epoch": 0.1865671641791045,
      "grad_norm": 0.4098423065809189,
      "learning_rate": 4.775455250463507e-05,
      "loss": 0.5749,
      "step": 100
    },
    {
      "epoch": 0.1958955223880597,
      "grad_norm": 0.43057914859393537,
      "learning_rate": 4.744258218451135e-05,
      "loss": 0.5846,
      "step": 105
    },
    {
      "epoch": 0.20522388059701493,
      "grad_norm": 0.3079441344852731,
      "learning_rate": 4.71116207210228e-05,
      "loss": 0.5758,
      "step": 110
    },
    {
      "epoch": 0.21455223880597016,
      "grad_norm": 0.35622829843104137,
      "learning_rate": 4.676198328582288e-05,
      "loss": 0.5712,
      "step": 115
    },
    {
      "epoch": 0.22388059701492538,
      "grad_norm": 0.3303782414556567,
      "learning_rate": 4.6394002835526535e-05,
      "loss": 0.582,
      "step": 120
    },
    {
      "epoch": 0.2332089552238806,
      "grad_norm": 0.36692461693115874,
      "learning_rate": 4.6008029794638596e-05,
      "loss": 0.5665,
      "step": 125
    },
    {
      "epoch": 0.24253731343283583,
      "grad_norm": 0.3438548029811277,
      "learning_rate": 4.560443172184763e-05,
      "loss": 0.5691,
      "step": 130
    },
    {
      "epoch": 0.251865671641791,
      "grad_norm": 0.3148363976766754,
      "learning_rate": 4.5183592960003104e-05,
      "loss": 0.5619,
      "step": 135
    },
    {
      "epoch": 0.26119402985074625,
      "grad_norm": 0.27079202033744576,
      "learning_rate": 4.4745914270109055e-05,
      "loss": 0.5671,
      "step": 140
    },
    {
      "epoch": 0.27052238805970147,
      "grad_norm": 0.35294951179615874,
      "learning_rate": 4.429181244968301e-05,
      "loss": 0.5663,
      "step": 145
    },
    {
      "epoch": 0.2798507462686567,
      "grad_norm": 0.2830160917424769,
      "learning_rate": 4.38217199358434e-05,
      "loss": 0.5602,
      "step": 150
    },
    {
      "epoch": 0.2891791044776119,
      "grad_norm": 0.35651413739101817,
      "learning_rate": 4.3336084393503545e-05,
      "loss": 0.5569,
      "step": 155
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 0.3354623584514313,
      "learning_rate": 4.283536828906436e-05,
      "loss": 0.5692,
      "step": 160
    },
    {
      "epoch": 0.30783582089552236,
      "grad_norm": 0.31607235976679265,
      "learning_rate": 4.2320048450011684e-05,
      "loss": 0.5626,
      "step": 165
    },
    {
      "epoch": 0.31716417910447764,
      "grad_norm": 0.28829206615383307,
      "learning_rate": 4.179061561083777e-05,
      "loss": 0.5547,
      "step": 170
    },
    {
      "epoch": 0.32649253731343286,
      "grad_norm": 0.3376743915281689,
      "learning_rate": 4.124757394571914e-05,
      "loss": 0.561,
      "step": 175
    },
    {
      "epoch": 0.3358208955223881,
      "grad_norm": 0.270382329131434,
      "learning_rate": 4.069144058839605e-05,
      "loss": 0.5621,
      "step": 180
    },
    {
      "epoch": 0.3451492537313433,
      "grad_norm": 0.2973987443207836,
      "learning_rate": 4.012274513971061e-05,
      "loss": 0.5661,
      "step": 185
    },
    {
      "epoch": 0.35447761194029853,
      "grad_norm": 0.28887092947840104,
      "learning_rate": 3.954202916327264e-05,
      "loss": 0.5542,
      "step": 190
    },
    {
      "epoch": 0.36380597014925375,
      "grad_norm": 0.2655355295288525,
      "learning_rate": 3.894984566973346e-05,
      "loss": 0.5541,
      "step": 195
    },
    {
      "epoch": 0.373134328358209,
      "grad_norm": 0.2811432888948243,
      "learning_rate": 3.834675859015876e-05,
      "loss": 0.5582,
      "step": 200
    },
    {
      "epoch": 0.3824626865671642,
      "grad_norm": 0.2515745875368655,
      "learning_rate": 3.77333422390021e-05,
      "loss": 0.5489,
      "step": 205
    },
    {
      "epoch": 0.3917910447761194,
      "grad_norm": 0.28289824813151215,
      "learning_rate": 3.711018076719034e-05,
      "loss": 0.556,
      "step": 210
    },
    {
      "epoch": 0.40111940298507465,
      "grad_norm": 0.267361883328445,
      "learning_rate": 3.647786760584194e-05,
      "loss": 0.5604,
      "step": 215
    },
    {
      "epoch": 0.41044776119402987,
      "grad_norm": 0.26327372469127713,
      "learning_rate": 3.583700490114776e-05,
      "loss": 0.5585,
      "step": 220
    },
    {
      "epoch": 0.4197761194029851,
      "grad_norm": 0.306243298391607,
      "learning_rate": 3.518820294095267e-05,
      "loss": 0.5545,
      "step": 225
    },
    {
      "epoch": 0.4291044776119403,
      "grad_norm": 0.26926366241042204,
      "learning_rate": 3.453207957358377e-05,
      "loss": 0.5464,
      "step": 230
    },
    {
      "epoch": 0.43843283582089554,
      "grad_norm": 0.2690060278556445,
      "learning_rate": 3.386925961947906e-05,
      "loss": 0.5475,
      "step": 235
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 0.2826334847052104,
      "learning_rate": 3.320037427617639e-05,
      "loss": 0.555,
      "step": 240
    },
    {
      "epoch": 0.457089552238806,
      "grad_norm": 0.26144344293288974,
      "learning_rate": 3.252606051722972e-05,
      "loss": 0.5535,
      "step": 245
    },
    {
      "epoch": 0.4664179104477612,
      "grad_norm": 0.2911683020725898,
      "learning_rate": 3.1846960485624886e-05,
      "loss": 0.5427,
      "step": 250
    },
    {
      "epoch": 0.47574626865671643,
      "grad_norm": 0.25802207095689544,
      "learning_rate": 3.1163720882272516e-05,
      "loss": 0.5505,
      "step": 255
    },
    {
      "epoch": 0.48507462686567165,
      "grad_norm": 0.2937192282456116,
      "learning_rate": 3.047699235016056e-05,
      "loss": 0.5428,
      "step": 260
    },
    {
      "epoch": 0.4944029850746269,
      "grad_norm": 0.3102482607702826,
      "learning_rate": 2.9787428854752736e-05,
      "loss": 0.5367,
      "step": 265
    },
    {
      "epoch": 0.503731343283582,
      "grad_norm": 0.2548120450864961,
      "learning_rate": 2.9095687061223058e-05,
      "loss": 0.5513,
      "step": 270
    },
    {
      "epoch": 0.5130597014925373,
      "grad_norm": 0.30245109922773245,
      "learning_rate": 2.8402425709119435e-05,
      "loss": 0.5504,
      "step": 275
    },
    {
      "epoch": 0.5223880597014925,
      "grad_norm": 0.27315442361776204,
      "learning_rate": 2.7708304985051868e-05,
      "loss": 0.5474,
      "step": 280
    },
    {
      "epoch": 0.5317164179104478,
      "grad_norm": 0.2469052900266435,
      "learning_rate": 2.7013985894002623e-05,
      "loss": 0.5353,
      "step": 285
    },
    {
      "epoch": 0.5410447761194029,
      "grad_norm": 0.2381522137991754,
      "learning_rate": 2.6320129629857093e-05,
      "loss": 0.5481,
      "step": 290
    },
    {
      "epoch": 0.5503731343283582,
      "grad_norm": 0.2435217695212293,
      "learning_rate": 2.56273969457547e-05,
      "loss": 0.537,
      "step": 295
    },
    {
      "epoch": 0.5597014925373134,
      "grad_norm": 0.23199546267694518,
      "learning_rate": 2.4936447524859625e-05,
      "loss": 0.5413,
      "step": 300
    },
    {
      "epoch": 0.5690298507462687,
      "grad_norm": 0.30029854485663093,
      "learning_rate": 2.4247939352150386e-05,
      "loss": 0.5365,
      "step": 305
    },
    {
      "epoch": 0.5783582089552238,
      "grad_norm": 0.25071572100454237,
      "learning_rate": 2.3562528087826573e-05,
      "loss": 0.5426,
      "step": 310
    },
    {
      "epoch": 0.5876865671641791,
      "grad_norm": 0.2361656180291179,
      "learning_rate": 2.2880866442929544e-05,
      "loss": 0.5396,
      "step": 315
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 0.23596745328071683,
      "learning_rate": 2.2203603557771447e-05,
      "loss": 0.5357,
      "step": 320
    },
    {
      "epoch": 0.6063432835820896,
      "grad_norm": 0.24184948453857802,
      "learning_rate": 2.153138438376473e-05,
      "loss": 0.5339,
      "step": 325
    },
    {
      "epoch": 0.6156716417910447,
      "grad_norm": 0.23369544200091724,
      "learning_rate": 2.0864849069240645e-05,
      "loss": 0.5386,
      "step": 330
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.2273042144154694,
      "learning_rate": 2.0204632349841667e-05,
      "loss": 0.5355,
      "step": 335
    },
    {
      "epoch": 0.6343283582089553,
      "grad_norm": 0.2094409370066005,
      "learning_rate": 1.9551362944068462e-05,
      "loss": 0.5377,
      "step": 340
    },
    {
      "epoch": 0.6436567164179104,
      "grad_norm": 0.22818911316518153,
      "learning_rate": 1.890566295455678e-05,
      "loss": 0.531,
      "step": 345
    },
    {
      "epoch": 0.6529850746268657,
      "grad_norm": 0.21863119932090663,
      "learning_rate": 1.8268147275654707e-05,
      "loss": 0.541,
      "step": 350
    },
    {
      "epoch": 0.6623134328358209,
      "grad_norm": 0.2106369913594231,
      "learning_rate": 1.7639423007864252e-05,
      "loss": 0.5278,
      "step": 355
    },
    {
      "epoch": 0.6716417910447762,
      "grad_norm": 0.21604757242141778,
      "learning_rate": 1.702008887970491e-05,
      "loss": 0.5287,
      "step": 360
    },
    {
      "epoch": 0.6809701492537313,
      "grad_norm": 0.21597143378114195,
      "learning_rate": 1.6410734677549872e-05,
      "loss": 0.5318,
      "step": 365
    },
    {
      "epoch": 0.6902985074626866,
      "grad_norm": 0.21420392632858606,
      "learning_rate": 1.58119406839777e-05,
      "loss": 0.5308,
      "step": 370
    },
    {
      "epoch": 0.6996268656716418,
      "grad_norm": 0.23220484507328432,
      "learning_rate": 1.5224277125174388e-05,
      "loss": 0.5337,
      "step": 375
    },
    {
      "epoch": 0.7089552238805971,
      "grad_norm": 0.21960291661563633,
      "learning_rate": 1.464830362791204e-05,
      "loss": 0.5359,
      "step": 380
    },
    {
      "epoch": 0.7182835820895522,
      "grad_norm": 0.20363391645776452,
      "learning_rate": 1.4084568686621314e-05,
      "loss": 0.5383,
      "step": 385
    },
    {
      "epoch": 0.7276119402985075,
      "grad_norm": 0.20430129686406656,
      "learning_rate": 1.3533609141065008e-05,
      "loss": 0.5366,
      "step": 390
    },
    {
      "epoch": 0.7369402985074627,
      "grad_norm": 0.21391864672286387,
      "learning_rate": 1.299594966511038e-05,
      "loss": 0.5338,
      "step": 395
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 0.20750528279220357,
      "learning_rate": 1.2472102267086904e-05,
      "loss": 0.5295,
      "step": 400
    },
    {
      "epoch": 0.7555970149253731,
      "grad_norm": 0.20652656803309716,
      "learning_rate": 1.1962565802205255e-05,
      "loss": 0.5362,
      "step": 405
    },
    {
      "epoch": 0.7649253731343284,
      "grad_norm": 0.19195629879165202,
      "learning_rate": 1.1467825497501954e-05,
      "loss": 0.5225,
      "step": 410
    },
    {
      "epoch": 0.7742537313432836,
      "grad_norm": 0.21151922246650445,
      "learning_rate": 1.0988352489762006e-05,
      "loss": 0.5384,
      "step": 415
    },
    {
      "epoch": 0.7835820895522388,
      "grad_norm": 0.204760450632927,
      "learning_rate": 1.052460337685951e-05,
      "loss": 0.5298,
      "step": 420
    },
    {
      "epoch": 0.792910447761194,
      "grad_norm": 0.19922336763600637,
      "learning_rate": 1.0077019782943584e-05,
      "loss": 0.5282,
      "step": 425
    },
    {
      "epoch": 0.8022388059701493,
      "grad_norm": 0.2104055047251735,
      "learning_rate": 9.646027937883622e-06,
      "loss": 0.529,
      "step": 430
    },
    {
      "epoch": 0.8115671641791045,
      "grad_norm": 0.18141387654521896,
      "learning_rate": 9.232038271374377e-06,
      "loss": 0.531,
      "step": 435
    },
    {
      "epoch": 0.8208955223880597,
      "grad_norm": 0.19879663171910475,
      "learning_rate": 8.835445022087426e-06,
      "loss": 0.5256,
      "step": 440
    },
    {
      "epoch": 0.8302238805970149,
      "grad_norm": 0.19565091593372,
      "learning_rate": 8.456625862241193e-06,
      "loss": 0.5358,
      "step": 445
    },
    {
      "epoch": 0.8395522388059702,
      "grad_norm": 0.187496803748416,
      "learning_rate": 8.095941537947057e-06,
      "loss": 0.5328,
      "step": 450
    },
    {
      "epoch": 0.8488805970149254,
      "grad_norm": 0.20371049579133477,
      "learning_rate": 7.753735525674059e-06,
      "loss": 0.5256,
      "step": 455
    },
    {
      "epoch": 0.8582089552238806,
      "grad_norm": 0.1964002704579327,
      "learning_rate": 7.430333705159286e-06,
      "loss": 0.5359,
      "step": 460
    },
    {
      "epoch": 0.8675373134328358,
      "grad_norm": 0.1848105783285991,
      "learning_rate": 7.126044049075548e-06,
      "loss": 0.5408,
      "step": 465
    },
    {
      "epoch": 0.8768656716417911,
      "grad_norm": 0.18246504825507467,
      "learning_rate": 6.8411563297516995e-06,
      "loss": 0.5208,
      "step": 470
    },
    {
      "epoch": 0.8861940298507462,
      "grad_norm": 0.17351842603547765,
      "learning_rate": 6.575941843225068e-06,
      "loss": 0.5246,
      "step": 475
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 0.20341591173191198,
      "learning_rate": 6.330653150888617e-06,
      "loss": 0.5292,
      "step": 480
    },
    {
      "epoch": 0.9048507462686567,
      "grad_norm": 0.20204818775308625,
      "learning_rate": 6.105523838979022e-06,
      "loss": 0.5373,
      "step": 485
    },
    {
      "epoch": 0.914179104477612,
      "grad_norm": 0.18421926232056562,
      "learning_rate": 5.900768296134551e-06,
      "loss": 0.5238,
      "step": 490
    },
    {
      "epoch": 0.9235074626865671,
      "grad_norm": 0.1788995157537539,
      "learning_rate": 5.7165815092346825e-06,
      "loss": 0.526,
      "step": 495
    },
    {
      "epoch": 0.9328358208955224,
      "grad_norm": 0.18233827133162475,
      "learning_rate": 5.553138877715833e-06,
      "loss": 0.5279,
      "step": 500
    },
    {
      "epoch": 0.9421641791044776,
      "grad_norm": 0.182345435097955,
      "learning_rate": 5.410596046540051e-06,
      "loss": 0.5229,
      "step": 505
    },
    {
      "epoch": 0.9514925373134329,
      "grad_norm": 0.19791728733223776,
      "learning_rate": 5.28908875797568e-06,
      "loss": 0.5235,
      "step": 510
    },
    {
      "epoch": 0.960820895522388,
      "grad_norm": 0.18416166735753317,
      "learning_rate": 5.1887327223312296e-06,
      "loss": 0.5246,
      "step": 515
    },
    {
      "epoch": 0.9701492537313433,
      "grad_norm": 0.19627109559115194,
      "learning_rate": 5.109623507765466e-06,
      "loss": 0.5289,
      "step": 520
    },
    {
      "epoch": 0.9794776119402985,
      "grad_norm": 0.19326569885491154,
      "learning_rate": 5.051836449278715e-06,
      "loss": 0.5311,
      "step": 525
    },
    {
      "epoch": 0.9888059701492538,
      "grad_norm": 0.19312098703576008,
      "learning_rate": 5.015426576972003e-06,
      "loss": 0.5298,
      "step": 530
    },
    {
      "epoch": 0.9981343283582089,
      "grad_norm": 0.17724401661078515,
      "learning_rate": 5.000428563642382e-06,
      "loss": 0.5299,
      "step": 535
    },
    {
      "epoch": 1.0,
      "step": 536,
      "total_flos": 488621249396736.0,
      "train_loss": 0.5610535256564617,
      "train_runtime": 6454.4726,
      "train_samples_per_second": 5.314,
      "train_steps_per_second": 0.083
    }
  ],
  "logging_steps": 5,
  "max_steps": 536,
  "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
}