{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9946814168705456,
  "eval_steps": 500,
  "global_step": 586,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0034038932028507603,
      "grad_norm": 15.375,
      "learning_rate": 1.9999856294471073e-05,
      "loss": 1.5065,
      "step": 1
    },
    {
      "epoch": 0.017019466014253803,
      "grad_norm": 2.8125,
      "learning_rate": 1.9996407568285402e-05,
      "loss": 1.0986,
      "step": 5
    },
    {
      "epoch": 0.03403893202850761,
      "grad_norm": 1.84375,
      "learning_rate": 1.9985632854254735e-05,
      "loss": 0.9487,
      "step": 10
    },
    {
      "epoch": 0.05105839804276141,
      "grad_norm": 1.671875,
      "learning_rate": 1.996768359939288e-05,
      "loss": 0.842,
      "step": 15
    },
    {
      "epoch": 0.06807786405701521,
      "grad_norm": 1.7578125,
      "learning_rate": 1.994257269999431e-05,
      "loss": 0.8323,
      "step": 20
    },
    {
      "epoch": 0.08509733007126902,
      "grad_norm": 1.5703125,
      "learning_rate": 1.991031819789732e-05,
      "loss": 0.8031,
      "step": 25
    },
    {
      "epoch": 0.10211679608552282,
      "grad_norm": 1.6796875,
      "learning_rate": 1.9870943267521144e-05,
      "loss": 0.7672,
      "step": 30
    },
    {
      "epoch": 0.11913626209977662,
      "grad_norm": 1.609375,
      "learning_rate": 1.9824476199215526e-05,
      "loss": 0.7578,
      "step": 35
    },
    {
      "epoch": 0.13615572811403043,
      "grad_norm": 1.6875,
      "learning_rate": 1.9770950378934433e-05,
      "loss": 0.7273,
      "step": 40
    },
    {
      "epoch": 0.15317519412828423,
      "grad_norm": 1.5625,
      "learning_rate": 1.9710404264248734e-05,
      "loss": 0.7031,
      "step": 45
    },
    {
      "epoch": 0.17019466014253803,
      "grad_norm": 1.625,
      "learning_rate": 1.964288135671494e-05,
      "loss": 0.699,
      "step": 50
    },
    {
      "epoch": 0.18721412615679184,
      "grad_norm": 1.609375,
      "learning_rate": 1.9568430170619953e-05,
      "loss": 0.6736,
      "step": 55
    },
    {
      "epoch": 0.20423359217104564,
      "grad_norm": 1.7734375,
      "learning_rate": 1.94871041981242e-05,
      "loss": 0.6702,
      "step": 60
    },
    {
      "epoch": 0.22125305818529944,
      "grad_norm": 1.640625,
      "learning_rate": 1.939896187082824e-05,
      "loss": 0.6768,
      "step": 65
    },
    {
      "epoch": 0.23827252419955325,
      "grad_norm": 1.6640625,
      "learning_rate": 1.9304066517790465e-05,
      "loss": 0.6614,
      "step": 70
    },
    {
      "epoch": 0.25529199021380705,
      "grad_norm": 1.6796875,
      "learning_rate": 1.9202486320026045e-05,
      "loss": 0.6553,
      "step": 75
    },
    {
      "epoch": 0.27231145622806086,
      "grad_norm": 1.7578125,
      "learning_rate": 1.9094294261519787e-05,
      "loss": 0.6698,
      "step": 80
    },
    {
      "epoch": 0.28933092224231466,
      "grad_norm": 1.5859375,
      "learning_rate": 1.897956807678813e-05,
      "loss": 0.6697,
      "step": 85
    },
    {
      "epoch": 0.30635038825656846,
      "grad_norm": 1.7890625,
      "learning_rate": 1.8858390195027986e-05,
      "loss": 0.6612,
      "step": 90
    },
    {
      "epoch": 0.32336985427082227,
      "grad_norm": 1.984375,
      "learning_rate": 1.873084768089246e-05,
      "loss": 0.6447,
      "step": 95
    },
    {
      "epoch": 0.34038932028507607,
      "grad_norm": 1.578125,
      "learning_rate": 1.8597032171936103e-05,
      "loss": 0.654,
      "step": 100
    },
    {
      "epoch": 0.3574087862993299,
      "grad_norm": 1.3984375,
      "learning_rate": 1.8457039812774574e-05,
      "loss": 0.6496,
      "step": 105
    },
    {
      "epoch": 0.3744282523135837,
      "grad_norm": 1.5703125,
      "learning_rate": 1.831097118600604e-05,
      "loss": 0.6488,
      "step": 110
    },
    {
      "epoch": 0.3914477183278375,
      "grad_norm": 1.9765625,
      "learning_rate": 1.8158931239943957e-05,
      "loss": 0.6268,
      "step": 115
    },
    {
      "epoch": 0.4084671843420913,
      "grad_norm": 1.6953125,
      "learning_rate": 1.800102921321316e-05,
      "loss": 0.6243,
      "step": 120
    },
    {
      "epoch": 0.4254866503563451,
      "grad_norm": 1.625,
      "learning_rate": 1.7837378556263368e-05,
      "loss": 0.6456,
      "step": 125
    },
    {
      "epoch": 0.4425061163705989,
      "grad_norm": 1.875,
      "learning_rate": 1.766809684985661e-05,
      "loss": 0.6363,
      "step": 130
    },
    {
      "epoch": 0.4595255823848527,
      "grad_norm": 1.6484375,
      "learning_rate": 1.7493305720587047e-05,
      "loss": 0.6448,
      "step": 135
    },
    {
      "epoch": 0.4765450483991065,
      "grad_norm": 1.546875,
      "learning_rate": 1.7313130753493917e-05,
      "loss": 0.6422,
      "step": 140
    },
    {
      "epoch": 0.4935645144133603,
      "grad_norm": 1.671875,
      "learning_rate": 1.7127701401830422e-05,
      "loss": 0.6118,
      "step": 145
    },
    {
      "epoch": 0.5105839804276141,
      "grad_norm": 1.734375,
      "learning_rate": 1.6937150894053306e-05,
      "loss": 0.6255,
      "step": 150
    },
    {
      "epoch": 0.5276034464418679,
      "grad_norm": 1.625,
      "learning_rate": 1.674161613810003e-05,
      "loss": 0.6451,
      "step": 155
    },
    {
      "epoch": 0.5446229124561217,
      "grad_norm": 1.6015625,
      "learning_rate": 1.6541237623022333e-05,
      "loss": 0.6268,
      "step": 160
    },
    {
      "epoch": 0.5616423784703755,
      "grad_norm": 1.5,
      "learning_rate": 1.6336159318046698e-05,
      "loss": 0.6077,
      "step": 165
    },
    {
      "epoch": 0.5786618444846293,
      "grad_norm": 1.5859375,
      "learning_rate": 1.612652856913449e-05,
      "loss": 0.6316,
      "step": 170
    },
    {
      "epoch": 0.5956813104988831,
      "grad_norm": 1.765625,
      "learning_rate": 1.5912495993115848e-05,
      "loss": 0.6167,
      "step": 175
    },
    {
      "epoch": 0.6127007765131369,
      "grad_norm": 1.8203125,
      "learning_rate": 1.5694215369473584e-05,
      "loss": 0.6155,
      "step": 180
    },
    {
      "epoch": 0.6297202425273907,
      "grad_norm": 1.6015625,
      "learning_rate": 1.547184352985472e-05,
      "loss": 0.603,
      "step": 185
    },
    {
      "epoch": 0.6467397085416445,
      "grad_norm": 1.6328125,
      "learning_rate": 1.5245540245389053e-05,
      "loss": 0.5883,
      "step": 190
    },
    {
      "epoch": 0.6637591745558983,
      "grad_norm": 1.65625,
      "learning_rate": 1.501546811189584e-05,
      "loss": 0.59,
      "step": 195
    },
    {
      "epoch": 0.6807786405701521,
      "grad_norm": 1.6171875,
      "learning_rate": 1.4781792433060884e-05,
      "loss": 0.6098,
      "step": 200
    },
    {
      "epoch": 0.6977981065844059,
      "grad_norm": 1.7109375,
      "learning_rate": 1.4544681101668099e-05,
      "loss": 0.5857,
      "step": 205
    },
    {
      "epoch": 0.7148175725986597,
      "grad_norm": 1.5859375,
      "learning_rate": 1.4304304478970839e-05,
      "loss": 0.6076,
      "step": 210
    },
    {
      "epoch": 0.7318370386129135,
      "grad_norm": 1.65625,
      "learning_rate": 1.4060835272289673e-05,
      "loss": 0.6014,
      "step": 215
    },
    {
      "epoch": 0.7488565046271674,
      "grad_norm": 1.6640625,
      "learning_rate": 1.381444841092452e-05,
      "loss": 0.5998,
      "step": 220
    },
    {
      "epoch": 0.7658759706414211,
      "grad_norm": 1.625,
      "learning_rate": 1.3565320920470348e-05,
      "loss": 0.5886,
      "step": 225
    },
    {
      "epoch": 0.782895436655675,
      "grad_norm": 1.6484375,
      "learning_rate": 1.3313631795626691e-05,
      "loss": 0.5963,
      "step": 230
    },
    {
      "epoch": 0.7999149026699287,
      "grad_norm": 1.734375,
      "learning_rate": 1.3059561871592413e-05,
      "loss": 0.5974,
      "step": 235
    },
    {
      "epoch": 0.8169343686841826,
      "grad_norm": 1.6171875,
      "learning_rate": 1.2803293694138077e-05,
      "loss": 0.5849,
      "step": 240
    },
    {
      "epoch": 0.8339538346984363,
      "grad_norm": 1.484375,
      "learning_rate": 1.254501138844931e-05,
      "loss": 0.5683,
      "step": 245
    },
    {
      "epoch": 0.8509733007126902,
      "grad_norm": 1.5703125,
      "learning_rate": 1.228490052683537e-05,
      "loss": 0.5866,
      "step": 250
    },
    {
      "epoch": 0.8679927667269439,
      "grad_norm": 1.546875,
      "learning_rate": 1.202314799539797e-05,
      "loss": 0.6027,
      "step": 255
    },
    {
      "epoch": 0.8850122327411978,
      "grad_norm": 1.5234375,
      "learning_rate": 1.1759941859756173e-05,
      "loss": 0.5975,
      "step": 260
    },
    {
      "epoch": 0.9020316987554515,
      "grad_norm": 1.546875,
      "learning_rate": 1.1495471229923805e-05,
      "loss": 0.5998,
      "step": 265
    },
    {
      "epoch": 0.9190511647697054,
      "grad_norm": 1.6640625,
      "learning_rate": 1.1229926124436506e-05,
      "loss": 0.5762,
      "step": 270
    },
    {
      "epoch": 0.9360706307839591,
      "grad_norm": 1.7109375,
      "learning_rate": 1.0963497333826002e-05,
      "loss": 0.5828,
      "step": 275
    },
    {
      "epoch": 0.953090096798213,
      "grad_norm": 1.671875,
      "learning_rate": 1.0696376283539704e-05,
      "loss": 0.5761,
      "step": 280
    },
    {
      "epoch": 0.9701095628124667,
      "grad_norm": 1.65625,
      "learning_rate": 1.042875489640415e-05,
      "loss": 0.6097,
      "step": 285
    },
    {
      "epoch": 0.9871290288267206,
      "grad_norm": 1.6953125,
      "learning_rate": 1.0160825454731072e-05,
      "loss": 0.5819,
      "step": 290
    },
    {
      "epoch": 1.0041484948409745,
      "grad_norm": 1.546875,
      "learning_rate": 9.892780462165172e-06,
      "loss": 0.5574,
      "step": 295
    },
    {
      "epoch": 1.0211679608552282,
      "grad_norm": 1.5625,
      "learning_rate": 9.624812505372907e-06,
      "loss": 0.579,
      "step": 300
    },
    {
      "epoch": 1.038187426869482,
      "grad_norm": 1.5234375,
      "learning_rate": 9.35711411567156e-06,
      "loss": 0.5607,
      "step": 305
    },
    {
      "epoch": 1.0552068928837357,
      "grad_norm": 1.6484375,
      "learning_rate": 9.089877630698159e-06,
      "loss": 0.5638,
      "step": 310
    },
    {
      "epoch": 1.0722263588979897,
      "grad_norm": 1.4921875,
      "learning_rate": 8.823295056217487e-06,
      "loss": 0.5816,
      "step": 315
    },
    {
      "epoch": 1.0892458249122434,
      "grad_norm": 1.5390625,
      "learning_rate": 8.557557928168568e-06,
      "loss": 0.5614,
      "step": 320
    },
    {
      "epoch": 1.1062652909264972,
      "grad_norm": 1.59375,
      "learning_rate": 8.292857175048715e-06,
      "loss": 0.5566,
      "step": 325
    },
    {
      "epoch": 1.123284756940751,
      "grad_norm": 1.5546875,
      "learning_rate": 8.029382980734e-06,
      "loss": 0.5872,
      "step": 330
    },
    {
      "epoch": 1.1403042229550047,
      "grad_norm": 1.640625,
      "learning_rate": 7.767324647834759e-06,
      "loss": 0.5495,
      "step": 335
    },
    {
      "epoch": 1.1573236889692586,
      "grad_norm": 1.5,
      "learning_rate": 7.506870461684215e-06,
      "loss": 0.5726,
      "step": 340
    },
    {
      "epoch": 1.1743431549835124,
      "grad_norm": 1.4921875,
      "learning_rate": 7.24820755505808e-06,
      "loss": 0.5722,
      "step": 345
    },
    {
      "epoch": 1.1913626209977661,
      "grad_norm": 1.5703125,
      "learning_rate": 6.991521773722186e-06,
      "loss": 0.5901,
      "step": 350
    },
    {
      "epoch": 1.20838208701202,
      "grad_norm": 1.78125,
      "learning_rate": 6.736997542904844e-06,
      "loss": 0.563,
      "step": 355
    },
    {
      "epoch": 1.2254015530262738,
      "grad_norm": 1.515625,
      "learning_rate": 6.484817734789839e-06,
      "loss": 0.5762,
      "step": 360
    },
    {
      "epoch": 1.2424210190405276,
      "grad_norm": 1.46875,
      "learning_rate": 6.235163537125257e-06,
      "loss": 0.5624,
      "step": 365
    },
    {
      "epoch": 1.2594404850547813,
      "grad_norm": 1.7265625,
      "learning_rate": 5.988214323042581e-06,
      "loss": 0.5574,
      "step": 370
    },
    {
      "epoch": 1.276459951069035,
      "grad_norm": 1.4921875,
      "learning_rate": 5.744147522179515e-06,
      "loss": 0.5576,
      "step": 375
    },
    {
      "epoch": 1.293479417083289,
      "grad_norm": 1.5546875,
      "learning_rate": 5.503138493199247e-06,
      "loss": 0.5682,
      "step": 380
    },
    {
      "epoch": 1.3104988830975428,
      "grad_norm": 1.46875,
      "learning_rate": 5.26536039779761e-06,
      "loss": 0.5874,
      "step": 385
    },
    {
      "epoch": 1.3275183491117966,
      "grad_norm": 1.625,
      "learning_rate": 5.030984076288805e-06,
      "loss": 0.5712,
      "step": 390
    },
    {
      "epoch": 1.3445378151260505,
      "grad_norm": 1.6171875,
      "learning_rate": 4.800177924858933e-06,
      "loss": 0.5752,
      "step": 395
    },
    {
      "epoch": 1.3615572811403043,
      "grad_norm": 1.59375,
      "learning_rate": 4.573107774575665e-06,
      "loss": 0.5645,
      "step": 400
    },
    {
      "epoch": 1.378576747154558,
      "grad_norm": 1.59375,
      "learning_rate": 4.34993677224086e-06,
      "loss": 0.5415,
      "step": 405
    },
    {
      "epoch": 1.3955962131688118,
      "grad_norm": 1.59375,
      "learning_rate": 4.130825263171833e-06,
      "loss": 0.5815,
      "step": 410
    },
    {
      "epoch": 1.4126156791830655,
      "grad_norm": 1.5625,
      "learning_rate": 3.915930675995424e-06,
      "loss": 0.5764,
      "step": 415
    },
    {
      "epoch": 1.4296351451973195,
      "grad_norm": 1.59375,
      "learning_rate": 3.7054074095376845e-06,
      "loss": 0.5807,
      "step": 420
    },
    {
      "epoch": 1.4466546112115732,
      "grad_norm": 1.625,
      "learning_rate": 3.4994067218904383e-06,
      "loss": 0.5731,
      "step": 425
    },
    {
      "epoch": 1.463674077225827,
      "grad_norm": 1.640625,
      "learning_rate": 3.2980766217343852e-06,
      "loss": 0.5633,
      "step": 430
    },
    {
      "epoch": 1.480693543240081,
      "grad_norm": 1.53125,
      "learning_rate": 3.101561761996912e-06,
      "loss": 0.5653,
      "step": 435
    },
    {
      "epoch": 1.4977130092543347,
      "grad_norm": 1.609375,
      "learning_rate": 2.910003335920918e-06,
      "loss": 0.5773,
      "step": 440
    },
    {
      "epoch": 1.5147324752685885,
      "grad_norm": 1.5703125,
      "learning_rate": 2.7235389756194097e-06,
      "loss": 0.5664,
      "step": 445
    },
    {
      "epoch": 1.5317519412828422,
      "grad_norm": 1.5546875,
      "learning_rate": 2.542302653188704e-06,
      "loss": 0.5732,
      "step": 450
    },
    {
      "epoch": 1.548771407297096,
      "grad_norm": 1.5625,
      "learning_rate": 2.3664245844513078e-06,
      "loss": 0.561,
      "step": 455
    },
    {
      "epoch": 1.5657908733113497,
      "grad_norm": 1.5625,
      "learning_rate": 2.1960311353976318e-06,
      "loss": 0.5793,
      "step": 460
    },
    {
      "epoch": 1.5828103393256037,
      "grad_norm": 1.5859375,
      "learning_rate": 2.031244731393739e-06,
      "loss": 0.5596,
      "step": 465
    },
    {
      "epoch": 1.5998298053398574,
      "grad_norm": 1.484375,
      "learning_rate": 1.8721837692204115e-06,
      "loss": 0.5539,
      "step": 470
    },
    {
      "epoch": 1.6168492713541114,
      "grad_norm": 1.5,
      "learning_rate": 1.7189625320066573e-06,
      "loss": 0.5489,
      "step": 475
    },
    {
      "epoch": 1.6338687373683651,
      "grad_norm": 1.484375,
      "learning_rate": 1.5716911071188611e-06,
      "loss": 0.5872,
      "step": 480
    },
    {
      "epoch": 1.6508882033826189,
      "grad_norm": 1.515625,
      "learning_rate": 1.4304753070645083e-06,
      "loss": 0.5538,
      "step": 485
    },
    {
      "epoch": 1.6679076693968726,
      "grad_norm": 1.453125,
      "learning_rate": 1.295416593467338e-06,
      "loss": 0.56,
      "step": 490
    },
    {
      "epoch": 1.6849271354111264,
      "grad_norm": 1.515625,
      "learning_rate": 1.166612004168568e-06,
      "loss": 0.569,
      "step": 495
    },
    {
      "epoch": 1.7019466014253801,
      "grad_norm": 1.75,
      "learning_rate": 1.0441540835065101e-06,
      "loss": 0.559,
      "step": 500
    },
    {
      "epoch": 1.718966067439634,
      "grad_norm": 1.5234375,
      "learning_rate": 9.28130815824746e-07,
      "loss": 0.5755,
      "step": 505
    },
    {
      "epoch": 1.7359855334538878,
      "grad_norm": 1.578125,
      "learning_rate": 8.186255622565642e-07,
      "loss": 0.5786,
      "step": 510
    },
    {
      "epoch": 1.7530049994681418,
      "grad_norm": 1.6015625,
      "learning_rate": 7.15717000831132e-07,
      "loss": 0.5783,
      "step": 515
    },
    {
      "epoch": 1.7700244654823956,
      "grad_norm": 1.46875,
      "learning_rate": 6.19479069944402e-07,
      "loss": 0.5747,
      "step": 520
    },
    {
      "epoch": 1.7870439314966493,
      "grad_norm": 1.5625,
      "learning_rate": 5.299809152353863e-07,
      "loss": 0.5789,
      "step": 525
    },
    {
      "epoch": 1.804063397510903,
      "grad_norm": 1.5546875,
      "learning_rate": 4.4728683990596267e-07,
      "loss": 0.5591,
      "step": 530
    },
    {
      "epoch": 1.8210828635251568,
      "grad_norm": 1.5234375,
      "learning_rate": 3.714562585198922e-07,
      "loss": 0.5722,
      "step": 535
    },
    {
      "epoch": 1.8381023295394106,
      "grad_norm": 1.609375,
      "learning_rate": 3.025436543142801e-07,
      "loss": 0.5578,
      "step": 540
    },
    {
      "epoch": 1.8551217955536645,
      "grad_norm": 1.6640625,
      "learning_rate": 2.4059854005410224e-07,
      "loss": 0.5662,
      "step": 545
    },
    {
      "epoch": 1.8721412615679183,
      "grad_norm": 1.5703125,
      "learning_rate": 1.856654224579635e-07,
      "loss": 0.5724,
      "step": 550
    },
    {
      "epoch": 1.8891607275821722,
      "grad_norm": 1.6796875,
      "learning_rate": 1.3778377022063083e-07,
      "loss": 0.5635,
      "step": 555
    },
    {
      "epoch": 1.906180193596426,
      "grad_norm": 1.6484375,
      "learning_rate": 9.698798565531464e-08,
      "loss": 0.5671,
      "step": 560
    },
    {
      "epoch": 1.9231996596106797,
      "grad_norm": 1.5234375,
      "learning_rate": 6.330737997607194e-08,
      "loss": 0.5511,
      "step": 565
    },
    {
      "epoch": 1.9402191256249335,
      "grad_norm": 1.546875,
      "learning_rate": 3.6766152238106866e-08,
      "loss": 0.5733,
      "step": 570
    },
    {
      "epoch": 1.9572385916391872,
      "grad_norm": 1.59375,
      "learning_rate": 1.738337195107187e-08,
      "loss": 0.5675,
      "step": 575
    },
    {
      "epoch": 1.974258057653441,
      "grad_norm": 1.5859375,
      "learning_rate": 5.172965377890915e-09,
      "loss": 0.5654,
      "step": 580
    },
    {
      "epoch": 1.991277523667695,
      "grad_norm": 1.53125,
      "learning_rate": 1.4370552892883916e-10,
      "loss": 0.5711,
      "step": 585
    },
    {
      "epoch": 1.9946814168705456,
      "step": 586,
      "total_flos": 2.7337919075608166e+17,
      "train_loss": 0.6114459749781638,
      "train_runtime": 7063.6073,
      "train_samples_per_second": 10.647,
      "train_steps_per_second": 0.083
    }
  ],
  "logging_steps": 5,
  "max_steps": 586,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "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": 2.7337919075608166e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}