{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1084,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0009225092250922509,
      "grad_norm": 9.870575577438405,
      "learning_rate": 1.8348623853211012e-07,
      "loss": 1.1607,
      "step": 1
    },
    {
      "epoch": 0.004612546125461255,
      "grad_norm": 7.65916956083968,
      "learning_rate": 9.174311926605506e-07,
      "loss": 1.1303,
      "step": 5
    },
    {
      "epoch": 0.00922509225092251,
      "grad_norm": 3.7726483117427345,
      "learning_rate": 1.8348623853211011e-06,
      "loss": 1.057,
      "step": 10
    },
    {
      "epoch": 0.013837638376383764,
      "grad_norm": 2.5379513268735505,
      "learning_rate": 2.7522935779816517e-06,
      "loss": 1.0238,
      "step": 15
    },
    {
      "epoch": 0.01845018450184502,
      "grad_norm": 2.0372936699866466,
      "learning_rate": 3.6697247706422022e-06,
      "loss": 0.9931,
      "step": 20
    },
    {
      "epoch": 0.023062730627306273,
      "grad_norm": 1.9015079532614876,
      "learning_rate": 4.587155963302753e-06,
      "loss": 0.9821,
      "step": 25
    },
    {
      "epoch": 0.027675276752767528,
      "grad_norm": 2.2588401749573546,
      "learning_rate": 5.504587155963303e-06,
      "loss": 0.9798,
      "step": 30
    },
    {
      "epoch": 0.03228782287822878,
      "grad_norm": 2.7788236168992237,
      "learning_rate": 6.422018348623854e-06,
      "loss": 0.9812,
      "step": 35
    },
    {
      "epoch": 0.03690036900369004,
      "grad_norm": 2.0946702560307022,
      "learning_rate": 7.3394495412844045e-06,
      "loss": 0.9807,
      "step": 40
    },
    {
      "epoch": 0.04151291512915129,
      "grad_norm": 2.5421232639973415,
      "learning_rate": 8.256880733944956e-06,
      "loss": 0.9867,
      "step": 45
    },
    {
      "epoch": 0.046125461254612546,
      "grad_norm": 2.346175802405767,
      "learning_rate": 9.174311926605506e-06,
      "loss": 0.9966,
      "step": 50
    },
    {
      "epoch": 0.0507380073800738,
      "grad_norm": 2.4887196180378783,
      "learning_rate": 1.0091743119266055e-05,
      "loss": 0.9908,
      "step": 55
    },
    {
      "epoch": 0.055350553505535055,
      "grad_norm": 2.0063521735753165,
      "learning_rate": 1.1009174311926607e-05,
      "loss": 0.9924,
      "step": 60
    },
    {
      "epoch": 0.05996309963099631,
      "grad_norm": 1.9753528261008033,
      "learning_rate": 1.1926605504587156e-05,
      "loss": 1.0028,
      "step": 65
    },
    {
      "epoch": 0.06457564575645756,
      "grad_norm": 2.8188122479797793,
      "learning_rate": 1.2844036697247708e-05,
      "loss": 0.9874,
      "step": 70
    },
    {
      "epoch": 0.06918819188191883,
      "grad_norm": 2.4283436777324243,
      "learning_rate": 1.3761467889908258e-05,
      "loss": 1.0038,
      "step": 75
    },
    {
      "epoch": 0.07380073800738007,
      "grad_norm": 2.2694140697845735,
      "learning_rate": 1.4678899082568809e-05,
      "loss": 1.0162,
      "step": 80
    },
    {
      "epoch": 0.07841328413284133,
      "grad_norm": 2.2296135287203533,
      "learning_rate": 1.559633027522936e-05,
      "loss": 1.0022,
      "step": 85
    },
    {
      "epoch": 0.08302583025830258,
      "grad_norm": 2.496334289275322,
      "learning_rate": 1.6513761467889912e-05,
      "loss": 0.9963,
      "step": 90
    },
    {
      "epoch": 0.08763837638376384,
      "grad_norm": 3.048918418901738,
      "learning_rate": 1.743119266055046e-05,
      "loss": 0.9874,
      "step": 95
    },
    {
      "epoch": 0.09225092250922509,
      "grad_norm": 2.219949213385871,
      "learning_rate": 1.834862385321101e-05,
      "loss": 1.0366,
      "step": 100
    },
    {
      "epoch": 0.09686346863468635,
      "grad_norm": 2.3152361863940305,
      "learning_rate": 1.9266055045871563e-05,
      "loss": 1.009,
      "step": 105
    },
    {
      "epoch": 0.1014760147601476,
      "grad_norm": 2.1557669545307925,
      "learning_rate": 1.9999948088910656e-05,
      "loss": 1.0126,
      "step": 110
    },
    {
      "epoch": 0.10608856088560886,
      "grad_norm": 2.5252882972323283,
      "learning_rate": 1.9998131257372878e-05,
      "loss": 1.0228,
      "step": 115
    },
    {
      "epoch": 0.11070110701107011,
      "grad_norm": 1.754976914882788,
      "learning_rate": 1.999371941029485e-05,
      "loss": 1.0132,
      "step": 120
    },
    {
      "epoch": 0.11531365313653137,
      "grad_norm": 1.8015448977018316,
      "learning_rate": 1.9986713692771732e-05,
      "loss": 1.0354,
      "step": 125
    },
    {
      "epoch": 0.11992619926199262,
      "grad_norm": 1.8342985482903624,
      "learning_rate": 1.9977115923137912e-05,
      "loss": 1.0299,
      "step": 130
    },
    {
      "epoch": 0.12453874538745388,
      "grad_norm": 1.7924526745821787,
      "learning_rate": 1.9964928592495046e-05,
      "loss": 1.0315,
      "step": 135
    },
    {
      "epoch": 0.12915129151291513,
      "grad_norm": 15.523610268563516,
      "learning_rate": 1.9950154864065497e-05,
      "loss": 1.0376,
      "step": 140
    },
    {
      "epoch": 0.13376383763837638,
      "grad_norm": 2.102764923940876,
      "learning_rate": 1.993279857237133e-05,
      "loss": 1.0436,
      "step": 145
    },
    {
      "epoch": 0.13837638376383765,
      "grad_norm": 2.1109629634915197,
      "learning_rate": 1.9912864222239045e-05,
      "loss": 1.0436,
      "step": 150
    },
    {
      "epoch": 0.1429889298892989,
      "grad_norm": 2.01098830232536,
      "learning_rate": 1.9890356987630362e-05,
      "loss": 1.0206,
      "step": 155
    },
    {
      "epoch": 0.14760147601476015,
      "grad_norm": 2.0368730493904157,
      "learning_rate": 1.986528271029931e-05,
      "loss": 1.0232,
      "step": 160
    },
    {
      "epoch": 0.1522140221402214,
      "grad_norm": 1.6688489601440277,
      "learning_rate": 1.9837647898276008e-05,
      "loss": 1.0236,
      "step": 165
    },
    {
      "epoch": 0.15682656826568267,
      "grad_norm": 2.0005420507660334,
      "learning_rate": 1.9807459724177497e-05,
      "loss": 1.0246,
      "step": 170
    },
    {
      "epoch": 0.16143911439114392,
      "grad_norm": 1.830621233659763,
      "learning_rate": 1.977472602334609e-05,
      "loss": 1.0322,
      "step": 175
    },
    {
      "epoch": 0.16605166051660517,
      "grad_norm": 2.0747249565016426,
      "learning_rate": 1.973945529181572e-05,
      "loss": 1.0356,
      "step": 180
    },
    {
      "epoch": 0.1706642066420664,
      "grad_norm": 1.8657819252798358,
      "learning_rate": 1.9701656684106764e-05,
      "loss": 1.0406,
      "step": 185
    },
    {
      "epoch": 0.1752767527675277,
      "grad_norm": 1.863882659265557,
      "learning_rate": 1.9661340010850025e-05,
      "loss": 1.0295,
      "step": 190
    },
    {
      "epoch": 0.17988929889298894,
      "grad_norm": 1.907383023598496,
      "learning_rate": 1.9618515736240353e-05,
      "loss": 1.0518,
      "step": 195
    },
    {
      "epoch": 0.18450184501845018,
      "grad_norm": 1.8946641894052252,
      "learning_rate": 1.9573194975320672e-05,
      "loss": 1.0382,
      "step": 200
    },
    {
      "epoch": 0.18911439114391143,
      "grad_norm": 1.8166789629573785,
      "learning_rate": 1.952538949109708e-05,
      "loss": 1.0443,
      "step": 205
    },
    {
      "epoch": 0.1937269372693727,
      "grad_norm": 1.7352343980788758,
      "learning_rate": 1.9475111691485737e-05,
      "loss": 1.0217,
      "step": 210
    },
    {
      "epoch": 0.19833948339483395,
      "grad_norm": 1.6974181401216584,
      "learning_rate": 1.9422374626092414e-05,
      "loss": 1.0305,
      "step": 215
    },
    {
      "epoch": 0.2029520295202952,
      "grad_norm": 1.5750696152336092,
      "learning_rate": 1.936719198282545e-05,
      "loss": 1.0293,
      "step": 220
    },
    {
      "epoch": 0.20756457564575645,
      "grad_norm": 1.623225736738918,
      "learning_rate": 1.930957808434307e-05,
      "loss": 1.0269,
      "step": 225
    },
    {
      "epoch": 0.21217712177121772,
      "grad_norm": 1.631576712487477,
      "learning_rate": 1.9249547884335917e-05,
      "loss": 1.0305,
      "step": 230
    },
    {
      "epoch": 0.21678966789667897,
      "grad_norm": 1.7005295416853186,
      "learning_rate": 1.9187116963645845e-05,
      "loss": 1.0356,
      "step": 235
    },
    {
      "epoch": 0.22140221402214022,
      "grad_norm": 1.7890847993052705,
      "learning_rate": 1.912230152622189e-05,
      "loss": 1.033,
      "step": 240
    },
    {
      "epoch": 0.22601476014760147,
      "grad_norm": 1.7968201186150952,
      "learning_rate": 1.9055118394914545e-05,
      "loss": 1.0367,
      "step": 245
    },
    {
      "epoch": 0.23062730627306274,
      "grad_norm": 1.5783711589062268,
      "learning_rate": 1.898558500710939e-05,
      "loss": 1.0329,
      "step": 250
    },
    {
      "epoch": 0.235239852398524,
      "grad_norm": 1.7821446933403304,
      "learning_rate": 1.891371941020121e-05,
      "loss": 1.0538,
      "step": 255
    },
    {
      "epoch": 0.23985239852398524,
      "grad_norm": 1.7571850494815153,
      "learning_rate": 1.88395402569098e-05,
      "loss": 1.0172,
      "step": 260
    },
    {
      "epoch": 0.2444649446494465,
      "grad_norm": 1.6777787820198717,
      "learning_rate": 1.8763066800438638e-05,
      "loss": 1.0262,
      "step": 265
    },
    {
      "epoch": 0.24907749077490776,
      "grad_norm": 1.7754650640017842,
      "learning_rate": 1.868431888947773e-05,
      "loss": 1.0378,
      "step": 270
    },
    {
      "epoch": 0.253690036900369,
      "grad_norm": 1.6204893672219858,
      "learning_rate": 1.860331696305188e-05,
      "loss": 1.0272,
      "step": 275
    },
    {
      "epoch": 0.25830258302583026,
      "grad_norm": 1.6082347328850755,
      "learning_rate": 1.852008204521572e-05,
      "loss": 1.0199,
      "step": 280
    },
    {
      "epoch": 0.2629151291512915,
      "grad_norm": 1.5512676811150408,
      "learning_rate": 1.8434635739596945e-05,
      "loss": 1.0365,
      "step": 285
    },
    {
      "epoch": 0.26752767527675275,
      "grad_norm": 1.7967133178474277,
      "learning_rate": 1.834700022378907e-05,
      "loss": 1.0541,
      "step": 290
    },
    {
      "epoch": 0.272140221402214,
      "grad_norm": 1.6009722961453385,
      "learning_rate": 1.825719824359524e-05,
      "loss": 1.0399,
      "step": 295
    },
    {
      "epoch": 0.2767527675276753,
      "grad_norm": 1.549904165894849,
      "learning_rate": 1.816525310712456e-05,
      "loss": 1.0096,
      "step": 300
    },
    {
      "epoch": 0.28136531365313655,
      "grad_norm": 1.515466383587945,
      "learning_rate": 1.8071188678742457e-05,
      "loss": 1.0166,
      "step": 305
    },
    {
      "epoch": 0.2859778597785978,
      "grad_norm": 1.5677132403955465,
      "learning_rate": 1.7975029372876706e-05,
      "loss": 1.0209,
      "step": 310
    },
    {
      "epoch": 0.29059040590405905,
      "grad_norm": 1.542178919337743,
      "learning_rate": 1.787680014768065e-05,
      "loss": 1.0434,
      "step": 315
    },
    {
      "epoch": 0.2952029520295203,
      "grad_norm": 1.6414967467646593,
      "learning_rate": 1.777652649855531e-05,
      "loss": 1.0184,
      "step": 320
    },
    {
      "epoch": 0.29981549815498154,
      "grad_norm": 1.6372383463858382,
      "learning_rate": 1.7674234451532065e-05,
      "loss": 1.0317,
      "step": 325
    },
    {
      "epoch": 0.3044280442804428,
      "grad_norm": 1.650440536079242,
      "learning_rate": 1.7569950556517566e-05,
      "loss": 1.052,
      "step": 330
    },
    {
      "epoch": 0.30904059040590404,
      "grad_norm": 1.553654913479007,
      "learning_rate": 1.7463701880402738e-05,
      "loss": 1.0205,
      "step": 335
    },
    {
      "epoch": 0.31365313653136534,
      "grad_norm": 1.6079774780306095,
      "learning_rate": 1.7355516000037555e-05,
      "loss": 1.0289,
      "step": 340
    },
    {
      "epoch": 0.3182656826568266,
      "grad_norm": 1.6788267575243834,
      "learning_rate": 1.7245420995073453e-05,
      "loss": 1.0452,
      "step": 345
    },
    {
      "epoch": 0.32287822878228783,
      "grad_norm": 1.5790256320964164,
      "learning_rate": 1.7133445440675268e-05,
      "loss": 1.028,
      "step": 350
    },
    {
      "epoch": 0.3274907749077491,
      "grad_norm": 1.5621840814508567,
      "learning_rate": 1.7019618400104572e-05,
      "loss": 1.0262,
      "step": 355
    },
    {
      "epoch": 0.33210332103321033,
      "grad_norm": 1.519849047216207,
      "learning_rate": 1.6903969417176244e-05,
      "loss": 1.0183,
      "step": 360
    },
    {
      "epoch": 0.3367158671586716,
      "grad_norm": 1.5037917545786645,
      "learning_rate": 1.6786528508590436e-05,
      "loss": 1.0334,
      "step": 365
    },
    {
      "epoch": 0.3413284132841328,
      "grad_norm": 1.750901636962616,
      "learning_rate": 1.666732615614169e-05,
      "loss": 1.0198,
      "step": 370
    },
    {
      "epoch": 0.3459409594095941,
      "grad_norm": 1.4681811166568075,
      "learning_rate": 1.6546393298807405e-05,
      "loss": 1.0181,
      "step": 375
    },
    {
      "epoch": 0.3505535055350554,
      "grad_norm": 1.9552415843064699,
      "learning_rate": 1.6423761324717636e-05,
      "loss": 1.0237,
      "step": 380
    },
    {
      "epoch": 0.3551660516605166,
      "grad_norm": 1.5359322834804126,
      "learning_rate": 1.6299462063008272e-05,
      "loss": 1.0164,
      "step": 385
    },
    {
      "epoch": 0.35977859778597787,
      "grad_norm": 1.538202612072466,
      "learning_rate": 1.61735277755598e-05,
      "loss": 1.0236,
      "step": 390
    },
    {
      "epoch": 0.3643911439114391,
      "grad_norm": 1.8749904513501805,
      "learning_rate": 1.6045991148623752e-05,
      "loss": 1.0168,
      "step": 395
    },
    {
      "epoch": 0.36900369003690037,
      "grad_norm": 1.4674737229917856,
      "learning_rate": 1.5916885284338937e-05,
      "loss": 1.0042,
      "step": 400
    },
    {
      "epoch": 0.3736162361623616,
      "grad_norm": 1.4565983698821008,
      "learning_rate": 1.5786243692139826e-05,
      "loss": 1.0269,
      "step": 405
    },
    {
      "epoch": 0.37822878228782286,
      "grad_norm": 1.5930896810188973,
      "learning_rate": 1.5654100280059155e-05,
      "loss": 1.0039,
      "step": 410
    },
    {
      "epoch": 0.3828413284132841,
      "grad_norm": 1.4721927126588246,
      "learning_rate": 1.5520489345927095e-05,
      "loss": 1.0036,
      "step": 415
    },
    {
      "epoch": 0.3874538745387454,
      "grad_norm": 1.5395462162350828,
      "learning_rate": 1.538544556846925e-05,
      "loss": 1.0253,
      "step": 420
    },
    {
      "epoch": 0.39206642066420666,
      "grad_norm": 1.4973959788085256,
      "learning_rate": 1.5249003998305787e-05,
      "loss": 1.0142,
      "step": 425
    },
    {
      "epoch": 0.3966789667896679,
      "grad_norm": 1.6021819916230071,
      "learning_rate": 1.5111200048854055e-05,
      "loss": 1.0203,
      "step": 430
    },
    {
      "epoch": 0.40129151291512916,
      "grad_norm": 1.5164653354921802,
      "learning_rate": 1.4972069487137024e-05,
      "loss": 1.0097,
      "step": 435
    },
    {
      "epoch": 0.4059040590405904,
      "grad_norm": 1.6120537324024575,
      "learning_rate": 1.4831648424499953e-05,
      "loss": 1.0156,
      "step": 440
    },
    {
      "epoch": 0.41051660516605165,
      "grad_norm": 1.4054842646806989,
      "learning_rate": 1.4689973307237687e-05,
      "loss": 1.0282,
      "step": 445
    },
    {
      "epoch": 0.4151291512915129,
      "grad_norm": 1.4671822665223906,
      "learning_rate": 1.4547080907135024e-05,
      "loss": 1.0125,
      "step": 450
    },
    {
      "epoch": 0.41974169741697415,
      "grad_norm": 1.464262893049375,
      "learning_rate": 1.4403008311922593e-05,
      "loss": 1.0106,
      "step": 455
    },
    {
      "epoch": 0.42435424354243545,
      "grad_norm": 1.443587186675441,
      "learning_rate": 1.4257792915650728e-05,
      "loss": 1.013,
      "step": 460
    },
    {
      "epoch": 0.4289667896678967,
      "grad_norm": 1.3761731833498907,
      "learning_rate": 1.4111472408983843e-05,
      "loss": 1.0073,
      "step": 465
    },
    {
      "epoch": 0.43357933579335795,
      "grad_norm": 1.529603561908737,
      "learning_rate": 1.3964084769417823e-05,
      "loss": 1.0296,
      "step": 470
    },
    {
      "epoch": 0.4381918819188192,
      "grad_norm": 1.6733724515060537,
      "learning_rate": 1.3815668251422953e-05,
      "loss": 0.9951,
      "step": 475
    },
    {
      "epoch": 0.44280442804428044,
      "grad_norm": 1.4450871420978533,
      "learning_rate": 1.3666261376514978e-05,
      "loss": 0.9927,
      "step": 480
    },
    {
      "epoch": 0.4474169741697417,
      "grad_norm": 1.4561810147046073,
      "learning_rate": 1.3515902923256832e-05,
      "loss": 1.0082,
      "step": 485
    },
    {
      "epoch": 0.45202952029520294,
      "grad_norm": 1.5238257427571544,
      "learning_rate": 1.3364631917193671e-05,
      "loss": 1.0025,
      "step": 490
    },
    {
      "epoch": 0.4566420664206642,
      "grad_norm": 1.5528476103643793,
      "learning_rate": 1.321248762072377e-05,
      "loss": 1.0031,
      "step": 495
    },
    {
      "epoch": 0.4612546125461255,
      "grad_norm": 1.4687610879413946,
      "learning_rate": 1.3059509522907998e-05,
      "loss": 0.9982,
      "step": 500
    },
    {
      "epoch": 0.46586715867158673,
      "grad_norm": 1.4093011733734708,
      "learning_rate": 1.2905737329220394e-05,
      "loss": 0.9966,
      "step": 505
    },
    {
      "epoch": 0.470479704797048,
      "grad_norm": 1.44715856596638,
      "learning_rate": 1.2751210951242636e-05,
      "loss": 1.0052,
      "step": 510
    },
    {
      "epoch": 0.47509225092250923,
      "grad_norm": 1.3996570148561756,
      "learning_rate": 1.2595970496304975e-05,
      "loss": 0.9942,
      "step": 515
    },
    {
      "epoch": 0.4797047970479705,
      "grad_norm": 1.4325383073010012,
      "learning_rate": 1.2440056257076376e-05,
      "loss": 1.0081,
      "step": 520
    },
    {
      "epoch": 0.4843173431734317,
      "grad_norm": 1.4383492965547426,
      "learning_rate": 1.2283508701106559e-05,
      "loss": 0.9788,
      "step": 525
    },
    {
      "epoch": 0.488929889298893,
      "grad_norm": 1.4099081664313255,
      "learning_rate": 1.2126368460322637e-05,
      "loss": 1.0083,
      "step": 530
    },
    {
      "epoch": 0.4935424354243542,
      "grad_norm": 1.3412375300770663,
      "learning_rate": 1.1968676320483103e-05,
      "loss": 0.9684,
      "step": 535
    },
    {
      "epoch": 0.4981549815498155,
      "grad_norm": 1.4515358137974845,
      "learning_rate": 1.1810473210591882e-05,
      "loss": 0.9852,
      "step": 540
    },
    {
      "epoch": 0.5027675276752768,
      "grad_norm": 1.4597275022169718,
      "learning_rate": 1.1651800192275197e-05,
      "loss": 0.9793,
      "step": 545
    },
    {
      "epoch": 0.507380073800738,
      "grad_norm": 1.3868232464146542,
      "learning_rate": 1.1492698449124042e-05,
      "loss": 0.9907,
      "step": 550
    },
    {
      "epoch": 0.5119926199261993,
      "grad_norm": 1.4023901770546143,
      "learning_rate": 1.1333209276004959e-05,
      "loss": 0.9967,
      "step": 555
    },
    {
      "epoch": 0.5166051660516605,
      "grad_norm": 1.403043028756451,
      "learning_rate": 1.1173374068341962e-05,
      "loss": 1.0058,
      "step": 560
    },
    {
      "epoch": 0.5212177121771218,
      "grad_norm": 1.5463650175486758,
      "learning_rate": 1.1013234311372353e-05,
      "loss": 0.9843,
      "step": 565
    },
    {
      "epoch": 0.525830258302583,
      "grad_norm": 1.566717033090475,
      "learning_rate": 1.0852831569379217e-05,
      "loss": 0.9911,
      "step": 570
    },
    {
      "epoch": 0.5304428044280443,
      "grad_norm": 1.4828461511508186,
      "learning_rate": 1.0692207474903421e-05,
      "loss": 0.9909,
      "step": 575
    },
    {
      "epoch": 0.5350553505535055,
      "grad_norm": 2.8521203782189253,
      "learning_rate": 1.0531403717937888e-05,
      "loss": 0.9927,
      "step": 580
    },
    {
      "epoch": 0.5396678966789668,
      "grad_norm": 1.512595456828133,
      "learning_rate": 1.037046203510694e-05,
      "loss": 0.979,
      "step": 585
    },
    {
      "epoch": 0.544280442804428,
      "grad_norm": 1.5368674133781381,
      "learning_rate": 1.0209424198833571e-05,
      "loss": 0.9877,
      "step": 590
    },
    {
      "epoch": 0.5488929889298892,
      "grad_norm": 1.6413121795161139,
      "learning_rate": 1.0048332006497406e-05,
      "loss": 0.9954,
      "step": 595
    },
    {
      "epoch": 0.5535055350553506,
      "grad_norm": 1.4636611051453832,
      "learning_rate": 9.887227269586184e-06,
      "loss": 0.9801,
      "step": 600
    },
    {
      "epoch": 0.5581180811808119,
      "grad_norm": 1.4876078711300504,
      "learning_rate": 9.7261518028436e-06,
      "loss": 0.9896,
      "step": 605
    },
    {
      "epoch": 0.5627306273062731,
      "grad_norm": 1.4565652096222934,
      "learning_rate": 9.565147413416266e-06,
      "loss": 0.9782,
      "step": 610
    },
    {
      "epoch": 0.5673431734317343,
      "grad_norm": 1.3376736240483553,
      "learning_rate": 9.404255890002677e-06,
      "loss": 0.996,
      "step": 615
    },
    {
      "epoch": 0.5719557195571956,
      "grad_norm": 1.266054229637192,
      "learning_rate": 9.243518992006944e-06,
      "loss": 0.974,
      "step": 620
    },
    {
      "epoch": 0.5765682656826568,
      "grad_norm": 1.3235752371132756,
      "learning_rate": 9.082978438700138e-06,
      "loss": 0.9679,
      "step": 625
    },
    {
      "epoch": 0.5811808118081181,
      "grad_norm": 1.4334968823366236,
      "learning_rate": 8.922675898392072e-06,
      "loss": 0.9666,
      "step": 630
    },
    {
      "epoch": 0.5857933579335793,
      "grad_norm": 1.373620518219274,
      "learning_rate": 8.762652977616258e-06,
      "loss": 0.9862,
      "step": 635
    },
    {
      "epoch": 0.5904059040590406,
      "grad_norm": 1.4317910304850499,
      "learning_rate": 8.602951210330942e-06,
      "loss": 0.9662,
      "step": 640
    },
    {
      "epoch": 0.5950184501845018,
      "grad_norm": 1.3805470019673511,
      "learning_rate": 8.443612047138965e-06,
      "loss": 0.9836,
      "step": 645
    },
    {
      "epoch": 0.5996309963099631,
      "grad_norm": 1.4344389139852456,
      "learning_rate": 8.284676844529258e-06,
      "loss": 0.9458,
      "step": 650
    },
    {
      "epoch": 0.6042435424354243,
      "grad_norm": 1.2923675142084592,
      "learning_rate": 8.126186854142752e-06,
      "loss": 0.953,
      "step": 655
    },
    {
      "epoch": 0.6088560885608856,
      "grad_norm": 1.3520769915175057,
      "learning_rate": 7.968183212065537e-06,
      "loss": 0.965,
      "step": 660
    },
    {
      "epoch": 0.6134686346863468,
      "grad_norm": 1.3540469095893106,
      "learning_rate": 7.81070692815195e-06,
      "loss": 0.962,
      "step": 665
    },
    {
      "epoch": 0.6180811808118081,
      "grad_norm": 1.4140291089073116,
      "learning_rate": 7.6537988753805e-06,
      "loss": 0.9566,
      "step": 670
    },
    {
      "epoch": 0.6226937269372693,
      "grad_norm": 1.3299616725752847,
      "learning_rate": 7.497499779245268e-06,
      "loss": 0.965,
      "step": 675
    },
    {
      "epoch": 0.6273062730627307,
      "grad_norm": 1.3769480821952969,
      "learning_rate": 7.3418502071856004e-06,
      "loss": 0.9736,
      "step": 680
    },
    {
      "epoch": 0.6319188191881919,
      "grad_norm": 1.3550016015001538,
      "learning_rate": 7.186890558056836e-06,
      "loss": 0.9528,
      "step": 685
    },
    {
      "epoch": 0.6365313653136532,
      "grad_norm": 1.4444985355599935,
      "learning_rate": 7.0326610516447825e-06,
      "loss": 0.9518,
      "step": 690
    },
    {
      "epoch": 0.6411439114391144,
      "grad_norm": 1.3752886192742222,
      "learning_rate": 6.879201718226658e-06,
      "loss": 0.9437,
      "step": 695
    },
    {
      "epoch": 0.6457564575645757,
      "grad_norm": 1.2728918848531818,
      "learning_rate": 6.7265523881812335e-06,
      "loss": 0.9564,
      "step": 700
    },
    {
      "epoch": 0.6503690036900369,
      "grad_norm": 1.2776578000630632,
      "learning_rate": 6.574752681650864e-06,
      "loss": 0.9431,
      "step": 705
    },
    {
      "epoch": 0.6549815498154982,
      "grad_norm": 1.2934037917676169,
      "learning_rate": 6.423841998258069e-06,
      "loss": 0.9491,
      "step": 710
    },
    {
      "epoch": 0.6595940959409594,
      "grad_norm": 1.3237574013597018,
      "learning_rate": 6.273859506879365e-06,
      "loss": 0.9681,
      "step": 715
    },
    {
      "epoch": 0.6642066420664207,
      "grad_norm": 1.3217921823273022,
      "learning_rate": 6.124844135478971e-06,
      "loss": 0.9649,
      "step": 720
    },
    {
      "epoch": 0.6688191881918819,
      "grad_norm": 1.2737386646079398,
      "learning_rate": 5.976834561005069e-06,
      "loss": 0.9606,
      "step": 725
    },
    {
      "epoch": 0.6734317343173432,
      "grad_norm": 1.3636332749577897,
      "learning_rate": 5.829869199351188e-06,
      "loss": 0.9491,
      "step": 730
    },
    {
      "epoch": 0.6780442804428044,
      "grad_norm": 1.308891744452198,
      "learning_rate": 5.68398619538536e-06,
      "loss": 0.9431,
      "step": 735
    },
    {
      "epoch": 0.6826568265682657,
      "grad_norm": 1.3623472799172087,
      "learning_rate": 5.53922341304961e-06,
      "loss": 0.938,
      "step": 740
    },
    {
      "epoch": 0.6872693726937269,
      "grad_norm": 1.3779395937417078,
      "learning_rate": 5.39561842553239e-06,
      "loss": 0.9765,
      "step": 745
    },
    {
      "epoch": 0.6918819188191881,
      "grad_norm": 1.314594616549353,
      "learning_rate": 5.2532085055164205e-06,
      "loss": 0.9495,
      "step": 750
    },
    {
      "epoch": 0.6964944649446494,
      "grad_norm": 1.2958945545107472,
      "learning_rate": 5.112030615504601e-06,
      "loss": 0.9526,
      "step": 755
    },
    {
      "epoch": 0.7011070110701108,
      "grad_norm": 1.268037176857953,
      "learning_rate": 4.972121398226371e-06,
      "loss": 0.9389,
      "step": 760
    },
    {
      "epoch": 0.705719557195572,
      "grad_norm": 1.3213967171034264,
      "learning_rate": 4.833517167127077e-06,
      "loss": 0.9541,
      "step": 765
    },
    {
      "epoch": 0.7103321033210332,
      "grad_norm": 1.303739776653261,
      "learning_rate": 4.6962538969428416e-06,
      "loss": 0.9523,
      "step": 770
    },
    {
      "epoch": 0.7149446494464945,
      "grad_norm": 1.2626443878609368,
      "learning_rate": 4.560367214363295e-06,
      "loss": 0.9463,
      "step": 775
    },
    {
      "epoch": 0.7195571955719557,
      "grad_norm": 1.4159625474171884,
      "learning_rate": 4.425892388784681e-06,
      "loss": 0.9486,
      "step": 780
    },
    {
      "epoch": 0.724169741697417,
      "grad_norm": 1.4010891439452202,
      "learning_rate": 4.292864323155684e-06,
      "loss": 0.9353,
      "step": 785
    },
    {
      "epoch": 0.7287822878228782,
      "grad_norm": 1.3287864388096529,
      "learning_rate": 4.161317544918345e-06,
      "loss": 0.9389,
      "step": 790
    },
    {
      "epoch": 0.7333948339483395,
      "grad_norm": 1.306056357380541,
      "learning_rate": 4.031286197046493e-06,
      "loss": 0.9423,
      "step": 795
    },
    {
      "epoch": 0.7380073800738007,
      "grad_norm": 1.3004419319996845,
      "learning_rate": 3.902804029183907e-06,
      "loss": 0.9448,
      "step": 800
    },
    {
      "epoch": 0.742619926199262,
      "grad_norm": 1.3193243689915886,
      "learning_rate": 3.775904388884618e-06,
      "loss": 0.9427,
      "step": 805
    },
    {
      "epoch": 0.7472324723247232,
      "grad_norm": 1.296148330228006,
      "learning_rate": 3.650620212957524e-06,
      "loss": 0.9412,
      "step": 810
    },
    {
      "epoch": 0.7518450184501845,
      "grad_norm": 1.4326318299445335,
      "learning_rate": 3.5269840189176616e-06,
      "loss": 0.9289,
      "step": 815
    },
    {
      "epoch": 0.7564575645756457,
      "grad_norm": 1.3055827977236605,
      "learning_rate": 3.405027896546277e-06,
      "loss": 0.926,
      "step": 820
    },
    {
      "epoch": 0.761070110701107,
      "grad_norm": 1.327865337759123,
      "learning_rate": 3.2847834995619067e-06,
      "loss": 0.9453,
      "step": 825
    },
    {
      "epoch": 0.7656826568265682,
      "grad_norm": 1.2543373923102779,
      "learning_rate": 3.1662820374046776e-06,
      "loss": 0.9357,
      "step": 830
    },
    {
      "epoch": 0.7702952029520295,
      "grad_norm": 1.311192918139976,
      "learning_rate": 3.0495542671358745e-06,
      "loss": 0.9299,
      "step": 835
    },
    {
      "epoch": 0.7749077490774908,
      "grad_norm": 1.3013405217794898,
      "learning_rate": 2.934630485454948e-06,
      "loss": 0.9363,
      "step": 840
    },
    {
      "epoch": 0.7795202952029521,
      "grad_norm": 1.3139133572420352,
      "learning_rate": 2.8215405208360237e-06,
      "loss": 0.9399,
      "step": 845
    },
    {
      "epoch": 0.7841328413284133,
      "grad_norm": 1.325489126281033,
      "learning_rate": 2.7103137257858867e-06,
      "loss": 0.9359,
      "step": 850
    },
    {
      "epoch": 0.7887453874538746,
      "grad_norm": 1.3557387383978536,
      "learning_rate": 2.600978969225558e-06,
      "loss": 0.9408,
      "step": 855
    },
    {
      "epoch": 0.7933579335793358,
      "grad_norm": 1.2968402542167223,
      "learning_rate": 2.493564628997369e-06,
      "loss": 0.9473,
      "step": 860
    },
    {
      "epoch": 0.7979704797047971,
      "grad_norm": 1.307351413531213,
      "learning_rate": 2.3880985844994674e-06,
      "loss": 0.9288,
      "step": 865
    },
    {
      "epoch": 0.8025830258302583,
      "grad_norm": 1.301711812307168,
      "learning_rate": 2.284608209449746e-06,
      "loss": 0.9367,
      "step": 870
    },
    {
      "epoch": 0.8071955719557196,
      "grad_norm": 1.3331160835163591,
      "learning_rate": 2.183120364780975e-06,
      "loss": 0.9394,
      "step": 875
    },
    {
      "epoch": 0.8118081180811808,
      "grad_norm": 1.2683359981117004,
      "learning_rate": 2.083661391669043e-06,
      "loss": 0.9255,
      "step": 880
    },
    {
      "epoch": 0.816420664206642,
      "grad_norm": 1.2993174833270567,
      "learning_rate": 1.986257104696121e-06,
      "loss": 0.9368,
      "step": 885
    },
    {
      "epoch": 0.8210332103321033,
      "grad_norm": 1.3745325636574788,
      "learning_rate": 1.8909327851504633e-06,
      "loss": 0.9527,
      "step": 890
    },
    {
      "epoch": 0.8256457564575646,
      "grad_norm": 1.339181816674666,
      "learning_rate": 1.7977131744646724e-06,
      "loss": 0.9331,
      "step": 895
    },
    {
      "epoch": 0.8302583025830258,
      "grad_norm": 1.327094422276523,
      "learning_rate": 1.7066224677940313e-06,
      "loss": 0.9437,
      "step": 900
    },
    {
      "epoch": 0.834870848708487,
      "grad_norm": 1.2652133382811892,
      "learning_rate": 1.6176843077366755e-06,
      "loss": 0.9139,
      "step": 905
    },
    {
      "epoch": 0.8394833948339483,
      "grad_norm": 1.2892238897012769,
      "learning_rate": 1.5309217781971419e-06,
      "loss": 0.9135,
      "step": 910
    },
    {
      "epoch": 0.8440959409594095,
      "grad_norm": 1.3102845800852012,
      "learning_rate": 1.446357398394934e-06,
      "loss": 0.923,
      "step": 915
    },
    {
      "epoch": 0.8487084870848709,
      "grad_norm": 1.3117325038831478,
      "learning_rate": 1.3640131170196758e-06,
      "loss": 0.9159,
      "step": 920
    },
    {
      "epoch": 0.8533210332103321,
      "grad_norm": 5.1971689200456375,
      "learning_rate": 1.2839103065343084e-06,
      "loss": 0.9295,
      "step": 925
    },
    {
      "epoch": 0.8579335793357934,
      "grad_norm": 1.314904030690533,
      "learning_rate": 1.2060697576278812e-06,
      "loss": 0.9336,
      "step": 930
    },
    {
      "epoch": 0.8625461254612546,
      "grad_norm": 1.2791863891720787,
      "learning_rate": 1.1305116738193211e-06,
      "loss": 0.9346,
      "step": 935
    },
    {
      "epoch": 0.8671586715867159,
      "grad_norm": 1.2890531276807027,
      "learning_rate": 1.0572556662136036e-06,
      "loss": 0.9302,
      "step": 940
    },
    {
      "epoch": 0.8717712177121771,
      "grad_norm": 1.2742267464237356,
      "learning_rate": 9.863207484116987e-07,
      "loss": 0.9181,
      "step": 945
    },
    {
      "epoch": 0.8763837638376384,
      "grad_norm": 1.2644327659379573,
      "learning_rate": 9.177253315755796e-07,
      "loss": 0.9379,
      "step": 950
    },
    {
      "epoch": 0.8809963099630996,
      "grad_norm": 1.2421396307830366,
      "learning_rate": 8.514872196496182e-07,
      "loss": 0.935,
      "step": 955
    },
    {
      "epoch": 0.8856088560885609,
      "grad_norm": 1.2203962112131188,
      "learning_rate": 7.876236047395525e-07,
      "loss": 0.9264,
      "step": 960
    },
    {
      "epoch": 0.8902214022140221,
      "grad_norm": 1.2994699310922848,
      "learning_rate": 7.26151062650291e-07,
      "loss": 0.934,
      "step": 965
    },
    {
      "epoch": 0.8948339483394834,
      "grad_norm": 1.3062465605195615,
      "learning_rate": 6.670855485836525e-07,
      "loss": 0.9285,
      "step": 970
    },
    {
      "epoch": 0.8994464944649446,
      "grad_norm": 1.2690114831578048,
      "learning_rate": 6.104423929971948e-07,
      "loss": 0.933,
      "step": 975
    },
    {
      "epoch": 0.9040590405904059,
      "grad_norm": 1.2785142678323616,
      "learning_rate": 5.562362976251901e-07,
      "loss": 0.9379,
      "step": 980
    },
    {
      "epoch": 0.9086715867158671,
      "grad_norm": 1.2131026491304173,
      "learning_rate": 5.044813316627994e-07,
      "loss": 0.9276,
      "step": 985
    },
    {
      "epoch": 0.9132841328413284,
      "grad_norm": 1.300615964359426,
      "learning_rate": 4.5519092811439627e-07,
      "loss": 0.9255,
      "step": 990
    },
    {
      "epoch": 0.9178966789667896,
      "grad_norm": 1.3050095257612062,
      "learning_rate": 4.083778803070504e-07,
      "loss": 0.9437,
      "step": 995
    },
    {
      "epoch": 0.922509225092251,
      "grad_norm": 1.2410745191815336,
      "learning_rate": 3.6405433856999684e-07,
      "loss": 0.957,
      "step": 1000
    },
    {
      "epoch": 0.9271217712177122,
      "grad_norm": 1.3063199127205747,
      "learning_rate": 3.2223180708102933e-07,
      "loss": 0.9324,
      "step": 1005
    },
    {
      "epoch": 0.9317343173431735,
      "grad_norm": 1.3222002154954164,
      "learning_rate": 2.829211408805932e-07,
      "loss": 0.9176,
      "step": 1010
    },
    {
      "epoch": 0.9363468634686347,
      "grad_norm": 1.2578620181074625,
      "learning_rate": 2.461325430543482e-07,
      "loss": 0.9322,
      "step": 1015
    },
    {
      "epoch": 0.940959409594096,
      "grad_norm": 1.2337296124585635,
      "learning_rate": 2.1187556208496885e-07,
      "loss": 0.9218,
      "step": 1020
    },
    {
      "epoch": 0.9455719557195572,
      "grad_norm": 1.2291744208349304,
      "learning_rate": 1.8015908937382587e-07,
      "loss": 0.9331,
      "step": 1025
    },
    {
      "epoch": 0.9501845018450185,
      "grad_norm": 1.2819268220912796,
      "learning_rate": 1.5099135693322776e-07,
      "loss": 0.9337,
      "step": 1030
    },
    {
      "epoch": 0.9547970479704797,
      "grad_norm": 1.2311873702844616,
      "learning_rate": 1.2437993524979984e-07,
      "loss": 0.9034,
      "step": 1035
    },
    {
      "epoch": 0.959409594095941,
      "grad_norm": 1.306307352211591,
      "learning_rate": 1.0033173131956175e-07,
      "loss": 0.9213,
      "step": 1040
    },
    {
      "epoch": 0.9640221402214022,
      "grad_norm": 1.2302162082382204,
      "learning_rate": 7.885298685522235e-08,
      "loss": 0.9153,
      "step": 1045
    },
    {
      "epoch": 0.9686346863468634,
      "grad_norm": 1.2949224700302513,
      "learning_rate": 5.99492766661347e-08,
      "loss": 0.9207,
      "step": 1050
    },
    {
      "epoch": 0.9732472324723247,
      "grad_norm": 1.2979324069801879,
      "learning_rate": 4.362550721136338e-08,
      "loss": 0.9254,
      "step": 1055
    },
    {
      "epoch": 0.977859778597786,
      "grad_norm": 1.2464202848956238,
      "learning_rate": 2.988591532620322e-08,
      "loss": 0.9155,
      "step": 1060
    },
    {
      "epoch": 0.9824723247232472,
      "grad_norm": 1.2826213715670498,
      "learning_rate": 1.8734067122514464e-08,
      "loss": 0.9229,
      "step": 1065
    },
    {
      "epoch": 0.9870848708487084,
      "grad_norm": 1.2838870564000777,
      "learning_rate": 1.0172857063137643e-08,
      "loss": 0.9254,
      "step": 1070
    },
    {
      "epoch": 0.9916974169741697,
      "grad_norm": 1.2809568528218904,
      "learning_rate": 4.204507210633368e-09,
      "loss": 0.9466,
      "step": 1075
    },
    {
      "epoch": 0.996309963099631,
      "grad_norm": 1.2838585438818646,
      "learning_rate": 8.30566650548148e-10,
      "loss": 0.9403,
      "step": 1080
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.9415282607078552,
      "eval_runtime": 141.608,
      "eval_samples_per_second": 108.398,
      "eval_steps_per_second": 1.695,
      "step": 1084
    },
    {
      "epoch": 1.0,
      "step": 1084,
      "total_flos": 453935093514240.0,
      "train_loss": 0.9830407258329357,
      "train_runtime": 11998.8497,
      "train_samples_per_second": 11.558,
      "train_steps_per_second": 0.09
    }
  ],
  "logging_steps": 5,
  "max_steps": 1084,
  "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": 453935093514240.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}