{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.999793431109275,
  "eval_steps": 500,
  "global_step": 1210,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0008262755629002272,
      "grad_norm": 23.554044233313164,
      "learning_rate": 8.264462809917357e-08,
      "loss": 1.4094,
      "step": 1
    },
    {
      "epoch": 0.004131377814501136,
      "grad_norm": 23.487881173685214,
      "learning_rate": 4.132231404958678e-07,
      "loss": 1.44,
      "step": 5
    },
    {
      "epoch": 0.008262755629002272,
      "grad_norm": 8.324804089201631,
      "learning_rate": 8.264462809917356e-07,
      "loss": 1.3352,
      "step": 10
    },
    {
      "epoch": 0.01239413344350341,
      "grad_norm": 10.482872108095675,
      "learning_rate": 1.2396694214876035e-06,
      "loss": 1.1699,
      "step": 15
    },
    {
      "epoch": 0.016525511258004544,
      "grad_norm": 3.0017247032336667,
      "learning_rate": 1.6528925619834712e-06,
      "loss": 1.0235,
      "step": 20
    },
    {
      "epoch": 0.02065688907250568,
      "grad_norm": 2.760976256425166,
      "learning_rate": 2.066115702479339e-06,
      "loss": 0.9675,
      "step": 25
    },
    {
      "epoch": 0.02478826688700682,
      "grad_norm": 2.376505582612684,
      "learning_rate": 2.479338842975207e-06,
      "loss": 0.9406,
      "step": 30
    },
    {
      "epoch": 0.028919644701507954,
      "grad_norm": 2.2154227436349765,
      "learning_rate": 2.8925619834710743e-06,
      "loss": 0.9302,
      "step": 35
    },
    {
      "epoch": 0.03305102251600909,
      "grad_norm": 2.3739150674432445,
      "learning_rate": 3.3057851239669424e-06,
      "loss": 0.8994,
      "step": 40
    },
    {
      "epoch": 0.03718240033051023,
      "grad_norm": 2.2109809527940554,
      "learning_rate": 3.71900826446281e-06,
      "loss": 0.8982,
      "step": 45
    },
    {
      "epoch": 0.04131377814501136,
      "grad_norm": 2.28822301545997,
      "learning_rate": 4.132231404958678e-06,
      "loss": 0.8837,
      "step": 50
    },
    {
      "epoch": 0.0454451559595125,
      "grad_norm": 2.2559023546514343,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 0.8754,
      "step": 55
    },
    {
      "epoch": 0.04957653377401364,
      "grad_norm": 2.2137771037984684,
      "learning_rate": 4.958677685950414e-06,
      "loss": 0.8582,
      "step": 60
    },
    {
      "epoch": 0.05370791158851477,
      "grad_norm": 2.373934998656087,
      "learning_rate": 5.371900826446281e-06,
      "loss": 0.8593,
      "step": 65
    },
    {
      "epoch": 0.05783928940301591,
      "grad_norm": 2.467630003052232,
      "learning_rate": 5.785123966942149e-06,
      "loss": 0.856,
      "step": 70
    },
    {
      "epoch": 0.06197066721751704,
      "grad_norm": 2.424099124838778,
      "learning_rate": 6.198347107438017e-06,
      "loss": 0.853,
      "step": 75
    },
    {
      "epoch": 0.06610204503201818,
      "grad_norm": 2.362990499630114,
      "learning_rate": 6.611570247933885e-06,
      "loss": 0.8383,
      "step": 80
    },
    {
      "epoch": 0.07023342284651932,
      "grad_norm": 2.399473277862941,
      "learning_rate": 7.0247933884297525e-06,
      "loss": 0.8318,
      "step": 85
    },
    {
      "epoch": 0.07436480066102046,
      "grad_norm": 2.5700793481674755,
      "learning_rate": 7.43801652892562e-06,
      "loss": 0.8303,
      "step": 90
    },
    {
      "epoch": 0.07849617847552158,
      "grad_norm": 2.3911560791916466,
      "learning_rate": 7.851239669421489e-06,
      "loss": 0.8283,
      "step": 95
    },
    {
      "epoch": 0.08262755629002272,
      "grad_norm": 2.5970628827400057,
      "learning_rate": 8.264462809917356e-06,
      "loss": 0.8249,
      "step": 100
    },
    {
      "epoch": 0.08675893410452386,
      "grad_norm": 2.3382293960099143,
      "learning_rate": 8.677685950413224e-06,
      "loss": 0.8186,
      "step": 105
    },
    {
      "epoch": 0.090890311919025,
      "grad_norm": 2.423365953954326,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.8202,
      "step": 110
    },
    {
      "epoch": 0.09502168973352614,
      "grad_norm": 2.5282725418752694,
      "learning_rate": 9.50413223140496e-06,
      "loss": 0.8122,
      "step": 115
    },
    {
      "epoch": 0.09915306754802727,
      "grad_norm": 2.3639158309512522,
      "learning_rate": 9.917355371900828e-06,
      "loss": 0.8127,
      "step": 120
    },
    {
      "epoch": 0.1032844453625284,
      "grad_norm": 2.844448120421469,
      "learning_rate": 9.999667111219573e-06,
      "loss": 0.7962,
      "step": 125
    },
    {
      "epoch": 0.10741582317702954,
      "grad_norm": 3.443216907128371,
      "learning_rate": 9.998314826517564e-06,
      "loss": 0.8031,
      "step": 130
    },
    {
      "epoch": 0.11154720099153068,
      "grad_norm": 2.681450835682187,
      "learning_rate": 9.995922621477252e-06,
      "loss": 0.805,
      "step": 135
    },
    {
      "epoch": 0.11567857880603181,
      "grad_norm": 2.41859239043235,
      "learning_rate": 9.99249099380692e-06,
      "loss": 0.7891,
      "step": 140
    },
    {
      "epoch": 0.11980995662053295,
      "grad_norm": 2.2320583844472375,
      "learning_rate": 9.988020657471078e-06,
      "loss": 0.7945,
      "step": 145
    },
    {
      "epoch": 0.12394133443503408,
      "grad_norm": 2.6404700133557064,
      "learning_rate": 9.98251254254193e-06,
      "loss": 0.7789,
      "step": 150
    },
    {
      "epoch": 0.12807271224953523,
      "grad_norm": 2.4700012252546855,
      "learning_rate": 9.97596779500586e-06,
      "loss": 0.7776,
      "step": 155
    },
    {
      "epoch": 0.13220409006403636,
      "grad_norm": 2.534567680863635,
      "learning_rate": 9.968387776525009e-06,
      "loss": 0.7665,
      "step": 160
    },
    {
      "epoch": 0.13633546787853748,
      "grad_norm": 2.3195785880081274,
      "learning_rate": 9.959774064153977e-06,
      "loss": 0.7633,
      "step": 165
    },
    {
      "epoch": 0.14046684569303863,
      "grad_norm": 2.6303462252799124,
      "learning_rate": 9.950128450011706e-06,
      "loss": 0.7587,
      "step": 170
    },
    {
      "epoch": 0.14459822350753976,
      "grad_norm": 2.4452240563950687,
      "learning_rate": 9.939452940908627e-06,
      "loss": 0.7627,
      "step": 175
    },
    {
      "epoch": 0.1487296013220409,
      "grad_norm": 2.3656117475844294,
      "learning_rate": 9.927749757929125e-06,
      "loss": 0.7546,
      "step": 180
    },
    {
      "epoch": 0.15286097913654204,
      "grad_norm": 2.307155404533025,
      "learning_rate": 9.915021335969452e-06,
      "loss": 0.744,
      "step": 185
    },
    {
      "epoch": 0.15699235695104316,
      "grad_norm": 2.3398594948394127,
      "learning_rate": 9.901270323231114e-06,
      "loss": 0.7423,
      "step": 190
    },
    {
      "epoch": 0.1611237347655443,
      "grad_norm": 2.3156292065612214,
      "learning_rate": 9.886499580669917e-06,
      "loss": 0.7396,
      "step": 195
    },
    {
      "epoch": 0.16525511258004544,
      "grad_norm": 2.574743453191487,
      "learning_rate": 9.870712181400726e-06,
      "loss": 0.7335,
      "step": 200
    },
    {
      "epoch": 0.1693864903945466,
      "grad_norm": 2.339577898310028,
      "learning_rate": 9.853911410058097e-06,
      "loss": 0.7398,
      "step": 205
    },
    {
      "epoch": 0.17351786820904772,
      "grad_norm": 2.489900169912843,
      "learning_rate": 9.836100762112887e-06,
      "loss": 0.7423,
      "step": 210
    },
    {
      "epoch": 0.17764924602354884,
      "grad_norm": 2.5114661447059365,
      "learning_rate": 9.817283943145014e-06,
      "loss": 0.7273,
      "step": 215
    },
    {
      "epoch": 0.18178062383805,
      "grad_norm": 2.4765147278897097,
      "learning_rate": 9.797464868072489e-06,
      "loss": 0.7171,
      "step": 220
    },
    {
      "epoch": 0.18591200165255112,
      "grad_norm": 2.35952144225102,
      "learning_rate": 9.776647660336905e-06,
      "loss": 0.7165,
      "step": 225
    },
    {
      "epoch": 0.19004337946705227,
      "grad_norm": 2.3928640553968115,
      "learning_rate": 9.754836651045538e-06,
      "loss": 0.7228,
      "step": 230
    },
    {
      "epoch": 0.1941747572815534,
      "grad_norm": 2.359655168950183,
      "learning_rate": 9.732036378070243e-06,
      "loss": 0.6999,
      "step": 235
    },
    {
      "epoch": 0.19830613509605455,
      "grad_norm": 2.7166804592398854,
      "learning_rate": 9.708251585103324e-06,
      "loss": 0.6924,
      "step": 240
    },
    {
      "epoch": 0.20243751291055567,
      "grad_norm": 2.6360301220005415,
      "learning_rate": 9.683487220670595e-06,
      "loss": 0.7038,
      "step": 245
    },
    {
      "epoch": 0.2065688907250568,
      "grad_norm": 2.358814454829314,
      "learning_rate": 9.657748437101819e-06,
      "loss": 0.7076,
      "step": 250
    },
    {
      "epoch": 0.21070026853955795,
      "grad_norm": 2.250240362766412,
      "learning_rate": 9.631040589458742e-06,
      "loss": 0.6833,
      "step": 255
    },
    {
      "epoch": 0.21483164635405907,
      "grad_norm": 2.4709639248779083,
      "learning_rate": 9.603369234420944e-06,
      "loss": 0.6751,
      "step": 260
    },
    {
      "epoch": 0.21896302416856023,
      "grad_norm": 2.315474134768072,
      "learning_rate": 9.574740129129767e-06,
      "loss": 0.6988,
      "step": 265
    },
    {
      "epoch": 0.22309440198306135,
      "grad_norm": 2.2090103807004975,
      "learning_rate": 9.545159229990493e-06,
      "loss": 0.6854,
      "step": 270
    },
    {
      "epoch": 0.22722577979756248,
      "grad_norm": 2.366702578463947,
      "learning_rate": 9.514632691433108e-06,
      "loss": 0.6657,
      "step": 275
    },
    {
      "epoch": 0.23135715761206363,
      "grad_norm": 2.298899028385364,
      "learning_rate": 9.483166864631837e-06,
      "loss": 0.663,
      "step": 280
    },
    {
      "epoch": 0.23548853542656475,
      "grad_norm": 2.2714246758736394,
      "learning_rate": 9.450768296183764e-06,
      "loss": 0.6655,
      "step": 285
    },
    {
      "epoch": 0.2396199132410659,
      "grad_norm": 2.1782463852503438,
      "learning_rate": 9.417443726746776e-06,
      "loss": 0.6678,
      "step": 290
    },
    {
      "epoch": 0.24375129105556703,
      "grad_norm": 2.206125943355958,
      "learning_rate": 9.383200089637143e-06,
      "loss": 0.6579,
      "step": 295
    },
    {
      "epoch": 0.24788266887006816,
      "grad_norm": 2.4822099384962133,
      "learning_rate": 9.348044509387022e-06,
      "loss": 0.6688,
      "step": 300
    },
    {
      "epoch": 0.2520140466845693,
      "grad_norm": 2.6347462586981827,
      "learning_rate": 9.311984300262151e-06,
      "loss": 0.6723,
      "step": 305
    },
    {
      "epoch": 0.25614542449907046,
      "grad_norm": 2.3969886300206915,
      "learning_rate": 9.275026964740101e-06,
      "loss": 0.6495,
      "step": 310
    },
    {
      "epoch": 0.2602768023135716,
      "grad_norm": 2.2255060581251733,
      "learning_rate": 9.237180191949347e-06,
      "loss": 0.6548,
      "step": 315
    },
    {
      "epoch": 0.2644081801280727,
      "grad_norm": 2.4470079917886585,
      "learning_rate": 9.198451856069514e-06,
      "loss": 0.6402,
      "step": 320
    },
    {
      "epoch": 0.26853955794257384,
      "grad_norm": 2.232598978818489,
      "learning_rate": 9.158850014693123e-06,
      "loss": 0.6428,
      "step": 325
    },
    {
      "epoch": 0.27267093575707496,
      "grad_norm": 2.2780042831807807,
      "learning_rate": 9.118382907149164e-06,
      "loss": 0.6449,
      "step": 330
    },
    {
      "epoch": 0.27680231357157614,
      "grad_norm": 2.250070488730356,
      "learning_rate": 9.077058952788888e-06,
      "loss": 0.6248,
      "step": 335
    },
    {
      "epoch": 0.28093369138607727,
      "grad_norm": 2.257817990210915,
      "learning_rate": 9.034886749234112e-06,
      "loss": 0.6287,
      "step": 340
    },
    {
      "epoch": 0.2850650692005784,
      "grad_norm": 2.2753575010897373,
      "learning_rate": 8.991875070588449e-06,
      "loss": 0.631,
      "step": 345
    },
    {
      "epoch": 0.2891964470150795,
      "grad_norm": 2.5652909583991534,
      "learning_rate": 8.948032865611823e-06,
      "loss": 0.6232,
      "step": 350
    },
    {
      "epoch": 0.29332782482958064,
      "grad_norm": 2.3365773173303843,
      "learning_rate": 8.90336925585864e-06,
      "loss": 0.608,
      "step": 355
    },
    {
      "epoch": 0.2974592026440818,
      "grad_norm": 2.27141916228481,
      "learning_rate": 8.857893533780015e-06,
      "loss": 0.6048,
      "step": 360
    },
    {
      "epoch": 0.30159058045858295,
      "grad_norm": 2.121899092032811,
      "learning_rate": 8.811615160790428e-06,
      "loss": 0.6192,
      "step": 365
    },
    {
      "epoch": 0.30572195827308407,
      "grad_norm": 2.210463830947181,
      "learning_rate": 8.764543765299245e-06,
      "loss": 0.611,
      "step": 370
    },
    {
      "epoch": 0.3098533360875852,
      "grad_norm": 2.2581243380183835,
      "learning_rate": 8.716689140707488e-06,
      "loss": 0.6116,
      "step": 375
    },
    {
      "epoch": 0.3139847139020863,
      "grad_norm": 2.291459985719934,
      "learning_rate": 8.668061243370273e-06,
      "loss": 0.6094,
      "step": 380
    },
    {
      "epoch": 0.3181160917165875,
      "grad_norm": 2.242064952311342,
      "learning_rate": 8.61867019052535e-06,
      "loss": 0.6076,
      "step": 385
    },
    {
      "epoch": 0.3222474695310886,
      "grad_norm": 2.3770181909891304,
      "learning_rate": 8.568526258188172e-06,
      "loss": 0.6047,
      "step": 390
    },
    {
      "epoch": 0.32637884734558975,
      "grad_norm": 2.4304957747194593,
      "learning_rate": 8.517639879013918e-06,
      "loss": 0.6037,
      "step": 395
    },
    {
      "epoch": 0.3305102251600909,
      "grad_norm": 2.159929931604691,
      "learning_rate": 8.466021640126946e-06,
      "loss": 0.5924,
      "step": 400
    },
    {
      "epoch": 0.334641602974592,
      "grad_norm": 2.181588993237923,
      "learning_rate": 8.413682280918094e-06,
      "loss": 0.5859,
      "step": 405
    },
    {
      "epoch": 0.3387729807890932,
      "grad_norm": 2.239848195768108,
      "learning_rate": 8.3606326908103e-06,
      "loss": 0.5808,
      "step": 410
    },
    {
      "epoch": 0.3429043586035943,
      "grad_norm": 2.2395305817254165,
      "learning_rate": 8.306883906993022e-06,
      "loss": 0.5805,
      "step": 415
    },
    {
      "epoch": 0.34703573641809543,
      "grad_norm": 2.276771540411031,
      "learning_rate": 8.25244711212589e-06,
      "loss": 0.5771,
      "step": 420
    },
    {
      "epoch": 0.35116711423259656,
      "grad_norm": 2.2443733616231674,
      "learning_rate": 8.197333632012123e-06,
      "loss": 0.5724,
      "step": 425
    },
    {
      "epoch": 0.3552984920470977,
      "grad_norm": 2.266086543452868,
      "learning_rate": 8.141554933242135e-06,
      "loss": 0.5766,
      "step": 430
    },
    {
      "epoch": 0.35942986986159886,
      "grad_norm": 2.3440174185212426,
      "learning_rate": 8.08512262080787e-06,
      "loss": 0.5704,
      "step": 435
    },
    {
      "epoch": 0.3635612476761,
      "grad_norm": 2.231383713024198,
      "learning_rate": 8.028048435688333e-06,
      "loss": 0.5672,
      "step": 440
    },
    {
      "epoch": 0.3676926254906011,
      "grad_norm": 2.210345864732212,
      "learning_rate": 7.970344252406832e-06,
      "loss": 0.5654,
      "step": 445
    },
    {
      "epoch": 0.37182400330510224,
      "grad_norm": 2.201487752141703,
      "learning_rate": 7.912022076560426e-06,
      "loss": 0.5608,
      "step": 450
    },
    {
      "epoch": 0.37595538111960336,
      "grad_norm": 2.2126765922615017,
      "learning_rate": 7.853094042322121e-06,
      "loss": 0.5676,
      "step": 455
    },
    {
      "epoch": 0.38008675893410454,
      "grad_norm": 2.368648057415258,
      "learning_rate": 7.7935724099163e-06,
      "loss": 0.5611,
      "step": 460
    },
    {
      "epoch": 0.38421813674860567,
      "grad_norm": 2.249897659892324,
      "learning_rate": 7.733469563067928e-06,
      "loss": 0.5632,
      "step": 465
    },
    {
      "epoch": 0.3883495145631068,
      "grad_norm": 2.257196050360634,
      "learning_rate": 7.67279800642607e-06,
      "loss": 0.5414,
      "step": 470
    },
    {
      "epoch": 0.3924808923776079,
      "grad_norm": 2.1223073675520125,
      "learning_rate": 7.611570362962247e-06,
      "loss": 0.5421,
      "step": 475
    },
    {
      "epoch": 0.3966122701921091,
      "grad_norm": 2.209118984554094,
      "learning_rate": 7.549799371344175e-06,
      "loss": 0.5475,
      "step": 480
    },
    {
      "epoch": 0.4007436480066102,
      "grad_norm": 2.070603624522537,
      "learning_rate": 7.487497883285428e-06,
      "loss": 0.5477,
      "step": 485
    },
    {
      "epoch": 0.40487502582111135,
      "grad_norm": 2.2718749779596097,
      "learning_rate": 7.424678860871584e-06,
      "loss": 0.5384,
      "step": 490
    },
    {
      "epoch": 0.40900640363561247,
      "grad_norm": 2.347289860613089,
      "learning_rate": 7.361355373863415e-06,
      "loss": 0.5407,
      "step": 495
    },
    {
      "epoch": 0.4131377814501136,
      "grad_norm": 2.491186132317311,
      "learning_rate": 7.297540596977663e-06,
      "loss": 0.5393,
      "step": 500
    },
    {
      "epoch": 0.4172691592646148,
      "grad_norm": 2.3320063032579177,
      "learning_rate": 7.233247807145989e-06,
      "loss": 0.5311,
      "step": 505
    },
    {
      "epoch": 0.4214005370791159,
      "grad_norm": 2.122601227745636,
      "learning_rate": 7.168490380752648e-06,
      "loss": 0.537,
      "step": 510
    },
    {
      "epoch": 0.425531914893617,
      "grad_norm": 2.2214499757842865,
      "learning_rate": 7.103281790851483e-06,
      "loss": 0.5338,
      "step": 515
    },
    {
      "epoch": 0.42966329270811815,
      "grad_norm": 2.1654002355826494,
      "learning_rate": 7.037635604362786e-06,
      "loss": 0.5188,
      "step": 520
    },
    {
      "epoch": 0.4337946705226193,
      "grad_norm": 2.1599888645905847,
      "learning_rate": 6.971565479250659e-06,
      "loss": 0.5321,
      "step": 525
    },
    {
      "epoch": 0.43792604833712045,
      "grad_norm": 2.4521064066537,
      "learning_rate": 6.905085161681408e-06,
      "loss": 0.5137,
      "step": 530
    },
    {
      "epoch": 0.4420574261516216,
      "grad_norm": 2.0735269808191563,
      "learning_rate": 6.838208483163601e-06,
      "loss": 0.5181,
      "step": 535
    },
    {
      "epoch": 0.4461888039661227,
      "grad_norm": 2.1964532327451307,
      "learning_rate": 6.770949357670358e-06,
      "loss": 0.5079,
      "step": 540
    },
    {
      "epoch": 0.45032018178062383,
      "grad_norm": 2.1721038727423996,
      "learning_rate": 6.703321778744495e-06,
      "loss": 0.5169,
      "step": 545
    },
    {
      "epoch": 0.45445155959512495,
      "grad_norm": 2.083001885275941,
      "learning_rate": 6.635339816587109e-06,
      "loss": 0.5189,
      "step": 550
    },
    {
      "epoch": 0.45858293740962613,
      "grad_norm": 2.2406805358253523,
      "learning_rate": 6.5670176151302136e-06,
      "loss": 0.5056,
      "step": 555
    },
    {
      "epoch": 0.46271431522412726,
      "grad_norm": 2.362515256614418,
      "learning_rate": 6.4983693890940335e-06,
      "loss": 0.5068,
      "step": 560
    },
    {
      "epoch": 0.4668456930386284,
      "grad_norm": 2.073329405987291,
      "learning_rate": 6.4294094210295725e-06,
      "loss": 0.5146,
      "step": 565
    },
    {
      "epoch": 0.4709770708531295,
      "grad_norm": 2.0802359451245276,
      "learning_rate": 6.360152058347068e-06,
      "loss": 0.4993,
      "step": 570
    },
    {
      "epoch": 0.47510844866763063,
      "grad_norm": 2.205382348988315,
      "learning_rate": 6.290611710330957e-06,
      "loss": 0.5056,
      "step": 575
    },
    {
      "epoch": 0.4792398264821318,
      "grad_norm": 2.106317834822583,
      "learning_rate": 6.2208028451419575e-06,
      "loss": 0.4939,
      "step": 580
    },
    {
      "epoch": 0.48337120429663294,
      "grad_norm": 2.0997265512356953,
      "learning_rate": 6.150739986806911e-06,
      "loss": 0.4904,
      "step": 585
    },
    {
      "epoch": 0.48750258211113406,
      "grad_norm": 2.087397537314153,
      "learning_rate": 6.0804377121969985e-06,
      "loss": 0.4903,
      "step": 590
    },
    {
      "epoch": 0.4916339599256352,
      "grad_norm": 2.1772365057315834,
      "learning_rate": 6.009910647994956e-06,
      "loss": 0.5043,
      "step": 595
    },
    {
      "epoch": 0.4957653377401363,
      "grad_norm": 2.0357765671221872,
      "learning_rate": 5.939173467651942e-06,
      "loss": 0.4916,
      "step": 600
    },
    {
      "epoch": 0.4998967155546375,
      "grad_norm": 2.140953972933528,
      "learning_rate": 5.8682408883346535e-06,
      "loss": 0.4881,
      "step": 605
    },
    {
      "epoch": 0.5040280933691386,
      "grad_norm": 2.133259953600996,
      "learning_rate": 5.7971276678633625e-06,
      "loss": 0.4925,
      "step": 610
    },
    {
      "epoch": 0.5081594711836398,
      "grad_norm": 2.0897840178036753,
      "learning_rate": 5.725848601641492e-06,
      "loss": 0.4909,
      "step": 615
    },
    {
      "epoch": 0.5122908489981409,
      "grad_norm": 2.1033344267926553,
      "learning_rate": 5.654418519577369e-06,
      "loss": 0.4808,
      "step": 620
    },
    {
      "epoch": 0.516422226812642,
      "grad_norm": 2.0931483592499927,
      "learning_rate": 5.5828522829987965e-06,
      "loss": 0.4697,
      "step": 625
    },
    {
      "epoch": 0.5205536046271432,
      "grad_norm": 2.0194919084584444,
      "learning_rate": 5.511164781561096e-06,
      "loss": 0.479,
      "step": 630
    },
    {
      "epoch": 0.5246849824416443,
      "grad_norm": 2.070247481858063,
      "learning_rate": 5.439370930149252e-06,
      "loss": 0.4705,
      "step": 635
    },
    {
      "epoch": 0.5288163602561454,
      "grad_norm": 2.060542153916691,
      "learning_rate": 5.367485665774802e-06,
      "loss": 0.4689,
      "step": 640
    },
    {
      "epoch": 0.5329477380706465,
      "grad_norm": 2.171769190139748,
      "learning_rate": 5.295523944468137e-06,
      "loss": 0.4678,
      "step": 645
    },
    {
      "epoch": 0.5370791158851477,
      "grad_norm": 2.437387629266701,
      "learning_rate": 5.223500738166837e-06,
      "loss": 0.4723,
      "step": 650
    },
    {
      "epoch": 0.5412104936996488,
      "grad_norm": 2.139213174417897,
      "learning_rate": 5.1514310316006835e-06,
      "loss": 0.472,
      "step": 655
    },
    {
      "epoch": 0.5453418715141499,
      "grad_norm": 2.083008247659044,
      "learning_rate": 5.07932981917404e-06,
      "loss": 0.4684,
      "step": 660
    },
    {
      "epoch": 0.5494732493286512,
      "grad_norm": 2.161242770803335,
      "learning_rate": 5.007212101846194e-06,
      "loss": 0.4587,
      "step": 665
    },
    {
      "epoch": 0.5536046271431523,
      "grad_norm": 2.0500783532272604,
      "learning_rate": 4.935092884010347e-06,
      "loss": 0.4611,
      "step": 670
    },
    {
      "epoch": 0.5577360049576534,
      "grad_norm": 2.1439435869325116,
      "learning_rate": 4.8629871703718844e-06,
      "loss": 0.4622,
      "step": 675
    },
    {
      "epoch": 0.5618673827721545,
      "grad_norm": 2.066743306657874,
      "learning_rate": 4.7909099628265946e-06,
      "loss": 0.4672,
      "step": 680
    },
    {
      "epoch": 0.5659987605866557,
      "grad_norm": 2.0461947863660574,
      "learning_rate": 4.718876257339444e-06,
      "loss": 0.4512,
      "step": 685
    },
    {
      "epoch": 0.5701301384011568,
      "grad_norm": 2.13397891500544,
      "learning_rate": 4.646901040824622e-06,
      "loss": 0.4568,
      "step": 690
    },
    {
      "epoch": 0.5742615162156579,
      "grad_norm": 2.038909744378958,
      "learning_rate": 4.574999288027437e-06,
      "loss": 0.4505,
      "step": 695
    },
    {
      "epoch": 0.578392894030159,
      "grad_norm": 2.061119461177199,
      "learning_rate": 4.503185958408767e-06,
      "loss": 0.4536,
      "step": 700
    },
    {
      "epoch": 0.5825242718446602,
      "grad_norm": 2.147709730965502,
      "learning_rate": 4.431475993032673e-06,
      "loss": 0.4567,
      "step": 705
    },
    {
      "epoch": 0.5866556496591613,
      "grad_norm": 2.0471799173895078,
      "learning_rate": 4.359884311457857e-06,
      "loss": 0.452,
      "step": 710
    },
    {
      "epoch": 0.5907870274736625,
      "grad_norm": 2.096762090340259,
      "learning_rate": 4.2884258086335755e-06,
      "loss": 0.4407,
      "step": 715
    },
    {
      "epoch": 0.5949184052881636,
      "grad_norm": 2.1776523990414436,
      "learning_rate": 4.217115351800693e-06,
      "loss": 0.4364,
      "step": 720
    },
    {
      "epoch": 0.5990497831026648,
      "grad_norm": 2.148545496231866,
      "learning_rate": 4.145967777398481e-06,
      "loss": 0.4407,
      "step": 725
    },
    {
      "epoch": 0.6031811609171659,
      "grad_norm": 2.0382917606214326,
      "learning_rate": 4.074997887977843e-06,
      "loss": 0.4339,
      "step": 730
    },
    {
      "epoch": 0.607312538731667,
      "grad_norm": 2.096587532898138,
      "learning_rate": 4.004220449121574e-06,
      "loss": 0.4337,
      "step": 735
    },
    {
      "epoch": 0.6114439165461681,
      "grad_norm": 2.0260734306608272,
      "learning_rate": 3.933650186372329e-06,
      "loss": 0.4297,
      "step": 740
    },
    {
      "epoch": 0.6155752943606693,
      "grad_norm": 2.0747974830236116,
      "learning_rate": 3.863301782168896e-06,
      "loss": 0.4362,
      "step": 745
    },
    {
      "epoch": 0.6197066721751704,
      "grad_norm": 2.1542278120116154,
      "learning_rate": 3.7931898727914723e-06,
      "loss": 0.4413,
      "step": 750
    },
    {
      "epoch": 0.6238380499896715,
      "grad_norm": 1.9775710411271903,
      "learning_rate": 3.7233290453165127e-06,
      "loss": 0.4273,
      "step": 755
    },
    {
      "epoch": 0.6279694278041726,
      "grad_norm": 2.034389521374345,
      "learning_rate": 3.6537338345818273e-06,
      "loss": 0.4312,
      "step": 760
    },
    {
      "epoch": 0.6321008056186739,
      "grad_norm": 2.059316170572532,
      "learning_rate": 3.5844187201625567e-06,
      "loss": 0.4301,
      "step": 765
    },
    {
      "epoch": 0.636232183433175,
      "grad_norm": 2.0806349809281235,
      "learning_rate": 3.5153981233586277e-06,
      "loss": 0.4234,
      "step": 770
    },
    {
      "epoch": 0.6403635612476761,
      "grad_norm": 2.1474742343641933,
      "learning_rate": 3.446686404194337e-06,
      "loss": 0.4228,
      "step": 775
    },
    {
      "epoch": 0.6444949390621773,
      "grad_norm": 2.141080032663376,
      "learning_rate": 3.3782978584307035e-06,
      "loss": 0.4179,
      "step": 780
    },
    {
      "epoch": 0.6486263168766784,
      "grad_norm": 2.0266388142571876,
      "learning_rate": 3.310246714591162e-06,
      "loss": 0.4197,
      "step": 785
    },
    {
      "epoch": 0.6527576946911795,
      "grad_norm": 2.1015063906736873,
      "learning_rate": 3.2425471310012645e-06,
      "loss": 0.4224,
      "step": 790
    },
    {
      "epoch": 0.6568890725056806,
      "grad_norm": 2.0688211913933126,
      "learning_rate": 3.1752131928429787e-06,
      "loss": 0.4148,
      "step": 795
    },
    {
      "epoch": 0.6610204503201818,
      "grad_norm": 2.020342446426012,
      "learning_rate": 3.1082589092242116e-06,
      "loss": 0.4209,
      "step": 800
    },
    {
      "epoch": 0.6651518281346829,
      "grad_norm": 2.050265450999902,
      "learning_rate": 3.041698210264149e-06,
      "loss": 0.4086,
      "step": 805
    },
    {
      "epoch": 0.669283205949184,
      "grad_norm": 2.0414704146915117,
      "learning_rate": 2.9755449441950434e-06,
      "loss": 0.424,
      "step": 810
    },
    {
      "epoch": 0.6734145837636852,
      "grad_norm": 2.0309738852976924,
      "learning_rate": 2.9098128744810245e-06,
      "loss": 0.4136,
      "step": 815
    },
    {
      "epoch": 0.6775459615781864,
      "grad_norm": 2.123876449787772,
      "learning_rate": 2.84451567695456e-06,
      "loss": 0.4213,
      "step": 820
    },
    {
      "epoch": 0.6816773393926875,
      "grad_norm": 2.1356187469686043,
      "learning_rate": 2.7796669369711294e-06,
      "loss": 0.4216,
      "step": 825
    },
    {
      "epoch": 0.6858087172071886,
      "grad_norm": 2.06757675819696,
      "learning_rate": 2.715280146582752e-06,
      "loss": 0.4192,
      "step": 830
    },
    {
      "epoch": 0.6899400950216897,
      "grad_norm": 2.0954350411098916,
      "learning_rate": 2.651368701730889e-06,
      "loss": 0.4077,
      "step": 835
    },
    {
      "epoch": 0.6940714728361909,
      "grad_norm": 1.9490264079119275,
      "learning_rate": 2.5879458994593786e-06,
      "loss": 0.4018,
      "step": 840
    },
    {
      "epoch": 0.698202850650692,
      "grad_norm": 1.9862703131789983,
      "learning_rate": 2.5250249351479206e-06,
      "loss": 0.4127,
      "step": 845
    },
    {
      "epoch": 0.7023342284651931,
      "grad_norm": 1.9907644914298972,
      "learning_rate": 2.4626188997667224e-06,
      "loss": 0.4057,
      "step": 850
    },
    {
      "epoch": 0.7064656062796942,
      "grad_norm": 2.0178385478283,
      "learning_rate": 2.400740777152874e-06,
      "loss": 0.405,
      "step": 855
    },
    {
      "epoch": 0.7105969840941954,
      "grad_norm": 2.0040936929718627,
      "learning_rate": 2.3394034413090015e-06,
      "loss": 0.4054,
      "step": 860
    },
    {
      "epoch": 0.7147283619086966,
      "grad_norm": 2.0032870121182516,
      "learning_rate": 2.278619653724781e-06,
      "loss": 0.4025,
      "step": 865
    },
    {
      "epoch": 0.7188597397231977,
      "grad_norm": 2.0315969629289867,
      "learning_rate": 2.218402060721845e-06,
      "loss": 0.3929,
      "step": 870
    },
    {
      "epoch": 0.7229911175376988,
      "grad_norm": 2.0727282353845626,
      "learning_rate": 2.1587631908226812e-06,
      "loss": 0.3986,
      "step": 875
    },
    {
      "epoch": 0.7271224953522,
      "grad_norm": 1.9152876057461066,
      "learning_rate": 2.09971545214401e-06,
      "loss": 0.3915,
      "step": 880
    },
    {
      "epoch": 0.7312538731667011,
      "grad_norm": 2.006764866181322,
      "learning_rate": 2.04127112981522e-06,
      "loss": 0.3972,
      "step": 885
    },
    {
      "epoch": 0.7353852509812022,
      "grad_norm": 2.009533139724546,
      "learning_rate": 1.9834423834224014e-06,
      "loss": 0.3952,
      "step": 890
    },
    {
      "epoch": 0.7395166287957033,
      "grad_norm": 1.9136982154678623,
      "learning_rate": 1.926241244478496e-06,
      "loss": 0.3894,
      "step": 895
    },
    {
      "epoch": 0.7436480066102045,
      "grad_norm": 1.9940424884511188,
      "learning_rate": 1.8696796139200852e-06,
      "loss": 0.3964,
      "step": 900
    },
    {
      "epoch": 0.7477793844247056,
      "grad_norm": 1.9232507318610614,
      "learning_rate": 1.8137692596313528e-06,
      "loss": 0.39,
      "step": 905
    },
    {
      "epoch": 0.7519107622392067,
      "grad_norm": 1.9669358048444339,
      "learning_rate": 1.7585218139957205e-06,
      "loss": 0.4089,
      "step": 910
    },
    {
      "epoch": 0.756042140053708,
      "grad_norm": 2.0275135426829007,
      "learning_rate": 1.7039487714756953e-06,
      "loss": 0.3897,
      "step": 915
    },
    {
      "epoch": 0.7601735178682091,
      "grad_norm": 1.9145558326260135,
      "learning_rate": 1.6500614862213866e-06,
      "loss": 0.3845,
      "step": 920
    },
    {
      "epoch": 0.7643048956827102,
      "grad_norm": 1.946476089351215,
      "learning_rate": 1.596871169708235e-06,
      "loss": 0.3827,
      "step": 925
    },
    {
      "epoch": 0.7684362734972113,
      "grad_norm": 2.061234713938133,
      "learning_rate": 1.54438888840441e-06,
      "loss": 0.3865,
      "step": 930
    },
    {
      "epoch": 0.7725676513117125,
      "grad_norm": 2.0046087896210993,
      "learning_rate": 1.4926255614683931e-06,
      "loss": 0.3919,
      "step": 935
    },
    {
      "epoch": 0.7766990291262136,
      "grad_norm": 2.013100573926538,
      "learning_rate": 1.4415919584771999e-06,
      "loss": 0.3791,
      "step": 940
    },
    {
      "epoch": 0.7808304069407147,
      "grad_norm": 2.052665994161596,
      "learning_rate": 1.3912986971857168e-06,
      "loss": 0.3813,
      "step": 945
    },
    {
      "epoch": 0.7849617847552158,
      "grad_norm": 1.9665758334089818,
      "learning_rate": 1.3417562413176405e-06,
      "loss": 0.3743,
      "step": 950
    },
    {
      "epoch": 0.789093162569717,
      "grad_norm": 1.9653082426188921,
      "learning_rate": 1.292974898388456e-06,
      "loss": 0.3884,
      "step": 955
    },
    {
      "epoch": 0.7932245403842182,
      "grad_norm": 2.1523771874487707,
      "learning_rate": 1.2449648175609115e-06,
      "loss": 0.383,
      "step": 960
    },
    {
      "epoch": 0.7973559181987193,
      "grad_norm": 1.8532885315782601,
      "learning_rate": 1.1977359875334454e-06,
      "loss": 0.3812,
      "step": 965
    },
    {
      "epoch": 0.8014872960132204,
      "grad_norm": 2.0616460511941477,
      "learning_rate": 1.1512982344619904e-06,
      "loss": 0.3742,
      "step": 970
    },
    {
      "epoch": 0.8056186738277216,
      "grad_norm": 2.000103259130858,
      "learning_rate": 1.1056612199156093e-06,
      "loss": 0.3795,
      "step": 975
    },
    {
      "epoch": 0.8097500516422227,
      "grad_norm": 2.0764477975903404,
      "learning_rate": 1.0608344388663576e-06,
      "loss": 0.3732,
      "step": 980
    },
    {
      "epoch": 0.8138814294567238,
      "grad_norm": 2.0844020934084315,
      "learning_rate": 1.01682721771382e-06,
      "loss": 0.3738,
      "step": 985
    },
    {
      "epoch": 0.8180128072712249,
      "grad_norm": 2.050710610881507,
      "learning_rate": 9.73648712344707e-07,
      "loss": 0.373,
      "step": 990
    },
    {
      "epoch": 0.8221441850857261,
      "grad_norm": 2.068871529971891,
      "learning_rate": 9.313079062279429e-07,
      "loss": 0.3852,
      "step": 995
    },
    {
      "epoch": 0.8262755629002272,
      "grad_norm": 1.9622479084933468,
      "learning_rate": 8.898136085456127e-07,
      "loss": 0.3839,
      "step": 1000
    },
    {
      "epoch": 0.8304069407147283,
      "grad_norm": 2.0489239754877913,
      "learning_rate": 8.491744523601741e-07,
      "loss": 0.3789,
      "step": 1005
    },
    {
      "epoch": 0.8345383185292296,
      "grad_norm": 1.97405744731925,
      "learning_rate": 8.093988928183127e-07,
      "loss": 0.3659,
      "step": 1010
    },
    {
      "epoch": 0.8386696963437307,
      "grad_norm": 2.0649700701955966,
      "learning_rate": 7.70495205391818e-07,
      "loss": 0.3684,
      "step": 1015
    },
    {
      "epoch": 0.8428010741582318,
      "grad_norm": 1.9711762707705944,
      "learning_rate": 7.324714841558322e-07,
      "loss": 0.3728,
      "step": 1020
    },
    {
      "epoch": 0.8469324519727329,
      "grad_norm": 2.089515078127623,
      "learning_rate": 6.953356401048466e-07,
      "loss": 0.361,
      "step": 1025
    },
    {
      "epoch": 0.851063829787234,
      "grad_norm": 2.0969140211573536,
      "learning_rate": 6.590953995067812e-07,
      "loss": 0.3621,
      "step": 1030
    },
    {
      "epoch": 0.8551952076017352,
      "grad_norm": 2.014897877209678,
      "learning_rate": 6.237583022955079e-07,
      "loss": 0.377,
      "step": 1035
    },
    {
      "epoch": 0.8593265854162363,
      "grad_norm": 2.0616530107992617,
      "learning_rate": 5.89331700502136e-07,
      "loss": 0.373,
      "step": 1040
    },
    {
      "epoch": 0.8634579632307374,
      "grad_norm": 2.0017552200191164,
      "learning_rate": 5.558227567253832e-07,
      "loss": 0.3738,
      "step": 1045
    },
    {
      "epoch": 0.8675893410452385,
      "grad_norm": 2.0884355733341415,
      "learning_rate": 5.232384426413772e-07,
      "loss": 0.3625,
      "step": 1050
    },
    {
      "epoch": 0.8717207188597397,
      "grad_norm": 2.111748293695631,
      "learning_rate": 4.915855375531642e-07,
      "loss": 0.3789,
      "step": 1055
    },
    {
      "epoch": 0.8758520966742409,
      "grad_norm": 2.0363025293506305,
      "learning_rate": 4.608706269802471e-07,
      "loss": 0.3615,
      "step": 1060
    },
    {
      "epoch": 0.879983474488742,
      "grad_norm": 2.0861363712497742,
      "learning_rate": 4.3110010128843747e-07,
      "loss": 0.3803,
      "step": 1065
    },
    {
      "epoch": 0.8841148523032432,
      "grad_norm": 1.9794223337927541,
      "learning_rate": 4.022801543603194e-07,
      "loss": 0.37,
      "step": 1070
    },
    {
      "epoch": 0.8882462301177443,
      "grad_norm": 1.976468279453303,
      "learning_rate": 3.744167823065814e-07,
      "loss": 0.3511,
      "step": 1075
    },
    {
      "epoch": 0.8923776079322454,
      "grad_norm": 2.0814002530931583,
      "learning_rate": 3.4751578221850377e-07,
      "loss": 0.3611,
      "step": 1080
    },
    {
      "epoch": 0.8965089857467465,
      "grad_norm": 2.0984819483954844,
      "learning_rate": 3.2158275096184744e-07,
      "loss": 0.3562,
      "step": 1085
    },
    {
      "epoch": 0.9006403635612477,
      "grad_norm": 2.144295712901401,
      "learning_rate": 2.966230840124007e-07,
      "loss": 0.3674,
      "step": 1090
    },
    {
      "epoch": 0.9047717413757488,
      "grad_norm": 2.118969417696939,
      "learning_rate": 2.72641974333433e-07,
      "loss": 0.3622,
      "step": 1095
    },
    {
      "epoch": 0.9089031191902499,
      "grad_norm": 2.0003217874891694,
      "learning_rate": 2.4964441129527337e-07,
      "loss": 0.3697,
      "step": 1100
    },
    {
      "epoch": 0.913034497004751,
      "grad_norm": 2.052789066548848,
      "learning_rate": 2.2763517963725169e-07,
      "loss": 0.3709,
      "step": 1105
    },
    {
      "epoch": 0.9171658748192523,
      "grad_norm": 1.9424275069967558,
      "learning_rate": 2.0661885847221398e-07,
      "loss": 0.3672,
      "step": 1110
    },
    {
      "epoch": 0.9212972526337534,
      "grad_norm": 2.030858945131136,
      "learning_rate": 1.8659982033381928e-07,
      "loss": 0.3675,
      "step": 1115
    },
    {
      "epoch": 0.9254286304482545,
      "grad_norm": 2.0600092854113132,
      "learning_rate": 1.6758223026681507e-07,
      "loss": 0.3636,
      "step": 1120
    },
    {
      "epoch": 0.9295600082627556,
      "grad_norm": 2.0750384756960374,
      "learning_rate": 1.4957004496048256e-07,
      "loss": 0.3656,
      "step": 1125
    },
    {
      "epoch": 0.9336913860772568,
      "grad_norm": 2.006693676746082,
      "learning_rate": 1.3256701192542853e-07,
      "loss": 0.3676,
      "step": 1130
    },
    {
      "epoch": 0.9378227638917579,
      "grad_norm": 2.114218265869402,
      "learning_rate": 1.1657666871390471e-07,
      "loss": 0.3658,
      "step": 1135
    },
    {
      "epoch": 0.941954141706259,
      "grad_norm": 2.0597089907860537,
      "learning_rate": 1.0160234218380305e-07,
      "loss": 0.3573,
      "step": 1140
    },
    {
      "epoch": 0.9460855195207601,
      "grad_norm": 2.0898079361017334,
      "learning_rate": 8.76471478064872e-08,
      "loss": 0.3724,
      "step": 1145
    },
    {
      "epoch": 0.9502168973352613,
      "grad_norm": 2.0371025770189632,
      "learning_rate": 7.471398901860772e-08,
      "loss": 0.3625,
      "step": 1150
    },
    {
      "epoch": 0.9543482751497624,
      "grad_norm": 1.938944230136909,
      "learning_rate": 6.280555661802857e-08,
      "loss": 0.3634,
      "step": 1155
    },
    {
      "epoch": 0.9584796529642636,
      "grad_norm": 2.1345661558712967,
      "learning_rate": 5.192432820399718e-08,
      "loss": 0.3674,
      "step": 1160
    },
    {
      "epoch": 0.9626110307787648,
      "grad_norm": 1.9759349002447268,
      "learning_rate": 4.207256766166845e-08,
      "loss": 0.3621,
      "step": 1165
    },
    {
      "epoch": 0.9667424085932659,
      "grad_norm": 2.0037876312026213,
      "learning_rate": 3.3252324691093185e-08,
      "loss": 0.3645,
      "step": 1170
    },
    {
      "epoch": 0.970873786407767,
      "grad_norm": 2.059268789116945,
      "learning_rate": 2.546543438077087e-08,
      "loss": 0.3569,
      "step": 1175
    },
    {
      "epoch": 0.9750051642222681,
      "grad_norm": 2.1019547322975916,
      "learning_rate": 1.8713516825851207e-08,
      "loss": 0.3604,
      "step": 1180
    },
    {
      "epoch": 0.9791365420367693,
      "grad_norm": 2.067476136669,
      "learning_rate": 1.2997976791065403e-08,
      "loss": 0.3611,
      "step": 1185
    },
    {
      "epoch": 0.9832679198512704,
      "grad_norm": 1.970947575751658,
      "learning_rate": 8.32000341846162e-09,
      "loss": 0.3626,
      "step": 1190
    },
    {
      "epoch": 0.9873992976657715,
      "grad_norm": 2.0213770050103816,
      "learning_rate": 4.6805699799967744e-09,
      "loss": 0.3577,
      "step": 1195
    },
    {
      "epoch": 0.9915306754802726,
      "grad_norm": 2.0031463180853915,
      "learning_rate": 2.0804336750429588e-09,
      "loss": 0.3594,
      "step": 1200
    },
    {
      "epoch": 0.9956620532947738,
      "grad_norm": 2.0391505951802165,
      "learning_rate": 5.201354728517905e-10,
      "loss": 0.3624,
      "step": 1205
    },
    {
      "epoch": 0.999793431109275,
      "grad_norm": 2.0674337207917235,
      "learning_rate": 0.0,
      "loss": 0.3717,
      "step": 1210
    },
    {
      "epoch": 0.999793431109275,
      "eval_loss": 0.3491726517677307,
      "eval_runtime": 125.1809,
      "eval_samples_per_second": 3.115,
      "eval_steps_per_second": 0.783,
      "step": 1210
    },
    {
      "epoch": 0.999793431109275,
      "step": 1210,
      "total_flos": 253297038458880.0,
      "train_loss": 0.5434080246066259,
      "train_runtime": 27449.6475,
      "train_samples_per_second": 1.411,
      "train_steps_per_second": 0.044
    }
  ],
  "logging_steps": 5,
  "max_steps": 1210,
  "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": 253297038458880.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}