{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.936619718309859,
  "eval_steps": 500,
  "global_step": 1251,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.011737089201877934,
      "grad_norm": 1.6926082666941107,
      "learning_rate": 9.237540571428572e-06,
      "loss": 4.1191,
      "step": 5
    },
    {
      "epoch": 0.023474178403755867,
      "grad_norm": 1.0523580952405567,
      "learning_rate": 2.0784466285714287e-05,
      "loss": 3.9165,
      "step": 10
    },
    {
      "epoch": 0.035211267605633804,
      "grad_norm": 0.9348982322725803,
      "learning_rate": 3.2331392000000005e-05,
      "loss": 3.0751,
      "step": 15
    },
    {
      "epoch": 0.046948356807511735,
      "grad_norm": 0.963142977085111,
      "learning_rate": 4.3878317714285716e-05,
      "loss": 2.3471,
      "step": 20
    },
    {
      "epoch": 0.05868544600938967,
      "grad_norm": 0.41999800522000985,
      "learning_rate": 5.542524342857144e-05,
      "loss": 1.9358,
      "step": 25
    },
    {
      "epoch": 0.07042253521126761,
      "grad_norm": 0.44353131458171546,
      "learning_rate": 6.697216914285716e-05,
      "loss": 1.8648,
      "step": 30
    },
    {
      "epoch": 0.08215962441314555,
      "grad_norm": 1.3157403937206136,
      "learning_rate": 7.851909485714286e-05,
      "loss": 1.874,
      "step": 35
    },
    {
      "epoch": 0.09389671361502347,
      "grad_norm": 0.30804693051898663,
      "learning_rate": 8.082693104266452e-05,
      "loss": 1.8383,
      "step": 40
    },
    {
      "epoch": 0.1056338028169014,
      "grad_norm": 0.27001251328292725,
      "learning_rate": 8.082063867481342e-05,
      "loss": 1.4036,
      "step": 45
    },
    {
      "epoch": 0.11737089201877934,
      "grad_norm": 0.3485335678060715,
      "learning_rate": 8.080950709070086e-05,
      "loss": 1.6265,
      "step": 50
    },
    {
      "epoch": 0.12910798122065728,
      "grad_norm": 0.25009903952840973,
      "learning_rate": 8.079353806798678e-05,
      "loss": 1.5537,
      "step": 55
    },
    {
      "epoch": 0.14084507042253522,
      "grad_norm": 0.283472870261176,
      "learning_rate": 8.07727341568466e-05,
      "loss": 1.5242,
      "step": 60
    },
    {
      "epoch": 0.15258215962441316,
      "grad_norm": 0.2603618862170815,
      "learning_rate": 8.074709867956397e-05,
      "loss": 1.52,
      "step": 65
    },
    {
      "epoch": 0.1643192488262911,
      "grad_norm": 0.26748765076963404,
      "learning_rate": 8.071663573000018e-05,
      "loss": 1.4147,
      "step": 70
    },
    {
      "epoch": 0.176056338028169,
      "grad_norm": 0.27233224635502085,
      "learning_rate": 8.068135017294043e-05,
      "loss": 1.4277,
      "step": 75
    },
    {
      "epoch": 0.18779342723004694,
      "grad_norm": 0.27807424769416844,
      "learning_rate": 8.064124764331693e-05,
      "loss": 1.5202,
      "step": 80
    },
    {
      "epoch": 0.19953051643192488,
      "grad_norm": 0.237501219160123,
      "learning_rate": 8.059633454530899e-05,
      "loss": 1.5265,
      "step": 85
    },
    {
      "epoch": 0.2112676056338028,
      "grad_norm": 0.26393741964527623,
      "learning_rate": 8.054661805132042e-05,
      "loss": 1.4661,
      "step": 90
    },
    {
      "epoch": 0.22300469483568075,
      "grad_norm": 0.307134230448153,
      "learning_rate": 8.049210610083398e-05,
      "loss": 1.5281,
      "step": 95
    },
    {
      "epoch": 0.2347417840375587,
      "grad_norm": 0.30034381561230195,
      "learning_rate": 8.043280739914361e-05,
      "loss": 1.4272,
      "step": 100
    },
    {
      "epoch": 0.24647887323943662,
      "grad_norm": 0.25199182793601094,
      "learning_rate": 8.036873141596414e-05,
      "loss": 1.385,
      "step": 105
    },
    {
      "epoch": 0.25821596244131456,
      "grad_norm": 0.25027218831718984,
      "learning_rate": 8.029988838391911e-05,
      "loss": 1.4313,
      "step": 110
    },
    {
      "epoch": 0.2699530516431925,
      "grad_norm": 1.0436457556233485,
      "learning_rate": 8.02262892969066e-05,
      "loss": 1.4076,
      "step": 115
    },
    {
      "epoch": 0.28169014084507044,
      "grad_norm": 0.25287945938151374,
      "learning_rate": 8.014794590834359e-05,
      "loss": 1.4006,
      "step": 120
    },
    {
      "epoch": 0.2934272300469484,
      "grad_norm": 0.3426017772114307,
      "learning_rate": 8.0064870729289e-05,
      "loss": 1.3478,
      "step": 125
    },
    {
      "epoch": 0.3051643192488263,
      "grad_norm": 0.24548095078180857,
      "learning_rate": 7.997707702644572e-05,
      "loss": 1.4508,
      "step": 130
    },
    {
      "epoch": 0.31690140845070425,
      "grad_norm": 0.3257280042891666,
      "learning_rate": 7.988457882004198e-05,
      "loss": 1.3507,
      "step": 135
    },
    {
      "epoch": 0.3286384976525822,
      "grad_norm": 0.2648447775178709,
      "learning_rate": 7.978739088159242e-05,
      "loss": 1.3959,
      "step": 140
    },
    {
      "epoch": 0.3403755868544601,
      "grad_norm": 0.3253574365810539,
      "learning_rate": 7.968552873153908e-05,
      "loss": 1.5608,
      "step": 145
    },
    {
      "epoch": 0.352112676056338,
      "grad_norm": 0.2683217685759068,
      "learning_rate": 7.957900863677296e-05,
      "loss": 1.3742,
      "step": 150
    },
    {
      "epoch": 0.36384976525821594,
      "grad_norm": 0.27823469209609253,
      "learning_rate": 7.94678476080362e-05,
      "loss": 1.4693,
      "step": 155
    },
    {
      "epoch": 0.3755868544600939,
      "grad_norm": 0.31396185033997276,
      "learning_rate": 7.935206339720553e-05,
      "loss": 1.4293,
      "step": 160
    },
    {
      "epoch": 0.3873239436619718,
      "grad_norm": 0.23711923917761538,
      "learning_rate": 7.923167449445751e-05,
      "loss": 1.3614,
      "step": 165
    },
    {
      "epoch": 0.39906103286384975,
      "grad_norm": 0.23098090222216835,
      "learning_rate": 7.910670012531556e-05,
      "loss": 1.4348,
      "step": 170
    },
    {
      "epoch": 0.4107981220657277,
      "grad_norm": 0.24726241172120542,
      "learning_rate": 7.897716024757991e-05,
      "loss": 1.3991,
      "step": 175
    },
    {
      "epoch": 0.4225352112676056,
      "grad_norm": 0.2529598203448259,
      "learning_rate": 7.884307554814028e-05,
      "loss": 1.3889,
      "step": 180
    },
    {
      "epoch": 0.43427230046948356,
      "grad_norm": 0.2626838974550443,
      "learning_rate": 7.870446743967242e-05,
      "loss": 1.3164,
      "step": 185
    },
    {
      "epoch": 0.4460093896713615,
      "grad_norm": 0.22340503189330538,
      "learning_rate": 7.856135805721846e-05,
      "loss": 1.3403,
      "step": 190
    },
    {
      "epoch": 0.45774647887323944,
      "grad_norm": 0.2643635528374707,
      "learning_rate": 7.841377025465227e-05,
      "loss": 1.3128,
      "step": 195
    },
    {
      "epoch": 0.4694835680751174,
      "grad_norm": 0.23702382238947953,
      "learning_rate": 7.826172760102954e-05,
      "loss": 1.3577,
      "step": 200
    },
    {
      "epoch": 0.4812206572769953,
      "grad_norm": 0.3089050115416261,
      "learning_rate": 7.810525437682417e-05,
      "loss": 1.3627,
      "step": 205
    },
    {
      "epoch": 0.49295774647887325,
      "grad_norm": 0.2459426412330782,
      "learning_rate": 7.794437557005058e-05,
      "loss": 1.361,
      "step": 210
    },
    {
      "epoch": 0.5046948356807511,
      "grad_norm": 0.2474645086201422,
      "learning_rate": 7.777911687227339e-05,
      "loss": 1.35,
      "step": 215
    },
    {
      "epoch": 0.5164319248826291,
      "grad_norm": 0.2776269910713455,
      "learning_rate": 7.760950467450457e-05,
      "loss": 1.3038,
      "step": 220
    },
    {
      "epoch": 0.528169014084507,
      "grad_norm": 0.23093548062400274,
      "learning_rate": 7.743556606298887e-05,
      "loss": 1.2999,
      "step": 225
    },
    {
      "epoch": 0.539906103286385,
      "grad_norm": 0.21029555812416517,
      "learning_rate": 7.725732881487837e-05,
      "loss": 1.3563,
      "step": 230
    },
    {
      "epoch": 0.5516431924882629,
      "grad_norm": 0.2380837640644916,
      "learning_rate": 7.707482139379649e-05,
      "loss": 1.4117,
      "step": 235
    },
    {
      "epoch": 0.5633802816901409,
      "grad_norm": 0.2501691778817746,
      "learning_rate": 7.688807294529259e-05,
      "loss": 1.3747,
      "step": 240
    },
    {
      "epoch": 0.5751173708920188,
      "grad_norm": 0.2531422248522496,
      "learning_rate": 7.669711329218748e-05,
      "loss": 1.3142,
      "step": 245
    },
    {
      "epoch": 0.5868544600938967,
      "grad_norm": 0.3071530417375628,
      "learning_rate": 7.6501972929811e-05,
      "loss": 1.3518,
      "step": 250
    },
    {
      "epoch": 0.5985915492957746,
      "grad_norm": 0.2462933779162881,
      "learning_rate": 7.630268302113184e-05,
      "loss": 1.3381,
      "step": 255
    },
    {
      "epoch": 0.6103286384976526,
      "grad_norm": 0.231440592355059,
      "learning_rate": 7.609927539178122e-05,
      "loss": 1.2967,
      "step": 260
    },
    {
      "epoch": 0.6220657276995305,
      "grad_norm": 0.26883215485587464,
      "learning_rate": 7.589178252497026e-05,
      "loss": 1.4204,
      "step": 265
    },
    {
      "epoch": 0.6338028169014085,
      "grad_norm": 1.5499941567046591,
      "learning_rate": 7.568023755630275e-05,
      "loss": 1.2625,
      "step": 270
    },
    {
      "epoch": 0.6455399061032864,
      "grad_norm": 0.23176606216751602,
      "learning_rate": 7.546467426848345e-05,
      "loss": 1.3575,
      "step": 275
    },
    {
      "epoch": 0.6572769953051644,
      "grad_norm": 0.29787341630319236,
      "learning_rate": 7.524512708592318e-05,
      "loss": 1.3631,
      "step": 280
    },
    {
      "epoch": 0.6690140845070423,
      "grad_norm": 0.2414950159288888,
      "learning_rate": 7.502163106924143e-05,
      "loss": 1.3421,
      "step": 285
    },
    {
      "epoch": 0.6807511737089202,
      "grad_norm": 0.2609054694114428,
      "learning_rate": 7.479422190966731e-05,
      "loss": 1.2775,
      "step": 290
    },
    {
      "epoch": 0.6924882629107981,
      "grad_norm": 0.24342075714638725,
      "learning_rate": 7.456293592333991e-05,
      "loss": 1.2399,
      "step": 295
    },
    {
      "epoch": 0.704225352112676,
      "grad_norm": 0.2451126010921149,
      "learning_rate": 7.432781004550864e-05,
      "loss": 1.1849,
      "step": 300
    },
    {
      "epoch": 0.715962441314554,
      "grad_norm": 0.38381425009051096,
      "learning_rate": 7.408888182463507e-05,
      "loss": 1.2887,
      "step": 305
    },
    {
      "epoch": 0.7276995305164319,
      "grad_norm": 0.23592915827608676,
      "learning_rate": 7.38461894163964e-05,
      "loss": 1.3505,
      "step": 310
    },
    {
      "epoch": 0.7394366197183099,
      "grad_norm": 0.2698206782214951,
      "learning_rate": 7.35997715775923e-05,
      "loss": 1.3215,
      "step": 315
    },
    {
      "epoch": 0.7511737089201878,
      "grad_norm": 0.2453301887513675,
      "learning_rate": 7.334966765995569e-05,
      "loss": 1.2296,
      "step": 320
    },
    {
      "epoch": 0.7629107981220657,
      "grad_norm": 0.21952063652943296,
      "learning_rate": 7.309591760386829e-05,
      "loss": 1.3247,
      "step": 325
    },
    {
      "epoch": 0.7746478873239436,
      "grad_norm": 0.28692324866840146,
      "learning_rate": 7.28385619319825e-05,
      "loss": 1.2973,
      "step": 330
    },
    {
      "epoch": 0.7863849765258216,
      "grad_norm": 0.24556875494678596,
      "learning_rate": 7.257764174275004e-05,
      "loss": 1.1292,
      "step": 335
    },
    {
      "epoch": 0.7981220657276995,
      "grad_norm": 0.21102775812196028,
      "learning_rate": 7.231319870385874e-05,
      "loss": 1.1703,
      "step": 340
    },
    {
      "epoch": 0.8098591549295775,
      "grad_norm": 0.22043483877271167,
      "learning_rate": 7.204527504557848e-05,
      "loss": 1.1836,
      "step": 345
    },
    {
      "epoch": 0.8215962441314554,
      "grad_norm": 0.23022946831058466,
      "learning_rate": 7.177391355401707e-05,
      "loss": 1.3726,
      "step": 350
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.2246896571930163,
      "learning_rate": 7.149915756428768e-05,
      "loss": 1.3799,
      "step": 355
    },
    {
      "epoch": 0.8450704225352113,
      "grad_norm": 0.2523792044709292,
      "learning_rate": 7.122105095358836e-05,
      "loss": 1.3122,
      "step": 360
    },
    {
      "epoch": 0.8568075117370892,
      "grad_norm": 0.23559138914799244,
      "learning_rate": 7.0939638134195e-05,
      "loss": 1.2995,
      "step": 365
    },
    {
      "epoch": 0.8685446009389671,
      "grad_norm": 0.2597889434863669,
      "learning_rate": 7.065496404636905e-05,
      "loss": 1.369,
      "step": 370
    },
    {
      "epoch": 0.8802816901408451,
      "grad_norm": 0.24829748827668763,
      "learning_rate": 7.036707415118069e-05,
      "loss": 1.2793,
      "step": 375
    },
    {
      "epoch": 0.892018779342723,
      "grad_norm": 0.6851956453951893,
      "learning_rate": 7.007601442324886e-05,
      "loss": 1.3262,
      "step": 380
    },
    {
      "epoch": 0.903755868544601,
      "grad_norm": 0.24078658149556825,
      "learning_rate": 6.978183134339949e-05,
      "loss": 1.3209,
      "step": 385
    },
    {
      "epoch": 0.9154929577464789,
      "grad_norm": 0.24167241236562123,
      "learning_rate": 6.948457189124259e-05,
      "loss": 1.3779,
      "step": 390
    },
    {
      "epoch": 0.9272300469483568,
      "grad_norm": 0.24575130336431836,
      "learning_rate": 6.918428353766991e-05,
      "loss": 1.3325,
      "step": 395
    },
    {
      "epoch": 0.9389671361502347,
      "grad_norm": 0.2326601273504395,
      "learning_rate": 6.888101423727413e-05,
      "loss": 1.2331,
      "step": 400
    },
    {
      "epoch": 0.9507042253521126,
      "grad_norm": 0.22591521946130988,
      "learning_rate": 6.857481242069052e-05,
      "loss": 1.2541,
      "step": 405
    },
    {
      "epoch": 0.9624413145539906,
      "grad_norm": 0.2308511630846233,
      "learning_rate": 6.82657269868631e-05,
      "loss": 1.2298,
      "step": 410
    },
    {
      "epoch": 0.9741784037558685,
      "grad_norm": 0.25182225001294944,
      "learning_rate": 6.795380729523547e-05,
      "loss": 1.301,
      "step": 415
    },
    {
      "epoch": 0.9859154929577465,
      "grad_norm": 0.2605089795549262,
      "learning_rate": 6.76391031578685e-05,
      "loss": 1.3731,
      "step": 420
    },
    {
      "epoch": 0.9976525821596244,
      "grad_norm": 0.22892211041701857,
      "learning_rate": 6.732166483148549e-05,
      "loss": 1.1978,
      "step": 425
    },
    {
      "epoch": 0.9976525821596244,
      "eval_loss": 1.6044145822525024,
      "eval_runtime": 50.0629,
      "eval_samples_per_second": 1.458,
      "eval_steps_per_second": 0.2,
      "step": 425
    },
    {
      "epoch": 1.0093896713615023,
      "grad_norm": 0.24910638304822852,
      "learning_rate": 6.70015430094465e-05,
      "loss": 1.123,
      "step": 430
    },
    {
      "epoch": 1.0211267605633803,
      "grad_norm": 0.33357404745827773,
      "learning_rate": 6.667878881365277e-05,
      "loss": 1.1185,
      "step": 435
    },
    {
      "epoch": 1.0328638497652582,
      "grad_norm": 0.29008454147641954,
      "learning_rate": 6.635345378638292e-05,
      "loss": 1.1284,
      "step": 440
    },
    {
      "epoch": 1.0446009389671362,
      "grad_norm": 0.2670483776495452,
      "learning_rate": 6.602558988206174e-05,
      "loss": 1.2331,
      "step": 445
    },
    {
      "epoch": 1.056338028169014,
      "grad_norm": 0.2430183857087479,
      "learning_rate": 6.56952494589635e-05,
      "loss": 1.0354,
      "step": 450
    },
    {
      "epoch": 1.068075117370892,
      "grad_norm": 0.25159392522233937,
      "learning_rate": 6.536248527085044e-05,
      "loss": 1.095,
      "step": 455
    },
    {
      "epoch": 1.07981220657277,
      "grad_norm": 0.25129604646154363,
      "learning_rate": 6.502735045854836e-05,
      "loss": 1.1631,
      "step": 460
    },
    {
      "epoch": 1.091549295774648,
      "grad_norm": 0.2375727576097195,
      "learning_rate": 6.468989854146017e-05,
      "loss": 1.056,
      "step": 465
    },
    {
      "epoch": 1.1032863849765258,
      "grad_norm": 0.25705262164871134,
      "learning_rate": 6.435018340901922e-05,
      "loss": 1.1019,
      "step": 470
    },
    {
      "epoch": 1.1150234741784038,
      "grad_norm": 0.2632363761365353,
      "learning_rate": 6.400825931208336e-05,
      "loss": 1.0925,
      "step": 475
    },
    {
      "epoch": 1.1267605633802817,
      "grad_norm": 0.25529732934150606,
      "learning_rate": 6.366418085427133e-05,
      "loss": 1.0742,
      "step": 480
    },
    {
      "epoch": 1.1384976525821595,
      "grad_norm": 0.24911618052576018,
      "learning_rate": 6.331800298324286e-05,
      "loss": 1.1257,
      "step": 485
    },
    {
      "epoch": 1.1502347417840375,
      "grad_norm": 0.2713693928567898,
      "learning_rate": 6.29697809819238e-05,
      "loss": 1.1852,
      "step": 490
    },
    {
      "epoch": 1.1619718309859155,
      "grad_norm": 0.25598293691462287,
      "learning_rate": 6.26195704596776e-05,
      "loss": 1.147,
      "step": 495
    },
    {
      "epoch": 1.1737089201877935,
      "grad_norm": 0.2506209145241536,
      "learning_rate": 6.2267427343425e-05,
      "loss": 1.0602,
      "step": 500
    },
    {
      "epoch": 1.1737089201877935,
      "eval_loss": 1.5600786209106445,
      "eval_runtime": 49.8763,
      "eval_samples_per_second": 1.464,
      "eval_steps_per_second": 0.2,
      "step": 500
    },
    {
      "epoch": 1.1854460093896715,
      "grad_norm": 0.24306048622973317,
      "learning_rate": 6.19134078687125e-05,
      "loss": 1.0243,
      "step": 505
    },
    {
      "epoch": 1.1971830985915493,
      "grad_norm": 0.26647005420227804,
      "learning_rate": 6.155756857073197e-05,
      "loss": 1.1676,
      "step": 510
    },
    {
      "epoch": 1.2089201877934272,
      "grad_norm": 0.28346081980702725,
      "learning_rate": 6.119996627529222e-05,
      "loss": 1.1091,
      "step": 515
    },
    {
      "epoch": 1.2206572769953052,
      "grad_norm": 0.2523758910052274,
      "learning_rate": 6.084065808974423e-05,
      "loss": 1.1431,
      "step": 520
    },
    {
      "epoch": 1.232394366197183,
      "grad_norm": 0.24326518180467072,
      "learning_rate": 6.047970139386133e-05,
      "loss": 1.0999,
      "step": 525
    },
    {
      "epoch": 1.244131455399061,
      "grad_norm": 0.2517876789791905,
      "learning_rate": 6.011715383067593e-05,
      "loss": 1.1285,
      "step": 530
    },
    {
      "epoch": 1.255868544600939,
      "grad_norm": 0.25670486692069977,
      "learning_rate": 5.975307329727426e-05,
      "loss": 1.0623,
      "step": 535
    },
    {
      "epoch": 1.267605633802817,
      "grad_norm": 0.2831263117979734,
      "learning_rate": 5.938751793555044e-05,
      "loss": 1.1384,
      "step": 540
    },
    {
      "epoch": 1.2793427230046948,
      "grad_norm": 0.2514294700936478,
      "learning_rate": 5.902054612292151e-05,
      "loss": 1.1115,
      "step": 545
    },
    {
      "epoch": 1.2910798122065728,
      "grad_norm": 0.2830207243631822,
      "learning_rate": 5.865221646300485e-05,
      "loss": 1.1742,
      "step": 550
    },
    {
      "epoch": 1.3028169014084507,
      "grad_norm": 0.27398863839678606,
      "learning_rate": 5.828258777625944e-05,
      "loss": 1.1052,
      "step": 555
    },
    {
      "epoch": 1.3145539906103285,
      "grad_norm": 0.25895033553118535,
      "learning_rate": 5.791171909059259e-05,
      "loss": 1.2186,
      "step": 560
    },
    {
      "epoch": 1.3262910798122065,
      "grad_norm": 0.258186822849988,
      "learning_rate": 5.7539669631933385e-05,
      "loss": 1.1201,
      "step": 565
    },
    {
      "epoch": 1.3380281690140845,
      "grad_norm": 0.25129851535157616,
      "learning_rate": 5.71664988147746e-05,
      "loss": 1.1031,
      "step": 570
    },
    {
      "epoch": 1.3497652582159625,
      "grad_norm": 0.25982425225157424,
      "learning_rate": 5.6792266232684574e-05,
      "loss": 1.0259,
      "step": 575
    },
    {
      "epoch": 1.3615023474178405,
      "grad_norm": 0.26067740399731576,
      "learning_rate": 5.6417031648790324e-05,
      "loss": 1.2319,
      "step": 580
    },
    {
      "epoch": 1.3732394366197183,
      "grad_norm": 0.28203855843767645,
      "learning_rate": 5.6040854986233706e-05,
      "loss": 1.1523,
      "step": 585
    },
    {
      "epoch": 1.3849765258215962,
      "grad_norm": 0.2616034024922836,
      "learning_rate": 5.5663796318601946e-05,
      "loss": 1.0862,
      "step": 590
    },
    {
      "epoch": 1.3967136150234742,
      "grad_norm": 0.2557988750857973,
      "learning_rate": 5.528591586033421e-05,
      "loss": 1.1261,
      "step": 595
    },
    {
      "epoch": 1.408450704225352,
      "grad_norm": 0.2583387594345229,
      "learning_rate": 5.490727395710563e-05,
      "loss": 1.1638,
      "step": 600
    },
    {
      "epoch": 1.42018779342723,
      "grad_norm": 0.2824039637269599,
      "learning_rate": 5.452793107619043e-05,
      "loss": 1.1064,
      "step": 605
    },
    {
      "epoch": 1.431924882629108,
      "grad_norm": 0.30378896985216747,
      "learning_rate": 5.4147947796805546e-05,
      "loss": 1.1587,
      "step": 610
    },
    {
      "epoch": 1.443661971830986,
      "grad_norm": 0.26055491232191014,
      "learning_rate": 5.376738480043643e-05,
      "loss": 1.1737,
      "step": 615
    },
    {
      "epoch": 1.455399061032864,
      "grad_norm": 0.28775987096455746,
      "learning_rate": 5.338630286114656e-05,
      "loss": 1.0302,
      "step": 620
    },
    {
      "epoch": 1.4671361502347418,
      "grad_norm": 0.27434540717537476,
      "learning_rate": 5.300476283587205e-05,
      "loss": 1.096,
      "step": 625
    },
    {
      "epoch": 1.4788732394366197,
      "grad_norm": 0.2687633406815332,
      "learning_rate": 5.262282565470313e-05,
      "loss": 1.1784,
      "step": 630
    },
    {
      "epoch": 1.4906103286384975,
      "grad_norm": 0.30497461697387473,
      "learning_rate": 5.224055231115389e-05,
      "loss": 1.0557,
      "step": 635
    },
    {
      "epoch": 1.5023474178403755,
      "grad_norm": 0.2401646558924421,
      "learning_rate": 5.185800385242196e-05,
      "loss": 1.2208,
      "step": 640
    },
    {
      "epoch": 1.5140845070422535,
      "grad_norm": 0.28302252582641413,
      "learning_rate": 5.147524136963947e-05,
      "loss": 1.1629,
      "step": 645
    },
    {
      "epoch": 1.5258215962441315,
      "grad_norm": 0.2733981765032047,
      "learning_rate": 5.1092325988117195e-05,
      "loss": 1.1179,
      "step": 650
    },
    {
      "epoch": 1.5375586854460095,
      "grad_norm": 0.3543210495906027,
      "learning_rate": 5.0709318857583076e-05,
      "loss": 1.1171,
      "step": 655
    },
    {
      "epoch": 1.5492957746478875,
      "grad_norm": 0.2678160500016965,
      "learning_rate": 5.032628114241694e-05,
      "loss": 1.1378,
      "step": 660
    },
    {
      "epoch": 1.5610328638497653,
      "grad_norm": 0.2841104290261492,
      "learning_rate": 4.994327401188283e-05,
      "loss": 1.1317,
      "step": 665
    },
    {
      "epoch": 1.5727699530516432,
      "grad_norm": 0.23756976664123616,
      "learning_rate": 4.956035863036055e-05,
      "loss": 1.1054,
      "step": 670
    },
    {
      "epoch": 1.584507042253521,
      "grad_norm": 0.299042726296967,
      "learning_rate": 4.9177596147578067e-05,
      "loss": 1.14,
      "step": 675
    },
    {
      "epoch": 1.596244131455399,
      "grad_norm": 0.24219892354659045,
      "learning_rate": 4.879504768884612e-05,
      "loss": 1.2392,
      "step": 680
    },
    {
      "epoch": 1.607981220657277,
      "grad_norm": 0.2659120257163632,
      "learning_rate": 4.841277434529689e-05,
      "loss": 1.076,
      "step": 685
    },
    {
      "epoch": 1.619718309859155,
      "grad_norm": 0.2535490633175144,
      "learning_rate": 4.803083716412797e-05,
      "loss": 1.1433,
      "step": 690
    },
    {
      "epoch": 1.631455399061033,
      "grad_norm": 0.2601946247918616,
      "learning_rate": 4.764929713885346e-05,
      "loss": 1.1151,
      "step": 695
    },
    {
      "epoch": 1.6431924882629108,
      "grad_norm": 0.2510924947348151,
      "learning_rate": 4.726821519956359e-05,
      "loss": 1.1195,
      "step": 700
    },
    {
      "epoch": 1.6549295774647887,
      "grad_norm": 0.2717853110326184,
      "learning_rate": 4.688765220319449e-05,
      "loss": 1.0717,
      "step": 705
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.29413144596709895,
      "learning_rate": 4.65076689238096e-05,
      "loss": 1.1293,
      "step": 710
    },
    {
      "epoch": 1.6784037558685445,
      "grad_norm": 0.260973856516902,
      "learning_rate": 4.61283260428944e-05,
      "loss": 1.0877,
      "step": 715
    },
    {
      "epoch": 1.6901408450704225,
      "grad_norm": 0.24688922048601705,
      "learning_rate": 4.574968413966582e-05,
      "loss": 1.074,
      "step": 720
    },
    {
      "epoch": 1.7018779342723005,
      "grad_norm": 0.29518134961553794,
      "learning_rate": 4.537180368139808e-05,
      "loss": 1.2313,
      "step": 725
    },
    {
      "epoch": 1.7136150234741785,
      "grad_norm": 0.2789615734091057,
      "learning_rate": 4.499474501376632e-05,
      "loss": 1.0704,
      "step": 730
    },
    {
      "epoch": 1.7253521126760565,
      "grad_norm": 0.2658926149309144,
      "learning_rate": 4.461856835120969e-05,
      "loss": 1.0596,
      "step": 735
    },
    {
      "epoch": 1.7370892018779343,
      "grad_norm": 0.2572838262933699,
      "learning_rate": 4.424333376731544e-05,
      "loss": 1.1131,
      "step": 740
    },
    {
      "epoch": 1.7488262910798122,
      "grad_norm": 0.2771724916201255,
      "learning_rate": 4.386910118522542e-05,
      "loss": 1.1371,
      "step": 745
    },
    {
      "epoch": 1.76056338028169,
      "grad_norm": 0.27676486532377537,
      "learning_rate": 4.349593036806664e-05,
      "loss": 1.0967,
      "step": 750
    },
    {
      "epoch": 1.772300469483568,
      "grad_norm": 0.26672550277720775,
      "learning_rate": 4.3123880909407424e-05,
      "loss": 1.0533,
      "step": 755
    },
    {
      "epoch": 1.784037558685446,
      "grad_norm": 0.23662667298158832,
      "learning_rate": 4.275301222374057e-05,
      "loss": 1.0783,
      "step": 760
    },
    {
      "epoch": 1.795774647887324,
      "grad_norm": 0.29383896298290885,
      "learning_rate": 4.238338353699518e-05,
      "loss": 1.17,
      "step": 765
    },
    {
      "epoch": 1.807511737089202,
      "grad_norm": 0.28154071670327757,
      "learning_rate": 4.201505387707851e-05,
      "loss": 1.1453,
      "step": 770
    },
    {
      "epoch": 1.8192488262910798,
      "grad_norm": 0.25700361275731554,
      "learning_rate": 4.1648082064449574e-05,
      "loss": 1.082,
      "step": 775
    },
    {
      "epoch": 1.8309859154929577,
      "grad_norm": 0.2826122217432797,
      "learning_rate": 4.128252670272575e-05,
      "loss": 1.0474,
      "step": 780
    },
    {
      "epoch": 1.8427230046948355,
      "grad_norm": 0.2620920138050905,
      "learning_rate": 4.091844616932409e-05,
      "loss": 1.1321,
      "step": 785
    },
    {
      "epoch": 1.8544600938967135,
      "grad_norm": 0.29154525087785954,
      "learning_rate": 4.055589860613869e-05,
      "loss": 1.1328,
      "step": 790
    },
    {
      "epoch": 1.8661971830985915,
      "grad_norm": 0.2619470877476919,
      "learning_rate": 4.01949419102558e-05,
      "loss": 1.0941,
      "step": 795
    },
    {
      "epoch": 1.8779342723004695,
      "grad_norm": 0.24822033480152877,
      "learning_rate": 3.983563372470781e-05,
      "loss": 0.9679,
      "step": 800
    },
    {
      "epoch": 1.8896713615023475,
      "grad_norm": 0.28224783656544505,
      "learning_rate": 3.9478031429268066e-05,
      "loss": 1.144,
      "step": 805
    },
    {
      "epoch": 1.9014084507042255,
      "grad_norm": 0.30516183723950785,
      "learning_rate": 3.912219213128753e-05,
      "loss": 1.1058,
      "step": 810
    },
    {
      "epoch": 1.9131455399061033,
      "grad_norm": 0.30562651002624985,
      "learning_rate": 3.876817265657502e-05,
      "loss": 1.1647,
      "step": 815
    },
    {
      "epoch": 1.9248826291079812,
      "grad_norm": 0.2768262798277747,
      "learning_rate": 3.8416029540322404e-05,
      "loss": 1.1273,
      "step": 820
    },
    {
      "epoch": 1.936619718309859,
      "grad_norm": 0.2842898850002602,
      "learning_rate": 3.806581901807623e-05,
      "loss": 1.1984,
      "step": 825
    },
    {
      "epoch": 1.948356807511737,
      "grad_norm": 0.31329090267496584,
      "learning_rate": 3.771759701675715e-05,
      "loss": 1.1333,
      "step": 830
    },
    {
      "epoch": 1.960093896713615,
      "grad_norm": 0.3381345989557091,
      "learning_rate": 3.7371419145728696e-05,
      "loss": 1.0973,
      "step": 835
    },
    {
      "epoch": 1.971830985915493,
      "grad_norm": 0.2970165089739574,
      "learning_rate": 3.702734068791666e-05,
      "loss": 1.1552,
      "step": 840
    },
    {
      "epoch": 1.983568075117371,
      "grad_norm": 0.26229774431947867,
      "learning_rate": 3.6685416590980806e-05,
      "loss": 1.0432,
      "step": 845
    },
    {
      "epoch": 1.995305164319249,
      "grad_norm": 0.3020477793486203,
      "learning_rate": 3.634570145853985e-05,
      "loss": 1.1401,
      "step": 850
    },
    {
      "epoch": 1.995305164319249,
      "eval_loss": 1.5306791067123413,
      "eval_runtime": 51.0959,
      "eval_samples_per_second": 1.429,
      "eval_steps_per_second": 0.196,
      "step": 850
    },
    {
      "epoch": 2.007042253521127,
      "grad_norm": 0.24436801124041757,
      "learning_rate": 3.600824954145166e-05,
      "loss": 0.9757,
      "step": 855
    },
    {
      "epoch": 2.0187793427230045,
      "grad_norm": 0.3332860055486253,
      "learning_rate": 3.567311472914957e-05,
      "loss": 0.9992,
      "step": 860
    },
    {
      "epoch": 2.0305164319248825,
      "grad_norm": 0.2915663428907029,
      "learning_rate": 3.5340350541036516e-05,
      "loss": 0.9287,
      "step": 865
    },
    {
      "epoch": 2.0422535211267605,
      "grad_norm": 0.31218559752172226,
      "learning_rate": 3.501001011793828e-05,
      "loss": 0.8744,
      "step": 870
    },
    {
      "epoch": 2.0539906103286385,
      "grad_norm": 0.3014812239987984,
      "learning_rate": 3.46821462136171e-05,
      "loss": 0.7871,
      "step": 875
    },
    {
      "epoch": 2.0657276995305165,
      "grad_norm": 0.2919198189363655,
      "learning_rate": 3.4356811186347255e-05,
      "loss": 0.8375,
      "step": 880
    },
    {
      "epoch": 2.0774647887323945,
      "grad_norm": 0.30297421951074827,
      "learning_rate": 3.4034056990553525e-05,
      "loss": 0.9539,
      "step": 885
    },
    {
      "epoch": 2.0892018779342725,
      "grad_norm": 0.3004785475460889,
      "learning_rate": 3.3713935168514526e-05,
      "loss": 0.8916,
      "step": 890
    },
    {
      "epoch": 2.10093896713615,
      "grad_norm": 0.3063305424781025,
      "learning_rate": 3.339649684213153e-05,
      "loss": 0.8923,
      "step": 895
    },
    {
      "epoch": 2.112676056338028,
      "grad_norm": 0.3087837541879336,
      "learning_rate": 3.308179270476455e-05,
      "loss": 0.9808,
      "step": 900
    },
    {
      "epoch": 2.124413145539906,
      "grad_norm": 0.3069084480862884,
      "learning_rate": 3.276987301313693e-05,
      "loss": 0.9499,
      "step": 905
    },
    {
      "epoch": 2.136150234741784,
      "grad_norm": 0.31772408652498,
      "learning_rate": 3.246078757930949e-05,
      "loss": 0.9302,
      "step": 910
    },
    {
      "epoch": 2.147887323943662,
      "grad_norm": 0.32823640884020344,
      "learning_rate": 3.21545857627259e-05,
      "loss": 0.8977,
      "step": 915
    },
    {
      "epoch": 2.15962441314554,
      "grad_norm": 0.2986661375144512,
      "learning_rate": 3.1851316462330086e-05,
      "loss": 0.9632,
      "step": 920
    },
    {
      "epoch": 2.171361502347418,
      "grad_norm": 0.3044796130319778,
      "learning_rate": 3.155102810875744e-05,
      "loss": 0.9244,
      "step": 925
    },
    {
      "epoch": 2.183098591549296,
      "grad_norm": 0.29391562913639346,
      "learning_rate": 3.125376865660053e-05,
      "loss": 0.9065,
      "step": 930
    },
    {
      "epoch": 2.1948356807511735,
      "grad_norm": 0.33464998182357986,
      "learning_rate": 3.095958557675115e-05,
      "loss": 1.0646,
      "step": 935
    },
    {
      "epoch": 2.2065727699530515,
      "grad_norm": 0.3097450215029285,
      "learning_rate": 3.0668525848819334e-05,
      "loss": 0.909,
      "step": 940
    },
    {
      "epoch": 2.2183098591549295,
      "grad_norm": 0.31619368356156724,
      "learning_rate": 3.0380635953630954e-05,
      "loss": 0.942,
      "step": 945
    },
    {
      "epoch": 2.2300469483568075,
      "grad_norm": 0.5488021620297451,
      "learning_rate": 3.009596186580501e-05,
      "loss": 0.9037,
      "step": 950
    },
    {
      "epoch": 2.2417840375586855,
      "grad_norm": 0.3007995564402941,
      "learning_rate": 2.981454904641166e-05,
      "loss": 0.841,
      "step": 955
    },
    {
      "epoch": 2.2535211267605635,
      "grad_norm": 1.0143926976429458,
      "learning_rate": 2.953644243571233e-05,
      "loss": 0.8931,
      "step": 960
    },
    {
      "epoch": 2.2652582159624415,
      "grad_norm": 0.32847505150062767,
      "learning_rate": 2.926168644598294e-05,
      "loss": 0.8939,
      "step": 965
    },
    {
      "epoch": 2.276995305164319,
      "grad_norm": 0.29868200806169454,
      "learning_rate": 2.899032495442154e-05,
      "loss": 0.877,
      "step": 970
    },
    {
      "epoch": 2.288732394366197,
      "grad_norm": 0.3247314434075676,
      "learning_rate": 2.8722401296141272e-05,
      "loss": 0.9554,
      "step": 975
    },
    {
      "epoch": 2.300469483568075,
      "grad_norm": 0.2963994274935962,
      "learning_rate": 2.845795825724998e-05,
      "loss": 0.8366,
      "step": 980
    },
    {
      "epoch": 2.312206572769953,
      "grad_norm": 0.3018384484277062,
      "learning_rate": 2.819703806801753e-05,
      "loss": 0.9772,
      "step": 985
    },
    {
      "epoch": 2.323943661971831,
      "grad_norm": 0.3138452072040482,
      "learning_rate": 2.7939682396131733e-05,
      "loss": 0.8705,
      "step": 990
    },
    {
      "epoch": 2.335680751173709,
      "grad_norm": 0.33495040627076267,
      "learning_rate": 2.7685932340044337e-05,
      "loss": 0.9398,
      "step": 995
    },
    {
      "epoch": 2.347417840375587,
      "grad_norm": 0.3426786719758913,
      "learning_rate": 2.743582842240771e-05,
      "loss": 0.9528,
      "step": 1000
    },
    {
      "epoch": 2.347417840375587,
      "eval_loss": 1.5564570426940918,
      "eval_runtime": 49.772,
      "eval_samples_per_second": 1.467,
      "eval_steps_per_second": 0.201,
      "step": 1000
    },
    {
      "epoch": 2.359154929577465,
      "grad_norm": 0.32715481854051526,
      "learning_rate": 2.7189410583603632e-05,
      "loss": 0.9294,
      "step": 1005
    },
    {
      "epoch": 2.370892018779343,
      "grad_norm": 0.31904903206727697,
      "learning_rate": 2.6946718175364948e-05,
      "loss": 0.9377,
      "step": 1010
    },
    {
      "epoch": 2.3826291079812205,
      "grad_norm": 0.3246876100683689,
      "learning_rate": 2.6707789954491374e-05,
      "loss": 0.9389,
      "step": 1015
    },
    {
      "epoch": 2.3943661971830985,
      "grad_norm": 0.33657822136933685,
      "learning_rate": 2.647266407666012e-05,
      "loss": 0.8748,
      "step": 1020
    },
    {
      "epoch": 2.4061032863849765,
      "grad_norm": 0.33075950831011175,
      "learning_rate": 2.62413780903327e-05,
      "loss": 0.9191,
      "step": 1025
    },
    {
      "epoch": 2.4178403755868545,
      "grad_norm": 0.3091789648897354,
      "learning_rate": 2.6013968930758587e-05,
      "loss": 0.8542,
      "step": 1030
    },
    {
      "epoch": 2.4295774647887325,
      "grad_norm": 0.3017557854181004,
      "learning_rate": 2.579047291407683e-05,
      "loss": 0.9691,
      "step": 1035
    },
    {
      "epoch": 2.4413145539906105,
      "grad_norm": 0.3315662930965317,
      "learning_rate": 2.557092573151657e-05,
      "loss": 0.9335,
      "step": 1040
    },
    {
      "epoch": 2.453051643192488,
      "grad_norm": 0.28405648990544513,
      "learning_rate": 2.5355362443697268e-05,
      "loss": 0.8608,
      "step": 1045
    },
    {
      "epoch": 2.464788732394366,
      "grad_norm": 0.3272098198440713,
      "learning_rate": 2.5143817475029763e-05,
      "loss": 0.945,
      "step": 1050
    },
    {
      "epoch": 2.476525821596244,
      "grad_norm": 15.473459249206623,
      "learning_rate": 2.4936324608218807e-05,
      "loss": 0.862,
      "step": 1055
    },
    {
      "epoch": 2.488262910798122,
      "grad_norm": 0.3214221685785962,
      "learning_rate": 2.4732916978868173e-05,
      "loss": 0.8372,
      "step": 1060
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.30527745666926165,
      "learning_rate": 2.453362707018903e-05,
      "loss": 0.9094,
      "step": 1065
    },
    {
      "epoch": 2.511737089201878,
      "grad_norm": 0.337431832474658,
      "learning_rate": 2.4338486707812528e-05,
      "loss": 0.8946,
      "step": 1070
    },
    {
      "epoch": 2.523474178403756,
      "grad_norm": 0.3322583740691652,
      "learning_rate": 2.414752705470744e-05,
      "loss": 0.949,
      "step": 1075
    },
    {
      "epoch": 2.535211267605634,
      "grad_norm": 0.3183558468618594,
      "learning_rate": 2.396077860620353e-05,
      "loss": 0.8969,
      "step": 1080
    },
    {
      "epoch": 2.546948356807512,
      "grad_norm": 0.3155273530496357,
      "learning_rate": 2.377827118512165e-05,
      "loss": 0.9395,
      "step": 1085
    },
    {
      "epoch": 2.5586854460093895,
      "grad_norm": 2.4170816896900025,
      "learning_rate": 2.3600033937011134e-05,
      "loss": 0.8337,
      "step": 1090
    },
    {
      "epoch": 2.5704225352112675,
      "grad_norm": 0.31607546736128544,
      "learning_rate": 2.342609532549545e-05,
      "loss": 0.856,
      "step": 1095
    },
    {
      "epoch": 2.5821596244131455,
      "grad_norm": 0.33297406422032433,
      "learning_rate": 2.3256483127726626e-05,
      "loss": 0.8664,
      "step": 1100
    },
    {
      "epoch": 2.5938967136150235,
      "grad_norm": 0.3237147629308357,
      "learning_rate": 2.309122442994944e-05,
      "loss": 0.8719,
      "step": 1105
    },
    {
      "epoch": 2.6056338028169015,
      "grad_norm": 0.3234191361238395,
      "learning_rate": 2.2930345623175854e-05,
      "loss": 0.9487,
      "step": 1110
    },
    {
      "epoch": 2.6173708920187795,
      "grad_norm": 0.3323839887038311,
      "learning_rate": 2.277387239897046e-05,
      "loss": 0.9493,
      "step": 1115
    },
    {
      "epoch": 2.629107981220657,
      "grad_norm": 0.32228378417361714,
      "learning_rate": 2.262182974534775e-05,
      "loss": 0.8866,
      "step": 1120
    },
    {
      "epoch": 2.640845070422535,
      "grad_norm": 0.3079696625324963,
      "learning_rate": 2.247424194278155e-05,
      "loss": 0.8054,
      "step": 1125
    },
    {
      "epoch": 2.652582159624413,
      "grad_norm": 0.3097789983846638,
      "learning_rate": 2.2331132560327617e-05,
      "loss": 0.9464,
      "step": 1130
    },
    {
      "epoch": 2.664319248826291,
      "grad_norm": 0.29487479541027356,
      "learning_rate": 2.2192524451859738e-05,
      "loss": 0.9063,
      "step": 1135
    },
    {
      "epoch": 2.676056338028169,
      "grad_norm": 0.33006135046655793,
      "learning_rate": 2.205843975242011e-05,
      "loss": 0.8962,
      "step": 1140
    },
    {
      "epoch": 2.687793427230047,
      "grad_norm": 0.3529330284339669,
      "learning_rate": 2.1928899874684446e-05,
      "loss": 0.9958,
      "step": 1145
    },
    {
      "epoch": 2.699530516431925,
      "grad_norm": 0.36979938140870244,
      "learning_rate": 2.1803925505542516e-05,
      "loss": 0.9466,
      "step": 1150
    },
    {
      "epoch": 2.711267605633803,
      "grad_norm": 0.30449528011099597,
      "learning_rate": 2.168353660279449e-05,
      "loss": 0.8422,
      "step": 1155
    },
    {
      "epoch": 2.723004694835681,
      "grad_norm": 0.32449892199883407,
      "learning_rate": 2.156775239196383e-05,
      "loss": 0.9596,
      "step": 1160
    },
    {
      "epoch": 2.7347417840375585,
      "grad_norm": 0.3082491729387956,
      "learning_rate": 2.145659136322706e-05,
      "loss": 0.8846,
      "step": 1165
    },
    {
      "epoch": 2.7464788732394365,
      "grad_norm": 0.3542469454397538,
      "learning_rate": 2.1350071268460934e-05,
      "loss": 0.9162,
      "step": 1170
    },
    {
      "epoch": 2.7582159624413145,
      "grad_norm": 0.35364030378555145,
      "learning_rate": 2.12482091184076e-05,
      "loss": 0.9548,
      "step": 1175
    },
    {
      "epoch": 2.7699530516431925,
      "grad_norm": 0.534630830850989,
      "learning_rate": 2.115102117995803e-05,
      "loss": 0.9527,
      "step": 1180
    },
    {
      "epoch": 2.7816901408450705,
      "grad_norm": 0.3638700943894012,
      "learning_rate": 2.1058522973554294e-05,
      "loss": 0.8624,
      "step": 1185
    },
    {
      "epoch": 2.7934272300469485,
      "grad_norm": 0.33210932308224567,
      "learning_rate": 2.0970729270711017e-05,
      "loss": 0.8602,
      "step": 1190
    },
    {
      "epoch": 2.805164319248826,
      "grad_norm": 0.3271576610317457,
      "learning_rate": 2.0887654091656426e-05,
      "loss": 0.8537,
      "step": 1195
    },
    {
      "epoch": 2.816901408450704,
      "grad_norm": 0.3445497480228464,
      "learning_rate": 2.0809310703093418e-05,
      "loss": 0.805,
      "step": 1200
    },
    {
      "epoch": 2.828638497652582,
      "grad_norm": 0.32705975575158863,
      "learning_rate": 2.073571161608091e-05,
      "loss": 0.955,
      "step": 1205
    },
    {
      "epoch": 2.84037558685446,
      "grad_norm": 0.34134786803880485,
      "learning_rate": 2.066686858403588e-05,
      "loss": 0.907,
      "step": 1210
    },
    {
      "epoch": 2.852112676056338,
      "grad_norm": 0.31253357487176564,
      "learning_rate": 2.0602792600856414e-05,
      "loss": 0.9051,
      "step": 1215
    },
    {
      "epoch": 2.863849765258216,
      "grad_norm": 0.3325144520719963,
      "learning_rate": 2.0543493899166034e-05,
      "loss": 0.9146,
      "step": 1220
    },
    {
      "epoch": 2.875586854460094,
      "grad_norm": 0.3293615881469167,
      "learning_rate": 2.04889819486796e-05,
      "loss": 0.8378,
      "step": 1225
    },
    {
      "epoch": 2.887323943661972,
      "grad_norm": 0.33386869136878455,
      "learning_rate": 2.0439265454691023e-05,
      "loss": 0.9188,
      "step": 1230
    },
    {
      "epoch": 2.89906103286385,
      "grad_norm": 0.32510298043861213,
      "learning_rate": 2.0394352356683095e-05,
      "loss": 0.9314,
      "step": 1235
    },
    {
      "epoch": 2.910798122065728,
      "grad_norm": 0.3244639716190284,
      "learning_rate": 2.0354249827059588e-05,
      "loss": 0.8556,
      "step": 1240
    },
    {
      "epoch": 2.9225352112676055,
      "grad_norm": 0.3442224853352613,
      "learning_rate": 2.0318964269999837e-05,
      "loss": 0.8086,
      "step": 1245
    },
    {
      "epoch": 2.9342723004694835,
      "grad_norm": 0.3402662522070222,
      "learning_rate": 2.0288501320436047e-05,
      "loss": 0.8865,
      "step": 1250
    },
    {
      "epoch": 2.936619718309859,
      "eval_loss": 1.5202196836471558,
      "eval_runtime": 49.8279,
      "eval_samples_per_second": 1.465,
      "eval_steps_per_second": 0.201,
      "step": 1251
    }
  ],
  "logging_steps": 5,
  "max_steps": 1278,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1297580487081984.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}