{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 50,
  "global_step": 213,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.014222222222222223,
      "grad_norm": 10.21201439945407,
      "learning_rate": 0.0,
      "loss": 0.9819135069847107,
      "step": 1
    },
    {
      "epoch": 0.028444444444444446,
      "grad_norm": 10.607308805754391,
      "learning_rate": 4.5454545454545457e-07,
      "loss": 0.992728054523468,
      "step": 2
    },
    {
      "epoch": 0.042666666666666665,
      "grad_norm": 8.137000657744727,
      "learning_rate": 9.090909090909091e-07,
      "loss": 0.88365638256073,
      "step": 3
    },
    {
      "epoch": 0.05688888888888889,
      "grad_norm": 5.015580274302158,
      "learning_rate": 1.3636363636363636e-06,
      "loss": 0.8155229091644287,
      "step": 4
    },
    {
      "epoch": 0.07111111111111111,
      "grad_norm": 3.513429105458463,
      "learning_rate": 1.8181818181818183e-06,
      "loss": 0.7148971557617188,
      "step": 5
    },
    {
      "epoch": 0.08533333333333333,
      "grad_norm": 2.890137665957255,
      "learning_rate": 2.2727272727272728e-06,
      "loss": 0.7388776540756226,
      "step": 6
    },
    {
      "epoch": 0.09955555555555555,
      "grad_norm": 2.6650307199214267,
      "learning_rate": 2.7272727272727272e-06,
      "loss": 0.6954053044319153,
      "step": 7
    },
    {
      "epoch": 0.11377777777777778,
      "grad_norm": 2.1027438786791453,
      "learning_rate": 3.181818181818182e-06,
      "loss": 0.6385300159454346,
      "step": 8
    },
    {
      "epoch": 0.128,
      "grad_norm": 1.9707186974912183,
      "learning_rate": 3.6363636363636366e-06,
      "loss": 0.5974307060241699,
      "step": 9
    },
    {
      "epoch": 0.14222222222222222,
      "grad_norm": 1.7892988232740288,
      "learning_rate": 4.0909090909090915e-06,
      "loss": 0.5823870897293091,
      "step": 10
    },
    {
      "epoch": 0.15644444444444444,
      "grad_norm": 1.8117663888596973,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 0.5595569610595703,
      "step": 11
    },
    {
      "epoch": 0.17066666666666666,
      "grad_norm": 1.6892201331978909,
      "learning_rate": 5e-06,
      "loss": 0.5447146892547607,
      "step": 12
    },
    {
      "epoch": 0.18488888888888888,
      "grad_norm": 1.5301037695464916,
      "learning_rate": 5.4545454545454545e-06,
      "loss": 0.5179443359375,
      "step": 13
    },
    {
      "epoch": 0.1991111111111111,
      "grad_norm": 1.2533624125991738,
      "learning_rate": 5.90909090909091e-06,
      "loss": 0.5241127014160156,
      "step": 14
    },
    {
      "epoch": 0.21333333333333335,
      "grad_norm": 1.2394418596122698,
      "learning_rate": 6.363636363636364e-06,
      "loss": 0.4842333197593689,
      "step": 15
    },
    {
      "epoch": 0.22755555555555557,
      "grad_norm": 1.1420735285454937,
      "learning_rate": 6.818181818181818e-06,
      "loss": 0.5132038593292236,
      "step": 16
    },
    {
      "epoch": 0.24177777777777779,
      "grad_norm": 1.0423494529114559,
      "learning_rate": 7.272727272727273e-06,
      "loss": 0.4917834401130676,
      "step": 17
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.9878207423648759,
      "learning_rate": 7.727272727272727e-06,
      "loss": 0.4813205897808075,
      "step": 18
    },
    {
      "epoch": 0.2702222222222222,
      "grad_norm": 0.9711754260021513,
      "learning_rate": 8.181818181818183e-06,
      "loss": 0.4883510172367096,
      "step": 19
    },
    {
      "epoch": 0.28444444444444444,
      "grad_norm": 0.9163050690645712,
      "learning_rate": 8.636363636363637e-06,
      "loss": 0.4871647357940674,
      "step": 20
    },
    {
      "epoch": 0.2986666666666667,
      "grad_norm": 0.8404612592234425,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.46865272521972656,
      "step": 21
    },
    {
      "epoch": 0.3128888888888889,
      "grad_norm": 0.8673877914303936,
      "learning_rate": 9.545454545454547e-06,
      "loss": 0.4601229727268219,
      "step": 22
    },
    {
      "epoch": 0.32711111111111113,
      "grad_norm": 0.8211864748561771,
      "learning_rate": 1e-05,
      "loss": 0.45921051502227783,
      "step": 23
    },
    {
      "epoch": 0.3413333333333333,
      "grad_norm": 0.8242615873645178,
      "learning_rate": 9.999323662872998e-06,
      "loss": 0.46318507194519043,
      "step": 24
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 0.7839407977142685,
      "learning_rate": 9.99729483446475e-06,
      "loss": 0.4633931815624237,
      "step": 25
    },
    {
      "epoch": 0.36977777777777776,
      "grad_norm": 0.7111762322751543,
      "learning_rate": 9.993914063644053e-06,
      "loss": 0.45212602615356445,
      "step": 26
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.718221639194032,
      "learning_rate": 9.989182265027232e-06,
      "loss": 0.44105663895606995,
      "step": 27
    },
    {
      "epoch": 0.3982222222222222,
      "grad_norm": 0.7040716895578701,
      "learning_rate": 9.98310071873072e-06,
      "loss": 0.4427238404750824,
      "step": 28
    },
    {
      "epoch": 0.41244444444444445,
      "grad_norm": 0.7489850512171834,
      "learning_rate": 9.975671070024741e-06,
      "loss": 0.443437397480011,
      "step": 29
    },
    {
      "epoch": 0.4266666666666667,
      "grad_norm": 0.7002493913122244,
      "learning_rate": 9.966895328888195e-06,
      "loss": 0.4632475674152374,
      "step": 30
    },
    {
      "epoch": 0.4408888888888889,
      "grad_norm": 0.6824040285805486,
      "learning_rate": 9.956775869464901e-06,
      "loss": 0.4354594349861145,
      "step": 31
    },
    {
      "epoch": 0.45511111111111113,
      "grad_norm": 0.6597936132766806,
      "learning_rate": 9.945315429421307e-06,
      "loss": 0.429914653301239,
      "step": 32
    },
    {
      "epoch": 0.4693333333333333,
      "grad_norm": 0.6352588389251546,
      "learning_rate": 9.932517109205849e-06,
      "loss": 0.43607252836227417,
      "step": 33
    },
    {
      "epoch": 0.48355555555555557,
      "grad_norm": 0.6921974704448327,
      "learning_rate": 9.918384371210178e-06,
      "loss": 0.45671695470809937,
      "step": 34
    },
    {
      "epoch": 0.49777777777777776,
      "grad_norm": 0.6491660902599252,
      "learning_rate": 9.902921038832456e-06,
      "loss": 0.4388338327407837,
      "step": 35
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.6096750069184391,
      "learning_rate": 9.886131295443003e-06,
      "loss": 0.4141705334186554,
      "step": 36
    },
    {
      "epoch": 0.5262222222222223,
      "grad_norm": 0.6503340768273712,
      "learning_rate": 9.868019683252543e-06,
      "loss": 0.4218006432056427,
      "step": 37
    },
    {
      "epoch": 0.5404444444444444,
      "grad_norm": 0.714483452396674,
      "learning_rate": 9.848591102083375e-06,
      "loss": 0.4457612931728363,
      "step": 38
    },
    {
      "epoch": 0.5546666666666666,
      "grad_norm": 0.6231964594158875,
      "learning_rate": 9.82785080804381e-06,
      "loss": 0.4331759214401245,
      "step": 39
    },
    {
      "epoch": 0.5688888888888889,
      "grad_norm": 0.6572593116495337,
      "learning_rate": 9.805804412106197e-06,
      "loss": 0.4507564902305603,
      "step": 40
    },
    {
      "epoch": 0.5831111111111111,
      "grad_norm": 0.673487925488406,
      "learning_rate": 9.782457878588977e-06,
      "loss": 0.4254251718521118,
      "step": 41
    },
    {
      "epoch": 0.5973333333333334,
      "grad_norm": 0.5846423543296639,
      "learning_rate": 9.75781752354311e-06,
      "loss": 0.4125726819038391,
      "step": 42
    },
    {
      "epoch": 0.6115555555555555,
      "grad_norm": 0.598724301403679,
      "learning_rate": 9.731890013043367e-06,
      "loss": 0.4325122535228729,
      "step": 43
    },
    {
      "epoch": 0.6257777777777778,
      "grad_norm": 0.6218044129681044,
      "learning_rate": 9.704682361384941e-06,
      "loss": 0.4131222367286682,
      "step": 44
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.617001419027673,
      "learning_rate": 9.676201929185809e-06,
      "loss": 0.42036178708076477,
      "step": 45
    },
    {
      "epoch": 0.6542222222222223,
      "grad_norm": 0.5809155918543004,
      "learning_rate": 9.646456421395447e-06,
      "loss": 0.4182787537574768,
      "step": 46
    },
    {
      "epoch": 0.6684444444444444,
      "grad_norm": 0.5850555211488232,
      "learning_rate": 9.615453885210368e-06,
      "loss": 0.4234634041786194,
      "step": 47
    },
    {
      "epoch": 0.6826666666666666,
      "grad_norm": 0.5511675603941021,
      "learning_rate": 9.583202707897075e-06,
      "loss": 0.40563416481018066,
      "step": 48
    },
    {
      "epoch": 0.6968888888888889,
      "grad_norm": 0.6111190985848377,
      "learning_rate": 9.549711614523007e-06,
      "loss": 0.3980520963668823,
      "step": 49
    },
    {
      "epoch": 0.7111111111111111,
      "grad_norm": 0.5736293779202373,
      "learning_rate": 9.514989665596114e-06,
      "loss": 0.4036375880241394,
      "step": 50
    },
    {
      "epoch": 0.7111111111111111,
      "eval_loss": 0.41823333501815796,
      "eval_runtime": 28.2382,
      "eval_samples_per_second": 7.083,
      "eval_steps_per_second": 1.771,
      "step": 50
    },
    {
      "epoch": 0.7253333333333334,
      "grad_norm": 0.5785272249494463,
      "learning_rate": 9.479046254613673e-06,
      "loss": 0.41144663095474243,
      "step": 51
    },
    {
      "epoch": 0.7395555555555555,
      "grad_norm": 0.5517717884994647,
      "learning_rate": 9.441891105521005e-06,
      "loss": 0.38760095834732056,
      "step": 52
    },
    {
      "epoch": 0.7537777777777778,
      "grad_norm": 0.5648730368507402,
      "learning_rate": 9.40353427008083e-06,
      "loss": 0.4036053717136383,
      "step": 53
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.6019960018903892,
      "learning_rate": 9.3639861251539e-06,
      "loss": 0.39992809295654297,
      "step": 54
    },
    {
      "epoch": 0.7822222222222223,
      "grad_norm": 0.5943571857589287,
      "learning_rate": 9.323257369891702e-06,
      "loss": 0.420696496963501,
      "step": 55
    },
    {
      "epoch": 0.7964444444444444,
      "grad_norm": 0.5800482521740772,
      "learning_rate": 9.281359022841966e-06,
      "loss": 0.41333717107772827,
      "step": 56
    },
    {
      "epoch": 0.8106666666666666,
      "grad_norm": 0.5896438771006596,
      "learning_rate": 9.238302418967757e-06,
      "loss": 0.41053399443626404,
      "step": 57
    },
    {
      "epoch": 0.8248888888888889,
      "grad_norm": 0.5828843912613246,
      "learning_rate": 9.194099206580981e-06,
      "loss": 0.4167237877845764,
      "step": 58
    },
    {
      "epoch": 0.8391111111111111,
      "grad_norm": 0.636505663941606,
      "learning_rate": 9.14876134419111e-06,
      "loss": 0.4126594066619873,
      "step": 59
    },
    {
      "epoch": 0.8533333333333334,
      "grad_norm": 0.5954156377152778,
      "learning_rate": 9.102301097269974e-06,
      "loss": 0.40986335277557373,
      "step": 60
    },
    {
      "epoch": 0.8675555555555555,
      "grad_norm": 0.5663259824764864,
      "learning_rate": 9.05473103493355e-06,
      "loss": 0.39707833528518677,
      "step": 61
    },
    {
      "epoch": 0.8817777777777778,
      "grad_norm": 0.6233708113516392,
      "learning_rate": 9.006064026541549e-06,
      "loss": 0.4146142601966858,
      "step": 62
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.583882066947991,
      "learning_rate": 8.956313238215824e-06,
      "loss": 0.3761865794658661,
      "step": 63
    },
    {
      "epoch": 0.9102222222222223,
      "grad_norm": 0.559228243852746,
      "learning_rate": 8.905492129278478e-06,
      "loss": 0.397402286529541,
      "step": 64
    },
    {
      "epoch": 0.9244444444444444,
      "grad_norm": 0.5630501047235328,
      "learning_rate": 8.85361444861063e-06,
      "loss": 0.40299445390701294,
      "step": 65
    },
    {
      "epoch": 0.9386666666666666,
      "grad_norm": 0.5778346041245924,
      "learning_rate": 8.800694230932885e-06,
      "loss": 0.4066828489303589,
      "step": 66
    },
    {
      "epoch": 0.9528888888888889,
      "grad_norm": 0.5735220314699014,
      "learning_rate": 8.74674579300843e-06,
      "loss": 0.39874544739723206,
      "step": 67
    },
    {
      "epoch": 0.9671111111111111,
      "grad_norm": 0.5774583972046927,
      "learning_rate": 8.691783729769874e-06,
      "loss": 0.38473668694496155,
      "step": 68
    },
    {
      "epoch": 0.9813333333333333,
      "grad_norm": 0.6169263233631689,
      "learning_rate": 8.635822910370793e-06,
      "loss": 0.4083672761917114,
      "step": 69
    },
    {
      "epoch": 0.9955555555555555,
      "grad_norm": 0.5815111675806847,
      "learning_rate": 8.578878474163115e-06,
      "loss": 0.3794967234134674,
      "step": 70
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.9420572347818393,
      "learning_rate": 8.520965826601394e-06,
      "loss": 0.40614083409309387,
      "step": 71
    },
    {
      "epoch": 1.0142222222222221,
      "grad_norm": 0.6281039360207515,
      "learning_rate": 8.462100635075097e-06,
      "loss": 0.36958712339401245,
      "step": 72
    },
    {
      "epoch": 1.0284444444444445,
      "grad_norm": 0.6233244816927183,
      "learning_rate": 8.40229882467003e-06,
      "loss": 0.3494463562965393,
      "step": 73
    },
    {
      "epoch": 1.0426666666666666,
      "grad_norm": 0.5621513032586688,
      "learning_rate": 8.341576573860049e-06,
      "loss": 0.3302980065345764,
      "step": 74
    },
    {
      "epoch": 1.056888888888889,
      "grad_norm": 0.5754725580485192,
      "learning_rate": 8.279950310130218e-06,
      "loss": 0.35172200202941895,
      "step": 75
    },
    {
      "epoch": 1.0711111111111111,
      "grad_norm": 0.5740681468075599,
      "learning_rate": 8.2174367055326e-06,
      "loss": 0.3394201993942261,
      "step": 76
    },
    {
      "epoch": 1.0853333333333333,
      "grad_norm": 0.590460680669288,
      "learning_rate": 8.154052672175888e-06,
      "loss": 0.33312541246414185,
      "step": 77
    },
    {
      "epoch": 1.0995555555555556,
      "grad_norm": 0.5842296281151235,
      "learning_rate": 8.08981535765009e-06,
      "loss": 0.35658037662506104,
      "step": 78
    },
    {
      "epoch": 1.1137777777777778,
      "grad_norm": 0.6088529350096018,
      "learning_rate": 8.024742140387506e-06,
      "loss": 0.3449564278125763,
      "step": 79
    },
    {
      "epoch": 1.1280000000000001,
      "grad_norm": 0.5744146541082055,
      "learning_rate": 7.95885062496126e-06,
      "loss": 0.3357444107532501,
      "step": 80
    },
    {
      "epoch": 1.1422222222222222,
      "grad_norm": 0.5909574288907287,
      "learning_rate": 7.892158637322647e-06,
      "loss": 0.3170369267463684,
      "step": 81
    },
    {
      "epoch": 1.1564444444444444,
      "grad_norm": 0.6136816332174048,
      "learning_rate": 7.824684219978591e-06,
      "loss": 0.3318628668785095,
      "step": 82
    },
    {
      "epoch": 1.1706666666666667,
      "grad_norm": 0.5568422531541258,
      "learning_rate": 7.756445627110523e-06,
      "loss": 0.34082144498825073,
      "step": 83
    },
    {
      "epoch": 1.1848888888888889,
      "grad_norm": 0.6082214692469516,
      "learning_rate": 7.68746131963598e-06,
      "loss": 0.3453400135040283,
      "step": 84
    },
    {
      "epoch": 1.199111111111111,
      "grad_norm": 0.5993208838733233,
      "learning_rate": 7.6177499602143e-06,
      "loss": 0.3420029580593109,
      "step": 85
    },
    {
      "epoch": 1.2133333333333334,
      "grad_norm": 0.5493880460077399,
      "learning_rate": 7.547330408197695e-06,
      "loss": 0.3330039381980896,
      "step": 86
    },
    {
      "epoch": 1.2275555555555555,
      "grad_norm": 0.6151717226672565,
      "learning_rate": 7.476221714529167e-06,
      "loss": 0.3428732454776764,
      "step": 87
    },
    {
      "epoch": 1.2417777777777779,
      "grad_norm": 0.6359199055571157,
      "learning_rate": 7.404443116588548e-06,
      "loss": 0.3575543761253357,
      "step": 88
    },
    {
      "epoch": 1.256,
      "grad_norm": 0.5764223898323285,
      "learning_rate": 7.332014032988123e-06,
      "loss": 0.3414611518383026,
      "step": 89
    },
    {
      "epoch": 1.2702222222222221,
      "grad_norm": 0.5701344890989715,
      "learning_rate": 7.2589540583192165e-06,
      "loss": 0.3387766480445862,
      "step": 90
    },
    {
      "epoch": 1.2844444444444445,
      "grad_norm": 0.5603334736031269,
      "learning_rate": 7.185282957851175e-06,
      "loss": 0.32765448093414307,
      "step": 91
    },
    {
      "epoch": 1.2986666666666666,
      "grad_norm": 0.5637583167829048,
      "learning_rate": 7.111020662184174e-06,
      "loss": 0.3432917296886444,
      "step": 92
    },
    {
      "epoch": 1.3128888888888888,
      "grad_norm": 0.5811322458912195,
      "learning_rate": 7.036187261857289e-06,
      "loss": 0.3334134817123413,
      "step": 93
    },
    {
      "epoch": 1.3271111111111111,
      "grad_norm": 0.6069998855060962,
      "learning_rate": 6.960803001913315e-06,
      "loss": 0.3528696894645691,
      "step": 94
    },
    {
      "epoch": 1.3413333333333333,
      "grad_norm": 0.5846537103757758,
      "learning_rate": 6.884888276421766e-06,
      "loss": 0.3428453207015991,
      "step": 95
    },
    {
      "epoch": 1.3555555555555556,
      "grad_norm": 0.5679194938574699,
      "learning_rate": 6.8084636229615786e-06,
      "loss": 0.33261382579803467,
      "step": 96
    },
    {
      "epoch": 1.3697777777777778,
      "grad_norm": 0.5611361976965846,
      "learning_rate": 6.731549717064975e-06,
      "loss": 0.33263495564460754,
      "step": 97
    },
    {
      "epoch": 1.384,
      "grad_norm": 0.5958752887464758,
      "learning_rate": 6.654167366624009e-06,
      "loss": 0.3409455418586731,
      "step": 98
    },
    {
      "epoch": 1.3982222222222223,
      "grad_norm": 0.5821747639433966,
      "learning_rate": 6.576337506261314e-06,
      "loss": 0.34175097942352295,
      "step": 99
    },
    {
      "epoch": 1.4124444444444444,
      "grad_norm": 0.5474416142301471,
      "learning_rate": 6.498081191666549e-06,
      "loss": 0.3422796130180359,
      "step": 100
    },
    {
      "epoch": 1.4124444444444444,
      "eval_loss": 0.39906859397888184,
      "eval_runtime": 28.0052,
      "eval_samples_per_second": 7.142,
      "eval_steps_per_second": 1.785,
      "step": 100
    },
    {
      "epoch": 1.4266666666666667,
      "grad_norm": 0.5470999547878086,
      "learning_rate": 6.419419593900109e-06,
      "loss": 0.3311368227005005,
      "step": 101
    },
    {
      "epoch": 1.4408888888888889,
      "grad_norm": 0.5417978810198306,
      "learning_rate": 6.340373993665607e-06,
      "loss": 0.32347068190574646,
      "step": 102
    },
    {
      "epoch": 1.455111111111111,
      "grad_norm": 0.5408071080127892,
      "learning_rate": 6.2609657755527135e-06,
      "loss": 0.32194703817367554,
      "step": 103
    },
    {
      "epoch": 1.4693333333333334,
      "grad_norm": 0.5610127885481652,
      "learning_rate": 6.1812164222518626e-06,
      "loss": 0.3477523922920227,
      "step": 104
    },
    {
      "epoch": 1.4835555555555555,
      "grad_norm": 0.6035742435126064,
      "learning_rate": 6.101147508742456e-06,
      "loss": 0.33583635091781616,
      "step": 105
    },
    {
      "epoch": 1.4977777777777779,
      "grad_norm": 0.5482179611450086,
      "learning_rate": 6.020780696456059e-06,
      "loss": 0.3305099904537201,
      "step": 106
    },
    {
      "epoch": 1.512,
      "grad_norm": 0.5549975206757557,
      "learning_rate": 5.940137727416247e-06,
      "loss": 0.3392476439476013,
      "step": 107
    },
    {
      "epoch": 1.5262222222222221,
      "grad_norm": 0.5452500081271687,
      "learning_rate": 5.859240418356614e-06,
      "loss": 0.3281456232070923,
      "step": 108
    },
    {
      "epoch": 1.5404444444444443,
      "grad_norm": 0.569343792950831,
      "learning_rate": 5.778110654818602e-06,
      "loss": 0.3557584285736084,
      "step": 109
    },
    {
      "epoch": 1.5546666666666666,
      "grad_norm": 0.5609073891735213,
      "learning_rate": 5.696770385230679e-06,
      "loss": 0.3375227153301239,
      "step": 110
    },
    {
      "epoch": 1.568888888888889,
      "grad_norm": 0.5979889868516021,
      "learning_rate": 5.615241614970546e-06,
      "loss": 0.33440524339675903,
      "step": 111
    },
    {
      "epoch": 1.5831111111111111,
      "grad_norm": 0.5752429374657604,
      "learning_rate": 5.533546400411899e-06,
      "loss": 0.3354341983795166,
      "step": 112
    },
    {
      "epoch": 1.5973333333333333,
      "grad_norm": 0.5583834084640693,
      "learning_rate": 5.4517068429574215e-06,
      "loss": 0.34512901306152344,
      "step": 113
    },
    {
      "epoch": 1.6115555555555554,
      "grad_norm": 0.5366739842708482,
      "learning_rate": 5.3697450830595775e-06,
      "loss": 0.3282182812690735,
      "step": 114
    },
    {
      "epoch": 1.6257777777777778,
      "grad_norm": 0.5923091160755782,
      "learning_rate": 5.287683294230855e-06,
      "loss": 0.3325689136981964,
      "step": 115
    },
    {
      "epoch": 1.6400000000000001,
      "grad_norm": 0.5603788239590529,
      "learning_rate": 5.20554367704505e-06,
      "loss": 0.3167092204093933,
      "step": 116
    },
    {
      "epoch": 1.6542222222222223,
      "grad_norm": 0.5675689042120237,
      "learning_rate": 5.123348453131242e-06,
      "loss": 0.3419051766395569,
      "step": 117
    },
    {
      "epoch": 1.6684444444444444,
      "grad_norm": 0.5456542923932385,
      "learning_rate": 5.041119859162068e-06,
      "loss": 0.333318293094635,
      "step": 118
    },
    {
      "epoch": 1.6826666666666665,
      "grad_norm": 0.570188952777049,
      "learning_rate": 4.958880140837934e-06,
      "loss": 0.33905792236328125,
      "step": 119
    },
    {
      "epoch": 1.696888888888889,
      "grad_norm": 0.5534588670481234,
      "learning_rate": 4.876651546868759e-06,
      "loss": 0.32569897174835205,
      "step": 120
    },
    {
      "epoch": 1.7111111111111112,
      "grad_norm": 0.5546097096941289,
      "learning_rate": 4.794456322954953e-06,
      "loss": 0.3532940745353699,
      "step": 121
    },
    {
      "epoch": 1.7253333333333334,
      "grad_norm": 0.5857002004886875,
      "learning_rate": 4.7123167057691446e-06,
      "loss": 0.35066288709640503,
      "step": 122
    },
    {
      "epoch": 1.7395555555555555,
      "grad_norm": 0.5351298499201909,
      "learning_rate": 4.630254916940424e-06,
      "loss": 0.35048598051071167,
      "step": 123
    },
    {
      "epoch": 1.7537777777777777,
      "grad_norm": 0.5572795301130913,
      "learning_rate": 4.548293157042581e-06,
      "loss": 0.3389701843261719,
      "step": 124
    },
    {
      "epoch": 1.768,
      "grad_norm": 0.5503673971003669,
      "learning_rate": 4.466453599588103e-06,
      "loss": 0.3477008044719696,
      "step": 125
    },
    {
      "epoch": 1.7822222222222224,
      "grad_norm": 0.5485965256926039,
      "learning_rate": 4.384758385029457e-06,
      "loss": 0.3188989758491516,
      "step": 126
    },
    {
      "epoch": 1.7964444444444445,
      "grad_norm": 0.5566326955710196,
      "learning_rate": 4.3032296147693225e-06,
      "loss": 0.3296913504600525,
      "step": 127
    },
    {
      "epoch": 1.8106666666666666,
      "grad_norm": 0.5592929839051507,
      "learning_rate": 4.2218893451814005e-06,
      "loss": 0.3278958797454834,
      "step": 128
    },
    {
      "epoch": 1.8248888888888888,
      "grad_norm": 0.5372208136274655,
      "learning_rate": 4.140759581643386e-06,
      "loss": 0.3338126838207245,
      "step": 129
    },
    {
      "epoch": 1.8391111111111111,
      "grad_norm": 0.5506377641318588,
      "learning_rate": 4.059862272583755e-06,
      "loss": 0.33511388301849365,
      "step": 130
    },
    {
      "epoch": 1.8533333333333335,
      "grad_norm": 0.5509814672788427,
      "learning_rate": 3.979219303543942e-06,
      "loss": 0.3190721869468689,
      "step": 131
    },
    {
      "epoch": 1.8675555555555556,
      "grad_norm": 0.5437024683532695,
      "learning_rate": 3.898852491257547e-06,
      "loss": 0.3397029638290405,
      "step": 132
    },
    {
      "epoch": 1.8817777777777778,
      "grad_norm": 0.5577533834862016,
      "learning_rate": 3.818783577748138e-06,
      "loss": 0.3368285894393921,
      "step": 133
    },
    {
      "epoch": 1.896,
      "grad_norm": 0.5654570173541145,
      "learning_rate": 3.7390342244472886e-06,
      "loss": 0.34375038743019104,
      "step": 134
    },
    {
      "epoch": 1.9102222222222223,
      "grad_norm": 0.5435541441710147,
      "learning_rate": 3.659626006334395e-06,
      "loss": 0.32158786058425903,
      "step": 135
    },
    {
      "epoch": 1.9244444444444444,
      "grad_norm": 0.5517597326060792,
      "learning_rate": 3.5805804060998926e-06,
      "loss": 0.32734596729278564,
      "step": 136
    },
    {
      "epoch": 1.9386666666666668,
      "grad_norm": 0.5529004589065376,
      "learning_rate": 3.501918808333453e-06,
      "loss": 0.3495728671550751,
      "step": 137
    },
    {
      "epoch": 1.952888888888889,
      "grad_norm": 0.5497949696155701,
      "learning_rate": 3.4236624937386874e-06,
      "loss": 0.3545406758785248,
      "step": 138
    },
    {
      "epoch": 1.967111111111111,
      "grad_norm": 0.50810534397043,
      "learning_rate": 3.3458326333759927e-06,
      "loss": 0.3283476233482361,
      "step": 139
    },
    {
      "epoch": 1.9813333333333332,
      "grad_norm": 0.5410931225650443,
      "learning_rate": 3.268450282935026e-06,
      "loss": 0.3362443447113037,
      "step": 140
    },
    {
      "epoch": 1.9955555555555555,
      "grad_norm": 0.5526133533239165,
      "learning_rate": 3.1915363770384223e-06,
      "loss": 0.33009064197540283,
      "step": 141
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.8788381822845209,
      "learning_rate": 3.1151117235782346e-06,
      "loss": 0.34658753871917725,
      "step": 142
    },
    {
      "epoch": 2.014222222222222,
      "grad_norm": 0.6321149682877057,
      "learning_rate": 3.0391969980866874e-06,
      "loss": 0.2953590750694275,
      "step": 143
    },
    {
      "epoch": 2.0284444444444443,
      "grad_norm": 0.5718010363312446,
      "learning_rate": 2.963812738142713e-06,
      "loss": 0.28190308809280396,
      "step": 144
    },
    {
      "epoch": 2.042666666666667,
      "grad_norm": 0.5860488240075866,
      "learning_rate": 2.8889793378158284e-06,
      "loss": 0.29579877853393555,
      "step": 145
    },
    {
      "epoch": 2.056888888888889,
      "grad_norm": 0.5087999304214683,
      "learning_rate": 2.814717042148827e-06,
      "loss": 0.2849843502044678,
      "step": 146
    },
    {
      "epoch": 2.071111111111111,
      "grad_norm": 0.5097817184428313,
      "learning_rate": 2.7410459416807856e-06,
      "loss": 0.2797221541404724,
      "step": 147
    },
    {
      "epoch": 2.0853333333333333,
      "grad_norm": 0.5476441083873779,
      "learning_rate": 2.6679859670118785e-06,
      "loss": 0.2898455262184143,
      "step": 148
    },
    {
      "epoch": 2.0995555555555554,
      "grad_norm": 0.5438939155777798,
      "learning_rate": 2.5955568834114523e-06,
      "loss": 0.27163612842559814,
      "step": 149
    },
    {
      "epoch": 2.113777777777778,
      "grad_norm": 0.5562699539267528,
      "learning_rate": 2.523778285470835e-06,
      "loss": 0.28451889753341675,
      "step": 150
    },
    {
      "epoch": 2.113777777777778,
      "eval_loss": 0.4041806757450104,
      "eval_runtime": 27.9066,
      "eval_samples_per_second": 7.167,
      "eval_steps_per_second": 1.792,
      "step": 150
    },
    {
      "epoch": 2.128,
      "grad_norm": 0.5511756829107596,
      "learning_rate": 2.452669591802307e-06,
      "loss": 0.2696727514266968,
      "step": 151
    },
    {
      "epoch": 2.1422222222222222,
      "grad_norm": 0.5230762023952198,
      "learning_rate": 2.3822500397857016e-06,
      "loss": 0.2786106467247009,
      "step": 152
    },
    {
      "epoch": 2.1564444444444444,
      "grad_norm": 0.5376649924382556,
      "learning_rate": 2.3125386803640188e-06,
      "loss": 0.29131805896759033,
      "step": 153
    },
    {
      "epoch": 2.1706666666666665,
      "grad_norm": 0.5176483788211893,
      "learning_rate": 2.243554372889479e-06,
      "loss": 0.2755666971206665,
      "step": 154
    },
    {
      "epoch": 2.1848888888888887,
      "grad_norm": 0.5045947338221831,
      "learning_rate": 2.1753157800214107e-06,
      "loss": 0.27331122756004333,
      "step": 155
    },
    {
      "epoch": 2.1991111111111112,
      "grad_norm": 0.5193109465079893,
      "learning_rate": 2.1078413626773547e-06,
      "loss": 0.28715455532073975,
      "step": 156
    },
    {
      "epoch": 2.2133333333333334,
      "grad_norm": 0.5212127411265028,
      "learning_rate": 2.0411493750387423e-06,
      "loss": 0.26138654351234436,
      "step": 157
    },
    {
      "epoch": 2.2275555555555555,
      "grad_norm": 0.5448628823500382,
      "learning_rate": 1.9752578596124955e-06,
      "loss": 0.29015523195266724,
      "step": 158
    },
    {
      "epoch": 2.2417777777777776,
      "grad_norm": 0.5501585558078743,
      "learning_rate": 1.9101846423499113e-06,
      "loss": 0.2733340561389923,
      "step": 159
    },
    {
      "epoch": 2.2560000000000002,
      "grad_norm": 0.5315657420322433,
      "learning_rate": 1.8459473278241125e-06,
      "loss": 0.2742033302783966,
      "step": 160
    },
    {
      "epoch": 2.2702222222222224,
      "grad_norm": 0.4902572304429622,
      "learning_rate": 1.7825632944674016e-06,
      "loss": 0.2752198874950409,
      "step": 161
    },
    {
      "epoch": 2.2844444444444445,
      "grad_norm": 0.534591959936222,
      "learning_rate": 1.7200496898697832e-06,
      "loss": 0.28458085656166077,
      "step": 162
    },
    {
      "epoch": 2.2986666666666666,
      "grad_norm": 0.5106425495762753,
      "learning_rate": 1.6584234261399535e-06,
      "loss": 0.277709424495697,
      "step": 163
    },
    {
      "epoch": 2.3128888888888888,
      "grad_norm": 0.5539379653737809,
      "learning_rate": 1.5977011753299726e-06,
      "loss": 0.2714066505432129,
      "step": 164
    },
    {
      "epoch": 2.327111111111111,
      "grad_norm": 0.5579822508546088,
      "learning_rate": 1.5378993649249053e-06,
      "loss": 0.2929142117500305,
      "step": 165
    },
    {
      "epoch": 2.3413333333333335,
      "grad_norm": 0.5128512129047608,
      "learning_rate": 1.4790341733986085e-06,
      "loss": 0.274400919675827,
      "step": 166
    },
    {
      "epoch": 2.3555555555555556,
      "grad_norm": 0.5180316098734932,
      "learning_rate": 1.4211215258368866e-06,
      "loss": 0.2633841633796692,
      "step": 167
    },
    {
      "epoch": 2.3697777777777778,
      "grad_norm": 0.5226866839878248,
      "learning_rate": 1.3641770896292083e-06,
      "loss": 0.271608829498291,
      "step": 168
    },
    {
      "epoch": 2.384,
      "grad_norm": 0.5213295087948763,
      "learning_rate": 1.3082162702301276e-06,
      "loss": 0.2772142291069031,
      "step": 169
    },
    {
      "epoch": 2.398222222222222,
      "grad_norm": 0.5200525913463248,
      "learning_rate": 1.2532542069915722e-06,
      "loss": 0.2833596467971802,
      "step": 170
    },
    {
      "epoch": 2.4124444444444446,
      "grad_norm": 0.5158405462335311,
      "learning_rate": 1.1993057690671174e-06,
      "loss": 0.27554380893707275,
      "step": 171
    },
    {
      "epoch": 2.4266666666666667,
      "grad_norm": 0.5070004563853152,
      "learning_rate": 1.1463855513893695e-06,
      "loss": 0.291002094745636,
      "step": 172
    },
    {
      "epoch": 2.440888888888889,
      "grad_norm": 0.5227423053894303,
      "learning_rate": 1.0945078707215224e-06,
      "loss": 0.2779809832572937,
      "step": 173
    },
    {
      "epoch": 2.455111111111111,
      "grad_norm": 0.530364411528684,
      "learning_rate": 1.0436867617841768e-06,
      "loss": 0.2763213515281677,
      "step": 174
    },
    {
      "epoch": 2.469333333333333,
      "grad_norm": 0.5210228993927588,
      "learning_rate": 9.939359734584552e-07,
      "loss": 0.28753769397735596,
      "step": 175
    },
    {
      "epoch": 2.4835555555555557,
      "grad_norm": 0.5214828400537481,
      "learning_rate": 9.452689650664515e-07,
      "loss": 0.2837107181549072,
      "step": 176
    },
    {
      "epoch": 2.497777777777778,
      "grad_norm": 0.5036225380576258,
      "learning_rate": 8.976989027300265e-07,
      "loss": 0.2816150188446045,
      "step": 177
    },
    {
      "epoch": 2.512,
      "grad_norm": 0.5488666037756793,
      "learning_rate": 8.512386558088919e-07,
      "loss": 0.2889126241207123,
      "step": 178
    },
    {
      "epoch": 2.526222222222222,
      "grad_norm": 0.5181437270813625,
      "learning_rate": 8.059007934190194e-07,
      "loss": 0.2788556218147278,
      "step": 179
    },
    {
      "epoch": 2.5404444444444443,
      "grad_norm": 0.5160804538602589,
      "learning_rate": 7.61697581032243e-07,
      "loss": 0.29226237535476685,
      "step": 180
    },
    {
      "epoch": 2.554666666666667,
      "grad_norm": 0.5105467594147919,
      "learning_rate": 7.186409771580355e-07,
      "loss": 0.2915460765361786,
      "step": 181
    },
    {
      "epoch": 2.568888888888889,
      "grad_norm": 0.49398572364663706,
      "learning_rate": 6.76742630108298e-07,
      "loss": 0.2790933847427368,
      "step": 182
    },
    {
      "epoch": 2.583111111111111,
      "grad_norm": 1.14219599430935e+31,
      "learning_rate": 6.360138748461015e-07,
      "loss": 0.27808326482772827,
      "step": 183
    },
    {
      "epoch": 2.5973333333333333,
      "grad_norm": 0.4983345243585802,
      "learning_rate": 5.964657299191712e-07,
      "loss": 0.2690117359161377,
      "step": 184
    },
    {
      "epoch": 2.6115555555555554,
      "grad_norm": 0.5069008538717243,
      "learning_rate": 5.581088944789953e-07,
      "loss": 0.27599871158599854,
      "step": 185
    },
    {
      "epoch": 2.6257777777777775,
      "grad_norm": 0.521476062094092,
      "learning_rate": 5.20953745386329e-07,
      "loss": 0.2919716238975525,
      "step": 186
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.510790594456331,
      "learning_rate": 4.850103344038853e-07,
      "loss": 0.28502631187438965,
      "step": 187
    },
    {
      "epoch": 2.6542222222222223,
      "grad_norm": 0.49201617380837526,
      "learning_rate": 4.502883854769935e-07,
      "loss": 0.26374831795692444,
      "step": 188
    },
    {
      "epoch": 2.6684444444444444,
      "grad_norm": 0.5211970358874106,
      "learning_rate": 4.167972921029262e-07,
      "loss": 0.28197193145751953,
      "step": 189
    },
    {
      "epoch": 2.6826666666666665,
      "grad_norm": 0.5370264524554923,
      "learning_rate": 3.8454611478963235e-07,
      "loss": 0.2938516139984131,
      "step": 190
    },
    {
      "epoch": 2.696888888888889,
      "grad_norm": 0.5232192741343314,
      "learning_rate": 3.535435786045538e-07,
      "loss": 0.27452725172042847,
      "step": 191
    },
    {
      "epoch": 2.7111111111111112,
      "grad_norm": 0.5024016728659283,
      "learning_rate": 3.237980708141919e-07,
      "loss": 0.2790898084640503,
      "step": 192
    },
    {
      "epoch": 2.7253333333333334,
      "grad_norm": 0.5309491933991259,
      "learning_rate": 2.9531763861505967e-07,
      "loss": 0.2802097201347351,
      "step": 193
    },
    {
      "epoch": 2.7395555555555555,
      "grad_norm": 0.5117668670583797,
      "learning_rate": 2.681099869566328e-07,
      "loss": 0.2703718841075897,
      "step": 194
    },
    {
      "epoch": 2.7537777777777777,
      "grad_norm": 0.5093979232137983,
      "learning_rate": 2.4218247645689306e-07,
      "loss": 0.2852793335914612,
      "step": 195
    },
    {
      "epoch": 2.768,
      "grad_norm": 0.49789356969379145,
      "learning_rate": 2.175421214110235e-07,
      "loss": 0.27227064967155457,
      "step": 196
    },
    {
      "epoch": 2.7822222222222224,
      "grad_norm": 0.5255831952638869,
      "learning_rate": 1.941955878938029e-07,
      "loss": 0.28222954273223877,
      "step": 197
    },
    {
      "epoch": 2.7964444444444445,
      "grad_norm": 0.5060979808451155,
      "learning_rate": 1.721491919561913e-07,
      "loss": 0.2887798249721527,
      "step": 198
    },
    {
      "epoch": 2.8106666666666666,
      "grad_norm": 0.5024415391860155,
      "learning_rate": 1.514088979166256e-07,
      "loss": 0.2730322480201721,
      "step": 199
    },
    {
      "epoch": 2.824888888888889,
      "grad_norm": 0.5071775702854565,
      "learning_rate": 1.3198031674745814e-07,
      "loss": 0.28655844926834106,
      "step": 200
    },
    {
      "epoch": 2.824888888888889,
      "eval_loss": 0.3998975455760956,
      "eval_runtime": 27.4547,
      "eval_samples_per_second": 7.285,
      "eval_steps_per_second": 1.821,
      "step": 200
    },
    {
      "epoch": 2.8391111111111114,
      "grad_norm": 0.5098506733931571,
      "learning_rate": 1.138687045569975e-07,
      "loss": 0.28616365790367126,
      "step": 201
    },
    {
      "epoch": 2.8533333333333335,
      "grad_norm": 0.5014028056158423,
      "learning_rate": 9.70789611675449e-08,
      "loss": 0.28517472743988037,
      "step": 202
    },
    {
      "epoch": 2.8675555555555556,
      "grad_norm": 0.5046489746973516,
      "learning_rate": 8.161562878982399e-08,
      "loss": 0.27302050590515137,
      "step": 203
    },
    {
      "epoch": 2.8817777777777778,
      "grad_norm": 0.5244675639408302,
      "learning_rate": 6.74828907941516e-08,
      "loss": 0.28563594818115234,
      "step": 204
    },
    {
      "epoch": 2.896,
      "grad_norm": 0.5147937769432819,
      "learning_rate": 5.468457057869358e-08,
      "loss": 0.2887130379676819,
      "step": 205
    },
    {
      "epoch": 2.910222222222222,
      "grad_norm": 0.49008819658339253,
      "learning_rate": 4.322413053509944e-08,
      "loss": 0.2721640467643738,
      "step": 206
    },
    {
      "epoch": 2.924444444444444,
      "grad_norm": 0.5051230135404199,
      "learning_rate": 3.31046711118066e-08,
      "loss": 0.27188998460769653,
      "step": 207
    },
    {
      "epoch": 2.9386666666666668,
      "grad_norm": 0.5046747206900266,
      "learning_rate": 2.4328929975260262e-08,
      "loss": 0.2795502245426178,
      "step": 208
    },
    {
      "epoch": 2.952888888888889,
      "grad_norm": 0.5035617947795644,
      "learning_rate": 1.6899281269279756e-08,
      "loss": 0.2727120816707611,
      "step": 209
    },
    {
      "epoch": 2.967111111111111,
      "grad_norm": 0.5078812650660502,
      "learning_rate": 1.0817734972768946e-08,
      "loss": 0.2795644998550415,
      "step": 210
    },
    {
      "epoch": 2.981333333333333,
      "grad_norm": 0.5269813851443549,
      "learning_rate": 6.085936355947897e-09,
      "loss": 0.28946471214294434,
      "step": 211
    },
    {
      "epoch": 2.9955555555555557,
      "grad_norm": 0.503969003244228,
      "learning_rate": 2.7051655352494654e-09,
      "loss": 0.26888906955718994,
      "step": 212
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.9730892144000087,
      "learning_rate": 6.763371270035457e-10,
      "loss": 0.33630549907684326,
      "step": 213
    },
    {
      "epoch": 3.0,
      "step": 213,
      "total_flos": 62893478699008.0,
      "train_loss": 0.36721718521185326,
      "train_runtime": 28478.4394,
      "train_samples_per_second": 1.896,
      "train_steps_per_second": 0.007
    }
  ],
  "logging_steps": 1,
  "max_steps": 213,
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 62893478699008.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}