{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 231,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.030303030303030304,
      "grad_norm": 6.389679464143325,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.8216,
      "step": 1
    },
    {
      "epoch": 0.06060606060606061,
      "grad_norm": 6.0870029465685995,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 0.794,
      "step": 2
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 6.33447423819787,
      "learning_rate": 2.5e-06,
      "loss": 0.8213,
      "step": 3
    },
    {
      "epoch": 0.12121212121212122,
      "grad_norm": 5.620528894398241,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.7913,
      "step": 4
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 4.316178832830539,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.7675,
      "step": 5
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 2.339462965558909,
      "learning_rate": 5e-06,
      "loss": 0.7282,
      "step": 6
    },
    {
      "epoch": 0.21212121212121213,
      "grad_norm": 1.9581073288072453,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.7226,
      "step": 7
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 3.6598363210046525,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.7141,
      "step": 8
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 4.008832019540338,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.7258,
      "step": 9
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 3.8023588513331323,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.6918,
      "step": 10
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 3.60329148202554,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.6673,
      "step": 11
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 3.0412468601087648,
      "learning_rate": 1e-05,
      "loss": 0.6527,
      "step": 12
    },
    {
      "epoch": 0.3939393939393939,
      "grad_norm": 1.7821663971794524,
      "learning_rate": 1.0833333333333334e-05,
      "loss": 0.6604,
      "step": 13
    },
    {
      "epoch": 0.42424242424242425,
      "grad_norm": 1.469929373790711,
      "learning_rate": 1.1666666666666668e-05,
      "loss": 0.6126,
      "step": 14
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 1.7088130414515645,
      "learning_rate": 1.25e-05,
      "loss": 0.6031,
      "step": 15
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 1.4345672873378874,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.6213,
      "step": 16
    },
    {
      "epoch": 0.5151515151515151,
      "grad_norm": 0.9537217329798097,
      "learning_rate": 1.416666666666667e-05,
      "loss": 0.5851,
      "step": 17
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 1.0359334660741335,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.5827,
      "step": 18
    },
    {
      "epoch": 0.5757575757575758,
      "grad_norm": 0.8612991371038484,
      "learning_rate": 1.5833333333333333e-05,
      "loss": 0.5626,
      "step": 19
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 0.8568176137969006,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.5622,
      "step": 20
    },
    {
      "epoch": 0.6363636363636364,
      "grad_norm": 0.8522777322963402,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 0.5481,
      "step": 21
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.7453158889642924,
      "learning_rate": 1.8333333333333333e-05,
      "loss": 0.5595,
      "step": 22
    },
    {
      "epoch": 0.696969696969697,
      "grad_norm": 0.9322760926453303,
      "learning_rate": 1.916666666666667e-05,
      "loss": 0.5446,
      "step": 23
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.7311347675851437,
      "learning_rate": 2e-05,
      "loss": 0.5457,
      "step": 24
    },
    {
      "epoch": 0.7575757575757576,
      "grad_norm": 0.7724040724529722,
      "learning_rate": 1.999884834944106e-05,
      "loss": 0.5346,
      "step": 25
    },
    {
      "epoch": 0.7878787878787878,
      "grad_norm": 0.7913425312973258,
      "learning_rate": 1.9995393663024054e-05,
      "loss": 0.5221,
      "step": 26
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 0.5435818443653805,
      "learning_rate": 1.9989636736467278e-05,
      "loss": 0.5274,
      "step": 27
    },
    {
      "epoch": 0.8484848484848485,
      "grad_norm": 0.7246903580516652,
      "learning_rate": 1.9981578895764272e-05,
      "loss": 0.5231,
      "step": 28
    },
    {
      "epoch": 0.8787878787878788,
      "grad_norm": 0.5172030157917001,
      "learning_rate": 1.9971221996878395e-05,
      "loss": 0.5221,
      "step": 29
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.7810391897625707,
      "learning_rate": 1.9958568425315316e-05,
      "loss": 0.5332,
      "step": 30
    },
    {
      "epoch": 0.9393939393939394,
      "grad_norm": 0.5497698223387213,
      "learning_rate": 1.9943621095573588e-05,
      "loss": 0.5204,
      "step": 31
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.5947629897020668,
      "learning_rate": 1.9926383450473344e-05,
      "loss": 0.5183,
      "step": 32
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.5330072862604534,
      "learning_rate": 1.9906859460363307e-05,
      "loss": 0.5098,
      "step": 33
    },
    {
      "epoch": 1.0303030303030303,
      "grad_norm": 0.4721557957467901,
      "learning_rate": 1.9885053622206305e-05,
      "loss": 0.4833,
      "step": 34
    },
    {
      "epoch": 1.0606060606060606,
      "grad_norm": 0.44696690995060356,
      "learning_rate": 1.986097095854347e-05,
      "loss": 0.4789,
      "step": 35
    },
    {
      "epoch": 1.0909090909090908,
      "grad_norm": 0.4715290430220381,
      "learning_rate": 1.9834617016337424e-05,
      "loss": 0.481,
      "step": 36
    },
    {
      "epoch": 1.121212121212121,
      "grad_norm": 0.484211942279594,
      "learning_rate": 1.9805997865694616e-05,
      "loss": 0.4686,
      "step": 37
    },
    {
      "epoch": 1.1515151515151516,
      "grad_norm": 0.4224809049475823,
      "learning_rate": 1.9775120098467212e-05,
      "loss": 0.488,
      "step": 38
    },
    {
      "epoch": 1.1818181818181819,
      "grad_norm": 0.4767478991829516,
      "learning_rate": 1.9741990826734793e-05,
      "loss": 0.4706,
      "step": 39
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 0.41475242363404724,
      "learning_rate": 1.970661768116622e-05,
      "loss": 0.4909,
      "step": 40
    },
    {
      "epoch": 1.2424242424242424,
      "grad_norm": 0.4493742206439678,
      "learning_rate": 1.9669008809262064e-05,
      "loss": 0.4816,
      "step": 41
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 0.4858174428866086,
      "learning_rate": 1.9629172873477995e-05,
      "loss": 0.4856,
      "step": 42
    },
    {
      "epoch": 1.303030303030303,
      "grad_norm": 0.5370196524998,
      "learning_rate": 1.9587119049229558e-05,
      "loss": 0.4848,
      "step": 43
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.3865029753757071,
      "learning_rate": 1.954285702277879e-05,
      "loss": 0.4734,
      "step": 44
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.5428346734484963,
      "learning_rate": 1.9496396989003195e-05,
      "loss": 0.4844,
      "step": 45
    },
    {
      "epoch": 1.393939393939394,
      "grad_norm": 0.49276277542024816,
      "learning_rate": 1.944774964904754e-05,
      "loss": 0.4597,
      "step": 46
    },
    {
      "epoch": 1.4242424242424243,
      "grad_norm": 0.5019994974591285,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 0.4662,
      "step": 47
    },
    {
      "epoch": 1.4545454545454546,
      "grad_norm": 0.46095938340959086,
      "learning_rate": 1.9343938371606714e-05,
      "loss": 0.4398,
      "step": 48
    },
    {
      "epoch": 1.4848484848484849,
      "grad_norm": 0.5436027541613718,
      "learning_rate": 1.9288798344984673e-05,
      "loss": 0.4594,
      "step": 49
    },
    {
      "epoch": 1.5151515151515151,
      "grad_norm": 0.512382069946856,
      "learning_rate": 1.9231518828401458e-05,
      "loss": 0.4627,
      "step": 50
    },
    {
      "epoch": 1.5454545454545454,
      "grad_norm": 0.3993417396610489,
      "learning_rate": 1.917211301505453e-05,
      "loss": 0.4531,
      "step": 51
    },
    {
      "epoch": 1.5757575757575757,
      "grad_norm": 0.4553369444771597,
      "learning_rate": 1.911059458789152e-05,
      "loss": 0.451,
      "step": 52
    },
    {
      "epoch": 1.606060606060606,
      "grad_norm": 0.4296801768417054,
      "learning_rate": 1.9046977716458627e-05,
      "loss": 0.4679,
      "step": 53
    },
    {
      "epoch": 1.6363636363636362,
      "grad_norm": 0.39723792447393597,
      "learning_rate": 1.8981277053636963e-05,
      "loss": 0.4649,
      "step": 54
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.43610761040124835,
      "learning_rate": 1.891350773226754e-05,
      "loss": 0.4432,
      "step": 55
    },
    {
      "epoch": 1.696969696969697,
      "grad_norm": 0.368349776002198,
      "learning_rate": 1.8843685361665724e-05,
      "loss": 0.454,
      "step": 56
    },
    {
      "epoch": 1.7272727272727273,
      "grad_norm": 0.4389169614250592,
      "learning_rate": 1.8771826024025944e-05,
      "loss": 0.4644,
      "step": 57
    },
    {
      "epoch": 1.7575757575757576,
      "grad_norm": 0.39930247566261223,
      "learning_rate": 1.8697946270717468e-05,
      "loss": 0.4628,
      "step": 58
    },
    {
      "epoch": 1.7878787878787878,
      "grad_norm": 0.42246137547786905,
      "learning_rate": 1.8622063118472135e-05,
      "loss": 0.4599,
      "step": 59
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.4991758233281764,
      "learning_rate": 1.8544194045464888e-05,
      "loss": 0.4597,
      "step": 60
    },
    {
      "epoch": 1.8484848484848486,
      "grad_norm": 0.3500385967202033,
      "learning_rate": 1.8464356987288012e-05,
      "loss": 0.4659,
      "step": 61
    },
    {
      "epoch": 1.878787878787879,
      "grad_norm": 0.5338413731056074,
      "learning_rate": 1.8382570332820045e-05,
      "loss": 0.4653,
      "step": 62
    },
    {
      "epoch": 1.9090909090909092,
      "grad_norm": 0.39315339859730924,
      "learning_rate": 1.8298852919990254e-05,
      "loss": 0.4712,
      "step": 63
    },
    {
      "epoch": 1.9393939393939394,
      "grad_norm": 0.49055759389219106,
      "learning_rate": 1.821322403143969e-05,
      "loss": 0.4659,
      "step": 64
    },
    {
      "epoch": 1.9696969696969697,
      "grad_norm": 0.4998865137451953,
      "learning_rate": 1.812570339007983e-05,
      "loss": 0.4617,
      "step": 65
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.5281523703208337,
      "learning_rate": 1.8036311154549783e-05,
      "loss": 0.4556,
      "step": 66
    },
    {
      "epoch": 2.0303030303030303,
      "grad_norm": 0.5183352728640479,
      "learning_rate": 1.7945067914573147e-05,
      "loss": 0.4028,
      "step": 67
    },
    {
      "epoch": 2.0606060606060606,
      "grad_norm": 0.43323389056033274,
      "learning_rate": 1.7851994686215592e-05,
      "loss": 0.4072,
      "step": 68
    },
    {
      "epoch": 2.090909090909091,
      "grad_norm": 0.490538354979083,
      "learning_rate": 1.77571129070442e-05,
      "loss": 0.4112,
      "step": 69
    },
    {
      "epoch": 2.121212121212121,
      "grad_norm": 0.48720992951592584,
      "learning_rate": 1.766044443118978e-05,
      "loss": 0.424,
      "step": 70
    },
    {
      "epoch": 2.1515151515151514,
      "grad_norm": 0.51892933006533,
      "learning_rate": 1.7562011524313187e-05,
      "loss": 0.4197,
      "step": 71
    },
    {
      "epoch": 2.1818181818181817,
      "grad_norm": 0.43316356012449,
      "learning_rate": 1.7461836858476858e-05,
      "loss": 0.412,
      "step": 72
    },
    {
      "epoch": 2.212121212121212,
      "grad_norm": 0.5011583477381106,
      "learning_rate": 1.7359943506922775e-05,
      "loss": 0.4163,
      "step": 73
    },
    {
      "epoch": 2.242424242424242,
      "grad_norm": 0.37391566522456676,
      "learning_rate": 1.725635493875799e-05,
      "loss": 0.3957,
      "step": 74
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.46126379346949903,
      "learning_rate": 1.7151095013548996e-05,
      "loss": 0.3962,
      "step": 75
    },
    {
      "epoch": 2.303030303030303,
      "grad_norm": 0.5138856949785989,
      "learning_rate": 1.7044187975826126e-05,
      "loss": 0.4197,
      "step": 76
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.5122522819282725,
      "learning_rate": 1.693565844949933e-05,
      "loss": 0.4174,
      "step": 77
    },
    {
      "epoch": 2.3636363636363638,
      "grad_norm": 0.4269939619340365,
      "learning_rate": 1.6825531432186545e-05,
      "loss": 0.4096,
      "step": 78
    },
    {
      "epoch": 2.393939393939394,
      "grad_norm": 0.3884930259980605,
      "learning_rate": 1.671383228945597e-05,
      "loss": 0.4013,
      "step": 79
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 0.4330282401428612,
      "learning_rate": 1.6600586748983642e-05,
      "loss": 0.4135,
      "step": 80
    },
    {
      "epoch": 2.4545454545454546,
      "grad_norm": 0.49867447378146007,
      "learning_rate": 1.648582089462756e-05,
      "loss": 0.4058,
      "step": 81
    },
    {
      "epoch": 2.484848484848485,
      "grad_norm": 0.4000948733390901,
      "learning_rate": 1.6369561160419783e-05,
      "loss": 0.3965,
      "step": 82
    },
    {
      "epoch": 2.515151515151515,
      "grad_norm": 0.40127339644535653,
      "learning_rate": 1.625183432447789e-05,
      "loss": 0.3977,
      "step": 83
    },
    {
      "epoch": 2.5454545454545454,
      "grad_norm": 0.4163281768705279,
      "learning_rate": 1.6132667502837164e-05,
      "loss": 0.4097,
      "step": 84
    },
    {
      "epoch": 2.5757575757575757,
      "grad_norm": 0.4130162141154564,
      "learning_rate": 1.6012088143204953e-05,
      "loss": 0.4036,
      "step": 85
    },
    {
      "epoch": 2.606060606060606,
      "grad_norm": 0.38323571037782045,
      "learning_rate": 1.589012401863864e-05,
      "loss": 0.4025,
      "step": 86
    },
    {
      "epoch": 2.6363636363636362,
      "grad_norm": 0.4359390222033282,
      "learning_rate": 1.5766803221148676e-05,
      "loss": 0.3929,
      "step": 87
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.390478372661262,
      "learning_rate": 1.5642154155228124e-05,
      "loss": 0.4087,
      "step": 88
    },
    {
      "epoch": 2.6969696969696972,
      "grad_norm": 0.3847582060805819,
      "learning_rate": 1.5516205531310272e-05,
      "loss": 0.4036,
      "step": 89
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.4010141318395045,
      "learning_rate": 1.538898635915576e-05,
      "loss": 0.4014,
      "step": 90
    },
    {
      "epoch": 2.757575757575758,
      "grad_norm": 0.39998137279337886,
      "learning_rate": 1.526052594117071e-05,
      "loss": 0.3989,
      "step": 91
    },
    {
      "epoch": 2.787878787878788,
      "grad_norm": 0.383782592700116,
      "learning_rate": 1.513085386565758e-05,
      "loss": 0.3953,
      "step": 92
    },
    {
      "epoch": 2.8181818181818183,
      "grad_norm": 0.33484545487270734,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.3982,
      "step": 93
    },
    {
      "epoch": 2.8484848484848486,
      "grad_norm": 0.4174189055080555,
      "learning_rate": 1.4867994483783485e-05,
      "loss": 0.4105,
      "step": 94
    },
    {
      "epoch": 2.878787878787879,
      "grad_norm": 0.3688575408583084,
      "learning_rate": 1.4734867721853341e-05,
      "loss": 0.397,
      "step": 95
    },
    {
      "epoch": 2.909090909090909,
      "grad_norm": 0.3911223572397203,
      "learning_rate": 1.4600650377311523e-05,
      "loss": 0.4092,
      "step": 96
    },
    {
      "epoch": 2.9393939393939394,
      "grad_norm": 0.393205250447062,
      "learning_rate": 1.4465373364454001e-05,
      "loss": 0.4152,
      "step": 97
    },
    {
      "epoch": 2.9696969696969697,
      "grad_norm": 0.33163654704896206,
      "learning_rate": 1.4329067841650274e-05,
      "loss": 0.4021,
      "step": 98
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.35852210007782426,
      "learning_rate": 1.4191765204166643e-05,
      "loss": 0.3881,
      "step": 99
    },
    {
      "epoch": 3.0303030303030303,
      "grad_norm": 0.4276929900256796,
      "learning_rate": 1.4053497076934948e-05,
      "loss": 0.3615,
      "step": 100
    },
    {
      "epoch": 3.0606060606060606,
      "grad_norm": 0.36943088418479125,
      "learning_rate": 1.3914295307268396e-05,
      "loss": 0.3647,
      "step": 101
    },
    {
      "epoch": 3.090909090909091,
      "grad_norm": 0.416864423936772,
      "learning_rate": 1.3774191957526144e-05,
      "loss": 0.3607,
      "step": 102
    },
    {
      "epoch": 3.121212121212121,
      "grad_norm": 0.454946907994254,
      "learning_rate": 1.3633219297728415e-05,
      "loss": 0.3542,
      "step": 103
    },
    {
      "epoch": 3.1515151515151514,
      "grad_norm": 0.4105877635581912,
      "learning_rate": 1.3491409798123687e-05,
      "loss": 0.3622,
      "step": 104
    },
    {
      "epoch": 3.1818181818181817,
      "grad_norm": 0.39877498537617606,
      "learning_rate": 1.3348796121709862e-05,
      "loss": 0.3641,
      "step": 105
    },
    {
      "epoch": 3.212121212121212,
      "grad_norm": 0.40857691652727696,
      "learning_rate": 1.3205411116710973e-05,
      "loss": 0.3357,
      "step": 106
    },
    {
      "epoch": 3.242424242424242,
      "grad_norm": 0.3774293129860823,
      "learning_rate": 1.3061287809011243e-05,
      "loss": 0.3528,
      "step": 107
    },
    {
      "epoch": 3.2727272727272725,
      "grad_norm": 0.36457845009271533,
      "learning_rate": 1.291645939454825e-05,
      "loss": 0.3467,
      "step": 108
    },
    {
      "epoch": 3.303030303030303,
      "grad_norm": 0.43697595376571186,
      "learning_rate": 1.277095923166689e-05,
      "loss": 0.3548,
      "step": 109
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.3279772156769785,
      "learning_rate": 1.2624820833435939e-05,
      "loss": 0.3628,
      "step": 110
    },
    {
      "epoch": 3.3636363636363638,
      "grad_norm": 0.3674180825534724,
      "learning_rate": 1.2478077859929e-05,
      "loss": 0.3445,
      "step": 111
    },
    {
      "epoch": 3.393939393939394,
      "grad_norm": 0.36677402712205043,
      "learning_rate": 1.2330764110471567e-05,
      "loss": 0.3518,
      "step": 112
    },
    {
      "epoch": 3.4242424242424243,
      "grad_norm": 0.32484641950884563,
      "learning_rate": 1.2182913515856016e-05,
      "loss": 0.349,
      "step": 113
    },
    {
      "epoch": 3.4545454545454546,
      "grad_norm": 0.32762570869276525,
      "learning_rate": 1.2034560130526341e-05,
      "loss": 0.3568,
      "step": 114
    },
    {
      "epoch": 3.484848484848485,
      "grad_norm": 0.3732488392977816,
      "learning_rate": 1.1885738124734359e-05,
      "loss": 0.3621,
      "step": 115
    },
    {
      "epoch": 3.515151515151515,
      "grad_norm": 0.3085942762155234,
      "learning_rate": 1.1736481776669307e-05,
      "loss": 0.3653,
      "step": 116
    },
    {
      "epoch": 3.5454545454545454,
      "grad_norm": 0.33005914795893343,
      "learning_rate": 1.1586825464562515e-05,
      "loss": 0.3434,
      "step": 117
    },
    {
      "epoch": 3.5757575757575757,
      "grad_norm": 0.37800192242609815,
      "learning_rate": 1.1436803658769082e-05,
      "loss": 0.3474,
      "step": 118
    },
    {
      "epoch": 3.606060606060606,
      "grad_norm": 0.2913441889881699,
      "learning_rate": 1.1286450913828313e-05,
      "loss": 0.3526,
      "step": 119
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.32804396327735513,
      "learning_rate": 1.113580186050475e-05,
      "loss": 0.3551,
      "step": 120
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.33516462116193874,
      "learning_rate": 1.0984891197811686e-05,
      "loss": 0.3568,
      "step": 121
    },
    {
      "epoch": 3.6969696969696972,
      "grad_norm": 0.32908995740085434,
      "learning_rate": 1.0833753685018935e-05,
      "loss": 0.3449,
      "step": 122
    },
    {
      "epoch": 3.7272727272727275,
      "grad_norm": 0.32147000359122696,
      "learning_rate": 1.0682424133646712e-05,
      "loss": 0.362,
      "step": 123
    },
    {
      "epoch": 3.757575757575758,
      "grad_norm": 0.3096197644123865,
      "learning_rate": 1.0530937399447496e-05,
      "loss": 0.3625,
      "step": 124
    },
    {
      "epoch": 3.787878787878788,
      "grad_norm": 0.31487973380195133,
      "learning_rate": 1.0379328374377715e-05,
      "loss": 0.3571,
      "step": 125
    },
    {
      "epoch": 3.8181818181818183,
      "grad_norm": 0.2896993260466313,
      "learning_rate": 1.0227631978561057e-05,
      "loss": 0.3547,
      "step": 126
    },
    {
      "epoch": 3.8484848484848486,
      "grad_norm": 0.30987572421685594,
      "learning_rate": 1.0075883152245334e-05,
      "loss": 0.3623,
      "step": 127
    },
    {
      "epoch": 3.878787878787879,
      "grad_norm": 0.2989010699414137,
      "learning_rate": 9.92411684775467e-06,
      "loss": 0.342,
      "step": 128
    },
    {
      "epoch": 3.909090909090909,
      "grad_norm": 0.33270560821661216,
      "learning_rate": 9.772368021438943e-06,
      "loss": 0.3482,
      "step": 129
    },
    {
      "epoch": 3.9393939393939394,
      "grad_norm": 0.30590884996875034,
      "learning_rate": 9.620671625622287e-06,
      "loss": 0.3522,
      "step": 130
    },
    {
      "epoch": 3.9696969696969697,
      "grad_norm": 0.3413263440780111,
      "learning_rate": 9.469062600552509e-06,
      "loss": 0.349,
      "step": 131
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.3221210837028033,
      "learning_rate": 9.317575866353293e-06,
      "loss": 0.3557,
      "step": 132
    },
    {
      "epoch": 4.03030303030303,
      "grad_norm": 0.4210306757150854,
      "learning_rate": 9.166246314981066e-06,
      "loss": 0.3101,
      "step": 133
    },
    {
      "epoch": 4.0606060606060606,
      "grad_norm": 0.35741966288239124,
      "learning_rate": 9.015108802188314e-06,
      "loss": 0.309,
      "step": 134
    },
    {
      "epoch": 4.090909090909091,
      "grad_norm": 0.4083522228049602,
      "learning_rate": 8.86419813949525e-06,
      "loss": 0.3245,
      "step": 135
    },
    {
      "epoch": 4.121212121212121,
      "grad_norm": 0.43594044169127666,
      "learning_rate": 8.71354908617169e-06,
      "loss": 0.3191,
      "step": 136
    },
    {
      "epoch": 4.151515151515151,
      "grad_norm": 0.4148276318838158,
      "learning_rate": 8.56319634123092e-06,
      "loss": 0.3044,
      "step": 137
    },
    {
      "epoch": 4.181818181818182,
      "grad_norm": 0.3631741823106515,
      "learning_rate": 8.413174535437486e-06,
      "loss": 0.3031,
      "step": 138
    },
    {
      "epoch": 4.212121212121212,
      "grad_norm": 0.3539162628389827,
      "learning_rate": 8.263518223330698e-06,
      "loss": 0.3132,
      "step": 139
    },
    {
      "epoch": 4.242424242424242,
      "grad_norm": 0.39308778047792964,
      "learning_rate": 8.114261875265643e-06,
      "loss": 0.309,
      "step": 140
    },
    {
      "epoch": 4.2727272727272725,
      "grad_norm": 0.4137786335768427,
      "learning_rate": 7.965439869473664e-06,
      "loss": 0.3141,
      "step": 141
    },
    {
      "epoch": 4.303030303030303,
      "grad_norm": 0.33053991835814944,
      "learning_rate": 7.817086484143987e-06,
      "loss": 0.3123,
      "step": 142
    },
    {
      "epoch": 4.333333333333333,
      "grad_norm": 0.360604137311516,
      "learning_rate": 7.669235889528436e-06,
      "loss": 0.299,
      "step": 143
    },
    {
      "epoch": 4.363636363636363,
      "grad_norm": 0.3886434471042827,
      "learning_rate": 7.521922140071003e-06,
      "loss": 0.3163,
      "step": 144
    },
    {
      "epoch": 4.393939393939394,
      "grad_norm": 0.3369677782711241,
      "learning_rate": 7.375179166564062e-06,
      "loss": 0.3012,
      "step": 145
    },
    {
      "epoch": 4.424242424242424,
      "grad_norm": 0.3192004662092612,
      "learning_rate": 7.2290407683331154e-06,
      "loss": 0.3124,
      "step": 146
    },
    {
      "epoch": 4.454545454545454,
      "grad_norm": 0.31553914305354386,
      "learning_rate": 7.0835406054517505e-06,
      "loss": 0.2999,
      "step": 147
    },
    {
      "epoch": 4.484848484848484,
      "grad_norm": 0.32409683962722596,
      "learning_rate": 6.93871219098876e-06,
      "loss": 0.3166,
      "step": 148
    },
    {
      "epoch": 4.515151515151516,
      "grad_norm": 0.29279884654404215,
      "learning_rate": 6.79458888328903e-06,
      "loss": 0.3134,
      "step": 149
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 0.3272960594603835,
      "learning_rate": 6.651203878290139e-06,
      "loss": 0.3171,
      "step": 150
    },
    {
      "epoch": 4.575757575757576,
      "grad_norm": 0.3186069661608066,
      "learning_rate": 6.508590201876317e-06,
      "loss": 0.3101,
      "step": 151
    },
    {
      "epoch": 4.606060606060606,
      "grad_norm": 0.2648439505655438,
      "learning_rate": 6.366780702271589e-06,
      "loss": 0.3043,
      "step": 152
    },
    {
      "epoch": 4.636363636363637,
      "grad_norm": 0.2788087197565788,
      "learning_rate": 6.225808042473857e-06,
      "loss": 0.31,
      "step": 153
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.2503203391643529,
      "learning_rate": 6.085704692731609e-06,
      "loss": 0.2977,
      "step": 154
    },
    {
      "epoch": 4.696969696969697,
      "grad_norm": 0.30809314794877196,
      "learning_rate": 5.946502923065054e-06,
      "loss": 0.3174,
      "step": 155
    },
    {
      "epoch": 4.7272727272727275,
      "grad_norm": 0.3045080386845063,
      "learning_rate": 5.8082347958333625e-06,
      "loss": 0.3138,
      "step": 156
    },
    {
      "epoch": 4.757575757575758,
      "grad_norm": 0.2797657437245348,
      "learning_rate": 5.670932158349732e-06,
      "loss": 0.3201,
      "step": 157
    },
    {
      "epoch": 4.787878787878788,
      "grad_norm": 0.2638418194723567,
      "learning_rate": 5.534626635546e-06,
      "loss": 0.3055,
      "step": 158
    },
    {
      "epoch": 4.818181818181818,
      "grad_norm": 0.27116179637904714,
      "learning_rate": 5.399349622688479e-06,
      "loss": 0.3047,
      "step": 159
    },
    {
      "epoch": 4.848484848484849,
      "grad_norm": 0.2609842736087905,
      "learning_rate": 5.2651322781466606e-06,
      "loss": 0.3079,
      "step": 160
    },
    {
      "epoch": 4.878787878787879,
      "grad_norm": 0.2637332780667072,
      "learning_rate": 5.132005516216512e-06,
      "loss": 0.3043,
      "step": 161
    },
    {
      "epoch": 4.909090909090909,
      "grad_norm": 0.2548209552892343,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.3132,
      "step": 162
    },
    {
      "epoch": 4.9393939393939394,
      "grad_norm": 0.26407785000721573,
      "learning_rate": 4.869146134342426e-06,
      "loss": 0.3133,
      "step": 163
    },
    {
      "epoch": 4.96969696969697,
      "grad_norm": 0.26320423865848713,
      "learning_rate": 4.739474058829288e-06,
      "loss": 0.3133,
      "step": 164
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.25260839130565943,
      "learning_rate": 4.611013640844245e-06,
      "loss": 0.3062,
      "step": 165
    },
    {
      "epoch": 5.03030303030303,
      "grad_norm": 0.35045364828201614,
      "learning_rate": 4.483794468689728e-06,
      "loss": 0.2825,
      "step": 166
    },
    {
      "epoch": 5.0606060606060606,
      "grad_norm": 0.2878497108222042,
      "learning_rate": 4.357845844771881e-06,
      "loss": 0.2881,
      "step": 167
    },
    {
      "epoch": 5.090909090909091,
      "grad_norm": 0.2692364179021947,
      "learning_rate": 4.2331967788513295e-06,
      "loss": 0.2864,
      "step": 168
    },
    {
      "epoch": 5.121212121212121,
      "grad_norm": 0.3397140513642043,
      "learning_rate": 4.109875981361363e-06,
      "loss": 0.2857,
      "step": 169
    },
    {
      "epoch": 5.151515151515151,
      "grad_norm": 0.33923809826546575,
      "learning_rate": 3.987911856795047e-06,
      "loss": 0.285,
      "step": 170
    },
    {
      "epoch": 5.181818181818182,
      "grad_norm": 0.2797575682833663,
      "learning_rate": 3.867332497162836e-06,
      "loss": 0.2824,
      "step": 171
    },
    {
      "epoch": 5.212121212121212,
      "grad_norm": 0.2653710601957683,
      "learning_rate": 3.748165675522113e-06,
      "loss": 0.2711,
      "step": 172
    },
    {
      "epoch": 5.242424242424242,
      "grad_norm": 0.27260257926718856,
      "learning_rate": 3.630438839580217e-06,
      "loss": 0.2724,
      "step": 173
    },
    {
      "epoch": 5.2727272727272725,
      "grad_norm": 0.27852051116246085,
      "learning_rate": 3.5141791053724405e-06,
      "loss": 0.2873,
      "step": 174
    },
    {
      "epoch": 5.303030303030303,
      "grad_norm": 0.2856413126687512,
      "learning_rate": 3.399413251016359e-06,
      "loss": 0.2735,
      "step": 175
    },
    {
      "epoch": 5.333333333333333,
      "grad_norm": 0.2593430557542982,
      "learning_rate": 3.2861677105440335e-06,
      "loss": 0.2687,
      "step": 176
    },
    {
      "epoch": 5.363636363636363,
      "grad_norm": 0.27770413574926345,
      "learning_rate": 3.174468567813461e-06,
      "loss": 0.2745,
      "step": 177
    },
    {
      "epoch": 5.393939393939394,
      "grad_norm": 0.28604505936188646,
      "learning_rate": 3.0643415505006733e-06,
      "loss": 0.2758,
      "step": 178
    },
    {
      "epoch": 5.424242424242424,
      "grad_norm": 0.2670316366650927,
      "learning_rate": 2.9558120241738786e-06,
      "loss": 0.278,
      "step": 179
    },
    {
      "epoch": 5.454545454545454,
      "grad_norm": 0.26192988499195347,
      "learning_rate": 2.8489049864510053e-06,
      "loss": 0.2713,
      "step": 180
    },
    {
      "epoch": 5.484848484848484,
      "grad_norm": 0.2579041025278341,
      "learning_rate": 2.7436450612420098e-06,
      "loss": 0.2758,
      "step": 181
    },
    {
      "epoch": 5.515151515151516,
      "grad_norm": 0.2581571699654079,
      "learning_rate": 2.640056493077231e-06,
      "loss": 0.2709,
      "step": 182
    },
    {
      "epoch": 5.545454545454545,
      "grad_norm": 0.26358679708819666,
      "learning_rate": 2.5381631415231455e-06,
      "loss": 0.2793,
      "step": 183
    },
    {
      "epoch": 5.575757575757576,
      "grad_norm": 0.26036766758260393,
      "learning_rate": 2.4379884756868167e-06,
      "loss": 0.2733,
      "step": 184
    },
    {
      "epoch": 5.606060606060606,
      "grad_norm": 0.24900549026134164,
      "learning_rate": 2.339555568810221e-06,
      "loss": 0.2745,
      "step": 185
    },
    {
      "epoch": 5.636363636363637,
      "grad_norm": 0.24263829872831869,
      "learning_rate": 2.2428870929558012e-06,
      "loss": 0.2646,
      "step": 186
    },
    {
      "epoch": 5.666666666666667,
      "grad_norm": 0.23173769595787047,
      "learning_rate": 2.1480053137844115e-06,
      "loss": 0.268,
      "step": 187
    },
    {
      "epoch": 5.696969696969697,
      "grad_norm": 0.24743207197013373,
      "learning_rate": 2.054932085426856e-06,
      "loss": 0.2747,
      "step": 188
    },
    {
      "epoch": 5.7272727272727275,
      "grad_norm": 0.24678627068896922,
      "learning_rate": 1.963688845450218e-06,
      "loss": 0.2784,
      "step": 189
    },
    {
      "epoch": 5.757575757575758,
      "grad_norm": 0.24292786245623466,
      "learning_rate": 1.8742966099201699e-06,
      "loss": 0.2743,
      "step": 190
    },
    {
      "epoch": 5.787878787878788,
      "grad_norm": 0.2301052425255644,
      "learning_rate": 1.7867759685603115e-06,
      "loss": 0.2755,
      "step": 191
    },
    {
      "epoch": 5.818181818181818,
      "grad_norm": 0.24436621048715285,
      "learning_rate": 1.7011470800097496e-06,
      "loss": 0.2802,
      "step": 192
    },
    {
      "epoch": 5.848484848484849,
      "grad_norm": 0.24233892802925538,
      "learning_rate": 1.6174296671799571e-06,
      "loss": 0.2783,
      "step": 193
    },
    {
      "epoch": 5.878787878787879,
      "grad_norm": 0.2326648485384654,
      "learning_rate": 1.5356430127119915e-06,
      "loss": 0.2794,
      "step": 194
    },
    {
      "epoch": 5.909090909090909,
      "grad_norm": 0.23967329818495606,
      "learning_rate": 1.4558059545351144e-06,
      "loss": 0.2922,
      "step": 195
    },
    {
      "epoch": 5.9393939393939394,
      "grad_norm": 0.23136254304772033,
      "learning_rate": 1.3779368815278648e-06,
      "loss": 0.2665,
      "step": 196
    },
    {
      "epoch": 5.96969696969697,
      "grad_norm": 0.22442234807503147,
      "learning_rate": 1.302053729282533e-06,
      "loss": 0.2768,
      "step": 197
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.2248897102418934,
      "learning_rate": 1.2281739759740575e-06,
      "loss": 0.2786,
      "step": 198
    },
    {
      "epoch": 6.03030303030303,
      "grad_norm": 0.27551238016074,
      "learning_rate": 1.156314638334277e-06,
      "loss": 0.2628,
      "step": 199
    },
    {
      "epoch": 6.0606060606060606,
      "grad_norm": 0.30409342282998103,
      "learning_rate": 1.086492267732462e-06,
      "loss": 0.269,
      "step": 200
    },
    {
      "epoch": 6.090909090909091,
      "grad_norm": 0.24172010021563448,
      "learning_rate": 1.01872294636304e-06,
      "loss": 0.2532,
      "step": 201
    },
    {
      "epoch": 6.121212121212121,
      "grad_norm": 0.22075170600714575,
      "learning_rate": 9.530222835413739e-07,
      "loss": 0.263,
      "step": 202
    },
    {
      "epoch": 6.151515151515151,
      "grad_norm": 0.23075253241023455,
      "learning_rate": 8.894054121084839e-07,
      "loss": 0.2652,
      "step": 203
    },
    {
      "epoch": 6.181818181818182,
      "grad_norm": 0.22700574128593443,
      "learning_rate": 8.278869849454718e-07,
      "loss": 0.2495,
      "step": 204
    },
    {
      "epoch": 6.212121212121212,
      "grad_norm": 0.23962162206606205,
      "learning_rate": 7.684811715985429e-07,
      "loss": 0.2635,
      "step": 205
    },
    {
      "epoch": 6.242424242424242,
      "grad_norm": 0.33390979469076315,
      "learning_rate": 7.1120165501533e-07,
      "loss": 0.2691,
      "step": 206
    },
    {
      "epoch": 6.2727272727272725,
      "grad_norm": 0.24819559014018444,
      "learning_rate": 6.560616283932897e-07,
      "loss": 0.2705,
      "step": 207
    },
    {
      "epoch": 6.303030303030303,
      "grad_norm": 0.2600890340128434,
      "learning_rate": 6.030737921409169e-07,
      "loss": 0.2672,
      "step": 208
    },
    {
      "epoch": 6.333333333333333,
      "grad_norm": 0.3911203830207476,
      "learning_rate": 5.522503509524591e-07,
      "loss": 0.2619,
      "step": 209
    },
    {
      "epoch": 6.363636363636363,
      "grad_norm": 0.22909261281290824,
      "learning_rate": 5.036030109968082e-07,
      "loss": 0.2707,
      "step": 210
    },
    {
      "epoch": 6.393939393939394,
      "grad_norm": 0.21607504208098888,
      "learning_rate": 4.5714297722121105e-07,
      "loss": 0.2582,
      "step": 211
    },
    {
      "epoch": 6.424242424242424,
      "grad_norm": 0.2306256207445463,
      "learning_rate": 4.128809507704445e-07,
      "loss": 0.2684,
      "step": 212
    },
    {
      "epoch": 6.454545454545454,
      "grad_norm": 0.22395608903910025,
      "learning_rate": 3.708271265220087e-07,
      "loss": 0.2573,
      "step": 213
    },
    {
      "epoch": 6.484848484848484,
      "grad_norm": 0.22022957019820769,
      "learning_rate": 3.309911907379393e-07,
      "loss": 0.2557,
      "step": 214
    },
    {
      "epoch": 6.515151515151516,
      "grad_norm": 0.22210833297225904,
      "learning_rate": 2.9338231883378365e-07,
      "loss": 0.2543,
      "step": 215
    },
    {
      "epoch": 6.545454545454545,
      "grad_norm": 0.23110809543643812,
      "learning_rate": 2.5800917326521013e-07,
      "loss": 0.2662,
      "step": 216
    },
    {
      "epoch": 6.575757575757576,
      "grad_norm": 0.22296248460503695,
      "learning_rate": 2.248799015327907e-07,
      "loss": 0.256,
      "step": 217
    },
    {
      "epoch": 6.606060606060606,
      "grad_norm": 0.20990335785125108,
      "learning_rate": 1.9400213430538773e-07,
      "loss": 0.2561,
      "step": 218
    },
    {
      "epoch": 6.636363636363637,
      "grad_norm": 0.2165769544705027,
      "learning_rate": 1.6538298366257975e-07,
      "loss": 0.263,
      "step": 219
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.21781489536538548,
      "learning_rate": 1.3902904145653094e-07,
      "loss": 0.2535,
      "step": 220
    },
    {
      "epoch": 6.696969696969697,
      "grad_norm": 0.20910433160702532,
      "learning_rate": 1.1494637779369766e-07,
      "loss": 0.2555,
      "step": 221
    },
    {
      "epoch": 6.7272727272727275,
      "grad_norm": 0.2236679596736996,
      "learning_rate": 9.314053963669245e-08,
      "loss": 0.2644,
      "step": 222
    },
    {
      "epoch": 6.757575757575758,
      "grad_norm": 0.27223019451133684,
      "learning_rate": 7.361654952665608e-08,
      "loss": 0.2707,
      "step": 223
    },
    {
      "epoch": 6.787878787878788,
      "grad_norm": 0.21385109093961496,
      "learning_rate": 5.637890442641403e-08,
      "loss": 0.265,
      "step": 224
    },
    {
      "epoch": 6.818181818181818,
      "grad_norm": 0.22145945419878782,
      "learning_rate": 4.143157468468717e-08,
      "loss": 0.2608,
      "step": 225
    },
    {
      "epoch": 6.848484848484849,
      "grad_norm": 0.21308961257858408,
      "learning_rate": 2.8778003121607834e-08,
      "loss": 0.2551,
      "step": 226
    },
    {
      "epoch": 6.878787878787879,
      "grad_norm": 0.2135751235389027,
      "learning_rate": 1.8421104235727406e-08,
      "loss": 0.2647,
      "step": 227
    },
    {
      "epoch": 6.909090909090909,
      "grad_norm": 0.21280976371458274,
      "learning_rate": 1.0363263532724433e-08,
      "loss": 0.2552,
      "step": 228
    },
    {
      "epoch": 6.9393939393939394,
      "grad_norm": 0.2128839202522525,
      "learning_rate": 4.606336975948589e-09,
      "loss": 0.2584,
      "step": 229
    },
    {
      "epoch": 6.96969696969697,
      "grad_norm": 0.2185104258110578,
      "learning_rate": 1.1516505589381777e-09,
      "loss": 0.2668,
      "step": 230
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.21354802243051257,
      "learning_rate": 0.0,
      "loss": 0.2646,
      "step": 231
    },
    {
      "epoch": 7.0,
      "step": 231,
      "total_flos": 7.881049861301207e+17,
      "train_loss": 0.3857915287281012,
      "train_runtime": 8985.043,
      "train_samples_per_second": 2.462,
      "train_steps_per_second": 0.026
    }
  ],
  "logging_steps": 1,
  "max_steps": 231,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "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": 7.881049861301207e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}