{
  "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": 6.452846450046593,
      "learning_rate": 8.333333333333333e-07,
      "loss": 1.0307,
      "step": 1
    },
    {
      "epoch": 0.06060606060606061,
      "grad_norm": 6.447988993134701,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.0308,
      "step": 2
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 6.351162539222836,
      "learning_rate": 2.5e-06,
      "loss": 1.0333,
      "step": 3
    },
    {
      "epoch": 0.12121212121212122,
      "grad_norm": 5.911792055461034,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.0195,
      "step": 4
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 4.519972683999297,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.9659,
      "step": 5
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 2.7514490775345006,
      "learning_rate": 5e-06,
      "loss": 0.95,
      "step": 6
    },
    {
      "epoch": 0.21212121212121213,
      "grad_norm": 2.5352290803313773,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.9384,
      "step": 7
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 3.8079896345967805,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.9136,
      "step": 8
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 4.33457964113686,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.9315,
      "step": 9
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 4.1431055386394675,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.9189,
      "step": 10
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 3.3104727522137947,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.8751,
      "step": 11
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 3.1514950318584503,
      "learning_rate": 1e-05,
      "loss": 0.8654,
      "step": 12
    },
    {
      "epoch": 0.3939393939393939,
      "grad_norm": 2.1539397920806564,
      "learning_rate": 1.0833333333333334e-05,
      "loss": 0.8387,
      "step": 13
    },
    {
      "epoch": 0.42424242424242425,
      "grad_norm": 1.5633645076135247,
      "learning_rate": 1.1666666666666668e-05,
      "loss": 0.8166,
      "step": 14
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 1.586277742812334,
      "learning_rate": 1.25e-05,
      "loss": 0.8037,
      "step": 15
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 1.5205656225238042,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.8047,
      "step": 16
    },
    {
      "epoch": 0.5151515151515151,
      "grad_norm": 1.2988751499636568,
      "learning_rate": 1.416666666666667e-05,
      "loss": 0.773,
      "step": 17
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 1.3033114764609133,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.7643,
      "step": 18
    },
    {
      "epoch": 0.5757575757575758,
      "grad_norm": 1.1501278524525895,
      "learning_rate": 1.5833333333333333e-05,
      "loss": 0.7628,
      "step": 19
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 1.0002786784231719,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.7616,
      "step": 20
    },
    {
      "epoch": 0.6363636363636364,
      "grad_norm": 0.9956781698574805,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 0.7289,
      "step": 21
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.9469024414634009,
      "learning_rate": 1.8333333333333333e-05,
      "loss": 0.7479,
      "step": 22
    },
    {
      "epoch": 0.696969696969697,
      "grad_norm": 0.9142125317289609,
      "learning_rate": 1.916666666666667e-05,
      "loss": 0.7534,
      "step": 23
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.8510305159919446,
      "learning_rate": 2e-05,
      "loss": 0.7232,
      "step": 24
    },
    {
      "epoch": 0.7575757575757576,
      "grad_norm": 0.8476097184464643,
      "learning_rate": 1.999884834944106e-05,
      "loss": 0.725,
      "step": 25
    },
    {
      "epoch": 0.7878787878787878,
      "grad_norm": 0.748052493231264,
      "learning_rate": 1.9995393663024054e-05,
      "loss": 0.7033,
      "step": 26
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 0.7162077987175882,
      "learning_rate": 1.9989636736467278e-05,
      "loss": 0.7135,
      "step": 27
    },
    {
      "epoch": 0.8484848484848485,
      "grad_norm": 0.7130681625855151,
      "learning_rate": 1.9981578895764272e-05,
      "loss": 0.7033,
      "step": 28
    },
    {
      "epoch": 0.8787878787878788,
      "grad_norm": 0.583626636803816,
      "learning_rate": 1.9971221996878395e-05,
      "loss": 0.6981,
      "step": 29
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.7338641572349641,
      "learning_rate": 1.9958568425315316e-05,
      "loss": 0.7126,
      "step": 30
    },
    {
      "epoch": 0.9393939393939394,
      "grad_norm": 0.7368636561596066,
      "learning_rate": 1.9943621095573588e-05,
      "loss": 0.6924,
      "step": 31
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.5793422348650669,
      "learning_rate": 1.9926383450473344e-05,
      "loss": 0.7029,
      "step": 32
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.9344765304856428,
      "learning_rate": 1.9906859460363307e-05,
      "loss": 0.6869,
      "step": 33
    },
    {
      "epoch": 1.0303030303030303,
      "grad_norm": 0.5948581986537889,
      "learning_rate": 1.9885053622206305e-05,
      "loss": 0.6433,
      "step": 34
    },
    {
      "epoch": 1.0606060606060606,
      "grad_norm": 0.7501650332446725,
      "learning_rate": 1.986097095854347e-05,
      "loss": 0.6518,
      "step": 35
    },
    {
      "epoch": 1.0909090909090908,
      "grad_norm": 0.738853656081209,
      "learning_rate": 1.9834617016337424e-05,
      "loss": 0.6527,
      "step": 36
    },
    {
      "epoch": 1.121212121212121,
      "grad_norm": 0.6124937052587867,
      "learning_rate": 1.9805997865694616e-05,
      "loss": 0.6412,
      "step": 37
    },
    {
      "epoch": 1.1515151515151516,
      "grad_norm": 0.7334688057639257,
      "learning_rate": 1.9775120098467212e-05,
      "loss": 0.6292,
      "step": 38
    },
    {
      "epoch": 1.1818181818181819,
      "grad_norm": 0.7187966341917992,
      "learning_rate": 1.9741990826734793e-05,
      "loss": 0.6428,
      "step": 39
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 0.5913252847846683,
      "learning_rate": 1.970661768116622e-05,
      "loss": 0.6364,
      "step": 40
    },
    {
      "epoch": 1.2424242424242424,
      "grad_norm": 0.7267662136237013,
      "learning_rate": 1.9669008809262064e-05,
      "loss": 0.6415,
      "step": 41
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 0.752493545629968,
      "learning_rate": 1.9629172873477995e-05,
      "loss": 0.6382,
      "step": 42
    },
    {
      "epoch": 1.303030303030303,
      "grad_norm": 0.6781778070413828,
      "learning_rate": 1.9587119049229558e-05,
      "loss": 0.6432,
      "step": 43
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.7000171568671039,
      "learning_rate": 1.954285702277879e-05,
      "loss": 0.6356,
      "step": 44
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.6115147887330554,
      "learning_rate": 1.9496396989003195e-05,
      "loss": 0.6229,
      "step": 45
    },
    {
      "epoch": 1.393939393939394,
      "grad_norm": 0.5311763795410067,
      "learning_rate": 1.944774964904754e-05,
      "loss": 0.6393,
      "step": 46
    },
    {
      "epoch": 1.4242424242424243,
      "grad_norm": 0.6544532904774394,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 0.619,
      "step": 47
    },
    {
      "epoch": 1.4545454545454546,
      "grad_norm": 0.6036809626033666,
      "learning_rate": 1.9343938371606714e-05,
      "loss": 0.6419,
      "step": 48
    },
    {
      "epoch": 1.4848484848484849,
      "grad_norm": 0.5249145378403873,
      "learning_rate": 1.9288798344984673e-05,
      "loss": 0.6201,
      "step": 49
    },
    {
      "epoch": 1.5151515151515151,
      "grad_norm": 0.9088639627765884,
      "learning_rate": 1.9231518828401458e-05,
      "loss": 0.6174,
      "step": 50
    },
    {
      "epoch": 1.5454545454545454,
      "grad_norm": 0.6254832909759365,
      "learning_rate": 1.917211301505453e-05,
      "loss": 0.6309,
      "step": 51
    },
    {
      "epoch": 1.5757575757575757,
      "grad_norm": 0.6394269138218432,
      "learning_rate": 1.911059458789152e-05,
      "loss": 0.6284,
      "step": 52
    },
    {
      "epoch": 1.606060606060606,
      "grad_norm": 0.571461974164157,
      "learning_rate": 1.9046977716458627e-05,
      "loss": 0.6222,
      "step": 53
    },
    {
      "epoch": 1.6363636363636362,
      "grad_norm": 0.5776233901420704,
      "learning_rate": 1.8981277053636963e-05,
      "loss": 0.6228,
      "step": 54
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.49407204558896484,
      "learning_rate": 1.891350773226754e-05,
      "loss": 0.6168,
      "step": 55
    },
    {
      "epoch": 1.696969696969697,
      "grad_norm": 0.6125202438051,
      "learning_rate": 1.8843685361665724e-05,
      "loss": 0.6292,
      "step": 56
    },
    {
      "epoch": 1.7272727272727273,
      "grad_norm": 0.5318692780440976,
      "learning_rate": 1.8771826024025944e-05,
      "loss": 0.61,
      "step": 57
    },
    {
      "epoch": 1.7575757575757576,
      "grad_norm": 0.5625788845807446,
      "learning_rate": 1.8697946270717468e-05,
      "loss": 0.6074,
      "step": 58
    },
    {
      "epoch": 1.7878787878787878,
      "grad_norm": 0.5839134897393666,
      "learning_rate": 1.8622063118472135e-05,
      "loss": 0.6178,
      "step": 59
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.5049923051878747,
      "learning_rate": 1.8544194045464888e-05,
      "loss": 0.6289,
      "step": 60
    },
    {
      "epoch": 1.8484848484848486,
      "grad_norm": 0.6269346793822278,
      "learning_rate": 1.8464356987288012e-05,
      "loss": 0.6211,
      "step": 61
    },
    {
      "epoch": 1.878787878787879,
      "grad_norm": 0.6044311170625676,
      "learning_rate": 1.8382570332820045e-05,
      "loss": 0.612,
      "step": 62
    },
    {
      "epoch": 1.9090909090909092,
      "grad_norm": 0.5084061377988478,
      "learning_rate": 1.8298852919990254e-05,
      "loss": 0.5993,
      "step": 63
    },
    {
      "epoch": 1.9393939393939394,
      "grad_norm": 0.591129466708481,
      "learning_rate": 1.821322403143969e-05,
      "loss": 0.6138,
      "step": 64
    },
    {
      "epoch": 1.9696969696969697,
      "grad_norm": 0.4850661968922958,
      "learning_rate": 1.812570339007983e-05,
      "loss": 0.6046,
      "step": 65
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.552538409179087,
      "learning_rate": 1.8036311154549783e-05,
      "loss": 0.6211,
      "step": 66
    },
    {
      "epoch": 2.0303030303030303,
      "grad_norm": 0.5823597726587708,
      "learning_rate": 1.7945067914573147e-05,
      "loss": 0.5528,
      "step": 67
    },
    {
      "epoch": 2.0606060606060606,
      "grad_norm": 0.5327036513046745,
      "learning_rate": 1.7851994686215592e-05,
      "loss": 0.567,
      "step": 68
    },
    {
      "epoch": 2.090909090909091,
      "grad_norm": 0.5725226908709942,
      "learning_rate": 1.77571129070442e-05,
      "loss": 0.5528,
      "step": 69
    },
    {
      "epoch": 2.121212121212121,
      "grad_norm": 0.5411050256261067,
      "learning_rate": 1.766044443118978e-05,
      "loss": 0.5465,
      "step": 70
    },
    {
      "epoch": 2.1515151515151514,
      "grad_norm": 0.5864645668462136,
      "learning_rate": 1.7562011524313187e-05,
      "loss": 0.5643,
      "step": 71
    },
    {
      "epoch": 2.1818181818181817,
      "grad_norm": 0.4328342018518578,
      "learning_rate": 1.7461836858476858e-05,
      "loss": 0.5518,
      "step": 72
    },
    {
      "epoch": 2.212121212121212,
      "grad_norm": 0.5255376359172979,
      "learning_rate": 1.7359943506922775e-05,
      "loss": 0.5379,
      "step": 73
    },
    {
      "epoch": 2.242424242424242,
      "grad_norm": 0.4551787562347675,
      "learning_rate": 1.725635493875799e-05,
      "loss": 0.5446,
      "step": 74
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.4807786266269301,
      "learning_rate": 1.7151095013548996e-05,
      "loss": 0.5386,
      "step": 75
    },
    {
      "epoch": 2.303030303030303,
      "grad_norm": 0.5120047461759244,
      "learning_rate": 1.7044187975826126e-05,
      "loss": 0.5548,
      "step": 76
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.5564208359614328,
      "learning_rate": 1.693565844949933e-05,
      "loss": 0.5543,
      "step": 77
    },
    {
      "epoch": 2.3636363636363638,
      "grad_norm": 0.559226052995087,
      "learning_rate": 1.6825531432186545e-05,
      "loss": 0.5421,
      "step": 78
    },
    {
      "epoch": 2.393939393939394,
      "grad_norm": 0.5214337047649399,
      "learning_rate": 1.671383228945597e-05,
      "loss": 0.559,
      "step": 79
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 0.5237447973428928,
      "learning_rate": 1.6600586748983642e-05,
      "loss": 0.5434,
      "step": 80
    },
    {
      "epoch": 2.4545454545454546,
      "grad_norm": 0.5991211230364246,
      "learning_rate": 1.648582089462756e-05,
      "loss": 0.5542,
      "step": 81
    },
    {
      "epoch": 2.484848484848485,
      "grad_norm": 0.47736503501954347,
      "learning_rate": 1.6369561160419783e-05,
      "loss": 0.5429,
      "step": 82
    },
    {
      "epoch": 2.515151515151515,
      "grad_norm": 0.5843353479827899,
      "learning_rate": 1.625183432447789e-05,
      "loss": 0.5375,
      "step": 83
    },
    {
      "epoch": 2.5454545454545454,
      "grad_norm": 0.5918072260921807,
      "learning_rate": 1.6132667502837164e-05,
      "loss": 0.5438,
      "step": 84
    },
    {
      "epoch": 2.5757575757575757,
      "grad_norm": 0.619926190067825,
      "learning_rate": 1.6012088143204953e-05,
      "loss": 0.5272,
      "step": 85
    },
    {
      "epoch": 2.606060606060606,
      "grad_norm": 0.5874894738609688,
      "learning_rate": 1.589012401863864e-05,
      "loss": 0.5376,
      "step": 86
    },
    {
      "epoch": 2.6363636363636362,
      "grad_norm": 0.5668555326682011,
      "learning_rate": 1.5766803221148676e-05,
      "loss": 0.5371,
      "step": 87
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.47413878181336,
      "learning_rate": 1.5642154155228124e-05,
      "loss": 0.5546,
      "step": 88
    },
    {
      "epoch": 2.6969696969696972,
      "grad_norm": 0.5582886449667891,
      "learning_rate": 1.5516205531310272e-05,
      "loss": 0.5611,
      "step": 89
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.45585217417408236,
      "learning_rate": 1.538898635915576e-05,
      "loss": 0.5369,
      "step": 90
    },
    {
      "epoch": 2.757575757575758,
      "grad_norm": 0.45245002656033984,
      "learning_rate": 1.526052594117071e-05,
      "loss": 0.5363,
      "step": 91
    },
    {
      "epoch": 2.787878787878788,
      "grad_norm": 0.4798081085255577,
      "learning_rate": 1.513085386565758e-05,
      "loss": 0.543,
      "step": 92
    },
    {
      "epoch": 2.8181818181818183,
      "grad_norm": 0.4862186474350233,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.5511,
      "step": 93
    },
    {
      "epoch": 2.8484848484848486,
      "grad_norm": 0.5102713100923337,
      "learning_rate": 1.4867994483783485e-05,
      "loss": 0.5449,
      "step": 94
    },
    {
      "epoch": 2.878787878787879,
      "grad_norm": 0.4245003658121503,
      "learning_rate": 1.4734867721853341e-05,
      "loss": 0.5353,
      "step": 95
    },
    {
      "epoch": 2.909090909090909,
      "grad_norm": 0.5474492718653113,
      "learning_rate": 1.4600650377311523e-05,
      "loss": 0.5403,
      "step": 96
    },
    {
      "epoch": 2.9393939393939394,
      "grad_norm": 0.41125771140947087,
      "learning_rate": 1.4465373364454001e-05,
      "loss": 0.5484,
      "step": 97
    },
    {
      "epoch": 2.9696969696969697,
      "grad_norm": 0.498597958282166,
      "learning_rate": 1.4329067841650274e-05,
      "loss": 0.5505,
      "step": 98
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.37535772204817613,
      "learning_rate": 1.4191765204166643e-05,
      "loss": 0.5251,
      "step": 99
    },
    {
      "epoch": 3.0303030303030303,
      "grad_norm": 0.6628620935507199,
      "learning_rate": 1.4053497076934948e-05,
      "loss": 0.4955,
      "step": 100
    },
    {
      "epoch": 3.0606060606060606,
      "grad_norm": 0.4824429937431021,
      "learning_rate": 1.3914295307268396e-05,
      "loss": 0.4877,
      "step": 101
    },
    {
      "epoch": 3.090909090909091,
      "grad_norm": 0.5952378627723972,
      "learning_rate": 1.3774191957526144e-05,
      "loss": 0.4976,
      "step": 102
    },
    {
      "epoch": 3.121212121212121,
      "grad_norm": 0.6352242534374526,
      "learning_rate": 1.3633219297728415e-05,
      "loss": 0.4837,
      "step": 103
    },
    {
      "epoch": 3.1515151515151514,
      "grad_norm": 0.5507724256152463,
      "learning_rate": 1.3491409798123687e-05,
      "loss": 0.4671,
      "step": 104
    },
    {
      "epoch": 3.1818181818181817,
      "grad_norm": 0.6027653508439558,
      "learning_rate": 1.3348796121709862e-05,
      "loss": 0.4718,
      "step": 105
    },
    {
      "epoch": 3.212121212121212,
      "grad_norm": 0.5985116836930497,
      "learning_rate": 1.3205411116710973e-05,
      "loss": 0.4723,
      "step": 106
    },
    {
      "epoch": 3.242424242424242,
      "grad_norm": 0.5199312081895848,
      "learning_rate": 1.3061287809011243e-05,
      "loss": 0.4778,
      "step": 107
    },
    {
      "epoch": 3.2727272727272725,
      "grad_norm": 0.6235943757883602,
      "learning_rate": 1.291645939454825e-05,
      "loss": 0.4763,
      "step": 108
    },
    {
      "epoch": 3.303030303030303,
      "grad_norm": 0.5403024442345382,
      "learning_rate": 1.277095923166689e-05,
      "loss": 0.4666,
      "step": 109
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.4604390671063723,
      "learning_rate": 1.2624820833435939e-05,
      "loss": 0.4814,
      "step": 110
    },
    {
      "epoch": 3.3636363636363638,
      "grad_norm": 0.5009759037686559,
      "learning_rate": 1.2478077859929e-05,
      "loss": 0.4947,
      "step": 111
    },
    {
      "epoch": 3.393939393939394,
      "grad_norm": 0.47167011145975224,
      "learning_rate": 1.2330764110471567e-05,
      "loss": 0.4839,
      "step": 112
    },
    {
      "epoch": 3.4242424242424243,
      "grad_norm": 0.5108218724689051,
      "learning_rate": 1.2182913515856016e-05,
      "loss": 0.4794,
      "step": 113
    },
    {
      "epoch": 3.4545454545454546,
      "grad_norm": 0.5610307246423525,
      "learning_rate": 1.2034560130526341e-05,
      "loss": 0.4866,
      "step": 114
    },
    {
      "epoch": 3.484848484848485,
      "grad_norm": 0.5202637827384432,
      "learning_rate": 1.1885738124734359e-05,
      "loss": 0.4801,
      "step": 115
    },
    {
      "epoch": 3.515151515151515,
      "grad_norm": 0.4554159008240091,
      "learning_rate": 1.1736481776669307e-05,
      "loss": 0.4717,
      "step": 116
    },
    {
      "epoch": 3.5454545454545454,
      "grad_norm": 0.5301547775626723,
      "learning_rate": 1.1586825464562515e-05,
      "loss": 0.4725,
      "step": 117
    },
    {
      "epoch": 3.5757575757575757,
      "grad_norm": 0.4522686511024801,
      "learning_rate": 1.1436803658769082e-05,
      "loss": 0.4808,
      "step": 118
    },
    {
      "epoch": 3.606060606060606,
      "grad_norm": 0.5579148074441338,
      "learning_rate": 1.1286450913828313e-05,
      "loss": 0.4825,
      "step": 119
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.452891993649687,
      "learning_rate": 1.113580186050475e-05,
      "loss": 0.4709,
      "step": 120
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.4935369412714626,
      "learning_rate": 1.0984891197811686e-05,
      "loss": 0.4807,
      "step": 121
    },
    {
      "epoch": 3.6969696969696972,
      "grad_norm": 0.48691440322998636,
      "learning_rate": 1.0833753685018935e-05,
      "loss": 0.4755,
      "step": 122
    },
    {
      "epoch": 3.7272727272727275,
      "grad_norm": 0.45151952548854407,
      "learning_rate": 1.0682424133646712e-05,
      "loss": 0.4698,
      "step": 123
    },
    {
      "epoch": 3.757575757575758,
      "grad_norm": 0.46005587127994213,
      "learning_rate": 1.0530937399447496e-05,
      "loss": 0.4704,
      "step": 124
    },
    {
      "epoch": 3.787878787878788,
      "grad_norm": 0.42665263281083105,
      "learning_rate": 1.0379328374377715e-05,
      "loss": 0.4687,
      "step": 125
    },
    {
      "epoch": 3.8181818181818183,
      "grad_norm": 0.4037490959459256,
      "learning_rate": 1.0227631978561057e-05,
      "loss": 0.4829,
      "step": 126
    },
    {
      "epoch": 3.8484848484848486,
      "grad_norm": 0.37672564002456316,
      "learning_rate": 1.0075883152245334e-05,
      "loss": 0.4768,
      "step": 127
    },
    {
      "epoch": 3.878787878787879,
      "grad_norm": 0.426352203255549,
      "learning_rate": 9.92411684775467e-06,
      "loss": 0.478,
      "step": 128
    },
    {
      "epoch": 3.909090909090909,
      "grad_norm": 0.4211175231835601,
      "learning_rate": 9.772368021438943e-06,
      "loss": 0.4775,
      "step": 129
    },
    {
      "epoch": 3.9393939393939394,
      "grad_norm": 0.43524365176680996,
      "learning_rate": 9.620671625622287e-06,
      "loss": 0.4713,
      "step": 130
    },
    {
      "epoch": 3.9696969696969697,
      "grad_norm": 0.3843736026999104,
      "learning_rate": 9.469062600552509e-06,
      "loss": 0.4925,
      "step": 131
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.48339179181993097,
      "learning_rate": 9.317575866353293e-06,
      "loss": 0.4665,
      "step": 132
    },
    {
      "epoch": 4.03030303030303,
      "grad_norm": 0.5279634781072566,
      "learning_rate": 9.166246314981066e-06,
      "loss": 0.4211,
      "step": 133
    },
    {
      "epoch": 4.0606060606060606,
      "grad_norm": 0.4436470190302797,
      "learning_rate": 9.015108802188314e-06,
      "loss": 0.4231,
      "step": 134
    },
    {
      "epoch": 4.090909090909091,
      "grad_norm": 0.5278592288444699,
      "learning_rate": 8.86419813949525e-06,
      "loss": 0.4263,
      "step": 135
    },
    {
      "epoch": 4.121212121212121,
      "grad_norm": 0.5534022673014775,
      "learning_rate": 8.71354908617169e-06,
      "loss": 0.4287,
      "step": 136
    },
    {
      "epoch": 4.151515151515151,
      "grad_norm": 0.5933759838821644,
      "learning_rate": 8.56319634123092e-06,
      "loss": 0.4264,
      "step": 137
    },
    {
      "epoch": 4.181818181818182,
      "grad_norm": 0.449904194974319,
      "learning_rate": 8.413174535437486e-06,
      "loss": 0.4243,
      "step": 138
    },
    {
      "epoch": 4.212121212121212,
      "grad_norm": 0.5631240554354834,
      "learning_rate": 8.263518223330698e-06,
      "loss": 0.4243,
      "step": 139
    },
    {
      "epoch": 4.242424242424242,
      "grad_norm": 0.505305514935825,
      "learning_rate": 8.114261875265643e-06,
      "loss": 0.4127,
      "step": 140
    },
    {
      "epoch": 4.2727272727272725,
      "grad_norm": 0.46731380560841523,
      "learning_rate": 7.965439869473664e-06,
      "loss": 0.418,
      "step": 141
    },
    {
      "epoch": 4.303030303030303,
      "grad_norm": 0.5005375778086428,
      "learning_rate": 7.817086484143987e-06,
      "loss": 0.4178,
      "step": 142
    },
    {
      "epoch": 4.333333333333333,
      "grad_norm": 0.45044163897671435,
      "learning_rate": 7.669235889528436e-06,
      "loss": 0.4312,
      "step": 143
    },
    {
      "epoch": 4.363636363636363,
      "grad_norm": 0.455701381035119,
      "learning_rate": 7.521922140071003e-06,
      "loss": 0.417,
      "step": 144
    },
    {
      "epoch": 4.393939393939394,
      "grad_norm": 0.45329982124885987,
      "learning_rate": 7.375179166564062e-06,
      "loss": 0.419,
      "step": 145
    },
    {
      "epoch": 4.424242424242424,
      "grad_norm": 0.3708224603144236,
      "learning_rate": 7.2290407683331154e-06,
      "loss": 0.4189,
      "step": 146
    },
    {
      "epoch": 4.454545454545454,
      "grad_norm": 0.47660756617939,
      "learning_rate": 7.0835406054517505e-06,
      "loss": 0.4275,
      "step": 147
    },
    {
      "epoch": 4.484848484848484,
      "grad_norm": 0.39472193757784907,
      "learning_rate": 6.93871219098876e-06,
      "loss": 0.4099,
      "step": 148
    },
    {
      "epoch": 4.515151515151516,
      "grad_norm": 0.39541770613408694,
      "learning_rate": 6.79458888328903e-06,
      "loss": 0.4098,
      "step": 149
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 0.4110844497041246,
      "learning_rate": 6.651203878290139e-06,
      "loss": 0.4238,
      "step": 150
    },
    {
      "epoch": 4.575757575757576,
      "grad_norm": 0.3742360711006438,
      "learning_rate": 6.508590201876317e-06,
      "loss": 0.4163,
      "step": 151
    },
    {
      "epoch": 4.606060606060606,
      "grad_norm": 0.3721597153644368,
      "learning_rate": 6.366780702271589e-06,
      "loss": 0.4063,
      "step": 152
    },
    {
      "epoch": 4.636363636363637,
      "grad_norm": 0.3836288611704139,
      "learning_rate": 6.225808042473857e-06,
      "loss": 0.4149,
      "step": 153
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.3571604785404966,
      "learning_rate": 6.085704692731609e-06,
      "loss": 0.4238,
      "step": 154
    },
    {
      "epoch": 4.696969696969697,
      "grad_norm": 0.38605645692446383,
      "learning_rate": 5.946502923065054e-06,
      "loss": 0.4302,
      "step": 155
    },
    {
      "epoch": 4.7272727272727275,
      "grad_norm": 0.40013558195546156,
      "learning_rate": 5.8082347958333625e-06,
      "loss": 0.4183,
      "step": 156
    },
    {
      "epoch": 4.757575757575758,
      "grad_norm": 0.34858846420807044,
      "learning_rate": 5.670932158349732e-06,
      "loss": 0.4201,
      "step": 157
    },
    {
      "epoch": 4.787878787878788,
      "grad_norm": 0.3887049926810658,
      "learning_rate": 5.534626635546e-06,
      "loss": 0.4211,
      "step": 158
    },
    {
      "epoch": 4.818181818181818,
      "grad_norm": 0.3438733953480103,
      "learning_rate": 5.399349622688479e-06,
      "loss": 0.4116,
      "step": 159
    },
    {
      "epoch": 4.848484848484849,
      "grad_norm": 0.33483745130691606,
      "learning_rate": 5.2651322781466606e-06,
      "loss": 0.4259,
      "step": 160
    },
    {
      "epoch": 4.878787878787879,
      "grad_norm": 0.370460838336697,
      "learning_rate": 5.132005516216512e-06,
      "loss": 0.4075,
      "step": 161
    },
    {
      "epoch": 4.909090909090909,
      "grad_norm": 0.3554053268678673,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.4169,
      "step": 162
    },
    {
      "epoch": 4.9393939393939394,
      "grad_norm": 0.3510178818538633,
      "learning_rate": 4.869146134342426e-06,
      "loss": 0.415,
      "step": 163
    },
    {
      "epoch": 4.96969696969697,
      "grad_norm": 0.3624143498739495,
      "learning_rate": 4.739474058829288e-06,
      "loss": 0.4007,
      "step": 164
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.39043974508826756,
      "learning_rate": 4.611013640844245e-06,
      "loss": 0.4081,
      "step": 165
    },
    {
      "epoch": 5.03030303030303,
      "grad_norm": 0.500321819203352,
      "learning_rate": 4.483794468689728e-06,
      "loss": 0.3724,
      "step": 166
    },
    {
      "epoch": 5.0606060606060606,
      "grad_norm": 0.38742563105200084,
      "learning_rate": 4.357845844771881e-06,
      "loss": 0.394,
      "step": 167
    },
    {
      "epoch": 5.090909090909091,
      "grad_norm": 0.35659271286189903,
      "learning_rate": 4.2331967788513295e-06,
      "loss": 0.376,
      "step": 168
    },
    {
      "epoch": 5.121212121212121,
      "grad_norm": 0.45526536046458815,
      "learning_rate": 4.109875981361363e-06,
      "loss": 0.3772,
      "step": 169
    },
    {
      "epoch": 5.151515151515151,
      "grad_norm": 0.4774266907639722,
      "learning_rate": 3.987911856795047e-06,
      "loss": 0.3732,
      "step": 170
    },
    {
      "epoch": 5.181818181818182,
      "grad_norm": 0.4069366748162923,
      "learning_rate": 3.867332497162836e-06,
      "loss": 0.372,
      "step": 171
    },
    {
      "epoch": 5.212121212121212,
      "grad_norm": 0.37203909060573703,
      "learning_rate": 3.748165675522113e-06,
      "loss": 0.3659,
      "step": 172
    },
    {
      "epoch": 5.242424242424242,
      "grad_norm": 0.3906778938941006,
      "learning_rate": 3.630438839580217e-06,
      "loss": 0.3786,
      "step": 173
    },
    {
      "epoch": 5.2727272727272725,
      "grad_norm": 0.3999060713767614,
      "learning_rate": 3.5141791053724405e-06,
      "loss": 0.3801,
      "step": 174
    },
    {
      "epoch": 5.303030303030303,
      "grad_norm": 0.34112453053862585,
      "learning_rate": 3.399413251016359e-06,
      "loss": 0.3742,
      "step": 175
    },
    {
      "epoch": 5.333333333333333,
      "grad_norm": 0.34403607203881237,
      "learning_rate": 3.2861677105440335e-06,
      "loss": 0.3754,
      "step": 176
    },
    {
      "epoch": 5.363636363636363,
      "grad_norm": 0.37937130590116647,
      "learning_rate": 3.174468567813461e-06,
      "loss": 0.3691,
      "step": 177
    },
    {
      "epoch": 5.393939393939394,
      "grad_norm": 0.356627496816996,
      "learning_rate": 3.0643415505006733e-06,
      "loss": 0.3734,
      "step": 178
    },
    {
      "epoch": 5.424242424242424,
      "grad_norm": 0.3430578999175932,
      "learning_rate": 2.9558120241738786e-06,
      "loss": 0.3901,
      "step": 179
    },
    {
      "epoch": 5.454545454545454,
      "grad_norm": 0.321735770236341,
      "learning_rate": 2.8489049864510053e-06,
      "loss": 0.3726,
      "step": 180
    },
    {
      "epoch": 5.484848484848484,
      "grad_norm": 0.33497008802507583,
      "learning_rate": 2.7436450612420098e-06,
      "loss": 0.3679,
      "step": 181
    },
    {
      "epoch": 5.515151515151516,
      "grad_norm": 0.32466689244689567,
      "learning_rate": 2.640056493077231e-06,
      "loss": 0.3817,
      "step": 182
    },
    {
      "epoch": 5.545454545454545,
      "grad_norm": 0.3420941785601393,
      "learning_rate": 2.5381631415231455e-06,
      "loss": 0.3722,
      "step": 183
    },
    {
      "epoch": 5.575757575757576,
      "grad_norm": 0.3336814209841103,
      "learning_rate": 2.4379884756868167e-06,
      "loss": 0.3688,
      "step": 184
    },
    {
      "epoch": 5.606060606060606,
      "grad_norm": 0.29928315276603257,
      "learning_rate": 2.339555568810221e-06,
      "loss": 0.3709,
      "step": 185
    },
    {
      "epoch": 5.636363636363637,
      "grad_norm": 0.33021550031238484,
      "learning_rate": 2.2428870929558012e-06,
      "loss": 0.3673,
      "step": 186
    },
    {
      "epoch": 5.666666666666667,
      "grad_norm": 0.3329496276601939,
      "learning_rate": 2.1480053137844115e-06,
      "loss": 0.3673,
      "step": 187
    },
    {
      "epoch": 5.696969696969697,
      "grad_norm": 0.31928302381605705,
      "learning_rate": 2.054932085426856e-06,
      "loss": 0.3718,
      "step": 188
    },
    {
      "epoch": 5.7272727272727275,
      "grad_norm": 0.3032372254800453,
      "learning_rate": 1.963688845450218e-06,
      "loss": 0.3657,
      "step": 189
    },
    {
      "epoch": 5.757575757575758,
      "grad_norm": 0.3219618396876488,
      "learning_rate": 1.8742966099201699e-06,
      "loss": 0.3609,
      "step": 190
    },
    {
      "epoch": 5.787878787878788,
      "grad_norm": 0.32551348825992976,
      "learning_rate": 1.7867759685603115e-06,
      "loss": 0.3837,
      "step": 191
    },
    {
      "epoch": 5.818181818181818,
      "grad_norm": 0.3110112847612315,
      "learning_rate": 1.7011470800097496e-06,
      "loss": 0.3704,
      "step": 192
    },
    {
      "epoch": 5.848484848484849,
      "grad_norm": 0.3053061517978929,
      "learning_rate": 1.6174296671799571e-06,
      "loss": 0.3634,
      "step": 193
    },
    {
      "epoch": 5.878787878787879,
      "grad_norm": 0.3145203036232622,
      "learning_rate": 1.5356430127119915e-06,
      "loss": 0.3814,
      "step": 194
    },
    {
      "epoch": 5.909090909090909,
      "grad_norm": 0.3135295868777817,
      "learning_rate": 1.4558059545351144e-06,
      "loss": 0.3682,
      "step": 195
    },
    {
      "epoch": 5.9393939393939394,
      "grad_norm": 0.3192500727816256,
      "learning_rate": 1.3779368815278648e-06,
      "loss": 0.3826,
      "step": 196
    },
    {
      "epoch": 5.96969696969697,
      "grad_norm": 0.3091885793694121,
      "learning_rate": 1.302053729282533e-06,
      "loss": 0.3709,
      "step": 197
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.29303232009963087,
      "learning_rate": 1.2281739759740575e-06,
      "loss": 0.3689,
      "step": 198
    },
    {
      "epoch": 6.03030303030303,
      "grad_norm": 0.3748100152999876,
      "learning_rate": 1.156314638334277e-06,
      "loss": 0.3499,
      "step": 199
    },
    {
      "epoch": 6.0606060606060606,
      "grad_norm": 0.3462028800468288,
      "learning_rate": 1.086492267732462e-06,
      "loss": 0.3513,
      "step": 200
    },
    {
      "epoch": 6.090909090909091,
      "grad_norm": 0.32649780791357175,
      "learning_rate": 1.01872294636304e-06,
      "loss": 0.3503,
      "step": 201
    },
    {
      "epoch": 6.121212121212121,
      "grad_norm": 0.3007300980231762,
      "learning_rate": 9.530222835413739e-07,
      "loss": 0.351,
      "step": 202
    },
    {
      "epoch": 6.151515151515151,
      "grad_norm": 0.3093638830508263,
      "learning_rate": 8.894054121084839e-07,
      "loss": 0.361,
      "step": 203
    },
    {
      "epoch": 6.181818181818182,
      "grad_norm": 0.30002724913492534,
      "learning_rate": 8.278869849454718e-07,
      "loss": 0.3441,
      "step": 204
    },
    {
      "epoch": 6.212121212121212,
      "grad_norm": 0.3246993604878837,
      "learning_rate": 7.684811715985429e-07,
      "loss": 0.3628,
      "step": 205
    },
    {
      "epoch": 6.242424242424242,
      "grad_norm": 0.34286891467174785,
      "learning_rate": 7.1120165501533e-07,
      "loss": 0.3544,
      "step": 206
    },
    {
      "epoch": 6.2727272727272725,
      "grad_norm": 0.3025701003519288,
      "learning_rate": 6.560616283932897e-07,
      "loss": 0.3343,
      "step": 207
    },
    {
      "epoch": 6.303030303030303,
      "grad_norm": 0.31448866070115905,
      "learning_rate": 6.030737921409169e-07,
      "loss": 0.3705,
      "step": 208
    },
    {
      "epoch": 6.333333333333333,
      "grad_norm": 0.32306310994716475,
      "learning_rate": 5.522503509524591e-07,
      "loss": 0.347,
      "step": 209
    },
    {
      "epoch": 6.363636363636363,
      "grad_norm": 0.30789964747641346,
      "learning_rate": 5.036030109968082e-07,
      "loss": 0.3705,
      "step": 210
    },
    {
      "epoch": 6.393939393939394,
      "grad_norm": 0.2927540363777904,
      "learning_rate": 4.5714297722121105e-07,
      "loss": 0.3478,
      "step": 211
    },
    {
      "epoch": 6.424242424242424,
      "grad_norm": 0.3054264576098276,
      "learning_rate": 4.128809507704445e-07,
      "loss": 0.3501,
      "step": 212
    },
    {
      "epoch": 6.454545454545454,
      "grad_norm": 0.29901723931255275,
      "learning_rate": 3.708271265220087e-07,
      "loss": 0.3487,
      "step": 213
    },
    {
      "epoch": 6.484848484848484,
      "grad_norm": 0.2915343495455346,
      "learning_rate": 3.309911907379393e-07,
      "loss": 0.3538,
      "step": 214
    },
    {
      "epoch": 6.515151515151516,
      "grad_norm": 0.30431834436921035,
      "learning_rate": 2.9338231883378365e-07,
      "loss": 0.3436,
      "step": 215
    },
    {
      "epoch": 6.545454545454545,
      "grad_norm": 0.3025501634594259,
      "learning_rate": 2.5800917326521013e-07,
      "loss": 0.3525,
      "step": 216
    },
    {
      "epoch": 6.575757575757576,
      "grad_norm": 0.2802045175778162,
      "learning_rate": 2.248799015327907e-07,
      "loss": 0.3544,
      "step": 217
    },
    {
      "epoch": 6.606060606060606,
      "grad_norm": 0.2795898117152192,
      "learning_rate": 1.9400213430538773e-07,
      "loss": 0.3577,
      "step": 218
    },
    {
      "epoch": 6.636363636363637,
      "grad_norm": 0.299850524815354,
      "learning_rate": 1.6538298366257975e-07,
      "loss": 0.3524,
      "step": 219
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.2794865746177794,
      "learning_rate": 1.3902904145653094e-07,
      "loss": 0.3524,
      "step": 220
    },
    {
      "epoch": 6.696969696969697,
      "grad_norm": 0.2912233798526373,
      "learning_rate": 1.1494637779369766e-07,
      "loss": 0.3626,
      "step": 221
    },
    {
      "epoch": 6.7272727272727275,
      "grad_norm": 0.279644199462283,
      "learning_rate": 9.314053963669245e-08,
      "loss": 0.3459,
      "step": 222
    },
    {
      "epoch": 6.757575757575758,
      "grad_norm": 0.2974543807510508,
      "learning_rate": 7.361654952665608e-08,
      "loss": 0.3479,
      "step": 223
    },
    {
      "epoch": 6.787878787878788,
      "grad_norm": 0.28632527564498134,
      "learning_rate": 5.637890442641403e-08,
      "loss": 0.353,
      "step": 224
    },
    {
      "epoch": 6.818181818181818,
      "grad_norm": 0.28611178659862363,
      "learning_rate": 4.143157468468717e-08,
      "loss": 0.353,
      "step": 225
    },
    {
      "epoch": 6.848484848484849,
      "grad_norm": 0.277315210541598,
      "learning_rate": 2.8778003121607834e-08,
      "loss": 0.3589,
      "step": 226
    },
    {
      "epoch": 6.878787878787879,
      "grad_norm": 0.2904200156243484,
      "learning_rate": 1.8421104235727406e-08,
      "loss": 0.3458,
      "step": 227
    },
    {
      "epoch": 6.909090909090909,
      "grad_norm": 0.31573361996908356,
      "learning_rate": 1.0363263532724433e-08,
      "loss": 0.354,
      "step": 228
    },
    {
      "epoch": 6.9393939393939394,
      "grad_norm": 0.2748504776523421,
      "learning_rate": 4.606336975948589e-09,
      "loss": 0.3614,
      "step": 229
    },
    {
      "epoch": 6.96969696969697,
      "grad_norm": 0.2753748072512228,
      "learning_rate": 1.1516505589381777e-09,
      "loss": 0.3533,
      "step": 230
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.284217986264471,
      "learning_rate": 0.0,
      "loss": 0.3485,
      "step": 231
    },
    {
      "epoch": 7.0,
      "step": 231,
      "total_flos": 6.360617938898125e+17,
      "train_loss": 0.5168484355206098,
      "train_runtime": 9139.7772,
      "train_samples_per_second": 2.42,
      "train_steps_per_second": 0.025
    }
  ],
  "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": 6.360617938898125e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}