{
  "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": 5.911785646138455,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.9579,
      "step": 1
    },
    {
      "epoch": 0.06060606060606061,
      "grad_norm": 6.10908942133586,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 0.9654,
      "step": 2
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 6.197197677031692,
      "learning_rate": 2.5e-06,
      "loss": 0.9885,
      "step": 3
    },
    {
      "epoch": 0.12121212121212122,
      "grad_norm": 5.749689497868795,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.9753,
      "step": 4
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 4.592248143944737,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.96,
      "step": 5
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 2.2347773835210307,
      "learning_rate": 5e-06,
      "loss": 0.8921,
      "step": 6
    },
    {
      "epoch": 0.21212121212121213,
      "grad_norm": 1.9953836480371383,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.9136,
      "step": 7
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 4.2302736259558,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.9094,
      "step": 8
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 4.547720386392645,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.9034,
      "step": 9
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 4.674591518721441,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.8838,
      "step": 10
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 4.012859208477787,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.8604,
      "step": 11
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 3.6891785632827623,
      "learning_rate": 1e-05,
      "loss": 0.8573,
      "step": 12
    },
    {
      "epoch": 0.3939393939393939,
      "grad_norm": 2.89726497068403,
      "learning_rate": 1.0833333333333334e-05,
      "loss": 0.8381,
      "step": 13
    },
    {
      "epoch": 0.42424242424242425,
      "grad_norm": 2.4048811114072817,
      "learning_rate": 1.1666666666666668e-05,
      "loss": 0.8238,
      "step": 14
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 2.2154229419910707,
      "learning_rate": 1.25e-05,
      "loss": 0.7985,
      "step": 15
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 1.9983567192589893,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.7599,
      "step": 16
    },
    {
      "epoch": 0.5151515151515151,
      "grad_norm": 1.9983619696833574,
      "learning_rate": 1.416666666666667e-05,
      "loss": 0.7807,
      "step": 17
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 1.5544093781893962,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.7701,
      "step": 18
    },
    {
      "epoch": 0.5757575757575758,
      "grad_norm": 1.4709111937234272,
      "learning_rate": 1.5833333333333333e-05,
      "loss": 0.7659,
      "step": 19
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 1.7444180719999556,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.7429,
      "step": 20
    },
    {
      "epoch": 0.6363636363636364,
      "grad_norm": 1.0899829386656736,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 0.7372,
      "step": 21
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.9532403114419696,
      "learning_rate": 1.8333333333333333e-05,
      "loss": 0.7352,
      "step": 22
    },
    {
      "epoch": 0.696969696969697,
      "grad_norm": 1.174539049795516,
      "learning_rate": 1.916666666666667e-05,
      "loss": 0.7296,
      "step": 23
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.9380022736210596,
      "learning_rate": 2e-05,
      "loss": 0.7313,
      "step": 24
    },
    {
      "epoch": 0.7575757575757576,
      "grad_norm": 0.8350122920364286,
      "learning_rate": 1.999884834944106e-05,
      "loss": 0.7378,
      "step": 25
    },
    {
      "epoch": 0.7878787878787878,
      "grad_norm": 0.9456327142613427,
      "learning_rate": 1.9995393663024054e-05,
      "loss": 0.6765,
      "step": 26
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 0.9694464777247638,
      "learning_rate": 1.9989636736467278e-05,
      "loss": 0.6828,
      "step": 27
    },
    {
      "epoch": 0.8484848484848485,
      "grad_norm": 0.7530578694907432,
      "learning_rate": 1.9981578895764272e-05,
      "loss": 0.7136,
      "step": 28
    },
    {
      "epoch": 0.8787878787878788,
      "grad_norm": 0.8511098107957615,
      "learning_rate": 1.9971221996878395e-05,
      "loss": 0.6958,
      "step": 29
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.667890809308289,
      "learning_rate": 1.9958568425315316e-05,
      "loss": 0.6586,
      "step": 30
    },
    {
      "epoch": 0.9393939393939394,
      "grad_norm": 0.7356080299741005,
      "learning_rate": 1.9943621095573588e-05,
      "loss": 0.6855,
      "step": 31
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.6332824448847514,
      "learning_rate": 1.9926383450473344e-05,
      "loss": 0.7067,
      "step": 32
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.7080487792283089,
      "learning_rate": 1.9906859460363307e-05,
      "loss": 0.6877,
      "step": 33
    },
    {
      "epoch": 1.0303030303030303,
      "grad_norm": 0.6310947145612089,
      "learning_rate": 1.9885053622206305e-05,
      "loss": 0.6305,
      "step": 34
    },
    {
      "epoch": 1.0606060606060606,
      "grad_norm": 0.640042576708934,
      "learning_rate": 1.986097095854347e-05,
      "loss": 0.6323,
      "step": 35
    },
    {
      "epoch": 1.0909090909090908,
      "grad_norm": 0.640164068301215,
      "learning_rate": 1.9834617016337424e-05,
      "loss": 0.6349,
      "step": 36
    },
    {
      "epoch": 1.121212121212121,
      "grad_norm": 0.549357059558253,
      "learning_rate": 1.9805997865694616e-05,
      "loss": 0.6136,
      "step": 37
    },
    {
      "epoch": 1.1515151515151516,
      "grad_norm": 0.66174818771124,
      "learning_rate": 1.9775120098467212e-05,
      "loss": 0.6008,
      "step": 38
    },
    {
      "epoch": 1.1818181818181819,
      "grad_norm": 0.5990297488533517,
      "learning_rate": 1.9741990826734793e-05,
      "loss": 0.632,
      "step": 39
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 0.6781669626264255,
      "learning_rate": 1.970661768116622e-05,
      "loss": 0.639,
      "step": 40
    },
    {
      "epoch": 1.2424242424242424,
      "grad_norm": 0.7505390162384353,
      "learning_rate": 1.9669008809262064e-05,
      "loss": 0.6426,
      "step": 41
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 0.6402496689451667,
      "learning_rate": 1.9629172873477995e-05,
      "loss": 0.6159,
      "step": 42
    },
    {
      "epoch": 1.303030303030303,
      "grad_norm": 0.7584087100559281,
      "learning_rate": 1.9587119049229558e-05,
      "loss": 0.6325,
      "step": 43
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.7610174569360902,
      "learning_rate": 1.954285702277879e-05,
      "loss": 0.5866,
      "step": 44
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.552985862012376,
      "learning_rate": 1.9496396989003195e-05,
      "loss": 0.6254,
      "step": 45
    },
    {
      "epoch": 1.393939393939394,
      "grad_norm": 0.6067437167609007,
      "learning_rate": 1.944774964904754e-05,
      "loss": 0.6024,
      "step": 46
    },
    {
      "epoch": 1.4242424242424243,
      "grad_norm": 0.6630494500038795,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 0.5994,
      "step": 47
    },
    {
      "epoch": 1.4545454545454546,
      "grad_norm": 0.6002506826599772,
      "learning_rate": 1.9343938371606714e-05,
      "loss": 0.5908,
      "step": 48
    },
    {
      "epoch": 1.4848484848484849,
      "grad_norm": 0.6411086640849772,
      "learning_rate": 1.9288798344984673e-05,
      "loss": 0.6174,
      "step": 49
    },
    {
      "epoch": 1.5151515151515151,
      "grad_norm": 0.568865884266851,
      "learning_rate": 1.9231518828401458e-05,
      "loss": 0.6066,
      "step": 50
    },
    {
      "epoch": 1.5454545454545454,
      "grad_norm": 0.6146457864970508,
      "learning_rate": 1.917211301505453e-05,
      "loss": 0.602,
      "step": 51
    },
    {
      "epoch": 1.5757575757575757,
      "grad_norm": 0.6220271594747446,
      "learning_rate": 1.911059458789152e-05,
      "loss": 0.6077,
      "step": 52
    },
    {
      "epoch": 1.606060606060606,
      "grad_norm": 0.6278182053074658,
      "learning_rate": 1.9046977716458627e-05,
      "loss": 0.573,
      "step": 53
    },
    {
      "epoch": 1.6363636363636362,
      "grad_norm": 0.6591979405862676,
      "learning_rate": 1.8981277053636963e-05,
      "loss": 0.6192,
      "step": 54
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.6131363177797416,
      "learning_rate": 1.891350773226754e-05,
      "loss": 0.5774,
      "step": 55
    },
    {
      "epoch": 1.696969696969697,
      "grad_norm": 0.5522686347365365,
      "learning_rate": 1.8843685361665724e-05,
      "loss": 0.5925,
      "step": 56
    },
    {
      "epoch": 1.7272727272727273,
      "grad_norm": 0.598252056486832,
      "learning_rate": 1.8771826024025944e-05,
      "loss": 0.5825,
      "step": 57
    },
    {
      "epoch": 1.7575757575757576,
      "grad_norm": 0.6536266089283411,
      "learning_rate": 1.8697946270717468e-05,
      "loss": 0.6159,
      "step": 58
    },
    {
      "epoch": 1.7878787878787878,
      "grad_norm": 0.5605761527608285,
      "learning_rate": 1.8622063118472135e-05,
      "loss": 0.6132,
      "step": 59
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.6016985675490889,
      "learning_rate": 1.8544194045464888e-05,
      "loss": 0.6238,
      "step": 60
    },
    {
      "epoch": 1.8484848484848486,
      "grad_norm": 0.581872970147029,
      "learning_rate": 1.8464356987288012e-05,
      "loss": 0.5684,
      "step": 61
    },
    {
      "epoch": 1.878787878787879,
      "grad_norm": 0.5081062881102312,
      "learning_rate": 1.8382570332820045e-05,
      "loss": 0.5897,
      "step": 62
    },
    {
      "epoch": 1.9090909090909092,
      "grad_norm": 0.8677787227010229,
      "learning_rate": 1.8298852919990254e-05,
      "loss": 0.6231,
      "step": 63
    },
    {
      "epoch": 1.9393939393939394,
      "grad_norm": 0.5155282934124168,
      "learning_rate": 1.821322403143969e-05,
      "loss": 0.5911,
      "step": 64
    },
    {
      "epoch": 1.9696969696969697,
      "grad_norm": 0.6518788036258234,
      "learning_rate": 1.812570339007983e-05,
      "loss": 0.5612,
      "step": 65
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.5569420873070471,
      "learning_rate": 1.8036311154549783e-05,
      "loss": 0.573,
      "step": 66
    },
    {
      "epoch": 2.0303030303030303,
      "grad_norm": 0.7444593210152036,
      "learning_rate": 1.7945067914573147e-05,
      "loss": 0.5028,
      "step": 67
    },
    {
      "epoch": 2.0606060606060606,
      "grad_norm": 0.656278741432818,
      "learning_rate": 1.7851994686215592e-05,
      "loss": 0.48,
      "step": 68
    },
    {
      "epoch": 2.090909090909091,
      "grad_norm": 0.7817377743059554,
      "learning_rate": 1.77571129070442e-05,
      "loss": 0.4891,
      "step": 69
    },
    {
      "epoch": 2.121212121212121,
      "grad_norm": 0.8191187293597691,
      "learning_rate": 1.766044443118978e-05,
      "loss": 0.48,
      "step": 70
    },
    {
      "epoch": 2.1515151515151514,
      "grad_norm": 0.7074719594212241,
      "learning_rate": 1.7562011524313187e-05,
      "loss": 0.4837,
      "step": 71
    },
    {
      "epoch": 2.1818181818181817,
      "grad_norm": 0.7068344144936601,
      "learning_rate": 1.7461836858476858e-05,
      "loss": 0.4691,
      "step": 72
    },
    {
      "epoch": 2.212121212121212,
      "grad_norm": 0.6317026713493528,
      "learning_rate": 1.7359943506922775e-05,
      "loss": 0.494,
      "step": 73
    },
    {
      "epoch": 2.242424242424242,
      "grad_norm": 0.6582705938264748,
      "learning_rate": 1.725635493875799e-05,
      "loss": 0.4724,
      "step": 74
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.6655616652631317,
      "learning_rate": 1.7151095013548996e-05,
      "loss": 0.4647,
      "step": 75
    },
    {
      "epoch": 2.303030303030303,
      "grad_norm": 0.6334381860960375,
      "learning_rate": 1.7044187975826126e-05,
      "loss": 0.4595,
      "step": 76
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.6430136883957006,
      "learning_rate": 1.693565844949933e-05,
      "loss": 0.4355,
      "step": 77
    },
    {
      "epoch": 2.3636363636363638,
      "grad_norm": 0.5882174360375881,
      "learning_rate": 1.6825531432186545e-05,
      "loss": 0.456,
      "step": 78
    },
    {
      "epoch": 2.393939393939394,
      "grad_norm": 0.5382465082419056,
      "learning_rate": 1.671383228945597e-05,
      "loss": 0.4649,
      "step": 79
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 0.6701569710390706,
      "learning_rate": 1.6600586748983642e-05,
      "loss": 0.4554,
      "step": 80
    },
    {
      "epoch": 2.4545454545454546,
      "grad_norm": 0.563211809443065,
      "learning_rate": 1.648582089462756e-05,
      "loss": 0.4563,
      "step": 81
    },
    {
      "epoch": 2.484848484848485,
      "grad_norm": 0.5059898615689822,
      "learning_rate": 1.6369561160419783e-05,
      "loss": 0.4575,
      "step": 82
    },
    {
      "epoch": 2.515151515151515,
      "grad_norm": 0.6931720899552953,
      "learning_rate": 1.625183432447789e-05,
      "loss": 0.4659,
      "step": 83
    },
    {
      "epoch": 2.5454545454545454,
      "grad_norm": 0.5919809502878728,
      "learning_rate": 1.6132667502837164e-05,
      "loss": 0.4636,
      "step": 84
    },
    {
      "epoch": 2.5757575757575757,
      "grad_norm": 0.5746580610339881,
      "learning_rate": 1.6012088143204953e-05,
      "loss": 0.4648,
      "step": 85
    },
    {
      "epoch": 2.606060606060606,
      "grad_norm": 0.7774355085257642,
      "learning_rate": 1.589012401863864e-05,
      "loss": 0.4688,
      "step": 86
    },
    {
      "epoch": 2.6363636363636362,
      "grad_norm": 0.5232218241002032,
      "learning_rate": 1.5766803221148676e-05,
      "loss": 0.4587,
      "step": 87
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.5780629454806447,
      "learning_rate": 1.5642154155228124e-05,
      "loss": 0.4828,
      "step": 88
    },
    {
      "epoch": 2.6969696969696972,
      "grad_norm": 0.53255062206722,
      "learning_rate": 1.5516205531310272e-05,
      "loss": 0.4993,
      "step": 89
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.5362812178898061,
      "learning_rate": 1.538898635915576e-05,
      "loss": 0.458,
      "step": 90
    },
    {
      "epoch": 2.757575757575758,
      "grad_norm": 0.5149152165149797,
      "learning_rate": 1.526052594117071e-05,
      "loss": 0.4689,
      "step": 91
    },
    {
      "epoch": 2.787878787878788,
      "grad_norm": 0.5696635528999822,
      "learning_rate": 1.513085386565758e-05,
      "loss": 0.4512,
      "step": 92
    },
    {
      "epoch": 2.8181818181818183,
      "grad_norm": 0.5405758768802251,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.4711,
      "step": 93
    },
    {
      "epoch": 2.8484848484848486,
      "grad_norm": 0.5184418557894883,
      "learning_rate": 1.4867994483783485e-05,
      "loss": 0.4679,
      "step": 94
    },
    {
      "epoch": 2.878787878787879,
      "grad_norm": 0.5425615358693525,
      "learning_rate": 1.4734867721853341e-05,
      "loss": 0.4835,
      "step": 95
    },
    {
      "epoch": 2.909090909090909,
      "grad_norm": 0.4479454753558774,
      "learning_rate": 1.4600650377311523e-05,
      "loss": 0.4499,
      "step": 96
    },
    {
      "epoch": 2.9393939393939394,
      "grad_norm": 0.5011469630297277,
      "learning_rate": 1.4465373364454001e-05,
      "loss": 0.4734,
      "step": 97
    },
    {
      "epoch": 2.9696969696969697,
      "grad_norm": 0.5179472172153285,
      "learning_rate": 1.4329067841650274e-05,
      "loss": 0.4602,
      "step": 98
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.5325069335238541,
      "learning_rate": 1.4191765204166643e-05,
      "loss": 0.4668,
      "step": 99
    },
    {
      "epoch": 3.0303030303030303,
      "grad_norm": 0.8144945387236661,
      "learning_rate": 1.4053497076934948e-05,
      "loss": 0.3897,
      "step": 100
    },
    {
      "epoch": 3.0606060606060606,
      "grad_norm": 0.5396003779478772,
      "learning_rate": 1.3914295307268396e-05,
      "loss": 0.3593,
      "step": 101
    },
    {
      "epoch": 3.090909090909091,
      "grad_norm": 0.8877649278485439,
      "learning_rate": 1.3774191957526144e-05,
      "loss": 0.356,
      "step": 102
    },
    {
      "epoch": 3.121212121212121,
      "grad_norm": 0.8139278746547827,
      "learning_rate": 1.3633219297728415e-05,
      "loss": 0.3493,
      "step": 103
    },
    {
      "epoch": 3.1515151515151514,
      "grad_norm": 0.6043020699036067,
      "learning_rate": 1.3491409798123687e-05,
      "loss": 0.3515,
      "step": 104
    },
    {
      "epoch": 3.1818181818181817,
      "grad_norm": 0.6222277215037485,
      "learning_rate": 1.3348796121709862e-05,
      "loss": 0.3457,
      "step": 105
    },
    {
      "epoch": 3.212121212121212,
      "grad_norm": 0.7168149567153531,
      "learning_rate": 1.3205411116710973e-05,
      "loss": 0.3449,
      "step": 106
    },
    {
      "epoch": 3.242424242424242,
      "grad_norm": 0.637751994781398,
      "learning_rate": 1.3061287809011243e-05,
      "loss": 0.3353,
      "step": 107
    },
    {
      "epoch": 3.2727272727272725,
      "grad_norm": 0.5216171180664114,
      "learning_rate": 1.291645939454825e-05,
      "loss": 0.3466,
      "step": 108
    },
    {
      "epoch": 3.303030303030303,
      "grad_norm": 0.6202849800394322,
      "learning_rate": 1.277095923166689e-05,
      "loss": 0.3475,
      "step": 109
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.5646348158339131,
      "learning_rate": 1.2624820833435939e-05,
      "loss": 0.3441,
      "step": 110
    },
    {
      "epoch": 3.3636363636363638,
      "grad_norm": 0.5565926648442473,
      "learning_rate": 1.2478077859929e-05,
      "loss": 0.3437,
      "step": 111
    },
    {
      "epoch": 3.393939393939394,
      "grad_norm": 0.5226883189864145,
      "learning_rate": 1.2330764110471567e-05,
      "loss": 0.3477,
      "step": 112
    },
    {
      "epoch": 3.4242424242424243,
      "grad_norm": 0.5827789798312649,
      "learning_rate": 1.2182913515856016e-05,
      "loss": 0.353,
      "step": 113
    },
    {
      "epoch": 3.4545454545454546,
      "grad_norm": 0.5243200724910204,
      "learning_rate": 1.2034560130526341e-05,
      "loss": 0.3455,
      "step": 114
    },
    {
      "epoch": 3.484848484848485,
      "grad_norm": 0.5536480431071154,
      "learning_rate": 1.1885738124734359e-05,
      "loss": 0.3334,
      "step": 115
    },
    {
      "epoch": 3.515151515151515,
      "grad_norm": 0.548972350231462,
      "learning_rate": 1.1736481776669307e-05,
      "loss": 0.3573,
      "step": 116
    },
    {
      "epoch": 3.5454545454545454,
      "grad_norm": 0.48270896792290413,
      "learning_rate": 1.1586825464562515e-05,
      "loss": 0.3439,
      "step": 117
    },
    {
      "epoch": 3.5757575757575757,
      "grad_norm": 0.567077706262168,
      "learning_rate": 1.1436803658769082e-05,
      "loss": 0.35,
      "step": 118
    },
    {
      "epoch": 3.606060606060606,
      "grad_norm": 0.5144124137290013,
      "learning_rate": 1.1286450913828313e-05,
      "loss": 0.3548,
      "step": 119
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.4436209498744881,
      "learning_rate": 1.113580186050475e-05,
      "loss": 0.3188,
      "step": 120
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.49291165195789344,
      "learning_rate": 1.0984891197811686e-05,
      "loss": 0.3405,
      "step": 121
    },
    {
      "epoch": 3.6969696969696972,
      "grad_norm": 0.5208762603388012,
      "learning_rate": 1.0833753685018935e-05,
      "loss": 0.3529,
      "step": 122
    },
    {
      "epoch": 3.7272727272727275,
      "grad_norm": 0.47607641340543305,
      "learning_rate": 1.0682424133646712e-05,
      "loss": 0.3501,
      "step": 123
    },
    {
      "epoch": 3.757575757575758,
      "grad_norm": 0.5457373945672145,
      "learning_rate": 1.0530937399447496e-05,
      "loss": 0.3236,
      "step": 124
    },
    {
      "epoch": 3.787878787878788,
      "grad_norm": 0.4467862989799738,
      "learning_rate": 1.0379328374377715e-05,
      "loss": 0.3445,
      "step": 125
    },
    {
      "epoch": 3.8181818181818183,
      "grad_norm": 0.512414810086636,
      "learning_rate": 1.0227631978561057e-05,
      "loss": 0.3496,
      "step": 126
    },
    {
      "epoch": 3.8484848484848486,
      "grad_norm": 0.46230476080548805,
      "learning_rate": 1.0075883152245334e-05,
      "loss": 0.336,
      "step": 127
    },
    {
      "epoch": 3.878787878787879,
      "grad_norm": 0.4537534517758466,
      "learning_rate": 9.92411684775467e-06,
      "loss": 0.3491,
      "step": 128
    },
    {
      "epoch": 3.909090909090909,
      "grad_norm": 0.5000872440393878,
      "learning_rate": 9.772368021438943e-06,
      "loss": 0.3578,
      "step": 129
    },
    {
      "epoch": 3.9393939393939394,
      "grad_norm": 0.4387456725437304,
      "learning_rate": 9.620671625622287e-06,
      "loss": 0.3367,
      "step": 130
    },
    {
      "epoch": 3.9696969696969697,
      "grad_norm": 0.41342425086989293,
      "learning_rate": 9.469062600552509e-06,
      "loss": 0.3234,
      "step": 131
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.476628783740727,
      "learning_rate": 9.317575866353293e-06,
      "loss": 0.343,
      "step": 132
    },
    {
      "epoch": 4.03030303030303,
      "grad_norm": 0.7814733734576866,
      "learning_rate": 9.166246314981066e-06,
      "loss": 0.2675,
      "step": 133
    },
    {
      "epoch": 4.0606060606060606,
      "grad_norm": 0.5196478172042786,
      "learning_rate": 9.015108802188314e-06,
      "loss": 0.2616,
      "step": 134
    },
    {
      "epoch": 4.090909090909091,
      "grad_norm": 0.5333998505254128,
      "learning_rate": 8.86419813949525e-06,
      "loss": 0.2555,
      "step": 135
    },
    {
      "epoch": 4.121212121212121,
      "grad_norm": 0.6935125106454675,
      "learning_rate": 8.71354908617169e-06,
      "loss": 0.2325,
      "step": 136
    },
    {
      "epoch": 4.151515151515151,
      "grad_norm": 0.671247769837199,
      "learning_rate": 8.56319634123092e-06,
      "loss": 0.2482,
      "step": 137
    },
    {
      "epoch": 4.181818181818182,
      "grad_norm": 0.5271953207775844,
      "learning_rate": 8.413174535437486e-06,
      "loss": 0.2354,
      "step": 138
    },
    {
      "epoch": 4.212121212121212,
      "grad_norm": 0.49792813980403605,
      "learning_rate": 8.263518223330698e-06,
      "loss": 0.2227,
      "step": 139
    },
    {
      "epoch": 4.242424242424242,
      "grad_norm": 0.5545812268318249,
      "learning_rate": 8.114261875265643e-06,
      "loss": 0.2437,
      "step": 140
    },
    {
      "epoch": 4.2727272727272725,
      "grad_norm": 0.5189292334554685,
      "learning_rate": 7.965439869473664e-06,
      "loss": 0.2397,
      "step": 141
    },
    {
      "epoch": 4.303030303030303,
      "grad_norm": 0.5026506405209156,
      "learning_rate": 7.817086484143987e-06,
      "loss": 0.2571,
      "step": 142
    },
    {
      "epoch": 4.333333333333333,
      "grad_norm": 0.5293946227145233,
      "learning_rate": 7.669235889528436e-06,
      "loss": 0.2554,
      "step": 143
    },
    {
      "epoch": 4.363636363636363,
      "grad_norm": 0.4901035203574336,
      "learning_rate": 7.521922140071003e-06,
      "loss": 0.2317,
      "step": 144
    },
    {
      "epoch": 4.393939393939394,
      "grad_norm": 0.5133407183206936,
      "learning_rate": 7.375179166564062e-06,
      "loss": 0.2366,
      "step": 145
    },
    {
      "epoch": 4.424242424242424,
      "grad_norm": 0.49042273995374663,
      "learning_rate": 7.2290407683331154e-06,
      "loss": 0.2403,
      "step": 146
    },
    {
      "epoch": 4.454545454545454,
      "grad_norm": 0.4646179177875379,
      "learning_rate": 7.0835406054517505e-06,
      "loss": 0.2384,
      "step": 147
    },
    {
      "epoch": 4.484848484848484,
      "grad_norm": 0.490995589807223,
      "learning_rate": 6.93871219098876e-06,
      "loss": 0.2515,
      "step": 148
    },
    {
      "epoch": 4.515151515151516,
      "grad_norm": 0.4481752571310081,
      "learning_rate": 6.79458888328903e-06,
      "loss": 0.2499,
      "step": 149
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 0.45248717972131547,
      "learning_rate": 6.651203878290139e-06,
      "loss": 0.2694,
      "step": 150
    },
    {
      "epoch": 4.575757575757576,
      "grad_norm": 0.4484116826981272,
      "learning_rate": 6.508590201876317e-06,
      "loss": 0.2548,
      "step": 151
    },
    {
      "epoch": 4.606060606060606,
      "grad_norm": 0.44786119710270594,
      "learning_rate": 6.366780702271589e-06,
      "loss": 0.2418,
      "step": 152
    },
    {
      "epoch": 4.636363636363637,
      "grad_norm": 0.44442314205150696,
      "learning_rate": 6.225808042473857e-06,
      "loss": 0.2529,
      "step": 153
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.42936394333813915,
      "learning_rate": 6.085704692731609e-06,
      "loss": 0.2473,
      "step": 154
    },
    {
      "epoch": 4.696969696969697,
      "grad_norm": 0.4762809627623751,
      "learning_rate": 5.946502923065054e-06,
      "loss": 0.2457,
      "step": 155
    },
    {
      "epoch": 4.7272727272727275,
      "grad_norm": 0.4516342117700447,
      "learning_rate": 5.8082347958333625e-06,
      "loss": 0.2281,
      "step": 156
    },
    {
      "epoch": 4.757575757575758,
      "grad_norm": 0.4330609382571187,
      "learning_rate": 5.670932158349732e-06,
      "loss": 0.2391,
      "step": 157
    },
    {
      "epoch": 4.787878787878788,
      "grad_norm": 0.4333909958224515,
      "learning_rate": 5.534626635546e-06,
      "loss": 0.2557,
      "step": 158
    },
    {
      "epoch": 4.818181818181818,
      "grad_norm": 0.46172425631051084,
      "learning_rate": 5.399349622688479e-06,
      "loss": 0.2373,
      "step": 159
    },
    {
      "epoch": 4.848484848484849,
      "grad_norm": 0.42412717241351194,
      "learning_rate": 5.2651322781466606e-06,
      "loss": 0.2534,
      "step": 160
    },
    {
      "epoch": 4.878787878787879,
      "grad_norm": 0.4248323005409566,
      "learning_rate": 5.132005516216512e-06,
      "loss": 0.2364,
      "step": 161
    },
    {
      "epoch": 4.909090909090909,
      "grad_norm": 0.4647219154562407,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.2512,
      "step": 162
    },
    {
      "epoch": 4.9393939393939394,
      "grad_norm": 0.4404885328600592,
      "learning_rate": 4.869146134342426e-06,
      "loss": 0.242,
      "step": 163
    },
    {
      "epoch": 4.96969696969697,
      "grad_norm": 0.4543486789472826,
      "learning_rate": 4.739474058829288e-06,
      "loss": 0.2373,
      "step": 164
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.40078315050081065,
      "learning_rate": 4.611013640844245e-06,
      "loss": 0.236,
      "step": 165
    },
    {
      "epoch": 5.03030303030303,
      "grad_norm": 0.6398244861809473,
      "learning_rate": 4.483794468689728e-06,
      "loss": 0.1825,
      "step": 166
    },
    {
      "epoch": 5.0606060606060606,
      "grad_norm": 0.5956245505192638,
      "learning_rate": 4.357845844771881e-06,
      "loss": 0.181,
      "step": 167
    },
    {
      "epoch": 5.090909090909091,
      "grad_norm": 0.4252211478908562,
      "learning_rate": 4.2331967788513295e-06,
      "loss": 0.1868,
      "step": 168
    },
    {
      "epoch": 5.121212121212121,
      "grad_norm": 0.48920975374837017,
      "learning_rate": 4.109875981361363e-06,
      "loss": 0.1711,
      "step": 169
    },
    {
      "epoch": 5.151515151515151,
      "grad_norm": 0.6712567453092636,
      "learning_rate": 3.987911856795047e-06,
      "loss": 0.1803,
      "step": 170
    },
    {
      "epoch": 5.181818181818182,
      "grad_norm": 0.6538285831990102,
      "learning_rate": 3.867332497162836e-06,
      "loss": 0.1779,
      "step": 171
    },
    {
      "epoch": 5.212121212121212,
      "grad_norm": 0.5155599061675556,
      "learning_rate": 3.748165675522113e-06,
      "loss": 0.1795,
      "step": 172
    },
    {
      "epoch": 5.242424242424242,
      "grad_norm": 0.4526422533609826,
      "learning_rate": 3.630438839580217e-06,
      "loss": 0.1779,
      "step": 173
    },
    {
      "epoch": 5.2727272727272725,
      "grad_norm": 0.4655761671450797,
      "learning_rate": 3.5141791053724405e-06,
      "loss": 0.1775,
      "step": 174
    },
    {
      "epoch": 5.303030303030303,
      "grad_norm": 0.47024523733481727,
      "learning_rate": 3.399413251016359e-06,
      "loss": 0.1572,
      "step": 175
    },
    {
      "epoch": 5.333333333333333,
      "grad_norm": 0.4803636326659098,
      "learning_rate": 3.2861677105440335e-06,
      "loss": 0.1785,
      "step": 176
    },
    {
      "epoch": 5.363636363636363,
      "grad_norm": 0.43273478141674415,
      "learning_rate": 3.174468567813461e-06,
      "loss": 0.183,
      "step": 177
    },
    {
      "epoch": 5.393939393939394,
      "grad_norm": 0.4411064451388547,
      "learning_rate": 3.0643415505006733e-06,
      "loss": 0.1647,
      "step": 178
    },
    {
      "epoch": 5.424242424242424,
      "grad_norm": 0.4364241978900685,
      "learning_rate": 2.9558120241738786e-06,
      "loss": 0.1755,
      "step": 179
    },
    {
      "epoch": 5.454545454545454,
      "grad_norm": 0.4379715790156301,
      "learning_rate": 2.8489049864510053e-06,
      "loss": 0.1903,
      "step": 180
    },
    {
      "epoch": 5.484848484848484,
      "grad_norm": 0.4636494810477295,
      "learning_rate": 2.7436450612420098e-06,
      "loss": 0.1726,
      "step": 181
    },
    {
      "epoch": 5.515151515151516,
      "grad_norm": 0.41243404665522404,
      "learning_rate": 2.640056493077231e-06,
      "loss": 0.1766,
      "step": 182
    },
    {
      "epoch": 5.545454545454545,
      "grad_norm": 0.44104390693869927,
      "learning_rate": 2.5381631415231455e-06,
      "loss": 0.1956,
      "step": 183
    },
    {
      "epoch": 5.575757575757576,
      "grad_norm": 0.432638721185701,
      "learning_rate": 2.4379884756868167e-06,
      "loss": 0.172,
      "step": 184
    },
    {
      "epoch": 5.606060606060606,
      "grad_norm": 0.43404913166177717,
      "learning_rate": 2.339555568810221e-06,
      "loss": 0.1701,
      "step": 185
    },
    {
      "epoch": 5.636363636363637,
      "grad_norm": 0.4025137455219994,
      "learning_rate": 2.2428870929558012e-06,
      "loss": 0.1771,
      "step": 186
    },
    {
      "epoch": 5.666666666666667,
      "grad_norm": 0.41144620163391915,
      "learning_rate": 2.1480053137844115e-06,
      "loss": 0.1753,
      "step": 187
    },
    {
      "epoch": 5.696969696969697,
      "grad_norm": 0.39948488132969573,
      "learning_rate": 2.054932085426856e-06,
      "loss": 0.1722,
      "step": 188
    },
    {
      "epoch": 5.7272727272727275,
      "grad_norm": 0.4209563447043757,
      "learning_rate": 1.963688845450218e-06,
      "loss": 0.1954,
      "step": 189
    },
    {
      "epoch": 5.757575757575758,
      "grad_norm": 0.395511485565277,
      "learning_rate": 1.8742966099201699e-06,
      "loss": 0.1851,
      "step": 190
    },
    {
      "epoch": 5.787878787878788,
      "grad_norm": 0.42205592811179804,
      "learning_rate": 1.7867759685603115e-06,
      "loss": 0.1735,
      "step": 191
    },
    {
      "epoch": 5.818181818181818,
      "grad_norm": 0.36937503213957223,
      "learning_rate": 1.7011470800097496e-06,
      "loss": 0.2093,
      "step": 192
    },
    {
      "epoch": 5.848484848484849,
      "grad_norm": 0.409697260191631,
      "learning_rate": 1.6174296671799571e-06,
      "loss": 0.1612,
      "step": 193
    },
    {
      "epoch": 5.878787878787879,
      "grad_norm": 0.4009937685501776,
      "learning_rate": 1.5356430127119915e-06,
      "loss": 0.1766,
      "step": 194
    },
    {
      "epoch": 5.909090909090909,
      "grad_norm": 0.37426211969441875,
      "learning_rate": 1.4558059545351144e-06,
      "loss": 0.1749,
      "step": 195
    },
    {
      "epoch": 5.9393939393939394,
      "grad_norm": 0.39087360862483345,
      "learning_rate": 1.3779368815278648e-06,
      "loss": 0.1806,
      "step": 196
    },
    {
      "epoch": 5.96969696969697,
      "grad_norm": 0.4171806327561228,
      "learning_rate": 1.302053729282533e-06,
      "loss": 0.1699,
      "step": 197
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.38389635808561096,
      "learning_rate": 1.2281739759740575e-06,
      "loss": 0.1765,
      "step": 198
    },
    {
      "epoch": 6.03030303030303,
      "grad_norm": 0.4605752777305765,
      "learning_rate": 1.156314638334277e-06,
      "loss": 0.1561,
      "step": 199
    },
    {
      "epoch": 6.0606060606060606,
      "grad_norm": 0.42960047473456353,
      "learning_rate": 1.086492267732462e-06,
      "loss": 0.1375,
      "step": 200
    },
    {
      "epoch": 6.090909090909091,
      "grad_norm": 0.4589389179823451,
      "learning_rate": 1.01872294636304e-06,
      "loss": 0.14,
      "step": 201
    },
    {
      "epoch": 6.121212121212121,
      "grad_norm": 0.3921248698033923,
      "learning_rate": 9.530222835413739e-07,
      "loss": 0.1574,
      "step": 202
    },
    {
      "epoch": 6.151515151515151,
      "grad_norm": 0.33922312146919553,
      "learning_rate": 8.894054121084839e-07,
      "loss": 0.1486,
      "step": 203
    },
    {
      "epoch": 6.181818181818182,
      "grad_norm": 0.3490823688216288,
      "learning_rate": 8.278869849454718e-07,
      "loss": 0.1366,
      "step": 204
    },
    {
      "epoch": 6.212121212121212,
      "grad_norm": 0.36703451049993513,
      "learning_rate": 7.684811715985429e-07,
      "loss": 0.1477,
      "step": 205
    },
    {
      "epoch": 6.242424242424242,
      "grad_norm": 0.40089957684640243,
      "learning_rate": 7.1120165501533e-07,
      "loss": 0.1519,
      "step": 206
    },
    {
      "epoch": 6.2727272727272725,
      "grad_norm": 0.38159839250036487,
      "learning_rate": 6.560616283932897e-07,
      "loss": 0.1554,
      "step": 207
    },
    {
      "epoch": 6.303030303030303,
      "grad_norm": 0.4231466315889625,
      "learning_rate": 6.030737921409169e-07,
      "loss": 0.1473,
      "step": 208
    },
    {
      "epoch": 6.333333333333333,
      "grad_norm": 0.3780150822266596,
      "learning_rate": 5.522503509524591e-07,
      "loss": 0.1345,
      "step": 209
    },
    {
      "epoch": 6.363636363636363,
      "grad_norm": 0.4263629460032663,
      "learning_rate": 5.036030109968082e-07,
      "loss": 0.1637,
      "step": 210
    },
    {
      "epoch": 6.393939393939394,
      "grad_norm": 0.381616730352659,
      "learning_rate": 4.5714297722121105e-07,
      "loss": 0.1445,
      "step": 211
    },
    {
      "epoch": 6.424242424242424,
      "grad_norm": 0.3666003560631986,
      "learning_rate": 4.128809507704445e-07,
      "loss": 0.1423,
      "step": 212
    },
    {
      "epoch": 6.454545454545454,
      "grad_norm": 0.37537715839109853,
      "learning_rate": 3.708271265220087e-07,
      "loss": 0.1406,
      "step": 213
    },
    {
      "epoch": 6.484848484848484,
      "grad_norm": 0.35890121413550524,
      "learning_rate": 3.309911907379393e-07,
      "loss": 0.1486,
      "step": 214
    },
    {
      "epoch": 6.515151515151516,
      "grad_norm": 0.3616812476930965,
      "learning_rate": 2.9338231883378365e-07,
      "loss": 0.1448,
      "step": 215
    },
    {
      "epoch": 6.545454545454545,
      "grad_norm": 0.3496192361731834,
      "learning_rate": 2.5800917326521013e-07,
      "loss": 0.1554,
      "step": 216
    },
    {
      "epoch": 6.575757575757576,
      "grad_norm": 0.35159497931952804,
      "learning_rate": 2.248799015327907e-07,
      "loss": 0.167,
      "step": 217
    },
    {
      "epoch": 6.606060606060606,
      "grad_norm": 0.37861842843551974,
      "learning_rate": 1.9400213430538773e-07,
      "loss": 0.1502,
      "step": 218
    },
    {
      "epoch": 6.636363636363637,
      "grad_norm": 0.38111446611123034,
      "learning_rate": 1.6538298366257975e-07,
      "loss": 0.1535,
      "step": 219
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.3675257466844091,
      "learning_rate": 1.3902904145653094e-07,
      "loss": 0.1372,
      "step": 220
    },
    {
      "epoch": 6.696969696969697,
      "grad_norm": 0.35174952729723796,
      "learning_rate": 1.1494637779369766e-07,
      "loss": 0.1454,
      "step": 221
    },
    {
      "epoch": 6.7272727272727275,
      "grad_norm": 0.3920175542582645,
      "learning_rate": 9.314053963669245e-08,
      "loss": 0.1677,
      "step": 222
    },
    {
      "epoch": 6.757575757575758,
      "grad_norm": 0.37163219362706756,
      "learning_rate": 7.361654952665608e-08,
      "loss": 0.1482,
      "step": 223
    },
    {
      "epoch": 6.787878787878788,
      "grad_norm": 0.35550074485597094,
      "learning_rate": 5.637890442641403e-08,
      "loss": 0.1601,
      "step": 224
    },
    {
      "epoch": 6.818181818181818,
      "grad_norm": 0.3566396940906455,
      "learning_rate": 4.143157468468717e-08,
      "loss": 0.1553,
      "step": 225
    },
    {
      "epoch": 6.848484848484849,
      "grad_norm": 0.3695021023233243,
      "learning_rate": 2.8778003121607834e-08,
      "loss": 0.1557,
      "step": 226
    },
    {
      "epoch": 6.878787878787879,
      "grad_norm": 0.34252158509950414,
      "learning_rate": 1.8421104235727406e-08,
      "loss": 0.1431,
      "step": 227
    },
    {
      "epoch": 6.909090909090909,
      "grad_norm": 0.3511628228100589,
      "learning_rate": 1.0363263532724433e-08,
      "loss": 0.144,
      "step": 228
    },
    {
      "epoch": 6.9393939393939394,
      "grad_norm": 0.34268597297268333,
      "learning_rate": 4.606336975948589e-09,
      "loss": 0.1441,
      "step": 229
    },
    {
      "epoch": 6.96969696969697,
      "grad_norm": 0.3734690753342464,
      "learning_rate": 1.1516505589381777e-09,
      "loss": 0.1499,
      "step": 230
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.35324584839751283,
      "learning_rate": 0.0,
      "loss": 0.1582,
      "step": 231
    },
    {
      "epoch": 7.0,
      "step": 231,
      "total_flos": 3.503726093611827e+17,
      "train_loss": 0.3997843897703922,
      "train_runtime": 6633.8315,
      "train_samples_per_second": 3.334,
      "train_steps_per_second": 0.035
    }
  ],
  "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": 3.503726093611827e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}