{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 16.93658536585366,
  "eval_steps": 500,
  "global_step": 136,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.11707317073170732,
      "grad_norm": 3.068251132965088,
      "learning_rate": 7.142857142857143e-07,
      "loss": 0.7283,
      "step": 1
    },
    {
      "epoch": 0.23414634146341465,
      "grad_norm": 3.0890955924987793,
      "learning_rate": 1.4285714285714286e-06,
      "loss": 0.7468,
      "step": 2
    },
    {
      "epoch": 0.35121951219512193,
      "grad_norm": 2.848057746887207,
      "learning_rate": 2.1428571428571427e-06,
      "loss": 0.7287,
      "step": 3
    },
    {
      "epoch": 0.4682926829268293,
      "grad_norm": 2.817434549331665,
      "learning_rate": 2.8571428571428573e-06,
      "loss": 0.7216,
      "step": 4
    },
    {
      "epoch": 0.5853658536585366,
      "grad_norm": 2.3008267879486084,
      "learning_rate": 3.5714285714285718e-06,
      "loss": 0.6971,
      "step": 5
    },
    {
      "epoch": 0.7024390243902439,
      "grad_norm": 2.112135887145996,
      "learning_rate": 4.2857142857142855e-06,
      "loss": 0.7117,
      "step": 6
    },
    {
      "epoch": 0.8195121951219512,
      "grad_norm": 1.6284232139587402,
      "learning_rate": 5e-06,
      "loss": 0.6701,
      "step": 7
    },
    {
      "epoch": 0.9365853658536586,
      "grad_norm": 1.4377871751785278,
      "learning_rate": 5.7142857142857145e-06,
      "loss": 0.6756,
      "step": 8
    },
    {
      "epoch": 1.1170731707317074,
      "grad_norm": 2.7016568183898926,
      "learning_rate": 6.4285714285714295e-06,
      "loss": 1.2659,
      "step": 9
    },
    {
      "epoch": 1.2341463414634146,
      "grad_norm": 1.4025566577911377,
      "learning_rate": 7.1428571428571436e-06,
      "loss": 0.629,
      "step": 10
    },
    {
      "epoch": 1.3512195121951218,
      "grad_norm": 1.0815410614013672,
      "learning_rate": 7.857142857142858e-06,
      "loss": 0.5818,
      "step": 11
    },
    {
      "epoch": 1.4682926829268292,
      "grad_norm": 1.4897761344909668,
      "learning_rate": 8.571428571428571e-06,
      "loss": 0.5669,
      "step": 12
    },
    {
      "epoch": 1.5853658536585367,
      "grad_norm": 1.0598845481872559,
      "learning_rate": 9.285714285714288e-06,
      "loss": 0.5562,
      "step": 13
    },
    {
      "epoch": 1.7024390243902439,
      "grad_norm": 0.800104022026062,
      "learning_rate": 1e-05,
      "loss": 0.5349,
      "step": 14
    },
    {
      "epoch": 1.819512195121951,
      "grad_norm": 0.7887536883354187,
      "learning_rate": 9.998342337571566e-06,
      "loss": 0.5225,
      "step": 15
    },
    {
      "epoch": 1.9365853658536585,
      "grad_norm": 1.0338839292526245,
      "learning_rate": 9.993370449424153e-06,
      "loss": 0.5276,
      "step": 16
    },
    {
      "epoch": 2.1170731707317074,
      "grad_norm": 1.6621471643447876,
      "learning_rate": 9.985087632242634e-06,
      "loss": 1.014,
      "step": 17
    },
    {
      "epoch": 2.234146341463415,
      "grad_norm": 0.6436282992362976,
      "learning_rate": 9.973499378072947e-06,
      "loss": 0.4925,
      "step": 18
    },
    {
      "epoch": 2.351219512195122,
      "grad_norm": 0.5632502436637878,
      "learning_rate": 9.958613370680507e-06,
      "loss": 0.4723,
      "step": 19
    },
    {
      "epoch": 2.4682926829268292,
      "grad_norm": 0.5272106528282166,
      "learning_rate": 9.940439480455386e-06,
      "loss": 0.4697,
      "step": 20
    },
    {
      "epoch": 2.5853658536585367,
      "grad_norm": 0.5168065428733826,
      "learning_rate": 9.918989757867584e-06,
      "loss": 0.4653,
      "step": 21
    },
    {
      "epoch": 2.7024390243902436,
      "grad_norm": 0.545330286026001,
      "learning_rate": 9.89427842547679e-06,
      "loss": 0.4437,
      "step": 22
    },
    {
      "epoch": 2.819512195121951,
      "grad_norm": 0.44697803258895874,
      "learning_rate": 9.866321868501914e-06,
      "loss": 0.4645,
      "step": 23
    },
    {
      "epoch": 2.9365853658536585,
      "grad_norm": 0.36716875433921814,
      "learning_rate": 9.835138623956603e-06,
      "loss": 0.4637,
      "step": 24
    },
    {
      "epoch": 3.1170731707317074,
      "grad_norm": 0.8350607752799988,
      "learning_rate": 9.80074936835801e-06,
      "loss": 0.8482,
      "step": 25
    },
    {
      "epoch": 3.234146341463415,
      "grad_norm": 0.34782156348228455,
      "learning_rate": 9.763176904016914e-06,
      "loss": 0.4335,
      "step": 26
    },
    {
      "epoch": 3.351219512195122,
      "grad_norm": 0.38511568307876587,
      "learning_rate": 9.722446143918307e-06,
      "loss": 0.4296,
      "step": 27
    },
    {
      "epoch": 3.4682926829268292,
      "grad_norm": 0.3624374270439148,
      "learning_rate": 9.678584095202468e-06,
      "loss": 0.4104,
      "step": 28
    },
    {
      "epoch": 3.5853658536585367,
      "grad_norm": 0.32947778701782227,
      "learning_rate": 9.631619841257477e-06,
      "loss": 0.3993,
      "step": 29
    },
    {
      "epoch": 3.7024390243902436,
      "grad_norm": 0.34152206778526306,
      "learning_rate": 9.581584522435025e-06,
      "loss": 0.4123,
      "step": 30
    },
    {
      "epoch": 3.819512195121951,
      "grad_norm": 0.36183634400367737,
      "learning_rate": 9.528511315402358e-06,
      "loss": 0.4022,
      "step": 31
    },
    {
      "epoch": 3.9365853658536585,
      "grad_norm": 0.350755512714386,
      "learning_rate": 9.472435411143979e-06,
      "loss": 0.4049,
      "step": 32
    },
    {
      "epoch": 4.117073170731707,
      "grad_norm": 0.6949635744094849,
      "learning_rate": 9.413393991627737e-06,
      "loss": 0.7805,
      "step": 33
    },
    {
      "epoch": 4.234146341463415,
      "grad_norm": 0.28440478444099426,
      "learning_rate": 9.351426205150778e-06,
      "loss": 0.3658,
      "step": 34
    },
    {
      "epoch": 4.351219512195122,
      "grad_norm": 0.35889554023742676,
      "learning_rate": 9.286573140381663e-06,
      "loss": 0.3743,
      "step": 35
    },
    {
      "epoch": 4.46829268292683,
      "grad_norm": 0.3411248028278351,
      "learning_rate": 9.218877799115929e-06,
      "loss": 0.3684,
      "step": 36
    },
    {
      "epoch": 4.585365853658536,
      "grad_norm": 0.3742453455924988,
      "learning_rate": 9.148385067763094e-06,
      "loss": 0.3754,
      "step": 37
    },
    {
      "epoch": 4.702439024390244,
      "grad_norm": 0.30102893710136414,
      "learning_rate": 9.075141687584056e-06,
      "loss": 0.3595,
      "step": 38
    },
    {
      "epoch": 4.819512195121951,
      "grad_norm": 0.3214716911315918,
      "learning_rate": 8.999196223698599e-06,
      "loss": 0.3792,
      "step": 39
    },
    {
      "epoch": 4.9365853658536585,
      "grad_norm": 0.29708611965179443,
      "learning_rate": 8.920599032883553e-06,
      "loss": 0.3585,
      "step": 40
    },
    {
      "epoch": 5.117073170731707,
      "grad_norm": 0.6357777714729309,
      "learning_rate": 8.839402230183e-06,
      "loss": 0.7275,
      "step": 41
    },
    {
      "epoch": 5.234146341463415,
      "grad_norm": 0.29909756779670715,
      "learning_rate": 8.755659654352599e-06,
      "loss": 0.3258,
      "step": 42
    },
    {
      "epoch": 5.351219512195122,
      "grad_norm": 0.3142697513103485,
      "learning_rate": 8.669426832160997e-06,
      "loss": 0.3419,
      "step": 43
    },
    {
      "epoch": 5.46829268292683,
      "grad_norm": 0.31918543577194214,
      "learning_rate": 8.580760941571968e-06,
      "loss": 0.3464,
      "step": 44
    },
    {
      "epoch": 5.585365853658536,
      "grad_norm": 0.29149070382118225,
      "learning_rate": 8.489720773831717e-06,
      "loss": 0.3103,
      "step": 45
    },
    {
      "epoch": 5.702439024390244,
      "grad_norm": 0.3091978132724762,
      "learning_rate": 8.396366694486466e-06,
      "loss": 0.323,
      "step": 46
    },
    {
      "epoch": 5.819512195121951,
      "grad_norm": 0.33374419808387756,
      "learning_rate": 8.30076060335616e-06,
      "loss": 0.3192,
      "step": 47
    },
    {
      "epoch": 5.9365853658536585,
      "grad_norm": 0.2958971858024597,
      "learning_rate": 8.202965893490877e-06,
      "loss": 0.3166,
      "step": 48
    },
    {
      "epoch": 6.117073170731707,
      "grad_norm": 0.2610955536365509,
      "learning_rate": 8.103047409137114e-06,
      "loss": 0.6359,
      "step": 49
    },
    {
      "epoch": 6.234146341463415,
      "grad_norm": 0.30614060163497925,
      "learning_rate": 8.001071402741843e-06,
      "loss": 0.2911,
      "step": 50
    },
    {
      "epoch": 6.351219512195122,
      "grad_norm": 0.28601980209350586,
      "learning_rate": 7.897105491022819e-06,
      "loss": 0.2911,
      "step": 51
    },
    {
      "epoch": 6.46829268292683,
      "grad_norm": 0.288555771112442,
      "learning_rate": 7.791218610134324e-06,
      "loss": 0.2929,
      "step": 52
    },
    {
      "epoch": 6.585365853658536,
      "grad_norm": 0.273262619972229,
      "learning_rate": 7.683480969958005e-06,
      "loss": 0.2856,
      "step": 53
    },
    {
      "epoch": 6.702439024390244,
      "grad_norm": 0.27246272563934326,
      "learning_rate": 7.5739640075491546e-06,
      "loss": 0.2911,
      "step": 54
    },
    {
      "epoch": 6.819512195121951,
      "grad_norm": 0.24324610829353333,
      "learning_rate": 7.462740339769323e-06,
      "loss": 0.2754,
      "step": 55
    },
    {
      "epoch": 6.9365853658536585,
      "grad_norm": 0.28536200523376465,
      "learning_rate": 7.349883715136601e-06,
      "loss": 0.284,
      "step": 56
    },
    {
      "epoch": 7.117073170731707,
      "grad_norm": 0.5948038697242737,
      "learning_rate": 7.235468964925571e-06,
      "loss": 0.5049,
      "step": 57
    },
    {
      "epoch": 7.234146341463415,
      "grad_norm": 0.3587504029273987,
      "learning_rate": 7.119571953549305e-06,
      "loss": 0.2436,
      "step": 58
    },
    {
      "epoch": 7.351219512195122,
      "grad_norm": 0.2773548364639282,
      "learning_rate": 7.002269528256334e-06,
      "loss": 0.2551,
      "step": 59
    },
    {
      "epoch": 7.46829268292683,
      "grad_norm": 0.3290187120437622,
      "learning_rate": 6.883639468175926e-06,
      "loss": 0.2624,
      "step": 60
    },
    {
      "epoch": 7.585365853658536,
      "grad_norm": 0.25596076250076294,
      "learning_rate": 6.763760432745475e-06,
      "loss": 0.2434,
      "step": 61
    },
    {
      "epoch": 7.702439024390244,
      "grad_norm": 0.3228106200695038,
      "learning_rate": 6.6427119095541745e-06,
      "loss": 0.2461,
      "step": 62
    },
    {
      "epoch": 7.819512195121951,
      "grad_norm": 0.2985459268093109,
      "learning_rate": 6.520574161637591e-06,
      "loss": 0.2454,
      "step": 63
    },
    {
      "epoch": 7.9365853658536585,
      "grad_norm": 0.2825384736061096,
      "learning_rate": 6.397428174258048e-06,
      "loss": 0.2453,
      "step": 64
    },
    {
      "epoch": 8.117073170731707,
      "grad_norm": 0.5776787400245667,
      "learning_rate": 6.273355601206143e-06,
      "loss": 0.4516,
      "step": 65
    },
    {
      "epoch": 8.234146341463415,
      "grad_norm": 0.2711504101753235,
      "learning_rate": 6.148438710658979e-06,
      "loss": 0.224,
      "step": 66
    },
    {
      "epoch": 8.351219512195122,
      "grad_norm": 0.29715049266815186,
      "learning_rate": 6.022760330631006e-06,
      "loss": 0.2096,
      "step": 67
    },
    {
      "epoch": 8.46829268292683,
      "grad_norm": 0.27093052864074707,
      "learning_rate": 5.896403794053679e-06,
      "loss": 0.2175,
      "step": 68
    },
    {
      "epoch": 8.585365853658537,
      "grad_norm": 0.29388418793678284,
      "learning_rate": 5.76945288352031e-06,
      "loss": 0.2162,
      "step": 69
    },
    {
      "epoch": 8.702439024390245,
      "grad_norm": 0.248213529586792,
      "learning_rate": 5.641991775732756e-06,
      "loss": 0.1999,
      "step": 70
    },
    {
      "epoch": 8.819512195121952,
      "grad_norm": 0.29978156089782715,
      "learning_rate": 5.514104985686802e-06,
      "loss": 0.2095,
      "step": 71
    },
    {
      "epoch": 8.93658536585366,
      "grad_norm": 0.2769979238510132,
      "learning_rate": 5.385877310633233e-06,
      "loss": 0.189,
      "step": 72
    },
    {
      "epoch": 9.117073170731707,
      "grad_norm": 0.6451863050460815,
      "learning_rate": 5.257393773851733e-06,
      "loss": 0.4253,
      "step": 73
    },
    {
      "epoch": 9.234146341463415,
      "grad_norm": 0.2627960443496704,
      "learning_rate": 5.1287395682749444e-06,
      "loss": 0.1792,
      "step": 74
    },
    {
      "epoch": 9.351219512195122,
      "grad_norm": 0.3619653284549713,
      "learning_rate": 5e-06,
      "loss": 0.1752,
      "step": 75
    },
    {
      "epoch": 9.46829268292683,
      "grad_norm": 0.3499554991722107,
      "learning_rate": 4.871260431725058e-06,
      "loss": 0.1834,
      "step": 76
    },
    {
      "epoch": 9.585365853658537,
      "grad_norm": 0.2961571216583252,
      "learning_rate": 4.742606226148268e-06,
      "loss": 0.1824,
      "step": 77
    },
    {
      "epoch": 9.702439024390245,
      "grad_norm": 0.29610928893089294,
      "learning_rate": 4.614122689366769e-06,
      "loss": 0.1682,
      "step": 78
    },
    {
      "epoch": 9.819512195121952,
      "grad_norm": 0.29082462191581726,
      "learning_rate": 4.485895014313198e-06,
      "loss": 0.1788,
      "step": 79
    },
    {
      "epoch": 9.93658536585366,
      "grad_norm": 0.3001067638397217,
      "learning_rate": 4.358008224267245e-06,
      "loss": 0.173,
      "step": 80
    },
    {
      "epoch": 10.117073170731707,
      "grad_norm": 0.5978091955184937,
      "learning_rate": 4.230547116479691e-06,
      "loss": 0.3196,
      "step": 81
    },
    {
      "epoch": 10.234146341463415,
      "grad_norm": 0.25669512152671814,
      "learning_rate": 4.103596205946323e-06,
      "loss": 0.159,
      "step": 82
    },
    {
      "epoch": 10.351219512195122,
      "grad_norm": 0.38418713212013245,
      "learning_rate": 3.977239669368998e-06,
      "loss": 0.1509,
      "step": 83
    },
    {
      "epoch": 10.46829268292683,
      "grad_norm": 0.3050716519355774,
      "learning_rate": 3.851561289341023e-06,
      "loss": 0.1467,
      "step": 84
    },
    {
      "epoch": 10.585365853658537,
      "grad_norm": 0.3000875413417816,
      "learning_rate": 3.726644398793857e-06,
      "loss": 0.1411,
      "step": 85
    },
    {
      "epoch": 10.702439024390245,
      "grad_norm": 0.29455339908599854,
      "learning_rate": 3.6025718257419532e-06,
      "loss": 0.1442,
      "step": 86
    },
    {
      "epoch": 10.819512195121952,
      "grad_norm": 0.29129162430763245,
      "learning_rate": 3.4794258383624115e-06,
      "loss": 0.1408,
      "step": 87
    },
    {
      "epoch": 10.93658536585366,
      "grad_norm": 0.28606802225112915,
      "learning_rate": 3.3572880904458267e-06,
      "loss": 0.1523,
      "step": 88
    },
    {
      "epoch": 11.117073170731707,
      "grad_norm": 0.6090349555015564,
      "learning_rate": 3.236239567254526e-06,
      "loss": 0.274,
      "step": 89
    },
    {
      "epoch": 11.234146341463415,
      "grad_norm": 0.38308948278427124,
      "learning_rate": 3.116360531824074e-06,
      "loss": 0.1233,
      "step": 90
    },
    {
      "epoch": 11.351219512195122,
      "grad_norm": 0.2933948338031769,
      "learning_rate": 2.997730471743667e-06,
      "loss": 0.1207,
      "step": 91
    },
    {
      "epoch": 11.46829268292683,
      "grad_norm": 0.37544092535972595,
      "learning_rate": 2.880428046450697e-06,
      "loss": 0.1231,
      "step": 92
    },
    {
      "epoch": 11.585365853658537,
      "grad_norm": 0.26281094551086426,
      "learning_rate": 2.7645310350744296e-06,
      "loss": 0.1253,
      "step": 93
    },
    {
      "epoch": 11.702439024390245,
      "grad_norm": 0.34619009494781494,
      "learning_rate": 2.6501162848634023e-06,
      "loss": 0.1226,
      "step": 94
    },
    {
      "epoch": 11.819512195121952,
      "grad_norm": 0.2849329113960266,
      "learning_rate": 2.537259660230679e-06,
      "loss": 0.1231,
      "step": 95
    },
    {
      "epoch": 11.93658536585366,
      "grad_norm": 0.2921876609325409,
      "learning_rate": 2.426035992450848e-06,
      "loss": 0.1233,
      "step": 96
    },
    {
      "epoch": 12.117073170731707,
      "grad_norm": 0.4465767741203308,
      "learning_rate": 2.316519030041998e-06,
      "loss": 0.2171,
      "step": 97
    },
    {
      "epoch": 12.234146341463415,
      "grad_norm": 0.4664257764816284,
      "learning_rate": 2.2087813898656775e-06,
      "loss": 0.1071,
      "step": 98
    },
    {
      "epoch": 12.351219512195122,
      "grad_norm": 0.27322566509246826,
      "learning_rate": 2.102894508977182e-06,
      "loss": 0.1039,
      "step": 99
    },
    {
      "epoch": 12.46829268292683,
      "grad_norm": 0.42056962847709656,
      "learning_rate": 1.9989285972581595e-06,
      "loss": 0.1033,
      "step": 100
    },
    {
      "epoch": 12.585365853658537,
      "grad_norm": 0.44652435183525085,
      "learning_rate": 1.896952590862886e-06,
      "loss": 0.1052,
      "step": 101
    },
    {
      "epoch": 12.702439024390245,
      "grad_norm": 0.2609389126300812,
      "learning_rate": 1.7970341065091246e-06,
      "loss": 0.1051,
      "step": 102
    },
    {
      "epoch": 12.819512195121952,
      "grad_norm": 0.3649502098560333,
      "learning_rate": 1.699239396643841e-06,
      "loss": 0.1059,
      "step": 103
    },
    {
      "epoch": 12.93658536585366,
      "grad_norm": 0.40414321422576904,
      "learning_rate": 1.6036333055135345e-06,
      "loss": 0.1015,
      "step": 104
    },
    {
      "epoch": 13.117073170731707,
      "grad_norm": 0.6658644080162048,
      "learning_rate": 1.5102792261682813e-06,
      "loss": 0.1962,
      "step": 105
    },
    {
      "epoch": 13.234146341463415,
      "grad_norm": 0.2903091013431549,
      "learning_rate": 1.4192390584280347e-06,
      "loss": 0.0981,
      "step": 106
    },
    {
      "epoch": 13.351219512195122,
      "grad_norm": 0.6606068015098572,
      "learning_rate": 1.330573167839005e-06,
      "loss": 0.1001,
      "step": 107
    },
    {
      "epoch": 13.46829268292683,
      "grad_norm": 0.3750366270542145,
      "learning_rate": 1.2443403456474017e-06,
      "loss": 0.0932,
      "step": 108
    },
    {
      "epoch": 13.585365853658537,
      "grad_norm": 0.21313251554965973,
      "learning_rate": 1.1605977698170001e-06,
      "loss": 0.0858,
      "step": 109
    },
    {
      "epoch": 13.702439024390245,
      "grad_norm": 0.33804988861083984,
      "learning_rate": 1.0794009671164484e-06,
      "loss": 0.0965,
      "step": 110
    },
    {
      "epoch": 13.819512195121952,
      "grad_norm": 0.40509283542633057,
      "learning_rate": 1.0008037763014033e-06,
      "loss": 0.0858,
      "step": 111
    },
    {
      "epoch": 13.93658536585366,
      "grad_norm": 0.29031136631965637,
      "learning_rate": 9.248583124159438e-07,
      "loss": 0.0887,
      "step": 112
    },
    {
      "epoch": 14.117073170731707,
      "grad_norm": 0.4998484253883362,
      "learning_rate": 8.516149322369055e-07,
      "loss": 0.1671,
      "step": 113
    },
    {
      "epoch": 14.234146341463415,
      "grad_norm": 0.20616880059242249,
      "learning_rate": 7.811222008840719e-07,
      "loss": 0.0857,
      "step": 114
    },
    {
      "epoch": 14.351219512195122,
      "grad_norm": 0.22407670319080353,
      "learning_rate": 7.13426859618338e-07,
      "loss": 0.0879,
      "step": 115
    },
    {
      "epoch": 14.46829268292683,
      "grad_norm": 0.2743969261646271,
      "learning_rate": 6.485737948492237e-07,
      "loss": 0.0853,
      "step": 116
    },
    {
      "epoch": 14.585365853658537,
      "grad_norm": 0.2952362596988678,
      "learning_rate": 5.866060083722624e-07,
      "loss": 0.0756,
      "step": 117
    },
    {
      "epoch": 14.702439024390245,
      "grad_norm": 0.2747920751571655,
      "learning_rate": 5.275645888560233e-07,
      "loss": 0.0821,
      "step": 118
    },
    {
      "epoch": 14.819512195121952,
      "grad_norm": 0.21030597388744354,
      "learning_rate": 4.71488684597643e-07,
      "loss": 0.0822,
      "step": 119
    },
    {
      "epoch": 14.93658536585366,
      "grad_norm": 0.19308297336101532,
      "learning_rate": 4.184154775649768e-07,
      "loss": 0.0831,
      "step": 120
    },
    {
      "epoch": 15.117073170731707,
      "grad_norm": 0.4680771231651306,
      "learning_rate": 3.683801587425251e-07,
      "loss": 0.1536,
      "step": 121
    },
    {
      "epoch": 15.234146341463415,
      "grad_norm": 0.23554736375808716,
      "learning_rate": 3.214159047975324e-07,
      "loss": 0.0759,
      "step": 122
    },
    {
      "epoch": 15.351219512195122,
      "grad_norm": 0.2371598780155182,
      "learning_rate": 2.7755385608169374e-07,
      "loss": 0.0799,
      "step": 123
    },
    {
      "epoch": 15.46829268292683,
      "grad_norm": 0.23220933973789215,
      "learning_rate": 2.368230959830875e-07,
      "loss": 0.0906,
      "step": 124
    },
    {
      "epoch": 15.585365853658537,
      "grad_norm": 0.20104503631591797,
      "learning_rate": 1.992506316419912e-07,
      "loss": 0.0743,
      "step": 125
    },
    {
      "epoch": 15.702439024390245,
      "grad_norm": 0.2180931121110916,
      "learning_rate": 1.6486137604339813e-07,
      "loss": 0.0817,
      "step": 126
    },
    {
      "epoch": 15.819512195121952,
      "grad_norm": 0.19550880789756775,
      "learning_rate": 1.3367813149808728e-07,
      "loss": 0.0761,
      "step": 127
    },
    {
      "epoch": 15.93658536585366,
      "grad_norm": 0.19533437490463257,
      "learning_rate": 1.0572157452321097e-07,
      "loss": 0.0782,
      "step": 128
    },
    {
      "epoch": 16.117073170731707,
      "grad_norm": 0.45543140172958374,
      "learning_rate": 8.101024213241826e-08,
      "loss": 0.1505,
      "step": 129
    },
    {
      "epoch": 16.234146341463415,
      "grad_norm": 0.20492489635944366,
      "learning_rate": 5.9560519544614725e-08,
      "loss": 0.0716,
      "step": 130
    },
    {
      "epoch": 16.351219512195122,
      "grad_norm": 0.16599130630493164,
      "learning_rate": 4.138662931949255e-08,
      "loss": 0.0768,
      "step": 131
    },
    {
      "epoch": 16.46829268292683,
      "grad_norm": 0.17992687225341797,
      "learning_rate": 2.6500621927054716e-08,
      "loss": 0.0777,
      "step": 132
    },
    {
      "epoch": 16.585365853658537,
      "grad_norm": 0.16531269252300262,
      "learning_rate": 1.4912367757366485e-08,
      "loss": 0.0802,
      "step": 133
    },
    {
      "epoch": 16.702439024390245,
      "grad_norm": 0.18542739748954773,
      "learning_rate": 6.629550575847355e-09,
      "loss": 0.0754,
      "step": 134
    },
    {
      "epoch": 16.819512195121952,
      "grad_norm": 0.16598491370677948,
      "learning_rate": 1.657662428434792e-09,
      "loss": 0.0773,
      "step": 135
    },
    {
      "epoch": 16.93658536585366,
      "grad_norm": 0.18370650708675385,
      "learning_rate": 0.0,
      "loss": 0.0795,
      "step": 136
    },
    {
      "epoch": 16.93658536585366,
      "step": 136,
      "total_flos": 487121640751104.0,
      "train_loss": 0.29537003558567343,
      "train_runtime": 25552.5852,
      "train_samples_per_second": 0.546,
      "train_steps_per_second": 0.005
    }
  ],
  "logging_steps": 1,
  "max_steps": 136,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 17,
  "save_steps": 10000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 487121640751104.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}