{
  "best_metric": 0.05524912104010582,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.2656924609764198,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.001328462304882099,
      "grad_norm": 1.390708565711975,
      "learning_rate": 1e-05,
      "loss": 1.2451,
      "step": 1
    },
    {
      "epoch": 0.001328462304882099,
      "eval_loss": 0.9099801778793335,
      "eval_runtime": 42.1616,
      "eval_samples_per_second": 30.075,
      "eval_steps_per_second": 7.519,
      "step": 1
    },
    {
      "epoch": 0.002656924609764198,
      "grad_norm": 1.187966227531433,
      "learning_rate": 2e-05,
      "loss": 0.5518,
      "step": 2
    },
    {
      "epoch": 0.003985386914646297,
      "grad_norm": 1.1299062967300415,
      "learning_rate": 3e-05,
      "loss": 0.8168,
      "step": 3
    },
    {
      "epoch": 0.005313849219528396,
      "grad_norm": 1.1839226484298706,
      "learning_rate": 4e-05,
      "loss": 0.996,
      "step": 4
    },
    {
      "epoch": 0.006642311524410495,
      "grad_norm": 1.2297836542129517,
      "learning_rate": 5e-05,
      "loss": 0.6306,
      "step": 5
    },
    {
      "epoch": 0.007970773829292593,
      "grad_norm": 1.2770973443984985,
      "learning_rate": 6e-05,
      "loss": 0.5992,
      "step": 6
    },
    {
      "epoch": 0.009299236134174693,
      "grad_norm": 1.62348473072052,
      "learning_rate": 7e-05,
      "loss": 0.6057,
      "step": 7
    },
    {
      "epoch": 0.010627698439056792,
      "grad_norm": 1.5257402658462524,
      "learning_rate": 8e-05,
      "loss": 0.436,
      "step": 8
    },
    {
      "epoch": 0.01195616074393889,
      "grad_norm": 2.8088972568511963,
      "learning_rate": 9e-05,
      "loss": 0.2978,
      "step": 9
    },
    {
      "epoch": 0.01328462304882099,
      "grad_norm": 1.6363927125930786,
      "learning_rate": 0.0001,
      "loss": 0.3407,
      "step": 10
    },
    {
      "epoch": 0.014613085353703089,
      "grad_norm": 1.5792278051376343,
      "learning_rate": 9.999316524962345e-05,
      "loss": 0.3065,
      "step": 11
    },
    {
      "epoch": 0.015941547658585187,
      "grad_norm": 1.292655110359192,
      "learning_rate": 9.997266286704631e-05,
      "loss": 0.2605,
      "step": 12
    },
    {
      "epoch": 0.017270009963467288,
      "grad_norm": 1.5605067014694214,
      "learning_rate": 9.993849845741524e-05,
      "loss": 0.1858,
      "step": 13
    },
    {
      "epoch": 0.018598472268349386,
      "grad_norm": 1.3306726217269897,
      "learning_rate": 9.989068136093873e-05,
      "loss": 0.1557,
      "step": 14
    },
    {
      "epoch": 0.019926934573231483,
      "grad_norm": 1.0045218467712402,
      "learning_rate": 9.98292246503335e-05,
      "loss": 0.1284,
      "step": 15
    },
    {
      "epoch": 0.021255396878113585,
      "grad_norm": 1.1914656162261963,
      "learning_rate": 9.975414512725057e-05,
      "loss": 0.1766,
      "step": 16
    },
    {
      "epoch": 0.022583859182995682,
      "grad_norm": 0.8283082246780396,
      "learning_rate": 9.966546331768191e-05,
      "loss": 0.1709,
      "step": 17
    },
    {
      "epoch": 0.02391232148787778,
      "grad_norm": 1.2841575145721436,
      "learning_rate": 9.956320346634876e-05,
      "loss": 0.1776,
      "step": 18
    },
    {
      "epoch": 0.02524078379275988,
      "grad_norm": 0.9893993139266968,
      "learning_rate": 9.944739353007344e-05,
      "loss": 0.1167,
      "step": 19
    },
    {
      "epoch": 0.02656924609764198,
      "grad_norm": 1.0702087879180908,
      "learning_rate": 9.931806517013612e-05,
      "loss": 0.1376,
      "step": 20
    },
    {
      "epoch": 0.027897708402524077,
      "grad_norm": 0.7615195512771606,
      "learning_rate": 9.917525374361912e-05,
      "loss": 0.1101,
      "step": 21
    },
    {
      "epoch": 0.029226170707406178,
      "grad_norm": 2.079369306564331,
      "learning_rate": 9.901899829374047e-05,
      "loss": 0.1718,
      "step": 22
    },
    {
      "epoch": 0.030554633012288276,
      "grad_norm": 1.0243886709213257,
      "learning_rate": 9.884934153917997e-05,
      "loss": 0.1273,
      "step": 23
    },
    {
      "epoch": 0.031883095317170373,
      "grad_norm": 1.1541416645050049,
      "learning_rate": 9.86663298624003e-05,
      "loss": 0.1004,
      "step": 24
    },
    {
      "epoch": 0.033211557622052475,
      "grad_norm": 0.69359290599823,
      "learning_rate": 9.847001329696653e-05,
      "loss": 0.0843,
      "step": 25
    },
    {
      "epoch": 0.034540019926934576,
      "grad_norm": 0.9377409815788269,
      "learning_rate": 9.826044551386744e-05,
      "loss": 0.1395,
      "step": 26
    },
    {
      "epoch": 0.03586848223181667,
      "grad_norm": 0.8794754147529602,
      "learning_rate": 9.803768380684242e-05,
      "loss": 0.1262,
      "step": 27
    },
    {
      "epoch": 0.03719694453669877,
      "grad_norm": 0.4305816888809204,
      "learning_rate": 9.780178907671789e-05,
      "loss": 0.0771,
      "step": 28
    },
    {
      "epoch": 0.03852540684158087,
      "grad_norm": 1.1237362623214722,
      "learning_rate": 9.755282581475769e-05,
      "loss": 0.0987,
      "step": 29
    },
    {
      "epoch": 0.03985386914646297,
      "grad_norm": 0.7229164838790894,
      "learning_rate": 9.729086208503174e-05,
      "loss": 0.0779,
      "step": 30
    },
    {
      "epoch": 0.04118233145134507,
      "grad_norm": 0.7091485261917114,
      "learning_rate": 9.701596950580806e-05,
      "loss": 0.084,
      "step": 31
    },
    {
      "epoch": 0.04251079375622717,
      "grad_norm": 0.6118314266204834,
      "learning_rate": 9.672822322997305e-05,
      "loss": 0.0714,
      "step": 32
    },
    {
      "epoch": 0.043839256061109264,
      "grad_norm": 0.5375179648399353,
      "learning_rate": 9.642770192448536e-05,
      "loss": 0.0596,
      "step": 33
    },
    {
      "epoch": 0.045167718365991365,
      "grad_norm": 0.5086399912834167,
      "learning_rate": 9.611448774886924e-05,
      "loss": 0.0522,
      "step": 34
    },
    {
      "epoch": 0.046496180670873466,
      "grad_norm": 1.0448681116104126,
      "learning_rate": 9.578866633275288e-05,
      "loss": 0.1023,
      "step": 35
    },
    {
      "epoch": 0.04782464297575556,
      "grad_norm": 0.978196918964386,
      "learning_rate": 9.545032675245813e-05,
      "loss": 0.0924,
      "step": 36
    },
    {
      "epoch": 0.04915310528063766,
      "grad_norm": 0.9906176924705505,
      "learning_rate": 9.509956150664796e-05,
      "loss": 0.0918,
      "step": 37
    },
    {
      "epoch": 0.05048156758551976,
      "grad_norm": 1.5893501043319702,
      "learning_rate": 9.473646649103818e-05,
      "loss": 0.1491,
      "step": 38
    },
    {
      "epoch": 0.05181002989040186,
      "grad_norm": 1.0811928510665894,
      "learning_rate": 9.43611409721806e-05,
      "loss": 0.1473,
      "step": 39
    },
    {
      "epoch": 0.05313849219528396,
      "grad_norm": 0.9489266276359558,
      "learning_rate": 9.397368756032445e-05,
      "loss": 0.1182,
      "step": 40
    },
    {
      "epoch": 0.05446695450016606,
      "grad_norm": 0.5963587760925293,
      "learning_rate": 9.357421218136386e-05,
      "loss": 0.0576,
      "step": 41
    },
    {
      "epoch": 0.055795416805048154,
      "grad_norm": 0.49168843030929565,
      "learning_rate": 9.316282404787871e-05,
      "loss": 0.0498,
      "step": 42
    },
    {
      "epoch": 0.057123879109930255,
      "grad_norm": 0.6792750954627991,
      "learning_rate": 9.273963562927695e-05,
      "loss": 0.072,
      "step": 43
    },
    {
      "epoch": 0.058452341414812356,
      "grad_norm": 0.5808904767036438,
      "learning_rate": 9.230476262104677e-05,
      "loss": 0.0641,
      "step": 44
    },
    {
      "epoch": 0.05978080371969445,
      "grad_norm": 0.9768387079238892,
      "learning_rate": 9.185832391312644e-05,
      "loss": 0.1122,
      "step": 45
    },
    {
      "epoch": 0.06110926602457655,
      "grad_norm": 1.0769027471542358,
      "learning_rate": 9.140044155740101e-05,
      "loss": 0.1555,
      "step": 46
    },
    {
      "epoch": 0.06243772832945865,
      "grad_norm": 0.6411651372909546,
      "learning_rate": 9.093124073433463e-05,
      "loss": 0.0541,
      "step": 47
    },
    {
      "epoch": 0.06376619063434075,
      "grad_norm": 1.018444538116455,
      "learning_rate": 9.045084971874738e-05,
      "loss": 0.0848,
      "step": 48
    },
    {
      "epoch": 0.06509465293922286,
      "grad_norm": 1.6416921615600586,
      "learning_rate": 8.995939984474624e-05,
      "loss": 0.1044,
      "step": 49
    },
    {
      "epoch": 0.06642311524410495,
      "grad_norm": 1.6693891286849976,
      "learning_rate": 8.945702546981969e-05,
      "loss": 0.1841,
      "step": 50
    },
    {
      "epoch": 0.06642311524410495,
      "eval_loss": 0.11820535361766815,
      "eval_runtime": 42.7529,
      "eval_samples_per_second": 29.659,
      "eval_steps_per_second": 7.415,
      "step": 50
    },
    {
      "epoch": 0.06775157754898704,
      "grad_norm": 1.100433588027954,
      "learning_rate": 8.894386393810563e-05,
      "loss": 0.7106,
      "step": 51
    },
    {
      "epoch": 0.06908003985386915,
      "grad_norm": 0.6428046822547913,
      "learning_rate": 8.842005554284296e-05,
      "loss": 0.1295,
      "step": 52
    },
    {
      "epoch": 0.07040850215875125,
      "grad_norm": 0.7475671768188477,
      "learning_rate": 8.788574348801675e-05,
      "loss": 0.5294,
      "step": 53
    },
    {
      "epoch": 0.07173696446363334,
      "grad_norm": 0.6037783622741699,
      "learning_rate": 8.73410738492077e-05,
      "loss": 0.4785,
      "step": 54
    },
    {
      "epoch": 0.07306542676851545,
      "grad_norm": 0.7302393317222595,
      "learning_rate": 8.678619553365659e-05,
      "loss": 0.1237,
      "step": 55
    },
    {
      "epoch": 0.07439388907339754,
      "grad_norm": 0.6783948540687561,
      "learning_rate": 8.622126023955446e-05,
      "loss": 0.1043,
      "step": 56
    },
    {
      "epoch": 0.07572235137827964,
      "grad_norm": 0.44472363591194153,
      "learning_rate": 8.564642241456986e-05,
      "loss": 0.1446,
      "step": 57
    },
    {
      "epoch": 0.07705081368316175,
      "grad_norm": 0.5291113257408142,
      "learning_rate": 8.506183921362443e-05,
      "loss": 0.156,
      "step": 58
    },
    {
      "epoch": 0.07837927598804384,
      "grad_norm": 1.0741983652114868,
      "learning_rate": 8.44676704559283e-05,
      "loss": 0.1011,
      "step": 59
    },
    {
      "epoch": 0.07970773829292593,
      "grad_norm": 0.4258630573749542,
      "learning_rate": 8.386407858128706e-05,
      "loss": 0.0577,
      "step": 60
    },
    {
      "epoch": 0.08103620059780804,
      "grad_norm": 0.44316762685775757,
      "learning_rate": 8.32512286056924e-05,
      "loss": 0.0536,
      "step": 61
    },
    {
      "epoch": 0.08236466290269014,
      "grad_norm": 0.5236944556236267,
      "learning_rate": 8.262928807620843e-05,
      "loss": 0.1124,
      "step": 62
    },
    {
      "epoch": 0.08369312520757223,
      "grad_norm": 0.482776939868927,
      "learning_rate": 8.199842702516583e-05,
      "loss": 0.1314,
      "step": 63
    },
    {
      "epoch": 0.08502158751245434,
      "grad_norm": 0.4929942488670349,
      "learning_rate": 8.135881792367686e-05,
      "loss": 0.0703,
      "step": 64
    },
    {
      "epoch": 0.08635004981733643,
      "grad_norm": 0.41983386874198914,
      "learning_rate": 8.07106356344834e-05,
      "loss": 0.0488,
      "step": 65
    },
    {
      "epoch": 0.08767851212221853,
      "grad_norm": 0.46827030181884766,
      "learning_rate": 8.005405736415126e-05,
      "loss": 0.0433,
      "step": 66
    },
    {
      "epoch": 0.08900697442710064,
      "grad_norm": 0.27043086290359497,
      "learning_rate": 7.938926261462366e-05,
      "loss": 0.0476,
      "step": 67
    },
    {
      "epoch": 0.09033543673198273,
      "grad_norm": 0.37928521633148193,
      "learning_rate": 7.871643313414718e-05,
      "loss": 0.1005,
      "step": 68
    },
    {
      "epoch": 0.09166389903686482,
      "grad_norm": 0.3404730558395386,
      "learning_rate": 7.803575286758364e-05,
      "loss": 0.0543,
      "step": 69
    },
    {
      "epoch": 0.09299236134174693,
      "grad_norm": 0.4588457942008972,
      "learning_rate": 7.734740790612136e-05,
      "loss": 0.0518,
      "step": 70
    },
    {
      "epoch": 0.09432082364662903,
      "grad_norm": 0.4835526645183563,
      "learning_rate": 7.66515864363997e-05,
      "loss": 0.0541,
      "step": 71
    },
    {
      "epoch": 0.09564928595151112,
      "grad_norm": 0.5752511024475098,
      "learning_rate": 7.594847868906076e-05,
      "loss": 0.0364,
      "step": 72
    },
    {
      "epoch": 0.09697774825639323,
      "grad_norm": 0.5717897415161133,
      "learning_rate": 7.52382768867422e-05,
      "loss": 0.0464,
      "step": 73
    },
    {
      "epoch": 0.09830621056127532,
      "grad_norm": 0.4291505515575409,
      "learning_rate": 7.452117519152542e-05,
      "loss": 0.0459,
      "step": 74
    },
    {
      "epoch": 0.09963467286615742,
      "grad_norm": 0.3689728081226349,
      "learning_rate": 7.379736965185368e-05,
      "loss": 0.0342,
      "step": 75
    },
    {
      "epoch": 0.10096313517103953,
      "grad_norm": 0.3441990315914154,
      "learning_rate": 7.30670581489344e-05,
      "loss": 0.0441,
      "step": 76
    },
    {
      "epoch": 0.10229159747592162,
      "grad_norm": 0.3409755229949951,
      "learning_rate": 7.233044034264034e-05,
      "loss": 0.0537,
      "step": 77
    },
    {
      "epoch": 0.10362005978080371,
      "grad_norm": 0.5653753280639648,
      "learning_rate": 7.158771761692464e-05,
      "loss": 0.0513,
      "step": 78
    },
    {
      "epoch": 0.10494852208568582,
      "grad_norm": 0.4996849596500397,
      "learning_rate": 7.083909302476453e-05,
      "loss": 0.0436,
      "step": 79
    },
    {
      "epoch": 0.10627698439056792,
      "grad_norm": 0.4425507187843323,
      "learning_rate": 7.008477123264848e-05,
      "loss": 0.0356,
      "step": 80
    },
    {
      "epoch": 0.10760544669545001,
      "grad_norm": 0.40278002619743347,
      "learning_rate": 6.932495846462261e-05,
      "loss": 0.0318,
      "step": 81
    },
    {
      "epoch": 0.10893390900033212,
      "grad_norm": 0.35948845744132996,
      "learning_rate": 6.855986244591104e-05,
      "loss": 0.0333,
      "step": 82
    },
    {
      "epoch": 0.11026237130521421,
      "grad_norm": 0.3643316924571991,
      "learning_rate": 6.778969234612584e-05,
      "loss": 0.0325,
      "step": 83
    },
    {
      "epoch": 0.11159083361009631,
      "grad_norm": 0.49635252356529236,
      "learning_rate": 6.701465872208216e-05,
      "loss": 0.0318,
      "step": 84
    },
    {
      "epoch": 0.11291929591497842,
      "grad_norm": 0.43887701630592346,
      "learning_rate": 6.623497346023418e-05,
      "loss": 0.024,
      "step": 85
    },
    {
      "epoch": 0.11424775821986051,
      "grad_norm": 0.45004895329475403,
      "learning_rate": 6.545084971874738e-05,
      "loss": 0.0426,
      "step": 86
    },
    {
      "epoch": 0.1155762205247426,
      "grad_norm": 1.0898586511611938,
      "learning_rate": 6.466250186922325e-05,
      "loss": 0.0678,
      "step": 87
    },
    {
      "epoch": 0.11690468282962471,
      "grad_norm": 0.6138445734977722,
      "learning_rate": 6.387014543809223e-05,
      "loss": 0.0466,
      "step": 88
    },
    {
      "epoch": 0.1182331451345068,
      "grad_norm": 0.7598215341567993,
      "learning_rate": 6.307399704769099e-05,
      "loss": 0.0486,
      "step": 89
    },
    {
      "epoch": 0.1195616074393889,
      "grad_norm": 0.7069115042686462,
      "learning_rate": 6.227427435703997e-05,
      "loss": 0.0429,
      "step": 90
    },
    {
      "epoch": 0.12089006974427101,
      "grad_norm": 0.5428783297538757,
      "learning_rate": 6.147119600233758e-05,
      "loss": 0.0455,
      "step": 91
    },
    {
      "epoch": 0.1222185320491531,
      "grad_norm": 0.2889843285083771,
      "learning_rate": 6.066498153718735e-05,
      "loss": 0.0238,
      "step": 92
    },
    {
      "epoch": 0.12354699435403521,
      "grad_norm": 0.2400159388780594,
      "learning_rate": 5.985585137257401e-05,
      "loss": 0.0128,
      "step": 93
    },
    {
      "epoch": 0.1248754566589173,
      "grad_norm": 0.31455761194229126,
      "learning_rate": 5.90440267166055e-05,
      "loss": 0.0128,
      "step": 94
    },
    {
      "epoch": 0.1262039189637994,
      "grad_norm": 0.6864827871322632,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 0.0467,
      "step": 95
    },
    {
      "epoch": 0.1275323812686815,
      "grad_norm": 0.515399158000946,
      "learning_rate": 5.74131823855921e-05,
      "loss": 0.0621,
      "step": 96
    },
    {
      "epoch": 0.1288608435735636,
      "grad_norm": 0.35708871483802795,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 0.0318,
      "step": 97
    },
    {
      "epoch": 0.1301893058784457,
      "grad_norm": 0.3164196014404297,
      "learning_rate": 5.577423184847932e-05,
      "loss": 0.0222,
      "step": 98
    },
    {
      "epoch": 0.1315177681833278,
      "grad_norm": 1.3855934143066406,
      "learning_rate": 5.495227651252315e-05,
      "loss": 0.0558,
      "step": 99
    },
    {
      "epoch": 0.1328462304882099,
      "grad_norm": 1.2984070777893066,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 0.1026,
      "step": 100
    },
    {
      "epoch": 0.1328462304882099,
      "eval_loss": 0.08204830437898636,
      "eval_runtime": 42.7886,
      "eval_samples_per_second": 29.634,
      "eval_steps_per_second": 7.409,
      "step": 100
    },
    {
      "epoch": 0.134174692793092,
      "grad_norm": 1.0639034509658813,
      "learning_rate": 5.330452921628497e-05,
      "loss": 0.7418,
      "step": 101
    },
    {
      "epoch": 0.1355031550979741,
      "grad_norm": 2.082030773162842,
      "learning_rate": 5.247918773366112e-05,
      "loss": 0.1444,
      "step": 102
    },
    {
      "epoch": 0.13683161740285618,
      "grad_norm": 0.6157129406929016,
      "learning_rate": 5.165316846586541e-05,
      "loss": 0.5029,
      "step": 103
    },
    {
      "epoch": 0.1381600797077383,
      "grad_norm": 0.3814556896686554,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 0.3842,
      "step": 104
    },
    {
      "epoch": 0.1394885420126204,
      "grad_norm": 0.6556084752082825,
      "learning_rate": 5e-05,
      "loss": 0.0657,
      "step": 105
    },
    {
      "epoch": 0.1408170043175025,
      "grad_norm": 0.4281993508338928,
      "learning_rate": 4.917330276168208e-05,
      "loss": 0.0942,
      "step": 106
    },
    {
      "epoch": 0.1421454666223846,
      "grad_norm": 0.5273692011833191,
      "learning_rate": 4.834683153413459e-05,
      "loss": 0.1102,
      "step": 107
    },
    {
      "epoch": 0.14347392892726668,
      "grad_norm": 0.8424544930458069,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 0.0599,
      "step": 108
    },
    {
      "epoch": 0.14480239123214877,
      "grad_norm": 0.2574545741081238,
      "learning_rate": 4.669547078371504e-05,
      "loss": 0.0261,
      "step": 109
    },
    {
      "epoch": 0.1461308535370309,
      "grad_norm": 0.23395954072475433,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 0.031,
      "step": 110
    },
    {
      "epoch": 0.147459315841913,
      "grad_norm": 0.33223798871040344,
      "learning_rate": 4.504772348747687e-05,
      "loss": 0.0446,
      "step": 111
    },
    {
      "epoch": 0.14878777814679509,
      "grad_norm": 0.32616204023361206,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 0.0762,
      "step": 112
    },
    {
      "epoch": 0.15011624045167718,
      "grad_norm": 0.4155626893043518,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 0.0487,
      "step": 113
    },
    {
      "epoch": 0.15144470275655927,
      "grad_norm": 0.2637746334075928,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 0.0237,
      "step": 114
    },
    {
      "epoch": 0.15277316506144137,
      "grad_norm": 0.18656152486801147,
      "learning_rate": 4.17702704859633e-05,
      "loss": 0.0183,
      "step": 115
    },
    {
      "epoch": 0.1541016273663235,
      "grad_norm": 0.15489180386066437,
      "learning_rate": 4.095597328339452e-05,
      "loss": 0.0202,
      "step": 116
    },
    {
      "epoch": 0.15543008967120558,
      "grad_norm": 0.2893701195716858,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 0.0974,
      "step": 117
    },
    {
      "epoch": 0.15675855197608768,
      "grad_norm": 0.19944243133068085,
      "learning_rate": 3.933501846281267e-05,
      "loss": 0.0371,
      "step": 118
    },
    {
      "epoch": 0.15808701428096977,
      "grad_norm": 0.2897372841835022,
      "learning_rate": 3.852880399766243e-05,
      "loss": 0.0299,
      "step": 119
    },
    {
      "epoch": 0.15941547658585187,
      "grad_norm": 0.3380316495895386,
      "learning_rate": 3.772572564296005e-05,
      "loss": 0.0349,
      "step": 120
    },
    {
      "epoch": 0.16074393889073396,
      "grad_norm": 0.2889164686203003,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 0.0219,
      "step": 121
    },
    {
      "epoch": 0.16207240119561608,
      "grad_norm": 0.6124722957611084,
      "learning_rate": 3.612985456190778e-05,
      "loss": 0.0394,
      "step": 122
    },
    {
      "epoch": 0.16340086350049818,
      "grad_norm": 0.2512660026550293,
      "learning_rate": 3.533749813077677e-05,
      "loss": 0.0257,
      "step": 123
    },
    {
      "epoch": 0.16472932580538027,
      "grad_norm": 0.1772724837064743,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 0.0254,
      "step": 124
    },
    {
      "epoch": 0.16605778811026237,
      "grad_norm": 0.1958891600370407,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 0.0322,
      "step": 125
    },
    {
      "epoch": 0.16738625041514446,
      "grad_norm": 0.24720436334609985,
      "learning_rate": 3.298534127791785e-05,
      "loss": 0.0404,
      "step": 126
    },
    {
      "epoch": 0.16871471272002658,
      "grad_norm": 0.21848170459270477,
      "learning_rate": 3.221030765387417e-05,
      "loss": 0.0314,
      "step": 127
    },
    {
      "epoch": 0.17004317502490868,
      "grad_norm": 0.16766457259655,
      "learning_rate": 3.144013755408895e-05,
      "loss": 0.0295,
      "step": 128
    },
    {
      "epoch": 0.17137163732979077,
      "grad_norm": 0.17886987328529358,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 0.0161,
      "step": 129
    },
    {
      "epoch": 0.17270009963467287,
      "grad_norm": 0.1716434359550476,
      "learning_rate": 2.991522876735154e-05,
      "loss": 0.0199,
      "step": 130
    },
    {
      "epoch": 0.17402856193955496,
      "grad_norm": 0.26915857195854187,
      "learning_rate": 2.916090697523549e-05,
      "loss": 0.0199,
      "step": 131
    },
    {
      "epoch": 0.17535702424443705,
      "grad_norm": 0.3413120210170746,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 0.0256,
      "step": 132
    },
    {
      "epoch": 0.17668548654931918,
      "grad_norm": 0.30393096804618835,
      "learning_rate": 2.766955965735968e-05,
      "loss": 0.0229,
      "step": 133
    },
    {
      "epoch": 0.17801394885420127,
      "grad_norm": 0.31084251403808594,
      "learning_rate": 2.693294185106562e-05,
      "loss": 0.0213,
      "step": 134
    },
    {
      "epoch": 0.17934241115908336,
      "grad_norm": 0.2636372447013855,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 0.0264,
      "step": 135
    },
    {
      "epoch": 0.18067087346396546,
      "grad_norm": 0.35219112038612366,
      "learning_rate": 2.547882480847461e-05,
      "loss": 0.0282,
      "step": 136
    },
    {
      "epoch": 0.18199933576884755,
      "grad_norm": 0.35717225074768066,
      "learning_rate": 2.476172311325783e-05,
      "loss": 0.0395,
      "step": 137
    },
    {
      "epoch": 0.18332779807372965,
      "grad_norm": 0.3106258809566498,
      "learning_rate": 2.405152131093926e-05,
      "loss": 0.0262,
      "step": 138
    },
    {
      "epoch": 0.18465626037861177,
      "grad_norm": 0.4700450897216797,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 0.0302,
      "step": 139
    },
    {
      "epoch": 0.18598472268349386,
      "grad_norm": 0.4046179950237274,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 0.0245,
      "step": 140
    },
    {
      "epoch": 0.18731318498837596,
      "grad_norm": 0.501435399055481,
      "learning_rate": 2.196424713241637e-05,
      "loss": 0.0247,
      "step": 141
    },
    {
      "epoch": 0.18864164729325805,
      "grad_norm": 0.11471497267484665,
      "learning_rate": 2.128356686585282e-05,
      "loss": 0.0107,
      "step": 142
    },
    {
      "epoch": 0.18997010959814015,
      "grad_norm": 0.09452979266643524,
      "learning_rate": 2.061073738537635e-05,
      "loss": 0.0069,
      "step": 143
    },
    {
      "epoch": 0.19129857190302224,
      "grad_norm": 0.13501185178756714,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 0.0052,
      "step": 144
    },
    {
      "epoch": 0.19262703420790436,
      "grad_norm": 0.2529805600643158,
      "learning_rate": 1.928936436551661e-05,
      "loss": 0.0119,
      "step": 145
    },
    {
      "epoch": 0.19395549651278646,
      "grad_norm": 0.3246322274208069,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 0.0209,
      "step": 146
    },
    {
      "epoch": 0.19528395881766855,
      "grad_norm": 0.3176211416721344,
      "learning_rate": 1.800157297483417e-05,
      "loss": 0.017,
      "step": 147
    },
    {
      "epoch": 0.19661242112255065,
      "grad_norm": 0.21707482635974884,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 0.0186,
      "step": 148
    },
    {
      "epoch": 0.19794088342743274,
      "grad_norm": 1.2247023582458496,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 0.0241,
      "step": 149
    },
    {
      "epoch": 0.19926934573231483,
      "grad_norm": 0.7454701066017151,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 0.0825,
      "step": 150
    },
    {
      "epoch": 0.19926934573231483,
      "eval_loss": 0.06576347351074219,
      "eval_runtime": 42.7975,
      "eval_samples_per_second": 29.628,
      "eval_steps_per_second": 7.407,
      "step": 150
    },
    {
      "epoch": 0.20059780803719696,
      "grad_norm": 1.1259056329727173,
      "learning_rate": 1.553232954407171e-05,
      "loss": 0.6334,
      "step": 151
    },
    {
      "epoch": 0.20192627034207905,
      "grad_norm": 4.155270576477051,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 0.188,
      "step": 152
    },
    {
      "epoch": 0.20325473264696114,
      "grad_norm": 1.1752597093582153,
      "learning_rate": 1.435357758543015e-05,
      "loss": 0.4346,
      "step": 153
    },
    {
      "epoch": 0.20458319495184324,
      "grad_norm": 0.7669537663459778,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 0.4315,
      "step": 154
    },
    {
      "epoch": 0.20591165725672533,
      "grad_norm": 0.3515520691871643,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 0.0545,
      "step": 155
    },
    {
      "epoch": 0.20724011956160743,
      "grad_norm": 0.30736786127090454,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 0.0524,
      "step": 156
    },
    {
      "epoch": 0.20856858186648955,
      "grad_norm": 0.34879329800605774,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 0.0802,
      "step": 157
    },
    {
      "epoch": 0.20989704417137164,
      "grad_norm": 0.22239454090595245,
      "learning_rate": 1.157994445715706e-05,
      "loss": 0.0553,
      "step": 158
    },
    {
      "epoch": 0.21122550647625374,
      "grad_norm": 0.26180335879325867,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 0.0228,
      "step": 159
    },
    {
      "epoch": 0.21255396878113583,
      "grad_norm": 0.2421458512544632,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 0.0267,
      "step": 160
    },
    {
      "epoch": 0.21388243108601793,
      "grad_norm": 0.11997482925653458,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 0.0335,
      "step": 161
    },
    {
      "epoch": 0.21521089339090002,
      "grad_norm": 0.24232330918312073,
      "learning_rate": 9.549150281252633e-06,
      "loss": 0.0788,
      "step": 162
    },
    {
      "epoch": 0.21653935569578214,
      "grad_norm": 0.13641518354415894,
      "learning_rate": 9.068759265665384e-06,
      "loss": 0.0284,
      "step": 163
    },
    {
      "epoch": 0.21786781800066424,
      "grad_norm": 0.20858359336853027,
      "learning_rate": 8.599558442598998e-06,
      "loss": 0.0164,
      "step": 164
    },
    {
      "epoch": 0.21919628030554633,
      "grad_norm": 0.2087399959564209,
      "learning_rate": 8.141676086873572e-06,
      "loss": 0.0175,
      "step": 165
    },
    {
      "epoch": 0.22052474261042843,
      "grad_norm": 0.14577652513980865,
      "learning_rate": 7.695237378953223e-06,
      "loss": 0.0216,
      "step": 166
    },
    {
      "epoch": 0.22185320491531052,
      "grad_norm": 0.1438261717557907,
      "learning_rate": 7.260364370723044e-06,
      "loss": 0.0515,
      "step": 167
    },
    {
      "epoch": 0.22318166722019261,
      "grad_norm": 0.28253060579299927,
      "learning_rate": 6.837175952121306e-06,
      "loss": 0.023,
      "step": 168
    },
    {
      "epoch": 0.22451012952507474,
      "grad_norm": 0.31980425119400024,
      "learning_rate": 6.425787818636131e-06,
      "loss": 0.0173,
      "step": 169
    },
    {
      "epoch": 0.22583859182995683,
      "grad_norm": 0.2921467125415802,
      "learning_rate": 6.026312439675552e-06,
      "loss": 0.0141,
      "step": 170
    },
    {
      "epoch": 0.22716705413483892,
      "grad_norm": 0.3821288049221039,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 0.01,
      "step": 171
    },
    {
      "epoch": 0.22849551643972102,
      "grad_norm": 0.32927316427230835,
      "learning_rate": 5.263533508961827e-06,
      "loss": 0.0156,
      "step": 172
    },
    {
      "epoch": 0.2298239787446031,
      "grad_norm": 0.21407172083854675,
      "learning_rate": 4.900438493352055e-06,
      "loss": 0.0153,
      "step": 173
    },
    {
      "epoch": 0.2311524410494852,
      "grad_norm": 0.17321035265922546,
      "learning_rate": 4.549673247541875e-06,
      "loss": 0.0154,
      "step": 174
    },
    {
      "epoch": 0.23248090335436733,
      "grad_norm": 0.18879427015781403,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 0.0211,
      "step": 175
    },
    {
      "epoch": 0.23380936565924942,
      "grad_norm": 0.24369116127490997,
      "learning_rate": 3.885512251130763e-06,
      "loss": 0.0314,
      "step": 176
    },
    {
      "epoch": 0.23513782796413152,
      "grad_norm": 0.22556409239768982,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 0.0251,
      "step": 177
    },
    {
      "epoch": 0.2364662902690136,
      "grad_norm": 0.12515389919281006,
      "learning_rate": 3.271776770026963e-06,
      "loss": 0.0175,
      "step": 178
    },
    {
      "epoch": 0.2377947525738957,
      "grad_norm": 0.20194387435913086,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 0.0125,
      "step": 179
    },
    {
      "epoch": 0.2391232148787778,
      "grad_norm": 0.28396233916282654,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 0.0143,
      "step": 180
    },
    {
      "epoch": 0.24045167718365992,
      "grad_norm": 0.25012674927711487,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 0.0215,
      "step": 181
    },
    {
      "epoch": 0.24178013948854202,
      "grad_norm": 0.2569197118282318,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 0.0194,
      "step": 182
    },
    {
      "epoch": 0.2431086017934241,
      "grad_norm": 0.28262314200401306,
      "learning_rate": 1.962316193157593e-06,
      "loss": 0.0186,
      "step": 183
    },
    {
      "epoch": 0.2444370640983062,
      "grad_norm": 0.2157091349363327,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 0.0099,
      "step": 184
    },
    {
      "epoch": 0.2457655264031883,
      "grad_norm": 0.1421394646167755,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 0.0126,
      "step": 185
    },
    {
      "epoch": 0.24709398870807042,
      "grad_norm": 0.19622422754764557,
      "learning_rate": 1.333670137599713e-06,
      "loss": 0.014,
      "step": 186
    },
    {
      "epoch": 0.24842245101295252,
      "grad_norm": 0.180411234498024,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 0.0241,
      "step": 187
    },
    {
      "epoch": 0.2497509133178346,
      "grad_norm": 0.13559560477733612,
      "learning_rate": 9.810017062595322e-07,
      "loss": 0.0187,
      "step": 188
    },
    {
      "epoch": 0.2510793756227167,
      "grad_norm": 0.12288277596235275,
      "learning_rate": 8.247462563808817e-07,
      "loss": 0.011,
      "step": 189
    },
    {
      "epoch": 0.2524078379275988,
      "grad_norm": 0.13741442561149597,
      "learning_rate": 6.819348298638839e-07,
      "loss": 0.0156,
      "step": 190
    },
    {
      "epoch": 0.2537363002324809,
      "grad_norm": 0.0944119468331337,
      "learning_rate": 5.526064699265753e-07,
      "loss": 0.0131,
      "step": 191
    },
    {
      "epoch": 0.255064762537363,
      "grad_norm": 0.1023307666182518,
      "learning_rate": 4.367965336512403e-07,
      "loss": 0.0174,
      "step": 192
    },
    {
      "epoch": 0.2563932248422451,
      "grad_norm": 0.0828242376446724,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 0.0086,
      "step": 193
    },
    {
      "epoch": 0.2577216871471272,
      "grad_norm": 0.0731416568160057,
      "learning_rate": 2.458548727494292e-07,
      "loss": 0.0043,
      "step": 194
    },
    {
      "epoch": 0.2590501494520093,
      "grad_norm": 0.11668583005666733,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 0.0099,
      "step": 195
    },
    {
      "epoch": 0.2603786117568914,
      "grad_norm": 0.17430779337882996,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 0.0116,
      "step": 196
    },
    {
      "epoch": 0.2617070740617735,
      "grad_norm": 0.10339223593473434,
      "learning_rate": 6.150154258476315e-08,
      "loss": 0.008,
      "step": 197
    },
    {
      "epoch": 0.2630355363666556,
      "grad_norm": 0.11887497454881668,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 0.016,
      "step": 198
    },
    {
      "epoch": 0.2643639986715377,
      "grad_norm": 0.15798938274383545,
      "learning_rate": 6.834750376549792e-09,
      "loss": 0.0127,
      "step": 199
    },
    {
      "epoch": 0.2656924609764198,
      "grad_norm": 0.32944944500923157,
      "learning_rate": 0.0,
      "loss": 0.04,
      "step": 200
    },
    {
      "epoch": 0.2656924609764198,
      "eval_loss": 0.05524912104010582,
      "eval_runtime": 42.7753,
      "eval_samples_per_second": 29.643,
      "eval_steps_per_second": 7.411,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.226038350249984e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}