{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9690721649484537,
  "eval_steps": 500,
  "global_step": 180,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.016494845360824743,
      "grad_norm": 6.510887010044826,
      "learning_rate": 4.444444444444444e-06,
      "loss": 1.1235,
      "step": 1
    },
    {
      "epoch": 0.032989690721649485,
      "grad_norm": 6.5268184982716475,
      "learning_rate": 8.888888888888888e-06,
      "loss": 1.1258,
      "step": 2
    },
    {
      "epoch": 0.049484536082474224,
      "grad_norm": 4.24927908625525,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 1.0563,
      "step": 3
    },
    {
      "epoch": 0.06597938144329897,
      "grad_norm": 2.816200291689023,
      "learning_rate": 1.7777777777777777e-05,
      "loss": 1.009,
      "step": 4
    },
    {
      "epoch": 0.08247422680412371,
      "grad_norm": 6.493905326694169,
      "learning_rate": 2.2222222222222227e-05,
      "loss": 1.013,
      "step": 5
    },
    {
      "epoch": 0.09896907216494845,
      "grad_norm": 7.726008627313003,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 1.0706,
      "step": 6
    },
    {
      "epoch": 0.1154639175257732,
      "grad_norm": 4.516493110783658,
      "learning_rate": 3.111111111111112e-05,
      "loss": 0.9695,
      "step": 7
    },
    {
      "epoch": 0.13195876288659794,
      "grad_norm": 4.640178402684917,
      "learning_rate": 3.555555555555555e-05,
      "loss": 0.9444,
      "step": 8
    },
    {
      "epoch": 0.14845360824742268,
      "grad_norm": 2.268866723993877,
      "learning_rate": 4e-05,
      "loss": 0.905,
      "step": 9
    },
    {
      "epoch": 0.16494845360824742,
      "grad_norm": 3.54868006526656,
      "learning_rate": 4.444444444444445e-05,
      "loss": 0.8855,
      "step": 10
    },
    {
      "epoch": 0.18144329896907216,
      "grad_norm": 2.702694507430541,
      "learning_rate": 4.88888888888889e-05,
      "loss": 0.8814,
      "step": 11
    },
    {
      "epoch": 0.1979381443298969,
      "grad_norm": 2.4163522339799424,
      "learning_rate": 5.333333333333333e-05,
      "loss": 0.8649,
      "step": 12
    },
    {
      "epoch": 0.21443298969072164,
      "grad_norm": 1.7906871535399842,
      "learning_rate": 5.777777777777778e-05,
      "loss": 0.8351,
      "step": 13
    },
    {
      "epoch": 0.2309278350515464,
      "grad_norm": 2.073156178645725,
      "learning_rate": 6.222222222222223e-05,
      "loss": 0.8352,
      "step": 14
    },
    {
      "epoch": 0.24742268041237114,
      "grad_norm": 4.016806278073859,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.8279,
      "step": 15
    },
    {
      "epoch": 0.2639175257731959,
      "grad_norm": 1.5493063583627422,
      "learning_rate": 7.11111111111111e-05,
      "loss": 0.8198,
      "step": 16
    },
    {
      "epoch": 0.2804123711340206,
      "grad_norm": 3.1170782607697487,
      "learning_rate": 7.555555555555556e-05,
      "loss": 0.8298,
      "step": 17
    },
    {
      "epoch": 0.29690721649484536,
      "grad_norm": 1.754906513968234,
      "learning_rate": 8e-05,
      "loss": 0.813,
      "step": 18
    },
    {
      "epoch": 0.3134020618556701,
      "grad_norm": 2.9900814711935846,
      "learning_rate": 7.999247881794007e-05,
      "loss": 0.8145,
      "step": 19
    },
    {
      "epoch": 0.32989690721649484,
      "grad_norm": 2.4128887550257256,
      "learning_rate": 7.996991810016922e-05,
      "loss": 0.8023,
      "step": 20
    },
    {
      "epoch": 0.3463917525773196,
      "grad_norm": 1.573430056889768,
      "learning_rate": 7.993232633085074e-05,
      "loss": 0.7798,
      "step": 21
    },
    {
      "epoch": 0.3628865979381443,
      "grad_norm": 1.7272127576229301,
      "learning_rate": 7.987971764671168e-05,
      "loss": 0.7822,
      "step": 22
    },
    {
      "epoch": 0.37938144329896906,
      "grad_norm": 2.4051614980913256,
      "learning_rate": 7.981211183172663e-05,
      "loss": 0.7858,
      "step": 23
    },
    {
      "epoch": 0.3958762886597938,
      "grad_norm": 1.2504967615335536,
      "learning_rate": 7.972953430967773e-05,
      "loss": 0.773,
      "step": 24
    },
    {
      "epoch": 0.41237113402061853,
      "grad_norm": 1.6023332952213127,
      "learning_rate": 7.963201613459381e-05,
      "loss": 0.7793,
      "step": 25
    },
    {
      "epoch": 0.4288659793814433,
      "grad_norm": 1.6124516785601999,
      "learning_rate": 7.951959397907237e-05,
      "loss": 0.7753,
      "step": 26
    },
    {
      "epoch": 0.44536082474226807,
      "grad_norm": 1.0172041366242313,
      "learning_rate": 7.939231012048833e-05,
      "loss": 0.7498,
      "step": 27
    },
    {
      "epoch": 0.4618556701030928,
      "grad_norm": 2.063022239192394,
      "learning_rate": 7.925021242509539e-05,
      "loss": 0.7668,
      "step": 28
    },
    {
      "epoch": 0.47835051546391755,
      "grad_norm": 1.203906865110656,
      "learning_rate": 7.909335433002543e-05,
      "loss": 0.7522,
      "step": 29
    },
    {
      "epoch": 0.4948453608247423,
      "grad_norm": 1.8963235208653197,
      "learning_rate": 7.892179482319297e-05,
      "loss": 0.7562,
      "step": 30
    },
    {
      "epoch": 0.511340206185567,
      "grad_norm": 1.4857136251084346,
      "learning_rate": 7.873559842111225e-05,
      "loss": 0.758,
      "step": 31
    },
    {
      "epoch": 0.5278350515463918,
      "grad_norm": 1.324299893798851,
      "learning_rate": 7.853483514463521e-05,
      "loss": 0.7392,
      "step": 32
    },
    {
      "epoch": 0.5443298969072164,
      "grad_norm": 1.1480502056057977,
      "learning_rate": 7.831958049261956e-05,
      "loss": 0.7363,
      "step": 33
    },
    {
      "epoch": 0.5608247422680412,
      "grad_norm": 1.0730578557805914,
      "learning_rate": 7.808991541353662e-05,
      "loss": 0.7244,
      "step": 34
    },
    {
      "epoch": 0.5773195876288659,
      "grad_norm": 1.2205423026753601,
      "learning_rate": 7.784592627503004e-05,
      "loss": 0.7307,
      "step": 35
    },
    {
      "epoch": 0.5938144329896907,
      "grad_norm": 1.1023928563893555,
      "learning_rate": 7.758770483143634e-05,
      "loss": 0.7178,
      "step": 36
    },
    {
      "epoch": 0.6103092783505155,
      "grad_norm": 1.2767578646872828,
      "learning_rate": 7.731534818928004e-05,
      "loss": 0.7173,
      "step": 37
    },
    {
      "epoch": 0.6268041237113402,
      "grad_norm": 1.682499530220273,
      "learning_rate": 7.702895877075563e-05,
      "loss": 0.733,
      "step": 38
    },
    {
      "epoch": 0.643298969072165,
      "grad_norm": 1.2445811359694992,
      "learning_rate": 7.672864427521097e-05,
      "loss": 0.708,
      "step": 39
    },
    {
      "epoch": 0.6597938144329897,
      "grad_norm": 0.923390898153776,
      "learning_rate": 7.641451763864587e-05,
      "loss": 0.7117,
      "step": 40
    },
    {
      "epoch": 0.6762886597938145,
      "grad_norm": 1.6810293301142893,
      "learning_rate": 7.608669699124153e-05,
      "loss": 0.7164,
      "step": 41
    },
    {
      "epoch": 0.6927835051546392,
      "grad_norm": 0.8665817723805724,
      "learning_rate": 7.57453056129365e-05,
      "loss": 0.7015,
      "step": 42
    },
    {
      "epoch": 0.709278350515464,
      "grad_norm": 1.38899095974282,
      "learning_rate": 7.539047188706631e-05,
      "loss": 0.7138,
      "step": 43
    },
    {
      "epoch": 0.7257731958762886,
      "grad_norm": 0.8370635723618396,
      "learning_rate": 7.502232925208365e-05,
      "loss": 0.6963,
      "step": 44
    },
    {
      "epoch": 0.7422680412371134,
      "grad_norm": 1.2109419876620096,
      "learning_rate": 7.464101615137756e-05,
      "loss": 0.6998,
      "step": 45
    },
    {
      "epoch": 0.7587628865979381,
      "grad_norm": 0.7345771754907713,
      "learning_rate": 7.424667598121067e-05,
      "loss": 0.6895,
      "step": 46
    },
    {
      "epoch": 0.7752577319587629,
      "grad_norm": 0.866583046558764,
      "learning_rate": 7.383945703679365e-05,
      "loss": 0.6853,
      "step": 47
    },
    {
      "epoch": 0.7917525773195876,
      "grad_norm": 0.954971416511012,
      "learning_rate": 7.341951245651747e-05,
      "loss": 0.6924,
      "step": 48
    },
    {
      "epoch": 0.8082474226804124,
      "grad_norm": 0.9698764503436211,
      "learning_rate": 7.298700016436427e-05,
      "loss": 0.692,
      "step": 49
    },
    {
      "epoch": 0.8247422680412371,
      "grad_norm": 1.144548805435363,
      "learning_rate": 7.254208281051871e-05,
      "loss": 0.6928,
      "step": 50
    },
    {
      "epoch": 0.8412371134020619,
      "grad_norm": 0.8399338717507018,
      "learning_rate": 7.208492771020176e-05,
      "loss": 0.685,
      "step": 51
    },
    {
      "epoch": 0.8577319587628865,
      "grad_norm": 0.8519348258795473,
      "learning_rate": 7.161570678075038e-05,
      "loss": 0.679,
      "step": 52
    },
    {
      "epoch": 0.8742268041237113,
      "grad_norm": 0.8682024621779066,
      "learning_rate": 7.113459647696641e-05,
      "loss": 0.6855,
      "step": 53
    },
    {
      "epoch": 0.8907216494845361,
      "grad_norm": 0.9687444841568461,
      "learning_rate": 7.064177772475912e-05,
      "loss": 0.679,
      "step": 54
    },
    {
      "epoch": 0.9072164948453608,
      "grad_norm": 1.3660815254644363,
      "learning_rate": 7.013743585310642e-05,
      "loss": 0.675,
      "step": 55
    },
    {
      "epoch": 0.9237113402061856,
      "grad_norm": 0.546131036326861,
      "learning_rate": 6.96217605243602e-05,
      "loss": 0.6711,
      "step": 56
    },
    {
      "epoch": 0.9402061855670103,
      "grad_norm": 1.1202176159006922,
      "learning_rate": 6.909494566292195e-05,
      "loss": 0.6688,
      "step": 57
    },
    {
      "epoch": 0.9567010309278351,
      "grad_norm": 1.2767701093146733,
      "learning_rate": 6.855718938231597e-05,
      "loss": 0.6707,
      "step": 58
    },
    {
      "epoch": 0.9731958762886598,
      "grad_norm": 0.5478022664863074,
      "learning_rate": 6.800869391068674e-05,
      "loss": 0.661,
      "step": 59
    },
    {
      "epoch": 0.9896907216494846,
      "grad_norm": 1.0121489489469924,
      "learning_rate": 6.744966551474936e-05,
      "loss": 0.6668,
      "step": 60
    },
    {
      "epoch": 1.0061855670103093,
      "grad_norm": 1.4780025517613948,
      "learning_rate": 6.688031442222091e-05,
      "loss": 0.9023,
      "step": 61
    },
    {
      "epoch": 1.022680412371134,
      "grad_norm": 0.8312783075179766,
      "learning_rate": 6.630085474276256e-05,
      "loss": 0.6324,
      "step": 62
    },
    {
      "epoch": 1.0391752577319588,
      "grad_norm": 0.8089531767122732,
      "learning_rate": 6.571150438746157e-05,
      "loss": 0.6264,
      "step": 63
    },
    {
      "epoch": 1.0556701030927835,
      "grad_norm": 0.5525157311375302,
      "learning_rate": 6.511248498688396e-05,
      "loss": 0.6245,
      "step": 64
    },
    {
      "epoch": 1.0721649484536082,
      "grad_norm": 0.6641634808727606,
      "learning_rate": 6.450402180772811e-05,
      "loss": 0.6239,
      "step": 65
    },
    {
      "epoch": 1.088659793814433,
      "grad_norm": 0.6945453989138148,
      "learning_rate": 6.388634366811146e-05,
      "loss": 0.6218,
      "step": 66
    },
    {
      "epoch": 1.1051546391752578,
      "grad_norm": 0.467762743943545,
      "learning_rate": 6.325968285152107e-05,
      "loss": 0.6191,
      "step": 67
    },
    {
      "epoch": 1.1216494845360825,
      "grad_norm": 0.5960377140822906,
      "learning_rate": 6.262427501946155e-05,
      "loss": 0.6196,
      "step": 68
    },
    {
      "epoch": 1.1381443298969072,
      "grad_norm": 0.6012136643477979,
      "learning_rate": 6.198035912283225e-05,
      "loss": 0.6156,
      "step": 69
    },
    {
      "epoch": 1.1546391752577319,
      "grad_norm": 0.3830555698329603,
      "learning_rate": 6.132817731206766e-05,
      "loss": 0.6164,
      "step": 70
    },
    {
      "epoch": 1.1711340206185568,
      "grad_norm": 0.5255708619248807,
      "learning_rate": 6.0667974846074524e-05,
      "loss": 0.6116,
      "step": 71
    },
    {
      "epoch": 1.1876288659793814,
      "grad_norm": 0.34124652573381775,
      "learning_rate": 6.000000000000001e-05,
      "loss": 0.6112,
      "step": 72
    },
    {
      "epoch": 1.2041237113402061,
      "grad_norm": 0.36719307010241353,
      "learning_rate": 5.9324503971865545e-05,
      "loss": 0.6087,
      "step": 73
    },
    {
      "epoch": 1.220618556701031,
      "grad_norm": 0.3744904944376666,
      "learning_rate": 5.8641740788101566e-05,
      "loss": 0.6178,
      "step": 74
    },
    {
      "epoch": 1.2371134020618557,
      "grad_norm": 0.40861995730924344,
      "learning_rate": 5.79519672080185e-05,
      "loss": 0.6078,
      "step": 75
    },
    {
      "epoch": 1.2536082474226804,
      "grad_norm": 0.4191817335171689,
      "learning_rate": 5.7255442627250146e-05,
      "loss": 0.6093,
      "step": 76
    },
    {
      "epoch": 1.270103092783505,
      "grad_norm": 0.4279620084135946,
      "learning_rate": 5.6552428980205575e-05,
      "loss": 0.6036,
      "step": 77
    },
    {
      "epoch": 1.2865979381443298,
      "grad_norm": 0.43387459801090117,
      "learning_rate": 5.584319064156628e-05,
      "loss": 0.6122,
      "step": 78
    },
    {
      "epoch": 1.3030927835051547,
      "grad_norm": 0.36347036101765934,
      "learning_rate": 5.5127994326865706e-05,
      "loss": 0.6095,
      "step": 79
    },
    {
      "epoch": 1.3195876288659794,
      "grad_norm": 0.4126765641672846,
      "learning_rate": 5.440710899218842e-05,
      "loss": 0.6066,
      "step": 80
    },
    {
      "epoch": 1.3360824742268043,
      "grad_norm": 0.4342669224320681,
      "learning_rate": 5.368080573302676e-05,
      "loss": 0.6049,
      "step": 81
    },
    {
      "epoch": 1.352577319587629,
      "grad_norm": 0.5247434320363652,
      "learning_rate": 5.294935768233285e-05,
      "loss": 0.6096,
      "step": 82
    },
    {
      "epoch": 1.3690721649484536,
      "grad_norm": 0.5346000474514714,
      "learning_rate": 5.2213039907804535e-05,
      "loss": 0.6103,
      "step": 83
    },
    {
      "epoch": 1.3855670103092783,
      "grad_norm": 0.4704610145658556,
      "learning_rate": 5.1472129308443616e-05,
      "loss": 0.6031,
      "step": 84
    },
    {
      "epoch": 1.402061855670103,
      "grad_norm": 0.39776693880011527,
      "learning_rate": 5.07269045104255e-05,
      "loss": 0.6037,
      "step": 85
    },
    {
      "epoch": 1.418556701030928,
      "grad_norm": 0.4551461417385109,
      "learning_rate": 4.9977645762319255e-05,
      "loss": 0.5964,
      "step": 86
    },
    {
      "epoch": 1.4350515463917526,
      "grad_norm": 0.49160111934260997,
      "learning_rate": 4.922463482969761e-05,
      "loss": 0.6041,
      "step": 87
    },
    {
      "epoch": 1.4515463917525773,
      "grad_norm": 0.4569400444284779,
      "learning_rate": 4.846815488917644e-05,
      "loss": 0.6038,
      "step": 88
    },
    {
      "epoch": 1.4680412371134022,
      "grad_norm": 0.37114067381878696,
      "learning_rate": 4.7708490421923596e-05,
      "loss": 0.6002,
      "step": 89
    },
    {
      "epoch": 1.4845360824742269,
      "grad_norm": 0.28286469224476896,
      "learning_rate": 4.694592710667723e-05,
      "loss": 0.5999,
      "step": 90
    },
    {
      "epoch": 1.5010309278350515,
      "grad_norm": 0.2544980684276211,
      "learning_rate": 4.618075171231363e-05,
      "loss": 0.5975,
      "step": 91
    },
    {
      "epoch": 1.5175257731958762,
      "grad_norm": 0.3294943971867536,
      "learning_rate": 4.541325199000525e-05,
      "loss": 0.6029,
      "step": 92
    },
    {
      "epoch": 1.534020618556701,
      "grad_norm": 0.41072515728050296,
      "learning_rate": 4.464371656500921e-05,
      "loss": 0.5921,
      "step": 93
    },
    {
      "epoch": 1.5505154639175258,
      "grad_norm": 0.35351679633113386,
      "learning_rate": 4.387243482812717e-05,
      "loss": 0.5949,
      "step": 94
    },
    {
      "epoch": 1.5670103092783505,
      "grad_norm": 0.26711294822182974,
      "learning_rate": 4.309969682687724e-05,
      "loss": 0.5994,
      "step": 95
    },
    {
      "epoch": 1.5835051546391754,
      "grad_norm": 0.32816659581517543,
      "learning_rate": 4.2325793156419035e-05,
      "loss": 0.5932,
      "step": 96
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.3504348660738673,
      "learning_rate": 4.155101485027268e-05,
      "loss": 0.5987,
      "step": 97
    },
    {
      "epoch": 1.6164948453608248,
      "grad_norm": 0.31517917046437344,
      "learning_rate": 4.077565327087298e-05,
      "loss": 0.5971,
      "step": 98
    },
    {
      "epoch": 1.6329896907216495,
      "grad_norm": 0.2535047071826055,
      "learning_rate": 4e-05,
      "loss": 0.5948,
      "step": 99
    },
    {
      "epoch": 1.6494845360824741,
      "grad_norm": 0.2375012914581539,
      "learning_rate": 3.9224346729127034e-05,
      "loss": 0.5949,
      "step": 100
    },
    {
      "epoch": 1.6659793814432988,
      "grad_norm": 0.2623717476380207,
      "learning_rate": 3.844898514972733e-05,
      "loss": 0.5947,
      "step": 101
    },
    {
      "epoch": 1.6824742268041237,
      "grad_norm": 0.2512637608509031,
      "learning_rate": 3.767420684358097e-05,
      "loss": 0.5988,
      "step": 102
    },
    {
      "epoch": 1.6989690721649484,
      "grad_norm": 0.18029355738972075,
      "learning_rate": 3.690030317312277e-05,
      "loss": 0.5927,
      "step": 103
    },
    {
      "epoch": 1.7154639175257733,
      "grad_norm": 0.20901651442289526,
      "learning_rate": 3.612756517187284e-05,
      "loss": 0.5963,
      "step": 104
    },
    {
      "epoch": 1.731958762886598,
      "grad_norm": 0.23100136098283308,
      "learning_rate": 3.535628343499079e-05,
      "loss": 0.5957,
      "step": 105
    },
    {
      "epoch": 1.7484536082474227,
      "grad_norm": 0.18935083808209807,
      "learning_rate": 3.458674800999477e-05,
      "loss": 0.5926,
      "step": 106
    },
    {
      "epoch": 1.7649484536082474,
      "grad_norm": 0.16393645527535064,
      "learning_rate": 3.3819248287686386e-05,
      "loss": 0.5949,
      "step": 107
    },
    {
      "epoch": 1.781443298969072,
      "grad_norm": 0.190745251922808,
      "learning_rate": 3.305407289332279e-05,
      "loss": 0.5891,
      "step": 108
    },
    {
      "epoch": 1.797938144329897,
      "grad_norm": 0.16188706039772582,
      "learning_rate": 3.229150957807641e-05,
      "loss": 0.5892,
      "step": 109
    },
    {
      "epoch": 1.8144329896907216,
      "grad_norm": 0.18778980243793414,
      "learning_rate": 3.153184511082359e-05,
      "loss": 0.592,
      "step": 110
    },
    {
      "epoch": 1.8309278350515465,
      "grad_norm": 0.16922028732307035,
      "learning_rate": 3.07753651703024e-05,
      "loss": 0.5926,
      "step": 111
    },
    {
      "epoch": 1.8474226804123712,
      "grad_norm": 0.2075008066134838,
      "learning_rate": 3.0022354237680752e-05,
      "loss": 0.5915,
      "step": 112
    },
    {
      "epoch": 1.863917525773196,
      "grad_norm": 0.15945271606226596,
      "learning_rate": 2.9273095489574502e-05,
      "loss": 0.5917,
      "step": 113
    },
    {
      "epoch": 1.8804123711340206,
      "grad_norm": 0.20924714994332358,
      "learning_rate": 2.8527870691556404e-05,
      "loss": 0.589,
      "step": 114
    },
    {
      "epoch": 1.8969072164948453,
      "grad_norm": 0.15413768093132926,
      "learning_rate": 2.778696009219548e-05,
      "loss": 0.5895,
      "step": 115
    },
    {
      "epoch": 1.91340206185567,
      "grad_norm": 0.17340241637334963,
      "learning_rate": 2.7050642317667164e-05,
      "loss": 0.5908,
      "step": 116
    },
    {
      "epoch": 1.9298969072164949,
      "grad_norm": 0.13700935671950965,
      "learning_rate": 2.6319194266973256e-05,
      "loss": 0.5881,
      "step": 117
    },
    {
      "epoch": 1.9463917525773196,
      "grad_norm": 0.16121450857357886,
      "learning_rate": 2.5592891007811594e-05,
      "loss": 0.5856,
      "step": 118
    },
    {
      "epoch": 1.9628865979381445,
      "grad_norm": 0.13760640745884387,
      "learning_rate": 2.4872005673134307e-05,
      "loss": 0.5888,
      "step": 119
    },
    {
      "epoch": 1.9793814432989691,
      "grad_norm": 0.136153453631026,
      "learning_rate": 2.4156809358433728e-05,
      "loss": 0.5866,
      "step": 120
    },
    {
      "epoch": 1.9958762886597938,
      "grad_norm": 0.1740518146800063,
      "learning_rate": 2.3447571019794438e-05,
      "loss": 0.7221,
      "step": 121
    },
    {
      "epoch": 2.0123711340206185,
      "grad_norm": 0.1913924294175378,
      "learning_rate": 2.274455737274987e-05,
      "loss": 0.649,
      "step": 122
    },
    {
      "epoch": 2.028865979381443,
      "grad_norm": 0.16283149279005063,
      "learning_rate": 2.2048032791981515e-05,
      "loss": 0.5525,
      "step": 123
    },
    {
      "epoch": 2.045360824742268,
      "grad_norm": 0.14595606592277305,
      "learning_rate": 2.135825921189846e-05,
      "loss": 0.5518,
      "step": 124
    },
    {
      "epoch": 2.0618556701030926,
      "grad_norm": 0.17091723837273506,
      "learning_rate": 2.067549602813446e-05,
      "loss": 0.5537,
      "step": 125
    },
    {
      "epoch": 2.0783505154639177,
      "grad_norm": 0.1472669016069401,
      "learning_rate": 2.0000000000000012e-05,
      "loss": 0.5468,
      "step": 126
    },
    {
      "epoch": 2.0948453608247424,
      "grad_norm": 0.1383718741274944,
      "learning_rate": 1.9332025153925486e-05,
      "loss": 0.554,
      "step": 127
    },
    {
      "epoch": 2.111340206185567,
      "grad_norm": 0.1449682321168763,
      "learning_rate": 1.867182268793236e-05,
      "loss": 0.5481,
      "step": 128
    },
    {
      "epoch": 2.1278350515463917,
      "grad_norm": 0.14268562258623765,
      "learning_rate": 1.8019640877167763e-05,
      "loss": 0.551,
      "step": 129
    },
    {
      "epoch": 2.1443298969072164,
      "grad_norm": 0.14398336699894976,
      "learning_rate": 1.7375724980538465e-05,
      "loss": 0.5485,
      "step": 130
    },
    {
      "epoch": 2.160824742268041,
      "grad_norm": 0.13632304217235544,
      "learning_rate": 1.6740317148478932e-05,
      "loss": 0.5513,
      "step": 131
    },
    {
      "epoch": 2.177319587628866,
      "grad_norm": 0.13802596151410262,
      "learning_rate": 1.6113656331888563e-05,
      "loss": 0.546,
      "step": 132
    },
    {
      "epoch": 2.193814432989691,
      "grad_norm": 0.1367651968028799,
      "learning_rate": 1.5495978192271887e-05,
      "loss": 0.5507,
      "step": 133
    },
    {
      "epoch": 2.2103092783505156,
      "grad_norm": 0.14319896957050066,
      "learning_rate": 1.4887515013116067e-05,
      "loss": 0.5474,
      "step": 134
    },
    {
      "epoch": 2.2268041237113403,
      "grad_norm": 0.13563813749833395,
      "learning_rate": 1.4288495612538427e-05,
      "loss": 0.5466,
      "step": 135
    },
    {
      "epoch": 2.243298969072165,
      "grad_norm": 0.13299637580848364,
      "learning_rate": 1.369914525723746e-05,
      "loss": 0.5546,
      "step": 136
    },
    {
      "epoch": 2.2597938144329897,
      "grad_norm": 0.12840209113989406,
      "learning_rate": 1.3119685577779105e-05,
      "loss": 0.5443,
      "step": 137
    },
    {
      "epoch": 2.2762886597938143,
      "grad_norm": 0.12810536536017894,
      "learning_rate": 1.2550334485250661e-05,
      "loss": 0.5499,
      "step": 138
    },
    {
      "epoch": 2.292783505154639,
      "grad_norm": 0.11355419047013178,
      "learning_rate": 1.1991306089313261e-05,
      "loss": 0.5479,
      "step": 139
    },
    {
      "epoch": 2.3092783505154637,
      "grad_norm": 0.1295630499886211,
      "learning_rate": 1.1442810617684046e-05,
      "loss": 0.5489,
      "step": 140
    },
    {
      "epoch": 2.325773195876289,
      "grad_norm": 0.1221697199562323,
      "learning_rate": 1.0905054337078051e-05,
      "loss": 0.5499,
      "step": 141
    },
    {
      "epoch": 2.3422680412371135,
      "grad_norm": 0.10736730884852848,
      "learning_rate": 1.0378239475639823e-05,
      "loss": 0.5469,
      "step": 142
    },
    {
      "epoch": 2.358762886597938,
      "grad_norm": 0.11911805451017765,
      "learning_rate": 9.862564146893571e-06,
      "loss": 0.5533,
      "step": 143
    },
    {
      "epoch": 2.375257731958763,
      "grad_norm": 0.10719682452107784,
      "learning_rate": 9.358222275240884e-06,
      "loss": 0.5492,
      "step": 144
    },
    {
      "epoch": 2.3917525773195876,
      "grad_norm": 0.1062158910238879,
      "learning_rate": 8.8654035230336e-06,
      "loss": 0.549,
      "step": 145
    },
    {
      "epoch": 2.4082474226804123,
      "grad_norm": 0.10542623562823075,
      "learning_rate": 8.384293219249633e-06,
      "loss": 0.5472,
      "step": 146
    },
    {
      "epoch": 2.424742268041237,
      "grad_norm": 0.10737625408939123,
      "learning_rate": 7.915072289798247e-06,
      "loss": 0.5439,
      "step": 147
    },
    {
      "epoch": 2.441237113402062,
      "grad_norm": 0.113365548855691,
      "learning_rate": 7.457917189481301e-06,
      "loss": 0.5465,
      "step": 148
    },
    {
      "epoch": 2.4577319587628867,
      "grad_norm": 0.09803128955615437,
      "learning_rate": 7.0129998356357295e-06,
      "loss": 0.5454,
      "step": 149
    },
    {
      "epoch": 2.4742268041237114,
      "grad_norm": 0.10740236378002757,
      "learning_rate": 6.58048754348255e-06,
      "loss": 0.5532,
      "step": 150
    },
    {
      "epoch": 2.490721649484536,
      "grad_norm": 0.10615779921255894,
      "learning_rate": 6.160542963206357e-06,
      "loss": 0.5462,
      "step": 151
    },
    {
      "epoch": 2.507216494845361,
      "grad_norm": 0.10941467555755459,
      "learning_rate": 5.753324018789346e-06,
      "loss": 0.5467,
      "step": 152
    },
    {
      "epoch": 2.5237113402061855,
      "grad_norm": 0.09824679682670868,
      "learning_rate": 5.358983848622452e-06,
      "loss": 0.5464,
      "step": 153
    },
    {
      "epoch": 2.54020618556701,
      "grad_norm": 0.09719860472925167,
      "learning_rate": 4.97767074791637e-06,
      "loss": 0.5437,
      "step": 154
    },
    {
      "epoch": 2.556701030927835,
      "grad_norm": 0.09683396959072503,
      "learning_rate": 4.609528112933688e-06,
      "loss": 0.5477,
      "step": 155
    },
    {
      "epoch": 2.5731958762886595,
      "grad_norm": 0.09836422675496619,
      "learning_rate": 4.254694387063514e-06,
      "loss": 0.5469,
      "step": 156
    },
    {
      "epoch": 2.5896907216494847,
      "grad_norm": 0.0995681618956925,
      "learning_rate": 3.913303008758491e-06,
      "loss": 0.5482,
      "step": 157
    },
    {
      "epoch": 2.6061855670103093,
      "grad_norm": 0.09559938988699572,
      "learning_rate": 3.585482361354138e-06,
      "loss": 0.5463,
      "step": 158
    },
    {
      "epoch": 2.622680412371134,
      "grad_norm": 0.09220251838124616,
      "learning_rate": 3.2713557247890447e-06,
      "loss": 0.549,
      "step": 159
    },
    {
      "epoch": 2.6391752577319587,
      "grad_norm": 0.09182371190140297,
      "learning_rate": 2.9710412292443868e-06,
      "loss": 0.5447,
      "step": 160
    },
    {
      "epoch": 2.6556701030927834,
      "grad_norm": 0.09285117312965371,
      "learning_rate": 2.6846518107199782e-06,
      "loss": 0.5499,
      "step": 161
    },
    {
      "epoch": 2.6721649484536085,
      "grad_norm": 0.08888916031023086,
      "learning_rate": 2.4122951685636674e-06,
      "loss": 0.5476,
      "step": 162
    },
    {
      "epoch": 2.688659793814433,
      "grad_norm": 0.09108536955799308,
      "learning_rate": 2.1540737249699893e-06,
      "loss": 0.5468,
      "step": 163
    },
    {
      "epoch": 2.705154639175258,
      "grad_norm": 0.09161991297859239,
      "learning_rate": 1.9100845864633875e-06,
      "loss": 0.5447,
      "step": 164
    },
    {
      "epoch": 2.7216494845360826,
      "grad_norm": 0.08993683350116187,
      "learning_rate": 1.6804195073804442e-06,
      "loss": 0.547,
      "step": 165
    },
    {
      "epoch": 2.7381443298969073,
      "grad_norm": 0.08768348095869663,
      "learning_rate": 1.4651648553647869e-06,
      "loss": 0.548,
      "step": 166
    },
    {
      "epoch": 2.754639175257732,
      "grad_norm": 0.08908278856482474,
      "learning_rate": 1.2644015788877684e-06,
      "loss": 0.5424,
      "step": 167
    },
    {
      "epoch": 2.7711340206185566,
      "grad_norm": 0.08772182232275079,
      "learning_rate": 1.0782051768070477e-06,
      "loss": 0.5426,
      "step": 168
    },
    {
      "epoch": 2.7876288659793813,
      "grad_norm": 0.09014271400131238,
      "learning_rate": 9.066456699745774e-07,
      "loss": 0.5423,
      "step": 169
    },
    {
      "epoch": 2.804123711340206,
      "grad_norm": 0.08846808941818018,
      "learning_rate": 7.497875749046124e-07,
      "loss": 0.5491,
      "step": 170
    },
    {
      "epoch": 2.8206185567010307,
      "grad_norm": 0.09003919031348395,
      "learning_rate": 6.076898795116792e-07,
      "loss": 0.5466,
      "step": 171
    },
    {
      "epoch": 2.837113402061856,
      "grad_norm": 0.0860255355801196,
      "learning_rate": 4.804060209276396e-07,
      "loss": 0.5481,
      "step": 172
    },
    {
      "epoch": 2.8536082474226805,
      "grad_norm": 0.0863084170468407,
      "learning_rate": 3.679838654061874e-07,
      "loss": 0.5466,
      "step": 173
    },
    {
      "epoch": 2.870103092783505,
      "grad_norm": 0.08494132877549186,
      "learning_rate": 2.704656903222791e-07,
      "loss": 0.5481,
      "step": 174
    },
    {
      "epoch": 2.88659793814433,
      "grad_norm": 0.08696339412244179,
      "learning_rate": 1.8788816827336686e-07,
      "loss": 0.5488,
      "step": 175
    },
    {
      "epoch": 2.9030927835051545,
      "grad_norm": 0.08479502393635585,
      "learning_rate": 1.2028235328831906e-07,
      "loss": 0.5457,
      "step": 176
    },
    {
      "epoch": 2.9195876288659792,
      "grad_norm": 0.08544111428797167,
      "learning_rate": 6.767366914927298e-08,
      "loss": 0.5482,
      "step": 177
    },
    {
      "epoch": 2.9360824742268044,
      "grad_norm": 0.08423838312148532,
      "learning_rate": 3.0081899830798345e-08,
      "loss": 0.5468,
      "step": 178
    },
    {
      "epoch": 2.952577319587629,
      "grad_norm": 0.08547740551858984,
      "learning_rate": 7.521182059946342e-09,
      "loss": 0.5457,
      "step": 179
    },
    {
      "epoch": 2.9690721649484537,
      "grad_norm": 0.08570261529447104,
      "learning_rate": 0.0,
      "loss": 0.5454,
      "step": 180
    },
    {
      "epoch": 2.9690721649484537,
      "step": 180,
      "total_flos": 4.323339643798946e+18,
      "train_loss": 0.6483473062515259,
      "train_runtime": 10930.2571,
      "train_samples_per_second": 8.519,
      "train_steps_per_second": 0.016
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 180,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 4.323339643798946e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}