{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 231,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.030303030303030304,
      "grad_norm": 6.133312490270507,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.9888,
      "step": 1
    },
    {
      "epoch": 0.06060606060606061,
      "grad_norm": 6.312267353299651,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.0105,
      "step": 2
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 6.208055389022369,
      "learning_rate": 2.5e-06,
      "loss": 0.984,
      "step": 3
    },
    {
      "epoch": 0.12121212121212122,
      "grad_norm": 5.675461956735031,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.9786,
      "step": 4
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 4.66177066597852,
      "learning_rate": 4.166666666666667e-06,
      "loss": 1.0105,
      "step": 5
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 2.293639946463445,
      "learning_rate": 5e-06,
      "loss": 0.8871,
      "step": 6
    },
    {
      "epoch": 0.21212121212121213,
      "grad_norm": 1.9003604297261614,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.862,
      "step": 7
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 4.34057790860278,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.9222,
      "step": 8
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 4.481914051352927,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.88,
      "step": 9
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 4.54495940680815,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.873,
      "step": 10
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 3.7670378642227913,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.8166,
      "step": 11
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 3.8526941144194042,
      "learning_rate": 1e-05,
      "loss": 0.9013,
      "step": 12
    },
    {
      "epoch": 0.3939393939393939,
      "grad_norm": 2.461756702762672,
      "learning_rate": 1.0833333333333334e-05,
      "loss": 0.7934,
      "step": 13
    },
    {
      "epoch": 0.42424242424242425,
      "grad_norm": 2.020047328008979,
      "learning_rate": 1.1666666666666668e-05,
      "loss": 0.8134,
      "step": 14
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 2.339950318674221,
      "learning_rate": 1.25e-05,
      "loss": 0.7884,
      "step": 15
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 2.3738008055129898,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.7986,
      "step": 16
    },
    {
      "epoch": 0.5151515151515151,
      "grad_norm": 2.08368208039381,
      "learning_rate": 1.416666666666667e-05,
      "loss": 0.7553,
      "step": 17
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 1.5151847330974046,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.7819,
      "step": 18
    },
    {
      "epoch": 0.5757575757575758,
      "grad_norm": 1.4130438397580987,
      "learning_rate": 1.5833333333333333e-05,
      "loss": 0.757,
      "step": 19
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 1.5602523903739551,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.7117,
      "step": 20
    },
    {
      "epoch": 0.6363636363636364,
      "grad_norm": 1.1219040774993525,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 0.7214,
      "step": 21
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 1.225922434964794,
      "learning_rate": 1.8333333333333333e-05,
      "loss": 0.7587,
      "step": 22
    },
    {
      "epoch": 0.696969696969697,
      "grad_norm": 1.1494919772986516,
      "learning_rate": 1.916666666666667e-05,
      "loss": 0.6967,
      "step": 23
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 1.1189446177809441,
      "learning_rate": 2e-05,
      "loss": 0.7407,
      "step": 24
    },
    {
      "epoch": 0.7575757575757576,
      "grad_norm": 0.8390109807203933,
      "learning_rate": 1.999884834944106e-05,
      "loss": 0.741,
      "step": 25
    },
    {
      "epoch": 0.7878787878787878,
      "grad_norm": 1.007712076574321,
      "learning_rate": 1.9995393663024054e-05,
      "loss": 0.7249,
      "step": 26
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 0.7348503701924564,
      "learning_rate": 1.9989636736467278e-05,
      "loss": 0.6793,
      "step": 27
    },
    {
      "epoch": 0.8484848484848485,
      "grad_norm": 0.8109780365875735,
      "learning_rate": 1.9981578895764272e-05,
      "loss": 0.6822,
      "step": 28
    },
    {
      "epoch": 0.8787878787878788,
      "grad_norm": 0.9304526383993851,
      "learning_rate": 1.9971221996878395e-05,
      "loss": 0.7035,
      "step": 29
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.8529935025494286,
      "learning_rate": 1.9958568425315316e-05,
      "loss": 0.7038,
      "step": 30
    },
    {
      "epoch": 0.9393939393939394,
      "grad_norm": 0.7556881934557372,
      "learning_rate": 1.9943621095573588e-05,
      "loss": 0.6911,
      "step": 31
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.7735812382379526,
      "learning_rate": 1.9926383450473344e-05,
      "loss": 0.679,
      "step": 32
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.6923641401736521,
      "learning_rate": 1.9906859460363307e-05,
      "loss": 0.6661,
      "step": 33
    },
    {
      "epoch": 1.0303030303030303,
      "grad_norm": 0.6826580303946777,
      "learning_rate": 1.9885053622206305e-05,
      "loss": 0.6344,
      "step": 34
    },
    {
      "epoch": 1.0606060606060606,
      "grad_norm": 0.6878400013541692,
      "learning_rate": 1.986097095854347e-05,
      "loss": 0.6168,
      "step": 35
    },
    {
      "epoch": 1.0909090909090908,
      "grad_norm": 0.6308834404286537,
      "learning_rate": 1.9834617016337424e-05,
      "loss": 0.6138,
      "step": 36
    },
    {
      "epoch": 1.121212121212121,
      "grad_norm": 0.8855682900976705,
      "learning_rate": 1.9805997865694616e-05,
      "loss": 0.6027,
      "step": 37
    },
    {
      "epoch": 1.1515151515151516,
      "grad_norm": 0.650276286498742,
      "learning_rate": 1.9775120098467212e-05,
      "loss": 0.6063,
      "step": 38
    },
    {
      "epoch": 1.1818181818181819,
      "grad_norm": 0.9241353438375617,
      "learning_rate": 1.9741990826734793e-05,
      "loss": 0.6487,
      "step": 39
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 0.6131304185741522,
      "learning_rate": 1.970661768116622e-05,
      "loss": 0.6125,
      "step": 40
    },
    {
      "epoch": 1.2424242424242424,
      "grad_norm": 0.9638298028620941,
      "learning_rate": 1.9669008809262064e-05,
      "loss": 0.6375,
      "step": 41
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 0.6864083028962933,
      "learning_rate": 1.9629172873477995e-05,
      "loss": 0.6064,
      "step": 42
    },
    {
      "epoch": 1.303030303030303,
      "grad_norm": 0.6541942252236372,
      "learning_rate": 1.9587119049229558e-05,
      "loss": 0.5929,
      "step": 43
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.591616863460495,
      "learning_rate": 1.954285702277879e-05,
      "loss": 0.5862,
      "step": 44
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.5788113217386186,
      "learning_rate": 1.9496396989003195e-05,
      "loss": 0.6165,
      "step": 45
    },
    {
      "epoch": 1.393939393939394,
      "grad_norm": 0.5825389483651302,
      "learning_rate": 1.944774964904754e-05,
      "loss": 0.6011,
      "step": 46
    },
    {
      "epoch": 1.4242424242424243,
      "grad_norm": 0.5950620245349484,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 0.6157,
      "step": 47
    },
    {
      "epoch": 1.4545454545454546,
      "grad_norm": 0.548727099377894,
      "learning_rate": 1.9343938371606714e-05,
      "loss": 0.5794,
      "step": 48
    },
    {
      "epoch": 1.4848484848484849,
      "grad_norm": 0.5785459402189005,
      "learning_rate": 1.9288798344984673e-05,
      "loss": 0.6099,
      "step": 49
    },
    {
      "epoch": 1.5151515151515151,
      "grad_norm": 0.6969881421938213,
      "learning_rate": 1.9231518828401458e-05,
      "loss": 0.6181,
      "step": 50
    },
    {
      "epoch": 1.5454545454545454,
      "grad_norm": 0.6014346704940163,
      "learning_rate": 1.917211301505453e-05,
      "loss": 0.6061,
      "step": 51
    },
    {
      "epoch": 1.5757575757575757,
      "grad_norm": 0.557203444582592,
      "learning_rate": 1.911059458789152e-05,
      "loss": 0.577,
      "step": 52
    },
    {
      "epoch": 1.606060606060606,
      "grad_norm": 0.6238802409413811,
      "learning_rate": 1.9046977716458627e-05,
      "loss": 0.633,
      "step": 53
    },
    {
      "epoch": 1.6363636363636362,
      "grad_norm": 0.5908668210103087,
      "learning_rate": 1.8981277053636963e-05,
      "loss": 0.5868,
      "step": 54
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.6377646074733742,
      "learning_rate": 1.891350773226754e-05,
      "loss": 0.5837,
      "step": 55
    },
    {
      "epoch": 1.696969696969697,
      "grad_norm": 0.6706654613912634,
      "learning_rate": 1.8843685361665724e-05,
      "loss": 0.6058,
      "step": 56
    },
    {
      "epoch": 1.7272727272727273,
      "grad_norm": 0.7135082134395438,
      "learning_rate": 1.8771826024025944e-05,
      "loss": 0.6108,
      "step": 57
    },
    {
      "epoch": 1.7575757575757576,
      "grad_norm": 0.6613757784482449,
      "learning_rate": 1.8697946270717468e-05,
      "loss": 0.6089,
      "step": 58
    },
    {
      "epoch": 1.7878787878787878,
      "grad_norm": 0.5892114524066855,
      "learning_rate": 1.8622063118472135e-05,
      "loss": 0.5978,
      "step": 59
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.613752632598102,
      "learning_rate": 1.8544194045464888e-05,
      "loss": 0.5784,
      "step": 60
    },
    {
      "epoch": 1.8484848484848486,
      "grad_norm": 0.6533078225121903,
      "learning_rate": 1.8464356987288012e-05,
      "loss": 0.5652,
      "step": 61
    },
    {
      "epoch": 1.878787878787879,
      "grad_norm": 0.5268805797122355,
      "learning_rate": 1.8382570332820045e-05,
      "loss": 0.5551,
      "step": 62
    },
    {
      "epoch": 1.9090909090909092,
      "grad_norm": 0.5718376684652964,
      "learning_rate": 1.8298852919990254e-05,
      "loss": 0.5993,
      "step": 63
    },
    {
      "epoch": 1.9393939393939394,
      "grad_norm": 0.5681556224844309,
      "learning_rate": 1.821322403143969e-05,
      "loss": 0.6061,
      "step": 64
    },
    {
      "epoch": 1.9696969696969697,
      "grad_norm": 0.5160510923850339,
      "learning_rate": 1.812570339007983e-05,
      "loss": 0.539,
      "step": 65
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.5808021119660712,
      "learning_rate": 1.8036311154549783e-05,
      "loss": 0.5604,
      "step": 66
    },
    {
      "epoch": 2.0303030303030303,
      "grad_norm": 0.6969394228030072,
      "learning_rate": 1.7945067914573147e-05,
      "loss": 0.4949,
      "step": 67
    },
    {
      "epoch": 2.0606060606060606,
      "grad_norm": 0.5831869434691073,
      "learning_rate": 1.7851994686215592e-05,
      "loss": 0.4799,
      "step": 68
    },
    {
      "epoch": 2.090909090909091,
      "grad_norm": 0.8189090133243225,
      "learning_rate": 1.77571129070442e-05,
      "loss": 0.4757,
      "step": 69
    },
    {
      "epoch": 2.121212121212121,
      "grad_norm": 0.6924137717196174,
      "learning_rate": 1.766044443118978e-05,
      "loss": 0.45,
      "step": 70
    },
    {
      "epoch": 2.1515151515151514,
      "grad_norm": 0.5579284812500711,
      "learning_rate": 1.7562011524313187e-05,
      "loss": 0.4652,
      "step": 71
    },
    {
      "epoch": 2.1818181818181817,
      "grad_norm": 0.7280937667428176,
      "learning_rate": 1.7461836858476858e-05,
      "loss": 0.4853,
      "step": 72
    },
    {
      "epoch": 2.212121212121212,
      "grad_norm": 0.5542550746479272,
      "learning_rate": 1.7359943506922775e-05,
      "loss": 0.4642,
      "step": 73
    },
    {
      "epoch": 2.242424242424242,
      "grad_norm": 0.6427517726580945,
      "learning_rate": 1.725635493875799e-05,
      "loss": 0.4558,
      "step": 74
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.5769363799510141,
      "learning_rate": 1.7151095013548996e-05,
      "loss": 0.448,
      "step": 75
    },
    {
      "epoch": 2.303030303030303,
      "grad_norm": 0.6236420369243885,
      "learning_rate": 1.7044187975826126e-05,
      "loss": 0.4539,
      "step": 76
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.5495637430654123,
      "learning_rate": 1.693565844949933e-05,
      "loss": 0.4505,
      "step": 77
    },
    {
      "epoch": 2.3636363636363638,
      "grad_norm": 0.6462062247160016,
      "learning_rate": 1.6825531432186545e-05,
      "loss": 0.4683,
      "step": 78
    },
    {
      "epoch": 2.393939393939394,
      "grad_norm": 0.4938607242900449,
      "learning_rate": 1.671383228945597e-05,
      "loss": 0.4446,
      "step": 79
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 0.7982112275998522,
      "learning_rate": 1.6600586748983642e-05,
      "loss": 0.4697,
      "step": 80
    },
    {
      "epoch": 2.4545454545454546,
      "grad_norm": 0.5745468299818484,
      "learning_rate": 1.648582089462756e-05,
      "loss": 0.4585,
      "step": 81
    },
    {
      "epoch": 2.484848484848485,
      "grad_norm": 0.5564417346991974,
      "learning_rate": 1.6369561160419783e-05,
      "loss": 0.4643,
      "step": 82
    },
    {
      "epoch": 2.515151515151515,
      "grad_norm": 0.49905270942102536,
      "learning_rate": 1.625183432447789e-05,
      "loss": 0.4602,
      "step": 83
    },
    {
      "epoch": 2.5454545454545454,
      "grad_norm": 0.4796081226277374,
      "learning_rate": 1.6132667502837164e-05,
      "loss": 0.4655,
      "step": 84
    },
    {
      "epoch": 2.5757575757575757,
      "grad_norm": 0.4927738870874845,
      "learning_rate": 1.6012088143204953e-05,
      "loss": 0.4538,
      "step": 85
    },
    {
      "epoch": 2.606060606060606,
      "grad_norm": 0.5180099965159457,
      "learning_rate": 1.589012401863864e-05,
      "loss": 0.458,
      "step": 86
    },
    {
      "epoch": 2.6363636363636362,
      "grad_norm": 0.5296320750810951,
      "learning_rate": 1.5766803221148676e-05,
      "loss": 0.4493,
      "step": 87
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.5733112965300271,
      "learning_rate": 1.5642154155228124e-05,
      "loss": 0.4686,
      "step": 88
    },
    {
      "epoch": 2.6969696969696972,
      "grad_norm": 0.5741347142903017,
      "learning_rate": 1.5516205531310272e-05,
      "loss": 0.4583,
      "step": 89
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.48174318336790667,
      "learning_rate": 1.538898635915576e-05,
      "loss": 0.4425,
      "step": 90
    },
    {
      "epoch": 2.757575757575758,
      "grad_norm": 0.5590385352635306,
      "learning_rate": 1.526052594117071e-05,
      "loss": 0.4464,
      "step": 91
    },
    {
      "epoch": 2.787878787878788,
      "grad_norm": 0.485518620325785,
      "learning_rate": 1.513085386565758e-05,
      "loss": 0.4436,
      "step": 92
    },
    {
      "epoch": 2.8181818181818183,
      "grad_norm": 0.49380448272343497,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.4591,
      "step": 93
    },
    {
      "epoch": 2.8484848484848486,
      "grad_norm": 0.5571812552248251,
      "learning_rate": 1.4867994483783485e-05,
      "loss": 0.4721,
      "step": 94
    },
    {
      "epoch": 2.878787878787879,
      "grad_norm": 0.5230295468295674,
      "learning_rate": 1.4734867721853341e-05,
      "loss": 0.4649,
      "step": 95
    },
    {
      "epoch": 2.909090909090909,
      "grad_norm": 0.5089127017794696,
      "learning_rate": 1.4600650377311523e-05,
      "loss": 0.4814,
      "step": 96
    },
    {
      "epoch": 2.9393939393939394,
      "grad_norm": 0.5105539748113577,
      "learning_rate": 1.4465373364454001e-05,
      "loss": 0.4503,
      "step": 97
    },
    {
      "epoch": 2.9696969696969697,
      "grad_norm": 0.45069990743245353,
      "learning_rate": 1.4329067841650274e-05,
      "loss": 0.4387,
      "step": 98
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.5033861625442646,
      "learning_rate": 1.4191765204166643e-05,
      "loss": 0.4538,
      "step": 99
    },
    {
      "epoch": 3.0303030303030303,
      "grad_norm": 0.7754730175625122,
      "learning_rate": 1.4053497076934948e-05,
      "loss": 0.3659,
      "step": 100
    },
    {
      "epoch": 3.0606060606060606,
      "grad_norm": 0.5140533227089905,
      "learning_rate": 1.3914295307268396e-05,
      "loss": 0.3641,
      "step": 101
    },
    {
      "epoch": 3.090909090909091,
      "grad_norm": 0.9403979311007741,
      "learning_rate": 1.3774191957526144e-05,
      "loss": 0.3546,
      "step": 102
    },
    {
      "epoch": 3.121212121212121,
      "grad_norm": 0.8010927180149187,
      "learning_rate": 1.3633219297728415e-05,
      "loss": 0.3555,
      "step": 103
    },
    {
      "epoch": 3.1515151515151514,
      "grad_norm": 0.6970449984340281,
      "learning_rate": 1.3491409798123687e-05,
      "loss": 0.3413,
      "step": 104
    },
    {
      "epoch": 3.1818181818181817,
      "grad_norm": 0.6812848683097438,
      "learning_rate": 1.3348796121709862e-05,
      "loss": 0.3495,
      "step": 105
    },
    {
      "epoch": 3.212121212121212,
      "grad_norm": 0.7872828399093813,
      "learning_rate": 1.3205411116710973e-05,
      "loss": 0.3461,
      "step": 106
    },
    {
      "epoch": 3.242424242424242,
      "grad_norm": 0.6352287831348774,
      "learning_rate": 1.3061287809011243e-05,
      "loss": 0.3422,
      "step": 107
    },
    {
      "epoch": 3.2727272727272725,
      "grad_norm": 0.5371263304613854,
      "learning_rate": 1.291645939454825e-05,
      "loss": 0.3297,
      "step": 108
    },
    {
      "epoch": 3.303030303030303,
      "grad_norm": 0.6106232841146089,
      "learning_rate": 1.277095923166689e-05,
      "loss": 0.3369,
      "step": 109
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.5764495175042285,
      "learning_rate": 1.2624820833435939e-05,
      "loss": 0.3432,
      "step": 110
    },
    {
      "epoch": 3.3636363636363638,
      "grad_norm": 0.5189044479833499,
      "learning_rate": 1.2478077859929e-05,
      "loss": 0.3225,
      "step": 111
    },
    {
      "epoch": 3.393939393939394,
      "grad_norm": 0.5420009037776947,
      "learning_rate": 1.2330764110471567e-05,
      "loss": 0.3335,
      "step": 112
    },
    {
      "epoch": 3.4242424242424243,
      "grad_norm": 0.4862496515664332,
      "learning_rate": 1.2182913515856016e-05,
      "loss": 0.3318,
      "step": 113
    },
    {
      "epoch": 3.4545454545454546,
      "grad_norm": 0.5090167554828083,
      "learning_rate": 1.2034560130526341e-05,
      "loss": 0.3494,
      "step": 114
    },
    {
      "epoch": 3.484848484848485,
      "grad_norm": 0.4911623919328435,
      "learning_rate": 1.1885738124734359e-05,
      "loss": 0.3275,
      "step": 115
    },
    {
      "epoch": 3.515151515151515,
      "grad_norm": 0.5045605919032637,
      "learning_rate": 1.1736481776669307e-05,
      "loss": 0.3572,
      "step": 116
    },
    {
      "epoch": 3.5454545454545454,
      "grad_norm": 0.504947761060277,
      "learning_rate": 1.1586825464562515e-05,
      "loss": 0.3451,
      "step": 117
    },
    {
      "epoch": 3.5757575757575757,
      "grad_norm": 0.4882425876038978,
      "learning_rate": 1.1436803658769082e-05,
      "loss": 0.3544,
      "step": 118
    },
    {
      "epoch": 3.606060606060606,
      "grad_norm": 0.4850349784810182,
      "learning_rate": 1.1286450913828313e-05,
      "loss": 0.3193,
      "step": 119
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.5320436358006013,
      "learning_rate": 1.113580186050475e-05,
      "loss": 0.3453,
      "step": 120
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.4825025254463151,
      "learning_rate": 1.0984891197811686e-05,
      "loss": 0.3395,
      "step": 121
    },
    {
      "epoch": 3.6969696969696972,
      "grad_norm": 0.47238049685696304,
      "learning_rate": 1.0833753685018935e-05,
      "loss": 0.3281,
      "step": 122
    },
    {
      "epoch": 3.7272727272727275,
      "grad_norm": 0.5347752627285653,
      "learning_rate": 1.0682424133646712e-05,
      "loss": 0.3402,
      "step": 123
    },
    {
      "epoch": 3.757575757575758,
      "grad_norm": 0.4948810103681615,
      "learning_rate": 1.0530937399447496e-05,
      "loss": 0.3336,
      "step": 124
    },
    {
      "epoch": 3.787878787878788,
      "grad_norm": 0.5191359531791342,
      "learning_rate": 1.0379328374377715e-05,
      "loss": 0.347,
      "step": 125
    },
    {
      "epoch": 3.8181818181818183,
      "grad_norm": 0.4176553849951949,
      "learning_rate": 1.0227631978561057e-05,
      "loss": 0.3479,
      "step": 126
    },
    {
      "epoch": 3.8484848484848486,
      "grad_norm": 0.5299308743413071,
      "learning_rate": 1.0075883152245334e-05,
      "loss": 0.345,
      "step": 127
    },
    {
      "epoch": 3.878787878787879,
      "grad_norm": 0.46549440438780987,
      "learning_rate": 9.92411684775467e-06,
      "loss": 0.3178,
      "step": 128
    },
    {
      "epoch": 3.909090909090909,
      "grad_norm": 0.4758499117360986,
      "learning_rate": 9.772368021438943e-06,
      "loss": 0.3395,
      "step": 129
    },
    {
      "epoch": 3.9393939393939394,
      "grad_norm": 0.44324711245538406,
      "learning_rate": 9.620671625622287e-06,
      "loss": 0.3375,
      "step": 130
    },
    {
      "epoch": 3.9696969696969697,
      "grad_norm": 0.47116040531641634,
      "learning_rate": 9.469062600552509e-06,
      "loss": 0.3341,
      "step": 131
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.43818067708802083,
      "learning_rate": 9.317575866353293e-06,
      "loss": 0.3234,
      "step": 132
    },
    {
      "epoch": 4.03030303030303,
      "grad_norm": 0.737485769336963,
      "learning_rate": 9.166246314981066e-06,
      "loss": 0.2627,
      "step": 133
    },
    {
      "epoch": 4.0606060606060606,
      "grad_norm": 0.5529080776552364,
      "learning_rate": 9.015108802188314e-06,
      "loss": 0.26,
      "step": 134
    },
    {
      "epoch": 4.090909090909091,
      "grad_norm": 0.48080243212121565,
      "learning_rate": 8.86419813949525e-06,
      "loss": 0.2552,
      "step": 135
    },
    {
      "epoch": 4.121212121212121,
      "grad_norm": 0.7350512487406256,
      "learning_rate": 8.71354908617169e-06,
      "loss": 0.2405,
      "step": 136
    },
    {
      "epoch": 4.151515151515151,
      "grad_norm": 0.7784132067101827,
      "learning_rate": 8.56319634123092e-06,
      "loss": 0.2572,
      "step": 137
    },
    {
      "epoch": 4.181818181818182,
      "grad_norm": 0.46727652708176803,
      "learning_rate": 8.413174535437486e-06,
      "loss": 0.224,
      "step": 138
    },
    {
      "epoch": 4.212121212121212,
      "grad_norm": 0.512190716791503,
      "learning_rate": 8.263518223330698e-06,
      "loss": 0.2409,
      "step": 139
    },
    {
      "epoch": 4.242424242424242,
      "grad_norm": 0.6123196750859912,
      "learning_rate": 8.114261875265643e-06,
      "loss": 0.2375,
      "step": 140
    },
    {
      "epoch": 4.2727272727272725,
      "grad_norm": 0.5069870710442763,
      "learning_rate": 7.965439869473664e-06,
      "loss": 0.2547,
      "step": 141
    },
    {
      "epoch": 4.303030303030303,
      "grad_norm": 0.4849101223324498,
      "learning_rate": 7.817086484143987e-06,
      "loss": 0.2483,
      "step": 142
    },
    {
      "epoch": 4.333333333333333,
      "grad_norm": 0.48073299903189964,
      "learning_rate": 7.669235889528436e-06,
      "loss": 0.2383,
      "step": 143
    },
    {
      "epoch": 4.363636363636363,
      "grad_norm": 0.5040942298901941,
      "learning_rate": 7.521922140071003e-06,
      "loss": 0.2337,
      "step": 144
    },
    {
      "epoch": 4.393939393939394,
      "grad_norm": 0.5174058357260617,
      "learning_rate": 7.375179166564062e-06,
      "loss": 0.2416,
      "step": 145
    },
    {
      "epoch": 4.424242424242424,
      "grad_norm": 0.5263164387363304,
      "learning_rate": 7.2290407683331154e-06,
      "loss": 0.2352,
      "step": 146
    },
    {
      "epoch": 4.454545454545454,
      "grad_norm": 0.48585757237463617,
      "learning_rate": 7.0835406054517505e-06,
      "loss": 0.2248,
      "step": 147
    },
    {
      "epoch": 4.484848484848484,
      "grad_norm": 0.456902311560064,
      "learning_rate": 6.93871219098876e-06,
      "loss": 0.2446,
      "step": 148
    },
    {
      "epoch": 4.515151515151516,
      "grad_norm": 0.4647131059560458,
      "learning_rate": 6.79458888328903e-06,
      "loss": 0.2389,
      "step": 149
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 0.43264056127979783,
      "learning_rate": 6.651203878290139e-06,
      "loss": 0.2574,
      "step": 150
    },
    {
      "epoch": 4.575757575757576,
      "grad_norm": 0.45526716122106875,
      "learning_rate": 6.508590201876317e-06,
      "loss": 0.2376,
      "step": 151
    },
    {
      "epoch": 4.606060606060606,
      "grad_norm": 0.46874751338386467,
      "learning_rate": 6.366780702271589e-06,
      "loss": 0.2559,
      "step": 152
    },
    {
      "epoch": 4.636363636363637,
      "grad_norm": 0.457100921015383,
      "learning_rate": 6.225808042473857e-06,
      "loss": 0.2514,
      "step": 153
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.4605869515605553,
      "learning_rate": 6.085704692731609e-06,
      "loss": 0.2362,
      "step": 154
    },
    {
      "epoch": 4.696969696969697,
      "grad_norm": 0.4473331033802291,
      "learning_rate": 5.946502923065054e-06,
      "loss": 0.2323,
      "step": 155
    },
    {
      "epoch": 4.7272727272727275,
      "grad_norm": 0.42628612014612893,
      "learning_rate": 5.8082347958333625e-06,
      "loss": 0.2407,
      "step": 156
    },
    {
      "epoch": 4.757575757575758,
      "grad_norm": 0.41450115644914304,
      "learning_rate": 5.670932158349732e-06,
      "loss": 0.2424,
      "step": 157
    },
    {
      "epoch": 4.787878787878788,
      "grad_norm": 0.4055025538345423,
      "learning_rate": 5.534626635546e-06,
      "loss": 0.2375,
      "step": 158
    },
    {
      "epoch": 4.818181818181818,
      "grad_norm": 0.44529585756329015,
      "learning_rate": 5.399349622688479e-06,
      "loss": 0.2308,
      "step": 159
    },
    {
      "epoch": 4.848484848484849,
      "grad_norm": 0.4471917638100028,
      "learning_rate": 5.2651322781466606e-06,
      "loss": 0.234,
      "step": 160
    },
    {
      "epoch": 4.878787878787879,
      "grad_norm": 0.4223307232900756,
      "learning_rate": 5.132005516216512e-06,
      "loss": 0.2426,
      "step": 161
    },
    {
      "epoch": 4.909090909090909,
      "grad_norm": 0.4372133744429308,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.2258,
      "step": 162
    },
    {
      "epoch": 4.9393939393939394,
      "grad_norm": 0.4420625006970779,
      "learning_rate": 4.869146134342426e-06,
      "loss": 0.2457,
      "step": 163
    },
    {
      "epoch": 4.96969696969697,
      "grad_norm": 0.41903834567451254,
      "learning_rate": 4.739474058829288e-06,
      "loss": 0.2447,
      "step": 164
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.41220312098219003,
      "learning_rate": 4.611013640844245e-06,
      "loss": 0.2394,
      "step": 165
    },
    {
      "epoch": 5.03030303030303,
      "grad_norm": 0.6003100588562155,
      "learning_rate": 4.483794468689728e-06,
      "loss": 0.1735,
      "step": 166
    },
    {
      "epoch": 5.0606060606060606,
      "grad_norm": 0.5648306970051146,
      "learning_rate": 4.357845844771881e-06,
      "loss": 0.1903,
      "step": 167
    },
    {
      "epoch": 5.090909090909091,
      "grad_norm": 0.42700456514849605,
      "learning_rate": 4.2331967788513295e-06,
      "loss": 0.1752,
      "step": 168
    },
    {
      "epoch": 5.121212121212121,
      "grad_norm": 0.42202181300842884,
      "learning_rate": 4.109875981361363e-06,
      "loss": 0.1688,
      "step": 169
    },
    {
      "epoch": 5.151515151515151,
      "grad_norm": 0.5838409787441209,
      "learning_rate": 3.987911856795047e-06,
      "loss": 0.1782,
      "step": 170
    },
    {
      "epoch": 5.181818181818182,
      "grad_norm": 0.6457145473441536,
      "learning_rate": 3.867332497162836e-06,
      "loss": 0.1692,
      "step": 171
    },
    {
      "epoch": 5.212121212121212,
      "grad_norm": 0.5629897276540763,
      "learning_rate": 3.748165675522113e-06,
      "loss": 0.1818,
      "step": 172
    },
    {
      "epoch": 5.242424242424242,
      "grad_norm": 0.4578791065982633,
      "learning_rate": 3.630438839580217e-06,
      "loss": 0.1816,
      "step": 173
    },
    {
      "epoch": 5.2727272727272725,
      "grad_norm": 0.4244633490788637,
      "learning_rate": 3.5141791053724405e-06,
      "loss": 0.1691,
      "step": 174
    },
    {
      "epoch": 5.303030303030303,
      "grad_norm": 0.43707284617669917,
      "learning_rate": 3.399413251016359e-06,
      "loss": 0.1825,
      "step": 175
    },
    {
      "epoch": 5.333333333333333,
      "grad_norm": 0.4660244768094549,
      "learning_rate": 3.2861677105440335e-06,
      "loss": 0.1784,
      "step": 176
    },
    {
      "epoch": 5.363636363636363,
      "grad_norm": 0.45078401522152955,
      "learning_rate": 3.174468567813461e-06,
      "loss": 0.1698,
      "step": 177
    },
    {
      "epoch": 5.393939393939394,
      "grad_norm": 0.4465656518722392,
      "learning_rate": 3.0643415505006733e-06,
      "loss": 0.1855,
      "step": 178
    },
    {
      "epoch": 5.424242424242424,
      "grad_norm": 0.43625841204668764,
      "learning_rate": 2.9558120241738786e-06,
      "loss": 0.1835,
      "step": 179
    },
    {
      "epoch": 5.454545454545454,
      "grad_norm": 0.40736478162641304,
      "learning_rate": 2.8489049864510053e-06,
      "loss": 0.1801,
      "step": 180
    },
    {
      "epoch": 5.484848484848484,
      "grad_norm": 0.3747193980913415,
      "learning_rate": 2.7436450612420098e-06,
      "loss": 0.1727,
      "step": 181
    },
    {
      "epoch": 5.515151515151516,
      "grad_norm": 0.39608234441553114,
      "learning_rate": 2.640056493077231e-06,
      "loss": 0.176,
      "step": 182
    },
    {
      "epoch": 5.545454545454545,
      "grad_norm": 0.41959398355677646,
      "learning_rate": 2.5381631415231455e-06,
      "loss": 0.1879,
      "step": 183
    },
    {
      "epoch": 5.575757575757576,
      "grad_norm": 0.4254934910328529,
      "learning_rate": 2.4379884756868167e-06,
      "loss": 0.1791,
      "step": 184
    },
    {
      "epoch": 5.606060606060606,
      "grad_norm": 0.4124792487276637,
      "learning_rate": 2.339555568810221e-06,
      "loss": 0.1798,
      "step": 185
    },
    {
      "epoch": 5.636363636363637,
      "grad_norm": 0.40905699809266405,
      "learning_rate": 2.2428870929558012e-06,
      "loss": 0.1879,
      "step": 186
    },
    {
      "epoch": 5.666666666666667,
      "grad_norm": 0.439209399354481,
      "learning_rate": 2.1480053137844115e-06,
      "loss": 0.1757,
      "step": 187
    },
    {
      "epoch": 5.696969696969697,
      "grad_norm": 0.38866717217479696,
      "learning_rate": 2.054932085426856e-06,
      "loss": 0.177,
      "step": 188
    },
    {
      "epoch": 5.7272727272727275,
      "grad_norm": 0.3877198550366768,
      "learning_rate": 1.963688845450218e-06,
      "loss": 0.1691,
      "step": 189
    },
    {
      "epoch": 5.757575757575758,
      "grad_norm": 0.38937605220810684,
      "learning_rate": 1.8742966099201699e-06,
      "loss": 0.1729,
      "step": 190
    },
    {
      "epoch": 5.787878787878788,
      "grad_norm": 0.3931967989696538,
      "learning_rate": 1.7867759685603115e-06,
      "loss": 0.1687,
      "step": 191
    },
    {
      "epoch": 5.818181818181818,
      "grad_norm": 0.4063426644619139,
      "learning_rate": 1.7011470800097496e-06,
      "loss": 0.1811,
      "step": 192
    },
    {
      "epoch": 5.848484848484849,
      "grad_norm": 0.3987773436262393,
      "learning_rate": 1.6174296671799571e-06,
      "loss": 0.1808,
      "step": 193
    },
    {
      "epoch": 5.878787878787879,
      "grad_norm": 0.42485860526619407,
      "learning_rate": 1.5356430127119915e-06,
      "loss": 0.1692,
      "step": 194
    },
    {
      "epoch": 5.909090909090909,
      "grad_norm": 0.3865913121103735,
      "learning_rate": 1.4558059545351144e-06,
      "loss": 0.1809,
      "step": 195
    },
    {
      "epoch": 5.9393939393939394,
      "grad_norm": 0.4087920786517138,
      "learning_rate": 1.3779368815278648e-06,
      "loss": 0.1695,
      "step": 196
    },
    {
      "epoch": 5.96969696969697,
      "grad_norm": 0.3774266154893843,
      "learning_rate": 1.302053729282533e-06,
      "loss": 0.1795,
      "step": 197
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.4076951621470744,
      "learning_rate": 1.2281739759740575e-06,
      "loss": 0.1638,
      "step": 198
    },
    {
      "epoch": 6.03030303030303,
      "grad_norm": 0.45466301759345923,
      "learning_rate": 1.156314638334277e-06,
      "loss": 0.1534,
      "step": 199
    },
    {
      "epoch": 6.0606060606060606,
      "grad_norm": 0.4323237765282992,
      "learning_rate": 1.086492267732462e-06,
      "loss": 0.1432,
      "step": 200
    },
    {
      "epoch": 6.090909090909091,
      "grad_norm": 0.39898281711174816,
      "learning_rate": 1.01872294636304e-06,
      "loss": 0.1367,
      "step": 201
    },
    {
      "epoch": 6.121212121212121,
      "grad_norm": 0.3557934068122329,
      "learning_rate": 9.530222835413739e-07,
      "loss": 0.1561,
      "step": 202
    },
    {
      "epoch": 6.151515151515151,
      "grad_norm": 0.38265274333153265,
      "learning_rate": 8.894054121084839e-07,
      "loss": 0.1542,
      "step": 203
    },
    {
      "epoch": 6.181818181818182,
      "grad_norm": 0.33626439767237964,
      "learning_rate": 8.278869849454718e-07,
      "loss": 0.1508,
      "step": 204
    },
    {
      "epoch": 6.212121212121212,
      "grad_norm": 0.3480623394329382,
      "learning_rate": 7.684811715985429e-07,
      "loss": 0.1435,
      "step": 205
    },
    {
      "epoch": 6.242424242424242,
      "grad_norm": 0.3637611567267387,
      "learning_rate": 7.1120165501533e-07,
      "loss": 0.1436,
      "step": 206
    },
    {
      "epoch": 6.2727272727272725,
      "grad_norm": 0.3756095538107021,
      "learning_rate": 6.560616283932897e-07,
      "loss": 0.1441,
      "step": 207
    },
    {
      "epoch": 6.303030303030303,
      "grad_norm": 0.4233715419677722,
      "learning_rate": 6.030737921409169e-07,
      "loss": 0.1418,
      "step": 208
    },
    {
      "epoch": 6.333333333333333,
      "grad_norm": 0.3959442236720861,
      "learning_rate": 5.522503509524591e-07,
      "loss": 0.1446,
      "step": 209
    },
    {
      "epoch": 6.363636363636363,
      "grad_norm": 0.40746572697437256,
      "learning_rate": 5.036030109968082e-07,
      "loss": 0.164,
      "step": 210
    },
    {
      "epoch": 6.393939393939394,
      "grad_norm": 0.3579587142573846,
      "learning_rate": 4.5714297722121105e-07,
      "loss": 0.1457,
      "step": 211
    },
    {
      "epoch": 6.424242424242424,
      "grad_norm": 0.3917891786740035,
      "learning_rate": 4.128809507704445e-07,
      "loss": 0.1494,
      "step": 212
    },
    {
      "epoch": 6.454545454545454,
      "grad_norm": 0.3545953463168816,
      "learning_rate": 3.708271265220087e-07,
      "loss": 0.1574,
      "step": 213
    },
    {
      "epoch": 6.484848484848484,
      "grad_norm": 0.37736081342197736,
      "learning_rate": 3.309911907379393e-07,
      "loss": 0.1331,
      "step": 214
    },
    {
      "epoch": 6.515151515151516,
      "grad_norm": 0.3634374777293854,
      "learning_rate": 2.9338231883378365e-07,
      "loss": 0.1526,
      "step": 215
    },
    {
      "epoch": 6.545454545454545,
      "grad_norm": 0.3447428712429749,
      "learning_rate": 2.5800917326521013e-07,
      "loss": 0.1376,
      "step": 216
    },
    {
      "epoch": 6.575757575757576,
      "grad_norm": 0.32300415105179925,
      "learning_rate": 2.248799015327907e-07,
      "loss": 0.1438,
      "step": 217
    },
    {
      "epoch": 6.606060606060606,
      "grad_norm": 0.37766546336053336,
      "learning_rate": 1.9400213430538773e-07,
      "loss": 0.1377,
      "step": 218
    },
    {
      "epoch": 6.636363636363637,
      "grad_norm": 0.3711273450161272,
      "learning_rate": 1.6538298366257975e-07,
      "loss": 0.1427,
      "step": 219
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.3435986132162903,
      "learning_rate": 1.3902904145653094e-07,
      "loss": 0.1625,
      "step": 220
    },
    {
      "epoch": 6.696969696969697,
      "grad_norm": 0.3706546368143587,
      "learning_rate": 1.1494637779369766e-07,
      "loss": 0.1488,
      "step": 221
    },
    {
      "epoch": 6.7272727272727275,
      "grad_norm": 0.36327877692558397,
      "learning_rate": 9.314053963669245e-08,
      "loss": 0.1482,
      "step": 222
    },
    {
      "epoch": 6.757575757575758,
      "grad_norm": 0.35086010775673443,
      "learning_rate": 7.361654952665608e-08,
      "loss": 0.1475,
      "step": 223
    },
    {
      "epoch": 6.787878787878788,
      "grad_norm": 0.34960714594014297,
      "learning_rate": 5.637890442641403e-08,
      "loss": 0.1452,
      "step": 224
    },
    {
      "epoch": 6.818181818181818,
      "grad_norm": 0.33896142149612296,
      "learning_rate": 4.143157468468717e-08,
      "loss": 0.159,
      "step": 225
    },
    {
      "epoch": 6.848484848484849,
      "grad_norm": 0.368104373375551,
      "learning_rate": 2.8778003121607834e-08,
      "loss": 0.1412,
      "step": 226
    },
    {
      "epoch": 6.878787878787879,
      "grad_norm": 0.3548934214749664,
      "learning_rate": 1.8421104235727406e-08,
      "loss": 0.1603,
      "step": 227
    },
    {
      "epoch": 6.909090909090909,
      "grad_norm": 0.36193246085450276,
      "learning_rate": 1.0363263532724433e-08,
      "loss": 0.1396,
      "step": 228
    },
    {
      "epoch": 6.9393939393939394,
      "grad_norm": 0.37277919425792433,
      "learning_rate": 4.606336975948589e-09,
      "loss": 0.1491,
      "step": 229
    },
    {
      "epoch": 6.96969696969697,
      "grad_norm": 0.35083940737519065,
      "learning_rate": 1.1516505589381777e-09,
      "loss": 0.1509,
      "step": 230
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.3804132675555632,
      "learning_rate": 0.0,
      "loss": 0.1549,
      "step": 231
    },
    {
      "epoch": 7.0,
      "step": 231,
      "total_flos": 3.564705674208215e+17,
      "train_loss": 0.3959881691705613,
      "train_runtime": 6616.9312,
      "train_samples_per_second": 3.343,
      "train_steps_per_second": 0.035
    }
  ],
  "logging_steps": 1,
  "max_steps": 231,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.564705674208215e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}