{
  "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": 6.517928735676965,
      "learning_rate": 8.333333333333333e-07,
      "loss": 1.0455,
      "step": 1
    },
    {
      "epoch": 0.06060606060606061,
      "grad_norm": 6.5772655527337225,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.0538,
      "step": 2
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 6.234913448864819,
      "learning_rate": 2.5e-06,
      "loss": 1.0183,
      "step": 3
    },
    {
      "epoch": 0.12121212121212122,
      "grad_norm": 6.028555631511048,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.0403,
      "step": 4
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 4.614248102967521,
      "learning_rate": 4.166666666666667e-06,
      "loss": 1.0087,
      "step": 5
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 2.8093652372056996,
      "learning_rate": 5e-06,
      "loss": 0.9609,
      "step": 6
    },
    {
      "epoch": 0.21212121212121213,
      "grad_norm": 2.531229757877748,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.9351,
      "step": 7
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 3.954820632458033,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.9216,
      "step": 8
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 4.349233683778192,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.9455,
      "step": 9
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 3.9434657814129603,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.9028,
      "step": 10
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 3.319039548252818,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.8936,
      "step": 11
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 2.9062149806353275,
      "learning_rate": 1e-05,
      "loss": 0.8668,
      "step": 12
    },
    {
      "epoch": 0.3939393939393939,
      "grad_norm": 2.1578300199262115,
      "learning_rate": 1.0833333333333334e-05,
      "loss": 0.8486,
      "step": 13
    },
    {
      "epoch": 0.42424242424242425,
      "grad_norm": 1.3909657883291664,
      "learning_rate": 1.1666666666666668e-05,
      "loss": 0.8195,
      "step": 14
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 1.524438571353993,
      "learning_rate": 1.25e-05,
      "loss": 0.7993,
      "step": 15
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 1.3007009517951573,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.7693,
      "step": 16
    },
    {
      "epoch": 0.5151515151515151,
      "grad_norm": 1.1840573751827055,
      "learning_rate": 1.416666666666667e-05,
      "loss": 0.7884,
      "step": 17
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 1.072242504675073,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.7686,
      "step": 18
    },
    {
      "epoch": 0.5757575757575758,
      "grad_norm": 1.0713373045598065,
      "learning_rate": 1.5833333333333333e-05,
      "loss": 0.7386,
      "step": 19
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 1.0077579984518008,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.759,
      "step": 20
    },
    {
      "epoch": 0.6363636363636364,
      "grad_norm": 0.9558690379004416,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 0.7453,
      "step": 21
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.9329162555700646,
      "learning_rate": 1.8333333333333333e-05,
      "loss": 0.7493,
      "step": 22
    },
    {
      "epoch": 0.696969696969697,
      "grad_norm": 0.9683964184874808,
      "learning_rate": 1.916666666666667e-05,
      "loss": 0.7069,
      "step": 23
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.8805496812777832,
      "learning_rate": 2e-05,
      "loss": 0.7274,
      "step": 24
    },
    {
      "epoch": 0.7575757575757576,
      "grad_norm": 1.1577430567000497,
      "learning_rate": 1.999884834944106e-05,
      "loss": 0.7113,
      "step": 25
    },
    {
      "epoch": 0.7878787878787878,
      "grad_norm": 0.820818250052884,
      "learning_rate": 1.9995393663024054e-05,
      "loss": 0.7339,
      "step": 26
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 1.026572847248684,
      "learning_rate": 1.9989636736467278e-05,
      "loss": 0.7404,
      "step": 27
    },
    {
      "epoch": 0.8484848484848485,
      "grad_norm": 0.8949030890717823,
      "learning_rate": 1.9981578895764272e-05,
      "loss": 0.7325,
      "step": 28
    },
    {
      "epoch": 0.8787878787878788,
      "grad_norm": 0.6612521991913959,
      "learning_rate": 1.9971221996878395e-05,
      "loss": 0.698,
      "step": 29
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.893505399104731,
      "learning_rate": 1.9958568425315316e-05,
      "loss": 0.7074,
      "step": 30
    },
    {
      "epoch": 0.9393939393939394,
      "grad_norm": 0.8127090262942633,
      "learning_rate": 1.9943621095573588e-05,
      "loss": 0.6983,
      "step": 31
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.6426688453837587,
      "learning_rate": 1.9926383450473344e-05,
      "loss": 0.7091,
      "step": 32
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.7643160321623741,
      "learning_rate": 1.9906859460363307e-05,
      "loss": 0.7099,
      "step": 33
    },
    {
      "epoch": 1.0303030303030303,
      "grad_norm": 0.7663374263559303,
      "learning_rate": 1.9885053622206305e-05,
      "loss": 0.6775,
      "step": 34
    },
    {
      "epoch": 1.0606060606060606,
      "grad_norm": 0.5721050329168993,
      "learning_rate": 1.986097095854347e-05,
      "loss": 0.6599,
      "step": 35
    },
    {
      "epoch": 1.0909090909090908,
      "grad_norm": 0.6381280372319813,
      "learning_rate": 1.9834617016337424e-05,
      "loss": 0.6543,
      "step": 36
    },
    {
      "epoch": 1.121212121212121,
      "grad_norm": 0.589157622166952,
      "learning_rate": 1.9805997865694616e-05,
      "loss": 0.6522,
      "step": 37
    },
    {
      "epoch": 1.1515151515151516,
      "grad_norm": 0.4955649866020235,
      "learning_rate": 1.9775120098467212e-05,
      "loss": 0.6342,
      "step": 38
    },
    {
      "epoch": 1.1818181818181819,
      "grad_norm": 0.6041664505696033,
      "learning_rate": 1.9741990826734793e-05,
      "loss": 0.6499,
      "step": 39
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 0.5417305039931442,
      "learning_rate": 1.970661768116622e-05,
      "loss": 0.6368,
      "step": 40
    },
    {
      "epoch": 1.2424242424242424,
      "grad_norm": 0.5479708269864506,
      "learning_rate": 1.9669008809262064e-05,
      "loss": 0.6359,
      "step": 41
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 0.5822279252934454,
      "learning_rate": 1.9629172873477995e-05,
      "loss": 0.6415,
      "step": 42
    },
    {
      "epoch": 1.303030303030303,
      "grad_norm": 0.6394116074964873,
      "learning_rate": 1.9587119049229558e-05,
      "loss": 0.642,
      "step": 43
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.5849806700810495,
      "learning_rate": 1.954285702277879e-05,
      "loss": 0.6355,
      "step": 44
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.5747917658467485,
      "learning_rate": 1.9496396989003195e-05,
      "loss": 0.6293,
      "step": 45
    },
    {
      "epoch": 1.393939393939394,
      "grad_norm": 0.6951246814835008,
      "learning_rate": 1.944774964904754e-05,
      "loss": 0.634,
      "step": 46
    },
    {
      "epoch": 1.4242424242424243,
      "grad_norm": 0.49648204546835345,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 0.6254,
      "step": 47
    },
    {
      "epoch": 1.4545454545454546,
      "grad_norm": 0.690453417629691,
      "learning_rate": 1.9343938371606714e-05,
      "loss": 0.6221,
      "step": 48
    },
    {
      "epoch": 1.4848484848484849,
      "grad_norm": 0.5951412266996973,
      "learning_rate": 1.9288798344984673e-05,
      "loss": 0.6386,
      "step": 49
    },
    {
      "epoch": 1.5151515151515151,
      "grad_norm": 0.6617890971401159,
      "learning_rate": 1.9231518828401458e-05,
      "loss": 0.629,
      "step": 50
    },
    {
      "epoch": 1.5454545454545454,
      "grad_norm": 0.5430380521759081,
      "learning_rate": 1.917211301505453e-05,
      "loss": 0.6266,
      "step": 51
    },
    {
      "epoch": 1.5757575757575757,
      "grad_norm": 0.6603764327836058,
      "learning_rate": 1.911059458789152e-05,
      "loss": 0.6237,
      "step": 52
    },
    {
      "epoch": 1.606060606060606,
      "grad_norm": 0.5418638187576439,
      "learning_rate": 1.9046977716458627e-05,
      "loss": 0.6219,
      "step": 53
    },
    {
      "epoch": 1.6363636363636362,
      "grad_norm": 0.5945041110356534,
      "learning_rate": 1.8981277053636963e-05,
      "loss": 0.6221,
      "step": 54
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.5690684623748405,
      "learning_rate": 1.891350773226754e-05,
      "loss": 0.628,
      "step": 55
    },
    {
      "epoch": 1.696969696969697,
      "grad_norm": 0.6631928877862863,
      "learning_rate": 1.8843685361665724e-05,
      "loss": 0.6181,
      "step": 56
    },
    {
      "epoch": 1.7272727272727273,
      "grad_norm": 0.6689566033883234,
      "learning_rate": 1.8771826024025944e-05,
      "loss": 0.6269,
      "step": 57
    },
    {
      "epoch": 1.7575757575757576,
      "grad_norm": 0.7066091560013505,
      "learning_rate": 1.8697946270717468e-05,
      "loss": 0.6265,
      "step": 58
    },
    {
      "epoch": 1.7878787878787878,
      "grad_norm": 0.5811635487093934,
      "learning_rate": 1.8622063118472135e-05,
      "loss": 0.6516,
      "step": 59
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.7488418345694374,
      "learning_rate": 1.8544194045464888e-05,
      "loss": 0.6098,
      "step": 60
    },
    {
      "epoch": 1.8484848484848486,
      "grad_norm": 0.5457353046046373,
      "learning_rate": 1.8464356987288012e-05,
      "loss": 0.6041,
      "step": 61
    },
    {
      "epoch": 1.878787878787879,
      "grad_norm": 0.6736333769853714,
      "learning_rate": 1.8382570332820045e-05,
      "loss": 0.6145,
      "step": 62
    },
    {
      "epoch": 1.9090909090909092,
      "grad_norm": 0.6469957663674077,
      "learning_rate": 1.8298852919990254e-05,
      "loss": 0.6163,
      "step": 63
    },
    {
      "epoch": 1.9393939393939394,
      "grad_norm": 0.6803830814584926,
      "learning_rate": 1.821322403143969e-05,
      "loss": 0.6242,
      "step": 64
    },
    {
      "epoch": 1.9696969696969697,
      "grad_norm": 0.5663508233607352,
      "learning_rate": 1.812570339007983e-05,
      "loss": 0.6178,
      "step": 65
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.551652751302577,
      "learning_rate": 1.8036311154549783e-05,
      "loss": 0.6288,
      "step": 66
    },
    {
      "epoch": 2.0303030303030303,
      "grad_norm": 0.7771647721010193,
      "learning_rate": 1.7945067914573147e-05,
      "loss": 0.5725,
      "step": 67
    },
    {
      "epoch": 2.0606060606060606,
      "grad_norm": 0.564899996509134,
      "learning_rate": 1.7851994686215592e-05,
      "loss": 0.5746,
      "step": 68
    },
    {
      "epoch": 2.090909090909091,
      "grad_norm": 0.618249122036095,
      "learning_rate": 1.77571129070442e-05,
      "loss": 0.5565,
      "step": 69
    },
    {
      "epoch": 2.121212121212121,
      "grad_norm": 0.5944500659793406,
      "learning_rate": 1.766044443118978e-05,
      "loss": 0.5548,
      "step": 70
    },
    {
      "epoch": 2.1515151515151514,
      "grad_norm": 0.6128159952824492,
      "learning_rate": 1.7562011524313187e-05,
      "loss": 0.561,
      "step": 71
    },
    {
      "epoch": 2.1818181818181817,
      "grad_norm": 0.5327494294178395,
      "learning_rate": 1.7461836858476858e-05,
      "loss": 0.564,
      "step": 72
    },
    {
      "epoch": 2.212121212121212,
      "grad_norm": 0.6082360705069719,
      "learning_rate": 1.7359943506922775e-05,
      "loss": 0.564,
      "step": 73
    },
    {
      "epoch": 2.242424242424242,
      "grad_norm": 0.5671302488950932,
      "learning_rate": 1.725635493875799e-05,
      "loss": 0.5297,
      "step": 74
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.5915509386240322,
      "learning_rate": 1.7151095013548996e-05,
      "loss": 0.551,
      "step": 75
    },
    {
      "epoch": 2.303030303030303,
      "grad_norm": 0.6660479034674898,
      "learning_rate": 1.7044187975826126e-05,
      "loss": 0.5455,
      "step": 76
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.5266660301477659,
      "learning_rate": 1.693565844949933e-05,
      "loss": 0.5364,
      "step": 77
    },
    {
      "epoch": 2.3636363636363638,
      "grad_norm": 0.6247338879059353,
      "learning_rate": 1.6825531432186545e-05,
      "loss": 0.5404,
      "step": 78
    },
    {
      "epoch": 2.393939393939394,
      "grad_norm": 0.6222629637099215,
      "learning_rate": 1.671383228945597e-05,
      "loss": 0.5349,
      "step": 79
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 0.6189555444712931,
      "learning_rate": 1.6600586748983642e-05,
      "loss": 0.5378,
      "step": 80
    },
    {
      "epoch": 2.4545454545454546,
      "grad_norm": 0.802497560917399,
      "learning_rate": 1.648582089462756e-05,
      "loss": 0.5573,
      "step": 81
    },
    {
      "epoch": 2.484848484848485,
      "grad_norm": 0.6793514991600532,
      "learning_rate": 1.6369561160419783e-05,
      "loss": 0.5474,
      "step": 82
    },
    {
      "epoch": 2.515151515151515,
      "grad_norm": 0.6645605435800442,
      "learning_rate": 1.625183432447789e-05,
      "loss": 0.548,
      "step": 83
    },
    {
      "epoch": 2.5454545454545454,
      "grad_norm": 0.697403221351068,
      "learning_rate": 1.6132667502837164e-05,
      "loss": 0.5573,
      "step": 84
    },
    {
      "epoch": 2.5757575757575757,
      "grad_norm": 0.6119028025068409,
      "learning_rate": 1.6012088143204953e-05,
      "loss": 0.5383,
      "step": 85
    },
    {
      "epoch": 2.606060606060606,
      "grad_norm": 0.5185634339277824,
      "learning_rate": 1.589012401863864e-05,
      "loss": 0.5431,
      "step": 86
    },
    {
      "epoch": 2.6363636363636362,
      "grad_norm": 0.6140765208748267,
      "learning_rate": 1.5766803221148676e-05,
      "loss": 0.5516,
      "step": 87
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.4340062681908139,
      "learning_rate": 1.5642154155228124e-05,
      "loss": 0.5284,
      "step": 88
    },
    {
      "epoch": 2.6969696969696972,
      "grad_norm": 0.5424528614752664,
      "learning_rate": 1.5516205531310272e-05,
      "loss": 0.5537,
      "step": 89
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.5595676031125654,
      "learning_rate": 1.538898635915576e-05,
      "loss": 0.5486,
      "step": 90
    },
    {
      "epoch": 2.757575757575758,
      "grad_norm": 0.49369623901247756,
      "learning_rate": 1.526052594117071e-05,
      "loss": 0.5453,
      "step": 91
    },
    {
      "epoch": 2.787878787878788,
      "grad_norm": 0.5697187256144819,
      "learning_rate": 1.513085386565758e-05,
      "loss": 0.5546,
      "step": 92
    },
    {
      "epoch": 2.8181818181818183,
      "grad_norm": 0.4382602815757698,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.5382,
      "step": 93
    },
    {
      "epoch": 2.8484848484848486,
      "grad_norm": 0.725931420202846,
      "learning_rate": 1.4867994483783485e-05,
      "loss": 0.5484,
      "step": 94
    },
    {
      "epoch": 2.878787878787879,
      "grad_norm": 0.5526464700652515,
      "learning_rate": 1.4734867721853341e-05,
      "loss": 0.5417,
      "step": 95
    },
    {
      "epoch": 2.909090909090909,
      "grad_norm": 0.5197731640521036,
      "learning_rate": 1.4600650377311523e-05,
      "loss": 0.55,
      "step": 96
    },
    {
      "epoch": 2.9393939393939394,
      "grad_norm": 0.5682826825760908,
      "learning_rate": 1.4465373364454001e-05,
      "loss": 0.5603,
      "step": 97
    },
    {
      "epoch": 2.9696969696969697,
      "grad_norm": 0.466275094156658,
      "learning_rate": 1.4329067841650274e-05,
      "loss": 0.537,
      "step": 98
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.6352363204388607,
      "learning_rate": 1.4191765204166643e-05,
      "loss": 0.545,
      "step": 99
    },
    {
      "epoch": 3.0303030303030303,
      "grad_norm": 0.5776585338313645,
      "learning_rate": 1.4053497076934948e-05,
      "loss": 0.4968,
      "step": 100
    },
    {
      "epoch": 3.0606060606060606,
      "grad_norm": 0.4771059248611455,
      "learning_rate": 1.3914295307268396e-05,
      "loss": 0.4974,
      "step": 101
    },
    {
      "epoch": 3.090909090909091,
      "grad_norm": 0.5605350362601154,
      "learning_rate": 1.3774191957526144e-05,
      "loss": 0.4824,
      "step": 102
    },
    {
      "epoch": 3.121212121212121,
      "grad_norm": 0.48093206214966144,
      "learning_rate": 1.3633219297728415e-05,
      "loss": 0.4825,
      "step": 103
    },
    {
      "epoch": 3.1515151515151514,
      "grad_norm": 0.5695727515540727,
      "learning_rate": 1.3491409798123687e-05,
      "loss": 0.4787,
      "step": 104
    },
    {
      "epoch": 3.1818181818181817,
      "grad_norm": 0.552164375832746,
      "learning_rate": 1.3348796121709862e-05,
      "loss": 0.4878,
      "step": 105
    },
    {
      "epoch": 3.212121212121212,
      "grad_norm": 0.5159675769861567,
      "learning_rate": 1.3205411116710973e-05,
      "loss": 0.4722,
      "step": 106
    },
    {
      "epoch": 3.242424242424242,
      "grad_norm": 0.46240380219363764,
      "learning_rate": 1.3061287809011243e-05,
      "loss": 0.4791,
      "step": 107
    },
    {
      "epoch": 3.2727272727272725,
      "grad_norm": 0.5231843366499154,
      "learning_rate": 1.291645939454825e-05,
      "loss": 0.4789,
      "step": 108
    },
    {
      "epoch": 3.303030303030303,
      "grad_norm": 0.43961723636777644,
      "learning_rate": 1.277095923166689e-05,
      "loss": 0.478,
      "step": 109
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.4721485286866721,
      "learning_rate": 1.2624820833435939e-05,
      "loss": 0.473,
      "step": 110
    },
    {
      "epoch": 3.3636363636363638,
      "grad_norm": 0.5443639503195921,
      "learning_rate": 1.2478077859929e-05,
      "loss": 0.4672,
      "step": 111
    },
    {
      "epoch": 3.393939393939394,
      "grad_norm": 0.4662242306518608,
      "learning_rate": 1.2330764110471567e-05,
      "loss": 0.4831,
      "step": 112
    },
    {
      "epoch": 3.4242424242424243,
      "grad_norm": 0.5388879388421665,
      "learning_rate": 1.2182913515856016e-05,
      "loss": 0.4793,
      "step": 113
    },
    {
      "epoch": 3.4545454545454546,
      "grad_norm": 0.4907864746109935,
      "learning_rate": 1.2034560130526341e-05,
      "loss": 0.4806,
      "step": 114
    },
    {
      "epoch": 3.484848484848485,
      "grad_norm": 0.5382786784782878,
      "learning_rate": 1.1885738124734359e-05,
      "loss": 0.4818,
      "step": 115
    },
    {
      "epoch": 3.515151515151515,
      "grad_norm": 0.4325648604370947,
      "learning_rate": 1.1736481776669307e-05,
      "loss": 0.4796,
      "step": 116
    },
    {
      "epoch": 3.5454545454545454,
      "grad_norm": 0.4914914905055827,
      "learning_rate": 1.1586825464562515e-05,
      "loss": 0.4765,
      "step": 117
    },
    {
      "epoch": 3.5757575757575757,
      "grad_norm": 0.44194107125135146,
      "learning_rate": 1.1436803658769082e-05,
      "loss": 0.494,
      "step": 118
    },
    {
      "epoch": 3.606060606060606,
      "grad_norm": 0.4192917180919241,
      "learning_rate": 1.1286450913828313e-05,
      "loss": 0.4917,
      "step": 119
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.4610644445088454,
      "learning_rate": 1.113580186050475e-05,
      "loss": 0.4814,
      "step": 120
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.3902931908900223,
      "learning_rate": 1.0984891197811686e-05,
      "loss": 0.4721,
      "step": 121
    },
    {
      "epoch": 3.6969696969696972,
      "grad_norm": 0.4140702518302563,
      "learning_rate": 1.0833753685018935e-05,
      "loss": 0.4732,
      "step": 122
    },
    {
      "epoch": 3.7272727272727275,
      "grad_norm": 0.4101357605462889,
      "learning_rate": 1.0682424133646712e-05,
      "loss": 0.4734,
      "step": 123
    },
    {
      "epoch": 3.757575757575758,
      "grad_norm": 0.44911020645751626,
      "learning_rate": 1.0530937399447496e-05,
      "loss": 0.4874,
      "step": 124
    },
    {
      "epoch": 3.787878787878788,
      "grad_norm": 0.3818119445654494,
      "learning_rate": 1.0379328374377715e-05,
      "loss": 0.4815,
      "step": 125
    },
    {
      "epoch": 3.8181818181818183,
      "grad_norm": 0.5580474214383856,
      "learning_rate": 1.0227631978561057e-05,
      "loss": 0.4756,
      "step": 126
    },
    {
      "epoch": 3.8484848484848486,
      "grad_norm": 0.4116446618071981,
      "learning_rate": 1.0075883152245334e-05,
      "loss": 0.4741,
      "step": 127
    },
    {
      "epoch": 3.878787878787879,
      "grad_norm": 0.4943120264299029,
      "learning_rate": 9.92411684775467e-06,
      "loss": 0.4875,
      "step": 128
    },
    {
      "epoch": 3.909090909090909,
      "grad_norm": 0.4180939845391508,
      "learning_rate": 9.772368021438943e-06,
      "loss": 0.4724,
      "step": 129
    },
    {
      "epoch": 3.9393939393939394,
      "grad_norm": 0.4769942870235733,
      "learning_rate": 9.620671625622287e-06,
      "loss": 0.4768,
      "step": 130
    },
    {
      "epoch": 3.9696969696969697,
      "grad_norm": 0.453817923016501,
      "learning_rate": 9.469062600552509e-06,
      "loss": 0.4714,
      "step": 131
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.39927065617614704,
      "learning_rate": 9.317575866353293e-06,
      "loss": 0.4713,
      "step": 132
    },
    {
      "epoch": 4.03030303030303,
      "grad_norm": 0.5987689057987375,
      "learning_rate": 9.166246314981066e-06,
      "loss": 0.4255,
      "step": 133
    },
    {
      "epoch": 4.0606060606060606,
      "grad_norm": 0.4790300942770665,
      "learning_rate": 9.015108802188314e-06,
      "loss": 0.4365,
      "step": 134
    },
    {
      "epoch": 4.090909090909091,
      "grad_norm": 0.48980998436066,
      "learning_rate": 8.86419813949525e-06,
      "loss": 0.418,
      "step": 135
    },
    {
      "epoch": 4.121212121212121,
      "grad_norm": 0.599569588363402,
      "learning_rate": 8.71354908617169e-06,
      "loss": 0.4052,
      "step": 136
    },
    {
      "epoch": 4.151515151515151,
      "grad_norm": 0.5585790964609069,
      "learning_rate": 8.56319634123092e-06,
      "loss": 0.4292,
      "step": 137
    },
    {
      "epoch": 4.181818181818182,
      "grad_norm": 0.5102739750207342,
      "learning_rate": 8.413174535437486e-06,
      "loss": 0.4173,
      "step": 138
    },
    {
      "epoch": 4.212121212121212,
      "grad_norm": 0.5175867828824035,
      "learning_rate": 8.263518223330698e-06,
      "loss": 0.4198,
      "step": 139
    },
    {
      "epoch": 4.242424242424242,
      "grad_norm": 0.5196144980105121,
      "learning_rate": 8.114261875265643e-06,
      "loss": 0.4049,
      "step": 140
    },
    {
      "epoch": 4.2727272727272725,
      "grad_norm": 0.4870940451465083,
      "learning_rate": 7.965439869473664e-06,
      "loss": 0.4244,
      "step": 141
    },
    {
      "epoch": 4.303030303030303,
      "grad_norm": 0.46255414067156897,
      "learning_rate": 7.817086484143987e-06,
      "loss": 0.4143,
      "step": 142
    },
    {
      "epoch": 4.333333333333333,
      "grad_norm": 0.5286334558089502,
      "learning_rate": 7.669235889528436e-06,
      "loss": 0.4216,
      "step": 143
    },
    {
      "epoch": 4.363636363636363,
      "grad_norm": 0.462928845467462,
      "learning_rate": 7.521922140071003e-06,
      "loss": 0.409,
      "step": 144
    },
    {
      "epoch": 4.393939393939394,
      "grad_norm": 0.42566600796425524,
      "learning_rate": 7.375179166564062e-06,
      "loss": 0.4159,
      "step": 145
    },
    {
      "epoch": 4.424242424242424,
      "grad_norm": 0.4171665568759143,
      "learning_rate": 7.2290407683331154e-06,
      "loss": 0.4256,
      "step": 146
    },
    {
      "epoch": 4.454545454545454,
      "grad_norm": 0.46595276831246435,
      "learning_rate": 7.0835406054517505e-06,
      "loss": 0.4163,
      "step": 147
    },
    {
      "epoch": 4.484848484848484,
      "grad_norm": 0.3851926386621285,
      "learning_rate": 6.93871219098876e-06,
      "loss": 0.4221,
      "step": 148
    },
    {
      "epoch": 4.515151515151516,
      "grad_norm": 0.3900923201698187,
      "learning_rate": 6.79458888328903e-06,
      "loss": 0.4244,
      "step": 149
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 0.4034534199650892,
      "learning_rate": 6.651203878290139e-06,
      "loss": 0.4215,
      "step": 150
    },
    {
      "epoch": 4.575757575757576,
      "grad_norm": 0.3799715262816614,
      "learning_rate": 6.508590201876317e-06,
      "loss": 0.4051,
      "step": 151
    },
    {
      "epoch": 4.606060606060606,
      "grad_norm": 0.371365785193263,
      "learning_rate": 6.366780702271589e-06,
      "loss": 0.4069,
      "step": 152
    },
    {
      "epoch": 4.636363636363637,
      "grad_norm": 0.3706688070478696,
      "learning_rate": 6.225808042473857e-06,
      "loss": 0.4348,
      "step": 153
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.3667995438450157,
      "learning_rate": 6.085704692731609e-06,
      "loss": 0.4225,
      "step": 154
    },
    {
      "epoch": 4.696969696969697,
      "grad_norm": 0.394320750379463,
      "learning_rate": 5.946502923065054e-06,
      "loss": 0.4085,
      "step": 155
    },
    {
      "epoch": 4.7272727272727275,
      "grad_norm": 0.3476113946267737,
      "learning_rate": 5.8082347958333625e-06,
      "loss": 0.4122,
      "step": 156
    },
    {
      "epoch": 4.757575757575758,
      "grad_norm": 0.384396711169214,
      "learning_rate": 5.670932158349732e-06,
      "loss": 0.4227,
      "step": 157
    },
    {
      "epoch": 4.787878787878788,
      "grad_norm": 0.3792106407811517,
      "learning_rate": 5.534626635546e-06,
      "loss": 0.4203,
      "step": 158
    },
    {
      "epoch": 4.818181818181818,
      "grad_norm": 0.3556912896979165,
      "learning_rate": 5.399349622688479e-06,
      "loss": 0.4176,
      "step": 159
    },
    {
      "epoch": 4.848484848484849,
      "grad_norm": 0.35611015173815613,
      "learning_rate": 5.2651322781466606e-06,
      "loss": 0.4161,
      "step": 160
    },
    {
      "epoch": 4.878787878787879,
      "grad_norm": 0.3747879122931635,
      "learning_rate": 5.132005516216512e-06,
      "loss": 0.4154,
      "step": 161
    },
    {
      "epoch": 4.909090909090909,
      "grad_norm": 0.34767749347685095,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.4169,
      "step": 162
    },
    {
      "epoch": 4.9393939393939394,
      "grad_norm": 0.32095751420570984,
      "learning_rate": 4.869146134342426e-06,
      "loss": 0.4209,
      "step": 163
    },
    {
      "epoch": 4.96969696969697,
      "grad_norm": 0.34456110661136463,
      "learning_rate": 4.739474058829288e-06,
      "loss": 0.419,
      "step": 164
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.353567015392029,
      "learning_rate": 4.611013640844245e-06,
      "loss": 0.4143,
      "step": 165
    },
    {
      "epoch": 5.03030303030303,
      "grad_norm": 0.5382318520401967,
      "learning_rate": 4.483794468689728e-06,
      "loss": 0.382,
      "step": 166
    },
    {
      "epoch": 5.0606060606060606,
      "grad_norm": 0.39983247500309826,
      "learning_rate": 4.357845844771881e-06,
      "loss": 0.3721,
      "step": 167
    },
    {
      "epoch": 5.090909090909091,
      "grad_norm": 0.3475018833963907,
      "learning_rate": 4.2331967788513295e-06,
      "loss": 0.3705,
      "step": 168
    },
    {
      "epoch": 5.121212121212121,
      "grad_norm": 0.5189815394990832,
      "learning_rate": 4.109875981361363e-06,
      "loss": 0.3817,
      "step": 169
    },
    {
      "epoch": 5.151515151515151,
      "grad_norm": 0.554522540829979,
      "learning_rate": 3.987911856795047e-06,
      "loss": 0.3864,
      "step": 170
    },
    {
      "epoch": 5.181818181818182,
      "grad_norm": 0.41688108911198396,
      "learning_rate": 3.867332497162836e-06,
      "loss": 0.3863,
      "step": 171
    },
    {
      "epoch": 5.212121212121212,
      "grad_norm": 0.4027653897797602,
      "learning_rate": 3.748165675522113e-06,
      "loss": 0.3716,
      "step": 172
    },
    {
      "epoch": 5.242424242424242,
      "grad_norm": 0.41406725572850883,
      "learning_rate": 3.630438839580217e-06,
      "loss": 0.366,
      "step": 173
    },
    {
      "epoch": 5.2727272727272725,
      "grad_norm": 0.4079762390454095,
      "learning_rate": 3.5141791053724405e-06,
      "loss": 0.362,
      "step": 174
    },
    {
      "epoch": 5.303030303030303,
      "grad_norm": 0.3766526825355298,
      "learning_rate": 3.399413251016359e-06,
      "loss": 0.3841,
      "step": 175
    },
    {
      "epoch": 5.333333333333333,
      "grad_norm": 0.37011491054019086,
      "learning_rate": 3.2861677105440335e-06,
      "loss": 0.3759,
      "step": 176
    },
    {
      "epoch": 5.363636363636363,
      "grad_norm": 0.3838011792978329,
      "learning_rate": 3.174468567813461e-06,
      "loss": 0.3592,
      "step": 177
    },
    {
      "epoch": 5.393939393939394,
      "grad_norm": 0.39870925113344313,
      "learning_rate": 3.0643415505006733e-06,
      "loss": 0.3725,
      "step": 178
    },
    {
      "epoch": 5.424242424242424,
      "grad_norm": 0.3741463586176945,
      "learning_rate": 2.9558120241738786e-06,
      "loss": 0.377,
      "step": 179
    },
    {
      "epoch": 5.454545454545454,
      "grad_norm": 0.35712527597315996,
      "learning_rate": 2.8489049864510053e-06,
      "loss": 0.3648,
      "step": 180
    },
    {
      "epoch": 5.484848484848484,
      "grad_norm": 0.37516801168343306,
      "learning_rate": 2.7436450612420098e-06,
      "loss": 0.3746,
      "step": 181
    },
    {
      "epoch": 5.515151515151516,
      "grad_norm": 0.33612195811172774,
      "learning_rate": 2.640056493077231e-06,
      "loss": 0.3716,
      "step": 182
    },
    {
      "epoch": 5.545454545454545,
      "grad_norm": 0.36328449670893154,
      "learning_rate": 2.5381631415231455e-06,
      "loss": 0.3817,
      "step": 183
    },
    {
      "epoch": 5.575757575757576,
      "grad_norm": 0.3703073744878245,
      "learning_rate": 2.4379884756868167e-06,
      "loss": 0.3671,
      "step": 184
    },
    {
      "epoch": 5.606060606060606,
      "grad_norm": 0.3427111752725232,
      "learning_rate": 2.339555568810221e-06,
      "loss": 0.3816,
      "step": 185
    },
    {
      "epoch": 5.636363636363637,
      "grad_norm": 0.3496215218548327,
      "learning_rate": 2.2428870929558012e-06,
      "loss": 0.3672,
      "step": 186
    },
    {
      "epoch": 5.666666666666667,
      "grad_norm": 0.329336790888063,
      "learning_rate": 2.1480053137844115e-06,
      "loss": 0.3712,
      "step": 187
    },
    {
      "epoch": 5.696969696969697,
      "grad_norm": 0.3310711965914957,
      "learning_rate": 2.054932085426856e-06,
      "loss": 0.3797,
      "step": 188
    },
    {
      "epoch": 5.7272727272727275,
      "grad_norm": 0.3370368521698809,
      "learning_rate": 1.963688845450218e-06,
      "loss": 0.3582,
      "step": 189
    },
    {
      "epoch": 5.757575757575758,
      "grad_norm": 0.33096560047160006,
      "learning_rate": 1.8742966099201699e-06,
      "loss": 0.3692,
      "step": 190
    },
    {
      "epoch": 5.787878787878788,
      "grad_norm": 0.3010735438457465,
      "learning_rate": 1.7867759685603115e-06,
      "loss": 0.3709,
      "step": 191
    },
    {
      "epoch": 5.818181818181818,
      "grad_norm": 0.3126661813863104,
      "learning_rate": 1.7011470800097496e-06,
      "loss": 0.3823,
      "step": 192
    },
    {
      "epoch": 5.848484848484849,
      "grad_norm": 0.3175598268657734,
      "learning_rate": 1.6174296671799571e-06,
      "loss": 0.3683,
      "step": 193
    },
    {
      "epoch": 5.878787878787879,
      "grad_norm": 0.3236261627206911,
      "learning_rate": 1.5356430127119915e-06,
      "loss": 0.3752,
      "step": 194
    },
    {
      "epoch": 5.909090909090909,
      "grad_norm": 0.3332425708602334,
      "learning_rate": 1.4558059545351144e-06,
      "loss": 0.3658,
      "step": 195
    },
    {
      "epoch": 5.9393939393939394,
      "grad_norm": 0.3152879170071617,
      "learning_rate": 1.3779368815278648e-06,
      "loss": 0.3635,
      "step": 196
    },
    {
      "epoch": 5.96969696969697,
      "grad_norm": 0.30008582946052875,
      "learning_rate": 1.302053729282533e-06,
      "loss": 0.3679,
      "step": 197
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.2977174824295516,
      "learning_rate": 1.2281739759740575e-06,
      "loss": 0.3767,
      "step": 198
    },
    {
      "epoch": 6.03030303030303,
      "grad_norm": 0.3879029800082202,
      "learning_rate": 1.156314638334277e-06,
      "loss": 0.3407,
      "step": 199
    },
    {
      "epoch": 6.0606060606060606,
      "grad_norm": 0.3271155972576399,
      "learning_rate": 1.086492267732462e-06,
      "loss": 0.3591,
      "step": 200
    },
    {
      "epoch": 6.090909090909091,
      "grad_norm": 0.34645313121663107,
      "learning_rate": 1.01872294636304e-06,
      "loss": 0.3483,
      "step": 201
    },
    {
      "epoch": 6.121212121212121,
      "grad_norm": 0.28743262379825385,
      "learning_rate": 9.530222835413739e-07,
      "loss": 0.3508,
      "step": 202
    },
    {
      "epoch": 6.151515151515151,
      "grad_norm": 0.29565754895347895,
      "learning_rate": 8.894054121084839e-07,
      "loss": 0.3446,
      "step": 203
    },
    {
      "epoch": 6.181818181818182,
      "grad_norm": 0.30250529576379726,
      "learning_rate": 8.278869849454718e-07,
      "loss": 0.3476,
      "step": 204
    },
    {
      "epoch": 6.212121212121212,
      "grad_norm": 0.3163060170935283,
      "learning_rate": 7.684811715985429e-07,
      "loss": 0.3543,
      "step": 205
    },
    {
      "epoch": 6.242424242424242,
      "grad_norm": 0.3283811226544294,
      "learning_rate": 7.1120165501533e-07,
      "loss": 0.3575,
      "step": 206
    },
    {
      "epoch": 6.2727272727272725,
      "grad_norm": 0.29942592377487787,
      "learning_rate": 6.560616283932897e-07,
      "loss": 0.3483,
      "step": 207
    },
    {
      "epoch": 6.303030303030303,
      "grad_norm": 0.36862485605365164,
      "learning_rate": 6.030737921409169e-07,
      "loss": 0.3615,
      "step": 208
    },
    {
      "epoch": 6.333333333333333,
      "grad_norm": 0.2949393074036796,
      "learning_rate": 5.522503509524591e-07,
      "loss": 0.3553,
      "step": 209
    },
    {
      "epoch": 6.363636363636363,
      "grad_norm": 0.318252977076404,
      "learning_rate": 5.036030109968082e-07,
      "loss": 0.3525,
      "step": 210
    },
    {
      "epoch": 6.393939393939394,
      "grad_norm": 0.3017808653775596,
      "learning_rate": 4.5714297722121105e-07,
      "loss": 0.3586,
      "step": 211
    },
    {
      "epoch": 6.424242424242424,
      "grad_norm": 0.31274706811092523,
      "learning_rate": 4.128809507704445e-07,
      "loss": 0.371,
      "step": 212
    },
    {
      "epoch": 6.454545454545454,
      "grad_norm": 0.2982929444465025,
      "learning_rate": 3.708271265220087e-07,
      "loss": 0.3369,
      "step": 213
    },
    {
      "epoch": 6.484848484848484,
      "grad_norm": 0.2856931805359614,
      "learning_rate": 3.309911907379393e-07,
      "loss": 0.3441,
      "step": 214
    },
    {
      "epoch": 6.515151515151516,
      "grad_norm": 0.295130199159678,
      "learning_rate": 2.9338231883378365e-07,
      "loss": 0.343,
      "step": 215
    },
    {
      "epoch": 6.545454545454545,
      "grad_norm": 0.30638944417414304,
      "learning_rate": 2.5800917326521013e-07,
      "loss": 0.3446,
      "step": 216
    },
    {
      "epoch": 6.575757575757576,
      "grad_norm": 0.2957235766685017,
      "learning_rate": 2.248799015327907e-07,
      "loss": 0.3559,
      "step": 217
    },
    {
      "epoch": 6.606060606060606,
      "grad_norm": 0.303702542574936,
      "learning_rate": 1.9400213430538773e-07,
      "loss": 0.347,
      "step": 218
    },
    {
      "epoch": 6.636363636363637,
      "grad_norm": 0.2851016218651606,
      "learning_rate": 1.6538298366257975e-07,
      "loss": 0.3496,
      "step": 219
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.3034112904525307,
      "learning_rate": 1.3902904145653094e-07,
      "loss": 0.3466,
      "step": 220
    },
    {
      "epoch": 6.696969696969697,
      "grad_norm": 0.29969234719385984,
      "learning_rate": 1.1494637779369766e-07,
      "loss": 0.3533,
      "step": 221
    },
    {
      "epoch": 6.7272727272727275,
      "grad_norm": 0.2907347327578063,
      "learning_rate": 9.314053963669245e-08,
      "loss": 0.3477,
      "step": 222
    },
    {
      "epoch": 6.757575757575758,
      "grad_norm": 0.27725605270722475,
      "learning_rate": 7.361654952665608e-08,
      "loss": 0.3512,
      "step": 223
    },
    {
      "epoch": 6.787878787878788,
      "grad_norm": 0.2983284815250133,
      "learning_rate": 5.637890442641403e-08,
      "loss": 0.3547,
      "step": 224
    },
    {
      "epoch": 6.818181818181818,
      "grad_norm": 0.29664319153122787,
      "learning_rate": 4.143157468468717e-08,
      "loss": 0.3648,
      "step": 225
    },
    {
      "epoch": 6.848484848484849,
      "grad_norm": 0.291300393043478,
      "learning_rate": 2.8778003121607834e-08,
      "loss": 0.354,
      "step": 226
    },
    {
      "epoch": 6.878787878787879,
      "grad_norm": 0.28479952634254985,
      "learning_rate": 1.8421104235727406e-08,
      "loss": 0.3475,
      "step": 227
    },
    {
      "epoch": 6.909090909090909,
      "grad_norm": 0.30549072336004507,
      "learning_rate": 1.0363263532724433e-08,
      "loss": 0.3526,
      "step": 228
    },
    {
      "epoch": 6.9393939393939394,
      "grad_norm": 0.28929693249167693,
      "learning_rate": 4.606336975948589e-09,
      "loss": 0.3469,
      "step": 229
    },
    {
      "epoch": 6.96969696969697,
      "grad_norm": 0.2867666386627962,
      "learning_rate": 1.1516505589381777e-09,
      "loss": 0.3435,
      "step": 230
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.30195503222447084,
      "learning_rate": 0.0,
      "loss": 0.3469,
      "step": 231
    },
    {
      "epoch": 7.0,
      "step": 231,
      "total_flos": 6.103710830103101e+17,
      "train_loss": 0.5184378920695483,
      "train_runtime": 8807.9751,
      "train_samples_per_second": 2.511,
      "train_steps_per_second": 0.026
    }
  ],
  "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": 6.103710830103101e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}