{
  "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": 6.178154629057961,
      "learning_rate": 8.333333333333333e-07,
      "loss": 1.007,
      "step": 1
    },
    {
      "epoch": 0.06060606060606061,
      "grad_norm": 6.28653900665313,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.0303,
      "step": 2
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 6.273816342787373,
      "learning_rate": 2.5e-06,
      "loss": 1.0194,
      "step": 3
    },
    {
      "epoch": 0.12121212121212122,
      "grad_norm": 5.723058206659835,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.9779,
      "step": 4
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 4.598614994348905,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.9602,
      "step": 5
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 2.2211942909025346,
      "learning_rate": 5e-06,
      "loss": 0.8598,
      "step": 6
    },
    {
      "epoch": 0.21212121212121213,
      "grad_norm": 1.8862784210596273,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.8662,
      "step": 7
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 4.061354104270583,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.8482,
      "step": 8
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 4.687508511784006,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.8973,
      "step": 9
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 4.258446848867394,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.8198,
      "step": 10
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 3.704476660938787,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.7998,
      "step": 11
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 3.4032498324984553,
      "learning_rate": 1e-05,
      "loss": 0.8185,
      "step": 12
    },
    {
      "epoch": 0.3939393939393939,
      "grad_norm": 2.703046201871238,
      "learning_rate": 1.0833333333333334e-05,
      "loss": 0.8136,
      "step": 13
    },
    {
      "epoch": 0.42424242424242425,
      "grad_norm": 1.9332005859456105,
      "learning_rate": 1.1666666666666668e-05,
      "loss": 0.7605,
      "step": 14
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 2.3674479304461817,
      "learning_rate": 1.25e-05,
      "loss": 0.8122,
      "step": 15
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 2.273176091382673,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.7722,
      "step": 16
    },
    {
      "epoch": 0.5151515151515151,
      "grad_norm": 1.7848884281692476,
      "learning_rate": 1.416666666666667e-05,
      "loss": 0.7705,
      "step": 17
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 1.401094178105897,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.7355,
      "step": 18
    },
    {
      "epoch": 0.5757575757575758,
      "grad_norm": 1.3798113058163846,
      "learning_rate": 1.5833333333333333e-05,
      "loss": 0.7318,
      "step": 19
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 1.2530455357370824,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.7333,
      "step": 20
    },
    {
      "epoch": 0.6363636363636364,
      "grad_norm": 0.9054872899846163,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 0.6976,
      "step": 21
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 1.187524611111422,
      "learning_rate": 1.8333333333333333e-05,
      "loss": 0.7451,
      "step": 22
    },
    {
      "epoch": 0.696969696969697,
      "grad_norm": 0.9369018366477593,
      "learning_rate": 1.916666666666667e-05,
      "loss": 0.6862,
      "step": 23
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.7350676694753355,
      "learning_rate": 2e-05,
      "loss": 0.6998,
      "step": 24
    },
    {
      "epoch": 0.7575757575757576,
      "grad_norm": 0.9785056866258712,
      "learning_rate": 1.999884834944106e-05,
      "loss": 0.7269,
      "step": 25
    },
    {
      "epoch": 0.7878787878787878,
      "grad_norm": 0.7604738677798567,
      "learning_rate": 1.9995393663024054e-05,
      "loss": 0.7376,
      "step": 26
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 0.7517869786299339,
      "learning_rate": 1.9989636736467278e-05,
      "loss": 0.7032,
      "step": 27
    },
    {
      "epoch": 0.8484848484848485,
      "grad_norm": 0.7564633040817004,
      "learning_rate": 1.9981578895764272e-05,
      "loss": 0.6952,
      "step": 28
    },
    {
      "epoch": 0.8787878787878788,
      "grad_norm": 0.7691210866565935,
      "learning_rate": 1.9971221996878395e-05,
      "loss": 0.6582,
      "step": 29
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.6589400645565682,
      "learning_rate": 1.9958568425315316e-05,
      "loss": 0.7088,
      "step": 30
    },
    {
      "epoch": 0.9393939393939394,
      "grad_norm": 0.7196433078183322,
      "learning_rate": 1.9943621095573588e-05,
      "loss": 0.6307,
      "step": 31
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.6140005584813665,
      "learning_rate": 1.9926383450473344e-05,
      "loss": 0.7026,
      "step": 32
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.6014469079817645,
      "learning_rate": 1.9906859460363307e-05,
      "loss": 0.6698,
      "step": 33
    },
    {
      "epoch": 1.0303030303030303,
      "grad_norm": 0.681940582366705,
      "learning_rate": 1.9885053622206305e-05,
      "loss": 0.6121,
      "step": 34
    },
    {
      "epoch": 1.0606060606060606,
      "grad_norm": 0.6825046208715866,
      "learning_rate": 1.986097095854347e-05,
      "loss": 0.6452,
      "step": 35
    },
    {
      "epoch": 1.0909090909090908,
      "grad_norm": 0.7026444330315502,
      "learning_rate": 1.9834617016337424e-05,
      "loss": 0.6537,
      "step": 36
    },
    {
      "epoch": 1.121212121212121,
      "grad_norm": 0.5889574368007281,
      "learning_rate": 1.9805997865694616e-05,
      "loss": 0.5976,
      "step": 37
    },
    {
      "epoch": 1.1515151515151516,
      "grad_norm": 0.5827221515346869,
      "learning_rate": 1.9775120098467212e-05,
      "loss": 0.5953,
      "step": 38
    },
    {
      "epoch": 1.1818181818181819,
      "grad_norm": 0.6101627998564864,
      "learning_rate": 1.9741990826734793e-05,
      "loss": 0.5921,
      "step": 39
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 0.6600869490187646,
      "learning_rate": 1.970661768116622e-05,
      "loss": 0.6233,
      "step": 40
    },
    {
      "epoch": 1.2424242424242424,
      "grad_norm": 0.5371028471450293,
      "learning_rate": 1.9669008809262064e-05,
      "loss": 0.5892,
      "step": 41
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 0.618174170542928,
      "learning_rate": 1.9629172873477995e-05,
      "loss": 0.6127,
      "step": 42
    },
    {
      "epoch": 1.303030303030303,
      "grad_norm": 0.5824367665975636,
      "learning_rate": 1.9587119049229558e-05,
      "loss": 0.6246,
      "step": 43
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.5542115244990704,
      "learning_rate": 1.954285702277879e-05,
      "loss": 0.5724,
      "step": 44
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.5651329595022573,
      "learning_rate": 1.9496396989003195e-05,
      "loss": 0.5774,
      "step": 45
    },
    {
      "epoch": 1.393939393939394,
      "grad_norm": 0.6072044500339653,
      "learning_rate": 1.944774964904754e-05,
      "loss": 0.5939,
      "step": 46
    },
    {
      "epoch": 1.4242424242424243,
      "grad_norm": 0.5699775230166506,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 0.5796,
      "step": 47
    },
    {
      "epoch": 1.4545454545454546,
      "grad_norm": 0.6593729138569979,
      "learning_rate": 1.9343938371606714e-05,
      "loss": 0.5838,
      "step": 48
    },
    {
      "epoch": 1.4848484848484849,
      "grad_norm": 0.6426480880684551,
      "learning_rate": 1.9288798344984673e-05,
      "loss": 0.5969,
      "step": 49
    },
    {
      "epoch": 1.5151515151515151,
      "grad_norm": 0.49469059731465775,
      "learning_rate": 1.9231518828401458e-05,
      "loss": 0.5597,
      "step": 50
    },
    {
      "epoch": 1.5454545454545454,
      "grad_norm": 0.5522986825042939,
      "learning_rate": 1.917211301505453e-05,
      "loss": 0.567,
      "step": 51
    },
    {
      "epoch": 1.5757575757575757,
      "grad_norm": 0.631149896322368,
      "learning_rate": 1.911059458789152e-05,
      "loss": 0.5767,
      "step": 52
    },
    {
      "epoch": 1.606060606060606,
      "grad_norm": 0.5870062255982978,
      "learning_rate": 1.9046977716458627e-05,
      "loss": 0.6147,
      "step": 53
    },
    {
      "epoch": 1.6363636363636362,
      "grad_norm": 0.6051046286706993,
      "learning_rate": 1.8981277053636963e-05,
      "loss": 0.6031,
      "step": 54
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.5712969865860921,
      "learning_rate": 1.891350773226754e-05,
      "loss": 0.5903,
      "step": 55
    },
    {
      "epoch": 1.696969696969697,
      "grad_norm": 0.5232741331205144,
      "learning_rate": 1.8843685361665724e-05,
      "loss": 0.5769,
      "step": 56
    },
    {
      "epoch": 1.7272727272727273,
      "grad_norm": 0.5572771601054789,
      "learning_rate": 1.8771826024025944e-05,
      "loss": 0.581,
      "step": 57
    },
    {
      "epoch": 1.7575757575757576,
      "grad_norm": 0.5586579609572825,
      "learning_rate": 1.8697946270717468e-05,
      "loss": 0.6141,
      "step": 58
    },
    {
      "epoch": 1.7878787878787878,
      "grad_norm": 0.5911522177220985,
      "learning_rate": 1.8622063118472135e-05,
      "loss": 0.5882,
      "step": 59
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.5618260648959794,
      "learning_rate": 1.8544194045464888e-05,
      "loss": 0.5775,
      "step": 60
    },
    {
      "epoch": 1.8484848484848486,
      "grad_norm": 0.5886311604223228,
      "learning_rate": 1.8464356987288012e-05,
      "loss": 0.6034,
      "step": 61
    },
    {
      "epoch": 1.878787878787879,
      "grad_norm": 0.7357700877855867,
      "learning_rate": 1.8382570332820045e-05,
      "loss": 0.595,
      "step": 62
    },
    {
      "epoch": 1.9090909090909092,
      "grad_norm": 0.5071788667817216,
      "learning_rate": 1.8298852919990254e-05,
      "loss": 0.5876,
      "step": 63
    },
    {
      "epoch": 1.9393939393939394,
      "grad_norm": 0.651401481412067,
      "learning_rate": 1.821322403143969e-05,
      "loss": 0.5942,
      "step": 64
    },
    {
      "epoch": 1.9696969696969697,
      "grad_norm": 0.5279812650467501,
      "learning_rate": 1.812570339007983e-05,
      "loss": 0.5628,
      "step": 65
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.5441424656717774,
      "learning_rate": 1.8036311154549783e-05,
      "loss": 0.5624,
      "step": 66
    },
    {
      "epoch": 2.0303030303030303,
      "grad_norm": 0.7193542703628895,
      "learning_rate": 1.7945067914573147e-05,
      "loss": 0.5079,
      "step": 67
    },
    {
      "epoch": 2.0606060606060606,
      "grad_norm": 0.5661355778237616,
      "learning_rate": 1.7851994686215592e-05,
      "loss": 0.5073,
      "step": 68
    },
    {
      "epoch": 2.090909090909091,
      "grad_norm": 0.7349280665252821,
      "learning_rate": 1.77571129070442e-05,
      "loss": 0.4882,
      "step": 69
    },
    {
      "epoch": 2.121212121212121,
      "grad_norm": 0.8198312094724657,
      "learning_rate": 1.766044443118978e-05,
      "loss": 0.4903,
      "step": 70
    },
    {
      "epoch": 2.1515151515151514,
      "grad_norm": 0.6336227051044719,
      "learning_rate": 1.7562011524313187e-05,
      "loss": 0.4645,
      "step": 71
    },
    {
      "epoch": 2.1818181818181817,
      "grad_norm": 0.707497557832566,
      "learning_rate": 1.7461836858476858e-05,
      "loss": 0.4598,
      "step": 72
    },
    {
      "epoch": 2.212121212121212,
      "grad_norm": 0.690875672824615,
      "learning_rate": 1.7359943506922775e-05,
      "loss": 0.5067,
      "step": 73
    },
    {
      "epoch": 2.242424242424242,
      "grad_norm": 0.7223464037929024,
      "learning_rate": 1.725635493875799e-05,
      "loss": 0.4982,
      "step": 74
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.5891946297646339,
      "learning_rate": 1.7151095013548996e-05,
      "loss": 0.45,
      "step": 75
    },
    {
      "epoch": 2.303030303030303,
      "grad_norm": 0.6584628545720952,
      "learning_rate": 1.7044187975826126e-05,
      "loss": 0.4922,
      "step": 76
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.6376975641727657,
      "learning_rate": 1.693565844949933e-05,
      "loss": 0.4494,
      "step": 77
    },
    {
      "epoch": 2.3636363636363638,
      "grad_norm": 0.5872374510090461,
      "learning_rate": 1.6825531432186545e-05,
      "loss": 0.4684,
      "step": 78
    },
    {
      "epoch": 2.393939393939394,
      "grad_norm": 0.5700271680913076,
      "learning_rate": 1.671383228945597e-05,
      "loss": 0.4772,
      "step": 79
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 0.6365890766629746,
      "learning_rate": 1.6600586748983642e-05,
      "loss": 0.4664,
      "step": 80
    },
    {
      "epoch": 2.4545454545454546,
      "grad_norm": 0.5264422492217266,
      "learning_rate": 1.648582089462756e-05,
      "loss": 0.4741,
      "step": 81
    },
    {
      "epoch": 2.484848484848485,
      "grad_norm": 0.5117378477790195,
      "learning_rate": 1.6369561160419783e-05,
      "loss": 0.4523,
      "step": 82
    },
    {
      "epoch": 2.515151515151515,
      "grad_norm": 0.5593271124257587,
      "learning_rate": 1.625183432447789e-05,
      "loss": 0.4744,
      "step": 83
    },
    {
      "epoch": 2.5454545454545454,
      "grad_norm": 0.5105994374537374,
      "learning_rate": 1.6132667502837164e-05,
      "loss": 0.4784,
      "step": 84
    },
    {
      "epoch": 2.5757575757575757,
      "grad_norm": 0.5493961843212499,
      "learning_rate": 1.6012088143204953e-05,
      "loss": 0.4658,
      "step": 85
    },
    {
      "epoch": 2.606060606060606,
      "grad_norm": 0.4457576647974015,
      "learning_rate": 1.589012401863864e-05,
      "loss": 0.4414,
      "step": 86
    },
    {
      "epoch": 2.6363636363636362,
      "grad_norm": 0.5491001222681214,
      "learning_rate": 1.5766803221148676e-05,
      "loss": 0.4636,
      "step": 87
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.4620206642862961,
      "learning_rate": 1.5642154155228124e-05,
      "loss": 0.4499,
      "step": 88
    },
    {
      "epoch": 2.6969696969696972,
      "grad_norm": 0.5463875806513324,
      "learning_rate": 1.5516205531310272e-05,
      "loss": 0.4669,
      "step": 89
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.5324487825935355,
      "learning_rate": 1.538898635915576e-05,
      "loss": 0.4862,
      "step": 90
    },
    {
      "epoch": 2.757575757575758,
      "grad_norm": 0.5320888382047156,
      "learning_rate": 1.526052594117071e-05,
      "loss": 0.4832,
      "step": 91
    },
    {
      "epoch": 2.787878787878788,
      "grad_norm": 0.47545400565083745,
      "learning_rate": 1.513085386565758e-05,
      "loss": 0.459,
      "step": 92
    },
    {
      "epoch": 2.8181818181818183,
      "grad_norm": 0.4956748327325273,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.4672,
      "step": 93
    },
    {
      "epoch": 2.8484848484848486,
      "grad_norm": 0.4818428929042486,
      "learning_rate": 1.4867994483783485e-05,
      "loss": 0.4736,
      "step": 94
    },
    {
      "epoch": 2.878787878787879,
      "grad_norm": 0.45572505042372163,
      "learning_rate": 1.4734867721853341e-05,
      "loss": 0.4263,
      "step": 95
    },
    {
      "epoch": 2.909090909090909,
      "grad_norm": 0.45718893938191624,
      "learning_rate": 1.4600650377311523e-05,
      "loss": 0.4581,
      "step": 96
    },
    {
      "epoch": 2.9393939393939394,
      "grad_norm": 0.5252500921984969,
      "learning_rate": 1.4465373364454001e-05,
      "loss": 0.4522,
      "step": 97
    },
    {
      "epoch": 2.9696969696969697,
      "grad_norm": 0.44205305104404924,
      "learning_rate": 1.4329067841650274e-05,
      "loss": 0.4574,
      "step": 98
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.4967991506249366,
      "learning_rate": 1.4191765204166643e-05,
      "loss": 0.4578,
      "step": 99
    },
    {
      "epoch": 3.0303030303030303,
      "grad_norm": 0.7151425380787589,
      "learning_rate": 1.4053497076934948e-05,
      "loss": 0.3852,
      "step": 100
    },
    {
      "epoch": 3.0606060606060606,
      "grad_norm": 0.471105802595834,
      "learning_rate": 1.3914295307268396e-05,
      "loss": 0.3677,
      "step": 101
    },
    {
      "epoch": 3.090909090909091,
      "grad_norm": 0.7679519275837007,
      "learning_rate": 1.3774191957526144e-05,
      "loss": 0.3948,
      "step": 102
    },
    {
      "epoch": 3.121212121212121,
      "grad_norm": 0.6096107640433591,
      "learning_rate": 1.3633219297728415e-05,
      "loss": 0.3662,
      "step": 103
    },
    {
      "epoch": 3.1515151515151514,
      "grad_norm": 0.5773856422504393,
      "learning_rate": 1.3491409798123687e-05,
      "loss": 0.3753,
      "step": 104
    },
    {
      "epoch": 3.1818181818181817,
      "grad_norm": 0.5476940993617544,
      "learning_rate": 1.3348796121709862e-05,
      "loss": 0.362,
      "step": 105
    },
    {
      "epoch": 3.212121212121212,
      "grad_norm": 0.5724953897994525,
      "learning_rate": 1.3205411116710973e-05,
      "loss": 0.3661,
      "step": 106
    },
    {
      "epoch": 3.242424242424242,
      "grad_norm": 0.5468596703809803,
      "learning_rate": 1.3061287809011243e-05,
      "loss": 0.3776,
      "step": 107
    },
    {
      "epoch": 3.2727272727272725,
      "grad_norm": 0.4967275448672379,
      "learning_rate": 1.291645939454825e-05,
      "loss": 0.3683,
      "step": 108
    },
    {
      "epoch": 3.303030303030303,
      "grad_norm": 0.4858309339569281,
      "learning_rate": 1.277095923166689e-05,
      "loss": 0.3368,
      "step": 109
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.526136058163559,
      "learning_rate": 1.2624820833435939e-05,
      "loss": 0.3614,
      "step": 110
    },
    {
      "epoch": 3.3636363636363638,
      "grad_norm": 0.5242521835089266,
      "learning_rate": 1.2478077859929e-05,
      "loss": 0.3485,
      "step": 111
    },
    {
      "epoch": 3.393939393939394,
      "grad_norm": 0.46715442255298417,
      "learning_rate": 1.2330764110471567e-05,
      "loss": 0.3569,
      "step": 112
    },
    {
      "epoch": 3.4242424242424243,
      "grad_norm": 0.521996525600283,
      "learning_rate": 1.2182913515856016e-05,
      "loss": 0.3494,
      "step": 113
    },
    {
      "epoch": 3.4545454545454546,
      "grad_norm": 0.48253735550558496,
      "learning_rate": 1.2034560130526341e-05,
      "loss": 0.3589,
      "step": 114
    },
    {
      "epoch": 3.484848484848485,
      "grad_norm": 0.5029993797510168,
      "learning_rate": 1.1885738124734359e-05,
      "loss": 0.3447,
      "step": 115
    },
    {
      "epoch": 3.515151515151515,
      "grad_norm": 0.5138768414894639,
      "learning_rate": 1.1736481776669307e-05,
      "loss": 0.3591,
      "step": 116
    },
    {
      "epoch": 3.5454545454545454,
      "grad_norm": 0.48385439976246397,
      "learning_rate": 1.1586825464562515e-05,
      "loss": 0.3549,
      "step": 117
    },
    {
      "epoch": 3.5757575757575757,
      "grad_norm": 0.44492348913321644,
      "learning_rate": 1.1436803658769082e-05,
      "loss": 0.352,
      "step": 118
    },
    {
      "epoch": 3.606060606060606,
      "grad_norm": 0.47650167555492806,
      "learning_rate": 1.1286450913828313e-05,
      "loss": 0.3634,
      "step": 119
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.4981094336708157,
      "learning_rate": 1.113580186050475e-05,
      "loss": 0.3791,
      "step": 120
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.46245203620201497,
      "learning_rate": 1.0984891197811686e-05,
      "loss": 0.359,
      "step": 121
    },
    {
      "epoch": 3.6969696969696972,
      "grad_norm": 0.4580816984758574,
      "learning_rate": 1.0833753685018935e-05,
      "loss": 0.347,
      "step": 122
    },
    {
      "epoch": 3.7272727272727275,
      "grad_norm": 0.5440109738897202,
      "learning_rate": 1.0682424133646712e-05,
      "loss": 0.3684,
      "step": 123
    },
    {
      "epoch": 3.757575757575758,
      "grad_norm": 0.41250983402477176,
      "learning_rate": 1.0530937399447496e-05,
      "loss": 0.3593,
      "step": 124
    },
    {
      "epoch": 3.787878787878788,
      "grad_norm": 0.44601360069040347,
      "learning_rate": 1.0379328374377715e-05,
      "loss": 0.359,
      "step": 125
    },
    {
      "epoch": 3.8181818181818183,
      "grad_norm": 0.45845379753827253,
      "learning_rate": 1.0227631978561057e-05,
      "loss": 0.3538,
      "step": 126
    },
    {
      "epoch": 3.8484848484848486,
      "grad_norm": 0.41481474621847353,
      "learning_rate": 1.0075883152245334e-05,
      "loss": 0.3586,
      "step": 127
    },
    {
      "epoch": 3.878787878787879,
      "grad_norm": 0.46076221175425724,
      "learning_rate": 9.92411684775467e-06,
      "loss": 0.3622,
      "step": 128
    },
    {
      "epoch": 3.909090909090909,
      "grad_norm": 0.42561018999639955,
      "learning_rate": 9.772368021438943e-06,
      "loss": 0.3601,
      "step": 129
    },
    {
      "epoch": 3.9393939393939394,
      "grad_norm": 0.44968828727341753,
      "learning_rate": 9.620671625622287e-06,
      "loss": 0.3634,
      "step": 130
    },
    {
      "epoch": 3.9696969696969697,
      "grad_norm": 0.4439853935691237,
      "learning_rate": 9.469062600552509e-06,
      "loss": 0.3454,
      "step": 131
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.37933183430088063,
      "learning_rate": 9.317575866353293e-06,
      "loss": 0.3166,
      "step": 132
    },
    {
      "epoch": 4.03030303030303,
      "grad_norm": 0.6470797565309441,
      "learning_rate": 9.166246314981066e-06,
      "loss": 0.2862,
      "step": 133
    },
    {
      "epoch": 4.0606060606060606,
      "grad_norm": 0.5321084573601125,
      "learning_rate": 9.015108802188314e-06,
      "loss": 0.2779,
      "step": 134
    },
    {
      "epoch": 4.090909090909091,
      "grad_norm": 0.600907417607459,
      "learning_rate": 8.86419813949525e-06,
      "loss": 0.2699,
      "step": 135
    },
    {
      "epoch": 4.121212121212121,
      "grad_norm": 0.8014477123047449,
      "learning_rate": 8.71354908617169e-06,
      "loss": 0.2633,
      "step": 136
    },
    {
      "epoch": 4.151515151515151,
      "grad_norm": 0.5555063605976032,
      "learning_rate": 8.56319634123092e-06,
      "loss": 0.2703,
      "step": 137
    },
    {
      "epoch": 4.181818181818182,
      "grad_norm": 0.5308332788150736,
      "learning_rate": 8.413174535437486e-06,
      "loss": 0.2628,
      "step": 138
    },
    {
      "epoch": 4.212121212121212,
      "grad_norm": 0.5317391665358759,
      "learning_rate": 8.263518223330698e-06,
      "loss": 0.2591,
      "step": 139
    },
    {
      "epoch": 4.242424242424242,
      "grad_norm": 0.5198942678107813,
      "learning_rate": 8.114261875265643e-06,
      "loss": 0.256,
      "step": 140
    },
    {
      "epoch": 4.2727272727272725,
      "grad_norm": 0.45165028881457747,
      "learning_rate": 7.965439869473664e-06,
      "loss": 0.2682,
      "step": 141
    },
    {
      "epoch": 4.303030303030303,
      "grad_norm": 0.4742243535752558,
      "learning_rate": 7.817086484143987e-06,
      "loss": 0.2525,
      "step": 142
    },
    {
      "epoch": 4.333333333333333,
      "grad_norm": 0.5247665557321094,
      "learning_rate": 7.669235889528436e-06,
      "loss": 0.2803,
      "step": 143
    },
    {
      "epoch": 4.363636363636363,
      "grad_norm": 0.468161304839431,
      "learning_rate": 7.521922140071003e-06,
      "loss": 0.2702,
      "step": 144
    },
    {
      "epoch": 4.393939393939394,
      "grad_norm": 0.4367520474517696,
      "learning_rate": 7.375179166564062e-06,
      "loss": 0.2614,
      "step": 145
    },
    {
      "epoch": 4.424242424242424,
      "grad_norm": 0.43272659279914977,
      "learning_rate": 7.2290407683331154e-06,
      "loss": 0.2684,
      "step": 146
    },
    {
      "epoch": 4.454545454545454,
      "grad_norm": 0.451461445499779,
      "learning_rate": 7.0835406054517505e-06,
      "loss": 0.2389,
      "step": 147
    },
    {
      "epoch": 4.484848484848484,
      "grad_norm": 0.46223766456207516,
      "learning_rate": 6.93871219098876e-06,
      "loss": 0.2775,
      "step": 148
    },
    {
      "epoch": 4.515151515151516,
      "grad_norm": 0.4316623996416017,
      "learning_rate": 6.79458888328903e-06,
      "loss": 0.2571,
      "step": 149
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 0.43169400457445495,
      "learning_rate": 6.651203878290139e-06,
      "loss": 0.2714,
      "step": 150
    },
    {
      "epoch": 4.575757575757576,
      "grad_norm": 0.4541513588066467,
      "learning_rate": 6.508590201876317e-06,
      "loss": 0.2603,
      "step": 151
    },
    {
      "epoch": 4.606060606060606,
      "grad_norm": 0.48099367886536276,
      "learning_rate": 6.366780702271589e-06,
      "loss": 0.2676,
      "step": 152
    },
    {
      "epoch": 4.636363636363637,
      "grad_norm": 0.41330818264759017,
      "learning_rate": 6.225808042473857e-06,
      "loss": 0.2773,
      "step": 153
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.427828702181915,
      "learning_rate": 6.085704692731609e-06,
      "loss": 0.2554,
      "step": 154
    },
    {
      "epoch": 4.696969696969697,
      "grad_norm": 0.46021373503469426,
      "learning_rate": 5.946502923065054e-06,
      "loss": 0.271,
      "step": 155
    },
    {
      "epoch": 4.7272727272727275,
      "grad_norm": 0.4169479652529033,
      "learning_rate": 5.8082347958333625e-06,
      "loss": 0.2492,
      "step": 156
    },
    {
      "epoch": 4.757575757575758,
      "grad_norm": 0.44818995527879435,
      "learning_rate": 5.670932158349732e-06,
      "loss": 0.2758,
      "step": 157
    },
    {
      "epoch": 4.787878787878788,
      "grad_norm": 0.39226706771872666,
      "learning_rate": 5.534626635546e-06,
      "loss": 0.2825,
      "step": 158
    },
    {
      "epoch": 4.818181818181818,
      "grad_norm": 0.4380390923035052,
      "learning_rate": 5.399349622688479e-06,
      "loss": 0.2788,
      "step": 159
    },
    {
      "epoch": 4.848484848484849,
      "grad_norm": 0.41987431351535937,
      "learning_rate": 5.2651322781466606e-06,
      "loss": 0.2765,
      "step": 160
    },
    {
      "epoch": 4.878787878787879,
      "grad_norm": 0.3774052167025911,
      "learning_rate": 5.132005516216512e-06,
      "loss": 0.2499,
      "step": 161
    },
    {
      "epoch": 4.909090909090909,
      "grad_norm": 0.42548342616290336,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.2615,
      "step": 162
    },
    {
      "epoch": 4.9393939393939394,
      "grad_norm": 0.39250995778651426,
      "learning_rate": 4.869146134342426e-06,
      "loss": 0.2717,
      "step": 163
    },
    {
      "epoch": 4.96969696969697,
      "grad_norm": 0.4072283354257161,
      "learning_rate": 4.739474058829288e-06,
      "loss": 0.2738,
      "step": 164
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.4094664375634055,
      "learning_rate": 4.611013640844245e-06,
      "loss": 0.2725,
      "step": 165
    },
    {
      "epoch": 5.03030303030303,
      "grad_norm": 0.5808293128425991,
      "learning_rate": 4.483794468689728e-06,
      "loss": 0.2189,
      "step": 166
    },
    {
      "epoch": 5.0606060606060606,
      "grad_norm": 0.507800942854878,
      "learning_rate": 4.357845844771881e-06,
      "loss": 0.2062,
      "step": 167
    },
    {
      "epoch": 5.090909090909091,
      "grad_norm": 0.377948773593431,
      "learning_rate": 4.2331967788513295e-06,
      "loss": 0.1989,
      "step": 168
    },
    {
      "epoch": 5.121212121212121,
      "grad_norm": 0.5201581783870141,
      "learning_rate": 4.109875981361363e-06,
      "loss": 0.2203,
      "step": 169
    },
    {
      "epoch": 5.151515151515151,
      "grad_norm": 0.7080893793467118,
      "learning_rate": 3.987911856795047e-06,
      "loss": 0.2223,
      "step": 170
    },
    {
      "epoch": 5.181818181818182,
      "grad_norm": 0.5367083341843927,
      "learning_rate": 3.867332497162836e-06,
      "loss": 0.2113,
      "step": 171
    },
    {
      "epoch": 5.212121212121212,
      "grad_norm": 0.4526772381512374,
      "learning_rate": 3.748165675522113e-06,
      "loss": 0.2218,
      "step": 172
    },
    {
      "epoch": 5.242424242424242,
      "grad_norm": 0.41982276386549366,
      "learning_rate": 3.630438839580217e-06,
      "loss": 0.2039,
      "step": 173
    },
    {
      "epoch": 5.2727272727272725,
      "grad_norm": 0.48303251636543654,
      "learning_rate": 3.5141791053724405e-06,
      "loss": 0.1965,
      "step": 174
    },
    {
      "epoch": 5.303030303030303,
      "grad_norm": 0.49184963071814763,
      "learning_rate": 3.399413251016359e-06,
      "loss": 0.2098,
      "step": 175
    },
    {
      "epoch": 5.333333333333333,
      "grad_norm": 0.4383246513939706,
      "learning_rate": 3.2861677105440335e-06,
      "loss": 0.1956,
      "step": 176
    },
    {
      "epoch": 5.363636363636363,
      "grad_norm": 0.42401719435999935,
      "learning_rate": 3.174468567813461e-06,
      "loss": 0.1999,
      "step": 177
    },
    {
      "epoch": 5.393939393939394,
      "grad_norm": 0.43753467083807723,
      "learning_rate": 3.0643415505006733e-06,
      "loss": 0.2018,
      "step": 178
    },
    {
      "epoch": 5.424242424242424,
      "grad_norm": 0.4590529242795019,
      "learning_rate": 2.9558120241738786e-06,
      "loss": 0.2021,
      "step": 179
    },
    {
      "epoch": 5.454545454545454,
      "grad_norm": 0.45966124692622584,
      "learning_rate": 2.8489049864510053e-06,
      "loss": 0.2045,
      "step": 180
    },
    {
      "epoch": 5.484848484848484,
      "grad_norm": 0.4263517952530578,
      "learning_rate": 2.7436450612420098e-06,
      "loss": 0.1979,
      "step": 181
    },
    {
      "epoch": 5.515151515151516,
      "grad_norm": 0.4007470790502227,
      "learning_rate": 2.640056493077231e-06,
      "loss": 0.179,
      "step": 182
    },
    {
      "epoch": 5.545454545454545,
      "grad_norm": 0.4654799952190163,
      "learning_rate": 2.5381631415231455e-06,
      "loss": 0.1988,
      "step": 183
    },
    {
      "epoch": 5.575757575757576,
      "grad_norm": 0.3708808312907506,
      "learning_rate": 2.4379884756868167e-06,
      "loss": 0.2137,
      "step": 184
    },
    {
      "epoch": 5.606060606060606,
      "grad_norm": 0.3970173283908434,
      "learning_rate": 2.339555568810221e-06,
      "loss": 0.2087,
      "step": 185
    },
    {
      "epoch": 5.636363636363637,
      "grad_norm": 0.38981696143002886,
      "learning_rate": 2.2428870929558012e-06,
      "loss": 0.2211,
      "step": 186
    },
    {
      "epoch": 5.666666666666667,
      "grad_norm": 0.370544974405656,
      "learning_rate": 2.1480053137844115e-06,
      "loss": 0.1952,
      "step": 187
    },
    {
      "epoch": 5.696969696969697,
      "grad_norm": 0.38020180226795797,
      "learning_rate": 2.054932085426856e-06,
      "loss": 0.1933,
      "step": 188
    },
    {
      "epoch": 5.7272727272727275,
      "grad_norm": 0.37871324688172686,
      "learning_rate": 1.963688845450218e-06,
      "loss": 0.1987,
      "step": 189
    },
    {
      "epoch": 5.757575757575758,
      "grad_norm": 0.3526562165980149,
      "learning_rate": 1.8742966099201699e-06,
      "loss": 0.208,
      "step": 190
    },
    {
      "epoch": 5.787878787878788,
      "grad_norm": 0.3916864688039014,
      "learning_rate": 1.7867759685603115e-06,
      "loss": 0.2001,
      "step": 191
    },
    {
      "epoch": 5.818181818181818,
      "grad_norm": 0.37050014872864956,
      "learning_rate": 1.7011470800097496e-06,
      "loss": 0.1874,
      "step": 192
    },
    {
      "epoch": 5.848484848484849,
      "grad_norm": 0.36343600800947173,
      "learning_rate": 1.6174296671799571e-06,
      "loss": 0.1961,
      "step": 193
    },
    {
      "epoch": 5.878787878787879,
      "grad_norm": 0.36977567897105296,
      "learning_rate": 1.5356430127119915e-06,
      "loss": 0.2117,
      "step": 194
    },
    {
      "epoch": 5.909090909090909,
      "grad_norm": 0.3831845952090707,
      "learning_rate": 1.4558059545351144e-06,
      "loss": 0.1905,
      "step": 195
    },
    {
      "epoch": 5.9393939393939394,
      "grad_norm": 0.3683454341984337,
      "learning_rate": 1.3779368815278648e-06,
      "loss": 0.1998,
      "step": 196
    },
    {
      "epoch": 5.96969696969697,
      "grad_norm": 0.3636124019597834,
      "learning_rate": 1.302053729282533e-06,
      "loss": 0.2128,
      "step": 197
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.34895010552301764,
      "learning_rate": 1.2281739759740575e-06,
      "loss": 0.2011,
      "step": 198
    },
    {
      "epoch": 6.03030303030303,
      "grad_norm": 0.4952822859281516,
      "learning_rate": 1.156314638334277e-06,
      "loss": 0.1736,
      "step": 199
    },
    {
      "epoch": 6.0606060606060606,
      "grad_norm": 0.44105699706220314,
      "learning_rate": 1.086492267732462e-06,
      "loss": 0.1852,
      "step": 200
    },
    {
      "epoch": 6.090909090909091,
      "grad_norm": 0.42971076379832884,
      "learning_rate": 1.01872294636304e-06,
      "loss": 0.1624,
      "step": 201
    },
    {
      "epoch": 6.121212121212121,
      "grad_norm": 0.36375074266798585,
      "learning_rate": 9.530222835413739e-07,
      "loss": 0.1821,
      "step": 202
    },
    {
      "epoch": 6.151515151515151,
      "grad_norm": 0.37782756153116515,
      "learning_rate": 8.894054121084839e-07,
      "loss": 0.1711,
      "step": 203
    },
    {
      "epoch": 6.181818181818182,
      "grad_norm": 0.3556156958116767,
      "learning_rate": 8.278869849454718e-07,
      "loss": 0.1655,
      "step": 204
    },
    {
      "epoch": 6.212121212121212,
      "grad_norm": 0.3525958294428809,
      "learning_rate": 7.684811715985429e-07,
      "loss": 0.1711,
      "step": 205
    },
    {
      "epoch": 6.242424242424242,
      "grad_norm": 0.3655206933076983,
      "learning_rate": 7.1120165501533e-07,
      "loss": 0.1709,
      "step": 206
    },
    {
      "epoch": 6.2727272727272725,
      "grad_norm": 0.39633133056347575,
      "learning_rate": 6.560616283932897e-07,
      "loss": 0.1742,
      "step": 207
    },
    {
      "epoch": 6.303030303030303,
      "grad_norm": 0.4128930485722999,
      "learning_rate": 6.030737921409169e-07,
      "loss": 0.1746,
      "step": 208
    },
    {
      "epoch": 6.333333333333333,
      "grad_norm": 0.3865282518146205,
      "learning_rate": 5.522503509524591e-07,
      "loss": 0.1829,
      "step": 209
    },
    {
      "epoch": 6.363636363636363,
      "grad_norm": 0.38910019530506973,
      "learning_rate": 5.036030109968082e-07,
      "loss": 0.1778,
      "step": 210
    },
    {
      "epoch": 6.393939393939394,
      "grad_norm": 0.36553087202247425,
      "learning_rate": 4.5714297722121105e-07,
      "loss": 0.1766,
      "step": 211
    },
    {
      "epoch": 6.424242424242424,
      "grad_norm": 0.3830479052912986,
      "learning_rate": 4.128809507704445e-07,
      "loss": 0.1792,
      "step": 212
    },
    {
      "epoch": 6.454545454545454,
      "grad_norm": 0.3315584314354747,
      "learning_rate": 3.708271265220087e-07,
      "loss": 0.1811,
      "step": 213
    },
    {
      "epoch": 6.484848484848484,
      "grad_norm": 0.3372722028777837,
      "learning_rate": 3.309911907379393e-07,
      "loss": 0.1748,
      "step": 214
    },
    {
      "epoch": 6.515151515151516,
      "grad_norm": 0.36465545806454136,
      "learning_rate": 2.9338231883378365e-07,
      "loss": 0.1843,
      "step": 215
    },
    {
      "epoch": 6.545454545454545,
      "grad_norm": 0.3607512929480319,
      "learning_rate": 2.5800917326521013e-07,
      "loss": 0.1775,
      "step": 216
    },
    {
      "epoch": 6.575757575757576,
      "grad_norm": 0.3473293056603518,
      "learning_rate": 2.248799015327907e-07,
      "loss": 0.1768,
      "step": 217
    },
    {
      "epoch": 6.606060606060606,
      "grad_norm": 0.3513704435134903,
      "learning_rate": 1.9400213430538773e-07,
      "loss": 0.167,
      "step": 218
    },
    {
      "epoch": 6.636363636363637,
      "grad_norm": 0.34831123543752296,
      "learning_rate": 1.6538298366257975e-07,
      "loss": 0.1946,
      "step": 219
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.335850246644892,
      "learning_rate": 1.3902904145653094e-07,
      "loss": 0.1686,
      "step": 220
    },
    {
      "epoch": 6.696969696969697,
      "grad_norm": 0.34609792203824336,
      "learning_rate": 1.1494637779369766e-07,
      "loss": 0.1771,
      "step": 221
    },
    {
      "epoch": 6.7272727272727275,
      "grad_norm": 0.33204184845074686,
      "learning_rate": 9.314053963669245e-08,
      "loss": 0.1736,
      "step": 222
    },
    {
      "epoch": 6.757575757575758,
      "grad_norm": 0.34066508310965216,
      "learning_rate": 7.361654952665608e-08,
      "loss": 0.1802,
      "step": 223
    },
    {
      "epoch": 6.787878787878788,
      "grad_norm": 0.34084081572188607,
      "learning_rate": 5.637890442641403e-08,
      "loss": 0.1883,
      "step": 224
    },
    {
      "epoch": 6.818181818181818,
      "grad_norm": 0.34255691953224654,
      "learning_rate": 4.143157468468717e-08,
      "loss": 0.1768,
      "step": 225
    },
    {
      "epoch": 6.848484848484849,
      "grad_norm": 0.3578230924997704,
      "learning_rate": 2.8778003121607834e-08,
      "loss": 0.1764,
      "step": 226
    },
    {
      "epoch": 6.878787878787879,
      "grad_norm": 0.34423532992168493,
      "learning_rate": 1.8421104235727406e-08,
      "loss": 0.164,
      "step": 227
    },
    {
      "epoch": 6.909090909090909,
      "grad_norm": 0.3367312936055244,
      "learning_rate": 1.0363263532724433e-08,
      "loss": 0.1797,
      "step": 228
    },
    {
      "epoch": 6.9393939393939394,
      "grad_norm": 0.331945015492203,
      "learning_rate": 4.606336975948589e-09,
      "loss": 0.1835,
      "step": 229
    },
    {
      "epoch": 6.96969696969697,
      "grad_norm": 0.3552845337018321,
      "learning_rate": 1.1516505589381777e-09,
      "loss": 0.1829,
      "step": 230
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.3635769580343739,
      "learning_rate": 0.0,
      "loss": 0.1748,
      "step": 231
    },
    {
      "epoch": 7.0,
      "step": 231,
      "total_flos": 3.9882325361688576e+17,
      "train_loss": 0.408949723491421,
      "train_runtime": 7046.6368,
      "train_samples_per_second": 3.139,
      "train_steps_per_second": 0.033
    }
  ],
  "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": 3.9882325361688576e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}