{
  "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": 6.052227873486153,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.9657,
      "step": 1
    },
    {
      "epoch": 0.06060606060606061,
      "grad_norm": 6.362670215979786,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.0286,
      "step": 2
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 6.27256618731045,
      "learning_rate": 2.5e-06,
      "loss": 1.016,
      "step": 3
    },
    {
      "epoch": 0.12121212121212122,
      "grad_norm": 5.551305027563004,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.9702,
      "step": 4
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 4.310546099316045,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.9272,
      "step": 5
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 2.281474210936113,
      "learning_rate": 5e-06,
      "loss": 0.9133,
      "step": 6
    },
    {
      "epoch": 0.21212121212121213,
      "grad_norm": 1.828058852291928,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.8551,
      "step": 7
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 4.1017619886517815,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.8587,
      "step": 8
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 4.768414872886225,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.9228,
      "step": 9
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 4.639971275443633,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.8859,
      "step": 10
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 4.203822322437633,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.8653,
      "step": 11
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 3.387005886952801,
      "learning_rate": 1e-05,
      "loss": 0.8072,
      "step": 12
    },
    {
      "epoch": 0.3939393939393939,
      "grad_norm": 3.076916093375926,
      "learning_rate": 1.0833333333333334e-05,
      "loss": 0.869,
      "step": 13
    },
    {
      "epoch": 0.42424242424242425,
      "grad_norm": 2.1825346363817992,
      "learning_rate": 1.1666666666666668e-05,
      "loss": 0.7942,
      "step": 14
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 2.377651889730694,
      "learning_rate": 1.25e-05,
      "loss": 0.7828,
      "step": 15
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 2.270198363166352,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.7734,
      "step": 16
    },
    {
      "epoch": 0.5151515151515151,
      "grad_norm": 1.8910913070978497,
      "learning_rate": 1.416666666666667e-05,
      "loss": 0.7484,
      "step": 17
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 1.417014246940671,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.7301,
      "step": 18
    },
    {
      "epoch": 0.5757575757575758,
      "grad_norm": 1.5022509367239376,
      "learning_rate": 1.5833333333333333e-05,
      "loss": 0.7646,
      "step": 19
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 1.5205978519478327,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.7329,
      "step": 20
    },
    {
      "epoch": 0.6363636363636364,
      "grad_norm": 1.1229292449571018,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 0.7372,
      "step": 21
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 1.1355094443928555,
      "learning_rate": 1.8333333333333333e-05,
      "loss": 0.7414,
      "step": 22
    },
    {
      "epoch": 0.696969696969697,
      "grad_norm": 1.2370479393126448,
      "learning_rate": 1.916666666666667e-05,
      "loss": 0.6909,
      "step": 23
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.86353430094354,
      "learning_rate": 2e-05,
      "loss": 0.7183,
      "step": 24
    },
    {
      "epoch": 0.7575757575757576,
      "grad_norm": 0.8595883994411603,
      "learning_rate": 1.999884834944106e-05,
      "loss": 0.6877,
      "step": 25
    },
    {
      "epoch": 0.7878787878787878,
      "grad_norm": 1.0145395716772336,
      "learning_rate": 1.9995393663024054e-05,
      "loss": 0.6967,
      "step": 26
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 0.6955200009392845,
      "learning_rate": 1.9989636736467278e-05,
      "loss": 0.6724,
      "step": 27
    },
    {
      "epoch": 0.8484848484848485,
      "grad_norm": 0.831656892095645,
      "learning_rate": 1.9981578895764272e-05,
      "loss": 0.7216,
      "step": 28
    },
    {
      "epoch": 0.8787878787878788,
      "grad_norm": 0.8453944238831343,
      "learning_rate": 1.9971221996878395e-05,
      "loss": 0.6993,
      "step": 29
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 1.3866105605305914,
      "learning_rate": 1.9958568425315316e-05,
      "loss": 0.6825,
      "step": 30
    },
    {
      "epoch": 0.9393939393939394,
      "grad_norm": 0.9838002655935489,
      "learning_rate": 1.9943621095573588e-05,
      "loss": 0.6803,
      "step": 31
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.7154333520049099,
      "learning_rate": 1.9926383450473344e-05,
      "loss": 0.7148,
      "step": 32
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.8562415720147992,
      "learning_rate": 1.9906859460363307e-05,
      "loss": 0.6804,
      "step": 33
    },
    {
      "epoch": 1.0303030303030303,
      "grad_norm": 0.7509706394789555,
      "learning_rate": 1.9885053622206305e-05,
      "loss": 0.6109,
      "step": 34
    },
    {
      "epoch": 1.0606060606060606,
      "grad_norm": 0.722571663868852,
      "learning_rate": 1.986097095854347e-05,
      "loss": 0.6264,
      "step": 35
    },
    {
      "epoch": 1.0909090909090908,
      "grad_norm": 0.9048240545870672,
      "learning_rate": 1.9834617016337424e-05,
      "loss": 0.6438,
      "step": 36
    },
    {
      "epoch": 1.121212121212121,
      "grad_norm": 0.7354190941174,
      "learning_rate": 1.9805997865694616e-05,
      "loss": 0.6346,
      "step": 37
    },
    {
      "epoch": 1.1515151515151516,
      "grad_norm": 9.163666536353414,
      "learning_rate": 1.9775120098467212e-05,
      "loss": 0.6191,
      "step": 38
    },
    {
      "epoch": 1.1818181818181819,
      "grad_norm": 0.9915146385904654,
      "learning_rate": 1.9741990826734793e-05,
      "loss": 0.6598,
      "step": 39
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 1.820602192623448,
      "learning_rate": 1.970661768116622e-05,
      "loss": 0.6145,
      "step": 40
    },
    {
      "epoch": 1.2424242424242424,
      "grad_norm": 1.2677899930044216,
      "learning_rate": 1.9669008809262064e-05,
      "loss": 0.6512,
      "step": 41
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 0.8451002975947255,
      "learning_rate": 1.9629172873477995e-05,
      "loss": 0.631,
      "step": 42
    },
    {
      "epoch": 1.303030303030303,
      "grad_norm": 0.8887838656444853,
      "learning_rate": 1.9587119049229558e-05,
      "loss": 0.612,
      "step": 43
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.7554984494169769,
      "learning_rate": 1.954285702277879e-05,
      "loss": 0.5945,
      "step": 44
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.9058207710357586,
      "learning_rate": 1.9496396989003195e-05,
      "loss": 0.6153,
      "step": 45
    },
    {
      "epoch": 1.393939393939394,
      "grad_norm": 0.7284794128987869,
      "learning_rate": 1.944774964904754e-05,
      "loss": 0.601,
      "step": 46
    },
    {
      "epoch": 1.4242424242424243,
      "grad_norm": 0.9203848945391696,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 0.6374,
      "step": 47
    },
    {
      "epoch": 1.4545454545454546,
      "grad_norm": 0.58556192869173,
      "learning_rate": 1.9343938371606714e-05,
      "loss": 0.6012,
      "step": 48
    },
    {
      "epoch": 1.4848484848484849,
      "grad_norm": 0.7309027450598211,
      "learning_rate": 1.9288798344984673e-05,
      "loss": 0.5634,
      "step": 49
    },
    {
      "epoch": 1.5151515151515151,
      "grad_norm": 0.6448352263777719,
      "learning_rate": 1.9231518828401458e-05,
      "loss": 0.5865,
      "step": 50
    },
    {
      "epoch": 1.5454545454545454,
      "grad_norm": 0.6745598030954697,
      "learning_rate": 1.917211301505453e-05,
      "loss": 0.6076,
      "step": 51
    },
    {
      "epoch": 1.5757575757575757,
      "grad_norm": 0.6534619743791373,
      "learning_rate": 1.911059458789152e-05,
      "loss": 0.5773,
      "step": 52
    },
    {
      "epoch": 1.606060606060606,
      "grad_norm": 0.5962072226333365,
      "learning_rate": 1.9046977716458627e-05,
      "loss": 0.5532,
      "step": 53
    },
    {
      "epoch": 1.6363636363636362,
      "grad_norm": 0.7970010648258183,
      "learning_rate": 1.8981277053636963e-05,
      "loss": 0.5937,
      "step": 54
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.7020550385867285,
      "learning_rate": 1.891350773226754e-05,
      "loss": 0.5233,
      "step": 55
    },
    {
      "epoch": 1.696969696969697,
      "grad_norm": 0.7382721379877667,
      "learning_rate": 1.8843685361665724e-05,
      "loss": 0.5607,
      "step": 56
    },
    {
      "epoch": 1.7272727272727273,
      "grad_norm": 0.7377968511826565,
      "learning_rate": 1.8771826024025944e-05,
      "loss": 0.6075,
      "step": 57
    },
    {
      "epoch": 1.7575757575757576,
      "grad_norm": 0.7586948873823123,
      "learning_rate": 1.8697946270717468e-05,
      "loss": 0.5896,
      "step": 58
    },
    {
      "epoch": 1.7878787878787878,
      "grad_norm": 0.6090973907355298,
      "learning_rate": 1.8622063118472135e-05,
      "loss": 0.6107,
      "step": 59
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.8219141947639699,
      "learning_rate": 1.8544194045464888e-05,
      "loss": 0.603,
      "step": 60
    },
    {
      "epoch": 1.8484848484848486,
      "grad_norm": 0.5715029761163661,
      "learning_rate": 1.8464356987288012e-05,
      "loss": 0.6002,
      "step": 61
    },
    {
      "epoch": 1.878787878787879,
      "grad_norm": 0.6463254929837546,
      "learning_rate": 1.8382570332820045e-05,
      "loss": 0.602,
      "step": 62
    },
    {
      "epoch": 1.9090909090909092,
      "grad_norm": 0.6697688436278422,
      "learning_rate": 1.8298852919990254e-05,
      "loss": 0.5914,
      "step": 63
    },
    {
      "epoch": 1.9393939393939394,
      "grad_norm": 0.6388186071165237,
      "learning_rate": 1.821322403143969e-05,
      "loss": 0.567,
      "step": 64
    },
    {
      "epoch": 1.9696969696969697,
      "grad_norm": 0.5080064976565799,
      "learning_rate": 1.812570339007983e-05,
      "loss": 0.5485,
      "step": 65
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6775895220099829,
      "learning_rate": 1.8036311154549783e-05,
      "loss": 0.6045,
      "step": 66
    },
    {
      "epoch": 2.0303030303030303,
      "grad_norm": 0.7787072850255053,
      "learning_rate": 1.7945067914573147e-05,
      "loss": 0.5038,
      "step": 67
    },
    {
      "epoch": 2.0606060606060606,
      "grad_norm": 0.5561391354048083,
      "learning_rate": 1.7851994686215592e-05,
      "loss": 0.5111,
      "step": 68
    },
    {
      "epoch": 2.090909090909091,
      "grad_norm": 0.7729906505541877,
      "learning_rate": 1.77571129070442e-05,
      "loss": 0.4978,
      "step": 69
    },
    {
      "epoch": 2.121212121212121,
      "grad_norm": 0.6600184827448743,
      "learning_rate": 1.766044443118978e-05,
      "loss": 0.4744,
      "step": 70
    },
    {
      "epoch": 2.1515151515151514,
      "grad_norm": 0.5976935979064065,
      "learning_rate": 1.7562011524313187e-05,
      "loss": 0.4935,
      "step": 71
    },
    {
      "epoch": 2.1818181818181817,
      "grad_norm": 0.6550648208826424,
      "learning_rate": 1.7461836858476858e-05,
      "loss": 0.474,
      "step": 72
    },
    {
      "epoch": 2.212121212121212,
      "grad_norm": 0.5861127919520355,
      "learning_rate": 1.7359943506922775e-05,
      "loss": 0.4758,
      "step": 73
    },
    {
      "epoch": 2.242424242424242,
      "grad_norm": 0.6463509459141927,
      "learning_rate": 1.725635493875799e-05,
      "loss": 0.461,
      "step": 74
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 6.531561329076502,
      "learning_rate": 1.7151095013548996e-05,
      "loss": 0.5246,
      "step": 75
    },
    {
      "epoch": 2.303030303030303,
      "grad_norm": 0.8325348260346536,
      "learning_rate": 1.7044187975826126e-05,
      "loss": 0.4647,
      "step": 76
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.5971615295149254,
      "learning_rate": 1.693565844949933e-05,
      "loss": 0.4701,
      "step": 77
    },
    {
      "epoch": 2.3636363636363638,
      "grad_norm": 0.6947382001990526,
      "learning_rate": 1.6825531432186545e-05,
      "loss": 0.4553,
      "step": 78
    },
    {
      "epoch": 2.393939393939394,
      "grad_norm": 0.578307508624809,
      "learning_rate": 1.671383228945597e-05,
      "loss": 0.4647,
      "step": 79
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 0.6343615706936231,
      "learning_rate": 1.6600586748983642e-05,
      "loss": 0.4897,
      "step": 80
    },
    {
      "epoch": 2.4545454545454546,
      "grad_norm": 0.575484467333354,
      "learning_rate": 1.648582089462756e-05,
      "loss": 0.4499,
      "step": 81
    },
    {
      "epoch": 2.484848484848485,
      "grad_norm": 0.5811660633683511,
      "learning_rate": 1.6369561160419783e-05,
      "loss": 0.4801,
      "step": 82
    },
    {
      "epoch": 2.515151515151515,
      "grad_norm": 0.5201412597211769,
      "learning_rate": 1.625183432447789e-05,
      "loss": 0.4644,
      "step": 83
    },
    {
      "epoch": 2.5454545454545454,
      "grad_norm": 0.5567629728980261,
      "learning_rate": 1.6132667502837164e-05,
      "loss": 0.4684,
      "step": 84
    },
    {
      "epoch": 2.5757575757575757,
      "grad_norm": 0.48458200625904974,
      "learning_rate": 1.6012088143204953e-05,
      "loss": 0.4469,
      "step": 85
    },
    {
      "epoch": 2.606060606060606,
      "grad_norm": 0.5101140414768554,
      "learning_rate": 1.589012401863864e-05,
      "loss": 0.4587,
      "step": 86
    },
    {
      "epoch": 2.6363636363636362,
      "grad_norm": 0.6213498460677401,
      "learning_rate": 1.5766803221148676e-05,
      "loss": 0.4567,
      "step": 87
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.512617290196496,
      "learning_rate": 1.5642154155228124e-05,
      "loss": 0.4645,
      "step": 88
    },
    {
      "epoch": 2.6969696969696972,
      "grad_norm": 0.5357938542824119,
      "learning_rate": 1.5516205531310272e-05,
      "loss": 0.464,
      "step": 89
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.5541209703373594,
      "learning_rate": 1.538898635915576e-05,
      "loss": 0.4709,
      "step": 90
    },
    {
      "epoch": 2.757575757575758,
      "grad_norm": 0.6267850456948834,
      "learning_rate": 1.526052594117071e-05,
      "loss": 0.4607,
      "step": 91
    },
    {
      "epoch": 2.787878787878788,
      "grad_norm": 0.5063695092087909,
      "learning_rate": 1.513085386565758e-05,
      "loss": 0.4375,
      "step": 92
    },
    {
      "epoch": 2.8181818181818183,
      "grad_norm": 0.6424338892560242,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.4907,
      "step": 93
    },
    {
      "epoch": 2.8484848484848486,
      "grad_norm": 0.6007897045134489,
      "learning_rate": 1.4867994483783485e-05,
      "loss": 0.4578,
      "step": 94
    },
    {
      "epoch": 2.878787878787879,
      "grad_norm": 0.5295751498916507,
      "learning_rate": 1.4734867721853341e-05,
      "loss": 0.4586,
      "step": 95
    },
    {
      "epoch": 2.909090909090909,
      "grad_norm": 0.5109672827294545,
      "learning_rate": 1.4600650377311523e-05,
      "loss": 0.4652,
      "step": 96
    },
    {
      "epoch": 2.9393939393939394,
      "grad_norm": 0.5226599147068228,
      "learning_rate": 1.4465373364454001e-05,
      "loss": 0.4854,
      "step": 97
    },
    {
      "epoch": 2.9696969696969697,
      "grad_norm": 0.48151193604875586,
      "learning_rate": 1.4329067841650274e-05,
      "loss": 0.4442,
      "step": 98
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.5268264328981253,
      "learning_rate": 1.4191765204166643e-05,
      "loss": 0.4574,
      "step": 99
    },
    {
      "epoch": 3.0303030303030303,
      "grad_norm": 0.7333409738095137,
      "learning_rate": 1.4053497076934948e-05,
      "loss": 0.3676,
      "step": 100
    },
    {
      "epoch": 3.0606060606060606,
      "grad_norm": 0.5580098146125548,
      "learning_rate": 1.3914295307268396e-05,
      "loss": 0.344,
      "step": 101
    },
    {
      "epoch": 3.090909090909091,
      "grad_norm": 1.1144068953296025,
      "learning_rate": 1.3774191957526144e-05,
      "loss": 0.3549,
      "step": 102
    },
    {
      "epoch": 3.121212121212121,
      "grad_norm": 0.7874305394211059,
      "learning_rate": 1.3633219297728415e-05,
      "loss": 0.3831,
      "step": 103
    },
    {
      "epoch": 3.1515151515151514,
      "grad_norm": 0.627609324740529,
      "learning_rate": 1.3491409798123687e-05,
      "loss": 0.3586,
      "step": 104
    },
    {
      "epoch": 3.1818181818181817,
      "grad_norm": 0.7451665221692803,
      "learning_rate": 1.3348796121709862e-05,
      "loss": 0.365,
      "step": 105
    },
    {
      "epoch": 3.212121212121212,
      "grad_norm": 0.5783238514968136,
      "learning_rate": 1.3205411116710973e-05,
      "loss": 0.3285,
      "step": 106
    },
    {
      "epoch": 3.242424242424242,
      "grad_norm": 0.6493821155038698,
      "learning_rate": 1.3061287809011243e-05,
      "loss": 0.3564,
      "step": 107
    },
    {
      "epoch": 3.2727272727272725,
      "grad_norm": 0.5909427642148763,
      "learning_rate": 1.291645939454825e-05,
      "loss": 0.3701,
      "step": 108
    },
    {
      "epoch": 3.303030303030303,
      "grad_norm": 0.5691537809484646,
      "learning_rate": 1.277095923166689e-05,
      "loss": 0.3475,
      "step": 109
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.6125857648804229,
      "learning_rate": 1.2624820833435939e-05,
      "loss": 0.3441,
      "step": 110
    },
    {
      "epoch": 3.3636363636363638,
      "grad_norm": 0.5399287979892525,
      "learning_rate": 1.2478077859929e-05,
      "loss": 0.3429,
      "step": 111
    },
    {
      "epoch": 3.393939393939394,
      "grad_norm": 0.532304717350705,
      "learning_rate": 1.2330764110471567e-05,
      "loss": 0.3582,
      "step": 112
    },
    {
      "epoch": 3.4242424242424243,
      "grad_norm": 0.6285818368380538,
      "learning_rate": 1.2182913515856016e-05,
      "loss": 0.3677,
      "step": 113
    },
    {
      "epoch": 3.4545454545454546,
      "grad_norm": 0.5445078695122021,
      "learning_rate": 1.2034560130526341e-05,
      "loss": 0.355,
      "step": 114
    },
    {
      "epoch": 3.484848484848485,
      "grad_norm": 0.4753393311012243,
      "learning_rate": 1.1885738124734359e-05,
      "loss": 0.3499,
      "step": 115
    },
    {
      "epoch": 3.515151515151515,
      "grad_norm": 0.5622350341807907,
      "learning_rate": 1.1736481776669307e-05,
      "loss": 0.3527,
      "step": 116
    },
    {
      "epoch": 3.5454545454545454,
      "grad_norm": 0.5379081395616313,
      "learning_rate": 1.1586825464562515e-05,
      "loss": 0.3657,
      "step": 117
    },
    {
      "epoch": 3.5757575757575757,
      "grad_norm": 0.4995108088552664,
      "learning_rate": 1.1436803658769082e-05,
      "loss": 0.3585,
      "step": 118
    },
    {
      "epoch": 3.606060606060606,
      "grad_norm": 0.48360184923429705,
      "learning_rate": 1.1286450913828313e-05,
      "loss": 0.3497,
      "step": 119
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.44582932868286906,
      "learning_rate": 1.113580186050475e-05,
      "loss": 0.3326,
      "step": 120
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.518809299199542,
      "learning_rate": 1.0984891197811686e-05,
      "loss": 0.3342,
      "step": 121
    },
    {
      "epoch": 3.6969696969696972,
      "grad_norm": 0.4384423988224861,
      "learning_rate": 1.0833753685018935e-05,
      "loss": 0.3494,
      "step": 122
    },
    {
      "epoch": 3.7272727272727275,
      "grad_norm": 0.4883672354749273,
      "learning_rate": 1.0682424133646712e-05,
      "loss": 0.3554,
      "step": 123
    },
    {
      "epoch": 3.757575757575758,
      "grad_norm": 0.4725161791342924,
      "learning_rate": 1.0530937399447496e-05,
      "loss": 0.3498,
      "step": 124
    },
    {
      "epoch": 3.787878787878788,
      "grad_norm": 0.4955219779358546,
      "learning_rate": 1.0379328374377715e-05,
      "loss": 0.3638,
      "step": 125
    },
    {
      "epoch": 3.8181818181818183,
      "grad_norm": 0.45556408707800333,
      "learning_rate": 1.0227631978561057e-05,
      "loss": 0.3401,
      "step": 126
    },
    {
      "epoch": 3.8484848484848486,
      "grad_norm": 0.44708059029833863,
      "learning_rate": 1.0075883152245334e-05,
      "loss": 0.3281,
      "step": 127
    },
    {
      "epoch": 3.878787878787879,
      "grad_norm": 0.49858072321304076,
      "learning_rate": 9.92411684775467e-06,
      "loss": 0.3353,
      "step": 128
    },
    {
      "epoch": 3.909090909090909,
      "grad_norm": 0.46308230977631354,
      "learning_rate": 9.772368021438943e-06,
      "loss": 0.3416,
      "step": 129
    },
    {
      "epoch": 3.9393939393939394,
      "grad_norm": 0.439011599398423,
      "learning_rate": 9.620671625622287e-06,
      "loss": 0.3558,
      "step": 130
    },
    {
      "epoch": 3.9696969696969697,
      "grad_norm": 0.4796581412626963,
      "learning_rate": 9.469062600552509e-06,
      "loss": 0.3596,
      "step": 131
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.44500553237137347,
      "learning_rate": 9.317575866353293e-06,
      "loss": 0.3298,
      "step": 132
    },
    {
      "epoch": 4.03030303030303,
      "grad_norm": 0.7916348807214354,
      "learning_rate": 9.166246314981066e-06,
      "loss": 0.2743,
      "step": 133
    },
    {
      "epoch": 4.0606060606060606,
      "grad_norm": 0.60902762280639,
      "learning_rate": 9.015108802188314e-06,
      "loss": 0.2646,
      "step": 134
    },
    {
      "epoch": 4.090909090909091,
      "grad_norm": 0.6173010916967449,
      "learning_rate": 8.86419813949525e-06,
      "loss": 0.2751,
      "step": 135
    },
    {
      "epoch": 4.121212121212121,
      "grad_norm": 1.0664380257274437,
      "learning_rate": 8.71354908617169e-06,
      "loss": 0.2514,
      "step": 136
    },
    {
      "epoch": 4.151515151515151,
      "grad_norm": 0.6977911758520724,
      "learning_rate": 8.56319634123092e-06,
      "loss": 0.2476,
      "step": 137
    },
    {
      "epoch": 4.181818181818182,
      "grad_norm": 0.5495457389560459,
      "learning_rate": 8.413174535437486e-06,
      "loss": 0.2657,
      "step": 138
    },
    {
      "epoch": 4.212121212121212,
      "grad_norm": 0.6453462301023484,
      "learning_rate": 8.263518223330698e-06,
      "loss": 0.2471,
      "step": 139
    },
    {
      "epoch": 4.242424242424242,
      "grad_norm": 0.6000717757856203,
      "learning_rate": 8.114261875265643e-06,
      "loss": 0.2452,
      "step": 140
    },
    {
      "epoch": 4.2727272727272725,
      "grad_norm": 0.5853791884587315,
      "learning_rate": 7.965439869473664e-06,
      "loss": 0.2424,
      "step": 141
    },
    {
      "epoch": 4.303030303030303,
      "grad_norm": 0.4949055144283988,
      "learning_rate": 7.817086484143987e-06,
      "loss": 0.2407,
      "step": 142
    },
    {
      "epoch": 4.333333333333333,
      "grad_norm": 0.5259656386577625,
      "learning_rate": 7.669235889528436e-06,
      "loss": 0.2553,
      "step": 143
    },
    {
      "epoch": 4.363636363636363,
      "grad_norm": 0.5616448847963217,
      "learning_rate": 7.521922140071003e-06,
      "loss": 0.2567,
      "step": 144
    },
    {
      "epoch": 4.393939393939394,
      "grad_norm": 0.5489620060684373,
      "learning_rate": 7.375179166564062e-06,
      "loss": 0.2698,
      "step": 145
    },
    {
      "epoch": 4.424242424242424,
      "grad_norm": 0.4951207309745249,
      "learning_rate": 7.2290407683331154e-06,
      "loss": 0.2349,
      "step": 146
    },
    {
      "epoch": 4.454545454545454,
      "grad_norm": 0.5222202349048461,
      "learning_rate": 7.0835406054517505e-06,
      "loss": 0.2563,
      "step": 147
    },
    {
      "epoch": 4.484848484848484,
      "grad_norm": 0.45445412967945614,
      "learning_rate": 6.93871219098876e-06,
      "loss": 0.2435,
      "step": 148
    },
    {
      "epoch": 4.515151515151516,
      "grad_norm": 0.4754985770528552,
      "learning_rate": 6.79458888328903e-06,
      "loss": 0.2597,
      "step": 149
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 0.4849712994258572,
      "learning_rate": 6.651203878290139e-06,
      "loss": 0.2666,
      "step": 150
    },
    {
      "epoch": 4.575757575757576,
      "grad_norm": 0.4957945447959049,
      "learning_rate": 6.508590201876317e-06,
      "loss": 0.2335,
      "step": 151
    },
    {
      "epoch": 4.606060606060606,
      "grad_norm": 0.4512488516416016,
      "learning_rate": 6.366780702271589e-06,
      "loss": 0.2543,
      "step": 152
    },
    {
      "epoch": 4.636363636363637,
      "grad_norm": 0.46904157755741865,
      "learning_rate": 6.225808042473857e-06,
      "loss": 0.2356,
      "step": 153
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.4457959701389976,
      "learning_rate": 6.085704692731609e-06,
      "loss": 0.2336,
      "step": 154
    },
    {
      "epoch": 4.696969696969697,
      "grad_norm": 0.442092095363142,
      "learning_rate": 5.946502923065054e-06,
      "loss": 0.2511,
      "step": 155
    },
    {
      "epoch": 4.7272727272727275,
      "grad_norm": 0.4086136541040712,
      "learning_rate": 5.8082347958333625e-06,
      "loss": 0.2338,
      "step": 156
    },
    {
      "epoch": 4.757575757575758,
      "grad_norm": 0.47809023852583177,
      "learning_rate": 5.670932158349732e-06,
      "loss": 0.2455,
      "step": 157
    },
    {
      "epoch": 4.787878787878788,
      "grad_norm": 0.4374689109271414,
      "learning_rate": 5.534626635546e-06,
      "loss": 0.2558,
      "step": 158
    },
    {
      "epoch": 4.818181818181818,
      "grad_norm": 0.40084984671932344,
      "learning_rate": 5.399349622688479e-06,
      "loss": 0.2554,
      "step": 159
    },
    {
      "epoch": 4.848484848484849,
      "grad_norm": 0.42775889553766333,
      "learning_rate": 5.2651322781466606e-06,
      "loss": 0.2599,
      "step": 160
    },
    {
      "epoch": 4.878787878787879,
      "grad_norm": 0.4304572599706753,
      "learning_rate": 5.132005516216512e-06,
      "loss": 0.2271,
      "step": 161
    },
    {
      "epoch": 4.909090909090909,
      "grad_norm": 0.4328092907068007,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.2534,
      "step": 162
    },
    {
      "epoch": 4.9393939393939394,
      "grad_norm": 0.449208190158122,
      "learning_rate": 4.869146134342426e-06,
      "loss": 0.2359,
      "step": 163
    },
    {
      "epoch": 4.96969696969697,
      "grad_norm": 0.4259587298923869,
      "learning_rate": 4.739474058829288e-06,
      "loss": 0.2268,
      "step": 164
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.4327989608606859,
      "learning_rate": 4.611013640844245e-06,
      "loss": 0.2712,
      "step": 165
    },
    {
      "epoch": 5.03030303030303,
      "grad_norm": 0.6440784714122533,
      "learning_rate": 4.483794468689728e-06,
      "loss": 0.1872,
      "step": 166
    },
    {
      "epoch": 5.0606060606060606,
      "grad_norm": 0.5213295903483506,
      "learning_rate": 4.357845844771881e-06,
      "loss": 0.1949,
      "step": 167
    },
    {
      "epoch": 5.090909090909091,
      "grad_norm": 0.4067288066504832,
      "learning_rate": 4.2331967788513295e-06,
      "loss": 0.1675,
      "step": 168
    },
    {
      "epoch": 5.121212121212121,
      "grad_norm": 0.4318314444875272,
      "learning_rate": 4.109875981361363e-06,
      "loss": 0.1808,
      "step": 169
    },
    {
      "epoch": 5.151515151515151,
      "grad_norm": 0.574155313358438,
      "learning_rate": 3.987911856795047e-06,
      "loss": 0.1851,
      "step": 170
    },
    {
      "epoch": 5.181818181818182,
      "grad_norm": 0.6134834556882526,
      "learning_rate": 3.867332497162836e-06,
      "loss": 0.1946,
      "step": 171
    },
    {
      "epoch": 5.212121212121212,
      "grad_norm": 0.5399437068959797,
      "learning_rate": 3.748165675522113e-06,
      "loss": 0.1838,
      "step": 172
    },
    {
      "epoch": 5.242424242424242,
      "grad_norm": 0.4671969345159338,
      "learning_rate": 3.630438839580217e-06,
      "loss": 0.1836,
      "step": 173
    },
    {
      "epoch": 5.2727272727272725,
      "grad_norm": 0.44783217311485224,
      "learning_rate": 3.5141791053724405e-06,
      "loss": 0.1712,
      "step": 174
    },
    {
      "epoch": 5.303030303030303,
      "grad_norm": 0.4760893085494367,
      "learning_rate": 3.399413251016359e-06,
      "loss": 0.1971,
      "step": 175
    },
    {
      "epoch": 5.333333333333333,
      "grad_norm": 0.5022635648150996,
      "learning_rate": 3.2861677105440335e-06,
      "loss": 0.1919,
      "step": 176
    },
    {
      "epoch": 5.363636363636363,
      "grad_norm": 0.4741245452055244,
      "learning_rate": 3.174468567813461e-06,
      "loss": 0.1981,
      "step": 177
    },
    {
      "epoch": 5.393939393939394,
      "grad_norm": 0.4483622903355323,
      "learning_rate": 3.0643415505006733e-06,
      "loss": 0.1746,
      "step": 178
    },
    {
      "epoch": 5.424242424242424,
      "grad_norm": 0.41418908963117185,
      "learning_rate": 2.9558120241738786e-06,
      "loss": 0.1657,
      "step": 179
    },
    {
      "epoch": 5.454545454545454,
      "grad_norm": 0.42895718735600913,
      "learning_rate": 2.8489049864510053e-06,
      "loss": 0.1837,
      "step": 180
    },
    {
      "epoch": 5.484848484848484,
      "grad_norm": 0.4149206360555145,
      "learning_rate": 2.7436450612420098e-06,
      "loss": 0.1754,
      "step": 181
    },
    {
      "epoch": 5.515151515151516,
      "grad_norm": 0.4509850773612303,
      "learning_rate": 2.640056493077231e-06,
      "loss": 0.1713,
      "step": 182
    },
    {
      "epoch": 5.545454545454545,
      "grad_norm": 0.44597682148041917,
      "learning_rate": 2.5381631415231455e-06,
      "loss": 0.1828,
      "step": 183
    },
    {
      "epoch": 5.575757575757576,
      "grad_norm": 0.5069691828136953,
      "learning_rate": 2.4379884756868167e-06,
      "loss": 0.1753,
      "step": 184
    },
    {
      "epoch": 5.606060606060606,
      "grad_norm": 0.4274251593097272,
      "learning_rate": 2.339555568810221e-06,
      "loss": 0.1759,
      "step": 185
    },
    {
      "epoch": 5.636363636363637,
      "grad_norm": 0.4091109248111021,
      "learning_rate": 2.2428870929558012e-06,
      "loss": 0.1964,
      "step": 186
    },
    {
      "epoch": 5.666666666666667,
      "grad_norm": 0.4251122344646706,
      "learning_rate": 2.1480053137844115e-06,
      "loss": 0.1716,
      "step": 187
    },
    {
      "epoch": 5.696969696969697,
      "grad_norm": 0.4168250191215947,
      "learning_rate": 2.054932085426856e-06,
      "loss": 0.1902,
      "step": 188
    },
    {
      "epoch": 5.7272727272727275,
      "grad_norm": 0.38407706882610726,
      "learning_rate": 1.963688845450218e-06,
      "loss": 0.203,
      "step": 189
    },
    {
      "epoch": 5.757575757575758,
      "grad_norm": 0.3766126413942047,
      "learning_rate": 1.8742966099201699e-06,
      "loss": 0.1927,
      "step": 190
    },
    {
      "epoch": 5.787878787878788,
      "grad_norm": 0.40200661424515893,
      "learning_rate": 1.7867759685603115e-06,
      "loss": 0.1914,
      "step": 191
    },
    {
      "epoch": 5.818181818181818,
      "grad_norm": 0.43031121559762775,
      "learning_rate": 1.7011470800097496e-06,
      "loss": 0.1934,
      "step": 192
    },
    {
      "epoch": 5.848484848484849,
      "grad_norm": 0.39649528343163404,
      "learning_rate": 1.6174296671799571e-06,
      "loss": 0.1763,
      "step": 193
    },
    {
      "epoch": 5.878787878787879,
      "grad_norm": 0.3962495956735567,
      "learning_rate": 1.5356430127119915e-06,
      "loss": 0.1909,
      "step": 194
    },
    {
      "epoch": 5.909090909090909,
      "grad_norm": 0.39697974568712374,
      "learning_rate": 1.4558059545351144e-06,
      "loss": 0.1701,
      "step": 195
    },
    {
      "epoch": 5.9393939393939394,
      "grad_norm": 0.39470211643952285,
      "learning_rate": 1.3779368815278648e-06,
      "loss": 0.1763,
      "step": 196
    },
    {
      "epoch": 5.96969696969697,
      "grad_norm": 0.3730537257405863,
      "learning_rate": 1.302053729282533e-06,
      "loss": 0.1856,
      "step": 197
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.3655877119736336,
      "learning_rate": 1.2281739759740575e-06,
      "loss": 0.1809,
      "step": 198
    },
    {
      "epoch": 6.03030303030303,
      "grad_norm": 0.46932465887461633,
      "learning_rate": 1.156314638334277e-06,
      "loss": 0.1586,
      "step": 199
    },
    {
      "epoch": 6.0606060606060606,
      "grad_norm": 0.42463062411423014,
      "learning_rate": 1.086492267732462e-06,
      "loss": 0.1486,
      "step": 200
    },
    {
      "epoch": 6.090909090909091,
      "grad_norm": 0.4204943708532176,
      "learning_rate": 1.01872294636304e-06,
      "loss": 0.1513,
      "step": 201
    },
    {
      "epoch": 6.121212121212121,
      "grad_norm": 0.36018726985462884,
      "learning_rate": 9.530222835413739e-07,
      "loss": 0.1599,
      "step": 202
    },
    {
      "epoch": 6.151515151515151,
      "grad_norm": 0.35373121166428373,
      "learning_rate": 8.894054121084839e-07,
      "loss": 0.1644,
      "step": 203
    },
    {
      "epoch": 6.181818181818182,
      "grad_norm": 0.35169271529725715,
      "learning_rate": 8.278869849454718e-07,
      "loss": 0.1494,
      "step": 204
    },
    {
      "epoch": 6.212121212121212,
      "grad_norm": 0.35811646468553465,
      "learning_rate": 7.684811715985429e-07,
      "loss": 0.138,
      "step": 205
    },
    {
      "epoch": 6.242424242424242,
      "grad_norm": 0.3639858571082851,
      "learning_rate": 7.1120165501533e-07,
      "loss": 0.1569,
      "step": 206
    },
    {
      "epoch": 6.2727272727272725,
      "grad_norm": 0.3743286463550599,
      "learning_rate": 6.560616283932897e-07,
      "loss": 0.1457,
      "step": 207
    },
    {
      "epoch": 6.303030303030303,
      "grad_norm": 0.4003397073869862,
      "learning_rate": 6.030737921409169e-07,
      "loss": 0.1504,
      "step": 208
    },
    {
      "epoch": 6.333333333333333,
      "grad_norm": 0.3885280268411254,
      "learning_rate": 5.522503509524591e-07,
      "loss": 0.1568,
      "step": 209
    },
    {
      "epoch": 6.363636363636363,
      "grad_norm": 0.4339815078343308,
      "learning_rate": 5.036030109968082e-07,
      "loss": 0.1602,
      "step": 210
    },
    {
      "epoch": 6.393939393939394,
      "grad_norm": 0.37273744998799907,
      "learning_rate": 4.5714297722121105e-07,
      "loss": 0.1599,
      "step": 211
    },
    {
      "epoch": 6.424242424242424,
      "grad_norm": 0.3884401499255415,
      "learning_rate": 4.128809507704445e-07,
      "loss": 0.1613,
      "step": 212
    },
    {
      "epoch": 6.454545454545454,
      "grad_norm": 0.39025373548396086,
      "learning_rate": 3.708271265220087e-07,
      "loss": 0.1446,
      "step": 213
    },
    {
      "epoch": 6.484848484848484,
      "grad_norm": 0.38038411719558896,
      "learning_rate": 3.309911907379393e-07,
      "loss": 0.1607,
      "step": 214
    },
    {
      "epoch": 6.515151515151516,
      "grad_norm": 0.36219407321675157,
      "learning_rate": 2.9338231883378365e-07,
      "loss": 0.1596,
      "step": 215
    },
    {
      "epoch": 6.545454545454545,
      "grad_norm": 0.357532621610846,
      "learning_rate": 2.5800917326521013e-07,
      "loss": 0.1514,
      "step": 216
    },
    {
      "epoch": 6.575757575757576,
      "grad_norm": 0.36156522997268226,
      "learning_rate": 2.248799015327907e-07,
      "loss": 0.1485,
      "step": 217
    },
    {
      "epoch": 6.606060606060606,
      "grad_norm": 0.35043349546858527,
      "learning_rate": 1.9400213430538773e-07,
      "loss": 0.1593,
      "step": 218
    },
    {
      "epoch": 6.636363636363637,
      "grad_norm": 0.35578087691685345,
      "learning_rate": 1.6538298366257975e-07,
      "loss": 0.151,
      "step": 219
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.3500065944748345,
      "learning_rate": 1.3902904145653094e-07,
      "loss": 0.1515,
      "step": 220
    },
    {
      "epoch": 6.696969696969697,
      "grad_norm": 0.3658140029174053,
      "learning_rate": 1.1494637779369766e-07,
      "loss": 0.1684,
      "step": 221
    },
    {
      "epoch": 6.7272727272727275,
      "grad_norm": 0.34724457037975026,
      "learning_rate": 9.314053963669245e-08,
      "loss": 0.1499,
      "step": 222
    },
    {
      "epoch": 6.757575757575758,
      "grad_norm": 0.3470543939923173,
      "learning_rate": 7.361654952665608e-08,
      "loss": 0.1611,
      "step": 223
    },
    {
      "epoch": 6.787878787878788,
      "grad_norm": 0.3491607886314148,
      "learning_rate": 5.637890442641403e-08,
      "loss": 0.1529,
      "step": 224
    },
    {
      "epoch": 6.818181818181818,
      "grad_norm": 0.3495528937906726,
      "learning_rate": 4.143157468468717e-08,
      "loss": 0.1524,
      "step": 225
    },
    {
      "epoch": 6.848484848484849,
      "grad_norm": 0.361077124122224,
      "learning_rate": 2.8778003121607834e-08,
      "loss": 0.1455,
      "step": 226
    },
    {
      "epoch": 6.878787878787879,
      "grad_norm": 0.3600044599803596,
      "learning_rate": 1.8421104235727406e-08,
      "loss": 0.1522,
      "step": 227
    },
    {
      "epoch": 6.909090909090909,
      "grad_norm": 0.3563659365013138,
      "learning_rate": 1.0363263532724433e-08,
      "loss": 0.1563,
      "step": 228
    },
    {
      "epoch": 6.9393939393939394,
      "grad_norm": 0.34423336022087025,
      "learning_rate": 4.606336975948589e-09,
      "loss": 0.1519,
      "step": 229
    },
    {
      "epoch": 6.96969696969697,
      "grad_norm": 0.3684818421630321,
      "learning_rate": 1.1516505589381777e-09,
      "loss": 0.1468,
      "step": 230
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.3622153271453113,
      "learning_rate": 0.0,
      "loss": 0.1541,
      "step": 231
    },
    {
      "epoch": 7.0,
      "step": 231,
      "total_flos": 3.4978494670281114e+17,
      "train_loss": 0.40140248718973875,
      "train_runtime": 6986.064,
      "train_samples_per_second": 3.166,
      "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": 3.4978494670281114e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}