{
  "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": 6.151421402706166,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.7998,
      "step": 1
    },
    {
      "epoch": 0.06060606060606061,
      "grad_norm": 6.237735357456026,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 0.8186,
      "step": 2
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 6.399283070753372,
      "learning_rate": 2.5e-06,
      "loss": 0.8187,
      "step": 3
    },
    {
      "epoch": 0.12121212121212122,
      "grad_norm": 5.759595140787336,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.8162,
      "step": 4
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 4.218610349318276,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.7538,
      "step": 5
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 2.288392339502158,
      "learning_rate": 5e-06,
      "loss": 0.7282,
      "step": 6
    },
    {
      "epoch": 0.21212121212121213,
      "grad_norm": 1.956752921975325,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.7132,
      "step": 7
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 3.6207898885642313,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.7113,
      "step": 8
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 4.025191746863598,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.7157,
      "step": 9
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 4.224907614185296,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.7245,
      "step": 10
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 3.864398058320154,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.6871,
      "step": 11
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 3.1180475313969445,
      "learning_rate": 1e-05,
      "loss": 0.6687,
      "step": 12
    },
    {
      "epoch": 0.3939393939393939,
      "grad_norm": 1.8961383494899933,
      "learning_rate": 1.0833333333333334e-05,
      "loss": 0.6603,
      "step": 13
    },
    {
      "epoch": 0.42424242424242425,
      "grad_norm": 1.579248318709875,
      "learning_rate": 1.1666666666666668e-05,
      "loss": 0.6206,
      "step": 14
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 1.7518819778439398,
      "learning_rate": 1.25e-05,
      "loss": 0.5991,
      "step": 15
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 1.4163661479522904,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.6135,
      "step": 16
    },
    {
      "epoch": 0.5151515151515151,
      "grad_norm": 0.9506906604196159,
      "learning_rate": 1.416666666666667e-05,
      "loss": 0.5988,
      "step": 17
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 0.9861887944080775,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.5839,
      "step": 18
    },
    {
      "epoch": 0.5757575757575758,
      "grad_norm": 0.8891403433298952,
      "learning_rate": 1.5833333333333333e-05,
      "loss": 0.576,
      "step": 19
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 0.9063819488681859,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.555,
      "step": 20
    },
    {
      "epoch": 0.6363636363636364,
      "grad_norm": 0.7343907815248991,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 0.5673,
      "step": 21
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.7204905208298097,
      "learning_rate": 1.8333333333333333e-05,
      "loss": 0.5472,
      "step": 22
    },
    {
      "epoch": 0.696969696969697,
      "grad_norm": 0.8919786890702169,
      "learning_rate": 1.916666666666667e-05,
      "loss": 0.565,
      "step": 23
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.6596792449367622,
      "learning_rate": 2e-05,
      "loss": 0.5334,
      "step": 24
    },
    {
      "epoch": 0.7575757575757576,
      "grad_norm": 0.7765534813514482,
      "learning_rate": 1.999884834944106e-05,
      "loss": 0.5497,
      "step": 25
    },
    {
      "epoch": 0.7878787878787878,
      "grad_norm": 0.9255785265202623,
      "learning_rate": 1.9995393663024054e-05,
      "loss": 0.557,
      "step": 26
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 0.6165457719339829,
      "learning_rate": 1.9989636736467278e-05,
      "loss": 0.5466,
      "step": 27
    },
    {
      "epoch": 0.8484848484848485,
      "grad_norm": 0.6730469961799309,
      "learning_rate": 1.9981578895764272e-05,
      "loss": 0.5217,
      "step": 28
    },
    {
      "epoch": 0.8787878787878788,
      "grad_norm": 0.5705578034571377,
      "learning_rate": 1.9971221996878395e-05,
      "loss": 0.5198,
      "step": 29
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.6244909826811817,
      "learning_rate": 1.9958568425315316e-05,
      "loss": 0.5208,
      "step": 30
    },
    {
      "epoch": 0.9393939393939394,
      "grad_norm": 0.502565756839221,
      "learning_rate": 1.9943621095573588e-05,
      "loss": 0.5136,
      "step": 31
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.5477752259868545,
      "learning_rate": 1.9926383450473344e-05,
      "loss": 0.4964,
      "step": 32
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.6075097423890373,
      "learning_rate": 1.9906859460363307e-05,
      "loss": 0.5148,
      "step": 33
    },
    {
      "epoch": 1.0303030303030303,
      "grad_norm": 0.4157826969679044,
      "learning_rate": 1.9885053622206305e-05,
      "loss": 0.4962,
      "step": 34
    },
    {
      "epoch": 1.0606060606060606,
      "grad_norm": 0.5241443563220115,
      "learning_rate": 1.986097095854347e-05,
      "loss": 0.4884,
      "step": 35
    },
    {
      "epoch": 1.0909090909090908,
      "grad_norm": 0.5007376523299666,
      "learning_rate": 1.9834617016337424e-05,
      "loss": 0.4809,
      "step": 36
    },
    {
      "epoch": 1.121212121212121,
      "grad_norm": 0.48022903853330573,
      "learning_rate": 1.9805997865694616e-05,
      "loss": 0.4919,
      "step": 37
    },
    {
      "epoch": 1.1515151515151516,
      "grad_norm": 0.5091703647218788,
      "learning_rate": 1.9775120098467212e-05,
      "loss": 0.4996,
      "step": 38
    },
    {
      "epoch": 1.1818181818181819,
      "grad_norm": 0.49844719482219074,
      "learning_rate": 1.9741990826734793e-05,
      "loss": 0.5006,
      "step": 39
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 0.4620938596174139,
      "learning_rate": 1.970661768116622e-05,
      "loss": 0.4941,
      "step": 40
    },
    {
      "epoch": 1.2424242424242424,
      "grad_norm": 0.4642081466020492,
      "learning_rate": 1.9669008809262064e-05,
      "loss": 0.4888,
      "step": 41
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 0.41752939490937213,
      "learning_rate": 1.9629172873477995e-05,
      "loss": 0.4838,
      "step": 42
    },
    {
      "epoch": 1.303030303030303,
      "grad_norm": 0.4592532116628272,
      "learning_rate": 1.9587119049229558e-05,
      "loss": 0.4745,
      "step": 43
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.36345227917664,
      "learning_rate": 1.954285702277879e-05,
      "loss": 0.4696,
      "step": 44
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.48994228690228747,
      "learning_rate": 1.9496396989003195e-05,
      "loss": 0.476,
      "step": 45
    },
    {
      "epoch": 1.393939393939394,
      "grad_norm": 0.40827778400579373,
      "learning_rate": 1.944774964904754e-05,
      "loss": 0.4626,
      "step": 46
    },
    {
      "epoch": 1.4242424242424243,
      "grad_norm": 0.4348999298734601,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 0.4744,
      "step": 47
    },
    {
      "epoch": 1.4545454545454546,
      "grad_norm": 0.406942585427329,
      "learning_rate": 1.9343938371606714e-05,
      "loss": 0.4736,
      "step": 48
    },
    {
      "epoch": 1.4848484848484849,
      "grad_norm": 0.428196464412865,
      "learning_rate": 1.9288798344984673e-05,
      "loss": 0.4701,
      "step": 49
    },
    {
      "epoch": 1.5151515151515151,
      "grad_norm": 0.33648544518508816,
      "learning_rate": 1.9231518828401458e-05,
      "loss": 0.4484,
      "step": 50
    },
    {
      "epoch": 1.5454545454545454,
      "grad_norm": 0.4321125716716293,
      "learning_rate": 1.917211301505453e-05,
      "loss": 0.4615,
      "step": 51
    },
    {
      "epoch": 1.5757575757575757,
      "grad_norm": 0.4799905829943536,
      "learning_rate": 1.911059458789152e-05,
      "loss": 0.4693,
      "step": 52
    },
    {
      "epoch": 1.606060606060606,
      "grad_norm": 0.3737715812752541,
      "learning_rate": 1.9046977716458627e-05,
      "loss": 0.4565,
      "step": 53
    },
    {
      "epoch": 1.6363636363636362,
      "grad_norm": 0.5075991030024265,
      "learning_rate": 1.8981277053636963e-05,
      "loss": 0.4558,
      "step": 54
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.33761564405769073,
      "learning_rate": 1.891350773226754e-05,
      "loss": 0.4445,
      "step": 55
    },
    {
      "epoch": 1.696969696969697,
      "grad_norm": 0.4894269837986009,
      "learning_rate": 1.8843685361665724e-05,
      "loss": 0.47,
      "step": 56
    },
    {
      "epoch": 1.7272727272727273,
      "grad_norm": 0.34225109412748267,
      "learning_rate": 1.8771826024025944e-05,
      "loss": 0.4624,
      "step": 57
    },
    {
      "epoch": 1.7575757575757576,
      "grad_norm": 0.38141480004074546,
      "learning_rate": 1.8697946270717468e-05,
      "loss": 0.4702,
      "step": 58
    },
    {
      "epoch": 1.7878787878787878,
      "grad_norm": 0.3662141474152526,
      "learning_rate": 1.8622063118472135e-05,
      "loss": 0.4515,
      "step": 59
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.378056072731251,
      "learning_rate": 1.8544194045464888e-05,
      "loss": 0.446,
      "step": 60
    },
    {
      "epoch": 1.8484848484848486,
      "grad_norm": 0.3354669449273175,
      "learning_rate": 1.8464356987288012e-05,
      "loss": 0.4533,
      "step": 61
    },
    {
      "epoch": 1.878787878787879,
      "grad_norm": 0.36515031931175357,
      "learning_rate": 1.8382570332820045e-05,
      "loss": 0.4545,
      "step": 62
    },
    {
      "epoch": 1.9090909090909092,
      "grad_norm": 0.33953063703732606,
      "learning_rate": 1.8298852919990254e-05,
      "loss": 0.4476,
      "step": 63
    },
    {
      "epoch": 1.9393939393939394,
      "grad_norm": 0.34154245914908793,
      "learning_rate": 1.821322403143969e-05,
      "loss": 0.455,
      "step": 64
    },
    {
      "epoch": 1.9696969696969697,
      "grad_norm": 0.35681534199639064,
      "learning_rate": 1.812570339007983e-05,
      "loss": 0.4405,
      "step": 65
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.3557302042867996,
      "learning_rate": 1.8036311154549783e-05,
      "loss": 0.4495,
      "step": 66
    },
    {
      "epoch": 2.0303030303030303,
      "grad_norm": 0.4313228876696512,
      "learning_rate": 1.7945067914573147e-05,
      "loss": 0.4151,
      "step": 67
    },
    {
      "epoch": 2.0606060606060606,
      "grad_norm": 0.39636883963516584,
      "learning_rate": 1.7851994686215592e-05,
      "loss": 0.4031,
      "step": 68
    },
    {
      "epoch": 2.090909090909091,
      "grad_norm": 0.3845619057177181,
      "learning_rate": 1.77571129070442e-05,
      "loss": 0.4141,
      "step": 69
    },
    {
      "epoch": 2.121212121212121,
      "grad_norm": 0.5153948126182881,
      "learning_rate": 1.766044443118978e-05,
      "loss": 0.4177,
      "step": 70
    },
    {
      "epoch": 2.1515151515151514,
      "grad_norm": 0.42410031473121124,
      "learning_rate": 1.7562011524313187e-05,
      "loss": 0.4096,
      "step": 71
    },
    {
      "epoch": 2.1818181818181817,
      "grad_norm": 0.5035844255487422,
      "learning_rate": 1.7461836858476858e-05,
      "loss": 0.4116,
      "step": 72
    },
    {
      "epoch": 2.212121212121212,
      "grad_norm": 0.4236441837203945,
      "learning_rate": 1.7359943506922775e-05,
      "loss": 0.3974,
      "step": 73
    },
    {
      "epoch": 2.242424242424242,
      "grad_norm": 0.4101595523864697,
      "learning_rate": 1.725635493875799e-05,
      "loss": 0.4124,
      "step": 74
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.4721485058084954,
      "learning_rate": 1.7151095013548996e-05,
      "loss": 0.4082,
      "step": 75
    },
    {
      "epoch": 2.303030303030303,
      "grad_norm": 0.4761526244186826,
      "learning_rate": 1.7044187975826126e-05,
      "loss": 0.418,
      "step": 76
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.41216573070732854,
      "learning_rate": 1.693565844949933e-05,
      "loss": 0.396,
      "step": 77
    },
    {
      "epoch": 2.3636363636363638,
      "grad_norm": 0.474040525484892,
      "learning_rate": 1.6825531432186545e-05,
      "loss": 0.41,
      "step": 78
    },
    {
      "epoch": 2.393939393939394,
      "grad_norm": 0.40414908877842,
      "learning_rate": 1.671383228945597e-05,
      "loss": 0.4008,
      "step": 79
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 0.4146014048772213,
      "learning_rate": 1.6600586748983642e-05,
      "loss": 0.3971,
      "step": 80
    },
    {
      "epoch": 2.4545454545454546,
      "grad_norm": 0.3624299424037976,
      "learning_rate": 1.648582089462756e-05,
      "loss": 0.3935,
      "step": 81
    },
    {
      "epoch": 2.484848484848485,
      "grad_norm": 0.4168075058440944,
      "learning_rate": 1.6369561160419783e-05,
      "loss": 0.4112,
      "step": 82
    },
    {
      "epoch": 2.515151515151515,
      "grad_norm": 0.47398991064221935,
      "learning_rate": 1.625183432447789e-05,
      "loss": 0.4092,
      "step": 83
    },
    {
      "epoch": 2.5454545454545454,
      "grad_norm": 0.40347095969699964,
      "learning_rate": 1.6132667502837164e-05,
      "loss": 0.4165,
      "step": 84
    },
    {
      "epoch": 2.5757575757575757,
      "grad_norm": 0.3557086750971042,
      "learning_rate": 1.6012088143204953e-05,
      "loss": 0.4045,
      "step": 85
    },
    {
      "epoch": 2.606060606060606,
      "grad_norm": 0.415076932935183,
      "learning_rate": 1.589012401863864e-05,
      "loss": 0.4003,
      "step": 86
    },
    {
      "epoch": 2.6363636363636362,
      "grad_norm": 0.3737327671495727,
      "learning_rate": 1.5766803221148676e-05,
      "loss": 0.3896,
      "step": 87
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.49349366802821937,
      "learning_rate": 1.5642154155228124e-05,
      "loss": 0.4163,
      "step": 88
    },
    {
      "epoch": 2.6969696969696972,
      "grad_norm": 0.33074390100988704,
      "learning_rate": 1.5516205531310272e-05,
      "loss": 0.4167,
      "step": 89
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.39422229006862003,
      "learning_rate": 1.538898635915576e-05,
      "loss": 0.4012,
      "step": 90
    },
    {
      "epoch": 2.757575757575758,
      "grad_norm": 0.3939338183410774,
      "learning_rate": 1.526052594117071e-05,
      "loss": 0.4141,
      "step": 91
    },
    {
      "epoch": 2.787878787878788,
      "grad_norm": 0.3692298796321192,
      "learning_rate": 1.513085386565758e-05,
      "loss": 0.4046,
      "step": 92
    },
    {
      "epoch": 2.8181818181818183,
      "grad_norm": 0.3440273759316456,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.3957,
      "step": 93
    },
    {
      "epoch": 2.8484848484848486,
      "grad_norm": 0.3458608176784509,
      "learning_rate": 1.4867994483783485e-05,
      "loss": 0.4035,
      "step": 94
    },
    {
      "epoch": 2.878787878787879,
      "grad_norm": 0.3732245764603725,
      "learning_rate": 1.4734867721853341e-05,
      "loss": 0.397,
      "step": 95
    },
    {
      "epoch": 2.909090909090909,
      "grad_norm": 0.3684184338596461,
      "learning_rate": 1.4600650377311523e-05,
      "loss": 0.4022,
      "step": 96
    },
    {
      "epoch": 2.9393939393939394,
      "grad_norm": 0.38431691405769114,
      "learning_rate": 1.4465373364454001e-05,
      "loss": 0.4015,
      "step": 97
    },
    {
      "epoch": 2.9696969696969697,
      "grad_norm": 0.3479261992677546,
      "learning_rate": 1.4329067841650274e-05,
      "loss": 0.3894,
      "step": 98
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.36150361244581664,
      "learning_rate": 1.4191765204166643e-05,
      "loss": 0.3946,
      "step": 99
    },
    {
      "epoch": 3.0303030303030303,
      "grad_norm": 0.4442679395592465,
      "learning_rate": 1.4053497076934948e-05,
      "loss": 0.373,
      "step": 100
    },
    {
      "epoch": 3.0606060606060606,
      "grad_norm": 0.3628146695830824,
      "learning_rate": 1.3914295307268396e-05,
      "loss": 0.353,
      "step": 101
    },
    {
      "epoch": 3.090909090909091,
      "grad_norm": 0.4531727956128449,
      "learning_rate": 1.3774191957526144e-05,
      "loss": 0.3676,
      "step": 102
    },
    {
      "epoch": 3.121212121212121,
      "grad_norm": 0.35221087152357233,
      "learning_rate": 1.3633219297728415e-05,
      "loss": 0.3528,
      "step": 103
    },
    {
      "epoch": 3.1515151515151514,
      "grad_norm": 0.4273364673141065,
      "learning_rate": 1.3491409798123687e-05,
      "loss": 0.3465,
      "step": 104
    },
    {
      "epoch": 3.1818181818181817,
      "grad_norm": 0.35895019056256633,
      "learning_rate": 1.3348796121709862e-05,
      "loss": 0.3617,
      "step": 105
    },
    {
      "epoch": 3.212121212121212,
      "grad_norm": 0.4299571084726519,
      "learning_rate": 1.3205411116710973e-05,
      "loss": 0.3566,
      "step": 106
    },
    {
      "epoch": 3.242424242424242,
      "grad_norm": 0.37012229987430967,
      "learning_rate": 1.3061287809011243e-05,
      "loss": 0.3569,
      "step": 107
    },
    {
      "epoch": 3.2727272727272725,
      "grad_norm": 0.4228821755548108,
      "learning_rate": 1.291645939454825e-05,
      "loss": 0.3534,
      "step": 108
    },
    {
      "epoch": 3.303030303030303,
      "grad_norm": 0.38901074618569526,
      "learning_rate": 1.277095923166689e-05,
      "loss": 0.3521,
      "step": 109
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.41850393250190165,
      "learning_rate": 1.2624820833435939e-05,
      "loss": 0.3532,
      "step": 110
    },
    {
      "epoch": 3.3636363636363638,
      "grad_norm": 0.4071425765346372,
      "learning_rate": 1.2478077859929e-05,
      "loss": 0.3552,
      "step": 111
    },
    {
      "epoch": 3.393939393939394,
      "grad_norm": 0.44585089951747325,
      "learning_rate": 1.2330764110471567e-05,
      "loss": 0.3482,
      "step": 112
    },
    {
      "epoch": 3.4242424242424243,
      "grad_norm": 0.3498613175154098,
      "learning_rate": 1.2182913515856016e-05,
      "loss": 0.3553,
      "step": 113
    },
    {
      "epoch": 3.4545454545454546,
      "grad_norm": 0.4421138588736285,
      "learning_rate": 1.2034560130526341e-05,
      "loss": 0.3702,
      "step": 114
    },
    {
      "epoch": 3.484848484848485,
      "grad_norm": 0.37129944371711043,
      "learning_rate": 1.1885738124734359e-05,
      "loss": 0.3518,
      "step": 115
    },
    {
      "epoch": 3.515151515151515,
      "grad_norm": 0.33350435392683414,
      "learning_rate": 1.1736481776669307e-05,
      "loss": 0.3485,
      "step": 116
    },
    {
      "epoch": 3.5454545454545454,
      "grad_norm": 0.37720568554538864,
      "learning_rate": 1.1586825464562515e-05,
      "loss": 0.3598,
      "step": 117
    },
    {
      "epoch": 3.5757575757575757,
      "grad_norm": 0.3057940091518938,
      "learning_rate": 1.1436803658769082e-05,
      "loss": 0.3677,
      "step": 118
    },
    {
      "epoch": 3.606060606060606,
      "grad_norm": 0.34257414118199886,
      "learning_rate": 1.1286450913828313e-05,
      "loss": 0.3575,
      "step": 119
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.30350348272535654,
      "learning_rate": 1.113580186050475e-05,
      "loss": 0.3508,
      "step": 120
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.2936155582176901,
      "learning_rate": 1.0984891197811686e-05,
      "loss": 0.3497,
      "step": 121
    },
    {
      "epoch": 3.6969696969696972,
      "grad_norm": 0.3065122902676449,
      "learning_rate": 1.0833753685018935e-05,
      "loss": 0.353,
      "step": 122
    },
    {
      "epoch": 3.7272727272727275,
      "grad_norm": 0.29671648483026797,
      "learning_rate": 1.0682424133646712e-05,
      "loss": 0.347,
      "step": 123
    },
    {
      "epoch": 3.757575757575758,
      "grad_norm": 0.3225797453481725,
      "learning_rate": 1.0530937399447496e-05,
      "loss": 0.3428,
      "step": 124
    },
    {
      "epoch": 3.787878787878788,
      "grad_norm": 0.30679168072111707,
      "learning_rate": 1.0379328374377715e-05,
      "loss": 0.3578,
      "step": 125
    },
    {
      "epoch": 3.8181818181818183,
      "grad_norm": 0.32321827283917637,
      "learning_rate": 1.0227631978561057e-05,
      "loss": 0.3453,
      "step": 126
    },
    {
      "epoch": 3.8484848484848486,
      "grad_norm": 0.3864236941699214,
      "learning_rate": 1.0075883152245334e-05,
      "loss": 0.3486,
      "step": 127
    },
    {
      "epoch": 3.878787878787879,
      "grad_norm": 0.3447204983973778,
      "learning_rate": 9.92411684775467e-06,
      "loss": 0.3559,
      "step": 128
    },
    {
      "epoch": 3.909090909090909,
      "grad_norm": 0.3289693834917111,
      "learning_rate": 9.772368021438943e-06,
      "loss": 0.3465,
      "step": 129
    },
    {
      "epoch": 3.9393939393939394,
      "grad_norm": 0.36322472159775204,
      "learning_rate": 9.620671625622287e-06,
      "loss": 0.3523,
      "step": 130
    },
    {
      "epoch": 3.9696969696969697,
      "grad_norm": 0.2978087468646819,
      "learning_rate": 9.469062600552509e-06,
      "loss": 0.3611,
      "step": 131
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.3156051152104095,
      "learning_rate": 9.317575866353293e-06,
      "loss": 0.3538,
      "step": 132
    },
    {
      "epoch": 4.03030303030303,
      "grad_norm": 0.41521005976798286,
      "learning_rate": 9.166246314981066e-06,
      "loss": 0.3238,
      "step": 133
    },
    {
      "epoch": 4.0606060606060606,
      "grad_norm": 0.3137260699741458,
      "learning_rate": 9.015108802188314e-06,
      "loss": 0.303,
      "step": 134
    },
    {
      "epoch": 4.090909090909091,
      "grad_norm": 0.40638571097134396,
      "learning_rate": 8.86419813949525e-06,
      "loss": 0.3082,
      "step": 135
    },
    {
      "epoch": 4.121212121212121,
      "grad_norm": 0.4098742663452661,
      "learning_rate": 8.71354908617169e-06,
      "loss": 0.3063,
      "step": 136
    },
    {
      "epoch": 4.151515151515151,
      "grad_norm": 0.39465773619929556,
      "learning_rate": 8.56319634123092e-06,
      "loss": 0.3133,
      "step": 137
    },
    {
      "epoch": 4.181818181818182,
      "grad_norm": 0.4428632875359087,
      "learning_rate": 8.413174535437486e-06,
      "loss": 0.31,
      "step": 138
    },
    {
      "epoch": 4.212121212121212,
      "grad_norm": 0.35458304653653905,
      "learning_rate": 8.263518223330698e-06,
      "loss": 0.2998,
      "step": 139
    },
    {
      "epoch": 4.242424242424242,
      "grad_norm": 0.3872277898042097,
      "learning_rate": 8.114261875265643e-06,
      "loss": 0.3152,
      "step": 140
    },
    {
      "epoch": 4.2727272727272725,
      "grad_norm": 0.39453005895293025,
      "learning_rate": 7.965439869473664e-06,
      "loss": 0.3203,
      "step": 141
    },
    {
      "epoch": 4.303030303030303,
      "grad_norm": 0.32784812316437645,
      "learning_rate": 7.817086484143987e-06,
      "loss": 0.3058,
      "step": 142
    },
    {
      "epoch": 4.333333333333333,
      "grad_norm": 0.35449516529956865,
      "learning_rate": 7.669235889528436e-06,
      "loss": 0.2906,
      "step": 143
    },
    {
      "epoch": 4.363636363636363,
      "grad_norm": 0.36693212185087015,
      "learning_rate": 7.521922140071003e-06,
      "loss": 0.3169,
      "step": 144
    },
    {
      "epoch": 4.393939393939394,
      "grad_norm": 0.30558336634663924,
      "learning_rate": 7.375179166564062e-06,
      "loss": 0.3017,
      "step": 145
    },
    {
      "epoch": 4.424242424242424,
      "grad_norm": 0.3417662137519071,
      "learning_rate": 7.2290407683331154e-06,
      "loss": 0.3022,
      "step": 146
    },
    {
      "epoch": 4.454545454545454,
      "grad_norm": 0.32702244113746887,
      "learning_rate": 7.0835406054517505e-06,
      "loss": 0.3192,
      "step": 147
    },
    {
      "epoch": 4.484848484848484,
      "grad_norm": 0.30083332294153053,
      "learning_rate": 6.93871219098876e-06,
      "loss": 0.3138,
      "step": 148
    },
    {
      "epoch": 4.515151515151516,
      "grad_norm": 0.30354496003888176,
      "learning_rate": 6.79458888328903e-06,
      "loss": 0.3081,
      "step": 149
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 0.32230083799564085,
      "learning_rate": 6.651203878290139e-06,
      "loss": 0.3105,
      "step": 150
    },
    {
      "epoch": 4.575757575757576,
      "grad_norm": 0.3060650930378369,
      "learning_rate": 6.508590201876317e-06,
      "loss": 0.3093,
      "step": 151
    },
    {
      "epoch": 4.606060606060606,
      "grad_norm": 0.2741744452769089,
      "learning_rate": 6.366780702271589e-06,
      "loss": 0.318,
      "step": 152
    },
    {
      "epoch": 4.636363636363637,
      "grad_norm": 0.25032587402962564,
      "learning_rate": 6.225808042473857e-06,
      "loss": 0.3061,
      "step": 153
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.3247430534867803,
      "learning_rate": 6.085704692731609e-06,
      "loss": 0.322,
      "step": 154
    },
    {
      "epoch": 4.696969696969697,
      "grad_norm": 0.2798354730102713,
      "learning_rate": 5.946502923065054e-06,
      "loss": 0.2973,
      "step": 155
    },
    {
      "epoch": 4.7272727272727275,
      "grad_norm": 0.3027303442803266,
      "learning_rate": 5.8082347958333625e-06,
      "loss": 0.316,
      "step": 156
    },
    {
      "epoch": 4.757575757575758,
      "grad_norm": 0.30488175262561684,
      "learning_rate": 5.670932158349732e-06,
      "loss": 0.3107,
      "step": 157
    },
    {
      "epoch": 4.787878787878788,
      "grad_norm": 0.27412556416733463,
      "learning_rate": 5.534626635546e-06,
      "loss": 0.3193,
      "step": 158
    },
    {
      "epoch": 4.818181818181818,
      "grad_norm": 0.29034152939008845,
      "learning_rate": 5.399349622688479e-06,
      "loss": 0.3102,
      "step": 159
    },
    {
      "epoch": 4.848484848484849,
      "grad_norm": 0.2720838047760481,
      "learning_rate": 5.2651322781466606e-06,
      "loss": 0.3186,
      "step": 160
    },
    {
      "epoch": 4.878787878787879,
      "grad_norm": 0.2782632381130198,
      "learning_rate": 5.132005516216512e-06,
      "loss": 0.3066,
      "step": 161
    },
    {
      "epoch": 4.909090909090909,
      "grad_norm": 0.2713858719228216,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.3111,
      "step": 162
    },
    {
      "epoch": 4.9393939393939394,
      "grad_norm": 0.25933497877965916,
      "learning_rate": 4.869146134342426e-06,
      "loss": 0.3081,
      "step": 163
    },
    {
      "epoch": 4.96969696969697,
      "grad_norm": 0.26317448361244816,
      "learning_rate": 4.739474058829288e-06,
      "loss": 0.2974,
      "step": 164
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.2786833041082574,
      "learning_rate": 4.611013640844245e-06,
      "loss": 0.2986,
      "step": 165
    },
    {
      "epoch": 5.03030303030303,
      "grad_norm": 0.34889607160623665,
      "learning_rate": 4.483794468689728e-06,
      "loss": 0.2747,
      "step": 166
    },
    {
      "epoch": 5.0606060606060606,
      "grad_norm": 0.2917693211777931,
      "learning_rate": 4.357845844771881e-06,
      "loss": 0.2775,
      "step": 167
    },
    {
      "epoch": 5.090909090909091,
      "grad_norm": 0.25274948875446884,
      "learning_rate": 4.2331967788513295e-06,
      "loss": 0.273,
      "step": 168
    },
    {
      "epoch": 5.121212121212121,
      "grad_norm": 0.2971920066426912,
      "learning_rate": 4.109875981361363e-06,
      "loss": 0.2789,
      "step": 169
    },
    {
      "epoch": 5.151515151515151,
      "grad_norm": 0.3897761744728906,
      "learning_rate": 3.987911856795047e-06,
      "loss": 0.2952,
      "step": 170
    },
    {
      "epoch": 5.181818181818182,
      "grad_norm": 0.2808252022207057,
      "learning_rate": 3.867332497162836e-06,
      "loss": 0.2726,
      "step": 171
    },
    {
      "epoch": 5.212121212121212,
      "grad_norm": 0.2886653672913755,
      "learning_rate": 3.748165675522113e-06,
      "loss": 0.2811,
      "step": 172
    },
    {
      "epoch": 5.242424242424242,
      "grad_norm": 0.30649821582422276,
      "learning_rate": 3.630438839580217e-06,
      "loss": 0.2844,
      "step": 173
    },
    {
      "epoch": 5.2727272727272725,
      "grad_norm": 0.266335957166941,
      "learning_rate": 3.5141791053724405e-06,
      "loss": 0.2761,
      "step": 174
    },
    {
      "epoch": 5.303030303030303,
      "grad_norm": 0.2624133974035284,
      "learning_rate": 3.399413251016359e-06,
      "loss": 0.2772,
      "step": 175
    },
    {
      "epoch": 5.333333333333333,
      "grad_norm": 0.28043984636902863,
      "learning_rate": 3.2861677105440335e-06,
      "loss": 0.279,
      "step": 176
    },
    {
      "epoch": 5.363636363636363,
      "grad_norm": 0.2844999612712147,
      "learning_rate": 3.174468567813461e-06,
      "loss": 0.276,
      "step": 177
    },
    {
      "epoch": 5.393939393939394,
      "grad_norm": 0.30274480640982454,
      "learning_rate": 3.0643415505006733e-06,
      "loss": 0.2844,
      "step": 178
    },
    {
      "epoch": 5.424242424242424,
      "grad_norm": 0.26964532079225034,
      "learning_rate": 2.9558120241738786e-06,
      "loss": 0.2729,
      "step": 179
    },
    {
      "epoch": 5.454545454545454,
      "grad_norm": 0.2558873145763057,
      "learning_rate": 2.8489049864510053e-06,
      "loss": 0.2751,
      "step": 180
    },
    {
      "epoch": 5.484848484848484,
      "grad_norm": 0.26603815939850806,
      "learning_rate": 2.7436450612420098e-06,
      "loss": 0.2836,
      "step": 181
    },
    {
      "epoch": 5.515151515151516,
      "grad_norm": 0.25162264845600446,
      "learning_rate": 2.640056493077231e-06,
      "loss": 0.2745,
      "step": 182
    },
    {
      "epoch": 5.545454545454545,
      "grad_norm": 0.25314063618339583,
      "learning_rate": 2.5381631415231455e-06,
      "loss": 0.2697,
      "step": 183
    },
    {
      "epoch": 5.575757575757576,
      "grad_norm": 0.24334307169277472,
      "learning_rate": 2.4379884756868167e-06,
      "loss": 0.2786,
      "step": 184
    },
    {
      "epoch": 5.606060606060606,
      "grad_norm": 0.23953331746205406,
      "learning_rate": 2.339555568810221e-06,
      "loss": 0.2702,
      "step": 185
    },
    {
      "epoch": 5.636363636363637,
      "grad_norm": 0.23340459550911002,
      "learning_rate": 2.2428870929558012e-06,
      "loss": 0.2687,
      "step": 186
    },
    {
      "epoch": 5.666666666666667,
      "grad_norm": 0.240436605273711,
      "learning_rate": 2.1480053137844115e-06,
      "loss": 0.2806,
      "step": 187
    },
    {
      "epoch": 5.696969696969697,
      "grad_norm": 0.23603381903331488,
      "learning_rate": 2.054932085426856e-06,
      "loss": 0.2738,
      "step": 188
    },
    {
      "epoch": 5.7272727272727275,
      "grad_norm": 0.2393826481228628,
      "learning_rate": 1.963688845450218e-06,
      "loss": 0.2809,
      "step": 189
    },
    {
      "epoch": 5.757575757575758,
      "grad_norm": 0.23585837806283824,
      "learning_rate": 1.8742966099201699e-06,
      "loss": 0.2712,
      "step": 190
    },
    {
      "epoch": 5.787878787878788,
      "grad_norm": 0.23414628338304533,
      "learning_rate": 1.7867759685603115e-06,
      "loss": 0.2832,
      "step": 191
    },
    {
      "epoch": 5.818181818181818,
      "grad_norm": 0.25646584558495067,
      "learning_rate": 1.7011470800097496e-06,
      "loss": 0.2794,
      "step": 192
    },
    {
      "epoch": 5.848484848484849,
      "grad_norm": 0.21951214477534955,
      "learning_rate": 1.6174296671799571e-06,
      "loss": 0.2751,
      "step": 193
    },
    {
      "epoch": 5.878787878787879,
      "grad_norm": 0.2228098678984023,
      "learning_rate": 1.5356430127119915e-06,
      "loss": 0.2698,
      "step": 194
    },
    {
      "epoch": 5.909090909090909,
      "grad_norm": 0.23975068101149252,
      "learning_rate": 1.4558059545351144e-06,
      "loss": 0.2773,
      "step": 195
    },
    {
      "epoch": 5.9393939393939394,
      "grad_norm": 0.23260124182213104,
      "learning_rate": 1.3779368815278648e-06,
      "loss": 0.2839,
      "step": 196
    },
    {
      "epoch": 5.96969696969697,
      "grad_norm": 0.22885848304589662,
      "learning_rate": 1.302053729282533e-06,
      "loss": 0.2804,
      "step": 197
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.2209029525053189,
      "learning_rate": 1.2281739759740575e-06,
      "loss": 0.2733,
      "step": 198
    },
    {
      "epoch": 6.03030303030303,
      "grad_norm": 0.2798330521366615,
      "learning_rate": 1.156314638334277e-06,
      "loss": 0.2654,
      "step": 199
    },
    {
      "epoch": 6.0606060606060606,
      "grad_norm": 0.24220094078655102,
      "learning_rate": 1.086492267732462e-06,
      "loss": 0.2582,
      "step": 200
    },
    {
      "epoch": 6.090909090909091,
      "grad_norm": 0.24911212920851225,
      "learning_rate": 1.01872294636304e-06,
      "loss": 0.257,
      "step": 201
    },
    {
      "epoch": 6.121212121212121,
      "grad_norm": 0.22376083440951192,
      "learning_rate": 9.530222835413739e-07,
      "loss": 0.2546,
      "step": 202
    },
    {
      "epoch": 6.151515151515151,
      "grad_norm": 0.22447225262876716,
      "learning_rate": 8.894054121084839e-07,
      "loss": 0.2568,
      "step": 203
    },
    {
      "epoch": 6.181818181818182,
      "grad_norm": 0.2228866645644093,
      "learning_rate": 8.278869849454718e-07,
      "loss": 0.2561,
      "step": 204
    },
    {
      "epoch": 6.212121212121212,
      "grad_norm": 0.24171597997329175,
      "learning_rate": 7.684811715985429e-07,
      "loss": 0.2544,
      "step": 205
    },
    {
      "epoch": 6.242424242424242,
      "grad_norm": 0.2594047028493119,
      "learning_rate": 7.1120165501533e-07,
      "loss": 0.2657,
      "step": 206
    },
    {
      "epoch": 6.2727272727272725,
      "grad_norm": 0.23483901334260437,
      "learning_rate": 6.560616283932897e-07,
      "loss": 0.2615,
      "step": 207
    },
    {
      "epoch": 6.303030303030303,
      "grad_norm": 0.2448722498623381,
      "learning_rate": 6.030737921409169e-07,
      "loss": 0.2602,
      "step": 208
    },
    {
      "epoch": 6.333333333333333,
      "grad_norm": 0.23934720127394496,
      "learning_rate": 5.522503509524591e-07,
      "loss": 0.2593,
      "step": 209
    },
    {
      "epoch": 6.363636363636363,
      "grad_norm": 0.24211507943580907,
      "learning_rate": 5.036030109968082e-07,
      "loss": 0.271,
      "step": 210
    },
    {
      "epoch": 6.393939393939394,
      "grad_norm": 0.2201760229724123,
      "learning_rate": 4.5714297722121105e-07,
      "loss": 0.2652,
      "step": 211
    },
    {
      "epoch": 6.424242424242424,
      "grad_norm": 0.21518165421575836,
      "learning_rate": 4.128809507704445e-07,
      "loss": 0.2657,
      "step": 212
    },
    {
      "epoch": 6.454545454545454,
      "grad_norm": 0.21509575521045643,
      "learning_rate": 3.708271265220087e-07,
      "loss": 0.2642,
      "step": 213
    },
    {
      "epoch": 6.484848484848484,
      "grad_norm": 0.23495249631178464,
      "learning_rate": 3.309911907379393e-07,
      "loss": 0.2736,
      "step": 214
    },
    {
      "epoch": 6.515151515151516,
      "grad_norm": 0.21090198006069721,
      "learning_rate": 2.9338231883378365e-07,
      "loss": 0.2563,
      "step": 215
    },
    {
      "epoch": 6.545454545454545,
      "grad_norm": 0.226384758620243,
      "learning_rate": 2.5800917326521013e-07,
      "loss": 0.2627,
      "step": 216
    },
    {
      "epoch": 6.575757575757576,
      "grad_norm": 0.22083017422244458,
      "learning_rate": 2.248799015327907e-07,
      "loss": 0.2523,
      "step": 217
    },
    {
      "epoch": 6.606060606060606,
      "grad_norm": 0.2213528960557884,
      "learning_rate": 1.9400213430538773e-07,
      "loss": 0.2632,
      "step": 218
    },
    {
      "epoch": 6.636363636363637,
      "grad_norm": 0.21492208877373387,
      "learning_rate": 1.6538298366257975e-07,
      "loss": 0.2585,
      "step": 219
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.21958064699470892,
      "learning_rate": 1.3902904145653094e-07,
      "loss": 0.2591,
      "step": 220
    },
    {
      "epoch": 6.696969696969697,
      "grad_norm": 0.21386216720479398,
      "learning_rate": 1.1494637779369766e-07,
      "loss": 0.2585,
      "step": 221
    },
    {
      "epoch": 6.7272727272727275,
      "grad_norm": 0.21349457269575486,
      "learning_rate": 9.314053963669245e-08,
      "loss": 0.2706,
      "step": 222
    },
    {
      "epoch": 6.757575757575758,
      "grad_norm": 0.2166601846916257,
      "learning_rate": 7.361654952665608e-08,
      "loss": 0.2694,
      "step": 223
    },
    {
      "epoch": 6.787878787878788,
      "grad_norm": 0.21162380851857757,
      "learning_rate": 5.637890442641403e-08,
      "loss": 0.2627,
      "step": 224
    },
    {
      "epoch": 6.818181818181818,
      "grad_norm": 0.21501451315368403,
      "learning_rate": 4.143157468468717e-08,
      "loss": 0.2682,
      "step": 225
    },
    {
      "epoch": 6.848484848484849,
      "grad_norm": 0.21131935974668115,
      "learning_rate": 2.8778003121607834e-08,
      "loss": 0.2613,
      "step": 226
    },
    {
      "epoch": 6.878787878787879,
      "grad_norm": 0.22082179540522404,
      "learning_rate": 1.8421104235727406e-08,
      "loss": 0.2633,
      "step": 227
    },
    {
      "epoch": 6.909090909090909,
      "grad_norm": 0.2100106661850552,
      "learning_rate": 1.0363263532724433e-08,
      "loss": 0.2623,
      "step": 228
    },
    {
      "epoch": 6.9393939393939394,
      "grad_norm": 0.21210969956284953,
      "learning_rate": 4.606336975948589e-09,
      "loss": 0.2626,
      "step": 229
    },
    {
      "epoch": 6.96969696969697,
      "grad_norm": 0.21136927249714363,
      "learning_rate": 1.1516505589381777e-09,
      "loss": 0.2653,
      "step": 230
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.21462942924377515,
      "learning_rate": 0.0,
      "loss": 0.2624,
      "step": 231
    },
    {
      "epoch": 7.0,
      "step": 231,
      "total_flos": 7.881011783731773e+17,
      "train_loss": 0.3864664324434289,
      "train_runtime": 8865.023,
      "train_samples_per_second": 2.495,
      "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": 7.881011783731773e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}