{
  "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": 6.391723498659803,
      "learning_rate": 8.333333333333333e-07,
      "loss": 1.0079,
      "step": 1
    },
    {
      "epoch": 0.06060606060606061,
      "grad_norm": 6.445240390692674,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.0204,
      "step": 2
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 6.382757287715354,
      "learning_rate": 2.5e-06,
      "loss": 1.0341,
      "step": 3
    },
    {
      "epoch": 0.12121212121212122,
      "grad_norm": 5.91272615348115,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.0061,
      "step": 4
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 4.722008058167302,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.9998,
      "step": 5
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 2.823573088690271,
      "learning_rate": 5e-06,
      "loss": 0.9534,
      "step": 6
    },
    {
      "epoch": 0.21212121212121213,
      "grad_norm": 2.5666390159658707,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.94,
      "step": 7
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 3.848402404280175,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.908,
      "step": 8
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 4.202645262587812,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.9232,
      "step": 9
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 3.9915170290882944,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.9167,
      "step": 10
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 3.23573372272586,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.852,
      "step": 11
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 2.9616255593701313,
      "learning_rate": 1e-05,
      "loss": 0.8476,
      "step": 12
    },
    {
      "epoch": 0.3939393939393939,
      "grad_norm": 2.0802017021336674,
      "learning_rate": 1.0833333333333334e-05,
      "loss": 0.8395,
      "step": 13
    },
    {
      "epoch": 0.42424242424242425,
      "grad_norm": 1.423032978782169,
      "learning_rate": 1.1666666666666668e-05,
      "loss": 0.8106,
      "step": 14
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 1.5233590243168056,
      "learning_rate": 1.25e-05,
      "loss": 0.7919,
      "step": 15
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 1.4027323424688252,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.7876,
      "step": 16
    },
    {
      "epoch": 0.5151515151515151,
      "grad_norm": 1.1436169964554868,
      "learning_rate": 1.416666666666667e-05,
      "loss": 0.7588,
      "step": 17
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 1.2584398803137398,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.7569,
      "step": 18
    },
    {
      "epoch": 0.5757575757575758,
      "grad_norm": 1.1708685062880144,
      "learning_rate": 1.5833333333333333e-05,
      "loss": 0.7448,
      "step": 19
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 0.9803925680829616,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.7378,
      "step": 20
    },
    {
      "epoch": 0.6363636363636364,
      "grad_norm": 0.9949928486317755,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 0.7286,
      "step": 21
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 1.030026417900517,
      "learning_rate": 1.8333333333333333e-05,
      "loss": 0.718,
      "step": 22
    },
    {
      "epoch": 0.696969696969697,
      "grad_norm": 1.1471899394360263,
      "learning_rate": 1.916666666666667e-05,
      "loss": 0.7148,
      "step": 23
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.9013002354547361,
      "learning_rate": 2e-05,
      "loss": 0.7179,
      "step": 24
    },
    {
      "epoch": 0.7575757575757576,
      "grad_norm": 1.1125959175909579,
      "learning_rate": 1.999884834944106e-05,
      "loss": 0.7161,
      "step": 25
    },
    {
      "epoch": 0.7878787878787878,
      "grad_norm": 0.9665027218020549,
      "learning_rate": 1.9995393663024054e-05,
      "loss": 0.7173,
      "step": 26
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 0.808324047733547,
      "learning_rate": 1.9989636736467278e-05,
      "loss": 0.7003,
      "step": 27
    },
    {
      "epoch": 0.8484848484848485,
      "grad_norm": 0.8598266261667703,
      "learning_rate": 1.9981578895764272e-05,
      "loss": 0.7005,
      "step": 28
    },
    {
      "epoch": 0.8787878787878788,
      "grad_norm": 0.8011804891864853,
      "learning_rate": 1.9971221996878395e-05,
      "loss": 0.7101,
      "step": 29
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.7777553110619772,
      "learning_rate": 1.9958568425315316e-05,
      "loss": 0.6925,
      "step": 30
    },
    {
      "epoch": 0.9393939393939394,
      "grad_norm": 0.8604670869538524,
      "learning_rate": 1.9943621095573588e-05,
      "loss": 0.6904,
      "step": 31
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.8621398383957887,
      "learning_rate": 1.9926383450473344e-05,
      "loss": 0.6751,
      "step": 32
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.7853612174091144,
      "learning_rate": 1.9906859460363307e-05,
      "loss": 0.6868,
      "step": 33
    },
    {
      "epoch": 1.0303030303030303,
      "grad_norm": 0.8046166369612455,
      "learning_rate": 1.9885053622206305e-05,
      "loss": 0.6497,
      "step": 34
    },
    {
      "epoch": 1.0606060606060606,
      "grad_norm": 0.6604470778365054,
      "learning_rate": 1.986097095854347e-05,
      "loss": 0.6293,
      "step": 35
    },
    {
      "epoch": 1.0909090909090908,
      "grad_norm": 0.6294942264260563,
      "learning_rate": 1.9834617016337424e-05,
      "loss": 0.6486,
      "step": 36
    },
    {
      "epoch": 1.121212121212121,
      "grad_norm": 0.5787782281521954,
      "learning_rate": 1.9805997865694616e-05,
      "loss": 0.6449,
      "step": 37
    },
    {
      "epoch": 1.1515151515151516,
      "grad_norm": 0.6035558760505482,
      "learning_rate": 1.9775120098467212e-05,
      "loss": 0.6322,
      "step": 38
    },
    {
      "epoch": 1.1818181818181819,
      "grad_norm": 0.5880242754158563,
      "learning_rate": 1.9741990826734793e-05,
      "loss": 0.6351,
      "step": 39
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 0.5329493991869941,
      "learning_rate": 1.970661768116622e-05,
      "loss": 0.6373,
      "step": 40
    },
    {
      "epoch": 1.2424242424242424,
      "grad_norm": 0.5487060567189378,
      "learning_rate": 1.9669008809262064e-05,
      "loss": 0.6254,
      "step": 41
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 0.577330341379163,
      "learning_rate": 1.9629172873477995e-05,
      "loss": 0.6145,
      "step": 42
    },
    {
      "epoch": 1.303030303030303,
      "grad_norm": 0.548723052022649,
      "learning_rate": 1.9587119049229558e-05,
      "loss": 0.631,
      "step": 43
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.4741961315186296,
      "learning_rate": 1.954285702277879e-05,
      "loss": 0.6315,
      "step": 44
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.5322002017205556,
      "learning_rate": 1.9496396989003195e-05,
      "loss": 0.6315,
      "step": 45
    },
    {
      "epoch": 1.393939393939394,
      "grad_norm": 0.5448230086028936,
      "learning_rate": 1.944774964904754e-05,
      "loss": 0.6338,
      "step": 46
    },
    {
      "epoch": 1.4242424242424243,
      "grad_norm": 0.5721457532205242,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 0.613,
      "step": 47
    },
    {
      "epoch": 1.4545454545454546,
      "grad_norm": 0.5011929282794981,
      "learning_rate": 1.9343938371606714e-05,
      "loss": 0.6268,
      "step": 48
    },
    {
      "epoch": 1.4848484848484849,
      "grad_norm": 0.6888065111036699,
      "learning_rate": 1.9288798344984673e-05,
      "loss": 0.6102,
      "step": 49
    },
    {
      "epoch": 1.5151515151515151,
      "grad_norm": 0.5819664946902224,
      "learning_rate": 1.9231518828401458e-05,
      "loss": 0.6165,
      "step": 50
    },
    {
      "epoch": 1.5454545454545454,
      "grad_norm": 0.4977271048283543,
      "learning_rate": 1.917211301505453e-05,
      "loss": 0.6135,
      "step": 51
    },
    {
      "epoch": 1.5757575757575757,
      "grad_norm": 0.5483360027543468,
      "learning_rate": 1.911059458789152e-05,
      "loss": 0.6035,
      "step": 52
    },
    {
      "epoch": 1.606060606060606,
      "grad_norm": 0.5938360983425195,
      "learning_rate": 1.9046977716458627e-05,
      "loss": 0.6082,
      "step": 53
    },
    {
      "epoch": 1.6363636363636362,
      "grad_norm": 0.639778467707558,
      "learning_rate": 1.8981277053636963e-05,
      "loss": 0.6192,
      "step": 54
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.6299987538556902,
      "learning_rate": 1.891350773226754e-05,
      "loss": 0.6187,
      "step": 55
    },
    {
      "epoch": 1.696969696969697,
      "grad_norm": 0.7462238042791106,
      "learning_rate": 1.8843685361665724e-05,
      "loss": 0.6339,
      "step": 56
    },
    {
      "epoch": 1.7272727272727273,
      "grad_norm": 0.4587409403241806,
      "learning_rate": 1.8771826024025944e-05,
      "loss": 0.6019,
      "step": 57
    },
    {
      "epoch": 1.7575757575757576,
      "grad_norm": 0.6309678302007972,
      "learning_rate": 1.8697946270717468e-05,
      "loss": 0.6206,
      "step": 58
    },
    {
      "epoch": 1.7878787878787878,
      "grad_norm": 0.5964975297698822,
      "learning_rate": 1.8622063118472135e-05,
      "loss": 0.6138,
      "step": 59
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.5091224937984986,
      "learning_rate": 1.8544194045464888e-05,
      "loss": 0.6237,
      "step": 60
    },
    {
      "epoch": 1.8484848484848486,
      "grad_norm": 0.43784547796166895,
      "learning_rate": 1.8464356987288012e-05,
      "loss": 0.6072,
      "step": 61
    },
    {
      "epoch": 1.878787878787879,
      "grad_norm": 0.47486738520964883,
      "learning_rate": 1.8382570332820045e-05,
      "loss": 0.6039,
      "step": 62
    },
    {
      "epoch": 1.9090909090909092,
      "grad_norm": 0.44865590799121363,
      "learning_rate": 1.8298852919990254e-05,
      "loss": 0.6092,
      "step": 63
    },
    {
      "epoch": 1.9393939393939394,
      "grad_norm": 0.5375632964773143,
      "learning_rate": 1.821322403143969e-05,
      "loss": 0.5939,
      "step": 64
    },
    {
      "epoch": 1.9696969696969697,
      "grad_norm": 0.608651292228407,
      "learning_rate": 1.812570339007983e-05,
      "loss": 0.6129,
      "step": 65
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.44024856323173045,
      "learning_rate": 1.8036311154549783e-05,
      "loss": 0.5893,
      "step": 66
    },
    {
      "epoch": 2.0303030303030303,
      "grad_norm": 0.6980844762428723,
      "learning_rate": 1.7945067914573147e-05,
      "loss": 0.5474,
      "step": 67
    },
    {
      "epoch": 2.0606060606060606,
      "grad_norm": 0.5182683022073494,
      "learning_rate": 1.7851994686215592e-05,
      "loss": 0.5725,
      "step": 68
    },
    {
      "epoch": 2.090909090909091,
      "grad_norm": 0.5404396715701119,
      "learning_rate": 1.77571129070442e-05,
      "loss": 0.5526,
      "step": 69
    },
    {
      "epoch": 2.121212121212121,
      "grad_norm": 0.5305063939153388,
      "learning_rate": 1.766044443118978e-05,
      "loss": 0.5536,
      "step": 70
    },
    {
      "epoch": 2.1515151515151514,
      "grad_norm": 0.5644734582249971,
      "learning_rate": 1.7562011524313187e-05,
      "loss": 0.5355,
      "step": 71
    },
    {
      "epoch": 2.1818181818181817,
      "grad_norm": 0.5202130970874336,
      "learning_rate": 1.7461836858476858e-05,
      "loss": 0.5384,
      "step": 72
    },
    {
      "epoch": 2.212121212121212,
      "grad_norm": 0.5357956217064614,
      "learning_rate": 1.7359943506922775e-05,
      "loss": 0.5442,
      "step": 73
    },
    {
      "epoch": 2.242424242424242,
      "grad_norm": 0.565767866530521,
      "learning_rate": 1.725635493875799e-05,
      "loss": 0.5276,
      "step": 74
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.5149462165435431,
      "learning_rate": 1.7151095013548996e-05,
      "loss": 0.5282,
      "step": 75
    },
    {
      "epoch": 2.303030303030303,
      "grad_norm": 0.6836826507439627,
      "learning_rate": 1.7044187975826126e-05,
      "loss": 0.5528,
      "step": 76
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.4830928504569663,
      "learning_rate": 1.693565844949933e-05,
      "loss": 0.5341,
      "step": 77
    },
    {
      "epoch": 2.3636363636363638,
      "grad_norm": 0.5788000592263608,
      "learning_rate": 1.6825531432186545e-05,
      "loss": 0.5385,
      "step": 78
    },
    {
      "epoch": 2.393939393939394,
      "grad_norm": 0.5726173587627938,
      "learning_rate": 1.671383228945597e-05,
      "loss": 0.5422,
      "step": 79
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 0.6643959541973968,
      "learning_rate": 1.6600586748983642e-05,
      "loss": 0.5548,
      "step": 80
    },
    {
      "epoch": 2.4545454545454546,
      "grad_norm": 0.48281206376363367,
      "learning_rate": 1.648582089462756e-05,
      "loss": 0.5416,
      "step": 81
    },
    {
      "epoch": 2.484848484848485,
      "grad_norm": 0.6009102492415257,
      "learning_rate": 1.6369561160419783e-05,
      "loss": 0.5322,
      "step": 82
    },
    {
      "epoch": 2.515151515151515,
      "grad_norm": 0.4893746393385918,
      "learning_rate": 1.625183432447789e-05,
      "loss": 0.5486,
      "step": 83
    },
    {
      "epoch": 2.5454545454545454,
      "grad_norm": 0.509521231101079,
      "learning_rate": 1.6132667502837164e-05,
      "loss": 0.5324,
      "step": 84
    },
    {
      "epoch": 2.5757575757575757,
      "grad_norm": 0.545869239651775,
      "learning_rate": 1.6012088143204953e-05,
      "loss": 0.5347,
      "step": 85
    },
    {
      "epoch": 2.606060606060606,
      "grad_norm": 0.5248012196294125,
      "learning_rate": 1.589012401863864e-05,
      "loss": 0.5474,
      "step": 86
    },
    {
      "epoch": 2.6363636363636362,
      "grad_norm": 0.5691007796569792,
      "learning_rate": 1.5766803221148676e-05,
      "loss": 0.5328,
      "step": 87
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.48088594253053496,
      "learning_rate": 1.5642154155228124e-05,
      "loss": 0.5281,
      "step": 88
    },
    {
      "epoch": 2.6969696969696972,
      "grad_norm": 0.4847231502014893,
      "learning_rate": 1.5516205531310272e-05,
      "loss": 0.5366,
      "step": 89
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.4579402446027877,
      "learning_rate": 1.538898635915576e-05,
      "loss": 0.5356,
      "step": 90
    },
    {
      "epoch": 2.757575757575758,
      "grad_norm": 0.5022857211635094,
      "learning_rate": 1.526052594117071e-05,
      "loss": 0.5257,
      "step": 91
    },
    {
      "epoch": 2.787878787878788,
      "grad_norm": 0.5197458111165683,
      "learning_rate": 1.513085386565758e-05,
      "loss": 0.5441,
      "step": 92
    },
    {
      "epoch": 2.8181818181818183,
      "grad_norm": 0.4443806849341786,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.5375,
      "step": 93
    },
    {
      "epoch": 2.8484848484848486,
      "grad_norm": 0.4415612167940715,
      "learning_rate": 1.4867994483783485e-05,
      "loss": 0.5198,
      "step": 94
    },
    {
      "epoch": 2.878787878787879,
      "grad_norm": 0.4729466207945027,
      "learning_rate": 1.4734867721853341e-05,
      "loss": 0.5327,
      "step": 95
    },
    {
      "epoch": 2.909090909090909,
      "grad_norm": 0.3927504975139151,
      "learning_rate": 1.4600650377311523e-05,
      "loss": 0.5323,
      "step": 96
    },
    {
      "epoch": 2.9393939393939394,
      "grad_norm": 0.4932818281771704,
      "learning_rate": 1.4465373364454001e-05,
      "loss": 0.54,
      "step": 97
    },
    {
      "epoch": 2.9696969696969697,
      "grad_norm": 0.395042020106255,
      "learning_rate": 1.4329067841650274e-05,
      "loss": 0.5499,
      "step": 98
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.4842433679584853,
      "learning_rate": 1.4191765204166643e-05,
      "loss": 0.5219,
      "step": 99
    },
    {
      "epoch": 3.0303030303030303,
      "grad_norm": 0.517778147081366,
      "learning_rate": 1.4053497076934948e-05,
      "loss": 0.4792,
      "step": 100
    },
    {
      "epoch": 3.0606060606060606,
      "grad_norm": 0.49787957784324754,
      "learning_rate": 1.3914295307268396e-05,
      "loss": 0.4846,
      "step": 101
    },
    {
      "epoch": 3.090909090909091,
      "grad_norm": 0.5333577077714253,
      "learning_rate": 1.3774191957526144e-05,
      "loss": 0.4834,
      "step": 102
    },
    {
      "epoch": 3.121212121212121,
      "grad_norm": 0.5553372294102208,
      "learning_rate": 1.3633219297728415e-05,
      "loss": 0.4893,
      "step": 103
    },
    {
      "epoch": 3.1515151515151514,
      "grad_norm": 0.5225451939758264,
      "learning_rate": 1.3491409798123687e-05,
      "loss": 0.4755,
      "step": 104
    },
    {
      "epoch": 3.1818181818181817,
      "grad_norm": 0.5147983075558922,
      "learning_rate": 1.3348796121709862e-05,
      "loss": 0.4688,
      "step": 105
    },
    {
      "epoch": 3.212121212121212,
      "grad_norm": 0.596538636274444,
      "learning_rate": 1.3205411116710973e-05,
      "loss": 0.482,
      "step": 106
    },
    {
      "epoch": 3.242424242424242,
      "grad_norm": 0.5648728417875709,
      "learning_rate": 1.3061287809011243e-05,
      "loss": 0.4887,
      "step": 107
    },
    {
      "epoch": 3.2727272727272725,
      "grad_norm": 0.47558400405296414,
      "learning_rate": 1.291645939454825e-05,
      "loss": 0.4782,
      "step": 108
    },
    {
      "epoch": 3.303030303030303,
      "grad_norm": 0.5186604404259834,
      "learning_rate": 1.277095923166689e-05,
      "loss": 0.4662,
      "step": 109
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.5236731465454446,
      "learning_rate": 1.2624820833435939e-05,
      "loss": 0.4643,
      "step": 110
    },
    {
      "epoch": 3.3636363636363638,
      "grad_norm": 0.4607134576648133,
      "learning_rate": 1.2478077859929e-05,
      "loss": 0.4704,
      "step": 111
    },
    {
      "epoch": 3.393939393939394,
      "grad_norm": 0.5092626471840158,
      "learning_rate": 1.2330764110471567e-05,
      "loss": 0.4814,
      "step": 112
    },
    {
      "epoch": 3.4242424242424243,
      "grad_norm": 0.47207207280474506,
      "learning_rate": 1.2182913515856016e-05,
      "loss": 0.4565,
      "step": 113
    },
    {
      "epoch": 3.4545454545454546,
      "grad_norm": 0.4374242509606414,
      "learning_rate": 1.2034560130526341e-05,
      "loss": 0.4731,
      "step": 114
    },
    {
      "epoch": 3.484848484848485,
      "grad_norm": 0.4730060091153659,
      "learning_rate": 1.1885738124734359e-05,
      "loss": 0.4683,
      "step": 115
    },
    {
      "epoch": 3.515151515151515,
      "grad_norm": 0.4221262453746083,
      "learning_rate": 1.1736481776669307e-05,
      "loss": 0.4663,
      "step": 116
    },
    {
      "epoch": 3.5454545454545454,
      "grad_norm": 0.5101200302479499,
      "learning_rate": 1.1586825464562515e-05,
      "loss": 0.4648,
      "step": 117
    },
    {
      "epoch": 3.5757575757575757,
      "grad_norm": 0.5103410919800033,
      "learning_rate": 1.1436803658769082e-05,
      "loss": 0.4779,
      "step": 118
    },
    {
      "epoch": 3.606060606060606,
      "grad_norm": 0.4450626730517202,
      "learning_rate": 1.1286450913828313e-05,
      "loss": 0.4752,
      "step": 119
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.4968592992971222,
      "learning_rate": 1.113580186050475e-05,
      "loss": 0.4741,
      "step": 120
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.4011648197454992,
      "learning_rate": 1.0984891197811686e-05,
      "loss": 0.4829,
      "step": 121
    },
    {
      "epoch": 3.6969696969696972,
      "grad_norm": 0.5273690586575496,
      "learning_rate": 1.0833753685018935e-05,
      "loss": 0.4693,
      "step": 122
    },
    {
      "epoch": 3.7272727272727275,
      "grad_norm": 0.43414871426684687,
      "learning_rate": 1.0682424133646712e-05,
      "loss": 0.4684,
      "step": 123
    },
    {
      "epoch": 3.757575757575758,
      "grad_norm": 0.4014074263052897,
      "learning_rate": 1.0530937399447496e-05,
      "loss": 0.4812,
      "step": 124
    },
    {
      "epoch": 3.787878787878788,
      "grad_norm": 0.4547211648900433,
      "learning_rate": 1.0379328374377715e-05,
      "loss": 0.4629,
      "step": 125
    },
    {
      "epoch": 3.8181818181818183,
      "grad_norm": 0.3961979664531267,
      "learning_rate": 1.0227631978561057e-05,
      "loss": 0.475,
      "step": 126
    },
    {
      "epoch": 3.8484848484848486,
      "grad_norm": 0.43542193984738736,
      "learning_rate": 1.0075883152245334e-05,
      "loss": 0.4597,
      "step": 127
    },
    {
      "epoch": 3.878787878787879,
      "grad_norm": 0.40945658123125683,
      "learning_rate": 9.92411684775467e-06,
      "loss": 0.4751,
      "step": 128
    },
    {
      "epoch": 3.909090909090909,
      "grad_norm": 0.37125307540613495,
      "learning_rate": 9.772368021438943e-06,
      "loss": 0.4788,
      "step": 129
    },
    {
      "epoch": 3.9393939393939394,
      "grad_norm": 0.43624019011061105,
      "learning_rate": 9.620671625622287e-06,
      "loss": 0.4658,
      "step": 130
    },
    {
      "epoch": 3.9696969696969697,
      "grad_norm": 0.38438890329969516,
      "learning_rate": 9.469062600552509e-06,
      "loss": 0.4693,
      "step": 131
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.46071327617582064,
      "learning_rate": 9.317575866353293e-06,
      "loss": 0.4703,
      "step": 132
    },
    {
      "epoch": 4.03030303030303,
      "grad_norm": 0.549915464553488,
      "learning_rate": 9.166246314981066e-06,
      "loss": 0.4316,
      "step": 133
    },
    {
      "epoch": 4.0606060606060606,
      "grad_norm": 0.49705236619568727,
      "learning_rate": 9.015108802188314e-06,
      "loss": 0.4099,
      "step": 134
    },
    {
      "epoch": 4.090909090909091,
      "grad_norm": 0.4289325156966301,
      "learning_rate": 8.86419813949525e-06,
      "loss": 0.4261,
      "step": 135
    },
    {
      "epoch": 4.121212121212121,
      "grad_norm": 0.5693436632007958,
      "learning_rate": 8.71354908617169e-06,
      "loss": 0.4191,
      "step": 136
    },
    {
      "epoch": 4.151515151515151,
      "grad_norm": 0.5387261802907803,
      "learning_rate": 8.56319634123092e-06,
      "loss": 0.422,
      "step": 137
    },
    {
      "epoch": 4.181818181818182,
      "grad_norm": 0.4309658522579619,
      "learning_rate": 8.413174535437486e-06,
      "loss": 0.4106,
      "step": 138
    },
    {
      "epoch": 4.212121212121212,
      "grad_norm": 0.5246341763714903,
      "learning_rate": 8.263518223330698e-06,
      "loss": 0.4288,
      "step": 139
    },
    {
      "epoch": 4.242424242424242,
      "grad_norm": 0.44678572175146675,
      "learning_rate": 8.114261875265643e-06,
      "loss": 0.4174,
      "step": 140
    },
    {
      "epoch": 4.2727272727272725,
      "grad_norm": 0.41094038943781347,
      "learning_rate": 7.965439869473664e-06,
      "loss": 0.4077,
      "step": 141
    },
    {
      "epoch": 4.303030303030303,
      "grad_norm": 0.40324815976350814,
      "learning_rate": 7.817086484143987e-06,
      "loss": 0.4168,
      "step": 142
    },
    {
      "epoch": 4.333333333333333,
      "grad_norm": 0.4677095167961872,
      "learning_rate": 7.669235889528436e-06,
      "loss": 0.4228,
      "step": 143
    },
    {
      "epoch": 4.363636363636363,
      "grad_norm": 0.4231620470214181,
      "learning_rate": 7.521922140071003e-06,
      "loss": 0.406,
      "step": 144
    },
    {
      "epoch": 4.393939393939394,
      "grad_norm": 0.38472876180976645,
      "learning_rate": 7.375179166564062e-06,
      "loss": 0.419,
      "step": 145
    },
    {
      "epoch": 4.424242424242424,
      "grad_norm": 0.39815218699138305,
      "learning_rate": 7.2290407683331154e-06,
      "loss": 0.4217,
      "step": 146
    },
    {
      "epoch": 4.454545454545454,
      "grad_norm": 0.38973334842621504,
      "learning_rate": 7.0835406054517505e-06,
      "loss": 0.407,
      "step": 147
    },
    {
      "epoch": 4.484848484848484,
      "grad_norm": 0.357129065814448,
      "learning_rate": 6.93871219098876e-06,
      "loss": 0.4161,
      "step": 148
    },
    {
      "epoch": 4.515151515151516,
      "grad_norm": 0.37574480732292936,
      "learning_rate": 6.79458888328903e-06,
      "loss": 0.4127,
      "step": 149
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 0.4304590573047278,
      "learning_rate": 6.651203878290139e-06,
      "loss": 0.4136,
      "step": 150
    },
    {
      "epoch": 4.575757575757576,
      "grad_norm": 0.37455970862907545,
      "learning_rate": 6.508590201876317e-06,
      "loss": 0.4126,
      "step": 151
    },
    {
      "epoch": 4.606060606060606,
      "grad_norm": 0.35140746880501045,
      "learning_rate": 6.366780702271589e-06,
      "loss": 0.402,
      "step": 152
    },
    {
      "epoch": 4.636363636363637,
      "grad_norm": 0.35468098669148884,
      "learning_rate": 6.225808042473857e-06,
      "loss": 0.4148,
      "step": 153
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.35870638919737535,
      "learning_rate": 6.085704692731609e-06,
      "loss": 0.4173,
      "step": 154
    },
    {
      "epoch": 4.696969696969697,
      "grad_norm": 0.3404383800514381,
      "learning_rate": 5.946502923065054e-06,
      "loss": 0.4148,
      "step": 155
    },
    {
      "epoch": 4.7272727272727275,
      "grad_norm": 0.3414680860961153,
      "learning_rate": 5.8082347958333625e-06,
      "loss": 0.4187,
      "step": 156
    },
    {
      "epoch": 4.757575757575758,
      "grad_norm": 0.3612773359243806,
      "learning_rate": 5.670932158349732e-06,
      "loss": 0.4156,
      "step": 157
    },
    {
      "epoch": 4.787878787878788,
      "grad_norm": 0.3455389798706968,
      "learning_rate": 5.534626635546e-06,
      "loss": 0.4121,
      "step": 158
    },
    {
      "epoch": 4.818181818181818,
      "grad_norm": 0.3469521902127527,
      "learning_rate": 5.399349622688479e-06,
      "loss": 0.42,
      "step": 159
    },
    {
      "epoch": 4.848484848484849,
      "grad_norm": 0.3287731553618517,
      "learning_rate": 5.2651322781466606e-06,
      "loss": 0.4236,
      "step": 160
    },
    {
      "epoch": 4.878787878787879,
      "grad_norm": 0.3271765785486276,
      "learning_rate": 5.132005516216512e-06,
      "loss": 0.4135,
      "step": 161
    },
    {
      "epoch": 4.909090909090909,
      "grad_norm": 0.31239522699154915,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.4097,
      "step": 162
    },
    {
      "epoch": 4.9393939393939394,
      "grad_norm": 0.3156681272030559,
      "learning_rate": 4.869146134342426e-06,
      "loss": 0.3954,
      "step": 163
    },
    {
      "epoch": 4.96969696969697,
      "grad_norm": 0.36382188933769277,
      "learning_rate": 4.739474058829288e-06,
      "loss": 0.4229,
      "step": 164
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.34045399086841477,
      "learning_rate": 4.611013640844245e-06,
      "loss": 0.4149,
      "step": 165
    },
    {
      "epoch": 5.03030303030303,
      "grad_norm": 0.5044756046722096,
      "learning_rate": 4.483794468689728e-06,
      "loss": 0.3844,
      "step": 166
    },
    {
      "epoch": 5.0606060606060606,
      "grad_norm": 0.41237791230202236,
      "learning_rate": 4.357845844771881e-06,
      "loss": 0.3671,
      "step": 167
    },
    {
      "epoch": 5.090909090909091,
      "grad_norm": 0.34955916182422403,
      "learning_rate": 4.2331967788513295e-06,
      "loss": 0.373,
      "step": 168
    },
    {
      "epoch": 5.121212121212121,
      "grad_norm": 0.4438250960546701,
      "learning_rate": 4.109875981361363e-06,
      "loss": 0.3564,
      "step": 169
    },
    {
      "epoch": 5.151515151515151,
      "grad_norm": 0.5378833215453719,
      "learning_rate": 3.987911856795047e-06,
      "loss": 0.3675,
      "step": 170
    },
    {
      "epoch": 5.181818181818182,
      "grad_norm": 0.42285744976151174,
      "learning_rate": 3.867332497162836e-06,
      "loss": 0.3746,
      "step": 171
    },
    {
      "epoch": 5.212121212121212,
      "grad_norm": 0.3830432468541377,
      "learning_rate": 3.748165675522113e-06,
      "loss": 0.3703,
      "step": 172
    },
    {
      "epoch": 5.242424242424242,
      "grad_norm": 0.3858055962114919,
      "learning_rate": 3.630438839580217e-06,
      "loss": 0.3783,
      "step": 173
    },
    {
      "epoch": 5.2727272727272725,
      "grad_norm": 0.3917855532263057,
      "learning_rate": 3.5141791053724405e-06,
      "loss": 0.3592,
      "step": 174
    },
    {
      "epoch": 5.303030303030303,
      "grad_norm": 0.419316148500706,
      "learning_rate": 3.399413251016359e-06,
      "loss": 0.3706,
      "step": 175
    },
    {
      "epoch": 5.333333333333333,
      "grad_norm": 0.38280771968727684,
      "learning_rate": 3.2861677105440335e-06,
      "loss": 0.369,
      "step": 176
    },
    {
      "epoch": 5.363636363636363,
      "grad_norm": 0.37161552848703766,
      "learning_rate": 3.174468567813461e-06,
      "loss": 0.3683,
      "step": 177
    },
    {
      "epoch": 5.393939393939394,
      "grad_norm": 0.3839626609743638,
      "learning_rate": 3.0643415505006733e-06,
      "loss": 0.373,
      "step": 178
    },
    {
      "epoch": 5.424242424242424,
      "grad_norm": 0.3737777443133059,
      "learning_rate": 2.9558120241738786e-06,
      "loss": 0.3751,
      "step": 179
    },
    {
      "epoch": 5.454545454545454,
      "grad_norm": 0.3699742548934759,
      "learning_rate": 2.8489049864510053e-06,
      "loss": 0.3705,
      "step": 180
    },
    {
      "epoch": 5.484848484848484,
      "grad_norm": 0.3621054260712193,
      "learning_rate": 2.7436450612420098e-06,
      "loss": 0.3692,
      "step": 181
    },
    {
      "epoch": 5.515151515151516,
      "grad_norm": 0.36441907661210643,
      "learning_rate": 2.640056493077231e-06,
      "loss": 0.3767,
      "step": 182
    },
    {
      "epoch": 5.545454545454545,
      "grad_norm": 0.34402215120323093,
      "learning_rate": 2.5381631415231455e-06,
      "loss": 0.3759,
      "step": 183
    },
    {
      "epoch": 5.575757575757576,
      "grad_norm": 0.32795833240845856,
      "learning_rate": 2.4379884756868167e-06,
      "loss": 0.3789,
      "step": 184
    },
    {
      "epoch": 5.606060606060606,
      "grad_norm": 0.32215801470209426,
      "learning_rate": 2.339555568810221e-06,
      "loss": 0.3711,
      "step": 185
    },
    {
      "epoch": 5.636363636363637,
      "grad_norm": 0.33932527796232764,
      "learning_rate": 2.2428870929558012e-06,
      "loss": 0.3806,
      "step": 186
    },
    {
      "epoch": 5.666666666666667,
      "grad_norm": 0.33010171544059436,
      "learning_rate": 2.1480053137844115e-06,
      "loss": 0.3793,
      "step": 187
    },
    {
      "epoch": 5.696969696969697,
      "grad_norm": 0.3347552964327516,
      "learning_rate": 2.054932085426856e-06,
      "loss": 0.388,
      "step": 188
    },
    {
      "epoch": 5.7272727272727275,
      "grad_norm": 0.30073887916372427,
      "learning_rate": 1.963688845450218e-06,
      "loss": 0.3744,
      "step": 189
    },
    {
      "epoch": 5.757575757575758,
      "grad_norm": 0.3131014609633843,
      "learning_rate": 1.8742966099201699e-06,
      "loss": 0.3641,
      "step": 190
    },
    {
      "epoch": 5.787878787878788,
      "grad_norm": 0.3325293489769436,
      "learning_rate": 1.7867759685603115e-06,
      "loss": 0.3624,
      "step": 191
    },
    {
      "epoch": 5.818181818181818,
      "grad_norm": 0.30487109929123163,
      "learning_rate": 1.7011470800097496e-06,
      "loss": 0.3752,
      "step": 192
    },
    {
      "epoch": 5.848484848484849,
      "grad_norm": 0.2984494414418286,
      "learning_rate": 1.6174296671799571e-06,
      "loss": 0.3569,
      "step": 193
    },
    {
      "epoch": 5.878787878787879,
      "grad_norm": 0.30215385875684886,
      "learning_rate": 1.5356430127119915e-06,
      "loss": 0.3644,
      "step": 194
    },
    {
      "epoch": 5.909090909090909,
      "grad_norm": 0.2944822474056789,
      "learning_rate": 1.4558059545351144e-06,
      "loss": 0.3733,
      "step": 195
    },
    {
      "epoch": 5.9393939393939394,
      "grad_norm": 0.2901537050818472,
      "learning_rate": 1.3779368815278648e-06,
      "loss": 0.3736,
      "step": 196
    },
    {
      "epoch": 5.96969696969697,
      "grad_norm": 0.3021704491376866,
      "learning_rate": 1.302053729282533e-06,
      "loss": 0.3587,
      "step": 197
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.33659568526850164,
      "learning_rate": 1.2281739759740575e-06,
      "loss": 0.374,
      "step": 198
    },
    {
      "epoch": 6.03030303030303,
      "grad_norm": 0.3961446442740241,
      "learning_rate": 1.156314638334277e-06,
      "loss": 0.3456,
      "step": 199
    },
    {
      "epoch": 6.0606060606060606,
      "grad_norm": 0.3578301159030496,
      "learning_rate": 1.086492267732462e-06,
      "loss": 0.3346,
      "step": 200
    },
    {
      "epoch": 6.090909090909091,
      "grad_norm": 0.32368117334038604,
      "learning_rate": 1.01872294636304e-06,
      "loss": 0.3551,
      "step": 201
    },
    {
      "epoch": 6.121212121212121,
      "grad_norm": 0.30832624564558136,
      "learning_rate": 9.530222835413739e-07,
      "loss": 0.3485,
      "step": 202
    },
    {
      "epoch": 6.151515151515151,
      "grad_norm": 0.3084039883834137,
      "learning_rate": 8.894054121084839e-07,
      "loss": 0.3435,
      "step": 203
    },
    {
      "epoch": 6.181818181818182,
      "grad_norm": 0.3022004078228386,
      "learning_rate": 8.278869849454718e-07,
      "loss": 0.3489,
      "step": 204
    },
    {
      "epoch": 6.212121212121212,
      "grad_norm": 0.3351879674470543,
      "learning_rate": 7.684811715985429e-07,
      "loss": 0.3423,
      "step": 205
    },
    {
      "epoch": 6.242424242424242,
      "grad_norm": 0.3223254087806397,
      "learning_rate": 7.1120165501533e-07,
      "loss": 0.3576,
      "step": 206
    },
    {
      "epoch": 6.2727272727272725,
      "grad_norm": 0.3249796694501032,
      "learning_rate": 6.560616283932897e-07,
      "loss": 0.3596,
      "step": 207
    },
    {
      "epoch": 6.303030303030303,
      "grad_norm": 0.3383471916015826,
      "learning_rate": 6.030737921409169e-07,
      "loss": 0.3579,
      "step": 208
    },
    {
      "epoch": 6.333333333333333,
      "grad_norm": 0.3210379088186702,
      "learning_rate": 5.522503509524591e-07,
      "loss": 0.3473,
      "step": 209
    },
    {
      "epoch": 6.363636363636363,
      "grad_norm": 0.32482561399174104,
      "learning_rate": 5.036030109968082e-07,
      "loss": 0.3467,
      "step": 210
    },
    {
      "epoch": 6.393939393939394,
      "grad_norm": 0.29307954683423065,
      "learning_rate": 4.5714297722121105e-07,
      "loss": 0.349,
      "step": 211
    },
    {
      "epoch": 6.424242424242424,
      "grad_norm": 0.28564159916131726,
      "learning_rate": 4.128809507704445e-07,
      "loss": 0.3458,
      "step": 212
    },
    {
      "epoch": 6.454545454545454,
      "grad_norm": 0.2794555238624641,
      "learning_rate": 3.708271265220087e-07,
      "loss": 0.3478,
      "step": 213
    },
    {
      "epoch": 6.484848484848484,
      "grad_norm": 0.2739868217444457,
      "learning_rate": 3.309911907379393e-07,
      "loss": 0.3588,
      "step": 214
    },
    {
      "epoch": 6.515151515151516,
      "grad_norm": 0.29345881187170175,
      "learning_rate": 2.9338231883378365e-07,
      "loss": 0.3528,
      "step": 215
    },
    {
      "epoch": 6.545454545454545,
      "grad_norm": 0.30010189778050617,
      "learning_rate": 2.5800917326521013e-07,
      "loss": 0.3544,
      "step": 216
    },
    {
      "epoch": 6.575757575757576,
      "grad_norm": 0.28752501262154156,
      "learning_rate": 2.248799015327907e-07,
      "loss": 0.3378,
      "step": 217
    },
    {
      "epoch": 6.606060606060606,
      "grad_norm": 0.3072149247627371,
      "learning_rate": 1.9400213430538773e-07,
      "loss": 0.336,
      "step": 218
    },
    {
      "epoch": 6.636363636363637,
      "grad_norm": 0.2713972189812151,
      "learning_rate": 1.6538298366257975e-07,
      "loss": 0.3443,
      "step": 219
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.2954197120252145,
      "learning_rate": 1.3902904145653094e-07,
      "loss": 0.3542,
      "step": 220
    },
    {
      "epoch": 6.696969696969697,
      "grad_norm": 0.28950931222260234,
      "learning_rate": 1.1494637779369766e-07,
      "loss": 0.3524,
      "step": 221
    },
    {
      "epoch": 6.7272727272727275,
      "grad_norm": 0.2751344749931131,
      "learning_rate": 9.314053963669245e-08,
      "loss": 0.3538,
      "step": 222
    },
    {
      "epoch": 6.757575757575758,
      "grad_norm": 0.30131842899827077,
      "learning_rate": 7.361654952665608e-08,
      "loss": 0.335,
      "step": 223
    },
    {
      "epoch": 6.787878787878788,
      "grad_norm": 0.27821946794035596,
      "learning_rate": 5.637890442641403e-08,
      "loss": 0.3757,
      "step": 224
    },
    {
      "epoch": 6.818181818181818,
      "grad_norm": 0.27764319306193624,
      "learning_rate": 4.143157468468717e-08,
      "loss": 0.3583,
      "step": 225
    },
    {
      "epoch": 6.848484848484849,
      "grad_norm": 0.3249780204193289,
      "learning_rate": 2.8778003121607834e-08,
      "loss": 0.3441,
      "step": 226
    },
    {
      "epoch": 6.878787878787879,
      "grad_norm": 0.2852587967210579,
      "learning_rate": 1.8421104235727406e-08,
      "loss": 0.3502,
      "step": 227
    },
    {
      "epoch": 6.909090909090909,
      "grad_norm": 0.28096930566095124,
      "learning_rate": 1.0363263532724433e-08,
      "loss": 0.3478,
      "step": 228
    },
    {
      "epoch": 6.9393939393939394,
      "grad_norm": 0.2910326513319612,
      "learning_rate": 4.606336975948589e-09,
      "loss": 0.3506,
      "step": 229
    },
    {
      "epoch": 6.96969696969697,
      "grad_norm": 0.2693333849629141,
      "learning_rate": 1.1516505589381777e-09,
      "loss": 0.3536,
      "step": 230
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.2762734213295961,
      "learning_rate": 0.0,
      "loss": 0.3553,
      "step": 231
    },
    {
      "epoch": 7.0,
      "step": 231,
      "total_flos": 6.488329161885614e+17,
      "train_loss": 0.511818146808839,
      "train_runtime": 8828.5494,
      "train_samples_per_second": 2.506,
      "train_steps_per_second": 0.026
    }
  ],
  "logging_steps": 1,
  "max_steps": 231,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 6.488329161885614e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}