{
  "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": 6.550661001900527,
      "learning_rate": 8.333333333333333e-07,
      "loss": 1.0429,
      "step": 1
    },
    {
      "epoch": 0.06060606060606061,
      "grad_norm": 6.183749689123775,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.0003,
      "step": 2
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 6.466678211734568,
      "learning_rate": 2.5e-06,
      "loss": 1.0332,
      "step": 3
    },
    {
      "epoch": 0.12121212121212122,
      "grad_norm": 5.972356265910597,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.025,
      "step": 4
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 4.488788085292552,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.9813,
      "step": 5
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 2.7297884733970386,
      "learning_rate": 5e-06,
      "loss": 0.952,
      "step": 6
    },
    {
      "epoch": 0.21212121212121213,
      "grad_norm": 2.496688797372574,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.9619,
      "step": 7
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 3.940545302245823,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.9387,
      "step": 8
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 4.159694963497642,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.9172,
      "step": 9
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 3.980967838362522,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.9165,
      "step": 10
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 3.3773524652480518,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.888,
      "step": 11
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 2.9560510226841794,
      "learning_rate": 1e-05,
      "loss": 0.8618,
      "step": 12
    },
    {
      "epoch": 0.3939393939393939,
      "grad_norm": 2.151270422200052,
      "learning_rate": 1.0833333333333334e-05,
      "loss": 0.8329,
      "step": 13
    },
    {
      "epoch": 0.42424242424242425,
      "grad_norm": 1.4937495937234244,
      "learning_rate": 1.1666666666666668e-05,
      "loss": 0.833,
      "step": 14
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 1.6566254057494072,
      "learning_rate": 1.25e-05,
      "loss": 0.7998,
      "step": 15
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 1.4454767442858956,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.8136,
      "step": 16
    },
    {
      "epoch": 0.5151515151515151,
      "grad_norm": 1.2419339199204398,
      "learning_rate": 1.416666666666667e-05,
      "loss": 0.7754,
      "step": 17
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 1.1887600541454704,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.7654,
      "step": 18
    },
    {
      "epoch": 0.5757575757575758,
      "grad_norm": 1.047597154433694,
      "learning_rate": 1.5833333333333333e-05,
      "loss": 0.7813,
      "step": 19
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 0.9639300303336714,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.7689,
      "step": 20
    },
    {
      "epoch": 0.6363636363636364,
      "grad_norm": 1.0150061663263692,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 0.7439,
      "step": 21
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 1.002534046998058,
      "learning_rate": 1.8333333333333333e-05,
      "loss": 0.7339,
      "step": 22
    },
    {
      "epoch": 0.696969696969697,
      "grad_norm": 0.8642181699244971,
      "learning_rate": 1.916666666666667e-05,
      "loss": 0.7104,
      "step": 23
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.9131770262733899,
      "learning_rate": 2e-05,
      "loss": 0.7183,
      "step": 24
    },
    {
      "epoch": 0.7575757575757576,
      "grad_norm": 0.8304776818717265,
      "learning_rate": 1.999884834944106e-05,
      "loss": 0.7381,
      "step": 25
    },
    {
      "epoch": 0.7878787878787878,
      "grad_norm": 0.8542376737773135,
      "learning_rate": 1.9995393663024054e-05,
      "loss": 0.7308,
      "step": 26
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 0.979281672915267,
      "learning_rate": 1.9989636736467278e-05,
      "loss": 0.702,
      "step": 27
    },
    {
      "epoch": 0.8484848484848485,
      "grad_norm": 0.7890801995780901,
      "learning_rate": 1.9981578895764272e-05,
      "loss": 0.7045,
      "step": 28
    },
    {
      "epoch": 0.8787878787878788,
      "grad_norm": 0.9021982049183424,
      "learning_rate": 1.9971221996878395e-05,
      "loss": 0.7001,
      "step": 29
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.8802448377399086,
      "learning_rate": 1.9958568425315316e-05,
      "loss": 0.7177,
      "step": 30
    },
    {
      "epoch": 0.9393939393939394,
      "grad_norm": 0.756194075464945,
      "learning_rate": 1.9943621095573588e-05,
      "loss": 0.7075,
      "step": 31
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.7972388977669735,
      "learning_rate": 1.9926383450473344e-05,
      "loss": 0.7008,
      "step": 32
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.6523681490101999,
      "learning_rate": 1.9906859460363307e-05,
      "loss": 0.6961,
      "step": 33
    },
    {
      "epoch": 1.0303030303030303,
      "grad_norm": 0.6046427320252548,
      "learning_rate": 1.9885053622206305e-05,
      "loss": 0.672,
      "step": 34
    },
    {
      "epoch": 1.0606060606060606,
      "grad_norm": 0.5728512382523253,
      "learning_rate": 1.986097095854347e-05,
      "loss": 0.6499,
      "step": 35
    },
    {
      "epoch": 1.0909090909090908,
      "grad_norm": 0.6441263157306794,
      "learning_rate": 1.9834617016337424e-05,
      "loss": 0.6398,
      "step": 36
    },
    {
      "epoch": 1.121212121212121,
      "grad_norm": 0.7250815731649851,
      "learning_rate": 1.9805997865694616e-05,
      "loss": 0.6458,
      "step": 37
    },
    {
      "epoch": 1.1515151515151516,
      "grad_norm": 0.5648787565713874,
      "learning_rate": 1.9775120098467212e-05,
      "loss": 0.6513,
      "step": 38
    },
    {
      "epoch": 1.1818181818181819,
      "grad_norm": 0.6618347481804527,
      "learning_rate": 1.9741990826734793e-05,
      "loss": 0.6575,
      "step": 39
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 0.6525414347292371,
      "learning_rate": 1.970661768116622e-05,
      "loss": 0.6474,
      "step": 40
    },
    {
      "epoch": 1.2424242424242424,
      "grad_norm": 0.5466846849498529,
      "learning_rate": 1.9669008809262064e-05,
      "loss": 0.6294,
      "step": 41
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 0.5349503407643637,
      "learning_rate": 1.9629172873477995e-05,
      "loss": 0.6395,
      "step": 42
    },
    {
      "epoch": 1.303030303030303,
      "grad_norm": 0.7006568418975037,
      "learning_rate": 1.9587119049229558e-05,
      "loss": 0.633,
      "step": 43
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.6309416888263866,
      "learning_rate": 1.954285702277879e-05,
      "loss": 0.6333,
      "step": 44
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.5219707273032846,
      "learning_rate": 1.9496396989003195e-05,
      "loss": 0.642,
      "step": 45
    },
    {
      "epoch": 1.393939393939394,
      "grad_norm": 0.5836074369317269,
      "learning_rate": 1.944774964904754e-05,
      "loss": 0.6534,
      "step": 46
    },
    {
      "epoch": 1.4242424242424243,
      "grad_norm": 0.5261881958409415,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 0.6228,
      "step": 47
    },
    {
      "epoch": 1.4545454545454546,
      "grad_norm": 0.5141375309869938,
      "learning_rate": 1.9343938371606714e-05,
      "loss": 0.6322,
      "step": 48
    },
    {
      "epoch": 1.4848484848484849,
      "grad_norm": 0.5899543300004795,
      "learning_rate": 1.9288798344984673e-05,
      "loss": 0.622,
      "step": 49
    },
    {
      "epoch": 1.5151515151515151,
      "grad_norm": 0.46491756803591755,
      "learning_rate": 1.9231518828401458e-05,
      "loss": 0.6277,
      "step": 50
    },
    {
      "epoch": 1.5454545454545454,
      "grad_norm": 0.6980721570667355,
      "learning_rate": 1.917211301505453e-05,
      "loss": 0.6332,
      "step": 51
    },
    {
      "epoch": 1.5757575757575757,
      "grad_norm": 0.7474545948655761,
      "learning_rate": 1.911059458789152e-05,
      "loss": 0.63,
      "step": 52
    },
    {
      "epoch": 1.606060606060606,
      "grad_norm": 0.5283125000371448,
      "learning_rate": 1.9046977716458627e-05,
      "loss": 0.6088,
      "step": 53
    },
    {
      "epoch": 1.6363636363636362,
      "grad_norm": 0.8442674047428046,
      "learning_rate": 1.8981277053636963e-05,
      "loss": 0.6262,
      "step": 54
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.6778644547078695,
      "learning_rate": 1.891350773226754e-05,
      "loss": 0.626,
      "step": 55
    },
    {
      "epoch": 1.696969696969697,
      "grad_norm": 0.5647718365118083,
      "learning_rate": 1.8843685361665724e-05,
      "loss": 0.6249,
      "step": 56
    },
    {
      "epoch": 1.7272727272727273,
      "grad_norm": 0.8321637711182053,
      "learning_rate": 1.8771826024025944e-05,
      "loss": 0.6217,
      "step": 57
    },
    {
      "epoch": 1.7575757575757576,
      "grad_norm": 0.5984562415271433,
      "learning_rate": 1.8697946270717468e-05,
      "loss": 0.6164,
      "step": 58
    },
    {
      "epoch": 1.7878787878787878,
      "grad_norm": 0.5362871503036654,
      "learning_rate": 1.8622063118472135e-05,
      "loss": 0.6134,
      "step": 59
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.694115726189206,
      "learning_rate": 1.8544194045464888e-05,
      "loss": 0.6289,
      "step": 60
    },
    {
      "epoch": 1.8484848484848486,
      "grad_norm": 0.4257323314171309,
      "learning_rate": 1.8464356987288012e-05,
      "loss": 0.6278,
      "step": 61
    },
    {
      "epoch": 1.878787878787879,
      "grad_norm": 0.6310792898943932,
      "learning_rate": 1.8382570332820045e-05,
      "loss": 0.6098,
      "step": 62
    },
    {
      "epoch": 1.9090909090909092,
      "grad_norm": 0.6327564395897852,
      "learning_rate": 1.8298852919990254e-05,
      "loss": 0.6304,
      "step": 63
    },
    {
      "epoch": 1.9393939393939394,
      "grad_norm": 0.5553665615377129,
      "learning_rate": 1.821322403143969e-05,
      "loss": 0.6351,
      "step": 64
    },
    {
      "epoch": 1.9696969696969697,
      "grad_norm": 0.5797777082750551,
      "learning_rate": 1.812570339007983e-05,
      "loss": 0.6241,
      "step": 65
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.5739316866857709,
      "learning_rate": 1.8036311154549783e-05,
      "loss": 0.6093,
      "step": 66
    },
    {
      "epoch": 2.0303030303030303,
      "grad_norm": 0.48121668015908287,
      "learning_rate": 1.7945067914573147e-05,
      "loss": 0.5747,
      "step": 67
    },
    {
      "epoch": 2.0606060606060606,
      "grad_norm": 0.5460930793951939,
      "learning_rate": 1.7851994686215592e-05,
      "loss": 0.5642,
      "step": 68
    },
    {
      "epoch": 2.090909090909091,
      "grad_norm": 0.5001197171400812,
      "learning_rate": 1.77571129070442e-05,
      "loss": 0.5608,
      "step": 69
    },
    {
      "epoch": 2.121212121212121,
      "grad_norm": 0.5298712565505415,
      "learning_rate": 1.766044443118978e-05,
      "loss": 0.5581,
      "step": 70
    },
    {
      "epoch": 2.1515151515151514,
      "grad_norm": 0.4966754793133404,
      "learning_rate": 1.7562011524313187e-05,
      "loss": 0.5395,
      "step": 71
    },
    {
      "epoch": 2.1818181818181817,
      "grad_norm": 0.5277085590864523,
      "learning_rate": 1.7461836858476858e-05,
      "loss": 0.5571,
      "step": 72
    },
    {
      "epoch": 2.212121212121212,
      "grad_norm": 0.5488445042923799,
      "learning_rate": 1.7359943506922775e-05,
      "loss": 0.5551,
      "step": 73
    },
    {
      "epoch": 2.242424242424242,
      "grad_norm": 0.5428582959844105,
      "learning_rate": 1.725635493875799e-05,
      "loss": 0.5408,
      "step": 74
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.5528520907823309,
      "learning_rate": 1.7151095013548996e-05,
      "loss": 0.5619,
      "step": 75
    },
    {
      "epoch": 2.303030303030303,
      "grad_norm": 0.5477695631744043,
      "learning_rate": 1.7044187975826126e-05,
      "loss": 0.5702,
      "step": 76
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.564006664627722,
      "learning_rate": 1.693565844949933e-05,
      "loss": 0.5558,
      "step": 77
    },
    {
      "epoch": 2.3636363636363638,
      "grad_norm": 0.5709183289059574,
      "learning_rate": 1.6825531432186545e-05,
      "loss": 0.5526,
      "step": 78
    },
    {
      "epoch": 2.393939393939394,
      "grad_norm": 0.4942184477441698,
      "learning_rate": 1.671383228945597e-05,
      "loss": 0.5583,
      "step": 79
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 0.6064793204713993,
      "learning_rate": 1.6600586748983642e-05,
      "loss": 0.5429,
      "step": 80
    },
    {
      "epoch": 2.4545454545454546,
      "grad_norm": 0.5630848920199718,
      "learning_rate": 1.648582089462756e-05,
      "loss": 0.5501,
      "step": 81
    },
    {
      "epoch": 2.484848484848485,
      "grad_norm": 0.5334325309353739,
      "learning_rate": 1.6369561160419783e-05,
      "loss": 0.5442,
      "step": 82
    },
    {
      "epoch": 2.515151515151515,
      "grad_norm": 0.707356267328791,
      "learning_rate": 1.625183432447789e-05,
      "loss": 0.5391,
      "step": 83
    },
    {
      "epoch": 2.5454545454545454,
      "grad_norm": 0.4985316010348116,
      "learning_rate": 1.6132667502837164e-05,
      "loss": 0.5588,
      "step": 84
    },
    {
      "epoch": 2.5757575757575757,
      "grad_norm": 0.5057913097761766,
      "learning_rate": 1.6012088143204953e-05,
      "loss": 0.5402,
      "step": 85
    },
    {
      "epoch": 2.606060606060606,
      "grad_norm": 0.5246827700768846,
      "learning_rate": 1.589012401863864e-05,
      "loss": 0.5444,
      "step": 86
    },
    {
      "epoch": 2.6363636363636362,
      "grad_norm": 0.509187918334316,
      "learning_rate": 1.5766803221148676e-05,
      "loss": 0.5437,
      "step": 87
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.44422082002239893,
      "learning_rate": 1.5642154155228124e-05,
      "loss": 0.5426,
      "step": 88
    },
    {
      "epoch": 2.6969696969696972,
      "grad_norm": 0.4920191158504586,
      "learning_rate": 1.5516205531310272e-05,
      "loss": 0.5473,
      "step": 89
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.48412504756311026,
      "learning_rate": 1.538898635915576e-05,
      "loss": 0.5465,
      "step": 90
    },
    {
      "epoch": 2.757575757575758,
      "grad_norm": 0.4530324304992249,
      "learning_rate": 1.526052594117071e-05,
      "loss": 0.5598,
      "step": 91
    },
    {
      "epoch": 2.787878787878788,
      "grad_norm": 0.5320237683268654,
      "learning_rate": 1.513085386565758e-05,
      "loss": 0.5622,
      "step": 92
    },
    {
      "epoch": 2.8181818181818183,
      "grad_norm": 0.4793867024944811,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.5243,
      "step": 93
    },
    {
      "epoch": 2.8484848484848486,
      "grad_norm": 0.4707390992093123,
      "learning_rate": 1.4867994483783485e-05,
      "loss": 0.5332,
      "step": 94
    },
    {
      "epoch": 2.878787878787879,
      "grad_norm": 0.6016765535088383,
      "learning_rate": 1.4734867721853341e-05,
      "loss": 0.5392,
      "step": 95
    },
    {
      "epoch": 2.909090909090909,
      "grad_norm": 0.5013462820403698,
      "learning_rate": 1.4600650377311523e-05,
      "loss": 0.5541,
      "step": 96
    },
    {
      "epoch": 2.9393939393939394,
      "grad_norm": 0.5276946510916265,
      "learning_rate": 1.4465373364454001e-05,
      "loss": 0.5374,
      "step": 97
    },
    {
      "epoch": 2.9696969696969697,
      "grad_norm": 0.4834947386949571,
      "learning_rate": 1.4329067841650274e-05,
      "loss": 0.5539,
      "step": 98
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.45297067920820455,
      "learning_rate": 1.4191765204166643e-05,
      "loss": 0.5261,
      "step": 99
    },
    {
      "epoch": 3.0303030303030303,
      "grad_norm": 0.5481169418218338,
      "learning_rate": 1.4053497076934948e-05,
      "loss": 0.5029,
      "step": 100
    },
    {
      "epoch": 3.0606060606060606,
      "grad_norm": 0.45899411605263574,
      "learning_rate": 1.3914295307268396e-05,
      "loss": 0.4882,
      "step": 101
    },
    {
      "epoch": 3.090909090909091,
      "grad_norm": 0.5174227385611566,
      "learning_rate": 1.3774191957526144e-05,
      "loss": 0.5109,
      "step": 102
    },
    {
      "epoch": 3.121212121212121,
      "grad_norm": 0.5426679323795233,
      "learning_rate": 1.3633219297728415e-05,
      "loss": 0.4902,
      "step": 103
    },
    {
      "epoch": 3.1515151515151514,
      "grad_norm": 0.5054394678723684,
      "learning_rate": 1.3491409798123687e-05,
      "loss": 0.4992,
      "step": 104
    },
    {
      "epoch": 3.1818181818181817,
      "grad_norm": 0.4868841687624966,
      "learning_rate": 1.3348796121709862e-05,
      "loss": 0.4754,
      "step": 105
    },
    {
      "epoch": 3.212121212121212,
      "grad_norm": 0.5013997004103273,
      "learning_rate": 1.3205411116710973e-05,
      "loss": 0.4717,
      "step": 106
    },
    {
      "epoch": 3.242424242424242,
      "grad_norm": 0.4517701958173837,
      "learning_rate": 1.3061287809011243e-05,
      "loss": 0.4753,
      "step": 107
    },
    {
      "epoch": 3.2727272727272725,
      "grad_norm": 0.4534115658280367,
      "learning_rate": 1.291645939454825e-05,
      "loss": 0.4817,
      "step": 108
    },
    {
      "epoch": 3.303030303030303,
      "grad_norm": 0.5585801528522834,
      "learning_rate": 1.277095923166689e-05,
      "loss": 0.4774,
      "step": 109
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.4498126532871032,
      "learning_rate": 1.2624820833435939e-05,
      "loss": 0.4604,
      "step": 110
    },
    {
      "epoch": 3.3636363636363638,
      "grad_norm": 0.4996368715048995,
      "learning_rate": 1.2478077859929e-05,
      "loss": 0.4778,
      "step": 111
    },
    {
      "epoch": 3.393939393939394,
      "grad_norm": 0.4980889806196696,
      "learning_rate": 1.2330764110471567e-05,
      "loss": 0.4903,
      "step": 112
    },
    {
      "epoch": 3.4242424242424243,
      "grad_norm": 0.48778762901817224,
      "learning_rate": 1.2182913515856016e-05,
      "loss": 0.4618,
      "step": 113
    },
    {
      "epoch": 3.4545454545454546,
      "grad_norm": 0.5664283015045994,
      "learning_rate": 1.2034560130526341e-05,
      "loss": 0.487,
      "step": 114
    },
    {
      "epoch": 3.484848484848485,
      "grad_norm": 0.4639978945171482,
      "learning_rate": 1.1885738124734359e-05,
      "loss": 0.4735,
      "step": 115
    },
    {
      "epoch": 3.515151515151515,
      "grad_norm": 0.5077293093403913,
      "learning_rate": 1.1736481776669307e-05,
      "loss": 0.4808,
      "step": 116
    },
    {
      "epoch": 3.5454545454545454,
      "grad_norm": 0.4169454288938421,
      "learning_rate": 1.1586825464562515e-05,
      "loss": 0.4669,
      "step": 117
    },
    {
      "epoch": 3.5757575757575757,
      "grad_norm": 0.44858098053776396,
      "learning_rate": 1.1436803658769082e-05,
      "loss": 0.476,
      "step": 118
    },
    {
      "epoch": 3.606060606060606,
      "grad_norm": 0.43444367089232144,
      "learning_rate": 1.1286450913828313e-05,
      "loss": 0.4728,
      "step": 119
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.46353379019353347,
      "learning_rate": 1.113580186050475e-05,
      "loss": 0.4783,
      "step": 120
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.4146543948691328,
      "learning_rate": 1.0984891197811686e-05,
      "loss": 0.487,
      "step": 121
    },
    {
      "epoch": 3.6969696969696972,
      "grad_norm": 0.440575475306285,
      "learning_rate": 1.0833753685018935e-05,
      "loss": 0.4798,
      "step": 122
    },
    {
      "epoch": 3.7272727272727275,
      "grad_norm": 0.44743831588295174,
      "learning_rate": 1.0682424133646712e-05,
      "loss": 0.4797,
      "step": 123
    },
    {
      "epoch": 3.757575757575758,
      "grad_norm": 0.43950461720911943,
      "learning_rate": 1.0530937399447496e-05,
      "loss": 0.4724,
      "step": 124
    },
    {
      "epoch": 3.787878787878788,
      "grad_norm": 0.3935680055615317,
      "learning_rate": 1.0379328374377715e-05,
      "loss": 0.4929,
      "step": 125
    },
    {
      "epoch": 3.8181818181818183,
      "grad_norm": 0.4670683819033576,
      "learning_rate": 1.0227631978561057e-05,
      "loss": 0.4918,
      "step": 126
    },
    {
      "epoch": 3.8484848484848486,
      "grad_norm": 0.3749075477068874,
      "learning_rate": 1.0075883152245334e-05,
      "loss": 0.473,
      "step": 127
    },
    {
      "epoch": 3.878787878787879,
      "grad_norm": 0.40782846624465935,
      "learning_rate": 9.92411684775467e-06,
      "loss": 0.4797,
      "step": 128
    },
    {
      "epoch": 3.909090909090909,
      "grad_norm": 0.3754701050008607,
      "learning_rate": 9.772368021438943e-06,
      "loss": 0.4803,
      "step": 129
    },
    {
      "epoch": 3.9393939393939394,
      "grad_norm": 0.4061074797288187,
      "learning_rate": 9.620671625622287e-06,
      "loss": 0.4704,
      "step": 130
    },
    {
      "epoch": 3.9696969696969697,
      "grad_norm": 0.404198184819985,
      "learning_rate": 9.469062600552509e-06,
      "loss": 0.4801,
      "step": 131
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.37575321539029605,
      "learning_rate": 9.317575866353293e-06,
      "loss": 0.4926,
      "step": 132
    },
    {
      "epoch": 4.03030303030303,
      "grad_norm": 0.5367885583971738,
      "learning_rate": 9.166246314981066e-06,
      "loss": 0.4255,
      "step": 133
    },
    {
      "epoch": 4.0606060606060606,
      "grad_norm": 0.4134811701043575,
      "learning_rate": 9.015108802188314e-06,
      "loss": 0.4255,
      "step": 134
    },
    {
      "epoch": 4.090909090909091,
      "grad_norm": 0.47090973879169806,
      "learning_rate": 8.86419813949525e-06,
      "loss": 0.4273,
      "step": 135
    },
    {
      "epoch": 4.121212121212121,
      "grad_norm": 0.514546447847447,
      "learning_rate": 8.71354908617169e-06,
      "loss": 0.4309,
      "step": 136
    },
    {
      "epoch": 4.151515151515151,
      "grad_norm": 0.5420751186949593,
      "learning_rate": 8.56319634123092e-06,
      "loss": 0.4202,
      "step": 137
    },
    {
      "epoch": 4.181818181818182,
      "grad_norm": 0.43931623053415303,
      "learning_rate": 8.413174535437486e-06,
      "loss": 0.4259,
      "step": 138
    },
    {
      "epoch": 4.212121212121212,
      "grad_norm": 0.5299447901602566,
      "learning_rate": 8.263518223330698e-06,
      "loss": 0.418,
      "step": 139
    },
    {
      "epoch": 4.242424242424242,
      "grad_norm": 0.4228893366358471,
      "learning_rate": 8.114261875265643e-06,
      "loss": 0.4119,
      "step": 140
    },
    {
      "epoch": 4.2727272727272725,
      "grad_norm": 0.4873011039417733,
      "learning_rate": 7.965439869473664e-06,
      "loss": 0.4445,
      "step": 141
    },
    {
      "epoch": 4.303030303030303,
      "grad_norm": 0.4417462254949149,
      "learning_rate": 7.817086484143987e-06,
      "loss": 0.4169,
      "step": 142
    },
    {
      "epoch": 4.333333333333333,
      "grad_norm": 0.4736957029714781,
      "learning_rate": 7.669235889528436e-06,
      "loss": 0.4339,
      "step": 143
    },
    {
      "epoch": 4.363636363636363,
      "grad_norm": 0.4370429919562305,
      "learning_rate": 7.521922140071003e-06,
      "loss": 0.4285,
      "step": 144
    },
    {
      "epoch": 4.393939393939394,
      "grad_norm": 0.4135913653729827,
      "learning_rate": 7.375179166564062e-06,
      "loss": 0.4132,
      "step": 145
    },
    {
      "epoch": 4.424242424242424,
      "grad_norm": 0.4349107770811642,
      "learning_rate": 7.2290407683331154e-06,
      "loss": 0.4034,
      "step": 146
    },
    {
      "epoch": 4.454545454545454,
      "grad_norm": 0.3689601811014803,
      "learning_rate": 7.0835406054517505e-06,
      "loss": 0.4167,
      "step": 147
    },
    {
      "epoch": 4.484848484848484,
      "grad_norm": 0.3919731493822646,
      "learning_rate": 6.93871219098876e-06,
      "loss": 0.4174,
      "step": 148
    },
    {
      "epoch": 4.515151515151516,
      "grad_norm": 0.3631359951340099,
      "learning_rate": 6.79458888328903e-06,
      "loss": 0.4119,
      "step": 149
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 0.3469055471359543,
      "learning_rate": 6.651203878290139e-06,
      "loss": 0.4118,
      "step": 150
    },
    {
      "epoch": 4.575757575757576,
      "grad_norm": 0.39077206938761855,
      "learning_rate": 6.508590201876317e-06,
      "loss": 0.4188,
      "step": 151
    },
    {
      "epoch": 4.606060606060606,
      "grad_norm": 0.35271378862736025,
      "learning_rate": 6.366780702271589e-06,
      "loss": 0.4074,
      "step": 152
    },
    {
      "epoch": 4.636363636363637,
      "grad_norm": 0.34874293152330893,
      "learning_rate": 6.225808042473857e-06,
      "loss": 0.4064,
      "step": 153
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.38610950294613744,
      "learning_rate": 6.085704692731609e-06,
      "loss": 0.4241,
      "step": 154
    },
    {
      "epoch": 4.696969696969697,
      "grad_norm": 0.340956715725986,
      "learning_rate": 5.946502923065054e-06,
      "loss": 0.4148,
      "step": 155
    },
    {
      "epoch": 4.7272727272727275,
      "grad_norm": 0.360923407917326,
      "learning_rate": 5.8082347958333625e-06,
      "loss": 0.4336,
      "step": 156
    },
    {
      "epoch": 4.757575757575758,
      "grad_norm": 0.35801997038496647,
      "learning_rate": 5.670932158349732e-06,
      "loss": 0.4263,
      "step": 157
    },
    {
      "epoch": 4.787878787878788,
      "grad_norm": 0.3950131131053667,
      "learning_rate": 5.534626635546e-06,
      "loss": 0.4386,
      "step": 158
    },
    {
      "epoch": 4.818181818181818,
      "grad_norm": 0.34407417429007875,
      "learning_rate": 5.399349622688479e-06,
      "loss": 0.4166,
      "step": 159
    },
    {
      "epoch": 4.848484848484849,
      "grad_norm": 0.33879406082435315,
      "learning_rate": 5.2651322781466606e-06,
      "loss": 0.4166,
      "step": 160
    },
    {
      "epoch": 4.878787878787879,
      "grad_norm": 0.37232220420006634,
      "learning_rate": 5.132005516216512e-06,
      "loss": 0.4241,
      "step": 161
    },
    {
      "epoch": 4.909090909090909,
      "grad_norm": 0.3607654704574992,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.4243,
      "step": 162
    },
    {
      "epoch": 4.9393939393939394,
      "grad_norm": 0.3414234600200536,
      "learning_rate": 4.869146134342426e-06,
      "loss": 0.4164,
      "step": 163
    },
    {
      "epoch": 4.96969696969697,
      "grad_norm": 0.3611583852859241,
      "learning_rate": 4.739474058829288e-06,
      "loss": 0.4213,
      "step": 164
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.36398905056957803,
      "learning_rate": 4.611013640844245e-06,
      "loss": 0.4093,
      "step": 165
    },
    {
      "epoch": 5.03030303030303,
      "grad_norm": 0.4948488800459289,
      "learning_rate": 4.483794468689728e-06,
      "loss": 0.3722,
      "step": 166
    },
    {
      "epoch": 5.0606060606060606,
      "grad_norm": 0.42211907809724325,
      "learning_rate": 4.357845844771881e-06,
      "loss": 0.3781,
      "step": 167
    },
    {
      "epoch": 5.090909090909091,
      "grad_norm": 0.36460880907743887,
      "learning_rate": 4.2331967788513295e-06,
      "loss": 0.3759,
      "step": 168
    },
    {
      "epoch": 5.121212121212121,
      "grad_norm": 0.45739180654643363,
      "learning_rate": 4.109875981361363e-06,
      "loss": 0.3886,
      "step": 169
    },
    {
      "epoch": 5.151515151515151,
      "grad_norm": 0.5110271991894739,
      "learning_rate": 3.987911856795047e-06,
      "loss": 0.3875,
      "step": 170
    },
    {
      "epoch": 5.181818181818182,
      "grad_norm": 0.4199897658711862,
      "learning_rate": 3.867332497162836e-06,
      "loss": 0.3761,
      "step": 171
    },
    {
      "epoch": 5.212121212121212,
      "grad_norm": 0.39558607930024225,
      "learning_rate": 3.748165675522113e-06,
      "loss": 0.3621,
      "step": 172
    },
    {
      "epoch": 5.242424242424242,
      "grad_norm": 0.4385916903985669,
      "learning_rate": 3.630438839580217e-06,
      "loss": 0.381,
      "step": 173
    },
    {
      "epoch": 5.2727272727272725,
      "grad_norm": 0.41362111804249807,
      "learning_rate": 3.5141791053724405e-06,
      "loss": 0.3747,
      "step": 174
    },
    {
      "epoch": 5.303030303030303,
      "grad_norm": 0.37632673730530525,
      "learning_rate": 3.399413251016359e-06,
      "loss": 0.3687,
      "step": 175
    },
    {
      "epoch": 5.333333333333333,
      "grad_norm": 0.348843813837047,
      "learning_rate": 3.2861677105440335e-06,
      "loss": 0.3757,
      "step": 176
    },
    {
      "epoch": 5.363636363636363,
      "grad_norm": 0.3801989643253093,
      "learning_rate": 3.174468567813461e-06,
      "loss": 0.3621,
      "step": 177
    },
    {
      "epoch": 5.393939393939394,
      "grad_norm": 0.41817834614685806,
      "learning_rate": 3.0643415505006733e-06,
      "loss": 0.3727,
      "step": 178
    },
    {
      "epoch": 5.424242424242424,
      "grad_norm": 0.38782415740688625,
      "learning_rate": 2.9558120241738786e-06,
      "loss": 0.3728,
      "step": 179
    },
    {
      "epoch": 5.454545454545454,
      "grad_norm": 0.39344313975000894,
      "learning_rate": 2.8489049864510053e-06,
      "loss": 0.3766,
      "step": 180
    },
    {
      "epoch": 5.484848484848484,
      "grad_norm": 0.3632628595301423,
      "learning_rate": 2.7436450612420098e-06,
      "loss": 0.3855,
      "step": 181
    },
    {
      "epoch": 5.515151515151516,
      "grad_norm": 0.37361035634295486,
      "learning_rate": 2.640056493077231e-06,
      "loss": 0.3867,
      "step": 182
    },
    {
      "epoch": 5.545454545454545,
      "grad_norm": 0.3695133015491995,
      "learning_rate": 2.5381631415231455e-06,
      "loss": 0.3733,
      "step": 183
    },
    {
      "epoch": 5.575757575757576,
      "grad_norm": 0.33131189012936435,
      "learning_rate": 2.4379884756868167e-06,
      "loss": 0.3736,
      "step": 184
    },
    {
      "epoch": 5.606060606060606,
      "grad_norm": 0.3071447272503216,
      "learning_rate": 2.339555568810221e-06,
      "loss": 0.3731,
      "step": 185
    },
    {
      "epoch": 5.636363636363637,
      "grad_norm": 0.3527068589774687,
      "learning_rate": 2.2428870929558012e-06,
      "loss": 0.3793,
      "step": 186
    },
    {
      "epoch": 5.666666666666667,
      "grad_norm": 0.34907167074247203,
      "learning_rate": 2.1480053137844115e-06,
      "loss": 0.3725,
      "step": 187
    },
    {
      "epoch": 5.696969696969697,
      "grad_norm": 0.33860261612720205,
      "learning_rate": 2.054932085426856e-06,
      "loss": 0.3907,
      "step": 188
    },
    {
      "epoch": 5.7272727272727275,
      "grad_norm": 0.29967223205264326,
      "learning_rate": 1.963688845450218e-06,
      "loss": 0.3777,
      "step": 189
    },
    {
      "epoch": 5.757575757575758,
      "grad_norm": 0.2945296124408129,
      "learning_rate": 1.8742966099201699e-06,
      "loss": 0.3823,
      "step": 190
    },
    {
      "epoch": 5.787878787878788,
      "grad_norm": 0.3115231018245738,
      "learning_rate": 1.7867759685603115e-06,
      "loss": 0.3599,
      "step": 191
    },
    {
      "epoch": 5.818181818181818,
      "grad_norm": 0.3054364488267583,
      "learning_rate": 1.7011470800097496e-06,
      "loss": 0.3601,
      "step": 192
    },
    {
      "epoch": 5.848484848484849,
      "grad_norm": 0.3032155039356207,
      "learning_rate": 1.6174296671799571e-06,
      "loss": 0.3751,
      "step": 193
    },
    {
      "epoch": 5.878787878787879,
      "grad_norm": 0.3172682225304549,
      "learning_rate": 1.5356430127119915e-06,
      "loss": 0.3851,
      "step": 194
    },
    {
      "epoch": 5.909090909090909,
      "grad_norm": 0.29648563210294665,
      "learning_rate": 1.4558059545351144e-06,
      "loss": 0.3751,
      "step": 195
    },
    {
      "epoch": 5.9393939393939394,
      "grad_norm": 0.31808432178394425,
      "learning_rate": 1.3779368815278648e-06,
      "loss": 0.3821,
      "step": 196
    },
    {
      "epoch": 5.96969696969697,
      "grad_norm": 0.30505296033012536,
      "learning_rate": 1.302053729282533e-06,
      "loss": 0.3771,
      "step": 197
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.3064904032183375,
      "learning_rate": 1.2281739759740575e-06,
      "loss": 0.3717,
      "step": 198
    },
    {
      "epoch": 6.03030303030303,
      "grad_norm": 0.3967787185171751,
      "learning_rate": 1.156314638334277e-06,
      "loss": 0.354,
      "step": 199
    },
    {
      "epoch": 6.0606060606060606,
      "grad_norm": 0.3437958942707737,
      "learning_rate": 1.086492267732462e-06,
      "loss": 0.3511,
      "step": 200
    },
    {
      "epoch": 6.090909090909091,
      "grad_norm": 0.31356655980843884,
      "learning_rate": 1.01872294636304e-06,
      "loss": 0.3533,
      "step": 201
    },
    {
      "epoch": 6.121212121212121,
      "grad_norm": 0.2886267653194944,
      "learning_rate": 9.530222835413739e-07,
      "loss": 0.3615,
      "step": 202
    },
    {
      "epoch": 6.151515151515151,
      "grad_norm": 0.3116161477520254,
      "learning_rate": 8.894054121084839e-07,
      "loss": 0.3486,
      "step": 203
    },
    {
      "epoch": 6.181818181818182,
      "grad_norm": 0.33419488923552587,
      "learning_rate": 8.278869849454718e-07,
      "loss": 0.3575,
      "step": 204
    },
    {
      "epoch": 6.212121212121212,
      "grad_norm": 0.32103937428496787,
      "learning_rate": 7.684811715985429e-07,
      "loss": 0.3412,
      "step": 205
    },
    {
      "epoch": 6.242424242424242,
      "grad_norm": 0.32052991478008325,
      "learning_rate": 7.1120165501533e-07,
      "loss": 0.3565,
      "step": 206
    },
    {
      "epoch": 6.2727272727272725,
      "grad_norm": 0.31054666814946663,
      "learning_rate": 6.560616283932897e-07,
      "loss": 0.3458,
      "step": 207
    },
    {
      "epoch": 6.303030303030303,
      "grad_norm": 0.3165690321551478,
      "learning_rate": 6.030737921409169e-07,
      "loss": 0.3544,
      "step": 208
    },
    {
      "epoch": 6.333333333333333,
      "grad_norm": 0.2954322262678854,
      "learning_rate": 5.522503509524591e-07,
      "loss": 0.3648,
      "step": 209
    },
    {
      "epoch": 6.363636363636363,
      "grad_norm": 0.31093404660265733,
      "learning_rate": 5.036030109968082e-07,
      "loss": 0.356,
      "step": 210
    },
    {
      "epoch": 6.393939393939394,
      "grad_norm": 0.29907287442459096,
      "learning_rate": 4.5714297722121105e-07,
      "loss": 0.3557,
      "step": 211
    },
    {
      "epoch": 6.424242424242424,
      "grad_norm": 0.2953002467444844,
      "learning_rate": 4.128809507704445e-07,
      "loss": 0.3532,
      "step": 212
    },
    {
      "epoch": 6.454545454545454,
      "grad_norm": 0.28570512731817044,
      "learning_rate": 3.708271265220087e-07,
      "loss": 0.3583,
      "step": 213
    },
    {
      "epoch": 6.484848484848484,
      "grad_norm": 0.2861735599470133,
      "learning_rate": 3.309911907379393e-07,
      "loss": 0.3627,
      "step": 214
    },
    {
      "epoch": 6.515151515151516,
      "grad_norm": 0.28915862953538735,
      "learning_rate": 2.9338231883378365e-07,
      "loss": 0.3596,
      "step": 215
    },
    {
      "epoch": 6.545454545454545,
      "grad_norm": 0.2976155820262455,
      "learning_rate": 2.5800917326521013e-07,
      "loss": 0.3528,
      "step": 216
    },
    {
      "epoch": 6.575757575757576,
      "grad_norm": 0.28593077903301745,
      "learning_rate": 2.248799015327907e-07,
      "loss": 0.3518,
      "step": 217
    },
    {
      "epoch": 6.606060606060606,
      "grad_norm": 0.2960154774334222,
      "learning_rate": 1.9400213430538773e-07,
      "loss": 0.3461,
      "step": 218
    },
    {
      "epoch": 6.636363636363637,
      "grad_norm": 0.2932117694224571,
      "learning_rate": 1.6538298366257975e-07,
      "loss": 0.3507,
      "step": 219
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.3019685222454801,
      "learning_rate": 1.3902904145653094e-07,
      "loss": 0.35,
      "step": 220
    },
    {
      "epoch": 6.696969696969697,
      "grad_norm": 0.3012454025850701,
      "learning_rate": 1.1494637779369766e-07,
      "loss": 0.3566,
      "step": 221
    },
    {
      "epoch": 6.7272727272727275,
      "grad_norm": 0.2829397007387986,
      "learning_rate": 9.314053963669245e-08,
      "loss": 0.35,
      "step": 222
    },
    {
      "epoch": 6.757575757575758,
      "grad_norm": 0.2895958625182457,
      "learning_rate": 7.361654952665608e-08,
      "loss": 0.3503,
      "step": 223
    },
    {
      "epoch": 6.787878787878788,
      "grad_norm": 0.29334862596602407,
      "learning_rate": 5.637890442641403e-08,
      "loss": 0.3513,
      "step": 224
    },
    {
      "epoch": 6.818181818181818,
      "grad_norm": 0.2964239759812529,
      "learning_rate": 4.143157468468717e-08,
      "loss": 0.3596,
      "step": 225
    },
    {
      "epoch": 6.848484848484849,
      "grad_norm": 0.2945301367659038,
      "learning_rate": 2.8778003121607834e-08,
      "loss": 0.3558,
      "step": 226
    },
    {
      "epoch": 6.878787878787879,
      "grad_norm": 0.2869431206167948,
      "learning_rate": 1.8421104235727406e-08,
      "loss": 0.3598,
      "step": 227
    },
    {
      "epoch": 6.909090909090909,
      "grad_norm": 0.28064804251871384,
      "learning_rate": 1.0363263532724433e-08,
      "loss": 0.3507,
      "step": 228
    },
    {
      "epoch": 6.9393939393939394,
      "grad_norm": 0.2875195405711857,
      "learning_rate": 4.606336975948589e-09,
      "loss": 0.355,
      "step": 229
    },
    {
      "epoch": 6.96969696969697,
      "grad_norm": 0.2951749764292389,
      "learning_rate": 1.1516505589381777e-09,
      "loss": 0.3499,
      "step": 230
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.2854545515700229,
      "learning_rate": 0.0,
      "loss": 0.3548,
      "step": 231
    },
    {
      "epoch": 7.0,
      "step": 231,
      "total_flos": 6.232439949602324e+17,
      "train_loss": 0.519667737024687,
      "train_runtime": 9192.6216,
      "train_samples_per_second": 2.406,
      "train_steps_per_second": 0.025
    }
  ],
  "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": 6.232439949602324e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}