{
  "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": 5.792642134780677,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.9515,
      "step": 1
    },
    {
      "epoch": 0.06060606060606061,
      "grad_norm": 6.382571164842765,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.0119,
      "step": 2
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 6.02036402240374,
      "learning_rate": 2.5e-06,
      "loss": 0.9728,
      "step": 3
    },
    {
      "epoch": 0.12121212121212122,
      "grad_norm": 5.818167379345609,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.0002,
      "step": 4
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 4.482978716814255,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.9005,
      "step": 5
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 2.2752195885468782,
      "learning_rate": 5e-06,
      "loss": 0.8823,
      "step": 6
    },
    {
      "epoch": 0.21212121212121213,
      "grad_norm": 2.007755284237358,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.8801,
      "step": 7
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 3.9023239936341967,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.8614,
      "step": 8
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 4.581155266714506,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.8856,
      "step": 9
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 4.4563639683308,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.8762,
      "step": 10
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 4.036284479049859,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.8418,
      "step": 11
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 3.5386084914228837,
      "learning_rate": 1e-05,
      "loss": 0.838,
      "step": 12
    },
    {
      "epoch": 0.3939393939393939,
      "grad_norm": 2.8027022946203317,
      "learning_rate": 1.0833333333333334e-05,
      "loss": 0.8285,
      "step": 13
    },
    {
      "epoch": 0.42424242424242425,
      "grad_norm": 1.9204210335828813,
      "learning_rate": 1.1666666666666668e-05,
      "loss": 0.7153,
      "step": 14
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 2.3107282485007232,
      "learning_rate": 1.25e-05,
      "loss": 0.7976,
      "step": 15
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 2.320207582587203,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.7645,
      "step": 16
    },
    {
      "epoch": 0.5151515151515151,
      "grad_norm": 1.990131101081182,
      "learning_rate": 1.416666666666667e-05,
      "loss": 0.752,
      "step": 17
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 1.522500998675586,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.7225,
      "step": 18
    },
    {
      "epoch": 0.5757575757575758,
      "grad_norm": 1.3644809685735186,
      "learning_rate": 1.5833333333333333e-05,
      "loss": 0.6978,
      "step": 19
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 1.5223710023668546,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.7086,
      "step": 20
    },
    {
      "epoch": 0.6363636363636364,
      "grad_norm": 1.3179520412189247,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 0.7108,
      "step": 21
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 1.032339489893466,
      "learning_rate": 1.8333333333333333e-05,
      "loss": 0.7159,
      "step": 22
    },
    {
      "epoch": 0.696969696969697,
      "grad_norm": 1.2588368994602428,
      "learning_rate": 1.916666666666667e-05,
      "loss": 0.7061,
      "step": 23
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.901121939524265,
      "learning_rate": 2e-05,
      "loss": 0.6796,
      "step": 24
    },
    {
      "epoch": 0.7575757575757576,
      "grad_norm": 0.7528128278120151,
      "learning_rate": 1.999884834944106e-05,
      "loss": 0.6993,
      "step": 25
    },
    {
      "epoch": 0.7878787878787878,
      "grad_norm": 0.9518503285882077,
      "learning_rate": 1.9995393663024054e-05,
      "loss": 0.6822,
      "step": 26
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 0.7573600517508814,
      "learning_rate": 1.9989636736467278e-05,
      "loss": 0.6453,
      "step": 27
    },
    {
      "epoch": 0.8484848484848485,
      "grad_norm": 0.9702144938750634,
      "learning_rate": 1.9981578895764272e-05,
      "loss": 0.6884,
      "step": 28
    },
    {
      "epoch": 0.8787878787878788,
      "grad_norm": 0.6893913942952646,
      "learning_rate": 1.9971221996878395e-05,
      "loss": 0.6516,
      "step": 29
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.8481568719894347,
      "learning_rate": 1.9958568425315316e-05,
      "loss": 0.6851,
      "step": 30
    },
    {
      "epoch": 0.9393939393939394,
      "grad_norm": 0.6860057833899889,
      "learning_rate": 1.9943621095573588e-05,
      "loss": 0.6661,
      "step": 31
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.6698001122649556,
      "learning_rate": 1.9926383450473344e-05,
      "loss": 0.6725,
      "step": 32
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.6344887887916567,
      "learning_rate": 1.9906859460363307e-05,
      "loss": 0.6535,
      "step": 33
    },
    {
      "epoch": 1.0303030303030303,
      "grad_norm": 0.6820742340475162,
      "learning_rate": 1.9885053622206305e-05,
      "loss": 0.5658,
      "step": 34
    },
    {
      "epoch": 1.0606060606060606,
      "grad_norm": 0.6198289923893283,
      "learning_rate": 1.986097095854347e-05,
      "loss": 0.5686,
      "step": 35
    },
    {
      "epoch": 1.0909090909090908,
      "grad_norm": 0.6329798983022055,
      "learning_rate": 1.9834617016337424e-05,
      "loss": 0.5851,
      "step": 36
    },
    {
      "epoch": 1.121212121212121,
      "grad_norm": 0.7775710564857078,
      "learning_rate": 1.9805997865694616e-05,
      "loss": 0.5805,
      "step": 37
    },
    {
      "epoch": 1.1515151515151516,
      "grad_norm": 0.5892214749020198,
      "learning_rate": 1.9775120098467212e-05,
      "loss": 0.5727,
      "step": 38
    },
    {
      "epoch": 1.1818181818181819,
      "grad_norm": 0.7142145316951859,
      "learning_rate": 1.9741990826734793e-05,
      "loss": 0.5988,
      "step": 39
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 0.7240893818720315,
      "learning_rate": 1.970661768116622e-05,
      "loss": 0.6047,
      "step": 40
    },
    {
      "epoch": 1.2424242424242424,
      "grad_norm": 1.3045892213155967,
      "learning_rate": 1.9669008809262064e-05,
      "loss": 0.5963,
      "step": 41
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 1.570873477257169,
      "learning_rate": 1.9629172873477995e-05,
      "loss": 0.5686,
      "step": 42
    },
    {
      "epoch": 1.303030303030303,
      "grad_norm": 0.8582320942548944,
      "learning_rate": 1.9587119049229558e-05,
      "loss": 0.5732,
      "step": 43
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.6072012475804945,
      "learning_rate": 1.954285702277879e-05,
      "loss": 0.5893,
      "step": 44
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.746084738873446,
      "learning_rate": 1.9496396989003195e-05,
      "loss": 0.5564,
      "step": 45
    },
    {
      "epoch": 1.393939393939394,
      "grad_norm": 0.7519606714903169,
      "learning_rate": 1.944774964904754e-05,
      "loss": 0.5808,
      "step": 46
    },
    {
      "epoch": 1.4242424242424243,
      "grad_norm": 2.5218024682528775,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 0.6085,
      "step": 47
    },
    {
      "epoch": 1.4545454545454546,
      "grad_norm": 0.7953052370583591,
      "learning_rate": 1.9343938371606714e-05,
      "loss": 0.5248,
      "step": 48
    },
    {
      "epoch": 1.4848484848484849,
      "grad_norm": 0.6817458558899578,
      "learning_rate": 1.9288798344984673e-05,
      "loss": 0.5508,
      "step": 49
    },
    {
      "epoch": 1.5151515151515151,
      "grad_norm": 0.7926918339374145,
      "learning_rate": 1.9231518828401458e-05,
      "loss": 0.577,
      "step": 50
    },
    {
      "epoch": 1.5454545454545454,
      "grad_norm": 0.8365570273338034,
      "learning_rate": 1.917211301505453e-05,
      "loss": 0.5833,
      "step": 51
    },
    {
      "epoch": 1.5757575757575757,
      "grad_norm": 0.6280132538822605,
      "learning_rate": 1.911059458789152e-05,
      "loss": 0.5598,
      "step": 52
    },
    {
      "epoch": 1.606060606060606,
      "grad_norm": 0.7504572682561476,
      "learning_rate": 1.9046977716458627e-05,
      "loss": 0.5543,
      "step": 53
    },
    {
      "epoch": 1.6363636363636362,
      "grad_norm": 0.6649032045086283,
      "learning_rate": 1.8981277053636963e-05,
      "loss": 0.5485,
      "step": 54
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.6158934656649441,
      "learning_rate": 1.891350773226754e-05,
      "loss": 0.552,
      "step": 55
    },
    {
      "epoch": 1.696969696969697,
      "grad_norm": 0.5578253751825053,
      "learning_rate": 1.8843685361665724e-05,
      "loss": 0.537,
      "step": 56
    },
    {
      "epoch": 1.7272727272727273,
      "grad_norm": 0.6329112649004469,
      "learning_rate": 1.8771826024025944e-05,
      "loss": 0.537,
      "step": 57
    },
    {
      "epoch": 1.7575757575757576,
      "grad_norm": 0.6850791050938211,
      "learning_rate": 1.8697946270717468e-05,
      "loss": 0.5955,
      "step": 58
    },
    {
      "epoch": 1.7878787878787878,
      "grad_norm": 0.6033750071365439,
      "learning_rate": 1.8622063118472135e-05,
      "loss": 0.5586,
      "step": 59
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.673718575643345,
      "learning_rate": 1.8544194045464888e-05,
      "loss": 0.5666,
      "step": 60
    },
    {
      "epoch": 1.8484848484848486,
      "grad_norm": 0.5624746929294268,
      "learning_rate": 1.8464356987288012e-05,
      "loss": 0.5598,
      "step": 61
    },
    {
      "epoch": 1.878787878787879,
      "grad_norm": 0.6211999431525035,
      "learning_rate": 1.8382570332820045e-05,
      "loss": 0.5824,
      "step": 62
    },
    {
      "epoch": 1.9090909090909092,
      "grad_norm": 0.4958772347998872,
      "learning_rate": 1.8298852919990254e-05,
      "loss": 0.5464,
      "step": 63
    },
    {
      "epoch": 1.9393939393939394,
      "grad_norm": 0.534752866759772,
      "learning_rate": 1.821322403143969e-05,
      "loss": 0.5312,
      "step": 64
    },
    {
      "epoch": 1.9696969696969697,
      "grad_norm": 0.5621388930667433,
      "learning_rate": 1.812570339007983e-05,
      "loss": 0.5498,
      "step": 65
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.5381122674752019,
      "learning_rate": 1.8036311154549783e-05,
      "loss": 0.5427,
      "step": 66
    },
    {
      "epoch": 2.0303030303030303,
      "grad_norm": 0.7578168001157617,
      "learning_rate": 1.7945067914573147e-05,
      "loss": 0.5067,
      "step": 67
    },
    {
      "epoch": 2.0606060606060606,
      "grad_norm": 0.6446804403966812,
      "learning_rate": 1.7851994686215592e-05,
      "loss": 0.4755,
      "step": 68
    },
    {
      "epoch": 2.090909090909091,
      "grad_norm": 0.6754071384034545,
      "learning_rate": 1.77571129070442e-05,
      "loss": 0.449,
      "step": 69
    },
    {
      "epoch": 2.121212121212121,
      "grad_norm": 0.7433159460035733,
      "learning_rate": 1.766044443118978e-05,
      "loss": 0.4546,
      "step": 70
    },
    {
      "epoch": 2.1515151515151514,
      "grad_norm": 0.6286732134939782,
      "learning_rate": 1.7562011524313187e-05,
      "loss": 0.4522,
      "step": 71
    },
    {
      "epoch": 2.1818181818181817,
      "grad_norm": 0.6931242528492747,
      "learning_rate": 1.7461836858476858e-05,
      "loss": 0.4363,
      "step": 72
    },
    {
      "epoch": 2.212121212121212,
      "grad_norm": 0.6228303093581359,
      "learning_rate": 1.7359943506922775e-05,
      "loss": 0.4394,
      "step": 73
    },
    {
      "epoch": 2.242424242424242,
      "grad_norm": 0.6566264031206333,
      "learning_rate": 1.725635493875799e-05,
      "loss": 0.4371,
      "step": 74
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.5927282396336451,
      "learning_rate": 1.7151095013548996e-05,
      "loss": 0.4303,
      "step": 75
    },
    {
      "epoch": 2.303030303030303,
      "grad_norm": 0.7264098258400549,
      "learning_rate": 1.7044187975826126e-05,
      "loss": 0.4378,
      "step": 76
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.5494574539456141,
      "learning_rate": 1.693565844949933e-05,
      "loss": 0.4364,
      "step": 77
    },
    {
      "epoch": 2.3636363636363638,
      "grad_norm": 0.586478823644888,
      "learning_rate": 1.6825531432186545e-05,
      "loss": 0.4187,
      "step": 78
    },
    {
      "epoch": 2.393939393939394,
      "grad_norm": 0.5934153390296679,
      "learning_rate": 1.671383228945597e-05,
      "loss": 0.4262,
      "step": 79
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 0.502863696614169,
      "learning_rate": 1.6600586748983642e-05,
      "loss": 0.4201,
      "step": 80
    },
    {
      "epoch": 2.4545454545454546,
      "grad_norm": 0.567123847809039,
      "learning_rate": 1.648582089462756e-05,
      "loss": 0.4397,
      "step": 81
    },
    {
      "epoch": 2.484848484848485,
      "grad_norm": 0.5616400197858715,
      "learning_rate": 1.6369561160419783e-05,
      "loss": 0.4339,
      "step": 82
    },
    {
      "epoch": 2.515151515151515,
      "grad_norm": 0.5155130770072307,
      "learning_rate": 1.625183432447789e-05,
      "loss": 0.4391,
      "step": 83
    },
    {
      "epoch": 2.5454545454545454,
      "grad_norm": 0.5529983074318092,
      "learning_rate": 1.6132667502837164e-05,
      "loss": 0.4264,
      "step": 84
    },
    {
      "epoch": 2.5757575757575757,
      "grad_norm": 0.5181135590344216,
      "learning_rate": 1.6012088143204953e-05,
      "loss": 0.4381,
      "step": 85
    },
    {
      "epoch": 2.606060606060606,
      "grad_norm": 0.500195563035196,
      "learning_rate": 1.589012401863864e-05,
      "loss": 0.411,
      "step": 86
    },
    {
      "epoch": 2.6363636363636362,
      "grad_norm": 0.6387564048552598,
      "learning_rate": 1.5766803221148676e-05,
      "loss": 0.406,
      "step": 87
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.5102855942894612,
      "learning_rate": 1.5642154155228124e-05,
      "loss": 0.4489,
      "step": 88
    },
    {
      "epoch": 2.6969696969696972,
      "grad_norm": 0.626186159413465,
      "learning_rate": 1.5516205531310272e-05,
      "loss": 0.4373,
      "step": 89
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.533918204813247,
      "learning_rate": 1.538898635915576e-05,
      "loss": 0.4364,
      "step": 90
    },
    {
      "epoch": 2.757575757575758,
      "grad_norm": 0.5826020156535429,
      "learning_rate": 1.526052594117071e-05,
      "loss": 0.4633,
      "step": 91
    },
    {
      "epoch": 2.787878787878788,
      "grad_norm": 0.5268233366892546,
      "learning_rate": 1.513085386565758e-05,
      "loss": 0.4288,
      "step": 92
    },
    {
      "epoch": 2.8181818181818183,
      "grad_norm": 0.4989207490329867,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.4158,
      "step": 93
    },
    {
      "epoch": 2.8484848484848486,
      "grad_norm": 0.5392513240813872,
      "learning_rate": 1.4867994483783485e-05,
      "loss": 0.4493,
      "step": 94
    },
    {
      "epoch": 2.878787878787879,
      "grad_norm": 0.5412547971942594,
      "learning_rate": 1.4734867721853341e-05,
      "loss": 0.4333,
      "step": 95
    },
    {
      "epoch": 2.909090909090909,
      "grad_norm": 0.47787701781989034,
      "learning_rate": 1.4600650377311523e-05,
      "loss": 0.436,
      "step": 96
    },
    {
      "epoch": 2.9393939393939394,
      "grad_norm": 0.4738010394162418,
      "learning_rate": 1.4465373364454001e-05,
      "loss": 0.4434,
      "step": 97
    },
    {
      "epoch": 2.9696969696969697,
      "grad_norm": 0.4759665355689762,
      "learning_rate": 1.4329067841650274e-05,
      "loss": 0.4292,
      "step": 98
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.5347888280085701,
      "learning_rate": 1.4191765204166643e-05,
      "loss": 0.4458,
      "step": 99
    },
    {
      "epoch": 3.0303030303030303,
      "grad_norm": 0.7578730318403921,
      "learning_rate": 1.4053497076934948e-05,
      "loss": 0.3589,
      "step": 100
    },
    {
      "epoch": 3.0606060606060606,
      "grad_norm": 0.5590827351296772,
      "learning_rate": 1.3914295307268396e-05,
      "loss": 0.3522,
      "step": 101
    },
    {
      "epoch": 3.090909090909091,
      "grad_norm": 0.7663017214456678,
      "learning_rate": 1.3774191957526144e-05,
      "loss": 0.3412,
      "step": 102
    },
    {
      "epoch": 3.121212121212121,
      "grad_norm": 0.7255267139613365,
      "learning_rate": 1.3633219297728415e-05,
      "loss": 0.3138,
      "step": 103
    },
    {
      "epoch": 3.1515151515151514,
      "grad_norm": 0.5953528568597565,
      "learning_rate": 1.3491409798123687e-05,
      "loss": 0.3325,
      "step": 104
    },
    {
      "epoch": 3.1818181818181817,
      "grad_norm": 0.5958107979593327,
      "learning_rate": 1.3348796121709862e-05,
      "loss": 0.3488,
      "step": 105
    },
    {
      "epoch": 3.212121212121212,
      "grad_norm": 0.6097699089574262,
      "learning_rate": 1.3205411116710973e-05,
      "loss": 0.3366,
      "step": 106
    },
    {
      "epoch": 3.242424242424242,
      "grad_norm": 0.525552536797333,
      "learning_rate": 1.3061287809011243e-05,
      "loss": 0.3354,
      "step": 107
    },
    {
      "epoch": 3.2727272727272725,
      "grad_norm": 0.5107168278136667,
      "learning_rate": 1.291645939454825e-05,
      "loss": 0.3255,
      "step": 108
    },
    {
      "epoch": 3.303030303030303,
      "grad_norm": 0.49562286761158575,
      "learning_rate": 1.277095923166689e-05,
      "loss": 0.3226,
      "step": 109
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.552564963270535,
      "learning_rate": 1.2624820833435939e-05,
      "loss": 0.3262,
      "step": 110
    },
    {
      "epoch": 3.3636363636363638,
      "grad_norm": 0.529860518763179,
      "learning_rate": 1.2478077859929e-05,
      "loss": 0.3223,
      "step": 111
    },
    {
      "epoch": 3.393939393939394,
      "grad_norm": 0.46705898403530477,
      "learning_rate": 1.2330764110471567e-05,
      "loss": 0.3148,
      "step": 112
    },
    {
      "epoch": 3.4242424242424243,
      "grad_norm": 0.5282623236299218,
      "learning_rate": 1.2182913515856016e-05,
      "loss": 0.3136,
      "step": 113
    },
    {
      "epoch": 3.4545454545454546,
      "grad_norm": 0.4736109046736109,
      "learning_rate": 1.2034560130526341e-05,
      "loss": 0.3209,
      "step": 114
    },
    {
      "epoch": 3.484848484848485,
      "grad_norm": 0.49962304598092333,
      "learning_rate": 1.1885738124734359e-05,
      "loss": 0.3027,
      "step": 115
    },
    {
      "epoch": 3.515151515151515,
      "grad_norm": 0.4433582001575731,
      "learning_rate": 1.1736481776669307e-05,
      "loss": 0.3217,
      "step": 116
    },
    {
      "epoch": 3.5454545454545454,
      "grad_norm": 0.5171221122870995,
      "learning_rate": 1.1586825464562515e-05,
      "loss": 0.3321,
      "step": 117
    },
    {
      "epoch": 3.5757575757575757,
      "grad_norm": 0.456601993523051,
      "learning_rate": 1.1436803658769082e-05,
      "loss": 0.3274,
      "step": 118
    },
    {
      "epoch": 3.606060606060606,
      "grad_norm": 0.48920924636407565,
      "learning_rate": 1.1286450913828313e-05,
      "loss": 0.332,
      "step": 119
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.48147073171548743,
      "learning_rate": 1.113580186050475e-05,
      "loss": 0.3165,
      "step": 120
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.5082342266188418,
      "learning_rate": 1.0984891197811686e-05,
      "loss": 0.3147,
      "step": 121
    },
    {
      "epoch": 3.6969696969696972,
      "grad_norm": 0.49947475460015794,
      "learning_rate": 1.0833753685018935e-05,
      "loss": 0.3195,
      "step": 122
    },
    {
      "epoch": 3.7272727272727275,
      "grad_norm": 0.4902554646083638,
      "learning_rate": 1.0682424133646712e-05,
      "loss": 0.328,
      "step": 123
    },
    {
      "epoch": 3.757575757575758,
      "grad_norm": 0.509057304468142,
      "learning_rate": 1.0530937399447496e-05,
      "loss": 0.3366,
      "step": 124
    },
    {
      "epoch": 3.787878787878788,
      "grad_norm": 0.5628445523146441,
      "learning_rate": 1.0379328374377715e-05,
      "loss": 0.3351,
      "step": 125
    },
    {
      "epoch": 3.8181818181818183,
      "grad_norm": 0.4412802225534068,
      "learning_rate": 1.0227631978561057e-05,
      "loss": 0.3174,
      "step": 126
    },
    {
      "epoch": 3.8484848484848486,
      "grad_norm": 0.45045981739898905,
      "learning_rate": 1.0075883152245334e-05,
      "loss": 0.3253,
      "step": 127
    },
    {
      "epoch": 3.878787878787879,
      "grad_norm": 0.4952983553162171,
      "learning_rate": 9.92411684775467e-06,
      "loss": 0.317,
      "step": 128
    },
    {
      "epoch": 3.909090909090909,
      "grad_norm": 0.4670258222227986,
      "learning_rate": 9.772368021438943e-06,
      "loss": 0.3304,
      "step": 129
    },
    {
      "epoch": 3.9393939393939394,
      "grad_norm": 0.45909455329725174,
      "learning_rate": 9.620671625622287e-06,
      "loss": 0.3136,
      "step": 130
    },
    {
      "epoch": 3.9696969696969697,
      "grad_norm": 0.5183779437816718,
      "learning_rate": 9.469062600552509e-06,
      "loss": 0.3305,
      "step": 131
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.43030496359434006,
      "learning_rate": 9.317575866353293e-06,
      "loss": 0.3233,
      "step": 132
    },
    {
      "epoch": 4.03030303030303,
      "grad_norm": 0.6551297453698638,
      "learning_rate": 9.166246314981066e-06,
      "loss": 0.245,
      "step": 133
    },
    {
      "epoch": 4.0606060606060606,
      "grad_norm": 0.5346447826258135,
      "learning_rate": 9.015108802188314e-06,
      "loss": 0.2484,
      "step": 134
    },
    {
      "epoch": 4.090909090909091,
      "grad_norm": 0.5129787144325263,
      "learning_rate": 8.86419813949525e-06,
      "loss": 0.2442,
      "step": 135
    },
    {
      "epoch": 4.121212121212121,
      "grad_norm": 0.8307262099481713,
      "learning_rate": 8.71354908617169e-06,
      "loss": 0.2453,
      "step": 136
    },
    {
      "epoch": 4.151515151515151,
      "grad_norm": 0.5648300076341694,
      "learning_rate": 8.56319634123092e-06,
      "loss": 0.2147,
      "step": 137
    },
    {
      "epoch": 4.181818181818182,
      "grad_norm": 0.5175008944201702,
      "learning_rate": 8.413174535437486e-06,
      "loss": 0.2384,
      "step": 138
    },
    {
      "epoch": 4.212121212121212,
      "grad_norm": 0.5434304177721216,
      "learning_rate": 8.263518223330698e-06,
      "loss": 0.2385,
      "step": 139
    },
    {
      "epoch": 4.242424242424242,
      "grad_norm": 0.5656126676103714,
      "learning_rate": 8.114261875265643e-06,
      "loss": 0.2382,
      "step": 140
    },
    {
      "epoch": 4.2727272727272725,
      "grad_norm": 0.47834029227182495,
      "learning_rate": 7.965439869473664e-06,
      "loss": 0.248,
      "step": 141
    },
    {
      "epoch": 4.303030303030303,
      "grad_norm": 0.45321812692904295,
      "learning_rate": 7.817086484143987e-06,
      "loss": 0.2108,
      "step": 142
    },
    {
      "epoch": 4.333333333333333,
      "grad_norm": 0.5017557961334287,
      "learning_rate": 7.669235889528436e-06,
      "loss": 0.2351,
      "step": 143
    },
    {
      "epoch": 4.363636363636363,
      "grad_norm": 0.5127175862774535,
      "learning_rate": 7.521922140071003e-06,
      "loss": 0.2381,
      "step": 144
    },
    {
      "epoch": 4.393939393939394,
      "grad_norm": 0.5064799180026548,
      "learning_rate": 7.375179166564062e-06,
      "loss": 0.2502,
      "step": 145
    },
    {
      "epoch": 4.424242424242424,
      "grad_norm": 0.4606421137621233,
      "learning_rate": 7.2290407683331154e-06,
      "loss": 0.2264,
      "step": 146
    },
    {
      "epoch": 4.454545454545454,
      "grad_norm": 0.42303923096285145,
      "learning_rate": 7.0835406054517505e-06,
      "loss": 0.2203,
      "step": 147
    },
    {
      "epoch": 4.484848484848484,
      "grad_norm": 0.4624770716910183,
      "learning_rate": 6.93871219098876e-06,
      "loss": 0.2414,
      "step": 148
    },
    {
      "epoch": 4.515151515151516,
      "grad_norm": 0.48359586920480535,
      "learning_rate": 6.79458888328903e-06,
      "loss": 0.2336,
      "step": 149
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 0.48813375601319753,
      "learning_rate": 6.651203878290139e-06,
      "loss": 0.2345,
      "step": 150
    },
    {
      "epoch": 4.575757575757576,
      "grad_norm": 0.41950624414364907,
      "learning_rate": 6.508590201876317e-06,
      "loss": 0.2406,
      "step": 151
    },
    {
      "epoch": 4.606060606060606,
      "grad_norm": 0.4293523600653929,
      "learning_rate": 6.366780702271589e-06,
      "loss": 0.2327,
      "step": 152
    },
    {
      "epoch": 4.636363636363637,
      "grad_norm": 0.4705142148220202,
      "learning_rate": 6.225808042473857e-06,
      "loss": 0.2444,
      "step": 153
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.4409469777504741,
      "learning_rate": 6.085704692731609e-06,
      "loss": 0.2397,
      "step": 154
    },
    {
      "epoch": 4.696969696969697,
      "grad_norm": 0.4612034712726125,
      "learning_rate": 5.946502923065054e-06,
      "loss": 0.2437,
      "step": 155
    },
    {
      "epoch": 4.7272727272727275,
      "grad_norm": 0.4276319004208902,
      "learning_rate": 5.8082347958333625e-06,
      "loss": 0.2404,
      "step": 156
    },
    {
      "epoch": 4.757575757575758,
      "grad_norm": 0.4035332466902113,
      "learning_rate": 5.670932158349732e-06,
      "loss": 0.2312,
      "step": 157
    },
    {
      "epoch": 4.787878787878788,
      "grad_norm": 0.4551405541985991,
      "learning_rate": 5.534626635546e-06,
      "loss": 0.231,
      "step": 158
    },
    {
      "epoch": 4.818181818181818,
      "grad_norm": 0.3978436746485399,
      "learning_rate": 5.399349622688479e-06,
      "loss": 0.2224,
      "step": 159
    },
    {
      "epoch": 4.848484848484849,
      "grad_norm": 0.4428461598197092,
      "learning_rate": 5.2651322781466606e-06,
      "loss": 0.2407,
      "step": 160
    },
    {
      "epoch": 4.878787878787879,
      "grad_norm": 0.42484770040407305,
      "learning_rate": 5.132005516216512e-06,
      "loss": 0.2271,
      "step": 161
    },
    {
      "epoch": 4.909090909090909,
      "grad_norm": 0.4221865203466915,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.23,
      "step": 162
    },
    {
      "epoch": 4.9393939393939394,
      "grad_norm": 0.405149774500086,
      "learning_rate": 4.869146134342426e-06,
      "loss": 0.2186,
      "step": 163
    },
    {
      "epoch": 4.96969696969697,
      "grad_norm": 0.40843472424969607,
      "learning_rate": 4.739474058829288e-06,
      "loss": 0.2252,
      "step": 164
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.4142427866189157,
      "learning_rate": 4.611013640844245e-06,
      "loss": 0.228,
      "step": 165
    },
    {
      "epoch": 5.03030303030303,
      "grad_norm": 0.5742115594923108,
      "learning_rate": 4.483794468689728e-06,
      "loss": 0.1915,
      "step": 166
    },
    {
      "epoch": 5.0606060606060606,
      "grad_norm": 0.5166207814269285,
      "learning_rate": 4.357845844771881e-06,
      "loss": 0.1897,
      "step": 167
    },
    {
      "epoch": 5.090909090909091,
      "grad_norm": 0.4165284747216062,
      "learning_rate": 4.2331967788513295e-06,
      "loss": 0.1721,
      "step": 168
    },
    {
      "epoch": 5.121212121212121,
      "grad_norm": 0.44388278472804926,
      "learning_rate": 4.109875981361363e-06,
      "loss": 0.1687,
      "step": 169
    },
    {
      "epoch": 5.151515151515151,
      "grad_norm": 0.5783207903922752,
      "learning_rate": 3.987911856795047e-06,
      "loss": 0.1955,
      "step": 170
    },
    {
      "epoch": 5.181818181818182,
      "grad_norm": 0.60783392075912,
      "learning_rate": 3.867332497162836e-06,
      "loss": 0.1877,
      "step": 171
    },
    {
      "epoch": 5.212121212121212,
      "grad_norm": 0.4271940085585141,
      "learning_rate": 3.748165675522113e-06,
      "loss": 0.1708,
      "step": 172
    },
    {
      "epoch": 5.242424242424242,
      "grad_norm": 0.40343616710144975,
      "learning_rate": 3.630438839580217e-06,
      "loss": 0.1678,
      "step": 173
    },
    {
      "epoch": 5.2727272727272725,
      "grad_norm": 0.4297988621112416,
      "learning_rate": 3.5141791053724405e-06,
      "loss": 0.1632,
      "step": 174
    },
    {
      "epoch": 5.303030303030303,
      "grad_norm": 0.44367455864660416,
      "learning_rate": 3.399413251016359e-06,
      "loss": 0.1876,
      "step": 175
    },
    {
      "epoch": 5.333333333333333,
      "grad_norm": 0.46142274579507064,
      "learning_rate": 3.2861677105440335e-06,
      "loss": 0.174,
      "step": 176
    },
    {
      "epoch": 5.363636363636363,
      "grad_norm": 0.4162071086041392,
      "learning_rate": 3.174468567813461e-06,
      "loss": 0.17,
      "step": 177
    },
    {
      "epoch": 5.393939393939394,
      "grad_norm": 0.3894461320822898,
      "learning_rate": 3.0643415505006733e-06,
      "loss": 0.1676,
      "step": 178
    },
    {
      "epoch": 5.424242424242424,
      "grad_norm": 0.39903454988143183,
      "learning_rate": 2.9558120241738786e-06,
      "loss": 0.1678,
      "step": 179
    },
    {
      "epoch": 5.454545454545454,
      "grad_norm": 0.43314054163119214,
      "learning_rate": 2.8489049864510053e-06,
      "loss": 0.1676,
      "step": 180
    },
    {
      "epoch": 5.484848484848484,
      "grad_norm": 0.42598643749346615,
      "learning_rate": 2.7436450612420098e-06,
      "loss": 0.1732,
      "step": 181
    },
    {
      "epoch": 5.515151515151516,
      "grad_norm": 0.41292734228892786,
      "learning_rate": 2.640056493077231e-06,
      "loss": 0.1679,
      "step": 182
    },
    {
      "epoch": 5.545454545454545,
      "grad_norm": 0.4220043665147572,
      "learning_rate": 2.5381631415231455e-06,
      "loss": 0.1801,
      "step": 183
    },
    {
      "epoch": 5.575757575757576,
      "grad_norm": 0.3841188506361115,
      "learning_rate": 2.4379884756868167e-06,
      "loss": 0.1559,
      "step": 184
    },
    {
      "epoch": 5.606060606060606,
      "grad_norm": 0.3929211402827616,
      "learning_rate": 2.339555568810221e-06,
      "loss": 0.1643,
      "step": 185
    },
    {
      "epoch": 5.636363636363637,
      "grad_norm": 0.38247231160015316,
      "learning_rate": 2.2428870929558012e-06,
      "loss": 0.1684,
      "step": 186
    },
    {
      "epoch": 5.666666666666667,
      "grad_norm": 0.3783424280815931,
      "learning_rate": 2.1480053137844115e-06,
      "loss": 0.1767,
      "step": 187
    },
    {
      "epoch": 5.696969696969697,
      "grad_norm": 0.3963870273429583,
      "learning_rate": 2.054932085426856e-06,
      "loss": 0.1663,
      "step": 188
    },
    {
      "epoch": 5.7272727272727275,
      "grad_norm": 0.3752808343226345,
      "learning_rate": 1.963688845450218e-06,
      "loss": 0.167,
      "step": 189
    },
    {
      "epoch": 5.757575757575758,
      "grad_norm": 0.3633282258887765,
      "learning_rate": 1.8742966099201699e-06,
      "loss": 0.1717,
      "step": 190
    },
    {
      "epoch": 5.787878787878788,
      "grad_norm": 0.36221110762082537,
      "learning_rate": 1.7867759685603115e-06,
      "loss": 0.1642,
      "step": 191
    },
    {
      "epoch": 5.818181818181818,
      "grad_norm": 0.3934225315456543,
      "learning_rate": 1.7011470800097496e-06,
      "loss": 0.1627,
      "step": 192
    },
    {
      "epoch": 5.848484848484849,
      "grad_norm": 0.4353556265864982,
      "learning_rate": 1.6174296671799571e-06,
      "loss": 0.1599,
      "step": 193
    },
    {
      "epoch": 5.878787878787879,
      "grad_norm": 0.40032835131099176,
      "learning_rate": 1.5356430127119915e-06,
      "loss": 0.1769,
      "step": 194
    },
    {
      "epoch": 5.909090909090909,
      "grad_norm": 0.38135206747532385,
      "learning_rate": 1.4558059545351144e-06,
      "loss": 0.1739,
      "step": 195
    },
    {
      "epoch": 5.9393939393939394,
      "grad_norm": 0.36420081321251285,
      "learning_rate": 1.3779368815278648e-06,
      "loss": 0.186,
      "step": 196
    },
    {
      "epoch": 5.96969696969697,
      "grad_norm": 0.3808807238458524,
      "learning_rate": 1.302053729282533e-06,
      "loss": 0.1635,
      "step": 197
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.3664803305584857,
      "learning_rate": 1.2281739759740575e-06,
      "loss": 0.1567,
      "step": 198
    },
    {
      "epoch": 6.03030303030303,
      "grad_norm": 0.49626825440614986,
      "learning_rate": 1.156314638334277e-06,
      "loss": 0.1431,
      "step": 199
    },
    {
      "epoch": 6.0606060606060606,
      "grad_norm": 0.4239002010356736,
      "learning_rate": 1.086492267732462e-06,
      "loss": 0.1336,
      "step": 200
    },
    {
      "epoch": 6.090909090909091,
      "grad_norm": 0.4416525859119522,
      "learning_rate": 1.01872294636304e-06,
      "loss": 0.1379,
      "step": 201
    },
    {
      "epoch": 6.121212121212121,
      "grad_norm": 0.3711931095261706,
      "learning_rate": 9.530222835413739e-07,
      "loss": 0.1457,
      "step": 202
    },
    {
      "epoch": 6.151515151515151,
      "grad_norm": 0.3447887072353519,
      "learning_rate": 8.894054121084839e-07,
      "loss": 0.1483,
      "step": 203
    },
    {
      "epoch": 6.181818181818182,
      "grad_norm": 0.3289966890609169,
      "learning_rate": 8.278869849454718e-07,
      "loss": 0.1364,
      "step": 204
    },
    {
      "epoch": 6.212121212121212,
      "grad_norm": 0.36254988244086095,
      "learning_rate": 7.684811715985429e-07,
      "loss": 0.1496,
      "step": 205
    },
    {
      "epoch": 6.242424242424242,
      "grad_norm": 0.3632904904272957,
      "learning_rate": 7.1120165501533e-07,
      "loss": 0.1436,
      "step": 206
    },
    {
      "epoch": 6.2727272727272725,
      "grad_norm": 0.37269285787256123,
      "learning_rate": 6.560616283932897e-07,
      "loss": 0.1387,
      "step": 207
    },
    {
      "epoch": 6.303030303030303,
      "grad_norm": 0.46235743024257353,
      "learning_rate": 6.030737921409169e-07,
      "loss": 0.1595,
      "step": 208
    },
    {
      "epoch": 6.333333333333333,
      "grad_norm": 0.3777015762464913,
      "learning_rate": 5.522503509524591e-07,
      "loss": 0.1383,
      "step": 209
    },
    {
      "epoch": 6.363636363636363,
      "grad_norm": 0.4213699560133867,
      "learning_rate": 5.036030109968082e-07,
      "loss": 0.1438,
      "step": 210
    },
    {
      "epoch": 6.393939393939394,
      "grad_norm": 0.36921531650339,
      "learning_rate": 4.5714297722121105e-07,
      "loss": 0.149,
      "step": 211
    },
    {
      "epoch": 6.424242424242424,
      "grad_norm": 0.4071554197055839,
      "learning_rate": 4.128809507704445e-07,
      "loss": 0.154,
      "step": 212
    },
    {
      "epoch": 6.454545454545454,
      "grad_norm": 0.3652257992581415,
      "learning_rate": 3.708271265220087e-07,
      "loss": 0.1491,
      "step": 213
    },
    {
      "epoch": 6.484848484848484,
      "grad_norm": 0.33809394138141063,
      "learning_rate": 3.309911907379393e-07,
      "loss": 0.1402,
      "step": 214
    },
    {
      "epoch": 6.515151515151516,
      "grad_norm": 0.3470365859188612,
      "learning_rate": 2.9338231883378365e-07,
      "loss": 0.1506,
      "step": 215
    },
    {
      "epoch": 6.545454545454545,
      "grad_norm": 0.34658124746044666,
      "learning_rate": 2.5800917326521013e-07,
      "loss": 0.1476,
      "step": 216
    },
    {
      "epoch": 6.575757575757576,
      "grad_norm": 0.3316918566486324,
      "learning_rate": 2.248799015327907e-07,
      "loss": 0.1376,
      "step": 217
    },
    {
      "epoch": 6.606060606060606,
      "grad_norm": 0.35672439409490786,
      "learning_rate": 1.9400213430538773e-07,
      "loss": 0.1429,
      "step": 218
    },
    {
      "epoch": 6.636363636363637,
      "grad_norm": 0.3631603267271598,
      "learning_rate": 1.6538298366257975e-07,
      "loss": 0.1533,
      "step": 219
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.3567061213195218,
      "learning_rate": 1.3902904145653094e-07,
      "loss": 0.1351,
      "step": 220
    },
    {
      "epoch": 6.696969696969697,
      "grad_norm": 0.34827383931170836,
      "learning_rate": 1.1494637779369766e-07,
      "loss": 0.1368,
      "step": 221
    },
    {
      "epoch": 6.7272727272727275,
      "grad_norm": 0.34287094175806077,
      "learning_rate": 9.314053963669245e-08,
      "loss": 0.1576,
      "step": 222
    },
    {
      "epoch": 6.757575757575758,
      "grad_norm": 0.3482358604532515,
      "learning_rate": 7.361654952665608e-08,
      "loss": 0.1517,
      "step": 223
    },
    {
      "epoch": 6.787878787878788,
      "grad_norm": 0.3356226829450221,
      "learning_rate": 5.637890442641403e-08,
      "loss": 0.1391,
      "step": 224
    },
    {
      "epoch": 6.818181818181818,
      "grad_norm": 0.3456342111460825,
      "learning_rate": 4.143157468468717e-08,
      "loss": 0.1446,
      "step": 225
    },
    {
      "epoch": 6.848484848484849,
      "grad_norm": 0.3836659500224973,
      "learning_rate": 2.8778003121607834e-08,
      "loss": 0.1412,
      "step": 226
    },
    {
      "epoch": 6.878787878787879,
      "grad_norm": 0.3561494017126276,
      "learning_rate": 1.8421104235727406e-08,
      "loss": 0.1443,
      "step": 227
    },
    {
      "epoch": 6.909090909090909,
      "grad_norm": 0.3480245489945403,
      "learning_rate": 1.0363263532724433e-08,
      "loss": 0.1637,
      "step": 228
    },
    {
      "epoch": 6.9393939393939394,
      "grad_norm": 0.3311151848847929,
      "learning_rate": 4.606336975948589e-09,
      "loss": 0.1516,
      "step": 229
    },
    {
      "epoch": 6.96969696969697,
      "grad_norm": 0.33221424604486316,
      "learning_rate": 1.1516505589381777e-09,
      "loss": 0.1555,
      "step": 230
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.33957128528435504,
      "learning_rate": 0.0,
      "loss": 0.1402,
      "step": 231
    },
    {
      "epoch": 7.0,
      "step": 231,
      "total_flos": 3.450799819629527e+17,
      "train_loss": 0.3807417738747287,
      "train_runtime": 6577.9886,
      "train_samples_per_second": 3.363,
      "train_steps_per_second": 0.035
    }
  ],
  "logging_steps": 1,
  "max_steps": 231,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.450799819629527e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}