{
  "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": 6.563540926047164,
      "learning_rate": 8.333333333333333e-07,
      "loss": 1.0428,
      "step": 1
    },
    {
      "epoch": 0.06060606060606061,
      "grad_norm": 6.580449611407564,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.0483,
      "step": 2
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 6.210885833052012,
      "learning_rate": 2.5e-06,
      "loss": 1.0076,
      "step": 3
    },
    {
      "epoch": 0.12121212121212122,
      "grad_norm": 6.036397259814275,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.0567,
      "step": 4
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 4.603774097681534,
      "learning_rate": 4.166666666666667e-06,
      "loss": 1.0084,
      "step": 5
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 2.737755371478083,
      "learning_rate": 5e-06,
      "loss": 0.965,
      "step": 6
    },
    {
      "epoch": 0.21212121212121213,
      "grad_norm": 2.5619445574428448,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.9429,
      "step": 7
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 3.8525147643115134,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.9282,
      "step": 8
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 4.209095398851915,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.9393,
      "step": 9
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 3.800581953224376,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.9139,
      "step": 10
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 3.476621176143793,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.8747,
      "step": 11
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 2.9503268969909007,
      "learning_rate": 1e-05,
      "loss": 0.8748,
      "step": 12
    },
    {
      "epoch": 0.3939393939393939,
      "grad_norm": 2.153814926870141,
      "learning_rate": 1.0833333333333334e-05,
      "loss": 0.8607,
      "step": 13
    },
    {
      "epoch": 0.42424242424242425,
      "grad_norm": 1.4726286662048231,
      "learning_rate": 1.1666666666666668e-05,
      "loss": 0.8153,
      "step": 14
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 1.5827019390839245,
      "learning_rate": 1.25e-05,
      "loss": 0.8089,
      "step": 15
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 1.370989555312634,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.7918,
      "step": 16
    },
    {
      "epoch": 0.5151515151515151,
      "grad_norm": 1.0938787558192071,
      "learning_rate": 1.416666666666667e-05,
      "loss": 0.7809,
      "step": 17
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 1.1189356582259942,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.7785,
      "step": 18
    },
    {
      "epoch": 0.5757575757575758,
      "grad_norm": 0.9962255721627962,
      "learning_rate": 1.5833333333333333e-05,
      "loss": 0.7742,
      "step": 19
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 1.044035841940071,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.7509,
      "step": 20
    },
    {
      "epoch": 0.6363636363636364,
      "grad_norm": 0.9721149401473398,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 0.7512,
      "step": 21
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.8933857358288726,
      "learning_rate": 1.8333333333333333e-05,
      "loss": 0.7381,
      "step": 22
    },
    {
      "epoch": 0.696969696969697,
      "grad_norm": 0.9947922049252935,
      "learning_rate": 1.916666666666667e-05,
      "loss": 0.7335,
      "step": 23
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.8824730582284096,
      "learning_rate": 2e-05,
      "loss": 0.7342,
      "step": 24
    },
    {
      "epoch": 0.7575757575757576,
      "grad_norm": 0.9463986654416301,
      "learning_rate": 1.999884834944106e-05,
      "loss": 0.7162,
      "step": 25
    },
    {
      "epoch": 0.7878787878787878,
      "grad_norm": 0.8632572070636885,
      "learning_rate": 1.9995393663024054e-05,
      "loss": 0.7068,
      "step": 26
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 0.706944869585806,
      "learning_rate": 1.9989636736467278e-05,
      "loss": 0.7179,
      "step": 27
    },
    {
      "epoch": 0.8484848484848485,
      "grad_norm": 0.71187796312837,
      "learning_rate": 1.9981578895764272e-05,
      "loss": 0.6952,
      "step": 28
    },
    {
      "epoch": 0.8787878787878788,
      "grad_norm": 0.7235082612231346,
      "learning_rate": 1.9971221996878395e-05,
      "loss": 0.7062,
      "step": 29
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.687345657184883,
      "learning_rate": 1.9958568425315316e-05,
      "loss": 0.6895,
      "step": 30
    },
    {
      "epoch": 0.9393939393939394,
      "grad_norm": 0.637332917341387,
      "learning_rate": 1.9943621095573588e-05,
      "loss": 0.7066,
      "step": 31
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.8211307421728865,
      "learning_rate": 1.9926383450473344e-05,
      "loss": 0.6939,
      "step": 32
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.6568867965417567,
      "learning_rate": 1.9906859460363307e-05,
      "loss": 0.6849,
      "step": 33
    },
    {
      "epoch": 1.0303030303030303,
      "grad_norm": 0.6316770232176342,
      "learning_rate": 1.9885053622206305e-05,
      "loss": 0.6666,
      "step": 34
    },
    {
      "epoch": 1.0606060606060606,
      "grad_norm": 0.598548781481323,
      "learning_rate": 1.986097095854347e-05,
      "loss": 0.6617,
      "step": 35
    },
    {
      "epoch": 1.0909090909090908,
      "grad_norm": 0.541888329561449,
      "learning_rate": 1.9834617016337424e-05,
      "loss": 0.6661,
      "step": 36
    },
    {
      "epoch": 1.121212121212121,
      "grad_norm": 0.548704296621704,
      "learning_rate": 1.9805997865694616e-05,
      "loss": 0.6541,
      "step": 37
    },
    {
      "epoch": 1.1515151515151516,
      "grad_norm": 0.7232814085495041,
      "learning_rate": 1.9775120098467212e-05,
      "loss": 0.6482,
      "step": 38
    },
    {
      "epoch": 1.1818181818181819,
      "grad_norm": 0.517966369427666,
      "learning_rate": 1.9741990826734793e-05,
      "loss": 0.6538,
      "step": 39
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 0.6716665600989004,
      "learning_rate": 1.970661768116622e-05,
      "loss": 0.6636,
      "step": 40
    },
    {
      "epoch": 1.2424242424242424,
      "grad_norm": 0.6847972723529687,
      "learning_rate": 1.9669008809262064e-05,
      "loss": 0.6434,
      "step": 41
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 0.5804576446224823,
      "learning_rate": 1.9629172873477995e-05,
      "loss": 0.653,
      "step": 42
    },
    {
      "epoch": 1.303030303030303,
      "grad_norm": 0.6500106007023674,
      "learning_rate": 1.9587119049229558e-05,
      "loss": 0.6362,
      "step": 43
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.694521890601108,
      "learning_rate": 1.954285702277879e-05,
      "loss": 0.6295,
      "step": 44
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.6074904864106893,
      "learning_rate": 1.9496396989003195e-05,
      "loss": 0.6233,
      "step": 45
    },
    {
      "epoch": 1.393939393939394,
      "grad_norm": 0.7774754447607893,
      "learning_rate": 1.944774964904754e-05,
      "loss": 0.6275,
      "step": 46
    },
    {
      "epoch": 1.4242424242424243,
      "grad_norm": 0.6113642022042228,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 0.6271,
      "step": 47
    },
    {
      "epoch": 1.4545454545454546,
      "grad_norm": 0.6760030831909414,
      "learning_rate": 1.9343938371606714e-05,
      "loss": 0.6125,
      "step": 48
    },
    {
      "epoch": 1.4848484848484849,
      "grad_norm": 0.6651014861881179,
      "learning_rate": 1.9288798344984673e-05,
      "loss": 0.6398,
      "step": 49
    },
    {
      "epoch": 1.5151515151515151,
      "grad_norm": 0.8828627870381741,
      "learning_rate": 1.9231518828401458e-05,
      "loss": 0.6255,
      "step": 50
    },
    {
      "epoch": 1.5454545454545454,
      "grad_norm": 0.759276830653897,
      "learning_rate": 1.917211301505453e-05,
      "loss": 0.6259,
      "step": 51
    },
    {
      "epoch": 1.5757575757575757,
      "grad_norm": 0.5517216460140265,
      "learning_rate": 1.911059458789152e-05,
      "loss": 0.6235,
      "step": 52
    },
    {
      "epoch": 1.606060606060606,
      "grad_norm": 0.7568131950239245,
      "learning_rate": 1.9046977716458627e-05,
      "loss": 0.6017,
      "step": 53
    },
    {
      "epoch": 1.6363636363636362,
      "grad_norm": 0.6210935288846776,
      "learning_rate": 1.8981277053636963e-05,
      "loss": 0.6113,
      "step": 54
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.529751177762721,
      "learning_rate": 1.891350773226754e-05,
      "loss": 0.6267,
      "step": 55
    },
    {
      "epoch": 1.696969696969697,
      "grad_norm": 0.6804806242920128,
      "learning_rate": 1.8843685361665724e-05,
      "loss": 0.6301,
      "step": 56
    },
    {
      "epoch": 1.7272727272727273,
      "grad_norm": 0.5234947176657136,
      "learning_rate": 1.8771826024025944e-05,
      "loss": 0.6296,
      "step": 57
    },
    {
      "epoch": 1.7575757575757576,
      "grad_norm": 0.630989209194213,
      "learning_rate": 1.8697946270717468e-05,
      "loss": 0.6242,
      "step": 58
    },
    {
      "epoch": 1.7878787878787878,
      "grad_norm": 0.5781907314015904,
      "learning_rate": 1.8622063118472135e-05,
      "loss": 0.6238,
      "step": 59
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.5200010563062794,
      "learning_rate": 1.8544194045464888e-05,
      "loss": 0.6159,
      "step": 60
    },
    {
      "epoch": 1.8484848484848486,
      "grad_norm": 0.5319459955427939,
      "learning_rate": 1.8464356987288012e-05,
      "loss": 0.608,
      "step": 61
    },
    {
      "epoch": 1.878787878787879,
      "grad_norm": 0.5163593309425878,
      "learning_rate": 1.8382570332820045e-05,
      "loss": 0.616,
      "step": 62
    },
    {
      "epoch": 1.9090909090909092,
      "grad_norm": 0.5585332117774429,
      "learning_rate": 1.8298852919990254e-05,
      "loss": 0.6191,
      "step": 63
    },
    {
      "epoch": 1.9393939393939394,
      "grad_norm": 0.5974666940232033,
      "learning_rate": 1.821322403143969e-05,
      "loss": 0.6365,
      "step": 64
    },
    {
      "epoch": 1.9696969696969697,
      "grad_norm": 0.7254714115326361,
      "learning_rate": 1.812570339007983e-05,
      "loss": 0.6143,
      "step": 65
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6048682162832721,
      "learning_rate": 1.8036311154549783e-05,
      "loss": 0.6266,
      "step": 66
    },
    {
      "epoch": 2.0303030303030303,
      "grad_norm": 0.5811332386679446,
      "learning_rate": 1.7945067914573147e-05,
      "loss": 0.5654,
      "step": 67
    },
    {
      "epoch": 2.0606060606060606,
      "grad_norm": 0.7078463221059967,
      "learning_rate": 1.7851994686215592e-05,
      "loss": 0.5782,
      "step": 68
    },
    {
      "epoch": 2.090909090909091,
      "grad_norm": 0.6467886516514096,
      "learning_rate": 1.77571129070442e-05,
      "loss": 0.5641,
      "step": 69
    },
    {
      "epoch": 2.121212121212121,
      "grad_norm": 0.8327454008647746,
      "learning_rate": 1.766044443118978e-05,
      "loss": 0.5633,
      "step": 70
    },
    {
      "epoch": 2.1515151515151514,
      "grad_norm": 0.914036765125269,
      "learning_rate": 1.7562011524313187e-05,
      "loss": 0.5498,
      "step": 71
    },
    {
      "epoch": 2.1818181818181817,
      "grad_norm": 0.7579034291454978,
      "learning_rate": 1.7461836858476858e-05,
      "loss": 0.5583,
      "step": 72
    },
    {
      "epoch": 2.212121212121212,
      "grad_norm": 0.5951617380041966,
      "learning_rate": 1.7359943506922775e-05,
      "loss": 0.558,
      "step": 73
    },
    {
      "epoch": 2.242424242424242,
      "grad_norm": 0.6656364633288391,
      "learning_rate": 1.725635493875799e-05,
      "loss": 0.5502,
      "step": 74
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.5498251841183976,
      "learning_rate": 1.7151095013548996e-05,
      "loss": 0.5432,
      "step": 75
    },
    {
      "epoch": 2.303030303030303,
      "grad_norm": 0.5181474131631552,
      "learning_rate": 1.7044187975826126e-05,
      "loss": 0.5341,
      "step": 76
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.6056649893501447,
      "learning_rate": 1.693565844949933e-05,
      "loss": 0.553,
      "step": 77
    },
    {
      "epoch": 2.3636363636363638,
      "grad_norm": 0.5505110256865079,
      "learning_rate": 1.6825531432186545e-05,
      "loss": 0.5423,
      "step": 78
    },
    {
      "epoch": 2.393939393939394,
      "grad_norm": 0.5420722140172576,
      "learning_rate": 1.671383228945597e-05,
      "loss": 0.5507,
      "step": 79
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 0.6144593157841388,
      "learning_rate": 1.6600586748983642e-05,
      "loss": 0.5455,
      "step": 80
    },
    {
      "epoch": 2.4545454545454546,
      "grad_norm": 0.6164703057994342,
      "learning_rate": 1.648582089462756e-05,
      "loss": 0.5592,
      "step": 81
    },
    {
      "epoch": 2.484848484848485,
      "grad_norm": 0.5526863546676946,
      "learning_rate": 1.6369561160419783e-05,
      "loss": 0.5533,
      "step": 82
    },
    {
      "epoch": 2.515151515151515,
      "grad_norm": 0.6382238950572803,
      "learning_rate": 1.625183432447789e-05,
      "loss": 0.5527,
      "step": 83
    },
    {
      "epoch": 2.5454545454545454,
      "grad_norm": 0.4457817952252059,
      "learning_rate": 1.6132667502837164e-05,
      "loss": 0.5506,
      "step": 84
    },
    {
      "epoch": 2.5757575757575757,
      "grad_norm": 0.5262250955813433,
      "learning_rate": 1.6012088143204953e-05,
      "loss": 0.5718,
      "step": 85
    },
    {
      "epoch": 2.606060606060606,
      "grad_norm": 0.4829041397558836,
      "learning_rate": 1.589012401863864e-05,
      "loss": 0.5567,
      "step": 86
    },
    {
      "epoch": 2.6363636363636362,
      "grad_norm": 0.43069510677405,
      "learning_rate": 1.5766803221148676e-05,
      "loss": 0.5429,
      "step": 87
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.5992851670087732,
      "learning_rate": 1.5642154155228124e-05,
      "loss": 0.55,
      "step": 88
    },
    {
      "epoch": 2.6969696969696972,
      "grad_norm": 0.48987744304682884,
      "learning_rate": 1.5516205531310272e-05,
      "loss": 0.5369,
      "step": 89
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.49423593619418865,
      "learning_rate": 1.538898635915576e-05,
      "loss": 0.5514,
      "step": 90
    },
    {
      "epoch": 2.757575757575758,
      "grad_norm": 0.38480659724740207,
      "learning_rate": 1.526052594117071e-05,
      "loss": 0.5428,
      "step": 91
    },
    {
      "epoch": 2.787878787878788,
      "grad_norm": 0.4631999660083842,
      "learning_rate": 1.513085386565758e-05,
      "loss": 0.5442,
      "step": 92
    },
    {
      "epoch": 2.8181818181818183,
      "grad_norm": 0.43383032995287757,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.5629,
      "step": 93
    },
    {
      "epoch": 2.8484848484848486,
      "grad_norm": 0.4930078476777237,
      "learning_rate": 1.4867994483783485e-05,
      "loss": 0.5399,
      "step": 94
    },
    {
      "epoch": 2.878787878787879,
      "grad_norm": 0.38813626716680266,
      "learning_rate": 1.4734867721853341e-05,
      "loss": 0.5388,
      "step": 95
    },
    {
      "epoch": 2.909090909090909,
      "grad_norm": 0.4163833019799601,
      "learning_rate": 1.4600650377311523e-05,
      "loss": 0.5412,
      "step": 96
    },
    {
      "epoch": 2.9393939393939394,
      "grad_norm": 0.46066144305840756,
      "learning_rate": 1.4465373364454001e-05,
      "loss": 0.5496,
      "step": 97
    },
    {
      "epoch": 2.9696969696969697,
      "grad_norm": 0.3898310359926936,
      "learning_rate": 1.4329067841650274e-05,
      "loss": 0.5333,
      "step": 98
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.5066752775101956,
      "learning_rate": 1.4191765204166643e-05,
      "loss": 0.5419,
      "step": 99
    },
    {
      "epoch": 3.0303030303030303,
      "grad_norm": 0.5474043513124414,
      "learning_rate": 1.4053497076934948e-05,
      "loss": 0.497,
      "step": 100
    },
    {
      "epoch": 3.0606060606060606,
      "grad_norm": 0.49901750517264826,
      "learning_rate": 1.3914295307268396e-05,
      "loss": 0.4843,
      "step": 101
    },
    {
      "epoch": 3.090909090909091,
      "grad_norm": 0.5068420885440028,
      "learning_rate": 1.3774191957526144e-05,
      "loss": 0.4917,
      "step": 102
    },
    {
      "epoch": 3.121212121212121,
      "grad_norm": 0.5572281263009282,
      "learning_rate": 1.3633219297728415e-05,
      "loss": 0.5018,
      "step": 103
    },
    {
      "epoch": 3.1515151515151514,
      "grad_norm": 0.43663824894920084,
      "learning_rate": 1.3491409798123687e-05,
      "loss": 0.4821,
      "step": 104
    },
    {
      "epoch": 3.1818181818181817,
      "grad_norm": 0.5393537880378538,
      "learning_rate": 1.3348796121709862e-05,
      "loss": 0.4955,
      "step": 105
    },
    {
      "epoch": 3.212121212121212,
      "grad_norm": 0.5387099450604667,
      "learning_rate": 1.3205411116710973e-05,
      "loss": 0.4774,
      "step": 106
    },
    {
      "epoch": 3.242424242424242,
      "grad_norm": 0.5054134734775204,
      "learning_rate": 1.3061287809011243e-05,
      "loss": 0.5004,
      "step": 107
    },
    {
      "epoch": 3.2727272727272725,
      "grad_norm": 0.5514702124191662,
      "learning_rate": 1.291645939454825e-05,
      "loss": 0.4782,
      "step": 108
    },
    {
      "epoch": 3.303030303030303,
      "grad_norm": 0.5001894829891724,
      "learning_rate": 1.277095923166689e-05,
      "loss": 0.4872,
      "step": 109
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.5400539892428468,
      "learning_rate": 1.2624820833435939e-05,
      "loss": 0.4892,
      "step": 110
    },
    {
      "epoch": 3.3636363636363638,
      "grad_norm": 0.5145544328425247,
      "learning_rate": 1.2478077859929e-05,
      "loss": 0.4859,
      "step": 111
    },
    {
      "epoch": 3.393939393939394,
      "grad_norm": 0.4110011582979585,
      "learning_rate": 1.2330764110471567e-05,
      "loss": 0.4866,
      "step": 112
    },
    {
      "epoch": 3.4242424242424243,
      "grad_norm": 0.5040490080891203,
      "learning_rate": 1.2182913515856016e-05,
      "loss": 0.4801,
      "step": 113
    },
    {
      "epoch": 3.4545454545454546,
      "grad_norm": 0.460311106110748,
      "learning_rate": 1.2034560130526341e-05,
      "loss": 0.4867,
      "step": 114
    },
    {
      "epoch": 3.484848484848485,
      "grad_norm": 0.4950344381182238,
      "learning_rate": 1.1885738124734359e-05,
      "loss": 0.4721,
      "step": 115
    },
    {
      "epoch": 3.515151515151515,
      "grad_norm": 0.4636731269112049,
      "learning_rate": 1.1736481776669307e-05,
      "loss": 0.4864,
      "step": 116
    },
    {
      "epoch": 3.5454545454545454,
      "grad_norm": 0.39781010318904547,
      "learning_rate": 1.1586825464562515e-05,
      "loss": 0.4764,
      "step": 117
    },
    {
      "epoch": 3.5757575757575757,
      "grad_norm": 0.47344015190765665,
      "learning_rate": 1.1436803658769082e-05,
      "loss": 0.4949,
      "step": 118
    },
    {
      "epoch": 3.606060606060606,
      "grad_norm": 0.45748213076954986,
      "learning_rate": 1.1286450913828313e-05,
      "loss": 0.4979,
      "step": 119
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.4543919473808373,
      "learning_rate": 1.113580186050475e-05,
      "loss": 0.4752,
      "step": 120
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.41379006149200803,
      "learning_rate": 1.0984891197811686e-05,
      "loss": 0.4753,
      "step": 121
    },
    {
      "epoch": 3.6969696969696972,
      "grad_norm": 0.43943952235155676,
      "learning_rate": 1.0833753685018935e-05,
      "loss": 0.4758,
      "step": 122
    },
    {
      "epoch": 3.7272727272727275,
      "grad_norm": 0.3985989707644989,
      "learning_rate": 1.0682424133646712e-05,
      "loss": 0.4852,
      "step": 123
    },
    {
      "epoch": 3.757575757575758,
      "grad_norm": 0.37866491127879987,
      "learning_rate": 1.0530937399447496e-05,
      "loss": 0.4738,
      "step": 124
    },
    {
      "epoch": 3.787878787878788,
      "grad_norm": 0.391346474863689,
      "learning_rate": 1.0379328374377715e-05,
      "loss": 0.4866,
      "step": 125
    },
    {
      "epoch": 3.8181818181818183,
      "grad_norm": 0.3625474315447914,
      "learning_rate": 1.0227631978561057e-05,
      "loss": 0.4787,
      "step": 126
    },
    {
      "epoch": 3.8484848484848486,
      "grad_norm": 0.3845329743925931,
      "learning_rate": 1.0075883152245334e-05,
      "loss": 0.4798,
      "step": 127
    },
    {
      "epoch": 3.878787878787879,
      "grad_norm": 0.383433574379574,
      "learning_rate": 9.92411684775467e-06,
      "loss": 0.4787,
      "step": 128
    },
    {
      "epoch": 3.909090909090909,
      "grad_norm": 0.4115392188921613,
      "learning_rate": 9.772368021438943e-06,
      "loss": 0.4862,
      "step": 129
    },
    {
      "epoch": 3.9393939393939394,
      "grad_norm": 0.389354832046105,
      "learning_rate": 9.620671625622287e-06,
      "loss": 0.4801,
      "step": 130
    },
    {
      "epoch": 3.9696969696969697,
      "grad_norm": 0.4485920448439754,
      "learning_rate": 9.469062600552509e-06,
      "loss": 0.4889,
      "step": 131
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.35020551865579047,
      "learning_rate": 9.317575866353293e-06,
      "loss": 0.4613,
      "step": 132
    },
    {
      "epoch": 4.03030303030303,
      "grad_norm": 0.5607546876543795,
      "learning_rate": 9.166246314981066e-06,
      "loss": 0.4256,
      "step": 133
    },
    {
      "epoch": 4.0606060606060606,
      "grad_norm": 0.4557755084880889,
      "learning_rate": 9.015108802188314e-06,
      "loss": 0.4333,
      "step": 134
    },
    {
      "epoch": 4.090909090909091,
      "grad_norm": 0.4725633025186906,
      "learning_rate": 8.86419813949525e-06,
      "loss": 0.4265,
      "step": 135
    },
    {
      "epoch": 4.121212121212121,
      "grad_norm": 0.573363372087563,
      "learning_rate": 8.71354908617169e-06,
      "loss": 0.4209,
      "step": 136
    },
    {
      "epoch": 4.151515151515151,
      "grad_norm": 0.46044001223143577,
      "learning_rate": 8.56319634123092e-06,
      "loss": 0.4283,
      "step": 137
    },
    {
      "epoch": 4.181818181818182,
      "grad_norm": 0.4467152469171767,
      "learning_rate": 8.413174535437486e-06,
      "loss": 0.4189,
      "step": 138
    },
    {
      "epoch": 4.212121212121212,
      "grad_norm": 0.4867813595382334,
      "learning_rate": 8.263518223330698e-06,
      "loss": 0.4273,
      "step": 139
    },
    {
      "epoch": 4.242424242424242,
      "grad_norm": 0.4177932775745287,
      "learning_rate": 8.114261875265643e-06,
      "loss": 0.428,
      "step": 140
    },
    {
      "epoch": 4.2727272727272725,
      "grad_norm": 0.411671150540731,
      "learning_rate": 7.965439869473664e-06,
      "loss": 0.4232,
      "step": 141
    },
    {
      "epoch": 4.303030303030303,
      "grad_norm": 0.47085390143038414,
      "learning_rate": 7.817086484143987e-06,
      "loss": 0.433,
      "step": 142
    },
    {
      "epoch": 4.333333333333333,
      "grad_norm": 0.3958790763883933,
      "learning_rate": 7.669235889528436e-06,
      "loss": 0.4196,
      "step": 143
    },
    {
      "epoch": 4.363636363636363,
      "grad_norm": 0.41826411109606854,
      "learning_rate": 7.521922140071003e-06,
      "loss": 0.4225,
      "step": 144
    },
    {
      "epoch": 4.393939393939394,
      "grad_norm": 0.40191901886516435,
      "learning_rate": 7.375179166564062e-06,
      "loss": 0.4231,
      "step": 145
    },
    {
      "epoch": 4.424242424242424,
      "grad_norm": 0.4269248963055479,
      "learning_rate": 7.2290407683331154e-06,
      "loss": 0.4206,
      "step": 146
    },
    {
      "epoch": 4.454545454545454,
      "grad_norm": 0.42115014136542406,
      "learning_rate": 7.0835406054517505e-06,
      "loss": 0.4261,
      "step": 147
    },
    {
      "epoch": 4.484848484848484,
      "grad_norm": 0.39326005037866735,
      "learning_rate": 6.93871219098876e-06,
      "loss": 0.4203,
      "step": 148
    },
    {
      "epoch": 4.515151515151516,
      "grad_norm": 0.39609232779775183,
      "learning_rate": 6.79458888328903e-06,
      "loss": 0.4216,
      "step": 149
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 0.3712986313940363,
      "learning_rate": 6.651203878290139e-06,
      "loss": 0.4265,
      "step": 150
    },
    {
      "epoch": 4.575757575757576,
      "grad_norm": 0.392633527257338,
      "learning_rate": 6.508590201876317e-06,
      "loss": 0.4179,
      "step": 151
    },
    {
      "epoch": 4.606060606060606,
      "grad_norm": 0.37590428954350047,
      "learning_rate": 6.366780702271589e-06,
      "loss": 0.4196,
      "step": 152
    },
    {
      "epoch": 4.636363636363637,
      "grad_norm": 0.37092433982418843,
      "learning_rate": 6.225808042473857e-06,
      "loss": 0.4309,
      "step": 153
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.3787184593441365,
      "learning_rate": 6.085704692731609e-06,
      "loss": 0.4318,
      "step": 154
    },
    {
      "epoch": 4.696969696969697,
      "grad_norm": 0.7699843144622641,
      "learning_rate": 5.946502923065054e-06,
      "loss": 0.4474,
      "step": 155
    },
    {
      "epoch": 4.7272727272727275,
      "grad_norm": 0.35041173464519193,
      "learning_rate": 5.8082347958333625e-06,
      "loss": 0.4195,
      "step": 156
    },
    {
      "epoch": 4.757575757575758,
      "grad_norm": 0.37410267572131256,
      "learning_rate": 5.670932158349732e-06,
      "loss": 0.4247,
      "step": 157
    },
    {
      "epoch": 4.787878787878788,
      "grad_norm": 0.34432142617929,
      "learning_rate": 5.534626635546e-06,
      "loss": 0.4328,
      "step": 158
    },
    {
      "epoch": 4.818181818181818,
      "grad_norm": 0.3466828538594678,
      "learning_rate": 5.399349622688479e-06,
      "loss": 0.4291,
      "step": 159
    },
    {
      "epoch": 4.848484848484849,
      "grad_norm": 0.3550750415579987,
      "learning_rate": 5.2651322781466606e-06,
      "loss": 0.4338,
      "step": 160
    },
    {
      "epoch": 4.878787878787879,
      "grad_norm": 0.3444707493311806,
      "learning_rate": 5.132005516216512e-06,
      "loss": 0.4129,
      "step": 161
    },
    {
      "epoch": 4.909090909090909,
      "grad_norm": 0.32768268988127786,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.4238,
      "step": 162
    },
    {
      "epoch": 4.9393939393939394,
      "grad_norm": 0.34960894126683545,
      "learning_rate": 4.869146134342426e-06,
      "loss": 0.4307,
      "step": 163
    },
    {
      "epoch": 4.96969696969697,
      "grad_norm": 0.330195615528033,
      "learning_rate": 4.739474058829288e-06,
      "loss": 0.4196,
      "step": 164
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.3459410278415265,
      "learning_rate": 4.611013640844245e-06,
      "loss": 0.4253,
      "step": 165
    },
    {
      "epoch": 5.03030303030303,
      "grad_norm": 0.48189255968384775,
      "learning_rate": 4.483794468689728e-06,
      "loss": 0.3872,
      "step": 166
    },
    {
      "epoch": 5.0606060606060606,
      "grad_norm": 0.4069317192144402,
      "learning_rate": 4.357845844771881e-06,
      "loss": 0.3822,
      "step": 167
    },
    {
      "epoch": 5.090909090909091,
      "grad_norm": 0.3381651303644385,
      "learning_rate": 4.2331967788513295e-06,
      "loss": 0.3778,
      "step": 168
    },
    {
      "epoch": 5.121212121212121,
      "grad_norm": 0.476940791131885,
      "learning_rate": 4.109875981361363e-06,
      "loss": 0.3863,
      "step": 169
    },
    {
      "epoch": 5.151515151515151,
      "grad_norm": 0.5094452924087921,
      "learning_rate": 3.987911856795047e-06,
      "loss": 0.3793,
      "step": 170
    },
    {
      "epoch": 5.181818181818182,
      "grad_norm": 0.3661104380813191,
      "learning_rate": 3.867332497162836e-06,
      "loss": 0.3765,
      "step": 171
    },
    {
      "epoch": 5.212121212121212,
      "grad_norm": 0.3393190292607263,
      "learning_rate": 3.748165675522113e-06,
      "loss": 0.3809,
      "step": 172
    },
    {
      "epoch": 5.242424242424242,
      "grad_norm": 0.3566253442002397,
      "learning_rate": 3.630438839580217e-06,
      "loss": 0.3858,
      "step": 173
    },
    {
      "epoch": 5.2727272727272725,
      "grad_norm": 0.38325860340550916,
      "learning_rate": 3.5141791053724405e-06,
      "loss": 0.38,
      "step": 174
    },
    {
      "epoch": 5.303030303030303,
      "grad_norm": 0.34337264051552197,
      "learning_rate": 3.399413251016359e-06,
      "loss": 0.3745,
      "step": 175
    },
    {
      "epoch": 5.333333333333333,
      "grad_norm": 0.33113896425624495,
      "learning_rate": 3.2861677105440335e-06,
      "loss": 0.3728,
      "step": 176
    },
    {
      "epoch": 5.363636363636363,
      "grad_norm": 0.35482288831996844,
      "learning_rate": 3.174468567813461e-06,
      "loss": 0.3706,
      "step": 177
    },
    {
      "epoch": 5.393939393939394,
      "grad_norm": 0.5621907966741517,
      "learning_rate": 3.0643415505006733e-06,
      "loss": 0.3814,
      "step": 178
    },
    {
      "epoch": 5.424242424242424,
      "grad_norm": 0.33758999301051723,
      "learning_rate": 2.9558120241738786e-06,
      "loss": 0.3639,
      "step": 179
    },
    {
      "epoch": 5.454545454545454,
      "grad_norm": 0.34667774663189316,
      "learning_rate": 2.8489049864510053e-06,
      "loss": 0.3807,
      "step": 180
    },
    {
      "epoch": 5.484848484848484,
      "grad_norm": 0.3132018666400906,
      "learning_rate": 2.7436450612420098e-06,
      "loss": 0.3798,
      "step": 181
    },
    {
      "epoch": 5.515151515151516,
      "grad_norm": 0.3371484635436534,
      "learning_rate": 2.640056493077231e-06,
      "loss": 0.3881,
      "step": 182
    },
    {
      "epoch": 5.545454545454545,
      "grad_norm": 0.3105879958077527,
      "learning_rate": 2.5381631415231455e-06,
      "loss": 0.3847,
      "step": 183
    },
    {
      "epoch": 5.575757575757576,
      "grad_norm": 0.3343505128008297,
      "learning_rate": 2.4379884756868167e-06,
      "loss": 0.3719,
      "step": 184
    },
    {
      "epoch": 5.606060606060606,
      "grad_norm": 0.33297737666263527,
      "learning_rate": 2.339555568810221e-06,
      "loss": 0.3977,
      "step": 185
    },
    {
      "epoch": 5.636363636363637,
      "grad_norm": 0.31346210814602987,
      "learning_rate": 2.2428870929558012e-06,
      "loss": 0.3835,
      "step": 186
    },
    {
      "epoch": 5.666666666666667,
      "grad_norm": 0.3268734785098672,
      "learning_rate": 2.1480053137844115e-06,
      "loss": 0.3852,
      "step": 187
    },
    {
      "epoch": 5.696969696969697,
      "grad_norm": 0.31547113405684213,
      "learning_rate": 2.054932085426856e-06,
      "loss": 0.3858,
      "step": 188
    },
    {
      "epoch": 5.7272727272727275,
      "grad_norm": 0.30699942166401356,
      "learning_rate": 1.963688845450218e-06,
      "loss": 0.3876,
      "step": 189
    },
    {
      "epoch": 5.757575757575758,
      "grad_norm": 0.3004604356148264,
      "learning_rate": 1.8742966099201699e-06,
      "loss": 0.3715,
      "step": 190
    },
    {
      "epoch": 5.787878787878788,
      "grad_norm": 0.30110377420404033,
      "learning_rate": 1.7867759685603115e-06,
      "loss": 0.3797,
      "step": 191
    },
    {
      "epoch": 5.818181818181818,
      "grad_norm": 0.3020524997749085,
      "learning_rate": 1.7011470800097496e-06,
      "loss": 0.3922,
      "step": 192
    },
    {
      "epoch": 5.848484848484849,
      "grad_norm": 0.30549617753019875,
      "learning_rate": 1.6174296671799571e-06,
      "loss": 0.3737,
      "step": 193
    },
    {
      "epoch": 5.878787878787879,
      "grad_norm": 0.37524600057770147,
      "learning_rate": 1.5356430127119915e-06,
      "loss": 0.3838,
      "step": 194
    },
    {
      "epoch": 5.909090909090909,
      "grad_norm": 0.3042161760921001,
      "learning_rate": 1.4558059545351144e-06,
      "loss": 0.375,
      "step": 195
    },
    {
      "epoch": 5.9393939393939394,
      "grad_norm": 0.30041972160383384,
      "learning_rate": 1.3779368815278648e-06,
      "loss": 0.3841,
      "step": 196
    },
    {
      "epoch": 5.96969696969697,
      "grad_norm": 0.29667721012490295,
      "learning_rate": 1.302053729282533e-06,
      "loss": 0.3955,
      "step": 197
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.29460086955248166,
      "learning_rate": 1.2281739759740575e-06,
      "loss": 0.3862,
      "step": 198
    },
    {
      "epoch": 6.03030303030303,
      "grad_norm": 0.39066495345577096,
      "learning_rate": 1.156314638334277e-06,
      "loss": 0.3559,
      "step": 199
    },
    {
      "epoch": 6.0606060606060606,
      "grad_norm": 0.3297082732903188,
      "learning_rate": 1.086492267732462e-06,
      "loss": 0.3621,
      "step": 200
    },
    {
      "epoch": 6.090909090909091,
      "grad_norm": 0.3218756416031739,
      "learning_rate": 1.01872294636304e-06,
      "loss": 0.3664,
      "step": 201
    },
    {
      "epoch": 6.121212121212121,
      "grad_norm": 0.29542877990945426,
      "learning_rate": 9.530222835413739e-07,
      "loss": 0.3701,
      "step": 202
    },
    {
      "epoch": 6.151515151515151,
      "grad_norm": 0.3226379281831477,
      "learning_rate": 8.894054121084839e-07,
      "loss": 0.3594,
      "step": 203
    },
    {
      "epoch": 6.181818181818182,
      "grad_norm": 0.29876651406377464,
      "learning_rate": 8.278869849454718e-07,
      "loss": 0.3478,
      "step": 204
    },
    {
      "epoch": 6.212121212121212,
      "grad_norm": 0.31595442268640583,
      "learning_rate": 7.684811715985429e-07,
      "loss": 0.3601,
      "step": 205
    },
    {
      "epoch": 6.242424242424242,
      "grad_norm": 0.3327410989346359,
      "learning_rate": 7.1120165501533e-07,
      "loss": 0.357,
      "step": 206
    },
    {
      "epoch": 6.2727272727272725,
      "grad_norm": 0.33021799323368684,
      "learning_rate": 6.560616283932897e-07,
      "loss": 0.3621,
      "step": 207
    },
    {
      "epoch": 6.303030303030303,
      "grad_norm": 0.3369077443848601,
      "learning_rate": 6.030737921409169e-07,
      "loss": 0.3553,
      "step": 208
    },
    {
      "epoch": 6.333333333333333,
      "grad_norm": 0.3402979609634558,
      "learning_rate": 5.522503509524591e-07,
      "loss": 0.3624,
      "step": 209
    },
    {
      "epoch": 6.363636363636363,
      "grad_norm": 0.32636505512382785,
      "learning_rate": 5.036030109968082e-07,
      "loss": 0.3728,
      "step": 210
    },
    {
      "epoch": 6.393939393939394,
      "grad_norm": 0.31061296784676257,
      "learning_rate": 4.5714297722121105e-07,
      "loss": 0.3594,
      "step": 211
    },
    {
      "epoch": 6.424242424242424,
      "grad_norm": 0.29914957991433055,
      "learning_rate": 4.128809507704445e-07,
      "loss": 0.3534,
      "step": 212
    },
    {
      "epoch": 6.454545454545454,
      "grad_norm": 0.3188455054527916,
      "learning_rate": 3.708271265220087e-07,
      "loss": 0.3545,
      "step": 213
    },
    {
      "epoch": 6.484848484848484,
      "grad_norm": 0.28950522233899323,
      "learning_rate": 3.309911907379393e-07,
      "loss": 0.368,
      "step": 214
    },
    {
      "epoch": 6.515151515151516,
      "grad_norm": 0.3069425529194652,
      "learning_rate": 2.9338231883378365e-07,
      "loss": 0.3614,
      "step": 215
    },
    {
      "epoch": 6.545454545454545,
      "grad_norm": 0.3003951738033514,
      "learning_rate": 2.5800917326521013e-07,
      "loss": 0.3576,
      "step": 216
    },
    {
      "epoch": 6.575757575757576,
      "grad_norm": 0.27217818372962993,
      "learning_rate": 2.248799015327907e-07,
      "loss": 0.3552,
      "step": 217
    },
    {
      "epoch": 6.606060606060606,
      "grad_norm": 0.29692820758872923,
      "learning_rate": 1.9400213430538773e-07,
      "loss": 0.3474,
      "step": 218
    },
    {
      "epoch": 6.636363636363637,
      "grad_norm": 0.29355347763625134,
      "learning_rate": 1.6538298366257975e-07,
      "loss": 0.3589,
      "step": 219
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.2905088604483449,
      "learning_rate": 1.3902904145653094e-07,
      "loss": 0.3627,
      "step": 220
    },
    {
      "epoch": 6.696969696969697,
      "grad_norm": 0.28660185842852404,
      "learning_rate": 1.1494637779369766e-07,
      "loss": 0.3625,
      "step": 221
    },
    {
      "epoch": 6.7272727272727275,
      "grad_norm": 0.2813520036309213,
      "learning_rate": 9.314053963669245e-08,
      "loss": 0.3597,
      "step": 222
    },
    {
      "epoch": 6.757575757575758,
      "grad_norm": 0.28485100572086103,
      "learning_rate": 7.361654952665608e-08,
      "loss": 0.3674,
      "step": 223
    },
    {
      "epoch": 6.787878787878788,
      "grad_norm": 0.2919101968416061,
      "learning_rate": 5.637890442641403e-08,
      "loss": 0.3565,
      "step": 224
    },
    {
      "epoch": 6.818181818181818,
      "grad_norm": 0.2881055002821337,
      "learning_rate": 4.143157468468717e-08,
      "loss": 0.3538,
      "step": 225
    },
    {
      "epoch": 6.848484848484849,
      "grad_norm": 0.2801943229513225,
      "learning_rate": 2.8778003121607834e-08,
      "loss": 0.3784,
      "step": 226
    },
    {
      "epoch": 6.878787878787879,
      "grad_norm": 0.2867035584776558,
      "learning_rate": 1.8421104235727406e-08,
      "loss": 0.3586,
      "step": 227
    },
    {
      "epoch": 6.909090909090909,
      "grad_norm": 0.27538587725464836,
      "learning_rate": 1.0363263532724433e-08,
      "loss": 0.3749,
      "step": 228
    },
    {
      "epoch": 6.9393939393939394,
      "grad_norm": 0.2861165909576405,
      "learning_rate": 4.606336975948589e-09,
      "loss": 0.3569,
      "step": 229
    },
    {
      "epoch": 6.96969696969697,
      "grad_norm": 0.2809624315847679,
      "learning_rate": 1.1516505589381777e-09,
      "loss": 0.3542,
      "step": 230
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.28930700143285026,
      "learning_rate": 0.0,
      "loss": 0.3549,
      "step": 231
    },
    {
      "epoch": 7.0,
      "step": 231,
      "total_flos": 6.415454425086689e+17,
      "train_loss": 0.5228371008650049,
      "train_runtime": 8840.4399,
      "train_samples_per_second": 2.502,
      "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": 6.415454425086689e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}