{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9968652037617555,
  "eval_steps": 500,
  "global_step": 212,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004702194357366771,
      "grad_norm": 2.658230781555176,
      "learning_rate": 5.0000000000000004e-08,
      "loss": 0.8839,
      "step": 1
    },
    {
      "epoch": 0.009404388714733543,
      "grad_norm": 2.568281650543213,
      "learning_rate": 1.0000000000000001e-07,
      "loss": 0.8864,
      "step": 2
    },
    {
      "epoch": 0.014106583072100314,
      "grad_norm": 2.142171859741211,
      "learning_rate": 1.5000000000000002e-07,
      "loss": 0.8549,
      "step": 3
    },
    {
      "epoch": 0.018808777429467086,
      "grad_norm": 2.254922866821289,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 0.8955,
      "step": 4
    },
    {
      "epoch": 0.023510971786833857,
      "grad_norm": 2.330303430557251,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 0.8506,
      "step": 5
    },
    {
      "epoch": 0.02821316614420063,
      "grad_norm": 2.304783344268799,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 0.8773,
      "step": 6
    },
    {
      "epoch": 0.032915360501567396,
      "grad_norm": 2.2242815494537354,
      "learning_rate": 3.5000000000000004e-07,
      "loss": 0.7988,
      "step": 7
    },
    {
      "epoch": 0.03761755485893417,
      "grad_norm": 2.267207622528076,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 0.8603,
      "step": 8
    },
    {
      "epoch": 0.04231974921630094,
      "grad_norm": 2.4641835689544678,
      "learning_rate": 4.5000000000000003e-07,
      "loss": 0.8803,
      "step": 9
    },
    {
      "epoch": 0.047021943573667714,
      "grad_norm": 2.6719448566436768,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.8245,
      "step": 10
    },
    {
      "epoch": 0.05172413793103448,
      "grad_norm": 2.502462148666382,
      "learning_rate": 5.5e-07,
      "loss": 0.856,
      "step": 11
    },
    {
      "epoch": 0.05642633228840126,
      "grad_norm": 2.3637027740478516,
      "learning_rate": 6.000000000000001e-07,
      "loss": 0.8867,
      "step": 12
    },
    {
      "epoch": 0.061128526645768025,
      "grad_norm": 2.157748222351074,
      "learning_rate": 6.5e-07,
      "loss": 0.8473,
      "step": 13
    },
    {
      "epoch": 0.06583072100313479,
      "grad_norm": 2.151967763900757,
      "learning_rate": 7.000000000000001e-07,
      "loss": 0.8424,
      "step": 14
    },
    {
      "epoch": 0.07053291536050156,
      "grad_norm": 2.308736562728882,
      "learning_rate": 7.5e-07,
      "loss": 0.8623,
      "step": 15
    },
    {
      "epoch": 0.07523510971786834,
      "grad_norm": 2.247464179992676,
      "learning_rate": 8.000000000000001e-07,
      "loss": 0.8209,
      "step": 16
    },
    {
      "epoch": 0.07993730407523511,
      "grad_norm": 2.0756967067718506,
      "learning_rate": 8.500000000000001e-07,
      "loss": 0.8815,
      "step": 17
    },
    {
      "epoch": 0.08463949843260188,
      "grad_norm": 2.0295917987823486,
      "learning_rate": 9.000000000000001e-07,
      "loss": 0.8166,
      "step": 18
    },
    {
      "epoch": 0.08934169278996865,
      "grad_norm": 2.0582873821258545,
      "learning_rate": 9.500000000000001e-07,
      "loss": 0.8118,
      "step": 19
    },
    {
      "epoch": 0.09404388714733543,
      "grad_norm": 2.036282777786255,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.8054,
      "step": 20
    },
    {
      "epoch": 0.0987460815047022,
      "grad_norm": 1.8246358633041382,
      "learning_rate": 1.0500000000000001e-06,
      "loss": 0.8301,
      "step": 21
    },
    {
      "epoch": 0.10344827586206896,
      "grad_norm": 1.9658900499343872,
      "learning_rate": 1.1e-06,
      "loss": 0.7796,
      "step": 22
    },
    {
      "epoch": 0.10815047021943573,
      "grad_norm": 2.4345409870147705,
      "learning_rate": 1.1500000000000002e-06,
      "loss": 0.7841,
      "step": 23
    },
    {
      "epoch": 0.11285266457680251,
      "grad_norm": 2.0551528930664062,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 0.7503,
      "step": 24
    },
    {
      "epoch": 0.11755485893416928,
      "grad_norm": 1.8872517347335815,
      "learning_rate": 1.25e-06,
      "loss": 0.7527,
      "step": 25
    },
    {
      "epoch": 0.12225705329153605,
      "grad_norm": 1.7877418994903564,
      "learning_rate": 1.3e-06,
      "loss": 0.7346,
      "step": 26
    },
    {
      "epoch": 0.12695924764890282,
      "grad_norm": 1.773200511932373,
      "learning_rate": 1.3500000000000002e-06,
      "loss": 0.7255,
      "step": 27
    },
    {
      "epoch": 0.13166144200626959,
      "grad_norm": 1.3603609800338745,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 0.707,
      "step": 28
    },
    {
      "epoch": 0.13636363636363635,
      "grad_norm": 1.5466724634170532,
      "learning_rate": 1.45e-06,
      "loss": 0.7433,
      "step": 29
    },
    {
      "epoch": 0.14106583072100312,
      "grad_norm": 1.2189550399780273,
      "learning_rate": 1.5e-06,
      "loss": 0.7345,
      "step": 30
    },
    {
      "epoch": 0.14576802507836992,
      "grad_norm": 1.0862981081008911,
      "learning_rate": 1.5500000000000002e-06,
      "loss": 0.7405,
      "step": 31
    },
    {
      "epoch": 0.15047021943573669,
      "grad_norm": 1.024996042251587,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 0.7333,
      "step": 32
    },
    {
      "epoch": 0.15517241379310345,
      "grad_norm": 0.9142002463340759,
      "learning_rate": 1.6500000000000003e-06,
      "loss": 0.6663,
      "step": 33
    },
    {
      "epoch": 0.15987460815047022,
      "grad_norm": 0.9509531259536743,
      "learning_rate": 1.7000000000000002e-06,
      "loss": 0.6502,
      "step": 34
    },
    {
      "epoch": 0.164576802507837,
      "grad_norm": 1.0602822303771973,
      "learning_rate": 1.75e-06,
      "loss": 0.7449,
      "step": 35
    },
    {
      "epoch": 0.16927899686520376,
      "grad_norm": 0.9628905057907104,
      "learning_rate": 1.8000000000000001e-06,
      "loss": 0.6664,
      "step": 36
    },
    {
      "epoch": 0.17398119122257052,
      "grad_norm": 0.8804571628570557,
      "learning_rate": 1.85e-06,
      "loss": 0.693,
      "step": 37
    },
    {
      "epoch": 0.1786833855799373,
      "grad_norm": 0.8594156503677368,
      "learning_rate": 1.9000000000000002e-06,
      "loss": 0.7439,
      "step": 38
    },
    {
      "epoch": 0.1833855799373041,
      "grad_norm": 1.092551827430725,
      "learning_rate": 1.9500000000000004e-06,
      "loss": 0.6681,
      "step": 39
    },
    {
      "epoch": 0.18808777429467086,
      "grad_norm": 0.9434524774551392,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.6981,
      "step": 40
    },
    {
      "epoch": 0.19278996865203762,
      "grad_norm": 0.7898667454719543,
      "learning_rate": 2.05e-06,
      "loss": 0.6602,
      "step": 41
    },
    {
      "epoch": 0.1974921630094044,
      "grad_norm": 0.843589186668396,
      "learning_rate": 2.1000000000000002e-06,
      "loss": 0.7026,
      "step": 42
    },
    {
      "epoch": 0.20219435736677116,
      "grad_norm": 0.7910621166229248,
      "learning_rate": 2.15e-06,
      "loss": 0.6721,
      "step": 43
    },
    {
      "epoch": 0.20689655172413793,
      "grad_norm": 0.7569350004196167,
      "learning_rate": 2.2e-06,
      "loss": 0.6601,
      "step": 44
    },
    {
      "epoch": 0.2115987460815047,
      "grad_norm": 0.9156045913696289,
      "learning_rate": 2.25e-06,
      "loss": 0.6407,
      "step": 45
    },
    {
      "epoch": 0.21630094043887146,
      "grad_norm": 0.7267714738845825,
      "learning_rate": 2.3000000000000004e-06,
      "loss": 0.6459,
      "step": 46
    },
    {
      "epoch": 0.22100313479623823,
      "grad_norm": 0.6335541009902954,
      "learning_rate": 2.35e-06,
      "loss": 0.6244,
      "step": 47
    },
    {
      "epoch": 0.22570532915360503,
      "grad_norm": 0.6519566178321838,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.6137,
      "step": 48
    },
    {
      "epoch": 0.2304075235109718,
      "grad_norm": 0.7315823435783386,
      "learning_rate": 2.4500000000000003e-06,
      "loss": 0.6603,
      "step": 49
    },
    {
      "epoch": 0.23510971786833856,
      "grad_norm": 0.684124767780304,
      "learning_rate": 2.5e-06,
      "loss": 0.6726,
      "step": 50
    },
    {
      "epoch": 0.23981191222570533,
      "grad_norm": 0.6655241250991821,
      "learning_rate": 2.55e-06,
      "loss": 0.6716,
      "step": 51
    },
    {
      "epoch": 0.2445141065830721,
      "grad_norm": 0.6261876821517944,
      "learning_rate": 2.6e-06,
      "loss": 0.6267,
      "step": 52
    },
    {
      "epoch": 0.24921630094043887,
      "grad_norm": 0.8665097951889038,
      "learning_rate": 2.6500000000000005e-06,
      "loss": 0.6291,
      "step": 53
    },
    {
      "epoch": 0.25391849529780564,
      "grad_norm": 0.6287381052970886,
      "learning_rate": 2.7000000000000004e-06,
      "loss": 0.6223,
      "step": 54
    },
    {
      "epoch": 0.25862068965517243,
      "grad_norm": 0.661234438419342,
      "learning_rate": 2.7500000000000004e-06,
      "loss": 0.6655,
      "step": 55
    },
    {
      "epoch": 0.26332288401253917,
      "grad_norm": 0.6680541634559631,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.629,
      "step": 56
    },
    {
      "epoch": 0.26802507836990597,
      "grad_norm": 0.6534361243247986,
      "learning_rate": 2.85e-06,
      "loss": 0.6305,
      "step": 57
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 0.5792294144630432,
      "learning_rate": 2.9e-06,
      "loss": 0.6034,
      "step": 58
    },
    {
      "epoch": 0.2774294670846395,
      "grad_norm": 0.6130394339561462,
      "learning_rate": 2.95e-06,
      "loss": 0.6418,
      "step": 59
    },
    {
      "epoch": 0.28213166144200624,
      "grad_norm": 0.6305087804794312,
      "learning_rate": 3e-06,
      "loss": 0.6687,
      "step": 60
    },
    {
      "epoch": 0.28683385579937304,
      "grad_norm": 0.5708836913108826,
      "learning_rate": 3.05e-06,
      "loss": 0.5886,
      "step": 61
    },
    {
      "epoch": 0.29153605015673983,
      "grad_norm": 0.5345668196678162,
      "learning_rate": 3.1000000000000004e-06,
      "loss": 0.585,
      "step": 62
    },
    {
      "epoch": 0.2962382445141066,
      "grad_norm": 0.6531270742416382,
      "learning_rate": 3.1500000000000003e-06,
      "loss": 0.6274,
      "step": 63
    },
    {
      "epoch": 0.30094043887147337,
      "grad_norm": 0.5387993454933167,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.6393,
      "step": 64
    },
    {
      "epoch": 0.3056426332288401,
      "grad_norm": 0.5837404131889343,
      "learning_rate": 3.2500000000000002e-06,
      "loss": 0.5805,
      "step": 65
    },
    {
      "epoch": 0.3103448275862069,
      "grad_norm": 0.668530285358429,
      "learning_rate": 3.3000000000000006e-06,
      "loss": 0.5982,
      "step": 66
    },
    {
      "epoch": 0.31504702194357365,
      "grad_norm": 0.5391478538513184,
      "learning_rate": 3.3500000000000005e-06,
      "loss": 0.627,
      "step": 67
    },
    {
      "epoch": 0.31974921630094044,
      "grad_norm": 0.5848458409309387,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 0.5944,
      "step": 68
    },
    {
      "epoch": 0.32445141065830724,
      "grad_norm": 0.5370898246765137,
      "learning_rate": 3.45e-06,
      "loss": 0.5803,
      "step": 69
    },
    {
      "epoch": 0.329153605015674,
      "grad_norm": 0.5453668832778931,
      "learning_rate": 3.5e-06,
      "loss": 0.6497,
      "step": 70
    },
    {
      "epoch": 0.3338557993730408,
      "grad_norm": 0.5806694626808167,
      "learning_rate": 3.5500000000000003e-06,
      "loss": 0.6194,
      "step": 71
    },
    {
      "epoch": 0.3385579937304075,
      "grad_norm": 0.4973330497741699,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 0.598,
      "step": 72
    },
    {
      "epoch": 0.3432601880877743,
      "grad_norm": 0.5627994537353516,
      "learning_rate": 3.65e-06,
      "loss": 0.5883,
      "step": 73
    },
    {
      "epoch": 0.34796238244514105,
      "grad_norm": 0.7145661115646362,
      "learning_rate": 3.7e-06,
      "loss": 0.6033,
      "step": 74
    },
    {
      "epoch": 0.35266457680250785,
      "grad_norm": 0.49009037017822266,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.5978,
      "step": 75
    },
    {
      "epoch": 0.3573667711598746,
      "grad_norm": 0.5787577629089355,
      "learning_rate": 3.8000000000000005e-06,
      "loss": 0.5981,
      "step": 76
    },
    {
      "epoch": 0.3620689655172414,
      "grad_norm": 0.5676442384719849,
      "learning_rate": 3.85e-06,
      "loss": 0.6101,
      "step": 77
    },
    {
      "epoch": 0.3667711598746082,
      "grad_norm": 0.49533793330192566,
      "learning_rate": 3.900000000000001e-06,
      "loss": 0.5564,
      "step": 78
    },
    {
      "epoch": 0.3714733542319749,
      "grad_norm": 0.6033879518508911,
      "learning_rate": 3.95e-06,
      "loss": 0.6175,
      "step": 79
    },
    {
      "epoch": 0.3761755485893417,
      "grad_norm": 0.502393901348114,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.5394,
      "step": 80
    },
    {
      "epoch": 0.38087774294670845,
      "grad_norm": 0.49090903997421265,
      "learning_rate": 4.05e-06,
      "loss": 0.5998,
      "step": 81
    },
    {
      "epoch": 0.38557993730407525,
      "grad_norm": 0.8119765520095825,
      "learning_rate": 4.1e-06,
      "loss": 0.5704,
      "step": 82
    },
    {
      "epoch": 0.390282131661442,
      "grad_norm": 0.561651349067688,
      "learning_rate": 4.15e-06,
      "loss": 0.5784,
      "step": 83
    },
    {
      "epoch": 0.3949843260188088,
      "grad_norm": 0.5209910273551941,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 0.5832,
      "step": 84
    },
    {
      "epoch": 0.3996865203761755,
      "grad_norm": 0.5189244151115417,
      "learning_rate": 4.25e-06,
      "loss": 0.6044,
      "step": 85
    },
    {
      "epoch": 0.4043887147335423,
      "grad_norm": 0.5183352828025818,
      "learning_rate": 4.3e-06,
      "loss": 0.6107,
      "step": 86
    },
    {
      "epoch": 0.4090909090909091,
      "grad_norm": 0.505588710308075,
      "learning_rate": 4.350000000000001e-06,
      "loss": 0.5637,
      "step": 87
    },
    {
      "epoch": 0.41379310344827586,
      "grad_norm": 0.5937759280204773,
      "learning_rate": 4.4e-06,
      "loss": 0.59,
      "step": 88
    },
    {
      "epoch": 0.41849529780564265,
      "grad_norm": 0.6988442540168762,
      "learning_rate": 4.450000000000001e-06,
      "loss": 0.5878,
      "step": 89
    },
    {
      "epoch": 0.4231974921630094,
      "grad_norm": 0.5292773842811584,
      "learning_rate": 4.5e-06,
      "loss": 0.5675,
      "step": 90
    },
    {
      "epoch": 0.4278996865203762,
      "grad_norm": 0.5222378373146057,
      "learning_rate": 4.5500000000000005e-06,
      "loss": 0.5941,
      "step": 91
    },
    {
      "epoch": 0.43260188087774293,
      "grad_norm": 0.5161038637161255,
      "learning_rate": 4.600000000000001e-06,
      "loss": 0.5974,
      "step": 92
    },
    {
      "epoch": 0.4373040752351097,
      "grad_norm": 0.4662926495075226,
      "learning_rate": 4.65e-06,
      "loss": 0.5258,
      "step": 93
    },
    {
      "epoch": 0.44200626959247646,
      "grad_norm": 0.5594052672386169,
      "learning_rate": 4.7e-06,
      "loss": 0.5452,
      "step": 94
    },
    {
      "epoch": 0.44670846394984326,
      "grad_norm": 0.5721049904823303,
      "learning_rate": 4.75e-06,
      "loss": 0.5486,
      "step": 95
    },
    {
      "epoch": 0.45141065830721006,
      "grad_norm": 0.47619950771331787,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.5495,
      "step": 96
    },
    {
      "epoch": 0.4561128526645768,
      "grad_norm": 0.5395726561546326,
      "learning_rate": 4.85e-06,
      "loss": 0.5543,
      "step": 97
    },
    {
      "epoch": 0.4608150470219436,
      "grad_norm": 0.5578956007957458,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.5709,
      "step": 98
    },
    {
      "epoch": 0.46551724137931033,
      "grad_norm": 0.565992534160614,
      "learning_rate": 4.95e-06,
      "loss": 0.5527,
      "step": 99
    },
    {
      "epoch": 0.4702194357366771,
      "grad_norm": 0.5071743726730347,
      "learning_rate": 5e-06,
      "loss": 0.5778,
      "step": 100
    },
    {
      "epoch": 0.47492163009404387,
      "grad_norm": 0.504084050655365,
      "learning_rate": 4.9999910183883085e-06,
      "loss": 0.5575,
      "step": 101
    },
    {
      "epoch": 0.47962382445141066,
      "grad_norm": 0.5246759653091431,
      "learning_rate": 4.999964073617768e-06,
      "loss": 0.5666,
      "step": 102
    },
    {
      "epoch": 0.4843260188087774,
      "grad_norm": 0.464106023311615,
      "learning_rate": 4.999919165881985e-06,
      "loss": 0.5541,
      "step": 103
    },
    {
      "epoch": 0.4890282131661442,
      "grad_norm": 0.48115217685699463,
      "learning_rate": 4.999856295503635e-06,
      "loss": 0.5646,
      "step": 104
    },
    {
      "epoch": 0.493730407523511,
      "grad_norm": 0.48950889706611633,
      "learning_rate": 4.9997754629344596e-06,
      "loss": 0.5424,
      "step": 105
    },
    {
      "epoch": 0.49843260188087773,
      "grad_norm": 0.691669762134552,
      "learning_rate": 4.999676668755263e-06,
      "loss": 0.5315,
      "step": 106
    },
    {
      "epoch": 0.5031347962382445,
      "grad_norm": 0.49993181228637695,
      "learning_rate": 4.999559913675912e-06,
      "loss": 0.5638,
      "step": 107
    },
    {
      "epoch": 0.5078369905956113,
      "grad_norm": 0.5278601050376892,
      "learning_rate": 4.999425198535325e-06,
      "loss": 0.5441,
      "step": 108
    },
    {
      "epoch": 0.512539184952978,
      "grad_norm": 0.5409471392631531,
      "learning_rate": 4.999272524301469e-06,
      "loss": 0.5474,
      "step": 109
    },
    {
      "epoch": 0.5172413793103449,
      "grad_norm": 0.5190234780311584,
      "learning_rate": 4.9991018920713505e-06,
      "loss": 0.5638,
      "step": 110
    },
    {
      "epoch": 0.5219435736677116,
      "grad_norm": 0.5064226388931274,
      "learning_rate": 4.9989133030710154e-06,
      "loss": 0.5474,
      "step": 111
    },
    {
      "epoch": 0.5266457680250783,
      "grad_norm": 0.584128201007843,
      "learning_rate": 4.9987067586555275e-06,
      "loss": 0.546,
      "step": 112
    },
    {
      "epoch": 0.5313479623824452,
      "grad_norm": 0.5325257778167725,
      "learning_rate": 4.998482260308969e-06,
      "loss": 0.5643,
      "step": 113
    },
    {
      "epoch": 0.5360501567398119,
      "grad_norm": 0.5349085330963135,
      "learning_rate": 4.998239809644427e-06,
      "loss": 0.5585,
      "step": 114
    },
    {
      "epoch": 0.5407523510971787,
      "grad_norm": 0.5952833890914917,
      "learning_rate": 4.9979794084039755e-06,
      "loss": 0.5711,
      "step": 115
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 0.48140987753868103,
      "learning_rate": 4.997701058458677e-06,
      "loss": 0.5319,
      "step": 116
    },
    {
      "epoch": 0.5501567398119123,
      "grad_norm": 0.6170802712440491,
      "learning_rate": 4.997404761808554e-06,
      "loss": 0.5659,
      "step": 117
    },
    {
      "epoch": 0.554858934169279,
      "grad_norm": 0.506353497505188,
      "learning_rate": 4.9970905205825845e-06,
      "loss": 0.5721,
      "step": 118
    },
    {
      "epoch": 0.5595611285266457,
      "grad_norm": 0.5027800798416138,
      "learning_rate": 4.996758337038683e-06,
      "loss": 0.5618,
      "step": 119
    },
    {
      "epoch": 0.5642633228840125,
      "grad_norm": 0.5272043943405151,
      "learning_rate": 4.996408213563684e-06,
      "loss": 0.5548,
      "step": 120
    },
    {
      "epoch": 0.5689655172413793,
      "grad_norm": 0.4655597507953644,
      "learning_rate": 4.996040152673326e-06,
      "loss": 0.522,
      "step": 121
    },
    {
      "epoch": 0.5736677115987461,
      "grad_norm": 0.49152645468711853,
      "learning_rate": 4.995654157012233e-06,
      "loss": 0.5353,
      "step": 122
    },
    {
      "epoch": 0.5783699059561128,
      "grad_norm": 0.5532233119010925,
      "learning_rate": 4.995250229353895e-06,
      "loss": 0.529,
      "step": 123
    },
    {
      "epoch": 0.5830721003134797,
      "grad_norm": 0.4706387221813202,
      "learning_rate": 4.99482837260065e-06,
      "loss": 0.5438,
      "step": 124
    },
    {
      "epoch": 0.5877742946708464,
      "grad_norm": 0.4653231203556061,
      "learning_rate": 4.99438858978366e-06,
      "loss": 0.5735,
      "step": 125
    },
    {
      "epoch": 0.5924764890282131,
      "grad_norm": 0.7020223736763,
      "learning_rate": 4.993930884062892e-06,
      "loss": 0.5557,
      "step": 126
    },
    {
      "epoch": 0.5971786833855799,
      "grad_norm": 0.6187618970870972,
      "learning_rate": 4.993455258727094e-06,
      "loss": 0.5557,
      "step": 127
    },
    {
      "epoch": 0.6018808777429467,
      "grad_norm": 0.5230109691619873,
      "learning_rate": 4.992961717193773e-06,
      "loss": 0.5575,
      "step": 128
    },
    {
      "epoch": 0.6065830721003135,
      "grad_norm": 1.7068945169448853,
      "learning_rate": 4.9924502630091655e-06,
      "loss": 0.565,
      "step": 129
    },
    {
      "epoch": 0.6112852664576802,
      "grad_norm": 0.5217270255088806,
      "learning_rate": 4.99192089984822e-06,
      "loss": 0.549,
      "step": 130
    },
    {
      "epoch": 0.6159874608150471,
      "grad_norm": 0.5645350217819214,
      "learning_rate": 4.9913736315145614e-06,
      "loss": 0.5568,
      "step": 131
    },
    {
      "epoch": 0.6206896551724138,
      "grad_norm": 0.5037111639976501,
      "learning_rate": 4.990808461940474e-06,
      "loss": 0.5375,
      "step": 132
    },
    {
      "epoch": 0.6253918495297806,
      "grad_norm": 0.5933152437210083,
      "learning_rate": 4.990225395186862e-06,
      "loss": 0.5443,
      "step": 133
    },
    {
      "epoch": 0.6300940438871473,
      "grad_norm": 0.532664954662323,
      "learning_rate": 4.9896244354432314e-06,
      "loss": 0.5411,
      "step": 134
    },
    {
      "epoch": 0.6347962382445141,
      "grad_norm": 1.2658472061157227,
      "learning_rate": 4.98900558702765e-06,
      "loss": 0.5506,
      "step": 135
    },
    {
      "epoch": 0.6394984326018809,
      "grad_norm": 0.4731752872467041,
      "learning_rate": 4.9883688543867225e-06,
      "loss": 0.5609,
      "step": 136
    },
    {
      "epoch": 0.6442006269592476,
      "grad_norm": 0.542219340801239,
      "learning_rate": 4.987714242095558e-06,
      "loss": 0.5306,
      "step": 137
    },
    {
      "epoch": 0.6489028213166145,
      "grad_norm": 0.502611517906189,
      "learning_rate": 4.9870417548577355e-06,
      "loss": 0.5353,
      "step": 138
    },
    {
      "epoch": 0.6536050156739812,
      "grad_norm": 0.4692250192165375,
      "learning_rate": 4.9863513975052696e-06,
      "loss": 0.5434,
      "step": 139
    },
    {
      "epoch": 0.658307210031348,
      "grad_norm": 0.5601999163627625,
      "learning_rate": 4.985643174998578e-06,
      "loss": 0.5654,
      "step": 140
    },
    {
      "epoch": 0.6630094043887147,
      "grad_norm": 0.5011409521102905,
      "learning_rate": 4.984917092426445e-06,
      "loss": 0.5518,
      "step": 141
    },
    {
      "epoch": 0.6677115987460815,
      "grad_norm": 0.5076945424079895,
      "learning_rate": 4.984173155005982e-06,
      "loss": 0.5407,
      "step": 142
    },
    {
      "epoch": 0.6724137931034483,
      "grad_norm": 0.4602852165699005,
      "learning_rate": 4.983411368082597e-06,
      "loss": 0.5465,
      "step": 143
    },
    {
      "epoch": 0.677115987460815,
      "grad_norm": 0.7861032485961914,
      "learning_rate": 4.982631737129948e-06,
      "loss": 0.529,
      "step": 144
    },
    {
      "epoch": 0.6818181818181818,
      "grad_norm": 0.48418697714805603,
      "learning_rate": 4.98183426774991e-06,
      "loss": 0.5684,
      "step": 145
    },
    {
      "epoch": 0.6865203761755486,
      "grad_norm": 0.5176032781600952,
      "learning_rate": 4.981018965672529e-06,
      "loss": 0.5686,
      "step": 146
    },
    {
      "epoch": 0.6912225705329154,
      "grad_norm": 0.5493810772895813,
      "learning_rate": 4.98018583675599e-06,
      "loss": 0.5496,
      "step": 147
    },
    {
      "epoch": 0.6959247648902821,
      "grad_norm": 0.6066202521324158,
      "learning_rate": 4.979334886986562e-06,
      "loss": 0.5503,
      "step": 148
    },
    {
      "epoch": 0.700626959247649,
      "grad_norm": 0.6061354875564575,
      "learning_rate": 4.978466122478567e-06,
      "loss": 0.5648,
      "step": 149
    },
    {
      "epoch": 0.7053291536050157,
      "grad_norm": 0.48728325963020325,
      "learning_rate": 4.97757954947433e-06,
      "loss": 0.5327,
      "step": 150
    },
    {
      "epoch": 0.7100313479623824,
      "grad_norm": 0.5088427662849426,
      "learning_rate": 4.976675174344132e-06,
      "loss": 0.5469,
      "step": 151
    },
    {
      "epoch": 0.7147335423197492,
      "grad_norm": 1.0665165185928345,
      "learning_rate": 4.975753003586172e-06,
      "loss": 0.5268,
      "step": 152
    },
    {
      "epoch": 0.719435736677116,
      "grad_norm": 0.528404712677002,
      "learning_rate": 4.974813043826513e-06,
      "loss": 0.52,
      "step": 153
    },
    {
      "epoch": 0.7241379310344828,
      "grad_norm": 0.5167615413665771,
      "learning_rate": 4.973855301819039e-06,
      "loss": 0.5376,
      "step": 154
    },
    {
      "epoch": 0.7288401253918495,
      "grad_norm": 0.4985736906528473,
      "learning_rate": 4.972879784445402e-06,
      "loss": 0.5192,
      "step": 155
    },
    {
      "epoch": 0.7335423197492164,
      "grad_norm": 0.5043352842330933,
      "learning_rate": 4.971886498714978e-06,
      "loss": 0.5413,
      "step": 156
    },
    {
      "epoch": 0.7382445141065831,
      "grad_norm": 0.5240288376808167,
      "learning_rate": 4.97087545176481e-06,
      "loss": 0.5184,
      "step": 157
    },
    {
      "epoch": 0.7429467084639498,
      "grad_norm": 0.8111078143119812,
      "learning_rate": 4.9698466508595655e-06,
      "loss": 0.5467,
      "step": 158
    },
    {
      "epoch": 0.7476489028213166,
      "grad_norm": 0.48973363637924194,
      "learning_rate": 4.9688001033914756e-06,
      "loss": 0.5341,
      "step": 159
    },
    {
      "epoch": 0.7523510971786834,
      "grad_norm": 0.5965595841407776,
      "learning_rate": 4.967735816880286e-06,
      "loss": 0.5452,
      "step": 160
    },
    {
      "epoch": 0.7570532915360502,
      "grad_norm": 0.5201161503791809,
      "learning_rate": 4.966653798973205e-06,
      "loss": 0.5246,
      "step": 161
    },
    {
      "epoch": 0.7617554858934169,
      "grad_norm": 0.5854963660240173,
      "learning_rate": 4.965554057444842e-06,
      "loss": 0.5178,
      "step": 162
    },
    {
      "epoch": 0.7664576802507836,
      "grad_norm": 0.5000066161155701,
      "learning_rate": 4.964436600197161e-06,
      "loss": 0.5413,
      "step": 163
    },
    {
      "epoch": 0.7711598746081505,
      "grad_norm": 0.49025341868400574,
      "learning_rate": 4.963301435259413e-06,
      "loss": 0.5111,
      "step": 164
    },
    {
      "epoch": 0.7758620689655172,
      "grad_norm": 0.5398334860801697,
      "learning_rate": 4.962148570788088e-06,
      "loss": 0.5313,
      "step": 165
    },
    {
      "epoch": 0.780564263322884,
      "grad_norm": 0.6545300483703613,
      "learning_rate": 4.96097801506685e-06,
      "loss": 0.52,
      "step": 166
    },
    {
      "epoch": 0.7852664576802508,
      "grad_norm": 0.6166910529136658,
      "learning_rate": 4.959789776506482e-06,
      "loss": 0.5288,
      "step": 167
    },
    {
      "epoch": 0.7899686520376176,
      "grad_norm": 0.5018202662467957,
      "learning_rate": 4.958583863644821e-06,
      "loss": 0.5517,
      "step": 168
    },
    {
      "epoch": 0.7946708463949843,
      "grad_norm": 0.5005012154579163,
      "learning_rate": 4.9573602851466985e-06,
      "loss": 0.517,
      "step": 169
    },
    {
      "epoch": 0.799373040752351,
      "grad_norm": 0.51115882396698,
      "learning_rate": 4.9561190498038815e-06,
      "loss": 0.5404,
      "step": 170
    },
    {
      "epoch": 0.8040752351097179,
      "grad_norm": 0.533501148223877,
      "learning_rate": 4.954860166535005e-06,
      "loss": 0.5304,
      "step": 171
    },
    {
      "epoch": 0.8087774294670846,
      "grad_norm": 0.4484115242958069,
      "learning_rate": 4.95358364438551e-06,
      "loss": 0.5355,
      "step": 172
    },
    {
      "epoch": 0.8134796238244514,
      "grad_norm": 0.48717421293258667,
      "learning_rate": 4.952289492527576e-06,
      "loss": 0.5608,
      "step": 173
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 0.780621349811554,
      "learning_rate": 4.9509777202600605e-06,
      "loss": 0.4935,
      "step": 174
    },
    {
      "epoch": 0.822884012539185,
      "grad_norm": 0.49938517808914185,
      "learning_rate": 4.949648337008425e-06,
      "loss": 0.5409,
      "step": 175
    },
    {
      "epoch": 0.8275862068965517,
      "grad_norm": 0.5497399568557739,
      "learning_rate": 4.948301352324674e-06,
      "loss": 0.5343,
      "step": 176
    },
    {
      "epoch": 0.8322884012539185,
      "grad_norm": 0.49721062183380127,
      "learning_rate": 4.946936775887281e-06,
      "loss": 0.53,
      "step": 177
    },
    {
      "epoch": 0.8369905956112853,
      "grad_norm": 0.48117756843566895,
      "learning_rate": 4.945554617501124e-06,
      "loss": 0.5238,
      "step": 178
    },
    {
      "epoch": 0.841692789968652,
      "grad_norm": 0.49232199788093567,
      "learning_rate": 4.944154887097411e-06,
      "loss": 0.5554,
      "step": 179
    },
    {
      "epoch": 0.8463949843260188,
      "grad_norm": 0.49600890278816223,
      "learning_rate": 4.942737594733608e-06,
      "loss": 0.5265,
      "step": 180
    },
    {
      "epoch": 0.8510971786833855,
      "grad_norm": 0.502390444278717,
      "learning_rate": 4.941302750593373e-06,
      "loss": 0.5427,
      "step": 181
    },
    {
      "epoch": 0.8557993730407524,
      "grad_norm": 0.4632537066936493,
      "learning_rate": 4.939850364986475e-06,
      "loss": 0.4842,
      "step": 182
    },
    {
      "epoch": 0.8605015673981191,
      "grad_norm": 0.5433545708656311,
      "learning_rate": 4.938380448348725e-06,
      "loss": 0.4928,
      "step": 183
    },
    {
      "epoch": 0.8652037617554859,
      "grad_norm": 0.5086951851844788,
      "learning_rate": 4.9368930112419e-06,
      "loss": 0.5333,
      "step": 184
    },
    {
      "epoch": 0.8699059561128527,
      "grad_norm": 0.47532474994659424,
      "learning_rate": 4.935388064353665e-06,
      "loss": 0.5357,
      "step": 185
    },
    {
      "epoch": 0.8746081504702194,
      "grad_norm": 0.49426794052124023,
      "learning_rate": 4.9338656184975e-06,
      "loss": 0.5316,
      "step": 186
    },
    {
      "epoch": 0.8793103448275862,
      "grad_norm": 0.49909746646881104,
      "learning_rate": 4.932325684612618e-06,
      "loss": 0.543,
      "step": 187
    },
    {
      "epoch": 0.8840125391849529,
      "grad_norm": 0.5019994378089905,
      "learning_rate": 4.93076827376389e-06,
      "loss": 0.5464,
      "step": 188
    },
    {
      "epoch": 0.8887147335423198,
      "grad_norm": 0.5011009573936462,
      "learning_rate": 4.9291933971417635e-06,
      "loss": 0.543,
      "step": 189
    },
    {
      "epoch": 0.8934169278996865,
      "grad_norm": 0.4558373987674713,
      "learning_rate": 4.9276010660621835e-06,
      "loss": 0.5298,
      "step": 190
    },
    {
      "epoch": 0.8981191222570533,
      "grad_norm": 0.5847108960151672,
      "learning_rate": 4.925991291966508e-06,
      "loss": 0.5089,
      "step": 191
    },
    {
      "epoch": 0.9028213166144201,
      "grad_norm": 0.5339157581329346,
      "learning_rate": 4.92436408642143e-06,
      "loss": 0.5466,
      "step": 192
    },
    {
      "epoch": 0.9075235109717869,
      "grad_norm": 0.5474189519882202,
      "learning_rate": 4.9227194611188934e-06,
      "loss": 0.5224,
      "step": 193
    },
    {
      "epoch": 0.9122257053291536,
      "grad_norm": 0.4983315169811249,
      "learning_rate": 4.921057427876007e-06,
      "loss": 0.4919,
      "step": 194
    },
    {
      "epoch": 0.9169278996865203,
      "grad_norm": 0.5565828680992126,
      "learning_rate": 4.919377998634959e-06,
      "loss": 0.5499,
      "step": 195
    },
    {
      "epoch": 0.9216300940438872,
      "grad_norm": 0.9578695893287659,
      "learning_rate": 4.917681185462934e-06,
      "loss": 0.5356,
      "step": 196
    },
    {
      "epoch": 0.9263322884012539,
      "grad_norm": 0.5272259712219238,
      "learning_rate": 4.915967000552028e-06,
      "loss": 0.526,
      "step": 197
    },
    {
      "epoch": 0.9310344827586207,
      "grad_norm": 1.1794627904891968,
      "learning_rate": 4.914235456219154e-06,
      "loss": 0.5193,
      "step": 198
    },
    {
      "epoch": 0.9357366771159875,
      "grad_norm": 0.49176672101020813,
      "learning_rate": 4.912486564905959e-06,
      "loss": 0.5477,
      "step": 199
    },
    {
      "epoch": 0.9404388714733543,
      "grad_norm": 0.48104146122932434,
      "learning_rate": 4.910720339178735e-06,
      "loss": 0.5303,
      "step": 200
    },
    {
      "epoch": 0.945141065830721,
      "grad_norm": 0.48861178755760193,
      "learning_rate": 4.908936791728323e-06,
      "loss": 0.5353,
      "step": 201
    },
    {
      "epoch": 0.9498432601880877,
      "grad_norm": 0.5078539848327637,
      "learning_rate": 4.907135935370027e-06,
      "loss": 0.5336,
      "step": 202
    },
    {
      "epoch": 0.9545454545454546,
      "grad_norm": 0.5397623777389526,
      "learning_rate": 4.905317783043523e-06,
      "loss": 0.5422,
      "step": 203
    },
    {
      "epoch": 0.9592476489028213,
      "grad_norm": 0.5742900371551514,
      "learning_rate": 4.9034823478127605e-06,
      "loss": 0.5225,
      "step": 204
    },
    {
      "epoch": 0.9639498432601881,
      "grad_norm": 0.4920111298561096,
      "learning_rate": 4.901629642865872e-06,
      "loss": 0.5001,
      "step": 205
    },
    {
      "epoch": 0.9686520376175548,
      "grad_norm": 0.5811502933502197,
      "learning_rate": 4.89975968151508e-06,
      "loss": 0.5237,
      "step": 206
    },
    {
      "epoch": 0.9733542319749217,
      "grad_norm": 0.9640730619430542,
      "learning_rate": 4.8978724771965965e-06,
      "loss": 0.5136,
      "step": 207
    },
    {
      "epoch": 0.9780564263322884,
      "grad_norm": 0.5389460921287537,
      "learning_rate": 4.895968043470532e-06,
      "loss": 0.5306,
      "step": 208
    },
    {
      "epoch": 0.9827586206896551,
      "grad_norm": 0.6080204248428345,
      "learning_rate": 4.894046394020794e-06,
      "loss": 0.4957,
      "step": 209
    },
    {
      "epoch": 0.987460815047022,
      "grad_norm": 0.5144377946853638,
      "learning_rate": 4.892107542654988e-06,
      "loss": 0.5275,
      "step": 210
    },
    {
      "epoch": 0.9921630094043887,
      "grad_norm": 0.4928152859210968,
      "learning_rate": 4.890151503304325e-06,
      "loss": 0.5496,
      "step": 211
    },
    {
      "epoch": 0.9968652037617555,
      "grad_norm": 0.4954489469528198,
      "learning_rate": 4.88817829002351e-06,
      "loss": 0.5247,
      "step": 212
    }
  ],
  "logging_steps": 1,
  "max_steps": 1272,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 212,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.3638348518765625e+19,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}