{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.9968652037617556,
  "eval_steps": 500,
  "global_step": 848,
  "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
    },
    {
      "epoch": 1.0047021943573669,
      "grad_norm": 1.0198856592178345,
      "learning_rate": 4.886187916990653e-06,
      "loss": 1.0557,
      "step": 213
    },
    {
      "epoch": 1.0094043887147335,
      "grad_norm": 0.4713587462902069,
      "learning_rate": 4.884180398507163e-06,
      "loss": 0.5031,
      "step": 214
    },
    {
      "epoch": 1.0141065830721003,
      "grad_norm": 0.46703630685806274,
      "learning_rate": 4.882155748997636e-06,
      "loss": 0.4952,
      "step": 215
    },
    {
      "epoch": 1.0188087774294672,
      "grad_norm": 0.48665955662727356,
      "learning_rate": 4.8801139830097685e-06,
      "loss": 0.5024,
      "step": 216
    },
    {
      "epoch": 1.0235109717868338,
      "grad_norm": 0.5477427840232849,
      "learning_rate": 4.878055115214238e-06,
      "loss": 0.5124,
      "step": 217
    },
    {
      "epoch": 1.0282131661442007,
      "grad_norm": 0.4587133228778839,
      "learning_rate": 4.875979160404607e-06,
      "loss": 0.5098,
      "step": 218
    },
    {
      "epoch": 1.0329153605015673,
      "grad_norm": 0.5163841247558594,
      "learning_rate": 4.873886133497209e-06,
      "loss": 0.5187,
      "step": 219
    },
    {
      "epoch": 1.0376175548589341,
      "grad_norm": 0.4965426027774811,
      "learning_rate": 4.87177604953105e-06,
      "loss": 0.5148,
      "step": 220
    },
    {
      "epoch": 1.042319749216301,
      "grad_norm": 0.5206760764122009,
      "learning_rate": 4.869648923667694e-06,
      "loss": 0.4681,
      "step": 221
    },
    {
      "epoch": 1.0470219435736676,
      "grad_norm": 2.3893775939941406,
      "learning_rate": 4.867504771191154e-06,
      "loss": 0.4942,
      "step": 222
    },
    {
      "epoch": 1.0517241379310345,
      "grad_norm": 0.4828566312789917,
      "learning_rate": 4.865343607507788e-06,
      "loss": 0.5019,
      "step": 223
    },
    {
      "epoch": 1.0564263322884013,
      "grad_norm": 0.5338858962059021,
      "learning_rate": 4.86316544814618e-06,
      "loss": 0.5172,
      "step": 224
    },
    {
      "epoch": 1.061128526645768,
      "grad_norm": 0.5287072062492371,
      "learning_rate": 4.860970308757038e-06,
      "loss": 0.462,
      "step": 225
    },
    {
      "epoch": 1.0658307210031348,
      "grad_norm": 0.4908182621002197,
      "learning_rate": 4.858758205113072e-06,
      "loss": 0.4908,
      "step": 226
    },
    {
      "epoch": 1.0705329153605017,
      "grad_norm": 0.4617731273174286,
      "learning_rate": 4.856529153108888e-06,
      "loss": 0.5229,
      "step": 227
    },
    {
      "epoch": 1.0752351097178683,
      "grad_norm": 0.8010035753250122,
      "learning_rate": 4.854283168760868e-06,
      "loss": 0.5016,
      "step": 228
    },
    {
      "epoch": 1.0799373040752351,
      "grad_norm": 0.5060340762138367,
      "learning_rate": 4.85202026820706e-06,
      "loss": 0.4996,
      "step": 229
    },
    {
      "epoch": 1.084639498432602,
      "grad_norm": 0.5193985104560852,
      "learning_rate": 4.84974046770706e-06,
      "loss": 0.5361,
      "step": 230
    },
    {
      "epoch": 1.0893416927899686,
      "grad_norm": 0.4898582100868225,
      "learning_rate": 4.847443783641893e-06,
      "loss": 0.4466,
      "step": 231
    },
    {
      "epoch": 1.0940438871473355,
      "grad_norm": 0.49226534366607666,
      "learning_rate": 4.845130232513901e-06,
      "loss": 0.4921,
      "step": 232
    },
    {
      "epoch": 1.098746081504702,
      "grad_norm": 0.531612753868103,
      "learning_rate": 4.842799830946615e-06,
      "loss": 0.4883,
      "step": 233
    },
    {
      "epoch": 1.103448275862069,
      "grad_norm": 0.48687735199928284,
      "learning_rate": 4.840452595684646e-06,
      "loss": 0.4845,
      "step": 234
    },
    {
      "epoch": 1.1081504702194358,
      "grad_norm": 0.5371798276901245,
      "learning_rate": 4.83808854359356e-06,
      "loss": 0.4788,
      "step": 235
    },
    {
      "epoch": 1.1128526645768024,
      "grad_norm": 0.4869842529296875,
      "learning_rate": 4.835707691659753e-06,
      "loss": 0.4821,
      "step": 236
    },
    {
      "epoch": 1.1175548589341693,
      "grad_norm": 0.6263965368270874,
      "learning_rate": 4.8333100569903365e-06,
      "loss": 0.4936,
      "step": 237
    },
    {
      "epoch": 1.1222570532915361,
      "grad_norm": 0.5065476298332214,
      "learning_rate": 4.8308956568130094e-06,
      "loss": 0.5136,
      "step": 238
    },
    {
      "epoch": 1.1269592476489028,
      "grad_norm": 0.5552216172218323,
      "learning_rate": 4.828464508475934e-06,
      "loss": 0.5072,
      "step": 239
    },
    {
      "epoch": 1.1316614420062696,
      "grad_norm": 0.5835661888122559,
      "learning_rate": 4.826016629447616e-06,
      "loss": 0.509,
      "step": 240
    },
    {
      "epoch": 1.1363636363636362,
      "grad_norm": 0.4653923511505127,
      "learning_rate": 4.823552037316775e-06,
      "loss": 0.4845,
      "step": 241
    },
    {
      "epoch": 1.141065830721003,
      "grad_norm": 0.6101334095001221,
      "learning_rate": 4.821070749792218e-06,
      "loss": 0.5401,
      "step": 242
    },
    {
      "epoch": 1.14576802507837,
      "grad_norm": 0.6516410708427429,
      "learning_rate": 4.818572784702713e-06,
      "loss": 0.5107,
      "step": 243
    },
    {
      "epoch": 1.1504702194357366,
      "grad_norm": 0.5194690823554993,
      "learning_rate": 4.816058159996863e-06,
      "loss": 0.521,
      "step": 244
    },
    {
      "epoch": 1.1551724137931034,
      "grad_norm": 0.4790078401565552,
      "learning_rate": 4.813526893742972e-06,
      "loss": 0.5038,
      "step": 245
    },
    {
      "epoch": 1.1598746081504703,
      "grad_norm": 0.4948838949203491,
      "learning_rate": 4.810979004128924e-06,
      "loss": 0.4966,
      "step": 246
    },
    {
      "epoch": 1.164576802507837,
      "grad_norm": 0.49500057101249695,
      "learning_rate": 4.808414509462042e-06,
      "loss": 0.5171,
      "step": 247
    },
    {
      "epoch": 1.1692789968652038,
      "grad_norm": 0.46993380784988403,
      "learning_rate": 4.80583342816896e-06,
      "loss": 0.484,
      "step": 248
    },
    {
      "epoch": 1.1739811912225706,
      "grad_norm": 0.5349460244178772,
      "learning_rate": 4.803235778795496e-06,
      "loss": 0.5232,
      "step": 249
    },
    {
      "epoch": 1.1786833855799372,
      "grad_norm": 0.7706441283226013,
      "learning_rate": 4.800621580006511e-06,
      "loss": 0.4677,
      "step": 250
    },
    {
      "epoch": 1.183385579937304,
      "grad_norm": 0.4796987473964691,
      "learning_rate": 4.797990850585782e-06,
      "loss": 0.5151,
      "step": 251
    },
    {
      "epoch": 1.188087774294671,
      "grad_norm": 1.0548434257507324,
      "learning_rate": 4.79534360943586e-06,
      "loss": 0.4954,
      "step": 252
    },
    {
      "epoch": 1.1927899686520376,
      "grad_norm": 0.6616916060447693,
      "learning_rate": 4.792679875577937e-06,
      "loss": 0.4803,
      "step": 253
    },
    {
      "epoch": 1.1974921630094044,
      "grad_norm": 0.5972627997398376,
      "learning_rate": 4.789999668151714e-06,
      "loss": 0.5145,
      "step": 254
    },
    {
      "epoch": 1.2021943573667713,
      "grad_norm": 0.46703872084617615,
      "learning_rate": 4.7873030064152545e-06,
      "loss": 0.4946,
      "step": 255
    },
    {
      "epoch": 1.206896551724138,
      "grad_norm": 0.477717787027359,
      "learning_rate": 4.784589909744856e-06,
      "loss": 0.4891,
      "step": 256
    },
    {
      "epoch": 1.2115987460815048,
      "grad_norm": 0.5048542022705078,
      "learning_rate": 4.7818603976349005e-06,
      "loss": 0.5018,
      "step": 257
    },
    {
      "epoch": 1.2163009404388714,
      "grad_norm": 0.4797464907169342,
      "learning_rate": 4.779114489697724e-06,
      "loss": 0.4978,
      "step": 258
    },
    {
      "epoch": 1.2210031347962382,
      "grad_norm": 0.5975466966629028,
      "learning_rate": 4.776352205663469e-06,
      "loss": 0.5009,
      "step": 259
    },
    {
      "epoch": 1.225705329153605,
      "grad_norm": 0.9671257734298706,
      "learning_rate": 4.773573565379947e-06,
      "loss": 0.5118,
      "step": 260
    },
    {
      "epoch": 1.2304075235109717,
      "grad_norm": 0.5943232178688049,
      "learning_rate": 4.770778588812489e-06,
      "loss": 0.4772,
      "step": 261
    },
    {
      "epoch": 1.2351097178683386,
      "grad_norm": 0.4887758493423462,
      "learning_rate": 4.7679672960438135e-06,
      "loss": 0.5038,
      "step": 262
    },
    {
      "epoch": 1.2398119122257054,
      "grad_norm": 0.6083370447158813,
      "learning_rate": 4.765139707273872e-06,
      "loss": 0.4916,
      "step": 263
    },
    {
      "epoch": 1.244514106583072,
      "grad_norm": 0.49385467171669006,
      "learning_rate": 4.762295842819707e-06,
      "loss": 0.5036,
      "step": 264
    },
    {
      "epoch": 1.249216300940439,
      "grad_norm": 0.5786624550819397,
      "learning_rate": 4.759435723115308e-06,
      "loss": 0.4692,
      "step": 265
    },
    {
      "epoch": 1.2539184952978055,
      "grad_norm": 0.48176029324531555,
      "learning_rate": 4.756559368711463e-06,
      "loss": 0.5065,
      "step": 266
    },
    {
      "epoch": 1.2586206896551724,
      "grad_norm": 0.5976678729057312,
      "learning_rate": 4.75366680027561e-06,
      "loss": 0.4879,
      "step": 267
    },
    {
      "epoch": 1.2633228840125392,
      "grad_norm": 0.5139954686164856,
      "learning_rate": 4.7507580385916906e-06,
      "loss": 0.4906,
      "step": 268
    },
    {
      "epoch": 1.2680250783699059,
      "grad_norm": 0.5675197839736938,
      "learning_rate": 4.747833104559999e-06,
      "loss": 0.4667,
      "step": 269
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 0.48431986570358276,
      "learning_rate": 4.744892019197033e-06,
      "loss": 0.4809,
      "step": 270
    },
    {
      "epoch": 1.2774294670846396,
      "grad_norm": 0.5017510056495667,
      "learning_rate": 4.74193480363534e-06,
      "loss": 0.4874,
      "step": 271
    },
    {
      "epoch": 1.2821316614420062,
      "grad_norm": 0.5467999577522278,
      "learning_rate": 4.738961479123373e-06,
      "loss": 0.497,
      "step": 272
    },
    {
      "epoch": 1.286833855799373,
      "grad_norm": 1.3385310173034668,
      "learning_rate": 4.735972067025326e-06,
      "loss": 0.5027,
      "step": 273
    },
    {
      "epoch": 1.29153605015674,
      "grad_norm": 0.5216737389564514,
      "learning_rate": 4.732966588820991e-06,
      "loss": 0.4953,
      "step": 274
    },
    {
      "epoch": 1.2962382445141065,
      "grad_norm": 0.5195981860160828,
      "learning_rate": 4.729945066105599e-06,
      "loss": 0.474,
      "step": 275
    },
    {
      "epoch": 1.3009404388714734,
      "grad_norm": 0.47055330872535706,
      "learning_rate": 4.726907520589664e-06,
      "loss": 0.4659,
      "step": 276
    },
    {
      "epoch": 1.3056426332288402,
      "grad_norm": 0.5389472246170044,
      "learning_rate": 4.72385397409883e-06,
      "loss": 0.5076,
      "step": 277
    },
    {
      "epoch": 1.3103448275862069,
      "grad_norm": 0.506074070930481,
      "learning_rate": 4.720784448573712e-06,
      "loss": 0.4993,
      "step": 278
    },
    {
      "epoch": 1.3150470219435737,
      "grad_norm": 0.5159019827842712,
      "learning_rate": 4.717698966069739e-06,
      "loss": 0.5292,
      "step": 279
    },
    {
      "epoch": 1.3197492163009406,
      "grad_norm": 0.5021179914474487,
      "learning_rate": 4.7145975487569965e-06,
      "loss": 0.5076,
      "step": 280
    },
    {
      "epoch": 1.3244514106583072,
      "grad_norm": 0.5160107016563416,
      "learning_rate": 4.711480218920064e-06,
      "loss": 0.4693,
      "step": 281
    },
    {
      "epoch": 1.329153605015674,
      "grad_norm": 0.5437175631523132,
      "learning_rate": 4.708346998957859e-06,
      "loss": 0.5179,
      "step": 282
    },
    {
      "epoch": 1.3338557993730409,
      "grad_norm": 0.5456722378730774,
      "learning_rate": 4.705197911383473e-06,
      "loss": 0.489,
      "step": 283
    },
    {
      "epoch": 1.3385579937304075,
      "grad_norm": 0.4638400375843048,
      "learning_rate": 4.7020329788240115e-06,
      "loss": 0.4752,
      "step": 284
    },
    {
      "epoch": 1.3432601880877744,
      "grad_norm": 0.5053242444992065,
      "learning_rate": 4.6988522240204325e-06,
      "loss": 0.4847,
      "step": 285
    },
    {
      "epoch": 1.347962382445141,
      "grad_norm": 0.5066615343093872,
      "learning_rate": 4.695655669827377e-06,
      "loss": 0.4992,
      "step": 286
    },
    {
      "epoch": 1.3526645768025078,
      "grad_norm": 0.5127485990524292,
      "learning_rate": 4.6924433392130135e-06,
      "loss": 0.4915,
      "step": 287
    },
    {
      "epoch": 1.3573667711598745,
      "grad_norm": 0.5612887144088745,
      "learning_rate": 4.689215255258866e-06,
      "loss": 0.5093,
      "step": 288
    },
    {
      "epoch": 1.3620689655172413,
      "grad_norm": 0.47511914372444153,
      "learning_rate": 4.685971441159653e-06,
      "loss": 0.4784,
      "step": 289
    },
    {
      "epoch": 1.3667711598746082,
      "grad_norm": 0.5002272129058838,
      "learning_rate": 4.682711920223115e-06,
      "loss": 0.4774,
      "step": 290
    },
    {
      "epoch": 1.3714733542319748,
      "grad_norm": 0.5161554217338562,
      "learning_rate": 4.679436715869856e-06,
      "loss": 0.4977,
      "step": 291
    },
    {
      "epoch": 1.3761755485893417,
      "grad_norm": 0.5877403020858765,
      "learning_rate": 4.676145851633166e-06,
      "loss": 0.515,
      "step": 292
    },
    {
      "epoch": 1.3808777429467085,
      "grad_norm": 0.47495776414871216,
      "learning_rate": 4.672839351158856e-06,
      "loss": 0.5002,
      "step": 293
    },
    {
      "epoch": 1.3855799373040751,
      "grad_norm": 0.5374031066894531,
      "learning_rate": 4.669517238205089e-06,
      "loss": 0.4841,
      "step": 294
    },
    {
      "epoch": 1.390282131661442,
      "grad_norm": 0.5107132792472839,
      "learning_rate": 4.666179536642208e-06,
      "loss": 0.4847,
      "step": 295
    },
    {
      "epoch": 1.3949843260188088,
      "grad_norm": 0.48953261971473694,
      "learning_rate": 4.662826270452565e-06,
      "loss": 0.4806,
      "step": 296
    },
    {
      "epoch": 1.3996865203761755,
      "grad_norm": 0.5226054191589355,
      "learning_rate": 4.659457463730347e-06,
      "loss": 0.4908,
      "step": 297
    },
    {
      "epoch": 1.4043887147335423,
      "grad_norm": 0.5019803643226624,
      "learning_rate": 4.6560731406814056e-06,
      "loss": 0.5054,
      "step": 298
    },
    {
      "epoch": 1.4090909090909092,
      "grad_norm": 0.5146315693855286,
      "learning_rate": 4.65267332562308e-06,
      "loss": 0.5076,
      "step": 299
    },
    {
      "epoch": 1.4137931034482758,
      "grad_norm": 0.5293256640434265,
      "learning_rate": 4.649258042984026e-06,
      "loss": 0.506,
      "step": 300
    },
    {
      "epoch": 1.4184952978056427,
      "grad_norm": 0.536259651184082,
      "learning_rate": 4.6458273173040395e-06,
      "loss": 0.4622,
      "step": 301
    },
    {
      "epoch": 1.4231974921630095,
      "grad_norm": 0.5373066067695618,
      "learning_rate": 4.642381173233874e-06,
      "loss": 0.4995,
      "step": 302
    },
    {
      "epoch": 1.4278996865203761,
      "grad_norm": 0.5316248536109924,
      "learning_rate": 4.638919635535073e-06,
      "loss": 0.4555,
      "step": 303
    },
    {
      "epoch": 1.432601880877743,
      "grad_norm": 1.0662360191345215,
      "learning_rate": 4.635442729079788e-06,
      "loss": 0.4831,
      "step": 304
    },
    {
      "epoch": 1.4373040752351098,
      "grad_norm": 0.6498283743858337,
      "learning_rate": 4.6319504788505956e-06,
      "loss": 0.4774,
      "step": 305
    },
    {
      "epoch": 1.4420062695924765,
      "grad_norm": 0.4853154122829437,
      "learning_rate": 4.628442909940325e-06,
      "loss": 0.4897,
      "step": 306
    },
    {
      "epoch": 1.4467084639498433,
      "grad_norm": 0.8411519527435303,
      "learning_rate": 4.624920047551874e-06,
      "loss": 0.5064,
      "step": 307
    },
    {
      "epoch": 1.4514106583072102,
      "grad_norm": 0.547363817691803,
      "learning_rate": 4.621381916998029e-06,
      "loss": 0.4733,
      "step": 308
    },
    {
      "epoch": 1.4561128526645768,
      "grad_norm": 0.5782042145729065,
      "learning_rate": 4.6178285437012806e-06,
      "loss": 0.5085,
      "step": 309
    },
    {
      "epoch": 1.4608150470219436,
      "grad_norm": 0.5078956484794617,
      "learning_rate": 4.6142599531936435e-06,
      "loss": 0.4715,
      "step": 310
    },
    {
      "epoch": 1.4655172413793103,
      "grad_norm": 0.49163371324539185,
      "learning_rate": 4.610676171116475e-06,
      "loss": 0.492,
      "step": 311
    },
    {
      "epoch": 1.4702194357366771,
      "grad_norm": 0.5490444898605347,
      "learning_rate": 4.607077223220286e-06,
      "loss": 0.4954,
      "step": 312
    },
    {
      "epoch": 1.4749216300940438,
      "grad_norm": 0.5161562561988831,
      "learning_rate": 4.603463135364556e-06,
      "loss": 0.4659,
      "step": 313
    },
    {
      "epoch": 1.4796238244514106,
      "grad_norm": 0.4908446669578552,
      "learning_rate": 4.5998339335175555e-06,
      "loss": 0.4884,
      "step": 314
    },
    {
      "epoch": 1.4843260188087775,
      "grad_norm": 0.5703722834587097,
      "learning_rate": 4.596189643756147e-06,
      "loss": 0.4659,
      "step": 315
    },
    {
      "epoch": 1.489028213166144,
      "grad_norm": 0.6954686045646667,
      "learning_rate": 4.592530292265609e-06,
      "loss": 0.4834,
      "step": 316
    },
    {
      "epoch": 1.493730407523511,
      "grad_norm": 0.498858243227005,
      "learning_rate": 4.58885590533944e-06,
      "loss": 0.4925,
      "step": 317
    },
    {
      "epoch": 1.4984326018808778,
      "grad_norm": 0.4805036783218384,
      "learning_rate": 4.585166509379173e-06,
      "loss": 0.5152,
      "step": 318
    },
    {
      "epoch": 1.5031347962382444,
      "grad_norm": 0.5069772005081177,
      "learning_rate": 4.581462130894186e-06,
      "loss": 0.4926,
      "step": 319
    },
    {
      "epoch": 1.5078369905956113,
      "grad_norm": 0.5127174854278564,
      "learning_rate": 4.57774279650151e-06,
      "loss": 0.4839,
      "step": 320
    },
    {
      "epoch": 1.5125391849529781,
      "grad_norm": 0.5314028263092041,
      "learning_rate": 4.574008532925638e-06,
      "loss": 0.5101,
      "step": 321
    },
    {
      "epoch": 1.5172413793103448,
      "grad_norm": 0.4849459230899811,
      "learning_rate": 4.570259366998336e-06,
      "loss": 0.495,
      "step": 322
    },
    {
      "epoch": 1.5219435736677116,
      "grad_norm": 0.5260595679283142,
      "learning_rate": 4.566495325658445e-06,
      "loss": 0.5232,
      "step": 323
    },
    {
      "epoch": 1.5266457680250785,
      "grad_norm": 0.4890941083431244,
      "learning_rate": 4.5627164359516915e-06,
      "loss": 0.5034,
      "step": 324
    },
    {
      "epoch": 1.531347962382445,
      "grad_norm": 0.516516387462616,
      "learning_rate": 4.558922725030491e-06,
      "loss": 0.4773,
      "step": 325
    },
    {
      "epoch": 1.536050156739812,
      "grad_norm": 0.622726321220398,
      "learning_rate": 4.555114220153755e-06,
      "loss": 0.4306,
      "step": 326
    },
    {
      "epoch": 1.5407523510971788,
      "grad_norm": 0.50007164478302,
      "learning_rate": 4.551290948686693e-06,
      "loss": 0.5134,
      "step": 327
    },
    {
      "epoch": 1.5454545454545454,
      "grad_norm": 0.669569194316864,
      "learning_rate": 4.547452938100615e-06,
      "loss": 0.5173,
      "step": 328
    },
    {
      "epoch": 1.5501567398119123,
      "grad_norm": 0.5498544573783875,
      "learning_rate": 4.54360021597274e-06,
      "loss": 0.4946,
      "step": 329
    },
    {
      "epoch": 1.5548589341692791,
      "grad_norm": 0.5294123888015747,
      "learning_rate": 4.539732809985989e-06,
      "loss": 0.4853,
      "step": 330
    },
    {
      "epoch": 1.5595611285266457,
      "grad_norm": 0.5199066996574402,
      "learning_rate": 4.535850747928796e-06,
      "loss": 0.4989,
      "step": 331
    },
    {
      "epoch": 1.5642633228840124,
      "grad_norm": 0.5348268151283264,
      "learning_rate": 4.531954057694897e-06,
      "loss": 0.4823,
      "step": 332
    },
    {
      "epoch": 1.5689655172413794,
      "grad_norm": 0.47487759590148926,
      "learning_rate": 4.5280427672831414e-06,
      "loss": 0.4894,
      "step": 333
    },
    {
      "epoch": 1.573667711598746,
      "grad_norm": 0.4653705060482025,
      "learning_rate": 4.524116904797281e-06,
      "loss": 0.4656,
      "step": 334
    },
    {
      "epoch": 1.5783699059561127,
      "grad_norm": 0.5206255912780762,
      "learning_rate": 4.520176498445774e-06,
      "loss": 0.4782,
      "step": 335
    },
    {
      "epoch": 1.5830721003134798,
      "grad_norm": 0.7349448800086975,
      "learning_rate": 4.516221576541581e-06,
      "loss": 0.4781,
      "step": 336
    },
    {
      "epoch": 1.5877742946708464,
      "grad_norm": 0.4842517077922821,
      "learning_rate": 4.512252167501959e-06,
      "loss": 0.4781,
      "step": 337
    },
    {
      "epoch": 1.592476489028213,
      "grad_norm": 0.5287013053894043,
      "learning_rate": 4.508268299848262e-06,
      "loss": 0.4841,
      "step": 338
    },
    {
      "epoch": 1.59717868338558,
      "grad_norm": 0.5405049324035645,
      "learning_rate": 4.50427000220573e-06,
      "loss": 0.4986,
      "step": 339
    },
    {
      "epoch": 1.6018808777429467,
      "grad_norm": 0.5042012333869934,
      "learning_rate": 4.50025730330329e-06,
      "loss": 0.4757,
      "step": 340
    },
    {
      "epoch": 1.6065830721003134,
      "grad_norm": 0.5087094902992249,
      "learning_rate": 4.4962302319733445e-06,
      "loss": 0.4927,
      "step": 341
    },
    {
      "epoch": 1.6112852664576802,
      "grad_norm": 0.5267127156257629,
      "learning_rate": 4.492188817151565e-06,
      "loss": 0.5282,
      "step": 342
    },
    {
      "epoch": 1.615987460815047,
      "grad_norm": 0.5424356460571289,
      "learning_rate": 4.488133087876688e-06,
      "loss": 0.4664,
      "step": 343
    },
    {
      "epoch": 1.6206896551724137,
      "grad_norm": 0.7226302027702332,
      "learning_rate": 4.484063073290301e-06,
      "loss": 0.4648,
      "step": 344
    },
    {
      "epoch": 1.6253918495297806,
      "grad_norm": 0.5150883197784424,
      "learning_rate": 4.479978802636637e-06,
      "loss": 0.4995,
      "step": 345
    },
    {
      "epoch": 1.6300940438871474,
      "grad_norm": 0.6959285140037537,
      "learning_rate": 4.475880305262362e-06,
      "loss": 0.5036,
      "step": 346
    },
    {
      "epoch": 1.634796238244514,
      "grad_norm": 0.48316431045532227,
      "learning_rate": 4.471767610616366e-06,
      "loss": 0.4956,
      "step": 347
    },
    {
      "epoch": 1.6394984326018809,
      "grad_norm": 0.5120992064476013,
      "learning_rate": 4.467640748249549e-06,
      "loss": 0.47,
      "step": 348
    },
    {
      "epoch": 1.6442006269592477,
      "grad_norm": 0.495301753282547,
      "learning_rate": 4.4634997478146125e-06,
      "loss": 0.4889,
      "step": 349
    },
    {
      "epoch": 1.6489028213166144,
      "grad_norm": 0.5113374590873718,
      "learning_rate": 4.459344639065842e-06,
      "loss": 0.4814,
      "step": 350
    },
    {
      "epoch": 1.6536050156739812,
      "grad_norm": 0.5123418569564819,
      "learning_rate": 4.455175451858897e-06,
      "loss": 0.4897,
      "step": 351
    },
    {
      "epoch": 1.658307210031348,
      "grad_norm": 0.4938543438911438,
      "learning_rate": 4.450992216150592e-06,
      "loss": 0.4981,
      "step": 352
    },
    {
      "epoch": 1.6630094043887147,
      "grad_norm": 0.6186752319335938,
      "learning_rate": 4.446794961998689e-06,
      "loss": 0.4669,
      "step": 353
    },
    {
      "epoch": 1.6677115987460815,
      "grad_norm": 0.46507078409194946,
      "learning_rate": 4.442583719561671e-06,
      "loss": 0.4927,
      "step": 354
    },
    {
      "epoch": 1.6724137931034484,
      "grad_norm": 0.6733965873718262,
      "learning_rate": 4.438358519098536e-06,
      "loss": 0.5021,
      "step": 355
    },
    {
      "epoch": 1.677115987460815,
      "grad_norm": 0.49244073033332825,
      "learning_rate": 4.4341193909685685e-06,
      "loss": 0.485,
      "step": 356
    },
    {
      "epoch": 1.6818181818181817,
      "grad_norm": 0.5992756485939026,
      "learning_rate": 4.429866365631134e-06,
      "loss": 0.4906,
      "step": 357
    },
    {
      "epoch": 1.6865203761755487,
      "grad_norm": 0.5802706480026245,
      "learning_rate": 4.425599473645447e-06,
      "loss": 0.4813,
      "step": 358
    },
    {
      "epoch": 1.6912225705329154,
      "grad_norm": 0.5721625685691833,
      "learning_rate": 4.421318745670364e-06,
      "loss": 0.4815,
      "step": 359
    },
    {
      "epoch": 1.695924764890282,
      "grad_norm": 0.4937971234321594,
      "learning_rate": 4.4170242124641524e-06,
      "loss": 0.4596,
      "step": 360
    },
    {
      "epoch": 1.700626959247649,
      "grad_norm": 0.49360817670822144,
      "learning_rate": 4.412715904884277e-06,
      "loss": 0.4893,
      "step": 361
    },
    {
      "epoch": 1.7053291536050157,
      "grad_norm": 0.5080047845840454,
      "learning_rate": 4.4083938538871735e-06,
      "loss": 0.4682,
      "step": 362
    },
    {
      "epoch": 1.7100313479623823,
      "grad_norm": 0.5415318608283997,
      "learning_rate": 4.4040580905280295e-06,
      "loss": 0.487,
      "step": 363
    },
    {
      "epoch": 1.7147335423197492,
      "grad_norm": 0.607445240020752,
      "learning_rate": 4.3997086459605586e-06,
      "loss": 0.4837,
      "step": 364
    },
    {
      "epoch": 1.719435736677116,
      "grad_norm": 0.47159233689308167,
      "learning_rate": 4.395345551436779e-06,
      "loss": 0.5081,
      "step": 365
    },
    {
      "epoch": 1.7241379310344827,
      "grad_norm": 0.516380786895752,
      "learning_rate": 4.390968838306788e-06,
      "loss": 0.4634,
      "step": 366
    },
    {
      "epoch": 1.7288401253918495,
      "grad_norm": 0.4957869350910187,
      "learning_rate": 4.386578538018535e-06,
      "loss": 0.4618,
      "step": 367
    },
    {
      "epoch": 1.7335423197492164,
      "grad_norm": 0.5057672262191772,
      "learning_rate": 4.382174682117598e-06,
      "loss": 0.5083,
      "step": 368
    },
    {
      "epoch": 1.738244514106583,
      "grad_norm": 0.5154942274093628,
      "learning_rate": 4.377757302246956e-06,
      "loss": 0.444,
      "step": 369
    },
    {
      "epoch": 1.7429467084639498,
      "grad_norm": 0.51052325963974,
      "learning_rate": 4.373326430146762e-06,
      "loss": 0.5009,
      "step": 370
    },
    {
      "epoch": 1.7476489028213167,
      "grad_norm": 0.4869065582752228,
      "learning_rate": 4.368882097654113e-06,
      "loss": 0.4956,
      "step": 371
    },
    {
      "epoch": 1.7523510971786833,
      "grad_norm": 0.46982306241989136,
      "learning_rate": 4.364424336702825e-06,
      "loss": 0.4701,
      "step": 372
    },
    {
      "epoch": 1.7570532915360502,
      "grad_norm": 0.5106102824211121,
      "learning_rate": 4.3599531793232e-06,
      "loss": 0.4879,
      "step": 373
    },
    {
      "epoch": 1.761755485893417,
      "grad_norm": 0.6152806282043457,
      "learning_rate": 4.355468657641797e-06,
      "loss": 0.4825,
      "step": 374
    },
    {
      "epoch": 1.7664576802507836,
      "grad_norm": 0.523330807685852,
      "learning_rate": 4.3509708038812035e-06,
      "loss": 0.4863,
      "step": 375
    },
    {
      "epoch": 1.7711598746081505,
      "grad_norm": 0.505370020866394,
      "learning_rate": 4.346459650359798e-06,
      "loss": 0.4845,
      "step": 376
    },
    {
      "epoch": 1.7758620689655173,
      "grad_norm": 4.284893989562988,
      "learning_rate": 4.341935229491525e-06,
      "loss": 0.4543,
      "step": 377
    },
    {
      "epoch": 1.780564263322884,
      "grad_norm": 0.4744650423526764,
      "learning_rate": 4.337397573785659e-06,
      "loss": 0.5041,
      "step": 378
    },
    {
      "epoch": 1.7852664576802508,
      "grad_norm": 0.8564308285713196,
      "learning_rate": 4.332846715846566e-06,
      "loss": 0.4704,
      "step": 379
    },
    {
      "epoch": 1.7899686520376177,
      "grad_norm": 0.573800265789032,
      "learning_rate": 4.328282688373479e-06,
      "loss": 0.491,
      "step": 380
    },
    {
      "epoch": 1.7946708463949843,
      "grad_norm": 0.5891814827919006,
      "learning_rate": 4.323705524160258e-06,
      "loss": 0.4892,
      "step": 381
    },
    {
      "epoch": 1.799373040752351,
      "grad_norm": 0.5569722056388855,
      "learning_rate": 4.319115256095149e-06,
      "loss": 0.4657,
      "step": 382
    },
    {
      "epoch": 1.804075235109718,
      "grad_norm": 0.4922447204589844,
      "learning_rate": 4.314511917160557e-06,
      "loss": 0.4669,
      "step": 383
    },
    {
      "epoch": 1.8087774294670846,
      "grad_norm": 0.49718669056892395,
      "learning_rate": 4.3098955404328045e-06,
      "loss": 0.4593,
      "step": 384
    },
    {
      "epoch": 1.8134796238244513,
      "grad_norm": 0.544978141784668,
      "learning_rate": 4.305266159081895e-06,
      "loss": 0.4806,
      "step": 385
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.5504943132400513,
      "learning_rate": 4.3006238063712725e-06,
      "loss": 0.4663,
      "step": 386
    },
    {
      "epoch": 1.822884012539185,
      "grad_norm": 0.5914635062217712,
      "learning_rate": 4.295968515657583e-06,
      "loss": 0.5008,
      "step": 387
    },
    {
      "epoch": 1.8275862068965516,
      "grad_norm": 0.6797587275505066,
      "learning_rate": 4.29130032039044e-06,
      "loss": 0.4837,
      "step": 388
    },
    {
      "epoch": 1.8322884012539185,
      "grad_norm": 0.49545228481292725,
      "learning_rate": 4.2866192541121755e-06,
      "loss": 0.4741,
      "step": 389
    },
    {
      "epoch": 1.8369905956112853,
      "grad_norm": 0.5231736898422241,
      "learning_rate": 4.281925350457606e-06,
      "loss": 0.4763,
      "step": 390
    },
    {
      "epoch": 1.841692789968652,
      "grad_norm": 0.4704969525337219,
      "learning_rate": 4.277218643153787e-06,
      "loss": 0.4778,
      "step": 391
    },
    {
      "epoch": 1.8463949843260188,
      "grad_norm": 0.5036218762397766,
      "learning_rate": 4.272499166019771e-06,
      "loss": 0.4778,
      "step": 392
    },
    {
      "epoch": 1.8510971786833856,
      "grad_norm": 0.5395018458366394,
      "learning_rate": 4.267766952966369e-06,
      "loss": 0.4659,
      "step": 393
    },
    {
      "epoch": 1.8557993730407523,
      "grad_norm": 0.5380336046218872,
      "learning_rate": 4.2630220379959006e-06,
      "loss": 0.4433,
      "step": 394
    },
    {
      "epoch": 1.8605015673981191,
      "grad_norm": 0.5290525555610657,
      "learning_rate": 4.258264455201953e-06,
      "loss": 0.4663,
      "step": 395
    },
    {
      "epoch": 1.865203761755486,
      "grad_norm": 1.051685094833374,
      "learning_rate": 4.2534942387691335e-06,
      "loss": 0.4906,
      "step": 396
    },
    {
      "epoch": 1.8699059561128526,
      "grad_norm": 0.5118448734283447,
      "learning_rate": 4.248711422972829e-06,
      "loss": 0.477,
      "step": 397
    },
    {
      "epoch": 1.8746081504702194,
      "grad_norm": 0.6070737242698669,
      "learning_rate": 4.243916042178954e-06,
      "loss": 0.4602,
      "step": 398
    },
    {
      "epoch": 1.8793103448275863,
      "grad_norm": 0.827873170375824,
      "learning_rate": 4.239108130843709e-06,
      "loss": 0.4671,
      "step": 399
    },
    {
      "epoch": 1.884012539184953,
      "grad_norm": 0.5535209774971008,
      "learning_rate": 4.234287723513326e-06,
      "loss": 0.4933,
      "step": 400
    },
    {
      "epoch": 1.8887147335423198,
      "grad_norm": 0.5281742811203003,
      "learning_rate": 4.229454854823827e-06,
      "loss": 0.4904,
      "step": 401
    },
    {
      "epoch": 1.8934169278996866,
      "grad_norm": 0.5530508756637573,
      "learning_rate": 4.224609559500772e-06,
      "loss": 0.5034,
      "step": 402
    },
    {
      "epoch": 1.8981191222570533,
      "grad_norm": 0.47475168108940125,
      "learning_rate": 4.21975187235901e-06,
      "loss": 0.454,
      "step": 403
    },
    {
      "epoch": 1.90282131661442,
      "grad_norm": 0.5026959776878357,
      "learning_rate": 4.21488182830243e-06,
      "loss": 0.4833,
      "step": 404
    },
    {
      "epoch": 1.907523510971787,
      "grad_norm": 0.520931601524353,
      "learning_rate": 4.209999462323706e-06,
      "loss": 0.4587,
      "step": 405
    },
    {
      "epoch": 1.9122257053291536,
      "grad_norm": 0.584832489490509,
      "learning_rate": 4.20510480950405e-06,
      "loss": 0.4924,
      "step": 406
    },
    {
      "epoch": 1.9169278996865202,
      "grad_norm": 0.5310719609260559,
      "learning_rate": 4.200197905012961e-06,
      "loss": 0.4525,
      "step": 407
    },
    {
      "epoch": 1.9216300940438873,
      "grad_norm": 0.48304978013038635,
      "learning_rate": 4.195278784107965e-06,
      "loss": 0.4703,
      "step": 408
    },
    {
      "epoch": 1.926332288401254,
      "grad_norm": 0.4856540560722351,
      "learning_rate": 4.19034748213437e-06,
      "loss": 0.472,
      "step": 409
    },
    {
      "epoch": 1.9310344827586206,
      "grad_norm": 0.500950038433075,
      "learning_rate": 4.185404034525008e-06,
      "loss": 0.4641,
      "step": 410
    },
    {
      "epoch": 1.9357366771159876,
      "grad_norm": 0.5088744163513184,
      "learning_rate": 4.180448476799981e-06,
      "loss": 0.5019,
      "step": 411
    },
    {
      "epoch": 1.9404388714733543,
      "grad_norm": 0.5200552940368652,
      "learning_rate": 4.175480844566404e-06,
      "loss": 0.4744,
      "step": 412
    },
    {
      "epoch": 1.9451410658307209,
      "grad_norm": 0.47198665142059326,
      "learning_rate": 4.170501173518152e-06,
      "loss": 0.4678,
      "step": 413
    },
    {
      "epoch": 1.9498432601880877,
      "grad_norm": 0.4798610806465149,
      "learning_rate": 4.165509499435604e-06,
      "loss": 0.4668,
      "step": 414
    },
    {
      "epoch": 1.9545454545454546,
      "grad_norm": 0.48917219042778015,
      "learning_rate": 4.16050585818538e-06,
      "loss": 0.4811,
      "step": 415
    },
    {
      "epoch": 1.9592476489028212,
      "grad_norm": 0.7600466012954712,
      "learning_rate": 4.155490285720092e-06,
      "loss": 0.504,
      "step": 416
    },
    {
      "epoch": 1.963949843260188,
      "grad_norm": 0.48938053846359253,
      "learning_rate": 4.150462818078079e-06,
      "loss": 0.4918,
      "step": 417
    },
    {
      "epoch": 1.968652037617555,
      "grad_norm": 0.6591557264328003,
      "learning_rate": 4.145423491383153e-06,
      "loss": 0.4864,
      "step": 418
    },
    {
      "epoch": 1.9733542319749215,
      "grad_norm": 0.5030936598777771,
      "learning_rate": 4.14037234184433e-06,
      "loss": 0.5032,
      "step": 419
    },
    {
      "epoch": 1.9780564263322884,
      "grad_norm": 0.46507832407951355,
      "learning_rate": 4.135309405755583e-06,
      "loss": 0.4885,
      "step": 420
    },
    {
      "epoch": 1.9827586206896552,
      "grad_norm": 0.6155677437782288,
      "learning_rate": 4.130234719495574e-06,
      "loss": 0.4721,
      "step": 421
    },
    {
      "epoch": 1.9874608150470219,
      "grad_norm": 0.5396779179573059,
      "learning_rate": 4.125148319527391e-06,
      "loss": 0.4628,
      "step": 422
    },
    {
      "epoch": 1.9921630094043887,
      "grad_norm": 0.5095715522766113,
      "learning_rate": 4.1200502423982904e-06,
      "loss": 0.4831,
      "step": 423
    },
    {
      "epoch": 1.9968652037617556,
      "grad_norm": 0.5792081356048584,
      "learning_rate": 4.1149405247394295e-06,
      "loss": 0.4878,
      "step": 424
    },
    {
      "epoch": 2.0047021943573666,
      "grad_norm": 1.0187262296676636,
      "learning_rate": 4.10981920326561e-06,
      "loss": 0.913,
      "step": 425
    },
    {
      "epoch": 2.0094043887147337,
      "grad_norm": 0.4980756938457489,
      "learning_rate": 4.104686314775009e-06,
      "loss": 0.4405,
      "step": 426
    },
    {
      "epoch": 2.0141065830721003,
      "grad_norm": 0.49685895442962646,
      "learning_rate": 4.099541896148914e-06,
      "loss": 0.4496,
      "step": 427
    },
    {
      "epoch": 2.018808777429467,
      "grad_norm": 0.5676024556159973,
      "learning_rate": 4.094385984351462e-06,
      "loss": 0.444,
      "step": 428
    },
    {
      "epoch": 2.023510971786834,
      "grad_norm": 0.9495695233345032,
      "learning_rate": 4.0892186164293715e-06,
      "loss": 0.4645,
      "step": 429
    },
    {
      "epoch": 2.0282131661442007,
      "grad_norm": 0.5299291610717773,
      "learning_rate": 4.0840398295116745e-06,
      "loss": 0.4527,
      "step": 430
    },
    {
      "epoch": 2.0329153605015673,
      "grad_norm": 1.048509955406189,
      "learning_rate": 4.078849660809456e-06,
      "loss": 0.4487,
      "step": 431
    },
    {
      "epoch": 2.0376175548589344,
      "grad_norm": 0.5003053545951843,
      "learning_rate": 4.073648147615579e-06,
      "loss": 0.4304,
      "step": 432
    },
    {
      "epoch": 2.042319749216301,
      "grad_norm": 0.5424538254737854,
      "learning_rate": 4.068435327304421e-06,
      "loss": 0.457,
      "step": 433
    },
    {
      "epoch": 2.0470219435736676,
      "grad_norm": 0.467619925737381,
      "learning_rate": 4.063211237331603e-06,
      "loss": 0.4549,
      "step": 434
    },
    {
      "epoch": 2.0517241379310347,
      "grad_norm": 0.5217760801315308,
      "learning_rate": 4.057975915233725e-06,
      "loss": 0.4385,
      "step": 435
    },
    {
      "epoch": 2.0564263322884013,
      "grad_norm": 0.673935055732727,
      "learning_rate": 4.052729398628089e-06,
      "loss": 0.4657,
      "step": 436
    },
    {
      "epoch": 2.061128526645768,
      "grad_norm": 0.5437114834785461,
      "learning_rate": 4.047471725212437e-06,
      "loss": 0.4614,
      "step": 437
    },
    {
      "epoch": 2.0658307210031346,
      "grad_norm": 0.5773066282272339,
      "learning_rate": 4.042202932764673e-06,
      "loss": 0.4476,
      "step": 438
    },
    {
      "epoch": 2.0705329153605017,
      "grad_norm": 0.4945369362831116,
      "learning_rate": 4.036923059142595e-06,
      "loss": 0.4487,
      "step": 439
    },
    {
      "epoch": 2.0752351097178683,
      "grad_norm": 0.49815094470977783,
      "learning_rate": 4.031632142283623e-06,
      "loss": 0.4424,
      "step": 440
    },
    {
      "epoch": 2.079937304075235,
      "grad_norm": 0.5249453186988831,
      "learning_rate": 4.026330220204524e-06,
      "loss": 0.451,
      "step": 441
    },
    {
      "epoch": 2.084639498432602,
      "grad_norm": 0.4735757112503052,
      "learning_rate": 4.021017331001146e-06,
      "loss": 0.4413,
      "step": 442
    },
    {
      "epoch": 2.0893416927899686,
      "grad_norm": 0.484040766954422,
      "learning_rate": 4.015693512848131e-06,
      "loss": 0.4487,
      "step": 443
    },
    {
      "epoch": 2.0940438871473352,
      "grad_norm": 0.5543732047080994,
      "learning_rate": 4.0103588039986556e-06,
      "loss": 0.452,
      "step": 444
    },
    {
      "epoch": 2.0987460815047023,
      "grad_norm": 0.5293512940406799,
      "learning_rate": 4.005013242784146e-06,
      "loss": 0.4605,
      "step": 445
    },
    {
      "epoch": 2.103448275862069,
      "grad_norm": 0.5147669911384583,
      "learning_rate": 3.999656867614006e-06,
      "loss": 0.4479,
      "step": 446
    },
    {
      "epoch": 2.1081504702194356,
      "grad_norm": 0.47510066628456116,
      "learning_rate": 3.994289716975341e-06,
      "loss": 0.4629,
      "step": 447
    },
    {
      "epoch": 2.1128526645768027,
      "grad_norm": 0.5011350512504578,
      "learning_rate": 3.988911829432682e-06,
      "loss": 0.4487,
      "step": 448
    },
    {
      "epoch": 2.1175548589341693,
      "grad_norm": 0.5059776902198792,
      "learning_rate": 3.983523243627706e-06,
      "loss": 0.4456,
      "step": 449
    },
    {
      "epoch": 2.122257053291536,
      "grad_norm": 0.4974157512187958,
      "learning_rate": 3.978123998278962e-06,
      "loss": 0.436,
      "step": 450
    },
    {
      "epoch": 2.126959247648903,
      "grad_norm": 0.645297646522522,
      "learning_rate": 3.97271413218159e-06,
      "loss": 0.4653,
      "step": 451
    },
    {
      "epoch": 2.1316614420062696,
      "grad_norm": 0.46000078320503235,
      "learning_rate": 3.9672936842070425e-06,
      "loss": 0.4618,
      "step": 452
    },
    {
      "epoch": 2.1363636363636362,
      "grad_norm": 0.54063880443573,
      "learning_rate": 3.9618626933028086e-06,
      "loss": 0.4662,
      "step": 453
    },
    {
      "epoch": 2.1410658307210033,
      "grad_norm": 0.4782627820968628,
      "learning_rate": 3.956421198492128e-06,
      "loss": 0.448,
      "step": 454
    },
    {
      "epoch": 2.14576802507837,
      "grad_norm": 0.6332294344902039,
      "learning_rate": 3.950969238873714e-06,
      "loss": 0.447,
      "step": 455
    },
    {
      "epoch": 2.1504702194357366,
      "grad_norm": 0.49692678451538086,
      "learning_rate": 3.9455068536214765e-06,
      "loss": 0.4758,
      "step": 456
    },
    {
      "epoch": 2.1551724137931036,
      "grad_norm": 0.5347000360488892,
      "learning_rate": 3.9400340819842335e-06,
      "loss": 0.4565,
      "step": 457
    },
    {
      "epoch": 2.1598746081504703,
      "grad_norm": 0.5034692883491516,
      "learning_rate": 3.934550963285432e-06,
      "loss": 0.4565,
      "step": 458
    },
    {
      "epoch": 2.164576802507837,
      "grad_norm": 0.4939899444580078,
      "learning_rate": 3.9290575369228664e-06,
      "loss": 0.4284,
      "step": 459
    },
    {
      "epoch": 2.169278996865204,
      "grad_norm": 0.5838932394981384,
      "learning_rate": 3.923553842368396e-06,
      "loss": 0.4667,
      "step": 460
    },
    {
      "epoch": 2.1739811912225706,
      "grad_norm": 0.4959818124771118,
      "learning_rate": 3.918039919167658e-06,
      "loss": 0.4335,
      "step": 461
    },
    {
      "epoch": 2.1786833855799372,
      "grad_norm": 0.5034835934638977,
      "learning_rate": 3.912515806939786e-06,
      "loss": 0.448,
      "step": 462
    },
    {
      "epoch": 2.183385579937304,
      "grad_norm": 0.526311993598938,
      "learning_rate": 3.906981545377124e-06,
      "loss": 0.4597,
      "step": 463
    },
    {
      "epoch": 2.188087774294671,
      "grad_norm": 0.5168907046318054,
      "learning_rate": 3.901437174244943e-06,
      "loss": 0.4299,
      "step": 464
    },
    {
      "epoch": 2.1927899686520376,
      "grad_norm": 0.5047626495361328,
      "learning_rate": 3.895882733381154e-06,
      "loss": 0.4477,
      "step": 465
    },
    {
      "epoch": 2.197492163009404,
      "grad_norm": 0.541617214679718,
      "learning_rate": 3.890318262696023e-06,
      "loss": 0.4664,
      "step": 466
    },
    {
      "epoch": 2.2021943573667713,
      "grad_norm": 0.49817150831222534,
      "learning_rate": 3.8847438021718805e-06,
      "loss": 0.4417,
      "step": 467
    },
    {
      "epoch": 2.206896551724138,
      "grad_norm": 0.5194390416145325,
      "learning_rate": 3.879159391862839e-06,
      "loss": 0.4631,
      "step": 468
    },
    {
      "epoch": 2.2115987460815045,
      "grad_norm": 0.5216783285140991,
      "learning_rate": 3.873565071894503e-06,
      "loss": 0.4344,
      "step": 469
    },
    {
      "epoch": 2.2163009404388716,
      "grad_norm": 0.5285272002220154,
      "learning_rate": 3.86796088246368e-06,
      "loss": 0.4318,
      "step": 470
    },
    {
      "epoch": 2.2210031347962382,
      "grad_norm": 0.7032628059387207,
      "learning_rate": 3.8623468638380905e-06,
      "loss": 0.4147,
      "step": 471
    },
    {
      "epoch": 2.225705329153605,
      "grad_norm": 0.5304352045059204,
      "learning_rate": 3.856723056356085e-06,
      "loss": 0.4587,
      "step": 472
    },
    {
      "epoch": 2.230407523510972,
      "grad_norm": 0.5279494524002075,
      "learning_rate": 3.851089500426346e-06,
      "loss": 0.4456,
      "step": 473
    },
    {
      "epoch": 2.2351097178683386,
      "grad_norm": 0.5182958841323853,
      "learning_rate": 3.845446236527605e-06,
      "loss": 0.4449,
      "step": 474
    },
    {
      "epoch": 2.239811912225705,
      "grad_norm": 0.5063281059265137,
      "learning_rate": 3.8397933052083445e-06,
      "loss": 0.4623,
      "step": 475
    },
    {
      "epoch": 2.2445141065830723,
      "grad_norm": 0.5559848546981812,
      "learning_rate": 3.834130747086512e-06,
      "loss": 0.4473,
      "step": 476
    },
    {
      "epoch": 2.249216300940439,
      "grad_norm": 0.5276119709014893,
      "learning_rate": 3.828458602849226e-06,
      "loss": 0.4468,
      "step": 477
    },
    {
      "epoch": 2.2539184952978055,
      "grad_norm": 0.5662305951118469,
      "learning_rate": 3.822776913252485e-06,
      "loss": 0.4504,
      "step": 478
    },
    {
      "epoch": 2.2586206896551726,
      "grad_norm": 0.5402324199676514,
      "learning_rate": 3.817085719120872e-06,
      "loss": 0.4017,
      "step": 479
    },
    {
      "epoch": 2.2633228840125392,
      "grad_norm": 0.7308655977249146,
      "learning_rate": 3.811385061347263e-06,
      "loss": 0.444,
      "step": 480
    },
    {
      "epoch": 2.268025078369906,
      "grad_norm": 0.5518504977226257,
      "learning_rate": 3.805674980892535e-06,
      "loss": 0.4693,
      "step": 481
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.6556211113929749,
      "learning_rate": 3.7999555187852667e-06,
      "loss": 0.4567,
      "step": 482
    },
    {
      "epoch": 2.2774294670846396,
      "grad_norm": 0.5011089444160461,
      "learning_rate": 3.7942267161214497e-06,
      "loss": 0.4449,
      "step": 483
    },
    {
      "epoch": 2.282131661442006,
      "grad_norm": 0.5480395555496216,
      "learning_rate": 3.7884886140641884e-06,
      "loss": 0.4678,
      "step": 484
    },
    {
      "epoch": 2.2868338557993733,
      "grad_norm": 0.5292133092880249,
      "learning_rate": 3.7827412538434062e-06,
      "loss": 0.4323,
      "step": 485
    },
    {
      "epoch": 2.29153605015674,
      "grad_norm": 0.49213892221450806,
      "learning_rate": 3.7769846767555495e-06,
      "loss": 0.4591,
      "step": 486
    },
    {
      "epoch": 2.2962382445141065,
      "grad_norm": 0.5282753109931946,
      "learning_rate": 3.7712189241632898e-06,
      "loss": 0.4549,
      "step": 487
    },
    {
      "epoch": 2.300940438871473,
      "grad_norm": 0.5287954211235046,
      "learning_rate": 3.7654440374952288e-06,
      "loss": 0.4419,
      "step": 488
    },
    {
      "epoch": 2.30564263322884,
      "grad_norm": 0.48144471645355225,
      "learning_rate": 3.7596600582455976e-06,
      "loss": 0.4196,
      "step": 489
    },
    {
      "epoch": 2.310344827586207,
      "grad_norm": 0.5843647122383118,
      "learning_rate": 3.75386702797396e-06,
      "loss": 0.4198,
      "step": 490
    },
    {
      "epoch": 2.3150470219435735,
      "grad_norm": 0.4764229953289032,
      "learning_rate": 3.7480649883049164e-06,
      "loss": 0.4254,
      "step": 491
    },
    {
      "epoch": 2.3197492163009406,
      "grad_norm": 0.5592864155769348,
      "learning_rate": 3.7422539809277993e-06,
      "loss": 0.4364,
      "step": 492
    },
    {
      "epoch": 2.324451410658307,
      "grad_norm": 0.5710645318031311,
      "learning_rate": 3.736434047596379e-06,
      "loss": 0.4435,
      "step": 493
    },
    {
      "epoch": 2.329153605015674,
      "grad_norm": 0.5412624478340149,
      "learning_rate": 3.73060523012856e-06,
      "loss": 0.4504,
      "step": 494
    },
    {
      "epoch": 2.333855799373041,
      "grad_norm": 0.49665457010269165,
      "learning_rate": 3.724767570406082e-06,
      "loss": 0.4677,
      "step": 495
    },
    {
      "epoch": 2.3385579937304075,
      "grad_norm": 0.5091891884803772,
      "learning_rate": 3.7189211103742206e-06,
      "loss": 0.4249,
      "step": 496
    },
    {
      "epoch": 2.343260188087774,
      "grad_norm": 0.4907149076461792,
      "learning_rate": 3.7130658920414818e-06,
      "loss": 0.46,
      "step": 497
    },
    {
      "epoch": 2.347962382445141,
      "grad_norm": 0.4705374240875244,
      "learning_rate": 3.7072019574793034e-06,
      "loss": 0.4715,
      "step": 498
    },
    {
      "epoch": 2.352664576802508,
      "grad_norm": 0.5139763355255127,
      "learning_rate": 3.701329348821752e-06,
      "loss": 0.4523,
      "step": 499
    },
    {
      "epoch": 2.3573667711598745,
      "grad_norm": 0.5120819211006165,
      "learning_rate": 3.695448108265221e-06,
      "loss": 0.4373,
      "step": 500
    },
    {
      "epoch": 2.3620689655172415,
      "grad_norm": 0.45697054266929626,
      "learning_rate": 3.6895582780681254e-06,
      "loss": 0.4337,
      "step": 501
    },
    {
      "epoch": 2.366771159874608,
      "grad_norm": 0.48620718717575073,
      "learning_rate": 3.683659900550598e-06,
      "loss": 0.4618,
      "step": 502
    },
    {
      "epoch": 2.371473354231975,
      "grad_norm": 0.5233144760131836,
      "learning_rate": 3.6777530180941894e-06,
      "loss": 0.4441,
      "step": 503
    },
    {
      "epoch": 2.376175548589342,
      "grad_norm": 0.5168657898902893,
      "learning_rate": 3.671837673141559e-06,
      "loss": 0.4309,
      "step": 504
    },
    {
      "epoch": 2.3808777429467085,
      "grad_norm": 0.5887295007705688,
      "learning_rate": 3.6659139081961707e-06,
      "loss": 0.4469,
      "step": 505
    },
    {
      "epoch": 2.385579937304075,
      "grad_norm": 0.5009861588478088,
      "learning_rate": 3.6599817658219916e-06,
      "loss": 0.4487,
      "step": 506
    },
    {
      "epoch": 2.3902821316614418,
      "grad_norm": 0.49702706933021545,
      "learning_rate": 3.6540412886431796e-06,
      "loss": 0.4605,
      "step": 507
    },
    {
      "epoch": 2.394984326018809,
      "grad_norm": 0.4844914376735687,
      "learning_rate": 3.648092519343783e-06,
      "loss": 0.4435,
      "step": 508
    },
    {
      "epoch": 2.3996865203761755,
      "grad_norm": 0.5124401450157166,
      "learning_rate": 3.642135500667431e-06,
      "loss": 0.4312,
      "step": 509
    },
    {
      "epoch": 2.4043887147335425,
      "grad_norm": 0.5017268657684326,
      "learning_rate": 3.6361702754170247e-06,
      "loss": 0.4465,
      "step": 510
    },
    {
      "epoch": 2.409090909090909,
      "grad_norm": 0.51381915807724,
      "learning_rate": 3.630196886454435e-06,
      "loss": 0.4288,
      "step": 511
    },
    {
      "epoch": 2.413793103448276,
      "grad_norm": 0.5275219082832336,
      "learning_rate": 3.62421537670019e-06,
      "loss": 0.4431,
      "step": 512
    },
    {
      "epoch": 2.4184952978056424,
      "grad_norm": 0.5186277627944946,
      "learning_rate": 3.618225789133167e-06,
      "loss": 0.4465,
      "step": 513
    },
    {
      "epoch": 2.4231974921630095,
      "grad_norm": 0.7465965151786804,
      "learning_rate": 3.612228166790287e-06,
      "loss": 0.445,
      "step": 514
    },
    {
      "epoch": 2.427899686520376,
      "grad_norm": 0.5389880537986755,
      "learning_rate": 3.606222552766201e-06,
      "loss": 0.4548,
      "step": 515
    },
    {
      "epoch": 2.4326018808777428,
      "grad_norm": 0.8773667216300964,
      "learning_rate": 3.6002089902129844e-06,
      "loss": 0.4482,
      "step": 516
    },
    {
      "epoch": 2.43730407523511,
      "grad_norm": 0.4819961190223694,
      "learning_rate": 3.5941875223398225e-06,
      "loss": 0.4387,
      "step": 517
    },
    {
      "epoch": 2.4420062695924765,
      "grad_norm": 0.5448062419891357,
      "learning_rate": 3.588158192412707e-06,
      "loss": 0.4347,
      "step": 518
    },
    {
      "epoch": 2.446708463949843,
      "grad_norm": 4.079322338104248,
      "learning_rate": 3.582121043754116e-06,
      "loss": 0.4383,
      "step": 519
    },
    {
      "epoch": 2.45141065830721,
      "grad_norm": 0.9761962890625,
      "learning_rate": 3.5760761197427097e-06,
      "loss": 0.4377,
      "step": 520
    },
    {
      "epoch": 2.456112852664577,
      "grad_norm": 0.5383471846580505,
      "learning_rate": 3.570023463813017e-06,
      "loss": 0.43,
      "step": 521
    },
    {
      "epoch": 2.4608150470219434,
      "grad_norm": 0.47819817066192627,
      "learning_rate": 3.5639631194551216e-06,
      "loss": 0.4526,
      "step": 522
    },
    {
      "epoch": 2.4655172413793105,
      "grad_norm": 0.5026260018348694,
      "learning_rate": 3.557895130214352e-06,
      "loss": 0.4445,
      "step": 523
    },
    {
      "epoch": 2.470219435736677,
      "grad_norm": 0.4826788008213043,
      "learning_rate": 3.5518195396909653e-06,
      "loss": 0.4634,
      "step": 524
    },
    {
      "epoch": 2.4749216300940438,
      "grad_norm": 0.5463977456092834,
      "learning_rate": 3.5457363915398384e-06,
      "loss": 0.4526,
      "step": 525
    },
    {
      "epoch": 2.479623824451411,
      "grad_norm": 0.5236705541610718,
      "learning_rate": 3.539645729470151e-06,
      "loss": 0.4383,
      "step": 526
    },
    {
      "epoch": 2.4843260188087775,
      "grad_norm": 0.5190213322639465,
      "learning_rate": 3.5335475972450715e-06,
      "loss": 0.4378,
      "step": 527
    },
    {
      "epoch": 2.489028213166144,
      "grad_norm": 0.7202126383781433,
      "learning_rate": 3.5274420386814458e-06,
      "loss": 0.4639,
      "step": 528
    },
    {
      "epoch": 2.493730407523511,
      "grad_norm": 0.6717228293418884,
      "learning_rate": 3.521329097649478e-06,
      "loss": 0.4458,
      "step": 529
    },
    {
      "epoch": 2.498432601880878,
      "grad_norm": 0.49824005365371704,
      "learning_rate": 3.515208818072418e-06,
      "loss": 0.4429,
      "step": 530
    },
    {
      "epoch": 2.5031347962382444,
      "grad_norm": 0.5616832971572876,
      "learning_rate": 3.509081243926247e-06,
      "loss": 0.4306,
      "step": 531
    },
    {
      "epoch": 2.507836990595611,
      "grad_norm": 0.5087556838989258,
      "learning_rate": 3.5029464192393557e-06,
      "loss": 0.4629,
      "step": 532
    },
    {
      "epoch": 2.512539184952978,
      "grad_norm": 0.5290883183479309,
      "learning_rate": 3.4968043880922363e-06,
      "loss": 0.414,
      "step": 533
    },
    {
      "epoch": 2.5172413793103448,
      "grad_norm": 0.5456948280334473,
      "learning_rate": 3.4906551946171603e-06,
      "loss": 0.4238,
      "step": 534
    },
    {
      "epoch": 2.521943573667712,
      "grad_norm": 0.49592944979667664,
      "learning_rate": 3.484498882997861e-06,
      "loss": 0.4224,
      "step": 535
    },
    {
      "epoch": 2.5266457680250785,
      "grad_norm": 0.4961735010147095,
      "learning_rate": 3.478335497469219e-06,
      "loss": 0.4507,
      "step": 536
    },
    {
      "epoch": 2.531347962382445,
      "grad_norm": 0.5094319581985474,
      "learning_rate": 3.472165082316943e-06,
      "loss": 0.4495,
      "step": 537
    },
    {
      "epoch": 2.5360501567398117,
      "grad_norm": 0.620319664478302,
      "learning_rate": 3.465987681877251e-06,
      "loss": 0.4421,
      "step": 538
    },
    {
      "epoch": 2.540752351097179,
      "grad_norm": 0.5389400720596313,
      "learning_rate": 3.4598033405365527e-06,
      "loss": 0.4547,
      "step": 539
    },
    {
      "epoch": 2.5454545454545454,
      "grad_norm": 0.7147080898284912,
      "learning_rate": 3.45361210273113e-06,
      "loss": 0.4478,
      "step": 540
    },
    {
      "epoch": 2.5501567398119125,
      "grad_norm": 0.5107600092887878,
      "learning_rate": 3.447414012946818e-06,
      "loss": 0.4349,
      "step": 541
    },
    {
      "epoch": 2.554858934169279,
      "grad_norm": 0.5057627558708191,
      "learning_rate": 3.4412091157186853e-06,
      "loss": 0.4502,
      "step": 542
    },
    {
      "epoch": 2.5595611285266457,
      "grad_norm": 0.5278921127319336,
      "learning_rate": 3.4349974556307146e-06,
      "loss": 0.4415,
      "step": 543
    },
    {
      "epoch": 2.5642633228840124,
      "grad_norm": 0.5077342987060547,
      "learning_rate": 3.4287790773154807e-06,
      "loss": 0.465,
      "step": 544
    },
    {
      "epoch": 2.5689655172413794,
      "grad_norm": 0.5148009061813354,
      "learning_rate": 3.4225540254538297e-06,
      "loss": 0.4613,
      "step": 545
    },
    {
      "epoch": 2.573667711598746,
      "grad_norm": 0.6661203503608704,
      "learning_rate": 3.416322344774562e-06,
      "loss": 0.4412,
      "step": 546
    },
    {
      "epoch": 2.5783699059561127,
      "grad_norm": 0.5129944086074829,
      "learning_rate": 3.4100840800541055e-06,
      "loss": 0.463,
      "step": 547
    },
    {
      "epoch": 2.58307210031348,
      "grad_norm": 1.101260781288147,
      "learning_rate": 3.4038392761161986e-06,
      "loss": 0.4528,
      "step": 548
    },
    {
      "epoch": 2.5877742946708464,
      "grad_norm": 0.9880437254905701,
      "learning_rate": 3.3975879778315634e-06,
      "loss": 0.4229,
      "step": 549
    },
    {
      "epoch": 2.592476489028213,
      "grad_norm": 0.4811207354068756,
      "learning_rate": 3.391330230117587e-06,
      "loss": 0.3961,
      "step": 550
    },
    {
      "epoch": 2.5971786833855797,
      "grad_norm": 0.5389580726623535,
      "learning_rate": 3.385066077937997e-06,
      "loss": 0.4654,
      "step": 551
    },
    {
      "epoch": 2.6018808777429467,
      "grad_norm": 0.5345444679260254,
      "learning_rate": 3.378795566302541e-06,
      "loss": 0.4249,
      "step": 552
    },
    {
      "epoch": 2.6065830721003134,
      "grad_norm": 0.5955337285995483,
      "learning_rate": 3.372518740266658e-06,
      "loss": 0.4436,
      "step": 553
    },
    {
      "epoch": 2.6112852664576804,
      "grad_norm": 0.5519671440124512,
      "learning_rate": 3.36623564493116e-06,
      "loss": 0.4559,
      "step": 554
    },
    {
      "epoch": 2.615987460815047,
      "grad_norm": 0.4986394941806793,
      "learning_rate": 3.3599463254419047e-06,
      "loss": 0.4598,
      "step": 555
    },
    {
      "epoch": 2.6206896551724137,
      "grad_norm": 0.5792630314826965,
      "learning_rate": 3.3536508269894724e-06,
      "loss": 0.4672,
      "step": 556
    },
    {
      "epoch": 2.6253918495297803,
      "grad_norm": 0.5608819127082825,
      "learning_rate": 3.347349194808842e-06,
      "loss": 0.454,
      "step": 557
    },
    {
      "epoch": 2.6300940438871474,
      "grad_norm": 0.5010476112365723,
      "learning_rate": 3.3410414741790625e-06,
      "loss": 0.4288,
      "step": 558
    },
    {
      "epoch": 2.634796238244514,
      "grad_norm": 0.5262719392776489,
      "learning_rate": 3.3347277104229332e-06,
      "loss": 0.4436,
      "step": 559
    },
    {
      "epoch": 2.639498432601881,
      "grad_norm": 0.5418802499771118,
      "learning_rate": 3.3284079489066728e-06,
      "loss": 0.4382,
      "step": 560
    },
    {
      "epoch": 2.6442006269592477,
      "grad_norm": 0.5096928477287292,
      "learning_rate": 3.3220822350395966e-06,
      "loss": 0.43,
      "step": 561
    },
    {
      "epoch": 2.6489028213166144,
      "grad_norm": 0.5615900158882141,
      "learning_rate": 3.31575061427379e-06,
      "loss": 0.4312,
      "step": 562
    },
    {
      "epoch": 2.653605015673981,
      "grad_norm": 0.5129327774047852,
      "learning_rate": 3.3094131321037783e-06,
      "loss": 0.4378,
      "step": 563
    },
    {
      "epoch": 2.658307210031348,
      "grad_norm": 0.4975152313709259,
      "learning_rate": 3.303069834066206e-06,
      "loss": 0.4331,
      "step": 564
    },
    {
      "epoch": 2.6630094043887147,
      "grad_norm": 0.4807284474372864,
      "learning_rate": 3.2967207657395055e-06,
      "loss": 0.4745,
      "step": 565
    },
    {
      "epoch": 2.6677115987460818,
      "grad_norm": 0.5497413873672485,
      "learning_rate": 3.2903659727435692e-06,
      "loss": 0.4618,
      "step": 566
    },
    {
      "epoch": 2.6724137931034484,
      "grad_norm": 0.6683759689331055,
      "learning_rate": 3.284005500739423e-06,
      "loss": 0.4575,
      "step": 567
    },
    {
      "epoch": 2.677115987460815,
      "grad_norm": 0.497938334941864,
      "learning_rate": 3.2776393954289e-06,
      "loss": 0.4293,
      "step": 568
    },
    {
      "epoch": 2.6818181818181817,
      "grad_norm": 0.6120790839195251,
      "learning_rate": 3.271267702554307e-06,
      "loss": 0.4202,
      "step": 569
    },
    {
      "epoch": 2.6865203761755487,
      "grad_norm": 0.5267842411994934,
      "learning_rate": 3.2648904678981032e-06,
      "loss": 0.4643,
      "step": 570
    },
    {
      "epoch": 2.6912225705329154,
      "grad_norm": 0.6027296185493469,
      "learning_rate": 3.2585077372825636e-06,
      "loss": 0.4127,
      "step": 571
    },
    {
      "epoch": 2.695924764890282,
      "grad_norm": 0.501907229423523,
      "learning_rate": 3.2521195565694543e-06,
      "loss": 0.4444,
      "step": 572
    },
    {
      "epoch": 2.700626959247649,
      "grad_norm": 0.5379273891448975,
      "learning_rate": 3.2457259716597023e-06,
      "loss": 0.4458,
      "step": 573
    },
    {
      "epoch": 2.7053291536050157,
      "grad_norm": 0.4921974241733551,
      "learning_rate": 3.2393270284930658e-06,
      "loss": 0.4538,
      "step": 574
    },
    {
      "epoch": 2.7100313479623823,
      "grad_norm": 0.5309234261512756,
      "learning_rate": 3.2329227730478026e-06,
      "loss": 0.4302,
      "step": 575
    },
    {
      "epoch": 2.714733542319749,
      "grad_norm": 0.5694313049316406,
      "learning_rate": 3.2265132513403415e-06,
      "loss": 0.42,
      "step": 576
    },
    {
      "epoch": 2.719435736677116,
      "grad_norm": 0.5232118964195251,
      "learning_rate": 3.22009850942495e-06,
      "loss": 0.4451,
      "step": 577
    },
    {
      "epoch": 2.7241379310344827,
      "grad_norm": 0.4894232749938965,
      "learning_rate": 3.213678593393405e-06,
      "loss": 0.4603,
      "step": 578
    },
    {
      "epoch": 2.7288401253918497,
      "grad_norm": 0.5060969591140747,
      "learning_rate": 3.207253549374662e-06,
      "loss": 0.4565,
      "step": 579
    },
    {
      "epoch": 2.7335423197492164,
      "grad_norm": 0.5979425311088562,
      "learning_rate": 3.200823423534519e-06,
      "loss": 0.4438,
      "step": 580
    },
    {
      "epoch": 2.738244514106583,
      "grad_norm": 0.5185328125953674,
      "learning_rate": 3.194388262075293e-06,
      "loss": 0.4475,
      "step": 581
    },
    {
      "epoch": 2.7429467084639496,
      "grad_norm": 0.4780375361442566,
      "learning_rate": 3.1879481112354804e-06,
      "loss": 0.449,
      "step": 582
    },
    {
      "epoch": 2.7476489028213167,
      "grad_norm": 0.5120851993560791,
      "learning_rate": 3.181503017289428e-06,
      "loss": 0.4108,
      "step": 583
    },
    {
      "epoch": 2.7523510971786833,
      "grad_norm": 0.5004064440727234,
      "learning_rate": 3.175053026547002e-06,
      "loss": 0.4154,
      "step": 584
    },
    {
      "epoch": 2.7570532915360504,
      "grad_norm": 0.5117692351341248,
      "learning_rate": 3.16859818535325e-06,
      "loss": 0.4577,
      "step": 585
    },
    {
      "epoch": 2.761755485893417,
      "grad_norm": 0.48328641057014465,
      "learning_rate": 3.1621385400880756e-06,
      "loss": 0.4523,
      "step": 586
    },
    {
      "epoch": 2.7664576802507836,
      "grad_norm": 0.6039829850196838,
      "learning_rate": 3.1556741371658984e-06,
      "loss": 0.4549,
      "step": 587
    },
    {
      "epoch": 2.7711598746081503,
      "grad_norm": 0.5623074769973755,
      "learning_rate": 3.1492050230353238e-06,
      "loss": 0.4461,
      "step": 588
    },
    {
      "epoch": 2.7758620689655173,
      "grad_norm": 0.5452325344085693,
      "learning_rate": 3.142731244178809e-06,
      "loss": 0.4202,
      "step": 589
    },
    {
      "epoch": 2.780564263322884,
      "grad_norm": 0.5552520155906677,
      "learning_rate": 3.1362528471123277e-06,
      "loss": 0.4033,
      "step": 590
    },
    {
      "epoch": 2.785266457680251,
      "grad_norm": 0.566600501537323,
      "learning_rate": 3.129769878385039e-06,
      "loss": 0.4086,
      "step": 591
    },
    {
      "epoch": 2.7899686520376177,
      "grad_norm": 0.4819733798503876,
      "learning_rate": 3.1232823845789473e-06,
      "loss": 0.4511,
      "step": 592
    },
    {
      "epoch": 2.7946708463949843,
      "grad_norm": 0.7382413744926453,
      "learning_rate": 3.1167904123085736e-06,
      "loss": 0.4563,
      "step": 593
    },
    {
      "epoch": 2.799373040752351,
      "grad_norm": 0.5518636703491211,
      "learning_rate": 3.110294008220617e-06,
      "loss": 0.4437,
      "step": 594
    },
    {
      "epoch": 2.804075235109718,
      "grad_norm": 0.48016345500946045,
      "learning_rate": 3.1037932189936205e-06,
      "loss": 0.4408,
      "step": 595
    },
    {
      "epoch": 2.8087774294670846,
      "grad_norm": 0.5128410458564758,
      "learning_rate": 3.097288091337635e-06,
      "loss": 0.4359,
      "step": 596
    },
    {
      "epoch": 2.8134796238244513,
      "grad_norm": 0.5333647727966309,
      "learning_rate": 3.0907786719938876e-06,
      "loss": 0.4251,
      "step": 597
    },
    {
      "epoch": 2.8181818181818183,
      "grad_norm": 0.46841663122177124,
      "learning_rate": 3.084265007734436e-06,
      "loss": 0.4357,
      "step": 598
    },
    {
      "epoch": 2.822884012539185,
      "grad_norm": 0.5592591762542725,
      "learning_rate": 3.0777471453618457e-06,
      "loss": 0.4596,
      "step": 599
    },
    {
      "epoch": 2.8275862068965516,
      "grad_norm": 0.5454399585723877,
      "learning_rate": 3.0712251317088426e-06,
      "loss": 0.4322,
      "step": 600
    },
    {
      "epoch": 2.8322884012539182,
      "grad_norm": 0.5095115303993225,
      "learning_rate": 3.064699013637983e-06,
      "loss": 0.451,
      "step": 601
    },
    {
      "epoch": 2.8369905956112853,
      "grad_norm": 0.56089848279953,
      "learning_rate": 3.0581688380413115e-06,
      "loss": 0.4352,
      "step": 602
    },
    {
      "epoch": 2.841692789968652,
      "grad_norm": 0.5011700391769409,
      "learning_rate": 3.0516346518400315e-06,
      "loss": 0.4521,
      "step": 603
    },
    {
      "epoch": 2.846394984326019,
      "grad_norm": 0.5560495853424072,
      "learning_rate": 3.0450965019841593e-06,
      "loss": 0.4522,
      "step": 604
    },
    {
      "epoch": 2.8510971786833856,
      "grad_norm": 0.5144981145858765,
      "learning_rate": 3.0385544354521957e-06,
      "loss": 0.4146,
      "step": 605
    },
    {
      "epoch": 2.8557993730407523,
      "grad_norm": 0.5166004300117493,
      "learning_rate": 3.0320084992507814e-06,
      "loss": 0.4423,
      "step": 606
    },
    {
      "epoch": 2.860501567398119,
      "grad_norm": 0.5098627805709839,
      "learning_rate": 3.0254587404143604e-06,
      "loss": 0.4793,
      "step": 607
    },
    {
      "epoch": 2.865203761755486,
      "grad_norm": 0.5003807544708252,
      "learning_rate": 3.0189052060048464e-06,
      "loss": 0.4409,
      "step": 608
    },
    {
      "epoch": 2.8699059561128526,
      "grad_norm": 0.5339270830154419,
      "learning_rate": 3.01234794311128e-06,
      "loss": 0.4376,
      "step": 609
    },
    {
      "epoch": 2.8746081504702197,
      "grad_norm": 0.5303454995155334,
      "learning_rate": 3.0057869988494925e-06,
      "loss": 0.4626,
      "step": 610
    },
    {
      "epoch": 2.8793103448275863,
      "grad_norm": 0.5876331329345703,
      "learning_rate": 2.999222420361767e-06,
      "loss": 0.4532,
      "step": 611
    },
    {
      "epoch": 2.884012539184953,
      "grad_norm": 0.5893226861953735,
      "learning_rate": 2.9926542548165e-06,
      "loss": 0.4654,
      "step": 612
    },
    {
      "epoch": 2.8887147335423196,
      "grad_norm": 0.5249570608139038,
      "learning_rate": 2.9860825494078605e-06,
      "loss": 0.4363,
      "step": 613
    },
    {
      "epoch": 2.8934169278996866,
      "grad_norm": 0.5067557692527771,
      "learning_rate": 2.979507351355454e-06,
      "loss": 0.4532,
      "step": 614
    },
    {
      "epoch": 2.8981191222570533,
      "grad_norm": 0.6186209917068481,
      "learning_rate": 2.972928707903981e-06,
      "loss": 0.4417,
      "step": 615
    },
    {
      "epoch": 2.9028213166144203,
      "grad_norm": 0.5927693843841553,
      "learning_rate": 2.966346666322898e-06,
      "loss": 0.4379,
      "step": 616
    },
    {
      "epoch": 2.907523510971787,
      "grad_norm": 0.6109799742698669,
      "learning_rate": 2.9597612739060775e-06,
      "loss": 0.4181,
      "step": 617
    },
    {
      "epoch": 2.9122257053291536,
      "grad_norm": 0.5325103998184204,
      "learning_rate": 2.9531725779714713e-06,
      "loss": 0.4494,
      "step": 618
    },
    {
      "epoch": 2.91692789968652,
      "grad_norm": 0.5224953293800354,
      "learning_rate": 2.9465806258607653e-06,
      "loss": 0.4558,
      "step": 619
    },
    {
      "epoch": 2.9216300940438873,
      "grad_norm": 0.5702230334281921,
      "learning_rate": 2.939985464939043e-06,
      "loss": 0.4587,
      "step": 620
    },
    {
      "epoch": 2.926332288401254,
      "grad_norm": 0.549354612827301,
      "learning_rate": 2.9333871425944434e-06,
      "loss": 0.4047,
      "step": 621
    },
    {
      "epoch": 2.9310344827586206,
      "grad_norm": 0.5000407099723816,
      "learning_rate": 2.926785706237822e-06,
      "loss": 0.4338,
      "step": 622
    },
    {
      "epoch": 2.9357366771159876,
      "grad_norm": 0.4803110361099243,
      "learning_rate": 2.920181203302409e-06,
      "loss": 0.4252,
      "step": 623
    },
    {
      "epoch": 2.9404388714733543,
      "grad_norm": 0.519071102142334,
      "learning_rate": 2.91357368124347e-06,
      "loss": 0.4226,
      "step": 624
    },
    {
      "epoch": 2.945141065830721,
      "grad_norm": 0.5117707252502441,
      "learning_rate": 2.906963187537962e-06,
      "loss": 0.4343,
      "step": 625
    },
    {
      "epoch": 2.9498432601880875,
      "grad_norm": 0.5518343448638916,
      "learning_rate": 2.9003497696841955e-06,
      "loss": 0.4136,
      "step": 626
    },
    {
      "epoch": 2.9545454545454546,
      "grad_norm": 0.4980529546737671,
      "learning_rate": 2.8937334752014913e-06,
      "loss": 0.4662,
      "step": 627
    },
    {
      "epoch": 2.959247648902821,
      "grad_norm": 0.5124077796936035,
      "learning_rate": 2.887114351629839e-06,
      "loss": 0.4424,
      "step": 628
    },
    {
      "epoch": 2.9639498432601883,
      "grad_norm": 0.5710898041725159,
      "learning_rate": 2.8804924465295575e-06,
      "loss": 0.4652,
      "step": 629
    },
    {
      "epoch": 2.968652037617555,
      "grad_norm": 0.5048180818557739,
      "learning_rate": 2.873867807480951e-06,
      "loss": 0.4725,
      "step": 630
    },
    {
      "epoch": 2.9733542319749215,
      "grad_norm": 0.5638993382453918,
      "learning_rate": 2.8672404820839676e-06,
      "loss": 0.4399,
      "step": 631
    },
    {
      "epoch": 2.978056426332288,
      "grad_norm": 0.593982458114624,
      "learning_rate": 2.8606105179578584e-06,
      "loss": 0.4479,
      "step": 632
    },
    {
      "epoch": 2.9827586206896552,
      "grad_norm": 0.5415588617324829,
      "learning_rate": 2.8539779627408332e-06,
      "loss": 0.4258,
      "step": 633
    },
    {
      "epoch": 2.987460815047022,
      "grad_norm": 0.6535062789916992,
      "learning_rate": 2.847342864089721e-06,
      "loss": 0.4433,
      "step": 634
    },
    {
      "epoch": 2.992163009404389,
      "grad_norm": 0.4936605989933014,
      "learning_rate": 2.8407052696796255e-06,
      "loss": 0.4299,
      "step": 635
    },
    {
      "epoch": 2.9968652037617556,
      "grad_norm": 0.5490683913230896,
      "learning_rate": 2.834065227203584e-06,
      "loss": 0.4488,
      "step": 636
    },
    {
      "epoch": 3.0047021943573666,
      "grad_norm": 0.5618948936462402,
      "learning_rate": 2.8274227843722213e-06,
      "loss": 0.866,
      "step": 637
    },
    {
      "epoch": 3.0094043887147337,
      "grad_norm": 0.5641200542449951,
      "learning_rate": 2.820777988913412e-06,
      "loss": 0.4171,
      "step": 638
    },
    {
      "epoch": 3.0141065830721003,
      "grad_norm": 0.5129956603050232,
      "learning_rate": 2.8141308885719337e-06,
      "loss": 0.4164,
      "step": 639
    },
    {
      "epoch": 3.018808777429467,
      "grad_norm": 0.587412416934967,
      "learning_rate": 2.8074815311091265e-06,
      "loss": 0.3903,
      "step": 640
    },
    {
      "epoch": 3.023510971786834,
      "grad_norm": 0.5251805782318115,
      "learning_rate": 2.8008299643025477e-06,
      "loss": 0.4326,
      "step": 641
    },
    {
      "epoch": 3.0282131661442007,
      "grad_norm": 0.5621777772903442,
      "learning_rate": 2.7941762359456294e-06,
      "loss": 0.4258,
      "step": 642
    },
    {
      "epoch": 3.0329153605015673,
      "grad_norm": 0.5080258846282959,
      "learning_rate": 2.787520393847334e-06,
      "loss": 0.4163,
      "step": 643
    },
    {
      "epoch": 3.0376175548589344,
      "grad_norm": 0.6336125731468201,
      "learning_rate": 2.780862485831814e-06,
      "loss": 0.4289,
      "step": 644
    },
    {
      "epoch": 3.042319749216301,
      "grad_norm": 0.5735598802566528,
      "learning_rate": 2.7742025597380644e-06,
      "loss": 0.4306,
      "step": 645
    },
    {
      "epoch": 3.0470219435736676,
      "grad_norm": 0.5653077960014343,
      "learning_rate": 2.7675406634195824e-06,
      "loss": 0.422,
      "step": 646
    },
    {
      "epoch": 3.0517241379310347,
      "grad_norm": 0.5047882199287415,
      "learning_rate": 2.7608768447440193e-06,
      "loss": 0.4076,
      "step": 647
    },
    {
      "epoch": 3.0564263322884013,
      "grad_norm": 1.2697781324386597,
      "learning_rate": 2.754211151592841e-06,
      "loss": 0.4115,
      "step": 648
    },
    {
      "epoch": 3.061128526645768,
      "grad_norm": 0.6861006617546082,
      "learning_rate": 2.7475436318609827e-06,
      "loss": 0.4309,
      "step": 649
    },
    {
      "epoch": 3.0658307210031346,
      "grad_norm": 0.5912951827049255,
      "learning_rate": 2.7408743334565006e-06,
      "loss": 0.4071,
      "step": 650
    },
    {
      "epoch": 3.0705329153605017,
      "grad_norm": 0.5096245408058167,
      "learning_rate": 2.734203304300235e-06,
      "loss": 0.4001,
      "step": 651
    },
    {
      "epoch": 3.0752351097178683,
      "grad_norm": 0.5158068537712097,
      "learning_rate": 2.7275305923254607e-06,
      "loss": 0.4115,
      "step": 652
    },
    {
      "epoch": 3.079937304075235,
      "grad_norm": 0.5201400518417358,
      "learning_rate": 2.720856245477544e-06,
      "loss": 0.4024,
      "step": 653
    },
    {
      "epoch": 3.084639498432602,
      "grad_norm": 0.5423036813735962,
      "learning_rate": 2.7141803117135978e-06,
      "loss": 0.3963,
      "step": 654
    },
    {
      "epoch": 3.0893416927899686,
      "grad_norm": 0.6557552218437195,
      "learning_rate": 2.7075028390021385e-06,
      "loss": 0.4,
      "step": 655
    },
    {
      "epoch": 3.0940438871473352,
      "grad_norm": 0.5141010880470276,
      "learning_rate": 2.7008238753227385e-06,
      "loss": 0.4056,
      "step": 656
    },
    {
      "epoch": 3.0987460815047023,
      "grad_norm": 0.504351794719696,
      "learning_rate": 2.694143468665685e-06,
      "loss": 0.4271,
      "step": 657
    },
    {
      "epoch": 3.103448275862069,
      "grad_norm": 0.5583532452583313,
      "learning_rate": 2.6874616670316338e-06,
      "loss": 0.3936,
      "step": 658
    },
    {
      "epoch": 3.1081504702194356,
      "grad_norm": 0.5661462545394897,
      "learning_rate": 2.6807785184312618e-06,
      "loss": 0.4136,
      "step": 659
    },
    {
      "epoch": 3.1128526645768027,
      "grad_norm": 0.5078592300415039,
      "learning_rate": 2.674094070884926e-06,
      "loss": 0.4138,
      "step": 660
    },
    {
      "epoch": 3.1175548589341693,
      "grad_norm": 0.579448401927948,
      "learning_rate": 2.6674083724223166e-06,
      "loss": 0.4333,
      "step": 661
    },
    {
      "epoch": 3.122257053291536,
      "grad_norm": 0.5355094075202942,
      "learning_rate": 2.6607214710821112e-06,
      "loss": 0.4041,
      "step": 662
    },
    {
      "epoch": 3.126959247648903,
      "grad_norm": 0.5290988683700562,
      "learning_rate": 2.6540334149116304e-06,
      "loss": 0.4181,
      "step": 663
    },
    {
      "epoch": 3.1316614420062696,
      "grad_norm": 0.5365299582481384,
      "learning_rate": 2.647344251966493e-06,
      "loss": 0.4152,
      "step": 664
    },
    {
      "epoch": 3.1363636363636362,
      "grad_norm": 0.5370200872421265,
      "learning_rate": 2.6406540303102714e-06,
      "loss": 0.4149,
      "step": 665
    },
    {
      "epoch": 3.1410658307210033,
      "grad_norm": 0.5290679335594177,
      "learning_rate": 2.6339627980141425e-06,
      "loss": 0.4172,
      "step": 666
    },
    {
      "epoch": 3.14576802507837,
      "grad_norm": 0.5143046379089355,
      "learning_rate": 2.6272706031565482e-06,
      "loss": 0.4021,
      "step": 667
    },
    {
      "epoch": 3.1504702194357366,
      "grad_norm": 0.6136928796768188,
      "learning_rate": 2.6205774938228433e-06,
      "loss": 0.3976,
      "step": 668
    },
    {
      "epoch": 3.1551724137931036,
      "grad_norm": 0.5212892889976501,
      "learning_rate": 2.6138835181049556e-06,
      "loss": 0.4225,
      "step": 669
    },
    {
      "epoch": 3.1598746081504703,
      "grad_norm": 0.5384306907653809,
      "learning_rate": 2.6071887241010374e-06,
      "loss": 0.405,
      "step": 670
    },
    {
      "epoch": 3.164576802507837,
      "grad_norm": 0.56938236951828,
      "learning_rate": 2.6004931599151223e-06,
      "loss": 0.3928,
      "step": 671
    },
    {
      "epoch": 3.169278996865204,
      "grad_norm": 0.5221254825592041,
      "learning_rate": 2.593796873656775e-06,
      "loss": 0.4348,
      "step": 672
    },
    {
      "epoch": 3.1739811912225706,
      "grad_norm": 0.5515516996383667,
      "learning_rate": 2.587099913440749e-06,
      "loss": 0.4141,
      "step": 673
    },
    {
      "epoch": 3.1786833855799372,
      "grad_norm": 0.5437397360801697,
      "learning_rate": 2.580402327386643e-06,
      "loss": 0.4026,
      "step": 674
    },
    {
      "epoch": 3.183385579937304,
      "grad_norm": 0.637416660785675,
      "learning_rate": 2.5737041636185496e-06,
      "loss": 0.4088,
      "step": 675
    },
    {
      "epoch": 3.188087774294671,
      "grad_norm": 0.5309600830078125,
      "learning_rate": 2.5670054702647146e-06,
      "loss": 0.3933,
      "step": 676
    },
    {
      "epoch": 3.1927899686520376,
      "grad_norm": 0.5138253569602966,
      "learning_rate": 2.5603062954571872e-06,
      "loss": 0.4172,
      "step": 677
    },
    {
      "epoch": 3.197492163009404,
      "grad_norm": 0.5631712675094604,
      "learning_rate": 2.553606687331477e-06,
      "loss": 0.4376,
      "step": 678
    },
    {
      "epoch": 3.2021943573667713,
      "grad_norm": 1.246191382408142,
      "learning_rate": 2.5469066940262073e-06,
      "loss": 0.4072,
      "step": 679
    },
    {
      "epoch": 3.206896551724138,
      "grad_norm": 0.5201125741004944,
      "learning_rate": 2.540206363682768e-06,
      "loss": 0.4004,
      "step": 680
    },
    {
      "epoch": 3.2115987460815045,
      "grad_norm": 0.49803441762924194,
      "learning_rate": 2.533505744444972e-06,
      "loss": 0.4178,
      "step": 681
    },
    {
      "epoch": 3.2163009404388716,
      "grad_norm": 0.49885454773902893,
      "learning_rate": 2.526804884458707e-06,
      "loss": 0.4089,
      "step": 682
    },
    {
      "epoch": 3.2210031347962382,
      "grad_norm": 0.527959942817688,
      "learning_rate": 2.520103831871591e-06,
      "loss": 0.4156,
      "step": 683
    },
    {
      "epoch": 3.225705329153605,
      "grad_norm": 0.5077382922172546,
      "learning_rate": 2.513402634832627e-06,
      "loss": 0.3946,
      "step": 684
    },
    {
      "epoch": 3.230407523510972,
      "grad_norm": 0.602326512336731,
      "learning_rate": 2.5067013414918523e-06,
      "loss": 0.4002,
      "step": 685
    },
    {
      "epoch": 3.2351097178683386,
      "grad_norm": 1.0339138507843018,
      "learning_rate": 2.5e-06,
      "loss": 0.4133,
      "step": 686
    },
    {
      "epoch": 3.239811912225705,
      "grad_norm": 0.5707440376281738,
      "learning_rate": 2.493298658508149e-06,
      "loss": 0.395,
      "step": 687
    },
    {
      "epoch": 3.2445141065830723,
      "grad_norm": 0.7559914588928223,
      "learning_rate": 2.4865973651673743e-06,
      "loss": 0.4128,
      "step": 688
    },
    {
      "epoch": 3.249216300940439,
      "grad_norm": 0.5365283489227295,
      "learning_rate": 2.4798961681284096e-06,
      "loss": 0.4371,
      "step": 689
    },
    {
      "epoch": 3.2539184952978055,
      "grad_norm": 0.5633559823036194,
      "learning_rate": 2.473195115541293e-06,
      "loss": 0.3945,
      "step": 690
    },
    {
      "epoch": 3.2586206896551726,
      "grad_norm": 0.502936840057373,
      "learning_rate": 2.466494255555029e-06,
      "loss": 0.4411,
      "step": 691
    },
    {
      "epoch": 3.2633228840125392,
      "grad_norm": 0.5298876166343689,
      "learning_rate": 2.459793636317233e-06,
      "loss": 0.4046,
      "step": 692
    },
    {
      "epoch": 3.268025078369906,
      "grad_norm": 0.6075465083122253,
      "learning_rate": 2.4530933059737936e-06,
      "loss": 0.4411,
      "step": 693
    },
    {
      "epoch": 3.2727272727272725,
      "grad_norm": 0.5472120046615601,
      "learning_rate": 2.4463933126685236e-06,
      "loss": 0.4055,
      "step": 694
    },
    {
      "epoch": 3.2774294670846396,
      "grad_norm": 0.5046771764755249,
      "learning_rate": 2.439693704542814e-06,
      "loss": 0.4015,
      "step": 695
    },
    {
      "epoch": 3.282131661442006,
      "grad_norm": 0.5108353495597839,
      "learning_rate": 2.432994529735286e-06,
      "loss": 0.4083,
      "step": 696
    },
    {
      "epoch": 3.2868338557993733,
      "grad_norm": 0.4968450963497162,
      "learning_rate": 2.4262958363814512e-06,
      "loss": 0.4185,
      "step": 697
    },
    {
      "epoch": 3.29153605015674,
      "grad_norm": 0.541350245475769,
      "learning_rate": 2.4195976726133574e-06,
      "loss": 0.4062,
      "step": 698
    },
    {
      "epoch": 3.2962382445141065,
      "grad_norm": 0.6516889929771423,
      "learning_rate": 2.4129000865592517e-06,
      "loss": 0.4059,
      "step": 699
    },
    {
      "epoch": 3.300940438871473,
      "grad_norm": 0.7150237560272217,
      "learning_rate": 2.4062031263432267e-06,
      "loss": 0.4259,
      "step": 700
    },
    {
      "epoch": 3.30564263322884,
      "grad_norm": 0.4952857792377472,
      "learning_rate": 2.3995068400848785e-06,
      "loss": 0.403,
      "step": 701
    },
    {
      "epoch": 3.310344827586207,
      "grad_norm": 0.6034581661224365,
      "learning_rate": 2.392811275898963e-06,
      "loss": 0.4192,
      "step": 702
    },
    {
      "epoch": 3.3150470219435735,
      "grad_norm": 0.7284024953842163,
      "learning_rate": 2.3861164818950448e-06,
      "loss": 0.3871,
      "step": 703
    },
    {
      "epoch": 3.3197492163009406,
      "grad_norm": 0.5211794972419739,
      "learning_rate": 2.379422506177157e-06,
      "loss": 0.4118,
      "step": 704
    },
    {
      "epoch": 3.324451410658307,
      "grad_norm": 0.5211071372032166,
      "learning_rate": 2.372729396843453e-06,
      "loss": 0.4139,
      "step": 705
    },
    {
      "epoch": 3.329153605015674,
      "grad_norm": 0.53047114610672,
      "learning_rate": 2.366037201985858e-06,
      "loss": 0.4169,
      "step": 706
    },
    {
      "epoch": 3.333855799373041,
      "grad_norm": 0.5385318994522095,
      "learning_rate": 2.3593459696897294e-06,
      "loss": 0.3949,
      "step": 707
    },
    {
      "epoch": 3.3385579937304075,
      "grad_norm": 0.5405430793762207,
      "learning_rate": 2.352655748033508e-06,
      "loss": 0.4126,
      "step": 708
    },
    {
      "epoch": 3.343260188087774,
      "grad_norm": 0.5243842005729675,
      "learning_rate": 2.3459665850883704e-06,
      "loss": 0.4342,
      "step": 709
    },
    {
      "epoch": 3.347962382445141,
      "grad_norm": 0.5552899241447449,
      "learning_rate": 2.33927852891789e-06,
      "loss": 0.4183,
      "step": 710
    },
    {
      "epoch": 3.352664576802508,
      "grad_norm": 0.7859904170036316,
      "learning_rate": 2.3325916275776834e-06,
      "loss": 0.4394,
      "step": 711
    },
    {
      "epoch": 3.3573667711598745,
      "grad_norm": 0.5205579400062561,
      "learning_rate": 2.3259059291150744e-06,
      "loss": 0.4004,
      "step": 712
    },
    {
      "epoch": 3.3620689655172415,
      "grad_norm": 0.5282108783721924,
      "learning_rate": 2.319221481568739e-06,
      "loss": 0.4205,
      "step": 713
    },
    {
      "epoch": 3.366771159874608,
      "grad_norm": 0.6452942490577698,
      "learning_rate": 2.3125383329683666e-06,
      "loss": 0.422,
      "step": 714
    },
    {
      "epoch": 3.371473354231975,
      "grad_norm": 0.5008051991462708,
      "learning_rate": 2.3058565313343152e-06,
      "loss": 0.3893,
      "step": 715
    },
    {
      "epoch": 3.376175548589342,
      "grad_norm": 0.6496668457984924,
      "learning_rate": 2.2991761246772623e-06,
      "loss": 0.4045,
      "step": 716
    },
    {
      "epoch": 3.3808777429467085,
      "grad_norm": 0.5258035659790039,
      "learning_rate": 2.2924971609978623e-06,
      "loss": 0.3973,
      "step": 717
    },
    {
      "epoch": 3.385579937304075,
      "grad_norm": 0.5373122096061707,
      "learning_rate": 2.285819688286403e-06,
      "loss": 0.3862,
      "step": 718
    },
    {
      "epoch": 3.3902821316614418,
      "grad_norm": 0.5864964723587036,
      "learning_rate": 2.2791437545224563e-06,
      "loss": 0.433,
      "step": 719
    },
    {
      "epoch": 3.394984326018809,
      "grad_norm": 0.5025304555892944,
      "learning_rate": 2.2724694076745397e-06,
      "loss": 0.4115,
      "step": 720
    },
    {
      "epoch": 3.3996865203761755,
      "grad_norm": 0.76083904504776,
      "learning_rate": 2.265796695699766e-06,
      "loss": 0.4155,
      "step": 721
    },
    {
      "epoch": 3.4043887147335425,
      "grad_norm": 0.5206389427185059,
      "learning_rate": 2.2591256665434998e-06,
      "loss": 0.4176,
      "step": 722
    },
    {
      "epoch": 3.409090909090909,
      "grad_norm": 0.559341311454773,
      "learning_rate": 2.252456368139019e-06,
      "loss": 0.414,
      "step": 723
    },
    {
      "epoch": 3.413793103448276,
      "grad_norm": 0.5281516313552856,
      "learning_rate": 2.245788848407159e-06,
      "loss": 0.4217,
      "step": 724
    },
    {
      "epoch": 3.4184952978056424,
      "grad_norm": 1.5208245515823364,
      "learning_rate": 2.2391231552559815e-06,
      "loss": 0.4209,
      "step": 725
    },
    {
      "epoch": 3.4231974921630095,
      "grad_norm": 0.5414798855781555,
      "learning_rate": 2.2324593365804184e-06,
      "loss": 0.3908,
      "step": 726
    },
    {
      "epoch": 3.427899686520376,
      "grad_norm": 0.587730348110199,
      "learning_rate": 2.225797440261936e-06,
      "loss": 0.4319,
      "step": 727
    },
    {
      "epoch": 3.4326018808777428,
      "grad_norm": 0.5572887659072876,
      "learning_rate": 2.219137514168187e-06,
      "loss": 0.3952,
      "step": 728
    },
    {
      "epoch": 3.43730407523511,
      "grad_norm": 0.589335560798645,
      "learning_rate": 2.212479606152667e-06,
      "loss": 0.4251,
      "step": 729
    },
    {
      "epoch": 3.4420062695924765,
      "grad_norm": 0.9469087719917297,
      "learning_rate": 2.205823764054372e-06,
      "loss": 0.4089,
      "step": 730
    },
    {
      "epoch": 3.446708463949843,
      "grad_norm": 0.5122615694999695,
      "learning_rate": 2.199170035697453e-06,
      "loss": 0.4083,
      "step": 731
    },
    {
      "epoch": 3.45141065830721,
      "grad_norm": 0.5385236144065857,
      "learning_rate": 2.1925184688908735e-06,
      "loss": 0.4127,
      "step": 732
    },
    {
      "epoch": 3.456112852664577,
      "grad_norm": 0.5631992220878601,
      "learning_rate": 2.185869111428067e-06,
      "loss": 0.4358,
      "step": 733
    },
    {
      "epoch": 3.4608150470219434,
      "grad_norm": 0.599616289138794,
      "learning_rate": 2.1792220110865885e-06,
      "loss": 0.4224,
      "step": 734
    },
    {
      "epoch": 3.4655172413793105,
      "grad_norm": 0.5281793475151062,
      "learning_rate": 2.1725772156277795e-06,
      "loss": 0.4028,
      "step": 735
    },
    {
      "epoch": 3.470219435736677,
      "grad_norm": 0.5679891705513,
      "learning_rate": 2.165934772796417e-06,
      "loss": 0.4127,
      "step": 736
    },
    {
      "epoch": 3.4749216300940438,
      "grad_norm": 0.58733731508255,
      "learning_rate": 2.159294730320374e-06,
      "loss": 0.4261,
      "step": 737
    },
    {
      "epoch": 3.479623824451411,
      "grad_norm": 0.5951966643333435,
      "learning_rate": 2.15265713591028e-06,
      "loss": 0.4132,
      "step": 738
    },
    {
      "epoch": 3.4843260188087775,
      "grad_norm": 0.5945015549659729,
      "learning_rate": 2.1460220372591676e-06,
      "loss": 0.4327,
      "step": 739
    },
    {
      "epoch": 3.489028213166144,
      "grad_norm": 0.5072157979011536,
      "learning_rate": 2.139389482042142e-06,
      "loss": 0.3988,
      "step": 740
    },
    {
      "epoch": 3.493730407523511,
      "grad_norm": 0.5170702338218689,
      "learning_rate": 2.1327595179160332e-06,
      "loss": 0.4157,
      "step": 741
    },
    {
      "epoch": 3.498432601880878,
      "grad_norm": 0.5536943674087524,
      "learning_rate": 2.1261321925190492e-06,
      "loss": 0.4061,
      "step": 742
    },
    {
      "epoch": 3.5031347962382444,
      "grad_norm": 0.5288745164871216,
      "learning_rate": 2.1195075534704433e-06,
      "loss": 0.383,
      "step": 743
    },
    {
      "epoch": 3.507836990595611,
      "grad_norm": 0.5845092535018921,
      "learning_rate": 2.1128856483701625e-06,
      "loss": 0.411,
      "step": 744
    },
    {
      "epoch": 3.512539184952978,
      "grad_norm": 0.5369064211845398,
      "learning_rate": 2.10626652479851e-06,
      "loss": 0.3936,
      "step": 745
    },
    {
      "epoch": 3.5172413793103448,
      "grad_norm": 0.5584667325019836,
      "learning_rate": 2.0996502303158057e-06,
      "loss": 0.4081,
      "step": 746
    },
    {
      "epoch": 3.521943573667712,
      "grad_norm": 0.537060022354126,
      "learning_rate": 2.0930368124620385e-06,
      "loss": 0.4087,
      "step": 747
    },
    {
      "epoch": 3.5266457680250785,
      "grad_norm": 0.5309715270996094,
      "learning_rate": 2.086426318756531e-06,
      "loss": 0.4255,
      "step": 748
    },
    {
      "epoch": 3.531347962382445,
      "grad_norm": 0.5657020807266235,
      "learning_rate": 2.0798187966975917e-06,
      "loss": 0.4317,
      "step": 749
    },
    {
      "epoch": 3.5360501567398117,
      "grad_norm": 0.5245983600616455,
      "learning_rate": 2.073214293762179e-06,
      "loss": 0.4203,
      "step": 750
    },
    {
      "epoch": 3.540752351097179,
      "grad_norm": 0.5540620684623718,
      "learning_rate": 2.0666128574055575e-06,
      "loss": 0.4178,
      "step": 751
    },
    {
      "epoch": 3.5454545454545454,
      "grad_norm": 0.5255053639411926,
      "learning_rate": 2.0600145350609585e-06,
      "loss": 0.4183,
      "step": 752
    },
    {
      "epoch": 3.5501567398119125,
      "grad_norm": 0.5604968070983887,
      "learning_rate": 2.053419374139235e-06,
      "loss": 0.4157,
      "step": 753
    },
    {
      "epoch": 3.554858934169279,
      "grad_norm": 0.5198661684989929,
      "learning_rate": 2.0468274220285295e-06,
      "loss": 0.4145,
      "step": 754
    },
    {
      "epoch": 3.5595611285266457,
      "grad_norm": 0.6241534352302551,
      "learning_rate": 2.0402387260939224e-06,
      "loss": 0.4101,
      "step": 755
    },
    {
      "epoch": 3.5642633228840124,
      "grad_norm": 0.5478257536888123,
      "learning_rate": 2.033653333677103e-06,
      "loss": 0.4211,
      "step": 756
    },
    {
      "epoch": 3.5689655172413794,
      "grad_norm": 0.544718861579895,
      "learning_rate": 2.02707129209602e-06,
      "loss": 0.403,
      "step": 757
    },
    {
      "epoch": 3.573667711598746,
      "grad_norm": 0.567952036857605,
      "learning_rate": 2.0204926486445463e-06,
      "loss": 0.4224,
      "step": 758
    },
    {
      "epoch": 3.5783699059561127,
      "grad_norm": 0.576141357421875,
      "learning_rate": 2.0139174505921403e-06,
      "loss": 0.4424,
      "step": 759
    },
    {
      "epoch": 3.58307210031348,
      "grad_norm": 0.5517438650131226,
      "learning_rate": 2.0073457451835e-06,
      "loss": 0.3991,
      "step": 760
    },
    {
      "epoch": 3.5877742946708464,
      "grad_norm": 0.49781471490859985,
      "learning_rate": 2.0007775796382335e-06,
      "loss": 0.4252,
      "step": 761
    },
    {
      "epoch": 3.592476489028213,
      "grad_norm": 0.5482198596000671,
      "learning_rate": 1.994213001150508e-06,
      "loss": 0.3928,
      "step": 762
    },
    {
      "epoch": 3.5971786833855797,
      "grad_norm": 0.5472637414932251,
      "learning_rate": 1.9876520568887207e-06,
      "loss": 0.4193,
      "step": 763
    },
    {
      "epoch": 3.6018808777429467,
      "grad_norm": 0.6185949444770813,
      "learning_rate": 1.981094793995155e-06,
      "loss": 0.4034,
      "step": 764
    },
    {
      "epoch": 3.6065830721003134,
      "grad_norm": 0.5806577801704407,
      "learning_rate": 1.974541259585641e-06,
      "loss": 0.4316,
      "step": 765
    },
    {
      "epoch": 3.6112852664576804,
      "grad_norm": 0.5567718744277954,
      "learning_rate": 1.9679915007492194e-06,
      "loss": 0.412,
      "step": 766
    },
    {
      "epoch": 3.615987460815047,
      "grad_norm": 0.7912635207176208,
      "learning_rate": 1.9614455645478047e-06,
      "loss": 0.4014,
      "step": 767
    },
    {
      "epoch": 3.6206896551724137,
      "grad_norm": 0.5250074863433838,
      "learning_rate": 1.9549034980158403e-06,
      "loss": 0.4146,
      "step": 768
    },
    {
      "epoch": 3.6253918495297803,
      "grad_norm": 0.5616881847381592,
      "learning_rate": 1.9483653481599697e-06,
      "loss": 0.4197,
      "step": 769
    },
    {
      "epoch": 3.6300940438871474,
      "grad_norm": 0.558824896812439,
      "learning_rate": 1.9418311619586897e-06,
      "loss": 0.4258,
      "step": 770
    },
    {
      "epoch": 3.634796238244514,
      "grad_norm": 0.5861754417419434,
      "learning_rate": 1.935300986362018e-06,
      "loss": 0.3986,
      "step": 771
    },
    {
      "epoch": 3.639498432601881,
      "grad_norm": 0.9162700176239014,
      "learning_rate": 1.9287748682911582e-06,
      "loss": 0.4302,
      "step": 772
    },
    {
      "epoch": 3.6442006269592477,
      "grad_norm": 0.9436094760894775,
      "learning_rate": 1.9222528546381543e-06,
      "loss": 0.4216,
      "step": 773
    },
    {
      "epoch": 3.6489028213166144,
      "grad_norm": 0.8662758469581604,
      "learning_rate": 1.9157349922655648e-06,
      "loss": 0.3978,
      "step": 774
    },
    {
      "epoch": 3.653605015673981,
      "grad_norm": 0.5423682332038879,
      "learning_rate": 1.909221328006114e-06,
      "loss": 0.409,
      "step": 775
    },
    {
      "epoch": 3.658307210031348,
      "grad_norm": 0.5691127777099609,
      "learning_rate": 1.9027119086623647e-06,
      "loss": 0.4215,
      "step": 776
    },
    {
      "epoch": 3.6630094043887147,
      "grad_norm": 0.5373672246932983,
      "learning_rate": 1.8962067810063806e-06,
      "loss": 0.4367,
      "step": 777
    },
    {
      "epoch": 3.6677115987460818,
      "grad_norm": 0.5069293975830078,
      "learning_rate": 1.8897059917793844e-06,
      "loss": 0.3924,
      "step": 778
    },
    {
      "epoch": 3.6724137931034484,
      "grad_norm": 0.5330380797386169,
      "learning_rate": 1.8832095876914268e-06,
      "loss": 0.4103,
      "step": 779
    },
    {
      "epoch": 3.677115987460815,
      "grad_norm": 0.5398195385932922,
      "learning_rate": 1.8767176154210537e-06,
      "loss": 0.4176,
      "step": 780
    },
    {
      "epoch": 3.6818181818181817,
      "grad_norm": 0.5592144727706909,
      "learning_rate": 1.8702301216149616e-06,
      "loss": 0.4074,
      "step": 781
    },
    {
      "epoch": 3.6865203761755487,
      "grad_norm": 0.5743839144706726,
      "learning_rate": 1.8637471528876727e-06,
      "loss": 0.4076,
      "step": 782
    },
    {
      "epoch": 3.6912225705329154,
      "grad_norm": 0.5384632349014282,
      "learning_rate": 1.8572687558211923e-06,
      "loss": 0.4189,
      "step": 783
    },
    {
      "epoch": 3.695924764890282,
      "grad_norm": 0.5259303450584412,
      "learning_rate": 1.850794976964677e-06,
      "loss": 0.3814,
      "step": 784
    },
    {
      "epoch": 3.700626959247649,
      "grad_norm": 0.5677003860473633,
      "learning_rate": 1.8443258628341026e-06,
      "loss": 0.4149,
      "step": 785
    },
    {
      "epoch": 3.7053291536050157,
      "grad_norm": 0.6973782181739807,
      "learning_rate": 1.837861459911925e-06,
      "loss": 0.4248,
      "step": 786
    },
    {
      "epoch": 3.7100313479623823,
      "grad_norm": 0.5952889323234558,
      "learning_rate": 1.8314018146467505e-06,
      "loss": 0.4051,
      "step": 787
    },
    {
      "epoch": 3.714733542319749,
      "grad_norm": 0.5266870856285095,
      "learning_rate": 1.8249469734529995e-06,
      "loss": 0.3879,
      "step": 788
    },
    {
      "epoch": 3.719435736677116,
      "grad_norm": 0.5926518440246582,
      "learning_rate": 1.818496982710572e-06,
      "loss": 0.4233,
      "step": 789
    },
    {
      "epoch": 3.7241379310344827,
      "grad_norm": 0.6534157395362854,
      "learning_rate": 1.81205188876452e-06,
      "loss": 0.4239,
      "step": 790
    },
    {
      "epoch": 3.7288401253918497,
      "grad_norm": 0.5505423545837402,
      "learning_rate": 1.8056117379247078e-06,
      "loss": 0.4267,
      "step": 791
    },
    {
      "epoch": 3.7335423197492164,
      "grad_norm": 0.626750111579895,
      "learning_rate": 1.7991765764654813e-06,
      "loss": 0.415,
      "step": 792
    },
    {
      "epoch": 3.738244514106583,
      "grad_norm": 0.5631802082061768,
      "learning_rate": 1.7927464506253394e-06,
      "loss": 0.4237,
      "step": 793
    },
    {
      "epoch": 3.7429467084639496,
      "grad_norm": 0.5428839325904846,
      "learning_rate": 1.7863214066065951e-06,
      "loss": 0.391,
      "step": 794
    },
    {
      "epoch": 3.7476489028213167,
      "grad_norm": 0.5595945119857788,
      "learning_rate": 1.779901490575051e-06,
      "loss": 0.4186,
      "step": 795
    },
    {
      "epoch": 3.7523510971786833,
      "grad_norm": 0.5219384431838989,
      "learning_rate": 1.7734867486596596e-06,
      "loss": 0.4204,
      "step": 796
    },
    {
      "epoch": 3.7570532915360504,
      "grad_norm": 0.5223650932312012,
      "learning_rate": 1.767077226952198e-06,
      "loss": 0.4267,
      "step": 797
    },
    {
      "epoch": 3.761755485893417,
      "grad_norm": 0.5161250829696655,
      "learning_rate": 1.7606729715069349e-06,
      "loss": 0.4146,
      "step": 798
    },
    {
      "epoch": 3.7664576802507836,
      "grad_norm": 0.5676782727241516,
      "learning_rate": 1.7542740283402981e-06,
      "loss": 0.4177,
      "step": 799
    },
    {
      "epoch": 3.7711598746081503,
      "grad_norm": 0.5766341090202332,
      "learning_rate": 1.7478804434305466e-06,
      "loss": 0.4,
      "step": 800
    },
    {
      "epoch": 3.7758620689655173,
      "grad_norm": 0.5547144412994385,
      "learning_rate": 1.741492262717438e-06,
      "loss": 0.4182,
      "step": 801
    },
    {
      "epoch": 3.780564263322884,
      "grad_norm": 0.5687820315361023,
      "learning_rate": 1.7351095321018974e-06,
      "loss": 0.3857,
      "step": 802
    },
    {
      "epoch": 3.785266457680251,
      "grad_norm": 0.5266862511634827,
      "learning_rate": 1.7287322974456933e-06,
      "loss": 0.3928,
      "step": 803
    },
    {
      "epoch": 3.7899686520376177,
      "grad_norm": 0.5501623153686523,
      "learning_rate": 1.7223606045711006e-06,
      "loss": 0.4165,
      "step": 804
    },
    {
      "epoch": 3.7946708463949843,
      "grad_norm": 0.5451024770736694,
      "learning_rate": 1.7159944992605774e-06,
      "loss": 0.4204,
      "step": 805
    },
    {
      "epoch": 3.799373040752351,
      "grad_norm": 0.5108196139335632,
      "learning_rate": 1.7096340272564318e-06,
      "loss": 0.409,
      "step": 806
    },
    {
      "epoch": 3.804075235109718,
      "grad_norm": 0.5510005950927734,
      "learning_rate": 1.7032792342604947e-06,
      "loss": 0.4005,
      "step": 807
    },
    {
      "epoch": 3.8087774294670846,
      "grad_norm": 0.5405788421630859,
      "learning_rate": 1.6969301659337944e-06,
      "loss": 0.4155,
      "step": 808
    },
    {
      "epoch": 3.8134796238244513,
      "grad_norm": 0.6032125353813171,
      "learning_rate": 1.6905868678962225e-06,
      "loss": 0.42,
      "step": 809
    },
    {
      "epoch": 3.8181818181818183,
      "grad_norm": 0.9348533749580383,
      "learning_rate": 1.684249385726211e-06,
      "loss": 0.4113,
      "step": 810
    },
    {
      "epoch": 3.822884012539185,
      "grad_norm": 0.5582144856452942,
      "learning_rate": 1.677917764960404e-06,
      "loss": 0.4012,
      "step": 811
    },
    {
      "epoch": 3.8275862068965516,
      "grad_norm": 0.5778011679649353,
      "learning_rate": 1.6715920510933277e-06,
      "loss": 0.4314,
      "step": 812
    },
    {
      "epoch": 3.8322884012539182,
      "grad_norm": 1.647594928741455,
      "learning_rate": 1.6652722895770676e-06,
      "loss": 0.4274,
      "step": 813
    },
    {
      "epoch": 3.8369905956112853,
      "grad_norm": 0.5681491494178772,
      "learning_rate": 1.6589585258209383e-06,
      "loss": 0.3784,
      "step": 814
    },
    {
      "epoch": 3.841692789968652,
      "grad_norm": 0.5122655630111694,
      "learning_rate": 1.6526508051911588e-06,
      "loss": 0.4186,
      "step": 815
    },
    {
      "epoch": 3.846394984326019,
      "grad_norm": 0.6520569920539856,
      "learning_rate": 1.6463491730105282e-06,
      "loss": 0.4057,
      "step": 816
    },
    {
      "epoch": 3.8510971786833856,
      "grad_norm": 0.5363179445266724,
      "learning_rate": 1.6400536745580955e-06,
      "loss": 0.3887,
      "step": 817
    },
    {
      "epoch": 3.8557993730407523,
      "grad_norm": 0.5956358909606934,
      "learning_rate": 1.6337643550688408e-06,
      "loss": 0.4171,
      "step": 818
    },
    {
      "epoch": 3.860501567398119,
      "grad_norm": 0.5232042670249939,
      "learning_rate": 1.627481259733343e-06,
      "loss": 0.3965,
      "step": 819
    },
    {
      "epoch": 3.865203761755486,
      "grad_norm": 0.5099393129348755,
      "learning_rate": 1.6212044336974598e-06,
      "loss": 0.3954,
      "step": 820
    },
    {
      "epoch": 3.8699059561128526,
      "grad_norm": 0.5676040053367615,
      "learning_rate": 1.614933922062003e-06,
      "loss": 0.4064,
      "step": 821
    },
    {
      "epoch": 3.8746081504702197,
      "grad_norm": 0.5794877409934998,
      "learning_rate": 1.6086697698824144e-06,
      "loss": 0.4017,
      "step": 822
    },
    {
      "epoch": 3.8793103448275863,
      "grad_norm": 0.572999119758606,
      "learning_rate": 1.6024120221684373e-06,
      "loss": 0.4002,
      "step": 823
    },
    {
      "epoch": 3.884012539184953,
      "grad_norm": 0.5625258088111877,
      "learning_rate": 1.5961607238838022e-06,
      "loss": 0.4086,
      "step": 824
    },
    {
      "epoch": 3.8887147335423196,
      "grad_norm": 0.5278126001358032,
      "learning_rate": 1.589915919945894e-06,
      "loss": 0.418,
      "step": 825
    },
    {
      "epoch": 3.8934169278996866,
      "grad_norm": 0.6338787078857422,
      "learning_rate": 1.5836776552254386e-06,
      "loss": 0.4353,
      "step": 826
    },
    {
      "epoch": 3.8981191222570533,
      "grad_norm": 0.6143558621406555,
      "learning_rate": 1.5774459745461711e-06,
      "loss": 0.4052,
      "step": 827
    },
    {
      "epoch": 3.9028213166144203,
      "grad_norm": 0.5196287631988525,
      "learning_rate": 1.5712209226845201e-06,
      "loss": 0.3821,
      "step": 828
    },
    {
      "epoch": 3.907523510971787,
      "grad_norm": 0.5631325840950012,
      "learning_rate": 1.565002544369286e-06,
      "loss": 0.4154,
      "step": 829
    },
    {
      "epoch": 3.9122257053291536,
      "grad_norm": 0.6087117195129395,
      "learning_rate": 1.5587908842813142e-06,
      "loss": 0.4056,
      "step": 830
    },
    {
      "epoch": 3.91692789968652,
      "grad_norm": 0.5318475365638733,
      "learning_rate": 1.5525859870531823e-06,
      "loss": 0.4212,
      "step": 831
    },
    {
      "epoch": 3.9216300940438873,
      "grad_norm": 0.5385299324989319,
      "learning_rate": 1.5463878972688707e-06,
      "loss": 0.4081,
      "step": 832
    },
    {
      "epoch": 3.926332288401254,
      "grad_norm": 0.5263051986694336,
      "learning_rate": 1.5401966594634483e-06,
      "loss": 0.4351,
      "step": 833
    },
    {
      "epoch": 3.9310344827586206,
      "grad_norm": 0.5557870864868164,
      "learning_rate": 1.5340123181227495e-06,
      "loss": 0.4235,
      "step": 834
    },
    {
      "epoch": 3.9357366771159876,
      "grad_norm": 0.5445315837860107,
      "learning_rate": 1.527834917683058e-06,
      "loss": 0.391,
      "step": 835
    },
    {
      "epoch": 3.9404388714733543,
      "grad_norm": 0.5401822924613953,
      "learning_rate": 1.5216645025307813e-06,
      "loss": 0.4057,
      "step": 836
    },
    {
      "epoch": 3.945141065830721,
      "grad_norm": 0.5524025559425354,
      "learning_rate": 1.5155011170021399e-06,
      "loss": 0.4134,
      "step": 837
    },
    {
      "epoch": 3.9498432601880875,
      "grad_norm": 0.5604265332221985,
      "learning_rate": 1.5093448053828402e-06,
      "loss": 0.4231,
      "step": 838
    },
    {
      "epoch": 3.9545454545454546,
      "grad_norm": 0.5822537541389465,
      "learning_rate": 1.503195611907764e-06,
      "loss": 0.4238,
      "step": 839
    },
    {
      "epoch": 3.959247648902821,
      "grad_norm": 0.5396215319633484,
      "learning_rate": 1.4970535807606453e-06,
      "loss": 0.3825,
      "step": 840
    },
    {
      "epoch": 3.9639498432601883,
      "grad_norm": 0.5705523490905762,
      "learning_rate": 1.4909187560737542e-06,
      "loss": 0.3951,
      "step": 841
    },
    {
      "epoch": 3.968652037617555,
      "grad_norm": 0.5342106819152832,
      "learning_rate": 1.4847911819275829e-06,
      "loss": 0.4068,
      "step": 842
    },
    {
      "epoch": 3.9733542319749215,
      "grad_norm": 0.5606650710105896,
      "learning_rate": 1.4786709023505224e-06,
      "loss": 0.397,
      "step": 843
    },
    {
      "epoch": 3.978056426332288,
      "grad_norm": 0.6344854235649109,
      "learning_rate": 1.4725579613185549e-06,
      "loss": 0.4226,
      "step": 844
    },
    {
      "epoch": 3.9827586206896552,
      "grad_norm": 0.6624012589454651,
      "learning_rate": 1.4664524027549291e-06,
      "loss": 0.4096,
      "step": 845
    },
    {
      "epoch": 3.987460815047022,
      "grad_norm": 0.5686039328575134,
      "learning_rate": 1.4603542705298493e-06,
      "loss": 0.395,
      "step": 846
    },
    {
      "epoch": 3.992163009404389,
      "grad_norm": 0.5838780999183655,
      "learning_rate": 1.4542636084601624e-06,
      "loss": 0.3682,
      "step": 847
    },
    {
      "epoch": 3.9968652037617556,
      "grad_norm": 0.555031418800354,
      "learning_rate": 1.4481804603090358e-06,
      "loss": 0.4111,
      "step": 848
    }
  ],
  "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": 5.461772594205793e+19,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}