{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9968652037617555,
  "eval_steps": 500,
  "global_step": 212,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004702194357366771,
      "grad_norm": 3.308954954147339,
      "learning_rate": 5.0000000000000004e-08,
      "loss": 0.9758,
      "step": 1
    },
    {
      "epoch": 0.009404388714733543,
      "grad_norm": 3.0945961475372314,
      "learning_rate": 1.0000000000000001e-07,
      "loss": 0.9784,
      "step": 2
    },
    {
      "epoch": 0.014106583072100314,
      "grad_norm": 2.6266396045684814,
      "learning_rate": 1.5000000000000002e-07,
      "loss": 0.9407,
      "step": 3
    },
    {
      "epoch": 0.018808777429467086,
      "grad_norm": 2.781346321105957,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 0.983,
      "step": 4
    },
    {
      "epoch": 0.023510971786833857,
      "grad_norm": 2.8360986709594727,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 0.941,
      "step": 5
    },
    {
      "epoch": 0.02821316614420063,
      "grad_norm": 2.814222812652588,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 0.9675,
      "step": 6
    },
    {
      "epoch": 0.032915360501567396,
      "grad_norm": 2.821528196334839,
      "learning_rate": 3.5000000000000004e-07,
      "loss": 0.8856,
      "step": 7
    },
    {
      "epoch": 0.03761755485893417,
      "grad_norm": 2.7562029361724854,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 0.95,
      "step": 8
    },
    {
      "epoch": 0.04231974921630094,
      "grad_norm": 2.9743621349334717,
      "learning_rate": 4.5000000000000003e-07,
      "loss": 0.9654,
      "step": 9
    },
    {
      "epoch": 0.047021943573667714,
      "grad_norm": 3.0444681644439697,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.9133,
      "step": 10
    },
    {
      "epoch": 0.05172413793103448,
      "grad_norm": 2.9696431159973145,
      "learning_rate": 5.5e-07,
      "loss": 0.9425,
      "step": 11
    },
    {
      "epoch": 0.05642633228840126,
      "grad_norm": 2.8478922843933105,
      "learning_rate": 6.000000000000001e-07,
      "loss": 0.9743,
      "step": 12
    },
    {
      "epoch": 0.061128526645768025,
      "grad_norm": 2.664818525314331,
      "learning_rate": 6.5e-07,
      "loss": 0.9276,
      "step": 13
    },
    {
      "epoch": 0.06583072100313479,
      "grad_norm": 2.727060556411743,
      "learning_rate": 7.000000000000001e-07,
      "loss": 0.9221,
      "step": 14
    },
    {
      "epoch": 0.07053291536050156,
      "grad_norm": 2.8166537284851074,
      "learning_rate": 7.5e-07,
      "loss": 0.9406,
      "step": 15
    },
    {
      "epoch": 0.07523510971786834,
      "grad_norm": 2.8412551879882812,
      "learning_rate": 8.000000000000001e-07,
      "loss": 0.901,
      "step": 16
    },
    {
      "epoch": 0.07993730407523511,
      "grad_norm": 2.6464271545410156,
      "learning_rate": 8.500000000000001e-07,
      "loss": 0.9586,
      "step": 17
    },
    {
      "epoch": 0.08463949843260188,
      "grad_norm": 2.506002187728882,
      "learning_rate": 9.000000000000001e-07,
      "loss": 0.8895,
      "step": 18
    },
    {
      "epoch": 0.08934169278996865,
      "grad_norm": 2.663196086883545,
      "learning_rate": 9.500000000000001e-07,
      "loss": 0.8833,
      "step": 19
    },
    {
      "epoch": 0.09404388714733543,
      "grad_norm": 2.5434982776641846,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.8737,
      "step": 20
    },
    {
      "epoch": 0.0987460815047022,
      "grad_norm": 2.2520358562469482,
      "learning_rate": 1.0500000000000001e-06,
      "loss": 0.8895,
      "step": 21
    },
    {
      "epoch": 0.10344827586206896,
      "grad_norm": 2.3946690559387207,
      "learning_rate": 1.1e-06,
      "loss": 0.8405,
      "step": 22
    },
    {
      "epoch": 0.10815047021943573,
      "grad_norm": 2.209516763687134,
      "learning_rate": 1.1500000000000002e-06,
      "loss": 0.8441,
      "step": 23
    },
    {
      "epoch": 0.11285266457680251,
      "grad_norm": 2.3084263801574707,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 0.8103,
      "step": 24
    },
    {
      "epoch": 0.11755485893416928,
      "grad_norm": 2.4108471870422363,
      "learning_rate": 1.25e-06,
      "loss": 0.808,
      "step": 25
    },
    {
      "epoch": 0.12225705329153605,
      "grad_norm": 2.116116762161255,
      "learning_rate": 1.3e-06,
      "loss": 0.7901,
      "step": 26
    },
    {
      "epoch": 0.12695924764890282,
      "grad_norm": 2.2457115650177,
      "learning_rate": 1.3500000000000002e-06,
      "loss": 0.7737,
      "step": 27
    },
    {
      "epoch": 0.13166144200626959,
      "grad_norm": 1.9436711072921753,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 0.7465,
      "step": 28
    },
    {
      "epoch": 0.13636363636363635,
      "grad_norm": 2.1678073406219482,
      "learning_rate": 1.45e-06,
      "loss": 0.7819,
      "step": 29
    },
    {
      "epoch": 0.14106583072100312,
      "grad_norm": 1.8869282007217407,
      "learning_rate": 1.5e-06,
      "loss": 0.7653,
      "step": 30
    },
    {
      "epoch": 0.14576802507836992,
      "grad_norm": 1.6957731246948242,
      "learning_rate": 1.5500000000000002e-06,
      "loss": 0.7659,
      "step": 31
    },
    {
      "epoch": 0.15047021943573669,
      "grad_norm": 1.4701424837112427,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 0.7553,
      "step": 32
    },
    {
      "epoch": 0.15517241379310345,
      "grad_norm": 1.3234986066818237,
      "learning_rate": 1.6500000000000003e-06,
      "loss": 0.6875,
      "step": 33
    },
    {
      "epoch": 0.15987460815047022,
      "grad_norm": 1.3325831890106201,
      "learning_rate": 1.7000000000000002e-06,
      "loss": 0.6681,
      "step": 34
    },
    {
      "epoch": 0.164576802507837,
      "grad_norm": 1.3700138330459595,
      "learning_rate": 1.75e-06,
      "loss": 0.7635,
      "step": 35
    },
    {
      "epoch": 0.16927899686520376,
      "grad_norm": 1.2316628694534302,
      "learning_rate": 1.8000000000000001e-06,
      "loss": 0.6776,
      "step": 36
    },
    {
      "epoch": 0.17398119122257052,
      "grad_norm": 1.170616865158081,
      "learning_rate": 1.85e-06,
      "loss": 0.7073,
      "step": 37
    },
    {
      "epoch": 0.1786833855799373,
      "grad_norm": 1.1797659397125244,
      "learning_rate": 1.9000000000000002e-06,
      "loss": 0.7572,
      "step": 38
    },
    {
      "epoch": 0.1833855799373041,
      "grad_norm": 1.0123047828674316,
      "learning_rate": 1.9500000000000004e-06,
      "loss": 0.6798,
      "step": 39
    },
    {
      "epoch": 0.18808777429467086,
      "grad_norm": 0.975169837474823,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.7088,
      "step": 40
    },
    {
      "epoch": 0.19278996865203762,
      "grad_norm": 0.891377866268158,
      "learning_rate": 2.05e-06,
      "loss": 0.6696,
      "step": 41
    },
    {
      "epoch": 0.1974921630094044,
      "grad_norm": 0.8997182250022888,
      "learning_rate": 2.1000000000000002e-06,
      "loss": 0.7135,
      "step": 42
    },
    {
      "epoch": 0.20219435736677116,
      "grad_norm": 0.9378617405891418,
      "learning_rate": 2.15e-06,
      "loss": 0.6827,
      "step": 43
    },
    {
      "epoch": 0.20689655172413793,
      "grad_norm": 0.8060838580131531,
      "learning_rate": 2.2e-06,
      "loss": 0.6704,
      "step": 44
    },
    {
      "epoch": 0.2115987460815047,
      "grad_norm": 0.7708892822265625,
      "learning_rate": 2.25e-06,
      "loss": 0.6499,
      "step": 45
    },
    {
      "epoch": 0.21630094043887146,
      "grad_norm": 0.7236770987510681,
      "learning_rate": 2.3000000000000004e-06,
      "loss": 0.657,
      "step": 46
    },
    {
      "epoch": 0.22100313479623823,
      "grad_norm": 0.6666117906570435,
      "learning_rate": 2.35e-06,
      "loss": 0.634,
      "step": 47
    },
    {
      "epoch": 0.22570532915360503,
      "grad_norm": 0.6475918292999268,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.6237,
      "step": 48
    },
    {
      "epoch": 0.2304075235109718,
      "grad_norm": 0.7277692556381226,
      "learning_rate": 2.4500000000000003e-06,
      "loss": 0.6707,
      "step": 49
    },
    {
      "epoch": 0.23510971786833856,
      "grad_norm": 0.6572731733322144,
      "learning_rate": 2.5e-06,
      "loss": 0.6812,
      "step": 50
    },
    {
      "epoch": 0.23981191222570533,
      "grad_norm": 0.6716486811637878,
      "learning_rate": 2.55e-06,
      "loss": 0.6803,
      "step": 51
    },
    {
      "epoch": 0.2445141065830721,
      "grad_norm": 0.6065218448638916,
      "learning_rate": 2.6e-06,
      "loss": 0.6353,
      "step": 52
    },
    {
      "epoch": 0.24921630094043887,
      "grad_norm": 0.6749820709228516,
      "learning_rate": 2.6500000000000005e-06,
      "loss": 0.6393,
      "step": 53
    },
    {
      "epoch": 0.25391849529780564,
      "grad_norm": 0.6187207102775574,
      "learning_rate": 2.7000000000000004e-06,
      "loss": 0.6318,
      "step": 54
    },
    {
      "epoch": 0.25862068965517243,
      "grad_norm": 0.6700596809387207,
      "learning_rate": 2.7500000000000004e-06,
      "loss": 0.6762,
      "step": 55
    },
    {
      "epoch": 0.26332288401253917,
      "grad_norm": 0.6460264921188354,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.6388,
      "step": 56
    },
    {
      "epoch": 0.26802507836990597,
      "grad_norm": 0.7263085842132568,
      "learning_rate": 2.85e-06,
      "loss": 0.6413,
      "step": 57
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 0.674784779548645,
      "learning_rate": 2.9e-06,
      "loss": 0.6134,
      "step": 58
    },
    {
      "epoch": 0.2774294670846395,
      "grad_norm": 0.6552851796150208,
      "learning_rate": 2.95e-06,
      "loss": 0.6508,
      "step": 59
    },
    {
      "epoch": 0.28213166144200624,
      "grad_norm": 0.7160854339599609,
      "learning_rate": 3e-06,
      "loss": 0.677,
      "step": 60
    },
    {
      "epoch": 0.28683385579937304,
      "grad_norm": 0.682802677154541,
      "learning_rate": 3.05e-06,
      "loss": 0.5983,
      "step": 61
    },
    {
      "epoch": 0.29153605015673983,
      "grad_norm": 0.5360945463180542,
      "learning_rate": 3.1000000000000004e-06,
      "loss": 0.5928,
      "step": 62
    },
    {
      "epoch": 0.2962382445141066,
      "grad_norm": 0.6875145435333252,
      "learning_rate": 3.1500000000000003e-06,
      "loss": 0.6366,
      "step": 63
    },
    {
      "epoch": 0.30094043887147337,
      "grad_norm": 0.5564191937446594,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.6466,
      "step": 64
    },
    {
      "epoch": 0.3056426332288401,
      "grad_norm": 0.8959600925445557,
      "learning_rate": 3.2500000000000002e-06,
      "loss": 0.5875,
      "step": 65
    },
    {
      "epoch": 0.3103448275862069,
      "grad_norm": 0.5513110160827637,
      "learning_rate": 3.3000000000000006e-06,
      "loss": 0.6042,
      "step": 66
    },
    {
      "epoch": 0.31504702194357365,
      "grad_norm": 0.5638250112533569,
      "learning_rate": 3.3500000000000005e-06,
      "loss": 0.6343,
      "step": 67
    },
    {
      "epoch": 0.31974921630094044,
      "grad_norm": 0.5723473429679871,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 0.6007,
      "step": 68
    },
    {
      "epoch": 0.32445141065830724,
      "grad_norm": 0.5193734765052795,
      "learning_rate": 3.45e-06,
      "loss": 0.5877,
      "step": 69
    },
    {
      "epoch": 0.329153605015674,
      "grad_norm": 0.7437443733215332,
      "learning_rate": 3.5e-06,
      "loss": 0.6579,
      "step": 70
    },
    {
      "epoch": 0.3338557993730408,
      "grad_norm": 0.5810403823852539,
      "learning_rate": 3.5500000000000003e-06,
      "loss": 0.625,
      "step": 71
    },
    {
      "epoch": 0.3385579937304075,
      "grad_norm": 0.5594515204429626,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 0.6039,
      "step": 72
    },
    {
      "epoch": 0.3432601880877743,
      "grad_norm": 0.5781659483909607,
      "learning_rate": 3.65e-06,
      "loss": 0.5936,
      "step": 73
    },
    {
      "epoch": 0.34796238244514105,
      "grad_norm": 0.5736168622970581,
      "learning_rate": 3.7e-06,
      "loss": 0.61,
      "step": 74
    },
    {
      "epoch": 0.35266457680250785,
      "grad_norm": 0.5082181096076965,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.6031,
      "step": 75
    },
    {
      "epoch": 0.3573667711598746,
      "grad_norm": 0.5188243985176086,
      "learning_rate": 3.8000000000000005e-06,
      "loss": 0.6048,
      "step": 76
    },
    {
      "epoch": 0.3620689655172414,
      "grad_norm": 0.678931713104248,
      "learning_rate": 3.85e-06,
      "loss": 0.615,
      "step": 77
    },
    {
      "epoch": 0.3667711598746082,
      "grad_norm": 0.5580578446388245,
      "learning_rate": 3.900000000000001e-06,
      "loss": 0.5622,
      "step": 78
    },
    {
      "epoch": 0.3714733542319749,
      "grad_norm": 0.5222985744476318,
      "learning_rate": 3.95e-06,
      "loss": 0.6233,
      "step": 79
    },
    {
      "epoch": 0.3761755485893417,
      "grad_norm": 0.6105766892433167,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.5448,
      "step": 80
    },
    {
      "epoch": 0.38087774294670845,
      "grad_norm": 0.4837906062602997,
      "learning_rate": 4.05e-06,
      "loss": 0.6058,
      "step": 81
    },
    {
      "epoch": 0.38557993730407525,
      "grad_norm": 0.5260081887245178,
      "learning_rate": 4.1e-06,
      "loss": 0.5753,
      "step": 82
    },
    {
      "epoch": 0.390282131661442,
      "grad_norm": 0.5424107909202576,
      "learning_rate": 4.15e-06,
      "loss": 0.5836,
      "step": 83
    },
    {
      "epoch": 0.3949843260188088,
      "grad_norm": 0.5307353734970093,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 0.5891,
      "step": 84
    },
    {
      "epoch": 0.3996865203761755,
      "grad_norm": 0.5606550574302673,
      "learning_rate": 4.25e-06,
      "loss": 0.6097,
      "step": 85
    },
    {
      "epoch": 0.4043887147335423,
      "grad_norm": 0.5380442142486572,
      "learning_rate": 4.3e-06,
      "loss": 0.6161,
      "step": 86
    },
    {
      "epoch": 0.4090909090909091,
      "grad_norm": 0.7110352516174316,
      "learning_rate": 4.350000000000001e-06,
      "loss": 0.5681,
      "step": 87
    },
    {
      "epoch": 0.41379310344827586,
      "grad_norm": 0.525680422782898,
      "learning_rate": 4.4e-06,
      "loss": 0.5942,
      "step": 88
    },
    {
      "epoch": 0.41849529780564265,
      "grad_norm": 0.5567077398300171,
      "learning_rate": 4.450000000000001e-06,
      "loss": 0.5929,
      "step": 89
    },
    {
      "epoch": 0.4231974921630094,
      "grad_norm": 0.5358735918998718,
      "learning_rate": 4.5e-06,
      "loss": 0.5717,
      "step": 90
    },
    {
      "epoch": 0.4278996865203762,
      "grad_norm": 0.48081454634666443,
      "learning_rate": 4.5500000000000005e-06,
      "loss": 0.5984,
      "step": 91
    },
    {
      "epoch": 0.43260188087774293,
      "grad_norm": 0.4900401830673218,
      "learning_rate": 4.600000000000001e-06,
      "loss": 0.6017,
      "step": 92
    },
    {
      "epoch": 0.4373040752351097,
      "grad_norm": 0.48369699716567993,
      "learning_rate": 4.65e-06,
      "loss": 0.5295,
      "step": 93
    },
    {
      "epoch": 0.44200626959247646,
      "grad_norm": 0.4992128312587738,
      "learning_rate": 4.7e-06,
      "loss": 0.5495,
      "step": 94
    },
    {
      "epoch": 0.44670846394984326,
      "grad_norm": 0.5139058828353882,
      "learning_rate": 4.75e-06,
      "loss": 0.5528,
      "step": 95
    },
    {
      "epoch": 0.45141065830721006,
      "grad_norm": 0.485097199678421,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.5528,
      "step": 96
    },
    {
      "epoch": 0.4561128526645768,
      "grad_norm": 0.5944838523864746,
      "learning_rate": 4.85e-06,
      "loss": 0.5577,
      "step": 97
    },
    {
      "epoch": 0.4608150470219436,
      "grad_norm": 0.5313893556594849,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.5744,
      "step": 98
    },
    {
      "epoch": 0.46551724137931033,
      "grad_norm": 0.49062076210975647,
      "learning_rate": 4.95e-06,
      "loss": 0.5563,
      "step": 99
    },
    {
      "epoch": 0.4702194357366771,
      "grad_norm": 0.5196151733398438,
      "learning_rate": 5e-06,
      "loss": 0.5817,
      "step": 100
    },
    {
      "epoch": 0.47492163009404387,
      "grad_norm": 0.507022500038147,
      "learning_rate": 4.9999910183883085e-06,
      "loss": 0.5608,
      "step": 101
    },
    {
      "epoch": 0.47962382445141066,
      "grad_norm": 0.494018018245697,
      "learning_rate": 4.999964073617768e-06,
      "loss": 0.57,
      "step": 102
    },
    {
      "epoch": 0.4843260188087774,
      "grad_norm": 0.48901236057281494,
      "learning_rate": 4.999919165881985e-06,
      "loss": 0.5568,
      "step": 103
    },
    {
      "epoch": 0.4890282131661442,
      "grad_norm": 0.5166047811508179,
      "learning_rate": 4.999856295503635e-06,
      "loss": 0.5676,
      "step": 104
    },
    {
      "epoch": 0.493730407523511,
      "grad_norm": 0.4952854812145233,
      "learning_rate": 4.9997754629344596e-06,
      "loss": 0.5454,
      "step": 105
    },
    {
      "epoch": 0.49843260188087773,
      "grad_norm": 0.48603224754333496,
      "learning_rate": 4.999676668755263e-06,
      "loss": 0.5351,
      "step": 106
    },
    {
      "epoch": 0.5031347962382445,
      "grad_norm": 0.4922390878200531,
      "learning_rate": 4.999559913675912e-06,
      "loss": 0.5672,
      "step": 107
    },
    {
      "epoch": 0.5078369905956113,
      "grad_norm": 0.4846023619174957,
      "learning_rate": 4.999425198535325e-06,
      "loss": 0.5472,
      "step": 108
    },
    {
      "epoch": 0.512539184952978,
      "grad_norm": 0.4974515736103058,
      "learning_rate": 4.999272524301469e-06,
      "loss": 0.5507,
      "step": 109
    },
    {
      "epoch": 0.5172413793103449,
      "grad_norm": 0.5394945740699768,
      "learning_rate": 4.9991018920713505e-06,
      "loss": 0.5667,
      "step": 110
    },
    {
      "epoch": 0.5219435736677116,
      "grad_norm": 0.5393733978271484,
      "learning_rate": 4.9989133030710154e-06,
      "loss": 0.5505,
      "step": 111
    },
    {
      "epoch": 0.5266457680250783,
      "grad_norm": 0.7595535516738892,
      "learning_rate": 4.9987067586555275e-06,
      "loss": 0.5484,
      "step": 112
    },
    {
      "epoch": 0.5313479623824452,
      "grad_norm": 0.5122578144073486,
      "learning_rate": 4.998482260308969e-06,
      "loss": 0.5669,
      "step": 113
    },
    {
      "epoch": 0.5360501567398119,
      "grad_norm": 0.5233834981918335,
      "learning_rate": 4.998239809644427e-06,
      "loss": 0.5613,
      "step": 114
    },
    {
      "epoch": 0.5407523510971787,
      "grad_norm": 3.08868408203125,
      "learning_rate": 4.9979794084039755e-06,
      "loss": 0.5746,
      "step": 115
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 0.49716249108314514,
      "learning_rate": 4.997701058458677e-06,
      "loss": 0.5349,
      "step": 116
    },
    {
      "epoch": 0.5501567398119123,
      "grad_norm": 0.5552968978881836,
      "learning_rate": 4.997404761808554e-06,
      "loss": 0.569,
      "step": 117
    },
    {
      "epoch": 0.554858934169279,
      "grad_norm": 0.5029860734939575,
      "learning_rate": 4.9970905205825845e-06,
      "loss": 0.5755,
      "step": 118
    },
    {
      "epoch": 0.5595611285266457,
      "grad_norm": 0.5006077885627747,
      "learning_rate": 4.996758337038683e-06,
      "loss": 0.5654,
      "step": 119
    },
    {
      "epoch": 0.5642633228840125,
      "grad_norm": 0.4900127053260803,
      "learning_rate": 4.996408213563684e-06,
      "loss": 0.5575,
      "step": 120
    },
    {
      "epoch": 0.5689655172413793,
      "grad_norm": 0.5142834186553955,
      "learning_rate": 4.996040152673326e-06,
      "loss": 0.5247,
      "step": 121
    },
    {
      "epoch": 0.5736677115987461,
      "grad_norm": 0.5097452998161316,
      "learning_rate": 4.995654157012233e-06,
      "loss": 0.538,
      "step": 122
    },
    {
      "epoch": 0.5783699059561128,
      "grad_norm": 0.5581697225570679,
      "learning_rate": 4.995250229353895e-06,
      "loss": 0.5324,
      "step": 123
    },
    {
      "epoch": 0.5830721003134797,
      "grad_norm": 0.5943841934204102,
      "learning_rate": 4.99482837260065e-06,
      "loss": 0.547,
      "step": 124
    },
    {
      "epoch": 0.5877742946708464,
      "grad_norm": 0.541631281375885,
      "learning_rate": 4.99438858978366e-06,
      "loss": 0.577,
      "step": 125
    },
    {
      "epoch": 0.5924764890282131,
      "grad_norm": 0.6008974313735962,
      "learning_rate": 4.993930884062892e-06,
      "loss": 0.5579,
      "step": 126
    },
    {
      "epoch": 0.5971786833855799,
      "grad_norm": 0.7009158730506897,
      "learning_rate": 4.993455258727094e-06,
      "loss": 0.5579,
      "step": 127
    },
    {
      "epoch": 0.6018808777429467,
      "grad_norm": 0.5942498445510864,
      "learning_rate": 4.992961717193773e-06,
      "loss": 0.5599,
      "step": 128
    },
    {
      "epoch": 0.6065830721003135,
      "grad_norm": 0.4936768710613251,
      "learning_rate": 4.9924502630091655e-06,
      "loss": 0.5676,
      "step": 129
    },
    {
      "epoch": 0.6112852664576802,
      "grad_norm": 0.5035172700881958,
      "learning_rate": 4.99192089984822e-06,
      "loss": 0.5519,
      "step": 130
    },
    {
      "epoch": 0.6159874608150471,
      "grad_norm": 0.8178919553756714,
      "learning_rate": 4.9913736315145614e-06,
      "loss": 0.5594,
      "step": 131
    },
    {
      "epoch": 0.6206896551724138,
      "grad_norm": 0.4977147877216339,
      "learning_rate": 4.990808461940474e-06,
      "loss": 0.54,
      "step": 132
    },
    {
      "epoch": 0.6253918495297806,
      "grad_norm": 0.7395517826080322,
      "learning_rate": 4.990225395186862e-06,
      "loss": 0.5467,
      "step": 133
    },
    {
      "epoch": 0.6300940438871473,
      "grad_norm": 1.1424250602722168,
      "learning_rate": 4.9896244354432314e-06,
      "loss": 0.5432,
      "step": 134
    },
    {
      "epoch": 0.6347962382445141,
      "grad_norm": 0.5002970695495605,
      "learning_rate": 4.98900558702765e-06,
      "loss": 0.5526,
      "step": 135
    },
    {
      "epoch": 0.6394984326018809,
      "grad_norm": 0.49464455246925354,
      "learning_rate": 4.9883688543867225e-06,
      "loss": 0.563,
      "step": 136
    },
    {
      "epoch": 0.6442006269592476,
      "grad_norm": 0.48660922050476074,
      "learning_rate": 4.987714242095558e-06,
      "loss": 0.5328,
      "step": 137
    },
    {
      "epoch": 0.6489028213166145,
      "grad_norm": 0.5086668133735657,
      "learning_rate": 4.9870417548577355e-06,
      "loss": 0.5373,
      "step": 138
    },
    {
      "epoch": 0.6536050156739812,
      "grad_norm": 0.4638510048389435,
      "learning_rate": 4.9863513975052696e-06,
      "loss": 0.5459,
      "step": 139
    },
    {
      "epoch": 0.658307210031348,
      "grad_norm": 0.5742124915122986,
      "learning_rate": 4.985643174998578e-06,
      "loss": 0.5681,
      "step": 140
    },
    {
      "epoch": 0.6630094043887147,
      "grad_norm": 0.5262806415557861,
      "learning_rate": 4.984917092426445e-06,
      "loss": 0.5549,
      "step": 141
    },
    {
      "epoch": 0.6677115987460815,
      "grad_norm": 0.4765531122684479,
      "learning_rate": 4.984173155005982e-06,
      "loss": 0.5435,
      "step": 142
    },
    {
      "epoch": 0.6724137931034483,
      "grad_norm": 0.4731754660606384,
      "learning_rate": 4.983411368082597e-06,
      "loss": 0.5484,
      "step": 143
    },
    {
      "epoch": 0.677115987460815,
      "grad_norm": 0.5015071630477905,
      "learning_rate": 4.982631737129948e-06,
      "loss": 0.5314,
      "step": 144
    },
    {
      "epoch": 0.6818181818181818,
      "grad_norm": 0.5013841986656189,
      "learning_rate": 4.98183426774991e-06,
      "loss": 0.5705,
      "step": 145
    },
    {
      "epoch": 0.6865203761755486,
      "grad_norm": 0.5398444533348083,
      "learning_rate": 4.981018965672529e-06,
      "loss": 0.5713,
      "step": 146
    },
    {
      "epoch": 0.6912225705329154,
      "grad_norm": 0.5247324109077454,
      "learning_rate": 4.98018583675599e-06,
      "loss": 0.5512,
      "step": 147
    },
    {
      "epoch": 0.6959247648902821,
      "grad_norm": 0.488935649394989,
      "learning_rate": 4.979334886986562e-06,
      "loss": 0.5532,
      "step": 148
    },
    {
      "epoch": 0.700626959247649,
      "grad_norm": 0.5056779384613037,
      "learning_rate": 4.978466122478567e-06,
      "loss": 0.5668,
      "step": 149
    },
    {
      "epoch": 0.7053291536050157,
      "grad_norm": 0.5102982521057129,
      "learning_rate": 4.97757954947433e-06,
      "loss": 0.5346,
      "step": 150
    },
    {
      "epoch": 0.7100313479623824,
      "grad_norm": 0.4989202320575714,
      "learning_rate": 4.976675174344132e-06,
      "loss": 0.5488,
      "step": 151
    },
    {
      "epoch": 0.7147335423197492,
      "grad_norm": 0.4996865689754486,
      "learning_rate": 4.975753003586172e-06,
      "loss": 0.528,
      "step": 152
    },
    {
      "epoch": 0.719435736677116,
      "grad_norm": 0.5011430978775024,
      "learning_rate": 4.974813043826513e-06,
      "loss": 0.5221,
      "step": 153
    },
    {
      "epoch": 0.7241379310344828,
      "grad_norm": 0.530082643032074,
      "learning_rate": 4.973855301819039e-06,
      "loss": 0.5398,
      "step": 154
    },
    {
      "epoch": 0.7288401253918495,
      "grad_norm": 0.5243843197822571,
      "learning_rate": 4.972879784445402e-06,
      "loss": 0.5215,
      "step": 155
    },
    {
      "epoch": 0.7335423197492164,
      "grad_norm": 0.5321754217147827,
      "learning_rate": 4.971886498714978e-06,
      "loss": 0.5432,
      "step": 156
    },
    {
      "epoch": 0.7382445141065831,
      "grad_norm": 0.5054258704185486,
      "learning_rate": 4.97087545176481e-06,
      "loss": 0.5197,
      "step": 157
    },
    {
      "epoch": 0.7429467084639498,
      "grad_norm": 0.5209529995918274,
      "learning_rate": 4.9698466508595655e-06,
      "loss": 0.5486,
      "step": 158
    },
    {
      "epoch": 0.7476489028213166,
      "grad_norm": 0.47057369351387024,
      "learning_rate": 4.9688001033914756e-06,
      "loss": 0.5358,
      "step": 159
    },
    {
      "epoch": 0.7523510971786834,
      "grad_norm": 0.5286029577255249,
      "learning_rate": 4.967735816880286e-06,
      "loss": 0.5471,
      "step": 160
    },
    {
      "epoch": 0.7570532915360502,
      "grad_norm": 0.7403479218482971,
      "learning_rate": 4.966653798973205e-06,
      "loss": 0.5264,
      "step": 161
    },
    {
      "epoch": 0.7617554858934169,
      "grad_norm": 0.5945819020271301,
      "learning_rate": 4.965554057444842e-06,
      "loss": 0.5194,
      "step": 162
    },
    {
      "epoch": 0.7664576802507836,
      "grad_norm": 0.48147913813591003,
      "learning_rate": 4.964436600197161e-06,
      "loss": 0.5425,
      "step": 163
    },
    {
      "epoch": 0.7711598746081505,
      "grad_norm": 0.5892655849456787,
      "learning_rate": 4.963301435259413e-06,
      "loss": 0.5134,
      "step": 164
    },
    {
      "epoch": 0.7758620689655172,
      "grad_norm": 0.546521782875061,
      "learning_rate": 4.962148570788088e-06,
      "loss": 0.5334,
      "step": 165
    },
    {
      "epoch": 0.780564263322884,
      "grad_norm": 0.5432559251785278,
      "learning_rate": 4.96097801506685e-06,
      "loss": 0.5214,
      "step": 166
    },
    {
      "epoch": 0.7852664576802508,
      "grad_norm": 0.4892060458660126,
      "learning_rate": 4.959789776506482e-06,
      "loss": 0.5307,
      "step": 167
    },
    {
      "epoch": 0.7899686520376176,
      "grad_norm": 0.5296177268028259,
      "learning_rate": 4.958583863644821e-06,
      "loss": 0.5539,
      "step": 168
    },
    {
      "epoch": 0.7946708463949843,
      "grad_norm": 0.4970422685146332,
      "learning_rate": 4.9573602851466985e-06,
      "loss": 0.5193,
      "step": 169
    },
    {
      "epoch": 0.799373040752351,
      "grad_norm": 0.5201449990272522,
      "learning_rate": 4.9561190498038815e-06,
      "loss": 0.542,
      "step": 170
    },
    {
      "epoch": 0.8040752351097179,
      "grad_norm": 0.6067378520965576,
      "learning_rate": 4.954860166535005e-06,
      "loss": 0.5319,
      "step": 171
    },
    {
      "epoch": 0.8087774294670846,
      "grad_norm": 0.4754677712917328,
      "learning_rate": 4.95358364438551e-06,
      "loss": 0.5366,
      "step": 172
    },
    {
      "epoch": 0.8134796238244514,
      "grad_norm": 0.4970039129257202,
      "learning_rate": 4.952289492527576e-06,
      "loss": 0.5626,
      "step": 173
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 0.5602061748504639,
      "learning_rate": 4.9509777202600605e-06,
      "loss": 0.4951,
      "step": 174
    },
    {
      "epoch": 0.822884012539185,
      "grad_norm": 0.5069930553436279,
      "learning_rate": 4.949648337008425e-06,
      "loss": 0.5419,
      "step": 175
    },
    {
      "epoch": 0.8275862068965517,
      "grad_norm": 2.1186537742614746,
      "learning_rate": 4.948301352324674e-06,
      "loss": 0.5361,
      "step": 176
    },
    {
      "epoch": 0.8322884012539185,
      "grad_norm": 0.4984792470932007,
      "learning_rate": 4.946936775887281e-06,
      "loss": 0.5315,
      "step": 177
    },
    {
      "epoch": 0.8369905956112853,
      "grad_norm": 0.49590349197387695,
      "learning_rate": 4.945554617501124e-06,
      "loss": 0.5258,
      "step": 178
    },
    {
      "epoch": 0.841692789968652,
      "grad_norm": 0.48108625411987305,
      "learning_rate": 4.944154887097411e-06,
      "loss": 0.5566,
      "step": 179
    },
    {
      "epoch": 0.8463949843260188,
      "grad_norm": 0.5167516469955444,
      "learning_rate": 4.942737594733608e-06,
      "loss": 0.5285,
      "step": 180
    },
    {
      "epoch": 0.8510971786833855,
      "grad_norm": 0.4649943709373474,
      "learning_rate": 4.941302750593373e-06,
      "loss": 0.5443,
      "step": 181
    },
    {
      "epoch": 0.8557993730407524,
      "grad_norm": 0.5184140205383301,
      "learning_rate": 4.939850364986475e-06,
      "loss": 0.4859,
      "step": 182
    },
    {
      "epoch": 0.8605015673981191,
      "grad_norm": 0.6746440529823303,
      "learning_rate": 4.938380448348725e-06,
      "loss": 0.4947,
      "step": 183
    },
    {
      "epoch": 0.8652037617554859,
      "grad_norm": 0.5150278806686401,
      "learning_rate": 4.9368930112419e-06,
      "loss": 0.5351,
      "step": 184
    },
    {
      "epoch": 0.8699059561128527,
      "grad_norm": 0.47614121437072754,
      "learning_rate": 4.935388064353665e-06,
      "loss": 0.5362,
      "step": 185
    },
    {
      "epoch": 0.8746081504702194,
      "grad_norm": 0.5044179558753967,
      "learning_rate": 4.9338656184975e-06,
      "loss": 0.532,
      "step": 186
    },
    {
      "epoch": 0.8793103448275862,
      "grad_norm": 0.4887774884700775,
      "learning_rate": 4.932325684612618e-06,
      "loss": 0.5448,
      "step": 187
    },
    {
      "epoch": 0.8840125391849529,
      "grad_norm": 0.49574583768844604,
      "learning_rate": 4.93076827376389e-06,
      "loss": 0.5485,
      "step": 188
    },
    {
      "epoch": 0.8887147335423198,
      "grad_norm": 0.5294933915138245,
      "learning_rate": 4.9291933971417635e-06,
      "loss": 0.5441,
      "step": 189
    },
    {
      "epoch": 0.8934169278996865,
      "grad_norm": 0.4594694972038269,
      "learning_rate": 4.9276010660621835e-06,
      "loss": 0.5322,
      "step": 190
    },
    {
      "epoch": 0.8981191222570533,
      "grad_norm": 0.47507232427597046,
      "learning_rate": 4.925991291966508e-06,
      "loss": 0.5103,
      "step": 191
    },
    {
      "epoch": 0.9028213166144201,
      "grad_norm": 0.4808727204799652,
      "learning_rate": 4.92436408642143e-06,
      "loss": 0.5479,
      "step": 192
    },
    {
      "epoch": 0.9075235109717869,
      "grad_norm": 0.481921523809433,
      "learning_rate": 4.9227194611188934e-06,
      "loss": 0.5242,
      "step": 193
    },
    {
      "epoch": 0.9122257053291536,
      "grad_norm": 0.6387396454811096,
      "learning_rate": 4.921057427876007e-06,
      "loss": 0.4937,
      "step": 194
    },
    {
      "epoch": 0.9169278996865203,
      "grad_norm": 0.5185182094573975,
      "learning_rate": 4.919377998634959e-06,
      "loss": 0.5515,
      "step": 195
    },
    {
      "epoch": 0.9216300940438872,
      "grad_norm": 0.5029557943344116,
      "learning_rate": 4.917681185462934e-06,
      "loss": 0.5371,
      "step": 196
    },
    {
      "epoch": 0.9263322884012539,
      "grad_norm": 0.5311322808265686,
      "learning_rate": 4.915967000552028e-06,
      "loss": 0.5275,
      "step": 197
    },
    {
      "epoch": 0.9310344827586207,
      "grad_norm": 0.46653756499290466,
      "learning_rate": 4.914235456219154e-06,
      "loss": 0.5208,
      "step": 198
    },
    {
      "epoch": 0.9357366771159875,
      "grad_norm": 0.4970494508743286,
      "learning_rate": 4.912486564905959e-06,
      "loss": 0.549,
      "step": 199
    },
    {
      "epoch": 0.9404388714733543,
      "grad_norm": 0.5315521359443665,
      "learning_rate": 4.910720339178735e-06,
      "loss": 0.5314,
      "step": 200
    },
    {
      "epoch": 0.945141065830721,
      "grad_norm": 0.5144534707069397,
      "learning_rate": 4.908936791728323e-06,
      "loss": 0.5372,
      "step": 201
    },
    {
      "epoch": 0.9498432601880877,
      "grad_norm": 0.6076200008392334,
      "learning_rate": 4.907135935370027e-06,
      "loss": 0.5353,
      "step": 202
    },
    {
      "epoch": 0.9545454545454546,
      "grad_norm": 0.5259833931922913,
      "learning_rate": 4.905317783043523e-06,
      "loss": 0.5442,
      "step": 203
    },
    {
      "epoch": 0.9592476489028213,
      "grad_norm": 0.4931281507015228,
      "learning_rate": 4.9034823478127605e-06,
      "loss": 0.5246,
      "step": 204
    },
    {
      "epoch": 0.9639498432601881,
      "grad_norm": 0.4909270405769348,
      "learning_rate": 4.901629642865872e-06,
      "loss": 0.5011,
      "step": 205
    },
    {
      "epoch": 0.9686520376175548,
      "grad_norm": 0.581427276134491,
      "learning_rate": 4.89975968151508e-06,
      "loss": 0.5253,
      "step": 206
    },
    {
      "epoch": 0.9733542319749217,
      "grad_norm": 0.5055932998657227,
      "learning_rate": 4.8978724771965965e-06,
      "loss": 0.5152,
      "step": 207
    },
    {
      "epoch": 0.9780564263322884,
      "grad_norm": 0.5337440371513367,
      "learning_rate": 4.895968043470532e-06,
      "loss": 0.5321,
      "step": 208
    },
    {
      "epoch": 0.9827586206896551,
      "grad_norm": 0.5171258449554443,
      "learning_rate": 4.894046394020794e-06,
      "loss": 0.4973,
      "step": 209
    },
    {
      "epoch": 0.987460815047022,
      "grad_norm": 0.4880523085594177,
      "learning_rate": 4.892107542654988e-06,
      "loss": 0.529,
      "step": 210
    },
    {
      "epoch": 0.9921630094043887,
      "grad_norm": 0.4594219923019409,
      "learning_rate": 4.890151503304325e-06,
      "loss": 0.5502,
      "step": 211
    },
    {
      "epoch": 0.9968652037617555,
      "grad_norm": 0.5071013569831848,
      "learning_rate": 4.88817829002351e-06,
      "loss": 0.5258,
      "step": 212
    }
  ],
  "logging_steps": 1,
  "max_steps": 1272,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 212,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.3638348518765625e+19,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}