{
  "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": 3.1606569290161133,
      "learning_rate": 5.0000000000000004e-08,
      "loss": 1.0072,
      "step": 1
    },
    {
      "epoch": 0.009404388714733543,
      "grad_norm": 3.2058725357055664,
      "learning_rate": 1.0000000000000001e-07,
      "loss": 1.0134,
      "step": 2
    },
    {
      "epoch": 0.014106583072100314,
      "grad_norm": 2.636291265487671,
      "learning_rate": 1.5000000000000002e-07,
      "loss": 0.9635,
      "step": 3
    },
    {
      "epoch": 0.018808777429467086,
      "grad_norm": 2.708746910095215,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 1.0068,
      "step": 4
    },
    {
      "epoch": 0.023510971786833857,
      "grad_norm": 2.8948426246643066,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 0.9608,
      "step": 5
    },
    {
      "epoch": 0.02821316614420063,
      "grad_norm": 2.8740086555480957,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 0.9896,
      "step": 6
    },
    {
      "epoch": 0.032915360501567396,
      "grad_norm": 2.8338170051574707,
      "learning_rate": 3.5000000000000004e-07,
      "loss": 0.9098,
      "step": 7
    },
    {
      "epoch": 0.03761755485893417,
      "grad_norm": 2.7783002853393555,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 0.9733,
      "step": 8
    },
    {
      "epoch": 0.04231974921630094,
      "grad_norm": 3.043574333190918,
      "learning_rate": 4.5000000000000003e-07,
      "loss": 0.9943,
      "step": 9
    },
    {
      "epoch": 0.047021943573667714,
      "grad_norm": 3.142383337020874,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.9475,
      "step": 10
    },
    {
      "epoch": 0.05172413793103448,
      "grad_norm": 2.9817280769348145,
      "learning_rate": 5.5e-07,
      "loss": 0.9701,
      "step": 11
    },
    {
      "epoch": 0.05642633228840126,
      "grad_norm": 2.95699405670166,
      "learning_rate": 6.000000000000001e-07,
      "loss": 0.9983,
      "step": 12
    },
    {
      "epoch": 0.061128526645768025,
      "grad_norm": 2.8782453536987305,
      "learning_rate": 6.5e-07,
      "loss": 0.9502,
      "step": 13
    },
    {
      "epoch": 0.06583072100313479,
      "grad_norm": 2.6715071201324463,
      "learning_rate": 7.000000000000001e-07,
      "loss": 0.9436,
      "step": 14
    },
    {
      "epoch": 0.07053291536050156,
      "grad_norm": 3.869649648666382,
      "learning_rate": 7.5e-07,
      "loss": 0.9692,
      "step": 15
    },
    {
      "epoch": 0.07523510971786834,
      "grad_norm": 3.060220956802368,
      "learning_rate": 8.000000000000001e-07,
      "loss": 0.9258,
      "step": 16
    },
    {
      "epoch": 0.07993730407523511,
      "grad_norm": 2.8922741413116455,
      "learning_rate": 8.500000000000001e-07,
      "loss": 0.9719,
      "step": 17
    },
    {
      "epoch": 0.08463949843260188,
      "grad_norm": 2.7857820987701416,
      "learning_rate": 9.000000000000001e-07,
      "loss": 0.9072,
      "step": 18
    },
    {
      "epoch": 0.08934169278996865,
      "grad_norm": 2.9753293991088867,
      "learning_rate": 9.500000000000001e-07,
      "loss": 0.9032,
      "step": 19
    },
    {
      "epoch": 0.09404388714733543,
      "grad_norm": 2.7989683151245117,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.8887,
      "step": 20
    },
    {
      "epoch": 0.0987460815047022,
      "grad_norm": 2.3953049182891846,
      "learning_rate": 1.0500000000000001e-06,
      "loss": 0.8968,
      "step": 21
    },
    {
      "epoch": 0.10344827586206896,
      "grad_norm": 2.643731117248535,
      "learning_rate": 1.1e-06,
      "loss": 0.8501,
      "step": 22
    },
    {
      "epoch": 0.10815047021943573,
      "grad_norm": 2.3679006099700928,
      "learning_rate": 1.1500000000000002e-06,
      "loss": 0.8476,
      "step": 23
    },
    {
      "epoch": 0.11285266457680251,
      "grad_norm": 2.5935540199279785,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 0.8095,
      "step": 24
    },
    {
      "epoch": 0.11755485893416928,
      "grad_norm": 2.510300636291504,
      "learning_rate": 1.25e-06,
      "loss": 0.8099,
      "step": 25
    },
    {
      "epoch": 0.12225705329153605,
      "grad_norm": 2.372344970703125,
      "learning_rate": 1.3e-06,
      "loss": 0.7869,
      "step": 26
    },
    {
      "epoch": 0.12695924764890282,
      "grad_norm": 2.303426504135132,
      "learning_rate": 1.3500000000000002e-06,
      "loss": 0.7758,
      "step": 27
    },
    {
      "epoch": 0.13166144200626959,
      "grad_norm": 1.9017939567565918,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 0.7498,
      "step": 28
    },
    {
      "epoch": 0.13636363636363635,
      "grad_norm": 1.8810580968856812,
      "learning_rate": 1.45e-06,
      "loss": 0.7878,
      "step": 29
    },
    {
      "epoch": 0.14106583072100312,
      "grad_norm": 1.7797424793243408,
      "learning_rate": 1.5e-06,
      "loss": 0.7747,
      "step": 30
    },
    {
      "epoch": 0.14576802507836992,
      "grad_norm": 1.5053879022598267,
      "learning_rate": 1.5500000000000002e-06,
      "loss": 0.7735,
      "step": 31
    },
    {
      "epoch": 0.15047021943573669,
      "grad_norm": 1.4909234046936035,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 0.7654,
      "step": 32
    },
    {
      "epoch": 0.15517241379310345,
      "grad_norm": 1.36083984375,
      "learning_rate": 1.6500000000000003e-06,
      "loss": 0.6895,
      "step": 33
    },
    {
      "epoch": 0.15987460815047022,
      "grad_norm": 1.536014199256897,
      "learning_rate": 1.7000000000000002e-06,
      "loss": 0.675,
      "step": 34
    },
    {
      "epoch": 0.164576802507837,
      "grad_norm": 1.3426779508590698,
      "learning_rate": 1.75e-06,
      "loss": 0.7652,
      "step": 35
    },
    {
      "epoch": 0.16927899686520376,
      "grad_norm": 1.4900612831115723,
      "learning_rate": 1.8000000000000001e-06,
      "loss": 0.6863,
      "step": 36
    },
    {
      "epoch": 0.17398119122257052,
      "grad_norm": 1.181241750717163,
      "learning_rate": 1.85e-06,
      "loss": 0.7136,
      "step": 37
    },
    {
      "epoch": 0.1786833855799373,
      "grad_norm": 1.461419701576233,
      "learning_rate": 1.9000000000000002e-06,
      "loss": 0.7606,
      "step": 38
    },
    {
      "epoch": 0.1833855799373041,
      "grad_norm": 1.04817795753479,
      "learning_rate": 1.9500000000000004e-06,
      "loss": 0.6829,
      "step": 39
    },
    {
      "epoch": 0.18808777429467086,
      "grad_norm": 1.0499993562698364,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.7144,
      "step": 40
    },
    {
      "epoch": 0.19278996865203762,
      "grad_norm": 0.9935064315795898,
      "learning_rate": 2.05e-06,
      "loss": 0.6736,
      "step": 41
    },
    {
      "epoch": 0.1974921630094044,
      "grad_norm": 0.9919099807739258,
      "learning_rate": 2.1000000000000002e-06,
      "loss": 0.7151,
      "step": 42
    },
    {
      "epoch": 0.20219435736677116,
      "grad_norm": 0.919556200504303,
      "learning_rate": 2.15e-06,
      "loss": 0.6847,
      "step": 43
    },
    {
      "epoch": 0.20689655172413793,
      "grad_norm": 1.4762015342712402,
      "learning_rate": 2.2e-06,
      "loss": 0.6694,
      "step": 44
    },
    {
      "epoch": 0.2115987460815047,
      "grad_norm": 0.9243163466453552,
      "learning_rate": 2.25e-06,
      "loss": 0.6489,
      "step": 45
    },
    {
      "epoch": 0.21630094043887146,
      "grad_norm": 0.7614469528198242,
      "learning_rate": 2.3000000000000004e-06,
      "loss": 0.6568,
      "step": 46
    },
    {
      "epoch": 0.22100313479623823,
      "grad_norm": 0.7543922662734985,
      "learning_rate": 2.35e-06,
      "loss": 0.6359,
      "step": 47
    },
    {
      "epoch": 0.22570532915360503,
      "grad_norm": 0.7558912038803101,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.6231,
      "step": 48
    },
    {
      "epoch": 0.2304075235109718,
      "grad_norm": 0.7822129130363464,
      "learning_rate": 2.4500000000000003e-06,
      "loss": 0.6691,
      "step": 49
    },
    {
      "epoch": 0.23510971786833856,
      "grad_norm": 0.8646999597549438,
      "learning_rate": 2.5e-06,
      "loss": 0.682,
      "step": 50
    },
    {
      "epoch": 0.23981191222570533,
      "grad_norm": 0.8824774622917175,
      "learning_rate": 2.55e-06,
      "loss": 0.6805,
      "step": 51
    },
    {
      "epoch": 0.2445141065830721,
      "grad_norm": 0.7697399258613586,
      "learning_rate": 2.6e-06,
      "loss": 0.6368,
      "step": 52
    },
    {
      "epoch": 0.24921630094043887,
      "grad_norm": 0.6522512435913086,
      "learning_rate": 2.6500000000000005e-06,
      "loss": 0.6367,
      "step": 53
    },
    {
      "epoch": 0.25391849529780564,
      "grad_norm": 0.6172305941581726,
      "learning_rate": 2.7000000000000004e-06,
      "loss": 0.6291,
      "step": 54
    },
    {
      "epoch": 0.25862068965517243,
      "grad_norm": 0.7860460877418518,
      "learning_rate": 2.7500000000000004e-06,
      "loss": 0.6736,
      "step": 55
    },
    {
      "epoch": 0.26332288401253917,
      "grad_norm": 0.6474862694740295,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.6365,
      "step": 56
    },
    {
      "epoch": 0.26802507836990597,
      "grad_norm": 0.6867114901542664,
      "learning_rate": 2.85e-06,
      "loss": 0.6397,
      "step": 57
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 0.7056852579116821,
      "learning_rate": 2.9e-06,
      "loss": 0.6138,
      "step": 58
    },
    {
      "epoch": 0.2774294670846395,
      "grad_norm": 0.6615664958953857,
      "learning_rate": 2.95e-06,
      "loss": 0.6482,
      "step": 59
    },
    {
      "epoch": 0.28213166144200624,
      "grad_norm": 0.6649022102355957,
      "learning_rate": 3e-06,
      "loss": 0.6745,
      "step": 60
    },
    {
      "epoch": 0.28683385579937304,
      "grad_norm": 0.850848913192749,
      "learning_rate": 3.05e-06,
      "loss": 0.5956,
      "step": 61
    },
    {
      "epoch": 0.29153605015673983,
      "grad_norm": 0.5983562469482422,
      "learning_rate": 3.1000000000000004e-06,
      "loss": 0.5894,
      "step": 62
    },
    {
      "epoch": 0.2962382445141066,
      "grad_norm": 0.6286782622337341,
      "learning_rate": 3.1500000000000003e-06,
      "loss": 0.6329,
      "step": 63
    },
    {
      "epoch": 0.30094043887147337,
      "grad_norm": 0.5919945240020752,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.6402,
      "step": 64
    },
    {
      "epoch": 0.3056426332288401,
      "grad_norm": 0.5632765889167786,
      "learning_rate": 3.2500000000000002e-06,
      "loss": 0.5862,
      "step": 65
    },
    {
      "epoch": 0.3103448275862069,
      "grad_norm": 0.7692590951919556,
      "learning_rate": 3.3000000000000006e-06,
      "loss": 0.6031,
      "step": 66
    },
    {
      "epoch": 0.31504702194357365,
      "grad_norm": 0.7313893437385559,
      "learning_rate": 3.3500000000000005e-06,
      "loss": 0.6312,
      "step": 67
    },
    {
      "epoch": 0.31974921630094044,
      "grad_norm": 0.6097120642662048,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 0.5986,
      "step": 68
    },
    {
      "epoch": 0.32445141065830724,
      "grad_norm": 0.5853808522224426,
      "learning_rate": 3.45e-06,
      "loss": 0.5847,
      "step": 69
    },
    {
      "epoch": 0.329153605015674,
      "grad_norm": 0.6093555092811584,
      "learning_rate": 3.5e-06,
      "loss": 0.6552,
      "step": 70
    },
    {
      "epoch": 0.3338557993730408,
      "grad_norm": 0.6106334328651428,
      "learning_rate": 3.5500000000000003e-06,
      "loss": 0.6196,
      "step": 71
    },
    {
      "epoch": 0.3385579937304075,
      "grad_norm": 0.9254828691482544,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 0.6005,
      "step": 72
    },
    {
      "epoch": 0.3432601880877743,
      "grad_norm": 0.5471694469451904,
      "learning_rate": 3.65e-06,
      "loss": 0.5907,
      "step": 73
    },
    {
      "epoch": 0.34796238244514105,
      "grad_norm": 0.6204228401184082,
      "learning_rate": 3.7e-06,
      "loss": 0.6079,
      "step": 74
    },
    {
      "epoch": 0.35266457680250785,
      "grad_norm": 0.52458256483078,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.6001,
      "step": 75
    },
    {
      "epoch": 0.3573667711598746,
      "grad_norm": 0.5356763601303101,
      "learning_rate": 3.8000000000000005e-06,
      "loss": 0.5987,
      "step": 76
    },
    {
      "epoch": 0.3620689655172414,
      "grad_norm": 0.5408467054367065,
      "learning_rate": 3.85e-06,
      "loss": 0.6104,
      "step": 77
    },
    {
      "epoch": 0.3667711598746082,
      "grad_norm": 0.5075871348381042,
      "learning_rate": 3.900000000000001e-06,
      "loss": 0.5569,
      "step": 78
    },
    {
      "epoch": 0.3714733542319749,
      "grad_norm": 0.8474109768867493,
      "learning_rate": 3.95e-06,
      "loss": 0.6195,
      "step": 79
    },
    {
      "epoch": 0.3761755485893417,
      "grad_norm": 0.4750897288322449,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.5399,
      "step": 80
    },
    {
      "epoch": 0.38087774294670845,
      "grad_norm": 0.5082002878189087,
      "learning_rate": 4.05e-06,
      "loss": 0.5997,
      "step": 81
    },
    {
      "epoch": 0.38557993730407525,
      "grad_norm": 0.5343796014785767,
      "learning_rate": 4.1e-06,
      "loss": 0.5704,
      "step": 82
    },
    {
      "epoch": 0.390282131661442,
      "grad_norm": 0.520311713218689,
      "learning_rate": 4.15e-06,
      "loss": 0.5818,
      "step": 83
    },
    {
      "epoch": 0.3949843260188088,
      "grad_norm": 0.5292978286743164,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 0.5852,
      "step": 84
    },
    {
      "epoch": 0.3996865203761755,
      "grad_norm": 0.539886474609375,
      "learning_rate": 4.25e-06,
      "loss": 0.6057,
      "step": 85
    },
    {
      "epoch": 0.4043887147335423,
      "grad_norm": 0.6468827128410339,
      "learning_rate": 4.3e-06,
      "loss": 0.6122,
      "step": 86
    },
    {
      "epoch": 0.4090909090909091,
      "grad_norm": 0.5537365078926086,
      "learning_rate": 4.350000000000001e-06,
      "loss": 0.5652,
      "step": 87
    },
    {
      "epoch": 0.41379310344827586,
      "grad_norm": 0.6226018667221069,
      "learning_rate": 4.4e-06,
      "loss": 0.5884,
      "step": 88
    },
    {
      "epoch": 0.41849529780564265,
      "grad_norm": 0.5016945004463196,
      "learning_rate": 4.450000000000001e-06,
      "loss": 0.5877,
      "step": 89
    },
    {
      "epoch": 0.4231974921630094,
      "grad_norm": 0.5059167146682739,
      "learning_rate": 4.5e-06,
      "loss": 0.5676,
      "step": 90
    },
    {
      "epoch": 0.4278996865203762,
      "grad_norm": 0.47521743178367615,
      "learning_rate": 4.5500000000000005e-06,
      "loss": 0.5929,
      "step": 91
    },
    {
      "epoch": 0.43260188087774293,
      "grad_norm": 0.531306266784668,
      "learning_rate": 4.600000000000001e-06,
      "loss": 0.5983,
      "step": 92
    },
    {
      "epoch": 0.4373040752351097,
      "grad_norm": 0.4965567886829376,
      "learning_rate": 4.65e-06,
      "loss": 0.5279,
      "step": 93
    },
    {
      "epoch": 0.44200626959247646,
      "grad_norm": 0.5125988125801086,
      "learning_rate": 4.7e-06,
      "loss": 0.5436,
      "step": 94
    },
    {
      "epoch": 0.44670846394984326,
      "grad_norm": 0.557763934135437,
      "learning_rate": 4.75e-06,
      "loss": 0.5496,
      "step": 95
    },
    {
      "epoch": 0.45141065830721006,
      "grad_norm": 0.6993274092674255,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.5498,
      "step": 96
    },
    {
      "epoch": 0.4561128526645768,
      "grad_norm": 0.5485453009605408,
      "learning_rate": 4.85e-06,
      "loss": 0.5552,
      "step": 97
    },
    {
      "epoch": 0.4608150470219436,
      "grad_norm": 1.9821522235870361,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.569,
      "step": 98
    },
    {
      "epoch": 0.46551724137931033,
      "grad_norm": 0.6074144840240479,
      "learning_rate": 4.95e-06,
      "loss": 0.5546,
      "step": 99
    },
    {
      "epoch": 0.4702194357366771,
      "grad_norm": 0.5404040813446045,
      "learning_rate": 5e-06,
      "loss": 0.5775,
      "step": 100
    },
    {
      "epoch": 0.47492163009404387,
      "grad_norm": 0.500438928604126,
      "learning_rate": 4.9999910183883085e-06,
      "loss": 0.5569,
      "step": 101
    },
    {
      "epoch": 0.47962382445141066,
      "grad_norm": 0.5036981701850891,
      "learning_rate": 4.999964073617768e-06,
      "loss": 0.5663,
      "step": 102
    },
    {
      "epoch": 0.4843260188087774,
      "grad_norm": 0.4537642300128937,
      "learning_rate": 4.999919165881985e-06,
      "loss": 0.5527,
      "step": 103
    },
    {
      "epoch": 0.4890282131661442,
      "grad_norm": 0.49653521180152893,
      "learning_rate": 4.999856295503635e-06,
      "loss": 0.563,
      "step": 104
    },
    {
      "epoch": 0.493730407523511,
      "grad_norm": 0.46847566962242126,
      "learning_rate": 4.9997754629344596e-06,
      "loss": 0.5425,
      "step": 105
    },
    {
      "epoch": 0.49843260188087773,
      "grad_norm": 0.5192411541938782,
      "learning_rate": 4.999676668755263e-06,
      "loss": 0.5315,
      "step": 106
    },
    {
      "epoch": 0.5031347962382445,
      "grad_norm": 0.5170287489891052,
      "learning_rate": 4.999559913675912e-06,
      "loss": 0.5627,
      "step": 107
    },
    {
      "epoch": 0.5078369905956113,
      "grad_norm": 0.47297438979148865,
      "learning_rate": 4.999425198535325e-06,
      "loss": 0.5432,
      "step": 108
    },
    {
      "epoch": 0.512539184952978,
      "grad_norm": 0.4873776137828827,
      "learning_rate": 4.999272524301469e-06,
      "loss": 0.5473,
      "step": 109
    },
    {
      "epoch": 0.5172413793103449,
      "grad_norm": 0.5432935357093811,
      "learning_rate": 4.9991018920713505e-06,
      "loss": 0.5642,
      "step": 110
    },
    {
      "epoch": 0.5219435736677116,
      "grad_norm": 0.4850105345249176,
      "learning_rate": 4.9989133030710154e-06,
      "loss": 0.548,
      "step": 111
    },
    {
      "epoch": 0.5266457680250783,
      "grad_norm": 0.9399585723876953,
      "learning_rate": 4.9987067586555275e-06,
      "loss": 0.5471,
      "step": 112
    },
    {
      "epoch": 0.5313479623824452,
      "grad_norm": 0.5167811512947083,
      "learning_rate": 4.998482260308969e-06,
      "loss": 0.5648,
      "step": 113
    },
    {
      "epoch": 0.5360501567398119,
      "grad_norm": 0.5069029927253723,
      "learning_rate": 4.998239809644427e-06,
      "loss": 0.5568,
      "step": 114
    },
    {
      "epoch": 0.5407523510971787,
      "grad_norm": 0.8738563656806946,
      "learning_rate": 4.9979794084039755e-06,
      "loss": 0.5719,
      "step": 115
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 0.5216553807258606,
      "learning_rate": 4.997701058458677e-06,
      "loss": 0.5309,
      "step": 116
    },
    {
      "epoch": 0.5501567398119123,
      "grad_norm": 0.9678344130516052,
      "learning_rate": 4.997404761808554e-06,
      "loss": 0.5645,
      "step": 117
    },
    {
      "epoch": 0.554858934169279,
      "grad_norm": 0.496598482131958,
      "learning_rate": 4.9970905205825845e-06,
      "loss": 0.5711,
      "step": 118
    },
    {
      "epoch": 0.5595611285266457,
      "grad_norm": 0.4745199680328369,
      "learning_rate": 4.996758337038683e-06,
      "loss": 0.5613,
      "step": 119
    },
    {
      "epoch": 0.5642633228840125,
      "grad_norm": 0.5595977902412415,
      "learning_rate": 4.996408213563684e-06,
      "loss": 0.5559,
      "step": 120
    },
    {
      "epoch": 0.5689655172413793,
      "grad_norm": 0.4743712544441223,
      "learning_rate": 4.996040152673326e-06,
      "loss": 0.5228,
      "step": 121
    },
    {
      "epoch": 0.5736677115987461,
      "grad_norm": 0.5418100953102112,
      "learning_rate": 4.995654157012233e-06,
      "loss": 0.536,
      "step": 122
    },
    {
      "epoch": 0.5783699059561128,
      "grad_norm": 0.521977424621582,
      "learning_rate": 4.995250229353895e-06,
      "loss": 0.5305,
      "step": 123
    },
    {
      "epoch": 0.5830721003134797,
      "grad_norm": 0.5062761902809143,
      "learning_rate": 4.99482837260065e-06,
      "loss": 0.5417,
      "step": 124
    },
    {
      "epoch": 0.5877742946708464,
      "grad_norm": 0.5895913243293762,
      "learning_rate": 4.99438858978366e-06,
      "loss": 0.573,
      "step": 125
    },
    {
      "epoch": 0.5924764890282131,
      "grad_norm": 0.5442466139793396,
      "learning_rate": 4.993930884062892e-06,
      "loss": 0.5563,
      "step": 126
    },
    {
      "epoch": 0.5971786833855799,
      "grad_norm": 0.5130571722984314,
      "learning_rate": 4.993455258727094e-06,
      "loss": 0.5549,
      "step": 127
    },
    {
      "epoch": 0.6018808777429467,
      "grad_norm": 0.5579081773757935,
      "learning_rate": 4.992961717193773e-06,
      "loss": 0.5554,
      "step": 128
    },
    {
      "epoch": 0.6065830721003135,
      "grad_norm": 0.6375890374183655,
      "learning_rate": 4.9924502630091655e-06,
      "loss": 0.5626,
      "step": 129
    },
    {
      "epoch": 0.6112852664576802,
      "grad_norm": 0.5129190683364868,
      "learning_rate": 4.99192089984822e-06,
      "loss": 0.5493,
      "step": 130
    },
    {
      "epoch": 0.6159874608150471,
      "grad_norm": 0.5293419361114502,
      "learning_rate": 4.9913736315145614e-06,
      "loss": 0.5565,
      "step": 131
    },
    {
      "epoch": 0.6206896551724138,
      "grad_norm": 0.6502572298049927,
      "learning_rate": 4.990808461940474e-06,
      "loss": 0.5358,
      "step": 132
    },
    {
      "epoch": 0.6253918495297806,
      "grad_norm": 0.5450296998023987,
      "learning_rate": 4.990225395186862e-06,
      "loss": 0.5421,
      "step": 133
    },
    {
      "epoch": 0.6300940438871473,
      "grad_norm": 0.45506399869918823,
      "learning_rate": 4.9896244354432314e-06,
      "loss": 0.5396,
      "step": 134
    },
    {
      "epoch": 0.6347962382445141,
      "grad_norm": 0.5095545649528503,
      "learning_rate": 4.98900558702765e-06,
      "loss": 0.5486,
      "step": 135
    },
    {
      "epoch": 0.6394984326018809,
      "grad_norm": 0.4836446940898895,
      "learning_rate": 4.9883688543867225e-06,
      "loss": 0.5596,
      "step": 136
    },
    {
      "epoch": 0.6442006269592476,
      "grad_norm": 0.5253512859344482,
      "learning_rate": 4.987714242095558e-06,
      "loss": 0.5308,
      "step": 137
    },
    {
      "epoch": 0.6489028213166145,
      "grad_norm": 0.8280164003372192,
      "learning_rate": 4.9870417548577355e-06,
      "loss": 0.5349,
      "step": 138
    },
    {
      "epoch": 0.6536050156739812,
      "grad_norm": 0.4729730188846588,
      "learning_rate": 4.9863513975052696e-06,
      "loss": 0.5416,
      "step": 139
    },
    {
      "epoch": 0.658307210031348,
      "grad_norm": 0.5932718515396118,
      "learning_rate": 4.985643174998578e-06,
      "loss": 0.5638,
      "step": 140
    },
    {
      "epoch": 0.6630094043887147,
      "grad_norm": 0.5187026262283325,
      "learning_rate": 4.984917092426445e-06,
      "loss": 0.5507,
      "step": 141
    },
    {
      "epoch": 0.6677115987460815,
      "grad_norm": 0.5024245977401733,
      "learning_rate": 4.984173155005982e-06,
      "loss": 0.5406,
      "step": 142
    },
    {
      "epoch": 0.6724137931034483,
      "grad_norm": 0.4735509157180786,
      "learning_rate": 4.983411368082597e-06,
      "loss": 0.5431,
      "step": 143
    },
    {
      "epoch": 0.677115987460815,
      "grad_norm": 0.5040024518966675,
      "learning_rate": 4.982631737129948e-06,
      "loss": 0.5291,
      "step": 144
    },
    {
      "epoch": 0.6818181818181818,
      "grad_norm": 0.47764894366264343,
      "learning_rate": 4.98183426774991e-06,
      "loss": 0.5677,
      "step": 145
    },
    {
      "epoch": 0.6865203761755486,
      "grad_norm": 0.5211489796638489,
      "learning_rate": 4.981018965672529e-06,
      "loss": 0.566,
      "step": 146
    },
    {
      "epoch": 0.6912225705329154,
      "grad_norm": 1.022007942199707,
      "learning_rate": 4.98018583675599e-06,
      "loss": 0.5476,
      "step": 147
    },
    {
      "epoch": 0.6959247648902821,
      "grad_norm": 0.5263912677764893,
      "learning_rate": 4.979334886986562e-06,
      "loss": 0.5473,
      "step": 148
    },
    {
      "epoch": 0.700626959247649,
      "grad_norm": 0.5014091730117798,
      "learning_rate": 4.978466122478567e-06,
      "loss": 0.5642,
      "step": 149
    },
    {
      "epoch": 0.7053291536050157,
      "grad_norm": 0.5003350973129272,
      "learning_rate": 4.97757954947433e-06,
      "loss": 0.5311,
      "step": 150
    },
    {
      "epoch": 0.7100313479623824,
      "grad_norm": 0.5010690093040466,
      "learning_rate": 4.976675174344132e-06,
      "loss": 0.5469,
      "step": 151
    },
    {
      "epoch": 0.7147335423197492,
      "grad_norm": 0.45779237151145935,
      "learning_rate": 4.975753003586172e-06,
      "loss": 0.5273,
      "step": 152
    },
    {
      "epoch": 0.719435736677116,
      "grad_norm": 0.6231539845466614,
      "learning_rate": 4.974813043826513e-06,
      "loss": 0.5182,
      "step": 153
    },
    {
      "epoch": 0.7241379310344828,
      "grad_norm": 0.5361394286155701,
      "learning_rate": 4.973855301819039e-06,
      "loss": 0.5372,
      "step": 154
    },
    {
      "epoch": 0.7288401253918495,
      "grad_norm": 0.5193538665771484,
      "learning_rate": 4.972879784445402e-06,
      "loss": 0.5201,
      "step": 155
    },
    {
      "epoch": 0.7335423197492164,
      "grad_norm": 0.47956809401512146,
      "learning_rate": 4.971886498714978e-06,
      "loss": 0.5402,
      "step": 156
    },
    {
      "epoch": 0.7382445141065831,
      "grad_norm": 0.5303016901016235,
      "learning_rate": 4.97087545176481e-06,
      "loss": 0.5174,
      "step": 157
    },
    {
      "epoch": 0.7429467084639498,
      "grad_norm": 0.5002286434173584,
      "learning_rate": 4.9698466508595655e-06,
      "loss": 0.5453,
      "step": 158
    },
    {
      "epoch": 0.7476489028213166,
      "grad_norm": 0.6070297360420227,
      "learning_rate": 4.9688001033914756e-06,
      "loss": 0.5327,
      "step": 159
    },
    {
      "epoch": 0.7523510971786834,
      "grad_norm": 0.5436793565750122,
      "learning_rate": 4.967735816880286e-06,
      "loss": 0.544,
      "step": 160
    },
    {
      "epoch": 0.7570532915360502,
      "grad_norm": 0.538012683391571,
      "learning_rate": 4.966653798973205e-06,
      "loss": 0.5233,
      "step": 161
    },
    {
      "epoch": 0.7617554858934169,
      "grad_norm": 0.4916169345378876,
      "learning_rate": 4.965554057444842e-06,
      "loss": 0.5168,
      "step": 162
    },
    {
      "epoch": 0.7664576802507836,
      "grad_norm": 0.48281437158584595,
      "learning_rate": 4.964436600197161e-06,
      "loss": 0.5393,
      "step": 163
    },
    {
      "epoch": 0.7711598746081505,
      "grad_norm": 0.5184990167617798,
      "learning_rate": 4.963301435259413e-06,
      "loss": 0.5085,
      "step": 164
    },
    {
      "epoch": 0.7758620689655172,
      "grad_norm": 0.4706438183784485,
      "learning_rate": 4.962148570788088e-06,
      "loss": 0.5299,
      "step": 165
    },
    {
      "epoch": 0.780564263322884,
      "grad_norm": 0.6550764441490173,
      "learning_rate": 4.96097801506685e-06,
      "loss": 0.5192,
      "step": 166
    },
    {
      "epoch": 0.7852664576802508,
      "grad_norm": 0.5386581420898438,
      "learning_rate": 4.959789776506482e-06,
      "loss": 0.5258,
      "step": 167
    },
    {
      "epoch": 0.7899686520376176,
      "grad_norm": 0.5060779452323914,
      "learning_rate": 4.958583863644821e-06,
      "loss": 0.5512,
      "step": 168
    },
    {
      "epoch": 0.7946708463949843,
      "grad_norm": 0.47050032019615173,
      "learning_rate": 4.9573602851466985e-06,
      "loss": 0.5176,
      "step": 169
    },
    {
      "epoch": 0.799373040752351,
      "grad_norm": 7.3139567375183105,
      "learning_rate": 4.9561190498038815e-06,
      "loss": 0.5381,
      "step": 170
    },
    {
      "epoch": 0.8040752351097179,
      "grad_norm": 0.620528519153595,
      "learning_rate": 4.954860166535005e-06,
      "loss": 0.5299,
      "step": 171
    },
    {
      "epoch": 0.8087774294670846,
      "grad_norm": 0.45067766308784485,
      "learning_rate": 4.95358364438551e-06,
      "loss": 0.5328,
      "step": 172
    },
    {
      "epoch": 0.8134796238244514,
      "grad_norm": 0.6771508455276489,
      "learning_rate": 4.952289492527576e-06,
      "loss": 0.5601,
      "step": 173
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 0.518925130367279,
      "learning_rate": 4.9509777202600605e-06,
      "loss": 0.494,
      "step": 174
    },
    {
      "epoch": 0.822884012539185,
      "grad_norm": 0.5191988945007324,
      "learning_rate": 4.949648337008425e-06,
      "loss": 0.5425,
      "step": 175
    },
    {
      "epoch": 0.8275862068965517,
      "grad_norm": 0.8600963354110718,
      "learning_rate": 4.948301352324674e-06,
      "loss": 0.5332,
      "step": 176
    },
    {
      "epoch": 0.8322884012539185,
      "grad_norm": 0.5405915379524231,
      "learning_rate": 4.946936775887281e-06,
      "loss": 0.5276,
      "step": 177
    },
    {
      "epoch": 0.8369905956112853,
      "grad_norm": 0.48730772733688354,
      "learning_rate": 4.945554617501124e-06,
      "loss": 0.5217,
      "step": 178
    },
    {
      "epoch": 0.841692789968652,
      "grad_norm": 0.5092865824699402,
      "learning_rate": 4.944154887097411e-06,
      "loss": 0.5534,
      "step": 179
    },
    {
      "epoch": 0.8463949843260188,
      "grad_norm": 0.4994933605194092,
      "learning_rate": 4.942737594733608e-06,
      "loss": 0.5242,
      "step": 180
    },
    {
      "epoch": 0.8510971786833855,
      "grad_norm": 0.4554043412208557,
      "learning_rate": 4.941302750593373e-06,
      "loss": 0.5424,
      "step": 181
    },
    {
      "epoch": 0.8557993730407524,
      "grad_norm": 0.4865265488624573,
      "learning_rate": 4.939850364986475e-06,
      "loss": 0.482,
      "step": 182
    },
    {
      "epoch": 0.8605015673981191,
      "grad_norm": 0.5013875365257263,
      "learning_rate": 4.938380448348725e-06,
      "loss": 0.4908,
      "step": 183
    },
    {
      "epoch": 0.8652037617554859,
      "grad_norm": 0.4997917115688324,
      "learning_rate": 4.9368930112419e-06,
      "loss": 0.5336,
      "step": 184
    },
    {
      "epoch": 0.8699059561128527,
      "grad_norm": 0.4783482551574707,
      "learning_rate": 4.935388064353665e-06,
      "loss": 0.5338,
      "step": 185
    },
    {
      "epoch": 0.8746081504702194,
      "grad_norm": 0.7221089005470276,
      "learning_rate": 4.9338656184975e-06,
      "loss": 0.5327,
      "step": 186
    },
    {
      "epoch": 0.8793103448275862,
      "grad_norm": 0.48115843534469604,
      "learning_rate": 4.932325684612618e-06,
      "loss": 0.5408,
      "step": 187
    },
    {
      "epoch": 0.8840125391849529,
      "grad_norm": 0.4940219223499298,
      "learning_rate": 4.93076827376389e-06,
      "loss": 0.5455,
      "step": 188
    },
    {
      "epoch": 0.8887147335423198,
      "grad_norm": 0.4754747450351715,
      "learning_rate": 4.9291933971417635e-06,
      "loss": 0.542,
      "step": 189
    },
    {
      "epoch": 0.8934169278996865,
      "grad_norm": 0.548713207244873,
      "learning_rate": 4.9276010660621835e-06,
      "loss": 0.5292,
      "step": 190
    },
    {
      "epoch": 0.8981191222570533,
      "grad_norm": 0.7292612195014954,
      "learning_rate": 4.925991291966508e-06,
      "loss": 0.5073,
      "step": 191
    },
    {
      "epoch": 0.9028213166144201,
      "grad_norm": 0.5254770517349243,
      "learning_rate": 4.92436408642143e-06,
      "loss": 0.5451,
      "step": 192
    },
    {
      "epoch": 0.9075235109717869,
      "grad_norm": 0.47938767075538635,
      "learning_rate": 4.9227194611188934e-06,
      "loss": 0.5204,
      "step": 193
    },
    {
      "epoch": 0.9122257053291536,
      "grad_norm": 0.6740232706069946,
      "learning_rate": 4.921057427876007e-06,
      "loss": 0.4928,
      "step": 194
    },
    {
      "epoch": 0.9169278996865203,
      "grad_norm": 0.5455343723297119,
      "learning_rate": 4.919377998634959e-06,
      "loss": 0.5468,
      "step": 195
    },
    {
      "epoch": 0.9216300940438872,
      "grad_norm": 0.5001958012580872,
      "learning_rate": 4.917681185462934e-06,
      "loss": 0.5339,
      "step": 196
    },
    {
      "epoch": 0.9263322884012539,
      "grad_norm": 0.5084257125854492,
      "learning_rate": 4.915967000552028e-06,
      "loss": 0.5259,
      "step": 197
    },
    {
      "epoch": 0.9310344827586207,
      "grad_norm": 0.4807164967060089,
      "learning_rate": 4.914235456219154e-06,
      "loss": 0.5204,
      "step": 198
    },
    {
      "epoch": 0.9357366771159875,
      "grad_norm": 0.6099370718002319,
      "learning_rate": 4.912486564905959e-06,
      "loss": 0.544,
      "step": 199
    },
    {
      "epoch": 0.9404388714733543,
      "grad_norm": 0.47461947798728943,
      "learning_rate": 4.910720339178735e-06,
      "loss": 0.5295,
      "step": 200
    },
    {
      "epoch": 0.945141065830721,
      "grad_norm": 0.500136137008667,
      "learning_rate": 4.908936791728323e-06,
      "loss": 0.5321,
      "step": 201
    },
    {
      "epoch": 0.9498432601880877,
      "grad_norm": 0.5235631465911865,
      "learning_rate": 4.907135935370027e-06,
      "loss": 0.5338,
      "step": 202
    },
    {
      "epoch": 0.9545454545454546,
      "grad_norm": 0.9285804629325867,
      "learning_rate": 4.905317783043523e-06,
      "loss": 0.5393,
      "step": 203
    },
    {
      "epoch": 0.9592476489028213,
      "grad_norm": 0.4834178388118744,
      "learning_rate": 4.9034823478127605e-06,
      "loss": 0.5211,
      "step": 204
    },
    {
      "epoch": 0.9639498432601881,
      "grad_norm": 0.4830580949783325,
      "learning_rate": 4.901629642865872e-06,
      "loss": 0.4986,
      "step": 205
    },
    {
      "epoch": 0.9686520376175548,
      "grad_norm": 0.49718615412712097,
      "learning_rate": 4.89975968151508e-06,
      "loss": 0.5204,
      "step": 206
    },
    {
      "epoch": 0.9733542319749217,
      "grad_norm": 0.5056726336479187,
      "learning_rate": 4.8978724771965965e-06,
      "loss": 0.5133,
      "step": 207
    },
    {
      "epoch": 0.9780564263322884,
      "grad_norm": 0.7357563376426697,
      "learning_rate": 4.895968043470532e-06,
      "loss": 0.5307,
      "step": 208
    },
    {
      "epoch": 0.9827586206896551,
      "grad_norm": 0.515610933303833,
      "learning_rate": 4.894046394020794e-06,
      "loss": 0.4955,
      "step": 209
    },
    {
      "epoch": 0.987460815047022,
      "grad_norm": 0.5124618411064148,
      "learning_rate": 4.892107542654988e-06,
      "loss": 0.526,
      "step": 210
    },
    {
      "epoch": 0.9921630094043887,
      "grad_norm": 0.5059565901756287,
      "learning_rate": 4.890151503304325e-06,
      "loss": 0.5473,
      "step": 211
    },
    {
      "epoch": 0.9968652037617555,
      "grad_norm": 0.4806717336177826,
      "learning_rate": 4.88817829002351e-06,
      "loss": 0.5212,
      "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
}