{
  "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": 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
    },
    {
      "epoch": 1.0047021943573669,
      "grad_norm": 0.9454345703125,
      "learning_rate": 4.886187916990653e-06,
      "loss": 1.0566,
      "step": 213
    },
    {
      "epoch": 1.0094043887147335,
      "grad_norm": 0.4871070086956024,
      "learning_rate": 4.884180398507163e-06,
      "loss": 0.503,
      "step": 214
    },
    {
      "epoch": 1.0141065830721003,
      "grad_norm": 0.45102012157440186,
      "learning_rate": 4.882155748997636e-06,
      "loss": 0.4954,
      "step": 215
    },
    {
      "epoch": 1.0188087774294672,
      "grad_norm": 0.49910685420036316,
      "learning_rate": 4.8801139830097685e-06,
      "loss": 0.5019,
      "step": 216
    },
    {
      "epoch": 1.0235109717868338,
      "grad_norm": 0.5155763030052185,
      "learning_rate": 4.878055115214238e-06,
      "loss": 0.5102,
      "step": 217
    },
    {
      "epoch": 1.0282131661442007,
      "grad_norm": 0.4567059874534607,
      "learning_rate": 4.875979160404607e-06,
      "loss": 0.5069,
      "step": 218
    },
    {
      "epoch": 1.0329153605015673,
      "grad_norm": 0.4782896935939789,
      "learning_rate": 4.873886133497209e-06,
      "loss": 0.5182,
      "step": 219
    },
    {
      "epoch": 1.0376175548589341,
      "grad_norm": 0.44995731115341187,
      "learning_rate": 4.87177604953105e-06,
      "loss": 0.513,
      "step": 220
    },
    {
      "epoch": 1.042319749216301,
      "grad_norm": 0.470059871673584,
      "learning_rate": 4.869648923667694e-06,
      "loss": 0.468,
      "step": 221
    },
    {
      "epoch": 1.0470219435736676,
      "grad_norm": 0.5356128215789795,
      "learning_rate": 4.867504771191154e-06,
      "loss": 0.4942,
      "step": 222
    },
    {
      "epoch": 1.0517241379310345,
      "grad_norm": 0.5137870907783508,
      "learning_rate": 4.865343607507788e-06,
      "loss": 0.5022,
      "step": 223
    },
    {
      "epoch": 1.0564263322884013,
      "grad_norm": 0.47419992089271545,
      "learning_rate": 4.86316544814618e-06,
      "loss": 0.5158,
      "step": 224
    },
    {
      "epoch": 1.061128526645768,
      "grad_norm": 0.49087393283843994,
      "learning_rate": 4.860970308757038e-06,
      "loss": 0.4605,
      "step": 225
    },
    {
      "epoch": 1.0658307210031348,
      "grad_norm": 0.4988348186016083,
      "learning_rate": 4.858758205113072e-06,
      "loss": 0.4912,
      "step": 226
    },
    {
      "epoch": 1.0705329153605017,
      "grad_norm": 0.44543248414993286,
      "learning_rate": 4.856529153108888e-06,
      "loss": 0.524,
      "step": 227
    },
    {
      "epoch": 1.0752351097178683,
      "grad_norm": 0.5953351259231567,
      "learning_rate": 4.854283168760868e-06,
      "loss": 0.5001,
      "step": 228
    },
    {
      "epoch": 1.0799373040752351,
      "grad_norm": 0.5012004375457764,
      "learning_rate": 4.85202026820706e-06,
      "loss": 0.4968,
      "step": 229
    },
    {
      "epoch": 1.084639498432602,
      "grad_norm": 0.5023937821388245,
      "learning_rate": 4.84974046770706e-06,
      "loss": 0.5345,
      "step": 230
    },
    {
      "epoch": 1.0893416927899686,
      "grad_norm": 0.4705684185028076,
      "learning_rate": 4.847443783641893e-06,
      "loss": 0.4459,
      "step": 231
    },
    {
      "epoch": 1.0940438871473355,
      "grad_norm": 0.5082476735115051,
      "learning_rate": 4.845130232513901e-06,
      "loss": 0.4905,
      "step": 232
    },
    {
      "epoch": 1.098746081504702,
      "grad_norm": 0.5283995866775513,
      "learning_rate": 4.842799830946615e-06,
      "loss": 0.4878,
      "step": 233
    },
    {
      "epoch": 1.103448275862069,
      "grad_norm": 0.6373623013496399,
      "learning_rate": 4.840452595684646e-06,
      "loss": 0.4867,
      "step": 234
    },
    {
      "epoch": 1.1081504702194358,
      "grad_norm": 0.4624481201171875,
      "learning_rate": 4.83808854359356e-06,
      "loss": 0.4793,
      "step": 235
    },
    {
      "epoch": 1.1128526645768024,
      "grad_norm": 0.4659098982810974,
      "learning_rate": 4.835707691659753e-06,
      "loss": 0.4827,
      "step": 236
    },
    {
      "epoch": 1.1175548589341693,
      "grad_norm": 0.4920850396156311,
      "learning_rate": 4.8333100569903365e-06,
      "loss": 0.4932,
      "step": 237
    },
    {
      "epoch": 1.1222570532915361,
      "grad_norm": 0.492286741733551,
      "learning_rate": 4.8308956568130094e-06,
      "loss": 0.5144,
      "step": 238
    },
    {
      "epoch": 1.1269592476489028,
      "grad_norm": 0.5429807901382446,
      "learning_rate": 4.828464508475934e-06,
      "loss": 0.5054,
      "step": 239
    },
    {
      "epoch": 1.1316614420062696,
      "grad_norm": 2.4671998023986816,
      "learning_rate": 4.826016629447616e-06,
      "loss": 0.5073,
      "step": 240
    },
    {
      "epoch": 1.1363636363636362,
      "grad_norm": 0.4593118131160736,
      "learning_rate": 4.823552037316775e-06,
      "loss": 0.4856,
      "step": 241
    },
    {
      "epoch": 1.141065830721003,
      "grad_norm": 0.6855646371841431,
      "learning_rate": 4.821070749792218e-06,
      "loss": 0.5388,
      "step": 242
    },
    {
      "epoch": 1.14576802507837,
      "grad_norm": 0.5722374320030212,
      "learning_rate": 4.818572784702713e-06,
      "loss": 0.51,
      "step": 243
    },
    {
      "epoch": 1.1504702194357366,
      "grad_norm": 0.4901357591152191,
      "learning_rate": 4.816058159996863e-06,
      "loss": 0.5201,
      "step": 244
    },
    {
      "epoch": 1.1551724137931034,
      "grad_norm": 0.4655209481716156,
      "learning_rate": 4.813526893742972e-06,
      "loss": 0.501,
      "step": 245
    },
    {
      "epoch": 1.1598746081504703,
      "grad_norm": 0.7608394622802734,
      "learning_rate": 4.810979004128924e-06,
      "loss": 0.4961,
      "step": 246
    },
    {
      "epoch": 1.164576802507837,
      "grad_norm": 0.4857081472873688,
      "learning_rate": 4.808414509462042e-06,
      "loss": 0.5174,
      "step": 247
    },
    {
      "epoch": 1.1692789968652038,
      "grad_norm": 0.46672946214675903,
      "learning_rate": 4.80583342816896e-06,
      "loss": 0.484,
      "step": 248
    },
    {
      "epoch": 1.1739811912225706,
      "grad_norm": 0.46982088685035706,
      "learning_rate": 4.803235778795496e-06,
      "loss": 0.5236,
      "step": 249
    },
    {
      "epoch": 1.1786833855799372,
      "grad_norm": 0.5086098909378052,
      "learning_rate": 4.800621580006511e-06,
      "loss": 0.4673,
      "step": 250
    },
    {
      "epoch": 1.183385579937304,
      "grad_norm": 0.45968860387802124,
      "learning_rate": 4.797990850585782e-06,
      "loss": 0.5151,
      "step": 251
    },
    {
      "epoch": 1.188087774294671,
      "grad_norm": 0.49544984102249146,
      "learning_rate": 4.79534360943586e-06,
      "loss": 0.494,
      "step": 252
    },
    {
      "epoch": 1.1927899686520376,
      "grad_norm": 0.531892716884613,
      "learning_rate": 4.792679875577937e-06,
      "loss": 0.4778,
      "step": 253
    },
    {
      "epoch": 1.1974921630094044,
      "grad_norm": 0.5013542175292969,
      "learning_rate": 4.789999668151714e-06,
      "loss": 0.5132,
      "step": 254
    },
    {
      "epoch": 1.2021943573667713,
      "grad_norm": 0.46963250637054443,
      "learning_rate": 4.7873030064152545e-06,
      "loss": 0.4938,
      "step": 255
    },
    {
      "epoch": 1.206896551724138,
      "grad_norm": 0.465285986661911,
      "learning_rate": 4.784589909744856e-06,
      "loss": 0.4898,
      "step": 256
    },
    {
      "epoch": 1.2115987460815048,
      "grad_norm": 0.5183936357498169,
      "learning_rate": 4.7818603976349005e-06,
      "loss": 0.5004,
      "step": 257
    },
    {
      "epoch": 1.2163009404388714,
      "grad_norm": 0.47324836254119873,
      "learning_rate": 4.779114489697724e-06,
      "loss": 0.4972,
      "step": 258
    },
    {
      "epoch": 1.2210031347962382,
      "grad_norm": 0.5208264589309692,
      "learning_rate": 4.776352205663469e-06,
      "loss": 0.5023,
      "step": 259
    },
    {
      "epoch": 1.225705329153605,
      "grad_norm": 0.5583804845809937,
      "learning_rate": 4.773573565379947e-06,
      "loss": 0.5099,
      "step": 260
    },
    {
      "epoch": 1.2304075235109717,
      "grad_norm": 0.5016160011291504,
      "learning_rate": 4.770778588812489e-06,
      "loss": 0.4765,
      "step": 261
    },
    {
      "epoch": 1.2351097178683386,
      "grad_norm": 0.50210040807724,
      "learning_rate": 4.7679672960438135e-06,
      "loss": 0.5029,
      "step": 262
    },
    {
      "epoch": 1.2398119122257054,
      "grad_norm": 0.6636150479316711,
      "learning_rate": 4.765139707273872e-06,
      "loss": 0.4909,
      "step": 263
    },
    {
      "epoch": 1.244514106583072,
      "grad_norm": 0.4798625111579895,
      "learning_rate": 4.762295842819707e-06,
      "loss": 0.5012,
      "step": 264
    },
    {
      "epoch": 1.249216300940439,
      "grad_norm": 0.5282374024391174,
      "learning_rate": 4.759435723115308e-06,
      "loss": 0.4681,
      "step": 265
    },
    {
      "epoch": 1.2539184952978055,
      "grad_norm": 0.5356930494308472,
      "learning_rate": 4.756559368711463e-06,
      "loss": 0.506,
      "step": 266
    },
    {
      "epoch": 1.2586206896551724,
      "grad_norm": 0.4857093095779419,
      "learning_rate": 4.75366680027561e-06,
      "loss": 0.4889,
      "step": 267
    },
    {
      "epoch": 1.2633228840125392,
      "grad_norm": 0.484018474817276,
      "learning_rate": 4.7507580385916906e-06,
      "loss": 0.4899,
      "step": 268
    },
    {
      "epoch": 1.2680250783699059,
      "grad_norm": 0.49720871448516846,
      "learning_rate": 4.747833104559999e-06,
      "loss": 0.4654,
      "step": 269
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 0.4631911516189575,
      "learning_rate": 4.744892019197033e-06,
      "loss": 0.4796,
      "step": 270
    },
    {
      "epoch": 1.2774294670846396,
      "grad_norm": 0.5116872787475586,
      "learning_rate": 4.74193480363534e-06,
      "loss": 0.4883,
      "step": 271
    },
    {
      "epoch": 1.2821316614420062,
      "grad_norm": 0.5275093913078308,
      "learning_rate": 4.738961479123373e-06,
      "loss": 0.496,
      "step": 272
    },
    {
      "epoch": 1.286833855799373,
      "grad_norm": 0.5001885890960693,
      "learning_rate": 4.735972067025326e-06,
      "loss": 0.5012,
      "step": 273
    },
    {
      "epoch": 1.29153605015674,
      "grad_norm": 0.5875861048698425,
      "learning_rate": 4.732966588820991e-06,
      "loss": 0.4951,
      "step": 274
    },
    {
      "epoch": 1.2962382445141065,
      "grad_norm": 0.4893011748790741,
      "learning_rate": 4.729945066105599e-06,
      "loss": 0.4742,
      "step": 275
    },
    {
      "epoch": 1.3009404388714734,
      "grad_norm": 0.4648543894290924,
      "learning_rate": 4.726907520589664e-06,
      "loss": 0.466,
      "step": 276
    },
    {
      "epoch": 1.3056426332288402,
      "grad_norm": 0.5300162434577942,
      "learning_rate": 4.72385397409883e-06,
      "loss": 0.5072,
      "step": 277
    },
    {
      "epoch": 1.3103448275862069,
      "grad_norm": 0.4667080044746399,
      "learning_rate": 4.720784448573712e-06,
      "loss": 0.4986,
      "step": 278
    },
    {
      "epoch": 1.3150470219435737,
      "grad_norm": 0.5278895497322083,
      "learning_rate": 4.717698966069739e-06,
      "loss": 0.5269,
      "step": 279
    },
    {
      "epoch": 1.3197492163009406,
      "grad_norm": 0.5325866937637329,
      "learning_rate": 4.7145975487569965e-06,
      "loss": 0.5074,
      "step": 280
    },
    {
      "epoch": 1.3244514106583072,
      "grad_norm": 0.500861644744873,
      "learning_rate": 4.711480218920064e-06,
      "loss": 0.4695,
      "step": 281
    },
    {
      "epoch": 1.329153605015674,
      "grad_norm": 0.5263222455978394,
      "learning_rate": 4.708346998957859e-06,
      "loss": 0.5173,
      "step": 282
    },
    {
      "epoch": 1.3338557993730409,
      "grad_norm": 0.622900128364563,
      "learning_rate": 4.705197911383473e-06,
      "loss": 0.4905,
      "step": 283
    },
    {
      "epoch": 1.3385579937304075,
      "grad_norm": 0.49273768067359924,
      "learning_rate": 4.7020329788240115e-06,
      "loss": 0.4743,
      "step": 284
    },
    {
      "epoch": 1.3432601880877744,
      "grad_norm": 0.49558964371681213,
      "learning_rate": 4.6988522240204325e-06,
      "loss": 0.4824,
      "step": 285
    },
    {
      "epoch": 1.347962382445141,
      "grad_norm": 0.4743976891040802,
      "learning_rate": 4.695655669827377e-06,
      "loss": 0.4977,
      "step": 286
    },
    {
      "epoch": 1.3526645768025078,
      "grad_norm": 0.49542659521102905,
      "learning_rate": 4.6924433392130135e-06,
      "loss": 0.4924,
      "step": 287
    },
    {
      "epoch": 1.3573667711598745,
      "grad_norm": 0.7385990619659424,
      "learning_rate": 4.689215255258866e-06,
      "loss": 0.5091,
      "step": 288
    },
    {
      "epoch": 1.3620689655172413,
      "grad_norm": 0.4826123118400574,
      "learning_rate": 4.685971441159653e-06,
      "loss": 0.4791,
      "step": 289
    },
    {
      "epoch": 1.3667711598746082,
      "grad_norm": 0.5389033555984497,
      "learning_rate": 4.682711920223115e-06,
      "loss": 0.4751,
      "step": 290
    },
    {
      "epoch": 1.3714733542319748,
      "grad_norm": 0.5059546232223511,
      "learning_rate": 4.679436715869856e-06,
      "loss": 0.499,
      "step": 291
    },
    {
      "epoch": 1.3761755485893417,
      "grad_norm": 0.5682849884033203,
      "learning_rate": 4.676145851633166e-06,
      "loss": 0.5143,
      "step": 292
    },
    {
      "epoch": 1.3808777429467085,
      "grad_norm": 0.4754337668418884,
      "learning_rate": 4.672839351158856e-06,
      "loss": 0.4997,
      "step": 293
    },
    {
      "epoch": 1.3855799373040751,
      "grad_norm": 0.5227643847465515,
      "learning_rate": 4.669517238205089e-06,
      "loss": 0.4834,
      "step": 294
    },
    {
      "epoch": 1.390282131661442,
      "grad_norm": 0.4954044222831726,
      "learning_rate": 4.666179536642208e-06,
      "loss": 0.483,
      "step": 295
    },
    {
      "epoch": 1.3949843260188088,
      "grad_norm": 0.4909021556377411,
      "learning_rate": 4.662826270452565e-06,
      "loss": 0.4808,
      "step": 296
    },
    {
      "epoch": 1.3996865203761755,
      "grad_norm": 0.4666971266269684,
      "learning_rate": 4.659457463730347e-06,
      "loss": 0.488,
      "step": 297
    },
    {
      "epoch": 1.4043887147335423,
      "grad_norm": 0.5064187049865723,
      "learning_rate": 4.6560731406814056e-06,
      "loss": 0.5046,
      "step": 298
    },
    {
      "epoch": 1.4090909090909092,
      "grad_norm": 0.4958318769931793,
      "learning_rate": 4.65267332562308e-06,
      "loss": 0.5102,
      "step": 299
    },
    {
      "epoch": 1.4137931034482758,
      "grad_norm": 0.5080632567405701,
      "learning_rate": 4.649258042984026e-06,
      "loss": 0.5055,
      "step": 300
    },
    {
      "epoch": 1.4184952978056427,
      "grad_norm": 0.46236541867256165,
      "learning_rate": 4.6458273173040395e-06,
      "loss": 0.4606,
      "step": 301
    },
    {
      "epoch": 1.4231974921630095,
      "grad_norm": 1.8524898290634155,
      "learning_rate": 4.642381173233874e-06,
      "loss": 0.5002,
      "step": 302
    },
    {
      "epoch": 1.4278996865203761,
      "grad_norm": 0.5202615261077881,
      "learning_rate": 4.638919635535073e-06,
      "loss": 0.4562,
      "step": 303
    },
    {
      "epoch": 1.432601880877743,
      "grad_norm": 0.5293647050857544,
      "learning_rate": 4.635442729079788e-06,
      "loss": 0.4806,
      "step": 304
    },
    {
      "epoch": 1.4373040752351098,
      "grad_norm": 0.5165356993675232,
      "learning_rate": 4.6319504788505956e-06,
      "loss": 0.4775,
      "step": 305
    },
    {
      "epoch": 1.4420062695924765,
      "grad_norm": 0.5092841386795044,
      "learning_rate": 4.628442909940325e-06,
      "loss": 0.4892,
      "step": 306
    },
    {
      "epoch": 1.4467084639498433,
      "grad_norm": 0.511424720287323,
      "learning_rate": 4.624920047551874e-06,
      "loss": 0.506,
      "step": 307
    },
    {
      "epoch": 1.4514106583072102,
      "grad_norm": 0.5631566643714905,
      "learning_rate": 4.621381916998029e-06,
      "loss": 0.4741,
      "step": 308
    },
    {
      "epoch": 1.4561128526645768,
      "grad_norm": 0.4748315215110779,
      "learning_rate": 4.6178285437012806e-06,
      "loss": 0.5084,
      "step": 309
    },
    {
      "epoch": 1.4608150470219436,
      "grad_norm": 0.47158119082450867,
      "learning_rate": 4.6142599531936435e-06,
      "loss": 0.4697,
      "step": 310
    },
    {
      "epoch": 1.4655172413793103,
      "grad_norm": 0.5358107089996338,
      "learning_rate": 4.610676171116475e-06,
      "loss": 0.491,
      "step": 311
    },
    {
      "epoch": 1.4702194357366771,
      "grad_norm": 0.47717440128326416,
      "learning_rate": 4.607077223220286e-06,
      "loss": 0.4948,
      "step": 312
    },
    {
      "epoch": 1.4749216300940438,
      "grad_norm": 0.5041193962097168,
      "learning_rate": 4.603463135364556e-06,
      "loss": 0.4648,
      "step": 313
    },
    {
      "epoch": 1.4796238244514106,
      "grad_norm": 0.9311274290084839,
      "learning_rate": 4.5998339335175555e-06,
      "loss": 0.4866,
      "step": 314
    },
    {
      "epoch": 1.4843260188087775,
      "grad_norm": 0.47408604621887207,
      "learning_rate": 4.596189643756147e-06,
      "loss": 0.4634,
      "step": 315
    },
    {
      "epoch": 1.489028213166144,
      "grad_norm": 0.5052632093429565,
      "learning_rate": 4.592530292265609e-06,
      "loss": 0.4843,
      "step": 316
    },
    {
      "epoch": 1.493730407523511,
      "grad_norm": 0.5100846886634827,
      "learning_rate": 4.58885590533944e-06,
      "loss": 0.4942,
      "step": 317
    },
    {
      "epoch": 1.4984326018808778,
      "grad_norm": 0.5132214426994324,
      "learning_rate": 4.585166509379173e-06,
      "loss": 0.5135,
      "step": 318
    },
    {
      "epoch": 1.5031347962382444,
      "grad_norm": 11.112855911254883,
      "learning_rate": 4.581462130894186e-06,
      "loss": 0.4933,
      "step": 319
    },
    {
      "epoch": 1.5078369905956113,
      "grad_norm": 0.4873805642127991,
      "learning_rate": 4.57774279650151e-06,
      "loss": 0.483,
      "step": 320
    },
    {
      "epoch": 1.5125391849529781,
      "grad_norm": 0.5026459693908691,
      "learning_rate": 4.574008532925638e-06,
      "loss": 0.5075,
      "step": 321
    },
    {
      "epoch": 1.5172413793103448,
      "grad_norm": 0.489947110414505,
      "learning_rate": 4.570259366998336e-06,
      "loss": 0.4954,
      "step": 322
    },
    {
      "epoch": 1.5219435736677116,
      "grad_norm": 0.48120853304862976,
      "learning_rate": 4.566495325658445e-06,
      "loss": 0.5221,
      "step": 323
    },
    {
      "epoch": 1.5266457680250785,
      "grad_norm": 0.4880066514015198,
      "learning_rate": 4.5627164359516915e-06,
      "loss": 0.5031,
      "step": 324
    },
    {
      "epoch": 1.531347962382445,
      "grad_norm": 0.5048410892486572,
      "learning_rate": 4.558922725030491e-06,
      "loss": 0.4757,
      "step": 325
    },
    {
      "epoch": 1.536050156739812,
      "grad_norm": 0.7033756375312805,
      "learning_rate": 4.555114220153755e-06,
      "loss": 0.4285,
      "step": 326
    },
    {
      "epoch": 1.5407523510971788,
      "grad_norm": 0.4716516435146332,
      "learning_rate": 4.551290948686693e-06,
      "loss": 0.5121,
      "step": 327
    },
    {
      "epoch": 1.5454545454545454,
      "grad_norm": 0.4782696068286896,
      "learning_rate": 4.547452938100615e-06,
      "loss": 0.5176,
      "step": 328
    },
    {
      "epoch": 1.5501567398119123,
      "grad_norm": 0.5119273066520691,
      "learning_rate": 4.54360021597274e-06,
      "loss": 0.4941,
      "step": 329
    },
    {
      "epoch": 1.5548589341692791,
      "grad_norm": 0.5010069608688354,
      "learning_rate": 4.539732809985989e-06,
      "loss": 0.4862,
      "step": 330
    },
    {
      "epoch": 1.5595611285266457,
      "grad_norm": 0.5129932165145874,
      "learning_rate": 4.535850747928796e-06,
      "loss": 0.4978,
      "step": 331
    },
    {
      "epoch": 1.5642633228840124,
      "grad_norm": 0.4957594573497772,
      "learning_rate": 4.531954057694897e-06,
      "loss": 0.4814,
      "step": 332
    },
    {
      "epoch": 1.5689655172413794,
      "grad_norm": 0.5642824172973633,
      "learning_rate": 4.5280427672831414e-06,
      "loss": 0.4888,
      "step": 333
    },
    {
      "epoch": 1.573667711598746,
      "grad_norm": 0.4562854468822479,
      "learning_rate": 4.524116904797281e-06,
      "loss": 0.4648,
      "step": 334
    },
    {
      "epoch": 1.5783699059561127,
      "grad_norm": 0.4849218428134918,
      "learning_rate": 4.520176498445774e-06,
      "loss": 0.476,
      "step": 335
    },
    {
      "epoch": 1.5830721003134798,
      "grad_norm": 0.5046947002410889,
      "learning_rate": 4.516221576541581e-06,
      "loss": 0.4776,
      "step": 336
    },
    {
      "epoch": 1.5877742946708464,
      "grad_norm": 0.48211777210235596,
      "learning_rate": 4.512252167501959e-06,
      "loss": 0.479,
      "step": 337
    },
    {
      "epoch": 1.592476489028213,
      "grad_norm": 0.4812171459197998,
      "learning_rate": 4.508268299848262e-06,
      "loss": 0.4849,
      "step": 338
    },
    {
      "epoch": 1.59717868338558,
      "grad_norm": 0.5865142345428467,
      "learning_rate": 4.50427000220573e-06,
      "loss": 0.499,
      "step": 339
    },
    {
      "epoch": 1.6018808777429467,
      "grad_norm": 0.49277785420417786,
      "learning_rate": 4.50025730330329e-06,
      "loss": 0.475,
      "step": 340
    },
    {
      "epoch": 1.6065830721003134,
      "grad_norm": 0.46771496534347534,
      "learning_rate": 4.4962302319733445e-06,
      "loss": 0.494,
      "step": 341
    },
    {
      "epoch": 1.6112852664576802,
      "grad_norm": 0.5189441442489624,
      "learning_rate": 4.492188817151565e-06,
      "loss": 0.5275,
      "step": 342
    },
    {
      "epoch": 1.615987460815047,
      "grad_norm": 0.48845574259757996,
      "learning_rate": 4.488133087876688e-06,
      "loss": 0.4676,
      "step": 343
    },
    {
      "epoch": 1.6206896551724137,
      "grad_norm": 0.47189632058143616,
      "learning_rate": 4.484063073290301e-06,
      "loss": 0.4642,
      "step": 344
    },
    {
      "epoch": 1.6253918495297806,
      "grad_norm": 0.5442587733268738,
      "learning_rate": 4.479978802636637e-06,
      "loss": 0.4981,
      "step": 345
    },
    {
      "epoch": 1.6300940438871474,
      "grad_norm": 0.5048685073852539,
      "learning_rate": 4.475880305262362e-06,
      "loss": 0.5037,
      "step": 346
    },
    {
      "epoch": 1.634796238244514,
      "grad_norm": 0.4781409800052643,
      "learning_rate": 4.471767610616366e-06,
      "loss": 0.4932,
      "step": 347
    },
    {
      "epoch": 1.6394984326018809,
      "grad_norm": 0.47388938069343567,
      "learning_rate": 4.467640748249549e-06,
      "loss": 0.4687,
      "step": 348
    },
    {
      "epoch": 1.6442006269592477,
      "grad_norm": 0.529712438583374,
      "learning_rate": 4.4634997478146125e-06,
      "loss": 0.487,
      "step": 349
    },
    {
      "epoch": 1.6489028213166144,
      "grad_norm": 0.5114791393280029,
      "learning_rate": 4.459344639065842e-06,
      "loss": 0.4809,
      "step": 350
    },
    {
      "epoch": 1.6536050156739812,
      "grad_norm": 0.45415258407592773,
      "learning_rate": 4.455175451858897e-06,
      "loss": 0.4901,
      "step": 351
    },
    {
      "epoch": 1.658307210031348,
      "grad_norm": 0.5842339396476746,
      "learning_rate": 4.450992216150592e-06,
      "loss": 0.499,
      "step": 352
    },
    {
      "epoch": 1.6630094043887147,
      "grad_norm": 0.48795560002326965,
      "learning_rate": 4.446794961998689e-06,
      "loss": 0.4659,
      "step": 353
    },
    {
      "epoch": 1.6677115987460815,
      "grad_norm": 0.5531855225563049,
      "learning_rate": 4.442583719561671e-06,
      "loss": 0.4923,
      "step": 354
    },
    {
      "epoch": 1.6724137931034484,
      "grad_norm": 0.5827644467353821,
      "learning_rate": 4.438358519098536e-06,
      "loss": 0.4991,
      "step": 355
    },
    {
      "epoch": 1.677115987460815,
      "grad_norm": 0.5260423421859741,
      "learning_rate": 4.4341193909685685e-06,
      "loss": 0.4843,
      "step": 356
    },
    {
      "epoch": 1.6818181818181817,
      "grad_norm": 0.4969344437122345,
      "learning_rate": 4.429866365631134e-06,
      "loss": 0.4915,
      "step": 357
    },
    {
      "epoch": 1.6865203761755487,
      "grad_norm": 0.4725005030632019,
      "learning_rate": 4.425599473645447e-06,
      "loss": 0.4804,
      "step": 358
    },
    {
      "epoch": 1.6912225705329154,
      "grad_norm": 0.47171467542648315,
      "learning_rate": 4.421318745670364e-06,
      "loss": 0.4823,
      "step": 359
    },
    {
      "epoch": 1.695924764890282,
      "grad_norm": 0.4839799106121063,
      "learning_rate": 4.4170242124641524e-06,
      "loss": 0.4585,
      "step": 360
    },
    {
      "epoch": 1.700626959247649,
      "grad_norm": 0.4786856472492218,
      "learning_rate": 4.412715904884277e-06,
      "loss": 0.49,
      "step": 361
    },
    {
      "epoch": 1.7053291536050157,
      "grad_norm": 0.49980080127716064,
      "learning_rate": 4.4083938538871735e-06,
      "loss": 0.4675,
      "step": 362
    },
    {
      "epoch": 1.7100313479623823,
      "grad_norm": 0.5201369524002075,
      "learning_rate": 4.4040580905280295e-06,
      "loss": 0.4862,
      "step": 363
    },
    {
      "epoch": 1.7147335423197492,
      "grad_norm": 0.7051575183868408,
      "learning_rate": 4.3997086459605586e-06,
      "loss": 0.4822,
      "step": 364
    },
    {
      "epoch": 1.719435736677116,
      "grad_norm": 0.48206666111946106,
      "learning_rate": 4.395345551436779e-06,
      "loss": 0.5076,
      "step": 365
    },
    {
      "epoch": 1.7241379310344827,
      "grad_norm": 0.4817257821559906,
      "learning_rate": 4.390968838306788e-06,
      "loss": 0.4623,
      "step": 366
    },
    {
      "epoch": 1.7288401253918495,
      "grad_norm": 0.5547840595245361,
      "learning_rate": 4.386578538018535e-06,
      "loss": 0.461,
      "step": 367
    },
    {
      "epoch": 1.7335423197492164,
      "grad_norm": 0.5085346698760986,
      "learning_rate": 4.382174682117598e-06,
      "loss": 0.5068,
      "step": 368
    },
    {
      "epoch": 1.738244514106583,
      "grad_norm": 0.4870692193508148,
      "learning_rate": 4.377757302246956e-06,
      "loss": 0.4403,
      "step": 369
    },
    {
      "epoch": 1.7429467084639498,
      "grad_norm": 0.49482715129852295,
      "learning_rate": 4.373326430146762e-06,
      "loss": 0.4986,
      "step": 370
    },
    {
      "epoch": 1.7476489028213167,
      "grad_norm": 0.5474854707717896,
      "learning_rate": 4.368882097654113e-06,
      "loss": 0.4938,
      "step": 371
    },
    {
      "epoch": 1.7523510971786833,
      "grad_norm": 0.5055244565010071,
      "learning_rate": 4.364424336702825e-06,
      "loss": 0.4711,
      "step": 372
    },
    {
      "epoch": 1.7570532915360502,
      "grad_norm": 0.48241329193115234,
      "learning_rate": 4.3599531793232e-06,
      "loss": 0.4856,
      "step": 373
    },
    {
      "epoch": 1.761755485893417,
      "grad_norm": 0.4932602047920227,
      "learning_rate": 4.355468657641797e-06,
      "loss": 0.4818,
      "step": 374
    },
    {
      "epoch": 1.7664576802507836,
      "grad_norm": 0.5512160658836365,
      "learning_rate": 4.3509708038812035e-06,
      "loss": 0.4864,
      "step": 375
    },
    {
      "epoch": 1.7711598746081505,
      "grad_norm": 0.47026327252388,
      "learning_rate": 4.346459650359798e-06,
      "loss": 0.4825,
      "step": 376
    },
    {
      "epoch": 1.7758620689655173,
      "grad_norm": 0.4831086993217468,
      "learning_rate": 4.341935229491525e-06,
      "loss": 0.4541,
      "step": 377
    },
    {
      "epoch": 1.780564263322884,
      "grad_norm": 0.5045217871665955,
      "learning_rate": 4.337397573785659e-06,
      "loss": 0.5025,
      "step": 378
    },
    {
      "epoch": 1.7852664576802508,
      "grad_norm": 0.5657753348350525,
      "learning_rate": 4.332846715846566e-06,
      "loss": 0.4698,
      "step": 379
    },
    {
      "epoch": 1.7899686520376177,
      "grad_norm": 0.49546748399734497,
      "learning_rate": 4.328282688373479e-06,
      "loss": 0.4911,
      "step": 380
    },
    {
      "epoch": 1.7946708463949843,
      "grad_norm": 0.5037291049957275,
      "learning_rate": 4.323705524160258e-06,
      "loss": 0.4877,
      "step": 381
    },
    {
      "epoch": 1.799373040752351,
      "grad_norm": 0.5256901383399963,
      "learning_rate": 4.319115256095149e-06,
      "loss": 0.4662,
      "step": 382
    },
    {
      "epoch": 1.804075235109718,
      "grad_norm": 0.4890702962875366,
      "learning_rate": 4.314511917160557e-06,
      "loss": 0.4683,
      "step": 383
    },
    {
      "epoch": 1.8087774294670846,
      "grad_norm": 0.4724109470844269,
      "learning_rate": 4.3098955404328045e-06,
      "loss": 0.4602,
      "step": 384
    },
    {
      "epoch": 1.8134796238244513,
      "grad_norm": 0.4933278560638428,
      "learning_rate": 4.305266159081895e-06,
      "loss": 0.4806,
      "step": 385
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.5068219304084778,
      "learning_rate": 4.3006238063712725e-06,
      "loss": 0.4647,
      "step": 386
    },
    {
      "epoch": 1.822884012539185,
      "grad_norm": 0.5293509364128113,
      "learning_rate": 4.295968515657583e-06,
      "loss": 0.4998,
      "step": 387
    },
    {
      "epoch": 1.8275862068965516,
      "grad_norm": 0.4775199294090271,
      "learning_rate": 4.29130032039044e-06,
      "loss": 0.4821,
      "step": 388
    },
    {
      "epoch": 1.8322884012539185,
      "grad_norm": 0.4914006292819977,
      "learning_rate": 4.2866192541121755e-06,
      "loss": 0.4735,
      "step": 389
    },
    {
      "epoch": 1.8369905956112853,
      "grad_norm": 0.5009908080101013,
      "learning_rate": 4.281925350457606e-06,
      "loss": 0.4741,
      "step": 390
    },
    {
      "epoch": 1.841692789968652,
      "grad_norm": 0.47211164236068726,
      "learning_rate": 4.277218643153787e-06,
      "loss": 0.4786,
      "step": 391
    },
    {
      "epoch": 1.8463949843260188,
      "grad_norm": 1.9644113779067993,
      "learning_rate": 4.272499166019771e-06,
      "loss": 0.4759,
      "step": 392
    },
    {
      "epoch": 1.8510971786833856,
      "grad_norm": 0.535971999168396,
      "learning_rate": 4.267766952966369e-06,
      "loss": 0.4665,
      "step": 393
    },
    {
      "epoch": 1.8557993730407523,
      "grad_norm": 0.4666787385940552,
      "learning_rate": 4.2630220379959006e-06,
      "loss": 0.4417,
      "step": 394
    },
    {
      "epoch": 1.8605015673981191,
      "grad_norm": 0.5976264476776123,
      "learning_rate": 4.258264455201953e-06,
      "loss": 0.4665,
      "step": 395
    },
    {
      "epoch": 1.865203761755486,
      "grad_norm": 0.4814331531524658,
      "learning_rate": 4.2534942387691335e-06,
      "loss": 0.4896,
      "step": 396
    },
    {
      "epoch": 1.8699059561128526,
      "grad_norm": 0.4929859936237335,
      "learning_rate": 4.248711422972829e-06,
      "loss": 0.4765,
      "step": 397
    },
    {
      "epoch": 1.8746081504702194,
      "grad_norm": 0.517914354801178,
      "learning_rate": 4.243916042178954e-06,
      "loss": 0.4601,
      "step": 398
    },
    {
      "epoch": 1.8793103448275863,
      "grad_norm": 0.47731271386146545,
      "learning_rate": 4.239108130843709e-06,
      "loss": 0.469,
      "step": 399
    },
    {
      "epoch": 1.884012539184953,
      "grad_norm": 0.4939954876899719,
      "learning_rate": 4.234287723513326e-06,
      "loss": 0.4929,
      "step": 400
    },
    {
      "epoch": 1.8887147335423198,
      "grad_norm": 0.48573923110961914,
      "learning_rate": 4.229454854823827e-06,
      "loss": 0.4913,
      "step": 401
    },
    {
      "epoch": 1.8934169278996866,
      "grad_norm": 0.5146409273147583,
      "learning_rate": 4.224609559500772e-06,
      "loss": 0.502,
      "step": 402
    },
    {
      "epoch": 1.8981191222570533,
      "grad_norm": 0.4884675443172455,
      "learning_rate": 4.21975187235901e-06,
      "loss": 0.4541,
      "step": 403
    },
    {
      "epoch": 1.90282131661442,
      "grad_norm": 0.4871810972690582,
      "learning_rate": 4.21488182830243e-06,
      "loss": 0.4811,
      "step": 404
    },
    {
      "epoch": 1.907523510971787,
      "grad_norm": 0.5089552402496338,
      "learning_rate": 4.209999462323706e-06,
      "loss": 0.4584,
      "step": 405
    },
    {
      "epoch": 1.9122257053291536,
      "grad_norm": 0.6191231608390808,
      "learning_rate": 4.20510480950405e-06,
      "loss": 0.4885,
      "step": 406
    },
    {
      "epoch": 1.9169278996865202,
      "grad_norm": 0.5512096285820007,
      "learning_rate": 4.200197905012961e-06,
      "loss": 0.4529,
      "step": 407
    },
    {
      "epoch": 1.9216300940438873,
      "grad_norm": 0.4743112027645111,
      "learning_rate": 4.195278784107965e-06,
      "loss": 0.4702,
      "step": 408
    },
    {
      "epoch": 1.926332288401254,
      "grad_norm": 0.4635118544101715,
      "learning_rate": 4.19034748213437e-06,
      "loss": 0.4718,
      "step": 409
    },
    {
      "epoch": 1.9310344827586206,
      "grad_norm": 0.48715919256210327,
      "learning_rate": 4.185404034525008e-06,
      "loss": 0.4638,
      "step": 410
    },
    {
      "epoch": 1.9357366771159876,
      "grad_norm": 0.5373724102973938,
      "learning_rate": 4.180448476799981e-06,
      "loss": 0.5009,
      "step": 411
    },
    {
      "epoch": 1.9404388714733543,
      "grad_norm": 0.4978715479373932,
      "learning_rate": 4.175480844566404e-06,
      "loss": 0.4726,
      "step": 412
    },
    {
      "epoch": 1.9451410658307209,
      "grad_norm": 0.44817060232162476,
      "learning_rate": 4.170501173518152e-06,
      "loss": 0.4683,
      "step": 413
    },
    {
      "epoch": 1.9498432601880877,
      "grad_norm": 0.48472973704338074,
      "learning_rate": 4.165509499435604e-06,
      "loss": 0.4662,
      "step": 414
    },
    {
      "epoch": 1.9545454545454546,
      "grad_norm": 0.6567174792289734,
      "learning_rate": 4.16050585818538e-06,
      "loss": 0.4801,
      "step": 415
    },
    {
      "epoch": 1.9592476489028212,
      "grad_norm": 0.5131425857543945,
      "learning_rate": 4.155490285720092e-06,
      "loss": 0.5036,
      "step": 416
    },
    {
      "epoch": 1.963949843260188,
      "grad_norm": 0.46051982045173645,
      "learning_rate": 4.150462818078079e-06,
      "loss": 0.4911,
      "step": 417
    },
    {
      "epoch": 1.968652037617555,
      "grad_norm": 0.5288883447647095,
      "learning_rate": 4.145423491383153e-06,
      "loss": 0.4871,
      "step": 418
    },
    {
      "epoch": 1.9733542319749215,
      "grad_norm": 0.5143817663192749,
      "learning_rate": 4.14037234184433e-06,
      "loss": 0.5027,
      "step": 419
    },
    {
      "epoch": 1.9780564263322884,
      "grad_norm": 0.46323707699775696,
      "learning_rate": 4.135309405755583e-06,
      "loss": 0.4876,
      "step": 420
    },
    {
      "epoch": 1.9827586206896552,
      "grad_norm": 0.5239706039428711,
      "learning_rate": 4.130234719495574e-06,
      "loss": 0.4702,
      "step": 421
    },
    {
      "epoch": 1.9874608150470219,
      "grad_norm": 0.538753867149353,
      "learning_rate": 4.125148319527391e-06,
      "loss": 0.4638,
      "step": 422
    },
    {
      "epoch": 1.9921630094043887,
      "grad_norm": 0.5180181860923767,
      "learning_rate": 4.1200502423982904e-06,
      "loss": 0.4841,
      "step": 423
    },
    {
      "epoch": 1.9968652037617556,
      "grad_norm": 0.6698167324066162,
      "learning_rate": 4.1149405247394295e-06,
      "loss": 0.4882,
      "step": 424
    },
    {
      "epoch": 2.0047021943573666,
      "grad_norm": 0.9728522896766663,
      "learning_rate": 4.10981920326561e-06,
      "loss": 0.9125,
      "step": 425
    },
    {
      "epoch": 2.0094043887147337,
      "grad_norm": 0.7356107831001282,
      "learning_rate": 4.104686314775009e-06,
      "loss": 0.4422,
      "step": 426
    },
    {
      "epoch": 2.0141065830721003,
      "grad_norm": 0.44414228200912476,
      "learning_rate": 4.099541896148914e-06,
      "loss": 0.4511,
      "step": 427
    },
    {
      "epoch": 2.018808777429467,
      "grad_norm": 0.5738011002540588,
      "learning_rate": 4.094385984351462e-06,
      "loss": 0.4457,
      "step": 428
    },
    {
      "epoch": 2.023510971786834,
      "grad_norm": 0.4643106460571289,
      "learning_rate": 4.0892186164293715e-06,
      "loss": 0.4644,
      "step": 429
    },
    {
      "epoch": 2.0282131661442007,
      "grad_norm": 0.5355309247970581,
      "learning_rate": 4.0840398295116745e-06,
      "loss": 0.4535,
      "step": 430
    },
    {
      "epoch": 2.0329153605015673,
      "grad_norm": 0.512458324432373,
      "learning_rate": 4.078849660809456e-06,
      "loss": 0.4481,
      "step": 431
    },
    {
      "epoch": 2.0376175548589344,
      "grad_norm": 0.5055253505706787,
      "learning_rate": 4.073648147615579e-06,
      "loss": 0.4309,
      "step": 432
    },
    {
      "epoch": 2.042319749216301,
      "grad_norm": 0.5128353834152222,
      "learning_rate": 4.068435327304421e-06,
      "loss": 0.4562,
      "step": 433
    },
    {
      "epoch": 2.0470219435736676,
      "grad_norm": 0.4432103633880615,
      "learning_rate": 4.063211237331603e-06,
      "loss": 0.4535,
      "step": 434
    },
    {
      "epoch": 2.0517241379310347,
      "grad_norm": 0.5092498660087585,
      "learning_rate": 4.057975915233725e-06,
      "loss": 0.4385,
      "step": 435
    },
    {
      "epoch": 2.0564263322884013,
      "grad_norm": 0.4798133671283722,
      "learning_rate": 4.052729398628089e-06,
      "loss": 0.466,
      "step": 436
    },
    {
      "epoch": 2.061128526645768,
      "grad_norm": 0.5094019770622253,
      "learning_rate": 4.047471725212437e-06,
      "loss": 0.4624,
      "step": 437
    },
    {
      "epoch": 2.0658307210031346,
      "grad_norm": 0.5814178586006165,
      "learning_rate": 4.042202932764673e-06,
      "loss": 0.4472,
      "step": 438
    },
    {
      "epoch": 2.0705329153605017,
      "grad_norm": 0.503394365310669,
      "learning_rate": 4.036923059142595e-06,
      "loss": 0.4481,
      "step": 439
    },
    {
      "epoch": 2.0752351097178683,
      "grad_norm": 0.5108861923217773,
      "learning_rate": 4.031632142283623e-06,
      "loss": 0.4416,
      "step": 440
    },
    {
      "epoch": 2.079937304075235,
      "grad_norm": 0.5303971171379089,
      "learning_rate": 4.026330220204524e-06,
      "loss": 0.4515,
      "step": 441
    },
    {
      "epoch": 2.084639498432602,
      "grad_norm": 0.45014286041259766,
      "learning_rate": 4.021017331001146e-06,
      "loss": 0.441,
      "step": 442
    },
    {
      "epoch": 2.0893416927899686,
      "grad_norm": 0.5371219515800476,
      "learning_rate": 4.015693512848131e-06,
      "loss": 0.4471,
      "step": 443
    },
    {
      "epoch": 2.0940438871473352,
      "grad_norm": 0.5105510354042053,
      "learning_rate": 4.0103588039986556e-06,
      "loss": 0.4534,
      "step": 444
    },
    {
      "epoch": 2.0987460815047023,
      "grad_norm": 0.4960611164569855,
      "learning_rate": 4.005013242784146e-06,
      "loss": 0.46,
      "step": 445
    },
    {
      "epoch": 2.103448275862069,
      "grad_norm": 0.500354528427124,
      "learning_rate": 3.999656867614006e-06,
      "loss": 0.45,
      "step": 446
    },
    {
      "epoch": 2.1081504702194356,
      "grad_norm": 0.4733876585960388,
      "learning_rate": 3.994289716975341e-06,
      "loss": 0.4644,
      "step": 447
    },
    {
      "epoch": 2.1128526645768027,
      "grad_norm": 0.5002915263175964,
      "learning_rate": 3.988911829432682e-06,
      "loss": 0.4493,
      "step": 448
    },
    {
      "epoch": 2.1175548589341693,
      "grad_norm": 0.48520293831825256,
      "learning_rate": 3.983523243627706e-06,
      "loss": 0.4458,
      "step": 449
    },
    {
      "epoch": 2.122257053291536,
      "grad_norm": 0.6339934468269348,
      "learning_rate": 3.978123998278962e-06,
      "loss": 0.4352,
      "step": 450
    },
    {
      "epoch": 2.126959247648903,
      "grad_norm": 1.172338843345642,
      "learning_rate": 3.97271413218159e-06,
      "loss": 0.4664,
      "step": 451
    },
    {
      "epoch": 2.1316614420062696,
      "grad_norm": 0.47842296957969666,
      "learning_rate": 3.9672936842070425e-06,
      "loss": 0.4604,
      "step": 452
    },
    {
      "epoch": 2.1363636363636362,
      "grad_norm": 0.506851077079773,
      "learning_rate": 3.9618626933028086e-06,
      "loss": 0.4674,
      "step": 453
    },
    {
      "epoch": 2.1410658307210033,
      "grad_norm": 0.4922677278518677,
      "learning_rate": 3.956421198492128e-06,
      "loss": 0.4476,
      "step": 454
    },
    {
      "epoch": 2.14576802507837,
      "grad_norm": 0.5307339429855347,
      "learning_rate": 3.950969238873714e-06,
      "loss": 0.4463,
      "step": 455
    },
    {
      "epoch": 2.1504702194357366,
      "grad_norm": 0.5131121873855591,
      "learning_rate": 3.9455068536214765e-06,
      "loss": 0.4779,
      "step": 456
    },
    {
      "epoch": 2.1551724137931036,
      "grad_norm": 0.5438089966773987,
      "learning_rate": 3.9400340819842335e-06,
      "loss": 0.4563,
      "step": 457
    },
    {
      "epoch": 2.1598746081504703,
      "grad_norm": 0.7426711916923523,
      "learning_rate": 3.934550963285432e-06,
      "loss": 0.4561,
      "step": 458
    },
    {
      "epoch": 2.164576802507837,
      "grad_norm": 0.482920378446579,
      "learning_rate": 3.9290575369228664e-06,
      "loss": 0.4293,
      "step": 459
    },
    {
      "epoch": 2.169278996865204,
      "grad_norm": 0.6583715081214905,
      "learning_rate": 3.923553842368396e-06,
      "loss": 0.4682,
      "step": 460
    },
    {
      "epoch": 2.1739811912225706,
      "grad_norm": 0.47901806235313416,
      "learning_rate": 3.918039919167658e-06,
      "loss": 0.4342,
      "step": 461
    },
    {
      "epoch": 2.1786833855799372,
      "grad_norm": 0.4929746389389038,
      "learning_rate": 3.912515806939786e-06,
      "loss": 0.4478,
      "step": 462
    },
    {
      "epoch": 2.183385579937304,
      "grad_norm": 0.48205333948135376,
      "learning_rate": 3.906981545377124e-06,
      "loss": 0.4595,
      "step": 463
    },
    {
      "epoch": 2.188087774294671,
      "grad_norm": 0.5059337019920349,
      "learning_rate": 3.901437174244943e-06,
      "loss": 0.4294,
      "step": 464
    },
    {
      "epoch": 2.1927899686520376,
      "grad_norm": 0.4752981662750244,
      "learning_rate": 3.895882733381154e-06,
      "loss": 0.448,
      "step": 465
    },
    {
      "epoch": 2.197492163009404,
      "grad_norm": 0.5249196290969849,
      "learning_rate": 3.890318262696023e-06,
      "loss": 0.4655,
      "step": 466
    },
    {
      "epoch": 2.2021943573667713,
      "grad_norm": 0.48044726252555847,
      "learning_rate": 3.8847438021718805e-06,
      "loss": 0.4413,
      "step": 467
    },
    {
      "epoch": 2.206896551724138,
      "grad_norm": 0.84516841173172,
      "learning_rate": 3.879159391862839e-06,
      "loss": 0.4645,
      "step": 468
    },
    {
      "epoch": 2.2115987460815045,
      "grad_norm": 0.5334392786026001,
      "learning_rate": 3.873565071894503e-06,
      "loss": 0.4347,
      "step": 469
    },
    {
      "epoch": 2.2163009404388716,
      "grad_norm": 0.5113687515258789,
      "learning_rate": 3.86796088246368e-06,
      "loss": 0.4314,
      "step": 470
    },
    {
      "epoch": 2.2210031347962382,
      "grad_norm": 0.5226101279258728,
      "learning_rate": 3.8623468638380905e-06,
      "loss": 0.418,
      "step": 471
    },
    {
      "epoch": 2.225705329153605,
      "grad_norm": 0.4901522099971771,
      "learning_rate": 3.856723056356085e-06,
      "loss": 0.4597,
      "step": 472
    },
    {
      "epoch": 2.230407523510972,
      "grad_norm": 0.5312012434005737,
      "learning_rate": 3.851089500426346e-06,
      "loss": 0.4444,
      "step": 473
    },
    {
      "epoch": 2.2351097178683386,
      "grad_norm": 0.5347906351089478,
      "learning_rate": 3.845446236527605e-06,
      "loss": 0.4447,
      "step": 474
    },
    {
      "epoch": 2.239811912225705,
      "grad_norm": 0.4781494438648224,
      "learning_rate": 3.8397933052083445e-06,
      "loss": 0.462,
      "step": 475
    },
    {
      "epoch": 2.2445141065830723,
      "grad_norm": 0.5215012431144714,
      "learning_rate": 3.834130747086512e-06,
      "loss": 0.4475,
      "step": 476
    },
    {
      "epoch": 2.249216300940439,
      "grad_norm": 0.5048666000366211,
      "learning_rate": 3.828458602849226e-06,
      "loss": 0.4483,
      "step": 477
    },
    {
      "epoch": 2.2539184952978055,
      "grad_norm": 0.5508173108100891,
      "learning_rate": 3.822776913252485e-06,
      "loss": 0.4511,
      "step": 478
    },
    {
      "epoch": 2.2586206896551726,
      "grad_norm": 0.5031043887138367,
      "learning_rate": 3.817085719120872e-06,
      "loss": 0.4019,
      "step": 479
    },
    {
      "epoch": 2.2633228840125392,
      "grad_norm": 0.508939802646637,
      "learning_rate": 3.811385061347263e-06,
      "loss": 0.4461,
      "step": 480
    },
    {
      "epoch": 2.268025078369906,
      "grad_norm": 0.5605170726776123,
      "learning_rate": 3.805674980892535e-06,
      "loss": 0.4695,
      "step": 481
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.5526806712150574,
      "learning_rate": 3.7999555187852667e-06,
      "loss": 0.4575,
      "step": 482
    },
    {
      "epoch": 2.2774294670846396,
      "grad_norm": 0.47659724950790405,
      "learning_rate": 3.7942267161214497e-06,
      "loss": 0.4433,
      "step": 483
    },
    {
      "epoch": 2.282131661442006,
      "grad_norm": 0.49713975191116333,
      "learning_rate": 3.7884886140641884e-06,
      "loss": 0.4692,
      "step": 484
    },
    {
      "epoch": 2.2868338557993733,
      "grad_norm": 0.48685988783836365,
      "learning_rate": 3.7827412538434062e-06,
      "loss": 0.4328,
      "step": 485
    },
    {
      "epoch": 2.29153605015674,
      "grad_norm": 0.5074832439422607,
      "learning_rate": 3.7769846767555495e-06,
      "loss": 0.4598,
      "step": 486
    },
    {
      "epoch": 2.2962382445141065,
      "grad_norm": 0.5333994030952454,
      "learning_rate": 3.7712189241632898e-06,
      "loss": 0.4554,
      "step": 487
    },
    {
      "epoch": 2.300940438871473,
      "grad_norm": 0.49985551834106445,
      "learning_rate": 3.7654440374952288e-06,
      "loss": 0.4421,
      "step": 488
    },
    {
      "epoch": 2.30564263322884,
      "grad_norm": 0.4791257679462433,
      "learning_rate": 3.7596600582455976e-06,
      "loss": 0.4187,
      "step": 489
    },
    {
      "epoch": 2.310344827586207,
      "grad_norm": 0.4951220154762268,
      "learning_rate": 3.75386702797396e-06,
      "loss": 0.4205,
      "step": 490
    },
    {
      "epoch": 2.3150470219435735,
      "grad_norm": 0.4765990674495697,
      "learning_rate": 3.7480649883049164e-06,
      "loss": 0.4251,
      "step": 491
    },
    {
      "epoch": 2.3197492163009406,
      "grad_norm": 0.5125405192375183,
      "learning_rate": 3.7422539809277993e-06,
      "loss": 0.4361,
      "step": 492
    },
    {
      "epoch": 2.324451410658307,
      "grad_norm": 0.5286112427711487,
      "learning_rate": 3.736434047596379e-06,
      "loss": 0.4423,
      "step": 493
    },
    {
      "epoch": 2.329153605015674,
      "grad_norm": 0.47961002588272095,
      "learning_rate": 3.73060523012856e-06,
      "loss": 0.453,
      "step": 494
    },
    {
      "epoch": 2.333855799373041,
      "grad_norm": 0.5857998728752136,
      "learning_rate": 3.724767570406082e-06,
      "loss": 0.4674,
      "step": 495
    },
    {
      "epoch": 2.3385579937304075,
      "grad_norm": 0.5348326563835144,
      "learning_rate": 3.7189211103742206e-06,
      "loss": 0.4267,
      "step": 496
    },
    {
      "epoch": 2.343260188087774,
      "grad_norm": 0.4718475937843323,
      "learning_rate": 3.7130658920414818e-06,
      "loss": 0.4619,
      "step": 497
    },
    {
      "epoch": 2.347962382445141,
      "grad_norm": 0.44225215911865234,
      "learning_rate": 3.7072019574793034e-06,
      "loss": 0.4712,
      "step": 498
    },
    {
      "epoch": 2.352664576802508,
      "grad_norm": 0.48492008447647095,
      "learning_rate": 3.701329348821752e-06,
      "loss": 0.4521,
      "step": 499
    },
    {
      "epoch": 2.3573667711598745,
      "grad_norm": 0.49741214513778687,
      "learning_rate": 3.695448108265221e-06,
      "loss": 0.4378,
      "step": 500
    },
    {
      "epoch": 2.3620689655172415,
      "grad_norm": 0.5086454749107361,
      "learning_rate": 3.6895582780681254e-06,
      "loss": 0.4349,
      "step": 501
    },
    {
      "epoch": 2.366771159874608,
      "grad_norm": 0.49111631512641907,
      "learning_rate": 3.683659900550598e-06,
      "loss": 0.4625,
      "step": 502
    },
    {
      "epoch": 2.371473354231975,
      "grad_norm": 0.5006322264671326,
      "learning_rate": 3.6777530180941894e-06,
      "loss": 0.4457,
      "step": 503
    },
    {
      "epoch": 2.376175548589342,
      "grad_norm": 0.5934097170829773,
      "learning_rate": 3.671837673141559e-06,
      "loss": 0.4306,
      "step": 504
    },
    {
      "epoch": 2.3808777429467085,
      "grad_norm": 0.626039981842041,
      "learning_rate": 3.6659139081961707e-06,
      "loss": 0.4464,
      "step": 505
    },
    {
      "epoch": 2.385579937304075,
      "grad_norm": 0.4751131236553192,
      "learning_rate": 3.6599817658219916e-06,
      "loss": 0.4508,
      "step": 506
    },
    {
      "epoch": 2.3902821316614418,
      "grad_norm": 1.4542276859283447,
      "learning_rate": 3.6540412886431796e-06,
      "loss": 0.4606,
      "step": 507
    },
    {
      "epoch": 2.394984326018809,
      "grad_norm": 0.5189768075942993,
      "learning_rate": 3.648092519343783e-06,
      "loss": 0.4435,
      "step": 508
    },
    {
      "epoch": 2.3996865203761755,
      "grad_norm": 1.4583938121795654,
      "learning_rate": 3.642135500667431e-06,
      "loss": 0.4314,
      "step": 509
    },
    {
      "epoch": 2.4043887147335425,
      "grad_norm": 0.5038107633590698,
      "learning_rate": 3.6361702754170247e-06,
      "loss": 0.4463,
      "step": 510
    },
    {
      "epoch": 2.409090909090909,
      "grad_norm": 0.5786447525024414,
      "learning_rate": 3.630196886454435e-06,
      "loss": 0.4281,
      "step": 511
    },
    {
      "epoch": 2.413793103448276,
      "grad_norm": 0.48684218525886536,
      "learning_rate": 3.62421537670019e-06,
      "loss": 0.4432,
      "step": 512
    },
    {
      "epoch": 2.4184952978056424,
      "grad_norm": 0.5117013454437256,
      "learning_rate": 3.618225789133167e-06,
      "loss": 0.4464,
      "step": 513
    },
    {
      "epoch": 2.4231974921630095,
      "grad_norm": 0.49249181151390076,
      "learning_rate": 3.612228166790287e-06,
      "loss": 0.4465,
      "step": 514
    },
    {
      "epoch": 2.427899686520376,
      "grad_norm": 0.5761134624481201,
      "learning_rate": 3.606222552766201e-06,
      "loss": 0.4539,
      "step": 515
    },
    {
      "epoch": 2.4326018808777428,
      "grad_norm": 0.4839339256286621,
      "learning_rate": 3.6002089902129844e-06,
      "loss": 0.4469,
      "step": 516
    },
    {
      "epoch": 2.43730407523511,
      "grad_norm": 0.4765976369380951,
      "learning_rate": 3.5941875223398225e-06,
      "loss": 0.4379,
      "step": 517
    },
    {
      "epoch": 2.4420062695924765,
      "grad_norm": 0.5239338874816895,
      "learning_rate": 3.588158192412707e-06,
      "loss": 0.4354,
      "step": 518
    },
    {
      "epoch": 2.446708463949843,
      "grad_norm": 0.48244595527648926,
      "learning_rate": 3.582121043754116e-06,
      "loss": 0.438,
      "step": 519
    },
    {
      "epoch": 2.45141065830721,
      "grad_norm": 0.4641244411468506,
      "learning_rate": 3.5760761197427097e-06,
      "loss": 0.438,
      "step": 520
    },
    {
      "epoch": 2.456112852664577,
      "grad_norm": 0.48468074202537537,
      "learning_rate": 3.570023463813017e-06,
      "loss": 0.4306,
      "step": 521
    },
    {
      "epoch": 2.4608150470219434,
      "grad_norm": 0.48626402020454407,
      "learning_rate": 3.5639631194551216e-06,
      "loss": 0.4531,
      "step": 522
    },
    {
      "epoch": 2.4655172413793105,
      "grad_norm": 0.5581764578819275,
      "learning_rate": 3.557895130214352e-06,
      "loss": 0.4451,
      "step": 523
    },
    {
      "epoch": 2.470219435736677,
      "grad_norm": 0.6739279627799988,
      "learning_rate": 3.5518195396909653e-06,
      "loss": 0.4636,
      "step": 524
    },
    {
      "epoch": 2.4749216300940438,
      "grad_norm": 0.550710916519165,
      "learning_rate": 3.5457363915398384e-06,
      "loss": 0.4513,
      "step": 525
    },
    {
      "epoch": 2.479623824451411,
      "grad_norm": 0.479632705450058,
      "learning_rate": 3.539645729470151e-06,
      "loss": 0.4387,
      "step": 526
    },
    {
      "epoch": 2.4843260188087775,
      "grad_norm": 0.48741331696510315,
      "learning_rate": 3.5335475972450715e-06,
      "loss": 0.4388,
      "step": 527
    },
    {
      "epoch": 2.489028213166144,
      "grad_norm": 0.4964964985847473,
      "learning_rate": 3.5274420386814458e-06,
      "loss": 0.4643,
      "step": 528
    },
    {
      "epoch": 2.493730407523511,
      "grad_norm": 0.5134934186935425,
      "learning_rate": 3.521329097649478e-06,
      "loss": 0.4454,
      "step": 529
    },
    {
      "epoch": 2.498432601880878,
      "grad_norm": 0.4962058961391449,
      "learning_rate": 3.515208818072418e-06,
      "loss": 0.4408,
      "step": 530
    },
    {
      "epoch": 2.5031347962382444,
      "grad_norm": 0.5611489415168762,
      "learning_rate": 3.509081243926247e-06,
      "loss": 0.4306,
      "step": 531
    },
    {
      "epoch": 2.507836990595611,
      "grad_norm": 0.7012472748756409,
      "learning_rate": 3.5029464192393557e-06,
      "loss": 0.4614,
      "step": 532
    },
    {
      "epoch": 2.512539184952978,
      "grad_norm": 0.5351004004478455,
      "learning_rate": 3.4968043880922363e-06,
      "loss": 0.4151,
      "step": 533
    },
    {
      "epoch": 2.5172413793103448,
      "grad_norm": 0.5087808966636658,
      "learning_rate": 3.4906551946171603e-06,
      "loss": 0.4242,
      "step": 534
    },
    {
      "epoch": 2.521943573667712,
      "grad_norm": 0.5459093451499939,
      "learning_rate": 3.484498882997861e-06,
      "loss": 0.4215,
      "step": 535
    },
    {
      "epoch": 2.5266457680250785,
      "grad_norm": 0.49804285168647766,
      "learning_rate": 3.478335497469219e-06,
      "loss": 0.4492,
      "step": 536
    },
    {
      "epoch": 2.531347962382445,
      "grad_norm": 0.4959704875946045,
      "learning_rate": 3.472165082316943e-06,
      "loss": 0.4511,
      "step": 537
    },
    {
      "epoch": 2.5360501567398117,
      "grad_norm": 0.5059382319450378,
      "learning_rate": 3.465987681877251e-06,
      "loss": 0.4419,
      "step": 538
    },
    {
      "epoch": 2.540752351097179,
      "grad_norm": 0.7398380637168884,
      "learning_rate": 3.4598033405365527e-06,
      "loss": 0.4548,
      "step": 539
    },
    {
      "epoch": 2.5454545454545454,
      "grad_norm": 0.5326687693595886,
      "learning_rate": 3.45361210273113e-06,
      "loss": 0.4473,
      "step": 540
    },
    {
      "epoch": 2.5501567398119125,
      "grad_norm": 0.5069761872291565,
      "learning_rate": 3.447414012946818e-06,
      "loss": 0.4343,
      "step": 541
    },
    {
      "epoch": 2.554858934169279,
      "grad_norm": 0.45915964245796204,
      "learning_rate": 3.4412091157186853e-06,
      "loss": 0.4499,
      "step": 542
    },
    {
      "epoch": 2.5595611285266457,
      "grad_norm": 0.5174360275268555,
      "learning_rate": 3.4349974556307146e-06,
      "loss": 0.44,
      "step": 543
    },
    {
      "epoch": 2.5642633228840124,
      "grad_norm": 0.5008105039596558,
      "learning_rate": 3.4287790773154807e-06,
      "loss": 0.4648,
      "step": 544
    },
    {
      "epoch": 2.5689655172413794,
      "grad_norm": 0.5628801584243774,
      "learning_rate": 3.4225540254538297e-06,
      "loss": 0.462,
      "step": 545
    },
    {
      "epoch": 2.573667711598746,
      "grad_norm": 0.9913654923439026,
      "learning_rate": 3.416322344774562e-06,
      "loss": 0.4403,
      "step": 546
    },
    {
      "epoch": 2.5783699059561127,
      "grad_norm": 0.5034172534942627,
      "learning_rate": 3.4100840800541055e-06,
      "loss": 0.4622,
      "step": 547
    },
    {
      "epoch": 2.58307210031348,
      "grad_norm": 0.495516836643219,
      "learning_rate": 3.4038392761161986e-06,
      "loss": 0.4523,
      "step": 548
    },
    {
      "epoch": 2.5877742946708464,
      "grad_norm": 0.48142367601394653,
      "learning_rate": 3.3975879778315634e-06,
      "loss": 0.4242,
      "step": 549
    },
    {
      "epoch": 2.592476489028213,
      "grad_norm": 0.4635900557041168,
      "learning_rate": 3.391330230117587e-06,
      "loss": 0.3949,
      "step": 550
    },
    {
      "epoch": 2.5971786833855797,
      "grad_norm": 0.4769044816493988,
      "learning_rate": 3.385066077937997e-06,
      "loss": 0.4651,
      "step": 551
    },
    {
      "epoch": 2.6018808777429467,
      "grad_norm": 1.059553861618042,
      "learning_rate": 3.378795566302541e-06,
      "loss": 0.4243,
      "step": 552
    },
    {
      "epoch": 2.6065830721003134,
      "grad_norm": 0.512134850025177,
      "learning_rate": 3.372518740266658e-06,
      "loss": 0.4435,
      "step": 553
    },
    {
      "epoch": 2.6112852664576804,
      "grad_norm": 0.5267173647880554,
      "learning_rate": 3.36623564493116e-06,
      "loss": 0.4558,
      "step": 554
    },
    {
      "epoch": 2.615987460815047,
      "grad_norm": 0.49343907833099365,
      "learning_rate": 3.3599463254419047e-06,
      "loss": 0.4598,
      "step": 555
    },
    {
      "epoch": 2.6206896551724137,
      "grad_norm": 0.5496839284896851,
      "learning_rate": 3.3536508269894724e-06,
      "loss": 0.4669,
      "step": 556
    },
    {
      "epoch": 2.6253918495297803,
      "grad_norm": 0.5957831740379333,
      "learning_rate": 3.347349194808842e-06,
      "loss": 0.4533,
      "step": 557
    },
    {
      "epoch": 2.6300940438871474,
      "grad_norm": 0.5049230456352234,
      "learning_rate": 3.3410414741790625e-06,
      "loss": 0.4293,
      "step": 558
    },
    {
      "epoch": 2.634796238244514,
      "grad_norm": 0.5167728066444397,
      "learning_rate": 3.3347277104229332e-06,
      "loss": 0.443,
      "step": 559
    },
    {
      "epoch": 2.639498432601881,
      "grad_norm": 0.6090758442878723,
      "learning_rate": 3.3284079489066728e-06,
      "loss": 0.4378,
      "step": 560
    },
    {
      "epoch": 2.6442006269592477,
      "grad_norm": 0.5165027379989624,
      "learning_rate": 3.3220822350395966e-06,
      "loss": 0.4302,
      "step": 561
    },
    {
      "epoch": 2.6489028213166144,
      "grad_norm": 0.5152680277824402,
      "learning_rate": 3.31575061427379e-06,
      "loss": 0.4311,
      "step": 562
    },
    {
      "epoch": 2.653605015673981,
      "grad_norm": 0.547235906124115,
      "learning_rate": 3.3094131321037783e-06,
      "loss": 0.4371,
      "step": 563
    },
    {
      "epoch": 2.658307210031348,
      "grad_norm": 0.521981418132782,
      "learning_rate": 3.303069834066206e-06,
      "loss": 0.4346,
      "step": 564
    },
    {
      "epoch": 2.6630094043887147,
      "grad_norm": 0.5127217769622803,
      "learning_rate": 3.2967207657395055e-06,
      "loss": 0.474,
      "step": 565
    },
    {
      "epoch": 2.6677115987460818,
      "grad_norm": 0.5210872888565063,
      "learning_rate": 3.2903659727435692e-06,
      "loss": 0.4622,
      "step": 566
    },
    {
      "epoch": 2.6724137931034484,
      "grad_norm": 0.5768873691558838,
      "learning_rate": 3.284005500739423e-06,
      "loss": 0.4556,
      "step": 567
    },
    {
      "epoch": 2.677115987460815,
      "grad_norm": 0.5305764675140381,
      "learning_rate": 3.2776393954289e-06,
      "loss": 0.429,
      "step": 568
    },
    {
      "epoch": 2.6818181818181817,
      "grad_norm": 0.5312129855155945,
      "learning_rate": 3.271267702554307e-06,
      "loss": 0.4208,
      "step": 569
    },
    {
      "epoch": 2.6865203761755487,
      "grad_norm": 0.5433884859085083,
      "learning_rate": 3.2648904678981032e-06,
      "loss": 0.4647,
      "step": 570
    },
    {
      "epoch": 2.6912225705329154,
      "grad_norm": 1.2331725358963013,
      "learning_rate": 3.2585077372825636e-06,
      "loss": 0.4126,
      "step": 571
    },
    {
      "epoch": 2.695924764890282,
      "grad_norm": 0.5495198369026184,
      "learning_rate": 3.2521195565694543e-06,
      "loss": 0.4453,
      "step": 572
    },
    {
      "epoch": 2.700626959247649,
      "grad_norm": 0.5230907201766968,
      "learning_rate": 3.2457259716597023e-06,
      "loss": 0.446,
      "step": 573
    },
    {
      "epoch": 2.7053291536050157,
      "grad_norm": 0.4807503819465637,
      "learning_rate": 3.2393270284930658e-06,
      "loss": 0.4547,
      "step": 574
    },
    {
      "epoch": 2.7100313479623823,
      "grad_norm": 0.5169614553451538,
      "learning_rate": 3.2329227730478026e-06,
      "loss": 0.4319,
      "step": 575
    },
    {
      "epoch": 2.714733542319749,
      "grad_norm": 0.502966046333313,
      "learning_rate": 3.2265132513403415e-06,
      "loss": 0.4196,
      "step": 576
    },
    {
      "epoch": 2.719435736677116,
      "grad_norm": 0.5387672781944275,
      "learning_rate": 3.22009850942495e-06,
      "loss": 0.4449,
      "step": 577
    },
    {
      "epoch": 2.7241379310344827,
      "grad_norm": 0.5503709316253662,
      "learning_rate": 3.213678593393405e-06,
      "loss": 0.4589,
      "step": 578
    },
    {
      "epoch": 2.7288401253918497,
      "grad_norm": 0.5165039300918579,
      "learning_rate": 3.207253549374662e-06,
      "loss": 0.4578,
      "step": 579
    },
    {
      "epoch": 2.7335423197492164,
      "grad_norm": 0.5894023180007935,
      "learning_rate": 3.200823423534519e-06,
      "loss": 0.4448,
      "step": 580
    },
    {
      "epoch": 2.738244514106583,
      "grad_norm": 0.5234156250953674,
      "learning_rate": 3.194388262075293e-06,
      "loss": 0.4504,
      "step": 581
    },
    {
      "epoch": 2.7429467084639496,
      "grad_norm": 0.47498077154159546,
      "learning_rate": 3.1879481112354804e-06,
      "loss": 0.4471,
      "step": 582
    },
    {
      "epoch": 2.7476489028213167,
      "grad_norm": 0.5213322043418884,
      "learning_rate": 3.181503017289428e-06,
      "loss": 0.4096,
      "step": 583
    },
    {
      "epoch": 2.7523510971786833,
      "grad_norm": 0.5031464695930481,
      "learning_rate": 3.175053026547002e-06,
      "loss": 0.416,
      "step": 584
    },
    {
      "epoch": 2.7570532915360504,
      "grad_norm": 0.7983574867248535,
      "learning_rate": 3.16859818535325e-06,
      "loss": 0.457,
      "step": 585
    },
    {
      "epoch": 2.761755485893417,
      "grad_norm": 0.47774994373321533,
      "learning_rate": 3.1621385400880756e-06,
      "loss": 0.4529,
      "step": 586
    },
    {
      "epoch": 2.7664576802507836,
      "grad_norm": 0.8216882348060608,
      "learning_rate": 3.1556741371658984e-06,
      "loss": 0.4559,
      "step": 587
    },
    {
      "epoch": 2.7711598746081503,
      "grad_norm": 0.5124049186706543,
      "learning_rate": 3.1492050230353238e-06,
      "loss": 0.4438,
      "step": 588
    },
    {
      "epoch": 2.7758620689655173,
      "grad_norm": 0.5410915017127991,
      "learning_rate": 3.142731244178809e-06,
      "loss": 0.4195,
      "step": 589
    },
    {
      "epoch": 2.780564263322884,
      "grad_norm": 0.5318175554275513,
      "learning_rate": 3.1362528471123277e-06,
      "loss": 0.4046,
      "step": 590
    },
    {
      "epoch": 2.785266457680251,
      "grad_norm": 0.6133676171302795,
      "learning_rate": 3.129769878385039e-06,
      "loss": 0.4098,
      "step": 591
    },
    {
      "epoch": 2.7899686520376177,
      "grad_norm": 0.4698888063430786,
      "learning_rate": 3.1232823845789473e-06,
      "loss": 0.4508,
      "step": 592
    },
    {
      "epoch": 2.7946708463949843,
      "grad_norm": 0.6980767250061035,
      "learning_rate": 3.1167904123085736e-06,
      "loss": 0.455,
      "step": 593
    },
    {
      "epoch": 2.799373040752351,
      "grad_norm": 0.5151284337043762,
      "learning_rate": 3.110294008220617e-06,
      "loss": 0.4431,
      "step": 594
    },
    {
      "epoch": 2.804075235109718,
      "grad_norm": 0.47901320457458496,
      "learning_rate": 3.1037932189936205e-06,
      "loss": 0.4406,
      "step": 595
    },
    {
      "epoch": 2.8087774294670846,
      "grad_norm": 0.5079891085624695,
      "learning_rate": 3.097288091337635e-06,
      "loss": 0.4351,
      "step": 596
    },
    {
      "epoch": 2.8134796238244513,
      "grad_norm": 0.5278874635696411,
      "learning_rate": 3.0907786719938876e-06,
      "loss": 0.4264,
      "step": 597
    },
    {
      "epoch": 2.8181818181818183,
      "grad_norm": 0.47123396396636963,
      "learning_rate": 3.084265007734436e-06,
      "loss": 0.434,
      "step": 598
    },
    {
      "epoch": 2.822884012539185,
      "grad_norm": 0.5229635834693909,
      "learning_rate": 3.0777471453618457e-06,
      "loss": 0.4602,
      "step": 599
    },
    {
      "epoch": 2.8275862068965516,
      "grad_norm": 0.47847074270248413,
      "learning_rate": 3.0712251317088426e-06,
      "loss": 0.4317,
      "step": 600
    },
    {
      "epoch": 2.8322884012539182,
      "grad_norm": 0.7754543423652649,
      "learning_rate": 3.064699013637983e-06,
      "loss": 0.4528,
      "step": 601
    },
    {
      "epoch": 2.8369905956112853,
      "grad_norm": 0.5581084489822388,
      "learning_rate": 3.0581688380413115e-06,
      "loss": 0.4369,
      "step": 602
    },
    {
      "epoch": 2.841692789968652,
      "grad_norm": 0.588622510433197,
      "learning_rate": 3.0516346518400315e-06,
      "loss": 0.4517,
      "step": 603
    },
    {
      "epoch": 2.846394984326019,
      "grad_norm": 0.565423846244812,
      "learning_rate": 3.0450965019841593e-06,
      "loss": 0.4517,
      "step": 604
    },
    {
      "epoch": 2.8510971786833856,
      "grad_norm": 0.47801777720451355,
      "learning_rate": 3.0385544354521957e-06,
      "loss": 0.4161,
      "step": 605
    },
    {
      "epoch": 2.8557993730407523,
      "grad_norm": 0.5034862756729126,
      "learning_rate": 3.0320084992507814e-06,
      "loss": 0.4428,
      "step": 606
    },
    {
      "epoch": 2.860501567398119,
      "grad_norm": 0.5339663624763489,
      "learning_rate": 3.0254587404143604e-06,
      "loss": 0.4792,
      "step": 607
    },
    {
      "epoch": 2.865203761755486,
      "grad_norm": 0.48184943199157715,
      "learning_rate": 3.0189052060048464e-06,
      "loss": 0.4409,
      "step": 608
    },
    {
      "epoch": 2.8699059561128526,
      "grad_norm": 0.5102176070213318,
      "learning_rate": 3.01234794311128e-06,
      "loss": 0.438,
      "step": 609
    },
    {
      "epoch": 2.8746081504702197,
      "grad_norm": 0.5111781358718872,
      "learning_rate": 3.0057869988494925e-06,
      "loss": 0.4617,
      "step": 610
    },
    {
      "epoch": 2.8793103448275863,
      "grad_norm": 0.5915101766586304,
      "learning_rate": 2.999222420361767e-06,
      "loss": 0.4532,
      "step": 611
    },
    {
      "epoch": 2.884012539184953,
      "grad_norm": 0.48898932337760925,
      "learning_rate": 2.9926542548165e-06,
      "loss": 0.4663,
      "step": 612
    },
    {
      "epoch": 2.8887147335423196,
      "grad_norm": 0.4943861961364746,
      "learning_rate": 2.9860825494078605e-06,
      "loss": 0.4354,
      "step": 613
    },
    {
      "epoch": 2.8934169278996866,
      "grad_norm": 0.5398025512695312,
      "learning_rate": 2.979507351355454e-06,
      "loss": 0.4546,
      "step": 614
    },
    {
      "epoch": 2.8981191222570533,
      "grad_norm": 0.545421302318573,
      "learning_rate": 2.972928707903981e-06,
      "loss": 0.4404,
      "step": 615
    },
    {
      "epoch": 2.9028213166144203,
      "grad_norm": 0.5370550751686096,
      "learning_rate": 2.966346666322898e-06,
      "loss": 0.4401,
      "step": 616
    },
    {
      "epoch": 2.907523510971787,
      "grad_norm": 0.5280672311782837,
      "learning_rate": 2.9597612739060775e-06,
      "loss": 0.4172,
      "step": 617
    },
    {
      "epoch": 2.9122257053291536,
      "grad_norm": 0.5043423175811768,
      "learning_rate": 2.9531725779714713e-06,
      "loss": 0.4487,
      "step": 618
    },
    {
      "epoch": 2.91692789968652,
      "grad_norm": 1.961200475692749,
      "learning_rate": 2.9465806258607653e-06,
      "loss": 0.4548,
      "step": 619
    },
    {
      "epoch": 2.9216300940438873,
      "grad_norm": 0.5286726355552673,
      "learning_rate": 2.939985464939043e-06,
      "loss": 0.4566,
      "step": 620
    },
    {
      "epoch": 2.926332288401254,
      "grad_norm": 0.5209453105926514,
      "learning_rate": 2.9333871425944434e-06,
      "loss": 0.4064,
      "step": 621
    },
    {
      "epoch": 2.9310344827586206,
      "grad_norm": 0.47711747884750366,
      "learning_rate": 2.926785706237822e-06,
      "loss": 0.4341,
      "step": 622
    },
    {
      "epoch": 2.9357366771159876,
      "grad_norm": 0.45926427841186523,
      "learning_rate": 2.920181203302409e-06,
      "loss": 0.4256,
      "step": 623
    },
    {
      "epoch": 2.9404388714733543,
      "grad_norm": 0.5624600648880005,
      "learning_rate": 2.91357368124347e-06,
      "loss": 0.4252,
      "step": 624
    },
    {
      "epoch": 2.945141065830721,
      "grad_norm": 0.5101850628852844,
      "learning_rate": 2.906963187537962e-06,
      "loss": 0.4352,
      "step": 625
    },
    {
      "epoch": 2.9498432601880875,
      "grad_norm": 0.5341358184814453,
      "learning_rate": 2.9003497696841955e-06,
      "loss": 0.4132,
      "step": 626
    },
    {
      "epoch": 2.9545454545454546,
      "grad_norm": 0.5917084217071533,
      "learning_rate": 2.8937334752014913e-06,
      "loss": 0.4693,
      "step": 627
    },
    {
      "epoch": 2.959247648902821,
      "grad_norm": 0.793695330619812,
      "learning_rate": 2.887114351629839e-06,
      "loss": 0.4431,
      "step": 628
    },
    {
      "epoch": 2.9639498432601883,
      "grad_norm": 0.5363728404045105,
      "learning_rate": 2.8804924465295575e-06,
      "loss": 0.4672,
      "step": 629
    },
    {
      "epoch": 2.968652037617555,
      "grad_norm": 0.4979572892189026,
      "learning_rate": 2.873867807480951e-06,
      "loss": 0.4723,
      "step": 630
    },
    {
      "epoch": 2.9733542319749215,
      "grad_norm": 0.5310130715370178,
      "learning_rate": 2.8672404820839676e-06,
      "loss": 0.4388,
      "step": 631
    },
    {
      "epoch": 2.978056426332288,
      "grad_norm": 0.530015766620636,
      "learning_rate": 2.8606105179578584e-06,
      "loss": 0.4466,
      "step": 632
    },
    {
      "epoch": 2.9827586206896552,
      "grad_norm": 0.5356627702713013,
      "learning_rate": 2.8539779627408332e-06,
      "loss": 0.4252,
      "step": 633
    },
    {
      "epoch": 2.987460815047022,
      "grad_norm": 0.5290245413780212,
      "learning_rate": 2.847342864089721e-06,
      "loss": 0.4453,
      "step": 634
    },
    {
      "epoch": 2.992163009404389,
      "grad_norm": 0.471682071685791,
      "learning_rate": 2.8407052696796255e-06,
      "loss": 0.43,
      "step": 635
    },
    {
      "epoch": 2.9968652037617556,
      "grad_norm": 0.5220829844474792,
      "learning_rate": 2.834065227203584e-06,
      "loss": 0.4494,
      "step": 636
    },
    {
      "epoch": 3.0047021943573666,
      "grad_norm": 0.4797399342060089,
      "learning_rate": 2.8274227843722213e-06,
      "loss": 0.8683,
      "step": 637
    },
    {
      "epoch": 3.0094043887147337,
      "grad_norm": 0.5463248491287231,
      "learning_rate": 2.820777988913412e-06,
      "loss": 0.4157,
      "step": 638
    },
    {
      "epoch": 3.0141065830721003,
      "grad_norm": 0.5081924200057983,
      "learning_rate": 2.8141308885719337e-06,
      "loss": 0.4169,
      "step": 639
    },
    {
      "epoch": 3.018808777429467,
      "grad_norm": 0.4916677474975586,
      "learning_rate": 2.8074815311091265e-06,
      "loss": 0.3898,
      "step": 640
    },
    {
      "epoch": 3.023510971786834,
      "grad_norm": 0.48858827352523804,
      "learning_rate": 2.8008299643025477e-06,
      "loss": 0.4319,
      "step": 641
    },
    {
      "epoch": 3.0282131661442007,
      "grad_norm": 0.49183058738708496,
      "learning_rate": 2.7941762359456294e-06,
      "loss": 0.4243,
      "step": 642
    },
    {
      "epoch": 3.0329153605015673,
      "grad_norm": 0.5068245530128479,
      "learning_rate": 2.787520393847334e-06,
      "loss": 0.4168,
      "step": 643
    },
    {
      "epoch": 3.0376175548589344,
      "grad_norm": 0.542245090007782,
      "learning_rate": 2.780862485831814e-06,
      "loss": 0.4289,
      "step": 644
    },
    {
      "epoch": 3.042319749216301,
      "grad_norm": 0.49114999175071716,
      "learning_rate": 2.7742025597380644e-06,
      "loss": 0.4337,
      "step": 645
    },
    {
      "epoch": 3.0470219435736676,
      "grad_norm": 0.4982999563217163,
      "learning_rate": 2.7675406634195824e-06,
      "loss": 0.4207,
      "step": 646
    },
    {
      "epoch": 3.0517241379310347,
      "grad_norm": 0.5352709293365479,
      "learning_rate": 2.7608768447440193e-06,
      "loss": 0.4087,
      "step": 647
    },
    {
      "epoch": 3.0564263322884013,
      "grad_norm": 0.5486279726028442,
      "learning_rate": 2.754211151592841e-06,
      "loss": 0.4129,
      "step": 648
    },
    {
      "epoch": 3.061128526645768,
      "grad_norm": 0.6048034429550171,
      "learning_rate": 2.7475436318609827e-06,
      "loss": 0.433,
      "step": 649
    },
    {
      "epoch": 3.0658307210031346,
      "grad_norm": 0.6576470136642456,
      "learning_rate": 2.7408743334565006e-06,
      "loss": 0.4086,
      "step": 650
    },
    {
      "epoch": 3.0705329153605017,
      "grad_norm": 0.49989938735961914,
      "learning_rate": 2.734203304300235e-06,
      "loss": 0.3999,
      "step": 651
    },
    {
      "epoch": 3.0752351097178683,
      "grad_norm": 0.5238141417503357,
      "learning_rate": 2.7275305923254607e-06,
      "loss": 0.4133,
      "step": 652
    },
    {
      "epoch": 3.079937304075235,
      "grad_norm": 0.5244804620742798,
      "learning_rate": 2.720856245477544e-06,
      "loss": 0.4016,
      "step": 653
    },
    {
      "epoch": 3.084639498432602,
      "grad_norm": 0.5036159753799438,
      "learning_rate": 2.7141803117135978e-06,
      "loss": 0.3972,
      "step": 654
    },
    {
      "epoch": 3.0893416927899686,
      "grad_norm": 0.5390443801879883,
      "learning_rate": 2.7075028390021385e-06,
      "loss": 0.3992,
      "step": 655
    },
    {
      "epoch": 3.0940438871473352,
      "grad_norm": 0.5226757526397705,
      "learning_rate": 2.7008238753227385e-06,
      "loss": 0.4074,
      "step": 656
    },
    {
      "epoch": 3.0987460815047023,
      "grad_norm": 0.48386913537979126,
      "learning_rate": 2.694143468665685e-06,
      "loss": 0.4284,
      "step": 657
    },
    {
      "epoch": 3.103448275862069,
      "grad_norm": 0.5081993341445923,
      "learning_rate": 2.6874616670316338e-06,
      "loss": 0.3952,
      "step": 658
    },
    {
      "epoch": 3.1081504702194356,
      "grad_norm": 0.538280189037323,
      "learning_rate": 2.6807785184312618e-06,
      "loss": 0.4136,
      "step": 659
    },
    {
      "epoch": 3.1128526645768027,
      "grad_norm": 0.7804566621780396,
      "learning_rate": 2.674094070884926e-06,
      "loss": 0.4131,
      "step": 660
    },
    {
      "epoch": 3.1175548589341693,
      "grad_norm": 0.6693199872970581,
      "learning_rate": 2.6674083724223166e-06,
      "loss": 0.4329,
      "step": 661
    },
    {
      "epoch": 3.122257053291536,
      "grad_norm": 0.5034769773483276,
      "learning_rate": 2.6607214710821112e-06,
      "loss": 0.4062,
      "step": 662
    },
    {
      "epoch": 3.126959247648903,
      "grad_norm": 0.5518231391906738,
      "learning_rate": 2.6540334149116304e-06,
      "loss": 0.4172,
      "step": 663
    },
    {
      "epoch": 3.1316614420062696,
      "grad_norm": 0.5797336101531982,
      "learning_rate": 2.647344251966493e-06,
      "loss": 0.4164,
      "step": 664
    },
    {
      "epoch": 3.1363636363636362,
      "grad_norm": 0.5404736399650574,
      "learning_rate": 2.6406540303102714e-06,
      "loss": 0.4157,
      "step": 665
    },
    {
      "epoch": 3.1410658307210033,
      "grad_norm": 0.5246729850769043,
      "learning_rate": 2.6339627980141425e-06,
      "loss": 0.4165,
      "step": 666
    },
    {
      "epoch": 3.14576802507837,
      "grad_norm": 0.5443553328514099,
      "learning_rate": 2.6272706031565482e-06,
      "loss": 0.4022,
      "step": 667
    },
    {
      "epoch": 3.1504702194357366,
      "grad_norm": 0.5127459168434143,
      "learning_rate": 2.6205774938228433e-06,
      "loss": 0.3983,
      "step": 668
    },
    {
      "epoch": 3.1551724137931036,
      "grad_norm": 0.5095480680465698,
      "learning_rate": 2.6138835181049556e-06,
      "loss": 0.4227,
      "step": 669
    },
    {
      "epoch": 3.1598746081504703,
      "grad_norm": 0.5238015651702881,
      "learning_rate": 2.6071887241010374e-06,
      "loss": 0.4056,
      "step": 670
    },
    {
      "epoch": 3.164576802507837,
      "grad_norm": 0.5659390687942505,
      "learning_rate": 2.6004931599151223e-06,
      "loss": 0.3933,
      "step": 671
    },
    {
      "epoch": 3.169278996865204,
      "grad_norm": 0.528191328048706,
      "learning_rate": 2.593796873656775e-06,
      "loss": 0.4356,
      "step": 672
    },
    {
      "epoch": 3.1739811912225706,
      "grad_norm": 1.1774086952209473,
      "learning_rate": 2.587099913440749e-06,
      "loss": 0.4149,
      "step": 673
    },
    {
      "epoch": 3.1786833855799372,
      "grad_norm": 0.5629571676254272,
      "learning_rate": 2.580402327386643e-06,
      "loss": 0.403,
      "step": 674
    },
    {
      "epoch": 3.183385579937304,
      "grad_norm": 1.1260513067245483,
      "learning_rate": 2.5737041636185496e-06,
      "loss": 0.4102,
      "step": 675
    },
    {
      "epoch": 3.188087774294671,
      "grad_norm": 0.6467511653900146,
      "learning_rate": 2.5670054702647146e-06,
      "loss": 0.3948,
      "step": 676
    },
    {
      "epoch": 3.1927899686520376,
      "grad_norm": 0.5177720785140991,
      "learning_rate": 2.5603062954571872e-06,
      "loss": 0.4188,
      "step": 677
    },
    {
      "epoch": 3.197492163009404,
      "grad_norm": 0.5086417198181152,
      "learning_rate": 2.553606687331477e-06,
      "loss": 0.4403,
      "step": 678
    },
    {
      "epoch": 3.2021943573667713,
      "grad_norm": 0.5762012600898743,
      "learning_rate": 2.5469066940262073e-06,
      "loss": 0.4084,
      "step": 679
    },
    {
      "epoch": 3.206896551724138,
      "grad_norm": 0.5122736692428589,
      "learning_rate": 2.540206363682768e-06,
      "loss": 0.4005,
      "step": 680
    },
    {
      "epoch": 3.2115987460815045,
      "grad_norm": 0.5179394483566284,
      "learning_rate": 2.533505744444972e-06,
      "loss": 0.419,
      "step": 681
    },
    {
      "epoch": 3.2163009404388716,
      "grad_norm": 0.5541443824768066,
      "learning_rate": 2.526804884458707e-06,
      "loss": 0.4112,
      "step": 682
    },
    {
      "epoch": 3.2210031347962382,
      "grad_norm": 0.5687317252159119,
      "learning_rate": 2.520103831871591e-06,
      "loss": 0.4145,
      "step": 683
    },
    {
      "epoch": 3.225705329153605,
      "grad_norm": 0.5060294270515442,
      "learning_rate": 2.513402634832627e-06,
      "loss": 0.3933,
      "step": 684
    },
    {
      "epoch": 3.230407523510972,
      "grad_norm": 0.6311008930206299,
      "learning_rate": 2.5067013414918523e-06,
      "loss": 0.401,
      "step": 685
    },
    {
      "epoch": 3.2351097178683386,
      "grad_norm": 0.5575832724571228,
      "learning_rate": 2.5e-06,
      "loss": 0.4127,
      "step": 686
    },
    {
      "epoch": 3.239811912225705,
      "grad_norm": 0.5105507373809814,
      "learning_rate": 2.493298658508149e-06,
      "loss": 0.3971,
      "step": 687
    },
    {
      "epoch": 3.2445141065830723,
      "grad_norm": 0.5813129544258118,
      "learning_rate": 2.4865973651673743e-06,
      "loss": 0.4136,
      "step": 688
    },
    {
      "epoch": 3.249216300940439,
      "grad_norm": 0.5921242833137512,
      "learning_rate": 2.4798961681284096e-06,
      "loss": 0.437,
      "step": 689
    },
    {
      "epoch": 3.2539184952978055,
      "grad_norm": 0.5654864311218262,
      "learning_rate": 2.473195115541293e-06,
      "loss": 0.3939,
      "step": 690
    },
    {
      "epoch": 3.2586206896551726,
      "grad_norm": 0.5103882551193237,
      "learning_rate": 2.466494255555029e-06,
      "loss": 0.4394,
      "step": 691
    },
    {
      "epoch": 3.2633228840125392,
      "grad_norm": 0.5423967242240906,
      "learning_rate": 2.459793636317233e-06,
      "loss": 0.4048,
      "step": 692
    },
    {
      "epoch": 3.268025078369906,
      "grad_norm": 0.6185951828956604,
      "learning_rate": 2.4530933059737936e-06,
      "loss": 0.4432,
      "step": 693
    },
    {
      "epoch": 3.2727272727272725,
      "grad_norm": 0.6062753796577454,
      "learning_rate": 2.4463933126685236e-06,
      "loss": 0.4061,
      "step": 694
    },
    {
      "epoch": 3.2774294670846396,
      "grad_norm": 0.5118281841278076,
      "learning_rate": 2.439693704542814e-06,
      "loss": 0.4008,
      "step": 695
    },
    {
      "epoch": 3.282131661442006,
      "grad_norm": 0.9080231785774231,
      "learning_rate": 2.432994529735286e-06,
      "loss": 0.409,
      "step": 696
    },
    {
      "epoch": 3.2868338557993733,
      "grad_norm": 0.550635814666748,
      "learning_rate": 2.4262958363814512e-06,
      "loss": 0.4202,
      "step": 697
    },
    {
      "epoch": 3.29153605015674,
      "grad_norm": 0.5728116631507874,
      "learning_rate": 2.4195976726133574e-06,
      "loss": 0.406,
      "step": 698
    },
    {
      "epoch": 3.2962382445141065,
      "grad_norm": 0.4995472729206085,
      "learning_rate": 2.4129000865592517e-06,
      "loss": 0.4063,
      "step": 699
    },
    {
      "epoch": 3.300940438871473,
      "grad_norm": 0.601259708404541,
      "learning_rate": 2.4062031263432267e-06,
      "loss": 0.4268,
      "step": 700
    },
    {
      "epoch": 3.30564263322884,
      "grad_norm": 0.570606529712677,
      "learning_rate": 2.3995068400848785e-06,
      "loss": 0.4034,
      "step": 701
    },
    {
      "epoch": 3.310344827586207,
      "grad_norm": 0.5638160705566406,
      "learning_rate": 2.392811275898963e-06,
      "loss": 0.4212,
      "step": 702
    },
    {
      "epoch": 3.3150470219435735,
      "grad_norm": 0.5354572534561157,
      "learning_rate": 2.3861164818950448e-06,
      "loss": 0.3893,
      "step": 703
    },
    {
      "epoch": 3.3197492163009406,
      "grad_norm": 0.5149163603782654,
      "learning_rate": 2.379422506177157e-06,
      "loss": 0.4126,
      "step": 704
    },
    {
      "epoch": 3.324451410658307,
      "grad_norm": 0.5132194757461548,
      "learning_rate": 2.372729396843453e-06,
      "loss": 0.4132,
      "step": 705
    },
    {
      "epoch": 3.329153605015674,
      "grad_norm": 0.5163543224334717,
      "learning_rate": 2.366037201985858e-06,
      "loss": 0.418,
      "step": 706
    },
    {
      "epoch": 3.333855799373041,
      "grad_norm": 0.5132508277893066,
      "learning_rate": 2.3593459696897294e-06,
      "loss": 0.3944,
      "step": 707
    },
    {
      "epoch": 3.3385579937304075,
      "grad_norm": 0.5490009188652039,
      "learning_rate": 2.352655748033508e-06,
      "loss": 0.414,
      "step": 708
    },
    {
      "epoch": 3.343260188087774,
      "grad_norm": 0.5879104733467102,
      "learning_rate": 2.3459665850883704e-06,
      "loss": 0.4344,
      "step": 709
    },
    {
      "epoch": 3.347962382445141,
      "grad_norm": 0.5451306700706482,
      "learning_rate": 2.33927852891789e-06,
      "loss": 0.4208,
      "step": 710
    },
    {
      "epoch": 3.352664576802508,
      "grad_norm": 0.5207070708274841,
      "learning_rate": 2.3325916275776834e-06,
      "loss": 0.4398,
      "step": 711
    },
    {
      "epoch": 3.3573667711598745,
      "grad_norm": 0.5440477132797241,
      "learning_rate": 2.3259059291150744e-06,
      "loss": 0.4015,
      "step": 712
    },
    {
      "epoch": 3.3620689655172415,
      "grad_norm": 0.5619958639144897,
      "learning_rate": 2.319221481568739e-06,
      "loss": 0.4196,
      "step": 713
    },
    {
      "epoch": 3.366771159874608,
      "grad_norm": 0.6007470488548279,
      "learning_rate": 2.3125383329683666e-06,
      "loss": 0.4217,
      "step": 714
    },
    {
      "epoch": 3.371473354231975,
      "grad_norm": 0.4972032904624939,
      "learning_rate": 2.3058565313343152e-06,
      "loss": 0.3904,
      "step": 715
    },
    {
      "epoch": 3.376175548589342,
      "grad_norm": 0.5420966148376465,
      "learning_rate": 2.2991761246772623e-06,
      "loss": 0.4048,
      "step": 716
    },
    {
      "epoch": 3.3808777429467085,
      "grad_norm": 0.520063042640686,
      "learning_rate": 2.2924971609978623e-06,
      "loss": 0.3965,
      "step": 717
    },
    {
      "epoch": 3.385579937304075,
      "grad_norm": 0.8903913497924805,
      "learning_rate": 2.285819688286403e-06,
      "loss": 0.3873,
      "step": 718
    },
    {
      "epoch": 3.3902821316614418,
      "grad_norm": 0.5380633473396301,
      "learning_rate": 2.2791437545224563e-06,
      "loss": 0.4335,
      "step": 719
    },
    {
      "epoch": 3.394984326018809,
      "grad_norm": 0.5058356523513794,
      "learning_rate": 2.2724694076745397e-06,
      "loss": 0.4134,
      "step": 720
    },
    {
      "epoch": 3.3996865203761755,
      "grad_norm": 0.5383400321006775,
      "learning_rate": 2.265796695699766e-06,
      "loss": 0.4154,
      "step": 721
    },
    {
      "epoch": 3.4043887147335425,
      "grad_norm": 0.5831345319747925,
      "learning_rate": 2.2591256665434998e-06,
      "loss": 0.4193,
      "step": 722
    },
    {
      "epoch": 3.409090909090909,
      "grad_norm": 0.5494023561477661,
      "learning_rate": 2.252456368139019e-06,
      "loss": 0.4137,
      "step": 723
    },
    {
      "epoch": 3.413793103448276,
      "grad_norm": 0.5735755562782288,
      "learning_rate": 2.245788848407159e-06,
      "loss": 0.4211,
      "step": 724
    },
    {
      "epoch": 3.4184952978056424,
      "grad_norm": 0.5244953036308289,
      "learning_rate": 2.2391231552559815e-06,
      "loss": 0.4194,
      "step": 725
    },
    {
      "epoch": 3.4231974921630095,
      "grad_norm": 0.5803194642066956,
      "learning_rate": 2.2324593365804184e-06,
      "loss": 0.3882,
      "step": 726
    },
    {
      "epoch": 3.427899686520376,
      "grad_norm": 0.5303656458854675,
      "learning_rate": 2.225797440261936e-06,
      "loss": 0.4336,
      "step": 727
    },
    {
      "epoch": 3.4326018808777428,
      "grad_norm": 0.6270896792411804,
      "learning_rate": 2.219137514168187e-06,
      "loss": 0.397,
      "step": 728
    },
    {
      "epoch": 3.43730407523511,
      "grad_norm": 0.5054409503936768,
      "learning_rate": 2.212479606152667e-06,
      "loss": 0.4261,
      "step": 729
    },
    {
      "epoch": 3.4420062695924765,
      "grad_norm": 0.5422618985176086,
      "learning_rate": 2.205823764054372e-06,
      "loss": 0.4105,
      "step": 730
    },
    {
      "epoch": 3.446708463949843,
      "grad_norm": 0.5200968980789185,
      "learning_rate": 2.199170035697453e-06,
      "loss": 0.4048,
      "step": 731
    },
    {
      "epoch": 3.45141065830721,
      "grad_norm": 0.5316998362541199,
      "learning_rate": 2.1925184688908735e-06,
      "loss": 0.4132,
      "step": 732
    },
    {
      "epoch": 3.456112852664577,
      "grad_norm": 0.5780388116836548,
      "learning_rate": 2.185869111428067e-06,
      "loss": 0.4381,
      "step": 733
    },
    {
      "epoch": 3.4608150470219434,
      "grad_norm": 0.5547174215316772,
      "learning_rate": 2.1792220110865885e-06,
      "loss": 0.4236,
      "step": 734
    },
    {
      "epoch": 3.4655172413793105,
      "grad_norm": 0.5188453197479248,
      "learning_rate": 2.1725772156277795e-06,
      "loss": 0.4052,
      "step": 735
    },
    {
      "epoch": 3.470219435736677,
      "grad_norm": 0.5145602822303772,
      "learning_rate": 2.165934772796417e-06,
      "loss": 0.412,
      "step": 736
    },
    {
      "epoch": 3.4749216300940438,
      "grad_norm": 0.5960094332695007,
      "learning_rate": 2.159294730320374e-06,
      "loss": 0.426,
      "step": 737
    },
    {
      "epoch": 3.479623824451411,
      "grad_norm": 0.7090360522270203,
      "learning_rate": 2.15265713591028e-06,
      "loss": 0.4133,
      "step": 738
    },
    {
      "epoch": 3.4843260188087775,
      "grad_norm": 0.5428952574729919,
      "learning_rate": 2.1460220372591676e-06,
      "loss": 0.4332,
      "step": 739
    },
    {
      "epoch": 3.489028213166144,
      "grad_norm": 0.6610196232795715,
      "learning_rate": 2.139389482042142e-06,
      "loss": 0.3985,
      "step": 740
    },
    {
      "epoch": 3.493730407523511,
      "grad_norm": 0.5409770607948303,
      "learning_rate": 2.1327595179160332e-06,
      "loss": 0.4148,
      "step": 741
    },
    {
      "epoch": 3.498432601880878,
      "grad_norm": 0.8822159171104431,
      "learning_rate": 2.1261321925190492e-06,
      "loss": 0.4071,
      "step": 742
    },
    {
      "epoch": 3.5031347962382444,
      "grad_norm": 0.5366957783699036,
      "learning_rate": 2.1195075534704433e-06,
      "loss": 0.3838,
      "step": 743
    },
    {
      "epoch": 3.507836990595611,
      "grad_norm": 0.5289701819419861,
      "learning_rate": 2.1128856483701625e-06,
      "loss": 0.4123,
      "step": 744
    },
    {
      "epoch": 3.512539184952978,
      "grad_norm": 0.5737835764884949,
      "learning_rate": 2.10626652479851e-06,
      "loss": 0.392,
      "step": 745
    },
    {
      "epoch": 3.5172413793103448,
      "grad_norm": 0.5381962060928345,
      "learning_rate": 2.0996502303158057e-06,
      "loss": 0.4088,
      "step": 746
    },
    {
      "epoch": 3.521943573667712,
      "grad_norm": 0.529466450214386,
      "learning_rate": 2.0930368124620385e-06,
      "loss": 0.4098,
      "step": 747
    },
    {
      "epoch": 3.5266457680250785,
      "grad_norm": 0.6686971783638,
      "learning_rate": 2.086426318756531e-06,
      "loss": 0.4273,
      "step": 748
    },
    {
      "epoch": 3.531347962382445,
      "grad_norm": 0.5246966481208801,
      "learning_rate": 2.0798187966975917e-06,
      "loss": 0.4318,
      "step": 749
    },
    {
      "epoch": 3.5360501567398117,
      "grad_norm": 0.5165736675262451,
      "learning_rate": 2.073214293762179e-06,
      "loss": 0.4212,
      "step": 750
    },
    {
      "epoch": 3.540752351097179,
      "grad_norm": 0.6821503043174744,
      "learning_rate": 2.0666128574055575e-06,
      "loss": 0.4199,
      "step": 751
    },
    {
      "epoch": 3.5454545454545454,
      "grad_norm": 0.5294732451438904,
      "learning_rate": 2.0600145350609585e-06,
      "loss": 0.4192,
      "step": 752
    },
    {
      "epoch": 3.5501567398119125,
      "grad_norm": 0.515800416469574,
      "learning_rate": 2.053419374139235e-06,
      "loss": 0.4172,
      "step": 753
    },
    {
      "epoch": 3.554858934169279,
      "grad_norm": 0.5241639614105225,
      "learning_rate": 2.0468274220285295e-06,
      "loss": 0.4138,
      "step": 754
    },
    {
      "epoch": 3.5595611285266457,
      "grad_norm": 0.546105146408081,
      "learning_rate": 2.0402387260939224e-06,
      "loss": 0.4123,
      "step": 755
    },
    {
      "epoch": 3.5642633228840124,
      "grad_norm": 0.5261510014533997,
      "learning_rate": 2.033653333677103e-06,
      "loss": 0.4225,
      "step": 756
    },
    {
      "epoch": 3.5689655172413794,
      "grad_norm": 0.5825217366218567,
      "learning_rate": 2.02707129209602e-06,
      "loss": 0.4042,
      "step": 757
    },
    {
      "epoch": 3.573667711598746,
      "grad_norm": 0.5916388034820557,
      "learning_rate": 2.0204926486445463e-06,
      "loss": 0.4222,
      "step": 758
    },
    {
      "epoch": 3.5783699059561127,
      "grad_norm": 0.5643376708030701,
      "learning_rate": 2.0139174505921403e-06,
      "loss": 0.4419,
      "step": 759
    },
    {
      "epoch": 3.58307210031348,
      "grad_norm": 0.5426534414291382,
      "learning_rate": 2.0073457451835e-06,
      "loss": 0.3985,
      "step": 760
    },
    {
      "epoch": 3.5877742946708464,
      "grad_norm": 0.48811203241348267,
      "learning_rate": 2.0007775796382335e-06,
      "loss": 0.4249,
      "step": 761
    },
    {
      "epoch": 3.592476489028213,
      "grad_norm": 0.5216817855834961,
      "learning_rate": 1.994213001150508e-06,
      "loss": 0.3931,
      "step": 762
    },
    {
      "epoch": 3.5971786833855797,
      "grad_norm": 0.5739433169364929,
      "learning_rate": 1.9876520568887207e-06,
      "loss": 0.42,
      "step": 763
    },
    {
      "epoch": 3.6018808777429467,
      "grad_norm": 0.5166419148445129,
      "learning_rate": 1.981094793995155e-06,
      "loss": 0.4041,
      "step": 764
    },
    {
      "epoch": 3.6065830721003134,
      "grad_norm": 0.6763928532600403,
      "learning_rate": 1.974541259585641e-06,
      "loss": 0.4319,
      "step": 765
    },
    {
      "epoch": 3.6112852664576804,
      "grad_norm": 0.5443664789199829,
      "learning_rate": 1.9679915007492194e-06,
      "loss": 0.4139,
      "step": 766
    },
    {
      "epoch": 3.615987460815047,
      "grad_norm": 0.6719280481338501,
      "learning_rate": 1.9614455645478047e-06,
      "loss": 0.4015,
      "step": 767
    },
    {
      "epoch": 3.6206896551724137,
      "grad_norm": 0.5685383677482605,
      "learning_rate": 1.9549034980158403e-06,
      "loss": 0.4153,
      "step": 768
    },
    {
      "epoch": 3.6253918495297803,
      "grad_norm": 0.5463993549346924,
      "learning_rate": 1.9483653481599697e-06,
      "loss": 0.4193,
      "step": 769
    },
    {
      "epoch": 3.6300940438871474,
      "grad_norm": 0.5228095054626465,
      "learning_rate": 1.9418311619586897e-06,
      "loss": 0.4268,
      "step": 770
    },
    {
      "epoch": 3.634796238244514,
      "grad_norm": 0.6472461223602295,
      "learning_rate": 1.935300986362018e-06,
      "loss": 0.3981,
      "step": 771
    },
    {
      "epoch": 3.639498432601881,
      "grad_norm": 0.61808842420578,
      "learning_rate": 1.9287748682911582e-06,
      "loss": 0.4313,
      "step": 772
    },
    {
      "epoch": 3.6442006269592477,
      "grad_norm": 0.5122710466384888,
      "learning_rate": 1.9222528546381543e-06,
      "loss": 0.4219,
      "step": 773
    },
    {
      "epoch": 3.6489028213166144,
      "grad_norm": 0.5540320873260498,
      "learning_rate": 1.9157349922655648e-06,
      "loss": 0.4001,
      "step": 774
    },
    {
      "epoch": 3.653605015673981,
      "grad_norm": 0.5066401958465576,
      "learning_rate": 1.909221328006114e-06,
      "loss": 0.4089,
      "step": 775
    },
    {
      "epoch": 3.658307210031348,
      "grad_norm": 0.5802583694458008,
      "learning_rate": 1.9027119086623647e-06,
      "loss": 0.4216,
      "step": 776
    },
    {
      "epoch": 3.6630094043887147,
      "grad_norm": 0.5735054016113281,
      "learning_rate": 1.8962067810063806e-06,
      "loss": 0.4372,
      "step": 777
    },
    {
      "epoch": 3.6677115987460818,
      "grad_norm": 0.5177802443504333,
      "learning_rate": 1.8897059917793844e-06,
      "loss": 0.3912,
      "step": 778
    },
    {
      "epoch": 3.6724137931034484,
      "grad_norm": 0.5790892243385315,
      "learning_rate": 1.8832095876914268e-06,
      "loss": 0.4096,
      "step": 779
    },
    {
      "epoch": 3.677115987460815,
      "grad_norm": 0.5386017560958862,
      "learning_rate": 1.8767176154210537e-06,
      "loss": 0.4191,
      "step": 780
    },
    {
      "epoch": 3.6818181818181817,
      "grad_norm": 0.5927474498748779,
      "learning_rate": 1.8702301216149616e-06,
      "loss": 0.4061,
      "step": 781
    },
    {
      "epoch": 3.6865203761755487,
      "grad_norm": 0.5609317421913147,
      "learning_rate": 1.8637471528876727e-06,
      "loss": 0.4067,
      "step": 782
    },
    {
      "epoch": 3.6912225705329154,
      "grad_norm": 0.6609043478965759,
      "learning_rate": 1.8572687558211923e-06,
      "loss": 0.4183,
      "step": 783
    },
    {
      "epoch": 3.695924764890282,
      "grad_norm": 0.5092527270317078,
      "learning_rate": 1.850794976964677e-06,
      "loss": 0.3827,
      "step": 784
    },
    {
      "epoch": 3.700626959247649,
      "grad_norm": 0.8918034434318542,
      "learning_rate": 1.8443258628341026e-06,
      "loss": 0.4144,
      "step": 785
    },
    {
      "epoch": 3.7053291536050157,
      "grad_norm": 0.5443233847618103,
      "learning_rate": 1.837861459911925e-06,
      "loss": 0.4246,
      "step": 786
    },
    {
      "epoch": 3.7100313479623823,
      "grad_norm": 0.6559080481529236,
      "learning_rate": 1.8314018146467505e-06,
      "loss": 0.4067,
      "step": 787
    },
    {
      "epoch": 3.714733542319749,
      "grad_norm": 0.5071741342544556,
      "learning_rate": 1.8249469734529995e-06,
      "loss": 0.3888,
      "step": 788
    },
    {
      "epoch": 3.719435736677116,
      "grad_norm": 0.5663676261901855,
      "learning_rate": 1.818496982710572e-06,
      "loss": 0.4256,
      "step": 789
    },
    {
      "epoch": 3.7241379310344827,
      "grad_norm": 0.5477777719497681,
      "learning_rate": 1.81205188876452e-06,
      "loss": 0.423,
      "step": 790
    },
    {
      "epoch": 3.7288401253918497,
      "grad_norm": 0.5709276795387268,
      "learning_rate": 1.8056117379247078e-06,
      "loss": 0.4265,
      "step": 791
    },
    {
      "epoch": 3.7335423197492164,
      "grad_norm": 0.49602681398391724,
      "learning_rate": 1.7991765764654813e-06,
      "loss": 0.4141,
      "step": 792
    },
    {
      "epoch": 3.738244514106583,
      "grad_norm": 0.5358700156211853,
      "learning_rate": 1.7927464506253394e-06,
      "loss": 0.4231,
      "step": 793
    },
    {
      "epoch": 3.7429467084639496,
      "grad_norm": 1.1592613458633423,
      "learning_rate": 1.7863214066065951e-06,
      "loss": 0.3929,
      "step": 794
    },
    {
      "epoch": 3.7476489028213167,
      "grad_norm": 0.5176786780357361,
      "learning_rate": 1.779901490575051e-06,
      "loss": 0.4201,
      "step": 795
    },
    {
      "epoch": 3.7523510971786833,
      "grad_norm": 0.5303675532341003,
      "learning_rate": 1.7734867486596596e-06,
      "loss": 0.4201,
      "step": 796
    },
    {
      "epoch": 3.7570532915360504,
      "grad_norm": 0.5633402466773987,
      "learning_rate": 1.767077226952198e-06,
      "loss": 0.4276,
      "step": 797
    },
    {
      "epoch": 3.761755485893417,
      "grad_norm": 0.6016635894775391,
      "learning_rate": 1.7606729715069349e-06,
      "loss": 0.4143,
      "step": 798
    },
    {
      "epoch": 3.7664576802507836,
      "grad_norm": 0.5202106237411499,
      "learning_rate": 1.7542740283402981e-06,
      "loss": 0.4195,
      "step": 799
    },
    {
      "epoch": 3.7711598746081503,
      "grad_norm": 0.6279420852661133,
      "learning_rate": 1.7478804434305466e-06,
      "loss": 0.4001,
      "step": 800
    },
    {
      "epoch": 3.7758620689655173,
      "grad_norm": 0.5253601670265198,
      "learning_rate": 1.741492262717438e-06,
      "loss": 0.4206,
      "step": 801
    },
    {
      "epoch": 3.780564263322884,
      "grad_norm": 0.5218167901039124,
      "learning_rate": 1.7351095321018974e-06,
      "loss": 0.387,
      "step": 802
    },
    {
      "epoch": 3.785266457680251,
      "grad_norm": 0.530846357345581,
      "learning_rate": 1.7287322974456933e-06,
      "loss": 0.3935,
      "step": 803
    },
    {
      "epoch": 3.7899686520376177,
      "grad_norm": 0.5487862825393677,
      "learning_rate": 1.7223606045711006e-06,
      "loss": 0.4168,
      "step": 804
    },
    {
      "epoch": 3.7946708463949843,
      "grad_norm": 0.5345083475112915,
      "learning_rate": 1.7159944992605774e-06,
      "loss": 0.4208,
      "step": 805
    },
    {
      "epoch": 3.799373040752351,
      "grad_norm": 0.5425072312355042,
      "learning_rate": 1.7096340272564318e-06,
      "loss": 0.4088,
      "step": 806
    },
    {
      "epoch": 3.804075235109718,
      "grad_norm": 0.5253011584281921,
      "learning_rate": 1.7032792342604947e-06,
      "loss": 0.3995,
      "step": 807
    },
    {
      "epoch": 3.8087774294670846,
      "grad_norm": 0.7746017575263977,
      "learning_rate": 1.6969301659337944e-06,
      "loss": 0.4145,
      "step": 808
    },
    {
      "epoch": 3.8134796238244513,
      "grad_norm": 0.7049569487571716,
      "learning_rate": 1.6905868678962225e-06,
      "loss": 0.4216,
      "step": 809
    },
    {
      "epoch": 3.8181818181818183,
      "grad_norm": 0.602180540561676,
      "learning_rate": 1.684249385726211e-06,
      "loss": 0.4134,
      "step": 810
    },
    {
      "epoch": 3.822884012539185,
      "grad_norm": 0.5291408896446228,
      "learning_rate": 1.677917764960404e-06,
      "loss": 0.402,
      "step": 811
    },
    {
      "epoch": 3.8275862068965516,
      "grad_norm": 0.5529280304908752,
      "learning_rate": 1.6715920510933277e-06,
      "loss": 0.4322,
      "step": 812
    },
    {
      "epoch": 3.8322884012539182,
      "grad_norm": 0.5989758968353271,
      "learning_rate": 1.6652722895770676e-06,
      "loss": 0.4275,
      "step": 813
    },
    {
      "epoch": 3.8369905956112853,
      "grad_norm": 0.5088624358177185,
      "learning_rate": 1.6589585258209383e-06,
      "loss": 0.378,
      "step": 814
    },
    {
      "epoch": 3.841692789968652,
      "grad_norm": 0.5167607665061951,
      "learning_rate": 1.6526508051911588e-06,
      "loss": 0.4221,
      "step": 815
    },
    {
      "epoch": 3.846394984326019,
      "grad_norm": 0.5582865476608276,
      "learning_rate": 1.6463491730105282e-06,
      "loss": 0.4091,
      "step": 816
    },
    {
      "epoch": 3.8510971786833856,
      "grad_norm": 0.5103083252906799,
      "learning_rate": 1.6400536745580955e-06,
      "loss": 0.3867,
      "step": 817
    },
    {
      "epoch": 3.8557993730407523,
      "grad_norm": 0.528692901134491,
      "learning_rate": 1.6337643550688408e-06,
      "loss": 0.4178,
      "step": 818
    },
    {
      "epoch": 3.860501567398119,
      "grad_norm": 0.5174258947372437,
      "learning_rate": 1.627481259733343e-06,
      "loss": 0.3989,
      "step": 819
    },
    {
      "epoch": 3.865203761755486,
      "grad_norm": 0.492735892534256,
      "learning_rate": 1.6212044336974598e-06,
      "loss": 0.3935,
      "step": 820
    },
    {
      "epoch": 3.8699059561128526,
      "grad_norm": 0.5810956954956055,
      "learning_rate": 1.614933922062003e-06,
      "loss": 0.4082,
      "step": 821
    },
    {
      "epoch": 3.8746081504702197,
      "grad_norm": 0.5235511660575867,
      "learning_rate": 1.6086697698824144e-06,
      "loss": 0.4026,
      "step": 822
    },
    {
      "epoch": 3.8793103448275863,
      "grad_norm": 0.5972744822502136,
      "learning_rate": 1.6024120221684373e-06,
      "loss": 0.4018,
      "step": 823
    },
    {
      "epoch": 3.884012539184953,
      "grad_norm": 0.5685083270072937,
      "learning_rate": 1.5961607238838022e-06,
      "loss": 0.4077,
      "step": 824
    },
    {
      "epoch": 3.8887147335423196,
      "grad_norm": 0.5427765250205994,
      "learning_rate": 1.589915919945894e-06,
      "loss": 0.4187,
      "step": 825
    },
    {
      "epoch": 3.8934169278996866,
      "grad_norm": 0.6297295093536377,
      "learning_rate": 1.5836776552254386e-06,
      "loss": 0.4367,
      "step": 826
    },
    {
      "epoch": 3.8981191222570533,
      "grad_norm": 0.6110124588012695,
      "learning_rate": 1.5774459745461711e-06,
      "loss": 0.4065,
      "step": 827
    },
    {
      "epoch": 3.9028213166144203,
      "grad_norm": 0.4981592297554016,
      "learning_rate": 1.5712209226845201e-06,
      "loss": 0.3836,
      "step": 828
    },
    {
      "epoch": 3.907523510971787,
      "grad_norm": 0.5722451210021973,
      "learning_rate": 1.565002544369286e-06,
      "loss": 0.4161,
      "step": 829
    },
    {
      "epoch": 3.9122257053291536,
      "grad_norm": 0.6718733310699463,
      "learning_rate": 1.5587908842813142e-06,
      "loss": 0.4053,
      "step": 830
    },
    {
      "epoch": 3.91692789968652,
      "grad_norm": 0.5070095658302307,
      "learning_rate": 1.5525859870531823e-06,
      "loss": 0.4198,
      "step": 831
    },
    {
      "epoch": 3.9216300940438873,
      "grad_norm": 0.5303407311439514,
      "learning_rate": 1.5463878972688707e-06,
      "loss": 0.4089,
      "step": 832
    },
    {
      "epoch": 3.926332288401254,
      "grad_norm": 0.5431908369064331,
      "learning_rate": 1.5401966594634483e-06,
      "loss": 0.4341,
      "step": 833
    },
    {
      "epoch": 3.9310344827586206,
      "grad_norm": 0.549174427986145,
      "learning_rate": 1.5340123181227495e-06,
      "loss": 0.4237,
      "step": 834
    },
    {
      "epoch": 3.9357366771159876,
      "grad_norm": 0.8902267217636108,
      "learning_rate": 1.527834917683058e-06,
      "loss": 0.3904,
      "step": 835
    },
    {
      "epoch": 3.9404388714733543,
      "grad_norm": 0.5055849552154541,
      "learning_rate": 1.5216645025307813e-06,
      "loss": 0.4058,
      "step": 836
    },
    {
      "epoch": 3.945141065830721,
      "grad_norm": 0.5319788455963135,
      "learning_rate": 1.5155011170021399e-06,
      "loss": 0.4153,
      "step": 837
    },
    {
      "epoch": 3.9498432601880875,
      "grad_norm": 0.5441375374794006,
      "learning_rate": 1.5093448053828402e-06,
      "loss": 0.4231,
      "step": 838
    },
    {
      "epoch": 3.9545454545454546,
      "grad_norm": 0.5940942764282227,
      "learning_rate": 1.503195611907764e-06,
      "loss": 0.4241,
      "step": 839
    },
    {
      "epoch": 3.959247648902821,
      "grad_norm": 0.5203325748443604,
      "learning_rate": 1.4970535807606453e-06,
      "loss": 0.3842,
      "step": 840
    },
    {
      "epoch": 3.9639498432601883,
      "grad_norm": 0.525404691696167,
      "learning_rate": 1.4909187560737542e-06,
      "loss": 0.3954,
      "step": 841
    },
    {
      "epoch": 3.968652037617555,
      "grad_norm": 0.5999636054039001,
      "learning_rate": 1.4847911819275829e-06,
      "loss": 0.4061,
      "step": 842
    },
    {
      "epoch": 3.9733542319749215,
      "grad_norm": 0.5253078937530518,
      "learning_rate": 1.4786709023505224e-06,
      "loss": 0.3969,
      "step": 843
    },
    {
      "epoch": 3.978056426332288,
      "grad_norm": 0.535467803478241,
      "learning_rate": 1.4725579613185549e-06,
      "loss": 0.4241,
      "step": 844
    },
    {
      "epoch": 3.9827586206896552,
      "grad_norm": 0.5458933711051941,
      "learning_rate": 1.4664524027549291e-06,
      "loss": 0.4102,
      "step": 845
    },
    {
      "epoch": 3.987460815047022,
      "grad_norm": 0.515102207660675,
      "learning_rate": 1.4603542705298493e-06,
      "loss": 0.3957,
      "step": 846
    },
    {
      "epoch": 3.992163009404389,
      "grad_norm": 0.572600245475769,
      "learning_rate": 1.4542636084601624e-06,
      "loss": 0.3686,
      "step": 847
    },
    {
      "epoch": 3.9968652037617556,
      "grad_norm": 0.520165205001831,
      "learning_rate": 1.4481804603090358e-06,
      "loss": 0.4109,
      "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
}