{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.993252361673415,
  "eval_steps": 500,
  "global_step": 925,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.005398110661268556,
      "grad_norm": 5.8547901781698295,
      "learning_rate": 4.301075268817205e-07,
      "loss": 0.9789,
      "step": 1
    },
    {
      "epoch": 0.010796221322537112,
      "grad_norm": 6.195900478096068,
      "learning_rate": 8.60215053763441e-07,
      "loss": 1.0263,
      "step": 2
    },
    {
      "epoch": 0.016194331983805668,
      "grad_norm": 5.844104649074486,
      "learning_rate": 1.2903225806451614e-06,
      "loss": 1.3526,
      "step": 3
    },
    {
      "epoch": 0.021592442645074223,
      "grad_norm": 7.368611794635692,
      "learning_rate": 1.720430107526882e-06,
      "loss": 1.049,
      "step": 4
    },
    {
      "epoch": 0.02699055330634278,
      "grad_norm": 5.672738441304507,
      "learning_rate": 2.1505376344086023e-06,
      "loss": 0.9437,
      "step": 5
    },
    {
      "epoch": 0.032388663967611336,
      "grad_norm": 4.775325426134859,
      "learning_rate": 2.580645161290323e-06,
      "loss": 0.9776,
      "step": 6
    },
    {
      "epoch": 0.037786774628879895,
      "grad_norm": 4.685418851624996,
      "learning_rate": 3.0107526881720433e-06,
      "loss": 1.1554,
      "step": 7
    },
    {
      "epoch": 0.043184885290148446,
      "grad_norm": 3.259634449429892,
      "learning_rate": 3.440860215053764e-06,
      "loss": 0.8474,
      "step": 8
    },
    {
      "epoch": 0.048582995951417005,
      "grad_norm": 3.2715012435354347,
      "learning_rate": 3.870967741935484e-06,
      "loss": 0.845,
      "step": 9
    },
    {
      "epoch": 0.05398110661268556,
      "grad_norm": 2.518263916966614,
      "learning_rate": 4.3010752688172045e-06,
      "loss": 0.8109,
      "step": 10
    },
    {
      "epoch": 0.059379217273954114,
      "grad_norm": 1.931792216640399,
      "learning_rate": 4.731182795698925e-06,
      "loss": 0.77,
      "step": 11
    },
    {
      "epoch": 0.06477732793522267,
      "grad_norm": 1.917478295407825,
      "learning_rate": 5.161290322580646e-06,
      "loss": 0.7055,
      "step": 12
    },
    {
      "epoch": 0.07017543859649122,
      "grad_norm": 1.9342362297971811,
      "learning_rate": 5.591397849462365e-06,
      "loss": 0.6897,
      "step": 13
    },
    {
      "epoch": 0.07557354925775979,
      "grad_norm": 2.335686007359441,
      "learning_rate": 6.021505376344087e-06,
      "loss": 0.8362,
      "step": 14
    },
    {
      "epoch": 0.08097165991902834,
      "grad_norm": 2.899927706834149,
      "learning_rate": 6.451612903225806e-06,
      "loss": 0.7419,
      "step": 15
    },
    {
      "epoch": 0.08636977058029689,
      "grad_norm": 2.320716820106864,
      "learning_rate": 6.881720430107528e-06,
      "loss": 0.7808,
      "step": 16
    },
    {
      "epoch": 0.09176788124156546,
      "grad_norm": 1.9233741520153684,
      "learning_rate": 7.311827956989248e-06,
      "loss": 0.7005,
      "step": 17
    },
    {
      "epoch": 0.09716599190283401,
      "grad_norm": 1.4890943029905404,
      "learning_rate": 7.741935483870968e-06,
      "loss": 0.737,
      "step": 18
    },
    {
      "epoch": 0.10256410256410256,
      "grad_norm": 1.6346157704701012,
      "learning_rate": 8.172043010752689e-06,
      "loss": 0.753,
      "step": 19
    },
    {
      "epoch": 0.10796221322537113,
      "grad_norm": 1.3923927391716204,
      "learning_rate": 8.602150537634409e-06,
      "loss": 0.662,
      "step": 20
    },
    {
      "epoch": 0.11336032388663968,
      "grad_norm": 1.4868588771935882,
      "learning_rate": 9.03225806451613e-06,
      "loss": 0.6965,
      "step": 21
    },
    {
      "epoch": 0.11875843454790823,
      "grad_norm": 1.5586239228768057,
      "learning_rate": 9.46236559139785e-06,
      "loss": 0.6918,
      "step": 22
    },
    {
      "epoch": 0.1241565452091768,
      "grad_norm": 1.6467617716036158,
      "learning_rate": 9.89247311827957e-06,
      "loss": 0.7206,
      "step": 23
    },
    {
      "epoch": 0.12955465587044535,
      "grad_norm": 1.5431165209998874,
      "learning_rate": 1.0322580645161291e-05,
      "loss": 0.7401,
      "step": 24
    },
    {
      "epoch": 0.1349527665317139,
      "grad_norm": 1.262214429142841,
      "learning_rate": 1.0752688172043012e-05,
      "loss": 0.7939,
      "step": 25
    },
    {
      "epoch": 0.14035087719298245,
      "grad_norm": 1.342395990639865,
      "learning_rate": 1.118279569892473e-05,
      "loss": 0.7074,
      "step": 26
    },
    {
      "epoch": 0.145748987854251,
      "grad_norm": 1.392051362880004,
      "learning_rate": 1.1612903225806453e-05,
      "loss": 0.6546,
      "step": 27
    },
    {
      "epoch": 0.15114709851551958,
      "grad_norm": 1.3560822319733044,
      "learning_rate": 1.2043010752688173e-05,
      "loss": 0.6785,
      "step": 28
    },
    {
      "epoch": 0.15654520917678813,
      "grad_norm": 1.2795645343244073,
      "learning_rate": 1.2473118279569894e-05,
      "loss": 0.6751,
      "step": 29
    },
    {
      "epoch": 0.16194331983805668,
      "grad_norm": 1.1855435088179302,
      "learning_rate": 1.2903225806451613e-05,
      "loss": 0.7002,
      "step": 30
    },
    {
      "epoch": 0.16734143049932523,
      "grad_norm": 1.1407557409125308,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.6543,
      "step": 31
    },
    {
      "epoch": 0.17273954116059378,
      "grad_norm": 1.2413150424813355,
      "learning_rate": 1.3763440860215056e-05,
      "loss": 0.9318,
      "step": 32
    },
    {
      "epoch": 0.17813765182186234,
      "grad_norm": 1.2499110019395852,
      "learning_rate": 1.4193548387096776e-05,
      "loss": 0.6477,
      "step": 33
    },
    {
      "epoch": 0.18353576248313092,
      "grad_norm": 1.4008520106194233,
      "learning_rate": 1.4623655913978497e-05,
      "loss": 0.711,
      "step": 34
    },
    {
      "epoch": 0.18893387314439947,
      "grad_norm": 1.1891772900416815,
      "learning_rate": 1.5053763440860215e-05,
      "loss": 0.6406,
      "step": 35
    },
    {
      "epoch": 0.19433198380566802,
      "grad_norm": 1.147501365646657,
      "learning_rate": 1.5483870967741936e-05,
      "loss": 0.6624,
      "step": 36
    },
    {
      "epoch": 0.19973009446693657,
      "grad_norm": 1.1751858805701079,
      "learning_rate": 1.5913978494623657e-05,
      "loss": 0.7633,
      "step": 37
    },
    {
      "epoch": 0.20512820512820512,
      "grad_norm": 1.1864282998965006,
      "learning_rate": 1.6344086021505377e-05,
      "loss": 0.7109,
      "step": 38
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 1.0272367928001183,
      "learning_rate": 1.6774193548387098e-05,
      "loss": 0.5869,
      "step": 39
    },
    {
      "epoch": 0.21592442645074225,
      "grad_norm": 1.2112435393460614,
      "learning_rate": 1.7204301075268818e-05,
      "loss": 0.6915,
      "step": 40
    },
    {
      "epoch": 0.2213225371120108,
      "grad_norm": 1.1750727844511175,
      "learning_rate": 1.763440860215054e-05,
      "loss": 0.6758,
      "step": 41
    },
    {
      "epoch": 0.22672064777327935,
      "grad_norm": 1.175639613071845,
      "learning_rate": 1.806451612903226e-05,
      "loss": 0.6514,
      "step": 42
    },
    {
      "epoch": 0.2321187584345479,
      "grad_norm": 1.1903426022335029,
      "learning_rate": 1.849462365591398e-05,
      "loss": 0.6711,
      "step": 43
    },
    {
      "epoch": 0.23751686909581646,
      "grad_norm": 1.13775486347414,
      "learning_rate": 1.89247311827957e-05,
      "loss": 0.6484,
      "step": 44
    },
    {
      "epoch": 0.242914979757085,
      "grad_norm": 1.1018523962648061,
      "learning_rate": 1.935483870967742e-05,
      "loss": 0.6413,
      "step": 45
    },
    {
      "epoch": 0.2483130904183536,
      "grad_norm": 1.1536291053738799,
      "learning_rate": 1.978494623655914e-05,
      "loss": 0.6382,
      "step": 46
    },
    {
      "epoch": 0.25371120107962214,
      "grad_norm": 1.2874232917542905,
      "learning_rate": 2.0215053763440862e-05,
      "loss": 0.716,
      "step": 47
    },
    {
      "epoch": 0.2591093117408907,
      "grad_norm": 1.0667615653454148,
      "learning_rate": 2.0645161290322582e-05,
      "loss": 0.6627,
      "step": 48
    },
    {
      "epoch": 0.26450742240215924,
      "grad_norm": 1.1268887691690699,
      "learning_rate": 2.1075268817204303e-05,
      "loss": 0.6586,
      "step": 49
    },
    {
      "epoch": 0.2699055330634278,
      "grad_norm": 1.1645388991541576,
      "learning_rate": 2.1505376344086024e-05,
      "loss": 0.6267,
      "step": 50
    },
    {
      "epoch": 0.27530364372469635,
      "grad_norm": 1.0147829233748185,
      "learning_rate": 2.193548387096774e-05,
      "loss": 0.6659,
      "step": 51
    },
    {
      "epoch": 0.2807017543859649,
      "grad_norm": 1.0582569825253139,
      "learning_rate": 2.236559139784946e-05,
      "loss": 0.64,
      "step": 52
    },
    {
      "epoch": 0.28609986504723345,
      "grad_norm": 1.1030609601483023,
      "learning_rate": 2.2795698924731185e-05,
      "loss": 0.6531,
      "step": 53
    },
    {
      "epoch": 0.291497975708502,
      "grad_norm": 1.0689716886960958,
      "learning_rate": 2.3225806451612906e-05,
      "loss": 0.6584,
      "step": 54
    },
    {
      "epoch": 0.2968960863697706,
      "grad_norm": 1.0511582991478239,
      "learning_rate": 2.3655913978494626e-05,
      "loss": 0.6263,
      "step": 55
    },
    {
      "epoch": 0.30229419703103916,
      "grad_norm": 1.096934265784751,
      "learning_rate": 2.4086021505376347e-05,
      "loss": 0.8141,
      "step": 56
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 1.0264643074156008,
      "learning_rate": 2.4516129032258067e-05,
      "loss": 0.6414,
      "step": 57
    },
    {
      "epoch": 0.31309041835357626,
      "grad_norm": 1.1419883911045894,
      "learning_rate": 2.4946236559139788e-05,
      "loss": 0.6248,
      "step": 58
    },
    {
      "epoch": 0.3184885290148448,
      "grad_norm": 1.119342396183509,
      "learning_rate": 2.537634408602151e-05,
      "loss": 0.6696,
      "step": 59
    },
    {
      "epoch": 0.32388663967611336,
      "grad_norm": 1.1524685035070938,
      "learning_rate": 2.5806451612903226e-05,
      "loss": 0.6926,
      "step": 60
    },
    {
      "epoch": 0.3292847503373819,
      "grad_norm": 1.1023493942950338,
      "learning_rate": 2.6236559139784946e-05,
      "loss": 0.7087,
      "step": 61
    },
    {
      "epoch": 0.33468286099865047,
      "grad_norm": 1.084679917232489,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 0.6711,
      "step": 62
    },
    {
      "epoch": 0.340080971659919,
      "grad_norm": 0.9988990454591712,
      "learning_rate": 2.7096774193548387e-05,
      "loss": 0.6966,
      "step": 63
    },
    {
      "epoch": 0.34547908232118757,
      "grad_norm": 1.0505889482507405,
      "learning_rate": 2.752688172043011e-05,
      "loss": 0.6916,
      "step": 64
    },
    {
      "epoch": 0.3508771929824561,
      "grad_norm": 1.061750405323965,
      "learning_rate": 2.795698924731183e-05,
      "loss": 0.6824,
      "step": 65
    },
    {
      "epoch": 0.3562753036437247,
      "grad_norm": 0.9984273646688296,
      "learning_rate": 2.8387096774193552e-05,
      "loss": 0.666,
      "step": 66
    },
    {
      "epoch": 0.3616734143049933,
      "grad_norm": 1.0894465182196065,
      "learning_rate": 2.8817204301075273e-05,
      "loss": 0.698,
      "step": 67
    },
    {
      "epoch": 0.36707152496626183,
      "grad_norm": 1.061814212390062,
      "learning_rate": 2.9247311827956993e-05,
      "loss": 0.6498,
      "step": 68
    },
    {
      "epoch": 0.3724696356275304,
      "grad_norm": 1.1225275383063609,
      "learning_rate": 2.9677419354838714e-05,
      "loss": 0.865,
      "step": 69
    },
    {
      "epoch": 0.37786774628879893,
      "grad_norm": 1.0101520932943957,
      "learning_rate": 3.010752688172043e-05,
      "loss": 0.635,
      "step": 70
    },
    {
      "epoch": 0.3832658569500675,
      "grad_norm": 1.1042840779826546,
      "learning_rate": 3.053763440860215e-05,
      "loss": 0.6602,
      "step": 71
    },
    {
      "epoch": 0.38866396761133604,
      "grad_norm": 0.9951373477683033,
      "learning_rate": 3.096774193548387e-05,
      "loss": 0.6181,
      "step": 72
    },
    {
      "epoch": 0.3940620782726046,
      "grad_norm": 1.0419770417033245,
      "learning_rate": 3.139784946236559e-05,
      "loss": 0.6591,
      "step": 73
    },
    {
      "epoch": 0.39946018893387314,
      "grad_norm": 1.1235300688912213,
      "learning_rate": 3.182795698924731e-05,
      "loss": 0.7067,
      "step": 74
    },
    {
      "epoch": 0.4048582995951417,
      "grad_norm": 1.032338142893847,
      "learning_rate": 3.2258064516129034e-05,
      "loss": 0.6232,
      "step": 75
    },
    {
      "epoch": 0.41025641025641024,
      "grad_norm": 1.0941631682251065,
      "learning_rate": 3.2688172043010754e-05,
      "loss": 0.7227,
      "step": 76
    },
    {
      "epoch": 0.4156545209176788,
      "grad_norm": 1.056284707482023,
      "learning_rate": 3.3118279569892475e-05,
      "loss": 0.7325,
      "step": 77
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 1.1196455822802613,
      "learning_rate": 3.3548387096774195e-05,
      "loss": 0.7219,
      "step": 78
    },
    {
      "epoch": 0.42645074224021595,
      "grad_norm": 0.9854925794840517,
      "learning_rate": 3.3978494623655916e-05,
      "loss": 0.6219,
      "step": 79
    },
    {
      "epoch": 0.4318488529014845,
      "grad_norm": 1.0865931108282347,
      "learning_rate": 3.4408602150537636e-05,
      "loss": 0.7059,
      "step": 80
    },
    {
      "epoch": 0.43724696356275305,
      "grad_norm": 0.9699541750890343,
      "learning_rate": 3.483870967741936e-05,
      "loss": 0.661,
      "step": 81
    },
    {
      "epoch": 0.4426450742240216,
      "grad_norm": 3.1013378557995916,
      "learning_rate": 3.526881720430108e-05,
      "loss": 0.6909,
      "step": 82
    },
    {
      "epoch": 0.44804318488529016,
      "grad_norm": 2.2538980648127542,
      "learning_rate": 3.56989247311828e-05,
      "loss": 0.6949,
      "step": 83
    },
    {
      "epoch": 0.4534412955465587,
      "grad_norm": 2.4075652108349894,
      "learning_rate": 3.612903225806452e-05,
      "loss": 0.659,
      "step": 84
    },
    {
      "epoch": 0.45883940620782726,
      "grad_norm": 7.56367856019384,
      "learning_rate": 3.655913978494624e-05,
      "loss": 0.828,
      "step": 85
    },
    {
      "epoch": 0.4642375168690958,
      "grad_norm": 2.895205886929977,
      "learning_rate": 3.698924731182796e-05,
      "loss": 0.9367,
      "step": 86
    },
    {
      "epoch": 0.46963562753036436,
      "grad_norm": 3.887496158116938,
      "learning_rate": 3.741935483870968e-05,
      "loss": 0.7459,
      "step": 87
    },
    {
      "epoch": 0.4750337381916329,
      "grad_norm": 2.274394089418949,
      "learning_rate": 3.78494623655914e-05,
      "loss": 0.7433,
      "step": 88
    },
    {
      "epoch": 0.48043184885290147,
      "grad_norm": 2.236395643430774,
      "learning_rate": 3.827956989247312e-05,
      "loss": 0.6513,
      "step": 89
    },
    {
      "epoch": 0.48582995951417,
      "grad_norm": 2.266924080857405,
      "learning_rate": 3.870967741935484e-05,
      "loss": 0.6671,
      "step": 90
    },
    {
      "epoch": 0.49122807017543857,
      "grad_norm": 1.2054447808759345,
      "learning_rate": 3.913978494623656e-05,
      "loss": 0.7066,
      "step": 91
    },
    {
      "epoch": 0.4966261808367072,
      "grad_norm": 60.63347204888335,
      "learning_rate": 3.956989247311828e-05,
      "loss": 0.6467,
      "step": 92
    },
    {
      "epoch": 0.5020242914979757,
      "grad_norm": 1.6457989694054762,
      "learning_rate": 4e-05,
      "loss": 0.7459,
      "step": 93
    },
    {
      "epoch": 0.5074224021592443,
      "grad_norm": 0.9244181817569254,
      "learning_rate": 3.999985742198083e-05,
      "loss": 0.7123,
      "step": 94
    },
    {
      "epoch": 0.5128205128205128,
      "grad_norm": 0.9024731727062596,
      "learning_rate": 3.999942968995616e-05,
      "loss": 0.6764,
      "step": 95
    },
    {
      "epoch": 0.5182186234817814,
      "grad_norm": 1.0014763698382203,
      "learning_rate": 3.99987168100245e-05,
      "loss": 0.7774,
      "step": 96
    },
    {
      "epoch": 0.5236167341430499,
      "grad_norm": 1.1017315016983185,
      "learning_rate": 3.9997718792349965e-05,
      "loss": 0.8379,
      "step": 97
    },
    {
      "epoch": 0.5290148448043185,
      "grad_norm": 0.9754394024215393,
      "learning_rate": 3.9996435651162085e-05,
      "loss": 0.7148,
      "step": 98
    },
    {
      "epoch": 0.5344129554655871,
      "grad_norm": 0.9585971972404963,
      "learning_rate": 3.999486740475564e-05,
      "loss": 0.699,
      "step": 99
    },
    {
      "epoch": 0.5398110661268556,
      "grad_norm": 1.1343058374086679,
      "learning_rate": 3.999301407549037e-05,
      "loss": 0.7095,
      "step": 100
    },
    {
      "epoch": 0.5452091767881242,
      "grad_norm": 1.06892995056798,
      "learning_rate": 3.9990875689790674e-05,
      "loss": 0.7254,
      "step": 101
    },
    {
      "epoch": 0.5506072874493927,
      "grad_norm": 1.0016896561256128,
      "learning_rate": 3.998845227814524e-05,
      "loss": 0.6704,
      "step": 102
    },
    {
      "epoch": 0.5560053981106613,
      "grad_norm": 1.0363960250258362,
      "learning_rate": 3.9985743875106584e-05,
      "loss": 0.6784,
      "step": 103
    },
    {
      "epoch": 0.5614035087719298,
      "grad_norm": 0.9134306930019722,
      "learning_rate": 3.9982750519290587e-05,
      "loss": 0.6061,
      "step": 104
    },
    {
      "epoch": 0.5668016194331984,
      "grad_norm": 1.010882912730621,
      "learning_rate": 3.997947225337592e-05,
      "loss": 0.797,
      "step": 105
    },
    {
      "epoch": 0.5721997300944669,
      "grad_norm": 1.0664043173199367,
      "learning_rate": 3.997590912410345e-05,
      "loss": 0.7478,
      "step": 106
    },
    {
      "epoch": 0.5775978407557355,
      "grad_norm": 0.8939136934200366,
      "learning_rate": 3.997206118227557e-05,
      "loss": 0.6959,
      "step": 107
    },
    {
      "epoch": 0.582995951417004,
      "grad_norm": 0.9519988505798721,
      "learning_rate": 3.996792848275546e-05,
      "loss": 0.7078,
      "step": 108
    },
    {
      "epoch": 0.5883940620782726,
      "grad_norm": 1.0090649283064343,
      "learning_rate": 3.996351108446635e-05,
      "loss": 0.8817,
      "step": 109
    },
    {
      "epoch": 0.5937921727395412,
      "grad_norm": 1.0131570568985884,
      "learning_rate": 3.9958809050390626e-05,
      "loss": 0.6692,
      "step": 110
    },
    {
      "epoch": 0.5991902834008097,
      "grad_norm": 1.0231276046162419,
      "learning_rate": 3.995382244756895e-05,
      "loss": 0.6858,
      "step": 111
    },
    {
      "epoch": 0.6045883940620783,
      "grad_norm": 1.064585741979068,
      "learning_rate": 3.994855134709931e-05,
      "loss": 0.7636,
      "step": 112
    },
    {
      "epoch": 0.6099865047233468,
      "grad_norm": 1.1130479996632003,
      "learning_rate": 3.994299582413603e-05,
      "loss": 0.762,
      "step": 113
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 1.0079939261061786,
      "learning_rate": 3.9937155957888646e-05,
      "loss": 0.7667,
      "step": 114
    },
    {
      "epoch": 0.6207827260458839,
      "grad_norm": 0.9882446367570781,
      "learning_rate": 3.9931031831620816e-05,
      "loss": 0.6575,
      "step": 115
    },
    {
      "epoch": 0.6261808367071525,
      "grad_norm": 0.9116587927342452,
      "learning_rate": 3.9924623532649124e-05,
      "loss": 0.6581,
      "step": 116
    },
    {
      "epoch": 0.631578947368421,
      "grad_norm": 1.048971008752528,
      "learning_rate": 3.991793115234182e-05,
      "loss": 0.7607,
      "step": 117
    },
    {
      "epoch": 0.6369770580296896,
      "grad_norm": 0.9629592207299862,
      "learning_rate": 3.991095478611755e-05,
      "loss": 0.6741,
      "step": 118
    },
    {
      "epoch": 0.6423751686909581,
      "grad_norm": 1.0602574291233269,
      "learning_rate": 3.990369453344394e-05,
      "loss": 0.7037,
      "step": 119
    },
    {
      "epoch": 0.6477732793522267,
      "grad_norm": 0.8986022296429939,
      "learning_rate": 3.989615049783625e-05,
      "loss": 0.7306,
      "step": 120
    },
    {
      "epoch": 0.6531713900134952,
      "grad_norm": 0.872613255131903,
      "learning_rate": 3.9888322786855846e-05,
      "loss": 0.6873,
      "step": 121
    },
    {
      "epoch": 0.6585695006747638,
      "grad_norm": 0.9970544195947879,
      "learning_rate": 3.9880211512108674e-05,
      "loss": 0.7597,
      "step": 122
    },
    {
      "epoch": 0.6639676113360324,
      "grad_norm": 0.9142253988898471,
      "learning_rate": 3.987181678924369e-05,
      "loss": 0.8754,
      "step": 123
    },
    {
      "epoch": 0.6693657219973009,
      "grad_norm": 0.8824260469924733,
      "learning_rate": 3.9863138737951175e-05,
      "loss": 0.6773,
      "step": 124
    },
    {
      "epoch": 0.6747638326585695,
      "grad_norm": 0.9378213861156586,
      "learning_rate": 3.985417748196108e-05,
      "loss": 0.6998,
      "step": 125
    },
    {
      "epoch": 0.680161943319838,
      "grad_norm": 0.9258545529411903,
      "learning_rate": 3.984493314904122e-05,
      "loss": 0.6635,
      "step": 126
    },
    {
      "epoch": 0.6855600539811066,
      "grad_norm": 0.942025251769537,
      "learning_rate": 3.983540587099545e-05,
      "loss": 0.7178,
      "step": 127
    },
    {
      "epoch": 0.6909581646423751,
      "grad_norm": 0.9024931118371414,
      "learning_rate": 3.982559578366182e-05,
      "loss": 0.8362,
      "step": 128
    },
    {
      "epoch": 0.6963562753036437,
      "grad_norm": 0.9148733733577575,
      "learning_rate": 3.98155030269106e-05,
      "loss": 0.6316,
      "step": 129
    },
    {
      "epoch": 0.7017543859649122,
      "grad_norm": 0.9421278891685188,
      "learning_rate": 3.980512774464235e-05,
      "loss": 0.7145,
      "step": 130
    },
    {
      "epoch": 0.7071524966261808,
      "grad_norm": 0.8421736450675255,
      "learning_rate": 3.979447008478575e-05,
      "loss": 0.6264,
      "step": 131
    },
    {
      "epoch": 0.7125506072874493,
      "grad_norm": 0.9448713733876393,
      "learning_rate": 3.978353019929562e-05,
      "loss": 0.7043,
      "step": 132
    },
    {
      "epoch": 0.717948717948718,
      "grad_norm": 1.0328624262343067,
      "learning_rate": 3.977230824415069e-05,
      "loss": 0.7935,
      "step": 133
    },
    {
      "epoch": 0.7233468286099866,
      "grad_norm": 0.955045550419709,
      "learning_rate": 3.9760804379351354e-05,
      "loss": 0.7184,
      "step": 134
    },
    {
      "epoch": 0.728744939271255,
      "grad_norm": 1.023688730851197,
      "learning_rate": 3.974901876891745e-05,
      "loss": 0.7233,
      "step": 135
    },
    {
      "epoch": 0.7341430499325237,
      "grad_norm": 0.9868216577806593,
      "learning_rate": 3.973695158088588e-05,
      "loss": 0.7037,
      "step": 136
    },
    {
      "epoch": 0.7395411605937922,
      "grad_norm": 0.9306844020240326,
      "learning_rate": 3.972460298730822e-05,
      "loss": 0.6593,
      "step": 137
    },
    {
      "epoch": 0.7449392712550608,
      "grad_norm": 1.0925332318325522,
      "learning_rate": 3.9711973164248255e-05,
      "loss": 0.7787,
      "step": 138
    },
    {
      "epoch": 0.7503373819163293,
      "grad_norm": 0.940936448827497,
      "learning_rate": 3.9699062291779516e-05,
      "loss": 0.6946,
      "step": 139
    },
    {
      "epoch": 0.7557354925775979,
      "grad_norm": 0.9658235331624828,
      "learning_rate": 3.9685870553982665e-05,
      "loss": 0.7574,
      "step": 140
    },
    {
      "epoch": 0.7611336032388664,
      "grad_norm": 0.9638079015407056,
      "learning_rate": 3.967239813894288e-05,
      "loss": 0.6656,
      "step": 141
    },
    {
      "epoch": 0.766531713900135,
      "grad_norm": 1.0035425179888644,
      "learning_rate": 3.965864523874718e-05,
      "loss": 0.7013,
      "step": 142
    },
    {
      "epoch": 0.7719298245614035,
      "grad_norm": 1.0201587225092634,
      "learning_rate": 3.964461204948171e-05,
      "loss": 0.6561,
      "step": 143
    },
    {
      "epoch": 0.7773279352226721,
      "grad_norm": 1.057632934842731,
      "learning_rate": 3.963029877122889e-05,
      "loss": 0.688,
      "step": 144
    },
    {
      "epoch": 0.7827260458839406,
      "grad_norm": 1.0799406339392832,
      "learning_rate": 3.961570560806461e-05,
      "loss": 0.7369,
      "step": 145
    },
    {
      "epoch": 0.7881241565452092,
      "grad_norm": 0.9807351275702001,
      "learning_rate": 3.96008327680553e-05,
      "loss": 0.755,
      "step": 146
    },
    {
      "epoch": 0.7935222672064778,
      "grad_norm": 0.8684222059026471,
      "learning_rate": 3.9585680463254965e-05,
      "loss": 0.6737,
      "step": 147
    },
    {
      "epoch": 0.7989203778677463,
      "grad_norm": 0.9291286386684726,
      "learning_rate": 3.9570248909702165e-05,
      "loss": 0.6869,
      "step": 148
    },
    {
      "epoch": 0.8043184885290149,
      "grad_norm": 0.9334466133420474,
      "learning_rate": 3.955453832741694e-05,
      "loss": 0.6368,
      "step": 149
    },
    {
      "epoch": 0.8097165991902834,
      "grad_norm": 0.9009663811050453,
      "learning_rate": 3.9538548940397654e-05,
      "loss": 0.9644,
      "step": 150
    },
    {
      "epoch": 0.815114709851552,
      "grad_norm": 1.0313471913166774,
      "learning_rate": 3.952228097661782e-05,
      "loss": 0.7582,
      "step": 151
    },
    {
      "epoch": 0.8205128205128205,
      "grad_norm": 0.9131908569576266,
      "learning_rate": 3.9505734668022844e-05,
      "loss": 0.7117,
      "step": 152
    },
    {
      "epoch": 0.8259109311740891,
      "grad_norm": 1.0165555550946475,
      "learning_rate": 3.948891025052672e-05,
      "loss": 0.6754,
      "step": 153
    },
    {
      "epoch": 0.8313090418353576,
      "grad_norm": 0.8870549778530541,
      "learning_rate": 3.947180796400865e-05,
      "loss": 0.6987,
      "step": 154
    },
    {
      "epoch": 0.8367071524966262,
      "grad_norm": 1.003808717947859,
      "learning_rate": 3.9454428052309654e-05,
      "loss": 0.6747,
      "step": 155
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 0.9111662680379845,
      "learning_rate": 3.943677076322908e-05,
      "loss": 0.787,
      "step": 156
    },
    {
      "epoch": 0.8475033738191633,
      "grad_norm": 0.9356146773942343,
      "learning_rate": 3.9418836348521045e-05,
      "loss": 0.7076,
      "step": 157
    },
    {
      "epoch": 0.8529014844804319,
      "grad_norm": 1.012979493517214,
      "learning_rate": 3.940062506389089e-05,
      "loss": 0.6868,
      "step": 158
    },
    {
      "epoch": 0.8582995951417004,
      "grad_norm": 0.9312587006321174,
      "learning_rate": 3.938213716899149e-05,
      "loss": 0.749,
      "step": 159
    },
    {
      "epoch": 0.863697705802969,
      "grad_norm": 0.9154261608012028,
      "learning_rate": 3.93633729274196e-05,
      "loss": 0.7358,
      "step": 160
    },
    {
      "epoch": 0.8690958164642375,
      "grad_norm": 0.9832755364311611,
      "learning_rate": 3.934433260671206e-05,
      "loss": 0.6628,
      "step": 161
    },
    {
      "epoch": 0.8744939271255061,
      "grad_norm": 0.9865516851317341,
      "learning_rate": 3.9325016478341986e-05,
      "loss": 0.7243,
      "step": 162
    },
    {
      "epoch": 0.8798920377867746,
      "grad_norm": 0.9580622704730626,
      "learning_rate": 3.9305424817714905e-05,
      "loss": 0.7155,
      "step": 163
    },
    {
      "epoch": 0.8852901484480432,
      "grad_norm": 0.886627340978247,
      "learning_rate": 3.928555790416485e-05,
      "loss": 0.8743,
      "step": 164
    },
    {
      "epoch": 0.8906882591093117,
      "grad_norm": 0.9369073382530939,
      "learning_rate": 3.926541602095033e-05,
      "loss": 0.8004,
      "step": 165
    },
    {
      "epoch": 0.8960863697705803,
      "grad_norm": 0.8967904070054883,
      "learning_rate": 3.9244999455250324e-05,
      "loss": 0.6837,
      "step": 166
    },
    {
      "epoch": 0.9014844804318488,
      "grad_norm": 0.8918453654014739,
      "learning_rate": 3.922430849816018e-05,
      "loss": 0.7069,
      "step": 167
    },
    {
      "epoch": 0.9068825910931174,
      "grad_norm": 0.9468000825480711,
      "learning_rate": 3.9203343444687475e-05,
      "loss": 0.6991,
      "step": 168
    },
    {
      "epoch": 0.9122807017543859,
      "grad_norm": 0.8690439611058536,
      "learning_rate": 3.918210459374778e-05,
      "loss": 0.6544,
      "step": 169
    },
    {
      "epoch": 0.9176788124156545,
      "grad_norm": 0.9203492753195474,
      "learning_rate": 3.916059224816043e-05,
      "loss": 0.8887,
      "step": 170
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 0.9400256924968273,
      "learning_rate": 3.913880671464418e-05,
      "loss": 0.6707,
      "step": 171
    },
    {
      "epoch": 0.9284750337381916,
      "grad_norm": 0.9929868868660712,
      "learning_rate": 3.911674830381286e-05,
      "loss": 0.746,
      "step": 172
    },
    {
      "epoch": 0.9338731443994602,
      "grad_norm": 0.8996290548917789,
      "learning_rate": 3.909441733017092e-05,
      "loss": 0.6638,
      "step": 173
    },
    {
      "epoch": 0.9392712550607287,
      "grad_norm": 1.0383078650677402,
      "learning_rate": 3.907181411210895e-05,
      "loss": 0.8804,
      "step": 174
    },
    {
      "epoch": 0.9446693657219973,
      "grad_norm": 0.8736475053968872,
      "learning_rate": 3.904893897189917e-05,
      "loss": 0.7383,
      "step": 175
    },
    {
      "epoch": 0.9500674763832658,
      "grad_norm": 0.8879797003351837,
      "learning_rate": 3.9025792235690787e-05,
      "loss": 0.686,
      "step": 176
    },
    {
      "epoch": 0.9554655870445344,
      "grad_norm": 0.8852717088467212,
      "learning_rate": 3.900237423350539e-05,
      "loss": 0.692,
      "step": 177
    },
    {
      "epoch": 0.9608636977058029,
      "grad_norm": 0.856393902188942,
      "learning_rate": 3.89786852992322e-05,
      "loss": 0.6664,
      "step": 178
    },
    {
      "epoch": 0.9662618083670715,
      "grad_norm": 0.9778728102030992,
      "learning_rate": 3.895472577062337e-05,
      "loss": 0.6959,
      "step": 179
    },
    {
      "epoch": 0.97165991902834,
      "grad_norm": 0.9807365431125,
      "learning_rate": 3.89304959892891e-05,
      "loss": 0.6826,
      "step": 180
    },
    {
      "epoch": 0.9770580296896086,
      "grad_norm": 0.9397890000073238,
      "learning_rate": 3.8905996300692806e-05,
      "loss": 0.8002,
      "step": 181
    },
    {
      "epoch": 0.9824561403508771,
      "grad_norm": 0.9314290560742274,
      "learning_rate": 3.888122705414621e-05,
      "loss": 0.6628,
      "step": 182
    },
    {
      "epoch": 0.9878542510121457,
      "grad_norm": 0.9348683094207633,
      "learning_rate": 3.885618860280433e-05,
      "loss": 0.7152,
      "step": 183
    },
    {
      "epoch": 0.9932523616734144,
      "grad_norm": 0.9283592796406461,
      "learning_rate": 3.883088130366042e-05,
      "loss": 0.6817,
      "step": 184
    },
    {
      "epoch": 0.9986504723346828,
      "grad_norm": 0.8957200626385793,
      "learning_rate": 3.880530551754095e-05,
      "loss": 0.6965,
      "step": 185
    },
    {
      "epoch": 1.0040485829959513,
      "grad_norm": 2.135429932537726,
      "learning_rate": 3.877946160910042e-05,
      "loss": 0.9106,
      "step": 186
    },
    {
      "epoch": 1.00944669365722,
      "grad_norm": 0.9324649166463199,
      "learning_rate": 3.8753349946816154e-05,
      "loss": 0.5466,
      "step": 187
    },
    {
      "epoch": 1.0148448043184886,
      "grad_norm": 0.889079188745322,
      "learning_rate": 3.8726970902983046e-05,
      "loss": 0.4034,
      "step": 188
    },
    {
      "epoch": 1.0202429149797572,
      "grad_norm": 0.9818393519416749,
      "learning_rate": 3.8700324853708304e-05,
      "loss": 0.3707,
      "step": 189
    },
    {
      "epoch": 1.0256410256410255,
      "grad_norm": 0.9989694298887677,
      "learning_rate": 3.867341217890599e-05,
      "loss": 0.3552,
      "step": 190
    },
    {
      "epoch": 1.0310391363022942,
      "grad_norm": 1.0964706430273843,
      "learning_rate": 3.864623326229172e-05,
      "loss": 0.3406,
      "step": 191
    },
    {
      "epoch": 1.0364372469635628,
      "grad_norm": 1.041322435377106,
      "learning_rate": 3.861878849137708e-05,
      "loss": 0.3507,
      "step": 192
    },
    {
      "epoch": 1.0418353576248314,
      "grad_norm": 1.0000512392448195,
      "learning_rate": 3.859107825746419e-05,
      "loss": 0.3974,
      "step": 193
    },
    {
      "epoch": 1.0472334682861,
      "grad_norm": 0.9147739098501848,
      "learning_rate": 3.8563102955640076e-05,
      "loss": 0.5635,
      "step": 194
    },
    {
      "epoch": 1.0526315789473684,
      "grad_norm": 0.8437219756968998,
      "learning_rate": 3.853486298477105e-05,
      "loss": 0.4385,
      "step": 195
    },
    {
      "epoch": 1.058029689608637,
      "grad_norm": 0.8292889033394908,
      "learning_rate": 3.850635874749701e-05,
      "loss": 0.3399,
      "step": 196
    },
    {
      "epoch": 1.0634278002699056,
      "grad_norm": 0.7709540617055263,
      "learning_rate": 3.8477590650225735e-05,
      "loss": 0.359,
      "step": 197
    },
    {
      "epoch": 1.0688259109311742,
      "grad_norm": 0.834537390618175,
      "learning_rate": 3.8448559103127065e-05,
      "loss": 0.5317,
      "step": 198
    },
    {
      "epoch": 1.0742240215924426,
      "grad_norm": 0.9152705040904082,
      "learning_rate": 3.841926452012704e-05,
      "loss": 0.3868,
      "step": 199
    },
    {
      "epoch": 1.0796221322537112,
      "grad_norm": 0.9881357548735162,
      "learning_rate": 3.838970731890202e-05,
      "loss": 0.391,
      "step": 200
    },
    {
      "epoch": 1.0850202429149798,
      "grad_norm": 0.9213113248556943,
      "learning_rate": 3.835988792087272e-05,
      "loss": 0.5105,
      "step": 201
    },
    {
      "epoch": 1.0904183535762484,
      "grad_norm": 0.9932763071736556,
      "learning_rate": 3.832980675119823e-05,
      "loss": 0.44,
      "step": 202
    },
    {
      "epoch": 1.0958164642375168,
      "grad_norm": 0.9979438465584454,
      "learning_rate": 3.8299464238769883e-05,
      "loss": 0.4094,
      "step": 203
    },
    {
      "epoch": 1.1012145748987854,
      "grad_norm": 0.7593268974099039,
      "learning_rate": 3.826886081620523e-05,
      "loss": 0.3163,
      "step": 204
    },
    {
      "epoch": 1.106612685560054,
      "grad_norm": 0.8549555976031787,
      "learning_rate": 3.82379969198418e-05,
      "loss": 0.5373,
      "step": 205
    },
    {
      "epoch": 1.1120107962213226,
      "grad_norm": 0.8485934704490189,
      "learning_rate": 3.8206872989730926e-05,
      "loss": 0.3291,
      "step": 206
    },
    {
      "epoch": 1.117408906882591,
      "grad_norm": 1.1808502925030266,
      "learning_rate": 3.817548946963142e-05,
      "loss": 0.3446,
      "step": 207
    },
    {
      "epoch": 1.1228070175438596,
      "grad_norm": 0.8718322044501718,
      "learning_rate": 3.81438468070033e-05,
      "loss": 0.3496,
      "step": 208
    },
    {
      "epoch": 1.1282051282051282,
      "grad_norm": 0.8269069896746979,
      "learning_rate": 3.811194545300139e-05,
      "loss": 0.3637,
      "step": 209
    },
    {
      "epoch": 1.1336032388663968,
      "grad_norm": 0.8241831789016307,
      "learning_rate": 3.807978586246887e-05,
      "loss": 0.329,
      "step": 210
    },
    {
      "epoch": 1.1390013495276654,
      "grad_norm": 0.8440307299715044,
      "learning_rate": 3.8047368493930814e-05,
      "loss": 0.3695,
      "step": 211
    },
    {
      "epoch": 1.1443994601889338,
      "grad_norm": 0.9057523221019308,
      "learning_rate": 3.8014693809587635e-05,
      "loss": 0.4349,
      "step": 212
    },
    {
      "epoch": 1.1497975708502024,
      "grad_norm": 0.7812716885128814,
      "learning_rate": 3.798176227530852e-05,
      "loss": 0.319,
      "step": 213
    },
    {
      "epoch": 1.155195681511471,
      "grad_norm": 0.754153255972741,
      "learning_rate": 3.794857436062476e-05,
      "loss": 0.3145,
      "step": 214
    },
    {
      "epoch": 1.1605937921727396,
      "grad_norm": 0.8752509690112213,
      "learning_rate": 3.791513053872306e-05,
      "loss": 0.3811,
      "step": 215
    },
    {
      "epoch": 1.165991902834008,
      "grad_norm": 0.8620173420159217,
      "learning_rate": 3.788143128643881e-05,
      "loss": 0.3402,
      "step": 216
    },
    {
      "epoch": 1.1713900134952766,
      "grad_norm": 0.8574889394982422,
      "learning_rate": 3.784747708424929e-05,
      "loss": 0.3523,
      "step": 217
    },
    {
      "epoch": 1.1767881241565452,
      "grad_norm": 0.9171855250942973,
      "learning_rate": 3.781326841626677e-05,
      "loss": 0.3733,
      "step": 218
    },
    {
      "epoch": 1.1821862348178138,
      "grad_norm": 0.8459848989881751,
      "learning_rate": 3.777880577023167e-05,
      "loss": 0.3391,
      "step": 219
    },
    {
      "epoch": 1.1875843454790824,
      "grad_norm": 1.7633085482731143,
      "learning_rate": 3.7744089637505565e-05,
      "loss": 0.6142,
      "step": 220
    },
    {
      "epoch": 1.1929824561403508,
      "grad_norm": 0.8882381136541697,
      "learning_rate": 3.7709120513064196e-05,
      "loss": 0.3793,
      "step": 221
    },
    {
      "epoch": 1.1983805668016194,
      "grad_norm": 0.9069761964495866,
      "learning_rate": 3.7673898895490435e-05,
      "loss": 0.3425,
      "step": 222
    },
    {
      "epoch": 1.203778677462888,
      "grad_norm": 0.9280379690711708,
      "learning_rate": 3.76384252869671e-05,
      "loss": 0.4051,
      "step": 223
    },
    {
      "epoch": 1.2091767881241566,
      "grad_norm": 0.8271806961908379,
      "learning_rate": 3.760270019326989e-05,
      "loss": 0.3556,
      "step": 224
    },
    {
      "epoch": 1.214574898785425,
      "grad_norm": 0.9136083147626765,
      "learning_rate": 3.7566724123760126e-05,
      "loss": 0.3273,
      "step": 225
    },
    {
      "epoch": 1.2199730094466936,
      "grad_norm": 0.9337801648538931,
      "learning_rate": 3.753049759137745e-05,
      "loss": 0.3721,
      "step": 226
    },
    {
      "epoch": 1.2253711201079622,
      "grad_norm": 0.8801415414022893,
      "learning_rate": 3.749402111263261e-05,
      "loss": 0.5007,
      "step": 227
    },
    {
      "epoch": 1.2307692307692308,
      "grad_norm": 0.8738904307010843,
      "learning_rate": 3.745729520760001e-05,
      "loss": 0.543,
      "step": 228
    },
    {
      "epoch": 1.2361673414304994,
      "grad_norm": 0.9215671196261831,
      "learning_rate": 3.7420320399910315e-05,
      "loss": 0.4912,
      "step": 229
    },
    {
      "epoch": 1.2415654520917678,
      "grad_norm": 0.7882682925458486,
      "learning_rate": 3.738309721674302e-05,
      "loss": 0.3524,
      "step": 230
    },
    {
      "epoch": 1.2469635627530364,
      "grad_norm": 0.9028612409679585,
      "learning_rate": 3.7345626188818906e-05,
      "loss": 0.4339,
      "step": 231
    },
    {
      "epoch": 1.252361673414305,
      "grad_norm": 0.8352453883537844,
      "learning_rate": 3.730790785039245e-05,
      "loss": 0.3874,
      "step": 232
    },
    {
      "epoch": 1.2577597840755734,
      "grad_norm": 0.8506796880369457,
      "learning_rate": 3.726994273924426e-05,
      "loss": 0.3473,
      "step": 233
    },
    {
      "epoch": 1.263157894736842,
      "grad_norm": 0.794622563861647,
      "learning_rate": 3.7231731396673365e-05,
      "loss": 0.3548,
      "step": 234
    },
    {
      "epoch": 1.2685560053981106,
      "grad_norm": 0.8169696872933733,
      "learning_rate": 3.7193274367489524e-05,
      "loss": 0.3366,
      "step": 235
    },
    {
      "epoch": 1.2739541160593792,
      "grad_norm": 0.9414753748124672,
      "learning_rate": 3.7154572200005446e-05,
      "loss": 0.3352,
      "step": 236
    },
    {
      "epoch": 1.2793522267206479,
      "grad_norm": 0.8524753438779967,
      "learning_rate": 3.711562544602895e-05,
      "loss": 0.3668,
      "step": 237
    },
    {
      "epoch": 1.2847503373819165,
      "grad_norm": 0.8494604337265534,
      "learning_rate": 3.707643466085516e-05,
      "loss": 0.3519,
      "step": 238
    },
    {
      "epoch": 1.2901484480431848,
      "grad_norm": 1.017755403367373,
      "learning_rate": 3.703700040325852e-05,
      "loss": 0.4038,
      "step": 239
    },
    {
      "epoch": 1.2955465587044535,
      "grad_norm": 0.7951998743639334,
      "learning_rate": 3.699732323548485e-05,
      "loss": 0.3344,
      "step": 240
    },
    {
      "epoch": 1.300944669365722,
      "grad_norm": 0.8451484427027826,
      "learning_rate": 3.695740372324337e-05,
      "loss": 0.3604,
      "step": 241
    },
    {
      "epoch": 1.3063427800269904,
      "grad_norm": 0.9474589070464138,
      "learning_rate": 3.691724243569857e-05,
      "loss": 0.3783,
      "step": 242
    },
    {
      "epoch": 1.311740890688259,
      "grad_norm": 0.9167363147150677,
      "learning_rate": 3.687683994546213e-05,
      "loss": 0.3336,
      "step": 243
    },
    {
      "epoch": 1.3171390013495277,
      "grad_norm": 0.8499771648824146,
      "learning_rate": 3.683619682858476e-05,
      "loss": 0.3706,
      "step": 244
    },
    {
      "epoch": 1.3225371120107963,
      "grad_norm": 0.8642561348231351,
      "learning_rate": 3.6795313664547965e-05,
      "loss": 0.3615,
      "step": 245
    },
    {
      "epoch": 1.3279352226720649,
      "grad_norm": 0.7524096422021259,
      "learning_rate": 3.675419103625579e-05,
      "loss": 0.328,
      "step": 246
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.8817200727149742,
      "learning_rate": 3.6712829530026535e-05,
      "loss": 0.3983,
      "step": 247
    },
    {
      "epoch": 1.3387314439946019,
      "grad_norm": 0.7586442964430375,
      "learning_rate": 3.6671229735584365e-05,
      "loss": 0.2935,
      "step": 248
    },
    {
      "epoch": 1.3441295546558705,
      "grad_norm": 0.9626722935944538,
      "learning_rate": 3.662939224605091e-05,
      "loss": 0.567,
      "step": 249
    },
    {
      "epoch": 1.349527665317139,
      "grad_norm": 0.7968343661737369,
      "learning_rate": 3.658731765793679e-05,
      "loss": 0.3566,
      "step": 250
    },
    {
      "epoch": 1.3549257759784075,
      "grad_norm": 0.899257227621249,
      "learning_rate": 3.6545006571133174e-05,
      "loss": 0.4005,
      "step": 251
    },
    {
      "epoch": 1.360323886639676,
      "grad_norm": 0.8787913231865052,
      "learning_rate": 3.650245958890314e-05,
      "loss": 0.4003,
      "step": 252
    },
    {
      "epoch": 1.3657219973009447,
      "grad_norm": 0.9125900215239395,
      "learning_rate": 3.645967731787313e-05,
      "loss": 0.4386,
      "step": 253
    },
    {
      "epoch": 1.3711201079622133,
      "grad_norm": 0.8331834311620439,
      "learning_rate": 3.64166603680243e-05,
      "loss": 0.3307,
      "step": 254
    },
    {
      "epoch": 1.376518218623482,
      "grad_norm": 0.8425535041307437,
      "learning_rate": 3.63734093526838e-05,
      "loss": 0.3347,
      "step": 255
    },
    {
      "epoch": 1.3819163292847503,
      "grad_norm": 0.8290080988756854,
      "learning_rate": 3.632992488851603e-05,
      "loss": 0.3312,
      "step": 256
    },
    {
      "epoch": 1.3873144399460189,
      "grad_norm": 0.9378460105642437,
      "learning_rate": 3.6286207595513884e-05,
      "loss": 0.4643,
      "step": 257
    },
    {
      "epoch": 1.3927125506072875,
      "grad_norm": 0.8267549903762161,
      "learning_rate": 3.624225809698984e-05,
      "loss": 0.3539,
      "step": 258
    },
    {
      "epoch": 1.3981106612685559,
      "grad_norm": 0.8768268582940559,
      "learning_rate": 3.619807701956717e-05,
      "loss": 0.6197,
      "step": 259
    },
    {
      "epoch": 1.4035087719298245,
      "grad_norm": 0.8906787557494711,
      "learning_rate": 3.61536649931709e-05,
      "loss": 0.3794,
      "step": 260
    },
    {
      "epoch": 1.408906882591093,
      "grad_norm": 0.8376396972815752,
      "learning_rate": 3.610902265101892e-05,
      "loss": 0.3374,
      "step": 261
    },
    {
      "epoch": 1.4143049932523617,
      "grad_norm": 0.9702245743999838,
      "learning_rate": 3.60641506296129e-05,
      "loss": 0.3682,
      "step": 262
    },
    {
      "epoch": 1.4197031039136303,
      "grad_norm": 0.8599060721053028,
      "learning_rate": 3.601904956872923e-05,
      "loss": 0.4175,
      "step": 263
    },
    {
      "epoch": 1.425101214574899,
      "grad_norm": 0.8379932139189499,
      "learning_rate": 3.5973720111409904e-05,
      "loss": 0.3271,
      "step": 264
    },
    {
      "epoch": 1.4304993252361673,
      "grad_norm": 0.8606949214353987,
      "learning_rate": 3.5928162903953347e-05,
      "loss": 0.3449,
      "step": 265
    },
    {
      "epoch": 1.435897435897436,
      "grad_norm": 0.8688526476474914,
      "learning_rate": 3.5882378595905195e-05,
      "loss": 0.4602,
      "step": 266
    },
    {
      "epoch": 1.4412955465587045,
      "grad_norm": 0.7995357785557587,
      "learning_rate": 3.583636784004904e-05,
      "loss": 0.5264,
      "step": 267
    },
    {
      "epoch": 1.446693657219973,
      "grad_norm": 0.8388725595998526,
      "learning_rate": 3.5790131292397135e-05,
      "loss": 0.3738,
      "step": 268
    },
    {
      "epoch": 1.4520917678812415,
      "grad_norm": 0.845571143136271,
      "learning_rate": 3.5743669612181004e-05,
      "loss": 0.3907,
      "step": 269
    },
    {
      "epoch": 1.45748987854251,
      "grad_norm": 0.8086293381970037,
      "learning_rate": 3.569698346184209e-05,
      "loss": 0.5639,
      "step": 270
    },
    {
      "epoch": 1.4628879892037787,
      "grad_norm": 0.9383168628054654,
      "learning_rate": 3.5650073507022286e-05,
      "loss": 0.3871,
      "step": 271
    },
    {
      "epoch": 1.4682860998650473,
      "grad_norm": 0.8681442872940774,
      "learning_rate": 3.560294041655442e-05,
      "loss": 0.397,
      "step": 272
    },
    {
      "epoch": 1.4736842105263157,
      "grad_norm": 0.8268737914017257,
      "learning_rate": 3.555558486245277e-05,
      "loss": 0.3486,
      "step": 273
    },
    {
      "epoch": 1.4790823211875843,
      "grad_norm": 0.8680158503834453,
      "learning_rate": 3.550800751990343e-05,
      "loss": 0.3518,
      "step": 274
    },
    {
      "epoch": 1.484480431848853,
      "grad_norm": 0.7711969794181421,
      "learning_rate": 3.546020906725474e-05,
      "loss": 0.3297,
      "step": 275
    },
    {
      "epoch": 1.4898785425101215,
      "grad_norm": 0.7746266967179702,
      "learning_rate": 3.5412190186007573e-05,
      "loss": 0.3296,
      "step": 276
    },
    {
      "epoch": 1.49527665317139,
      "grad_norm": 0.866167286472648,
      "learning_rate": 3.5363951560805615e-05,
      "loss": 0.4019,
      "step": 277
    },
    {
      "epoch": 1.5006747638326585,
      "grad_norm": 0.8349494942186633,
      "learning_rate": 3.5315493879425634e-05,
      "loss": 0.4084,
      "step": 278
    },
    {
      "epoch": 1.5060728744939271,
      "grad_norm": 0.7622596623778285,
      "learning_rate": 3.526681783276765e-05,
      "loss": 0.4291,
      "step": 279
    },
    {
      "epoch": 1.5114709851551957,
      "grad_norm": 0.8720420315217544,
      "learning_rate": 3.52179241148451e-05,
      "loss": 0.37,
      "step": 280
    },
    {
      "epoch": 1.5168690958164643,
      "grad_norm": 0.920897251888379,
      "learning_rate": 3.5168813422774923e-05,
      "loss": 0.3914,
      "step": 281
    },
    {
      "epoch": 1.522267206477733,
      "grad_norm": 0.8228842239801039,
      "learning_rate": 3.511948645676764e-05,
      "loss": 0.3532,
      "step": 282
    },
    {
      "epoch": 1.5276653171390013,
      "grad_norm": 0.7888080706333876,
      "learning_rate": 3.506994392011737e-05,
      "loss": 0.5181,
      "step": 283
    },
    {
      "epoch": 1.53306342780027,
      "grad_norm": 0.8705748415419389,
      "learning_rate": 3.5020186519191775e-05,
      "loss": 0.3917,
      "step": 284
    },
    {
      "epoch": 1.5384615384615383,
      "grad_norm": 0.8946816906145529,
      "learning_rate": 3.497021496342203e-05,
      "loss": 0.4199,
      "step": 285
    },
    {
      "epoch": 1.543859649122807,
      "grad_norm": 0.7859844266157671,
      "learning_rate": 3.492002996529267e-05,
      "loss": 0.3159,
      "step": 286
    },
    {
      "epoch": 1.5492577597840755,
      "grad_norm": 0.8271181630355136,
      "learning_rate": 3.486963224033146e-05,
      "loss": 0.5559,
      "step": 287
    },
    {
      "epoch": 1.5546558704453441,
      "grad_norm": 0.8337970276185057,
      "learning_rate": 3.4819022507099184e-05,
      "loss": 0.3672,
      "step": 288
    },
    {
      "epoch": 1.5600539811066128,
      "grad_norm": 0.9863741535233492,
      "learning_rate": 3.4768201487179395e-05,
      "loss": 0.3333,
      "step": 289
    },
    {
      "epoch": 1.5654520917678814,
      "grad_norm": 1.0399201093786894,
      "learning_rate": 3.471716990516812e-05,
      "loss": 0.4489,
      "step": 290
    },
    {
      "epoch": 1.5708502024291497,
      "grad_norm": 0.8592335687596742,
      "learning_rate": 3.466592848866356e-05,
      "loss": 0.3636,
      "step": 291
    },
    {
      "epoch": 1.5762483130904184,
      "grad_norm": 0.9016848617465185,
      "learning_rate": 3.4614477968255663e-05,
      "loss": 0.4327,
      "step": 292
    },
    {
      "epoch": 1.581646423751687,
      "grad_norm": 0.8819320121086326,
      "learning_rate": 3.456281907751577e-05,
      "loss": 0.351,
      "step": 293
    },
    {
      "epoch": 1.5870445344129553,
      "grad_norm": 0.8025348430113183,
      "learning_rate": 3.4510952552986114e-05,
      "loss": 0.3266,
      "step": 294
    },
    {
      "epoch": 1.592442645074224,
      "grad_norm": 0.8800970701171718,
      "learning_rate": 3.445887913416932e-05,
      "loss": 0.3621,
      "step": 295
    },
    {
      "epoch": 1.5978407557354926,
      "grad_norm": 0.8808067572947517,
      "learning_rate": 3.440659956351788e-05,
      "loss": 0.3376,
      "step": 296
    },
    {
      "epoch": 1.6032388663967612,
      "grad_norm": 1.1199823914819425,
      "learning_rate": 3.435411458642357e-05,
      "loss": 0.3798,
      "step": 297
    },
    {
      "epoch": 1.6086369770580298,
      "grad_norm": 1.0162058885262826,
      "learning_rate": 3.430142495120678e-05,
      "loss": 0.3566,
      "step": 298
    },
    {
      "epoch": 1.6140350877192984,
      "grad_norm": 0.8782900680409256,
      "learning_rate": 3.4248531409105896e-05,
      "loss": 0.3372,
      "step": 299
    },
    {
      "epoch": 1.6194331983805668,
      "grad_norm": 0.8000888023404262,
      "learning_rate": 3.419543471426657e-05,
      "loss": 0.3644,
      "step": 300
    },
    {
      "epoch": 1.6248313090418354,
      "grad_norm": 0.843844017800044,
      "learning_rate": 3.4142135623730954e-05,
      "loss": 0.385,
      "step": 301
    },
    {
      "epoch": 1.6302294197031038,
      "grad_norm": 0.8497338959321703,
      "learning_rate": 3.408863489742692e-05,
      "loss": 0.3938,
      "step": 302
    },
    {
      "epoch": 1.6356275303643724,
      "grad_norm": 0.774530442372388,
      "learning_rate": 3.403493329815724e-05,
      "loss": 0.3193,
      "step": 303
    },
    {
      "epoch": 1.641025641025641,
      "grad_norm": 0.8713896929383775,
      "learning_rate": 3.3981031591588665e-05,
      "loss": 0.352,
      "step": 304
    },
    {
      "epoch": 1.6464237516869096,
      "grad_norm": 0.8388021660084379,
      "learning_rate": 3.392693054624106e-05,
      "loss": 0.348,
      "step": 305
    },
    {
      "epoch": 1.6518218623481782,
      "grad_norm": 0.8436318600036422,
      "learning_rate": 3.387263093347641e-05,
      "loss": 0.367,
      "step": 306
    },
    {
      "epoch": 1.6572199730094468,
      "grad_norm": 0.8392694994613278,
      "learning_rate": 3.381813352748783e-05,
      "loss": 0.3621,
      "step": 307
    },
    {
      "epoch": 1.6626180836707154,
      "grad_norm": 0.8188401843130375,
      "learning_rate": 3.376343910528855e-05,
      "loss": 0.3918,
      "step": 308
    },
    {
      "epoch": 1.6680161943319838,
      "grad_norm": 0.8043575662262005,
      "learning_rate": 3.37085484467008e-05,
      "loss": 0.3546,
      "step": 309
    },
    {
      "epoch": 1.6734143049932524,
      "grad_norm": 0.8941674035447104,
      "learning_rate": 3.365346233434472e-05,
      "loss": 0.3941,
      "step": 310
    },
    {
      "epoch": 1.6788124156545208,
      "grad_norm": 0.8580597773997315,
      "learning_rate": 3.359818155362719e-05,
      "loss": 0.4437,
      "step": 311
    },
    {
      "epoch": 1.6842105263157894,
      "grad_norm": 0.9187515726131583,
      "learning_rate": 3.354270689273064e-05,
      "loss": 0.3914,
      "step": 312
    },
    {
      "epoch": 1.689608636977058,
      "grad_norm": 0.8126321379065723,
      "learning_rate": 3.348703914260178e-05,
      "loss": 0.3402,
      "step": 313
    },
    {
      "epoch": 1.6950067476383266,
      "grad_norm": 0.9023277271934033,
      "learning_rate": 3.3431179096940375e-05,
      "loss": 0.3702,
      "step": 314
    },
    {
      "epoch": 1.7004048582995952,
      "grad_norm": 0.9134916790344669,
      "learning_rate": 3.3375127552187885e-05,
      "loss": 0.399,
      "step": 315
    },
    {
      "epoch": 1.7058029689608638,
      "grad_norm": 0.7631514166634741,
      "learning_rate": 3.331888530751613e-05,
      "loss": 0.3151,
      "step": 316
    },
    {
      "epoch": 1.7112010796221324,
      "grad_norm": 0.9238661037906718,
      "learning_rate": 3.326245316481591e-05,
      "loss": 0.4818,
      "step": 317
    },
    {
      "epoch": 1.7165991902834008,
      "grad_norm": 0.7902959093954609,
      "learning_rate": 3.320583192868552e-05,
      "loss": 0.3446,
      "step": 318
    },
    {
      "epoch": 1.7219973009446694,
      "grad_norm": 0.8651279342148734,
      "learning_rate": 3.3149022406419335e-05,
      "loss": 0.3842,
      "step": 319
    },
    {
      "epoch": 1.7273954116059378,
      "grad_norm": 0.872911869425239,
      "learning_rate": 3.309202540799628e-05,
      "loss": 0.4333,
      "step": 320
    },
    {
      "epoch": 1.7327935222672064,
      "grad_norm": 0.819055468707437,
      "learning_rate": 3.303484174606824e-05,
      "loss": 0.3727,
      "step": 321
    },
    {
      "epoch": 1.738191632928475,
      "grad_norm": 0.7643068590729667,
      "learning_rate": 3.297747223594858e-05,
      "loss": 0.3481,
      "step": 322
    },
    {
      "epoch": 1.7435897435897436,
      "grad_norm": 0.827195670933345,
      "learning_rate": 3.291991769560038e-05,
      "loss": 0.355,
      "step": 323
    },
    {
      "epoch": 1.7489878542510122,
      "grad_norm": 0.883234143840346,
      "learning_rate": 3.286217894562489e-05,
      "loss": 0.4227,
      "step": 324
    },
    {
      "epoch": 1.7543859649122808,
      "grad_norm": 0.8280893038921171,
      "learning_rate": 3.280425680924976e-05,
      "loss": 0.3337,
      "step": 325
    },
    {
      "epoch": 1.7597840755735492,
      "grad_norm": 0.8012629303209649,
      "learning_rate": 3.274615211231735e-05,
      "loss": 0.3394,
      "step": 326
    },
    {
      "epoch": 1.7651821862348178,
      "grad_norm": 0.8237991067683915,
      "learning_rate": 3.268786568327291e-05,
      "loss": 0.3799,
      "step": 327
    },
    {
      "epoch": 1.7705802968960864,
      "grad_norm": 0.9040042326048502,
      "learning_rate": 3.262939835315281e-05,
      "loss": 0.4133,
      "step": 328
    },
    {
      "epoch": 1.7759784075573548,
      "grad_norm": 0.8013478185913319,
      "learning_rate": 3.2570750955572643e-05,
      "loss": 0.3828,
      "step": 329
    },
    {
      "epoch": 1.7813765182186234,
      "grad_norm": 0.7411195677522415,
      "learning_rate": 3.25119243267154e-05,
      "loss": 0.3108,
      "step": 330
    },
    {
      "epoch": 1.786774628879892,
      "grad_norm": 0.8217449129947968,
      "learning_rate": 3.2452919305319505e-05,
      "loss": 0.3499,
      "step": 331
    },
    {
      "epoch": 1.7921727395411606,
      "grad_norm": 0.8595509733708185,
      "learning_rate": 3.239373673266686e-05,
      "loss": 0.3666,
      "step": 332
    },
    {
      "epoch": 1.7975708502024292,
      "grad_norm": 0.8868926190097939,
      "learning_rate": 3.2334377452570866e-05,
      "loss": 0.4031,
      "step": 333
    },
    {
      "epoch": 1.8029689608636978,
      "grad_norm": 0.7623837903935355,
      "learning_rate": 3.227484231136437e-05,
      "loss": 0.3092,
      "step": 334
    },
    {
      "epoch": 1.8083670715249662,
      "grad_norm": 0.8231066068295543,
      "learning_rate": 3.2215132157887636e-05,
      "loss": 0.3553,
      "step": 335
    },
    {
      "epoch": 1.8137651821862348,
      "grad_norm": 0.8069543975515238,
      "learning_rate": 3.2155247843476204e-05,
      "loss": 0.353,
      "step": 336
    },
    {
      "epoch": 1.8191632928475032,
      "grad_norm": 0.7982465506547287,
      "learning_rate": 3.209519022194875e-05,
      "loss": 0.3698,
      "step": 337
    },
    {
      "epoch": 1.8245614035087718,
      "grad_norm": 0.8918760369983304,
      "learning_rate": 3.203496014959497e-05,
      "loss": 0.4392,
      "step": 338
    },
    {
      "epoch": 1.8299595141700404,
      "grad_norm": 0.9160858315953113,
      "learning_rate": 3.197455848516328e-05,
      "loss": 0.3949,
      "step": 339
    },
    {
      "epoch": 1.835357624831309,
      "grad_norm": 0.9222161272336913,
      "learning_rate": 3.191398608984867e-05,
      "loss": 0.3974,
      "step": 340
    },
    {
      "epoch": 1.8407557354925776,
      "grad_norm": 0.7564011344843046,
      "learning_rate": 3.185324382728034e-05,
      "loss": 0.3199,
      "step": 341
    },
    {
      "epoch": 1.8461538461538463,
      "grad_norm": 0.8272808795137335,
      "learning_rate": 3.179233256350944e-05,
      "loss": 0.3794,
      "step": 342
    },
    {
      "epoch": 1.8515519568151149,
      "grad_norm": 0.8965111151969724,
      "learning_rate": 3.173125316699671e-05,
      "loss": 0.5687,
      "step": 343
    },
    {
      "epoch": 1.8569500674763832,
      "grad_norm": 0.8269731890692298,
      "learning_rate": 3.1670006508600076e-05,
      "loss": 0.3222,
      "step": 344
    },
    {
      "epoch": 1.8623481781376519,
      "grad_norm": 0.9313561418996097,
      "learning_rate": 3.160859346156227e-05,
      "loss": 0.3745,
      "step": 345
    },
    {
      "epoch": 1.8677462887989202,
      "grad_norm": 0.8236123893552746,
      "learning_rate": 3.1547014901498344e-05,
      "loss": 0.3848,
      "step": 346
    },
    {
      "epoch": 1.8731443994601888,
      "grad_norm": 0.8881682095506358,
      "learning_rate": 3.148527170638322e-05,
      "loss": 0.437,
      "step": 347
    },
    {
      "epoch": 1.8785425101214575,
      "grad_norm": 0.816673941404442,
      "learning_rate": 3.1423364756539135e-05,
      "loss": 0.3823,
      "step": 348
    },
    {
      "epoch": 1.883940620782726,
      "grad_norm": 0.8353539656538781,
      "learning_rate": 3.136129493462312e-05,
      "loss": 0.3817,
      "step": 349
    },
    {
      "epoch": 1.8893387314439947,
      "grad_norm": 0.8623642136451524,
      "learning_rate": 3.12990631256144e-05,
      "loss": 0.3773,
      "step": 350
    },
    {
      "epoch": 1.8947368421052633,
      "grad_norm": 0.8785883532356465,
      "learning_rate": 3.1236670216801786e-05,
      "loss": 0.3782,
      "step": 351
    },
    {
      "epoch": 1.9001349527665317,
      "grad_norm": 0.9116852292901764,
      "learning_rate": 3.117411709777101e-05,
      "loss": 0.4707,
      "step": 352
    },
    {
      "epoch": 1.9055330634278003,
      "grad_norm": 0.8258672453724147,
      "learning_rate": 3.111140466039205e-05,
      "loss": 0.3295,
      "step": 353
    },
    {
      "epoch": 1.9109311740890689,
      "grad_norm": 0.8047035314518469,
      "learning_rate": 3.104853379880641e-05,
      "loss": 0.5049,
      "step": 354
    },
    {
      "epoch": 1.9163292847503373,
      "grad_norm": 0.8224201466858146,
      "learning_rate": 3.0985505409414395e-05,
      "loss": 0.3459,
      "step": 355
    },
    {
      "epoch": 1.9217273954116059,
      "grad_norm": 0.8194601649707134,
      "learning_rate": 3.09223203908623e-05,
      "loss": 0.5184,
      "step": 356
    },
    {
      "epoch": 1.9271255060728745,
      "grad_norm": 1.9213742164403345,
      "learning_rate": 3.085897964402958e-05,
      "loss": 0.4168,
      "step": 357
    },
    {
      "epoch": 1.932523616734143,
      "grad_norm": 0.7976219658784348,
      "learning_rate": 3.0795484072016066e-05,
      "loss": 0.3383,
      "step": 358
    },
    {
      "epoch": 1.9379217273954117,
      "grad_norm": 0.8829544955168173,
      "learning_rate": 3.073183458012906e-05,
      "loss": 0.4246,
      "step": 359
    },
    {
      "epoch": 1.9433198380566803,
      "grad_norm": 0.7999199853428338,
      "learning_rate": 3.0668032075870394e-05,
      "loss": 0.3316,
      "step": 360
    },
    {
      "epoch": 1.9487179487179487,
      "grad_norm": 0.7754126352865167,
      "learning_rate": 3.060407746892354e-05,
      "loss": 0.3532,
      "step": 361
    },
    {
      "epoch": 1.9541160593792173,
      "grad_norm": 0.8116757582096946,
      "learning_rate": 3.053997167114062e-05,
      "loss": 0.3592,
      "step": 362
    },
    {
      "epoch": 1.9595141700404857,
      "grad_norm": 0.780747731868162,
      "learning_rate": 3.0475715596529402e-05,
      "loss": 0.3831,
      "step": 363
    },
    {
      "epoch": 1.9649122807017543,
      "grad_norm": 0.7784545333341477,
      "learning_rate": 3.041131016124026e-05,
      "loss": 0.3469,
      "step": 364
    },
    {
      "epoch": 1.9703103913630229,
      "grad_norm": 0.780104268381351,
      "learning_rate": 3.0346756283553138e-05,
      "loss": 0.3173,
      "step": 365
    },
    {
      "epoch": 1.9757085020242915,
      "grad_norm": 0.868243693709775,
      "learning_rate": 3.0282054883864434e-05,
      "loss": 0.4914,
      "step": 366
    },
    {
      "epoch": 1.98110661268556,
      "grad_norm": 0.8092251365777564,
      "learning_rate": 3.0217206884673898e-05,
      "loss": 0.4763,
      "step": 367
    },
    {
      "epoch": 1.9865047233468287,
      "grad_norm": 0.9194927920738091,
      "learning_rate": 3.015221321057145e-05,
      "loss": 0.4446,
      "step": 368
    },
    {
      "epoch": 1.9919028340080973,
      "grad_norm": 0.8487891788474984,
      "learning_rate": 3.0087074788224016e-05,
      "loss": 0.5507,
      "step": 369
    },
    {
      "epoch": 1.9973009446693657,
      "grad_norm": 0.8108830919109473,
      "learning_rate": 3.0021792546362332e-05,
      "loss": 0.3367,
      "step": 370
    },
    {
      "epoch": 2.002699055330634,
      "grad_norm": 1.8729539106932147,
      "learning_rate": 2.9956367415767657e-05,
      "loss": 0.4987,
      "step": 371
    },
    {
      "epoch": 2.0080971659919027,
      "grad_norm": 0.7462995058419852,
      "learning_rate": 2.9890800329258554e-05,
      "loss": 0.1457,
      "step": 372
    },
    {
      "epoch": 2.0134952766531713,
      "grad_norm": 0.6392949061952808,
      "learning_rate": 2.982509222167755e-05,
      "loss": 0.1257,
      "step": 373
    },
    {
      "epoch": 2.01889338731444,
      "grad_norm": 0.7022587732356804,
      "learning_rate": 2.975924402987783e-05,
      "loss": 0.1274,
      "step": 374
    },
    {
      "epoch": 2.0242914979757085,
      "grad_norm": 0.77290317342361,
      "learning_rate": 2.969325669270987e-05,
      "loss": 0.1244,
      "step": 375
    },
    {
      "epoch": 2.029689608636977,
      "grad_norm": 0.7674422966027143,
      "learning_rate": 2.9627131151008046e-05,
      "loss": 0.1215,
      "step": 376
    },
    {
      "epoch": 2.0350877192982457,
      "grad_norm": 0.819257096752639,
      "learning_rate": 2.9560868347577235e-05,
      "loss": 0.1414,
      "step": 377
    },
    {
      "epoch": 2.0404858299595143,
      "grad_norm": 0.7484653317083918,
      "learning_rate": 2.9494469227179375e-05,
      "loss": 0.2481,
      "step": 378
    },
    {
      "epoch": 2.045883940620783,
      "grad_norm": 0.6741305314820067,
      "learning_rate": 2.9427934736519962e-05,
      "loss": 0.1336,
      "step": 379
    },
    {
      "epoch": 2.051282051282051,
      "grad_norm": 0.6164943712448056,
      "learning_rate": 2.9361265824234575e-05,
      "loss": 0.1222,
      "step": 380
    },
    {
      "epoch": 2.0566801619433197,
      "grad_norm": 0.6852123731952208,
      "learning_rate": 2.9294463440875375e-05,
      "loss": 0.1449,
      "step": 381
    },
    {
      "epoch": 2.0620782726045883,
      "grad_norm": 0.6244781256553339,
      "learning_rate": 2.9227528538897503e-05,
      "loss": 0.1723,
      "step": 382
    },
    {
      "epoch": 2.067476383265857,
      "grad_norm": 0.7978024679197341,
      "learning_rate": 2.916046207264554e-05,
      "loss": 0.1326,
      "step": 383
    },
    {
      "epoch": 2.0728744939271255,
      "grad_norm": 0.6147819567342936,
      "learning_rate": 2.9093264998339875e-05,
      "loss": 0.1294,
      "step": 384
    },
    {
      "epoch": 2.078272604588394,
      "grad_norm": 0.6787811592156624,
      "learning_rate": 2.9025938274063077e-05,
      "loss": 0.1497,
      "step": 385
    },
    {
      "epoch": 2.0836707152496627,
      "grad_norm": 0.7317408188133651,
      "learning_rate": 2.8958482859746242e-05,
      "loss": 0.1887,
      "step": 386
    },
    {
      "epoch": 2.0890688259109313,
      "grad_norm": 0.6723746847129357,
      "learning_rate": 2.889089971715532e-05,
      "loss": 0.139,
      "step": 387
    },
    {
      "epoch": 2.0944669365722,
      "grad_norm": 0.6818331924468926,
      "learning_rate": 2.8823189809877365e-05,
      "loss": 0.1342,
      "step": 388
    },
    {
      "epoch": 2.099865047233468,
      "grad_norm": 0.5980820718094894,
      "learning_rate": 2.8755354103306808e-05,
      "loss": 0.2167,
      "step": 389
    },
    {
      "epoch": 2.1052631578947367,
      "grad_norm": 0.6648551106554124,
      "learning_rate": 2.868739356463174e-05,
      "loss": 0.1224,
      "step": 390
    },
    {
      "epoch": 2.1106612685560053,
      "grad_norm": 0.7486161553352669,
      "learning_rate": 2.8619309162820048e-05,
      "loss": 0.1267,
      "step": 391
    },
    {
      "epoch": 2.116059379217274,
      "grad_norm": 0.6867674315095388,
      "learning_rate": 2.8551101868605644e-05,
      "loss": 0.2035,
      "step": 392
    },
    {
      "epoch": 2.1214574898785425,
      "grad_norm": 0.6878121126213758,
      "learning_rate": 2.8482772654474624e-05,
      "loss": 0.2658,
      "step": 393
    },
    {
      "epoch": 2.126855600539811,
      "grad_norm": 0.6678383543442121,
      "learning_rate": 2.841432249465138e-05,
      "loss": 0.1294,
      "step": 394
    },
    {
      "epoch": 2.1322537112010798,
      "grad_norm": 0.6372763697558534,
      "learning_rate": 2.8345752365084744e-05,
      "loss": 0.131,
      "step": 395
    },
    {
      "epoch": 2.1376518218623484,
      "grad_norm": 0.6334966824967888,
      "learning_rate": 2.8277063243434036e-05,
      "loss": 0.1268,
      "step": 396
    },
    {
      "epoch": 2.1430499325236165,
      "grad_norm": 0.6404580338280741,
      "learning_rate": 2.820825610905514e-05,
      "loss": 0.1153,
      "step": 397
    },
    {
      "epoch": 2.148448043184885,
      "grad_norm": 0.6466950437091068,
      "learning_rate": 2.8139331942986552e-05,
      "loss": 0.1215,
      "step": 398
    },
    {
      "epoch": 2.1538461538461537,
      "grad_norm": 0.6714487489560431,
      "learning_rate": 2.8070291727935385e-05,
      "loss": 0.1218,
      "step": 399
    },
    {
      "epoch": 2.1592442645074224,
      "grad_norm": 0.6006007696911629,
      "learning_rate": 2.8001136448263345e-05,
      "loss": 0.1183,
      "step": 400
    },
    {
      "epoch": 2.164642375168691,
      "grad_norm": 0.6386900747836864,
      "learning_rate": 2.7931867089972703e-05,
      "loss": 0.3012,
      "step": 401
    },
    {
      "epoch": 2.1700404858299596,
      "grad_norm": 0.6241422911928288,
      "learning_rate": 2.7862484640692265e-05,
      "loss": 0.1267,
      "step": 402
    },
    {
      "epoch": 2.175438596491228,
      "grad_norm": 0.698965066265789,
      "learning_rate": 2.7792990089663244e-05,
      "loss": 0.2367,
      "step": 403
    },
    {
      "epoch": 2.180836707152497,
      "grad_norm": 0.6577790993693683,
      "learning_rate": 2.772338442772517e-05,
      "loss": 0.1247,
      "step": 404
    },
    {
      "epoch": 2.1862348178137654,
      "grad_norm": 0.7057930356119323,
      "learning_rate": 2.7653668647301797e-05,
      "loss": 0.1416,
      "step": 405
    },
    {
      "epoch": 2.1916329284750335,
      "grad_norm": 0.6283432520769865,
      "learning_rate": 2.758384374238691e-05,
      "loss": 0.1171,
      "step": 406
    },
    {
      "epoch": 2.197031039136302,
      "grad_norm": 0.6758905973282627,
      "learning_rate": 2.751391070853017e-05,
      "loss": 0.1277,
      "step": 407
    },
    {
      "epoch": 2.2024291497975708,
      "grad_norm": 0.6936284196657481,
      "learning_rate": 2.744387054282293e-05,
      "loss": 0.1423,
      "step": 408
    },
    {
      "epoch": 2.2078272604588394,
      "grad_norm": 0.6341950745676999,
      "learning_rate": 2.737372424388398e-05,
      "loss": 0.2242,
      "step": 409
    },
    {
      "epoch": 2.213225371120108,
      "grad_norm": 0.6693004983980072,
      "learning_rate": 2.7303472811845373e-05,
      "loss": 0.1232,
      "step": 410
    },
    {
      "epoch": 2.2186234817813766,
      "grad_norm": 0.6629406429715882,
      "learning_rate": 2.72331172483381e-05,
      "loss": 0.1658,
      "step": 411
    },
    {
      "epoch": 2.224021592442645,
      "grad_norm": 0.6127381303453924,
      "learning_rate": 2.7162658556477856e-05,
      "loss": 0.1371,
      "step": 412
    },
    {
      "epoch": 2.229419703103914,
      "grad_norm": 0.679698173745626,
      "learning_rate": 2.7092097740850712e-05,
      "loss": 0.1316,
      "step": 413
    },
    {
      "epoch": 2.234817813765182,
      "grad_norm": 0.5950073990369081,
      "learning_rate": 2.7021435807498803e-05,
      "loss": 0.1207,
      "step": 414
    },
    {
      "epoch": 2.2402159244264506,
      "grad_norm": 0.6677093728594548,
      "learning_rate": 2.6950673763905976e-05,
      "loss": 0.1327,
      "step": 415
    },
    {
      "epoch": 2.245614035087719,
      "grad_norm": 0.9128432310383863,
      "learning_rate": 2.6879812618983427e-05,
      "loss": 0.1654,
      "step": 416
    },
    {
      "epoch": 2.251012145748988,
      "grad_norm": 0.6476391269524641,
      "learning_rate": 2.680885338305532e-05,
      "loss": 0.1275,
      "step": 417
    },
    {
      "epoch": 2.2564102564102564,
      "grad_norm": 0.5686092680345144,
      "learning_rate": 2.6737797067844403e-05,
      "loss": 0.1103,
      "step": 418
    },
    {
      "epoch": 2.261808367071525,
      "grad_norm": 0.7170335546600183,
      "learning_rate": 2.6666644686457527e-05,
      "loss": 0.2701,
      "step": 419
    },
    {
      "epoch": 2.2672064777327936,
      "grad_norm": 0.6208498042067502,
      "learning_rate": 2.6595397253371263e-05,
      "loss": 0.1189,
      "step": 420
    },
    {
      "epoch": 2.272604588394062,
      "grad_norm": 0.6398812216990059,
      "learning_rate": 2.652405578441739e-05,
      "loss": 0.1232,
      "step": 421
    },
    {
      "epoch": 2.278002699055331,
      "grad_norm": 0.7138464146301546,
      "learning_rate": 2.6452621296768444e-05,
      "loss": 0.1549,
      "step": 422
    },
    {
      "epoch": 2.283400809716599,
      "grad_norm": 0.6592603480537171,
      "learning_rate": 2.63810948089232e-05,
      "loss": 0.1369,
      "step": 423
    },
    {
      "epoch": 2.2887989203778676,
      "grad_norm": 0.6546216583647926,
      "learning_rate": 2.630947734069216e-05,
      "loss": 0.1276,
      "step": 424
    },
    {
      "epoch": 2.294197031039136,
      "grad_norm": 0.5970453793673439,
      "learning_rate": 2.623776991318298e-05,
      "loss": 0.1096,
      "step": 425
    },
    {
      "epoch": 2.299595141700405,
      "grad_norm": 0.6022630941533801,
      "learning_rate": 2.6165973548785985e-05,
      "loss": 0.1358,
      "step": 426
    },
    {
      "epoch": 2.3049932523616734,
      "grad_norm": 0.5754057954915315,
      "learning_rate": 2.6094089271159497e-05,
      "loss": 0.1076,
      "step": 427
    },
    {
      "epoch": 2.310391363022942,
      "grad_norm": 0.6455725563135731,
      "learning_rate": 2.6022118105215314e-05,
      "loss": 0.1219,
      "step": 428
    },
    {
      "epoch": 2.3157894736842106,
      "grad_norm": 0.6533947853819206,
      "learning_rate": 2.595006107710406e-05,
      "loss": 0.1349,
      "step": 429
    },
    {
      "epoch": 2.3211875843454792,
      "grad_norm": 0.5646029755765858,
      "learning_rate": 2.5877919214200575e-05,
      "loss": 0.1105,
      "step": 430
    },
    {
      "epoch": 2.326585695006748,
      "grad_norm": 0.6305035499092596,
      "learning_rate": 2.580569354508925e-05,
      "loss": 0.1233,
      "step": 431
    },
    {
      "epoch": 2.331983805668016,
      "grad_norm": 0.7303006034253096,
      "learning_rate": 2.5733385099549365e-05,
      "loss": 0.1488,
      "step": 432
    },
    {
      "epoch": 2.3373819163292846,
      "grad_norm": 0.6328447736616827,
      "learning_rate": 2.566099490854041e-05,
      "loss": 0.1229,
      "step": 433
    },
    {
      "epoch": 2.342780026990553,
      "grad_norm": 0.8566015327759697,
      "learning_rate": 2.5588524004187395e-05,
      "loss": 0.1509,
      "step": 434
    },
    {
      "epoch": 2.348178137651822,
      "grad_norm": 0.6624968711715733,
      "learning_rate": 2.5515973419766117e-05,
      "loss": 0.3453,
      "step": 435
    },
    {
      "epoch": 2.3535762483130904,
      "grad_norm": 0.6945361115581712,
      "learning_rate": 2.5443344189688442e-05,
      "loss": 0.2186,
      "step": 436
    },
    {
      "epoch": 2.358974358974359,
      "grad_norm": 0.6241503828356365,
      "learning_rate": 2.5370637349487537e-05,
      "loss": 0.1316,
      "step": 437
    },
    {
      "epoch": 2.3643724696356276,
      "grad_norm": 0.6349270911136394,
      "learning_rate": 2.5297853935803134e-05,
      "loss": 0.1438,
      "step": 438
    },
    {
      "epoch": 2.3697705802968962,
      "grad_norm": 0.6264630421591376,
      "learning_rate": 2.522499498636673e-05,
      "loss": 0.1287,
      "step": 439
    },
    {
      "epoch": 2.375168690958165,
      "grad_norm": 0.6494294150925718,
      "learning_rate": 2.5152061539986786e-05,
      "loss": 0.1237,
      "step": 440
    },
    {
      "epoch": 2.380566801619433,
      "grad_norm": 0.6590711569125003,
      "learning_rate": 2.5079054636533943e-05,
      "loss": 0.1477,
      "step": 441
    },
    {
      "epoch": 2.3859649122807016,
      "grad_norm": 0.6017175013652537,
      "learning_rate": 2.5005975316926155e-05,
      "loss": 0.1102,
      "step": 442
    },
    {
      "epoch": 2.3913630229419702,
      "grad_norm": 0.6640068999885245,
      "learning_rate": 2.4932824623113904e-05,
      "loss": 0.169,
      "step": 443
    },
    {
      "epoch": 2.396761133603239,
      "grad_norm": 0.5766339469065881,
      "learning_rate": 2.485960359806528e-05,
      "loss": 0.2012,
      "step": 444
    },
    {
      "epoch": 2.4021592442645074,
      "grad_norm": 0.7291306101333911,
      "learning_rate": 2.4786313285751158e-05,
      "loss": 0.1579,
      "step": 445
    },
    {
      "epoch": 2.407557354925776,
      "grad_norm": 0.6380383823229943,
      "learning_rate": 2.4712954731130295e-05,
      "loss": 0.1661,
      "step": 446
    },
    {
      "epoch": 2.4129554655870447,
      "grad_norm": 0.6659006653830499,
      "learning_rate": 2.4639528980134424e-05,
      "loss": 0.128,
      "step": 447
    },
    {
      "epoch": 2.4183535762483133,
      "grad_norm": 0.6099007883870388,
      "learning_rate": 2.4566037079653372e-05,
      "loss": 0.1083,
      "step": 448
    },
    {
      "epoch": 2.423751686909582,
      "grad_norm": 0.613692855683778,
      "learning_rate": 2.4492480077520084e-05,
      "loss": 0.1223,
      "step": 449
    },
    {
      "epoch": 2.42914979757085,
      "grad_norm": 0.6715299948015367,
      "learning_rate": 2.441885902249573e-05,
      "loss": 0.1248,
      "step": 450
    },
    {
      "epoch": 2.4345479082321186,
      "grad_norm": 0.610328390420646,
      "learning_rate": 2.434517496425474e-05,
      "loss": 0.1181,
      "step": 451
    },
    {
      "epoch": 2.4399460188933872,
      "grad_norm": 0.6479329176924008,
      "learning_rate": 2.4271428953369805e-05,
      "loss": 0.3185,
      "step": 452
    },
    {
      "epoch": 2.445344129554656,
      "grad_norm": 0.700649740705671,
      "learning_rate": 2.419762204129695e-05,
      "loss": 0.138,
      "step": 453
    },
    {
      "epoch": 2.4507422402159245,
      "grad_norm": 0.5760283063402104,
      "learning_rate": 2.412375528036051e-05,
      "loss": 0.1096,
      "step": 454
    },
    {
      "epoch": 2.456140350877193,
      "grad_norm": 0.6796450613672332,
      "learning_rate": 2.4049829723738127e-05,
      "loss": 0.1323,
      "step": 455
    },
    {
      "epoch": 2.4615384615384617,
      "grad_norm": 0.5519373684819371,
      "learning_rate": 2.3975846425445745e-05,
      "loss": 0.104,
      "step": 456
    },
    {
      "epoch": 2.4669365721997303,
      "grad_norm": 0.5941230543506764,
      "learning_rate": 2.390180644032257e-05,
      "loss": 0.1159,
      "step": 457
    },
    {
      "epoch": 2.472334682860999,
      "grad_norm": 0.6005019267580625,
      "learning_rate": 2.382771082401605e-05,
      "loss": 0.1127,
      "step": 458
    },
    {
      "epoch": 2.477732793522267,
      "grad_norm": 0.6231210450690562,
      "learning_rate": 2.3753560632966803e-05,
      "loss": 0.2765,
      "step": 459
    },
    {
      "epoch": 2.4831309041835357,
      "grad_norm": 0.6761993902876848,
      "learning_rate": 2.367935692439357e-05,
      "loss": 0.1311,
      "step": 460
    },
    {
      "epoch": 2.4885290148448043,
      "grad_norm": 0.6296854883290703,
      "learning_rate": 2.360510075627812e-05,
      "loss": 0.1246,
      "step": 461
    },
    {
      "epoch": 2.493927125506073,
      "grad_norm": 0.631923430747307,
      "learning_rate": 2.3530793187350196e-05,
      "loss": 0.1194,
      "step": 462
    },
    {
      "epoch": 2.4993252361673415,
      "grad_norm": 0.5962340880765427,
      "learning_rate": 2.3456435277072392e-05,
      "loss": 0.1114,
      "step": 463
    },
    {
      "epoch": 2.50472334682861,
      "grad_norm": 0.611293805000405,
      "learning_rate": 2.3382028085625074e-05,
      "loss": 0.1192,
      "step": 464
    },
    {
      "epoch": 2.5101214574898787,
      "grad_norm": 0.6017071362047015,
      "learning_rate": 2.3307572673891227e-05,
      "loss": 0.1186,
      "step": 465
    },
    {
      "epoch": 2.515519568151147,
      "grad_norm": 0.6608480467314211,
      "learning_rate": 2.3233070103441372e-05,
      "loss": 0.1269,
      "step": 466
    },
    {
      "epoch": 2.520917678812416,
      "grad_norm": 0.6286315055894832,
      "learning_rate": 2.3158521436518395e-05,
      "loss": 0.1439,
      "step": 467
    },
    {
      "epoch": 2.526315789473684,
      "grad_norm": 0.5941211446808222,
      "learning_rate": 2.3083927736022423e-05,
      "loss": 0.1163,
      "step": 468
    },
    {
      "epoch": 2.5317139001349527,
      "grad_norm": 0.6014018923806088,
      "learning_rate": 2.3009290065495663e-05,
      "loss": 0.1165,
      "step": 469
    },
    {
      "epoch": 2.5371120107962213,
      "grad_norm": 0.6364853987180407,
      "learning_rate": 2.2934609489107236e-05,
      "loss": 0.1201,
      "step": 470
    },
    {
      "epoch": 2.54251012145749,
      "grad_norm": 0.6044054475261765,
      "learning_rate": 2.285988707163801e-05,
      "loss": 0.1091,
      "step": 471
    },
    {
      "epoch": 2.5479082321187585,
      "grad_norm": 0.6087359767421359,
      "learning_rate": 2.2785123878465414e-05,
      "loss": 0.1148,
      "step": 472
    },
    {
      "epoch": 2.553306342780027,
      "grad_norm": 0.69235536929551,
      "learning_rate": 2.2710320975548238e-05,
      "loss": 0.1299,
      "step": 473
    },
    {
      "epoch": 2.5587044534412957,
      "grad_norm": 0.6088709234303951,
      "learning_rate": 2.2635479429411463e-05,
      "loss": 0.1171,
      "step": 474
    },
    {
      "epoch": 2.564102564102564,
      "grad_norm": 0.6626427945597948,
      "learning_rate": 2.256060030713102e-05,
      "loss": 0.1271,
      "step": 475
    },
    {
      "epoch": 2.569500674763833,
      "grad_norm": 0.6307597674316254,
      "learning_rate": 2.248568467631862e-05,
      "loss": 0.2658,
      "step": 476
    },
    {
      "epoch": 2.574898785425101,
      "grad_norm": 0.6601911176181631,
      "learning_rate": 2.2410733605106462e-05,
      "loss": 0.1292,
      "step": 477
    },
    {
      "epoch": 2.5802968960863697,
      "grad_norm": 0.7416111139812673,
      "learning_rate": 2.2335748162132093e-05,
      "loss": 0.146,
      "step": 478
    },
    {
      "epoch": 2.5856950067476383,
      "grad_norm": 0.589360245246151,
      "learning_rate": 2.2260729416523103e-05,
      "loss": 0.1111,
      "step": 479
    },
    {
      "epoch": 2.591093117408907,
      "grad_norm": 0.654143642296817,
      "learning_rate": 2.2185678437881898e-05,
      "loss": 0.1198,
      "step": 480
    },
    {
      "epoch": 2.5964912280701755,
      "grad_norm": 0.6608861416350591,
      "learning_rate": 2.2110596296270472e-05,
      "loss": 0.12,
      "step": 481
    },
    {
      "epoch": 2.601889338731444,
      "grad_norm": 0.6681912905025659,
      "learning_rate": 2.2035484062195124e-05,
      "loss": 0.1309,
      "step": 482
    },
    {
      "epoch": 2.6072874493927127,
      "grad_norm": 0.631378882096793,
      "learning_rate": 2.196034280659122e-05,
      "loss": 0.1181,
      "step": 483
    },
    {
      "epoch": 2.612685560053981,
      "grad_norm": 0.6527132040126822,
      "learning_rate": 2.188517360080788e-05,
      "loss": 0.1394,
      "step": 484
    },
    {
      "epoch": 2.6180836707152495,
      "grad_norm": 0.6227805441719897,
      "learning_rate": 2.180997751659276e-05,
      "loss": 0.1195,
      "step": 485
    },
    {
      "epoch": 2.623481781376518,
      "grad_norm": 0.6802847699912662,
      "learning_rate": 2.1734755626076733e-05,
      "loss": 0.1395,
      "step": 486
    },
    {
      "epoch": 2.6288798920377867,
      "grad_norm": 0.6888202624358717,
      "learning_rate": 2.1659509001758616e-05,
      "loss": 0.1325,
      "step": 487
    },
    {
      "epoch": 2.6342780026990553,
      "grad_norm": 0.593385764404308,
      "learning_rate": 2.158423871648988e-05,
      "loss": 0.1113,
      "step": 488
    },
    {
      "epoch": 2.639676113360324,
      "grad_norm": 0.6774506983126284,
      "learning_rate": 2.1508945843459322e-05,
      "loss": 0.1804,
      "step": 489
    },
    {
      "epoch": 2.6450742240215925,
      "grad_norm": 0.6654559325765059,
      "learning_rate": 2.1433631456177823e-05,
      "loss": 0.307,
      "step": 490
    },
    {
      "epoch": 2.650472334682861,
      "grad_norm": 0.5979358796010572,
      "learning_rate": 2.1358296628463008e-05,
      "loss": 0.1167,
      "step": 491
    },
    {
      "epoch": 2.6558704453441297,
      "grad_norm": 0.7087596924028456,
      "learning_rate": 2.1282942434423917e-05,
      "loss": 0.152,
      "step": 492
    },
    {
      "epoch": 2.661268556005398,
      "grad_norm": 0.5914031045078398,
      "learning_rate": 2.1207569948445724e-05,
      "loss": 0.2739,
      "step": 493
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.6479960919806412,
      "learning_rate": 2.11321802451744e-05,
      "loss": 0.1252,
      "step": 494
    },
    {
      "epoch": 2.672064777327935,
      "grad_norm": 0.6359038955310647,
      "learning_rate": 2.1056774399501413e-05,
      "loss": 0.1181,
      "step": 495
    },
    {
      "epoch": 2.6774628879892037,
      "grad_norm": 0.5845117483373671,
      "learning_rate": 2.0981353486548363e-05,
      "loss": 0.1045,
      "step": 496
    },
    {
      "epoch": 2.6828609986504723,
      "grad_norm": 0.6312095850180834,
      "learning_rate": 2.0905918581651692e-05,
      "loss": 0.1247,
      "step": 497
    },
    {
      "epoch": 2.688259109311741,
      "grad_norm": 0.7064036700451566,
      "learning_rate": 2.0830470760347326e-05,
      "loss": 0.1468,
      "step": 498
    },
    {
      "epoch": 2.6936572199730096,
      "grad_norm": 0.6425598356622161,
      "learning_rate": 2.0755011098355366e-05,
      "loss": 0.1269,
      "step": 499
    },
    {
      "epoch": 2.699055330634278,
      "grad_norm": 0.6089176429064123,
      "learning_rate": 2.067954067156472e-05,
      "loss": 0.1221,
      "step": 500
    },
    {
      "epoch": 2.7044534412955468,
      "grad_norm": 0.6709598340951207,
      "learning_rate": 2.060406055601778e-05,
      "loss": 0.1408,
      "step": 501
    },
    {
      "epoch": 2.709851551956815,
      "grad_norm": 0.6038478330611535,
      "learning_rate": 2.052857182789509e-05,
      "loss": 0.1069,
      "step": 502
    },
    {
      "epoch": 2.7152496626180835,
      "grad_norm": 0.6385055235648402,
      "learning_rate": 2.0453075563499975e-05,
      "loss": 0.114,
      "step": 503
    },
    {
      "epoch": 2.720647773279352,
      "grad_norm": 0.6961725778048614,
      "learning_rate": 2.0377572839243225e-05,
      "loss": 0.2043,
      "step": 504
    },
    {
      "epoch": 2.7260458839406208,
      "grad_norm": 0.5770500121000621,
      "learning_rate": 2.030206473162772e-05,
      "loss": 0.2564,
      "step": 505
    },
    {
      "epoch": 2.7314439946018894,
      "grad_norm": 0.5369215916183971,
      "learning_rate": 2.022655231723311e-05,
      "loss": 0.2308,
      "step": 506
    },
    {
      "epoch": 2.736842105263158,
      "grad_norm": 0.6775333454012917,
      "learning_rate": 2.0151036672700437e-05,
      "loss": 0.1325,
      "step": 507
    },
    {
      "epoch": 2.7422402159244266,
      "grad_norm": 0.627884242138156,
      "learning_rate": 2.0075518874716797e-05,
      "loss": 0.1085,
      "step": 508
    },
    {
      "epoch": 2.7476383265856947,
      "grad_norm": 0.7044141840918825,
      "learning_rate": 2e-05,
      "loss": 0.124,
      "step": 509
    },
    {
      "epoch": 2.753036437246964,
      "grad_norm": 0.6810850215625832,
      "learning_rate": 1.9924481125283203e-05,
      "loss": 0.5196,
      "step": 510
    },
    {
      "epoch": 2.758434547908232,
      "grad_norm": 0.6985941640179243,
      "learning_rate": 1.984896332729957e-05,
      "loss": 0.1378,
      "step": 511
    },
    {
      "epoch": 2.7638326585695006,
      "grad_norm": 0.6208939763658017,
      "learning_rate": 1.9773447682766894e-05,
      "loss": 0.1199,
      "step": 512
    },
    {
      "epoch": 2.769230769230769,
      "grad_norm": 0.6960421491040145,
      "learning_rate": 1.9697935268372285e-05,
      "loss": 0.1435,
      "step": 513
    },
    {
      "epoch": 2.7746288798920378,
      "grad_norm": 0.6479313802993825,
      "learning_rate": 1.9622427160756778e-05,
      "loss": 0.1189,
      "step": 514
    },
    {
      "epoch": 2.7800269905533064,
      "grad_norm": 0.6257589956418593,
      "learning_rate": 1.9546924436500028e-05,
      "loss": 0.1192,
      "step": 515
    },
    {
      "epoch": 2.785425101214575,
      "grad_norm": 0.612853830087127,
      "learning_rate": 1.9471428172104916e-05,
      "loss": 0.1295,
      "step": 516
    },
    {
      "epoch": 2.7908232118758436,
      "grad_norm": 0.6341457718175875,
      "learning_rate": 1.9395939443982228e-05,
      "loss": 0.2511,
      "step": 517
    },
    {
      "epoch": 2.7962213225371118,
      "grad_norm": 0.6179614260256036,
      "learning_rate": 1.9320459328435287e-05,
      "loss": 0.1408,
      "step": 518
    },
    {
      "epoch": 2.801619433198381,
      "grad_norm": 0.6465167698062426,
      "learning_rate": 1.924498890164464e-05,
      "loss": 0.2563,
      "step": 519
    },
    {
      "epoch": 2.807017543859649,
      "grad_norm": 0.6556794701046659,
      "learning_rate": 1.9169529239652678e-05,
      "loss": 0.2109,
      "step": 520
    },
    {
      "epoch": 2.8124156545209176,
      "grad_norm": 0.5822517488639858,
      "learning_rate": 1.9094081418348318e-05,
      "loss": 0.1149,
      "step": 521
    },
    {
      "epoch": 2.817813765182186,
      "grad_norm": 0.6311322525747073,
      "learning_rate": 1.901864651345164e-05,
      "loss": 0.1308,
      "step": 522
    },
    {
      "epoch": 2.823211875843455,
      "grad_norm": 0.5898219875972766,
      "learning_rate": 1.894322560049859e-05,
      "loss": 0.1253,
      "step": 523
    },
    {
      "epoch": 2.8286099865047234,
      "grad_norm": 0.5936183602331494,
      "learning_rate": 1.8867819754825602e-05,
      "loss": 0.1254,
      "step": 524
    },
    {
      "epoch": 2.834008097165992,
      "grad_norm": 0.6117641560994483,
      "learning_rate": 1.879243005155428e-05,
      "loss": 0.1232,
      "step": 525
    },
    {
      "epoch": 2.8394062078272606,
      "grad_norm": 0.6500880335763644,
      "learning_rate": 1.871705756557609e-05,
      "loss": 0.2541,
      "step": 526
    },
    {
      "epoch": 2.8448043184885288,
      "grad_norm": 0.6058813838626976,
      "learning_rate": 1.8641703371536995e-05,
      "loss": 0.1253,
      "step": 527
    },
    {
      "epoch": 2.850202429149798,
      "grad_norm": 0.6680465461222095,
      "learning_rate": 1.856636854382218e-05,
      "loss": 0.2805,
      "step": 528
    },
    {
      "epoch": 2.855600539811066,
      "grad_norm": 0.6134739260909959,
      "learning_rate": 1.849105415654068e-05,
      "loss": 0.1168,
      "step": 529
    },
    {
      "epoch": 2.8609986504723346,
      "grad_norm": 0.6617449814395124,
      "learning_rate": 1.841576128351013e-05,
      "loss": 0.1461,
      "step": 530
    },
    {
      "epoch": 2.866396761133603,
      "grad_norm": 0.6389843047318983,
      "learning_rate": 1.8340490998241387e-05,
      "loss": 0.128,
      "step": 531
    },
    {
      "epoch": 2.871794871794872,
      "grad_norm": 0.5738756222417001,
      "learning_rate": 1.8265244373923274e-05,
      "loss": 0.1077,
      "step": 532
    },
    {
      "epoch": 2.8771929824561404,
      "grad_norm": 0.5919186419431655,
      "learning_rate": 1.8190022483407246e-05,
      "loss": 0.1072,
      "step": 533
    },
    {
      "epoch": 2.882591093117409,
      "grad_norm": 0.683056671095403,
      "learning_rate": 1.8114826399192133e-05,
      "loss": 0.1212,
      "step": 534
    },
    {
      "epoch": 2.8879892037786776,
      "grad_norm": 0.6129763131639642,
      "learning_rate": 1.8039657193408788e-05,
      "loss": 0.1171,
      "step": 535
    },
    {
      "epoch": 2.893387314439946,
      "grad_norm": 0.5627763381998306,
      "learning_rate": 1.7964515937804875e-05,
      "loss": 0.1069,
      "step": 536
    },
    {
      "epoch": 2.898785425101215,
      "grad_norm": 0.602936257523276,
      "learning_rate": 1.7889403703729535e-05,
      "loss": 0.1075,
      "step": 537
    },
    {
      "epoch": 2.904183535762483,
      "grad_norm": 0.6225600024019384,
      "learning_rate": 1.7814321562118105e-05,
      "loss": 0.1192,
      "step": 538
    },
    {
      "epoch": 2.9095816464237516,
      "grad_norm": 0.6232613214009397,
      "learning_rate": 1.7739270583476904e-05,
      "loss": 0.1129,
      "step": 539
    },
    {
      "epoch": 2.91497975708502,
      "grad_norm": 0.7106160834112466,
      "learning_rate": 1.766425183786791e-05,
      "loss": 0.1325,
      "step": 540
    },
    {
      "epoch": 2.920377867746289,
      "grad_norm": 0.6215307665098612,
      "learning_rate": 1.758926639489354e-05,
      "loss": 0.1152,
      "step": 541
    },
    {
      "epoch": 2.9257759784075574,
      "grad_norm": 0.6493142964042798,
      "learning_rate": 1.7514315323681386e-05,
      "loss": 0.1351,
      "step": 542
    },
    {
      "epoch": 2.931174089068826,
      "grad_norm": 0.5928462030642684,
      "learning_rate": 1.7439399692868985e-05,
      "loss": 0.1127,
      "step": 543
    },
    {
      "epoch": 2.9365721997300946,
      "grad_norm": 0.7261667290131814,
      "learning_rate": 1.7364520570588543e-05,
      "loss": 0.1588,
      "step": 544
    },
    {
      "epoch": 2.941970310391363,
      "grad_norm": 0.630870393655033,
      "learning_rate": 1.7289679024451772e-05,
      "loss": 0.1088,
      "step": 545
    },
    {
      "epoch": 2.9473684210526314,
      "grad_norm": 0.6335205983688524,
      "learning_rate": 1.7214876121534592e-05,
      "loss": 0.0999,
      "step": 546
    },
    {
      "epoch": 2.9527665317139,
      "grad_norm": 0.6865001571083343,
      "learning_rate": 1.7140112928361996e-05,
      "loss": 0.1231,
      "step": 547
    },
    {
      "epoch": 2.9581646423751686,
      "grad_norm": 0.6326899562878843,
      "learning_rate": 1.7065390510892767e-05,
      "loss": 0.115,
      "step": 548
    },
    {
      "epoch": 2.9635627530364372,
      "grad_norm": 0.6095222445393947,
      "learning_rate": 1.699070993450434e-05,
      "loss": 0.1022,
      "step": 549
    },
    {
      "epoch": 2.968960863697706,
      "grad_norm": 0.671598010718977,
      "learning_rate": 1.6916072263977583e-05,
      "loss": 0.1718,
      "step": 550
    },
    {
      "epoch": 2.9743589743589745,
      "grad_norm": 0.5878250512418469,
      "learning_rate": 1.6841478563481612e-05,
      "loss": 0.0996,
      "step": 551
    },
    {
      "epoch": 2.979757085020243,
      "grad_norm": 0.7503089305640249,
      "learning_rate": 1.6766929896558638e-05,
      "loss": 0.2841,
      "step": 552
    },
    {
      "epoch": 2.9851551956815117,
      "grad_norm": 0.6989279699107632,
      "learning_rate": 1.6692427326108776e-05,
      "loss": 0.1517,
      "step": 553
    },
    {
      "epoch": 2.99055330634278,
      "grad_norm": 0.6049149193808451,
      "learning_rate": 1.6617971914374933e-05,
      "loss": 0.2593,
      "step": 554
    },
    {
      "epoch": 2.9959514170040484,
      "grad_norm": 0.6071511813101513,
      "learning_rate": 1.654356472292761e-05,
      "loss": 0.119,
      "step": 555
    },
    {
      "epoch": 3.001349527665317,
      "grad_norm": 2.191278573121276,
      "learning_rate": 1.6469206812649818e-05,
      "loss": 0.2184,
      "step": 556
    },
    {
      "epoch": 3.0067476383265856,
      "grad_norm": 0.3436912254691727,
      "learning_rate": 1.6394899243721887e-05,
      "loss": 0.0354,
      "step": 557
    },
    {
      "epoch": 3.0121457489878543,
      "grad_norm": 0.34881471094295025,
      "learning_rate": 1.632064307560644e-05,
      "loss": 0.0332,
      "step": 558
    },
    {
      "epoch": 3.017543859649123,
      "grad_norm": 0.3692217535663024,
      "learning_rate": 1.6246439367033207e-05,
      "loss": 0.0408,
      "step": 559
    },
    {
      "epoch": 3.0229419703103915,
      "grad_norm": 0.3280449705906347,
      "learning_rate": 1.617228917598396e-05,
      "loss": 0.0319,
      "step": 560
    },
    {
      "epoch": 3.02834008097166,
      "grad_norm": 0.33260821933202167,
      "learning_rate": 1.609819355967744e-05,
      "loss": 0.0337,
      "step": 561
    },
    {
      "epoch": 3.0337381916329287,
      "grad_norm": 0.405541501347011,
      "learning_rate": 1.602415357455426e-05,
      "loss": 0.0535,
      "step": 562
    },
    {
      "epoch": 3.039136302294197,
      "grad_norm": 0.47557751817634936,
      "learning_rate": 1.5950170276261876e-05,
      "loss": 0.0402,
      "step": 563
    },
    {
      "epoch": 3.0445344129554655,
      "grad_norm": 0.42113145547242353,
      "learning_rate": 1.587624471963949e-05,
      "loss": 0.0368,
      "step": 564
    },
    {
      "epoch": 3.049932523616734,
      "grad_norm": 0.4587759167556131,
      "learning_rate": 1.5802377958703054e-05,
      "loss": 0.0396,
      "step": 565
    },
    {
      "epoch": 3.0553306342780027,
      "grad_norm": 0.4093864304314231,
      "learning_rate": 1.5728571046630195e-05,
      "loss": 0.0328,
      "step": 566
    },
    {
      "epoch": 3.0607287449392713,
      "grad_norm": 0.5852288088536234,
      "learning_rate": 1.5654825035745268e-05,
      "loss": 0.0692,
      "step": 567
    },
    {
      "epoch": 3.06612685560054,
      "grad_norm": 0.4076610339068381,
      "learning_rate": 1.5581140977504273e-05,
      "loss": 0.0347,
      "step": 568
    },
    {
      "epoch": 3.0715249662618085,
      "grad_norm": 0.48848968423165967,
      "learning_rate": 1.550751992247993e-05,
      "loss": 0.036,
      "step": 569
    },
    {
      "epoch": 3.076923076923077,
      "grad_norm": 0.47155692827014656,
      "learning_rate": 1.5433962920346638e-05,
      "loss": 0.036,
      "step": 570
    },
    {
      "epoch": 3.0823211875843457,
      "grad_norm": 0.380639334036154,
      "learning_rate": 1.5360471019865583e-05,
      "loss": 0.0296,
      "step": 571
    },
    {
      "epoch": 3.087719298245614,
      "grad_norm": 0.43221784870918734,
      "learning_rate": 1.5287045268869712e-05,
      "loss": 0.0351,
      "step": 572
    },
    {
      "epoch": 3.0931174089068825,
      "grad_norm": 0.49905359599899657,
      "learning_rate": 1.5213686714248852e-05,
      "loss": 0.039,
      "step": 573
    },
    {
      "epoch": 3.098515519568151,
      "grad_norm": 0.41331595216842526,
      "learning_rate": 1.5140396401934725e-05,
      "loss": 0.0309,
      "step": 574
    },
    {
      "epoch": 3.1039136302294197,
      "grad_norm": 0.6043807596540545,
      "learning_rate": 1.5067175376886099e-05,
      "loss": 0.0274,
      "step": 575
    },
    {
      "epoch": 3.1093117408906883,
      "grad_norm": 0.5014002726682666,
      "learning_rate": 1.4994024683073848e-05,
      "loss": 0.0458,
      "step": 576
    },
    {
      "epoch": 3.114709851551957,
      "grad_norm": 0.40613279392800844,
      "learning_rate": 1.4920945363466062e-05,
      "loss": 0.0324,
      "step": 577
    },
    {
      "epoch": 3.1201079622132255,
      "grad_norm": 0.4886303787102555,
      "learning_rate": 1.4847938460013219e-05,
      "loss": 0.2548,
      "step": 578
    },
    {
      "epoch": 3.125506072874494,
      "grad_norm": 0.3825952135908576,
      "learning_rate": 1.4775005013633277e-05,
      "loss": 0.0319,
      "step": 579
    },
    {
      "epoch": 3.1309041835357623,
      "grad_norm": 0.3654848769695381,
      "learning_rate": 1.4702146064196875e-05,
      "loss": 0.0351,
      "step": 580
    },
    {
      "epoch": 3.136302294197031,
      "grad_norm": 0.38630959354240707,
      "learning_rate": 1.4629362650512464e-05,
      "loss": 0.0403,
      "step": 581
    },
    {
      "epoch": 3.1417004048582995,
      "grad_norm": 0.4225028608467292,
      "learning_rate": 1.4556655810311566e-05,
      "loss": 0.1626,
      "step": 582
    },
    {
      "epoch": 3.147098515519568,
      "grad_norm": 0.3827994232164537,
      "learning_rate": 1.4484026580233888e-05,
      "loss": 0.0344,
      "step": 583
    },
    {
      "epoch": 3.1524966261808367,
      "grad_norm": 0.3310555630534496,
      "learning_rate": 1.4411475995812613e-05,
      "loss": 0.0275,
      "step": 584
    },
    {
      "epoch": 3.1578947368421053,
      "grad_norm": 0.3632971332190463,
      "learning_rate": 1.4339005091459595e-05,
      "loss": 0.0839,
      "step": 585
    },
    {
      "epoch": 3.163292847503374,
      "grad_norm": 0.3602691812823302,
      "learning_rate": 1.4266614900450645e-05,
      "loss": 0.0291,
      "step": 586
    },
    {
      "epoch": 3.1686909581646425,
      "grad_norm": 0.40629829870227013,
      "learning_rate": 1.4194306454910757e-05,
      "loss": 0.0425,
      "step": 587
    },
    {
      "epoch": 3.174089068825911,
      "grad_norm": 0.3521781479995114,
      "learning_rate": 1.4122080785799423e-05,
      "loss": 0.0311,
      "step": 588
    },
    {
      "epoch": 3.1794871794871793,
      "grad_norm": 0.5300855991321622,
      "learning_rate": 1.4049938922895945e-05,
      "loss": 0.1332,
      "step": 589
    },
    {
      "epoch": 3.184885290148448,
      "grad_norm": 0.47747106483553625,
      "learning_rate": 1.3977881894784689e-05,
      "loss": 0.1839,
      "step": 590
    },
    {
      "epoch": 3.1902834008097165,
      "grad_norm": 0.5073414782013317,
      "learning_rate": 1.390591072884051e-05,
      "loss": 0.0332,
      "step": 591
    },
    {
      "epoch": 3.195681511470985,
      "grad_norm": 0.36797349854225664,
      "learning_rate": 1.383402645121402e-05,
      "loss": 0.0297,
      "step": 592
    },
    {
      "epoch": 3.2010796221322537,
      "grad_norm": 0.39930020559853535,
      "learning_rate": 1.3762230086817024e-05,
      "loss": 0.0311,
      "step": 593
    },
    {
      "epoch": 3.2064777327935223,
      "grad_norm": 0.41828193288392684,
      "learning_rate": 1.3690522659307846e-05,
      "loss": 0.1682,
      "step": 594
    },
    {
      "epoch": 3.211875843454791,
      "grad_norm": 0.4254248832162874,
      "learning_rate": 1.3618905191076806e-05,
      "loss": 0.0431,
      "step": 595
    },
    {
      "epoch": 3.2172739541160595,
      "grad_norm": 0.38086605557209824,
      "learning_rate": 1.3547378703231559e-05,
      "loss": 0.0275,
      "step": 596
    },
    {
      "epoch": 3.2226720647773277,
      "grad_norm": 0.4245476891916035,
      "learning_rate": 1.3475944215582619e-05,
      "loss": 0.0391,
      "step": 597
    },
    {
      "epoch": 3.2280701754385963,
      "grad_norm": 0.43518229539768466,
      "learning_rate": 1.340460274662874e-05,
      "loss": 0.0378,
      "step": 598
    },
    {
      "epoch": 3.233468286099865,
      "grad_norm": 0.3780375093099488,
      "learning_rate": 1.3333355313542478e-05,
      "loss": 0.0302,
      "step": 599
    },
    {
      "epoch": 3.2388663967611335,
      "grad_norm": 0.3629192307455684,
      "learning_rate": 1.3262202932155602e-05,
      "loss": 0.0344,
      "step": 600
    },
    {
      "epoch": 3.244264507422402,
      "grad_norm": 0.3699894264323716,
      "learning_rate": 1.3191146616944676e-05,
      "loss": 0.0289,
      "step": 601
    },
    {
      "epoch": 3.2496626180836707,
      "grad_norm": 0.35188609379499725,
      "learning_rate": 1.312018738101658e-05,
      "loss": 0.0339,
      "step": 602
    },
    {
      "epoch": 3.2550607287449393,
      "grad_norm": 0.4491270054593502,
      "learning_rate": 1.304932623609403e-05,
      "loss": 0.1386,
      "step": 603
    },
    {
      "epoch": 3.260458839406208,
      "grad_norm": 0.38745106173266436,
      "learning_rate": 1.2978564192501203e-05,
      "loss": 0.0402,
      "step": 604
    },
    {
      "epoch": 3.2658569500674766,
      "grad_norm": 0.3773073501230217,
      "learning_rate": 1.2907902259149287e-05,
      "loss": 0.0308,
      "step": 605
    },
    {
      "epoch": 3.2712550607287447,
      "grad_norm": 0.39839198674044035,
      "learning_rate": 1.2837341443522147e-05,
      "loss": 0.031,
      "step": 606
    },
    {
      "epoch": 3.2766531713900133,
      "grad_norm": 0.38676509058198444,
      "learning_rate": 1.2766882751661905e-05,
      "loss": 0.034,
      "step": 607
    },
    {
      "epoch": 3.282051282051282,
      "grad_norm": 0.37175755499128865,
      "learning_rate": 1.2696527188154639e-05,
      "loss": 0.0319,
      "step": 608
    },
    {
      "epoch": 3.2874493927125505,
      "grad_norm": 0.364975396498793,
      "learning_rate": 1.2626275756116027e-05,
      "loss": 0.0291,
      "step": 609
    },
    {
      "epoch": 3.292847503373819,
      "grad_norm": 0.4410768716884598,
      "learning_rate": 1.2556129457177084e-05,
      "loss": 0.0351,
      "step": 610
    },
    {
      "epoch": 3.2982456140350878,
      "grad_norm": 0.3686080508993296,
      "learning_rate": 1.2486089291469835e-05,
      "loss": 0.0304,
      "step": 611
    },
    {
      "epoch": 3.3036437246963564,
      "grad_norm": 0.3825335449989966,
      "learning_rate": 1.2416156257613098e-05,
      "loss": 0.0367,
      "step": 612
    },
    {
      "epoch": 3.309041835357625,
      "grad_norm": 0.34218361939907355,
      "learning_rate": 1.2346331352698206e-05,
      "loss": 0.0315,
      "step": 613
    },
    {
      "epoch": 3.3144399460188936,
      "grad_norm": 0.4478836143104432,
      "learning_rate": 1.227661557227483e-05,
      "loss": 0.0427,
      "step": 614
    },
    {
      "epoch": 3.3198380566801617,
      "grad_norm": 0.37407337511538385,
      "learning_rate": 1.2207009910336764e-05,
      "loss": 0.0327,
      "step": 615
    },
    {
      "epoch": 3.3252361673414303,
      "grad_norm": 0.35934207428825143,
      "learning_rate": 1.2137515359307737e-05,
      "loss": 0.0312,
      "step": 616
    },
    {
      "epoch": 3.330634278002699,
      "grad_norm": 0.43937992207581134,
      "learning_rate": 1.2068132910027299e-05,
      "loss": 0.0354,
      "step": 617
    },
    {
      "epoch": 3.3360323886639676,
      "grad_norm": 0.33294700463710736,
      "learning_rate": 1.1998863551736659e-05,
      "loss": 0.0283,
      "step": 618
    },
    {
      "epoch": 3.341430499325236,
      "grad_norm": 0.39301352026371505,
      "learning_rate": 1.192970827206462e-05,
      "loss": 0.0331,
      "step": 619
    },
    {
      "epoch": 3.3468286099865048,
      "grad_norm": 0.36498619637281343,
      "learning_rate": 1.1860668057013451e-05,
      "loss": 0.0328,
      "step": 620
    },
    {
      "epoch": 3.3522267206477734,
      "grad_norm": 0.3559476591751902,
      "learning_rate": 1.1791743890944869e-05,
      "loss": 0.0343,
      "step": 621
    },
    {
      "epoch": 3.357624831309042,
      "grad_norm": 0.43684774915670094,
      "learning_rate": 1.1722936756565969e-05,
      "loss": 0.2334,
      "step": 622
    },
    {
      "epoch": 3.3630229419703106,
      "grad_norm": 0.4400139001073612,
      "learning_rate": 1.1654247634915261e-05,
      "loss": 0.0323,
      "step": 623
    },
    {
      "epoch": 3.3684210526315788,
      "grad_norm": 0.3993508002078833,
      "learning_rate": 1.1585677505348618e-05,
      "loss": 0.1723,
      "step": 624
    },
    {
      "epoch": 3.3738191632928474,
      "grad_norm": 0.39545396795136123,
      "learning_rate": 1.151722734552538e-05,
      "loss": 0.136,
      "step": 625
    },
    {
      "epoch": 3.379217273954116,
      "grad_norm": 0.3784737302500327,
      "learning_rate": 1.1448898131394364e-05,
      "loss": 0.0385,
      "step": 626
    },
    {
      "epoch": 3.3846153846153846,
      "grad_norm": 0.3633996380704606,
      "learning_rate": 1.1380690837179955e-05,
      "loss": 0.0297,
      "step": 627
    },
    {
      "epoch": 3.390013495276653,
      "grad_norm": 0.38600677874076367,
      "learning_rate": 1.1312606435368266e-05,
      "loss": 0.0292,
      "step": 628
    },
    {
      "epoch": 3.395411605937922,
      "grad_norm": 0.3685107386955438,
      "learning_rate": 1.124464589669319e-05,
      "loss": 0.0283,
      "step": 629
    },
    {
      "epoch": 3.4008097165991904,
      "grad_norm": 0.3796999501062114,
      "learning_rate": 1.1176810190122644e-05,
      "loss": 0.0345,
      "step": 630
    },
    {
      "epoch": 3.406207827260459,
      "grad_norm": 0.39313860478281704,
      "learning_rate": 1.110910028284468e-05,
      "loss": 0.0315,
      "step": 631
    },
    {
      "epoch": 3.4116059379217276,
      "grad_norm": 0.38729735419763556,
      "learning_rate": 1.1041517140253761e-05,
      "loss": 0.0538,
      "step": 632
    },
    {
      "epoch": 3.417004048582996,
      "grad_norm": 0.3796093100412225,
      "learning_rate": 1.0974061725936935e-05,
      "loss": 0.0305,
      "step": 633
    },
    {
      "epoch": 3.4224021592442644,
      "grad_norm": 0.3654235584300773,
      "learning_rate": 1.0906735001660138e-05,
      "loss": 0.0283,
      "step": 634
    },
    {
      "epoch": 3.427800269905533,
      "grad_norm": 0.409937460602826,
      "learning_rate": 1.0839537927354466e-05,
      "loss": 0.0379,
      "step": 635
    },
    {
      "epoch": 3.4331983805668016,
      "grad_norm": 0.3917028036928748,
      "learning_rate": 1.0772471461102505e-05,
      "loss": 0.0269,
      "step": 636
    },
    {
      "epoch": 3.43859649122807,
      "grad_norm": 0.3625369529682563,
      "learning_rate": 1.070553655912463e-05,
      "loss": 0.0283,
      "step": 637
    },
    {
      "epoch": 3.443994601889339,
      "grad_norm": 0.38788057237068674,
      "learning_rate": 1.0638734175765433e-05,
      "loss": 0.0314,
      "step": 638
    },
    {
      "epoch": 3.4493927125506074,
      "grad_norm": 0.3847399681910376,
      "learning_rate": 1.0572065263480046e-05,
      "loss": 0.0324,
      "step": 639
    },
    {
      "epoch": 3.454790823211876,
      "grad_norm": 0.4382934253068173,
      "learning_rate": 1.0505530772820625e-05,
      "loss": 0.0894,
      "step": 640
    },
    {
      "epoch": 3.4601889338731446,
      "grad_norm": 0.419058031320498,
      "learning_rate": 1.0439131652422763e-05,
      "loss": 0.2127,
      "step": 641
    },
    {
      "epoch": 3.465587044534413,
      "grad_norm": 0.44649997683190956,
      "learning_rate": 1.037286884899196e-05,
      "loss": 0.1479,
      "step": 642
    },
    {
      "epoch": 3.4709851551956814,
      "grad_norm": 0.40480754300934557,
      "learning_rate": 1.030674330729014e-05,
      "loss": 0.0352,
      "step": 643
    },
    {
      "epoch": 3.47638326585695,
      "grad_norm": 0.30967102679189645,
      "learning_rate": 1.0240755970122173e-05,
      "loss": 0.0218,
      "step": 644
    },
    {
      "epoch": 3.4817813765182186,
      "grad_norm": 0.37838596103828964,
      "learning_rate": 1.0174907778322458e-05,
      "loss": 0.0288,
      "step": 645
    },
    {
      "epoch": 3.4871794871794872,
      "grad_norm": 0.4146619345330915,
      "learning_rate": 1.0109199670741447e-05,
      "loss": 0.0362,
      "step": 646
    },
    {
      "epoch": 3.492577597840756,
      "grad_norm": 0.4117765119433817,
      "learning_rate": 1.004363258423235e-05,
      "loss": 0.2213,
      "step": 647
    },
    {
      "epoch": 3.4979757085020244,
      "grad_norm": 0.34377868810679235,
      "learning_rate": 9.978207453637671e-06,
      "loss": 0.0269,
      "step": 648
    },
    {
      "epoch": 3.5033738191632926,
      "grad_norm": 0.38450416903251206,
      "learning_rate": 9.912925211775989e-06,
      "loss": 0.0284,
      "step": 649
    },
    {
      "epoch": 3.5087719298245617,
      "grad_norm": 0.3681687532508216,
      "learning_rate": 9.84778678942856e-06,
      "loss": 0.0319,
      "step": 650
    },
    {
      "epoch": 3.51417004048583,
      "grad_norm": 0.3964437963905117,
      "learning_rate": 9.782793115326112e-06,
      "loss": 0.0347,
      "step": 651
    },
    {
      "epoch": 3.5195681511470984,
      "grad_norm": 0.3655533421389497,
      "learning_rate": 9.717945116135568e-06,
      "loss": 0.0333,
      "step": 652
    },
    {
      "epoch": 3.524966261808367,
      "grad_norm": 0.3224001401289986,
      "learning_rate": 9.653243716446862e-06,
      "loss": 0.0276,
      "step": 653
    },
    {
      "epoch": 3.5303643724696356,
      "grad_norm": 0.3452946278683242,
      "learning_rate": 9.588689838759744e-06,
      "loss": 0.0281,
      "step": 654
    },
    {
      "epoch": 3.5357624831309042,
      "grad_norm": 0.36042207805510473,
      "learning_rate": 9.524284403470598e-06,
      "loss": 0.0299,
      "step": 655
    },
    {
      "epoch": 3.541160593792173,
      "grad_norm": 0.38055305365429554,
      "learning_rate": 9.460028328859379e-06,
      "loss": 0.0252,
      "step": 656
    },
    {
      "epoch": 3.5465587044534415,
      "grad_norm": 0.3771448822244628,
      "learning_rate": 9.395922531076462e-06,
      "loss": 0.0331,
      "step": 657
    },
    {
      "epoch": 3.5519568151147096,
      "grad_norm": 0.3462359501161187,
      "learning_rate": 9.331967924129615e-06,
      "loss": 0.0264,
      "step": 658
    },
    {
      "epoch": 3.5573549257759787,
      "grad_norm": 0.39941647689494786,
      "learning_rate": 9.268165419870947e-06,
      "loss": 0.0333,
      "step": 659
    },
    {
      "epoch": 3.562753036437247,
      "grad_norm": 0.4205128793287871,
      "learning_rate": 9.20451592798394e-06,
      "loss": 0.0307,
      "step": 660
    },
    {
      "epoch": 3.5681511470985154,
      "grad_norm": 0.42077068867710476,
      "learning_rate": 9.141020355970427e-06,
      "loss": 0.0325,
      "step": 661
    },
    {
      "epoch": 3.573549257759784,
      "grad_norm": 0.3622122750826694,
      "learning_rate": 9.077679609137714e-06,
      "loss": 0.0285,
      "step": 662
    },
    {
      "epoch": 3.5789473684210527,
      "grad_norm": 0.4360493505456192,
      "learning_rate": 9.014494590585603e-06,
      "loss": 0.0327,
      "step": 663
    },
    {
      "epoch": 3.5843454790823213,
      "grad_norm": 0.40930157708444925,
      "learning_rate": 8.95146620119359e-06,
      "loss": 0.1576,
      "step": 664
    },
    {
      "epoch": 3.58974358974359,
      "grad_norm": 0.38811598863254687,
      "learning_rate": 8.888595339607961e-06,
      "loss": 0.0302,
      "step": 665
    },
    {
      "epoch": 3.5951417004048585,
      "grad_norm": 0.36493375147907436,
      "learning_rate": 8.825882902228998e-06,
      "loss": 0.0342,
      "step": 666
    },
    {
      "epoch": 3.6005398110661266,
      "grad_norm": 0.3459341178116053,
      "learning_rate": 8.763329783198222e-06,
      "loss": 0.0306,
      "step": 667
    },
    {
      "epoch": 3.6059379217273952,
      "grad_norm": 0.39037878669889553,
      "learning_rate": 8.700936874385601e-06,
      "loss": 0.0316,
      "step": 668
    },
    {
      "epoch": 3.611336032388664,
      "grad_norm": 0.39416938896362874,
      "learning_rate": 8.638705065376887e-06,
      "loss": 0.0307,
      "step": 669
    },
    {
      "epoch": 3.6167341430499325,
      "grad_norm": 0.34015438708553125,
      "learning_rate": 8.576635243460868e-06,
      "loss": 0.0257,
      "step": 670
    },
    {
      "epoch": 3.622132253711201,
      "grad_norm": 0.3899592691796045,
      "learning_rate": 8.514728293616788e-06,
      "loss": 0.0295,
      "step": 671
    },
    {
      "epoch": 3.6275303643724697,
      "grad_norm": 0.43563748789771956,
      "learning_rate": 8.452985098501659e-06,
      "loss": 0.0586,
      "step": 672
    },
    {
      "epoch": 3.6329284750337383,
      "grad_norm": 0.3664811599053362,
      "learning_rate": 8.391406538437737e-06,
      "loss": 0.027,
      "step": 673
    },
    {
      "epoch": 3.638326585695007,
      "grad_norm": 0.3616556776179159,
      "learning_rate": 8.329993491399933e-06,
      "loss": 0.0317,
      "step": 674
    },
    {
      "epoch": 3.6437246963562755,
      "grad_norm": 0.3948186866229346,
      "learning_rate": 8.268746833003304e-06,
      "loss": 0.0287,
      "step": 675
    },
    {
      "epoch": 3.6491228070175437,
      "grad_norm": 0.4064305102650153,
      "learning_rate": 8.207667436490564e-06,
      "loss": 0.0318,
      "step": 676
    },
    {
      "epoch": 3.6545209176788123,
      "grad_norm": 0.3887421174083186,
      "learning_rate": 8.146756172719668e-06,
      "loss": 0.0337,
      "step": 677
    },
    {
      "epoch": 3.659919028340081,
      "grad_norm": 0.35086318209010237,
      "learning_rate": 8.086013910151334e-06,
      "loss": 0.0303,
      "step": 678
    },
    {
      "epoch": 3.6653171390013495,
      "grad_norm": 0.3875957934712957,
      "learning_rate": 8.025441514836715e-06,
      "loss": 0.0309,
      "step": 679
    },
    {
      "epoch": 3.670715249662618,
      "grad_norm": 0.36661840834858717,
      "learning_rate": 7.965039850405037e-06,
      "loss": 0.027,
      "step": 680
    },
    {
      "epoch": 3.6761133603238867,
      "grad_norm": 0.34118151803342256,
      "learning_rate": 7.904809778051252e-06,
      "loss": 0.0296,
      "step": 681
    },
    {
      "epoch": 3.6815114709851553,
      "grad_norm": 0.37213000529906315,
      "learning_rate": 7.84475215652381e-06,
      "loss": 0.1303,
      "step": 682
    },
    {
      "epoch": 3.686909581646424,
      "grad_norm": 0.3896509711111651,
      "learning_rate": 7.784867842112367e-06,
      "loss": 0.1687,
      "step": 683
    },
    {
      "epoch": 3.6923076923076925,
      "grad_norm": 0.3426839606336355,
      "learning_rate": 7.725157688635634e-06,
      "loss": 0.0312,
      "step": 684
    },
    {
      "epoch": 3.6977058029689607,
      "grad_norm": 0.35402053366618513,
      "learning_rate": 7.665622547429139e-06,
      "loss": 0.0293,
      "step": 685
    },
    {
      "epoch": 3.7031039136302293,
      "grad_norm": 0.3546158448227337,
      "learning_rate": 7.606263267333145e-06,
      "loss": 0.0249,
      "step": 686
    },
    {
      "epoch": 3.708502024291498,
      "grad_norm": 0.33157529931614793,
      "learning_rate": 7.547080694680495e-06,
      "loss": 0.0219,
      "step": 687
    },
    {
      "epoch": 3.7139001349527665,
      "grad_norm": 0.39512925735292526,
      "learning_rate": 7.4880756732846e-06,
      "loss": 0.0356,
      "step": 688
    },
    {
      "epoch": 3.719298245614035,
      "grad_norm": 0.3714005261676109,
      "learning_rate": 7.429249044427362e-06,
      "loss": 0.0352,
      "step": 689
    },
    {
      "epoch": 3.7246963562753037,
      "grad_norm": 0.3613957102307198,
      "learning_rate": 7.3706016468472045e-06,
      "loss": 0.0306,
      "step": 690
    },
    {
      "epoch": 3.7300944669365723,
      "grad_norm": 0.31816258283900495,
      "learning_rate": 7.312134316727093e-06,
      "loss": 0.0266,
      "step": 691
    },
    {
      "epoch": 3.7354925775978405,
      "grad_norm": 0.41925403759376656,
      "learning_rate": 7.25384788768265e-06,
      "loss": 0.1886,
      "step": 692
    },
    {
      "epoch": 3.7408906882591095,
      "grad_norm": 0.4045239613978412,
      "learning_rate": 7.195743190750241e-06,
      "loss": 0.0321,
      "step": 693
    },
    {
      "epoch": 3.7462887989203777,
      "grad_norm": 0.38175252409850824,
      "learning_rate": 7.137821054375114e-06,
      "loss": 0.0622,
      "step": 694
    },
    {
      "epoch": 3.7516869095816463,
      "grad_norm": 0.34465054751860885,
      "learning_rate": 7.080082304399625e-06,
      "loss": 0.0271,
      "step": 695
    },
    {
      "epoch": 3.757085020242915,
      "grad_norm": 0.33989881366203534,
      "learning_rate": 7.022527764051423e-06,
      "loss": 0.0286,
      "step": 696
    },
    {
      "epoch": 3.7624831309041835,
      "grad_norm": 0.3412805918296631,
      "learning_rate": 6.9651582539317565e-06,
      "loss": 0.027,
      "step": 697
    },
    {
      "epoch": 3.767881241565452,
      "grad_norm": 0.3334336573527621,
      "learning_rate": 6.907974592003732e-06,
      "loss": 0.0265,
      "step": 698
    },
    {
      "epoch": 3.7732793522267207,
      "grad_norm": 0.4270420508919924,
      "learning_rate": 6.850977593580675e-06,
      "loss": 0.034,
      "step": 699
    },
    {
      "epoch": 3.7786774628879893,
      "grad_norm": 0.3297327960112233,
      "learning_rate": 6.794168071314486e-06,
      "loss": 0.0238,
      "step": 700
    },
    {
      "epoch": 3.7840755735492575,
      "grad_norm": 0.36163451391560975,
      "learning_rate": 6.737546835184101e-06,
      "loss": 0.0279,
      "step": 701
    },
    {
      "epoch": 3.7894736842105265,
      "grad_norm": 0.4411471545722498,
      "learning_rate": 6.6811146924838725e-06,
      "loss": 0.0666,
      "step": 702
    },
    {
      "epoch": 3.7948717948717947,
      "grad_norm": 0.35871421378374524,
      "learning_rate": 6.6248724478121254e-06,
      "loss": 0.0291,
      "step": 703
    },
    {
      "epoch": 3.8002699055330633,
      "grad_norm": 0.36756705518749777,
      "learning_rate": 6.568820903059632e-06,
      "loss": 0.0367,
      "step": 704
    },
    {
      "epoch": 3.805668016194332,
      "grad_norm": 0.47277824056660106,
      "learning_rate": 6.512960857398227e-06,
      "loss": 0.0392,
      "step": 705
    },
    {
      "epoch": 3.8110661268556005,
      "grad_norm": 0.4010852296701039,
      "learning_rate": 6.457293107269371e-06,
      "loss": 0.0308,
      "step": 706
    },
    {
      "epoch": 3.816464237516869,
      "grad_norm": 0.3680019544802103,
      "learning_rate": 6.401818446372809e-06,
      "loss": 0.0489,
      "step": 707
    },
    {
      "epoch": 3.8218623481781377,
      "grad_norm": 0.3820591256150461,
      "learning_rate": 6.346537665655286e-06,
      "loss": 0.1109,
      "step": 708
    },
    {
      "epoch": 3.8272604588394064,
      "grad_norm": 0.29674819102746314,
      "learning_rate": 6.291451553299204e-06,
      "loss": 0.0221,
      "step": 709
    },
    {
      "epoch": 3.8326585695006745,
      "grad_norm": 0.38107534212823674,
      "learning_rate": 6.236560894711459e-06,
      "loss": 0.0317,
      "step": 710
    },
    {
      "epoch": 3.8380566801619436,
      "grad_norm": 0.4300573413232902,
      "learning_rate": 6.181866472512175e-06,
      "loss": 0.1339,
      "step": 711
    },
    {
      "epoch": 3.8434547908232117,
      "grad_norm": 0.4653463809534711,
      "learning_rate": 6.127369066523599e-06,
      "loss": 0.3112,
      "step": 712
    },
    {
      "epoch": 3.8488529014844803,
      "grad_norm": 0.3828688005837094,
      "learning_rate": 6.073069453758946e-06,
      "loss": 0.0259,
      "step": 713
    },
    {
      "epoch": 3.854251012145749,
      "grad_norm": 0.35459961651464167,
      "learning_rate": 6.018968408411341e-06,
      "loss": 0.0267,
      "step": 714
    },
    {
      "epoch": 3.8596491228070176,
      "grad_norm": 0.35643103172233276,
      "learning_rate": 5.965066701842766e-06,
      "loss": 0.1439,
      "step": 715
    },
    {
      "epoch": 3.865047233468286,
      "grad_norm": 0.34409226399099263,
      "learning_rate": 5.9113651025730835e-06,
      "loss": 0.0292,
      "step": 716
    },
    {
      "epoch": 3.8704453441295548,
      "grad_norm": 0.35933383547027126,
      "learning_rate": 5.857864376269051e-06,
      "loss": 0.0804,
      "step": 717
    },
    {
      "epoch": 3.8758434547908234,
      "grad_norm": 0.3739973480129211,
      "learning_rate": 5.804565285733432e-06,
      "loss": 0.035,
      "step": 718
    },
    {
      "epoch": 3.8812415654520915,
      "grad_norm": 0.3722765784519681,
      "learning_rate": 5.751468590894107e-06,
      "loss": 0.1128,
      "step": 719
    },
    {
      "epoch": 3.8866396761133606,
      "grad_norm": 0.3713792988536375,
      "learning_rate": 5.698575048793223e-06,
      "loss": 0.0277,
      "step": 720
    },
    {
      "epoch": 3.8920377867746287,
      "grad_norm": 0.4568426442621641,
      "learning_rate": 5.645885413576433e-06,
      "loss": 0.1226,
      "step": 721
    },
    {
      "epoch": 3.8974358974358974,
      "grad_norm": 0.37950864411226903,
      "learning_rate": 5.593400436482119e-06,
      "loss": 0.043,
      "step": 722
    },
    {
      "epoch": 3.902834008097166,
      "grad_norm": 0.3925818680423842,
      "learning_rate": 5.541120865830687e-06,
      "loss": 0.0307,
      "step": 723
    },
    {
      "epoch": 3.9082321187584346,
      "grad_norm": 0.36496534398830294,
      "learning_rate": 5.489047447013891e-06,
      "loss": 0.0312,
      "step": 724
    },
    {
      "epoch": 3.913630229419703,
      "grad_norm": 0.3411659648761055,
      "learning_rate": 5.4371809224842354e-06,
      "loss": 0.0249,
      "step": 725
    },
    {
      "epoch": 3.919028340080972,
      "grad_norm": 0.35092445194976446,
      "learning_rate": 5.3855220317443416e-06,
      "loss": 0.0274,
      "step": 726
    },
    {
      "epoch": 3.9244264507422404,
      "grad_norm": 0.3489498272274411,
      "learning_rate": 5.334071511336449e-06,
      "loss": 0.0252,
      "step": 727
    },
    {
      "epoch": 3.9298245614035086,
      "grad_norm": 0.36114128102871357,
      "learning_rate": 5.282830094831881e-06,
      "loss": 0.0294,
      "step": 728
    },
    {
      "epoch": 3.9352226720647776,
      "grad_norm": 0.3284749937819038,
      "learning_rate": 5.231798512820612e-06,
      "loss": 0.025,
      "step": 729
    },
    {
      "epoch": 3.9406207827260458,
      "grad_norm": 0.386218354934746,
      "learning_rate": 5.180977492900823e-06,
      "loss": 0.1581,
      "step": 730
    },
    {
      "epoch": 3.9460188933873144,
      "grad_norm": 0.3924605685249013,
      "learning_rate": 5.130367759668544e-06,
      "loss": 0.1108,
      "step": 731
    },
    {
      "epoch": 3.951417004048583,
      "grad_norm": 0.35950299892176835,
      "learning_rate": 5.07997003470734e-06,
      "loss": 0.0269,
      "step": 732
    },
    {
      "epoch": 3.9568151147098516,
      "grad_norm": 0.35019384697385253,
      "learning_rate": 5.029785036577976e-06,
      "loss": 0.0262,
      "step": 733
    },
    {
      "epoch": 3.96221322537112,
      "grad_norm": 0.3432029539479147,
      "learning_rate": 4.979813480808231e-06,
      "loss": 0.0254,
      "step": 734
    },
    {
      "epoch": 3.967611336032389,
      "grad_norm": 0.35144365016547774,
      "learning_rate": 4.930056079882632e-06,
      "loss": 0.0298,
      "step": 735
    },
    {
      "epoch": 3.9730094466936574,
      "grad_norm": 0.3541877009200062,
      "learning_rate": 4.880513543232361e-06,
      "loss": 0.0267,
      "step": 736
    },
    {
      "epoch": 3.9784075573549256,
      "grad_norm": 0.40028293095466,
      "learning_rate": 4.831186577225082e-06,
      "loss": 0.0359,
      "step": 737
    },
    {
      "epoch": 3.983805668016194,
      "grad_norm": 0.40302496611767596,
      "learning_rate": 4.782075885154909e-06,
      "loss": 0.0301,
      "step": 738
    },
    {
      "epoch": 3.989203778677463,
      "grad_norm": 0.4237077479286715,
      "learning_rate": 4.733182167232356e-06,
      "loss": 0.0346,
      "step": 739
    },
    {
      "epoch": 3.9946018893387314,
      "grad_norm": 0.3547714604476633,
      "learning_rate": 4.684506120574375e-06,
      "loss": 0.0258,
      "step": 740
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.102540355528973,
      "learning_rate": 4.636048439194392e-06,
      "loss": 0.0685,
      "step": 741
    },
    {
      "epoch": 4.005398110661268,
      "grad_norm": 0.15325932809564183,
      "learning_rate": 4.587809813992437e-06,
      "loss": 0.0121,
      "step": 742
    },
    {
      "epoch": 4.010796221322537,
      "grad_norm": 0.13530023614474354,
      "learning_rate": 4.53979093274526e-06,
      "loss": 0.008,
      "step": 743
    },
    {
      "epoch": 4.016194331983805,
      "grad_norm": 0.13502016529933805,
      "learning_rate": 4.4919924800965696e-06,
      "loss": 0.0087,
      "step": 744
    },
    {
      "epoch": 4.021592442645074,
      "grad_norm": 0.18254596131179956,
      "learning_rate": 4.444415137547238e-06,
      "loss": 0.0103,
      "step": 745
    },
    {
      "epoch": 4.026990553306343,
      "grad_norm": 0.17029406387858442,
      "learning_rate": 4.3970595834455845e-06,
      "loss": 0.0089,
      "step": 746
    },
    {
      "epoch": 4.032388663967612,
      "grad_norm": 0.2522421603518247,
      "learning_rate": 4.349926492977719e-06,
      "loss": 0.0828,
      "step": 747
    },
    {
      "epoch": 4.03778677462888,
      "grad_norm": 0.14277344736187908,
      "learning_rate": 4.303016538157907e-06,
      "loss": 0.0287,
      "step": 748
    },
    {
      "epoch": 4.043184885290149,
      "grad_norm": 0.2642123078469134,
      "learning_rate": 4.256330387818999e-06,
      "loss": 0.1838,
      "step": 749
    },
    {
      "epoch": 4.048582995951417,
      "grad_norm": 0.271440865327848,
      "learning_rate": 4.209868707602871e-06,
      "loss": 0.1105,
      "step": 750
    },
    {
      "epoch": 4.053981106612685,
      "grad_norm": 0.14472235533108238,
      "learning_rate": 4.163632159950965e-06,
      "loss": 0.0073,
      "step": 751
    },
    {
      "epoch": 4.059379217273954,
      "grad_norm": 0.20035179292259456,
      "learning_rate": 4.117621404094811e-06,
      "loss": 0.0085,
      "step": 752
    },
    {
      "epoch": 4.064777327935222,
      "grad_norm": 0.14362507519164897,
      "learning_rate": 4.07183709604666e-06,
      "loss": 0.008,
      "step": 753
    },
    {
      "epoch": 4.0701754385964914,
      "grad_norm": 0.15431708802792676,
      "learning_rate": 4.026279888590101e-06,
      "loss": 0.0092,
      "step": 754
    },
    {
      "epoch": 4.07557354925776,
      "grad_norm": 0.13999009793334247,
      "learning_rate": 3.980950431270776e-06,
      "loss": 0.0077,
      "step": 755
    },
    {
      "epoch": 4.080971659919029,
      "grad_norm": 0.15652426490564023,
      "learning_rate": 3.935849370387104e-06,
      "loss": 0.0079,
      "step": 756
    },
    {
      "epoch": 4.086369770580297,
      "grad_norm": 0.14077430214547057,
      "learning_rate": 3.89097734898108e-06,
      "loss": 0.0073,
      "step": 757
    },
    {
      "epoch": 4.091767881241566,
      "grad_norm": 0.149289451771641,
      "learning_rate": 3.846335006829103e-06,
      "loss": 0.0077,
      "step": 758
    },
    {
      "epoch": 4.097165991902834,
      "grad_norm": 0.12481268750534742,
      "learning_rate": 3.801922980432835e-06,
      "loss": 0.0079,
      "step": 759
    },
    {
      "epoch": 4.102564102564102,
      "grad_norm": 0.17656054970348686,
      "learning_rate": 3.7577419030101615e-06,
      "loss": 0.0088,
      "step": 760
    },
    {
      "epoch": 4.107962213225371,
      "grad_norm": 0.2610557267049597,
      "learning_rate": 3.7137924044861273e-06,
      "loss": 0.2539,
      "step": 761
    },
    {
      "epoch": 4.113360323886639,
      "grad_norm": 0.1838290356805346,
      "learning_rate": 3.670075111483975e-06,
      "loss": 0.0284,
      "step": 762
    },
    {
      "epoch": 4.1187584345479085,
      "grad_norm": 0.20610335600698715,
      "learning_rate": 3.626590647316206e-06,
      "loss": 0.0099,
      "step": 763
    },
    {
      "epoch": 4.124156545209177,
      "grad_norm": 0.1630204215332218,
      "learning_rate": 3.583339631975704e-06,
      "loss": 0.0083,
      "step": 764
    },
    {
      "epoch": 4.129554655870446,
      "grad_norm": 0.6353797431357658,
      "learning_rate": 3.5403226821268734e-06,
      "loss": 0.0068,
      "step": 765
    },
    {
      "epoch": 4.134952766531714,
      "grad_norm": 0.12893014852172485,
      "learning_rate": 3.497540411096869e-06,
      "loss": 0.0062,
      "step": 766
    },
    {
      "epoch": 4.140350877192983,
      "grad_norm": 0.15853943942991885,
      "learning_rate": 3.454993428866831e-06,
      "loss": 0.0078,
      "step": 767
    },
    {
      "epoch": 4.145748987854251,
      "grad_norm": 0.16815858876201803,
      "learning_rate": 3.4126823420632095e-06,
      "loss": 0.0085,
      "step": 768
    },
    {
      "epoch": 4.151147098515519,
      "grad_norm": 0.17322201023818573,
      "learning_rate": 3.3706077539490933e-06,
      "loss": 0.0499,
      "step": 769
    },
    {
      "epoch": 4.156545209176788,
      "grad_norm": 0.301026674845652,
      "learning_rate": 3.328770264415635e-06,
      "loss": 0.0338,
      "step": 770
    },
    {
      "epoch": 4.161943319838056,
      "grad_norm": 0.1778290826516132,
      "learning_rate": 3.287170469973466e-06,
      "loss": 0.0069,
      "step": 771
    },
    {
      "epoch": 4.1673414304993255,
      "grad_norm": 0.1656358924994505,
      "learning_rate": 3.2458089637442124e-06,
      "loss": 0.0079,
      "step": 772
    },
    {
      "epoch": 4.172739541160594,
      "grad_norm": 0.17053375220760866,
      "learning_rate": 3.204686335452043e-06,
      "loss": 0.0076,
      "step": 773
    },
    {
      "epoch": 4.178137651821863,
      "grad_norm": 0.1620063244322808,
      "learning_rate": 3.163803171415243e-06,
      "loss": 0.0073,
      "step": 774
    },
    {
      "epoch": 4.183535762483131,
      "grad_norm": 0.16390709001023246,
      "learning_rate": 3.1231600545378703e-06,
      "loss": 0.0069,
      "step": 775
    },
    {
      "epoch": 4.1889338731444,
      "grad_norm": 0.1894766297474302,
      "learning_rate": 3.0827575643014283e-06,
      "loss": 0.0092,
      "step": 776
    },
    {
      "epoch": 4.194331983805668,
      "grad_norm": 0.18523445493915092,
      "learning_rate": 3.0425962767566307e-06,
      "loss": 0.0074,
      "step": 777
    },
    {
      "epoch": 4.199730094466936,
      "grad_norm": 0.1846148108584596,
      "learning_rate": 3.0026767645151532e-06,
      "loss": 0.0104,
      "step": 778
    },
    {
      "epoch": 4.205128205128205,
      "grad_norm": 0.12575311717937954,
      "learning_rate": 2.9629995967414914e-06,
      "loss": 0.0054,
      "step": 779
    },
    {
      "epoch": 4.2105263157894735,
      "grad_norm": 0.18798854404124588,
      "learning_rate": 2.9235653391448448e-06,
      "loss": 0.0075,
      "step": 780
    },
    {
      "epoch": 4.2159244264507425,
      "grad_norm": 0.1627294340931128,
      "learning_rate": 2.8843745539710523e-06,
      "loss": 0.0085,
      "step": 781
    },
    {
      "epoch": 4.221322537112011,
      "grad_norm": 0.16326469476459096,
      "learning_rate": 2.8454277999945603e-06,
      "loss": 0.0072,
      "step": 782
    },
    {
      "epoch": 4.22672064777328,
      "grad_norm": 0.35487427857462644,
      "learning_rate": 2.806725632510472e-06,
      "loss": 0.2504,
      "step": 783
    },
    {
      "epoch": 4.232118758434548,
      "grad_norm": 0.18235633910436322,
      "learning_rate": 2.7682686033266337e-06,
      "loss": 0.007,
      "step": 784
    },
    {
      "epoch": 4.237516869095816,
      "grad_norm": 0.15004129859723508,
      "learning_rate": 2.730057260755743e-06,
      "loss": 0.0064,
      "step": 785
    },
    {
      "epoch": 4.242914979757085,
      "grad_norm": 0.27399522146899674,
      "learning_rate": 2.6920921496075545e-06,
      "loss": 0.1879,
      "step": 786
    },
    {
      "epoch": 4.248313090418353,
      "grad_norm": 0.17249001942659825,
      "learning_rate": 2.654373811181099e-06,
      "loss": 0.0074,
      "step": 787
    },
    {
      "epoch": 4.253711201079622,
      "grad_norm": 0.25235961908712856,
      "learning_rate": 2.6169027832569825e-06,
      "loss": 0.1351,
      "step": 788
    },
    {
      "epoch": 4.2591093117408905,
      "grad_norm": 0.13122161765075682,
      "learning_rate": 2.5796796000896882e-06,
      "loss": 0.0058,
      "step": 789
    },
    {
      "epoch": 4.2645074224021595,
      "grad_norm": 0.13879501710771264,
      "learning_rate": 2.5427047923999993e-06,
      "loss": 0.0066,
      "step": 790
    },
    {
      "epoch": 4.269905533063428,
      "grad_norm": 0.17241689078274697,
      "learning_rate": 2.505978887367393e-06,
      "loss": 0.0092,
      "step": 791
    },
    {
      "epoch": 4.275303643724697,
      "grad_norm": 0.2389164870342563,
      "learning_rate": 2.4695024086225507e-06,
      "loss": 0.2019,
      "step": 792
    },
    {
      "epoch": 4.280701754385965,
      "grad_norm": 0.20120093721542376,
      "learning_rate": 2.4332758762398846e-06,
      "loss": 0.0087,
      "step": 793
    },
    {
      "epoch": 4.286099865047233,
      "grad_norm": 0.1859035713215901,
      "learning_rate": 2.3972998067301113e-06,
      "loss": 0.0087,
      "step": 794
    },
    {
      "epoch": 4.291497975708502,
      "grad_norm": 0.24334555738366437,
      "learning_rate": 2.3615747130329013e-06,
      "loss": 0.0658,
      "step": 795
    },
    {
      "epoch": 4.29689608636977,
      "grad_norm": 0.11224893182038939,
      "learning_rate": 2.3261011045095706e-06,
      "loss": 0.0052,
      "step": 796
    },
    {
      "epoch": 4.302294197031039,
      "grad_norm": 0.18820534336508035,
      "learning_rate": 2.2908794869358044e-06,
      "loss": 0.0076,
      "step": 797
    },
    {
      "epoch": 4.3076923076923075,
      "grad_norm": 0.1754336848612205,
      "learning_rate": 2.25591036249444e-06,
      "loss": 0.0072,
      "step": 798
    },
    {
      "epoch": 4.3130904183535765,
      "grad_norm": 0.141399435578848,
      "learning_rate": 2.22119422976834e-06,
      "loss": 0.0063,
      "step": 799
    },
    {
      "epoch": 4.318488529014845,
      "grad_norm": 0.15027110262927373,
      "learning_rate": 2.186731583733235e-06,
      "loss": 0.0065,
      "step": 800
    },
    {
      "epoch": 4.323886639676114,
      "grad_norm": 0.1912745070493108,
      "learning_rate": 2.1525229157507142e-06,
      "loss": 0.0085,
      "step": 801
    },
    {
      "epoch": 4.329284750337382,
      "grad_norm": 0.17652416759862824,
      "learning_rate": 2.118568713561189e-06,
      "loss": 0.0097,
      "step": 802
    },
    {
      "epoch": 4.33468286099865,
      "grad_norm": 0.17840550642036743,
      "learning_rate": 2.084869461276946e-06,
      "loss": 0.0071,
      "step": 803
    },
    {
      "epoch": 4.340080971659919,
      "grad_norm": 0.12186032825306584,
      "learning_rate": 2.0514256393752465e-06,
      "loss": 0.006,
      "step": 804
    },
    {
      "epoch": 4.345479082321187,
      "grad_norm": 0.18333270183994504,
      "learning_rate": 2.018237724691483e-06,
      "loss": 0.0096,
      "step": 805
    },
    {
      "epoch": 4.350877192982456,
      "grad_norm": 0.15649804149725655,
      "learning_rate": 1.9853061904123662e-06,
      "loss": 0.0066,
      "step": 806
    },
    {
      "epoch": 4.3562753036437245,
      "grad_norm": 0.14219357084240056,
      "learning_rate": 1.9526315060691915e-06,
      "loss": 0.0066,
      "step": 807
    },
    {
      "epoch": 4.361673414304994,
      "grad_norm": 0.16645031776290997,
      "learning_rate": 1.9202141375311335e-06,
      "loss": 0.0069,
      "step": 808
    },
    {
      "epoch": 4.367071524966262,
      "grad_norm": 0.16646923239933129,
      "learning_rate": 1.8880545469986144e-06,
      "loss": 0.0076,
      "step": 809
    },
    {
      "epoch": 4.372469635627531,
      "grad_norm": 0.25569705111064256,
      "learning_rate": 1.8561531929967057e-06,
      "loss": 0.0084,
      "step": 810
    },
    {
      "epoch": 4.377867746288799,
      "grad_norm": 0.2319287519781518,
      "learning_rate": 1.8245105303685861e-06,
      "loss": 0.1204,
      "step": 811
    },
    {
      "epoch": 4.383265856950067,
      "grad_norm": 0.1978562124884058,
      "learning_rate": 1.7931270102690822e-06,
      "loss": 0.0095,
      "step": 812
    },
    {
      "epoch": 4.388663967611336,
      "grad_norm": 0.24941829732701262,
      "learning_rate": 1.7620030801581988e-06,
      "loss": 0.0067,
      "step": 813
    },
    {
      "epoch": 4.394062078272604,
      "grad_norm": 0.1802122557412193,
      "learning_rate": 1.7311391837947723e-06,
      "loss": 0.0081,
      "step": 814
    },
    {
      "epoch": 4.399460188933873,
      "grad_norm": 0.1494705404024646,
      "learning_rate": 1.7005357612301176e-06,
      "loss": 0.0067,
      "step": 815
    },
    {
      "epoch": 4.4048582995951415,
      "grad_norm": 0.1649691836281055,
      "learning_rate": 1.6701932488017791e-06,
      "loss": 0.0101,
      "step": 816
    },
    {
      "epoch": 4.410256410256411,
      "grad_norm": 0.20415445318740522,
      "learning_rate": 1.6401120791272829e-06,
      "loss": 0.0083,
      "step": 817
    },
    {
      "epoch": 4.415654520917679,
      "grad_norm": 0.2904972849430406,
      "learning_rate": 1.6102926810979912e-06,
      "loss": 0.0133,
      "step": 818
    },
    {
      "epoch": 4.421052631578947,
      "grad_norm": 0.17455678640910202,
      "learning_rate": 1.5807354798729678e-06,
      "loss": 0.0075,
      "step": 819
    },
    {
      "epoch": 4.426450742240216,
      "grad_norm": 0.20286904970203815,
      "learning_rate": 1.5514408968729399e-06,
      "loss": 0.0723,
      "step": 820
    },
    {
      "epoch": 4.431848852901484,
      "grad_norm": 0.141866348848916,
      "learning_rate": 1.5224093497742654e-06,
      "loss": 0.0066,
      "step": 821
    },
    {
      "epoch": 4.437246963562753,
      "grad_norm": 0.17699214194765778,
      "learning_rate": 1.493641252502993e-06,
      "loss": 0.0083,
      "step": 822
    },
    {
      "epoch": 4.442645074224021,
      "grad_norm": 0.2305318180070288,
      "learning_rate": 1.465137015228959e-06,
      "loss": 0.0629,
      "step": 823
    },
    {
      "epoch": 4.44804318488529,
      "grad_norm": 0.12688029124502123,
      "learning_rate": 1.4368970443599261e-06,
      "loss": 0.0062,
      "step": 824
    },
    {
      "epoch": 4.4534412955465585,
      "grad_norm": 0.14573022719889014,
      "learning_rate": 1.4089217425358115e-06,
      "loss": 0.0054,
      "step": 825
    },
    {
      "epoch": 4.458839406207828,
      "grad_norm": 0.1754574359245676,
      "learning_rate": 1.381211508622924e-06,
      "loss": 0.0069,
      "step": 826
    },
    {
      "epoch": 4.464237516869096,
      "grad_norm": 0.17767870785629436,
      "learning_rate": 1.3537667377082885e-06,
      "loss": 0.0074,
      "step": 827
    },
    {
      "epoch": 4.469635627530364,
      "grad_norm": 0.1679498826516171,
      "learning_rate": 1.3265878210940098e-06,
      "loss": 0.0078,
      "step": 828
    },
    {
      "epoch": 4.475033738191633,
      "grad_norm": 0.16389733624024713,
      "learning_rate": 1.2996751462917057e-06,
      "loss": 0.0071,
      "step": 829
    },
    {
      "epoch": 4.480431848852901,
      "grad_norm": 0.1597873001757078,
      "learning_rate": 1.2730290970169535e-06,
      "loss": 0.0066,
      "step": 830
    },
    {
      "epoch": 4.48582995951417,
      "grad_norm": 0.1357039185716906,
      "learning_rate": 1.2466500531838532e-06,
      "loss": 0.006,
      "step": 831
    },
    {
      "epoch": 4.491228070175438,
      "grad_norm": 0.2159108023936079,
      "learning_rate": 1.2205383908995815e-06,
      "loss": 0.2062,
      "step": 832
    },
    {
      "epoch": 4.496626180836707,
      "grad_norm": 0.14914685798519492,
      "learning_rate": 1.1946944824590513e-06,
      "loss": 0.0066,
      "step": 833
    },
    {
      "epoch": 4.502024291497976,
      "grad_norm": 0.1431530220204283,
      "learning_rate": 1.1691186963395861e-06,
      "loss": 0.0055,
      "step": 834
    },
    {
      "epoch": 4.507422402159245,
      "grad_norm": 0.1786336644306138,
      "learning_rate": 1.1438113971956776e-06,
      "loss": 0.0084,
      "step": 835
    },
    {
      "epoch": 4.512820512820513,
      "grad_norm": 0.17770533748705705,
      "learning_rate": 1.11877294585379e-06,
      "loss": 0.0075,
      "step": 836
    },
    {
      "epoch": 4.518218623481781,
      "grad_norm": 0.15129590466594908,
      "learning_rate": 1.0940036993071934e-06,
      "loss": 0.0081,
      "step": 837
    },
    {
      "epoch": 4.52361673414305,
      "grad_norm": 0.19105021139758796,
      "learning_rate": 1.0695040107109089e-06,
      "loss": 0.0065,
      "step": 838
    },
    {
      "epoch": 4.529014844804318,
      "grad_norm": 0.14979723404668785,
      "learning_rate": 1.045274229376636e-06,
      "loss": 0.006,
      "step": 839
    },
    {
      "epoch": 4.534412955465587,
      "grad_norm": 0.1846638960067091,
      "learning_rate": 1.0213147007678036e-06,
      "loss": 0.0075,
      "step": 840
    },
    {
      "epoch": 4.539811066126855,
      "grad_norm": 0.1454433452689002,
      "learning_rate": 9.97625766494621e-07,
      "loss": 0.0073,
      "step": 841
    },
    {
      "epoch": 4.545209176788124,
      "grad_norm": 0.18488942655661209,
      "learning_rate": 9.742077643092207e-07,
      "loss": 0.0096,
      "step": 842
    },
    {
      "epoch": 4.550607287449393,
      "grad_norm": 0.2006493023154337,
      "learning_rate": 9.51061028100837e-07,
      "loss": 0.0074,
      "step": 843
    },
    {
      "epoch": 4.556005398110662,
      "grad_norm": 0.18899821986914248,
      "learning_rate": 9.281858878910555e-07,
      "loss": 0.0091,
      "step": 844
    },
    {
      "epoch": 4.56140350877193,
      "grad_norm": 0.21143375715615836,
      "learning_rate": 9.055826698290881e-07,
      "loss": 0.0075,
      "step": 845
    },
    {
      "epoch": 4.566801619433198,
      "grad_norm": 0.15198860228197414,
      "learning_rate": 8.832516961871462e-07,
      "loss": 0.0053,
      "step": 846
    },
    {
      "epoch": 4.572199730094467,
      "grad_norm": 0.22751271824950478,
      "learning_rate": 8.611932853558236e-07,
      "loss": 0.0084,
      "step": 847
    },
    {
      "epoch": 4.577597840755735,
      "grad_norm": 0.16335646731283246,
      "learning_rate": 8.394077518395738e-07,
      "loss": 0.0079,
      "step": 848
    },
    {
      "epoch": 4.582995951417004,
      "grad_norm": 0.16417859671024335,
      "learning_rate": 8.178954062522226e-07,
      "loss": 0.0082,
      "step": 849
    },
    {
      "epoch": 4.588394062078272,
      "grad_norm": 0.1371732934582776,
      "learning_rate": 7.966565553125294e-07,
      "loss": 0.0059,
      "step": 850
    },
    {
      "epoch": 4.593792172739541,
      "grad_norm": 0.1718988181824814,
      "learning_rate": 7.756915018398214e-07,
      "loss": 0.0068,
      "step": 851
    },
    {
      "epoch": 4.59919028340081,
      "grad_norm": 0.15416993412420898,
      "learning_rate": 7.5500054474968e-07,
      "loss": 0.0062,
      "step": 852
    },
    {
      "epoch": 4.604588394062079,
      "grad_norm": 0.17416110080710895,
      "learning_rate": 7.345839790496745e-07,
      "loss": 0.0068,
      "step": 853
    },
    {
      "epoch": 4.609986504723347,
      "grad_norm": 0.16830986049559513,
      "learning_rate": 7.144420958351506e-07,
      "loss": 0.0075,
      "step": 854
    },
    {
      "epoch": 4.615384615384615,
      "grad_norm": 0.17367183744141607,
      "learning_rate": 6.945751822850955e-07,
      "loss": 0.009,
      "step": 855
    },
    {
      "epoch": 4.620782726045884,
      "grad_norm": 0.19142013846945674,
      "learning_rate": 6.749835216580214e-07,
      "loss": 0.0092,
      "step": 856
    },
    {
      "epoch": 4.626180836707152,
      "grad_norm": 0.2067536155528566,
      "learning_rate": 6.556673932879488e-07,
      "loss": 0.0072,
      "step": 857
    },
    {
      "epoch": 4.631578947368421,
      "grad_norm": 0.14797804422750227,
      "learning_rate": 6.366270725804069e-07,
      "loss": 0.0068,
      "step": 858
    },
    {
      "epoch": 4.636977058029689,
      "grad_norm": 0.19586304150902226,
      "learning_rate": 6.178628310085178e-07,
      "loss": 0.0103,
      "step": 859
    },
    {
      "epoch": 4.6423751686909585,
      "grad_norm": 0.24693313978740186,
      "learning_rate": 5.993749361091206e-07,
      "loss": 0.133,
      "step": 860
    },
    {
      "epoch": 4.647773279352227,
      "grad_norm": 0.15229371755436824,
      "learning_rate": 5.811636514789598e-07,
      "loss": 0.0068,
      "step": 861
    },
    {
      "epoch": 4.653171390013496,
      "grad_norm": 0.2247760397976527,
      "learning_rate": 5.632292367709236e-07,
      "loss": 0.1055,
      "step": 862
    },
    {
      "epoch": 4.658569500674764,
      "grad_norm": 0.14424338258050534,
      "learning_rate": 5.455719476903442e-07,
      "loss": 0.0063,
      "step": 863
    },
    {
      "epoch": 4.663967611336032,
      "grad_norm": 0.15937610135438382,
      "learning_rate": 5.281920359913528e-07,
      "loss": 0.0176,
      "step": 864
    },
    {
      "epoch": 4.669365721997301,
      "grad_norm": 0.1679601387250827,
      "learning_rate": 5.110897494732881e-07,
      "loss": 0.0064,
      "step": 865
    },
    {
      "epoch": 4.674763832658569,
      "grad_norm": 0.1526193669278446,
      "learning_rate": 4.942653319771618e-07,
      "loss": 0.0071,
      "step": 866
    },
    {
      "epoch": 4.680161943319838,
      "grad_norm": 0.13442330611159828,
      "learning_rate": 4.777190233821838e-07,
      "loss": 0.0061,
      "step": 867
    },
    {
      "epoch": 4.685560053981106,
      "grad_norm": 0.1636822162108528,
      "learning_rate": 4.6145105960235136e-07,
      "loss": 0.0071,
      "step": 868
    },
    {
      "epoch": 4.6909581646423755,
      "grad_norm": 0.1596200679398143,
      "learning_rate": 4.4546167258306296e-07,
      "loss": 0.0075,
      "step": 869
    },
    {
      "epoch": 4.696356275303644,
      "grad_norm": 0.16839642747934516,
      "learning_rate": 4.297510902978341e-07,
      "loss": 0.0083,
      "step": 870
    },
    {
      "epoch": 4.701754385964913,
      "grad_norm": 0.159063406884678,
      "learning_rate": 4.1431953674503586e-07,
      "loss": 0.0065,
      "step": 871
    },
    {
      "epoch": 4.707152496626181,
      "grad_norm": 0.16942375232844864,
      "learning_rate": 3.991672319447015e-07,
      "loss": 0.0073,
      "step": 872
    },
    {
      "epoch": 4.712550607287449,
      "grad_norm": 0.15555943118350962,
      "learning_rate": 3.842943919353914e-07,
      "loss": 0.0073,
      "step": 873
    },
    {
      "epoch": 4.717948717948718,
      "grad_norm": 0.25614964211315966,
      "learning_rate": 3.6970122877111104e-07,
      "loss": 0.2021,
      "step": 874
    },
    {
      "epoch": 4.723346828609986,
      "grad_norm": 0.19433096617350248,
      "learning_rate": 3.5538795051829336e-07,
      "loss": 0.0068,
      "step": 875
    },
    {
      "epoch": 4.728744939271255,
      "grad_norm": 0.16794302830857327,
      "learning_rate": 3.413547612528212e-07,
      "loss": 0.0068,
      "step": 876
    },
    {
      "epoch": 4.734143049932523,
      "grad_norm": 0.16974503840122634,
      "learning_rate": 3.2760186105712964e-07,
      "loss": 0.0083,
      "step": 877
    },
    {
      "epoch": 4.7395411605937925,
      "grad_norm": 0.13212568788120146,
      "learning_rate": 3.1412944601734165e-07,
      "loss": 0.0057,
      "step": 878
    },
    {
      "epoch": 4.744939271255061,
      "grad_norm": 0.1976283645393667,
      "learning_rate": 3.0093770822048786e-07,
      "loss": 0.0079,
      "step": 879
    },
    {
      "epoch": 4.75033738191633,
      "grad_norm": 0.15836118640095823,
      "learning_rate": 2.8802683575174907e-07,
      "loss": 0.0073,
      "step": 880
    },
    {
      "epoch": 4.755735492577598,
      "grad_norm": 0.14238697443763573,
      "learning_rate": 2.753970126917871e-07,
      "loss": 0.0055,
      "step": 881
    },
    {
      "epoch": 4.761133603238866,
      "grad_norm": 0.26842025770841327,
      "learning_rate": 2.6304841911412027e-07,
      "loss": 0.0182,
      "step": 882
    },
    {
      "epoch": 4.766531713900135,
      "grad_norm": 0.15757318285422225,
      "learning_rate": 2.509812310825499e-07,
      "loss": 0.0066,
      "step": 883
    },
    {
      "epoch": 4.771929824561403,
      "grad_norm": 0.1618576241796593,
      "learning_rate": 2.391956206486468e-07,
      "loss": 0.0071,
      "step": 884
    },
    {
      "epoch": 4.777327935222672,
      "grad_norm": 0.1946044303403527,
      "learning_rate": 2.2769175584931746e-07,
      "loss": 0.0153,
      "step": 885
    },
    {
      "epoch": 4.7827260458839405,
      "grad_norm": 0.2296158932271888,
      "learning_rate": 2.1646980070437973e-07,
      "loss": 0.0081,
      "step": 886
    },
    {
      "epoch": 4.7881241565452095,
      "grad_norm": 0.3387688201949446,
      "learning_rate": 2.0552991521425537e-07,
      "loss": 0.2726,
      "step": 887
    },
    {
      "epoch": 4.793522267206478,
      "grad_norm": 0.2482321920693236,
      "learning_rate": 1.9487225535765873e-07,
      "loss": 0.0211,
      "step": 888
    },
    {
      "epoch": 4.798920377867747,
      "grad_norm": 0.15086049070317792,
      "learning_rate": 1.8449697308939863e-07,
      "loss": 0.0062,
      "step": 889
    },
    {
      "epoch": 4.804318488529015,
      "grad_norm": 0.15240141938966198,
      "learning_rate": 1.7440421633818872e-07,
      "loss": 0.0062,
      "step": 890
    },
    {
      "epoch": 4.809716599190283,
      "grad_norm": 0.23578299035015984,
      "learning_rate": 1.6459412900455607e-07,
      "loss": 0.0125,
      "step": 891
    },
    {
      "epoch": 4.815114709851552,
      "grad_norm": 0.1470906953689497,
      "learning_rate": 1.5506685095878715e-07,
      "loss": 0.0065,
      "step": 892
    },
    {
      "epoch": 4.82051282051282,
      "grad_norm": 0.15212434477146647,
      "learning_rate": 1.4582251803892055e-07,
      "loss": 0.0065,
      "step": 893
    },
    {
      "epoch": 4.825910931174089,
      "grad_norm": 0.14747099109095105,
      "learning_rate": 1.3686126204882634e-07,
      "loss": 0.007,
      "step": 894
    },
    {
      "epoch": 4.8313090418353575,
      "grad_norm": 0.16444206201175393,
      "learning_rate": 1.2818321075631635e-07,
      "loss": 0.0059,
      "step": 895
    },
    {
      "epoch": 4.8367071524966265,
      "grad_norm": 0.15611273866535655,
      "learning_rate": 1.19788487891328e-07,
      "loss": 0.0075,
      "step": 896
    },
    {
      "epoch": 4.842105263157895,
      "grad_norm": 0.1781583785611643,
      "learning_rate": 1.1167721314415681e-07,
      "loss": 0.0079,
      "step": 897
    },
    {
      "epoch": 4.847503373819164,
      "grad_norm": 0.20606188899572905,
      "learning_rate": 1.038495021637509e-07,
      "loss": 0.0108,
      "step": 898
    },
    {
      "epoch": 4.852901484480432,
      "grad_norm": 0.17844153813529373,
      "learning_rate": 9.630546655606365e-08,
      "loss": 0.0087,
      "step": 899
    },
    {
      "epoch": 4.8582995951417,
      "grad_norm": 0.1560764016762354,
      "learning_rate": 8.904521388245713e-08,
      "loss": 0.0068,
      "step": 900
    },
    {
      "epoch": 4.863697705802969,
      "grad_norm": 0.1487169516366418,
      "learning_rate": 8.206884765818102e-08,
      "loss": 0.0068,
      "step": 901
    },
    {
      "epoch": 4.869095816464237,
      "grad_norm": 0.2463461690457611,
      "learning_rate": 7.537646735087833e-08,
      "loss": 0.1513,
      "step": 902
    },
    {
      "epoch": 4.874493927125506,
      "grad_norm": 0.2102978007590072,
      "learning_rate": 6.896816837918651e-08,
      "loss": 0.0084,
      "step": 903
    },
    {
      "epoch": 4.8798920377867745,
      "grad_norm": 0.20095267792043717,
      "learning_rate": 6.284404211135853e-08,
      "loss": 0.0065,
      "step": 904
    },
    {
      "epoch": 4.8852901484480435,
      "grad_norm": 0.15528097841618993,
      "learning_rate": 5.700417586397722e-08,
      "loss": 0.0065,
      "step": 905
    },
    {
      "epoch": 4.890688259109312,
      "grad_norm": 0.16940230756369326,
      "learning_rate": 5.144865290069634e-08,
      "loss": 0.0087,
      "step": 906
    },
    {
      "epoch": 4.896086369770581,
      "grad_norm": 0.1378627208589382,
      "learning_rate": 4.61775524310637e-08,
      "loss": 0.0078,
      "step": 907
    },
    {
      "epoch": 4.901484480431849,
      "grad_norm": 0.20709815634537396,
      "learning_rate": 4.119094960938208e-08,
      "loss": 0.0138,
      "step": 908
    },
    {
      "epoch": 4.906882591093117,
      "grad_norm": 0.2007039062473555,
      "learning_rate": 3.648891553365008e-08,
      "loss": 0.0083,
      "step": 909
    },
    {
      "epoch": 4.912280701754386,
      "grad_norm": 0.19819856548790862,
      "learning_rate": 3.20715172445385e-08,
      "loss": 0.1599,
      "step": 910
    },
    {
      "epoch": 4.917678812415654,
      "grad_norm": 0.1719664928908639,
      "learning_rate": 2.793881772443552e-08,
      "loss": 0.0071,
      "step": 911
    },
    {
      "epoch": 4.923076923076923,
      "grad_norm": 0.22465213945689566,
      "learning_rate": 2.4090875896551903e-08,
      "loss": 0.1584,
      "step": 912
    },
    {
      "epoch": 4.9284750337381915,
      "grad_norm": 0.16971530754399908,
      "learning_rate": 2.0527746624081635e-08,
      "loss": 0.0062,
      "step": 913
    },
    {
      "epoch": 4.933873144399461,
      "grad_norm": 0.16730279275755114,
      "learning_rate": 1.72494807094159e-08,
      "loss": 0.0064,
      "step": 914
    },
    {
      "epoch": 4.939271255060729,
      "grad_norm": 0.15712693233721395,
      "learning_rate": 1.425612489341921e-08,
      "loss": 0.0087,
      "step": 915
    },
    {
      "epoch": 4.944669365721998,
      "grad_norm": 0.15485321834215096,
      "learning_rate": 1.1547721854767712e-08,
      "loss": 0.0071,
      "step": 916
    },
    {
      "epoch": 4.950067476383266,
      "grad_norm": 0.1633322503676924,
      "learning_rate": 9.12431020933191e-09,
      "loss": 0.0079,
      "step": 917
    },
    {
      "epoch": 4.955465587044534,
      "grad_norm": 0.13338657253580138,
      "learning_rate": 6.985924509639308e-09,
      "loss": 0.0054,
      "step": 918
    },
    {
      "epoch": 4.960863697705803,
      "grad_norm": 0.1615631056618365,
      "learning_rate": 5.1325952443681544e-09,
      "loss": 0.0072,
      "step": 919
    },
    {
      "epoch": 4.966261808367071,
      "grad_norm": 0.18025545983832483,
      "learning_rate": 3.5643488379211167e-09,
      "loss": 0.0069,
      "step": 920
    },
    {
      "epoch": 4.97165991902834,
      "grad_norm": 0.19857679729061328,
      "learning_rate": 2.2812076500411397e-09,
      "loss": 0.1414,
      "step": 921
    },
    {
      "epoch": 4.9770580296896085,
      "grad_norm": 0.20801482450926878,
      "learning_rate": 1.283189975505028e-09,
      "loss": 0.0678,
      "step": 922
    },
    {
      "epoch": 4.982456140350877,
      "grad_norm": 0.15867502405159886,
      "learning_rate": 5.703100438481102e-10,
      "loss": 0.0065,
      "step": 923
    },
    {
      "epoch": 4.987854251012146,
      "grad_norm": 0.1615039310082322,
      "learning_rate": 1.4257801917549884e-10,
      "loss": 0.0068,
      "step": 924
    },
    {
      "epoch": 4.993252361673415,
      "grad_norm": 0.1853912335035784,
      "learning_rate": 0.0,
      "loss": 0.0084,
      "step": 925
    },
    {
      "epoch": 4.993252361673415,
      "step": 925,
      "total_flos": 1.797075321619333e+17,
      "train_loss": 0.26902934940078777,
      "train_runtime": 3110.9836,
      "train_samples_per_second": 38.091,
      "train_steps_per_second": 0.297
    }
  ],
  "logging_steps": 1,
  "max_steps": 925,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.797075321619333e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}