{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 6992,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 4.635512828826904,
      "learning_rate": 2.9995709382151033e-05,
      "loss": 6.0242,
      "step": 1
    },
    {
      "epoch": 0.0,
      "grad_norm": 4.262711524963379,
      "learning_rate": 2.9991418764302058e-05,
      "loss": 5.8131,
      "step": 2
    },
    {
      "epoch": 0.0,
      "grad_norm": 4.200323581695557,
      "learning_rate": 2.998712814645309e-05,
      "loss": 5.6225,
      "step": 3
    },
    {
      "epoch": 0.0,
      "grad_norm": 4.810133457183838,
      "learning_rate": 2.998283752860412e-05,
      "loss": 5.4304,
      "step": 4
    },
    {
      "epoch": 0.0,
      "grad_norm": 4.93181037902832,
      "learning_rate": 2.997854691075515e-05,
      "loss": 5.2684,
      "step": 5
    },
    {
      "epoch": 0.0,
      "grad_norm": 5.074693202972412,
      "learning_rate": 2.9974256292906177e-05,
      "loss": 5.0533,
      "step": 6
    },
    {
      "epoch": 0.0,
      "grad_norm": 4.967318534851074,
      "learning_rate": 2.996996567505721e-05,
      "loss": 4.9104,
      "step": 7
    },
    {
      "epoch": 0.0,
      "grad_norm": 5.453105926513672,
      "learning_rate": 2.9965675057208238e-05,
      "loss": 4.6064,
      "step": 8
    },
    {
      "epoch": 0.0,
      "grad_norm": 5.421672344207764,
      "learning_rate": 2.996138443935927e-05,
      "loss": 4.5984,
      "step": 9
    },
    {
      "epoch": 0.0,
      "grad_norm": 5.02965784072876,
      "learning_rate": 2.99570938215103e-05,
      "loss": 4.449,
      "step": 10
    },
    {
      "epoch": 0.0,
      "grad_norm": 5.782529830932617,
      "learning_rate": 2.9952803203661327e-05,
      "loss": 4.1912,
      "step": 11
    },
    {
      "epoch": 0.0,
      "grad_norm": 6.099738121032715,
      "learning_rate": 2.994851258581236e-05,
      "loss": 4.1198,
      "step": 12
    },
    {
      "epoch": 0.0,
      "grad_norm": 5.961019515991211,
      "learning_rate": 2.9944221967963388e-05,
      "loss": 4.0437,
      "step": 13
    },
    {
      "epoch": 0.0,
      "grad_norm": 10.967097282409668,
      "learning_rate": 2.9939931350114417e-05,
      "loss": 3.885,
      "step": 14
    },
    {
      "epoch": 0.0,
      "grad_norm": 8.03209114074707,
      "learning_rate": 2.9935640732265446e-05,
      "loss": 3.6933,
      "step": 15
    },
    {
      "epoch": 0.0,
      "grad_norm": 8.347334861755371,
      "learning_rate": 2.9931350114416478e-05,
      "loss": 3.859,
      "step": 16
    },
    {
      "epoch": 0.0,
      "grad_norm": 9.741003036499023,
      "learning_rate": 2.9927059496567503e-05,
      "loss": 3.4883,
      "step": 17
    },
    {
      "epoch": 0.01,
      "grad_norm": 11.666119575500488,
      "learning_rate": 2.9922768878718536e-05,
      "loss": 3.4243,
      "step": 18
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.64477252960205,
      "learning_rate": 2.9918478260869568e-05,
      "loss": 3.2306,
      "step": 19
    },
    {
      "epoch": 0.01,
      "grad_norm": 22.396427154541016,
      "learning_rate": 2.9914187643020596e-05,
      "loss": 3.3432,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 14.657906532287598,
      "learning_rate": 2.9909897025171625e-05,
      "loss": 3.2188,
      "step": 21
    },
    {
      "epoch": 0.01,
      "grad_norm": 11.172964096069336,
      "learning_rate": 2.9905606407322654e-05,
      "loss": 3.009,
      "step": 22
    },
    {
      "epoch": 0.01,
      "grad_norm": 30.494609832763672,
      "learning_rate": 2.9901315789473686e-05,
      "loss": 3.1648,
      "step": 23
    },
    {
      "epoch": 0.01,
      "grad_norm": 20.13739776611328,
      "learning_rate": 2.9897025171624715e-05,
      "loss": 2.8155,
      "step": 24
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.68649959564209,
      "learning_rate": 2.9892734553775744e-05,
      "loss": 2.8085,
      "step": 25
    },
    {
      "epoch": 0.01,
      "grad_norm": 13.222569465637207,
      "learning_rate": 2.9888443935926776e-05,
      "loss": 2.8157,
      "step": 26
    },
    {
      "epoch": 0.01,
      "grad_norm": 21.290855407714844,
      "learning_rate": 2.9884153318077805e-05,
      "loss": 2.6447,
      "step": 27
    },
    {
      "epoch": 0.01,
      "grad_norm": 20.91508674621582,
      "learning_rate": 2.9879862700228833e-05,
      "loss": 2.6235,
      "step": 28
    },
    {
      "epoch": 0.01,
      "grad_norm": 14.453760147094727,
      "learning_rate": 2.9875572082379862e-05,
      "loss": 2.6736,
      "step": 29
    },
    {
      "epoch": 0.01,
      "grad_norm": 15.384343147277832,
      "learning_rate": 2.9871281464530894e-05,
      "loss": 2.6873,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 11.872452735900879,
      "learning_rate": 2.9866990846681923e-05,
      "loss": 2.6731,
      "step": 31
    },
    {
      "epoch": 0.01,
      "grad_norm": 7.314385890960693,
      "learning_rate": 2.9862700228832952e-05,
      "loss": 2.2646,
      "step": 32
    },
    {
      "epoch": 0.01,
      "grad_norm": 8.31044864654541,
      "learning_rate": 2.985840961098398e-05,
      "loss": 2.5389,
      "step": 33
    },
    {
      "epoch": 0.01,
      "grad_norm": 11.034132957458496,
      "learning_rate": 2.9854118993135013e-05,
      "loss": 2.2643,
      "step": 34
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.82011604309082,
      "learning_rate": 2.9849828375286045e-05,
      "loss": 2.0751,
      "step": 35
    },
    {
      "epoch": 0.01,
      "grad_norm": 8.5283842086792,
      "learning_rate": 2.984553775743707e-05,
      "loss": 2.4416,
      "step": 36
    },
    {
      "epoch": 0.01,
      "grad_norm": 8.773575782775879,
      "learning_rate": 2.9841247139588102e-05,
      "loss": 2.417,
      "step": 37
    },
    {
      "epoch": 0.01,
      "grad_norm": 17.709125518798828,
      "learning_rate": 2.983695652173913e-05,
      "loss": 2.5134,
      "step": 38
    },
    {
      "epoch": 0.01,
      "grad_norm": 11.986088752746582,
      "learning_rate": 2.9832665903890163e-05,
      "loss": 2.5418,
      "step": 39
    },
    {
      "epoch": 0.01,
      "grad_norm": 8.830208778381348,
      "learning_rate": 2.982837528604119e-05,
      "loss": 2.4333,
      "step": 40
    },
    {
      "epoch": 0.01,
      "grad_norm": 8.927085876464844,
      "learning_rate": 2.982408466819222e-05,
      "loss": 2.3037,
      "step": 41
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.560665130615234,
      "learning_rate": 2.981979405034325e-05,
      "loss": 2.5077,
      "step": 42
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.039464950561523,
      "learning_rate": 2.981550343249428e-05,
      "loss": 2.2192,
      "step": 43
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.975371360778809,
      "learning_rate": 2.981121281464531e-05,
      "loss": 2.6494,
      "step": 44
    },
    {
      "epoch": 0.01,
      "grad_norm": 8.994516372680664,
      "learning_rate": 2.980692219679634e-05,
      "loss": 2.2144,
      "step": 45
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.323376655578613,
      "learning_rate": 2.980263157894737e-05,
      "loss": 2.2742,
      "step": 46
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.652460098266602,
      "learning_rate": 2.9798340961098397e-05,
      "loss": 2.1642,
      "step": 47
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.230737686157227,
      "learning_rate": 2.979405034324943e-05,
      "loss": 2.2399,
      "step": 48
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.706355094909668,
      "learning_rate": 2.9789759725400458e-05,
      "loss": 2.2218,
      "step": 49
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.502559661865234,
      "learning_rate": 2.978546910755149e-05,
      "loss": 2.0737,
      "step": 50
    },
    {
      "epoch": 0.01,
      "grad_norm": 11.392340660095215,
      "learning_rate": 2.9781178489702515e-05,
      "loss": 2.5131,
      "step": 51
    },
    {
      "epoch": 0.01,
      "grad_norm": 14.368525505065918,
      "learning_rate": 2.9776887871853548e-05,
      "loss": 1.8392,
      "step": 52
    },
    {
      "epoch": 0.02,
      "grad_norm": 12.000808715820312,
      "learning_rate": 2.977259725400458e-05,
      "loss": 2.0387,
      "step": 53
    },
    {
      "epoch": 0.02,
      "grad_norm": 10.566619873046875,
      "learning_rate": 2.9768306636155605e-05,
      "loss": 2.2464,
      "step": 54
    },
    {
      "epoch": 0.02,
      "grad_norm": 10.068693161010742,
      "learning_rate": 2.9764016018306637e-05,
      "loss": 2.0293,
      "step": 55
    },
    {
      "epoch": 0.02,
      "grad_norm": 12.956381797790527,
      "learning_rate": 2.9759725400457666e-05,
      "loss": 2.3687,
      "step": 56
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.81608772277832,
      "learning_rate": 2.9755434782608698e-05,
      "loss": 2.377,
      "step": 57
    },
    {
      "epoch": 0.02,
      "grad_norm": 11.117986679077148,
      "learning_rate": 2.9751144164759724e-05,
      "loss": 2.439,
      "step": 58
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.488683700561523,
      "learning_rate": 2.9746853546910756e-05,
      "loss": 1.8631,
      "step": 59
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.672380447387695,
      "learning_rate": 2.9742562929061788e-05,
      "loss": 2.0851,
      "step": 60
    },
    {
      "epoch": 0.02,
      "grad_norm": 12.200733184814453,
      "learning_rate": 2.9738272311212817e-05,
      "loss": 2.5484,
      "step": 61
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.255772590637207,
      "learning_rate": 2.9733981693363845e-05,
      "loss": 2.1288,
      "step": 62
    },
    {
      "epoch": 0.02,
      "grad_norm": 10.493276596069336,
      "learning_rate": 2.9729691075514874e-05,
      "loss": 2.2297,
      "step": 63
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.244399070739746,
      "learning_rate": 2.9725400457665906e-05,
      "loss": 2.4804,
      "step": 64
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.426789283752441,
      "learning_rate": 2.9721109839816935e-05,
      "loss": 2.1489,
      "step": 65
    },
    {
      "epoch": 0.02,
      "grad_norm": 11.235112190246582,
      "learning_rate": 2.9716819221967964e-05,
      "loss": 2.0655,
      "step": 66
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.057259559631348,
      "learning_rate": 2.9712528604118993e-05,
      "loss": 2.2147,
      "step": 67
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.027843475341797,
      "learning_rate": 2.9708237986270025e-05,
      "loss": 1.9357,
      "step": 68
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.6011962890625,
      "learning_rate": 2.9703947368421054e-05,
      "loss": 2.1391,
      "step": 69
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.837102890014648,
      "learning_rate": 2.9699656750572082e-05,
      "loss": 2.1937,
      "step": 70
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.933879852294922,
      "learning_rate": 2.9695366132723114e-05,
      "loss": 2.4057,
      "step": 71
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.57646369934082,
      "learning_rate": 2.9691075514874143e-05,
      "loss": 2.2708,
      "step": 72
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.571762084960938,
      "learning_rate": 2.9686784897025172e-05,
      "loss": 1.9102,
      "step": 73
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.619646072387695,
      "learning_rate": 2.96824942791762e-05,
      "loss": 1.9027,
      "step": 74
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.471129417419434,
      "learning_rate": 2.9678203661327233e-05,
      "loss": 2.2281,
      "step": 75
    },
    {
      "epoch": 0.02,
      "grad_norm": 12.377062797546387,
      "learning_rate": 2.967391304347826e-05,
      "loss": 2.3651,
      "step": 76
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.500938415527344,
      "learning_rate": 2.966962242562929e-05,
      "loss": 2.1969,
      "step": 77
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.627216339111328,
      "learning_rate": 2.9665331807780323e-05,
      "loss": 2.1825,
      "step": 78
    },
    {
      "epoch": 0.02,
      "grad_norm": 10.074145317077637,
      "learning_rate": 2.966104118993135e-05,
      "loss": 2.0536,
      "step": 79
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.057881355285645,
      "learning_rate": 2.965675057208238e-05,
      "loss": 1.9442,
      "step": 80
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.898153305053711,
      "learning_rate": 2.965245995423341e-05,
      "loss": 1.9726,
      "step": 81
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.749974250793457,
      "learning_rate": 2.964816933638444e-05,
      "loss": 1.8927,
      "step": 82
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.17689037322998,
      "learning_rate": 2.964387871853547e-05,
      "loss": 2.2203,
      "step": 83
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.687392234802246,
      "learning_rate": 2.96395881006865e-05,
      "loss": 2.314,
      "step": 84
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.856606483459473,
      "learning_rate": 2.9635297482837527e-05,
      "loss": 1.8409,
      "step": 85
    },
    {
      "epoch": 0.02,
      "grad_norm": 11.169034004211426,
      "learning_rate": 2.963100686498856e-05,
      "loss": 2.1135,
      "step": 86
    },
    {
      "epoch": 0.02,
      "grad_norm": 12.441506385803223,
      "learning_rate": 2.962671624713959e-05,
      "loss": 1.8989,
      "step": 87
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.590865135192871,
      "learning_rate": 2.9622425629290617e-05,
      "loss": 2.1097,
      "step": 88
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.905450820922852,
      "learning_rate": 2.961813501144165e-05,
      "loss": 1.9331,
      "step": 89
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.653109550476074,
      "learning_rate": 2.9613844393592678e-05,
      "loss": 2.4247,
      "step": 90
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.621190071105957,
      "learning_rate": 2.960955377574371e-05,
      "loss": 2.0484,
      "step": 91
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.247668266296387,
      "learning_rate": 2.9605263157894735e-05,
      "loss": 1.8694,
      "step": 92
    },
    {
      "epoch": 0.03,
      "grad_norm": 11.075560569763184,
      "learning_rate": 2.9600972540045768e-05,
      "loss": 1.7545,
      "step": 93
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.714777946472168,
      "learning_rate": 2.9596681922196796e-05,
      "loss": 1.961,
      "step": 94
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.880008697509766,
      "learning_rate": 2.9592391304347825e-05,
      "loss": 1.727,
      "step": 95
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.815738677978516,
      "learning_rate": 2.9588100686498857e-05,
      "loss": 1.6125,
      "step": 96
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.082423210144043,
      "learning_rate": 2.9583810068649886e-05,
      "loss": 2.1206,
      "step": 97
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.187947273254395,
      "learning_rate": 2.9579519450800918e-05,
      "loss": 2.1751,
      "step": 98
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.888360977172852,
      "learning_rate": 2.9575228832951944e-05,
      "loss": 1.8519,
      "step": 99
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.402551651000977,
      "learning_rate": 2.9570938215102976e-05,
      "loss": 1.8692,
      "step": 100
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.821449279785156,
      "learning_rate": 2.9566647597254005e-05,
      "loss": 2.0974,
      "step": 101
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.82812213897705,
      "learning_rate": 2.9562356979405037e-05,
      "loss": 2.039,
      "step": 102
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.911670684814453,
      "learning_rate": 2.9558066361556065e-05,
      "loss": 1.8222,
      "step": 103
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.715346336364746,
      "learning_rate": 2.9553775743707094e-05,
      "loss": 1.9212,
      "step": 104
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.142864227294922,
      "learning_rate": 2.9549485125858126e-05,
      "loss": 1.8101,
      "step": 105
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.892821311950684,
      "learning_rate": 2.9545194508009152e-05,
      "loss": 1.7335,
      "step": 106
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.593453407287598,
      "learning_rate": 2.9540903890160184e-05,
      "loss": 1.8757,
      "step": 107
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.683908462524414,
      "learning_rate": 2.9536613272311213e-05,
      "loss": 1.7976,
      "step": 108
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.749051094055176,
      "learning_rate": 2.9532322654462245e-05,
      "loss": 1.7942,
      "step": 109
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.038222312927246,
      "learning_rate": 2.952803203661327e-05,
      "loss": 1.9111,
      "step": 110
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.059216499328613,
      "learning_rate": 2.9523741418764302e-05,
      "loss": 1.7355,
      "step": 111
    },
    {
      "epoch": 0.03,
      "grad_norm": 12.66389274597168,
      "learning_rate": 2.9519450800915335e-05,
      "loss": 1.9505,
      "step": 112
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.865212440490723,
      "learning_rate": 2.9515160183066363e-05,
      "loss": 1.878,
      "step": 113
    },
    {
      "epoch": 0.03,
      "grad_norm": 11.242400169372559,
      "learning_rate": 2.9510869565217392e-05,
      "loss": 1.8862,
      "step": 114
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.39486312866211,
      "learning_rate": 2.950657894736842e-05,
      "loss": 2.164,
      "step": 115
    },
    {
      "epoch": 0.03,
      "grad_norm": 7.808732986450195,
      "learning_rate": 2.9502288329519453e-05,
      "loss": 1.5214,
      "step": 116
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.637123107910156,
      "learning_rate": 2.9497997711670482e-05,
      "loss": 1.8034,
      "step": 117
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.61000919342041,
      "learning_rate": 2.949370709382151e-05,
      "loss": 1.8804,
      "step": 118
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.153660774230957,
      "learning_rate": 2.948941647597254e-05,
      "loss": 1.757,
      "step": 119
    },
    {
      "epoch": 0.03,
      "grad_norm": 11.368372917175293,
      "learning_rate": 2.948512585812357e-05,
      "loss": 2.169,
      "step": 120
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.239725112915039,
      "learning_rate": 2.94808352402746e-05,
      "loss": 1.7511,
      "step": 121
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.863459587097168,
      "learning_rate": 2.947654462242563e-05,
      "loss": 1.6704,
      "step": 122
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.673696517944336,
      "learning_rate": 2.947225400457666e-05,
      "loss": 2.0968,
      "step": 123
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.66123104095459,
      "learning_rate": 2.946796338672769e-05,
      "loss": 2.0148,
      "step": 124
    },
    {
      "epoch": 0.04,
      "grad_norm": 11.710880279541016,
      "learning_rate": 2.946367276887872e-05,
      "loss": 1.8219,
      "step": 125
    },
    {
      "epoch": 0.04,
      "grad_norm": 13.148943901062012,
      "learning_rate": 2.9459382151029747e-05,
      "loss": 2.0196,
      "step": 126
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.845335960388184,
      "learning_rate": 2.945509153318078e-05,
      "loss": 1.9038,
      "step": 127
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.211142539978027,
      "learning_rate": 2.945080091533181e-05,
      "loss": 1.4489,
      "step": 128
    },
    {
      "epoch": 0.04,
      "grad_norm": 13.306626319885254,
      "learning_rate": 2.9446510297482837e-05,
      "loss": 2.1193,
      "step": 129
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.279850959777832,
      "learning_rate": 2.944221967963387e-05,
      "loss": 1.6406,
      "step": 130
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.513463020324707,
      "learning_rate": 2.9437929061784898e-05,
      "loss": 1.4844,
      "step": 131
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.108253479003906,
      "learning_rate": 2.9433638443935927e-05,
      "loss": 1.935,
      "step": 132
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.302982330322266,
      "learning_rate": 2.9429347826086956e-05,
      "loss": 1.9848,
      "step": 133
    },
    {
      "epoch": 0.04,
      "grad_norm": 11.062186241149902,
      "learning_rate": 2.9425057208237988e-05,
      "loss": 1.8631,
      "step": 134
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.099900245666504,
      "learning_rate": 2.9420766590389017e-05,
      "loss": 1.9532,
      "step": 135
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.016757011413574,
      "learning_rate": 2.9416475972540045e-05,
      "loss": 2.2666,
      "step": 136
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.943782806396484,
      "learning_rate": 2.9412185354691074e-05,
      "loss": 1.8332,
      "step": 137
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.009693145751953,
      "learning_rate": 2.9407894736842106e-05,
      "loss": 1.7901,
      "step": 138
    },
    {
      "epoch": 0.04,
      "grad_norm": 8.308258056640625,
      "learning_rate": 2.940360411899314e-05,
      "loss": 1.5903,
      "step": 139
    },
    {
      "epoch": 0.04,
      "grad_norm": 8.762723922729492,
      "learning_rate": 2.9399313501144164e-05,
      "loss": 1.933,
      "step": 140
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.699069023132324,
      "learning_rate": 2.9395022883295196e-05,
      "loss": 1.631,
      "step": 141
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.041118621826172,
      "learning_rate": 2.9390732265446225e-05,
      "loss": 1.4745,
      "step": 142
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.30391788482666,
      "learning_rate": 2.9386441647597257e-05,
      "loss": 1.8366,
      "step": 143
    },
    {
      "epoch": 0.04,
      "grad_norm": 12.04333209991455,
      "learning_rate": 2.9382151029748282e-05,
      "loss": 1.8336,
      "step": 144
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.09054946899414,
      "learning_rate": 2.9377860411899314e-05,
      "loss": 1.7866,
      "step": 145
    },
    {
      "epoch": 0.04,
      "grad_norm": 13.490518569946289,
      "learning_rate": 2.9373569794050347e-05,
      "loss": 2.1837,
      "step": 146
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.580062866210938,
      "learning_rate": 2.9369279176201372e-05,
      "loss": 1.9534,
      "step": 147
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.711172103881836,
      "learning_rate": 2.9364988558352404e-05,
      "loss": 1.7171,
      "step": 148
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.504125595092773,
      "learning_rate": 2.9360697940503433e-05,
      "loss": 1.5971,
      "step": 149
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.39430046081543,
      "learning_rate": 2.9356407322654465e-05,
      "loss": 1.8896,
      "step": 150
    },
    {
      "epoch": 0.04,
      "grad_norm": 8.143174171447754,
      "learning_rate": 2.935211670480549e-05,
      "loss": 1.7462,
      "step": 151
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.526236534118652,
      "learning_rate": 2.9347826086956523e-05,
      "loss": 2.1149,
      "step": 152
    },
    {
      "epoch": 0.04,
      "grad_norm": 8.047273635864258,
      "learning_rate": 2.934353546910755e-05,
      "loss": 1.7804,
      "step": 153
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.352036476135254,
      "learning_rate": 2.9339244851258583e-05,
      "loss": 1.6782,
      "step": 154
    },
    {
      "epoch": 0.04,
      "grad_norm": 8.56297492980957,
      "learning_rate": 2.9334954233409612e-05,
      "loss": 1.9109,
      "step": 155
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.026894569396973,
      "learning_rate": 2.933066361556064e-05,
      "loss": 1.9137,
      "step": 156
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.315428733825684,
      "learning_rate": 2.9326372997711673e-05,
      "loss": 1.9549,
      "step": 157
    },
    {
      "epoch": 0.05,
      "grad_norm": 10.20797348022461,
      "learning_rate": 2.93220823798627e-05,
      "loss": 1.947,
      "step": 158
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.273527145385742,
      "learning_rate": 2.931779176201373e-05,
      "loss": 1.8877,
      "step": 159
    },
    {
      "epoch": 0.05,
      "grad_norm": 10.411182403564453,
      "learning_rate": 2.931350114416476e-05,
      "loss": 2.0504,
      "step": 160
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.119146347045898,
      "learning_rate": 2.930921052631579e-05,
      "loss": 1.782,
      "step": 161
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.15261173248291,
      "learning_rate": 2.9304919908466817e-05,
      "loss": 1.5476,
      "step": 162
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.623429298400879,
      "learning_rate": 2.930062929061785e-05,
      "loss": 1.5317,
      "step": 163
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.154732704162598,
      "learning_rate": 2.929633867276888e-05,
      "loss": 1.6405,
      "step": 164
    },
    {
      "epoch": 0.05,
      "grad_norm": 11.059431076049805,
      "learning_rate": 2.929204805491991e-05,
      "loss": 1.8032,
      "step": 165
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.126934051513672,
      "learning_rate": 2.928775743707094e-05,
      "loss": 1.7491,
      "step": 166
    },
    {
      "epoch": 0.05,
      "grad_norm": 10.09865665435791,
      "learning_rate": 2.9283466819221968e-05,
      "loss": 1.8666,
      "step": 167
    },
    {
      "epoch": 0.05,
      "grad_norm": 10.668212890625,
      "learning_rate": 2.9279176201373e-05,
      "loss": 1.6601,
      "step": 168
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.207748413085938,
      "learning_rate": 2.927488558352403e-05,
      "loss": 1.6901,
      "step": 169
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.090766906738281,
      "learning_rate": 2.9270594965675057e-05,
      "loss": 1.809,
      "step": 170
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.604634284973145,
      "learning_rate": 2.9266304347826086e-05,
      "loss": 1.5706,
      "step": 171
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.801182746887207,
      "learning_rate": 2.9262013729977118e-05,
      "loss": 1.6983,
      "step": 172
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.632198333740234,
      "learning_rate": 2.9257723112128147e-05,
      "loss": 1.6291,
      "step": 173
    },
    {
      "epoch": 0.05,
      "grad_norm": 7.52984094619751,
      "learning_rate": 2.9253432494279176e-05,
      "loss": 1.6256,
      "step": 174
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.251009941101074,
      "learning_rate": 2.9249141876430208e-05,
      "loss": 1.4832,
      "step": 175
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.106322288513184,
      "learning_rate": 2.9244851258581237e-05,
      "loss": 1.7677,
      "step": 176
    },
    {
      "epoch": 0.05,
      "grad_norm": 10.289727210998535,
      "learning_rate": 2.9240560640732265e-05,
      "loss": 1.3136,
      "step": 177
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.70762825012207,
      "learning_rate": 2.9236270022883294e-05,
      "loss": 1.9693,
      "step": 178
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.674822807312012,
      "learning_rate": 2.9231979405034326e-05,
      "loss": 1.9052,
      "step": 179
    },
    {
      "epoch": 0.05,
      "grad_norm": 7.9198899269104,
      "learning_rate": 2.922768878718536e-05,
      "loss": 1.299,
      "step": 180
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.938362121582031,
      "learning_rate": 2.9223398169336384e-05,
      "loss": 1.724,
      "step": 181
    },
    {
      "epoch": 0.05,
      "grad_norm": 10.080430030822754,
      "learning_rate": 2.9219107551487416e-05,
      "loss": 1.8977,
      "step": 182
    },
    {
      "epoch": 0.05,
      "grad_norm": 11.959859848022461,
      "learning_rate": 2.9214816933638445e-05,
      "loss": 2.1547,
      "step": 183
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.7039794921875,
      "learning_rate": 2.9210526315789474e-05,
      "loss": 2.1874,
      "step": 184
    },
    {
      "epoch": 0.05,
      "grad_norm": 11.266067504882812,
      "learning_rate": 2.9206235697940502e-05,
      "loss": 1.5548,
      "step": 185
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.631067276000977,
      "learning_rate": 2.9201945080091535e-05,
      "loss": 1.9972,
      "step": 186
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.844793319702148,
      "learning_rate": 2.9197654462242563e-05,
      "loss": 1.6902,
      "step": 187
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.674718856811523,
      "learning_rate": 2.9193363844393592e-05,
      "loss": 1.5891,
      "step": 188
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.092727661132812,
      "learning_rate": 2.9189073226544624e-05,
      "loss": 1.8592,
      "step": 189
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.948922157287598,
      "learning_rate": 2.9184782608695653e-05,
      "loss": 1.5874,
      "step": 190
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.707277297973633,
      "learning_rate": 2.9180491990846685e-05,
      "loss": 2.122,
      "step": 191
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.720427513122559,
      "learning_rate": 2.917620137299771e-05,
      "loss": 1.9793,
      "step": 192
    },
    {
      "epoch": 0.06,
      "grad_norm": 10.036628723144531,
      "learning_rate": 2.9171910755148743e-05,
      "loss": 2.1604,
      "step": 193
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.703857421875,
      "learning_rate": 2.916762013729977e-05,
      "loss": 1.735,
      "step": 194
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.382275581359863,
      "learning_rate": 2.9163329519450804e-05,
      "loss": 1.3775,
      "step": 195
    },
    {
      "epoch": 0.06,
      "grad_norm": 10.020468711853027,
      "learning_rate": 2.915903890160183e-05,
      "loss": 1.5763,
      "step": 196
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.866108894348145,
      "learning_rate": 2.915474828375286e-05,
      "loss": 1.6297,
      "step": 197
    },
    {
      "epoch": 0.06,
      "grad_norm": 7.478298664093018,
      "learning_rate": 2.9150457665903893e-05,
      "loss": 1.4512,
      "step": 198
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.695619583129883,
      "learning_rate": 2.914616704805492e-05,
      "loss": 2.2167,
      "step": 199
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.859832763671875,
      "learning_rate": 2.914187643020595e-05,
      "loss": 1.7297,
      "step": 200
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.221556663513184,
      "learning_rate": 2.913758581235698e-05,
      "loss": 1.4514,
      "step": 201
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.9334135055542,
      "learning_rate": 2.9133295194508012e-05,
      "loss": 1.8826,
      "step": 202
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.83910846710205,
      "learning_rate": 2.9129004576659037e-05,
      "loss": 1.6077,
      "step": 203
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.432626724243164,
      "learning_rate": 2.912471395881007e-05,
      "loss": 1.4681,
      "step": 204
    },
    {
      "epoch": 0.06,
      "grad_norm": 10.136731147766113,
      "learning_rate": 2.9120423340961098e-05,
      "loss": 1.6358,
      "step": 205
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.665613174438477,
      "learning_rate": 2.911613272311213e-05,
      "loss": 1.7486,
      "step": 206
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.559819221496582,
      "learning_rate": 2.911184210526316e-05,
      "loss": 1.5329,
      "step": 207
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.54000473022461,
      "learning_rate": 2.9107551487414188e-05,
      "loss": 1.7634,
      "step": 208
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.584062576293945,
      "learning_rate": 2.910326086956522e-05,
      "loss": 1.6295,
      "step": 209
    },
    {
      "epoch": 0.06,
      "grad_norm": 15.094372749328613,
      "learning_rate": 2.909897025171625e-05,
      "loss": 1.7466,
      "step": 210
    },
    {
      "epoch": 0.06,
      "grad_norm": 11.30350112915039,
      "learning_rate": 2.9094679633867277e-05,
      "loss": 1.8303,
      "step": 211
    },
    {
      "epoch": 0.06,
      "grad_norm": 11.015275955200195,
      "learning_rate": 2.9090389016018306e-05,
      "loss": 2.1698,
      "step": 212
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.419334411621094,
      "learning_rate": 2.908609839816934e-05,
      "loss": 1.6318,
      "step": 213
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.208695411682129,
      "learning_rate": 2.9081807780320364e-05,
      "loss": 1.4457,
      "step": 214
    },
    {
      "epoch": 0.06,
      "grad_norm": 11.84401798248291,
      "learning_rate": 2.9077517162471396e-05,
      "loss": 1.888,
      "step": 215
    },
    {
      "epoch": 0.06,
      "grad_norm": 10.923035621643066,
      "learning_rate": 2.9073226544622428e-05,
      "loss": 1.5284,
      "step": 216
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.151190757751465,
      "learning_rate": 2.9068935926773457e-05,
      "loss": 1.7259,
      "step": 217
    },
    {
      "epoch": 0.06,
      "grad_norm": 10.143967628479004,
      "learning_rate": 2.9064645308924486e-05,
      "loss": 1.5389,
      "step": 218
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.866828918457031,
      "learning_rate": 2.9060354691075514e-05,
      "loss": 1.7653,
      "step": 219
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.750387191772461,
      "learning_rate": 2.9056064073226547e-05,
      "loss": 1.9439,
      "step": 220
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.95814323425293,
      "learning_rate": 2.9051773455377575e-05,
      "loss": 1.7061,
      "step": 221
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.123912811279297,
      "learning_rate": 2.9047482837528604e-05,
      "loss": 1.8644,
      "step": 222
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.334152221679688,
      "learning_rate": 2.9043192219679636e-05,
      "loss": 1.7821,
      "step": 223
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.544561386108398,
      "learning_rate": 2.9038901601830665e-05,
      "loss": 1.6721,
      "step": 224
    },
    {
      "epoch": 0.06,
      "grad_norm": 10.224461555480957,
      "learning_rate": 2.9034610983981694e-05,
      "loss": 1.3896,
      "step": 225
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.83475112915039,
      "learning_rate": 2.9030320366132723e-05,
      "loss": 1.8461,
      "step": 226
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.36566162109375,
      "learning_rate": 2.9026029748283755e-05,
      "loss": 1.8513,
      "step": 227
    },
    {
      "epoch": 0.07,
      "grad_norm": 11.508695602416992,
      "learning_rate": 2.9021739130434783e-05,
      "loss": 1.4862,
      "step": 228
    },
    {
      "epoch": 0.07,
      "grad_norm": 12.102386474609375,
      "learning_rate": 2.9017448512585812e-05,
      "loss": 1.9282,
      "step": 229
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.708913803100586,
      "learning_rate": 2.901315789473684e-05,
      "loss": 1.4771,
      "step": 230
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.158058166503906,
      "learning_rate": 2.9008867276887873e-05,
      "loss": 1.5567,
      "step": 231
    },
    {
      "epoch": 0.07,
      "grad_norm": 7.920383930206299,
      "learning_rate": 2.9004576659038905e-05,
      "loss": 1.7236,
      "step": 232
    },
    {
      "epoch": 0.07,
      "grad_norm": 10.36752986907959,
      "learning_rate": 2.900028604118993e-05,
      "loss": 1.3592,
      "step": 233
    },
    {
      "epoch": 0.07,
      "grad_norm": 11.605936050415039,
      "learning_rate": 2.8995995423340963e-05,
      "loss": 1.6096,
      "step": 234
    },
    {
      "epoch": 0.07,
      "grad_norm": 11.090018272399902,
      "learning_rate": 2.899170480549199e-05,
      "loss": 1.9197,
      "step": 235
    },
    {
      "epoch": 0.07,
      "grad_norm": 11.563491821289062,
      "learning_rate": 2.8987414187643024e-05,
      "loss": 1.7927,
      "step": 236
    },
    {
      "epoch": 0.07,
      "grad_norm": 11.237104415893555,
      "learning_rate": 2.898312356979405e-05,
      "loss": 2.0423,
      "step": 237
    },
    {
      "epoch": 0.07,
      "grad_norm": 12.092026710510254,
      "learning_rate": 2.897883295194508e-05,
      "loss": 1.6959,
      "step": 238
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.514479637145996,
      "learning_rate": 2.897454233409611e-05,
      "loss": 1.4793,
      "step": 239
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.548802375793457,
      "learning_rate": 2.897025171624714e-05,
      "loss": 1.7667,
      "step": 240
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.590652465820312,
      "learning_rate": 2.896596109839817e-05,
      "loss": 1.9237,
      "step": 241
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.465312957763672,
      "learning_rate": 2.89616704805492e-05,
      "loss": 1.4994,
      "step": 242
    },
    {
      "epoch": 0.07,
      "grad_norm": 11.648931503295898,
      "learning_rate": 2.8957379862700232e-05,
      "loss": 2.0277,
      "step": 243
    },
    {
      "epoch": 0.07,
      "grad_norm": 10.328200340270996,
      "learning_rate": 2.8953089244851257e-05,
      "loss": 1.9486,
      "step": 244
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.387717247009277,
      "learning_rate": 2.894879862700229e-05,
      "loss": 1.6182,
      "step": 245
    },
    {
      "epoch": 0.07,
      "grad_norm": 10.021469116210938,
      "learning_rate": 2.8944508009153318e-05,
      "loss": 1.6038,
      "step": 246
    },
    {
      "epoch": 0.07,
      "grad_norm": 14.117218017578125,
      "learning_rate": 2.894021739130435e-05,
      "loss": 1.7402,
      "step": 247
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.381802558898926,
      "learning_rate": 2.8935926773455376e-05,
      "loss": 1.7406,
      "step": 248
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.9871826171875,
      "learning_rate": 2.8931636155606408e-05,
      "loss": 1.7055,
      "step": 249
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.729340553283691,
      "learning_rate": 2.892734553775744e-05,
      "loss": 1.6221,
      "step": 250
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.504460334777832,
      "learning_rate": 2.8923054919908465e-05,
      "loss": 1.7229,
      "step": 251
    },
    {
      "epoch": 0.07,
      "grad_norm": 7.916970252990723,
      "learning_rate": 2.8918764302059498e-05,
      "loss": 1.6216,
      "step": 252
    },
    {
      "epoch": 0.07,
      "grad_norm": 12.042543411254883,
      "learning_rate": 2.8914473684210526e-05,
      "loss": 2.0739,
      "step": 253
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.228246688842773,
      "learning_rate": 2.891018306636156e-05,
      "loss": 1.5692,
      "step": 254
    },
    {
      "epoch": 0.07,
      "grad_norm": 10.877127647399902,
      "learning_rate": 2.8905892448512584e-05,
      "loss": 1.7809,
      "step": 255
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.196961402893066,
      "learning_rate": 2.8901601830663616e-05,
      "loss": 1.421,
      "step": 256
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.786067962646484,
      "learning_rate": 2.8897311212814645e-05,
      "loss": 2.0212,
      "step": 257
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.010289192199707,
      "learning_rate": 2.8893020594965677e-05,
      "loss": 1.58,
      "step": 258
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.008468627929688,
      "learning_rate": 2.8888729977116706e-05,
      "loss": 1.931,
      "step": 259
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.551482200622559,
      "learning_rate": 2.8884439359267735e-05,
      "loss": 1.5659,
      "step": 260
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.659241676330566,
      "learning_rate": 2.8880148741418767e-05,
      "loss": 1.8359,
      "step": 261
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.562986373901367,
      "learning_rate": 2.8875858123569795e-05,
      "loss": 1.5877,
      "step": 262
    },
    {
      "epoch": 0.08,
      "grad_norm": 7.784165382385254,
      "learning_rate": 2.8871567505720824e-05,
      "loss": 1.4274,
      "step": 263
    },
    {
      "epoch": 0.08,
      "grad_norm": 10.632110595703125,
      "learning_rate": 2.8867276887871853e-05,
      "loss": 1.8839,
      "step": 264
    },
    {
      "epoch": 0.08,
      "grad_norm": 7.901241779327393,
      "learning_rate": 2.8862986270022885e-05,
      "loss": 1.3047,
      "step": 265
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.041234970092773,
      "learning_rate": 2.8858695652173914e-05,
      "loss": 1.5787,
      "step": 266
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.147648811340332,
      "learning_rate": 2.8854405034324943e-05,
      "loss": 1.3792,
      "step": 267
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.332252502441406,
      "learning_rate": 2.8850114416475975e-05,
      "loss": 2.0235,
      "step": 268
    },
    {
      "epoch": 0.08,
      "grad_norm": 7.136639595031738,
      "learning_rate": 2.8845823798627004e-05,
      "loss": 1.5239,
      "step": 269
    },
    {
      "epoch": 0.08,
      "grad_norm": 21.16472053527832,
      "learning_rate": 2.8841533180778032e-05,
      "loss": 1.4307,
      "step": 270
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.72244930267334,
      "learning_rate": 2.883724256292906e-05,
      "loss": 1.7088,
      "step": 271
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.103704452514648,
      "learning_rate": 2.8832951945080093e-05,
      "loss": 1.8197,
      "step": 272
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.807337760925293,
      "learning_rate": 2.8828661327231122e-05,
      "loss": 1.8819,
      "step": 273
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.339899063110352,
      "learning_rate": 2.882437070938215e-05,
      "loss": 1.7452,
      "step": 274
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.583975791931152,
      "learning_rate": 2.8820080091533183e-05,
      "loss": 2.0151,
      "step": 275
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.562725067138672,
      "learning_rate": 2.8815789473684212e-05,
      "loss": 1.3846,
      "step": 276
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.248943328857422,
      "learning_rate": 2.881149885583524e-05,
      "loss": 1.6616,
      "step": 277
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.356354713439941,
      "learning_rate": 2.880720823798627e-05,
      "loss": 1.9607,
      "step": 278
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.86824893951416,
      "learning_rate": 2.88029176201373e-05,
      "loss": 1.5309,
      "step": 279
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.779897689819336,
      "learning_rate": 2.879862700228833e-05,
      "loss": 1.8248,
      "step": 280
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.615009307861328,
      "learning_rate": 2.879433638443936e-05,
      "loss": 1.8228,
      "step": 281
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.457856178283691,
      "learning_rate": 2.8790045766590388e-05,
      "loss": 1.4264,
      "step": 282
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.006612777709961,
      "learning_rate": 2.878575514874142e-05,
      "loss": 1.1521,
      "step": 283
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.314384460449219,
      "learning_rate": 2.8781464530892452e-05,
      "loss": 1.5094,
      "step": 284
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.521162033081055,
      "learning_rate": 2.8777173913043477e-05,
      "loss": 1.5872,
      "step": 285
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.152196884155273,
      "learning_rate": 2.877288329519451e-05,
      "loss": 1.8608,
      "step": 286
    },
    {
      "epoch": 0.08,
      "grad_norm": 10.426616668701172,
      "learning_rate": 2.876859267734554e-05,
      "loss": 1.3441,
      "step": 287
    },
    {
      "epoch": 0.08,
      "grad_norm": 10.633772850036621,
      "learning_rate": 2.876430205949657e-05,
      "loss": 2.0061,
      "step": 288
    },
    {
      "epoch": 0.08,
      "grad_norm": 11.46512508392334,
      "learning_rate": 2.8760011441647596e-05,
      "loss": 1.9394,
      "step": 289
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.515049934387207,
      "learning_rate": 2.8755720823798628e-05,
      "loss": 1.1297,
      "step": 290
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.623251914978027,
      "learning_rate": 2.8751430205949657e-05,
      "loss": 1.7703,
      "step": 291
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.966745376586914,
      "learning_rate": 2.8747139588100686e-05,
      "loss": 1.7564,
      "step": 292
    },
    {
      "epoch": 0.08,
      "grad_norm": 10.016712188720703,
      "learning_rate": 2.8742848970251718e-05,
      "loss": 1.4858,
      "step": 293
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.469600677490234,
      "learning_rate": 2.8738558352402747e-05,
      "loss": 1.7119,
      "step": 294
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.720199584960938,
      "learning_rate": 2.873426773455378e-05,
      "loss": 1.6103,
      "step": 295
    },
    {
      "epoch": 0.08,
      "grad_norm": 11.37694263458252,
      "learning_rate": 2.8729977116704804e-05,
      "loss": 1.9161,
      "step": 296
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.98237419128418,
      "learning_rate": 2.8725686498855836e-05,
      "loss": 1.5475,
      "step": 297
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.324840545654297,
      "learning_rate": 2.8721395881006865e-05,
      "loss": 1.4995,
      "step": 298
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.735411643981934,
      "learning_rate": 2.8717105263157897e-05,
      "loss": 1.3443,
      "step": 299
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.82810115814209,
      "learning_rate": 2.8712814645308926e-05,
      "loss": 1.7164,
      "step": 300
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.84345817565918,
      "learning_rate": 2.8708524027459955e-05,
      "loss": 1.4979,
      "step": 301
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.750595092773438,
      "learning_rate": 2.8704233409610987e-05,
      "loss": 1.506,
      "step": 302
    },
    {
      "epoch": 0.09,
      "grad_norm": 10.426441192626953,
      "learning_rate": 2.8699942791762012e-05,
      "loss": 1.6819,
      "step": 303
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.954815864562988,
      "learning_rate": 2.8695652173913044e-05,
      "loss": 1.628,
      "step": 304
    },
    {
      "epoch": 0.09,
      "grad_norm": 10.592462539672852,
      "learning_rate": 2.8691361556064073e-05,
      "loss": 1.6014,
      "step": 305
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.672579765319824,
      "learning_rate": 2.8687070938215105e-05,
      "loss": 1.7661,
      "step": 306
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.73773193359375,
      "learning_rate": 2.868278032036613e-05,
      "loss": 1.7942,
      "step": 307
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.554476737976074,
      "learning_rate": 2.8678489702517163e-05,
      "loss": 1.2383,
      "step": 308
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.73338508605957,
      "learning_rate": 2.8674199084668195e-05,
      "loss": 1.3876,
      "step": 309
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.714142799377441,
      "learning_rate": 2.8669908466819224e-05,
      "loss": 1.3542,
      "step": 310
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.723358154296875,
      "learning_rate": 2.8665617848970253e-05,
      "loss": 1.5286,
      "step": 311
    },
    {
      "epoch": 0.09,
      "grad_norm": 7.836288928985596,
      "learning_rate": 2.866132723112128e-05,
      "loss": 1.4391,
      "step": 312
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.699174880981445,
      "learning_rate": 2.8657036613272313e-05,
      "loss": 1.4725,
      "step": 313
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.351971626281738,
      "learning_rate": 2.8652745995423342e-05,
      "loss": 1.5725,
      "step": 314
    },
    {
      "epoch": 0.09,
      "grad_norm": 10.71751880645752,
      "learning_rate": 2.864845537757437e-05,
      "loss": 1.8646,
      "step": 315
    },
    {
      "epoch": 0.09,
      "grad_norm": 10.017102241516113,
      "learning_rate": 2.86441647597254e-05,
      "loss": 1.8136,
      "step": 316
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.702308654785156,
      "learning_rate": 2.8639874141876432e-05,
      "loss": 1.6153,
      "step": 317
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.080348014831543,
      "learning_rate": 2.863558352402746e-05,
      "loss": 1.6787,
      "step": 318
    },
    {
      "epoch": 0.09,
      "grad_norm": 10.15158462524414,
      "learning_rate": 2.863129290617849e-05,
      "loss": 1.8533,
      "step": 319
    },
    {
      "epoch": 0.09,
      "grad_norm": 11.515812873840332,
      "learning_rate": 2.862700228832952e-05,
      "loss": 1.6914,
      "step": 320
    },
    {
      "epoch": 0.09,
      "grad_norm": 10.957345962524414,
      "learning_rate": 2.862271167048055e-05,
      "loss": 1.8418,
      "step": 321
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.494919776916504,
      "learning_rate": 2.861842105263158e-05,
      "loss": 1.6915,
      "step": 322
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.075895309448242,
      "learning_rate": 2.8614130434782608e-05,
      "loss": 1.4509,
      "step": 323
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.955622673034668,
      "learning_rate": 2.860983981693364e-05,
      "loss": 1.4193,
      "step": 324
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.788787841796875,
      "learning_rate": 2.860554919908467e-05,
      "loss": 1.5942,
      "step": 325
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.214378356933594,
      "learning_rate": 2.8601258581235698e-05,
      "loss": 1.6611,
      "step": 326
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.840919494628906,
      "learning_rate": 2.859696796338673e-05,
      "loss": 1.3909,
      "step": 327
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.53697681427002,
      "learning_rate": 2.859267734553776e-05,
      "loss": 1.5233,
      "step": 328
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.874279975891113,
      "learning_rate": 2.8588386727688787e-05,
      "loss": 1.8105,
      "step": 329
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.041751861572266,
      "learning_rate": 2.8584096109839816e-05,
      "loss": 1.6273,
      "step": 330
    },
    {
      "epoch": 0.09,
      "grad_norm": 10.59816837310791,
      "learning_rate": 2.8579805491990848e-05,
      "loss": 1.6535,
      "step": 331
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.560877799987793,
      "learning_rate": 2.8575514874141877e-05,
      "loss": 1.5892,
      "step": 332
    },
    {
      "epoch": 0.1,
      "grad_norm": 10.164031028747559,
      "learning_rate": 2.8571224256292906e-05,
      "loss": 1.7095,
      "step": 333
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.156442642211914,
      "learning_rate": 2.8566933638443934e-05,
      "loss": 1.4494,
      "step": 334
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.107510566711426,
      "learning_rate": 2.8562643020594967e-05,
      "loss": 1.6335,
      "step": 335
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.041921615600586,
      "learning_rate": 2.8558352402746e-05,
      "loss": 1.3935,
      "step": 336
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.365302085876465,
      "learning_rate": 2.8554061784897024e-05,
      "loss": 1.5283,
      "step": 337
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.143136024475098,
      "learning_rate": 2.8549771167048056e-05,
      "loss": 1.3668,
      "step": 338
    },
    {
      "epoch": 0.1,
      "grad_norm": 12.083857536315918,
      "learning_rate": 2.8545480549199085e-05,
      "loss": 2.0169,
      "step": 339
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.328901290893555,
      "learning_rate": 2.8541189931350117e-05,
      "loss": 1.3423,
      "step": 340
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.404691696166992,
      "learning_rate": 2.8536899313501143e-05,
      "loss": 1.5616,
      "step": 341
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.255928993225098,
      "learning_rate": 2.8532608695652175e-05,
      "loss": 1.4001,
      "step": 342
    },
    {
      "epoch": 0.1,
      "grad_norm": 13.580086708068848,
      "learning_rate": 2.8528318077803207e-05,
      "loss": 1.5508,
      "step": 343
    },
    {
      "epoch": 0.1,
      "grad_norm": 10.312821388244629,
      "learning_rate": 2.8524027459954232e-05,
      "loss": 2.1192,
      "step": 344
    },
    {
      "epoch": 0.1,
      "grad_norm": 11.666476249694824,
      "learning_rate": 2.8519736842105264e-05,
      "loss": 1.8555,
      "step": 345
    },
    {
      "epoch": 0.1,
      "grad_norm": 11.251701354980469,
      "learning_rate": 2.8515446224256293e-05,
      "loss": 1.9988,
      "step": 346
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.151383399963379,
      "learning_rate": 2.8511155606407325e-05,
      "loss": 1.602,
      "step": 347
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.019693374633789,
      "learning_rate": 2.850686498855835e-05,
      "loss": 1.6737,
      "step": 348
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.890886306762695,
      "learning_rate": 2.8502574370709383e-05,
      "loss": 1.5644,
      "step": 349
    },
    {
      "epoch": 0.1,
      "grad_norm": 7.539163112640381,
      "learning_rate": 2.8498283752860412e-05,
      "loss": 1.6877,
      "step": 350
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.853219032287598,
      "learning_rate": 2.8493993135011444e-05,
      "loss": 1.6493,
      "step": 351
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.162408828735352,
      "learning_rate": 2.8489702517162473e-05,
      "loss": 1.8122,
      "step": 352
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.233209609985352,
      "learning_rate": 2.84854118993135e-05,
      "loss": 1.7432,
      "step": 353
    },
    {
      "epoch": 0.1,
      "grad_norm": 7.457662105560303,
      "learning_rate": 2.8481121281464534e-05,
      "loss": 1.689,
      "step": 354
    },
    {
      "epoch": 0.1,
      "grad_norm": 7.5489397048950195,
      "learning_rate": 2.8476830663615562e-05,
      "loss": 1.6797,
      "step": 355
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.018563270568848,
      "learning_rate": 2.847254004576659e-05,
      "loss": 1.7935,
      "step": 356
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.723430633544922,
      "learning_rate": 2.846824942791762e-05,
      "loss": 1.6617,
      "step": 357
    },
    {
      "epoch": 0.1,
      "grad_norm": 7.9907636642456055,
      "learning_rate": 2.8463958810068652e-05,
      "loss": 1.7775,
      "step": 358
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.556412696838379,
      "learning_rate": 2.8459668192219677e-05,
      "loss": 1.5166,
      "step": 359
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.768579483032227,
      "learning_rate": 2.845537757437071e-05,
      "loss": 1.4092,
      "step": 360
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.49852466583252,
      "learning_rate": 2.8451086956521742e-05,
      "loss": 1.5011,
      "step": 361
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.627189636230469,
      "learning_rate": 2.844679633867277e-05,
      "loss": 1.5046,
      "step": 362
    },
    {
      "epoch": 0.1,
      "grad_norm": 11.643542289733887,
      "learning_rate": 2.84425057208238e-05,
      "loss": 2.18,
      "step": 363
    },
    {
      "epoch": 0.1,
      "grad_norm": 10.84292221069336,
      "learning_rate": 2.8438215102974828e-05,
      "loss": 2.0504,
      "step": 364
    },
    {
      "epoch": 0.1,
      "grad_norm": 10.566253662109375,
      "learning_rate": 2.843392448512586e-05,
      "loss": 1.7854,
      "step": 365
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.069684028625488,
      "learning_rate": 2.842963386727689e-05,
      "loss": 1.7755,
      "step": 366
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.103590965270996,
      "learning_rate": 2.8425343249427918e-05,
      "loss": 1.5551,
      "step": 367
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.507101058959961,
      "learning_rate": 2.8421052631578946e-05,
      "loss": 1.3767,
      "step": 368
    },
    {
      "epoch": 0.11,
      "grad_norm": 10.058374404907227,
      "learning_rate": 2.841676201372998e-05,
      "loss": 1.5317,
      "step": 369
    },
    {
      "epoch": 0.11,
      "grad_norm": 11.702919960021973,
      "learning_rate": 2.8412471395881007e-05,
      "loss": 1.9569,
      "step": 370
    },
    {
      "epoch": 0.11,
      "grad_norm": 10.480443954467773,
      "learning_rate": 2.8408180778032036e-05,
      "loss": 1.8785,
      "step": 371
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.140588760375977,
      "learning_rate": 2.840389016018307e-05,
      "loss": 1.6059,
      "step": 372
    },
    {
      "epoch": 0.11,
      "grad_norm": 7.487027168273926,
      "learning_rate": 2.8399599542334097e-05,
      "loss": 1.6194,
      "step": 373
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.53647518157959,
      "learning_rate": 2.8395308924485126e-05,
      "loss": 1.826,
      "step": 374
    },
    {
      "epoch": 0.11,
      "grad_norm": 7.119506359100342,
      "learning_rate": 2.8391018306636155e-05,
      "loss": 1.5237,
      "step": 375
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.412140846252441,
      "learning_rate": 2.8386727688787187e-05,
      "loss": 1.2996,
      "step": 376
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.240374565124512,
      "learning_rate": 2.8382437070938216e-05,
      "loss": 1.8373,
      "step": 377
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.49670124053955,
      "learning_rate": 2.8378146453089244e-05,
      "loss": 1.649,
      "step": 378
    },
    {
      "epoch": 0.11,
      "grad_norm": 10.530879974365234,
      "learning_rate": 2.8373855835240276e-05,
      "loss": 1.9523,
      "step": 379
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.252795219421387,
      "learning_rate": 2.8369565217391305e-05,
      "loss": 1.3662,
      "step": 380
    },
    {
      "epoch": 0.11,
      "grad_norm": 10.121156692504883,
      "learning_rate": 2.8365274599542334e-05,
      "loss": 1.8007,
      "step": 381
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.110308647155762,
      "learning_rate": 2.8360983981693363e-05,
      "loss": 1.6888,
      "step": 382
    },
    {
      "epoch": 0.11,
      "grad_norm": 13.776968955993652,
      "learning_rate": 2.8356693363844395e-05,
      "loss": 1.5352,
      "step": 383
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.49256706237793,
      "learning_rate": 2.8352402745995424e-05,
      "loss": 1.5955,
      "step": 384
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.58328914642334,
      "learning_rate": 2.8348112128146452e-05,
      "loss": 1.6714,
      "step": 385
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.3377103805542,
      "learning_rate": 2.8343821510297485e-05,
      "loss": 1.6089,
      "step": 386
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.08781623840332,
      "learning_rate": 2.8339530892448513e-05,
      "loss": 1.3559,
      "step": 387
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.012792587280273,
      "learning_rate": 2.8335240274599546e-05,
      "loss": 1.5669,
      "step": 388
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.003541946411133,
      "learning_rate": 2.833094965675057e-05,
      "loss": 1.7334,
      "step": 389
    },
    {
      "epoch": 0.11,
      "grad_norm": 7.730262279510498,
      "learning_rate": 2.8326659038901603e-05,
      "loss": 1.648,
      "step": 390
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.558412551879883,
      "learning_rate": 2.8322368421052632e-05,
      "loss": 1.681,
      "step": 391
    },
    {
      "epoch": 0.11,
      "grad_norm": 7.935482978820801,
      "learning_rate": 2.8318077803203664e-05,
      "loss": 1.5282,
      "step": 392
    },
    {
      "epoch": 0.11,
      "grad_norm": 7.9830098152160645,
      "learning_rate": 2.831378718535469e-05,
      "loss": 1.7982,
      "step": 393
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.142667770385742,
      "learning_rate": 2.830949656750572e-05,
      "loss": 1.767,
      "step": 394
    },
    {
      "epoch": 0.11,
      "grad_norm": 7.926479339599609,
      "learning_rate": 2.8305205949656754e-05,
      "loss": 1.6133,
      "step": 395
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.327887535095215,
      "learning_rate": 2.830091533180778e-05,
      "loss": 1.3836,
      "step": 396
    },
    {
      "epoch": 0.11,
      "grad_norm": 7.93806266784668,
      "learning_rate": 2.829662471395881e-05,
      "loss": 1.5477,
      "step": 397
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.302078247070312,
      "learning_rate": 2.829233409610984e-05,
      "loss": 1.6192,
      "step": 398
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.20541000366211,
      "learning_rate": 2.8288043478260872e-05,
      "loss": 1.7687,
      "step": 399
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.753988265991211,
      "learning_rate": 2.8283752860411898e-05,
      "loss": 1.6086,
      "step": 400
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.764487266540527,
      "learning_rate": 2.827946224256293e-05,
      "loss": 1.6797,
      "step": 401
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.942675590515137,
      "learning_rate": 2.827517162471396e-05,
      "loss": 1.7611,
      "step": 402
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.709769248962402,
      "learning_rate": 2.827088100686499e-05,
      "loss": 1.6783,
      "step": 403
    },
    {
      "epoch": 0.12,
      "grad_norm": 12.350006103515625,
      "learning_rate": 2.826659038901602e-05,
      "loss": 1.5502,
      "step": 404
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.543091773986816,
      "learning_rate": 2.8262299771167048e-05,
      "loss": 1.5901,
      "step": 405
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.05308723449707,
      "learning_rate": 2.825800915331808e-05,
      "loss": 1.5929,
      "step": 406
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.204638481140137,
      "learning_rate": 2.825371853546911e-05,
      "loss": 1.4633,
      "step": 407
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.998212814331055,
      "learning_rate": 2.8249427917620138e-05,
      "loss": 1.7381,
      "step": 408
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.445399284362793,
      "learning_rate": 2.8245137299771167e-05,
      "loss": 1.8509,
      "step": 409
    },
    {
      "epoch": 0.12,
      "grad_norm": 7.8274736404418945,
      "learning_rate": 2.82408466819222e-05,
      "loss": 1.6,
      "step": 410
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.54792308807373,
      "learning_rate": 2.8236556064073224e-05,
      "loss": 1.5708,
      "step": 411
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.415521621704102,
      "learning_rate": 2.8232265446224256e-05,
      "loss": 1.4578,
      "step": 412
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.425575256347656,
      "learning_rate": 2.822797482837529e-05,
      "loss": 1.4646,
      "step": 413
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.011249542236328,
      "learning_rate": 2.8223684210526317e-05,
      "loss": 1.742,
      "step": 414
    },
    {
      "epoch": 0.12,
      "grad_norm": 7.653934478759766,
      "learning_rate": 2.8219393592677346e-05,
      "loss": 1.5293,
      "step": 415
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.901880264282227,
      "learning_rate": 2.8215102974828375e-05,
      "loss": 1.6321,
      "step": 416
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.286306381225586,
      "learning_rate": 2.8210812356979407e-05,
      "loss": 1.6388,
      "step": 417
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.622949600219727,
      "learning_rate": 2.8206521739130436e-05,
      "loss": 1.664,
      "step": 418
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.048681259155273,
      "learning_rate": 2.8202231121281464e-05,
      "loss": 1.4288,
      "step": 419
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.250710487365723,
      "learning_rate": 2.8197940503432497e-05,
      "loss": 1.5068,
      "step": 420
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.357898712158203,
      "learning_rate": 2.8193649885583525e-05,
      "loss": 1.489,
      "step": 421
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.524295806884766,
      "learning_rate": 2.8189359267734554e-05,
      "loss": 1.5555,
      "step": 422
    },
    {
      "epoch": 0.12,
      "grad_norm": 10.393868446350098,
      "learning_rate": 2.8185068649885583e-05,
      "loss": 1.8444,
      "step": 423
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.325084686279297,
      "learning_rate": 2.8180778032036615e-05,
      "loss": 1.4135,
      "step": 424
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.550622940063477,
      "learning_rate": 2.8176487414187644e-05,
      "loss": 1.4797,
      "step": 425
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.385437965393066,
      "learning_rate": 2.8172196796338673e-05,
      "loss": 1.6019,
      "step": 426
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.729028701782227,
      "learning_rate": 2.81679061784897e-05,
      "loss": 1.2063,
      "step": 427
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.351651191711426,
      "learning_rate": 2.8163615560640734e-05,
      "loss": 1.6706,
      "step": 428
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.507067680358887,
      "learning_rate": 2.8159324942791766e-05,
      "loss": 1.5326,
      "step": 429
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.760214805603027,
      "learning_rate": 2.815503432494279e-05,
      "loss": 1.6172,
      "step": 430
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.832893371582031,
      "learning_rate": 2.8150743707093823e-05,
      "loss": 1.4825,
      "step": 431
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.570629119873047,
      "learning_rate": 2.8146453089244852e-05,
      "loss": 1.4647,
      "step": 432
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.038146018981934,
      "learning_rate": 2.8142162471395884e-05,
      "loss": 1.4906,
      "step": 433
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.179692268371582,
      "learning_rate": 2.813787185354691e-05,
      "loss": 1.7981,
      "step": 434
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.589019775390625,
      "learning_rate": 2.8133581235697942e-05,
      "loss": 1.8932,
      "step": 435
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.438563346862793,
      "learning_rate": 2.812929061784897e-05,
      "loss": 1.5164,
      "step": 436
    },
    {
      "epoch": 0.12,
      "grad_norm": 7.712190628051758,
      "learning_rate": 2.8125e-05,
      "loss": 1.587,
      "step": 437
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.640191078186035,
      "learning_rate": 2.812070938215103e-05,
      "loss": 1.3491,
      "step": 438
    },
    {
      "epoch": 0.13,
      "grad_norm": 7.796834468841553,
      "learning_rate": 2.811641876430206e-05,
      "loss": 1.533,
      "step": 439
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.263230323791504,
      "learning_rate": 2.8112128146453092e-05,
      "loss": 1.4814,
      "step": 440
    },
    {
      "epoch": 0.13,
      "grad_norm": 7.591869354248047,
      "learning_rate": 2.8107837528604118e-05,
      "loss": 1.4957,
      "step": 441
    },
    {
      "epoch": 0.13,
      "grad_norm": 7.961359024047852,
      "learning_rate": 2.810354691075515e-05,
      "loss": 1.5566,
      "step": 442
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.94115924835205,
      "learning_rate": 2.809925629290618e-05,
      "loss": 1.4563,
      "step": 443
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.587944984436035,
      "learning_rate": 2.809496567505721e-05,
      "loss": 1.3679,
      "step": 444
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.851431846618652,
      "learning_rate": 2.8090675057208236e-05,
      "loss": 1.1945,
      "step": 445
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.290416717529297,
      "learning_rate": 2.8086384439359268e-05,
      "loss": 1.3487,
      "step": 446
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.352265357971191,
      "learning_rate": 2.80820938215103e-05,
      "loss": 1.4586,
      "step": 447
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.46808910369873,
      "learning_rate": 2.8077803203661326e-05,
      "loss": 1.5198,
      "step": 448
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.68099594116211,
      "learning_rate": 2.8073512585812358e-05,
      "loss": 1.8168,
      "step": 449
    },
    {
      "epoch": 0.13,
      "grad_norm": 11.353328704833984,
      "learning_rate": 2.8069221967963387e-05,
      "loss": 1.9091,
      "step": 450
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.7789888381958,
      "learning_rate": 2.806493135011442e-05,
      "loss": 1.6881,
      "step": 451
    },
    {
      "epoch": 0.13,
      "grad_norm": 10.931682586669922,
      "learning_rate": 2.8060640732265444e-05,
      "loss": 1.4275,
      "step": 452
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.01468563079834,
      "learning_rate": 2.8056350114416476e-05,
      "loss": 1.5962,
      "step": 453
    },
    {
      "epoch": 0.13,
      "grad_norm": 10.376738548278809,
      "learning_rate": 2.8052059496567505e-05,
      "loss": 1.6438,
      "step": 454
    },
    {
      "epoch": 0.13,
      "grad_norm": 10.467251777648926,
      "learning_rate": 2.8047768878718537e-05,
      "loss": 1.5796,
      "step": 455
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.759889602661133,
      "learning_rate": 2.8043478260869566e-05,
      "loss": 1.4211,
      "step": 456
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.308679580688477,
      "learning_rate": 2.8039187643020595e-05,
      "loss": 1.2328,
      "step": 457
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.23669147491455,
      "learning_rate": 2.8034897025171627e-05,
      "loss": 1.3872,
      "step": 458
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.825939178466797,
      "learning_rate": 2.8030606407322656e-05,
      "loss": 1.4965,
      "step": 459
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.422708511352539,
      "learning_rate": 2.8026315789473685e-05,
      "loss": 1.4893,
      "step": 460
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.007925987243652,
      "learning_rate": 2.8022025171624713e-05,
      "loss": 1.698,
      "step": 461
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.067597389221191,
      "learning_rate": 2.8017734553775746e-05,
      "loss": 1.4106,
      "step": 462
    },
    {
      "epoch": 0.13,
      "grad_norm": 7.50868034362793,
      "learning_rate": 2.8013443935926774e-05,
      "loss": 1.1772,
      "step": 463
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.10647201538086,
      "learning_rate": 2.8009153318077803e-05,
      "loss": 1.7043,
      "step": 464
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.774900436401367,
      "learning_rate": 2.8004862700228835e-05,
      "loss": 1.5012,
      "step": 465
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.257354736328125,
      "learning_rate": 2.8000572082379864e-05,
      "loss": 1.6215,
      "step": 466
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.386637687683105,
      "learning_rate": 2.7996281464530893e-05,
      "loss": 1.5138,
      "step": 467
    },
    {
      "epoch": 0.13,
      "grad_norm": 10.336791038513184,
      "learning_rate": 2.799199084668192e-05,
      "loss": 1.8383,
      "step": 468
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.72229290008545,
      "learning_rate": 2.7987700228832954e-05,
      "loss": 1.3901,
      "step": 469
    },
    {
      "epoch": 0.13,
      "grad_norm": 7.802308559417725,
      "learning_rate": 2.7983409610983982e-05,
      "loss": 1.6891,
      "step": 470
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.691417694091797,
      "learning_rate": 2.797911899313501e-05,
      "loss": 1.3613,
      "step": 471
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.308101654052734,
      "learning_rate": 2.7974828375286043e-05,
      "loss": 1.5225,
      "step": 472
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.286172866821289,
      "learning_rate": 2.7970537757437072e-05,
      "loss": 1.186,
      "step": 473
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.888103485107422,
      "learning_rate": 2.79662471395881e-05,
      "loss": 1.5228,
      "step": 474
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.277830123901367,
      "learning_rate": 2.796195652173913e-05,
      "loss": 1.6649,
      "step": 475
    },
    {
      "epoch": 0.14,
      "grad_norm": 10.21792221069336,
      "learning_rate": 2.7957665903890162e-05,
      "loss": 1.6197,
      "step": 476
    },
    {
      "epoch": 0.14,
      "grad_norm": 10.857234954833984,
      "learning_rate": 2.795337528604119e-05,
      "loss": 1.6978,
      "step": 477
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.541655540466309,
      "learning_rate": 2.794908466819222e-05,
      "loss": 1.5044,
      "step": 478
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.415315628051758,
      "learning_rate": 2.7944794050343248e-05,
      "loss": 1.839,
      "step": 479
    },
    {
      "epoch": 0.14,
      "grad_norm": 7.52134370803833,
      "learning_rate": 2.794050343249428e-05,
      "loss": 1.3303,
      "step": 480
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.852075576782227,
      "learning_rate": 2.7936212814645312e-05,
      "loss": 1.5964,
      "step": 481
    },
    {
      "epoch": 0.14,
      "grad_norm": 10.195798873901367,
      "learning_rate": 2.7931922196796338e-05,
      "loss": 1.6181,
      "step": 482
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.980742454528809,
      "learning_rate": 2.792763157894737e-05,
      "loss": 1.5698,
      "step": 483
    },
    {
      "epoch": 0.14,
      "grad_norm": 7.269761085510254,
      "learning_rate": 2.79233409610984e-05,
      "loss": 1.1187,
      "step": 484
    },
    {
      "epoch": 0.14,
      "grad_norm": 10.480413436889648,
      "learning_rate": 2.791905034324943e-05,
      "loss": 1.2166,
      "step": 485
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.4518404006958,
      "learning_rate": 2.7914759725400456e-05,
      "loss": 1.8096,
      "step": 486
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.00314998626709,
      "learning_rate": 2.791046910755149e-05,
      "loss": 1.7498,
      "step": 487
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.233830451965332,
      "learning_rate": 2.7906178489702517e-05,
      "loss": 1.6336,
      "step": 488
    },
    {
      "epoch": 0.14,
      "grad_norm": 7.957357406616211,
      "learning_rate": 2.7901887871853546e-05,
      "loss": 1.3371,
      "step": 489
    },
    {
      "epoch": 0.14,
      "grad_norm": 7.830927848815918,
      "learning_rate": 2.7897597254004578e-05,
      "loss": 1.3703,
      "step": 490
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.449617385864258,
      "learning_rate": 2.7893306636155607e-05,
      "loss": 1.3393,
      "step": 491
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.626500129699707,
      "learning_rate": 2.788901601830664e-05,
      "loss": 1.4369,
      "step": 492
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.729392051696777,
      "learning_rate": 2.7884725400457664e-05,
      "loss": 1.7288,
      "step": 493
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.956872940063477,
      "learning_rate": 2.7880434782608697e-05,
      "loss": 1.5276,
      "step": 494
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.511049270629883,
      "learning_rate": 2.7876144164759725e-05,
      "loss": 1.3991,
      "step": 495
    },
    {
      "epoch": 0.14,
      "grad_norm": 10.387262344360352,
      "learning_rate": 2.7871853546910758e-05,
      "loss": 1.3234,
      "step": 496
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.903482437133789,
      "learning_rate": 2.7867562929061786e-05,
      "loss": 1.6989,
      "step": 497
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.461954116821289,
      "learning_rate": 2.7863272311212815e-05,
      "loss": 1.65,
      "step": 498
    },
    {
      "epoch": 0.14,
      "grad_norm": 10.389571189880371,
      "learning_rate": 2.7858981693363847e-05,
      "loss": 1.8413,
      "step": 499
    },
    {
      "epoch": 0.14,
      "grad_norm": 10.60290241241455,
      "learning_rate": 2.7854691075514873e-05,
      "loss": 1.5862,
      "step": 500
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.807363510131836,
      "learning_rate": 2.7850400457665905e-05,
      "loss": 1.4559,
      "step": 501
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.981348037719727,
      "learning_rate": 2.7846109839816934e-05,
      "loss": 1.4053,
      "step": 502
    },
    {
      "epoch": 0.14,
      "grad_norm": 12.610603332519531,
      "learning_rate": 2.7841819221967966e-05,
      "loss": 1.7157,
      "step": 503
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.09223461151123,
      "learning_rate": 2.783752860411899e-05,
      "loss": 1.4528,
      "step": 504
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.06646728515625,
      "learning_rate": 2.7833237986270023e-05,
      "loss": 1.5101,
      "step": 505
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.109045028686523,
      "learning_rate": 2.7828947368421055e-05,
      "loss": 1.6101,
      "step": 506
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.117020606994629,
      "learning_rate": 2.7824656750572084e-05,
      "loss": 1.4038,
      "step": 507
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.422054290771484,
      "learning_rate": 2.7820366132723113e-05,
      "loss": 1.6549,
      "step": 508
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.228119850158691,
      "learning_rate": 2.781607551487414e-05,
      "loss": 1.4346,
      "step": 509
    },
    {
      "epoch": 0.15,
      "grad_norm": 11.748756408691406,
      "learning_rate": 2.7811784897025174e-05,
      "loss": 1.9742,
      "step": 510
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.695389747619629,
      "learning_rate": 2.7807494279176203e-05,
      "loss": 1.3546,
      "step": 511
    },
    {
      "epoch": 0.15,
      "grad_norm": 7.655311584472656,
      "learning_rate": 2.780320366132723e-05,
      "loss": 1.2804,
      "step": 512
    },
    {
      "epoch": 0.15,
      "grad_norm": 7.489044189453125,
      "learning_rate": 2.779891304347826e-05,
      "loss": 1.5239,
      "step": 513
    },
    {
      "epoch": 0.15,
      "grad_norm": 7.734864711761475,
      "learning_rate": 2.7794622425629292e-05,
      "loss": 1.516,
      "step": 514
    },
    {
      "epoch": 0.15,
      "grad_norm": 7.818589687347412,
      "learning_rate": 2.779033180778032e-05,
      "loss": 1.2432,
      "step": 515
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.193716049194336,
      "learning_rate": 2.778604118993135e-05,
      "loss": 1.3179,
      "step": 516
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.570880889892578,
      "learning_rate": 2.7781750572082382e-05,
      "loss": 1.2127,
      "step": 517
    },
    {
      "epoch": 0.15,
      "grad_norm": 10.106489181518555,
      "learning_rate": 2.777745995423341e-05,
      "loss": 1.4325,
      "step": 518
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.232028007507324,
      "learning_rate": 2.777316933638444e-05,
      "loss": 1.4359,
      "step": 519
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.200241088867188,
      "learning_rate": 2.7768878718535468e-05,
      "loss": 1.346,
      "step": 520
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.4496488571167,
      "learning_rate": 2.77645881006865e-05,
      "loss": 1.5329,
      "step": 521
    },
    {
      "epoch": 0.15,
      "grad_norm": 10.742775917053223,
      "learning_rate": 2.776029748283753e-05,
      "loss": 1.599,
      "step": 522
    },
    {
      "epoch": 0.15,
      "grad_norm": 10.676810264587402,
      "learning_rate": 2.7756006864988558e-05,
      "loss": 1.6097,
      "step": 523
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.049821853637695,
      "learning_rate": 2.775171624713959e-05,
      "loss": 1.6242,
      "step": 524
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.311460494995117,
      "learning_rate": 2.774742562929062e-05,
      "loss": 1.613,
      "step": 525
    },
    {
      "epoch": 0.15,
      "grad_norm": 10.265384674072266,
      "learning_rate": 2.7743135011441648e-05,
      "loss": 1.8818,
      "step": 526
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.79378890991211,
      "learning_rate": 2.7738844393592676e-05,
      "loss": 1.8994,
      "step": 527
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.35362434387207,
      "learning_rate": 2.773455377574371e-05,
      "loss": 1.4752,
      "step": 528
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.710142135620117,
      "learning_rate": 2.7730263157894737e-05,
      "loss": 1.3721,
      "step": 529
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.677940368652344,
      "learning_rate": 2.7725972540045766e-05,
      "loss": 1.3834,
      "step": 530
    },
    {
      "epoch": 0.15,
      "grad_norm": 6.871583461761475,
      "learning_rate": 2.7721681922196795e-05,
      "loss": 1.3109,
      "step": 531
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.098319053649902,
      "learning_rate": 2.7717391304347827e-05,
      "loss": 1.5228,
      "step": 532
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.34329891204834,
      "learning_rate": 2.771310068649886e-05,
      "loss": 1.6127,
      "step": 533
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.98932933807373,
      "learning_rate": 2.7708810068649885e-05,
      "loss": 1.3145,
      "step": 534
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.540974617004395,
      "learning_rate": 2.7704519450800917e-05,
      "loss": 1.638,
      "step": 535
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.818389892578125,
      "learning_rate": 2.7700228832951946e-05,
      "loss": 1.4179,
      "step": 536
    },
    {
      "epoch": 0.15,
      "grad_norm": 10.52511978149414,
      "learning_rate": 2.7695938215102978e-05,
      "loss": 1.481,
      "step": 537
    },
    {
      "epoch": 0.15,
      "grad_norm": 11.235359191894531,
      "learning_rate": 2.7691647597254003e-05,
      "loss": 1.6016,
      "step": 538
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.686572074890137,
      "learning_rate": 2.7687356979405035e-05,
      "loss": 1.7741,
      "step": 539
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.403671264648438,
      "learning_rate": 2.7683066361556067e-05,
      "loss": 1.4,
      "step": 540
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.480093002319336,
      "learning_rate": 2.7678775743707093e-05,
      "loss": 1.3759,
      "step": 541
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.01067066192627,
      "learning_rate": 2.7674485125858125e-05,
      "loss": 1.2315,
      "step": 542
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.176029205322266,
      "learning_rate": 2.7670194508009154e-05,
      "loss": 1.6042,
      "step": 543
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.995184898376465,
      "learning_rate": 2.7665903890160186e-05,
      "loss": 2.0453,
      "step": 544
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.106335639953613,
      "learning_rate": 2.766161327231121e-05,
      "loss": 1.4097,
      "step": 545
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.503254890441895,
      "learning_rate": 2.7657322654462243e-05,
      "loss": 1.2176,
      "step": 546
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.362741470336914,
      "learning_rate": 2.7653032036613272e-05,
      "loss": 1.1075,
      "step": 547
    },
    {
      "epoch": 0.16,
      "grad_norm": 7.65399169921875,
      "learning_rate": 2.7648741418764304e-05,
      "loss": 1.4227,
      "step": 548
    },
    {
      "epoch": 0.16,
      "grad_norm": 7.963006973266602,
      "learning_rate": 2.7644450800915333e-05,
      "loss": 1.3591,
      "step": 549
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.329169273376465,
      "learning_rate": 2.7640160183066362e-05,
      "loss": 1.6284,
      "step": 550
    },
    {
      "epoch": 0.16,
      "grad_norm": 7.6844916343688965,
      "learning_rate": 2.7635869565217394e-05,
      "loss": 1.4227,
      "step": 551
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.27899169921875,
      "learning_rate": 2.7631578947368423e-05,
      "loss": 1.5949,
      "step": 552
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.054556846618652,
      "learning_rate": 2.762728832951945e-05,
      "loss": 1.1804,
      "step": 553
    },
    {
      "epoch": 0.16,
      "grad_norm": 7.904552459716797,
      "learning_rate": 2.762299771167048e-05,
      "loss": 1.2641,
      "step": 554
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.288396835327148,
      "learning_rate": 2.7618707093821512e-05,
      "loss": 1.6544,
      "step": 555
    },
    {
      "epoch": 0.16,
      "grad_norm": 10.109333038330078,
      "learning_rate": 2.7614416475972538e-05,
      "loss": 1.5606,
      "step": 556
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.692341804504395,
      "learning_rate": 2.761012585812357e-05,
      "loss": 1.3516,
      "step": 557
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.641256332397461,
      "learning_rate": 2.7605835240274602e-05,
      "loss": 1.3218,
      "step": 558
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.773728370666504,
      "learning_rate": 2.760154462242563e-05,
      "loss": 1.1629,
      "step": 559
    },
    {
      "epoch": 0.16,
      "grad_norm": 12.425092697143555,
      "learning_rate": 2.759725400457666e-05,
      "loss": 1.296,
      "step": 560
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.03546142578125,
      "learning_rate": 2.759296338672769e-05,
      "loss": 1.6785,
      "step": 561
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.400121688842773,
      "learning_rate": 2.758867276887872e-05,
      "loss": 1.8052,
      "step": 562
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.625514030456543,
      "learning_rate": 2.758438215102975e-05,
      "loss": 1.3596,
      "step": 563
    },
    {
      "epoch": 0.16,
      "grad_norm": 10.135568618774414,
      "learning_rate": 2.7580091533180778e-05,
      "loss": 1.7402,
      "step": 564
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.666523933410645,
      "learning_rate": 2.7575800915331807e-05,
      "loss": 1.4847,
      "step": 565
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.05019760131836,
      "learning_rate": 2.757151029748284e-05,
      "loss": 1.4296,
      "step": 566
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.563323020935059,
      "learning_rate": 2.7567219679633868e-05,
      "loss": 1.6986,
      "step": 567
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.601005554199219,
      "learning_rate": 2.7562929061784897e-05,
      "loss": 1.8225,
      "step": 568
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.396276473999023,
      "learning_rate": 2.755863844393593e-05,
      "loss": 1.2789,
      "step": 569
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.241303443908691,
      "learning_rate": 2.7554347826086957e-05,
      "loss": 1.5369,
      "step": 570
    },
    {
      "epoch": 0.16,
      "grad_norm": 7.413241386413574,
      "learning_rate": 2.7550057208237986e-05,
      "loss": 1.015,
      "step": 571
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.75148868560791,
      "learning_rate": 2.7545766590389015e-05,
      "loss": 1.4863,
      "step": 572
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.57811450958252,
      "learning_rate": 2.7541475972540047e-05,
      "loss": 1.589,
      "step": 573
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.320711135864258,
      "learning_rate": 2.7537185354691076e-05,
      "loss": 1.5479,
      "step": 574
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.287501335144043,
      "learning_rate": 2.7532894736842105e-05,
      "loss": 1.1933,
      "step": 575
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.462252616882324,
      "learning_rate": 2.7528604118993137e-05,
      "loss": 1.4716,
      "step": 576
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.69751262664795,
      "learning_rate": 2.7524313501144166e-05,
      "loss": 1.789,
      "step": 577
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.301849365234375,
      "learning_rate": 2.7520022883295198e-05,
      "loss": 1.5647,
      "step": 578
    },
    {
      "epoch": 0.17,
      "grad_norm": 10.18299674987793,
      "learning_rate": 2.7515732265446223e-05,
      "loss": 1.4232,
      "step": 579
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.621585845947266,
      "learning_rate": 2.7511441647597255e-05,
      "loss": 1.6084,
      "step": 580
    },
    {
      "epoch": 0.17,
      "grad_norm": 7.825051784515381,
      "learning_rate": 2.7507151029748284e-05,
      "loss": 1.4396,
      "step": 581
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.34615707397461,
      "learning_rate": 2.7502860411899313e-05,
      "loss": 1.7879,
      "step": 582
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.576688766479492,
      "learning_rate": 2.7498569794050345e-05,
      "loss": 1.1719,
      "step": 583
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.739412307739258,
      "learning_rate": 2.7494279176201374e-05,
      "loss": 1.6384,
      "step": 584
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.907939910888672,
      "learning_rate": 2.7489988558352406e-05,
      "loss": 1.5864,
      "step": 585
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.430566787719727,
      "learning_rate": 2.748569794050343e-05,
      "loss": 1.5615,
      "step": 586
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.017333030700684,
      "learning_rate": 2.7481407322654463e-05,
      "loss": 1.4049,
      "step": 587
    },
    {
      "epoch": 0.17,
      "grad_norm": 10.431910514831543,
      "learning_rate": 2.7477116704805492e-05,
      "loss": 1.5352,
      "step": 588
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.034849166870117,
      "learning_rate": 2.7472826086956524e-05,
      "loss": 1.742,
      "step": 589
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.847888946533203,
      "learning_rate": 2.746853546910755e-05,
      "loss": 1.4581,
      "step": 590
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.845882415771484,
      "learning_rate": 2.7464244851258582e-05,
      "loss": 1.4175,
      "step": 591
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.636897087097168,
      "learning_rate": 2.7459954233409614e-05,
      "loss": 1.5935,
      "step": 592
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.328896522521973,
      "learning_rate": 2.745566361556064e-05,
      "loss": 1.4,
      "step": 593
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.179758071899414,
      "learning_rate": 2.745137299771167e-05,
      "loss": 1.4922,
      "step": 594
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.562376976013184,
      "learning_rate": 2.74470823798627e-05,
      "loss": 1.4231,
      "step": 595
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.743012428283691,
      "learning_rate": 2.7442791762013733e-05,
      "loss": 1.457,
      "step": 596
    },
    {
      "epoch": 0.17,
      "grad_norm": 10.17732048034668,
      "learning_rate": 2.7438501144164758e-05,
      "loss": 1.7462,
      "step": 597
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.325789451599121,
      "learning_rate": 2.743421052631579e-05,
      "loss": 1.6907,
      "step": 598
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.406829833984375,
      "learning_rate": 2.742991990846682e-05,
      "loss": 1.3209,
      "step": 599
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.432525634765625,
      "learning_rate": 2.742562929061785e-05,
      "loss": 1.827,
      "step": 600
    },
    {
      "epoch": 0.17,
      "grad_norm": 10.047086715698242,
      "learning_rate": 2.742133867276888e-05,
      "loss": 1.529,
      "step": 601
    },
    {
      "epoch": 0.17,
      "grad_norm": 10.171759605407715,
      "learning_rate": 2.741704805491991e-05,
      "loss": 1.5158,
      "step": 602
    },
    {
      "epoch": 0.17,
      "grad_norm": 7.706381320953369,
      "learning_rate": 2.741275743707094e-05,
      "loss": 1.468,
      "step": 603
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.306461334228516,
      "learning_rate": 2.740846681922197e-05,
      "loss": 1.7419,
      "step": 604
    },
    {
      "epoch": 0.17,
      "grad_norm": 10.386027336120605,
      "learning_rate": 2.7404176201372998e-05,
      "loss": 1.3714,
      "step": 605
    },
    {
      "epoch": 0.17,
      "grad_norm": 10.559309959411621,
      "learning_rate": 2.7399885583524027e-05,
      "loss": 1.453,
      "step": 606
    },
    {
      "epoch": 0.17,
      "grad_norm": 10.220772743225098,
      "learning_rate": 2.739559496567506e-05,
      "loss": 1.5189,
      "step": 607
    },
    {
      "epoch": 0.17,
      "grad_norm": 10.321388244628906,
      "learning_rate": 2.7391304347826085e-05,
      "loss": 1.783,
      "step": 608
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.188323020935059,
      "learning_rate": 2.7387013729977117e-05,
      "loss": 1.6611,
      "step": 609
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.511995315551758,
      "learning_rate": 2.738272311212815e-05,
      "loss": 1.4972,
      "step": 610
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.255911827087402,
      "learning_rate": 2.7378432494279178e-05,
      "loss": 1.6587,
      "step": 611
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.88244342803955,
      "learning_rate": 2.7374141876430206e-05,
      "loss": 1.8485,
      "step": 612
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.155802726745605,
      "learning_rate": 2.7369851258581235e-05,
      "loss": 1.3883,
      "step": 613
    },
    {
      "epoch": 0.18,
      "grad_norm": 10.00084114074707,
      "learning_rate": 2.7365560640732267e-05,
      "loss": 1.6237,
      "step": 614
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.468073844909668,
      "learning_rate": 2.7361270022883296e-05,
      "loss": 1.3699,
      "step": 615
    },
    {
      "epoch": 0.18,
      "grad_norm": 10.218586921691895,
      "learning_rate": 2.7356979405034325e-05,
      "loss": 1.4842,
      "step": 616
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.44926643371582,
      "learning_rate": 2.7352688787185357e-05,
      "loss": 1.5308,
      "step": 617
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.894407272338867,
      "learning_rate": 2.7348398169336386e-05,
      "loss": 1.4603,
      "step": 618
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.344779014587402,
      "learning_rate": 2.7344107551487415e-05,
      "loss": 1.4788,
      "step": 619
    },
    {
      "epoch": 0.18,
      "grad_norm": 10.086723327636719,
      "learning_rate": 2.7339816933638443e-05,
      "loss": 1.5314,
      "step": 620
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.903287887573242,
      "learning_rate": 2.7335526315789475e-05,
      "loss": 1.2463,
      "step": 621
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.488832473754883,
      "learning_rate": 2.7331235697940504e-05,
      "loss": 1.4667,
      "step": 622
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.60849666595459,
      "learning_rate": 2.7326945080091533e-05,
      "loss": 1.4363,
      "step": 623
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.344590187072754,
      "learning_rate": 2.7322654462242562e-05,
      "loss": 1.2329,
      "step": 624
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.900020599365234,
      "learning_rate": 2.7318363844393594e-05,
      "loss": 1.6858,
      "step": 625
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.698345184326172,
      "learning_rate": 2.7314073226544626e-05,
      "loss": 1.6498,
      "step": 626
    },
    {
      "epoch": 0.18,
      "grad_norm": 10.291359901428223,
      "learning_rate": 2.730978260869565e-05,
      "loss": 1.7879,
      "step": 627
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.507843017578125,
      "learning_rate": 2.7305491990846684e-05,
      "loss": 1.7091,
      "step": 628
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.516185760498047,
      "learning_rate": 2.7301201372997712e-05,
      "loss": 1.5036,
      "step": 629
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.028081893920898,
      "learning_rate": 2.7296910755148745e-05,
      "loss": 1.2918,
      "step": 630
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.034631729125977,
      "learning_rate": 2.729262013729977e-05,
      "loss": 1.7479,
      "step": 631
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.290336608886719,
      "learning_rate": 2.7288329519450802e-05,
      "loss": 1.8074,
      "step": 632
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.031563758850098,
      "learning_rate": 2.728403890160183e-05,
      "loss": 1.2922,
      "step": 633
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.388860702514648,
      "learning_rate": 2.727974828375286e-05,
      "loss": 1.2661,
      "step": 634
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.607510566711426,
      "learning_rate": 2.7275457665903892e-05,
      "loss": 1.5673,
      "step": 635
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.935588836669922,
      "learning_rate": 2.727116704805492e-05,
      "loss": 1.7149,
      "step": 636
    },
    {
      "epoch": 0.18,
      "grad_norm": 6.709554672241211,
      "learning_rate": 2.7266876430205953e-05,
      "loss": 1.3551,
      "step": 637
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.180497169494629,
      "learning_rate": 2.7262585812356978e-05,
      "loss": 1.473,
      "step": 638
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.450583457946777,
      "learning_rate": 2.725829519450801e-05,
      "loss": 1.2896,
      "step": 639
    },
    {
      "epoch": 0.18,
      "grad_norm": 7.966728210449219,
      "learning_rate": 2.725400457665904e-05,
      "loss": 1.405,
      "step": 640
    },
    {
      "epoch": 0.18,
      "grad_norm": 10.327277183532715,
      "learning_rate": 2.724971395881007e-05,
      "loss": 1.6359,
      "step": 641
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.075959205627441,
      "learning_rate": 2.7245423340961097e-05,
      "loss": 1.4942,
      "step": 642
    },
    {
      "epoch": 0.18,
      "grad_norm": 11.307262420654297,
      "learning_rate": 2.724113272311213e-05,
      "loss": 1.6728,
      "step": 643
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.996954917907715,
      "learning_rate": 2.723684210526316e-05,
      "loss": 1.2933,
      "step": 644
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.015362739562988,
      "learning_rate": 2.7232551487414186e-05,
      "loss": 1.2742,
      "step": 645
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.029914855957031,
      "learning_rate": 2.722826086956522e-05,
      "loss": 1.5412,
      "step": 646
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.644845962524414,
      "learning_rate": 2.7223970251716247e-05,
      "loss": 1.2903,
      "step": 647
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.480746269226074,
      "learning_rate": 2.721967963386728e-05,
      "loss": 1.2119,
      "step": 648
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.632102966308594,
      "learning_rate": 2.7215389016018305e-05,
      "loss": 1.7093,
      "step": 649
    },
    {
      "epoch": 0.19,
      "grad_norm": 10.944215774536133,
      "learning_rate": 2.7211098398169337e-05,
      "loss": 1.5005,
      "step": 650
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.142459869384766,
      "learning_rate": 2.7206807780320366e-05,
      "loss": 1.3795,
      "step": 651
    },
    {
      "epoch": 0.19,
      "grad_norm": 7.930737495422363,
      "learning_rate": 2.7202517162471398e-05,
      "loss": 1.2816,
      "step": 652
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.225480079650879,
      "learning_rate": 2.7198226544622427e-05,
      "loss": 1.3025,
      "step": 653
    },
    {
      "epoch": 0.19,
      "grad_norm": 10.08369255065918,
      "learning_rate": 2.7193935926773455e-05,
      "loss": 1.4614,
      "step": 654
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.468962669372559,
      "learning_rate": 2.7189645308924487e-05,
      "loss": 1.5527,
      "step": 655
    },
    {
      "epoch": 0.19,
      "grad_norm": 10.897669792175293,
      "learning_rate": 2.7185354691075516e-05,
      "loss": 1.3351,
      "step": 656
    },
    {
      "epoch": 0.19,
      "grad_norm": 10.082230567932129,
      "learning_rate": 2.7181064073226545e-05,
      "loss": 1.416,
      "step": 657
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.202457427978516,
      "learning_rate": 2.7176773455377574e-05,
      "loss": 1.7083,
      "step": 658
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.446024894714355,
      "learning_rate": 2.7172482837528606e-05,
      "loss": 1.126,
      "step": 659
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.490036010742188,
      "learning_rate": 2.7168192219679635e-05,
      "loss": 1.6602,
      "step": 660
    },
    {
      "epoch": 0.19,
      "grad_norm": 7.171084880828857,
      "learning_rate": 2.7163901601830663e-05,
      "loss": 1.2008,
      "step": 661
    },
    {
      "epoch": 0.19,
      "grad_norm": 7.329586982727051,
      "learning_rate": 2.7159610983981696e-05,
      "loss": 1.1091,
      "step": 662
    },
    {
      "epoch": 0.19,
      "grad_norm": 7.757038116455078,
      "learning_rate": 2.7155320366132724e-05,
      "loss": 1.4675,
      "step": 663
    },
    {
      "epoch": 0.19,
      "grad_norm": 10.028850555419922,
      "learning_rate": 2.7151029748283753e-05,
      "loss": 1.7565,
      "step": 664
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.8215970993042,
      "learning_rate": 2.7146739130434782e-05,
      "loss": 1.3142,
      "step": 665
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.601076126098633,
      "learning_rate": 2.7142448512585814e-05,
      "loss": 1.6848,
      "step": 666
    },
    {
      "epoch": 0.19,
      "grad_norm": 10.196307182312012,
      "learning_rate": 2.7138157894736843e-05,
      "loss": 1.879,
      "step": 667
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.428377151489258,
      "learning_rate": 2.713386727688787e-05,
      "loss": 1.6253,
      "step": 668
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.526021003723145,
      "learning_rate": 2.7129576659038904e-05,
      "loss": 1.2186,
      "step": 669
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.465749740600586,
      "learning_rate": 2.7125286041189933e-05,
      "loss": 1.4507,
      "step": 670
    },
    {
      "epoch": 0.19,
      "grad_norm": 11.018233299255371,
      "learning_rate": 2.712099542334096e-05,
      "loss": 1.5466,
      "step": 671
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.881890296936035,
      "learning_rate": 2.711670480549199e-05,
      "loss": 1.8524,
      "step": 672
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.963903427124023,
      "learning_rate": 2.7112414187643022e-05,
      "loss": 1.3904,
      "step": 673
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.687522888183594,
      "learning_rate": 2.710812356979405e-05,
      "loss": 1.1201,
      "step": 674
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.622756004333496,
      "learning_rate": 2.710383295194508e-05,
      "loss": 1.2027,
      "step": 675
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.560112953186035,
      "learning_rate": 2.709954233409611e-05,
      "loss": 1.7021,
      "step": 676
    },
    {
      "epoch": 0.19,
      "grad_norm": 7.986423015594482,
      "learning_rate": 2.709525171624714e-05,
      "loss": 1.2896,
      "step": 677
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.79885482788086,
      "learning_rate": 2.7090961098398173e-05,
      "loss": 1.608,
      "step": 678
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.915427207946777,
      "learning_rate": 2.7086670480549198e-05,
      "loss": 1.2062,
      "step": 679
    },
    {
      "epoch": 0.19,
      "grad_norm": 10.137808799743652,
      "learning_rate": 2.708237986270023e-05,
      "loss": 1.3713,
      "step": 680
    },
    {
      "epoch": 0.19,
      "grad_norm": 10.706426620483398,
      "learning_rate": 2.707808924485126e-05,
      "loss": 1.5678,
      "step": 681
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.411629676818848,
      "learning_rate": 2.707379862700229e-05,
      "loss": 1.2383,
      "step": 682
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.0653715133667,
      "learning_rate": 2.7069508009153317e-05,
      "loss": 1.2453,
      "step": 683
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.046063423156738,
      "learning_rate": 2.706521739130435e-05,
      "loss": 1.2131,
      "step": 684
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.591142654418945,
      "learning_rate": 2.7060926773455378e-05,
      "loss": 1.4589,
      "step": 685
    },
    {
      "epoch": 0.2,
      "grad_norm": 7.216357231140137,
      "learning_rate": 2.7056636155606406e-05,
      "loss": 1.0575,
      "step": 686
    },
    {
      "epoch": 0.2,
      "grad_norm": 11.10873031616211,
      "learning_rate": 2.705234553775744e-05,
      "loss": 1.6542,
      "step": 687
    },
    {
      "epoch": 0.2,
      "grad_norm": 11.519617080688477,
      "learning_rate": 2.7048054919908467e-05,
      "loss": 1.6119,
      "step": 688
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.64222240447998,
      "learning_rate": 2.70437643020595e-05,
      "loss": 1.3678,
      "step": 689
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.729552268981934,
      "learning_rate": 2.7039473684210525e-05,
      "loss": 1.2114,
      "step": 690
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.77514362335205,
      "learning_rate": 2.7035183066361557e-05,
      "loss": 1.6144,
      "step": 691
    },
    {
      "epoch": 0.2,
      "grad_norm": 10.376748085021973,
      "learning_rate": 2.7030892448512586e-05,
      "loss": 1.4726,
      "step": 692
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.687190055847168,
      "learning_rate": 2.7026601830663618e-05,
      "loss": 1.3794,
      "step": 693
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.168863296508789,
      "learning_rate": 2.7022311212814643e-05,
      "loss": 1.573,
      "step": 694
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.668581008911133,
      "learning_rate": 2.7018020594965675e-05,
      "loss": 1.5868,
      "step": 695
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.229246139526367,
      "learning_rate": 2.7013729977116708e-05,
      "loss": 1.1747,
      "step": 696
    },
    {
      "epoch": 0.2,
      "grad_norm": 7.377071857452393,
      "learning_rate": 2.7009439359267733e-05,
      "loss": 1.3782,
      "step": 697
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.349214553833008,
      "learning_rate": 2.7005148741418765e-05,
      "loss": 1.5642,
      "step": 698
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.807302474975586,
      "learning_rate": 2.7000858123569794e-05,
      "loss": 1.4453,
      "step": 699
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.276571273803711,
      "learning_rate": 2.6996567505720826e-05,
      "loss": 1.4524,
      "step": 700
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.345519065856934,
      "learning_rate": 2.699227688787185e-05,
      "loss": 1.4265,
      "step": 701
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.080941200256348,
      "learning_rate": 2.6987986270022884e-05,
      "loss": 1.5395,
      "step": 702
    },
    {
      "epoch": 0.2,
      "grad_norm": 7.670738697052002,
      "learning_rate": 2.6983695652173916e-05,
      "loss": 1.2509,
      "step": 703
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.05324935913086,
      "learning_rate": 2.6979405034324945e-05,
      "loss": 1.3125,
      "step": 704
    },
    {
      "epoch": 0.2,
      "grad_norm": 7.438810348510742,
      "learning_rate": 2.6975114416475973e-05,
      "loss": 1.5452,
      "step": 705
    },
    {
      "epoch": 0.2,
      "grad_norm": 7.855247974395752,
      "learning_rate": 2.6970823798627002e-05,
      "loss": 1.6452,
      "step": 706
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.98528003692627,
      "learning_rate": 2.6966533180778034e-05,
      "loss": 1.4263,
      "step": 707
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.596749305725098,
      "learning_rate": 2.6962242562929063e-05,
      "loss": 1.4357,
      "step": 708
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.83573055267334,
      "learning_rate": 2.6957951945080092e-05,
      "loss": 1.4635,
      "step": 709
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.183515548706055,
      "learning_rate": 2.695366132723112e-05,
      "loss": 1.6072,
      "step": 710
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.77737808227539,
      "learning_rate": 2.6949370709382153e-05,
      "loss": 1.5869,
      "step": 711
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.03734302520752,
      "learning_rate": 2.694508009153318e-05,
      "loss": 1.3849,
      "step": 712
    },
    {
      "epoch": 0.2,
      "grad_norm": 10.255045890808105,
      "learning_rate": 2.694078947368421e-05,
      "loss": 1.3344,
      "step": 713
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.329828262329102,
      "learning_rate": 2.6936498855835242e-05,
      "loss": 1.3703,
      "step": 714
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.793808937072754,
      "learning_rate": 2.693220823798627e-05,
      "loss": 1.7685,
      "step": 715
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.194530487060547,
      "learning_rate": 2.69279176201373e-05,
      "loss": 1.7428,
      "step": 716
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.332390308380127,
      "learning_rate": 2.692362700228833e-05,
      "loss": 1.2059,
      "step": 717
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.699625015258789,
      "learning_rate": 2.691933638443936e-05,
      "loss": 1.542,
      "step": 718
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.5438232421875,
      "learning_rate": 2.691504576659039e-05,
      "loss": 1.3377,
      "step": 719
    },
    {
      "epoch": 0.21,
      "grad_norm": 10.025996208190918,
      "learning_rate": 2.691075514874142e-05,
      "loss": 1.7135,
      "step": 720
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.530084609985352,
      "learning_rate": 2.690646453089245e-05,
      "loss": 1.1862,
      "step": 721
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.832947731018066,
      "learning_rate": 2.690217391304348e-05,
      "loss": 1.5505,
      "step": 722
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.562239646911621,
      "learning_rate": 2.6897883295194508e-05,
      "loss": 1.4249,
      "step": 723
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.007452011108398,
      "learning_rate": 2.6893592677345537e-05,
      "loss": 1.3717,
      "step": 724
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.98516845703125,
      "learning_rate": 2.688930205949657e-05,
      "loss": 1.7065,
      "step": 725
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.55573844909668,
      "learning_rate": 2.6885011441647598e-05,
      "loss": 1.4353,
      "step": 726
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.175204277038574,
      "learning_rate": 2.6880720823798627e-05,
      "loss": 1.2867,
      "step": 727
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.250204086303711,
      "learning_rate": 2.6876430205949655e-05,
      "loss": 1.3647,
      "step": 728
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.163433074951172,
      "learning_rate": 2.6872139588100687e-05,
      "loss": 1.4372,
      "step": 729
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.446063041687012,
      "learning_rate": 2.686784897025172e-05,
      "loss": 1.7786,
      "step": 730
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.554766654968262,
      "learning_rate": 2.6863558352402745e-05,
      "loss": 1.2214,
      "step": 731
    },
    {
      "epoch": 0.21,
      "grad_norm": 10.089883804321289,
      "learning_rate": 2.6859267734553777e-05,
      "loss": 1.451,
      "step": 732
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.008309364318848,
      "learning_rate": 2.6854977116704806e-05,
      "loss": 1.8082,
      "step": 733
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.668070793151855,
      "learning_rate": 2.6850686498855838e-05,
      "loss": 1.5068,
      "step": 734
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.68413782119751,
      "learning_rate": 2.6846395881006863e-05,
      "loss": 1.3725,
      "step": 735
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.327756881713867,
      "learning_rate": 2.6842105263157896e-05,
      "loss": 1.1768,
      "step": 736
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.70918083190918,
      "learning_rate": 2.6837814645308928e-05,
      "loss": 1.3466,
      "step": 737
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.751984596252441,
      "learning_rate": 2.6833524027459953e-05,
      "loss": 1.2856,
      "step": 738
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.377949714660645,
      "learning_rate": 2.6829233409610985e-05,
      "loss": 1.7009,
      "step": 739
    },
    {
      "epoch": 0.21,
      "grad_norm": 10.846776962280273,
      "learning_rate": 2.6824942791762014e-05,
      "loss": 1.2112,
      "step": 740
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.490901947021484,
      "learning_rate": 2.6820652173913046e-05,
      "loss": 1.3016,
      "step": 741
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.792441368103027,
      "learning_rate": 2.681636155606407e-05,
      "loss": 1.6748,
      "step": 742
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.022385597229004,
      "learning_rate": 2.6812070938215104e-05,
      "loss": 1.2802,
      "step": 743
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.235421180725098,
      "learning_rate": 2.6807780320366133e-05,
      "loss": 1.1258,
      "step": 744
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.907036781311035,
      "learning_rate": 2.6803489702517165e-05,
      "loss": 1.1832,
      "step": 745
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.597955703735352,
      "learning_rate": 2.6799199084668193e-05,
      "loss": 1.4996,
      "step": 746
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.799057006835938,
      "learning_rate": 2.6794908466819222e-05,
      "loss": 1.5921,
      "step": 747
    },
    {
      "epoch": 0.21,
      "grad_norm": 10.515116691589355,
      "learning_rate": 2.6790617848970254e-05,
      "loss": 1.8032,
      "step": 748
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.887858390808105,
      "learning_rate": 2.6786327231121283e-05,
      "loss": 1.3823,
      "step": 749
    },
    {
      "epoch": 0.21,
      "grad_norm": 10.281638145446777,
      "learning_rate": 2.6782036613272312e-05,
      "loss": 1.9007,
      "step": 750
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.96937370300293,
      "learning_rate": 2.677774599542334e-05,
      "loss": 1.3861,
      "step": 751
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.151745796203613,
      "learning_rate": 2.6773455377574373e-05,
      "loss": 1.2723,
      "step": 752
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.529725074768066,
      "learning_rate": 2.6769164759725398e-05,
      "loss": 1.5717,
      "step": 753
    },
    {
      "epoch": 0.22,
      "grad_norm": 7.393907070159912,
      "learning_rate": 2.676487414187643e-05,
      "loss": 1.1694,
      "step": 754
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.489725112915039,
      "learning_rate": 2.6760583524027463e-05,
      "loss": 1.5208,
      "step": 755
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.4804048538208,
      "learning_rate": 2.675629290617849e-05,
      "loss": 1.3057,
      "step": 756
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.823626518249512,
      "learning_rate": 2.675200228832952e-05,
      "loss": 1.4457,
      "step": 757
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.953473091125488,
      "learning_rate": 2.674771167048055e-05,
      "loss": 1.7318,
      "step": 758
    },
    {
      "epoch": 0.22,
      "grad_norm": 10.457352638244629,
      "learning_rate": 2.674342105263158e-05,
      "loss": 1.4098,
      "step": 759
    },
    {
      "epoch": 0.22,
      "grad_norm": 7.706247329711914,
      "learning_rate": 2.673913043478261e-05,
      "loss": 1.2625,
      "step": 760
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.632022857666016,
      "learning_rate": 2.673483981693364e-05,
      "loss": 1.5531,
      "step": 761
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.971721649169922,
      "learning_rate": 2.6730549199084667e-05,
      "loss": 1.4319,
      "step": 762
    },
    {
      "epoch": 0.22,
      "grad_norm": 7.860352516174316,
      "learning_rate": 2.67262585812357e-05,
      "loss": 1.2889,
      "step": 763
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.265642166137695,
      "learning_rate": 2.6721967963386728e-05,
      "loss": 1.2929,
      "step": 764
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.112139701843262,
      "learning_rate": 2.6717677345537757e-05,
      "loss": 1.632,
      "step": 765
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.007532119750977,
      "learning_rate": 2.671338672768879e-05,
      "loss": 1.4319,
      "step": 766
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.495431900024414,
      "learning_rate": 2.6709096109839818e-05,
      "loss": 1.3153,
      "step": 767
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.244626998901367,
      "learning_rate": 2.6704805491990847e-05,
      "loss": 1.2545,
      "step": 768
    },
    {
      "epoch": 0.22,
      "grad_norm": 10.376943588256836,
      "learning_rate": 2.6700514874141875e-05,
      "loss": 1.3685,
      "step": 769
    },
    {
      "epoch": 0.22,
      "grad_norm": 12.076695442199707,
      "learning_rate": 2.6696224256292908e-05,
      "loss": 1.494,
      "step": 770
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.511319160461426,
      "learning_rate": 2.6691933638443936e-05,
      "loss": 1.387,
      "step": 771
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.889599800109863,
      "learning_rate": 2.6687643020594965e-05,
      "loss": 1.5296,
      "step": 772
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.285345077514648,
      "learning_rate": 2.6683352402745997e-05,
      "loss": 1.578,
      "step": 773
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.711544036865234,
      "learning_rate": 2.6679061784897026e-05,
      "loss": 1.3207,
      "step": 774
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.246892929077148,
      "learning_rate": 2.6674771167048058e-05,
      "loss": 1.5279,
      "step": 775
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.31828498840332,
      "learning_rate": 2.6670480549199084e-05,
      "loss": 1.5191,
      "step": 776
    },
    {
      "epoch": 0.22,
      "grad_norm": 10.230610847473145,
      "learning_rate": 2.6666189931350116e-05,
      "loss": 1.9263,
      "step": 777
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.435502052307129,
      "learning_rate": 2.6661899313501145e-05,
      "loss": 1.5,
      "step": 778
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.353205680847168,
      "learning_rate": 2.6657608695652173e-05,
      "loss": 1.1672,
      "step": 779
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.82955265045166,
      "learning_rate": 2.6653318077803205e-05,
      "loss": 1.7323,
      "step": 780
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.477160453796387,
      "learning_rate": 2.6649027459954234e-05,
      "loss": 1.5587,
      "step": 781
    },
    {
      "epoch": 0.22,
      "grad_norm": 7.85951566696167,
      "learning_rate": 2.6644736842105266e-05,
      "loss": 1.4767,
      "step": 782
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.957002639770508,
      "learning_rate": 2.6640446224256292e-05,
      "loss": 1.2401,
      "step": 783
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.295096397399902,
      "learning_rate": 2.6636155606407324e-05,
      "loss": 1.7779,
      "step": 784
    },
    {
      "epoch": 0.22,
      "grad_norm": 10.103282928466797,
      "learning_rate": 2.6631864988558353e-05,
      "loss": 1.6127,
      "step": 785
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.313076972961426,
      "learning_rate": 2.6627574370709385e-05,
      "loss": 1.2176,
      "step": 786
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.133856773376465,
      "learning_rate": 2.662328375286041e-05,
      "loss": 1.2762,
      "step": 787
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.65207290649414,
      "learning_rate": 2.6618993135011442e-05,
      "loss": 1.6066,
      "step": 788
    },
    {
      "epoch": 0.23,
      "grad_norm": 10.28887939453125,
      "learning_rate": 2.6614702517162475e-05,
      "loss": 1.7862,
      "step": 789
    },
    {
      "epoch": 0.23,
      "grad_norm": 10.515134811401367,
      "learning_rate": 2.66104118993135e-05,
      "loss": 1.5071,
      "step": 790
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.611674308776855,
      "learning_rate": 2.6606121281464532e-05,
      "loss": 1.5397,
      "step": 791
    },
    {
      "epoch": 0.23,
      "grad_norm": 10.166763305664062,
      "learning_rate": 2.660183066361556e-05,
      "loss": 1.294,
      "step": 792
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.607548713684082,
      "learning_rate": 2.6597540045766593e-05,
      "loss": 1.3341,
      "step": 793
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.909729957580566,
      "learning_rate": 2.659324942791762e-05,
      "loss": 1.4576,
      "step": 794
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.688549995422363,
      "learning_rate": 2.658895881006865e-05,
      "loss": 1.3312,
      "step": 795
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.35120677947998,
      "learning_rate": 2.658466819221968e-05,
      "loss": 1.3555,
      "step": 796
    },
    {
      "epoch": 0.23,
      "grad_norm": 10.456376075744629,
      "learning_rate": 2.658037757437071e-05,
      "loss": 1.6144,
      "step": 797
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.537264823913574,
      "learning_rate": 2.657608695652174e-05,
      "loss": 1.5542,
      "step": 798
    },
    {
      "epoch": 0.23,
      "grad_norm": 10.017496109008789,
      "learning_rate": 2.657179633867277e-05,
      "loss": 1.5115,
      "step": 799
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.876407623291016,
      "learning_rate": 2.65675057208238e-05,
      "loss": 1.461,
      "step": 800
    },
    {
      "epoch": 0.23,
      "grad_norm": 10.00885009765625,
      "learning_rate": 2.656321510297483e-05,
      "loss": 1.514,
      "step": 801
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.821441650390625,
      "learning_rate": 2.655892448512586e-05,
      "loss": 1.4214,
      "step": 802
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.728100776672363,
      "learning_rate": 2.6554633867276887e-05,
      "loss": 1.4235,
      "step": 803
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.904654502868652,
      "learning_rate": 2.655034324942792e-05,
      "loss": 1.6595,
      "step": 804
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.324999809265137,
      "learning_rate": 2.6546052631578945e-05,
      "loss": 1.4202,
      "step": 805
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.588123321533203,
      "learning_rate": 2.6541762013729977e-05,
      "loss": 1.3708,
      "step": 806
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.563580513000488,
      "learning_rate": 2.653747139588101e-05,
      "loss": 1.3736,
      "step": 807
    },
    {
      "epoch": 0.23,
      "grad_norm": 7.648112773895264,
      "learning_rate": 2.6533180778032038e-05,
      "loss": 1.3695,
      "step": 808
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.872638702392578,
      "learning_rate": 2.6528890160183067e-05,
      "loss": 1.3779,
      "step": 809
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.684102058410645,
      "learning_rate": 2.6524599542334096e-05,
      "loss": 1.7289,
      "step": 810
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.175305366516113,
      "learning_rate": 2.6520308924485128e-05,
      "loss": 1.5158,
      "step": 811
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.868097305297852,
      "learning_rate": 2.6516018306636156e-05,
      "loss": 1.7496,
      "step": 812
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.471468925476074,
      "learning_rate": 2.6511727688787185e-05,
      "loss": 1.6277,
      "step": 813
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.10570240020752,
      "learning_rate": 2.6507437070938214e-05,
      "loss": 1.6749,
      "step": 814
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.941532135009766,
      "learning_rate": 2.6503146453089246e-05,
      "loss": 1.5023,
      "step": 815
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.747757911682129,
      "learning_rate": 2.6498855835240275e-05,
      "loss": 1.4065,
      "step": 816
    },
    {
      "epoch": 0.23,
      "grad_norm": 10.095454216003418,
      "learning_rate": 2.6494565217391304e-05,
      "loss": 1.1893,
      "step": 817
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.078791618347168,
      "learning_rate": 2.6490274599542336e-05,
      "loss": 1.3243,
      "step": 818
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.536273002624512,
      "learning_rate": 2.6485983981693365e-05,
      "loss": 1.4985,
      "step": 819
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.09534740447998,
      "learning_rate": 2.6481693363844393e-05,
      "loss": 1.3751,
      "step": 820
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.298988342285156,
      "learning_rate": 2.6477402745995422e-05,
      "loss": 1.444,
      "step": 821
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.982908725738525,
      "learning_rate": 2.6473112128146454e-05,
      "loss": 1.3142,
      "step": 822
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.453125,
      "learning_rate": 2.6468821510297486e-05,
      "loss": 1.2621,
      "step": 823
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.977948188781738,
      "learning_rate": 2.6464530892448512e-05,
      "loss": 1.2767,
      "step": 824
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.755941390991211,
      "learning_rate": 2.6460240274599544e-05,
      "loss": 1.4843,
      "step": 825
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.676729202270508,
      "learning_rate": 2.6455949656750573e-05,
      "loss": 1.3186,
      "step": 826
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.526939392089844,
      "learning_rate": 2.6451659038901605e-05,
      "loss": 1.7197,
      "step": 827
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.150410652160645,
      "learning_rate": 2.644736842105263e-05,
      "loss": 1.5229,
      "step": 828
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.937966823577881,
      "learning_rate": 2.6443077803203662e-05,
      "loss": 1.1528,
      "step": 829
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.968094825744629,
      "learning_rate": 2.643878718535469e-05,
      "loss": 1.2425,
      "step": 830
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.370745658874512,
      "learning_rate": 2.643449656750572e-05,
      "loss": 1.5296,
      "step": 831
    },
    {
      "epoch": 0.24,
      "grad_norm": 10.294328689575195,
      "learning_rate": 2.6430205949656752e-05,
      "loss": 1.5658,
      "step": 832
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.228590965270996,
      "learning_rate": 2.642591533180778e-05,
      "loss": 1.1133,
      "step": 833
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.272261619567871,
      "learning_rate": 2.6421624713958813e-05,
      "loss": 1.2029,
      "step": 834
    },
    {
      "epoch": 0.24,
      "grad_norm": 10.820258140563965,
      "learning_rate": 2.641733409610984e-05,
      "loss": 1.517,
      "step": 835
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.859976768493652,
      "learning_rate": 2.641304347826087e-05,
      "loss": 1.2971,
      "step": 836
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.648213386535645,
      "learning_rate": 2.64087528604119e-05,
      "loss": 1.1995,
      "step": 837
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.68163013458252,
      "learning_rate": 2.640446224256293e-05,
      "loss": 1.3573,
      "step": 838
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.964151382446289,
      "learning_rate": 2.6400171624713957e-05,
      "loss": 0.964,
      "step": 839
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.399771690368652,
      "learning_rate": 2.639588100686499e-05,
      "loss": 1.5236,
      "step": 840
    },
    {
      "epoch": 0.24,
      "grad_norm": 10.44339370727539,
      "learning_rate": 2.639159038901602e-05,
      "loss": 1.5872,
      "step": 841
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.737083435058594,
      "learning_rate": 2.6387299771167047e-05,
      "loss": 1.6335,
      "step": 842
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.311802864074707,
      "learning_rate": 2.638300915331808e-05,
      "loss": 1.254,
      "step": 843
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.597392082214355,
      "learning_rate": 2.6378718535469108e-05,
      "loss": 1.2899,
      "step": 844
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.301457405090332,
      "learning_rate": 2.637442791762014e-05,
      "loss": 1.1914,
      "step": 845
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.583595275878906,
      "learning_rate": 2.6370137299771165e-05,
      "loss": 1.6079,
      "step": 846
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.2188138961792,
      "learning_rate": 2.6365846681922197e-05,
      "loss": 1.4473,
      "step": 847
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.318650245666504,
      "learning_rate": 2.6361556064073226e-05,
      "loss": 1.409,
      "step": 848
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.837665557861328,
      "learning_rate": 2.6357265446224258e-05,
      "loss": 1.3829,
      "step": 849
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.132809638977051,
      "learning_rate": 2.6352974828375287e-05,
      "loss": 1.2159,
      "step": 850
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.9716339111328125,
      "learning_rate": 2.6348684210526316e-05,
      "loss": 1.5728,
      "step": 851
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.7947492599487305,
      "learning_rate": 2.6344393592677348e-05,
      "loss": 1.3865,
      "step": 852
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.062242031097412,
      "learning_rate": 2.6340102974828377e-05,
      "loss": 1.3908,
      "step": 853
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.27694320678711,
      "learning_rate": 2.6335812356979405e-05,
      "loss": 1.6951,
      "step": 854
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.277055263519287,
      "learning_rate": 2.6331521739130434e-05,
      "loss": 1.3148,
      "step": 855
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.7327470779418945,
      "learning_rate": 2.6327231121281466e-05,
      "loss": 1.2293,
      "step": 856
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.443324089050293,
      "learning_rate": 2.6322940503432495e-05,
      "loss": 1.2151,
      "step": 857
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.355253219604492,
      "learning_rate": 2.6318649885583524e-05,
      "loss": 1.5138,
      "step": 858
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.991596221923828,
      "learning_rate": 2.6314359267734556e-05,
      "loss": 1.4703,
      "step": 859
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.560808181762695,
      "learning_rate": 2.6310068649885585e-05,
      "loss": 1.1305,
      "step": 860
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.362797737121582,
      "learning_rate": 2.6305778032036614e-05,
      "loss": 1.2677,
      "step": 861
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.149765014648438,
      "learning_rate": 2.6301487414187642e-05,
      "loss": 1.2291,
      "step": 862
    },
    {
      "epoch": 0.25,
      "grad_norm": 10.845412254333496,
      "learning_rate": 2.6297196796338674e-05,
      "loss": 1.2982,
      "step": 863
    },
    {
      "epoch": 0.25,
      "grad_norm": 7.772393226623535,
      "learning_rate": 2.6292906178489703e-05,
      "loss": 1.272,
      "step": 864
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.768017768859863,
      "learning_rate": 2.6288615560640732e-05,
      "loss": 1.5599,
      "step": 865
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.420010566711426,
      "learning_rate": 2.6284324942791764e-05,
      "loss": 1.1816,
      "step": 866
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.9022798538208,
      "learning_rate": 2.6280034324942793e-05,
      "loss": 1.6228,
      "step": 867
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.873797416687012,
      "learning_rate": 2.6275743707093822e-05,
      "loss": 1.3512,
      "step": 868
    },
    {
      "epoch": 0.25,
      "grad_norm": 7.665680408477783,
      "learning_rate": 2.627145308924485e-05,
      "loss": 1.1994,
      "step": 869
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.856553077697754,
      "learning_rate": 2.6267162471395883e-05,
      "loss": 1.6137,
      "step": 870
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.588093757629395,
      "learning_rate": 2.626287185354691e-05,
      "loss": 1.1853,
      "step": 871
    },
    {
      "epoch": 0.25,
      "grad_norm": 7.5686163902282715,
      "learning_rate": 2.625858123569794e-05,
      "loss": 1.3191,
      "step": 872
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.717270851135254,
      "learning_rate": 2.625429061784897e-05,
      "loss": 1.556,
      "step": 873
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.204094886779785,
      "learning_rate": 2.625e-05,
      "loss": 1.4066,
      "step": 874
    },
    {
      "epoch": 0.25,
      "grad_norm": 7.699957370758057,
      "learning_rate": 2.6245709382151033e-05,
      "loss": 1.2625,
      "step": 875
    },
    {
      "epoch": 0.25,
      "grad_norm": 6.8762640953063965,
      "learning_rate": 2.624141876430206e-05,
      "loss": 1.1352,
      "step": 876
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.54216480255127,
      "learning_rate": 2.623712814645309e-05,
      "loss": 1.3852,
      "step": 877
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.11685848236084,
      "learning_rate": 2.623283752860412e-05,
      "loss": 1.0808,
      "step": 878
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.630313873291016,
      "learning_rate": 2.6228546910755152e-05,
      "loss": 1.2204,
      "step": 879
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.078722953796387,
      "learning_rate": 2.6224256292906177e-05,
      "loss": 1.3517,
      "step": 880
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.79880428314209,
      "learning_rate": 2.621996567505721e-05,
      "loss": 1.533,
      "step": 881
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.823820114135742,
      "learning_rate": 2.6215675057208238e-05,
      "loss": 1.3513,
      "step": 882
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.656079292297363,
      "learning_rate": 2.6211384439359267e-05,
      "loss": 1.3914,
      "step": 883
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.426806449890137,
      "learning_rate": 2.62070938215103e-05,
      "loss": 1.4989,
      "step": 884
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.181078910827637,
      "learning_rate": 2.6202803203661328e-05,
      "loss": 1.4191,
      "step": 885
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.820877075195312,
      "learning_rate": 2.619851258581236e-05,
      "loss": 1.0893,
      "step": 886
    },
    {
      "epoch": 0.25,
      "grad_norm": 10.27674388885498,
      "learning_rate": 2.6194221967963385e-05,
      "loss": 1.4704,
      "step": 887
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.574858665466309,
      "learning_rate": 2.6189931350114417e-05,
      "loss": 1.5897,
      "step": 888
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.53704833984375,
      "learning_rate": 2.6185640732265446e-05,
      "loss": 1.2895,
      "step": 889
    },
    {
      "epoch": 0.25,
      "grad_norm": 7.642971515655518,
      "learning_rate": 2.618135011441648e-05,
      "loss": 1.3837,
      "step": 890
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.106728553771973,
      "learning_rate": 2.6177059496567504e-05,
      "loss": 1.2967,
      "step": 891
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.707520484924316,
      "learning_rate": 2.6172768878718536e-05,
      "loss": 1.6252,
      "step": 892
    },
    {
      "epoch": 0.26,
      "grad_norm": 9.524190902709961,
      "learning_rate": 2.6168478260869568e-05,
      "loss": 1.4453,
      "step": 893
    },
    {
      "epoch": 0.26,
      "grad_norm": 7.845490455627441,
      "learning_rate": 2.6164187643020593e-05,
      "loss": 1.2811,
      "step": 894
    },
    {
      "epoch": 0.26,
      "grad_norm": 9.159387588500977,
      "learning_rate": 2.6159897025171626e-05,
      "loss": 1.409,
      "step": 895
    },
    {
      "epoch": 0.26,
      "grad_norm": 7.53691291809082,
      "learning_rate": 2.6155606407322654e-05,
      "loss": 1.3078,
      "step": 896
    },
    {
      "epoch": 0.26,
      "grad_norm": 9.199991226196289,
      "learning_rate": 2.6151315789473686e-05,
      "loss": 1.3968,
      "step": 897
    },
    {
      "epoch": 0.26,
      "grad_norm": 9.692501068115234,
      "learning_rate": 2.6147025171624712e-05,
      "loss": 1.3681,
      "step": 898
    },
    {
      "epoch": 0.26,
      "grad_norm": 10.840760231018066,
      "learning_rate": 2.6142734553775744e-05,
      "loss": 1.4837,
      "step": 899
    },
    {
      "epoch": 0.26,
      "grad_norm": 9.615287780761719,
      "learning_rate": 2.6138443935926776e-05,
      "loss": 1.153,
      "step": 900
    },
    {
      "epoch": 0.26,
      "grad_norm": 11.4331693649292,
      "learning_rate": 2.6134153318077805e-05,
      "loss": 1.2511,
      "step": 901
    },
    {
      "epoch": 0.26,
      "grad_norm": 12.56297779083252,
      "learning_rate": 2.6129862700228834e-05,
      "loss": 1.4341,
      "step": 902
    },
    {
      "epoch": 0.26,
      "grad_norm": 10.176109313964844,
      "learning_rate": 2.6125572082379862e-05,
      "loss": 1.1976,
      "step": 903
    },
    {
      "epoch": 0.26,
      "grad_norm": 10.312824249267578,
      "learning_rate": 2.6121281464530895e-05,
      "loss": 1.3812,
      "step": 904
    },
    {
      "epoch": 0.26,
      "grad_norm": 9.968500137329102,
      "learning_rate": 2.6116990846681923e-05,
      "loss": 1.3292,
      "step": 905
    },
    {
      "epoch": 0.26,
      "grad_norm": 10.515392303466797,
      "learning_rate": 2.6112700228832952e-05,
      "loss": 1.5927,
      "step": 906
    },
    {
      "epoch": 0.26,
      "grad_norm": 10.473443031311035,
      "learning_rate": 2.610840961098398e-05,
      "loss": 1.4326,
      "step": 907
    },
    {
      "epoch": 0.26,
      "grad_norm": 10.10062313079834,
      "learning_rate": 2.6104118993135013e-05,
      "loss": 1.4284,
      "step": 908
    },
    {
      "epoch": 0.26,
      "grad_norm": 13.7451753616333,
      "learning_rate": 2.6099828375286042e-05,
      "loss": 1.7868,
      "step": 909
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.696174621582031,
      "learning_rate": 2.609553775743707e-05,
      "loss": 1.1353,
      "step": 910
    },
    {
      "epoch": 0.26,
      "grad_norm": 7.880374908447266,
      "learning_rate": 2.6091247139588103e-05,
      "loss": 1.0382,
      "step": 911
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.875225067138672,
      "learning_rate": 2.608695652173913e-05,
      "loss": 1.1643,
      "step": 912
    },
    {
      "epoch": 0.26,
      "grad_norm": 7.7749247550964355,
      "learning_rate": 2.608266590389016e-05,
      "loss": 1.1447,
      "step": 913
    },
    {
      "epoch": 0.26,
      "grad_norm": 10.65074634552002,
      "learning_rate": 2.607837528604119e-05,
      "loss": 1.2644,
      "step": 914
    },
    {
      "epoch": 0.26,
      "grad_norm": 9.620262145996094,
      "learning_rate": 2.607408466819222e-05,
      "loss": 1.668,
      "step": 915
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.444010734558105,
      "learning_rate": 2.606979405034325e-05,
      "loss": 1.4725,
      "step": 916
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.710188865661621,
      "learning_rate": 2.606550343249428e-05,
      "loss": 1.5552,
      "step": 917
    },
    {
      "epoch": 0.26,
      "grad_norm": 10.831197738647461,
      "learning_rate": 2.606121281464531e-05,
      "loss": 1.6034,
      "step": 918
    },
    {
      "epoch": 0.26,
      "grad_norm": 10.13759708404541,
      "learning_rate": 2.605692219679634e-05,
      "loss": 1.0555,
      "step": 919
    },
    {
      "epoch": 0.26,
      "grad_norm": 9.175713539123535,
      "learning_rate": 2.605263157894737e-05,
      "loss": 1.2767,
      "step": 920
    },
    {
      "epoch": 0.26,
      "grad_norm": 7.387826442718506,
      "learning_rate": 2.6048340961098397e-05,
      "loss": 1.0568,
      "step": 921
    },
    {
      "epoch": 0.26,
      "grad_norm": 9.601868629455566,
      "learning_rate": 2.604405034324943e-05,
      "loss": 1.578,
      "step": 922
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.701878547668457,
      "learning_rate": 2.6039759725400458e-05,
      "loss": 1.486,
      "step": 923
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.450980186462402,
      "learning_rate": 2.6035469107551487e-05,
      "loss": 1.3934,
      "step": 924
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.329426765441895,
      "learning_rate": 2.6031178489702516e-05,
      "loss": 1.2754,
      "step": 925
    },
    {
      "epoch": 0.26,
      "grad_norm": 7.788487434387207,
      "learning_rate": 2.6026887871853548e-05,
      "loss": 1.0489,
      "step": 926
    },
    {
      "epoch": 0.27,
      "grad_norm": 10.402837753295898,
      "learning_rate": 2.602259725400458e-05,
      "loss": 1.487,
      "step": 927
    },
    {
      "epoch": 0.27,
      "grad_norm": 7.651103973388672,
      "learning_rate": 2.6018306636155605e-05,
      "loss": 1.3695,
      "step": 928
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.581132888793945,
      "learning_rate": 2.6014016018306638e-05,
      "loss": 1.6594,
      "step": 929
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.17892074584961,
      "learning_rate": 2.6009725400457666e-05,
      "loss": 1.5891,
      "step": 930
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.242602348327637,
      "learning_rate": 2.60054347826087e-05,
      "loss": 1.2397,
      "step": 931
    },
    {
      "epoch": 0.27,
      "grad_norm": 7.623937129974365,
      "learning_rate": 2.6001144164759724e-05,
      "loss": 1.2555,
      "step": 932
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.100603103637695,
      "learning_rate": 2.5996853546910756e-05,
      "loss": 1.2265,
      "step": 933
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.403155326843262,
      "learning_rate": 2.5992562929061788e-05,
      "loss": 1.2792,
      "step": 934
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.048784255981445,
      "learning_rate": 2.5988272311212814e-05,
      "loss": 1.5941,
      "step": 935
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.061111450195312,
      "learning_rate": 2.5983981693363846e-05,
      "loss": 1.4503,
      "step": 936
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.425355911254883,
      "learning_rate": 2.5979691075514874e-05,
      "loss": 1.3135,
      "step": 937
    },
    {
      "epoch": 0.27,
      "grad_norm": 7.993234634399414,
      "learning_rate": 2.5975400457665907e-05,
      "loss": 1.3449,
      "step": 938
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.822248458862305,
      "learning_rate": 2.5971109839816932e-05,
      "loss": 1.2058,
      "step": 939
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.437129974365234,
      "learning_rate": 2.5966819221967964e-05,
      "loss": 1.2206,
      "step": 940
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.871660232543945,
      "learning_rate": 2.5962528604118993e-05,
      "loss": 1.4596,
      "step": 941
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.384592056274414,
      "learning_rate": 2.5958237986270025e-05,
      "loss": 1.3168,
      "step": 942
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.543497085571289,
      "learning_rate": 2.5953947368421054e-05,
      "loss": 1.3973,
      "step": 943
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.770606994628906,
      "learning_rate": 2.5949656750572083e-05,
      "loss": 1.4777,
      "step": 944
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.465770721435547,
      "learning_rate": 2.5945366132723115e-05,
      "loss": 1.2012,
      "step": 945
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.79576587677002,
      "learning_rate": 2.5941075514874144e-05,
      "loss": 1.4603,
      "step": 946
    },
    {
      "epoch": 0.27,
      "grad_norm": 7.772630214691162,
      "learning_rate": 2.5936784897025172e-05,
      "loss": 1.4308,
      "step": 947
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.234416961669922,
      "learning_rate": 2.59324942791762e-05,
      "loss": 1.3327,
      "step": 948
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.109087944030762,
      "learning_rate": 2.5928203661327233e-05,
      "loss": 1.4196,
      "step": 949
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.199943542480469,
      "learning_rate": 2.592391304347826e-05,
      "loss": 1.6369,
      "step": 950
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.374139785766602,
      "learning_rate": 2.591962242562929e-05,
      "loss": 1.2285,
      "step": 951
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.24150276184082,
      "learning_rate": 2.5915331807780323e-05,
      "loss": 1.117,
      "step": 952
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.536650657653809,
      "learning_rate": 2.591104118993135e-05,
      "loss": 1.1041,
      "step": 953
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.652554512023926,
      "learning_rate": 2.590675057208238e-05,
      "loss": 1.2805,
      "step": 954
    },
    {
      "epoch": 0.27,
      "grad_norm": 10.050910949707031,
      "learning_rate": 2.590245995423341e-05,
      "loss": 1.4703,
      "step": 955
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.267804145812988,
      "learning_rate": 2.589816933638444e-05,
      "loss": 1.4239,
      "step": 956
    },
    {
      "epoch": 0.27,
      "grad_norm": 7.391750812530518,
      "learning_rate": 2.589387871853547e-05,
      "loss": 1.4491,
      "step": 957
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.495930671691895,
      "learning_rate": 2.58895881006865e-05,
      "loss": 1.156,
      "step": 958
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.142219543457031,
      "learning_rate": 2.5885297482837528e-05,
      "loss": 1.2461,
      "step": 959
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.015151977539062,
      "learning_rate": 2.588100686498856e-05,
      "loss": 1.2013,
      "step": 960
    },
    {
      "epoch": 0.27,
      "grad_norm": 7.377295017242432,
      "learning_rate": 2.587671624713959e-05,
      "loss": 1.0362,
      "step": 961
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.156243324279785,
      "learning_rate": 2.5872425629290617e-05,
      "loss": 1.4888,
      "step": 962
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.230779647827148,
      "learning_rate": 2.586813501144165e-05,
      "loss": 1.4112,
      "step": 963
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.809797286987305,
      "learning_rate": 2.5863844393592678e-05,
      "loss": 1.3778,
      "step": 964
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.639060974121094,
      "learning_rate": 2.5859553775743707e-05,
      "loss": 1.2252,
      "step": 965
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.97224235534668,
      "learning_rate": 2.5855263157894736e-05,
      "loss": 1.5233,
      "step": 966
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.486656188964844,
      "learning_rate": 2.5850972540045768e-05,
      "loss": 1.5144,
      "step": 967
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.242863655090332,
      "learning_rate": 2.5846681922196797e-05,
      "loss": 1.4416,
      "step": 968
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.781201362609863,
      "learning_rate": 2.5842391304347826e-05,
      "loss": 1.1063,
      "step": 969
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.582833290100098,
      "learning_rate": 2.5838100686498858e-05,
      "loss": 1.0988,
      "step": 970
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.837419033050537,
      "learning_rate": 2.5833810068649886e-05,
      "loss": 1.3314,
      "step": 971
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.807422637939453,
      "learning_rate": 2.582951945080092e-05,
      "loss": 1.3859,
      "step": 972
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.319928169250488,
      "learning_rate": 2.5825228832951944e-05,
      "loss": 1.3565,
      "step": 973
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.010677337646484,
      "learning_rate": 2.5820938215102976e-05,
      "loss": 1.4269,
      "step": 974
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.191767692565918,
      "learning_rate": 2.5816647597254005e-05,
      "loss": 1.5395,
      "step": 975
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.80266284942627,
      "learning_rate": 2.5812356979405034e-05,
      "loss": 1.4944,
      "step": 976
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.184325218200684,
      "learning_rate": 2.5808066361556066e-05,
      "loss": 1.2486,
      "step": 977
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.970193862915039,
      "learning_rate": 2.5803775743707095e-05,
      "loss": 1.3057,
      "step": 978
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.319640159606934,
      "learning_rate": 2.5799485125858127e-05,
      "loss": 1.4045,
      "step": 979
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.836695671081543,
      "learning_rate": 2.5795194508009152e-05,
      "loss": 1.3089,
      "step": 980
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.435751914978027,
      "learning_rate": 2.5790903890160184e-05,
      "loss": 1.311,
      "step": 981
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.068690299987793,
      "learning_rate": 2.5786613272311213e-05,
      "loss": 1.6868,
      "step": 982
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.13770866394043,
      "learning_rate": 2.5782322654462245e-05,
      "loss": 1.5963,
      "step": 983
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.724429607391357,
      "learning_rate": 2.577803203661327e-05,
      "loss": 1.4556,
      "step": 984
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.583844184875488,
      "learning_rate": 2.5773741418764303e-05,
      "loss": 1.4503,
      "step": 985
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.158873558044434,
      "learning_rate": 2.5769450800915335e-05,
      "loss": 1.3628,
      "step": 986
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.549899578094482,
      "learning_rate": 2.576516018306636e-05,
      "loss": 1.2796,
      "step": 987
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.432297706604004,
      "learning_rate": 2.5760869565217392e-05,
      "loss": 1.6099,
      "step": 988
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.539736270904541,
      "learning_rate": 2.575657894736842e-05,
      "loss": 1.3312,
      "step": 989
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.980632781982422,
      "learning_rate": 2.5752288329519453e-05,
      "loss": 1.3764,
      "step": 990
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.134516716003418,
      "learning_rate": 2.574799771167048e-05,
      "loss": 1.0205,
      "step": 991
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.90495777130127,
      "learning_rate": 2.574370709382151e-05,
      "loss": 1.597,
      "step": 992
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.75139331817627,
      "learning_rate": 2.573941647597254e-05,
      "loss": 1.2573,
      "step": 993
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.376870155334473,
      "learning_rate": 2.5735125858123572e-05,
      "loss": 1.8324,
      "step": 994
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.90761947631836,
      "learning_rate": 2.57308352402746e-05,
      "loss": 1.2191,
      "step": 995
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.86811351776123,
      "learning_rate": 2.572654462242563e-05,
      "loss": 1.4312,
      "step": 996
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.165155410766602,
      "learning_rate": 2.572225400457666e-05,
      "loss": 1.4856,
      "step": 997
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.089405059814453,
      "learning_rate": 2.571796338672769e-05,
      "loss": 1.3299,
      "step": 998
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.177542686462402,
      "learning_rate": 2.571367276887872e-05,
      "loss": 1.3916,
      "step": 999
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.137677192687988,
      "learning_rate": 2.5709382151029748e-05,
      "loss": 1.2121,
      "step": 1000
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.527697563171387,
      "learning_rate": 2.570509153318078e-05,
      "loss": 1.4384,
      "step": 1001
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.89223861694336,
      "learning_rate": 2.5700800915331805e-05,
      "loss": 1.1825,
      "step": 1002
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.23388671875,
      "learning_rate": 2.5696510297482838e-05,
      "loss": 1.8242,
      "step": 1003
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.942590713500977,
      "learning_rate": 2.569221967963387e-05,
      "loss": 1.5872,
      "step": 1004
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.871176719665527,
      "learning_rate": 2.56879290617849e-05,
      "loss": 1.4072,
      "step": 1005
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.461094856262207,
      "learning_rate": 2.5683638443935927e-05,
      "loss": 1.4736,
      "step": 1006
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.951471328735352,
      "learning_rate": 2.5679347826086956e-05,
      "loss": 1.2968,
      "step": 1007
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.7858171463012695,
      "learning_rate": 2.5675057208237988e-05,
      "loss": 1.3835,
      "step": 1008
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.943772792816162,
      "learning_rate": 2.5670766590389017e-05,
      "loss": 1.8021,
      "step": 1009
    },
    {
      "epoch": 0.29,
      "grad_norm": 6.024533748626709,
      "learning_rate": 2.5666475972540046e-05,
      "loss": 0.9488,
      "step": 1010
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.342346668243408,
      "learning_rate": 2.5662185354691074e-05,
      "loss": 1.1327,
      "step": 1011
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.056375503540039,
      "learning_rate": 2.5657894736842107e-05,
      "loss": 1.4397,
      "step": 1012
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.417347431182861,
      "learning_rate": 2.5653604118993135e-05,
      "loss": 1.2529,
      "step": 1013
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.64712905883789,
      "learning_rate": 2.5649313501144164e-05,
      "loss": 1.3253,
      "step": 1014
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.166894912719727,
      "learning_rate": 2.5645022883295196e-05,
      "loss": 1.5091,
      "step": 1015
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.005392074584961,
      "learning_rate": 2.5640732265446225e-05,
      "loss": 1.226,
      "step": 1016
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.414338111877441,
      "learning_rate": 2.5636441647597254e-05,
      "loss": 1.3174,
      "step": 1017
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.061562538146973,
      "learning_rate": 2.5632151029748283e-05,
      "loss": 1.4758,
      "step": 1018
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.46532917022705,
      "learning_rate": 2.5627860411899315e-05,
      "loss": 1.2356,
      "step": 1019
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.5079026222229,
      "learning_rate": 2.5623569794050347e-05,
      "loss": 1.2108,
      "step": 1020
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.118891716003418,
      "learning_rate": 2.5619279176201372e-05,
      "loss": 1.2809,
      "step": 1021
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.336339950561523,
      "learning_rate": 2.5614988558352404e-05,
      "loss": 1.3214,
      "step": 1022
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.880213737487793,
      "learning_rate": 2.5610697940503433e-05,
      "loss": 1.0074,
      "step": 1023
    },
    {
      "epoch": 0.29,
      "grad_norm": 10.576152801513672,
      "learning_rate": 2.5606407322654465e-05,
      "loss": 1.7207,
      "step": 1024
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.325366973876953,
      "learning_rate": 2.560211670480549e-05,
      "loss": 1.353,
      "step": 1025
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.758729457855225,
      "learning_rate": 2.5597826086956523e-05,
      "loss": 1.4028,
      "step": 1026
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.77721118927002,
      "learning_rate": 2.559353546910755e-05,
      "loss": 1.5293,
      "step": 1027
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.359360694885254,
      "learning_rate": 2.558924485125858e-05,
      "loss": 1.1269,
      "step": 1028
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.833972930908203,
      "learning_rate": 2.5584954233409613e-05,
      "loss": 1.3712,
      "step": 1029
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.514402389526367,
      "learning_rate": 2.558066361556064e-05,
      "loss": 1.204,
      "step": 1030
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.313460350036621,
      "learning_rate": 2.5576372997711674e-05,
      "loss": 1.5063,
      "step": 1031
    },
    {
      "epoch": 0.3,
      "grad_norm": 10.018709182739258,
      "learning_rate": 2.55720823798627e-05,
      "loss": 1.2818,
      "step": 1032
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.012324810028076,
      "learning_rate": 2.556779176201373e-05,
      "loss": 1.1244,
      "step": 1033
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.470826148986816,
      "learning_rate": 2.556350114416476e-05,
      "loss": 1.2546,
      "step": 1034
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.6571269035339355,
      "learning_rate": 2.5559210526315792e-05,
      "loss": 1.2892,
      "step": 1035
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.890312194824219,
      "learning_rate": 2.5554919908466817e-05,
      "loss": 1.521,
      "step": 1036
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.642049789428711,
      "learning_rate": 2.555062929061785e-05,
      "loss": 1.1408,
      "step": 1037
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.882678985595703,
      "learning_rate": 2.554633867276888e-05,
      "loss": 1.2596,
      "step": 1038
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.036952018737793,
      "learning_rate": 2.5542048054919907e-05,
      "loss": 1.331,
      "step": 1039
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.739363670349121,
      "learning_rate": 2.553775743707094e-05,
      "loss": 1.2239,
      "step": 1040
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.373655319213867,
      "learning_rate": 2.5533466819221968e-05,
      "loss": 1.2434,
      "step": 1041
    },
    {
      "epoch": 0.3,
      "grad_norm": 10.343050956726074,
      "learning_rate": 2.5529176201373e-05,
      "loss": 1.4349,
      "step": 1042
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.710387229919434,
      "learning_rate": 2.5524885583524025e-05,
      "loss": 1.4334,
      "step": 1043
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.289206504821777,
      "learning_rate": 2.5520594965675058e-05,
      "loss": 1.1788,
      "step": 1044
    },
    {
      "epoch": 0.3,
      "grad_norm": 10.75119400024414,
      "learning_rate": 2.5516304347826086e-05,
      "loss": 1.4688,
      "step": 1045
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.680585861206055,
      "learning_rate": 2.551201372997712e-05,
      "loss": 1.449,
      "step": 1046
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.56112813949585,
      "learning_rate": 2.5507723112128147e-05,
      "loss": 1.2321,
      "step": 1047
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.956106185913086,
      "learning_rate": 2.5503432494279176e-05,
      "loss": 1.3308,
      "step": 1048
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.079631805419922,
      "learning_rate": 2.5499141876430208e-05,
      "loss": 1.1529,
      "step": 1049
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.40312671661377,
      "learning_rate": 2.5494851258581237e-05,
      "loss": 1.5271,
      "step": 1050
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.621920585632324,
      "learning_rate": 2.5490560640732266e-05,
      "loss": 1.1845,
      "step": 1051
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.569252967834473,
      "learning_rate": 2.5486270022883295e-05,
      "loss": 1.3228,
      "step": 1052
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.784046173095703,
      "learning_rate": 2.5481979405034327e-05,
      "loss": 1.3341,
      "step": 1053
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.730287551879883,
      "learning_rate": 2.5477688787185355e-05,
      "loss": 1.2053,
      "step": 1054
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.498429298400879,
      "learning_rate": 2.5473398169336384e-05,
      "loss": 1.1694,
      "step": 1055
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.870926856994629,
      "learning_rate": 2.5469107551487416e-05,
      "loss": 1.4959,
      "step": 1056
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.9740214347839355,
      "learning_rate": 2.5464816933638445e-05,
      "loss": 1.4715,
      "step": 1057
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.216447830200195,
      "learning_rate": 2.5460526315789474e-05,
      "loss": 1.1949,
      "step": 1058
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.495555877685547,
      "learning_rate": 2.5456235697940503e-05,
      "loss": 1.5628,
      "step": 1059
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.50665283203125,
      "learning_rate": 2.5451945080091535e-05,
      "loss": 1.275,
      "step": 1060
    },
    {
      "epoch": 0.3,
      "grad_norm": 10.58385181427002,
      "learning_rate": 2.5447654462242564e-05,
      "loss": 1.6663,
      "step": 1061
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.590288162231445,
      "learning_rate": 2.5443363844393592e-05,
      "loss": 1.3047,
      "step": 1062
    },
    {
      "epoch": 0.3,
      "grad_norm": 10.212004661560059,
      "learning_rate": 2.5439073226544625e-05,
      "loss": 1.6188,
      "step": 1063
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.003531455993652,
      "learning_rate": 2.5434782608695653e-05,
      "loss": 1.2925,
      "step": 1064
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.413034439086914,
      "learning_rate": 2.5430491990846682e-05,
      "loss": 1.0456,
      "step": 1065
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.396681785583496,
      "learning_rate": 2.542620137299771e-05,
      "loss": 1.2743,
      "step": 1066
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.070167541503906,
      "learning_rate": 2.5421910755148743e-05,
      "loss": 1.4599,
      "step": 1067
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.388463973999023,
      "learning_rate": 2.5417620137299772e-05,
      "loss": 1.3985,
      "step": 1068
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.501354217529297,
      "learning_rate": 2.54133295194508e-05,
      "loss": 1.5014,
      "step": 1069
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.798341751098633,
      "learning_rate": 2.540903890160183e-05,
      "loss": 1.1759,
      "step": 1070
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.82196044921875,
      "learning_rate": 2.540474828375286e-05,
      "loss": 1.3665,
      "step": 1071
    },
    {
      "epoch": 0.31,
      "grad_norm": 10.14842700958252,
      "learning_rate": 2.5400457665903894e-05,
      "loss": 1.2128,
      "step": 1072
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.154882431030273,
      "learning_rate": 2.539616704805492e-05,
      "loss": 1.2736,
      "step": 1073
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.904618263244629,
      "learning_rate": 2.539187643020595e-05,
      "loss": 1.3211,
      "step": 1074
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.278619766235352,
      "learning_rate": 2.538758581235698e-05,
      "loss": 1.3248,
      "step": 1075
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.660553932189941,
      "learning_rate": 2.5383295194508012e-05,
      "loss": 1.2732,
      "step": 1076
    },
    {
      "epoch": 0.31,
      "grad_norm": 10.31358528137207,
      "learning_rate": 2.5379004576659037e-05,
      "loss": 1.4625,
      "step": 1077
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.699115753173828,
      "learning_rate": 2.537471395881007e-05,
      "loss": 1.1086,
      "step": 1078
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.902919292449951,
      "learning_rate": 2.53704233409611e-05,
      "loss": 1.0589,
      "step": 1079
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.76630973815918,
      "learning_rate": 2.5366132723112127e-05,
      "loss": 1.1594,
      "step": 1080
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.40144157409668,
      "learning_rate": 2.536184210526316e-05,
      "loss": 1.4975,
      "step": 1081
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.331470489501953,
      "learning_rate": 2.5357551487414188e-05,
      "loss": 1.1938,
      "step": 1082
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.63892126083374,
      "learning_rate": 2.535326086956522e-05,
      "loss": 1.3181,
      "step": 1083
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.823601722717285,
      "learning_rate": 2.5348970251716246e-05,
      "loss": 1.4152,
      "step": 1084
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.027241706848145,
      "learning_rate": 2.5344679633867278e-05,
      "loss": 1.2235,
      "step": 1085
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.12237548828125,
      "learning_rate": 2.5340389016018307e-05,
      "loss": 1.3252,
      "step": 1086
    },
    {
      "epoch": 0.31,
      "grad_norm": 10.51408576965332,
      "learning_rate": 2.533609839816934e-05,
      "loss": 1.6956,
      "step": 1087
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.42678451538086,
      "learning_rate": 2.5331807780320364e-05,
      "loss": 1.3338,
      "step": 1088
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.596070289611816,
      "learning_rate": 2.5327517162471396e-05,
      "loss": 1.2526,
      "step": 1089
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.929720401763916,
      "learning_rate": 2.532322654462243e-05,
      "loss": 1.1503,
      "step": 1090
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.295254707336426,
      "learning_rate": 2.5318935926773457e-05,
      "loss": 1.6479,
      "step": 1091
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.760801315307617,
      "learning_rate": 2.5314645308924486e-05,
      "loss": 1.492,
      "step": 1092
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.011282920837402,
      "learning_rate": 2.5310354691075515e-05,
      "loss": 1.1961,
      "step": 1093
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.240226745605469,
      "learning_rate": 2.5306064073226547e-05,
      "loss": 0.9983,
      "step": 1094
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.712852478027344,
      "learning_rate": 2.5301773455377572e-05,
      "loss": 1.1892,
      "step": 1095
    },
    {
      "epoch": 0.31,
      "grad_norm": 11.721887588500977,
      "learning_rate": 2.5297482837528604e-05,
      "loss": 1.4409,
      "step": 1096
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.104716300964355,
      "learning_rate": 2.5293192219679637e-05,
      "loss": 1.3911,
      "step": 1097
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.200728416442871,
      "learning_rate": 2.5288901601830665e-05,
      "loss": 1.0253,
      "step": 1098
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.017943382263184,
      "learning_rate": 2.5284610983981694e-05,
      "loss": 1.0505,
      "step": 1099
    },
    {
      "epoch": 0.31,
      "grad_norm": 10.048181533813477,
      "learning_rate": 2.5280320366132723e-05,
      "loss": 1.6158,
      "step": 1100
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.796588897705078,
      "learning_rate": 2.5276029748283755e-05,
      "loss": 1.4182,
      "step": 1101
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.351282119750977,
      "learning_rate": 2.5271739130434784e-05,
      "loss": 0.9927,
      "step": 1102
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.599355697631836,
      "learning_rate": 2.5267448512585813e-05,
      "loss": 1.0907,
      "step": 1103
    },
    {
      "epoch": 0.32,
      "grad_norm": 10.412554740905762,
      "learning_rate": 2.526315789473684e-05,
      "loss": 1.4171,
      "step": 1104
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.77164077758789,
      "learning_rate": 2.5258867276887873e-05,
      "loss": 1.3462,
      "step": 1105
    },
    {
      "epoch": 0.32,
      "grad_norm": 10.209586143493652,
      "learning_rate": 2.5254576659038902e-05,
      "loss": 1.2277,
      "step": 1106
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.908252716064453,
      "learning_rate": 2.525028604118993e-05,
      "loss": 1.5864,
      "step": 1107
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.8651933670043945,
      "learning_rate": 2.5245995423340963e-05,
      "loss": 1.3012,
      "step": 1108
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.069273948669434,
      "learning_rate": 2.5241704805491992e-05,
      "loss": 1.1916,
      "step": 1109
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.798157691955566,
      "learning_rate": 2.523741418764302e-05,
      "loss": 1.4687,
      "step": 1110
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.591025352478027,
      "learning_rate": 2.523312356979405e-05,
      "loss": 1.2547,
      "step": 1111
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.359914779663086,
      "learning_rate": 2.522883295194508e-05,
      "loss": 1.1862,
      "step": 1112
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.099730491638184,
      "learning_rate": 2.522454233409611e-05,
      "loss": 1.4387,
      "step": 1113
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.6268157958984375,
      "learning_rate": 2.522025171624714e-05,
      "loss": 1.2851,
      "step": 1114
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.440269470214844,
      "learning_rate": 2.521596109839817e-05,
      "loss": 1.4515,
      "step": 1115
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.150582313537598,
      "learning_rate": 2.52116704805492e-05,
      "loss": 1.4764,
      "step": 1116
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.502106666564941,
      "learning_rate": 2.520737986270023e-05,
      "loss": 1.1288,
      "step": 1117
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.629725456237793,
      "learning_rate": 2.5203089244851258e-05,
      "loss": 1.3621,
      "step": 1118
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.673270225524902,
      "learning_rate": 2.519879862700229e-05,
      "loss": 1.2357,
      "step": 1119
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.973155975341797,
      "learning_rate": 2.519450800915332e-05,
      "loss": 1.3026,
      "step": 1120
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.627095222473145,
      "learning_rate": 2.5190217391304347e-05,
      "loss": 1.1334,
      "step": 1121
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.878874778747559,
      "learning_rate": 2.5185926773455376e-05,
      "loss": 1.1232,
      "step": 1122
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.94869613647461,
      "learning_rate": 2.5181636155606408e-05,
      "loss": 1.6974,
      "step": 1123
    },
    {
      "epoch": 0.32,
      "grad_norm": 10.141195297241211,
      "learning_rate": 2.517734553775744e-05,
      "loss": 1.6349,
      "step": 1124
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.594303131103516,
      "learning_rate": 2.5173054919908466e-05,
      "loss": 1.6864,
      "step": 1125
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.13325023651123,
      "learning_rate": 2.5168764302059498e-05,
      "loss": 1.2408,
      "step": 1126
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.727703094482422,
      "learning_rate": 2.5164473684210527e-05,
      "loss": 1.5705,
      "step": 1127
    },
    {
      "epoch": 0.32,
      "grad_norm": 11.216474533081055,
      "learning_rate": 2.516018306636156e-05,
      "loss": 1.6617,
      "step": 1128
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.66234016418457,
      "learning_rate": 2.5155892448512584e-05,
      "loss": 1.2739,
      "step": 1129
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.064190864562988,
      "learning_rate": 2.5151601830663616e-05,
      "loss": 1.0814,
      "step": 1130
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.823426246643066,
      "learning_rate": 2.5147311212814645e-05,
      "loss": 1.7493,
      "step": 1131
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.140297889709473,
      "learning_rate": 2.5143020594965674e-05,
      "loss": 1.3828,
      "step": 1132
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.560859680175781,
      "learning_rate": 2.5138729977116706e-05,
      "loss": 1.3355,
      "step": 1133
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.429637908935547,
      "learning_rate": 2.5134439359267735e-05,
      "loss": 1.2483,
      "step": 1134
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.343366622924805,
      "learning_rate": 2.5130148741418767e-05,
      "loss": 1.6462,
      "step": 1135
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.374827861785889,
      "learning_rate": 2.5125858123569792e-05,
      "loss": 1.0401,
      "step": 1136
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.14588737487793,
      "learning_rate": 2.5121567505720825e-05,
      "loss": 1.1931,
      "step": 1137
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.2505521774292,
      "learning_rate": 2.5117276887871853e-05,
      "loss": 1.2012,
      "step": 1138
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.059476852416992,
      "learning_rate": 2.5112986270022885e-05,
      "loss": 1.1364,
      "step": 1139
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.84462308883667,
      "learning_rate": 2.5108695652173914e-05,
      "loss": 1.1497,
      "step": 1140
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.2100830078125,
      "learning_rate": 2.5104405034324943e-05,
      "loss": 1.0862,
      "step": 1141
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.621990203857422,
      "learning_rate": 2.5100114416475975e-05,
      "loss": 1.2482,
      "step": 1142
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.5615925788879395,
      "learning_rate": 2.5095823798627004e-05,
      "loss": 0.9791,
      "step": 1143
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.368245124816895,
      "learning_rate": 2.5091533180778033e-05,
      "loss": 1.5939,
      "step": 1144
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.730009078979492,
      "learning_rate": 2.508724256292906e-05,
      "loss": 1.2557,
      "step": 1145
    },
    {
      "epoch": 0.33,
      "grad_norm": 10.07455062866211,
      "learning_rate": 2.5082951945080094e-05,
      "loss": 1.5275,
      "step": 1146
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.566425323486328,
      "learning_rate": 2.507866132723112e-05,
      "loss": 1.1481,
      "step": 1147
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.685506820678711,
      "learning_rate": 2.507437070938215e-05,
      "loss": 1.4072,
      "step": 1148
    },
    {
      "epoch": 0.33,
      "grad_norm": 10.160717010498047,
      "learning_rate": 2.5070080091533183e-05,
      "loss": 1.4458,
      "step": 1149
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.259299278259277,
      "learning_rate": 2.5065789473684212e-05,
      "loss": 1.6841,
      "step": 1150
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.61259937286377,
      "learning_rate": 2.506149885583524e-05,
      "loss": 1.4721,
      "step": 1151
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.85054874420166,
      "learning_rate": 2.505720823798627e-05,
      "loss": 1.352,
      "step": 1152
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.074386596679688,
      "learning_rate": 2.5052917620137302e-05,
      "loss": 1.4526,
      "step": 1153
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.661362648010254,
      "learning_rate": 2.504862700228833e-05,
      "loss": 1.3133,
      "step": 1154
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.18763256072998,
      "learning_rate": 2.504433638443936e-05,
      "loss": 1.5691,
      "step": 1155
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.982461929321289,
      "learning_rate": 2.5040045766590388e-05,
      "loss": 1.3467,
      "step": 1156
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.239413261413574,
      "learning_rate": 2.503575514874142e-05,
      "loss": 1.4486,
      "step": 1157
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.908697605133057,
      "learning_rate": 2.503146453089245e-05,
      "loss": 1.4239,
      "step": 1158
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.104166030883789,
      "learning_rate": 2.5027173913043478e-05,
      "loss": 1.1813,
      "step": 1159
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.924987316131592,
      "learning_rate": 2.502288329519451e-05,
      "loss": 1.0608,
      "step": 1160
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.379941940307617,
      "learning_rate": 2.501859267734554e-05,
      "loss": 1.3629,
      "step": 1161
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.123723983764648,
      "learning_rate": 2.5014302059496567e-05,
      "loss": 1.3557,
      "step": 1162
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.908727645874023,
      "learning_rate": 2.5010011441647596e-05,
      "loss": 1.0705,
      "step": 1163
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.93742847442627,
      "learning_rate": 2.500572082379863e-05,
      "loss": 1.1082,
      "step": 1164
    },
    {
      "epoch": 0.33,
      "grad_norm": 10.610589981079102,
      "learning_rate": 2.5001430205949657e-05,
      "loss": 1.4131,
      "step": 1165
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.712764739990234,
      "learning_rate": 2.4997139588100686e-05,
      "loss": 1.1069,
      "step": 1166
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.058858871459961,
      "learning_rate": 2.4992848970251718e-05,
      "loss": 1.2948,
      "step": 1167
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.90199613571167,
      "learning_rate": 2.4988558352402747e-05,
      "loss": 1.1113,
      "step": 1168
    },
    {
      "epoch": 0.33,
      "grad_norm": 10.635540008544922,
      "learning_rate": 2.498426773455378e-05,
      "loss": 1.4564,
      "step": 1169
    },
    {
      "epoch": 0.33,
      "grad_norm": 10.573737144470215,
      "learning_rate": 2.4979977116704804e-05,
      "loss": 1.4524,
      "step": 1170
    },
    {
      "epoch": 0.33,
      "grad_norm": 13.97392749786377,
      "learning_rate": 2.4975686498855837e-05,
      "loss": 1.7515,
      "step": 1171
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.978637218475342,
      "learning_rate": 2.4971395881006865e-05,
      "loss": 0.9382,
      "step": 1172
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.72246265411377,
      "learning_rate": 2.4967105263157894e-05,
      "loss": 1.3806,
      "step": 1173
    },
    {
      "epoch": 0.34,
      "grad_norm": 11.520387649536133,
      "learning_rate": 2.4962814645308926e-05,
      "loss": 1.3977,
      "step": 1174
    },
    {
      "epoch": 0.34,
      "grad_norm": 10.564435958862305,
      "learning_rate": 2.4958524027459955e-05,
      "loss": 1.323,
      "step": 1175
    },
    {
      "epoch": 0.34,
      "grad_norm": 10.678526878356934,
      "learning_rate": 2.4954233409610987e-05,
      "loss": 1.7055,
      "step": 1176
    },
    {
      "epoch": 0.34,
      "grad_norm": 10.193828582763672,
      "learning_rate": 2.4949942791762013e-05,
      "loss": 1.4728,
      "step": 1177
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.24264144897461,
      "learning_rate": 2.4945652173913045e-05,
      "loss": 1.2945,
      "step": 1178
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.4100341796875,
      "learning_rate": 2.4941361556064073e-05,
      "loss": 1.5619,
      "step": 1179
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.36609172821045,
      "learning_rate": 2.4937070938215106e-05,
      "loss": 1.1721,
      "step": 1180
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.610232353210449,
      "learning_rate": 2.493278032036613e-05,
      "loss": 1.3444,
      "step": 1181
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.46660327911377,
      "learning_rate": 2.4928489702517163e-05,
      "loss": 1.6277,
      "step": 1182
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.28674602508545,
      "learning_rate": 2.4924199084668195e-05,
      "loss": 1.2887,
      "step": 1183
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.585891723632812,
      "learning_rate": 2.491990846681922e-05,
      "loss": 1.5058,
      "step": 1184
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.711112976074219,
      "learning_rate": 2.4915617848970253e-05,
      "loss": 1.364,
      "step": 1185
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.428664684295654,
      "learning_rate": 2.491132723112128e-05,
      "loss": 1.2072,
      "step": 1186
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.880519866943359,
      "learning_rate": 2.4907036613272314e-05,
      "loss": 1.3054,
      "step": 1187
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.168986320495605,
      "learning_rate": 2.490274599542334e-05,
      "loss": 1.3597,
      "step": 1188
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.571588039398193,
      "learning_rate": 2.489845537757437e-05,
      "loss": 1.4104,
      "step": 1189
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.356108665466309,
      "learning_rate": 2.48941647597254e-05,
      "loss": 1.1608,
      "step": 1190
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.247628211975098,
      "learning_rate": 2.4889874141876432e-05,
      "loss": 1.4549,
      "step": 1191
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.7991461753845215,
      "learning_rate": 2.488558352402746e-05,
      "loss": 1.1564,
      "step": 1192
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.154600620269775,
      "learning_rate": 2.488129290617849e-05,
      "loss": 1.1743,
      "step": 1193
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.795459747314453,
      "learning_rate": 2.4877002288329522e-05,
      "loss": 1.2105,
      "step": 1194
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.458930969238281,
      "learning_rate": 2.487271167048055e-05,
      "loss": 1.2433,
      "step": 1195
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.928001880645752,
      "learning_rate": 2.486842105263158e-05,
      "loss": 1.0818,
      "step": 1196
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.688636779785156,
      "learning_rate": 2.4864130434782608e-05,
      "loss": 1.4069,
      "step": 1197
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.312451362609863,
      "learning_rate": 2.485983981693364e-05,
      "loss": 1.1773,
      "step": 1198
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.169563293457031,
      "learning_rate": 2.4855549199084666e-05,
      "loss": 1.1708,
      "step": 1199
    },
    {
      "epoch": 0.34,
      "grad_norm": 10.276508331298828,
      "learning_rate": 2.4851258581235698e-05,
      "loss": 1.6543,
      "step": 1200
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.86247730255127,
      "learning_rate": 2.484696796338673e-05,
      "loss": 1.5157,
      "step": 1201
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.955330848693848,
      "learning_rate": 2.484267734553776e-05,
      "loss": 1.0957,
      "step": 1202
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.047369956970215,
      "learning_rate": 2.4838386727688788e-05,
      "loss": 1.135,
      "step": 1203
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.094965934753418,
      "learning_rate": 2.4834096109839816e-05,
      "loss": 1.055,
      "step": 1204
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.789654731750488,
      "learning_rate": 2.482980549199085e-05,
      "loss": 1.3808,
      "step": 1205
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.9882707595825195,
      "learning_rate": 2.4825514874141877e-05,
      "loss": 1.4581,
      "step": 1206
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.175067901611328,
      "learning_rate": 2.4821224256292906e-05,
      "loss": 1.4699,
      "step": 1207
    },
    {
      "epoch": 0.35,
      "grad_norm": 7.715549945831299,
      "learning_rate": 2.4816933638443935e-05,
      "loss": 1.2313,
      "step": 1208
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.82292366027832,
      "learning_rate": 2.4812643020594967e-05,
      "loss": 1.3491,
      "step": 1209
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.679081916809082,
      "learning_rate": 2.4808352402745996e-05,
      "loss": 1.3,
      "step": 1210
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.480063438415527,
      "learning_rate": 2.4804061784897025e-05,
      "loss": 0.9899,
      "step": 1211
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.44324016571045,
      "learning_rate": 2.4799771167048057e-05,
      "loss": 1.2802,
      "step": 1212
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.433918952941895,
      "learning_rate": 2.4795480549199085e-05,
      "loss": 1.2748,
      "step": 1213
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.200996398925781,
      "learning_rate": 2.4791189931350114e-05,
      "loss": 1.2814,
      "step": 1214
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.791327476501465,
      "learning_rate": 2.4786899313501143e-05,
      "loss": 1.4816,
      "step": 1215
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.12392520904541,
      "learning_rate": 2.4782608695652175e-05,
      "loss": 0.9496,
      "step": 1216
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.058268547058105,
      "learning_rate": 2.4778318077803207e-05,
      "loss": 1.0657,
      "step": 1217
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.287860870361328,
      "learning_rate": 2.4774027459954233e-05,
      "loss": 1.3242,
      "step": 1218
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.007309913635254,
      "learning_rate": 2.4769736842105265e-05,
      "loss": 1.3125,
      "step": 1219
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.979630470275879,
      "learning_rate": 2.4765446224256294e-05,
      "loss": 1.0767,
      "step": 1220
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.455676078796387,
      "learning_rate": 2.4761155606407326e-05,
      "loss": 1.4545,
      "step": 1221
    },
    {
      "epoch": 0.35,
      "grad_norm": 10.137964248657227,
      "learning_rate": 2.475686498855835e-05,
      "loss": 1.5658,
      "step": 1222
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.38525104522705,
      "learning_rate": 2.4752574370709383e-05,
      "loss": 1.3243,
      "step": 1223
    },
    {
      "epoch": 0.35,
      "grad_norm": 7.730583190917969,
      "learning_rate": 2.4748283752860412e-05,
      "loss": 1.1342,
      "step": 1224
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.241644859313965,
      "learning_rate": 2.474399313501144e-05,
      "loss": 1.2161,
      "step": 1225
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.795563697814941,
      "learning_rate": 2.4739702517162473e-05,
      "loss": 1.3466,
      "step": 1226
    },
    {
      "epoch": 0.35,
      "grad_norm": 7.525288105010986,
      "learning_rate": 2.4735411899313502e-05,
      "loss": 1.4963,
      "step": 1227
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.251082420349121,
      "learning_rate": 2.4731121281464534e-05,
      "loss": 1.1266,
      "step": 1228
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.439674377441406,
      "learning_rate": 2.472683066361556e-05,
      "loss": 1.0338,
      "step": 1229
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.983979225158691,
      "learning_rate": 2.472254004576659e-05,
      "loss": 1.6099,
      "step": 1230
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.698434829711914,
      "learning_rate": 2.471824942791762e-05,
      "loss": 1.6213,
      "step": 1231
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.603226661682129,
      "learning_rate": 2.4713958810068652e-05,
      "loss": 1.5249,
      "step": 1232
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.249344825744629,
      "learning_rate": 2.4709668192219678e-05,
      "loss": 1.0933,
      "step": 1233
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.643524169921875,
      "learning_rate": 2.470537757437071e-05,
      "loss": 1.2073,
      "step": 1234
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.731209754943848,
      "learning_rate": 2.4701086956521742e-05,
      "loss": 1.2141,
      "step": 1235
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.993982315063477,
      "learning_rate": 2.4696796338672767e-05,
      "loss": 1.3216,
      "step": 1236
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.99123764038086,
      "learning_rate": 2.46925057208238e-05,
      "loss": 1.4136,
      "step": 1237
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.950716018676758,
      "learning_rate": 2.468821510297483e-05,
      "loss": 1.3142,
      "step": 1238
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.027567863464355,
      "learning_rate": 2.468392448512586e-05,
      "loss": 1.1953,
      "step": 1239
    },
    {
      "epoch": 0.35,
      "grad_norm": 7.93207311630249,
      "learning_rate": 2.4679633867276886e-05,
      "loss": 0.9217,
      "step": 1240
    },
    {
      "epoch": 0.35,
      "grad_norm": 7.474572658538818,
      "learning_rate": 2.4675343249427918e-05,
      "loss": 0.9354,
      "step": 1241
    },
    {
      "epoch": 0.36,
      "grad_norm": 10.033756256103516,
      "learning_rate": 2.4671052631578947e-05,
      "loss": 1.4035,
      "step": 1242
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.002933502197266,
      "learning_rate": 2.466676201372998e-05,
      "loss": 0.9634,
      "step": 1243
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.01943302154541,
      "learning_rate": 2.4662471395881008e-05,
      "loss": 1.0456,
      "step": 1244
    },
    {
      "epoch": 0.36,
      "grad_norm": 10.665780067443848,
      "learning_rate": 2.4658180778032037e-05,
      "loss": 1.4181,
      "step": 1245
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.919100761413574,
      "learning_rate": 2.465389016018307e-05,
      "loss": 1.3317,
      "step": 1246
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.487162590026855,
      "learning_rate": 2.4649599542334097e-05,
      "loss": 1.3828,
      "step": 1247
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.3540678024292,
      "learning_rate": 2.4645308924485126e-05,
      "loss": 1.1949,
      "step": 1248
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.808308601379395,
      "learning_rate": 2.4641018306636155e-05,
      "loss": 1.156,
      "step": 1249
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.10568904876709,
      "learning_rate": 2.4636727688787187e-05,
      "loss": 1.387,
      "step": 1250
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.04727840423584,
      "learning_rate": 2.4632437070938212e-05,
      "loss": 1.1991,
      "step": 1251
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.775404930114746,
      "learning_rate": 2.4628146453089245e-05,
      "loss": 1.5236,
      "step": 1252
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.491841316223145,
      "learning_rate": 2.4623855835240277e-05,
      "loss": 1.1701,
      "step": 1253
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.250205039978027,
      "learning_rate": 2.4619565217391306e-05,
      "loss": 1.4561,
      "step": 1254
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.8140230178833,
      "learning_rate": 2.4615274599542334e-05,
      "loss": 1.1453,
      "step": 1255
    },
    {
      "epoch": 0.36,
      "grad_norm": 10.00520133972168,
      "learning_rate": 2.4610983981693363e-05,
      "loss": 1.4822,
      "step": 1256
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.018996238708496,
      "learning_rate": 2.4606693363844395e-05,
      "loss": 1.0052,
      "step": 1257
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.236218452453613,
      "learning_rate": 2.4602402745995424e-05,
      "loss": 1.6975,
      "step": 1258
    },
    {
      "epoch": 0.36,
      "grad_norm": 7.647527694702148,
      "learning_rate": 2.4598112128146453e-05,
      "loss": 1.2796,
      "step": 1259
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.628692626953125,
      "learning_rate": 2.4593821510297485e-05,
      "loss": 1.2854,
      "step": 1260
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.866012573242188,
      "learning_rate": 2.4589530892448514e-05,
      "loss": 1.3675,
      "step": 1261
    },
    {
      "epoch": 0.36,
      "grad_norm": 10.024090766906738,
      "learning_rate": 2.4585240274599542e-05,
      "loss": 1.3188,
      "step": 1262
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.781789779663086,
      "learning_rate": 2.458094965675057e-05,
      "loss": 1.4322,
      "step": 1263
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.879053115844727,
      "learning_rate": 2.4576659038901603e-05,
      "loss": 1.437,
      "step": 1264
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.417719841003418,
      "learning_rate": 2.4572368421052632e-05,
      "loss": 1.0245,
      "step": 1265
    },
    {
      "epoch": 0.36,
      "grad_norm": 11.67907428741455,
      "learning_rate": 2.456807780320366e-05,
      "loss": 1.1642,
      "step": 1266
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.326916694641113,
      "learning_rate": 2.456378718535469e-05,
      "loss": 1.547,
      "step": 1267
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.553854942321777,
      "learning_rate": 2.4559496567505722e-05,
      "loss": 1.4601,
      "step": 1268
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.98757266998291,
      "learning_rate": 2.4555205949656754e-05,
      "loss": 1.1181,
      "step": 1269
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.45922565460205,
      "learning_rate": 2.455091533180778e-05,
      "loss": 1.0405,
      "step": 1270
    },
    {
      "epoch": 0.36,
      "grad_norm": 10.106505393981934,
      "learning_rate": 2.454662471395881e-05,
      "loss": 1.4776,
      "step": 1271
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.02731990814209,
      "learning_rate": 2.454233409610984e-05,
      "loss": 1.6652,
      "step": 1272
    },
    {
      "epoch": 0.36,
      "grad_norm": 10.727587699890137,
      "learning_rate": 2.4538043478260872e-05,
      "loss": 1.2804,
      "step": 1273
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.161126136779785,
      "learning_rate": 2.4533752860411898e-05,
      "loss": 1.3514,
      "step": 1274
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.589289665222168,
      "learning_rate": 2.452946224256293e-05,
      "loss": 1.3186,
      "step": 1275
    },
    {
      "epoch": 0.36,
      "grad_norm": 7.296161651611328,
      "learning_rate": 2.452517162471396e-05,
      "loss": 1.2932,
      "step": 1276
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.590826034545898,
      "learning_rate": 2.4520881006864988e-05,
      "loss": 1.1958,
      "step": 1277
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.288223266601562,
      "learning_rate": 2.451659038901602e-05,
      "loss": 1.1643,
      "step": 1278
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.862256050109863,
      "learning_rate": 2.451229977116705e-05,
      "loss": 1.5331,
      "step": 1279
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.638525009155273,
      "learning_rate": 2.450800915331808e-05,
      "loss": 1.3783,
      "step": 1280
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.808732986450195,
      "learning_rate": 2.4503718535469106e-05,
      "loss": 1.4159,
      "step": 1281
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.5829496383667,
      "learning_rate": 2.4499427917620138e-05,
      "loss": 1.4327,
      "step": 1282
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.102815628051758,
      "learning_rate": 2.4495137299771167e-05,
      "loss": 1.4051,
      "step": 1283
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.11328125,
      "learning_rate": 2.44908466819222e-05,
      "loss": 1.4893,
      "step": 1284
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.082062721252441,
      "learning_rate": 2.4486556064073224e-05,
      "loss": 1.1809,
      "step": 1285
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.857675552368164,
      "learning_rate": 2.4482265446224257e-05,
      "loss": 1.3247,
      "step": 1286
    },
    {
      "epoch": 0.37,
      "grad_norm": 7.963157653808594,
      "learning_rate": 2.447797482837529e-05,
      "loss": 1.0194,
      "step": 1287
    },
    {
      "epoch": 0.37,
      "grad_norm": 7.50498628616333,
      "learning_rate": 2.4473684210526318e-05,
      "loss": 0.9924,
      "step": 1288
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.573609352111816,
      "learning_rate": 2.4469393592677346e-05,
      "loss": 1.6141,
      "step": 1289
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.218179702758789,
      "learning_rate": 2.4465102974828375e-05,
      "loss": 1.365,
      "step": 1290
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.033930778503418,
      "learning_rate": 2.4460812356979407e-05,
      "loss": 1.4356,
      "step": 1291
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.373608589172363,
      "learning_rate": 2.4456521739130433e-05,
      "loss": 1.4214,
      "step": 1292
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.133663177490234,
      "learning_rate": 2.4452231121281465e-05,
      "loss": 1.2398,
      "step": 1293
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.034195899963379,
      "learning_rate": 2.4447940503432497e-05,
      "loss": 1.3809,
      "step": 1294
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.040098190307617,
      "learning_rate": 2.4443649885583526e-05,
      "loss": 1.3916,
      "step": 1295
    },
    {
      "epoch": 0.37,
      "grad_norm": 7.705639362335205,
      "learning_rate": 2.4439359267734554e-05,
      "loss": 1.3414,
      "step": 1296
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.405379295349121,
      "learning_rate": 2.4435068649885583e-05,
      "loss": 1.5481,
      "step": 1297
    },
    {
      "epoch": 0.37,
      "grad_norm": 7.484687805175781,
      "learning_rate": 2.4430778032036615e-05,
      "loss": 1.2209,
      "step": 1298
    },
    {
      "epoch": 0.37,
      "grad_norm": 7.746870994567871,
      "learning_rate": 2.4426487414187644e-05,
      "loss": 1.2595,
      "step": 1299
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.075613975524902,
      "learning_rate": 2.4422196796338673e-05,
      "loss": 1.2402,
      "step": 1300
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.48637580871582,
      "learning_rate": 2.4417906178489702e-05,
      "loss": 1.4072,
      "step": 1301
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.419695854187012,
      "learning_rate": 2.4413615560640734e-05,
      "loss": 0.9975,
      "step": 1302
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.64478874206543,
      "learning_rate": 2.4409324942791763e-05,
      "loss": 1.2916,
      "step": 1303
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.342888832092285,
      "learning_rate": 2.440503432494279e-05,
      "loss": 1.3963,
      "step": 1304
    },
    {
      "epoch": 0.37,
      "grad_norm": 7.8090362548828125,
      "learning_rate": 2.4400743707093824e-05,
      "loss": 1.1569,
      "step": 1305
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.666332244873047,
      "learning_rate": 2.4396453089244852e-05,
      "loss": 1.4286,
      "step": 1306
    },
    {
      "epoch": 0.37,
      "grad_norm": 10.178092956542969,
      "learning_rate": 2.439216247139588e-05,
      "loss": 1.7242,
      "step": 1307
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.932698249816895,
      "learning_rate": 2.438787185354691e-05,
      "loss": 1.2023,
      "step": 1308
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.65407657623291,
      "learning_rate": 2.4383581235697942e-05,
      "loss": 1.299,
      "step": 1309
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.978166580200195,
      "learning_rate": 2.437929061784897e-05,
      "loss": 1.3771,
      "step": 1310
    },
    {
      "epoch": 0.38,
      "grad_norm": 7.935179710388184,
      "learning_rate": 2.4375e-05,
      "loss": 1.3486,
      "step": 1311
    },
    {
      "epoch": 0.38,
      "grad_norm": 7.84268045425415,
      "learning_rate": 2.4370709382151032e-05,
      "loss": 1.0263,
      "step": 1312
    },
    {
      "epoch": 0.38,
      "grad_norm": 7.6278228759765625,
      "learning_rate": 2.436641876430206e-05,
      "loss": 1.2904,
      "step": 1313
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.676783561706543,
      "learning_rate": 2.436212814645309e-05,
      "loss": 0.9318,
      "step": 1314
    },
    {
      "epoch": 0.38,
      "grad_norm": 7.841330528259277,
      "learning_rate": 2.4357837528604118e-05,
      "loss": 1.1714,
      "step": 1315
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.406579971313477,
      "learning_rate": 2.435354691075515e-05,
      "loss": 1.2257,
      "step": 1316
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.868280410766602,
      "learning_rate": 2.434925629290618e-05,
      "loss": 1.3,
      "step": 1317
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.387145042419434,
      "learning_rate": 2.4344965675057208e-05,
      "loss": 1.1957,
      "step": 1318
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.952069282531738,
      "learning_rate": 2.4340675057208236e-05,
      "loss": 1.0876,
      "step": 1319
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.023176193237305,
      "learning_rate": 2.433638443935927e-05,
      "loss": 1.3813,
      "step": 1320
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.673901557922363,
      "learning_rate": 2.43320938215103e-05,
      "loss": 1.179,
      "step": 1321
    },
    {
      "epoch": 0.38,
      "grad_norm": 11.203038215637207,
      "learning_rate": 2.4327803203661326e-05,
      "loss": 1.0637,
      "step": 1322
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.70146656036377,
      "learning_rate": 2.432351258581236e-05,
      "loss": 1.3067,
      "step": 1323
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.276034355163574,
      "learning_rate": 2.4319221967963387e-05,
      "loss": 1.3299,
      "step": 1324
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.101346015930176,
      "learning_rate": 2.431493135011442e-05,
      "loss": 1.3958,
      "step": 1325
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.510576248168945,
      "learning_rate": 2.4310640732265445e-05,
      "loss": 1.1045,
      "step": 1326
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.810791969299316,
      "learning_rate": 2.4306350114416477e-05,
      "loss": 1.2913,
      "step": 1327
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.729816436767578,
      "learning_rate": 2.4302059496567506e-05,
      "loss": 1.2609,
      "step": 1328
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.372541427612305,
      "learning_rate": 2.4297768878718534e-05,
      "loss": 1.4533,
      "step": 1329
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.771547317504883,
      "learning_rate": 2.4293478260869566e-05,
      "loss": 1.376,
      "step": 1330
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.162940979003906,
      "learning_rate": 2.4289187643020595e-05,
      "loss": 1.2187,
      "step": 1331
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.148791313171387,
      "learning_rate": 2.4284897025171627e-05,
      "loss": 1.3383,
      "step": 1332
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.023940086364746,
      "learning_rate": 2.4280606407322653e-05,
      "loss": 1.0984,
      "step": 1333
    },
    {
      "epoch": 0.38,
      "grad_norm": 7.9747843742370605,
      "learning_rate": 2.4276315789473685e-05,
      "loss": 1.1721,
      "step": 1334
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.558140754699707,
      "learning_rate": 2.4272025171624714e-05,
      "loss": 1.2776,
      "step": 1335
    },
    {
      "epoch": 0.38,
      "grad_norm": 11.381139755249023,
      "learning_rate": 2.4267734553775746e-05,
      "loss": 1.1888,
      "step": 1336
    },
    {
      "epoch": 0.38,
      "grad_norm": 10.013339042663574,
      "learning_rate": 2.4263443935926775e-05,
      "loss": 1.5181,
      "step": 1337
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.366135597229004,
      "learning_rate": 2.4259153318077803e-05,
      "loss": 1.1374,
      "step": 1338
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.258301734924316,
      "learning_rate": 2.4254862700228836e-05,
      "loss": 1.1101,
      "step": 1339
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.093093872070312,
      "learning_rate": 2.4250572082379864e-05,
      "loss": 1.0627,
      "step": 1340
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.383330345153809,
      "learning_rate": 2.4246281464530893e-05,
      "loss": 1.1353,
      "step": 1341
    },
    {
      "epoch": 0.38,
      "grad_norm": 10.459028244018555,
      "learning_rate": 2.4241990846681922e-05,
      "loss": 1.2926,
      "step": 1342
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.990187644958496,
      "learning_rate": 2.4237700228832954e-05,
      "loss": 1.4097,
      "step": 1343
    },
    {
      "epoch": 0.38,
      "grad_norm": 11.132209777832031,
      "learning_rate": 2.423340961098398e-05,
      "loss": 1.603,
      "step": 1344
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.05010986328125,
      "learning_rate": 2.422911899313501e-05,
      "loss": 1.1662,
      "step": 1345
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.722626686096191,
      "learning_rate": 2.4224828375286044e-05,
      "loss": 1.3975,
      "step": 1346
    },
    {
      "epoch": 0.39,
      "grad_norm": 11.301931381225586,
      "learning_rate": 2.4220537757437072e-05,
      "loss": 1.6098,
      "step": 1347
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.93022346496582,
      "learning_rate": 2.42162471395881e-05,
      "loss": 1.299,
      "step": 1348
    },
    {
      "epoch": 0.39,
      "grad_norm": 7.965664386749268,
      "learning_rate": 2.421195652173913e-05,
      "loss": 1.1455,
      "step": 1349
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.106415748596191,
      "learning_rate": 2.4207665903890162e-05,
      "loss": 1.146,
      "step": 1350
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.92429256439209,
      "learning_rate": 2.420337528604119e-05,
      "loss": 1.5329,
      "step": 1351
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.482048988342285,
      "learning_rate": 2.419908466819222e-05,
      "loss": 1.4407,
      "step": 1352
    },
    {
      "epoch": 0.39,
      "grad_norm": 7.7144904136657715,
      "learning_rate": 2.419479405034325e-05,
      "loss": 1.1205,
      "step": 1353
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.001365661621094,
      "learning_rate": 2.419050343249428e-05,
      "loss": 1.1997,
      "step": 1354
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.95022201538086,
      "learning_rate": 2.418621281464531e-05,
      "loss": 1.3606,
      "step": 1355
    },
    {
      "epoch": 0.39,
      "grad_norm": 7.322661876678467,
      "learning_rate": 2.4181922196796338e-05,
      "loss": 1.0748,
      "step": 1356
    },
    {
      "epoch": 0.39,
      "grad_norm": 7.953974723815918,
      "learning_rate": 2.417763157894737e-05,
      "loss": 1.0571,
      "step": 1357
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.130672454833984,
      "learning_rate": 2.41733409610984e-05,
      "loss": 1.2434,
      "step": 1358
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.41163444519043,
      "learning_rate": 2.4169050343249428e-05,
      "loss": 1.505,
      "step": 1359
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.805686950683594,
      "learning_rate": 2.4164759725400457e-05,
      "loss": 1.3325,
      "step": 1360
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.738990783691406,
      "learning_rate": 2.416046910755149e-05,
      "loss": 1.116,
      "step": 1361
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.374129295349121,
      "learning_rate": 2.4156178489702518e-05,
      "loss": 1.4156,
      "step": 1362
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.01900863647461,
      "learning_rate": 2.4151887871853546e-05,
      "loss": 1.3289,
      "step": 1363
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.995376586914062,
      "learning_rate": 2.414759725400458e-05,
      "loss": 1.7106,
      "step": 1364
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.745902061462402,
      "learning_rate": 2.4143306636155607e-05,
      "loss": 1.6264,
      "step": 1365
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.528644561767578,
      "learning_rate": 2.413901601830664e-05,
      "loss": 1.1799,
      "step": 1366
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.402424812316895,
      "learning_rate": 2.4134725400457665e-05,
      "loss": 1.2642,
      "step": 1367
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.838375091552734,
      "learning_rate": 2.4130434782608697e-05,
      "loss": 1.5943,
      "step": 1368
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.3024320602417,
      "learning_rate": 2.4126144164759726e-05,
      "loss": 1.3896,
      "step": 1369
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.33349323272705,
      "learning_rate": 2.4121853546910754e-05,
      "loss": 1.4866,
      "step": 1370
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.199357986450195,
      "learning_rate": 2.4117562929061787e-05,
      "loss": 1.2041,
      "step": 1371
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.57052230834961,
      "learning_rate": 2.4113272311212815e-05,
      "loss": 1.2684,
      "step": 1372
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.016642570495605,
      "learning_rate": 2.4108981693363848e-05,
      "loss": 0.9034,
      "step": 1373
    },
    {
      "epoch": 0.39,
      "grad_norm": 7.731619358062744,
      "learning_rate": 2.4104691075514873e-05,
      "loss": 0.9991,
      "step": 1374
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.050274848937988,
      "learning_rate": 2.4100400457665905e-05,
      "loss": 1.1917,
      "step": 1375
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.060944557189941,
      "learning_rate": 2.4096109839816934e-05,
      "loss": 1.1226,
      "step": 1376
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.330954551696777,
      "learning_rate": 2.4091819221967966e-05,
      "loss": 1.2132,
      "step": 1377
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.34937858581543,
      "learning_rate": 2.408752860411899e-05,
      "loss": 1.2008,
      "step": 1378
    },
    {
      "epoch": 0.39,
      "grad_norm": 10.294057846069336,
      "learning_rate": 2.4083237986270024e-05,
      "loss": 1.51,
      "step": 1379
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.819464683532715,
      "learning_rate": 2.4078947368421056e-05,
      "loss": 1.3065,
      "step": 1380
    },
    {
      "epoch": 0.4,
      "grad_norm": 11.210472106933594,
      "learning_rate": 2.407465675057208e-05,
      "loss": 1.3412,
      "step": 1381
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.68340015411377,
      "learning_rate": 2.4070366132723113e-05,
      "loss": 1.2003,
      "step": 1382
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.639593124389648,
      "learning_rate": 2.4066075514874142e-05,
      "loss": 1.1033,
      "step": 1383
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.134466171264648,
      "learning_rate": 2.4061784897025174e-05,
      "loss": 1.2797,
      "step": 1384
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.370076179504395,
      "learning_rate": 2.40574942791762e-05,
      "loss": 1.6027,
      "step": 1385
    },
    {
      "epoch": 0.4,
      "grad_norm": 10.915976524353027,
      "learning_rate": 2.405320366132723e-05,
      "loss": 1.5843,
      "step": 1386
    },
    {
      "epoch": 0.4,
      "grad_norm": 11.412590026855469,
      "learning_rate": 2.404891304347826e-05,
      "loss": 1.6253,
      "step": 1387
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.945582389831543,
      "learning_rate": 2.4044622425629293e-05,
      "loss": 1.3744,
      "step": 1388
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.530177116394043,
      "learning_rate": 2.404033180778032e-05,
      "loss": 0.8811,
      "step": 1389
    },
    {
      "epoch": 0.4,
      "grad_norm": 7.074493885040283,
      "learning_rate": 2.403604118993135e-05,
      "loss": 0.9554,
      "step": 1390
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.388182640075684,
      "learning_rate": 2.4031750572082382e-05,
      "loss": 1.1566,
      "step": 1391
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.897048950195312,
      "learning_rate": 2.402745995423341e-05,
      "loss": 1.041,
      "step": 1392
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.569944381713867,
      "learning_rate": 2.402316933638444e-05,
      "loss": 1.4443,
      "step": 1393
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.350441932678223,
      "learning_rate": 2.401887871853547e-05,
      "loss": 1.2276,
      "step": 1394
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.618452072143555,
      "learning_rate": 2.40145881006865e-05,
      "loss": 1.2577,
      "step": 1395
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.834698677062988,
      "learning_rate": 2.4010297482837526e-05,
      "loss": 1.4529,
      "step": 1396
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.187966346740723,
      "learning_rate": 2.4006006864988558e-05,
      "loss": 1.1598,
      "step": 1397
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.153619766235352,
      "learning_rate": 2.400171624713959e-05,
      "loss": 1.5876,
      "step": 1398
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.023094177246094,
      "learning_rate": 2.399742562929062e-05,
      "loss": 1.1627,
      "step": 1399
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.275665283203125,
      "learning_rate": 2.3993135011441648e-05,
      "loss": 1.2592,
      "step": 1400
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.038787841796875,
      "learning_rate": 2.3988844393592677e-05,
      "loss": 1.0274,
      "step": 1401
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.471991539001465,
      "learning_rate": 2.398455377574371e-05,
      "loss": 1.2368,
      "step": 1402
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.320343017578125,
      "learning_rate": 2.3980263157894738e-05,
      "loss": 1.0732,
      "step": 1403
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.846017837524414,
      "learning_rate": 2.3975972540045766e-05,
      "loss": 1.1362,
      "step": 1404
    },
    {
      "epoch": 0.4,
      "grad_norm": 11.177179336547852,
      "learning_rate": 2.3971681922196795e-05,
      "loss": 1.3263,
      "step": 1405
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.469406127929688,
      "learning_rate": 2.3967391304347827e-05,
      "loss": 1.2119,
      "step": 1406
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.226251602172852,
      "learning_rate": 2.3963100686498856e-05,
      "loss": 1.5254,
      "step": 1407
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.160237312316895,
      "learning_rate": 2.3958810068649885e-05,
      "loss": 1.2009,
      "step": 1408
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.037147521972656,
      "learning_rate": 2.3954519450800917e-05,
      "loss": 1.2019,
      "step": 1409
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.3545560836792,
      "learning_rate": 2.3950228832951946e-05,
      "loss": 1.0934,
      "step": 1410
    },
    {
      "epoch": 0.4,
      "grad_norm": 7.292141437530518,
      "learning_rate": 2.3945938215102975e-05,
      "loss": 1.057,
      "step": 1411
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.759928703308105,
      "learning_rate": 2.3941647597254003e-05,
      "loss": 1.2076,
      "step": 1412
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.506291389465332,
      "learning_rate": 2.3937356979405036e-05,
      "loss": 1.3331,
      "step": 1413
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.5624418258667,
      "learning_rate": 2.3933066361556068e-05,
      "loss": 1.3577,
      "step": 1414
    },
    {
      "epoch": 0.4,
      "grad_norm": 7.361660003662109,
      "learning_rate": 2.3928775743707093e-05,
      "loss": 0.9031,
      "step": 1415
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.643257141113281,
      "learning_rate": 2.3924485125858125e-05,
      "loss": 1.2242,
      "step": 1416
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.304159164428711,
      "learning_rate": 2.3920194508009154e-05,
      "loss": 1.1468,
      "step": 1417
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.026140213012695,
      "learning_rate": 2.3915903890160186e-05,
      "loss": 1.1251,
      "step": 1418
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.157430648803711,
      "learning_rate": 2.391161327231121e-05,
      "loss": 1.2867,
      "step": 1419
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.46143627166748,
      "learning_rate": 2.3907322654462244e-05,
      "loss": 1.1906,
      "step": 1420
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.104385375976562,
      "learning_rate": 2.3903032036613272e-05,
      "loss": 1.0685,
      "step": 1421
    },
    {
      "epoch": 0.41,
      "grad_norm": 7.670660018920898,
      "learning_rate": 2.38987414187643e-05,
      "loss": 1.0857,
      "step": 1422
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.949851989746094,
      "learning_rate": 2.3894450800915333e-05,
      "loss": 1.1966,
      "step": 1423
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.842041015625,
      "learning_rate": 2.3890160183066362e-05,
      "loss": 1.4296,
      "step": 1424
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.33974838256836,
      "learning_rate": 2.3885869565217394e-05,
      "loss": 1.6068,
      "step": 1425
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.253230094909668,
      "learning_rate": 2.388157894736842e-05,
      "loss": 1.0953,
      "step": 1426
    },
    {
      "epoch": 0.41,
      "grad_norm": 10.492298126220703,
      "learning_rate": 2.3877288329519452e-05,
      "loss": 1.6289,
      "step": 1427
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.839768409729004,
      "learning_rate": 2.387299771167048e-05,
      "loss": 1.2073,
      "step": 1428
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.604236602783203,
      "learning_rate": 2.3868707093821513e-05,
      "loss": 1.5861,
      "step": 1429
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.197074890136719,
      "learning_rate": 2.3864416475972538e-05,
      "loss": 1.3949,
      "step": 1430
    },
    {
      "epoch": 0.41,
      "grad_norm": 7.777657985687256,
      "learning_rate": 2.386012585812357e-05,
      "loss": 1.1687,
      "step": 1431
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.555583953857422,
      "learning_rate": 2.3855835240274602e-05,
      "loss": 1.2248,
      "step": 1432
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.336581230163574,
      "learning_rate": 2.3851544622425628e-05,
      "loss": 1.2914,
      "step": 1433
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.059797286987305,
      "learning_rate": 2.384725400457666e-05,
      "loss": 1.1301,
      "step": 1434
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.82107925415039,
      "learning_rate": 2.384296338672769e-05,
      "loss": 1.3305,
      "step": 1435
    },
    {
      "epoch": 0.41,
      "grad_norm": 6.868978023529053,
      "learning_rate": 2.383867276887872e-05,
      "loss": 1.1569,
      "step": 1436
    },
    {
      "epoch": 0.41,
      "grad_norm": 7.062217712402344,
      "learning_rate": 2.3834382151029746e-05,
      "loss": 1.2047,
      "step": 1437
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.001330375671387,
      "learning_rate": 2.383009153318078e-05,
      "loss": 1.5471,
      "step": 1438
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.982261657714844,
      "learning_rate": 2.3825800915331807e-05,
      "loss": 1.5064,
      "step": 1439
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.749502182006836,
      "learning_rate": 2.382151029748284e-05,
      "loss": 1.3297,
      "step": 1440
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.250558853149414,
      "learning_rate": 2.3817219679633868e-05,
      "loss": 1.2343,
      "step": 1441
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.083165168762207,
      "learning_rate": 2.3812929061784897e-05,
      "loss": 1.1752,
      "step": 1442
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.45124626159668,
      "learning_rate": 2.380863844393593e-05,
      "loss": 1.3145,
      "step": 1443
    },
    {
      "epoch": 0.41,
      "grad_norm": 7.276574611663818,
      "learning_rate": 2.3804347826086958e-05,
      "loss": 1.0405,
      "step": 1444
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.243409156799316,
      "learning_rate": 2.3800057208237987e-05,
      "loss": 1.2529,
      "step": 1445
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.41983699798584,
      "learning_rate": 2.3795766590389015e-05,
      "loss": 1.2946,
      "step": 1446
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.295051574707031,
      "learning_rate": 2.3791475972540048e-05,
      "loss": 1.2942,
      "step": 1447
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.93205738067627,
      "learning_rate": 2.3787185354691073e-05,
      "loss": 1.3353,
      "step": 1448
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.91295051574707,
      "learning_rate": 2.3782894736842105e-05,
      "loss": 1.3193,
      "step": 1449
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.965539932250977,
      "learning_rate": 2.3778604118993137e-05,
      "loss": 1.2203,
      "step": 1450
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.239827156066895,
      "learning_rate": 2.3774313501144166e-05,
      "loss": 1.2644,
      "step": 1451
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.137190818786621,
      "learning_rate": 2.3770022883295195e-05,
      "loss": 1.3303,
      "step": 1452
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.793137550354004,
      "learning_rate": 2.3765732265446224e-05,
      "loss": 1.3811,
      "step": 1453
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.519445419311523,
      "learning_rate": 2.3761441647597256e-05,
      "loss": 1.4462,
      "step": 1454
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.207239151000977,
      "learning_rate": 2.3757151029748284e-05,
      "loss": 1.3354,
      "step": 1455
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.833036422729492,
      "learning_rate": 2.3752860411899313e-05,
      "loss": 1.5372,
      "step": 1456
    },
    {
      "epoch": 0.42,
      "grad_norm": 10.283196449279785,
      "learning_rate": 2.3748569794050345e-05,
      "loss": 1.1521,
      "step": 1457
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.274222373962402,
      "learning_rate": 2.3744279176201374e-05,
      "loss": 1.4175,
      "step": 1458
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.186198234558105,
      "learning_rate": 2.3739988558352403e-05,
      "loss": 1.3154,
      "step": 1459
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.352169036865234,
      "learning_rate": 2.373569794050343e-05,
      "loss": 1.2017,
      "step": 1460
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.348889350891113,
      "learning_rate": 2.3731407322654464e-05,
      "loss": 1.2981,
      "step": 1461
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.70347785949707,
      "learning_rate": 2.3727116704805493e-05,
      "loss": 1.5687,
      "step": 1462
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.238858222961426,
      "learning_rate": 2.372282608695652e-05,
      "loss": 1.2132,
      "step": 1463
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.277292251586914,
      "learning_rate": 2.371853546910755e-05,
      "loss": 1.0626,
      "step": 1464
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.16200065612793,
      "learning_rate": 2.3714244851258582e-05,
      "loss": 1.5139,
      "step": 1465
    },
    {
      "epoch": 0.42,
      "grad_norm": 7.976869583129883,
      "learning_rate": 2.3709954233409614e-05,
      "loss": 1.353,
      "step": 1466
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.776309967041016,
      "learning_rate": 2.370566361556064e-05,
      "loss": 1.3921,
      "step": 1467
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.482604026794434,
      "learning_rate": 2.3701372997711672e-05,
      "loss": 1.5642,
      "step": 1468
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.380767822265625,
      "learning_rate": 2.36970823798627e-05,
      "loss": 1.3062,
      "step": 1469
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.498199462890625,
      "learning_rate": 2.3692791762013733e-05,
      "loss": 1.4111,
      "step": 1470
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.320439338684082,
      "learning_rate": 2.3688501144164758e-05,
      "loss": 1.3205,
      "step": 1471
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.131627082824707,
      "learning_rate": 2.368421052631579e-05,
      "loss": 1.4386,
      "step": 1472
    },
    {
      "epoch": 0.42,
      "grad_norm": 7.26507043838501,
      "learning_rate": 2.367991990846682e-05,
      "loss": 1.2513,
      "step": 1473
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.468502044677734,
      "learning_rate": 2.3675629290617848e-05,
      "loss": 1.0528,
      "step": 1474
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.420333862304688,
      "learning_rate": 2.367133867276888e-05,
      "loss": 1.3885,
      "step": 1475
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.6135892868042,
      "learning_rate": 2.366704805491991e-05,
      "loss": 1.2761,
      "step": 1476
    },
    {
      "epoch": 0.42,
      "grad_norm": 7.293272972106934,
      "learning_rate": 2.366275743707094e-05,
      "loss": 0.8257,
      "step": 1477
    },
    {
      "epoch": 0.42,
      "grad_norm": 7.460020065307617,
      "learning_rate": 2.3658466819221966e-05,
      "loss": 1.0788,
      "step": 1478
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.232573509216309,
      "learning_rate": 2.3654176201373e-05,
      "loss": 1.4714,
      "step": 1479
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.258026123046875,
      "learning_rate": 2.3649885583524027e-05,
      "loss": 1.1303,
      "step": 1480
    },
    {
      "epoch": 0.42,
      "grad_norm": 10.65665340423584,
      "learning_rate": 2.364559496567506e-05,
      "loss": 1.3665,
      "step": 1481
    },
    {
      "epoch": 0.42,
      "grad_norm": 11.075395584106445,
      "learning_rate": 2.3641304347826085e-05,
      "loss": 1.1579,
      "step": 1482
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.51037311553955,
      "learning_rate": 2.3637013729977117e-05,
      "loss": 1.1477,
      "step": 1483
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.741596221923828,
      "learning_rate": 2.363272311212815e-05,
      "loss": 1.14,
      "step": 1484
    },
    {
      "epoch": 0.42,
      "grad_norm": 11.098268508911133,
      "learning_rate": 2.3628432494279178e-05,
      "loss": 1.4584,
      "step": 1485
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.322772026062012,
      "learning_rate": 2.3624141876430207e-05,
      "loss": 1.26,
      "step": 1486
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.70244312286377,
      "learning_rate": 2.3619851258581236e-05,
      "loss": 1.3188,
      "step": 1487
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.393972396850586,
      "learning_rate": 2.3615560640732268e-05,
      "loss": 1.3285,
      "step": 1488
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.17244815826416,
      "learning_rate": 2.3611270022883293e-05,
      "loss": 1.2333,
      "step": 1489
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.737652778625488,
      "learning_rate": 2.3606979405034325e-05,
      "loss": 1.4713,
      "step": 1490
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.080338478088379,
      "learning_rate": 2.3602688787185357e-05,
      "loss": 1.0657,
      "step": 1491
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.210623741149902,
      "learning_rate": 2.3598398169336386e-05,
      "loss": 1.0281,
      "step": 1492
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.44229507446289,
      "learning_rate": 2.3594107551487415e-05,
      "loss": 1.3468,
      "step": 1493
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.192740440368652,
      "learning_rate": 2.3589816933638444e-05,
      "loss": 1.4332,
      "step": 1494
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.328465461730957,
      "learning_rate": 2.3585526315789476e-05,
      "loss": 1.2152,
      "step": 1495
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.111274719238281,
      "learning_rate": 2.3581235697940505e-05,
      "loss": 1.1548,
      "step": 1496
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.120012283325195,
      "learning_rate": 2.3576945080091533e-05,
      "loss": 1.2525,
      "step": 1497
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.653223991394043,
      "learning_rate": 2.3572654462242562e-05,
      "loss": 1.5593,
      "step": 1498
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.163555145263672,
      "learning_rate": 2.3568363844393594e-05,
      "loss": 1.1074,
      "step": 1499
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.843320846557617,
      "learning_rate": 2.3564073226544623e-05,
      "loss": 1.6794,
      "step": 1500
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.1924409866333,
      "learning_rate": 2.3559782608695652e-05,
      "loss": 1.1617,
      "step": 1501
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.335710525512695,
      "learning_rate": 2.3555491990846684e-05,
      "loss": 1.2226,
      "step": 1502
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.478708267211914,
      "learning_rate": 2.3551201372997713e-05,
      "loss": 1.3615,
      "step": 1503
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.279967308044434,
      "learning_rate": 2.354691075514874e-05,
      "loss": 1.2645,
      "step": 1504
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.7472562789917,
      "learning_rate": 2.354262013729977e-05,
      "loss": 1.4243,
      "step": 1505
    },
    {
      "epoch": 0.43,
      "grad_norm": 10.629935264587402,
      "learning_rate": 2.3538329519450802e-05,
      "loss": 1.7176,
      "step": 1506
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.287959098815918,
      "learning_rate": 2.353403890160183e-05,
      "loss": 1.2294,
      "step": 1507
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.736032485961914,
      "learning_rate": 2.352974828375286e-05,
      "loss": 1.3285,
      "step": 1508
    },
    {
      "epoch": 0.43,
      "grad_norm": 7.25554895401001,
      "learning_rate": 2.3525457665903892e-05,
      "loss": 1.0597,
      "step": 1509
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.317910194396973,
      "learning_rate": 2.352116704805492e-05,
      "loss": 1.2206,
      "step": 1510
    },
    {
      "epoch": 0.43,
      "grad_norm": 7.878781318664551,
      "learning_rate": 2.3516876430205953e-05,
      "loss": 1.3455,
      "step": 1511
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.274806022644043,
      "learning_rate": 2.351258581235698e-05,
      "loss": 1.1493,
      "step": 1512
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.328680038452148,
      "learning_rate": 2.350829519450801e-05,
      "loss": 1.0924,
      "step": 1513
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.172496795654297,
      "learning_rate": 2.350400457665904e-05,
      "loss": 1.3245,
      "step": 1514
    },
    {
      "epoch": 0.43,
      "grad_norm": 7.963389873504639,
      "learning_rate": 2.3499713958810068e-05,
      "loss": 1.1043,
      "step": 1515
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.72868537902832,
      "learning_rate": 2.3495423340961097e-05,
      "loss": 1.3213,
      "step": 1516
    },
    {
      "epoch": 0.43,
      "grad_norm": 11.010828018188477,
      "learning_rate": 2.349113272311213e-05,
      "loss": 1.3094,
      "step": 1517
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.427931785583496,
      "learning_rate": 2.348684210526316e-05,
      "loss": 1.4155,
      "step": 1518
    },
    {
      "epoch": 0.43,
      "grad_norm": 10.318192481994629,
      "learning_rate": 2.3482551487414187e-05,
      "loss": 1.3738,
      "step": 1519
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.963619232177734,
      "learning_rate": 2.347826086956522e-05,
      "loss": 1.1076,
      "step": 1520
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.685846328735352,
      "learning_rate": 2.3473970251716247e-05,
      "loss": 1.2785,
      "step": 1521
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.562762260437012,
      "learning_rate": 2.346967963386728e-05,
      "loss": 1.2463,
      "step": 1522
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.336532592773438,
      "learning_rate": 2.3465389016018305e-05,
      "loss": 1.1253,
      "step": 1523
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.443828582763672,
      "learning_rate": 2.3461098398169337e-05,
      "loss": 1.0381,
      "step": 1524
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.334905624389648,
      "learning_rate": 2.3456807780320366e-05,
      "loss": 1.1564,
      "step": 1525
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.118339538574219,
      "learning_rate": 2.3452517162471395e-05,
      "loss": 1.3133,
      "step": 1526
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.523225784301758,
      "learning_rate": 2.3448226544622427e-05,
      "loss": 1.266,
      "step": 1527
    },
    {
      "epoch": 0.44,
      "grad_norm": 10.204797744750977,
      "learning_rate": 2.3443935926773456e-05,
      "loss": 1.5447,
      "step": 1528
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.629857063293457,
      "learning_rate": 2.3439645308924488e-05,
      "loss": 1.2232,
      "step": 1529
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.885933876037598,
      "learning_rate": 2.3435354691075513e-05,
      "loss": 1.245,
      "step": 1530
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.119367599487305,
      "learning_rate": 2.3431064073226545e-05,
      "loss": 1.1223,
      "step": 1531
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.569046020507812,
      "learning_rate": 2.3426773455377574e-05,
      "loss": 1.3702,
      "step": 1532
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.757391929626465,
      "learning_rate": 2.3422482837528606e-05,
      "loss": 1.8398,
      "step": 1533
    },
    {
      "epoch": 0.44,
      "grad_norm": 7.991138935089111,
      "learning_rate": 2.3418192219679635e-05,
      "loss": 1.3525,
      "step": 1534
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.762513160705566,
      "learning_rate": 2.3413901601830664e-05,
      "loss": 1.3359,
      "step": 1535
    },
    {
      "epoch": 0.44,
      "grad_norm": 10.335434913635254,
      "learning_rate": 2.3409610983981696e-05,
      "loss": 1.1716,
      "step": 1536
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.94473934173584,
      "learning_rate": 2.3405320366132725e-05,
      "loss": 1.079,
      "step": 1537
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.816618919372559,
      "learning_rate": 2.3401029748283753e-05,
      "loss": 1.1536,
      "step": 1538
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.002206802368164,
      "learning_rate": 2.3396739130434782e-05,
      "loss": 1.268,
      "step": 1539
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.662298202514648,
      "learning_rate": 2.3392448512585814e-05,
      "loss": 1.3461,
      "step": 1540
    },
    {
      "epoch": 0.44,
      "grad_norm": 10.412079811096191,
      "learning_rate": 2.338815789473684e-05,
      "loss": 1.4632,
      "step": 1541
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.586483001708984,
      "learning_rate": 2.3383867276887872e-05,
      "loss": 1.5167,
      "step": 1542
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.466395378112793,
      "learning_rate": 2.3379576659038904e-05,
      "loss": 1.1714,
      "step": 1543
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.513823509216309,
      "learning_rate": 2.3375286041189933e-05,
      "loss": 1.2773,
      "step": 1544
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.646080017089844,
      "learning_rate": 2.337099542334096e-05,
      "loss": 1.1845,
      "step": 1545
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.359091758728027,
      "learning_rate": 2.336670480549199e-05,
      "loss": 1.3177,
      "step": 1546
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.305411338806152,
      "learning_rate": 2.3362414187643023e-05,
      "loss": 1.159,
      "step": 1547
    },
    {
      "epoch": 0.44,
      "grad_norm": 10.000143051147461,
      "learning_rate": 2.335812356979405e-05,
      "loss": 1.1044,
      "step": 1548
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.8117094039917,
      "learning_rate": 2.335383295194508e-05,
      "loss": 1.4982,
      "step": 1549
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.323227882385254,
      "learning_rate": 2.334954233409611e-05,
      "loss": 1.4575,
      "step": 1550
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.354408264160156,
      "learning_rate": 2.334525171624714e-05,
      "loss": 1.2588,
      "step": 1551
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.50368881225586,
      "learning_rate": 2.334096109839817e-05,
      "loss": 1.0109,
      "step": 1552
    },
    {
      "epoch": 0.44,
      "grad_norm": 10.090970039367676,
      "learning_rate": 2.33366704805492e-05,
      "loss": 1.4676,
      "step": 1553
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.197575569152832,
      "learning_rate": 2.333237986270023e-05,
      "loss": 1.4296,
      "step": 1554
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.461068153381348,
      "learning_rate": 2.332808924485126e-05,
      "loss": 1.3344,
      "step": 1555
    },
    {
      "epoch": 0.45,
      "grad_norm": 7.606350421905518,
      "learning_rate": 2.3323798627002288e-05,
      "loss": 1.157,
      "step": 1556
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.888628005981445,
      "learning_rate": 2.3319508009153317e-05,
      "loss": 1.1213,
      "step": 1557
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.940152168273926,
      "learning_rate": 2.331521739130435e-05,
      "loss": 1.3451,
      "step": 1558
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.469507217407227,
      "learning_rate": 2.3310926773455378e-05,
      "loss": 1.1574,
      "step": 1559
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.728058815002441,
      "learning_rate": 2.3306636155606407e-05,
      "loss": 1.2855,
      "step": 1560
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.956009864807129,
      "learning_rate": 2.330234553775744e-05,
      "loss": 0.8661,
      "step": 1561
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.254283905029297,
      "learning_rate": 2.3298054919908468e-05,
      "loss": 1.4645,
      "step": 1562
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.068936347961426,
      "learning_rate": 2.32937643020595e-05,
      "loss": 1.2349,
      "step": 1563
    },
    {
      "epoch": 0.45,
      "grad_norm": 10.539466857910156,
      "learning_rate": 2.3289473684210525e-05,
      "loss": 1.2601,
      "step": 1564
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.409708023071289,
      "learning_rate": 2.3285183066361557e-05,
      "loss": 1.0722,
      "step": 1565
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.638346672058105,
      "learning_rate": 2.3280892448512586e-05,
      "loss": 1.2543,
      "step": 1566
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.373037338256836,
      "learning_rate": 2.3276601830663615e-05,
      "loss": 1.4724,
      "step": 1567
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.4254789352417,
      "learning_rate": 2.3272311212814644e-05,
      "loss": 1.3384,
      "step": 1568
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.436665534973145,
      "learning_rate": 2.3268020594965676e-05,
      "loss": 1.5132,
      "step": 1569
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.459665298461914,
      "learning_rate": 2.3263729977116708e-05,
      "loss": 1.1472,
      "step": 1570
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.826032638549805,
      "learning_rate": 2.3259439359267733e-05,
      "loss": 1.2903,
      "step": 1571
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.004020690917969,
      "learning_rate": 2.3255148741418765e-05,
      "loss": 1.4222,
      "step": 1572
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.417764663696289,
      "learning_rate": 2.3250858123569794e-05,
      "loss": 1.1339,
      "step": 1573
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.473125457763672,
      "learning_rate": 2.3246567505720826e-05,
      "loss": 1.3293,
      "step": 1574
    },
    {
      "epoch": 0.45,
      "grad_norm": 7.595772743225098,
      "learning_rate": 2.3242276887871852e-05,
      "loss": 1.0717,
      "step": 1575
    },
    {
      "epoch": 0.45,
      "grad_norm": 7.648061275482178,
      "learning_rate": 2.3237986270022884e-05,
      "loss": 0.9319,
      "step": 1576
    },
    {
      "epoch": 0.45,
      "grad_norm": 10.227629661560059,
      "learning_rate": 2.3233695652173916e-05,
      "loss": 1.2729,
      "step": 1577
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.110548973083496,
      "learning_rate": 2.322940503432494e-05,
      "loss": 1.2656,
      "step": 1578
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.993183135986328,
      "learning_rate": 2.3225114416475974e-05,
      "loss": 1.2294,
      "step": 1579
    },
    {
      "epoch": 0.45,
      "grad_norm": 7.090939044952393,
      "learning_rate": 2.3220823798627002e-05,
      "loss": 1.021,
      "step": 1580
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.64580249786377,
      "learning_rate": 2.3216533180778035e-05,
      "loss": 1.2793,
      "step": 1581
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.000699996948242,
      "learning_rate": 2.321224256292906e-05,
      "loss": 1.3211,
      "step": 1582
    },
    {
      "epoch": 0.45,
      "grad_norm": 7.053379535675049,
      "learning_rate": 2.3207951945080092e-05,
      "loss": 1.1912,
      "step": 1583
    },
    {
      "epoch": 0.45,
      "grad_norm": 10.279417991638184,
      "learning_rate": 2.320366132723112e-05,
      "loss": 1.4488,
      "step": 1584
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.077784538269043,
      "learning_rate": 2.3199370709382153e-05,
      "loss": 1.1453,
      "step": 1585
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.773016929626465,
      "learning_rate": 2.3195080091533182e-05,
      "loss": 1.1198,
      "step": 1586
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.63460636138916,
      "learning_rate": 2.319078947368421e-05,
      "loss": 1.0544,
      "step": 1587
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.71286678314209,
      "learning_rate": 2.3186498855835243e-05,
      "loss": 1.1635,
      "step": 1588
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.21727466583252,
      "learning_rate": 2.318220823798627e-05,
      "loss": 1.38,
      "step": 1589
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.806991577148438,
      "learning_rate": 2.31779176201373e-05,
      "loss": 1.1062,
      "step": 1590
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.155527114868164,
      "learning_rate": 2.317362700228833e-05,
      "loss": 1.2349,
      "step": 1591
    },
    {
      "epoch": 0.46,
      "grad_norm": 7.316340923309326,
      "learning_rate": 2.316933638443936e-05,
      "loss": 0.7638,
      "step": 1592
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.659943580627441,
      "learning_rate": 2.3165045766590387e-05,
      "loss": 1.1853,
      "step": 1593
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.813520431518555,
      "learning_rate": 2.316075514874142e-05,
      "loss": 1.4322,
      "step": 1594
    },
    {
      "epoch": 0.46,
      "grad_norm": 7.801672458648682,
      "learning_rate": 2.315646453089245e-05,
      "loss": 1.1196,
      "step": 1595
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.470723152160645,
      "learning_rate": 2.315217391304348e-05,
      "loss": 1.3101,
      "step": 1596
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.638633728027344,
      "learning_rate": 2.314788329519451e-05,
      "loss": 1.0461,
      "step": 1597
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.282573699951172,
      "learning_rate": 2.3143592677345537e-05,
      "loss": 1.2249,
      "step": 1598
    },
    {
      "epoch": 0.46,
      "grad_norm": 12.078672409057617,
      "learning_rate": 2.313930205949657e-05,
      "loss": 1.575,
      "step": 1599
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.369669914245605,
      "learning_rate": 2.3135011441647598e-05,
      "loss": 1.1217,
      "step": 1600
    },
    {
      "epoch": 0.46,
      "grad_norm": 10.193273544311523,
      "learning_rate": 2.3130720823798627e-05,
      "loss": 1.385,
      "step": 1601
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.902130126953125,
      "learning_rate": 2.3126430205949656e-05,
      "loss": 1.3013,
      "step": 1602
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.053253173828125,
      "learning_rate": 2.3122139588100688e-05,
      "loss": 1.3588,
      "step": 1603
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.413948059082031,
      "learning_rate": 2.3117848970251717e-05,
      "loss": 1.1728,
      "step": 1604
    },
    {
      "epoch": 0.46,
      "grad_norm": 10.024253845214844,
      "learning_rate": 2.3113558352402745e-05,
      "loss": 1.3468,
      "step": 1605
    },
    {
      "epoch": 0.46,
      "grad_norm": 6.907523155212402,
      "learning_rate": 2.3109267734553777e-05,
      "loss": 0.9856,
      "step": 1606
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.771257400512695,
      "learning_rate": 2.3104977116704806e-05,
      "loss": 1.0424,
      "step": 1607
    },
    {
      "epoch": 0.46,
      "grad_norm": 10.07789421081543,
      "learning_rate": 2.3100686498855835e-05,
      "loss": 1.3329,
      "step": 1608
    },
    {
      "epoch": 0.46,
      "grad_norm": 7.559316635131836,
      "learning_rate": 2.3096395881006864e-05,
      "loss": 1.291,
      "step": 1609
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.47673225402832,
      "learning_rate": 2.3092105263157896e-05,
      "loss": 1.2689,
      "step": 1610
    },
    {
      "epoch": 0.46,
      "grad_norm": 7.403980731964111,
      "learning_rate": 2.3087814645308928e-05,
      "loss": 1.046,
      "step": 1611
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.600950241088867,
      "learning_rate": 2.3083524027459953e-05,
      "loss": 1.0703,
      "step": 1612
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.569875717163086,
      "learning_rate": 2.3079233409610986e-05,
      "loss": 1.1255,
      "step": 1613
    },
    {
      "epoch": 0.46,
      "grad_norm": 10.12997055053711,
      "learning_rate": 2.3074942791762014e-05,
      "loss": 1.2578,
      "step": 1614
    },
    {
      "epoch": 0.46,
      "grad_norm": 10.162885665893555,
      "learning_rate": 2.3070652173913047e-05,
      "loss": 1.234,
      "step": 1615
    },
    {
      "epoch": 0.46,
      "grad_norm": 7.93971061706543,
      "learning_rate": 2.3066361556064072e-05,
      "loss": 0.9553,
      "step": 1616
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.405335426330566,
      "learning_rate": 2.3062070938215104e-05,
      "loss": 1.3859,
      "step": 1617
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.421310424804688,
      "learning_rate": 2.3057780320366133e-05,
      "loss": 1.0926,
      "step": 1618
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.382375717163086,
      "learning_rate": 2.305348970251716e-05,
      "loss": 0.9399,
      "step": 1619
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.15948486328125,
      "learning_rate": 2.3049199084668194e-05,
      "loss": 1.4505,
      "step": 1620
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.814046859741211,
      "learning_rate": 2.3044908466819223e-05,
      "loss": 0.9856,
      "step": 1621
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.627074241638184,
      "learning_rate": 2.3040617848970255e-05,
      "loss": 1.349,
      "step": 1622
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.516486167907715,
      "learning_rate": 2.303632723112128e-05,
      "loss": 1.1915,
      "step": 1623
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.72075080871582,
      "learning_rate": 2.3032036613272312e-05,
      "loss": 1.2033,
      "step": 1624
    },
    {
      "epoch": 0.46,
      "grad_norm": 11.18325138092041,
      "learning_rate": 2.302774599542334e-05,
      "loss": 1.2994,
      "step": 1625
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.222212791442871,
      "learning_rate": 2.3023455377574373e-05,
      "loss": 1.2269,
      "step": 1626
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.96588134765625,
      "learning_rate": 2.30191647597254e-05,
      "loss": 1.5958,
      "step": 1627
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.952670097351074,
      "learning_rate": 2.301487414187643e-05,
      "loss": 1.1666,
      "step": 1628
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.964648246765137,
      "learning_rate": 2.3010583524027463e-05,
      "loss": 1.2652,
      "step": 1629
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.549927711486816,
      "learning_rate": 2.3006292906178488e-05,
      "loss": 1.2339,
      "step": 1630
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.86664342880249,
      "learning_rate": 2.300200228832952e-05,
      "loss": 1.2895,
      "step": 1631
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.924127578735352,
      "learning_rate": 2.299771167048055e-05,
      "loss": 1.2733,
      "step": 1632
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.034481048583984,
      "learning_rate": 2.299342105263158e-05,
      "loss": 1.2594,
      "step": 1633
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.503042221069336,
      "learning_rate": 2.2989130434782607e-05,
      "loss": 1.1939,
      "step": 1634
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.935110092163086,
      "learning_rate": 2.298483981693364e-05,
      "loss": 1.0546,
      "step": 1635
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.795965194702148,
      "learning_rate": 2.2980549199084668e-05,
      "loss": 1.2768,
      "step": 1636
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.024333953857422,
      "learning_rate": 2.29762585812357e-05,
      "loss": 1.0955,
      "step": 1637
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.787989616394043,
      "learning_rate": 2.297196796338673e-05,
      "loss": 1.6545,
      "step": 1638
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.520974159240723,
      "learning_rate": 2.2967677345537757e-05,
      "loss": 0.9123,
      "step": 1639
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.345480918884277,
      "learning_rate": 2.296338672768879e-05,
      "loss": 1.2796,
      "step": 1640
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.186126232147217,
      "learning_rate": 2.2959096109839818e-05,
      "loss": 0.8431,
      "step": 1641
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.361184120178223,
      "learning_rate": 2.2954805491990847e-05,
      "loss": 1.4357,
      "step": 1642
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.415900230407715,
      "learning_rate": 2.2950514874141876e-05,
      "loss": 1.1054,
      "step": 1643
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.558908462524414,
      "learning_rate": 2.2946224256292908e-05,
      "loss": 1.2839,
      "step": 1644
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.749784469604492,
      "learning_rate": 2.2941933638443933e-05,
      "loss": 1.2448,
      "step": 1645
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.153518676757812,
      "learning_rate": 2.2937643020594965e-05,
      "loss": 1.1404,
      "step": 1646
    },
    {
      "epoch": 0.47,
      "grad_norm": 10.492255210876465,
      "learning_rate": 2.2933352402745998e-05,
      "loss": 1.421,
      "step": 1647
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.747618675231934,
      "learning_rate": 2.2929061784897026e-05,
      "loss": 1.4087,
      "step": 1648
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.309370994567871,
      "learning_rate": 2.2924771167048055e-05,
      "loss": 1.2397,
      "step": 1649
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.79931926727295,
      "learning_rate": 2.2920480549199084e-05,
      "loss": 1.3496,
      "step": 1650
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.985559463500977,
      "learning_rate": 2.2916189931350116e-05,
      "loss": 1.1182,
      "step": 1651
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.381940841674805,
      "learning_rate": 2.2911899313501145e-05,
      "loss": 1.0647,
      "step": 1652
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.11472225189209,
      "learning_rate": 2.2907608695652174e-05,
      "loss": 1.1354,
      "step": 1653
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.868236064910889,
      "learning_rate": 2.2903318077803206e-05,
      "loss": 0.9615,
      "step": 1654
    },
    {
      "epoch": 0.47,
      "grad_norm": 10.499838829040527,
      "learning_rate": 2.2899027459954235e-05,
      "loss": 1.3577,
      "step": 1655
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.916804313659668,
      "learning_rate": 2.2894736842105263e-05,
      "loss": 1.3944,
      "step": 1656
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.581201553344727,
      "learning_rate": 2.2890446224256292e-05,
      "loss": 1.4577,
      "step": 1657
    },
    {
      "epoch": 0.47,
      "grad_norm": 10.742766380310059,
      "learning_rate": 2.2886155606407324e-05,
      "loss": 1.4189,
      "step": 1658
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.715123176574707,
      "learning_rate": 2.2881864988558353e-05,
      "loss": 1.3955,
      "step": 1659
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.724863052368164,
      "learning_rate": 2.2877574370709382e-05,
      "loss": 1.4059,
      "step": 1660
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.264730453491211,
      "learning_rate": 2.287328375286041e-05,
      "loss": 1.1103,
      "step": 1661
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.868542671203613,
      "learning_rate": 2.2868993135011443e-05,
      "loss": 1.1557,
      "step": 1662
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.071462631225586,
      "learning_rate": 2.2864702517162475e-05,
      "loss": 1.3348,
      "step": 1663
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.672259330749512,
      "learning_rate": 2.28604118993135e-05,
      "loss": 1.2869,
      "step": 1664
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.83873176574707,
      "learning_rate": 2.2856121281464532e-05,
      "loss": 1.3988,
      "step": 1665
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.405563354492188,
      "learning_rate": 2.285183066361556e-05,
      "loss": 1.3591,
      "step": 1666
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.799030303955078,
      "learning_rate": 2.2847540045766593e-05,
      "loss": 1.2546,
      "step": 1667
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.277997970581055,
      "learning_rate": 2.284324942791762e-05,
      "loss": 0.9324,
      "step": 1668
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.408809661865234,
      "learning_rate": 2.283895881006865e-05,
      "loss": 1.2105,
      "step": 1669
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.046985626220703,
      "learning_rate": 2.283466819221968e-05,
      "loss": 1.4227,
      "step": 1670
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.309938430786133,
      "learning_rate": 2.283037757437071e-05,
      "loss": 1.0991,
      "step": 1671
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.683304786682129,
      "learning_rate": 2.282608695652174e-05,
      "loss": 1.2208,
      "step": 1672
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.217636108398438,
      "learning_rate": 2.282179633867277e-05,
      "loss": 1.1511,
      "step": 1673
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.621679306030273,
      "learning_rate": 2.28175057208238e-05,
      "loss": 1.2189,
      "step": 1674
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.081939697265625,
      "learning_rate": 2.2813215102974827e-05,
      "loss": 0.9736,
      "step": 1675
    },
    {
      "epoch": 0.48,
      "grad_norm": 10.959156036376953,
      "learning_rate": 2.280892448512586e-05,
      "loss": 1.1892,
      "step": 1676
    },
    {
      "epoch": 0.48,
      "grad_norm": 10.01351261138916,
      "learning_rate": 2.2804633867276888e-05,
      "loss": 1.378,
      "step": 1677
    },
    {
      "epoch": 0.48,
      "grad_norm": 7.741083145141602,
      "learning_rate": 2.280034324942792e-05,
      "loss": 1.1976,
      "step": 1678
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.066365242004395,
      "learning_rate": 2.2796052631578945e-05,
      "loss": 1.0487,
      "step": 1679
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.079463958740234,
      "learning_rate": 2.2791762013729977e-05,
      "loss": 1.03,
      "step": 1680
    },
    {
      "epoch": 0.48,
      "grad_norm": 11.072271347045898,
      "learning_rate": 2.278747139588101e-05,
      "loss": 1.2092,
      "step": 1681
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.641533851623535,
      "learning_rate": 2.278318077803204e-05,
      "loss": 1.4444,
      "step": 1682
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.743915557861328,
      "learning_rate": 2.2778890160183067e-05,
      "loss": 0.9681,
      "step": 1683
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.499055862426758,
      "learning_rate": 2.2774599542334096e-05,
      "loss": 1.3324,
      "step": 1684
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.612502098083496,
      "learning_rate": 2.2770308924485128e-05,
      "loss": 1.2739,
      "step": 1685
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.890045166015625,
      "learning_rate": 2.2766018306636153e-05,
      "loss": 1.3744,
      "step": 1686
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.541918754577637,
      "learning_rate": 2.2761727688787186e-05,
      "loss": 1.3663,
      "step": 1687
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.16054630279541,
      "learning_rate": 2.2757437070938214e-05,
      "loss": 1.1082,
      "step": 1688
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.091609954833984,
      "learning_rate": 2.2753146453089247e-05,
      "loss": 1.1344,
      "step": 1689
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.458251953125,
      "learning_rate": 2.2748855835240275e-05,
      "loss": 1.228,
      "step": 1690
    },
    {
      "epoch": 0.48,
      "grad_norm": 10.517223358154297,
      "learning_rate": 2.2744565217391304e-05,
      "loss": 1.081,
      "step": 1691
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.296842575073242,
      "learning_rate": 2.2740274599542336e-05,
      "loss": 1.0767,
      "step": 1692
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.35966682434082,
      "learning_rate": 2.2735983981693365e-05,
      "loss": 1.2591,
      "step": 1693
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.021504402160645,
      "learning_rate": 2.2731693363844394e-05,
      "loss": 1.0929,
      "step": 1694
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.004847526550293,
      "learning_rate": 2.2727402745995423e-05,
      "loss": 1.0958,
      "step": 1695
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.843135833740234,
      "learning_rate": 2.2723112128146455e-05,
      "loss": 1.2846,
      "step": 1696
    },
    {
      "epoch": 0.49,
      "grad_norm": 7.5560221672058105,
      "learning_rate": 2.2718821510297483e-05,
      "loss": 0.871,
      "step": 1697
    },
    {
      "epoch": 0.49,
      "grad_norm": 10.070405960083008,
      "learning_rate": 2.2714530892448512e-05,
      "loss": 1.2432,
      "step": 1698
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.583444595336914,
      "learning_rate": 2.2710240274599544e-05,
      "loss": 1.4108,
      "step": 1699
    },
    {
      "epoch": 0.49,
      "grad_norm": 10.146078109741211,
      "learning_rate": 2.2705949656750573e-05,
      "loss": 1.1774,
      "step": 1700
    },
    {
      "epoch": 0.49,
      "grad_norm": 10.606247901916504,
      "learning_rate": 2.2701659038901602e-05,
      "loss": 1.2373,
      "step": 1701
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.760239601135254,
      "learning_rate": 2.269736842105263e-05,
      "loss": 1.2494,
      "step": 1702
    },
    {
      "epoch": 0.49,
      "grad_norm": 10.328970909118652,
      "learning_rate": 2.2693077803203663e-05,
      "loss": 1.3952,
      "step": 1703
    },
    {
      "epoch": 0.49,
      "grad_norm": 15.020970344543457,
      "learning_rate": 2.268878718535469e-05,
      "loss": 1.6676,
      "step": 1704
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.3358793258667,
      "learning_rate": 2.268449656750572e-05,
      "loss": 0.8649,
      "step": 1705
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.585587501525879,
      "learning_rate": 2.2680205949656753e-05,
      "loss": 1.1444,
      "step": 1706
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.351394653320312,
      "learning_rate": 2.267591533180778e-05,
      "loss": 1.1239,
      "step": 1707
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.965201377868652,
      "learning_rate": 2.2671624713958813e-05,
      "loss": 1.2742,
      "step": 1708
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.27197265625,
      "learning_rate": 2.266733409610984e-05,
      "loss": 1.1194,
      "step": 1709
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.04944133758545,
      "learning_rate": 2.266304347826087e-05,
      "loss": 0.8732,
      "step": 1710
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.792150497436523,
      "learning_rate": 2.26587528604119e-05,
      "loss": 1.1253,
      "step": 1711
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.72215747833252,
      "learning_rate": 2.265446224256293e-05,
      "loss": 1.014,
      "step": 1712
    },
    {
      "epoch": 0.49,
      "grad_norm": 11.412017822265625,
      "learning_rate": 2.2650171624713957e-05,
      "loss": 1.4195,
      "step": 1713
    },
    {
      "epoch": 0.49,
      "grad_norm": 6.746893882751465,
      "learning_rate": 2.264588100686499e-05,
      "loss": 0.8599,
      "step": 1714
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.354917526245117,
      "learning_rate": 2.264159038901602e-05,
      "loss": 1.2557,
      "step": 1715
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.321457862854004,
      "learning_rate": 2.2637299771167047e-05,
      "loss": 1.4941,
      "step": 1716
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.840455055236816,
      "learning_rate": 2.263300915331808e-05,
      "loss": 1.1555,
      "step": 1717
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.907840728759766,
      "learning_rate": 2.2628718535469108e-05,
      "loss": 1.311,
      "step": 1718
    },
    {
      "epoch": 0.49,
      "grad_norm": 7.54386568069458,
      "learning_rate": 2.262442791762014e-05,
      "loss": 0.913,
      "step": 1719
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.386448860168457,
      "learning_rate": 2.2620137299771165e-05,
      "loss": 0.9293,
      "step": 1720
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.661848068237305,
      "learning_rate": 2.2615846681922198e-05,
      "loss": 1.1228,
      "step": 1721
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.990273475646973,
      "learning_rate": 2.2611556064073226e-05,
      "loss": 1.1689,
      "step": 1722
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.665467262268066,
      "learning_rate": 2.2607265446224255e-05,
      "loss": 1.448,
      "step": 1723
    },
    {
      "epoch": 0.49,
      "grad_norm": 11.016701698303223,
      "learning_rate": 2.2602974828375287e-05,
      "loss": 1.0422,
      "step": 1724
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.635202407836914,
      "learning_rate": 2.2598684210526316e-05,
      "loss": 1.0084,
      "step": 1725
    },
    {
      "epoch": 0.49,
      "grad_norm": 7.417308807373047,
      "learning_rate": 2.2594393592677348e-05,
      "loss": 1.0566,
      "step": 1726
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.810213088989258,
      "learning_rate": 2.2590102974828374e-05,
      "loss": 1.0975,
      "step": 1727
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.824432373046875,
      "learning_rate": 2.2585812356979406e-05,
      "loss": 1.2073,
      "step": 1728
    },
    {
      "epoch": 0.49,
      "grad_norm": 11.824142456054688,
      "learning_rate": 2.2581521739130434e-05,
      "loss": 1.5631,
      "step": 1729
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.870973587036133,
      "learning_rate": 2.2577231121281467e-05,
      "loss": 1.0913,
      "step": 1730
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.540595054626465,
      "learning_rate": 2.2572940503432495e-05,
      "loss": 1.1692,
      "step": 1731
    },
    {
      "epoch": 0.5,
      "grad_norm": 11.168774604797363,
      "learning_rate": 2.2568649885583524e-05,
      "loss": 1.4247,
      "step": 1732
    },
    {
      "epoch": 0.5,
      "grad_norm": 10.752882957458496,
      "learning_rate": 2.2564359267734556e-05,
      "loss": 1.3847,
      "step": 1733
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.558868408203125,
      "learning_rate": 2.2560068649885585e-05,
      "loss": 1.2971,
      "step": 1734
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.694822311401367,
      "learning_rate": 2.2555778032036614e-05,
      "loss": 1.2184,
      "step": 1735
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.882486343383789,
      "learning_rate": 2.2551487414187643e-05,
      "loss": 0.9044,
      "step": 1736
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.602743148803711,
      "learning_rate": 2.2547196796338675e-05,
      "loss": 1.1263,
      "step": 1737
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.551252365112305,
      "learning_rate": 2.25429061784897e-05,
      "loss": 1.3499,
      "step": 1738
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.753378868103027,
      "learning_rate": 2.2538615560640732e-05,
      "loss": 1.0904,
      "step": 1739
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.857767105102539,
      "learning_rate": 2.2534324942791764e-05,
      "loss": 1.3323,
      "step": 1740
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.807705879211426,
      "learning_rate": 2.2530034324942793e-05,
      "loss": 1.0568,
      "step": 1741
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.36620044708252,
      "learning_rate": 2.2525743707093822e-05,
      "loss": 1.0966,
      "step": 1742
    },
    {
      "epoch": 0.5,
      "grad_norm": 10.410367012023926,
      "learning_rate": 2.252145308924485e-05,
      "loss": 1.236,
      "step": 1743
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.877896308898926,
      "learning_rate": 2.2517162471395883e-05,
      "loss": 0.9688,
      "step": 1744
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.445262908935547,
      "learning_rate": 2.2512871853546912e-05,
      "loss": 1.2386,
      "step": 1745
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.056120872497559,
      "learning_rate": 2.250858123569794e-05,
      "loss": 1.109,
      "step": 1746
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.643233299255371,
      "learning_rate": 2.250429061784897e-05,
      "loss": 1.3352,
      "step": 1747
    },
    {
      "epoch": 0.5,
      "grad_norm": 10.496377944946289,
      "learning_rate": 2.25e-05,
      "loss": 1.5072,
      "step": 1748
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.46739673614502,
      "learning_rate": 2.249570938215103e-05,
      "loss": 1.3112,
      "step": 1749
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.37043571472168,
      "learning_rate": 2.249141876430206e-05,
      "loss": 1.2561,
      "step": 1750
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.924062728881836,
      "learning_rate": 2.248712814645309e-05,
      "loss": 1.3819,
      "step": 1751
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.79951286315918,
      "learning_rate": 2.248283752860412e-05,
      "loss": 1.2067,
      "step": 1752
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.621842384338379,
      "learning_rate": 2.247854691075515e-05,
      "loss": 1.3184,
      "step": 1753
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.6490478515625,
      "learning_rate": 2.2474256292906177e-05,
      "loss": 1.1118,
      "step": 1754
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.803616523742676,
      "learning_rate": 2.246996567505721e-05,
      "loss": 1.2251,
      "step": 1755
    },
    {
      "epoch": 0.5,
      "grad_norm": 7.9771599769592285,
      "learning_rate": 2.246567505720824e-05,
      "loss": 1.0104,
      "step": 1756
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.340866088867188,
      "learning_rate": 2.2461384439359267e-05,
      "loss": 0.8333,
      "step": 1757
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.675475120544434,
      "learning_rate": 2.24570938215103e-05,
      "loss": 1.1501,
      "step": 1758
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.976713180541992,
      "learning_rate": 2.2452803203661328e-05,
      "loss": 1.4332,
      "step": 1759
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.324586868286133,
      "learning_rate": 2.244851258581236e-05,
      "loss": 1.2735,
      "step": 1760
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.389039039611816,
      "learning_rate": 2.2444221967963386e-05,
      "loss": 1.3965,
      "step": 1761
    },
    {
      "epoch": 0.5,
      "grad_norm": 7.7743425369262695,
      "learning_rate": 2.2439931350114418e-05,
      "loss": 1.0439,
      "step": 1762
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.806344032287598,
      "learning_rate": 2.2435640732265446e-05,
      "loss": 1.2115,
      "step": 1763
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.250657081604004,
      "learning_rate": 2.2431350114416475e-05,
      "loss": 1.3243,
      "step": 1764
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.528936386108398,
      "learning_rate": 2.2427059496567504e-05,
      "loss": 1.2346,
      "step": 1765
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.796095848083496,
      "learning_rate": 2.2422768878718536e-05,
      "loss": 1.3457,
      "step": 1766
    },
    {
      "epoch": 0.51,
      "grad_norm": 6.960087776184082,
      "learning_rate": 2.241847826086957e-05,
      "loss": 1.1521,
      "step": 1767
    },
    {
      "epoch": 0.51,
      "grad_norm": 6.853693962097168,
      "learning_rate": 2.2414187643020594e-05,
      "loss": 1.1208,
      "step": 1768
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.802952766418457,
      "learning_rate": 2.2409897025171626e-05,
      "loss": 1.188,
      "step": 1769
    },
    {
      "epoch": 0.51,
      "grad_norm": 7.594489574432373,
      "learning_rate": 2.2405606407322655e-05,
      "loss": 1.5319,
      "step": 1770
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.9003267288208,
      "learning_rate": 2.2401315789473687e-05,
      "loss": 1.1745,
      "step": 1771
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.605582237243652,
      "learning_rate": 2.2397025171624712e-05,
      "loss": 1.03,
      "step": 1772
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.687684059143066,
      "learning_rate": 2.2392734553775744e-05,
      "loss": 1.2067,
      "step": 1773
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.256669044494629,
      "learning_rate": 2.2388443935926776e-05,
      "loss": 0.9294,
      "step": 1774
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.63094711303711,
      "learning_rate": 2.2384153318077802e-05,
      "loss": 1.2256,
      "step": 1775
    },
    {
      "epoch": 0.51,
      "grad_norm": 7.976043224334717,
      "learning_rate": 2.2379862700228834e-05,
      "loss": 1.191,
      "step": 1776
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.502062797546387,
      "learning_rate": 2.2375572082379863e-05,
      "loss": 1.106,
      "step": 1777
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.629040718078613,
      "learning_rate": 2.2371281464530895e-05,
      "loss": 1.1471,
      "step": 1778
    },
    {
      "epoch": 0.51,
      "grad_norm": 10.833641052246094,
      "learning_rate": 2.236699084668192e-05,
      "loss": 1.3265,
      "step": 1779
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.588047981262207,
      "learning_rate": 2.2362700228832952e-05,
      "loss": 1.1669,
      "step": 1780
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.60359001159668,
      "learning_rate": 2.235840961098398e-05,
      "loss": 1.1454,
      "step": 1781
    },
    {
      "epoch": 0.51,
      "grad_norm": 10.990739822387695,
      "learning_rate": 2.2354118993135013e-05,
      "loss": 1.4634,
      "step": 1782
    },
    {
      "epoch": 0.51,
      "grad_norm": 10.82275390625,
      "learning_rate": 2.2349828375286042e-05,
      "loss": 1.3927,
      "step": 1783
    },
    {
      "epoch": 0.51,
      "grad_norm": 10.499259948730469,
      "learning_rate": 2.234553775743707e-05,
      "loss": 1.3967,
      "step": 1784
    },
    {
      "epoch": 0.51,
      "grad_norm": 10.63853645324707,
      "learning_rate": 2.2341247139588103e-05,
      "loss": 1.1029,
      "step": 1785
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.913352012634277,
      "learning_rate": 2.2336956521739132e-05,
      "loss": 1.507,
      "step": 1786
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.286478042602539,
      "learning_rate": 2.233266590389016e-05,
      "loss": 1.3428,
      "step": 1787
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.3721342086792,
      "learning_rate": 2.232837528604119e-05,
      "loss": 1.061,
      "step": 1788
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.989459037780762,
      "learning_rate": 2.232408466819222e-05,
      "loss": 1.2538,
      "step": 1789
    },
    {
      "epoch": 0.51,
      "grad_norm": 10.688898086547852,
      "learning_rate": 2.2319794050343247e-05,
      "loss": 1.0603,
      "step": 1790
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.396010398864746,
      "learning_rate": 2.231550343249428e-05,
      "loss": 1.0253,
      "step": 1791
    },
    {
      "epoch": 0.51,
      "grad_norm": 7.522763252258301,
      "learning_rate": 2.231121281464531e-05,
      "loss": 0.966,
      "step": 1792
    },
    {
      "epoch": 0.51,
      "grad_norm": 10.1019287109375,
      "learning_rate": 2.230692219679634e-05,
      "loss": 1.3999,
      "step": 1793
    },
    {
      "epoch": 0.51,
      "grad_norm": 7.7475266456604,
      "learning_rate": 2.230263157894737e-05,
      "loss": 1.3943,
      "step": 1794
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.222977638244629,
      "learning_rate": 2.2298340961098398e-05,
      "loss": 1.4151,
      "step": 1795
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.31693172454834,
      "learning_rate": 2.229405034324943e-05,
      "loss": 1.0597,
      "step": 1796
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.179215431213379,
      "learning_rate": 2.228975972540046e-05,
      "loss": 1.2418,
      "step": 1797
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.257237434387207,
      "learning_rate": 2.2285469107551487e-05,
      "loss": 0.9888,
      "step": 1798
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.153144836425781,
      "learning_rate": 2.2281178489702516e-05,
      "loss": 1.4073,
      "step": 1799
    },
    {
      "epoch": 0.51,
      "grad_norm": 7.4600982666015625,
      "learning_rate": 2.2276887871853548e-05,
      "loss": 0.9782,
      "step": 1800
    },
    {
      "epoch": 0.52,
      "grad_norm": 7.916541576385498,
      "learning_rate": 2.2272597254004577e-05,
      "loss": 1.159,
      "step": 1801
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.717126846313477,
      "learning_rate": 2.2268306636155606e-05,
      "loss": 1.1931,
      "step": 1802
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.714426040649414,
      "learning_rate": 2.2264016018306638e-05,
      "loss": 1.3754,
      "step": 1803
    },
    {
      "epoch": 0.52,
      "grad_norm": 10.216131210327148,
      "learning_rate": 2.2259725400457667e-05,
      "loss": 1.1484,
      "step": 1804
    },
    {
      "epoch": 0.52,
      "grad_norm": 10.28882122039795,
      "learning_rate": 2.2255434782608695e-05,
      "loss": 1.3258,
      "step": 1805
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.439464569091797,
      "learning_rate": 2.2251144164759724e-05,
      "loss": 1.3977,
      "step": 1806
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.396167755126953,
      "learning_rate": 2.2246853546910756e-05,
      "loss": 1.4009,
      "step": 1807
    },
    {
      "epoch": 0.52,
      "grad_norm": 7.737992763519287,
      "learning_rate": 2.224256292906179e-05,
      "loss": 1.2623,
      "step": 1808
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.101402282714844,
      "learning_rate": 2.2238272311212814e-05,
      "loss": 1.458,
      "step": 1809
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.564967155456543,
      "learning_rate": 2.2233981693363846e-05,
      "loss": 1.1027,
      "step": 1810
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.401267051696777,
      "learning_rate": 2.2229691075514875e-05,
      "loss": 0.9862,
      "step": 1811
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.633465766906738,
      "learning_rate": 2.2225400457665907e-05,
      "loss": 1.23,
      "step": 1812
    },
    {
      "epoch": 0.52,
      "grad_norm": 10.191815376281738,
      "learning_rate": 2.2221109839816932e-05,
      "loss": 1.1127,
      "step": 1813
    },
    {
      "epoch": 0.52,
      "grad_norm": 7.706086158752441,
      "learning_rate": 2.2216819221967964e-05,
      "loss": 0.9632,
      "step": 1814
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.303768157958984,
      "learning_rate": 2.2212528604118993e-05,
      "loss": 0.9504,
      "step": 1815
    },
    {
      "epoch": 0.52,
      "grad_norm": 10.180074691772461,
      "learning_rate": 2.2208237986270022e-05,
      "loss": 1.0633,
      "step": 1816
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.964677810668945,
      "learning_rate": 2.2203947368421054e-05,
      "loss": 1.4001,
      "step": 1817
    },
    {
      "epoch": 0.52,
      "grad_norm": 7.654510021209717,
      "learning_rate": 2.2199656750572083e-05,
      "loss": 1.0369,
      "step": 1818
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.303197860717773,
      "learning_rate": 2.2195366132723115e-05,
      "loss": 1.2726,
      "step": 1819
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.660150527954102,
      "learning_rate": 2.219107551487414e-05,
      "loss": 0.9767,
      "step": 1820
    },
    {
      "epoch": 0.52,
      "grad_norm": 10.521946907043457,
      "learning_rate": 2.2186784897025173e-05,
      "loss": 1.317,
      "step": 1821
    },
    {
      "epoch": 0.52,
      "grad_norm": 10.280237197875977,
      "learning_rate": 2.21824942791762e-05,
      "loss": 1.302,
      "step": 1822
    },
    {
      "epoch": 0.52,
      "grad_norm": 10.660154342651367,
      "learning_rate": 2.2178203661327234e-05,
      "loss": 1.4013,
      "step": 1823
    },
    {
      "epoch": 0.52,
      "grad_norm": 10.585969924926758,
      "learning_rate": 2.217391304347826e-05,
      "loss": 1.3267,
      "step": 1824
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.46932315826416,
      "learning_rate": 2.216962242562929e-05,
      "loss": 1.2359,
      "step": 1825
    },
    {
      "epoch": 0.52,
      "grad_norm": 11.864045143127441,
      "learning_rate": 2.2165331807780323e-05,
      "loss": 1.5453,
      "step": 1826
    },
    {
      "epoch": 0.52,
      "grad_norm": 10.174424171447754,
      "learning_rate": 2.216104118993135e-05,
      "loss": 1.1714,
      "step": 1827
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.30581283569336,
      "learning_rate": 2.215675057208238e-05,
      "loss": 1.2256,
      "step": 1828
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.124938011169434,
      "learning_rate": 2.215245995423341e-05,
      "loss": 1.3058,
      "step": 1829
    },
    {
      "epoch": 0.52,
      "grad_norm": 7.997518539428711,
      "learning_rate": 2.2148169336384442e-05,
      "loss": 1.1819,
      "step": 1830
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.968917846679688,
      "learning_rate": 2.2143878718535467e-05,
      "loss": 1.4171,
      "step": 1831
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.303597450256348,
      "learning_rate": 2.21395881006865e-05,
      "loss": 1.0651,
      "step": 1832
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.020977020263672,
      "learning_rate": 2.2135297482837528e-05,
      "loss": 0.999,
      "step": 1833
    },
    {
      "epoch": 0.52,
      "grad_norm": 10.523795127868652,
      "learning_rate": 2.213100686498856e-05,
      "loss": 1.1654,
      "step": 1834
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.899744987487793,
      "learning_rate": 2.212671624713959e-05,
      "loss": 1.2683,
      "step": 1835
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.98214054107666,
      "learning_rate": 2.2122425629290618e-05,
      "loss": 0.9926,
      "step": 1836
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.100053787231445,
      "learning_rate": 2.211813501144165e-05,
      "loss": 1.0374,
      "step": 1837
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.72776985168457,
      "learning_rate": 2.211384439359268e-05,
      "loss": 1.5318,
      "step": 1838
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.794493675231934,
      "learning_rate": 2.2109553775743707e-05,
      "loss": 1.1,
      "step": 1839
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.37671184539795,
      "learning_rate": 2.2105263157894736e-05,
      "loss": 1.0562,
      "step": 1840
    },
    {
      "epoch": 0.53,
      "grad_norm": 7.808015823364258,
      "learning_rate": 2.210097254004577e-05,
      "loss": 0.8847,
      "step": 1841
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.175644874572754,
      "learning_rate": 2.2096681922196794e-05,
      "loss": 1.527,
      "step": 1842
    },
    {
      "epoch": 0.53,
      "grad_norm": 7.95731782913208,
      "learning_rate": 2.2092391304347826e-05,
      "loss": 1.2384,
      "step": 1843
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.053457260131836,
      "learning_rate": 2.2088100686498858e-05,
      "loss": 1.4655,
      "step": 1844
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.427711486816406,
      "learning_rate": 2.2083810068649887e-05,
      "loss": 1.2248,
      "step": 1845
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.888148307800293,
      "learning_rate": 2.2079519450800916e-05,
      "loss": 1.3585,
      "step": 1846
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.449552536010742,
      "learning_rate": 2.2075228832951944e-05,
      "loss": 1.2309,
      "step": 1847
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.936347007751465,
      "learning_rate": 2.2070938215102976e-05,
      "loss": 1.3455,
      "step": 1848
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.431443214416504,
      "learning_rate": 2.2066647597254005e-05,
      "loss": 1.172,
      "step": 1849
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.189772605895996,
      "learning_rate": 2.2062356979405034e-05,
      "loss": 1.1464,
      "step": 1850
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.310342788696289,
      "learning_rate": 2.2058066361556066e-05,
      "loss": 1.2688,
      "step": 1851
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.572994232177734,
      "learning_rate": 2.2053775743707095e-05,
      "loss": 1.2116,
      "step": 1852
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.292588233947754,
      "learning_rate": 2.2049485125858124e-05,
      "loss": 1.2581,
      "step": 1853
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.470903396606445,
      "learning_rate": 2.2045194508009152e-05,
      "loss": 1.0894,
      "step": 1854
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.987678527832031,
      "learning_rate": 2.2040903890160185e-05,
      "loss": 1.5021,
      "step": 1855
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.505946159362793,
      "learning_rate": 2.2036613272311213e-05,
      "loss": 1.1892,
      "step": 1856
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.59296989440918,
      "learning_rate": 2.2032322654462242e-05,
      "loss": 1.4473,
      "step": 1857
    },
    {
      "epoch": 0.53,
      "grad_norm": 7.823615074157715,
      "learning_rate": 2.202803203661327e-05,
      "loss": 0.9165,
      "step": 1858
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.013300895690918,
      "learning_rate": 2.2023741418764303e-05,
      "loss": 1.3646,
      "step": 1859
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.896345138549805,
      "learning_rate": 2.2019450800915335e-05,
      "loss": 1.0996,
      "step": 1860
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.645607948303223,
      "learning_rate": 2.201516018306636e-05,
      "loss": 0.8644,
      "step": 1861
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.687626838684082,
      "learning_rate": 2.2010869565217393e-05,
      "loss": 0.887,
      "step": 1862
    },
    {
      "epoch": 0.53,
      "grad_norm": 10.690119743347168,
      "learning_rate": 2.200657894736842e-05,
      "loss": 1.3183,
      "step": 1863
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.361321449279785,
      "learning_rate": 2.2002288329519454e-05,
      "loss": 1.3161,
      "step": 1864
    },
    {
      "epoch": 0.53,
      "grad_norm": 10.281274795532227,
      "learning_rate": 2.199799771167048e-05,
      "loss": 1.4265,
      "step": 1865
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.622797012329102,
      "learning_rate": 2.199370709382151e-05,
      "loss": 1.3256,
      "step": 1866
    },
    {
      "epoch": 0.53,
      "grad_norm": 11.24836254119873,
      "learning_rate": 2.198941647597254e-05,
      "loss": 1.2944,
      "step": 1867
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.081978797912598,
      "learning_rate": 2.198512585812357e-05,
      "loss": 1.021,
      "step": 1868
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.716439247131348,
      "learning_rate": 2.19808352402746e-05,
      "loss": 1.2928,
      "step": 1869
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.750237464904785,
      "learning_rate": 2.197654462242563e-05,
      "loss": 1.3407,
      "step": 1870
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.843461036682129,
      "learning_rate": 2.1972254004576662e-05,
      "loss": 1.195,
      "step": 1871
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.61376953125,
      "learning_rate": 2.1967963386727687e-05,
      "loss": 1.1895,
      "step": 1872
    },
    {
      "epoch": 0.54,
      "grad_norm": 10.649840354919434,
      "learning_rate": 2.196367276887872e-05,
      "loss": 1.6845,
      "step": 1873
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.97451400756836,
      "learning_rate": 2.1959382151029748e-05,
      "loss": 1.136,
      "step": 1874
    },
    {
      "epoch": 0.54,
      "grad_norm": 10.086967468261719,
      "learning_rate": 2.195509153318078e-05,
      "loss": 1.3111,
      "step": 1875
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.95392894744873,
      "learning_rate": 2.1950800915331806e-05,
      "loss": 1.0321,
      "step": 1876
    },
    {
      "epoch": 0.54,
      "grad_norm": 7.904888153076172,
      "learning_rate": 2.1946510297482838e-05,
      "loss": 1.1564,
      "step": 1877
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.450617790222168,
      "learning_rate": 2.194221967963387e-05,
      "loss": 1.2575,
      "step": 1878
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.170722961425781,
      "learning_rate": 2.19379290617849e-05,
      "loss": 1.0407,
      "step": 1879
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.442112922668457,
      "learning_rate": 2.1933638443935928e-05,
      "loss": 1.2719,
      "step": 1880
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.413675308227539,
      "learning_rate": 2.1929347826086956e-05,
      "loss": 1.1307,
      "step": 1881
    },
    {
      "epoch": 0.54,
      "grad_norm": 11.167388916015625,
      "learning_rate": 2.192505720823799e-05,
      "loss": 1.8384,
      "step": 1882
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.069658279418945,
      "learning_rate": 2.1920766590389014e-05,
      "loss": 1.1637,
      "step": 1883
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.350651741027832,
      "learning_rate": 2.1916475972540046e-05,
      "loss": 1.2792,
      "step": 1884
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.218442916870117,
      "learning_rate": 2.1912185354691075e-05,
      "loss": 1.0882,
      "step": 1885
    },
    {
      "epoch": 0.54,
      "grad_norm": 10.116166114807129,
      "learning_rate": 2.1907894736842107e-05,
      "loss": 1.3364,
      "step": 1886
    },
    {
      "epoch": 0.54,
      "grad_norm": 7.011248588562012,
      "learning_rate": 2.1903604118993136e-05,
      "loss": 0.9083,
      "step": 1887
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.695975303649902,
      "learning_rate": 2.1899313501144164e-05,
      "loss": 1.2792,
      "step": 1888
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.591291427612305,
      "learning_rate": 2.1895022883295197e-05,
      "loss": 1.2621,
      "step": 1889
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.328107833862305,
      "learning_rate": 2.1890732265446225e-05,
      "loss": 0.9975,
      "step": 1890
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.86084270477295,
      "learning_rate": 2.1886441647597254e-05,
      "loss": 1.6399,
      "step": 1891
    },
    {
      "epoch": 0.54,
      "grad_norm": 7.860621929168701,
      "learning_rate": 2.1882151029748283e-05,
      "loss": 1.0072,
      "step": 1892
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.932562828063965,
      "learning_rate": 2.1877860411899315e-05,
      "loss": 1.2332,
      "step": 1893
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.443490982055664,
      "learning_rate": 2.1873569794050344e-05,
      "loss": 1.2122,
      "step": 1894
    },
    {
      "epoch": 0.54,
      "grad_norm": 7.739225387573242,
      "learning_rate": 2.1869279176201373e-05,
      "loss": 1.0659,
      "step": 1895
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.519488334655762,
      "learning_rate": 2.1864988558352405e-05,
      "loss": 1.5054,
      "step": 1896
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.8134765625,
      "learning_rate": 2.1860697940503434e-05,
      "loss": 1.3584,
      "step": 1897
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.263157844543457,
      "learning_rate": 2.1856407322654462e-05,
      "loss": 0.9603,
      "step": 1898
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.455352783203125,
      "learning_rate": 2.185211670480549e-05,
      "loss": 1.1942,
      "step": 1899
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.731051445007324,
      "learning_rate": 2.1847826086956523e-05,
      "loss": 1.1219,
      "step": 1900
    },
    {
      "epoch": 0.54,
      "grad_norm": 11.34709358215332,
      "learning_rate": 2.1843535469107552e-05,
      "loss": 1.1467,
      "step": 1901
    },
    {
      "epoch": 0.54,
      "grad_norm": 7.7822394371032715,
      "learning_rate": 2.183924485125858e-05,
      "loss": 1.0081,
      "step": 1902
    },
    {
      "epoch": 0.54,
      "grad_norm": 10.143342971801758,
      "learning_rate": 2.1834954233409613e-05,
      "loss": 1.25,
      "step": 1903
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.767507553100586,
      "learning_rate": 2.183066361556064e-05,
      "loss": 0.8448,
      "step": 1904
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.370316505432129,
      "learning_rate": 2.1826372997711674e-05,
      "loss": 1.0835,
      "step": 1905
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.88417911529541,
      "learning_rate": 2.18220823798627e-05,
      "loss": 1.1458,
      "step": 1906
    },
    {
      "epoch": 0.55,
      "grad_norm": 10.507073402404785,
      "learning_rate": 2.181779176201373e-05,
      "loss": 1.2188,
      "step": 1907
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.445672035217285,
      "learning_rate": 2.181350114416476e-05,
      "loss": 1.1862,
      "step": 1908
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.55842399597168,
      "learning_rate": 2.180921052631579e-05,
      "loss": 1.4181,
      "step": 1909
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.664935111999512,
      "learning_rate": 2.1804919908466818e-05,
      "loss": 1.1623,
      "step": 1910
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.136595726013184,
      "learning_rate": 2.180062929061785e-05,
      "loss": 1.3494,
      "step": 1911
    },
    {
      "epoch": 0.55,
      "grad_norm": 7.968899250030518,
      "learning_rate": 2.1796338672768882e-05,
      "loss": 0.9201,
      "step": 1912
    },
    {
      "epoch": 0.55,
      "grad_norm": 7.745423793792725,
      "learning_rate": 2.1792048054919907e-05,
      "loss": 1.2271,
      "step": 1913
    },
    {
      "epoch": 0.55,
      "grad_norm": 7.834641933441162,
      "learning_rate": 2.178775743707094e-05,
      "loss": 0.9795,
      "step": 1914
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.000736236572266,
      "learning_rate": 2.1783466819221968e-05,
      "loss": 1.1041,
      "step": 1915
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.695572853088379,
      "learning_rate": 2.1779176201373e-05,
      "loss": 1.176,
      "step": 1916
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.908391952514648,
      "learning_rate": 2.1774885583524026e-05,
      "loss": 1.1249,
      "step": 1917
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.691105842590332,
      "learning_rate": 2.1770594965675058e-05,
      "loss": 1.1521,
      "step": 1918
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.583081245422363,
      "learning_rate": 2.1766304347826087e-05,
      "loss": 1.1156,
      "step": 1919
    },
    {
      "epoch": 0.55,
      "grad_norm": 12.590463638305664,
      "learning_rate": 2.1762013729977116e-05,
      "loss": 1.264,
      "step": 1920
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.349695205688477,
      "learning_rate": 2.1757723112128148e-05,
      "loss": 1.1436,
      "step": 1921
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.978100776672363,
      "learning_rate": 2.1753432494279176e-05,
      "loss": 0.9863,
      "step": 1922
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.360097885131836,
      "learning_rate": 2.174914187643021e-05,
      "loss": 1.1557,
      "step": 1923
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.199853897094727,
      "learning_rate": 2.1744851258581234e-05,
      "loss": 1.3216,
      "step": 1924
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.022394180297852,
      "learning_rate": 2.1740560640732266e-05,
      "loss": 1.2244,
      "step": 1925
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.001893997192383,
      "learning_rate": 2.1736270022883295e-05,
      "loss": 1.3324,
      "step": 1926
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.297123908996582,
      "learning_rate": 2.1731979405034327e-05,
      "loss": 1.1158,
      "step": 1927
    },
    {
      "epoch": 0.55,
      "grad_norm": 10.891748428344727,
      "learning_rate": 2.1727688787185356e-05,
      "loss": 1.2127,
      "step": 1928
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.588907241821289,
      "learning_rate": 2.1723398169336385e-05,
      "loss": 1.2623,
      "step": 1929
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.479381561279297,
      "learning_rate": 2.1719107551487417e-05,
      "loss": 1.1961,
      "step": 1930
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.2621431350708,
      "learning_rate": 2.1714816933638446e-05,
      "loss": 1.3056,
      "step": 1931
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.670048713684082,
      "learning_rate": 2.1710526315789474e-05,
      "loss": 1.2837,
      "step": 1932
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.699110984802246,
      "learning_rate": 2.1706235697940503e-05,
      "loss": 1.2077,
      "step": 1933
    },
    {
      "epoch": 0.55,
      "grad_norm": 7.799557685852051,
      "learning_rate": 2.1701945080091535e-05,
      "loss": 0.8993,
      "step": 1934
    },
    {
      "epoch": 0.55,
      "grad_norm": 7.742098808288574,
      "learning_rate": 2.169765446224256e-05,
      "loss": 1.2505,
      "step": 1935
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.196280479431152,
      "learning_rate": 2.1693363844393593e-05,
      "loss": 1.3756,
      "step": 1936
    },
    {
      "epoch": 0.55,
      "grad_norm": 6.625796318054199,
      "learning_rate": 2.1689073226544625e-05,
      "loss": 0.8229,
      "step": 1937
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.674047470092773,
      "learning_rate": 2.1684782608695654e-05,
      "loss": 0.9824,
      "step": 1938
    },
    {
      "epoch": 0.55,
      "grad_norm": 7.702921390533447,
      "learning_rate": 2.1680491990846682e-05,
      "loss": 1.0809,
      "step": 1939
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.056543350219727,
      "learning_rate": 2.167620137299771e-05,
      "loss": 1.1688,
      "step": 1940
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.157732963562012,
      "learning_rate": 2.1671910755148743e-05,
      "loss": 1.201,
      "step": 1941
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.283062934875488,
      "learning_rate": 2.1667620137299772e-05,
      "loss": 1.4347,
      "step": 1942
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.770071029663086,
      "learning_rate": 2.16633295194508e-05,
      "loss": 0.9662,
      "step": 1943
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.866783142089844,
      "learning_rate": 2.165903890160183e-05,
      "loss": 1.1569,
      "step": 1944
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.816003799438477,
      "learning_rate": 2.1654748283752862e-05,
      "loss": 1.1886,
      "step": 1945
    },
    {
      "epoch": 0.56,
      "grad_norm": 11.375457763671875,
      "learning_rate": 2.165045766590389e-05,
      "loss": 1.307,
      "step": 1946
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.87993335723877,
      "learning_rate": 2.164616704805492e-05,
      "loss": 1.2194,
      "step": 1947
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.221055030822754,
      "learning_rate": 2.164187643020595e-05,
      "loss": 1.1166,
      "step": 1948
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.571791648864746,
      "learning_rate": 2.163758581235698e-05,
      "loss": 1.2822,
      "step": 1949
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.576422691345215,
      "learning_rate": 2.163329519450801e-05,
      "loss": 0.8849,
      "step": 1950
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.687113761901855,
      "learning_rate": 2.1629004576659038e-05,
      "loss": 1.1576,
      "step": 1951
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.230101585388184,
      "learning_rate": 2.162471395881007e-05,
      "loss": 0.8704,
      "step": 1952
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.570345878601074,
      "learning_rate": 2.16204233409611e-05,
      "loss": 0.9873,
      "step": 1953
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.308501243591309,
      "learning_rate": 2.1616132723112128e-05,
      "loss": 1.5126,
      "step": 1954
    },
    {
      "epoch": 0.56,
      "grad_norm": 7.877004146575928,
      "learning_rate": 2.161184210526316e-05,
      "loss": 1.0293,
      "step": 1955
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.860822677612305,
      "learning_rate": 2.160755148741419e-05,
      "loss": 1.1995,
      "step": 1956
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.284786224365234,
      "learning_rate": 2.160326086956522e-05,
      "loss": 1.1766,
      "step": 1957
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.226141929626465,
      "learning_rate": 2.1598970251716246e-05,
      "loss": 1.4969,
      "step": 1958
    },
    {
      "epoch": 0.56,
      "grad_norm": 11.073193550109863,
      "learning_rate": 2.1594679633867278e-05,
      "loss": 1.326,
      "step": 1959
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.278326988220215,
      "learning_rate": 2.1590389016018307e-05,
      "loss": 1.1496,
      "step": 1960
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.75911808013916,
      "learning_rate": 2.1586098398169336e-05,
      "loss": 1.302,
      "step": 1961
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.249210357666016,
      "learning_rate": 2.1581807780320364e-05,
      "loss": 1.4291,
      "step": 1962
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.406425476074219,
      "learning_rate": 2.1577517162471397e-05,
      "loss": 1.1631,
      "step": 1963
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.443185806274414,
      "learning_rate": 2.157322654462243e-05,
      "loss": 1.7803,
      "step": 1964
    },
    {
      "epoch": 0.56,
      "grad_norm": 6.937527656555176,
      "learning_rate": 2.1568935926773454e-05,
      "loss": 0.8683,
      "step": 1965
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.713139533996582,
      "learning_rate": 2.1564645308924486e-05,
      "loss": 1.0959,
      "step": 1966
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.732977867126465,
      "learning_rate": 2.1560354691075515e-05,
      "loss": 1.103,
      "step": 1967
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.105290412902832,
      "learning_rate": 2.1556064073226547e-05,
      "loss": 1.1155,
      "step": 1968
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.14785385131836,
      "learning_rate": 2.1551773455377573e-05,
      "loss": 1.0435,
      "step": 1969
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.170175552368164,
      "learning_rate": 2.1547482837528605e-05,
      "loss": 0.8512,
      "step": 1970
    },
    {
      "epoch": 0.56,
      "grad_norm": 12.869532585144043,
      "learning_rate": 2.1543192219679637e-05,
      "loss": 1.489,
      "step": 1971
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.479695320129395,
      "learning_rate": 2.1538901601830662e-05,
      "loss": 1.2053,
      "step": 1972
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.297906875610352,
      "learning_rate": 2.1534610983981694e-05,
      "loss": 1.0794,
      "step": 1973
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.7311429977417,
      "learning_rate": 2.1530320366132723e-05,
      "loss": 1.1172,
      "step": 1974
    },
    {
      "epoch": 0.56,
      "grad_norm": 7.909209251403809,
      "learning_rate": 2.1526029748283755e-05,
      "loss": 0.8978,
      "step": 1975
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.352812767028809,
      "learning_rate": 2.152173913043478e-05,
      "loss": 1.1804,
      "step": 1976
    },
    {
      "epoch": 0.57,
      "grad_norm": 10.12589168548584,
      "learning_rate": 2.1517448512585813e-05,
      "loss": 1.1711,
      "step": 1977
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.779499053955078,
      "learning_rate": 2.151315789473684e-05,
      "loss": 1.2041,
      "step": 1978
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.799803733825684,
      "learning_rate": 2.1508867276887874e-05,
      "loss": 1.0511,
      "step": 1979
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.467108726501465,
      "learning_rate": 2.1504576659038903e-05,
      "loss": 1.2735,
      "step": 1980
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.35204792022705,
      "learning_rate": 2.150028604118993e-05,
      "loss": 1.0897,
      "step": 1981
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.898812294006348,
      "learning_rate": 2.1495995423340963e-05,
      "loss": 1.101,
      "step": 1982
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.413460731506348,
      "learning_rate": 2.1491704805491992e-05,
      "loss": 1.1917,
      "step": 1983
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.989850997924805,
      "learning_rate": 2.148741418764302e-05,
      "loss": 1.2358,
      "step": 1984
    },
    {
      "epoch": 0.57,
      "grad_norm": 7.671233654022217,
      "learning_rate": 2.148312356979405e-05,
      "loss": 1.2357,
      "step": 1985
    },
    {
      "epoch": 0.57,
      "grad_norm": 10.515417098999023,
      "learning_rate": 2.1478832951945082e-05,
      "loss": 1.1486,
      "step": 1986
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.841854095458984,
      "learning_rate": 2.1474542334096107e-05,
      "loss": 1.3288,
      "step": 1987
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.639217376708984,
      "learning_rate": 2.147025171624714e-05,
      "loss": 1.1916,
      "step": 1988
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.984121322631836,
      "learning_rate": 2.146596109839817e-05,
      "loss": 1.2535,
      "step": 1989
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.637347221374512,
      "learning_rate": 2.14616704805492e-05,
      "loss": 1.0519,
      "step": 1990
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.064813613891602,
      "learning_rate": 2.145737986270023e-05,
      "loss": 1.2175,
      "step": 1991
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.587684631347656,
      "learning_rate": 2.1453089244851258e-05,
      "loss": 0.9991,
      "step": 1992
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.530813217163086,
      "learning_rate": 2.144879862700229e-05,
      "loss": 0.9494,
      "step": 1993
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.605618476867676,
      "learning_rate": 2.144450800915332e-05,
      "loss": 1.2244,
      "step": 1994
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.557844161987305,
      "learning_rate": 2.1440217391304348e-05,
      "loss": 1.0231,
      "step": 1995
    },
    {
      "epoch": 0.57,
      "grad_norm": 7.781949043273926,
      "learning_rate": 2.1435926773455376e-05,
      "loss": 1.192,
      "step": 1996
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.588788986206055,
      "learning_rate": 2.143163615560641e-05,
      "loss": 1.2064,
      "step": 1997
    },
    {
      "epoch": 0.57,
      "grad_norm": 10.51740550994873,
      "learning_rate": 2.1427345537757437e-05,
      "loss": 1.3746,
      "step": 1998
    },
    {
      "epoch": 0.57,
      "grad_norm": 10.21289348602295,
      "learning_rate": 2.1423054919908466e-05,
      "loss": 1.1652,
      "step": 1999
    },
    {
      "epoch": 0.57,
      "grad_norm": 7.317837715148926,
      "learning_rate": 2.1418764302059498e-05,
      "loss": 1.2338,
      "step": 2000
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.614209175109863,
      "learning_rate": 2.1414473684210527e-05,
      "loss": 1.1308,
      "step": 2001
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.713343620300293,
      "learning_rate": 2.1410183066361556e-05,
      "loss": 1.0924,
      "step": 2002
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.420449256896973,
      "learning_rate": 2.1405892448512585e-05,
      "loss": 1.2524,
      "step": 2003
    },
    {
      "epoch": 0.57,
      "grad_norm": 10.03909683227539,
      "learning_rate": 2.1401601830663617e-05,
      "loss": 1.0411,
      "step": 2004
    },
    {
      "epoch": 0.57,
      "grad_norm": 10.105528831481934,
      "learning_rate": 2.1397311212814645e-05,
      "loss": 1.0933,
      "step": 2005
    },
    {
      "epoch": 0.57,
      "grad_norm": 11.266928672790527,
      "learning_rate": 2.1393020594965674e-05,
      "loss": 1.2867,
      "step": 2006
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.959481239318848,
      "learning_rate": 2.1388729977116706e-05,
      "loss": 0.9824,
      "step": 2007
    },
    {
      "epoch": 0.57,
      "grad_norm": 11.12039852142334,
      "learning_rate": 2.1384439359267735e-05,
      "loss": 1.3624,
      "step": 2008
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.747989654541016,
      "learning_rate": 2.1380148741418767e-05,
      "loss": 0.9588,
      "step": 2009
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.255857467651367,
      "learning_rate": 2.1375858123569793e-05,
      "loss": 1.1861,
      "step": 2010
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.196959495544434,
      "learning_rate": 2.1371567505720825e-05,
      "loss": 1.1194,
      "step": 2011
    },
    {
      "epoch": 0.58,
      "grad_norm": 7.997098922729492,
      "learning_rate": 2.1367276887871854e-05,
      "loss": 1.2389,
      "step": 2012
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.245403289794922,
      "learning_rate": 2.1362986270022882e-05,
      "loss": 1.0143,
      "step": 2013
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.781437873840332,
      "learning_rate": 2.1358695652173915e-05,
      "loss": 1.1881,
      "step": 2014
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.052122116088867,
      "learning_rate": 2.1354405034324943e-05,
      "loss": 0.8558,
      "step": 2015
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.379697799682617,
      "learning_rate": 2.1350114416475975e-05,
      "loss": 1.1805,
      "step": 2016
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.204866409301758,
      "learning_rate": 2.1345823798627e-05,
      "loss": 0.9265,
      "step": 2017
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.82898235321045,
      "learning_rate": 2.1341533180778033e-05,
      "loss": 1.2062,
      "step": 2018
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.235489845275879,
      "learning_rate": 2.1337242562929062e-05,
      "loss": 1.1957,
      "step": 2019
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.028446197509766,
      "learning_rate": 2.1332951945080094e-05,
      "loss": 1.0049,
      "step": 2020
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.168365478515625,
      "learning_rate": 2.132866132723112e-05,
      "loss": 0.9962,
      "step": 2021
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.710518836975098,
      "learning_rate": 2.132437070938215e-05,
      "loss": 1.1069,
      "step": 2022
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.715204238891602,
      "learning_rate": 2.1320080091533184e-05,
      "loss": 1.3578,
      "step": 2023
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.091570854187012,
      "learning_rate": 2.1315789473684212e-05,
      "loss": 0.8876,
      "step": 2024
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.888350486755371,
      "learning_rate": 2.131149885583524e-05,
      "loss": 1.1702,
      "step": 2025
    },
    {
      "epoch": 0.58,
      "grad_norm": 11.44349479675293,
      "learning_rate": 2.130720823798627e-05,
      "loss": 1.715,
      "step": 2026
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.71235466003418,
      "learning_rate": 2.1302917620137302e-05,
      "loss": 1.2066,
      "step": 2027
    },
    {
      "epoch": 0.58,
      "grad_norm": 11.094636917114258,
      "learning_rate": 2.1298627002288327e-05,
      "loss": 0.91,
      "step": 2028
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.74800968170166,
      "learning_rate": 2.129433638443936e-05,
      "loss": 1.0944,
      "step": 2029
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.274219512939453,
      "learning_rate": 2.129004576659039e-05,
      "loss": 0.9759,
      "step": 2030
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.601633071899414,
      "learning_rate": 2.128575514874142e-05,
      "loss": 1.491,
      "step": 2031
    },
    {
      "epoch": 0.58,
      "grad_norm": 7.937924385070801,
      "learning_rate": 2.128146453089245e-05,
      "loss": 1.0133,
      "step": 2032
    },
    {
      "epoch": 0.58,
      "grad_norm": 7.562763214111328,
      "learning_rate": 2.1277173913043478e-05,
      "loss": 0.9443,
      "step": 2033
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.2804536819458,
      "learning_rate": 2.127288329519451e-05,
      "loss": 1.473,
      "step": 2034
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.761529922485352,
      "learning_rate": 2.126859267734554e-05,
      "loss": 1.2093,
      "step": 2035
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.06948471069336,
      "learning_rate": 2.1264302059496568e-05,
      "loss": 1.2422,
      "step": 2036
    },
    {
      "epoch": 0.58,
      "grad_norm": 7.157027244567871,
      "learning_rate": 2.1260011441647597e-05,
      "loss": 0.8158,
      "step": 2037
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.515137672424316,
      "learning_rate": 2.125572082379863e-05,
      "loss": 1.424,
      "step": 2038
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.041302680969238,
      "learning_rate": 2.1251430205949654e-05,
      "loss": 0.908,
      "step": 2039
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.615650177001953,
      "learning_rate": 2.1247139588100686e-05,
      "loss": 1.4213,
      "step": 2040
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.955256462097168,
      "learning_rate": 2.124284897025172e-05,
      "loss": 1.3073,
      "step": 2041
    },
    {
      "epoch": 0.58,
      "grad_norm": 7.355934143066406,
      "learning_rate": 2.1238558352402747e-05,
      "loss": 0.8669,
      "step": 2042
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.485574722290039,
      "learning_rate": 2.1234267734553776e-05,
      "loss": 1.4687,
      "step": 2043
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.188101768493652,
      "learning_rate": 2.1229977116704805e-05,
      "loss": 1.083,
      "step": 2044
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.055419921875,
      "learning_rate": 2.1225686498855837e-05,
      "loss": 1.189,
      "step": 2045
    },
    {
      "epoch": 0.59,
      "grad_norm": 8.69979190826416,
      "learning_rate": 2.1221395881006866e-05,
      "loss": 0.8412,
      "step": 2046
    },
    {
      "epoch": 0.59,
      "grad_norm": 8.15188217163086,
      "learning_rate": 2.1217105263157894e-05,
      "loss": 1.0282,
      "step": 2047
    },
    {
      "epoch": 0.59,
      "grad_norm": 11.337603569030762,
      "learning_rate": 2.1212814645308927e-05,
      "loss": 1.1298,
      "step": 2048
    },
    {
      "epoch": 0.59,
      "grad_norm": 11.604207038879395,
      "learning_rate": 2.1208524027459955e-05,
      "loss": 1.26,
      "step": 2049
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.798079490661621,
      "learning_rate": 2.1204233409610984e-05,
      "loss": 1.3073,
      "step": 2050
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.075469017028809,
      "learning_rate": 2.1199942791762013e-05,
      "loss": 1.1813,
      "step": 2051
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.19278335571289,
      "learning_rate": 2.1195652173913045e-05,
      "loss": 1.0175,
      "step": 2052
    },
    {
      "epoch": 0.59,
      "grad_norm": 8.641410827636719,
      "learning_rate": 2.1191361556064074e-05,
      "loss": 1.0907,
      "step": 2053
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.665210723876953,
      "learning_rate": 2.1187070938215103e-05,
      "loss": 1.5654,
      "step": 2054
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.677375793457031,
      "learning_rate": 2.118278032036613e-05,
      "loss": 1.3304,
      "step": 2055
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.193497657775879,
      "learning_rate": 2.1178489702517163e-05,
      "loss": 1.1239,
      "step": 2056
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.984728813171387,
      "learning_rate": 2.1174199084668196e-05,
      "loss": 1.3757,
      "step": 2057
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.717297554016113,
      "learning_rate": 2.116990846681922e-05,
      "loss": 1.147,
      "step": 2058
    },
    {
      "epoch": 0.59,
      "grad_norm": 8.830130577087402,
      "learning_rate": 2.1165617848970253e-05,
      "loss": 0.9154,
      "step": 2059
    },
    {
      "epoch": 0.59,
      "grad_norm": 8.209005355834961,
      "learning_rate": 2.1161327231121282e-05,
      "loss": 1.4371,
      "step": 2060
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.334105491638184,
      "learning_rate": 2.1157036613272314e-05,
      "loss": 1.055,
      "step": 2061
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.272869110107422,
      "learning_rate": 2.115274599542334e-05,
      "loss": 1.049,
      "step": 2062
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.7098388671875,
      "learning_rate": 2.114845537757437e-05,
      "loss": 1.1033,
      "step": 2063
    },
    {
      "epoch": 0.59,
      "grad_norm": 8.774618148803711,
      "learning_rate": 2.11441647597254e-05,
      "loss": 0.8308,
      "step": 2064
    },
    {
      "epoch": 0.59,
      "grad_norm": 8.327656745910645,
      "learning_rate": 2.113987414187643e-05,
      "loss": 1.0889,
      "step": 2065
    },
    {
      "epoch": 0.59,
      "grad_norm": 7.887519359588623,
      "learning_rate": 2.113558352402746e-05,
      "loss": 1.0136,
      "step": 2066
    },
    {
      "epoch": 0.59,
      "grad_norm": 8.921797752380371,
      "learning_rate": 2.113129290617849e-05,
      "loss": 1.0613,
      "step": 2067
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.537035942077637,
      "learning_rate": 2.1127002288329522e-05,
      "loss": 1.2971,
      "step": 2068
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.419792175292969,
      "learning_rate": 2.1122711670480548e-05,
      "loss": 1.504,
      "step": 2069
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.048227310180664,
      "learning_rate": 2.111842105263158e-05,
      "loss": 1.1806,
      "step": 2070
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.106603622436523,
      "learning_rate": 2.111413043478261e-05,
      "loss": 1.4529,
      "step": 2071
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.84475040435791,
      "learning_rate": 2.110983981693364e-05,
      "loss": 0.9912,
      "step": 2072
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.732037544250488,
      "learning_rate": 2.1105549199084666e-05,
      "loss": 1.0068,
      "step": 2073
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.757185935974121,
      "learning_rate": 2.1101258581235698e-05,
      "loss": 1.2628,
      "step": 2074
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.87586498260498,
      "learning_rate": 2.109696796338673e-05,
      "loss": 1.5249,
      "step": 2075
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.647394180297852,
      "learning_rate": 2.109267734553776e-05,
      "loss": 1.3032,
      "step": 2076
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.906692504882812,
      "learning_rate": 2.1088386727688788e-05,
      "loss": 1.1063,
      "step": 2077
    },
    {
      "epoch": 0.59,
      "grad_norm": 6.8456926345825195,
      "learning_rate": 2.1084096109839817e-05,
      "loss": 0.728,
      "step": 2078
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.565463066101074,
      "learning_rate": 2.107980549199085e-05,
      "loss": 1.2561,
      "step": 2079
    },
    {
      "epoch": 0.59,
      "grad_norm": 8.672847747802734,
      "learning_rate": 2.1075514874141874e-05,
      "loss": 1.2062,
      "step": 2080
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.219789505004883,
      "learning_rate": 2.1071224256292906e-05,
      "loss": 1.1501,
      "step": 2081
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.40511417388916,
      "learning_rate": 2.1066933638443935e-05,
      "loss": 1.0968,
      "step": 2082
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.463879585266113,
      "learning_rate": 2.1062643020594967e-05,
      "loss": 0.9518,
      "step": 2083
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.839200973510742,
      "learning_rate": 2.1058352402745996e-05,
      "loss": 1.1395,
      "step": 2084
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.488741874694824,
      "learning_rate": 2.1054061784897025e-05,
      "loss": 1.0916,
      "step": 2085
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.752121925354004,
      "learning_rate": 2.1049771167048057e-05,
      "loss": 1.1249,
      "step": 2086
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.362652778625488,
      "learning_rate": 2.1045480549199086e-05,
      "loss": 1.4336,
      "step": 2087
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.308162689208984,
      "learning_rate": 2.1041189931350115e-05,
      "loss": 0.9118,
      "step": 2088
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.194893836975098,
      "learning_rate": 2.1036899313501143e-05,
      "loss": 0.9576,
      "step": 2089
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.937129020690918,
      "learning_rate": 2.1032608695652175e-05,
      "loss": 1.1631,
      "step": 2090
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.854636192321777,
      "learning_rate": 2.1028318077803204e-05,
      "loss": 1.2521,
      "step": 2091
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.050986289978027,
      "learning_rate": 2.1024027459954233e-05,
      "loss": 1.081,
      "step": 2092
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.640084266662598,
      "learning_rate": 2.1019736842105265e-05,
      "loss": 1.0249,
      "step": 2093
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.013104438781738,
      "learning_rate": 2.1015446224256294e-05,
      "loss": 0.9533,
      "step": 2094
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.025662422180176,
      "learning_rate": 2.1011155606407323e-05,
      "loss": 1.1224,
      "step": 2095
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.688165664672852,
      "learning_rate": 2.100686498855835e-05,
      "loss": 1.0492,
      "step": 2096
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.700384140014648,
      "learning_rate": 2.1002574370709384e-05,
      "loss": 1.0879,
      "step": 2097
    },
    {
      "epoch": 0.6,
      "grad_norm": 11.119878768920898,
      "learning_rate": 2.0998283752860412e-05,
      "loss": 1.4303,
      "step": 2098
    },
    {
      "epoch": 0.6,
      "grad_norm": 7.985533714294434,
      "learning_rate": 2.099399313501144e-05,
      "loss": 1.0919,
      "step": 2099
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.751565933227539,
      "learning_rate": 2.0989702517162473e-05,
      "loss": 1.2801,
      "step": 2100
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.29434585571289,
      "learning_rate": 2.0985411899313502e-05,
      "loss": 1.0066,
      "step": 2101
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.124516487121582,
      "learning_rate": 2.0981121281464534e-05,
      "loss": 1.2844,
      "step": 2102
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.318306922912598,
      "learning_rate": 2.097683066361556e-05,
      "loss": 1.5056,
      "step": 2103
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.364325523376465,
      "learning_rate": 2.0972540045766592e-05,
      "loss": 1.052,
      "step": 2104
    },
    {
      "epoch": 0.6,
      "grad_norm": 7.576369285583496,
      "learning_rate": 2.096824942791762e-05,
      "loss": 1.0159,
      "step": 2105
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.611291885375977,
      "learning_rate": 2.096395881006865e-05,
      "loss": 1.1869,
      "step": 2106
    },
    {
      "epoch": 0.6,
      "grad_norm": 7.1158061027526855,
      "learning_rate": 2.0959668192219678e-05,
      "loss": 1.0589,
      "step": 2107
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.34353256225586,
      "learning_rate": 2.095537757437071e-05,
      "loss": 1.1575,
      "step": 2108
    },
    {
      "epoch": 0.6,
      "grad_norm": 7.847663879394531,
      "learning_rate": 2.0951086956521742e-05,
      "loss": 1.0464,
      "step": 2109
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.272878646850586,
      "learning_rate": 2.0946796338672768e-05,
      "loss": 1.2564,
      "step": 2110
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.065635681152344,
      "learning_rate": 2.09425057208238e-05,
      "loss": 1.5392,
      "step": 2111
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.396218299865723,
      "learning_rate": 2.093821510297483e-05,
      "loss": 1.1943,
      "step": 2112
    },
    {
      "epoch": 0.6,
      "grad_norm": 6.927637577056885,
      "learning_rate": 2.093392448512586e-05,
      "loss": 0.9002,
      "step": 2113
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.358610153198242,
      "learning_rate": 2.0929633867276886e-05,
      "loss": 1.3487,
      "step": 2114
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.768491744995117,
      "learning_rate": 2.092534324942792e-05,
      "loss": 1.263,
      "step": 2115
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.752141952514648,
      "learning_rate": 2.0921052631578947e-05,
      "loss": 1.1306,
      "step": 2116
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.661421775817871,
      "learning_rate": 2.0916762013729976e-05,
      "loss": 1.0151,
      "step": 2117
    },
    {
      "epoch": 0.61,
      "grad_norm": 10.647745132446289,
      "learning_rate": 2.0912471395881008e-05,
      "loss": 1.0757,
      "step": 2118
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.90361213684082,
      "learning_rate": 2.0908180778032037e-05,
      "loss": 1.1973,
      "step": 2119
    },
    {
      "epoch": 0.61,
      "grad_norm": 10.638439178466797,
      "learning_rate": 2.090389016018307e-05,
      "loss": 1.0734,
      "step": 2120
    },
    {
      "epoch": 0.61,
      "grad_norm": 10.39708423614502,
      "learning_rate": 2.0899599542334094e-05,
      "loss": 1.1697,
      "step": 2121
    },
    {
      "epoch": 0.61,
      "grad_norm": 11.243378639221191,
      "learning_rate": 2.0895308924485127e-05,
      "loss": 1.4658,
      "step": 2122
    },
    {
      "epoch": 0.61,
      "grad_norm": 10.966980934143066,
      "learning_rate": 2.0891018306636155e-05,
      "loss": 1.3194,
      "step": 2123
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.226202011108398,
      "learning_rate": 2.0886727688787187e-05,
      "loss": 1.2252,
      "step": 2124
    },
    {
      "epoch": 0.61,
      "grad_norm": 7.807674407958984,
      "learning_rate": 2.0882437070938213e-05,
      "loss": 0.936,
      "step": 2125
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.516277313232422,
      "learning_rate": 2.0878146453089245e-05,
      "loss": 1.1853,
      "step": 2126
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.409971237182617,
      "learning_rate": 2.0873855835240277e-05,
      "loss": 1.1543,
      "step": 2127
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.993494987487793,
      "learning_rate": 2.0869565217391306e-05,
      "loss": 1.4076,
      "step": 2128
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.260432243347168,
      "learning_rate": 2.0865274599542335e-05,
      "loss": 1.1792,
      "step": 2129
    },
    {
      "epoch": 0.61,
      "grad_norm": 7.629167556762695,
      "learning_rate": 2.0860983981693363e-05,
      "loss": 1.1718,
      "step": 2130
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.276948928833008,
      "learning_rate": 2.0856693363844396e-05,
      "loss": 1.2379,
      "step": 2131
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.391092300415039,
      "learning_rate": 2.085240274599542e-05,
      "loss": 1.0704,
      "step": 2132
    },
    {
      "epoch": 0.61,
      "grad_norm": 7.370236396789551,
      "learning_rate": 2.0848112128146453e-05,
      "loss": 1.2056,
      "step": 2133
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.195940971374512,
      "learning_rate": 2.0843821510297485e-05,
      "loss": 1.1338,
      "step": 2134
    },
    {
      "epoch": 0.61,
      "grad_norm": 7.95945930480957,
      "learning_rate": 2.0839530892448514e-05,
      "loss": 1.0615,
      "step": 2135
    },
    {
      "epoch": 0.61,
      "grad_norm": 7.3939056396484375,
      "learning_rate": 2.0835240274599543e-05,
      "loss": 0.9169,
      "step": 2136
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.680707931518555,
      "learning_rate": 2.083094965675057e-05,
      "loss": 1.3699,
      "step": 2137
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.372149467468262,
      "learning_rate": 2.0826659038901604e-05,
      "loss": 1.2526,
      "step": 2138
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.262450218200684,
      "learning_rate": 2.0822368421052633e-05,
      "loss": 1.2425,
      "step": 2139
    },
    {
      "epoch": 0.61,
      "grad_norm": 7.113325119018555,
      "learning_rate": 2.081807780320366e-05,
      "loss": 0.7992,
      "step": 2140
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.704121589660645,
      "learning_rate": 2.081378718535469e-05,
      "loss": 1.351,
      "step": 2141
    },
    {
      "epoch": 0.61,
      "grad_norm": 10.085105895996094,
      "learning_rate": 2.0809496567505722e-05,
      "loss": 1.2512,
      "step": 2142
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.638779640197754,
      "learning_rate": 2.080520594965675e-05,
      "loss": 1.1392,
      "step": 2143
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.373291015625,
      "learning_rate": 2.080091533180778e-05,
      "loss": 1.312,
      "step": 2144
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.3704833984375,
      "learning_rate": 2.0796624713958812e-05,
      "loss": 1.0496,
      "step": 2145
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.622896194458008,
      "learning_rate": 2.079233409610984e-05,
      "loss": 1.1862,
      "step": 2146
    },
    {
      "epoch": 0.61,
      "grad_norm": 10.123673439025879,
      "learning_rate": 2.078804347826087e-05,
      "loss": 1.1337,
      "step": 2147
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.257584571838379,
      "learning_rate": 2.0783752860411898e-05,
      "loss": 1.0569,
      "step": 2148
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.453557968139648,
      "learning_rate": 2.077946224256293e-05,
      "loss": 1.0547,
      "step": 2149
    },
    {
      "epoch": 0.61,
      "grad_norm": 11.368425369262695,
      "learning_rate": 2.077517162471396e-05,
      "loss": 1.3893,
      "step": 2150
    },
    {
      "epoch": 0.62,
      "grad_norm": 6.905810356140137,
      "learning_rate": 2.0770881006864988e-05,
      "loss": 0.8847,
      "step": 2151
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.193455696105957,
      "learning_rate": 2.076659038901602e-05,
      "loss": 1.1597,
      "step": 2152
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.975447654724121,
      "learning_rate": 2.076229977116705e-05,
      "loss": 1.4165,
      "step": 2153
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.44798469543457,
      "learning_rate": 2.075800915331808e-05,
      "loss": 0.8604,
      "step": 2154
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.232073783874512,
      "learning_rate": 2.0753718535469106e-05,
      "loss": 0.9972,
      "step": 2155
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.040815353393555,
      "learning_rate": 2.074942791762014e-05,
      "loss": 1.0944,
      "step": 2156
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.155389785766602,
      "learning_rate": 2.0745137299771167e-05,
      "loss": 0.9028,
      "step": 2157
    },
    {
      "epoch": 0.62,
      "grad_norm": 7.844616413116455,
      "learning_rate": 2.0740846681922196e-05,
      "loss": 1.0819,
      "step": 2158
    },
    {
      "epoch": 0.62,
      "grad_norm": 10.479866027832031,
      "learning_rate": 2.0736556064073225e-05,
      "loss": 1.247,
      "step": 2159
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.226712226867676,
      "learning_rate": 2.0732265446224257e-05,
      "loss": 1.2943,
      "step": 2160
    },
    {
      "epoch": 0.62,
      "grad_norm": 10.073928833007812,
      "learning_rate": 2.072797482837529e-05,
      "loss": 1.1224,
      "step": 2161
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.664060592651367,
      "learning_rate": 2.0723684210526315e-05,
      "loss": 1.1303,
      "step": 2162
    },
    {
      "epoch": 0.62,
      "grad_norm": 7.801047325134277,
      "learning_rate": 2.0719393592677347e-05,
      "loss": 0.9249,
      "step": 2163
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.232869148254395,
      "learning_rate": 2.0715102974828375e-05,
      "loss": 1.2656,
      "step": 2164
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.818890571594238,
      "learning_rate": 2.0710812356979408e-05,
      "loss": 1.2511,
      "step": 2165
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.902132987976074,
      "learning_rate": 2.0706521739130433e-05,
      "loss": 0.9535,
      "step": 2166
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.043094635009766,
      "learning_rate": 2.0702231121281465e-05,
      "loss": 1.0496,
      "step": 2167
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.18312931060791,
      "learning_rate": 2.0697940503432497e-05,
      "loss": 1.0329,
      "step": 2168
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.758755683898926,
      "learning_rate": 2.0693649885583523e-05,
      "loss": 0.928,
      "step": 2169
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.533914566040039,
      "learning_rate": 2.0689359267734555e-05,
      "loss": 0.9896,
      "step": 2170
    },
    {
      "epoch": 0.62,
      "grad_norm": 10.245708465576172,
      "learning_rate": 2.0685068649885584e-05,
      "loss": 0.9935,
      "step": 2171
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.271585464477539,
      "learning_rate": 2.0680778032036616e-05,
      "loss": 1.0515,
      "step": 2172
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.259747505187988,
      "learning_rate": 2.067648741418764e-05,
      "loss": 1.3014,
      "step": 2173
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.592206954956055,
      "learning_rate": 2.0672196796338673e-05,
      "loss": 1.0706,
      "step": 2174
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.149291038513184,
      "learning_rate": 2.0667906178489702e-05,
      "loss": 1.0591,
      "step": 2175
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.54188346862793,
      "learning_rate": 2.0663615560640734e-05,
      "loss": 1.1857,
      "step": 2176
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.516733169555664,
      "learning_rate": 2.0659324942791763e-05,
      "loss": 1.0539,
      "step": 2177
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.936399459838867,
      "learning_rate": 2.0655034324942792e-05,
      "loss": 1.0294,
      "step": 2178
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.519046783447266,
      "learning_rate": 2.0650743707093824e-05,
      "loss": 0.9475,
      "step": 2179
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.646597862243652,
      "learning_rate": 2.0646453089244853e-05,
      "loss": 1.376,
      "step": 2180
    },
    {
      "epoch": 0.62,
      "grad_norm": 10.472565650939941,
      "learning_rate": 2.064216247139588e-05,
      "loss": 0.9122,
      "step": 2181
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.792078971862793,
      "learning_rate": 2.063787185354691e-05,
      "loss": 1.2285,
      "step": 2182
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.674057006835938,
      "learning_rate": 2.0633581235697942e-05,
      "loss": 1.0658,
      "step": 2183
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.935258865356445,
      "learning_rate": 2.0629290617848968e-05,
      "loss": 0.9981,
      "step": 2184
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.39003849029541,
      "learning_rate": 2.0625e-05,
      "loss": 1.0571,
      "step": 2185
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.747468948364258,
      "learning_rate": 2.0620709382151032e-05,
      "loss": 0.9429,
      "step": 2186
    },
    {
      "epoch": 0.63,
      "grad_norm": 12.674715042114258,
      "learning_rate": 2.061641876430206e-05,
      "loss": 1.3938,
      "step": 2187
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.458878517150879,
      "learning_rate": 2.061212814645309e-05,
      "loss": 1.1022,
      "step": 2188
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.754960060119629,
      "learning_rate": 2.060783752860412e-05,
      "loss": 1.3661,
      "step": 2189
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.990206718444824,
      "learning_rate": 2.060354691075515e-05,
      "loss": 1.08,
      "step": 2190
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.211421012878418,
      "learning_rate": 2.059925629290618e-05,
      "loss": 1.1255,
      "step": 2191
    },
    {
      "epoch": 0.63,
      "grad_norm": 10.06490707397461,
      "learning_rate": 2.0594965675057208e-05,
      "loss": 1.3158,
      "step": 2192
    },
    {
      "epoch": 0.63,
      "grad_norm": 7.738185405731201,
      "learning_rate": 2.0590675057208237e-05,
      "loss": 0.8872,
      "step": 2193
    },
    {
      "epoch": 0.63,
      "grad_norm": 7.422226905822754,
      "learning_rate": 2.058638443935927e-05,
      "loss": 1.1407,
      "step": 2194
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.108756065368652,
      "learning_rate": 2.0582093821510298e-05,
      "loss": 1.0932,
      "step": 2195
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.866443634033203,
      "learning_rate": 2.0577803203661326e-05,
      "loss": 1.3087,
      "step": 2196
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.107356071472168,
      "learning_rate": 2.057351258581236e-05,
      "loss": 1.3113,
      "step": 2197
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.157082557678223,
      "learning_rate": 2.0569221967963387e-05,
      "loss": 1.143,
      "step": 2198
    },
    {
      "epoch": 0.63,
      "grad_norm": 7.636971473693848,
      "learning_rate": 2.0564931350114416e-05,
      "loss": 1.2641,
      "step": 2199
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.182448387145996,
      "learning_rate": 2.0560640732265445e-05,
      "loss": 1.1857,
      "step": 2200
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.556438446044922,
      "learning_rate": 2.0556350114416477e-05,
      "loss": 1.225,
      "step": 2201
    },
    {
      "epoch": 0.63,
      "grad_norm": 7.437405109405518,
      "learning_rate": 2.0552059496567506e-05,
      "loss": 0.8973,
      "step": 2202
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.594822883605957,
      "learning_rate": 2.0547768878718535e-05,
      "loss": 1.1858,
      "step": 2203
    },
    {
      "epoch": 0.63,
      "grad_norm": 10.339887619018555,
      "learning_rate": 2.0543478260869567e-05,
      "loss": 1.3691,
      "step": 2204
    },
    {
      "epoch": 0.63,
      "grad_norm": 7.668767929077148,
      "learning_rate": 2.0539187643020596e-05,
      "loss": 1.1917,
      "step": 2205
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.312116622924805,
      "learning_rate": 2.0534897025171628e-05,
      "loss": 1.2876,
      "step": 2206
    },
    {
      "epoch": 0.63,
      "grad_norm": 7.3700151443481445,
      "learning_rate": 2.0530606407322653e-05,
      "loss": 0.9395,
      "step": 2207
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.707545280456543,
      "learning_rate": 2.0526315789473685e-05,
      "loss": 0.936,
      "step": 2208
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.36646842956543,
      "learning_rate": 2.0522025171624714e-05,
      "loss": 1.0538,
      "step": 2209
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.574594497680664,
      "learning_rate": 2.0517734553775743e-05,
      "loss": 1.2534,
      "step": 2210
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.412030220031738,
      "learning_rate": 2.0513443935926775e-05,
      "loss": 0.9075,
      "step": 2211
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.844574928283691,
      "learning_rate": 2.0509153318077804e-05,
      "loss": 1.0429,
      "step": 2212
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.063969612121582,
      "learning_rate": 2.0504862700228836e-05,
      "loss": 1.1042,
      "step": 2213
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.998139381408691,
      "learning_rate": 2.050057208237986e-05,
      "loss": 1.2145,
      "step": 2214
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.991023063659668,
      "learning_rate": 2.0496281464530893e-05,
      "loss": 1.0048,
      "step": 2215
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.947741508483887,
      "learning_rate": 2.0491990846681922e-05,
      "loss": 1.3416,
      "step": 2216
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.170355796813965,
      "learning_rate": 2.0487700228832954e-05,
      "loss": 0.8943,
      "step": 2217
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.764482498168945,
      "learning_rate": 2.048340961098398e-05,
      "loss": 0.9741,
      "step": 2218
    },
    {
      "epoch": 0.63,
      "grad_norm": 11.503988265991211,
      "learning_rate": 2.0479118993135012e-05,
      "loss": 1.3174,
      "step": 2219
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.56205940246582,
      "learning_rate": 2.0474828375286044e-05,
      "loss": 1.0041,
      "step": 2220
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.355834007263184,
      "learning_rate": 2.0470537757437073e-05,
      "loss": 1.0502,
      "step": 2221
    },
    {
      "epoch": 0.64,
      "grad_norm": 11.749591827392578,
      "learning_rate": 2.04662471395881e-05,
      "loss": 1.0767,
      "step": 2222
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.654799461364746,
      "learning_rate": 2.046195652173913e-05,
      "loss": 1.28,
      "step": 2223
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.626363754272461,
      "learning_rate": 2.0457665903890162e-05,
      "loss": 0.9684,
      "step": 2224
    },
    {
      "epoch": 0.64,
      "grad_norm": 7.304511070251465,
      "learning_rate": 2.0453375286041188e-05,
      "loss": 0.9854,
      "step": 2225
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.409841537475586,
      "learning_rate": 2.044908466819222e-05,
      "loss": 1.0368,
      "step": 2226
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.896282196044922,
      "learning_rate": 2.044479405034325e-05,
      "loss": 1.0266,
      "step": 2227
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.600053787231445,
      "learning_rate": 2.044050343249428e-05,
      "loss": 1.1626,
      "step": 2228
    },
    {
      "epoch": 0.64,
      "grad_norm": 11.362189292907715,
      "learning_rate": 2.043621281464531e-05,
      "loss": 1.2641,
      "step": 2229
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.800779342651367,
      "learning_rate": 2.043192219679634e-05,
      "loss": 1.3698,
      "step": 2230
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.380990982055664,
      "learning_rate": 2.042763157894737e-05,
      "loss": 1.2739,
      "step": 2231
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.903562545776367,
      "learning_rate": 2.04233409610984e-05,
      "loss": 1.1876,
      "step": 2232
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.238883972167969,
      "learning_rate": 2.0419050343249428e-05,
      "loss": 1.1984,
      "step": 2233
    },
    {
      "epoch": 0.64,
      "grad_norm": 11.166067123413086,
      "learning_rate": 2.0414759725400457e-05,
      "loss": 1.215,
      "step": 2234
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.688907623291016,
      "learning_rate": 2.041046910755149e-05,
      "loss": 1.1086,
      "step": 2235
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.190163612365723,
      "learning_rate": 2.0406178489702514e-05,
      "loss": 0.9761,
      "step": 2236
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.373835563659668,
      "learning_rate": 2.0401887871853547e-05,
      "loss": 1.1574,
      "step": 2237
    },
    {
      "epoch": 0.64,
      "grad_norm": 10.375309944152832,
      "learning_rate": 2.039759725400458e-05,
      "loss": 1.4149,
      "step": 2238
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.415925979614258,
      "learning_rate": 2.0393306636155608e-05,
      "loss": 1.238,
      "step": 2239
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.052910804748535,
      "learning_rate": 2.0389016018306636e-05,
      "loss": 1.2935,
      "step": 2240
    },
    {
      "epoch": 0.64,
      "grad_norm": 10.221535682678223,
      "learning_rate": 2.0384725400457665e-05,
      "loss": 1.6285,
      "step": 2241
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.205775260925293,
      "learning_rate": 2.0380434782608697e-05,
      "loss": 1.2095,
      "step": 2242
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.49085521697998,
      "learning_rate": 2.0376144164759726e-05,
      "loss": 1.1456,
      "step": 2243
    },
    {
      "epoch": 0.64,
      "grad_norm": 10.231178283691406,
      "learning_rate": 2.0371853546910755e-05,
      "loss": 0.8561,
      "step": 2244
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.529679298400879,
      "learning_rate": 2.0367562929061787e-05,
      "loss": 0.9702,
      "step": 2245
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.257425308227539,
      "learning_rate": 2.0363272311212816e-05,
      "loss": 1.1129,
      "step": 2246
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.610989570617676,
      "learning_rate": 2.0358981693363848e-05,
      "loss": 1.0532,
      "step": 2247
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.476094245910645,
      "learning_rate": 2.0354691075514873e-05,
      "loss": 0.966,
      "step": 2248
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.928940773010254,
      "learning_rate": 2.0350400457665905e-05,
      "loss": 1.1766,
      "step": 2249
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.589329719543457,
      "learning_rate": 2.0346109839816934e-05,
      "loss": 1.0415,
      "step": 2250
    },
    {
      "epoch": 0.64,
      "grad_norm": 10.908214569091797,
      "learning_rate": 2.0341819221967963e-05,
      "loss": 1.4506,
      "step": 2251
    },
    {
      "epoch": 0.64,
      "grad_norm": 10.8147554397583,
      "learning_rate": 2.0337528604118992e-05,
      "loss": 1.2983,
      "step": 2252
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.463642120361328,
      "learning_rate": 2.0333237986270024e-05,
      "loss": 1.0882,
      "step": 2253
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.379366874694824,
      "learning_rate": 2.0328947368421056e-05,
      "loss": 0.9897,
      "step": 2254
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.763785362243652,
      "learning_rate": 2.032465675057208e-05,
      "loss": 1.098,
      "step": 2255
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.14033031463623,
      "learning_rate": 2.0320366132723114e-05,
      "loss": 1.1363,
      "step": 2256
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.258912086486816,
      "learning_rate": 2.0316075514874142e-05,
      "loss": 1.1787,
      "step": 2257
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.259736061096191,
      "learning_rate": 2.0311784897025174e-05,
      "loss": 1.1174,
      "step": 2258
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.554972648620605,
      "learning_rate": 2.03074942791762e-05,
      "loss": 1.0788,
      "step": 2259
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.994354248046875,
      "learning_rate": 2.0303203661327232e-05,
      "loss": 1.2247,
      "step": 2260
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.856390953063965,
      "learning_rate": 2.029891304347826e-05,
      "loss": 1.203,
      "step": 2261
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.640615463256836,
      "learning_rate": 2.029462242562929e-05,
      "loss": 1.0098,
      "step": 2262
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.442652702331543,
      "learning_rate": 2.0290331807780322e-05,
      "loss": 0.8984,
      "step": 2263
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.604318618774414,
      "learning_rate": 2.028604118993135e-05,
      "loss": 1.2527,
      "step": 2264
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.059185981750488,
      "learning_rate": 2.0281750572082383e-05,
      "loss": 1.1368,
      "step": 2265
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.449300765991211,
      "learning_rate": 2.0277459954233408e-05,
      "loss": 1.3336,
      "step": 2266
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.19174861907959,
      "learning_rate": 2.027316933638444e-05,
      "loss": 0.9381,
      "step": 2267
    },
    {
      "epoch": 0.65,
      "grad_norm": 7.822282791137695,
      "learning_rate": 2.026887871853547e-05,
      "loss": 1.2018,
      "step": 2268
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.758184432983398,
      "learning_rate": 2.02645881006865e-05,
      "loss": 1.0484,
      "step": 2269
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.32040786743164,
      "learning_rate": 2.0260297482837526e-05,
      "loss": 1.2123,
      "step": 2270
    },
    {
      "epoch": 0.65,
      "grad_norm": 7.464560031890869,
      "learning_rate": 2.025600686498856e-05,
      "loss": 1.0035,
      "step": 2271
    },
    {
      "epoch": 0.65,
      "grad_norm": 10.008631706237793,
      "learning_rate": 2.025171624713959e-05,
      "loss": 1.3727,
      "step": 2272
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.217758178710938,
      "learning_rate": 2.024742562929062e-05,
      "loss": 0.9811,
      "step": 2273
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.442007064819336,
      "learning_rate": 2.024313501144165e-05,
      "loss": 1.1444,
      "step": 2274
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.960583686828613,
      "learning_rate": 2.0238844393592677e-05,
      "loss": 1.1571,
      "step": 2275
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.48130989074707,
      "learning_rate": 2.023455377574371e-05,
      "loss": 1.0325,
      "step": 2276
    },
    {
      "epoch": 0.65,
      "grad_norm": 7.487373352050781,
      "learning_rate": 2.0230263157894735e-05,
      "loss": 0.8956,
      "step": 2277
    },
    {
      "epoch": 0.65,
      "grad_norm": 7.44172477722168,
      "learning_rate": 2.0225972540045767e-05,
      "loss": 0.9372,
      "step": 2278
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.773127555847168,
      "learning_rate": 2.0221681922196796e-05,
      "loss": 1.227,
      "step": 2279
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.245462417602539,
      "learning_rate": 2.0217391304347828e-05,
      "loss": 1.1996,
      "step": 2280
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.477191925048828,
      "learning_rate": 2.0213100686498856e-05,
      "loss": 0.8538,
      "step": 2281
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.498077392578125,
      "learning_rate": 2.0208810068649885e-05,
      "loss": 0.8543,
      "step": 2282
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.986138343811035,
      "learning_rate": 2.0204519450800917e-05,
      "loss": 1.061,
      "step": 2283
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.769342422485352,
      "learning_rate": 2.0200228832951946e-05,
      "loss": 1.0352,
      "step": 2284
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.644377708435059,
      "learning_rate": 2.0195938215102975e-05,
      "loss": 1.1557,
      "step": 2285
    },
    {
      "epoch": 0.65,
      "grad_norm": 11.438213348388672,
      "learning_rate": 2.0191647597254004e-05,
      "loss": 1.0977,
      "step": 2286
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.189215660095215,
      "learning_rate": 2.0187356979405036e-05,
      "loss": 0.9457,
      "step": 2287
    },
    {
      "epoch": 0.65,
      "grad_norm": 6.1353864669799805,
      "learning_rate": 2.0183066361556065e-05,
      "loss": 0.4329,
      "step": 2288
    },
    {
      "epoch": 0.65,
      "grad_norm": 11.20683765411377,
      "learning_rate": 2.0178775743707093e-05,
      "loss": 1.3686,
      "step": 2289
    },
    {
      "epoch": 0.66,
      "grad_norm": 10.71774673461914,
      "learning_rate": 2.0174485125858126e-05,
      "loss": 1.1133,
      "step": 2290
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.056065559387207,
      "learning_rate": 2.0170194508009154e-05,
      "loss": 0.834,
      "step": 2291
    },
    {
      "epoch": 0.66,
      "grad_norm": 11.788326263427734,
      "learning_rate": 2.0165903890160183e-05,
      "loss": 1.2508,
      "step": 2292
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.932804107666016,
      "learning_rate": 2.0161613272311212e-05,
      "loss": 1.5201,
      "step": 2293
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.786749839782715,
      "learning_rate": 2.0157322654462244e-05,
      "loss": 1.0267,
      "step": 2294
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.666580200195312,
      "learning_rate": 2.0153032036613273e-05,
      "loss": 1.1708,
      "step": 2295
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.67428970336914,
      "learning_rate": 2.01487414187643e-05,
      "loss": 0.9788,
      "step": 2296
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.6431303024292,
      "learning_rate": 2.0144450800915334e-05,
      "loss": 1.0471,
      "step": 2297
    },
    {
      "epoch": 0.66,
      "grad_norm": 7.736600399017334,
      "learning_rate": 2.0140160183066362e-05,
      "loss": 0.8203,
      "step": 2298
    },
    {
      "epoch": 0.66,
      "grad_norm": 10.18701457977295,
      "learning_rate": 2.0135869565217395e-05,
      "loss": 1.3264,
      "step": 2299
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.586442947387695,
      "learning_rate": 2.013157894736842e-05,
      "loss": 1.2081,
      "step": 2300
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.979724884033203,
      "learning_rate": 2.0127288329519452e-05,
      "loss": 1.2535,
      "step": 2301
    },
    {
      "epoch": 0.66,
      "grad_norm": 10.856149673461914,
      "learning_rate": 2.012299771167048e-05,
      "loss": 1.1634,
      "step": 2302
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.824821472167969,
      "learning_rate": 2.011870709382151e-05,
      "loss": 1.2392,
      "step": 2303
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.485241889953613,
      "learning_rate": 2.011441647597254e-05,
      "loss": 0.8693,
      "step": 2304
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.472940444946289,
      "learning_rate": 2.011012585812357e-05,
      "loss": 0.999,
      "step": 2305
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.946276664733887,
      "learning_rate": 2.0105835240274603e-05,
      "loss": 1.2392,
      "step": 2306
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.315030097961426,
      "learning_rate": 2.0101544622425628e-05,
      "loss": 0.996,
      "step": 2307
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.195350646972656,
      "learning_rate": 2.009725400457666e-05,
      "loss": 1.2306,
      "step": 2308
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.057229042053223,
      "learning_rate": 2.009296338672769e-05,
      "loss": 1.4311,
      "step": 2309
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.28529167175293,
      "learning_rate": 2.008867276887872e-05,
      "loss": 0.9281,
      "step": 2310
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.697407722473145,
      "learning_rate": 2.0084382151029747e-05,
      "loss": 1.0057,
      "step": 2311
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.405632972717285,
      "learning_rate": 2.008009153318078e-05,
      "loss": 1.4378,
      "step": 2312
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.110221862792969,
      "learning_rate": 2.0075800915331808e-05,
      "loss": 1.1495,
      "step": 2313
    },
    {
      "epoch": 0.66,
      "grad_norm": 7.584765434265137,
      "learning_rate": 2.0071510297482836e-05,
      "loss": 0.9162,
      "step": 2314
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.562564849853516,
      "learning_rate": 2.006721967963387e-05,
      "loss": 0.9467,
      "step": 2315
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.545437812805176,
      "learning_rate": 2.0062929061784897e-05,
      "loss": 1.062,
      "step": 2316
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.241497993469238,
      "learning_rate": 2.005863844393593e-05,
      "loss": 1.2093,
      "step": 2317
    },
    {
      "epoch": 0.66,
      "grad_norm": 10.11286449432373,
      "learning_rate": 2.0054347826086955e-05,
      "loss": 1.2354,
      "step": 2318
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.947376251220703,
      "learning_rate": 2.0050057208237987e-05,
      "loss": 1.1555,
      "step": 2319
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.167878150939941,
      "learning_rate": 2.0045766590389016e-05,
      "loss": 0.8684,
      "step": 2320
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.431657791137695,
      "learning_rate": 2.0041475972540048e-05,
      "loss": 1.1477,
      "step": 2321
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.346592903137207,
      "learning_rate": 2.0037185354691073e-05,
      "loss": 1.0756,
      "step": 2322
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.329963684082031,
      "learning_rate": 2.0032894736842105e-05,
      "loss": 0.8836,
      "step": 2323
    },
    {
      "epoch": 0.66,
      "grad_norm": 10.168521881103516,
      "learning_rate": 2.0028604118993138e-05,
      "loss": 1.2817,
      "step": 2324
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.805412292480469,
      "learning_rate": 2.0024313501144166e-05,
      "loss": 0.9593,
      "step": 2325
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.176719665527344,
      "learning_rate": 2.0020022883295195e-05,
      "loss": 0.9256,
      "step": 2326
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.037216186523438,
      "learning_rate": 2.0015732265446224e-05,
      "loss": 1.1016,
      "step": 2327
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.835389137268066,
      "learning_rate": 2.0011441647597256e-05,
      "loss": 1.1988,
      "step": 2328
    },
    {
      "epoch": 0.67,
      "grad_norm": 7.627316951751709,
      "learning_rate": 2.000715102974828e-05,
      "loss": 0.8773,
      "step": 2329
    },
    {
      "epoch": 0.67,
      "grad_norm": 11.330472946166992,
      "learning_rate": 2.0002860411899314e-05,
      "loss": 1.3436,
      "step": 2330
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.78010082244873,
      "learning_rate": 1.9998569794050346e-05,
      "loss": 0.9646,
      "step": 2331
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.319540023803711,
      "learning_rate": 1.9994279176201374e-05,
      "loss": 1.1373,
      "step": 2332
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.710355758666992,
      "learning_rate": 1.9989988558352403e-05,
      "loss": 1.2371,
      "step": 2333
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.212077140808105,
      "learning_rate": 1.9985697940503432e-05,
      "loss": 0.9714,
      "step": 2334
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.847437858581543,
      "learning_rate": 1.9981407322654464e-05,
      "loss": 1.0738,
      "step": 2335
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.038082122802734,
      "learning_rate": 1.9977116704805493e-05,
      "loss": 1.1797,
      "step": 2336
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.931341171264648,
      "learning_rate": 1.997282608695652e-05,
      "loss": 1.0681,
      "step": 2337
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.347209930419922,
      "learning_rate": 1.996853546910755e-05,
      "loss": 1.0247,
      "step": 2338
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.950143814086914,
      "learning_rate": 1.9964244851258583e-05,
      "loss": 1.0472,
      "step": 2339
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.463066101074219,
      "learning_rate": 1.995995423340961e-05,
      "loss": 1.0845,
      "step": 2340
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.918933868408203,
      "learning_rate": 1.995566361556064e-05,
      "loss": 1.355,
      "step": 2341
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.00894546508789,
      "learning_rate": 1.9951372997711672e-05,
      "loss": 0.8562,
      "step": 2342
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.099607467651367,
      "learning_rate": 1.99470823798627e-05,
      "loss": 0.9076,
      "step": 2343
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.967351913452148,
      "learning_rate": 1.994279176201373e-05,
      "loss": 0.95,
      "step": 2344
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.86395263671875,
      "learning_rate": 1.993850114416476e-05,
      "loss": 1.2044,
      "step": 2345
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.624259948730469,
      "learning_rate": 1.993421052631579e-05,
      "loss": 1.4937,
      "step": 2346
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.641493797302246,
      "learning_rate": 1.992991990846682e-05,
      "loss": 1.1664,
      "step": 2347
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.341193199157715,
      "learning_rate": 1.9925629290617848e-05,
      "loss": 1.0786,
      "step": 2348
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.933123588562012,
      "learning_rate": 1.992133867276888e-05,
      "loss": 1.1971,
      "step": 2349
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.406205177307129,
      "learning_rate": 1.991704805491991e-05,
      "loss": 1.466,
      "step": 2350
    },
    {
      "epoch": 0.67,
      "grad_norm": 7.680570125579834,
      "learning_rate": 1.991275743707094e-05,
      "loss": 0.8666,
      "step": 2351
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.177651405334473,
      "learning_rate": 1.9908466819221967e-05,
      "loss": 1.0843,
      "step": 2352
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.994171142578125,
      "learning_rate": 1.9904176201373e-05,
      "loss": 1.3342,
      "step": 2353
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.462268829345703,
      "learning_rate": 1.9899885583524028e-05,
      "loss": 1.1493,
      "step": 2354
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.827287673950195,
      "learning_rate": 1.9895594965675056e-05,
      "loss": 1.3075,
      "step": 2355
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.858643531799316,
      "learning_rate": 1.9891304347826085e-05,
      "loss": 1.3137,
      "step": 2356
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.442103385925293,
      "learning_rate": 1.9887013729977117e-05,
      "loss": 1.0156,
      "step": 2357
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.429743766784668,
      "learning_rate": 1.988272311212815e-05,
      "loss": 1.12,
      "step": 2358
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.051727294921875,
      "learning_rate": 1.9878432494279175e-05,
      "loss": 1.1413,
      "step": 2359
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.427308082580566,
      "learning_rate": 1.9874141876430207e-05,
      "loss": 1.2045,
      "step": 2360
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.644721984863281,
      "learning_rate": 1.9869851258581236e-05,
      "loss": 1.4366,
      "step": 2361
    },
    {
      "epoch": 0.68,
      "grad_norm": 7.993911266326904,
      "learning_rate": 1.9865560640732268e-05,
      "loss": 1.0485,
      "step": 2362
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.51898193359375,
      "learning_rate": 1.9861270022883293e-05,
      "loss": 1.2335,
      "step": 2363
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.08745002746582,
      "learning_rate": 1.9856979405034326e-05,
      "loss": 1.0672,
      "step": 2364
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.468375205993652,
      "learning_rate": 1.9852688787185358e-05,
      "loss": 1.0735,
      "step": 2365
    },
    {
      "epoch": 0.68,
      "grad_norm": 7.964073657989502,
      "learning_rate": 1.9848398169336383e-05,
      "loss": 1.038,
      "step": 2366
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.643996238708496,
      "learning_rate": 1.9844107551487415e-05,
      "loss": 1.1209,
      "step": 2367
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.548465728759766,
      "learning_rate": 1.9839816933638444e-05,
      "loss": 1.2388,
      "step": 2368
    },
    {
      "epoch": 0.68,
      "grad_norm": 10.564348220825195,
      "learning_rate": 1.9835526315789476e-05,
      "loss": 1.4411,
      "step": 2369
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.36286735534668,
      "learning_rate": 1.98312356979405e-05,
      "loss": 1.1672,
      "step": 2370
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.827733039855957,
      "learning_rate": 1.9826945080091534e-05,
      "loss": 1.4856,
      "step": 2371
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.963473320007324,
      "learning_rate": 1.9822654462242562e-05,
      "loss": 0.9815,
      "step": 2372
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.298812866210938,
      "learning_rate": 1.9818363844393595e-05,
      "loss": 1.0033,
      "step": 2373
    },
    {
      "epoch": 0.68,
      "grad_norm": 7.578042984008789,
      "learning_rate": 1.9814073226544623e-05,
      "loss": 1.0352,
      "step": 2374
    },
    {
      "epoch": 0.68,
      "grad_norm": 10.863141059875488,
      "learning_rate": 1.9809782608695652e-05,
      "loss": 1.0734,
      "step": 2375
    },
    {
      "epoch": 0.68,
      "grad_norm": 10.756592750549316,
      "learning_rate": 1.9805491990846684e-05,
      "loss": 1.3603,
      "step": 2376
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.858885765075684,
      "learning_rate": 1.9801201372997713e-05,
      "loss": 0.6397,
      "step": 2377
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.354283332824707,
      "learning_rate": 1.9796910755148742e-05,
      "loss": 1.2313,
      "step": 2378
    },
    {
      "epoch": 0.68,
      "grad_norm": 10.650423049926758,
      "learning_rate": 1.979262013729977e-05,
      "loss": 1.2583,
      "step": 2379
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.179424285888672,
      "learning_rate": 1.9788329519450803e-05,
      "loss": 1.1419,
      "step": 2380
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.642839431762695,
      "learning_rate": 1.9784038901601828e-05,
      "loss": 1.2686,
      "step": 2381
    },
    {
      "epoch": 0.68,
      "grad_norm": 7.6793646812438965,
      "learning_rate": 1.977974828375286e-05,
      "loss": 1.0472,
      "step": 2382
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.839323043823242,
      "learning_rate": 1.9775457665903892e-05,
      "loss": 1.3391,
      "step": 2383
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.746617317199707,
      "learning_rate": 1.977116704805492e-05,
      "loss": 1.2077,
      "step": 2384
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.517253875732422,
      "learning_rate": 1.976687643020595e-05,
      "loss": 0.8728,
      "step": 2385
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.364797592163086,
      "learning_rate": 1.976258581235698e-05,
      "loss": 1.2451,
      "step": 2386
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.547454833984375,
      "learning_rate": 1.975829519450801e-05,
      "loss": 1.151,
      "step": 2387
    },
    {
      "epoch": 0.68,
      "grad_norm": 7.6331095695495605,
      "learning_rate": 1.975400457665904e-05,
      "loss": 0.9652,
      "step": 2388
    },
    {
      "epoch": 0.68,
      "grad_norm": 11.37586784362793,
      "learning_rate": 1.974971395881007e-05,
      "loss": 1.3211,
      "step": 2389
    },
    {
      "epoch": 0.68,
      "grad_norm": 11.770246505737305,
      "learning_rate": 1.9745423340961097e-05,
      "loss": 1.4752,
      "step": 2390
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.619356155395508,
      "learning_rate": 1.974113272311213e-05,
      "loss": 1.4053,
      "step": 2391
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.411015510559082,
      "learning_rate": 1.9736842105263158e-05,
      "loss": 1.1664,
      "step": 2392
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.58008098602295,
      "learning_rate": 1.9732551487414187e-05,
      "loss": 1.1125,
      "step": 2393
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.968585014343262,
      "learning_rate": 1.972826086956522e-05,
      "loss": 1.1154,
      "step": 2394
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.26370620727539,
      "learning_rate": 1.9723970251716248e-05,
      "loss": 1.1691,
      "step": 2395
    },
    {
      "epoch": 0.69,
      "grad_norm": 7.678633689880371,
      "learning_rate": 1.9719679633867277e-05,
      "loss": 0.8724,
      "step": 2396
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.350838661193848,
      "learning_rate": 1.9715389016018305e-05,
      "loss": 0.9113,
      "step": 2397
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.200567245483398,
      "learning_rate": 1.9711098398169338e-05,
      "loss": 0.8809,
      "step": 2398
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.043484687805176,
      "learning_rate": 1.9706807780320366e-05,
      "loss": 1.0817,
      "step": 2399
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.50027084350586,
      "learning_rate": 1.9702517162471395e-05,
      "loss": 0.8939,
      "step": 2400
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.82008171081543,
      "learning_rate": 1.9698226544622427e-05,
      "loss": 1.2579,
      "step": 2401
    },
    {
      "epoch": 0.69,
      "grad_norm": 10.053723335266113,
      "learning_rate": 1.9693935926773456e-05,
      "loss": 1.164,
      "step": 2402
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.12060260772705,
      "learning_rate": 1.9689645308924488e-05,
      "loss": 0.9384,
      "step": 2403
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.901167869567871,
      "learning_rate": 1.9685354691075514e-05,
      "loss": 1.109,
      "step": 2404
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.169709205627441,
      "learning_rate": 1.9681064073226546e-05,
      "loss": 1.381,
      "step": 2405
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.790369033813477,
      "learning_rate": 1.9676773455377574e-05,
      "loss": 1.131,
      "step": 2406
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.024384498596191,
      "learning_rate": 1.9672482837528603e-05,
      "loss": 1.0487,
      "step": 2407
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.600336074829102,
      "learning_rate": 1.9668192219679635e-05,
      "loss": 1.0277,
      "step": 2408
    },
    {
      "epoch": 0.69,
      "grad_norm": 11.553108215332031,
      "learning_rate": 1.9663901601830664e-05,
      "loss": 1.0997,
      "step": 2409
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.540474891662598,
      "learning_rate": 1.9659610983981696e-05,
      "loss": 1.004,
      "step": 2410
    },
    {
      "epoch": 0.69,
      "grad_norm": 10.915793418884277,
      "learning_rate": 1.965532036613272e-05,
      "loss": 1.3838,
      "step": 2411
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.809085845947266,
      "learning_rate": 1.9651029748283754e-05,
      "loss": 1.1397,
      "step": 2412
    },
    {
      "epoch": 0.69,
      "grad_norm": 10.364714622497559,
      "learning_rate": 1.9646739130434783e-05,
      "loss": 1.2425,
      "step": 2413
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.270904541015625,
      "learning_rate": 1.9642448512585815e-05,
      "loss": 1.0459,
      "step": 2414
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.43685245513916,
      "learning_rate": 1.963815789473684e-05,
      "loss": 1.1709,
      "step": 2415
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.97892951965332,
      "learning_rate": 1.9633867276887872e-05,
      "loss": 0.9688,
      "step": 2416
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.369361877441406,
      "learning_rate": 1.9629576659038904e-05,
      "loss": 1.4883,
      "step": 2417
    },
    {
      "epoch": 0.69,
      "grad_norm": 10.247814178466797,
      "learning_rate": 1.9625286041189933e-05,
      "loss": 1.2236,
      "step": 2418
    },
    {
      "epoch": 0.69,
      "grad_norm": 10.082172393798828,
      "learning_rate": 1.9620995423340962e-05,
      "loss": 1.2265,
      "step": 2419
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.495353698730469,
      "learning_rate": 1.961670480549199e-05,
      "loss": 1.4157,
      "step": 2420
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.695307731628418,
      "learning_rate": 1.9612414187643023e-05,
      "loss": 1.0561,
      "step": 2421
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.251786231994629,
      "learning_rate": 1.9608123569794048e-05,
      "loss": 1.1198,
      "step": 2422
    },
    {
      "epoch": 0.69,
      "grad_norm": 7.915933132171631,
      "learning_rate": 1.960383295194508e-05,
      "loss": 1.0593,
      "step": 2423
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.137659072875977,
      "learning_rate": 1.959954233409611e-05,
      "loss": 1.0115,
      "step": 2424
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.613964080810547,
      "learning_rate": 1.959525171624714e-05,
      "loss": 1.2495,
      "step": 2425
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.045031547546387,
      "learning_rate": 1.959096109839817e-05,
      "loss": 1.1832,
      "step": 2426
    },
    {
      "epoch": 0.69,
      "grad_norm": 7.762406349182129,
      "learning_rate": 1.95866704805492e-05,
      "loss": 1.0719,
      "step": 2427
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.389142990112305,
      "learning_rate": 1.958237986270023e-05,
      "loss": 1.2699,
      "step": 2428
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.382674217224121,
      "learning_rate": 1.957808924485126e-05,
      "loss": 1.2774,
      "step": 2429
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.801231384277344,
      "learning_rate": 1.957379862700229e-05,
      "loss": 1.0567,
      "step": 2430
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.806095123291016,
      "learning_rate": 1.9569508009153317e-05,
      "loss": 1.3174,
      "step": 2431
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.43411922454834,
      "learning_rate": 1.956521739130435e-05,
      "loss": 0.9739,
      "step": 2432
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.108588218688965,
      "learning_rate": 1.9560926773455375e-05,
      "loss": 0.9199,
      "step": 2433
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.779991149902344,
      "learning_rate": 1.9556636155606407e-05,
      "loss": 0.967,
      "step": 2434
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.319437026977539,
      "learning_rate": 1.955234553775744e-05,
      "loss": 1.3314,
      "step": 2435
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.56344223022461,
      "learning_rate": 1.9548054919908468e-05,
      "loss": 1.0371,
      "step": 2436
    },
    {
      "epoch": 0.7,
      "grad_norm": 10.121199607849121,
      "learning_rate": 1.9543764302059497e-05,
      "loss": 1.1355,
      "step": 2437
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.621366500854492,
      "learning_rate": 1.9539473684210525e-05,
      "loss": 1.1594,
      "step": 2438
    },
    {
      "epoch": 0.7,
      "grad_norm": 10.054593086242676,
      "learning_rate": 1.9535183066361558e-05,
      "loss": 1.084,
      "step": 2439
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.424495697021484,
      "learning_rate": 1.9530892448512586e-05,
      "loss": 1.0958,
      "step": 2440
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.96681022644043,
      "learning_rate": 1.9526601830663615e-05,
      "loss": 1.1376,
      "step": 2441
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.197810173034668,
      "learning_rate": 1.9522311212814644e-05,
      "loss": 1.1704,
      "step": 2442
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.515543937683105,
      "learning_rate": 1.9518020594965676e-05,
      "loss": 0.9871,
      "step": 2443
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.73293399810791,
      "learning_rate": 1.9513729977116708e-05,
      "loss": 1.22,
      "step": 2444
    },
    {
      "epoch": 0.7,
      "grad_norm": 7.523548603057861,
      "learning_rate": 1.9509439359267734e-05,
      "loss": 0.8579,
      "step": 2445
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.210893630981445,
      "learning_rate": 1.9505148741418766e-05,
      "loss": 1.0666,
      "step": 2446
    },
    {
      "epoch": 0.7,
      "grad_norm": 7.820652008056641,
      "learning_rate": 1.9500858123569795e-05,
      "loss": 1.2288,
      "step": 2447
    },
    {
      "epoch": 0.7,
      "grad_norm": 7.922313213348389,
      "learning_rate": 1.9496567505720823e-05,
      "loss": 1.1147,
      "step": 2448
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.569446563720703,
      "learning_rate": 1.9492276887871852e-05,
      "loss": 1.1903,
      "step": 2449
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.225363731384277,
      "learning_rate": 1.9487986270022884e-05,
      "loss": 1.2157,
      "step": 2450
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.372381210327148,
      "learning_rate": 1.9483695652173916e-05,
      "loss": 1.2702,
      "step": 2451
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.420974731445312,
      "learning_rate": 1.9479405034324942e-05,
      "loss": 1.4587,
      "step": 2452
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.934778213500977,
      "learning_rate": 1.9475114416475974e-05,
      "loss": 1.4302,
      "step": 2453
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.482074737548828,
      "learning_rate": 1.9470823798627003e-05,
      "loss": 1.1006,
      "step": 2454
    },
    {
      "epoch": 0.7,
      "grad_norm": 7.829884052276611,
      "learning_rate": 1.9466533180778035e-05,
      "loss": 1.1241,
      "step": 2455
    },
    {
      "epoch": 0.7,
      "grad_norm": 10.12347412109375,
      "learning_rate": 1.946224256292906e-05,
      "loss": 0.8159,
      "step": 2456
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.303497314453125,
      "learning_rate": 1.9457951945080092e-05,
      "loss": 1.3549,
      "step": 2457
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.074723243713379,
      "learning_rate": 1.945366132723112e-05,
      "loss": 0.9796,
      "step": 2458
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.593305587768555,
      "learning_rate": 1.944937070938215e-05,
      "loss": 1.0156,
      "step": 2459
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.930562019348145,
      "learning_rate": 1.9445080091533182e-05,
      "loss": 0.9765,
      "step": 2460
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.386600494384766,
      "learning_rate": 1.944078947368421e-05,
      "loss": 1.1518,
      "step": 2461
    },
    {
      "epoch": 0.7,
      "grad_norm": 10.876976013183594,
      "learning_rate": 1.9436498855835243e-05,
      "loss": 1.2208,
      "step": 2462
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.179274559020996,
      "learning_rate": 1.943220823798627e-05,
      "loss": 0.8473,
      "step": 2463
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.907797813415527,
      "learning_rate": 1.94279176201373e-05,
      "loss": 1.1434,
      "step": 2464
    },
    {
      "epoch": 0.71,
      "grad_norm": 12.280512809753418,
      "learning_rate": 1.942362700228833e-05,
      "loss": 1.4368,
      "step": 2465
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.636651992797852,
      "learning_rate": 1.941933638443936e-05,
      "loss": 0.9767,
      "step": 2466
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.658784866333008,
      "learning_rate": 1.9415045766590387e-05,
      "loss": 0.9644,
      "step": 2467
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.097394943237305,
      "learning_rate": 1.941075514874142e-05,
      "loss": 0.9025,
      "step": 2468
    },
    {
      "epoch": 0.71,
      "grad_norm": 11.89475154876709,
      "learning_rate": 1.940646453089245e-05,
      "loss": 1.2886,
      "step": 2469
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.268407821655273,
      "learning_rate": 1.940217391304348e-05,
      "loss": 0.9651,
      "step": 2470
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.751604080200195,
      "learning_rate": 1.939788329519451e-05,
      "loss": 1.0763,
      "step": 2471
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.457395553588867,
      "learning_rate": 1.9393592677345537e-05,
      "loss": 0.9043,
      "step": 2472
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.714011192321777,
      "learning_rate": 1.938930205949657e-05,
      "loss": 0.9692,
      "step": 2473
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.668814659118652,
      "learning_rate": 1.9385011441647595e-05,
      "loss": 1.0284,
      "step": 2474
    },
    {
      "epoch": 0.71,
      "grad_norm": 13.269686698913574,
      "learning_rate": 1.9380720823798627e-05,
      "loss": 1.5412,
      "step": 2475
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.09195327758789,
      "learning_rate": 1.9376430205949656e-05,
      "loss": 0.8828,
      "step": 2476
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.99887752532959,
      "learning_rate": 1.9372139588100688e-05,
      "loss": 1.1266,
      "step": 2477
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.050641059875488,
      "learning_rate": 1.9367848970251717e-05,
      "loss": 0.8672,
      "step": 2478
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.21500301361084,
      "learning_rate": 1.9363558352402746e-05,
      "loss": 1.0729,
      "step": 2479
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.102481842041016,
      "learning_rate": 1.9359267734553778e-05,
      "loss": 1.1911,
      "step": 2480
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.278970718383789,
      "learning_rate": 1.9354977116704807e-05,
      "loss": 0.7449,
      "step": 2481
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.626091003417969,
      "learning_rate": 1.9350686498855835e-05,
      "loss": 1.1365,
      "step": 2482
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.949851989746094,
      "learning_rate": 1.9346395881006864e-05,
      "loss": 1.0653,
      "step": 2483
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.16932201385498,
      "learning_rate": 1.9342105263157896e-05,
      "loss": 0.9872,
      "step": 2484
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.62385368347168,
      "learning_rate": 1.9337814645308925e-05,
      "loss": 1.0783,
      "step": 2485
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.550857543945312,
      "learning_rate": 1.9333524027459954e-05,
      "loss": 1.2684,
      "step": 2486
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.422028541564941,
      "learning_rate": 1.9329233409610986e-05,
      "loss": 0.9019,
      "step": 2487
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.581709861755371,
      "learning_rate": 1.9324942791762015e-05,
      "loss": 1.0272,
      "step": 2488
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.081588745117188,
      "learning_rate": 1.9320652173913043e-05,
      "loss": 1.0151,
      "step": 2489
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.836494445800781,
      "learning_rate": 1.9316361556064072e-05,
      "loss": 0.8711,
      "step": 2490
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.66437816619873,
      "learning_rate": 1.9312070938215104e-05,
      "loss": 0.9395,
      "step": 2491
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.79299259185791,
      "learning_rate": 1.9307780320366133e-05,
      "loss": 0.7948,
      "step": 2492
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.855372428894043,
      "learning_rate": 1.9303489702517162e-05,
      "loss": 1.1007,
      "step": 2493
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.21322250366211,
      "learning_rate": 1.9299199084668194e-05,
      "loss": 0.8631,
      "step": 2494
    },
    {
      "epoch": 0.71,
      "grad_norm": 10.277508735656738,
      "learning_rate": 1.9294908466819223e-05,
      "loss": 1.1876,
      "step": 2495
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.104170799255371,
      "learning_rate": 1.9290617848970255e-05,
      "loss": 1.0572,
      "step": 2496
    },
    {
      "epoch": 0.71,
      "grad_norm": 13.10643196105957,
      "learning_rate": 1.928632723112128e-05,
      "loss": 1.5775,
      "step": 2497
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.110145568847656,
      "learning_rate": 1.9282036613272313e-05,
      "loss": 0.9785,
      "step": 2498
    },
    {
      "epoch": 0.71,
      "grad_norm": 7.77860164642334,
      "learning_rate": 1.927774599542334e-05,
      "loss": 0.939,
      "step": 2499
    },
    {
      "epoch": 0.72,
      "grad_norm": 10.14551830291748,
      "learning_rate": 1.927345537757437e-05,
      "loss": 1.3206,
      "step": 2500
    },
    {
      "epoch": 0.72,
      "grad_norm": 12.676076889038086,
      "learning_rate": 1.92691647597254e-05,
      "loss": 1.3357,
      "step": 2501
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.927719116210938,
      "learning_rate": 1.926487414187643e-05,
      "loss": 1.3232,
      "step": 2502
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.037900924682617,
      "learning_rate": 1.9260583524027463e-05,
      "loss": 1.0347,
      "step": 2503
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.319453239440918,
      "learning_rate": 1.925629290617849e-05,
      "loss": 1.0187,
      "step": 2504
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.605573654174805,
      "learning_rate": 1.925200228832952e-05,
      "loss": 1.2331,
      "step": 2505
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.671036720275879,
      "learning_rate": 1.924771167048055e-05,
      "loss": 1.2035,
      "step": 2506
    },
    {
      "epoch": 0.72,
      "grad_norm": 7.830979824066162,
      "learning_rate": 1.924342105263158e-05,
      "loss": 0.9673,
      "step": 2507
    },
    {
      "epoch": 0.72,
      "grad_norm": 10.420891761779785,
      "learning_rate": 1.9239130434782607e-05,
      "loss": 1.2931,
      "step": 2508
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.493072509765625,
      "learning_rate": 1.923483981693364e-05,
      "loss": 1.0137,
      "step": 2509
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.534939765930176,
      "learning_rate": 1.9230549199084668e-05,
      "loss": 0.9068,
      "step": 2510
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.904736518859863,
      "learning_rate": 1.9226258581235697e-05,
      "loss": 1.1091,
      "step": 2511
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.013572692871094,
      "learning_rate": 1.922196796338673e-05,
      "loss": 0.912,
      "step": 2512
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.072867393493652,
      "learning_rate": 1.9217677345537758e-05,
      "loss": 1.2602,
      "step": 2513
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.154093742370605,
      "learning_rate": 1.921338672768879e-05,
      "loss": 1.0201,
      "step": 2514
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.744938850402832,
      "learning_rate": 1.9209096109839815e-05,
      "loss": 1.1458,
      "step": 2515
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.348825454711914,
      "learning_rate": 1.9204805491990847e-05,
      "loss": 1.0297,
      "step": 2516
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.790005683898926,
      "learning_rate": 1.9200514874141876e-05,
      "loss": 1.063,
      "step": 2517
    },
    {
      "epoch": 0.72,
      "grad_norm": 11.721095085144043,
      "learning_rate": 1.9196224256292908e-05,
      "loss": 1.13,
      "step": 2518
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.300861358642578,
      "learning_rate": 1.9191933638443934e-05,
      "loss": 0.8934,
      "step": 2519
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.523219108581543,
      "learning_rate": 1.9187643020594966e-05,
      "loss": 0.9684,
      "step": 2520
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.683642387390137,
      "learning_rate": 1.9183352402745998e-05,
      "loss": 1.0538,
      "step": 2521
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.01683235168457,
      "learning_rate": 1.9179061784897027e-05,
      "loss": 0.9628,
      "step": 2522
    },
    {
      "epoch": 0.72,
      "grad_norm": 7.2914652824401855,
      "learning_rate": 1.9174771167048055e-05,
      "loss": 0.7033,
      "step": 2523
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.02861499786377,
      "learning_rate": 1.9170480549199084e-05,
      "loss": 0.8338,
      "step": 2524
    },
    {
      "epoch": 0.72,
      "grad_norm": 10.126641273498535,
      "learning_rate": 1.9166189931350116e-05,
      "loss": 1.1902,
      "step": 2525
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.803444862365723,
      "learning_rate": 1.9161899313501142e-05,
      "loss": 1.1011,
      "step": 2526
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.839347839355469,
      "learning_rate": 1.9157608695652174e-05,
      "loss": 1.0522,
      "step": 2527
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.325223922729492,
      "learning_rate": 1.9153318077803206e-05,
      "loss": 1.219,
      "step": 2528
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.535804748535156,
      "learning_rate": 1.9149027459954235e-05,
      "loss": 0.9931,
      "step": 2529
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.813097953796387,
      "learning_rate": 1.9144736842105264e-05,
      "loss": 0.8771,
      "step": 2530
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.58659839630127,
      "learning_rate": 1.9140446224256292e-05,
      "loss": 1.0303,
      "step": 2531
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.635456085205078,
      "learning_rate": 1.9136155606407325e-05,
      "loss": 1.0285,
      "step": 2532
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.991578102111816,
      "learning_rate": 1.9131864988558353e-05,
      "loss": 1.0554,
      "step": 2533
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.071062088012695,
      "learning_rate": 1.9127574370709382e-05,
      "loss": 1.0673,
      "step": 2534
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.62812328338623,
      "learning_rate": 1.912328375286041e-05,
      "loss": 1.0919,
      "step": 2535
    },
    {
      "epoch": 0.73,
      "grad_norm": 7.468624114990234,
      "learning_rate": 1.9118993135011443e-05,
      "loss": 0.8146,
      "step": 2536
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.206171035766602,
      "learning_rate": 1.9114702517162472e-05,
      "loss": 1.1295,
      "step": 2537
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.525646209716797,
      "learning_rate": 1.91104118993135e-05,
      "loss": 1.3128,
      "step": 2538
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.496814727783203,
      "learning_rate": 1.9106121281464533e-05,
      "loss": 1.0191,
      "step": 2539
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.425469398498535,
      "learning_rate": 1.910183066361556e-05,
      "loss": 1.268,
      "step": 2540
    },
    {
      "epoch": 0.73,
      "grad_norm": 7.101430416107178,
      "learning_rate": 1.909754004576659e-05,
      "loss": 0.8288,
      "step": 2541
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.26478385925293,
      "learning_rate": 1.909324942791762e-05,
      "loss": 0.9117,
      "step": 2542
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.921600341796875,
      "learning_rate": 1.908895881006865e-05,
      "loss": 1.3147,
      "step": 2543
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.800148963928223,
      "learning_rate": 1.908466819221968e-05,
      "loss": 1.1436,
      "step": 2544
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.311388969421387,
      "learning_rate": 1.908037757437071e-05,
      "loss": 0.9736,
      "step": 2545
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.316076278686523,
      "learning_rate": 1.907608695652174e-05,
      "loss": 0.9719,
      "step": 2546
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.000842094421387,
      "learning_rate": 1.907179633867277e-05,
      "loss": 1.0308,
      "step": 2547
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.87099552154541,
      "learning_rate": 1.9067505720823802e-05,
      "loss": 1.3402,
      "step": 2548
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.852824211120605,
      "learning_rate": 1.9063215102974827e-05,
      "loss": 1.2251,
      "step": 2549
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.490774154663086,
      "learning_rate": 1.905892448512586e-05,
      "loss": 0.7745,
      "step": 2550
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.485783576965332,
      "learning_rate": 1.9054633867276888e-05,
      "loss": 1.4443,
      "step": 2551
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.42505168914795,
      "learning_rate": 1.9050343249427917e-05,
      "loss": 0.9396,
      "step": 2552
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.281153678894043,
      "learning_rate": 1.9046052631578946e-05,
      "loss": 1.1538,
      "step": 2553
    },
    {
      "epoch": 0.73,
      "grad_norm": 7.801650524139404,
      "learning_rate": 1.9041762013729978e-05,
      "loss": 0.842,
      "step": 2554
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.858254432678223,
      "learning_rate": 1.903747139588101e-05,
      "loss": 1.0788,
      "step": 2555
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.263640403747559,
      "learning_rate": 1.9033180778032035e-05,
      "loss": 1.0237,
      "step": 2556
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.844314575195312,
      "learning_rate": 1.9028890160183067e-05,
      "loss": 1.4164,
      "step": 2557
    },
    {
      "epoch": 0.73,
      "grad_norm": 7.440629005432129,
      "learning_rate": 1.9024599542334096e-05,
      "loss": 0.874,
      "step": 2558
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.15101146697998,
      "learning_rate": 1.902030892448513e-05,
      "loss": 1.1855,
      "step": 2559
    },
    {
      "epoch": 0.73,
      "grad_norm": 7.89607572555542,
      "learning_rate": 1.9016018306636154e-05,
      "loss": 0.8347,
      "step": 2560
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.365269660949707,
      "learning_rate": 1.9011727688787186e-05,
      "loss": 1.2891,
      "step": 2561
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.910431861877441,
      "learning_rate": 1.9007437070938215e-05,
      "loss": 1.0829,
      "step": 2562
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.414385795593262,
      "learning_rate": 1.9003146453089243e-05,
      "loss": 1.0729,
      "step": 2563
    },
    {
      "epoch": 0.73,
      "grad_norm": 11.398804664611816,
      "learning_rate": 1.8998855835240276e-05,
      "loss": 1.4494,
      "step": 2564
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.427895545959473,
      "learning_rate": 1.8994565217391304e-05,
      "loss": 0.9459,
      "step": 2565
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.052424430847168,
      "learning_rate": 1.8990274599542337e-05,
      "loss": 1.0575,
      "step": 2566
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.641477584838867,
      "learning_rate": 1.8985983981693362e-05,
      "loss": 1.0919,
      "step": 2567
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.028521537780762,
      "learning_rate": 1.8981693363844394e-05,
      "loss": 1.0165,
      "step": 2568
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.408310890197754,
      "learning_rate": 1.8977402745995423e-05,
      "loss": 1.1676,
      "step": 2569
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.366040229797363,
      "learning_rate": 1.8973112128146455e-05,
      "loss": 1.1194,
      "step": 2570
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.971992492675781,
      "learning_rate": 1.8968821510297484e-05,
      "loss": 1.0539,
      "step": 2571
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.06114387512207,
      "learning_rate": 1.8964530892448513e-05,
      "loss": 1.0163,
      "step": 2572
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.988670349121094,
      "learning_rate": 1.8960240274599545e-05,
      "loss": 1.0297,
      "step": 2573
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.470457077026367,
      "learning_rate": 1.8955949656750573e-05,
      "loss": 0.9485,
      "step": 2574
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.652470588684082,
      "learning_rate": 1.8951659038901602e-05,
      "loss": 1.1597,
      "step": 2575
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.958063125610352,
      "learning_rate": 1.894736842105263e-05,
      "loss": 1.3988,
      "step": 2576
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.248319625854492,
      "learning_rate": 1.8943077803203663e-05,
      "loss": 1.0791,
      "step": 2577
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.106576919555664,
      "learning_rate": 1.893878718535469e-05,
      "loss": 1.1785,
      "step": 2578
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.481927871704102,
      "learning_rate": 1.893449656750572e-05,
      "loss": 1.0148,
      "step": 2579
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.416252136230469,
      "learning_rate": 1.8930205949656753e-05,
      "loss": 1.0448,
      "step": 2580
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.41560173034668,
      "learning_rate": 1.892591533180778e-05,
      "loss": 1.0477,
      "step": 2581
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.394129753112793,
      "learning_rate": 1.892162471395881e-05,
      "loss": 0.9718,
      "step": 2582
    },
    {
      "epoch": 0.74,
      "grad_norm": 7.814067363739014,
      "learning_rate": 1.891733409610984e-05,
      "loss": 0.7597,
      "step": 2583
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.573678970336914,
      "learning_rate": 1.891304347826087e-05,
      "loss": 1.1839,
      "step": 2584
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.720595359802246,
      "learning_rate": 1.89087528604119e-05,
      "loss": 1.3163,
      "step": 2585
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.662771224975586,
      "learning_rate": 1.890446224256293e-05,
      "loss": 0.8793,
      "step": 2586
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.156516075134277,
      "learning_rate": 1.8900171624713958e-05,
      "loss": 1.0649,
      "step": 2587
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.830260276794434,
      "learning_rate": 1.889588100686499e-05,
      "loss": 1.3595,
      "step": 2588
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.842612266540527,
      "learning_rate": 1.889159038901602e-05,
      "loss": 1.3127,
      "step": 2589
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.004518508911133,
      "learning_rate": 1.8887299771167047e-05,
      "loss": 0.791,
      "step": 2590
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.46159839630127,
      "learning_rate": 1.888300915331808e-05,
      "loss": 1.0763,
      "step": 2591
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.404709815979004,
      "learning_rate": 1.8878718535469108e-05,
      "loss": 1.0805,
      "step": 2592
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.421338081359863,
      "learning_rate": 1.8874427917620137e-05,
      "loss": 1.2548,
      "step": 2593
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.607651710510254,
      "learning_rate": 1.8870137299771166e-05,
      "loss": 1.2314,
      "step": 2594
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.027852058410645,
      "learning_rate": 1.8865846681922198e-05,
      "loss": 0.9529,
      "step": 2595
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.327431678771973,
      "learning_rate": 1.8861556064073227e-05,
      "loss": 1.162,
      "step": 2596
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.700788497924805,
      "learning_rate": 1.8857265446224255e-05,
      "loss": 1.2408,
      "step": 2597
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.374039649963379,
      "learning_rate": 1.8852974828375288e-05,
      "loss": 1.115,
      "step": 2598
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.619352340698242,
      "learning_rate": 1.8848684210526316e-05,
      "loss": 1.0719,
      "step": 2599
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.652650833129883,
      "learning_rate": 1.884439359267735e-05,
      "loss": 1.0184,
      "step": 2600
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.914128303527832,
      "learning_rate": 1.8840102974828374e-05,
      "loss": 1.37,
      "step": 2601
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.957221031188965,
      "learning_rate": 1.8835812356979406e-05,
      "loss": 1.1816,
      "step": 2602
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.956828117370605,
      "learning_rate": 1.8831521739130435e-05,
      "loss": 1.1326,
      "step": 2603
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.612574577331543,
      "learning_rate": 1.8827231121281464e-05,
      "loss": 0.9619,
      "step": 2604
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.865391731262207,
      "learning_rate": 1.8822940503432496e-05,
      "loss": 1.0499,
      "step": 2605
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.714326858520508,
      "learning_rate": 1.8818649885583525e-05,
      "loss": 1.0391,
      "step": 2606
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.581421852111816,
      "learning_rate": 1.8814359267734557e-05,
      "loss": 1.0204,
      "step": 2607
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.185111045837402,
      "learning_rate": 1.8810068649885582e-05,
      "loss": 1.0015,
      "step": 2608
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.517500877380371,
      "learning_rate": 1.8805778032036614e-05,
      "loss": 0.9711,
      "step": 2609
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.859618186950684,
      "learning_rate": 1.8801487414187643e-05,
      "loss": 1.0174,
      "step": 2610
    },
    {
      "epoch": 0.75,
      "grad_norm": 7.156263828277588,
      "learning_rate": 1.8797196796338675e-05,
      "loss": 0.8121,
      "step": 2611
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.95913314819336,
      "learning_rate": 1.87929061784897e-05,
      "loss": 1.1522,
      "step": 2612
    },
    {
      "epoch": 0.75,
      "grad_norm": 7.466686248779297,
      "learning_rate": 1.8788615560640733e-05,
      "loss": 0.883,
      "step": 2613
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.566696166992188,
      "learning_rate": 1.8784324942791765e-05,
      "loss": 1.1206,
      "step": 2614
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.486233711242676,
      "learning_rate": 1.8780034324942794e-05,
      "loss": 1.0718,
      "step": 2615
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.664968490600586,
      "learning_rate": 1.8775743707093822e-05,
      "loss": 0.9513,
      "step": 2616
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.688613891601562,
      "learning_rate": 1.877145308924485e-05,
      "loss": 1.0027,
      "step": 2617
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.05102825164795,
      "learning_rate": 1.8767162471395883e-05,
      "loss": 1.1885,
      "step": 2618
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.73543643951416,
      "learning_rate": 1.876287185354691e-05,
      "loss": 1.0659,
      "step": 2619
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.386043548583984,
      "learning_rate": 1.875858123569794e-05,
      "loss": 1.2505,
      "step": 2620
    },
    {
      "epoch": 0.75,
      "grad_norm": 12.936104774475098,
      "learning_rate": 1.875429061784897e-05,
      "loss": 1.1552,
      "step": 2621
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.49905014038086,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 1.1317,
      "step": 2622
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.787158012390137,
      "learning_rate": 1.874570938215103e-05,
      "loss": 0.9795,
      "step": 2623
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.00570297241211,
      "learning_rate": 1.874141876430206e-05,
      "loss": 1.0594,
      "step": 2624
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.770262718200684,
      "learning_rate": 1.873712814645309e-05,
      "loss": 1.1636,
      "step": 2625
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.585314750671387,
      "learning_rate": 1.873283752860412e-05,
      "loss": 1.1635,
      "step": 2626
    },
    {
      "epoch": 0.75,
      "grad_norm": 11.837164878845215,
      "learning_rate": 1.872854691075515e-05,
      "loss": 1.4008,
      "step": 2627
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.564339637756348,
      "learning_rate": 1.8724256292906178e-05,
      "loss": 1.2256,
      "step": 2628
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.10994815826416,
      "learning_rate": 1.871996567505721e-05,
      "loss": 1.3112,
      "step": 2629
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.335712432861328,
      "learning_rate": 1.8715675057208235e-05,
      "loss": 1.1398,
      "step": 2630
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.341137886047363,
      "learning_rate": 1.8711384439359267e-05,
      "loss": 0.8713,
      "step": 2631
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.362038612365723,
      "learning_rate": 1.87070938215103e-05,
      "loss": 1.4944,
      "step": 2632
    },
    {
      "epoch": 0.75,
      "grad_norm": 7.386731147766113,
      "learning_rate": 1.870280320366133e-05,
      "loss": 0.8924,
      "step": 2633
    },
    {
      "epoch": 0.75,
      "grad_norm": 7.392826557159424,
      "learning_rate": 1.8698512585812357e-05,
      "loss": 0.9767,
      "step": 2634
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.432779312133789,
      "learning_rate": 1.8694221967963386e-05,
      "loss": 1.155,
      "step": 2635
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.268701553344727,
      "learning_rate": 1.8689931350114418e-05,
      "loss": 1.2737,
      "step": 2636
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.36255931854248,
      "learning_rate": 1.8685640732265447e-05,
      "loss": 0.8501,
      "step": 2637
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.937508583068848,
      "learning_rate": 1.8681350114416476e-05,
      "loss": 1.2618,
      "step": 2638
    },
    {
      "epoch": 0.75,
      "grad_norm": 6.815885543823242,
      "learning_rate": 1.8677059496567504e-05,
      "loss": 1.0871,
      "step": 2639
    },
    {
      "epoch": 0.76,
      "grad_norm": 8.406804084777832,
      "learning_rate": 1.8672768878718537e-05,
      "loss": 1.1689,
      "step": 2640
    },
    {
      "epoch": 0.76,
      "grad_norm": 7.972434043884277,
      "learning_rate": 1.866847826086957e-05,
      "loss": 0.8999,
      "step": 2641
    },
    {
      "epoch": 0.76,
      "grad_norm": 11.70073127746582,
      "learning_rate": 1.8664187643020594e-05,
      "loss": 1.4695,
      "step": 2642
    },
    {
      "epoch": 0.76,
      "grad_norm": 8.955720901489258,
      "learning_rate": 1.8659897025171626e-05,
      "loss": 1.1373,
      "step": 2643
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.123139381408691,
      "learning_rate": 1.8655606407322655e-05,
      "loss": 1.0841,
      "step": 2644
    },
    {
      "epoch": 0.76,
      "grad_norm": 10.323467254638672,
      "learning_rate": 1.8651315789473684e-05,
      "loss": 1.1008,
      "step": 2645
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.747201919555664,
      "learning_rate": 1.8647025171624713e-05,
      "loss": 1.024,
      "step": 2646
    },
    {
      "epoch": 0.76,
      "grad_norm": 8.707141876220703,
      "learning_rate": 1.8642734553775745e-05,
      "loss": 1.0112,
      "step": 2647
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.492386817932129,
      "learning_rate": 1.8638443935926777e-05,
      "loss": 1.1169,
      "step": 2648
    },
    {
      "epoch": 0.76,
      "grad_norm": 8.571977615356445,
      "learning_rate": 1.8634153318077802e-05,
      "loss": 1.0638,
      "step": 2649
    },
    {
      "epoch": 0.76,
      "grad_norm": 10.807567596435547,
      "learning_rate": 1.8629862700228834e-05,
      "loss": 1.3333,
      "step": 2650
    },
    {
      "epoch": 0.76,
      "grad_norm": 8.413228034973145,
      "learning_rate": 1.8625572082379863e-05,
      "loss": 1.0567,
      "step": 2651
    },
    {
      "epoch": 0.76,
      "grad_norm": 12.146622657775879,
      "learning_rate": 1.8621281464530895e-05,
      "loss": 1.1632,
      "step": 2652
    },
    {
      "epoch": 0.76,
      "grad_norm": 8.922531127929688,
      "learning_rate": 1.861699084668192e-05,
      "loss": 0.8596,
      "step": 2653
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.631738662719727,
      "learning_rate": 1.8612700228832953e-05,
      "loss": 1.17,
      "step": 2654
    },
    {
      "epoch": 0.76,
      "grad_norm": 11.031031608581543,
      "learning_rate": 1.860840961098398e-05,
      "loss": 1.0493,
      "step": 2655
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.348603248596191,
      "learning_rate": 1.860411899313501e-05,
      "loss": 1.1414,
      "step": 2656
    },
    {
      "epoch": 0.76,
      "grad_norm": 10.053489685058594,
      "learning_rate": 1.8599828375286043e-05,
      "loss": 0.9839,
      "step": 2657
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.80137825012207,
      "learning_rate": 1.859553775743707e-05,
      "loss": 1.2351,
      "step": 2658
    },
    {
      "epoch": 0.76,
      "grad_norm": 10.180928230285645,
      "learning_rate": 1.8591247139588103e-05,
      "loss": 0.9798,
      "step": 2659
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.522939682006836,
      "learning_rate": 1.858695652173913e-05,
      "loss": 1.1515,
      "step": 2660
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.786197662353516,
      "learning_rate": 1.858266590389016e-05,
      "loss": 1.0778,
      "step": 2661
    },
    {
      "epoch": 0.76,
      "grad_norm": 10.08316707611084,
      "learning_rate": 1.857837528604119e-05,
      "loss": 1.0474,
      "step": 2662
    },
    {
      "epoch": 0.76,
      "grad_norm": 8.703034400939941,
      "learning_rate": 1.8574084668192222e-05,
      "loss": 0.894,
      "step": 2663
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.297806739807129,
      "learning_rate": 1.8569794050343247e-05,
      "loss": 1.0639,
      "step": 2664
    },
    {
      "epoch": 0.76,
      "grad_norm": 8.80936336517334,
      "learning_rate": 1.856550343249428e-05,
      "loss": 0.9902,
      "step": 2665
    },
    {
      "epoch": 0.76,
      "grad_norm": 8.111929893493652,
      "learning_rate": 1.856121281464531e-05,
      "loss": 1.1252,
      "step": 2666
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.989709854125977,
      "learning_rate": 1.855692219679634e-05,
      "loss": 1.1853,
      "step": 2667
    },
    {
      "epoch": 0.76,
      "grad_norm": 10.384547233581543,
      "learning_rate": 1.855263157894737e-05,
      "loss": 0.946,
      "step": 2668
    },
    {
      "epoch": 0.76,
      "grad_norm": 7.475389003753662,
      "learning_rate": 1.8548340961098398e-05,
      "loss": 0.704,
      "step": 2669
    },
    {
      "epoch": 0.76,
      "grad_norm": 10.378499031066895,
      "learning_rate": 1.854405034324943e-05,
      "loss": 1.3423,
      "step": 2670
    },
    {
      "epoch": 0.76,
      "grad_norm": 10.349032402038574,
      "learning_rate": 1.8539759725400455e-05,
      "loss": 1.032,
      "step": 2671
    },
    {
      "epoch": 0.76,
      "grad_norm": 8.30756950378418,
      "learning_rate": 1.8535469107551488e-05,
      "loss": 1.0812,
      "step": 2672
    },
    {
      "epoch": 0.76,
      "grad_norm": 11.054231643676758,
      "learning_rate": 1.8531178489702516e-05,
      "loss": 1.12,
      "step": 2673
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.871310234069824,
      "learning_rate": 1.852688787185355e-05,
      "loss": 0.9755,
      "step": 2674
    },
    {
      "epoch": 0.77,
      "grad_norm": 7.793604373931885,
      "learning_rate": 1.8522597254004577e-05,
      "loss": 0.9116,
      "step": 2675
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.23965072631836,
      "learning_rate": 1.8518306636155606e-05,
      "loss": 0.8213,
      "step": 2676
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.072052001953125,
      "learning_rate": 1.8514016018306638e-05,
      "loss": 1.0072,
      "step": 2677
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.18537712097168,
      "learning_rate": 1.8509725400457667e-05,
      "loss": 0.925,
      "step": 2678
    },
    {
      "epoch": 0.77,
      "grad_norm": 13.233726501464844,
      "learning_rate": 1.8505434782608696e-05,
      "loss": 1.1301,
      "step": 2679
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.48590087890625,
      "learning_rate": 1.8501144164759724e-05,
      "loss": 1.0966,
      "step": 2680
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.115750312805176,
      "learning_rate": 1.8496853546910757e-05,
      "loss": 1.0654,
      "step": 2681
    },
    {
      "epoch": 0.77,
      "grad_norm": 8.910179138183594,
      "learning_rate": 1.8492562929061785e-05,
      "loss": 0.8975,
      "step": 2682
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.317663192749023,
      "learning_rate": 1.8488272311212814e-05,
      "loss": 1.1109,
      "step": 2683
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.628530502319336,
      "learning_rate": 1.8483981693363846e-05,
      "loss": 1.0068,
      "step": 2684
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.652451515197754,
      "learning_rate": 1.8479691075514875e-05,
      "loss": 1.124,
      "step": 2685
    },
    {
      "epoch": 0.77,
      "grad_norm": 8.596827507019043,
      "learning_rate": 1.8475400457665904e-05,
      "loss": 0.9857,
      "step": 2686
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.025288581848145,
      "learning_rate": 1.8471109839816933e-05,
      "loss": 1.1873,
      "step": 2687
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.74854564666748,
      "learning_rate": 1.8466819221967965e-05,
      "loss": 1.0938,
      "step": 2688
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.616058349609375,
      "learning_rate": 1.8462528604118994e-05,
      "loss": 1.0013,
      "step": 2689
    },
    {
      "epoch": 0.77,
      "grad_norm": 8.085488319396973,
      "learning_rate": 1.8458237986270022e-05,
      "loss": 0.9281,
      "step": 2690
    },
    {
      "epoch": 0.77,
      "grad_norm": 7.250451564788818,
      "learning_rate": 1.8453947368421054e-05,
      "loss": 1.0076,
      "step": 2691
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.059502601623535,
      "learning_rate": 1.8449656750572083e-05,
      "loss": 0.9559,
      "step": 2692
    },
    {
      "epoch": 0.77,
      "grad_norm": 8.862520217895508,
      "learning_rate": 1.8445366132723115e-05,
      "loss": 1.0068,
      "step": 2693
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.757304191589355,
      "learning_rate": 1.844107551487414e-05,
      "loss": 1.0841,
      "step": 2694
    },
    {
      "epoch": 0.77,
      "grad_norm": 12.453340530395508,
      "learning_rate": 1.8436784897025173e-05,
      "loss": 1.4574,
      "step": 2695
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.447304725646973,
      "learning_rate": 1.8432494279176202e-05,
      "loss": 1.1475,
      "step": 2696
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.128987312316895,
      "learning_rate": 1.842820366132723e-05,
      "loss": 1.3936,
      "step": 2697
    },
    {
      "epoch": 0.77,
      "grad_norm": 8.295130729675293,
      "learning_rate": 1.842391304347826e-05,
      "loss": 1.0786,
      "step": 2698
    },
    {
      "epoch": 0.77,
      "grad_norm": 8.610480308532715,
      "learning_rate": 1.841962242562929e-05,
      "loss": 1.0077,
      "step": 2699
    },
    {
      "epoch": 0.77,
      "grad_norm": 7.891595840454102,
      "learning_rate": 1.8415331807780324e-05,
      "loss": 0.8981,
      "step": 2700
    },
    {
      "epoch": 0.77,
      "grad_norm": 8.902894020080566,
      "learning_rate": 1.841104118993135e-05,
      "loss": 0.9923,
      "step": 2701
    },
    {
      "epoch": 0.77,
      "grad_norm": 8.750622749328613,
      "learning_rate": 1.840675057208238e-05,
      "loss": 0.8989,
      "step": 2702
    },
    {
      "epoch": 0.77,
      "grad_norm": 8.808574676513672,
      "learning_rate": 1.840245995423341e-05,
      "loss": 0.848,
      "step": 2703
    },
    {
      "epoch": 0.77,
      "grad_norm": 8.473341941833496,
      "learning_rate": 1.8398169336384442e-05,
      "loss": 1.1411,
      "step": 2704
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.206725120544434,
      "learning_rate": 1.8393878718535467e-05,
      "loss": 1.0971,
      "step": 2705
    },
    {
      "epoch": 0.77,
      "grad_norm": 8.211832046508789,
      "learning_rate": 1.83895881006865e-05,
      "loss": 0.8216,
      "step": 2706
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.205649375915527,
      "learning_rate": 1.838529748283753e-05,
      "loss": 1.5003,
      "step": 2707
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.92562198638916,
      "learning_rate": 1.8381006864988557e-05,
      "loss": 1.094,
      "step": 2708
    },
    {
      "epoch": 0.77,
      "grad_norm": 8.90312671661377,
      "learning_rate": 1.837671624713959e-05,
      "loss": 1.1677,
      "step": 2709
    },
    {
      "epoch": 0.78,
      "grad_norm": 11.466680526733398,
      "learning_rate": 1.8372425629290618e-05,
      "loss": 1.1583,
      "step": 2710
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.08590030670166,
      "learning_rate": 1.836813501144165e-05,
      "loss": 1.0724,
      "step": 2711
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.549266815185547,
      "learning_rate": 1.8363844393592676e-05,
      "loss": 1.1368,
      "step": 2712
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.550378799438477,
      "learning_rate": 1.8359553775743708e-05,
      "loss": 1.1726,
      "step": 2713
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.870880126953125,
      "learning_rate": 1.8355263157894736e-05,
      "loss": 1.0843,
      "step": 2714
    },
    {
      "epoch": 0.78,
      "grad_norm": 8.207592964172363,
      "learning_rate": 1.835097254004577e-05,
      "loss": 1.1149,
      "step": 2715
    },
    {
      "epoch": 0.78,
      "grad_norm": 10.710161209106445,
      "learning_rate": 1.8346681922196794e-05,
      "loss": 1.188,
      "step": 2716
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.842384338378906,
      "learning_rate": 1.8342391304347826e-05,
      "loss": 1.2707,
      "step": 2717
    },
    {
      "epoch": 0.78,
      "grad_norm": 10.725979804992676,
      "learning_rate": 1.833810068649886e-05,
      "loss": 1.1073,
      "step": 2718
    },
    {
      "epoch": 0.78,
      "grad_norm": 8.74438190460205,
      "learning_rate": 1.8333810068649887e-05,
      "loss": 1.0223,
      "step": 2719
    },
    {
      "epoch": 0.78,
      "grad_norm": 10.196919441223145,
      "learning_rate": 1.8329519450800916e-05,
      "loss": 1.0389,
      "step": 2720
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.852926254272461,
      "learning_rate": 1.8325228832951945e-05,
      "loss": 1.3454,
      "step": 2721
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.271720886230469,
      "learning_rate": 1.8320938215102977e-05,
      "loss": 1.0826,
      "step": 2722
    },
    {
      "epoch": 0.78,
      "grad_norm": 8.834572792053223,
      "learning_rate": 1.8316647597254002e-05,
      "loss": 0.9662,
      "step": 2723
    },
    {
      "epoch": 0.78,
      "grad_norm": 8.742707252502441,
      "learning_rate": 1.8312356979405034e-05,
      "loss": 1.0828,
      "step": 2724
    },
    {
      "epoch": 0.78,
      "grad_norm": 10.010408401489258,
      "learning_rate": 1.8308066361556066e-05,
      "loss": 1.0087,
      "step": 2725
    },
    {
      "epoch": 0.78,
      "grad_norm": 10.239311218261719,
      "learning_rate": 1.8303775743707095e-05,
      "loss": 1.086,
      "step": 2726
    },
    {
      "epoch": 0.78,
      "grad_norm": 8.413912773132324,
      "learning_rate": 1.8299485125858124e-05,
      "loss": 1.0647,
      "step": 2727
    },
    {
      "epoch": 0.78,
      "grad_norm": 8.625919342041016,
      "learning_rate": 1.8295194508009153e-05,
      "loss": 1.1359,
      "step": 2728
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.683062553405762,
      "learning_rate": 1.8290903890160185e-05,
      "loss": 1.1086,
      "step": 2729
    },
    {
      "epoch": 0.78,
      "grad_norm": 8.346321105957031,
      "learning_rate": 1.8286613272311214e-05,
      "loss": 0.9379,
      "step": 2730
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.237349510192871,
      "learning_rate": 1.8282322654462242e-05,
      "loss": 0.968,
      "step": 2731
    },
    {
      "epoch": 0.78,
      "grad_norm": 8.374894142150879,
      "learning_rate": 1.827803203661327e-05,
      "loss": 0.9907,
      "step": 2732
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.64543342590332,
      "learning_rate": 1.8273741418764303e-05,
      "loss": 1.0265,
      "step": 2733
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.248924255371094,
      "learning_rate": 1.8269450800915332e-05,
      "loss": 1.1036,
      "step": 2734
    },
    {
      "epoch": 0.78,
      "grad_norm": 8.953285217285156,
      "learning_rate": 1.826516018306636e-05,
      "loss": 0.8402,
      "step": 2735
    },
    {
      "epoch": 0.78,
      "grad_norm": 11.745101928710938,
      "learning_rate": 1.8260869565217393e-05,
      "loss": 1.0951,
      "step": 2736
    },
    {
      "epoch": 0.78,
      "grad_norm": 11.503557205200195,
      "learning_rate": 1.8256578947368422e-05,
      "loss": 1.2745,
      "step": 2737
    },
    {
      "epoch": 0.78,
      "grad_norm": 7.832923412322998,
      "learning_rate": 1.825228832951945e-05,
      "loss": 0.9448,
      "step": 2738
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.32075023651123,
      "learning_rate": 1.824799771167048e-05,
      "loss": 1.1564,
      "step": 2739
    },
    {
      "epoch": 0.78,
      "grad_norm": 10.57994556427002,
      "learning_rate": 1.824370709382151e-05,
      "loss": 1.0131,
      "step": 2740
    },
    {
      "epoch": 0.78,
      "grad_norm": 10.00416374206543,
      "learning_rate": 1.823941647597254e-05,
      "loss": 1.0168,
      "step": 2741
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.534642219543457,
      "learning_rate": 1.823512585812357e-05,
      "loss": 1.1624,
      "step": 2742
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.412590980529785,
      "learning_rate": 1.82308352402746e-05,
      "loss": 1.0281,
      "step": 2743
    },
    {
      "epoch": 0.78,
      "grad_norm": 7.895086765289307,
      "learning_rate": 1.822654462242563e-05,
      "loss": 0.911,
      "step": 2744
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.005386352539062,
      "learning_rate": 1.8222254004576662e-05,
      "loss": 1.0018,
      "step": 2745
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.550594329833984,
      "learning_rate": 1.8217963386727688e-05,
      "loss": 0.9918,
      "step": 2746
    },
    {
      "epoch": 0.79,
      "grad_norm": 6.900699138641357,
      "learning_rate": 1.821367276887872e-05,
      "loss": 0.8453,
      "step": 2747
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.093343734741211,
      "learning_rate": 1.820938215102975e-05,
      "loss": 0.7882,
      "step": 2748
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.774897575378418,
      "learning_rate": 1.8205091533180777e-05,
      "loss": 1.0871,
      "step": 2749
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.677360534667969,
      "learning_rate": 1.8200800915331806e-05,
      "loss": 0.8507,
      "step": 2750
    },
    {
      "epoch": 0.79,
      "grad_norm": 11.09919548034668,
      "learning_rate": 1.8196510297482838e-05,
      "loss": 1.1706,
      "step": 2751
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.05200481414795,
      "learning_rate": 1.819221967963387e-05,
      "loss": 0.9521,
      "step": 2752
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.532218933105469,
      "learning_rate": 1.8187929061784896e-05,
      "loss": 1.1341,
      "step": 2753
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.517202377319336,
      "learning_rate": 1.8183638443935928e-05,
      "loss": 0.9918,
      "step": 2754
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.410207748413086,
      "learning_rate": 1.8179347826086957e-05,
      "loss": 1.1503,
      "step": 2755
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.514126777648926,
      "learning_rate": 1.817505720823799e-05,
      "loss": 0.9599,
      "step": 2756
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.519126892089844,
      "learning_rate": 1.8170766590389014e-05,
      "loss": 0.9525,
      "step": 2757
    },
    {
      "epoch": 0.79,
      "grad_norm": 11.100128173828125,
      "learning_rate": 1.8166475972540046e-05,
      "loss": 1.4367,
      "step": 2758
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.024754524230957,
      "learning_rate": 1.8162185354691075e-05,
      "loss": 0.8391,
      "step": 2759
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.76130485534668,
      "learning_rate": 1.8157894736842107e-05,
      "loss": 1.2622,
      "step": 2760
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.94499397277832,
      "learning_rate": 1.8153604118993136e-05,
      "loss": 1.3885,
      "step": 2761
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.015135765075684,
      "learning_rate": 1.8149313501144165e-05,
      "loss": 1.5093,
      "step": 2762
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.107895851135254,
      "learning_rate": 1.8145022883295197e-05,
      "loss": 0.9366,
      "step": 2763
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.041106224060059,
      "learning_rate": 1.8140732265446222e-05,
      "loss": 1.1956,
      "step": 2764
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.189181327819824,
      "learning_rate": 1.8136441647597254e-05,
      "loss": 1.0976,
      "step": 2765
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.372639656066895,
      "learning_rate": 1.8132151029748283e-05,
      "loss": 1.0334,
      "step": 2766
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.486791610717773,
      "learning_rate": 1.8127860411899315e-05,
      "loss": 1.115,
      "step": 2767
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.912895202636719,
      "learning_rate": 1.8123569794050344e-05,
      "loss": 0.761,
      "step": 2768
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.6273832321167,
      "learning_rate": 1.8119279176201373e-05,
      "loss": 0.8588,
      "step": 2769
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.135920524597168,
      "learning_rate": 1.8114988558352405e-05,
      "loss": 1.069,
      "step": 2770
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.210888862609863,
      "learning_rate": 1.8110697940503434e-05,
      "loss": 0.9714,
      "step": 2771
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.39043140411377,
      "learning_rate": 1.8106407322654463e-05,
      "loss": 1.0827,
      "step": 2772
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.327219009399414,
      "learning_rate": 1.810211670480549e-05,
      "loss": 1.0137,
      "step": 2773
    },
    {
      "epoch": 0.79,
      "grad_norm": 11.347169876098633,
      "learning_rate": 1.8097826086956524e-05,
      "loss": 1.0051,
      "step": 2774
    },
    {
      "epoch": 0.79,
      "grad_norm": 7.7931671142578125,
      "learning_rate": 1.809353546910755e-05,
      "loss": 0.9233,
      "step": 2775
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.85224437713623,
      "learning_rate": 1.808924485125858e-05,
      "loss": 1.0101,
      "step": 2776
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.604768753051758,
      "learning_rate": 1.8084954233409613e-05,
      "loss": 1.0067,
      "step": 2777
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.965160369873047,
      "learning_rate": 1.8080663615560642e-05,
      "loss": 0.8081,
      "step": 2778
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.196931838989258,
      "learning_rate": 1.807637299771167e-05,
      "loss": 1.1284,
      "step": 2779
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.866255760192871,
      "learning_rate": 1.80720823798627e-05,
      "loss": 1.2934,
      "step": 2780
    },
    {
      "epoch": 0.8,
      "grad_norm": 11.388153076171875,
      "learning_rate": 1.806779176201373e-05,
      "loss": 1.2407,
      "step": 2781
    },
    {
      "epoch": 0.8,
      "grad_norm": 7.777901649475098,
      "learning_rate": 1.806350114416476e-05,
      "loss": 0.78,
      "step": 2782
    },
    {
      "epoch": 0.8,
      "grad_norm": 12.280560493469238,
      "learning_rate": 1.805921052631579e-05,
      "loss": 1.2513,
      "step": 2783
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.953399658203125,
      "learning_rate": 1.8054919908466818e-05,
      "loss": 0.8567,
      "step": 2784
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.849992752075195,
      "learning_rate": 1.805062929061785e-05,
      "loss": 1.1959,
      "step": 2785
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.891160011291504,
      "learning_rate": 1.804633867276888e-05,
      "loss": 1.1964,
      "step": 2786
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.540994644165039,
      "learning_rate": 1.8042048054919908e-05,
      "loss": 1.0048,
      "step": 2787
    },
    {
      "epoch": 0.8,
      "grad_norm": 7.833096027374268,
      "learning_rate": 1.803775743707094e-05,
      "loss": 0.9079,
      "step": 2788
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.545397758483887,
      "learning_rate": 1.803346681922197e-05,
      "loss": 1.137,
      "step": 2789
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.271461486816406,
      "learning_rate": 1.8029176201372997e-05,
      "loss": 0.7457,
      "step": 2790
    },
    {
      "epoch": 0.8,
      "grad_norm": 7.652635097503662,
      "learning_rate": 1.8024885583524026e-05,
      "loss": 0.6924,
      "step": 2791
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.800633430480957,
      "learning_rate": 1.8020594965675058e-05,
      "loss": 0.8786,
      "step": 2792
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.03191089630127,
      "learning_rate": 1.8016304347826087e-05,
      "loss": 1.047,
      "step": 2793
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.993693351745605,
      "learning_rate": 1.8012013729977116e-05,
      "loss": 0.8651,
      "step": 2794
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.549789428710938,
      "learning_rate": 1.8007723112128148e-05,
      "loss": 0.8024,
      "step": 2795
    },
    {
      "epoch": 0.8,
      "grad_norm": 12.151975631713867,
      "learning_rate": 1.8003432494279177e-05,
      "loss": 1.3883,
      "step": 2796
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.671039581298828,
      "learning_rate": 1.799914187643021e-05,
      "loss": 0.9327,
      "step": 2797
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.68326187133789,
      "learning_rate": 1.7994851258581234e-05,
      "loss": 1.0669,
      "step": 2798
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.871553421020508,
      "learning_rate": 1.7990560640732266e-05,
      "loss": 0.989,
      "step": 2799
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.067877769470215,
      "learning_rate": 1.7986270022883295e-05,
      "loss": 1.0254,
      "step": 2800
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.065559387207031,
      "learning_rate": 1.7981979405034324e-05,
      "loss": 1.0136,
      "step": 2801
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.589637756347656,
      "learning_rate": 1.7977688787185356e-05,
      "loss": 1.0668,
      "step": 2802
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.31413745880127,
      "learning_rate": 1.7973398169336385e-05,
      "loss": 0.8831,
      "step": 2803
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.096830368041992,
      "learning_rate": 1.7969107551487417e-05,
      "loss": 0.702,
      "step": 2804
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.909378051757812,
      "learning_rate": 1.7964816933638442e-05,
      "loss": 1.2209,
      "step": 2805
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.658547401428223,
      "learning_rate": 1.7960526315789475e-05,
      "loss": 1.0153,
      "step": 2806
    },
    {
      "epoch": 0.8,
      "grad_norm": 11.248679161071777,
      "learning_rate": 1.7956235697940503e-05,
      "loss": 1.302,
      "step": 2807
    },
    {
      "epoch": 0.8,
      "grad_norm": 12.030311584472656,
      "learning_rate": 1.7951945080091536e-05,
      "loss": 1.3551,
      "step": 2808
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.535387992858887,
      "learning_rate": 1.794765446224256e-05,
      "loss": 1.0262,
      "step": 2809
    },
    {
      "epoch": 0.8,
      "grad_norm": 11.121785163879395,
      "learning_rate": 1.7943363844393593e-05,
      "loss": 1.2911,
      "step": 2810
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.692615509033203,
      "learning_rate": 1.7939073226544625e-05,
      "loss": 1.0072,
      "step": 2811
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.252633094787598,
      "learning_rate": 1.7934782608695654e-05,
      "loss": 1.1589,
      "step": 2812
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.307209968566895,
      "learning_rate": 1.7930491990846683e-05,
      "loss": 1.133,
      "step": 2813
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.116240501403809,
      "learning_rate": 1.792620137299771e-05,
      "loss": 0.8728,
      "step": 2814
    },
    {
      "epoch": 0.81,
      "grad_norm": 8.729217529296875,
      "learning_rate": 1.7921910755148744e-05,
      "loss": 1.0193,
      "step": 2815
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.098910331726074,
      "learning_rate": 1.791762013729977e-05,
      "loss": 1.291,
      "step": 2816
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.680808067321777,
      "learning_rate": 1.79133295194508e-05,
      "loss": 1.5833,
      "step": 2817
    },
    {
      "epoch": 0.81,
      "grad_norm": 8.461689949035645,
      "learning_rate": 1.790903890160183e-05,
      "loss": 1.0234,
      "step": 2818
    },
    {
      "epoch": 0.81,
      "grad_norm": 8.697273254394531,
      "learning_rate": 1.7904748283752862e-05,
      "loss": 1.3734,
      "step": 2819
    },
    {
      "epoch": 0.81,
      "grad_norm": 6.797390937805176,
      "learning_rate": 1.790045766590389e-05,
      "loss": 0.6928,
      "step": 2820
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.431440353393555,
      "learning_rate": 1.789616704805492e-05,
      "loss": 1.086,
      "step": 2821
    },
    {
      "epoch": 0.81,
      "grad_norm": 8.511075019836426,
      "learning_rate": 1.7891876430205952e-05,
      "loss": 1.0503,
      "step": 2822
    },
    {
      "epoch": 0.81,
      "grad_norm": 10.25367546081543,
      "learning_rate": 1.788758581235698e-05,
      "loss": 1.3238,
      "step": 2823
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.629301071166992,
      "learning_rate": 1.788329519450801e-05,
      "loss": 1.2745,
      "step": 2824
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.358258247375488,
      "learning_rate": 1.7879004576659038e-05,
      "loss": 1.1041,
      "step": 2825
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.950977325439453,
      "learning_rate": 1.787471395881007e-05,
      "loss": 1.126,
      "step": 2826
    },
    {
      "epoch": 0.81,
      "grad_norm": 8.868396759033203,
      "learning_rate": 1.7870423340961096e-05,
      "loss": 0.8953,
      "step": 2827
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.405526161193848,
      "learning_rate": 1.7866132723112128e-05,
      "loss": 1.1577,
      "step": 2828
    },
    {
      "epoch": 0.81,
      "grad_norm": 8.505644798278809,
      "learning_rate": 1.786184210526316e-05,
      "loss": 1.0532,
      "step": 2829
    },
    {
      "epoch": 0.81,
      "grad_norm": 8.55901050567627,
      "learning_rate": 1.785755148741419e-05,
      "loss": 1.044,
      "step": 2830
    },
    {
      "epoch": 0.81,
      "grad_norm": 11.979047775268555,
      "learning_rate": 1.7853260869565218e-05,
      "loss": 1.2484,
      "step": 2831
    },
    {
      "epoch": 0.81,
      "grad_norm": 8.85673713684082,
      "learning_rate": 1.7848970251716246e-05,
      "loss": 1.0276,
      "step": 2832
    },
    {
      "epoch": 0.81,
      "grad_norm": 8.023272514343262,
      "learning_rate": 1.784467963386728e-05,
      "loss": 0.9118,
      "step": 2833
    },
    {
      "epoch": 0.81,
      "grad_norm": 10.74399471282959,
      "learning_rate": 1.7840389016018307e-05,
      "loss": 1.2764,
      "step": 2834
    },
    {
      "epoch": 0.81,
      "grad_norm": 10.298340797424316,
      "learning_rate": 1.7836098398169336e-05,
      "loss": 1.1011,
      "step": 2835
    },
    {
      "epoch": 0.81,
      "grad_norm": 11.09740924835205,
      "learning_rate": 1.7831807780320365e-05,
      "loss": 1.3755,
      "step": 2836
    },
    {
      "epoch": 0.81,
      "grad_norm": 11.25870132446289,
      "learning_rate": 1.7827517162471397e-05,
      "loss": 1.5211,
      "step": 2837
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.853792190551758,
      "learning_rate": 1.782322654462243e-05,
      "loss": 1.031,
      "step": 2838
    },
    {
      "epoch": 0.81,
      "grad_norm": 8.365081787109375,
      "learning_rate": 1.7818935926773454e-05,
      "loss": 1.1502,
      "step": 2839
    },
    {
      "epoch": 0.81,
      "grad_norm": 7.9195556640625,
      "learning_rate": 1.7814645308924487e-05,
      "loss": 0.8347,
      "step": 2840
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.461267471313477,
      "learning_rate": 1.7810354691075515e-05,
      "loss": 0.959,
      "step": 2841
    },
    {
      "epoch": 0.81,
      "grad_norm": 7.771499156951904,
      "learning_rate": 1.7806064073226544e-05,
      "loss": 0.9366,
      "step": 2842
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.530256271362305,
      "learning_rate": 1.7801773455377573e-05,
      "loss": 1.1998,
      "step": 2843
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.167468070983887,
      "learning_rate": 1.7797482837528605e-05,
      "loss": 1.0393,
      "step": 2844
    },
    {
      "epoch": 0.81,
      "grad_norm": 10.303680419921875,
      "learning_rate": 1.7793192219679637e-05,
      "loss": 0.9507,
      "step": 2845
    },
    {
      "epoch": 0.81,
      "grad_norm": 8.955582618713379,
      "learning_rate": 1.7788901601830663e-05,
      "loss": 0.9256,
      "step": 2846
    },
    {
      "epoch": 0.81,
      "grad_norm": 12.868518829345703,
      "learning_rate": 1.7784610983981695e-05,
      "loss": 1.414,
      "step": 2847
    },
    {
      "epoch": 0.81,
      "grad_norm": 10.280289649963379,
      "learning_rate": 1.7780320366132724e-05,
      "loss": 1.2262,
      "step": 2848
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.211709022521973,
      "learning_rate": 1.7776029748283756e-05,
      "loss": 0.9237,
      "step": 2849
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.173371315002441,
      "learning_rate": 1.777173913043478e-05,
      "loss": 1.1584,
      "step": 2850
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.673914909362793,
      "learning_rate": 1.7767448512585813e-05,
      "loss": 0.9677,
      "step": 2851
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.421575546264648,
      "learning_rate": 1.7763157894736842e-05,
      "loss": 1.2843,
      "step": 2852
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.3233060836792,
      "learning_rate": 1.775886727688787e-05,
      "loss": 0.9641,
      "step": 2853
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.67033863067627,
      "learning_rate": 1.7754576659038903e-05,
      "loss": 0.9924,
      "step": 2854
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.184983253479004,
      "learning_rate": 1.775028604118993e-05,
      "loss": 0.8973,
      "step": 2855
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.033408164978027,
      "learning_rate": 1.7745995423340964e-05,
      "loss": 1.0007,
      "step": 2856
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.438945770263672,
      "learning_rate": 1.774170480549199e-05,
      "loss": 1.0585,
      "step": 2857
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.126609802246094,
      "learning_rate": 1.773741418764302e-05,
      "loss": 1.0277,
      "step": 2858
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.337931632995605,
      "learning_rate": 1.773312356979405e-05,
      "loss": 0.9812,
      "step": 2859
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.967913627624512,
      "learning_rate": 1.7728832951945082e-05,
      "loss": 1.3706,
      "step": 2860
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.8753023147583,
      "learning_rate": 1.7724542334096108e-05,
      "loss": 1.0067,
      "step": 2861
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.703413009643555,
      "learning_rate": 1.772025171624714e-05,
      "loss": 1.0373,
      "step": 2862
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.074624061584473,
      "learning_rate": 1.7715961098398172e-05,
      "loss": 1.3182,
      "step": 2863
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.192731857299805,
      "learning_rate": 1.77116704805492e-05,
      "loss": 0.897,
      "step": 2864
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.995240211486816,
      "learning_rate": 1.770737986270023e-05,
      "loss": 1.1138,
      "step": 2865
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.425854682922363,
      "learning_rate": 1.7703089244851258e-05,
      "loss": 1.1363,
      "step": 2866
    },
    {
      "epoch": 0.82,
      "grad_norm": 7.753910541534424,
      "learning_rate": 1.769879862700229e-05,
      "loss": 0.824,
      "step": 2867
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.238604545593262,
      "learning_rate": 1.7694508009153316e-05,
      "loss": 0.8746,
      "step": 2868
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.719847679138184,
      "learning_rate": 1.7690217391304348e-05,
      "loss": 1.1737,
      "step": 2869
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.411002159118652,
      "learning_rate": 1.7685926773455377e-05,
      "loss": 0.995,
      "step": 2870
    },
    {
      "epoch": 0.82,
      "grad_norm": 7.016213893890381,
      "learning_rate": 1.768163615560641e-05,
      "loss": 0.8299,
      "step": 2871
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.941311836242676,
      "learning_rate": 1.7677345537757438e-05,
      "loss": 1.0729,
      "step": 2872
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.35824203491211,
      "learning_rate": 1.7673054919908466e-05,
      "loss": 1.0629,
      "step": 2873
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.965079307556152,
      "learning_rate": 1.76687643020595e-05,
      "loss": 0.9683,
      "step": 2874
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.929749488830566,
      "learning_rate": 1.7664473684210527e-05,
      "loss": 0.8733,
      "step": 2875
    },
    {
      "epoch": 0.82,
      "grad_norm": 6.842812538146973,
      "learning_rate": 1.7660183066361556e-05,
      "loss": 0.6962,
      "step": 2876
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.920101165771484,
      "learning_rate": 1.7655892448512585e-05,
      "loss": 0.9707,
      "step": 2877
    },
    {
      "epoch": 0.82,
      "grad_norm": 7.8547139167785645,
      "learning_rate": 1.7651601830663617e-05,
      "loss": 0.8205,
      "step": 2878
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.265000343322754,
      "learning_rate": 1.7647311212814642e-05,
      "loss": 1.1583,
      "step": 2879
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.203521728515625,
      "learning_rate": 1.7643020594965675e-05,
      "loss": 0.9276,
      "step": 2880
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.345846176147461,
      "learning_rate": 1.7638729977116707e-05,
      "loss": 0.7885,
      "step": 2881
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.39890193939209,
      "learning_rate": 1.7634439359267736e-05,
      "loss": 1.0957,
      "step": 2882
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.122784614562988,
      "learning_rate": 1.7630148741418764e-05,
      "loss": 1.0031,
      "step": 2883
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.748235702514648,
      "learning_rate": 1.7625858123569793e-05,
      "loss": 0.785,
      "step": 2884
    },
    {
      "epoch": 0.83,
      "grad_norm": 7.878302097320557,
      "learning_rate": 1.7621567505720825e-05,
      "loss": 0.7446,
      "step": 2885
    },
    {
      "epoch": 0.83,
      "grad_norm": 12.413397789001465,
      "learning_rate": 1.7617276887871854e-05,
      "loss": 1.0913,
      "step": 2886
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.62625503540039,
      "learning_rate": 1.7612986270022883e-05,
      "loss": 0.729,
      "step": 2887
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.879027366638184,
      "learning_rate": 1.7608695652173915e-05,
      "loss": 1.1488,
      "step": 2888
    },
    {
      "epoch": 0.83,
      "grad_norm": 12.591888427734375,
      "learning_rate": 1.7604405034324944e-05,
      "loss": 1.2602,
      "step": 2889
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.228856086730957,
      "learning_rate": 1.7600114416475976e-05,
      "loss": 0.954,
      "step": 2890
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.087937355041504,
      "learning_rate": 1.7595823798627e-05,
      "loss": 0.8813,
      "step": 2891
    },
    {
      "epoch": 0.83,
      "grad_norm": 11.661734580993652,
      "learning_rate": 1.7591533180778033e-05,
      "loss": 1.2382,
      "step": 2892
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.572035789489746,
      "learning_rate": 1.7587242562929062e-05,
      "loss": 0.7731,
      "step": 2893
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.623916625976562,
      "learning_rate": 1.758295194508009e-05,
      "loss": 1.0266,
      "step": 2894
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.228483200073242,
      "learning_rate": 1.757866132723112e-05,
      "loss": 0.8791,
      "step": 2895
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.038342475891113,
      "learning_rate": 1.7574370709382152e-05,
      "loss": 0.9342,
      "step": 2896
    },
    {
      "epoch": 0.83,
      "grad_norm": 8.881660461425781,
      "learning_rate": 1.7570080091533184e-05,
      "loss": 1.0093,
      "step": 2897
    },
    {
      "epoch": 0.83,
      "grad_norm": 11.040688514709473,
      "learning_rate": 1.756578947368421e-05,
      "loss": 1.0726,
      "step": 2898
    },
    {
      "epoch": 0.83,
      "grad_norm": 7.727034568786621,
      "learning_rate": 1.756149885583524e-05,
      "loss": 0.9336,
      "step": 2899
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.47050666809082,
      "learning_rate": 1.755720823798627e-05,
      "loss": 1.1571,
      "step": 2900
    },
    {
      "epoch": 0.83,
      "grad_norm": 8.488235473632812,
      "learning_rate": 1.7552917620137302e-05,
      "loss": 0.917,
      "step": 2901
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.848282814025879,
      "learning_rate": 1.7548627002288328e-05,
      "loss": 1.1838,
      "step": 2902
    },
    {
      "epoch": 0.83,
      "grad_norm": 8.979755401611328,
      "learning_rate": 1.754433638443936e-05,
      "loss": 1.234,
      "step": 2903
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.087435722351074,
      "learning_rate": 1.754004576659039e-05,
      "loss": 1.1591,
      "step": 2904
    },
    {
      "epoch": 0.83,
      "grad_norm": 7.905564308166504,
      "learning_rate": 1.7535755148741417e-05,
      "loss": 0.9567,
      "step": 2905
    },
    {
      "epoch": 0.83,
      "grad_norm": 8.611047744750977,
      "learning_rate": 1.753146453089245e-05,
      "loss": 1.0315,
      "step": 2906
    },
    {
      "epoch": 0.83,
      "grad_norm": 8.980897903442383,
      "learning_rate": 1.752717391304348e-05,
      "loss": 1.0776,
      "step": 2907
    },
    {
      "epoch": 0.83,
      "grad_norm": 8.368359565734863,
      "learning_rate": 1.752288329519451e-05,
      "loss": 0.922,
      "step": 2908
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.533286094665527,
      "learning_rate": 1.7518592677345536e-05,
      "loss": 1.0402,
      "step": 2909
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.625426292419434,
      "learning_rate": 1.7514302059496568e-05,
      "loss": 1.0357,
      "step": 2910
    },
    {
      "epoch": 0.83,
      "grad_norm": 7.755105018615723,
      "learning_rate": 1.7510011441647597e-05,
      "loss": 0.9851,
      "step": 2911
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.236310005187988,
      "learning_rate": 1.750572082379863e-05,
      "loss": 1.1758,
      "step": 2912
    },
    {
      "epoch": 0.83,
      "grad_norm": 8.932103157043457,
      "learning_rate": 1.7501430205949654e-05,
      "loss": 0.7994,
      "step": 2913
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.338969230651855,
      "learning_rate": 1.7497139588100687e-05,
      "loss": 1.0408,
      "step": 2914
    },
    {
      "epoch": 0.83,
      "grad_norm": 8.797367095947266,
      "learning_rate": 1.749284897025172e-05,
      "loss": 1.1807,
      "step": 2915
    },
    {
      "epoch": 0.83,
      "grad_norm": 8.847677230834961,
      "learning_rate": 1.7488558352402747e-05,
      "loss": 0.8965,
      "step": 2916
    },
    {
      "epoch": 0.83,
      "grad_norm": 7.4715447425842285,
      "learning_rate": 1.7484267734553776e-05,
      "loss": 0.9171,
      "step": 2917
    },
    {
      "epoch": 0.83,
      "grad_norm": 8.547280311584473,
      "learning_rate": 1.7479977116704805e-05,
      "loss": 1.1989,
      "step": 2918
    },
    {
      "epoch": 0.83,
      "grad_norm": 8.58836841583252,
      "learning_rate": 1.7475686498855837e-05,
      "loss": 0.7985,
      "step": 2919
    },
    {
      "epoch": 0.84,
      "grad_norm": 12.026870727539062,
      "learning_rate": 1.7471395881006863e-05,
      "loss": 1.15,
      "step": 2920
    },
    {
      "epoch": 0.84,
      "grad_norm": 8.279645919799805,
      "learning_rate": 1.7467105263157895e-05,
      "loss": 0.9435,
      "step": 2921
    },
    {
      "epoch": 0.84,
      "grad_norm": 8.120729446411133,
      "learning_rate": 1.7462814645308927e-05,
      "loss": 0.8074,
      "step": 2922
    },
    {
      "epoch": 0.84,
      "grad_norm": 11.065886497497559,
      "learning_rate": 1.7458524027459956e-05,
      "loss": 1.1931,
      "step": 2923
    },
    {
      "epoch": 0.84,
      "grad_norm": 12.420193672180176,
      "learning_rate": 1.7454233409610984e-05,
      "loss": 1.3434,
      "step": 2924
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.91895866394043,
      "learning_rate": 1.7449942791762013e-05,
      "loss": 1.0267,
      "step": 2925
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.360223770141602,
      "learning_rate": 1.7445652173913045e-05,
      "loss": 0.9777,
      "step": 2926
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.885186195373535,
      "learning_rate": 1.7441361556064074e-05,
      "loss": 1.035,
      "step": 2927
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.056436538696289,
      "learning_rate": 1.7437070938215103e-05,
      "loss": 1.3135,
      "step": 2928
    },
    {
      "epoch": 0.84,
      "grad_norm": 11.684242248535156,
      "learning_rate": 1.743278032036613e-05,
      "loss": 1.1643,
      "step": 2929
    },
    {
      "epoch": 0.84,
      "grad_norm": 7.900113582611084,
      "learning_rate": 1.7428489702517164e-05,
      "loss": 0.8937,
      "step": 2930
    },
    {
      "epoch": 0.84,
      "grad_norm": 8.781208992004395,
      "learning_rate": 1.7424199084668193e-05,
      "loss": 0.6792,
      "step": 2931
    },
    {
      "epoch": 0.84,
      "grad_norm": 8.980362892150879,
      "learning_rate": 1.741990846681922e-05,
      "loss": 1.0824,
      "step": 2932
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.64148235321045,
      "learning_rate": 1.7415617848970253e-05,
      "loss": 1.1883,
      "step": 2933
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.602545738220215,
      "learning_rate": 1.7411327231121282e-05,
      "loss": 1.1871,
      "step": 2934
    },
    {
      "epoch": 0.84,
      "grad_norm": 12.133111000061035,
      "learning_rate": 1.740703661327231e-05,
      "loss": 1.5245,
      "step": 2935
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.329170227050781,
      "learning_rate": 1.740274599542334e-05,
      "loss": 1.2061,
      "step": 2936
    },
    {
      "epoch": 0.84,
      "grad_norm": 8.698854446411133,
      "learning_rate": 1.7398455377574372e-05,
      "loss": 0.9893,
      "step": 2937
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.945671081542969,
      "learning_rate": 1.73941647597254e-05,
      "loss": 1.381,
      "step": 2938
    },
    {
      "epoch": 0.84,
      "grad_norm": 8.756754875183105,
      "learning_rate": 1.738987414187643e-05,
      "loss": 0.9526,
      "step": 2939
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.349870681762695,
      "learning_rate": 1.738558352402746e-05,
      "loss": 1.3003,
      "step": 2940
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.652379035949707,
      "learning_rate": 1.738129290617849e-05,
      "loss": 1.1248,
      "step": 2941
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.066186904907227,
      "learning_rate": 1.7377002288329523e-05,
      "loss": 1.0749,
      "step": 2942
    },
    {
      "epoch": 0.84,
      "grad_norm": 7.665074825286865,
      "learning_rate": 1.7372711670480548e-05,
      "loss": 0.9764,
      "step": 2943
    },
    {
      "epoch": 0.84,
      "grad_norm": 6.827920436859131,
      "learning_rate": 1.736842105263158e-05,
      "loss": 0.9367,
      "step": 2944
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.536245346069336,
      "learning_rate": 1.736413043478261e-05,
      "loss": 1.0384,
      "step": 2945
    },
    {
      "epoch": 0.84,
      "grad_norm": 8.041460037231445,
      "learning_rate": 1.7359839816933638e-05,
      "loss": 0.9429,
      "step": 2946
    },
    {
      "epoch": 0.84,
      "grad_norm": 7.036596298217773,
      "learning_rate": 1.7355549199084666e-05,
      "loss": 0.8475,
      "step": 2947
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.844472885131836,
      "learning_rate": 1.73512585812357e-05,
      "loss": 1.017,
      "step": 2948
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.281776428222656,
      "learning_rate": 1.734696796338673e-05,
      "loss": 1.1786,
      "step": 2949
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.428088188171387,
      "learning_rate": 1.7342677345537756e-05,
      "loss": 1.0895,
      "step": 2950
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.562387466430664,
      "learning_rate": 1.7338386727688788e-05,
      "loss": 1.2424,
      "step": 2951
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.42207145690918,
      "learning_rate": 1.7334096109839817e-05,
      "loss": 1.0374,
      "step": 2952
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.31929874420166,
      "learning_rate": 1.732980549199085e-05,
      "loss": 1.0619,
      "step": 2953
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.772723197937012,
      "learning_rate": 1.7325514874141875e-05,
      "loss": 1.2016,
      "step": 2954
    },
    {
      "epoch": 0.85,
      "grad_norm": 10.2816743850708,
      "learning_rate": 1.7321224256292907e-05,
      "loss": 1.1468,
      "step": 2955
    },
    {
      "epoch": 0.85,
      "grad_norm": 10.321672439575195,
      "learning_rate": 1.7316933638443935e-05,
      "loss": 1.0652,
      "step": 2956
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.434853553771973,
      "learning_rate": 1.7312643020594968e-05,
      "loss": 1.1972,
      "step": 2957
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.571502685546875,
      "learning_rate": 1.7308352402745996e-05,
      "loss": 0.9115,
      "step": 2958
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.985664367675781,
      "learning_rate": 1.7304061784897025e-05,
      "loss": 1.0932,
      "step": 2959
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.816332817077637,
      "learning_rate": 1.7299771167048057e-05,
      "loss": 1.2125,
      "step": 2960
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.004415512084961,
      "learning_rate": 1.7295480549199083e-05,
      "loss": 1.018,
      "step": 2961
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.268218040466309,
      "learning_rate": 1.7291189931350115e-05,
      "loss": 1.0142,
      "step": 2962
    },
    {
      "epoch": 0.85,
      "grad_norm": 7.872369766235352,
      "learning_rate": 1.7286899313501144e-05,
      "loss": 0.9751,
      "step": 2963
    },
    {
      "epoch": 0.85,
      "grad_norm": 8.151576042175293,
      "learning_rate": 1.7282608695652176e-05,
      "loss": 0.9471,
      "step": 2964
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.960821151733398,
      "learning_rate": 1.7278318077803205e-05,
      "loss": 1.2366,
      "step": 2965
    },
    {
      "epoch": 0.85,
      "grad_norm": 8.932233810424805,
      "learning_rate": 1.7274027459954233e-05,
      "loss": 0.9703,
      "step": 2966
    },
    {
      "epoch": 0.85,
      "grad_norm": 8.246437072753906,
      "learning_rate": 1.7269736842105265e-05,
      "loss": 0.7479,
      "step": 2967
    },
    {
      "epoch": 0.85,
      "grad_norm": 11.880038261413574,
      "learning_rate": 1.7265446224256294e-05,
      "loss": 1.344,
      "step": 2968
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.149515151977539,
      "learning_rate": 1.7261155606407323e-05,
      "loss": 1.066,
      "step": 2969
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.845952033996582,
      "learning_rate": 1.7256864988558352e-05,
      "loss": 1.1182,
      "step": 2970
    },
    {
      "epoch": 0.85,
      "grad_norm": 10.139139175415039,
      "learning_rate": 1.7252574370709384e-05,
      "loss": 1.2844,
      "step": 2971
    },
    {
      "epoch": 0.85,
      "grad_norm": 8.441656112670898,
      "learning_rate": 1.724828375286041e-05,
      "loss": 0.835,
      "step": 2972
    },
    {
      "epoch": 0.85,
      "grad_norm": 10.01870346069336,
      "learning_rate": 1.724399313501144e-05,
      "loss": 1.1136,
      "step": 2973
    },
    {
      "epoch": 0.85,
      "grad_norm": 8.709890365600586,
      "learning_rate": 1.7239702517162474e-05,
      "loss": 1.0586,
      "step": 2974
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.84835147857666,
      "learning_rate": 1.7235411899313502e-05,
      "loss": 1.1279,
      "step": 2975
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.617959022521973,
      "learning_rate": 1.723112128146453e-05,
      "loss": 1.2462,
      "step": 2976
    },
    {
      "epoch": 0.85,
      "grad_norm": 8.675017356872559,
      "learning_rate": 1.722683066361556e-05,
      "loss": 0.9196,
      "step": 2977
    },
    {
      "epoch": 0.85,
      "grad_norm": 8.21192455291748,
      "learning_rate": 1.7222540045766592e-05,
      "loss": 1.1072,
      "step": 2978
    },
    {
      "epoch": 0.85,
      "grad_norm": 7.8568902015686035,
      "learning_rate": 1.721824942791762e-05,
      "loss": 1.0281,
      "step": 2979
    },
    {
      "epoch": 0.85,
      "grad_norm": 7.6694488525390625,
      "learning_rate": 1.721395881006865e-05,
      "loss": 0.902,
      "step": 2980
    },
    {
      "epoch": 0.85,
      "grad_norm": 8.478190422058105,
      "learning_rate": 1.720966819221968e-05,
      "loss": 0.9622,
      "step": 2981
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.412214279174805,
      "learning_rate": 1.720537757437071e-05,
      "loss": 1.0104,
      "step": 2982
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.078675270080566,
      "learning_rate": 1.7201086956521743e-05,
      "loss": 1.0179,
      "step": 2983
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.115436553955078,
      "learning_rate": 1.7196796338672768e-05,
      "loss": 1.1088,
      "step": 2984
    },
    {
      "epoch": 0.85,
      "grad_norm": 7.480807781219482,
      "learning_rate": 1.71925057208238e-05,
      "loss": 0.8795,
      "step": 2985
    },
    {
      "epoch": 0.85,
      "grad_norm": 8.407920837402344,
      "learning_rate": 1.718821510297483e-05,
      "loss": 0.9394,
      "step": 2986
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.731978416442871,
      "learning_rate": 1.7183924485125858e-05,
      "loss": 1.0389,
      "step": 2987
    },
    {
      "epoch": 0.85,
      "grad_norm": 8.553013801574707,
      "learning_rate": 1.7179633867276887e-05,
      "loss": 0.96,
      "step": 2988
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.58366870880127,
      "learning_rate": 1.717534324942792e-05,
      "loss": 1.2158,
      "step": 2989
    },
    {
      "epoch": 0.86,
      "grad_norm": 12.59498119354248,
      "learning_rate": 1.7171052631578947e-05,
      "loss": 0.9116,
      "step": 2990
    },
    {
      "epoch": 0.86,
      "grad_norm": 8.889739036560059,
      "learning_rate": 1.7166762013729976e-05,
      "loss": 0.7415,
      "step": 2991
    },
    {
      "epoch": 0.86,
      "grad_norm": 7.628733158111572,
      "learning_rate": 1.716247139588101e-05,
      "loss": 0.7047,
      "step": 2992
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.510764122009277,
      "learning_rate": 1.7158180778032037e-05,
      "loss": 0.9974,
      "step": 2993
    },
    {
      "epoch": 0.86,
      "grad_norm": 11.699542999267578,
      "learning_rate": 1.715389016018307e-05,
      "loss": 1.1202,
      "step": 2994
    },
    {
      "epoch": 0.86,
      "grad_norm": 11.28286075592041,
      "learning_rate": 1.7149599542334095e-05,
      "loss": 0.9915,
      "step": 2995
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.383990287780762,
      "learning_rate": 1.7145308924485127e-05,
      "loss": 0.9895,
      "step": 2996
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.332915306091309,
      "learning_rate": 1.7141018306636156e-05,
      "loss": 0.9564,
      "step": 2997
    },
    {
      "epoch": 0.86,
      "grad_norm": 8.677774429321289,
      "learning_rate": 1.7136727688787184e-05,
      "loss": 0.7396,
      "step": 2998
    },
    {
      "epoch": 0.86,
      "grad_norm": 10.549933433532715,
      "learning_rate": 1.7132437070938213e-05,
      "loss": 0.9943,
      "step": 2999
    },
    {
      "epoch": 0.86,
      "grad_norm": 10.350984573364258,
      "learning_rate": 1.7128146453089245e-05,
      "loss": 1.2945,
      "step": 3000
    },
    {
      "epoch": 0.86,
      "grad_norm": 11.026261329650879,
      "learning_rate": 1.7123855835240277e-05,
      "loss": 1.1736,
      "step": 3001
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.979823112487793,
      "learning_rate": 1.7119565217391303e-05,
      "loss": 1.0384,
      "step": 3002
    },
    {
      "epoch": 0.86,
      "grad_norm": 8.890838623046875,
      "learning_rate": 1.7115274599542335e-05,
      "loss": 1.1439,
      "step": 3003
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.048349380493164,
      "learning_rate": 1.7110983981693364e-05,
      "loss": 1.0301,
      "step": 3004
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.576517105102539,
      "learning_rate": 1.7106693363844396e-05,
      "loss": 0.9483,
      "step": 3005
    },
    {
      "epoch": 0.86,
      "grad_norm": 11.636025428771973,
      "learning_rate": 1.710240274599542e-05,
      "loss": 1.1323,
      "step": 3006
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.963531494140625,
      "learning_rate": 1.7098112128146453e-05,
      "loss": 1.1524,
      "step": 3007
    },
    {
      "epoch": 0.86,
      "grad_norm": 10.237224578857422,
      "learning_rate": 1.7093821510297486e-05,
      "loss": 1.2993,
      "step": 3008
    },
    {
      "epoch": 0.86,
      "grad_norm": 8.78339672088623,
      "learning_rate": 1.7089530892448514e-05,
      "loss": 0.8406,
      "step": 3009
    },
    {
      "epoch": 0.86,
      "grad_norm": 10.362381935119629,
      "learning_rate": 1.7085240274599543e-05,
      "loss": 1.0079,
      "step": 3010
    },
    {
      "epoch": 0.86,
      "grad_norm": 10.179791450500488,
      "learning_rate": 1.7080949656750572e-05,
      "loss": 1.2726,
      "step": 3011
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.268647193908691,
      "learning_rate": 1.7076659038901604e-05,
      "loss": 1.1014,
      "step": 3012
    },
    {
      "epoch": 0.86,
      "grad_norm": 11.005672454833984,
      "learning_rate": 1.707236842105263e-05,
      "loss": 1.3472,
      "step": 3013
    },
    {
      "epoch": 0.86,
      "grad_norm": 8.85457706451416,
      "learning_rate": 1.706807780320366e-05,
      "loss": 1.0802,
      "step": 3014
    },
    {
      "epoch": 0.86,
      "grad_norm": 10.423368453979492,
      "learning_rate": 1.706378718535469e-05,
      "loss": 1.0567,
      "step": 3015
    },
    {
      "epoch": 0.86,
      "grad_norm": 8.43641471862793,
      "learning_rate": 1.7059496567505723e-05,
      "loss": 1.0063,
      "step": 3016
    },
    {
      "epoch": 0.86,
      "grad_norm": 8.766290664672852,
      "learning_rate": 1.705520594965675e-05,
      "loss": 1.1333,
      "step": 3017
    },
    {
      "epoch": 0.86,
      "grad_norm": 8.754609107971191,
      "learning_rate": 1.705091533180778e-05,
      "loss": 1.0378,
      "step": 3018
    },
    {
      "epoch": 0.86,
      "grad_norm": 8.543985366821289,
      "learning_rate": 1.7046624713958812e-05,
      "loss": 1.1846,
      "step": 3019
    },
    {
      "epoch": 0.86,
      "grad_norm": 7.664855480194092,
      "learning_rate": 1.704233409610984e-05,
      "loss": 1.0171,
      "step": 3020
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.609197616577148,
      "learning_rate": 1.703804347826087e-05,
      "loss": 1.0862,
      "step": 3021
    },
    {
      "epoch": 0.86,
      "grad_norm": 8.459212303161621,
      "learning_rate": 1.70337528604119e-05,
      "loss": 0.9658,
      "step": 3022
    },
    {
      "epoch": 0.86,
      "grad_norm": 8.076066017150879,
      "learning_rate": 1.702946224256293e-05,
      "loss": 0.9613,
      "step": 3023
    },
    {
      "epoch": 0.86,
      "grad_norm": 8.68406867980957,
      "learning_rate": 1.7025171624713956e-05,
      "loss": 0.8566,
      "step": 3024
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.512566566467285,
      "learning_rate": 1.7020881006864988e-05,
      "loss": 0.8524,
      "step": 3025
    },
    {
      "epoch": 0.87,
      "grad_norm": 8.096138954162598,
      "learning_rate": 1.701659038901602e-05,
      "loss": 0.682,
      "step": 3026
    },
    {
      "epoch": 0.87,
      "grad_norm": 8.064615249633789,
      "learning_rate": 1.701229977116705e-05,
      "loss": 1.1461,
      "step": 3027
    },
    {
      "epoch": 0.87,
      "grad_norm": 8.561189651489258,
      "learning_rate": 1.7008009153318078e-05,
      "loss": 1.0293,
      "step": 3028
    },
    {
      "epoch": 0.87,
      "grad_norm": 7.9466447830200195,
      "learning_rate": 1.7003718535469107e-05,
      "loss": 0.8737,
      "step": 3029
    },
    {
      "epoch": 0.87,
      "grad_norm": 8.754287719726562,
      "learning_rate": 1.699942791762014e-05,
      "loss": 0.9984,
      "step": 3030
    },
    {
      "epoch": 0.87,
      "grad_norm": 11.927494049072266,
      "learning_rate": 1.6995137299771168e-05,
      "loss": 1.2555,
      "step": 3031
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.53317642211914,
      "learning_rate": 1.6990846681922196e-05,
      "loss": 1.0106,
      "step": 3032
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.192111015319824,
      "learning_rate": 1.6986556064073225e-05,
      "loss": 1.1324,
      "step": 3033
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.124427795410156,
      "learning_rate": 1.6982265446224257e-05,
      "loss": 0.9526,
      "step": 3034
    },
    {
      "epoch": 0.87,
      "grad_norm": 11.455728530883789,
      "learning_rate": 1.697797482837529e-05,
      "loss": 1.2767,
      "step": 3035
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.49399471282959,
      "learning_rate": 1.6973684210526315e-05,
      "loss": 0.9035,
      "step": 3036
    },
    {
      "epoch": 0.87,
      "grad_norm": 8.540650367736816,
      "learning_rate": 1.6969393592677347e-05,
      "loss": 1.011,
      "step": 3037
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.606213569641113,
      "learning_rate": 1.6965102974828376e-05,
      "loss": 1.1867,
      "step": 3038
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.003400802612305,
      "learning_rate": 1.6960812356979405e-05,
      "loss": 1.0107,
      "step": 3039
    },
    {
      "epoch": 0.87,
      "grad_norm": 8.365703582763672,
      "learning_rate": 1.6956521739130433e-05,
      "loss": 0.8612,
      "step": 3040
    },
    {
      "epoch": 0.87,
      "grad_norm": 8.514734268188477,
      "learning_rate": 1.6952231121281465e-05,
      "loss": 0.8481,
      "step": 3041
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.031387329101562,
      "learning_rate": 1.6947940503432498e-05,
      "loss": 0.8515,
      "step": 3042
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.178349494934082,
      "learning_rate": 1.6943649885583523e-05,
      "loss": 1.3057,
      "step": 3043
    },
    {
      "epoch": 0.87,
      "grad_norm": 8.990435600280762,
      "learning_rate": 1.6939359267734555e-05,
      "loss": 0.7206,
      "step": 3044
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.12608814239502,
      "learning_rate": 1.6935068649885584e-05,
      "loss": 0.9431,
      "step": 3045
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.1451416015625,
      "learning_rate": 1.6930778032036616e-05,
      "loss": 0.8445,
      "step": 3046
    },
    {
      "epoch": 0.87,
      "grad_norm": 8.905163764953613,
      "learning_rate": 1.692648741418764e-05,
      "loss": 1.088,
      "step": 3047
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.993717193603516,
      "learning_rate": 1.6922196796338674e-05,
      "loss": 1.2007,
      "step": 3048
    },
    {
      "epoch": 0.87,
      "grad_norm": 11.483057975769043,
      "learning_rate": 1.6917906178489702e-05,
      "loss": 1.0004,
      "step": 3049
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.861003875732422,
      "learning_rate": 1.691361556064073e-05,
      "loss": 1.049,
      "step": 3050
    },
    {
      "epoch": 0.87,
      "grad_norm": 8.530832290649414,
      "learning_rate": 1.6909324942791763e-05,
      "loss": 0.7936,
      "step": 3051
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.132295608520508,
      "learning_rate": 1.6905034324942792e-05,
      "loss": 0.6706,
      "step": 3052
    },
    {
      "epoch": 0.87,
      "grad_norm": 11.055174827575684,
      "learning_rate": 1.6900743707093824e-05,
      "loss": 0.957,
      "step": 3053
    },
    {
      "epoch": 0.87,
      "grad_norm": 11.633831977844238,
      "learning_rate": 1.689645308924485e-05,
      "loss": 1.2538,
      "step": 3054
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.796125411987305,
      "learning_rate": 1.6892162471395882e-05,
      "loss": 1.0641,
      "step": 3055
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.294535636901855,
      "learning_rate": 1.688787185354691e-05,
      "loss": 0.9855,
      "step": 3056
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.125809669494629,
      "learning_rate": 1.6883581235697943e-05,
      "loss": 0.9899,
      "step": 3057
    },
    {
      "epoch": 0.87,
      "grad_norm": 8.996604919433594,
      "learning_rate": 1.6879290617848968e-05,
      "loss": 0.9472,
      "step": 3058
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.40684986114502,
      "learning_rate": 1.6875e-05,
      "loss": 1.0968,
      "step": 3059
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.77299976348877,
      "learning_rate": 1.6870709382151032e-05,
      "loss": 1.3063,
      "step": 3060
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.238236427307129,
      "learning_rate": 1.686641876430206e-05,
      "loss": 1.0473,
      "step": 3061
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.003403663635254,
      "learning_rate": 1.686212814645309e-05,
      "loss": 1.09,
      "step": 3062
    },
    {
      "epoch": 0.88,
      "grad_norm": 11.033190727233887,
      "learning_rate": 1.685783752860412e-05,
      "loss": 1.3033,
      "step": 3063
    },
    {
      "epoch": 0.88,
      "grad_norm": 12.842031478881836,
      "learning_rate": 1.685354691075515e-05,
      "loss": 1.1193,
      "step": 3064
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.233874320983887,
      "learning_rate": 1.6849256292906176e-05,
      "loss": 0.924,
      "step": 3065
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.496210098266602,
      "learning_rate": 1.684496567505721e-05,
      "loss": 1.2856,
      "step": 3066
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.07898998260498,
      "learning_rate": 1.6840675057208237e-05,
      "loss": 0.8656,
      "step": 3067
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.06139850616455,
      "learning_rate": 1.683638443935927e-05,
      "loss": 0.9459,
      "step": 3068
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.527409553527832,
      "learning_rate": 1.6832093821510298e-05,
      "loss": 0.9977,
      "step": 3069
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.252501487731934,
      "learning_rate": 1.6827803203661327e-05,
      "loss": 1.0964,
      "step": 3070
    },
    {
      "epoch": 0.88,
      "grad_norm": 11.455581665039062,
      "learning_rate": 1.682351258581236e-05,
      "loss": 1.286,
      "step": 3071
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.602471351623535,
      "learning_rate": 1.6819221967963388e-05,
      "loss": 1.1101,
      "step": 3072
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.360219955444336,
      "learning_rate": 1.6814931350114417e-05,
      "loss": 0.9595,
      "step": 3073
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.391324043273926,
      "learning_rate": 1.6810640732265445e-05,
      "loss": 0.8692,
      "step": 3074
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.762292861938477,
      "learning_rate": 1.6806350114416477e-05,
      "loss": 1.068,
      "step": 3075
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.805365562438965,
      "learning_rate": 1.6802059496567503e-05,
      "loss": 1.0729,
      "step": 3076
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.31454086303711,
      "learning_rate": 1.6797768878718535e-05,
      "loss": 0.7873,
      "step": 3077
    },
    {
      "epoch": 0.88,
      "grad_norm": 11.559967041015625,
      "learning_rate": 1.6793478260869567e-05,
      "loss": 1.0795,
      "step": 3078
    },
    {
      "epoch": 0.88,
      "grad_norm": 11.343534469604492,
      "learning_rate": 1.6789187643020596e-05,
      "loss": 1.1779,
      "step": 3079
    },
    {
      "epoch": 0.88,
      "grad_norm": 7.209791660308838,
      "learning_rate": 1.6784897025171625e-05,
      "loss": 0.788,
      "step": 3080
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.449609756469727,
      "learning_rate": 1.6780606407322653e-05,
      "loss": 0.9448,
      "step": 3081
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.622262001037598,
      "learning_rate": 1.6776315789473686e-05,
      "loss": 0.788,
      "step": 3082
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.25172233581543,
      "learning_rate": 1.6772025171624714e-05,
      "loss": 0.9932,
      "step": 3083
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.239367485046387,
      "learning_rate": 1.6767734553775743e-05,
      "loss": 0.8375,
      "step": 3084
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.967516899108887,
      "learning_rate": 1.6763443935926775e-05,
      "loss": 0.8182,
      "step": 3085
    },
    {
      "epoch": 0.88,
      "grad_norm": 12.356708526611328,
      "learning_rate": 1.6759153318077804e-05,
      "loss": 1.0941,
      "step": 3086
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.643538475036621,
      "learning_rate": 1.6754862700228836e-05,
      "loss": 1.0097,
      "step": 3087
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.382425308227539,
      "learning_rate": 1.675057208237986e-05,
      "loss": 0.9607,
      "step": 3088
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.312171936035156,
      "learning_rate": 1.6746281464530894e-05,
      "loss": 1.2677,
      "step": 3089
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.411649703979492,
      "learning_rate": 1.6741990846681923e-05,
      "loss": 0.9854,
      "step": 3090
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.293243408203125,
      "learning_rate": 1.673770022883295e-05,
      "loss": 1.0462,
      "step": 3091
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.277262687683105,
      "learning_rate": 1.673340961098398e-05,
      "loss": 1.1064,
      "step": 3092
    },
    {
      "epoch": 0.88,
      "grad_norm": 12.135811805725098,
      "learning_rate": 1.6729118993135012e-05,
      "loss": 0.8687,
      "step": 3093
    },
    {
      "epoch": 0.89,
      "grad_norm": 12.491722106933594,
      "learning_rate": 1.6724828375286044e-05,
      "loss": 1.0355,
      "step": 3094
    },
    {
      "epoch": 0.89,
      "grad_norm": 7.733823299407959,
      "learning_rate": 1.672053775743707e-05,
      "loss": 0.7278,
      "step": 3095
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.659414291381836,
      "learning_rate": 1.6716247139588102e-05,
      "loss": 1.0786,
      "step": 3096
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.9557466506958,
      "learning_rate": 1.671195652173913e-05,
      "loss": 1.2499,
      "step": 3097
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.577733039855957,
      "learning_rate": 1.6707665903890163e-05,
      "loss": 0.9821,
      "step": 3098
    },
    {
      "epoch": 0.89,
      "grad_norm": 8.668160438537598,
      "learning_rate": 1.6703375286041188e-05,
      "loss": 0.8874,
      "step": 3099
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.793705940246582,
      "learning_rate": 1.669908466819222e-05,
      "loss": 0.9216,
      "step": 3100
    },
    {
      "epoch": 0.89,
      "grad_norm": 8.16748332977295,
      "learning_rate": 1.669479405034325e-05,
      "loss": 0.7774,
      "step": 3101
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.827296257019043,
      "learning_rate": 1.6690503432494278e-05,
      "loss": 0.965,
      "step": 3102
    },
    {
      "epoch": 0.89,
      "grad_norm": 11.990835189819336,
      "learning_rate": 1.668621281464531e-05,
      "loss": 0.8418,
      "step": 3103
    },
    {
      "epoch": 0.89,
      "grad_norm": 12.614843368530273,
      "learning_rate": 1.668192219679634e-05,
      "loss": 0.8315,
      "step": 3104
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.358224868774414,
      "learning_rate": 1.667763157894737e-05,
      "loss": 1.1432,
      "step": 3105
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.044696807861328,
      "learning_rate": 1.6673340961098396e-05,
      "loss": 1.0201,
      "step": 3106
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.316402435302734,
      "learning_rate": 1.666905034324943e-05,
      "loss": 0.946,
      "step": 3107
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.924317359924316,
      "learning_rate": 1.6664759725400457e-05,
      "loss": 1.1153,
      "step": 3108
    },
    {
      "epoch": 0.89,
      "grad_norm": 8.860552787780762,
      "learning_rate": 1.666046910755149e-05,
      "loss": 1.0023,
      "step": 3109
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.352194786071777,
      "learning_rate": 1.6656178489702515e-05,
      "loss": 0.8491,
      "step": 3110
    },
    {
      "epoch": 0.89,
      "grad_norm": 8.415346145629883,
      "learning_rate": 1.6651887871853547e-05,
      "loss": 0.6951,
      "step": 3111
    },
    {
      "epoch": 0.89,
      "grad_norm": 7.790891647338867,
      "learning_rate": 1.664759725400458e-05,
      "loss": 0.9119,
      "step": 3112
    },
    {
      "epoch": 0.89,
      "grad_norm": 8.764174461364746,
      "learning_rate": 1.6643306636155608e-05,
      "loss": 1.2364,
      "step": 3113
    },
    {
      "epoch": 0.89,
      "grad_norm": 8.090877532958984,
      "learning_rate": 1.6639016018306637e-05,
      "loss": 0.8456,
      "step": 3114
    },
    {
      "epoch": 0.89,
      "grad_norm": 8.34622859954834,
      "learning_rate": 1.6634725400457665e-05,
      "loss": 0.8178,
      "step": 3115
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.495920181274414,
      "learning_rate": 1.6630434782608698e-05,
      "loss": 1.2764,
      "step": 3116
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.262067794799805,
      "learning_rate": 1.6626144164759723e-05,
      "loss": 1.2381,
      "step": 3117
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.799932479858398,
      "learning_rate": 1.6621853546910755e-05,
      "loss": 1.1879,
      "step": 3118
    },
    {
      "epoch": 0.89,
      "grad_norm": 12.72973346710205,
      "learning_rate": 1.6617562929061787e-05,
      "loss": 1.2686,
      "step": 3119
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.531723976135254,
      "learning_rate": 1.6613272311212816e-05,
      "loss": 1.1524,
      "step": 3120
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.594547271728516,
      "learning_rate": 1.6608981693363845e-05,
      "loss": 1.0767,
      "step": 3121
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.052020072937012,
      "learning_rate": 1.6604691075514874e-05,
      "loss": 0.8348,
      "step": 3122
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.655191421508789,
      "learning_rate": 1.6600400457665906e-05,
      "loss": 1.0825,
      "step": 3123
    },
    {
      "epoch": 0.89,
      "grad_norm": 11.176694869995117,
      "learning_rate": 1.6596109839816935e-05,
      "loss": 1.1807,
      "step": 3124
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.434635162353516,
      "learning_rate": 1.6591819221967963e-05,
      "loss": 0.836,
      "step": 3125
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.345967292785645,
      "learning_rate": 1.6587528604118992e-05,
      "loss": 1.0104,
      "step": 3126
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.331924438476562,
      "learning_rate": 1.6583237986270024e-05,
      "loss": 1.0878,
      "step": 3127
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.606528282165527,
      "learning_rate": 1.6578947368421053e-05,
      "loss": 1.068,
      "step": 3128
    },
    {
      "epoch": 0.9,
      "grad_norm": 7.378060817718506,
      "learning_rate": 1.6574656750572082e-05,
      "loss": 0.7387,
      "step": 3129
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.276107788085938,
      "learning_rate": 1.6570366132723114e-05,
      "loss": 1.0986,
      "step": 3130
    },
    {
      "epoch": 0.9,
      "grad_norm": 7.8492631912231445,
      "learning_rate": 1.6566075514874143e-05,
      "loss": 1.0775,
      "step": 3131
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.005735397338867,
      "learning_rate": 1.656178489702517e-05,
      "loss": 1.2657,
      "step": 3132
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.524003982543945,
      "learning_rate": 1.65574942791762e-05,
      "loss": 1.1501,
      "step": 3133
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.842630386352539,
      "learning_rate": 1.6553203661327232e-05,
      "loss": 1.3252,
      "step": 3134
    },
    {
      "epoch": 0.9,
      "grad_norm": 8.478659629821777,
      "learning_rate": 1.654891304347826e-05,
      "loss": 0.8981,
      "step": 3135
    },
    {
      "epoch": 0.9,
      "grad_norm": 8.528156280517578,
      "learning_rate": 1.654462242562929e-05,
      "loss": 0.9483,
      "step": 3136
    },
    {
      "epoch": 0.9,
      "grad_norm": 8.961124420166016,
      "learning_rate": 1.6540331807780322e-05,
      "loss": 1.2126,
      "step": 3137
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.386552810668945,
      "learning_rate": 1.653604118993135e-05,
      "loss": 0.9615,
      "step": 3138
    },
    {
      "epoch": 0.9,
      "grad_norm": 8.276663780212402,
      "learning_rate": 1.6531750572082383e-05,
      "loss": 0.8203,
      "step": 3139
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.044401168823242,
      "learning_rate": 1.652745995423341e-05,
      "loss": 1.0648,
      "step": 3140
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.254098892211914,
      "learning_rate": 1.652316933638444e-05,
      "loss": 1.0039,
      "step": 3141
    },
    {
      "epoch": 0.9,
      "grad_norm": 7.073532581329346,
      "learning_rate": 1.651887871853547e-05,
      "loss": 0.7994,
      "step": 3142
    },
    {
      "epoch": 0.9,
      "grad_norm": 8.611063003540039,
      "learning_rate": 1.6514588100686498e-05,
      "loss": 1.0628,
      "step": 3143
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.973498344421387,
      "learning_rate": 1.6510297482837527e-05,
      "loss": 0.9726,
      "step": 3144
    },
    {
      "epoch": 0.9,
      "grad_norm": 8.836945533752441,
      "learning_rate": 1.650600686498856e-05,
      "loss": 0.9552,
      "step": 3145
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.754205703735352,
      "learning_rate": 1.650171624713959e-05,
      "loss": 0.8553,
      "step": 3146
    },
    {
      "epoch": 0.9,
      "grad_norm": 8.891960144042969,
      "learning_rate": 1.6497425629290616e-05,
      "loss": 0.9489,
      "step": 3147
    },
    {
      "epoch": 0.9,
      "grad_norm": 11.262051582336426,
      "learning_rate": 1.649313501144165e-05,
      "loss": 0.9542,
      "step": 3148
    },
    {
      "epoch": 0.9,
      "grad_norm": 7.30917501449585,
      "learning_rate": 1.6488844393592677e-05,
      "loss": 0.8196,
      "step": 3149
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.353693008422852,
      "learning_rate": 1.648455377574371e-05,
      "loss": 0.9017,
      "step": 3150
    },
    {
      "epoch": 0.9,
      "grad_norm": 11.002487182617188,
      "learning_rate": 1.6480263157894735e-05,
      "loss": 0.8348,
      "step": 3151
    },
    {
      "epoch": 0.9,
      "grad_norm": 10.591752052307129,
      "learning_rate": 1.6475972540045767e-05,
      "loss": 0.7949,
      "step": 3152
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.392352104187012,
      "learning_rate": 1.6471681922196796e-05,
      "loss": 0.8932,
      "step": 3153
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.849568367004395,
      "learning_rate": 1.6467391304347828e-05,
      "loss": 1.1319,
      "step": 3154
    },
    {
      "epoch": 0.9,
      "grad_norm": 11.07901668548584,
      "learning_rate": 1.6463100686498857e-05,
      "loss": 1.2838,
      "step": 3155
    },
    {
      "epoch": 0.9,
      "grad_norm": 10.085216522216797,
      "learning_rate": 1.6458810068649886e-05,
      "loss": 0.9914,
      "step": 3156
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.623714447021484,
      "learning_rate": 1.6454519450800918e-05,
      "loss": 0.9696,
      "step": 3157
    },
    {
      "epoch": 0.9,
      "grad_norm": 8.515621185302734,
      "learning_rate": 1.6450228832951943e-05,
      "loss": 0.8009,
      "step": 3158
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.103911399841309,
      "learning_rate": 1.6445938215102975e-05,
      "loss": 0.8285,
      "step": 3159
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.101774215698242,
      "learning_rate": 1.6441647597254004e-05,
      "loss": 0.7519,
      "step": 3160
    },
    {
      "epoch": 0.9,
      "grad_norm": 8.230019569396973,
      "learning_rate": 1.6437356979405036e-05,
      "loss": 0.9152,
      "step": 3161
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.370148658752441,
      "learning_rate": 1.6433066361556065e-05,
      "loss": 0.8063,
      "step": 3162
    },
    {
      "epoch": 0.9,
      "grad_norm": 10.779153823852539,
      "learning_rate": 1.6428775743707094e-05,
      "loss": 1.3406,
      "step": 3163
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.598963737487793,
      "learning_rate": 1.6424485125858126e-05,
      "loss": 0.8718,
      "step": 3164
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.773757934570312,
      "learning_rate": 1.6420194508009155e-05,
      "loss": 1.1121,
      "step": 3165
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.981141090393066,
      "learning_rate": 1.6415903890160183e-05,
      "loss": 0.881,
      "step": 3166
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.617985725402832,
      "learning_rate": 1.6411613272311212e-05,
      "loss": 1.0474,
      "step": 3167
    },
    {
      "epoch": 0.91,
      "grad_norm": 8.892661094665527,
      "learning_rate": 1.6407322654462244e-05,
      "loss": 1.0152,
      "step": 3168
    },
    {
      "epoch": 0.91,
      "grad_norm": 8.366165161132812,
      "learning_rate": 1.640303203661327e-05,
      "loss": 1.1588,
      "step": 3169
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.079965591430664,
      "learning_rate": 1.6398741418764302e-05,
      "loss": 0.9844,
      "step": 3170
    },
    {
      "epoch": 0.91,
      "grad_norm": 7.300412178039551,
      "learning_rate": 1.6394450800915334e-05,
      "loss": 0.7623,
      "step": 3171
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.555304527282715,
      "learning_rate": 1.6390160183066363e-05,
      "loss": 0.8854,
      "step": 3172
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.230918884277344,
      "learning_rate": 1.638586956521739e-05,
      "loss": 1.1895,
      "step": 3173
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.42735481262207,
      "learning_rate": 1.638157894736842e-05,
      "loss": 1.2141,
      "step": 3174
    },
    {
      "epoch": 0.91,
      "grad_norm": 11.082991600036621,
      "learning_rate": 1.6377288329519452e-05,
      "loss": 1.0251,
      "step": 3175
    },
    {
      "epoch": 0.91,
      "grad_norm": 8.29589557647705,
      "learning_rate": 1.637299771167048e-05,
      "loss": 1.1099,
      "step": 3176
    },
    {
      "epoch": 0.91,
      "grad_norm": 8.441349029541016,
      "learning_rate": 1.636870709382151e-05,
      "loss": 0.9099,
      "step": 3177
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.149102210998535,
      "learning_rate": 1.636441647597254e-05,
      "loss": 1.2221,
      "step": 3178
    },
    {
      "epoch": 0.91,
      "grad_norm": 11.996493339538574,
      "learning_rate": 1.636012585812357e-05,
      "loss": 1.2849,
      "step": 3179
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.45579719543457,
      "learning_rate": 1.6355835240274603e-05,
      "loss": 1.152,
      "step": 3180
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.456029891967773,
      "learning_rate": 1.635154462242563e-05,
      "loss": 1.0496,
      "step": 3181
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.552337646484375,
      "learning_rate": 1.634725400457666e-05,
      "loss": 0.7783,
      "step": 3182
    },
    {
      "epoch": 0.91,
      "grad_norm": 8.918163299560547,
      "learning_rate": 1.634296338672769e-05,
      "loss": 1.0866,
      "step": 3183
    },
    {
      "epoch": 0.91,
      "grad_norm": 8.836033821105957,
      "learning_rate": 1.6338672768878718e-05,
      "loss": 1.0125,
      "step": 3184
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.861259460449219,
      "learning_rate": 1.6334382151029747e-05,
      "loss": 1.1258,
      "step": 3185
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.794419288635254,
      "learning_rate": 1.633009153318078e-05,
      "loss": 1.0022,
      "step": 3186
    },
    {
      "epoch": 0.91,
      "grad_norm": 7.860095500946045,
      "learning_rate": 1.6325800915331808e-05,
      "loss": 0.9663,
      "step": 3187
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.200760841369629,
      "learning_rate": 1.6321510297482837e-05,
      "loss": 0.9666,
      "step": 3188
    },
    {
      "epoch": 0.91,
      "grad_norm": 7.224365234375,
      "learning_rate": 1.631721967963387e-05,
      "loss": 0.7837,
      "step": 3189
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.976065635681152,
      "learning_rate": 1.6312929061784898e-05,
      "loss": 1.1979,
      "step": 3190
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.322733879089355,
      "learning_rate": 1.630863844393593e-05,
      "loss": 0.828,
      "step": 3191
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.57866096496582,
      "learning_rate": 1.6304347826086955e-05,
      "loss": 0.9835,
      "step": 3192
    },
    {
      "epoch": 0.91,
      "grad_norm": 8.49717903137207,
      "learning_rate": 1.6300057208237987e-05,
      "loss": 0.9419,
      "step": 3193
    },
    {
      "epoch": 0.91,
      "grad_norm": 7.884578227996826,
      "learning_rate": 1.6295766590389016e-05,
      "loss": 0.6729,
      "step": 3194
    },
    {
      "epoch": 0.91,
      "grad_norm": 8.000340461730957,
      "learning_rate": 1.6291475972540045e-05,
      "loss": 0.8283,
      "step": 3195
    },
    {
      "epoch": 0.91,
      "grad_norm": 12.245285987854004,
      "learning_rate": 1.6287185354691074e-05,
      "loss": 1.363,
      "step": 3196
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.48442268371582,
      "learning_rate": 1.6282894736842106e-05,
      "loss": 0.944,
      "step": 3197
    },
    {
      "epoch": 0.91,
      "grad_norm": 8.397787094116211,
      "learning_rate": 1.6278604118993138e-05,
      "loss": 0.9394,
      "step": 3198
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.015027046203613,
      "learning_rate": 1.6274313501144163e-05,
      "loss": 0.8126,
      "step": 3199
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.795433044433594,
      "learning_rate": 1.6270022883295195e-05,
      "loss": 1.1504,
      "step": 3200
    },
    {
      "epoch": 0.92,
      "grad_norm": 8.850866317749023,
      "learning_rate": 1.6265732265446224e-05,
      "loss": 1.0913,
      "step": 3201
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.196776390075684,
      "learning_rate": 1.6261441647597256e-05,
      "loss": 0.9343,
      "step": 3202
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.463024139404297,
      "learning_rate": 1.6257151029748282e-05,
      "loss": 1.1631,
      "step": 3203
    },
    {
      "epoch": 0.92,
      "grad_norm": 12.322643280029297,
      "learning_rate": 1.6252860411899314e-05,
      "loss": 1.1612,
      "step": 3204
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.773918151855469,
      "learning_rate": 1.6248569794050346e-05,
      "loss": 1.2507,
      "step": 3205
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.977486610412598,
      "learning_rate": 1.6244279176201375e-05,
      "loss": 0.9428,
      "step": 3206
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.164650917053223,
      "learning_rate": 1.6239988558352404e-05,
      "loss": 0.9418,
      "step": 3207
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.065773010253906,
      "learning_rate": 1.6235697940503432e-05,
      "loss": 0.9398,
      "step": 3208
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.81015682220459,
      "learning_rate": 1.6231407322654464e-05,
      "loss": 0.8495,
      "step": 3209
    },
    {
      "epoch": 0.92,
      "grad_norm": 8.081978797912598,
      "learning_rate": 1.622711670480549e-05,
      "loss": 0.8546,
      "step": 3210
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.246928215026855,
      "learning_rate": 1.6222826086956522e-05,
      "loss": 0.8802,
      "step": 3211
    },
    {
      "epoch": 0.92,
      "grad_norm": 11.4380521774292,
      "learning_rate": 1.621853546910755e-05,
      "loss": 1.0022,
      "step": 3212
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.336760520935059,
      "learning_rate": 1.6214244851258583e-05,
      "loss": 1.0296,
      "step": 3213
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.019309997558594,
      "learning_rate": 1.6209954233409612e-05,
      "loss": 0.8747,
      "step": 3214
    },
    {
      "epoch": 0.92,
      "grad_norm": 8.206274032592773,
      "learning_rate": 1.620566361556064e-05,
      "loss": 0.9604,
      "step": 3215
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.94852066040039,
      "learning_rate": 1.6201372997711673e-05,
      "loss": 0.9252,
      "step": 3216
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.24874496459961,
      "learning_rate": 1.61970823798627e-05,
      "loss": 0.9816,
      "step": 3217
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.955107688903809,
      "learning_rate": 1.619279176201373e-05,
      "loss": 1.4561,
      "step": 3218
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.70648193359375,
      "learning_rate": 1.618850114416476e-05,
      "loss": 1.0059,
      "step": 3219
    },
    {
      "epoch": 0.92,
      "grad_norm": 11.166799545288086,
      "learning_rate": 1.618421052631579e-05,
      "loss": 1.0511,
      "step": 3220
    },
    {
      "epoch": 0.92,
      "grad_norm": 7.510572910308838,
      "learning_rate": 1.6179919908466816e-05,
      "loss": 0.8416,
      "step": 3221
    },
    {
      "epoch": 0.92,
      "grad_norm": 8.313467979431152,
      "learning_rate": 1.617562929061785e-05,
      "loss": 0.9154,
      "step": 3222
    },
    {
      "epoch": 0.92,
      "grad_norm": 8.768488883972168,
      "learning_rate": 1.617133867276888e-05,
      "loss": 0.7979,
      "step": 3223
    },
    {
      "epoch": 0.92,
      "grad_norm": 8.637980461120605,
      "learning_rate": 1.616704805491991e-05,
      "loss": 0.9436,
      "step": 3224
    },
    {
      "epoch": 0.92,
      "grad_norm": 11.256755828857422,
      "learning_rate": 1.616275743707094e-05,
      "loss": 1.0568,
      "step": 3225
    },
    {
      "epoch": 0.92,
      "grad_norm": 11.269331932067871,
      "learning_rate": 1.6158466819221967e-05,
      "loss": 1.2881,
      "step": 3226
    },
    {
      "epoch": 0.92,
      "grad_norm": 8.064380645751953,
      "learning_rate": 1.6154176201373e-05,
      "loss": 0.8013,
      "step": 3227
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.082005500793457,
      "learning_rate": 1.6149885583524028e-05,
      "loss": 1.0026,
      "step": 3228
    },
    {
      "epoch": 0.92,
      "grad_norm": 8.47630500793457,
      "learning_rate": 1.6145594965675057e-05,
      "loss": 1.1559,
      "step": 3229
    },
    {
      "epoch": 0.92,
      "grad_norm": 8.413256645202637,
      "learning_rate": 1.6141304347826086e-05,
      "loss": 0.9758,
      "step": 3230
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.38217830657959,
      "learning_rate": 1.6137013729977118e-05,
      "loss": 1.1137,
      "step": 3231
    },
    {
      "epoch": 0.92,
      "grad_norm": 8.091568946838379,
      "learning_rate": 1.613272311212815e-05,
      "loss": 1.0385,
      "step": 3232
    },
    {
      "epoch": 0.92,
      "grad_norm": 11.839828491210938,
      "learning_rate": 1.6128432494279175e-05,
      "loss": 1.0217,
      "step": 3233
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.954550743103027,
      "learning_rate": 1.6124141876430207e-05,
      "loss": 0.9288,
      "step": 3234
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.699662208557129,
      "learning_rate": 1.6119851258581236e-05,
      "loss": 1.166,
      "step": 3235
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.94688892364502,
      "learning_rate": 1.6115560640732265e-05,
      "loss": 0.8658,
      "step": 3236
    },
    {
      "epoch": 0.93,
      "grad_norm": 8.275080680847168,
      "learning_rate": 1.6111270022883294e-05,
      "loss": 0.801,
      "step": 3237
    },
    {
      "epoch": 0.93,
      "grad_norm": 7.638550758361816,
      "learning_rate": 1.6106979405034326e-05,
      "loss": 0.8034,
      "step": 3238
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.947699546813965,
      "learning_rate": 1.6102688787185358e-05,
      "loss": 1.234,
      "step": 3239
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.809272766113281,
      "learning_rate": 1.6098398169336383e-05,
      "loss": 1.0888,
      "step": 3240
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.042963027954102,
      "learning_rate": 1.6094107551487416e-05,
      "loss": 1.2458,
      "step": 3241
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.463010787963867,
      "learning_rate": 1.6089816933638444e-05,
      "loss": 0.8898,
      "step": 3242
    },
    {
      "epoch": 0.93,
      "grad_norm": 10.513991355895996,
      "learning_rate": 1.6085526315789476e-05,
      "loss": 1.0433,
      "step": 3243
    },
    {
      "epoch": 0.93,
      "grad_norm": 11.259737968444824,
      "learning_rate": 1.6081235697940502e-05,
      "loss": 1.1836,
      "step": 3244
    },
    {
      "epoch": 0.93,
      "grad_norm": 11.464911460876465,
      "learning_rate": 1.6076945080091534e-05,
      "loss": 1.3609,
      "step": 3245
    },
    {
      "epoch": 0.93,
      "grad_norm": 7.19293212890625,
      "learning_rate": 1.6072654462242563e-05,
      "loss": 0.8884,
      "step": 3246
    },
    {
      "epoch": 0.93,
      "grad_norm": 7.9146037101745605,
      "learning_rate": 1.606836384439359e-05,
      "loss": 0.7844,
      "step": 3247
    },
    {
      "epoch": 0.93,
      "grad_norm": 8.42154598236084,
      "learning_rate": 1.6064073226544624e-05,
      "loss": 0.9371,
      "step": 3248
    },
    {
      "epoch": 0.93,
      "grad_norm": 8.096179008483887,
      "learning_rate": 1.6059782608695652e-05,
      "loss": 0.8173,
      "step": 3249
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.047220230102539,
      "learning_rate": 1.6055491990846685e-05,
      "loss": 1.1358,
      "step": 3250
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.304205894470215,
      "learning_rate": 1.605120137299771e-05,
      "loss": 0.8731,
      "step": 3251
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.897394180297852,
      "learning_rate": 1.6046910755148742e-05,
      "loss": 1.0724,
      "step": 3252
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.924748420715332,
      "learning_rate": 1.604262013729977e-05,
      "loss": 0.9685,
      "step": 3253
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.466620445251465,
      "learning_rate": 1.6038329519450803e-05,
      "loss": 0.8894,
      "step": 3254
    },
    {
      "epoch": 0.93,
      "grad_norm": 8.961726188659668,
      "learning_rate": 1.603403890160183e-05,
      "loss": 0.9971,
      "step": 3255
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.545747756958008,
      "learning_rate": 1.602974828375286e-05,
      "loss": 0.8442,
      "step": 3256
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.017826080322266,
      "learning_rate": 1.6025457665903893e-05,
      "loss": 0.9271,
      "step": 3257
    },
    {
      "epoch": 0.93,
      "grad_norm": 7.684579372406006,
      "learning_rate": 1.602116704805492e-05,
      "loss": 0.547,
      "step": 3258
    },
    {
      "epoch": 0.93,
      "grad_norm": 8.903583526611328,
      "learning_rate": 1.601687643020595e-05,
      "loss": 0.9456,
      "step": 3259
    },
    {
      "epoch": 0.93,
      "grad_norm": 11.6437349319458,
      "learning_rate": 1.601258581235698e-05,
      "loss": 1.0433,
      "step": 3260
    },
    {
      "epoch": 0.93,
      "grad_norm": 8.919283866882324,
      "learning_rate": 1.600829519450801e-05,
      "loss": 0.9232,
      "step": 3261
    },
    {
      "epoch": 0.93,
      "grad_norm": 11.192386627197266,
      "learning_rate": 1.6004004576659037e-05,
      "loss": 1.0861,
      "step": 3262
    },
    {
      "epoch": 0.93,
      "grad_norm": 10.518370628356934,
      "learning_rate": 1.599971395881007e-05,
      "loss": 1.1146,
      "step": 3263
    },
    {
      "epoch": 0.93,
      "grad_norm": 10.677868843078613,
      "learning_rate": 1.5995423340961098e-05,
      "loss": 1.271,
      "step": 3264
    },
    {
      "epoch": 0.93,
      "grad_norm": 10.702676773071289,
      "learning_rate": 1.599113272311213e-05,
      "loss": 1.2026,
      "step": 3265
    },
    {
      "epoch": 0.93,
      "grad_norm": 8.851753234863281,
      "learning_rate": 1.598684210526316e-05,
      "loss": 0.9492,
      "step": 3266
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.764097213745117,
      "learning_rate": 1.5982551487414187e-05,
      "loss": 0.9993,
      "step": 3267
    },
    {
      "epoch": 0.93,
      "grad_norm": 7.402448654174805,
      "learning_rate": 1.597826086956522e-05,
      "loss": 0.7815,
      "step": 3268
    },
    {
      "epoch": 0.94,
      "grad_norm": 9.05008316040039,
      "learning_rate": 1.5973970251716248e-05,
      "loss": 0.7609,
      "step": 3269
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.128369331359863,
      "learning_rate": 1.5969679633867277e-05,
      "loss": 1.1545,
      "step": 3270
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.443517684936523,
      "learning_rate": 1.5965389016018306e-05,
      "loss": 1.0197,
      "step": 3271
    },
    {
      "epoch": 0.94,
      "grad_norm": 11.30739688873291,
      "learning_rate": 1.5961098398169338e-05,
      "loss": 1.1871,
      "step": 3272
    },
    {
      "epoch": 0.94,
      "grad_norm": 9.866897583007812,
      "learning_rate": 1.5956807780320367e-05,
      "loss": 0.9183,
      "step": 3273
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.537786483764648,
      "learning_rate": 1.5952517162471395e-05,
      "loss": 1.3296,
      "step": 3274
    },
    {
      "epoch": 0.94,
      "grad_norm": 8.425891876220703,
      "learning_rate": 1.5948226544622428e-05,
      "loss": 0.8341,
      "step": 3275
    },
    {
      "epoch": 0.94,
      "grad_norm": 8.038627624511719,
      "learning_rate": 1.5943935926773456e-05,
      "loss": 0.8743,
      "step": 3276
    },
    {
      "epoch": 0.94,
      "grad_norm": 8.650530815124512,
      "learning_rate": 1.5939645308924485e-05,
      "loss": 0.9197,
      "step": 3277
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.73486042022705,
      "learning_rate": 1.5935354691075514e-05,
      "loss": 1.2104,
      "step": 3278
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.41913890838623,
      "learning_rate": 1.5931064073226546e-05,
      "loss": 1.0029,
      "step": 3279
    },
    {
      "epoch": 0.94,
      "grad_norm": 126.80941772460938,
      "learning_rate": 1.5926773455377575e-05,
      "loss": 1.2048,
      "step": 3280
    },
    {
      "epoch": 0.94,
      "grad_norm": 11.67647933959961,
      "learning_rate": 1.5922482837528604e-05,
      "loss": 1.3911,
      "step": 3281
    },
    {
      "epoch": 0.94,
      "grad_norm": 11.207754135131836,
      "learning_rate": 1.5918192219679636e-05,
      "loss": 1.153,
      "step": 3282
    },
    {
      "epoch": 0.94,
      "grad_norm": 11.849954605102539,
      "learning_rate": 1.5913901601830664e-05,
      "loss": 1.009,
      "step": 3283
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.543679237365723,
      "learning_rate": 1.5909610983981697e-05,
      "loss": 0.8047,
      "step": 3284
    },
    {
      "epoch": 0.94,
      "grad_norm": 9.641677856445312,
      "learning_rate": 1.5905320366132722e-05,
      "loss": 0.6823,
      "step": 3285
    },
    {
      "epoch": 0.94,
      "grad_norm": 12.406120300292969,
      "learning_rate": 1.5901029748283754e-05,
      "loss": 1.2721,
      "step": 3286
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.462618827819824,
      "learning_rate": 1.5896739130434783e-05,
      "loss": 1.0017,
      "step": 3287
    },
    {
      "epoch": 0.94,
      "grad_norm": 8.184019088745117,
      "learning_rate": 1.589244851258581e-05,
      "loss": 0.9553,
      "step": 3288
    },
    {
      "epoch": 0.94,
      "grad_norm": 7.911831855773926,
      "learning_rate": 1.588815789473684e-05,
      "loss": 0.8236,
      "step": 3289
    },
    {
      "epoch": 0.94,
      "grad_norm": 8.127483367919922,
      "learning_rate": 1.5883867276887873e-05,
      "loss": 0.8052,
      "step": 3290
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.355116844177246,
      "learning_rate": 1.5879576659038905e-05,
      "loss": 0.9767,
      "step": 3291
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.917792320251465,
      "learning_rate": 1.587528604118993e-05,
      "loss": 1.0647,
      "step": 3292
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.53119945526123,
      "learning_rate": 1.5870995423340962e-05,
      "loss": 1.1415,
      "step": 3293
    },
    {
      "epoch": 0.94,
      "grad_norm": 9.361116409301758,
      "learning_rate": 1.586670480549199e-05,
      "loss": 1.0765,
      "step": 3294
    },
    {
      "epoch": 0.94,
      "grad_norm": 9.876181602478027,
      "learning_rate": 1.5862414187643023e-05,
      "loss": 1.2355,
      "step": 3295
    },
    {
      "epoch": 0.94,
      "grad_norm": 7.987552642822266,
      "learning_rate": 1.585812356979405e-05,
      "loss": 0.7288,
      "step": 3296
    },
    {
      "epoch": 0.94,
      "grad_norm": 16.646928787231445,
      "learning_rate": 1.585383295194508e-05,
      "loss": 1.0366,
      "step": 3297
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.816709518432617,
      "learning_rate": 1.584954233409611e-05,
      "loss": 1.3376,
      "step": 3298
    },
    {
      "epoch": 0.94,
      "grad_norm": 7.854706764221191,
      "learning_rate": 1.5845251716247138e-05,
      "loss": 0.9233,
      "step": 3299
    },
    {
      "epoch": 0.94,
      "grad_norm": 7.691450119018555,
      "learning_rate": 1.584096109839817e-05,
      "loss": 0.8012,
      "step": 3300
    },
    {
      "epoch": 0.94,
      "grad_norm": 9.379831314086914,
      "learning_rate": 1.58366704805492e-05,
      "loss": 0.8886,
      "step": 3301
    },
    {
      "epoch": 0.94,
      "grad_norm": 9.529094696044922,
      "learning_rate": 1.583237986270023e-05,
      "loss": 1.0188,
      "step": 3302
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.709331512451172,
      "learning_rate": 1.5828089244851257e-05,
      "loss": 1.0348,
      "step": 3303
    },
    {
      "epoch": 0.95,
      "grad_norm": 10.2379732131958,
      "learning_rate": 1.582379862700229e-05,
      "loss": 0.8678,
      "step": 3304
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.242757797241211,
      "learning_rate": 1.5819508009153318e-05,
      "loss": 1.1564,
      "step": 3305
    },
    {
      "epoch": 0.95,
      "grad_norm": 8.454890251159668,
      "learning_rate": 1.581521739130435e-05,
      "loss": 0.9922,
      "step": 3306
    },
    {
      "epoch": 0.95,
      "grad_norm": 8.57014274597168,
      "learning_rate": 1.5810926773455375e-05,
      "loss": 1.048,
      "step": 3307
    },
    {
      "epoch": 0.95,
      "grad_norm": 8.023551940917969,
      "learning_rate": 1.5806636155606407e-05,
      "loss": 0.8361,
      "step": 3308
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.4293794631958,
      "learning_rate": 1.580234553775744e-05,
      "loss": 1.1393,
      "step": 3309
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.56303882598877,
      "learning_rate": 1.5798054919908468e-05,
      "loss": 1.1764,
      "step": 3310
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.185225486755371,
      "learning_rate": 1.5793764302059497e-05,
      "loss": 0.869,
      "step": 3311
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.167194366455078,
      "learning_rate": 1.5789473684210526e-05,
      "loss": 0.9115,
      "step": 3312
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.10361385345459,
      "learning_rate": 1.5785183066361558e-05,
      "loss": 1.1696,
      "step": 3313
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.448793411254883,
      "learning_rate": 1.5780892448512583e-05,
      "loss": 0.8851,
      "step": 3314
    },
    {
      "epoch": 0.95,
      "grad_norm": 8.414605140686035,
      "learning_rate": 1.5776601830663616e-05,
      "loss": 0.7982,
      "step": 3315
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.551800727844238,
      "learning_rate": 1.5772311212814644e-05,
      "loss": 0.8693,
      "step": 3316
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.259560585021973,
      "learning_rate": 1.5768020594965676e-05,
      "loss": 0.8606,
      "step": 3317
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.371599197387695,
      "learning_rate": 1.5763729977116705e-05,
      "loss": 0.9186,
      "step": 3318
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.043257713317871,
      "learning_rate": 1.5759439359267734e-05,
      "loss": 0.7495,
      "step": 3319
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.614897727966309,
      "learning_rate": 1.5755148741418766e-05,
      "loss": 0.7925,
      "step": 3320
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.677974700927734,
      "learning_rate": 1.5750858123569795e-05,
      "loss": 1.086,
      "step": 3321
    },
    {
      "epoch": 0.95,
      "grad_norm": 11.341986656188965,
      "learning_rate": 1.5746567505720824e-05,
      "loss": 1.1042,
      "step": 3322
    },
    {
      "epoch": 0.95,
      "grad_norm": 10.899744033813477,
      "learning_rate": 1.5742276887871852e-05,
      "loss": 0.9006,
      "step": 3323
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.07625961303711,
      "learning_rate": 1.5737986270022885e-05,
      "loss": 0.9405,
      "step": 3324
    },
    {
      "epoch": 0.95,
      "grad_norm": 10.768837928771973,
      "learning_rate": 1.5733695652173913e-05,
      "loss": 1.0818,
      "step": 3325
    },
    {
      "epoch": 0.95,
      "grad_norm": 10.189533233642578,
      "learning_rate": 1.5729405034324942e-05,
      "loss": 1.2383,
      "step": 3326
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.930407524108887,
      "learning_rate": 1.5725114416475974e-05,
      "loss": 0.8071,
      "step": 3327
    },
    {
      "epoch": 0.95,
      "grad_norm": 7.634292125701904,
      "learning_rate": 1.5720823798627003e-05,
      "loss": 0.8753,
      "step": 3328
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.883581161499023,
      "learning_rate": 1.5716533180778032e-05,
      "loss": 0.8969,
      "step": 3329
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.58691120147705,
      "learning_rate": 1.571224256292906e-05,
      "loss": 1.2702,
      "step": 3330
    },
    {
      "epoch": 0.95,
      "grad_norm": 11.525691986083984,
      "learning_rate": 1.5707951945080093e-05,
      "loss": 0.9529,
      "step": 3331
    },
    {
      "epoch": 0.95,
      "grad_norm": 10.9685640335083,
      "learning_rate": 1.570366132723112e-05,
      "loss": 0.9871,
      "step": 3332
    },
    {
      "epoch": 0.95,
      "grad_norm": 10.324636459350586,
      "learning_rate": 1.569937070938215e-05,
      "loss": 1.0842,
      "step": 3333
    },
    {
      "epoch": 0.95,
      "grad_norm": 8.318097114562988,
      "learning_rate": 1.5695080091533182e-05,
      "loss": 0.8873,
      "step": 3334
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.906346321105957,
      "learning_rate": 1.569078947368421e-05,
      "loss": 0.9217,
      "step": 3335
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.697980880737305,
      "learning_rate": 1.5686498855835243e-05,
      "loss": 1.0565,
      "step": 3336
    },
    {
      "epoch": 0.95,
      "grad_norm": 8.229022026062012,
      "learning_rate": 1.568220823798627e-05,
      "loss": 0.7354,
      "step": 3337
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.6383695602417,
      "learning_rate": 1.56779176201373e-05,
      "loss": 1.015,
      "step": 3338
    },
    {
      "epoch": 0.96,
      "grad_norm": 10.653667449951172,
      "learning_rate": 1.567362700228833e-05,
      "loss": 1.5088,
      "step": 3339
    },
    {
      "epoch": 0.96,
      "grad_norm": 8.933518409729004,
      "learning_rate": 1.566933638443936e-05,
      "loss": 1.1612,
      "step": 3340
    },
    {
      "epoch": 0.96,
      "grad_norm": 10.447750091552734,
      "learning_rate": 1.5665045766590387e-05,
      "loss": 1.19,
      "step": 3341
    },
    {
      "epoch": 0.96,
      "grad_norm": 8.314963340759277,
      "learning_rate": 1.566075514874142e-05,
      "loss": 0.6001,
      "step": 3342
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.529887199401855,
      "learning_rate": 1.565646453089245e-05,
      "loss": 0.9961,
      "step": 3343
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.230010986328125,
      "learning_rate": 1.5652173913043477e-05,
      "loss": 0.7727,
      "step": 3344
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.142366409301758,
      "learning_rate": 1.564788329519451e-05,
      "loss": 0.8415,
      "step": 3345
    },
    {
      "epoch": 0.96,
      "grad_norm": 8.853986740112305,
      "learning_rate": 1.5643592677345538e-05,
      "loss": 1.0927,
      "step": 3346
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.466641426086426,
      "learning_rate": 1.563930205949657e-05,
      "loss": 1.0912,
      "step": 3347
    },
    {
      "epoch": 0.96,
      "grad_norm": 8.603524208068848,
      "learning_rate": 1.5635011441647595e-05,
      "loss": 0.9035,
      "step": 3348
    },
    {
      "epoch": 0.96,
      "grad_norm": 12.175256729125977,
      "learning_rate": 1.5630720823798628e-05,
      "loss": 1.3209,
      "step": 3349
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.855653762817383,
      "learning_rate": 1.5626430205949656e-05,
      "loss": 0.9801,
      "step": 3350
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.72407054901123,
      "learning_rate": 1.562213958810069e-05,
      "loss": 0.819,
      "step": 3351
    },
    {
      "epoch": 0.96,
      "grad_norm": 11.676117897033691,
      "learning_rate": 1.5617848970251717e-05,
      "loss": 1.2892,
      "step": 3352
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.289527893066406,
      "learning_rate": 1.5613558352402746e-05,
      "loss": 1.0813,
      "step": 3353
    },
    {
      "epoch": 0.96,
      "grad_norm": 10.125395774841309,
      "learning_rate": 1.5609267734553778e-05,
      "loss": 0.9076,
      "step": 3354
    },
    {
      "epoch": 0.96,
      "grad_norm": 10.02776050567627,
      "learning_rate": 1.5604977116704804e-05,
      "loss": 1.1723,
      "step": 3355
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.241559982299805,
      "learning_rate": 1.5600686498855836e-05,
      "loss": 0.8031,
      "step": 3356
    },
    {
      "epoch": 0.96,
      "grad_norm": 10.31569766998291,
      "learning_rate": 1.5596395881006864e-05,
      "loss": 0.8867,
      "step": 3357
    },
    {
      "epoch": 0.96,
      "grad_norm": 8.171711921691895,
      "learning_rate": 1.5592105263157897e-05,
      "loss": 1.0547,
      "step": 3358
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.938399314880371,
      "learning_rate": 1.5587814645308925e-05,
      "loss": 0.9277,
      "step": 3359
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.381302833557129,
      "learning_rate": 1.5583524027459954e-05,
      "loss": 0.9156,
      "step": 3360
    },
    {
      "epoch": 0.96,
      "grad_norm": 10.064274787902832,
      "learning_rate": 1.5579233409610986e-05,
      "loss": 0.897,
      "step": 3361
    },
    {
      "epoch": 0.96,
      "grad_norm": 8.624159812927246,
      "learning_rate": 1.5574942791762015e-05,
      "loss": 0.8896,
      "step": 3362
    },
    {
      "epoch": 0.96,
      "grad_norm": 10.71406078338623,
      "learning_rate": 1.5570652173913044e-05,
      "loss": 0.977,
      "step": 3363
    },
    {
      "epoch": 0.96,
      "grad_norm": 10.768048286437988,
      "learning_rate": 1.5566361556064073e-05,
      "loss": 1.2492,
      "step": 3364
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.973236083984375,
      "learning_rate": 1.5562070938215105e-05,
      "loss": 0.8089,
      "step": 3365
    },
    {
      "epoch": 0.96,
      "grad_norm": 8.230417251586914,
      "learning_rate": 1.555778032036613e-05,
      "loss": 0.8676,
      "step": 3366
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.360048294067383,
      "learning_rate": 1.5553489702517162e-05,
      "loss": 1.0663,
      "step": 3367
    },
    {
      "epoch": 0.96,
      "grad_norm": 8.129524230957031,
      "learning_rate": 1.5549199084668194e-05,
      "loss": 0.8034,
      "step": 3368
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.040216445922852,
      "learning_rate": 1.5544908466819223e-05,
      "loss": 0.9049,
      "step": 3369
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.51258373260498,
      "learning_rate": 1.5540617848970252e-05,
      "loss": 0.9196,
      "step": 3370
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.116039276123047,
      "learning_rate": 1.553632723112128e-05,
      "loss": 0.9509,
      "step": 3371
    },
    {
      "epoch": 0.96,
      "grad_norm": 7.990875244140625,
      "learning_rate": 1.5532036613272313e-05,
      "loss": 0.7907,
      "step": 3372
    },
    {
      "epoch": 0.96,
      "grad_norm": 10.509187698364258,
      "learning_rate": 1.552774599542334e-05,
      "loss": 1.0079,
      "step": 3373
    },
    {
      "epoch": 0.97,
      "grad_norm": 9.599989891052246,
      "learning_rate": 1.552345537757437e-05,
      "loss": 1.0179,
      "step": 3374
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.02453899383545,
      "learning_rate": 1.55191647597254e-05,
      "loss": 1.0442,
      "step": 3375
    },
    {
      "epoch": 0.97,
      "grad_norm": 9.042299270629883,
      "learning_rate": 1.551487414187643e-05,
      "loss": 0.9454,
      "step": 3376
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.32194995880127,
      "learning_rate": 1.5510583524027463e-05,
      "loss": 1.0094,
      "step": 3377
    },
    {
      "epoch": 0.97,
      "grad_norm": 8.976984024047852,
      "learning_rate": 1.550629290617849e-05,
      "loss": 0.8772,
      "step": 3378
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.98599910736084,
      "learning_rate": 1.550200228832952e-05,
      "loss": 0.8978,
      "step": 3379
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.110823631286621,
      "learning_rate": 1.549771167048055e-05,
      "loss": 1.0381,
      "step": 3380
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.129533767700195,
      "learning_rate": 1.549342105263158e-05,
      "loss": 1.132,
      "step": 3381
    },
    {
      "epoch": 0.97,
      "grad_norm": 9.173952102661133,
      "learning_rate": 1.5489130434782607e-05,
      "loss": 1.1657,
      "step": 3382
    },
    {
      "epoch": 0.97,
      "grad_norm": 9.895243644714355,
      "learning_rate": 1.548483981693364e-05,
      "loss": 0.975,
      "step": 3383
    },
    {
      "epoch": 0.97,
      "grad_norm": 12.187675476074219,
      "learning_rate": 1.5480549199084668e-05,
      "loss": 1.4564,
      "step": 3384
    },
    {
      "epoch": 0.97,
      "grad_norm": 11.068359375,
      "learning_rate": 1.5476258581235697e-05,
      "loss": 1.3282,
      "step": 3385
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.907172203063965,
      "learning_rate": 1.547196796338673e-05,
      "loss": 1.0901,
      "step": 3386
    },
    {
      "epoch": 0.97,
      "grad_norm": 8.23672103881836,
      "learning_rate": 1.5467677345537758e-05,
      "loss": 0.6549,
      "step": 3387
    },
    {
      "epoch": 0.97,
      "grad_norm": 12.269115447998047,
      "learning_rate": 1.546338672768879e-05,
      "loss": 1.1592,
      "step": 3388
    },
    {
      "epoch": 0.97,
      "grad_norm": 11.227182388305664,
      "learning_rate": 1.5459096109839815e-05,
      "loss": 0.8964,
      "step": 3389
    },
    {
      "epoch": 0.97,
      "grad_norm": 7.9429402351379395,
      "learning_rate": 1.5454805491990848e-05,
      "loss": 0.9423,
      "step": 3390
    },
    {
      "epoch": 0.97,
      "grad_norm": 8.185820579528809,
      "learning_rate": 1.5450514874141876e-05,
      "loss": 0.8259,
      "step": 3391
    },
    {
      "epoch": 0.97,
      "grad_norm": 11.577701568603516,
      "learning_rate": 1.5446224256292905e-05,
      "loss": 0.9428,
      "step": 3392
    },
    {
      "epoch": 0.97,
      "grad_norm": 9.812365531921387,
      "learning_rate": 1.5441933638443934e-05,
      "loss": 0.8602,
      "step": 3393
    },
    {
      "epoch": 0.97,
      "grad_norm": 8.900385856628418,
      "learning_rate": 1.5437643020594966e-05,
      "loss": 0.8757,
      "step": 3394
    },
    {
      "epoch": 0.97,
      "grad_norm": 8.60604190826416,
      "learning_rate": 1.5433352402745998e-05,
      "loss": 0.8252,
      "step": 3395
    },
    {
      "epoch": 0.97,
      "grad_norm": 8.94503402709961,
      "learning_rate": 1.5429061784897024e-05,
      "loss": 0.9966,
      "step": 3396
    },
    {
      "epoch": 0.97,
      "grad_norm": 7.996914863586426,
      "learning_rate": 1.5424771167048056e-05,
      "loss": 0.6504,
      "step": 3397
    },
    {
      "epoch": 0.97,
      "grad_norm": 8.389142036437988,
      "learning_rate": 1.5420480549199085e-05,
      "loss": 0.9272,
      "step": 3398
    },
    {
      "epoch": 0.97,
      "grad_norm": 9.937409400939941,
      "learning_rate": 1.5416189931350117e-05,
      "loss": 0.9442,
      "step": 3399
    },
    {
      "epoch": 0.97,
      "grad_norm": 12.677313804626465,
      "learning_rate": 1.5411899313501142e-05,
      "loss": 1.4754,
      "step": 3400
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.936812400817871,
      "learning_rate": 1.5407608695652174e-05,
      "loss": 0.9896,
      "step": 3401
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.202360153198242,
      "learning_rate": 1.5403318077803206e-05,
      "loss": 1.2344,
      "step": 3402
    },
    {
      "epoch": 0.97,
      "grad_norm": 8.682063102722168,
      "learning_rate": 1.5399027459954235e-05,
      "loss": 0.9496,
      "step": 3403
    },
    {
      "epoch": 0.97,
      "grad_norm": 9.442776679992676,
      "learning_rate": 1.5394736842105264e-05,
      "loss": 0.826,
      "step": 3404
    },
    {
      "epoch": 0.97,
      "grad_norm": 9.357182502746582,
      "learning_rate": 1.5390446224256293e-05,
      "loss": 1.0239,
      "step": 3405
    },
    {
      "epoch": 0.97,
      "grad_norm": 7.865712642669678,
      "learning_rate": 1.5386155606407325e-05,
      "loss": 0.9239,
      "step": 3406
    },
    {
      "epoch": 0.97,
      "grad_norm": 9.808480262756348,
      "learning_rate": 1.538186498855835e-05,
      "loss": 0.7902,
      "step": 3407
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.461572647094727,
      "learning_rate": 1.5377574370709382e-05,
      "loss": 1.0959,
      "step": 3408
    },
    {
      "epoch": 0.98,
      "grad_norm": 8.114765167236328,
      "learning_rate": 1.537328375286041e-05,
      "loss": 0.8807,
      "step": 3409
    },
    {
      "epoch": 0.98,
      "grad_norm": 8.87446117401123,
      "learning_rate": 1.5368993135011443e-05,
      "loss": 1.1426,
      "step": 3410
    },
    {
      "epoch": 0.98,
      "grad_norm": 11.126185417175293,
      "learning_rate": 1.5364702517162472e-05,
      "loss": 1.0115,
      "step": 3411
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.193495750427246,
      "learning_rate": 1.53604118993135e-05,
      "loss": 0.8523,
      "step": 3412
    },
    {
      "epoch": 0.98,
      "grad_norm": 8.045726776123047,
      "learning_rate": 1.5356121281464533e-05,
      "loss": 0.6918,
      "step": 3413
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.752435684204102,
      "learning_rate": 1.5351830663615562e-05,
      "loss": 1.0277,
      "step": 3414
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.08300495147705,
      "learning_rate": 1.534754004576659e-05,
      "loss": 0.9409,
      "step": 3415
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.115248680114746,
      "learning_rate": 1.534324942791762e-05,
      "loss": 0.916,
      "step": 3416
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.535004615783691,
      "learning_rate": 1.533895881006865e-05,
      "loss": 0.7505,
      "step": 3417
    },
    {
      "epoch": 0.98,
      "grad_norm": 7.965742588043213,
      "learning_rate": 1.5334668192219677e-05,
      "loss": 0.9348,
      "step": 3418
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.332640647888184,
      "learning_rate": 1.533037757437071e-05,
      "loss": 1.0244,
      "step": 3419
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.236135482788086,
      "learning_rate": 1.532608695652174e-05,
      "loss": 0.884,
      "step": 3420
    },
    {
      "epoch": 0.98,
      "grad_norm": 7.922878742218018,
      "learning_rate": 1.532179633867277e-05,
      "loss": 0.8087,
      "step": 3421
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.578164100646973,
      "learning_rate": 1.53175057208238e-05,
      "loss": 1.1014,
      "step": 3422
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.6139554977417,
      "learning_rate": 1.5313215102974827e-05,
      "loss": 0.9526,
      "step": 3423
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.56572151184082,
      "learning_rate": 1.530892448512586e-05,
      "loss": 0.8512,
      "step": 3424
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.807348251342773,
      "learning_rate": 1.530463386727689e-05,
      "loss": 0.9336,
      "step": 3425
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.48486328125,
      "learning_rate": 1.5300343249427917e-05,
      "loss": 0.7069,
      "step": 3426
    },
    {
      "epoch": 0.98,
      "grad_norm": 8.4706392288208,
      "learning_rate": 1.5296052631578946e-05,
      "loss": 1.0151,
      "step": 3427
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.076640129089355,
      "learning_rate": 1.5291762013729978e-05,
      "loss": 0.9837,
      "step": 3428
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.002256393432617,
      "learning_rate": 1.528747139588101e-05,
      "loss": 0.9437,
      "step": 3429
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.732633590698242,
      "learning_rate": 1.5283180778032036e-05,
      "loss": 0.8993,
      "step": 3430
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.509393692016602,
      "learning_rate": 1.5278890160183068e-05,
      "loss": 0.9594,
      "step": 3431
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.0447359085083,
      "learning_rate": 1.5274599542334097e-05,
      "loss": 0.873,
      "step": 3432
    },
    {
      "epoch": 0.98,
      "grad_norm": 12.14205551147461,
      "learning_rate": 1.5270308924485125e-05,
      "loss": 0.9494,
      "step": 3433
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.456064224243164,
      "learning_rate": 1.5266018306636154e-05,
      "loss": 1.0558,
      "step": 3434
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.385796546936035,
      "learning_rate": 1.5261727688787186e-05,
      "loss": 0.9873,
      "step": 3435
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.608353614807129,
      "learning_rate": 1.5257437070938213e-05,
      "loss": 0.8733,
      "step": 3436
    },
    {
      "epoch": 0.98,
      "grad_norm": 8.533060073852539,
      "learning_rate": 1.5253146453089245e-05,
      "loss": 0.7835,
      "step": 3437
    },
    {
      "epoch": 0.98,
      "grad_norm": 8.630105972290039,
      "learning_rate": 1.5248855835240276e-05,
      "loss": 0.8252,
      "step": 3438
    },
    {
      "epoch": 0.98,
      "grad_norm": 12.451923370361328,
      "learning_rate": 1.5244565217391305e-05,
      "loss": 1.3839,
      "step": 3439
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.088960647583008,
      "learning_rate": 1.5240274599542335e-05,
      "loss": 0.9431,
      "step": 3440
    },
    {
      "epoch": 0.98,
      "grad_norm": 11.612284660339355,
      "learning_rate": 1.5235983981693364e-05,
      "loss": 1.2605,
      "step": 3441
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.245793342590332,
      "learning_rate": 1.5231693363844394e-05,
      "loss": 1.1037,
      "step": 3442
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.433359146118164,
      "learning_rate": 1.5227402745995423e-05,
      "loss": 0.8219,
      "step": 3443
    },
    {
      "epoch": 0.99,
      "grad_norm": 9.079718589782715,
      "learning_rate": 1.5223112128146454e-05,
      "loss": 0.955,
      "step": 3444
    },
    {
      "epoch": 0.99,
      "grad_norm": 10.114764213562012,
      "learning_rate": 1.5218821510297484e-05,
      "loss": 1.1642,
      "step": 3445
    },
    {
      "epoch": 0.99,
      "grad_norm": 9.784605979919434,
      "learning_rate": 1.5214530892448513e-05,
      "loss": 0.8729,
      "step": 3446
    },
    {
      "epoch": 0.99,
      "grad_norm": 11.648603439331055,
      "learning_rate": 1.5210240274599543e-05,
      "loss": 1.0002,
      "step": 3447
    },
    {
      "epoch": 0.99,
      "grad_norm": 10.44534969329834,
      "learning_rate": 1.5205949656750572e-05,
      "loss": 1.2613,
      "step": 3448
    },
    {
      "epoch": 0.99,
      "grad_norm": 8.184114456176758,
      "learning_rate": 1.5201659038901603e-05,
      "loss": 1.0381,
      "step": 3449
    },
    {
      "epoch": 0.99,
      "grad_norm": 8.278879165649414,
      "learning_rate": 1.5197368421052631e-05,
      "loss": 0.8383,
      "step": 3450
    },
    {
      "epoch": 0.99,
      "grad_norm": 10.543951034545898,
      "learning_rate": 1.5193077803203662e-05,
      "loss": 1.1907,
      "step": 3451
    },
    {
      "epoch": 0.99,
      "grad_norm": 9.217336654663086,
      "learning_rate": 1.518878718535469e-05,
      "loss": 1.1273,
      "step": 3452
    },
    {
      "epoch": 0.99,
      "grad_norm": 8.791984558105469,
      "learning_rate": 1.5184496567505721e-05,
      "loss": 1.0187,
      "step": 3453
    },
    {
      "epoch": 0.99,
      "grad_norm": 9.331918716430664,
      "learning_rate": 1.5180205949656753e-05,
      "loss": 0.8774,
      "step": 3454
    },
    {
      "epoch": 0.99,
      "grad_norm": 9.732755661010742,
      "learning_rate": 1.517591533180778e-05,
      "loss": 1.2942,
      "step": 3455
    },
    {
      "epoch": 0.99,
      "grad_norm": 9.513022422790527,
      "learning_rate": 1.5171624713958812e-05,
      "loss": 1.0173,
      "step": 3456
    },
    {
      "epoch": 0.99,
      "grad_norm": 7.993738651275635,
      "learning_rate": 1.516733409610984e-05,
      "loss": 0.9076,
      "step": 3457
    },
    {
      "epoch": 0.99,
      "grad_norm": 10.794379234313965,
      "learning_rate": 1.516304347826087e-05,
      "loss": 1.0986,
      "step": 3458
    },
    {
      "epoch": 0.99,
      "grad_norm": 11.232872009277344,
      "learning_rate": 1.5158752860411899e-05,
      "loss": 1.0131,
      "step": 3459
    },
    {
      "epoch": 0.99,
      "grad_norm": 8.671607971191406,
      "learning_rate": 1.515446224256293e-05,
      "loss": 0.7477,
      "step": 3460
    },
    {
      "epoch": 0.99,
      "grad_norm": 9.410494804382324,
      "learning_rate": 1.5150171624713958e-05,
      "loss": 0.6941,
      "step": 3461
    },
    {
      "epoch": 0.99,
      "grad_norm": 9.8814058303833,
      "learning_rate": 1.5145881006864988e-05,
      "loss": 0.9058,
      "step": 3462
    },
    {
      "epoch": 0.99,
      "grad_norm": 8.776352882385254,
      "learning_rate": 1.514159038901602e-05,
      "loss": 0.9837,
      "step": 3463
    },
    {
      "epoch": 0.99,
      "grad_norm": 10.406237602233887,
      "learning_rate": 1.5137299771167048e-05,
      "loss": 0.9577,
      "step": 3464
    },
    {
      "epoch": 0.99,
      "grad_norm": 9.833565711975098,
      "learning_rate": 1.513300915331808e-05,
      "loss": 1.0214,
      "step": 3465
    },
    {
      "epoch": 0.99,
      "grad_norm": 7.488353729248047,
      "learning_rate": 1.5128718535469107e-05,
      "loss": 0.7898,
      "step": 3466
    },
    {
      "epoch": 0.99,
      "grad_norm": 9.927220344543457,
      "learning_rate": 1.5124427917620139e-05,
      "loss": 0.8986,
      "step": 3467
    },
    {
      "epoch": 0.99,
      "grad_norm": 9.887399673461914,
      "learning_rate": 1.5120137299771166e-05,
      "loss": 0.9155,
      "step": 3468
    },
    {
      "epoch": 0.99,
      "grad_norm": 9.517233848571777,
      "learning_rate": 1.5115846681922198e-05,
      "loss": 0.8387,
      "step": 3469
    },
    {
      "epoch": 0.99,
      "grad_norm": 9.98740291595459,
      "learning_rate": 1.5111556064073225e-05,
      "loss": 1.0168,
      "step": 3470
    },
    {
      "epoch": 0.99,
      "grad_norm": 10.123132705688477,
      "learning_rate": 1.5107265446224256e-05,
      "loss": 0.9903,
      "step": 3471
    },
    {
      "epoch": 0.99,
      "grad_norm": 8.628500938415527,
      "learning_rate": 1.5102974828375288e-05,
      "loss": 0.5802,
      "step": 3472
    },
    {
      "epoch": 0.99,
      "grad_norm": 9.079924583435059,
      "learning_rate": 1.5098684210526315e-05,
      "loss": 0.9983,
      "step": 3473
    },
    {
      "epoch": 0.99,
      "grad_norm": 9.199273109436035,
      "learning_rate": 1.5094393592677347e-05,
      "loss": 1.0829,
      "step": 3474
    },
    {
      "epoch": 0.99,
      "grad_norm": 9.835076332092285,
      "learning_rate": 1.5090102974828374e-05,
      "loss": 1.0947,
      "step": 3475
    },
    {
      "epoch": 0.99,
      "grad_norm": 10.03668212890625,
      "learning_rate": 1.5085812356979406e-05,
      "loss": 0.7865,
      "step": 3476
    },
    {
      "epoch": 0.99,
      "grad_norm": 10.380677223205566,
      "learning_rate": 1.5081521739130433e-05,
      "loss": 1.2301,
      "step": 3477
    },
    {
      "epoch": 0.99,
      "grad_norm": 12.602004051208496,
      "learning_rate": 1.5077231121281466e-05,
      "loss": 1.2464,
      "step": 3478
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.200499534606934,
      "learning_rate": 1.5072940503432496e-05,
      "loss": 0.9179,
      "step": 3479
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.837831497192383,
      "learning_rate": 1.5068649885583525e-05,
      "loss": 0.6146,
      "step": 3480
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.19485092163086,
      "learning_rate": 1.5064359267734555e-05,
      "loss": 0.7807,
      "step": 3481
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.346631050109863,
      "learning_rate": 1.5060068649885584e-05,
      "loss": 1.0513,
      "step": 3482
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.652012825012207,
      "learning_rate": 1.5055778032036615e-05,
      "loss": 0.9422,
      "step": 3483
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.3707857131958,
      "learning_rate": 1.5051487414187642e-05,
      "loss": 0.8258,
      "step": 3484
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.806299209594727,
      "learning_rate": 1.5047196796338674e-05,
      "loss": 0.8674,
      "step": 3485
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.051835060119629,
      "learning_rate": 1.50429061784897e-05,
      "loss": 1.0132,
      "step": 3486
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.189953804016113,
      "learning_rate": 1.5038615560640733e-05,
      "loss": 0.8535,
      "step": 3487
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.710848808288574,
      "learning_rate": 1.5034324942791763e-05,
      "loss": 0.8356,
      "step": 3488
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.190397262573242,
      "learning_rate": 1.5030034324942792e-05,
      "loss": 0.9986,
      "step": 3489
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.301618576049805,
      "learning_rate": 1.5025743707093823e-05,
      "loss": 0.9548,
      "step": 3490
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.565505027770996,
      "learning_rate": 1.5021453089244851e-05,
      "loss": 0.9822,
      "step": 3491
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.500279426574707,
      "learning_rate": 1.5017162471395882e-05,
      "loss": 0.8476,
      "step": 3492
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.195255279541016,
      "learning_rate": 1.501287185354691e-05,
      "loss": 0.8558,
      "step": 3493
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.99519157409668,
      "learning_rate": 1.5008581235697941e-05,
      "loss": 0.9316,
      "step": 3494
    },
    {
      "epoch": 1.0,
      "grad_norm": 7.147590637207031,
      "learning_rate": 1.500429061784897e-05,
      "loss": 0.7436,
      "step": 3495
    },
    {
      "epoch": 1.0,
      "grad_norm": 11.322429656982422,
      "learning_rate": 1.5e-05,
      "loss": 1.0799,
      "step": 3496
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.729336738586426,
      "learning_rate": 1.4995709382151029e-05,
      "loss": 0.9429,
      "step": 3497
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.235333442687988,
      "learning_rate": 1.499141876430206e-05,
      "loss": 0.9913,
      "step": 3498
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.02188777923584,
      "learning_rate": 1.4987128146453088e-05,
      "loss": 0.9376,
      "step": 3499
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.01506233215332,
      "learning_rate": 1.4982837528604119e-05,
      "loss": 0.6847,
      "step": 3500
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.915964126586914,
      "learning_rate": 1.497854691075515e-05,
      "loss": 1.0377,
      "step": 3501
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.767047882080078,
      "learning_rate": 1.497425629290618e-05,
      "loss": 1.0724,
      "step": 3502
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.36973762512207,
      "learning_rate": 1.4969965675057209e-05,
      "loss": 1.0589,
      "step": 3503
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.49530029296875,
      "learning_rate": 1.4965675057208239e-05,
      "loss": 0.9688,
      "step": 3504
    },
    {
      "epoch": 1.0,
      "grad_norm": 7.333240509033203,
      "learning_rate": 1.4961384439359268e-05,
      "loss": 0.7759,
      "step": 3505
    },
    {
      "epoch": 1.0,
      "grad_norm": 7.345729827880859,
      "learning_rate": 1.4957093821510298e-05,
      "loss": 0.5091,
      "step": 3506
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.10158634185791,
      "learning_rate": 1.4952803203661327e-05,
      "loss": 0.602,
      "step": 3507
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.788975715637207,
      "learning_rate": 1.4948512585812357e-05,
      "loss": 0.8844,
      "step": 3508
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.100739479064941,
      "learning_rate": 1.4944221967963388e-05,
      "loss": 0.7641,
      "step": 3509
    },
    {
      "epoch": 1.0,
      "grad_norm": 7.1900248527526855,
      "learning_rate": 1.4939931350114417e-05,
      "loss": 0.7606,
      "step": 3510
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.251717567443848,
      "learning_rate": 1.4935640732265447e-05,
      "loss": 0.792,
      "step": 3511
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.145462036132812,
      "learning_rate": 1.4931350114416476e-05,
      "loss": 1.0039,
      "step": 3512
    },
    {
      "epoch": 1.0,
      "grad_norm": 7.771153450012207,
      "learning_rate": 1.4927059496567506e-05,
      "loss": 0.8965,
      "step": 3513
    },
    {
      "epoch": 1.01,
      "grad_norm": 9.03454303741455,
      "learning_rate": 1.4922768878718535e-05,
      "loss": 0.8384,
      "step": 3514
    },
    {
      "epoch": 1.01,
      "grad_norm": 8.778411865234375,
      "learning_rate": 1.4918478260869566e-05,
      "loss": 0.8014,
      "step": 3515
    },
    {
      "epoch": 1.01,
      "grad_norm": 9.974286079406738,
      "learning_rate": 1.4914187643020594e-05,
      "loss": 0.7165,
      "step": 3516
    },
    {
      "epoch": 1.01,
      "grad_norm": 7.651914596557617,
      "learning_rate": 1.4909897025171625e-05,
      "loss": 0.7693,
      "step": 3517
    },
    {
      "epoch": 1.01,
      "grad_norm": 11.590794563293457,
      "learning_rate": 1.4905606407322655e-05,
      "loss": 0.8939,
      "step": 3518
    },
    {
      "epoch": 1.01,
      "grad_norm": 11.48332691192627,
      "learning_rate": 1.4901315789473686e-05,
      "loss": 0.8586,
      "step": 3519
    },
    {
      "epoch": 1.01,
      "grad_norm": 10.365842819213867,
      "learning_rate": 1.4897025171624715e-05,
      "loss": 0.8985,
      "step": 3520
    },
    {
      "epoch": 1.01,
      "grad_norm": 9.330724716186523,
      "learning_rate": 1.4892734553775745e-05,
      "loss": 0.8827,
      "step": 3521
    },
    {
      "epoch": 1.01,
      "grad_norm": 8.041494369506836,
      "learning_rate": 1.4888443935926774e-05,
      "loss": 0.7185,
      "step": 3522
    },
    {
      "epoch": 1.01,
      "grad_norm": 8.145724296569824,
      "learning_rate": 1.4884153318077803e-05,
      "loss": 0.7699,
      "step": 3523
    },
    {
      "epoch": 1.01,
      "grad_norm": 7.616659164428711,
      "learning_rate": 1.4879862700228833e-05,
      "loss": 0.6846,
      "step": 3524
    },
    {
      "epoch": 1.01,
      "grad_norm": 7.798200607299805,
      "learning_rate": 1.4875572082379862e-05,
      "loss": 0.7173,
      "step": 3525
    },
    {
      "epoch": 1.01,
      "grad_norm": 10.310019493103027,
      "learning_rate": 1.4871281464530894e-05,
      "loss": 0.7357,
      "step": 3526
    },
    {
      "epoch": 1.01,
      "grad_norm": 11.527127265930176,
      "learning_rate": 1.4866990846681923e-05,
      "loss": 0.9704,
      "step": 3527
    },
    {
      "epoch": 1.01,
      "grad_norm": 11.12434196472168,
      "learning_rate": 1.4862700228832953e-05,
      "loss": 0.918,
      "step": 3528
    },
    {
      "epoch": 1.01,
      "grad_norm": 11.516888618469238,
      "learning_rate": 1.4858409610983982e-05,
      "loss": 1.0105,
      "step": 3529
    },
    {
      "epoch": 1.01,
      "grad_norm": 7.838921546936035,
      "learning_rate": 1.4854118993135012e-05,
      "loss": 0.7692,
      "step": 3530
    },
    {
      "epoch": 1.01,
      "grad_norm": 9.907757759094238,
      "learning_rate": 1.4849828375286041e-05,
      "loss": 0.7239,
      "step": 3531
    },
    {
      "epoch": 1.01,
      "grad_norm": 9.133485794067383,
      "learning_rate": 1.4845537757437072e-05,
      "loss": 0.9211,
      "step": 3532
    },
    {
      "epoch": 1.01,
      "grad_norm": 11.23843002319336,
      "learning_rate": 1.48412471395881e-05,
      "loss": 1.0158,
      "step": 3533
    },
    {
      "epoch": 1.01,
      "grad_norm": 9.841829299926758,
      "learning_rate": 1.483695652173913e-05,
      "loss": 0.6117,
      "step": 3534
    },
    {
      "epoch": 1.01,
      "grad_norm": 10.754142761230469,
      "learning_rate": 1.4832665903890161e-05,
      "loss": 0.6776,
      "step": 3535
    },
    {
      "epoch": 1.01,
      "grad_norm": 13.35550308227539,
      "learning_rate": 1.482837528604119e-05,
      "loss": 0.8379,
      "step": 3536
    },
    {
      "epoch": 1.01,
      "grad_norm": 11.992549896240234,
      "learning_rate": 1.482408466819222e-05,
      "loss": 1.0228,
      "step": 3537
    },
    {
      "epoch": 1.01,
      "grad_norm": 11.021371841430664,
      "learning_rate": 1.481979405034325e-05,
      "loss": 0.7198,
      "step": 3538
    },
    {
      "epoch": 1.01,
      "grad_norm": 9.923230171203613,
      "learning_rate": 1.481550343249428e-05,
      "loss": 0.9317,
      "step": 3539
    },
    {
      "epoch": 1.01,
      "grad_norm": 8.832412719726562,
      "learning_rate": 1.4811212814645309e-05,
      "loss": 0.8334,
      "step": 3540
    },
    {
      "epoch": 1.01,
      "grad_norm": 9.36916732788086,
      "learning_rate": 1.4806922196796339e-05,
      "loss": 0.7994,
      "step": 3541
    },
    {
      "epoch": 1.01,
      "grad_norm": 9.33375358581543,
      "learning_rate": 1.4802631578947368e-05,
      "loss": 0.8682,
      "step": 3542
    },
    {
      "epoch": 1.01,
      "grad_norm": 10.338713645935059,
      "learning_rate": 1.4798340961098398e-05,
      "loss": 0.7042,
      "step": 3543
    },
    {
      "epoch": 1.01,
      "grad_norm": 9.743721008300781,
      "learning_rate": 1.4794050343249429e-05,
      "loss": 0.8741,
      "step": 3544
    },
    {
      "epoch": 1.01,
      "grad_norm": 6.583962917327881,
      "learning_rate": 1.4789759725400459e-05,
      "loss": 0.5621,
      "step": 3545
    },
    {
      "epoch": 1.01,
      "grad_norm": 9.400547981262207,
      "learning_rate": 1.4785469107551488e-05,
      "loss": 0.7926,
      "step": 3546
    },
    {
      "epoch": 1.01,
      "grad_norm": 10.095444679260254,
      "learning_rate": 1.4781178489702518e-05,
      "loss": 0.7835,
      "step": 3547
    },
    {
      "epoch": 1.01,
      "grad_norm": 10.908488273620605,
      "learning_rate": 1.4776887871853547e-05,
      "loss": 0.8011,
      "step": 3548
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.587437629699707,
      "learning_rate": 1.4772597254004576e-05,
      "loss": 0.876,
      "step": 3549
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.30736255645752,
      "learning_rate": 1.4768306636155606e-05,
      "loss": 1.0483,
      "step": 3550
    },
    {
      "epoch": 1.02,
      "grad_norm": 7.909506320953369,
      "learning_rate": 1.4764016018306635e-05,
      "loss": 0.7863,
      "step": 3551
    },
    {
      "epoch": 1.02,
      "grad_norm": 8.779753684997559,
      "learning_rate": 1.4759725400457667e-05,
      "loss": 0.998,
      "step": 3552
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.129301071166992,
      "learning_rate": 1.4755434782608696e-05,
      "loss": 0.7694,
      "step": 3553
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.928060531616211,
      "learning_rate": 1.4751144164759727e-05,
      "loss": 0.7666,
      "step": 3554
    },
    {
      "epoch": 1.02,
      "grad_norm": 10.944051742553711,
      "learning_rate": 1.4746853546910755e-05,
      "loss": 0.9608,
      "step": 3555
    },
    {
      "epoch": 1.02,
      "grad_norm": 7.6837239265441895,
      "learning_rate": 1.4742562929061786e-05,
      "loss": 0.5508,
      "step": 3556
    },
    {
      "epoch": 1.02,
      "grad_norm": 10.845468521118164,
      "learning_rate": 1.4738272311212815e-05,
      "loss": 0.8979,
      "step": 3557
    },
    {
      "epoch": 1.02,
      "grad_norm": 8.706075668334961,
      "learning_rate": 1.4733981693363845e-05,
      "loss": 0.6102,
      "step": 3558
    },
    {
      "epoch": 1.02,
      "grad_norm": 13.22905158996582,
      "learning_rate": 1.4729691075514874e-05,
      "loss": 1.0806,
      "step": 3559
    },
    {
      "epoch": 1.02,
      "grad_norm": 7.789815902709961,
      "learning_rate": 1.4725400457665904e-05,
      "loss": 0.5796,
      "step": 3560
    },
    {
      "epoch": 1.02,
      "grad_norm": 8.726973533630371,
      "learning_rate": 1.4721109839816935e-05,
      "loss": 0.7599,
      "step": 3561
    },
    {
      "epoch": 1.02,
      "grad_norm": 10.404717445373535,
      "learning_rate": 1.4716819221967963e-05,
      "loss": 0.8246,
      "step": 3562
    },
    {
      "epoch": 1.02,
      "grad_norm": 11.025307655334473,
      "learning_rate": 1.4712528604118994e-05,
      "loss": 0.8303,
      "step": 3563
    },
    {
      "epoch": 1.02,
      "grad_norm": 11.1351318359375,
      "learning_rate": 1.4708237986270023e-05,
      "loss": 0.8676,
      "step": 3564
    },
    {
      "epoch": 1.02,
      "grad_norm": 8.169400215148926,
      "learning_rate": 1.4703947368421053e-05,
      "loss": 0.5322,
      "step": 3565
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.377978324890137,
      "learning_rate": 1.4699656750572082e-05,
      "loss": 0.8039,
      "step": 3566
    },
    {
      "epoch": 1.02,
      "grad_norm": 10.150505065917969,
      "learning_rate": 1.4695366132723112e-05,
      "loss": 0.9011,
      "step": 3567
    },
    {
      "epoch": 1.02,
      "grad_norm": 8.142455101013184,
      "learning_rate": 1.4691075514874141e-05,
      "loss": 0.8546,
      "step": 3568
    },
    {
      "epoch": 1.02,
      "grad_norm": 8.890527725219727,
      "learning_rate": 1.4686784897025173e-05,
      "loss": 0.901,
      "step": 3569
    },
    {
      "epoch": 1.02,
      "grad_norm": 10.418173789978027,
      "learning_rate": 1.4682494279176202e-05,
      "loss": 0.8699,
      "step": 3570
    },
    {
      "epoch": 1.02,
      "grad_norm": 10.354543685913086,
      "learning_rate": 1.4678203661327233e-05,
      "loss": 1.0417,
      "step": 3571
    },
    {
      "epoch": 1.02,
      "grad_norm": 11.536067008972168,
      "learning_rate": 1.4673913043478261e-05,
      "loss": 0.8563,
      "step": 3572
    },
    {
      "epoch": 1.02,
      "grad_norm": 10.119474411010742,
      "learning_rate": 1.4669622425629292e-05,
      "loss": 1.0292,
      "step": 3573
    },
    {
      "epoch": 1.02,
      "grad_norm": 7.384093284606934,
      "learning_rate": 1.466533180778032e-05,
      "loss": 0.6153,
      "step": 3574
    },
    {
      "epoch": 1.02,
      "grad_norm": 10.39474105834961,
      "learning_rate": 1.466104118993135e-05,
      "loss": 1.0311,
      "step": 3575
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.86563777923584,
      "learning_rate": 1.465675057208238e-05,
      "loss": 0.743,
      "step": 3576
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.815017700195312,
      "learning_rate": 1.4652459954233409e-05,
      "loss": 0.8117,
      "step": 3577
    },
    {
      "epoch": 1.02,
      "grad_norm": 8.892045974731445,
      "learning_rate": 1.464816933638444e-05,
      "loss": 0.7091,
      "step": 3578
    },
    {
      "epoch": 1.02,
      "grad_norm": 11.2473726272583,
      "learning_rate": 1.464387871853547e-05,
      "loss": 0.8584,
      "step": 3579
    },
    {
      "epoch": 1.02,
      "grad_norm": 8.565177917480469,
      "learning_rate": 1.46395881006865e-05,
      "loss": 0.7814,
      "step": 3580
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.64526653289795,
      "learning_rate": 1.4635297482837529e-05,
      "loss": 0.9121,
      "step": 3581
    },
    {
      "epoch": 1.02,
      "grad_norm": 8.384381294250488,
      "learning_rate": 1.4631006864988559e-05,
      "loss": 0.6535,
      "step": 3582
    },
    {
      "epoch": 1.02,
      "grad_norm": 10.528741836547852,
      "learning_rate": 1.4626716247139588e-05,
      "loss": 0.5912,
      "step": 3583
    },
    {
      "epoch": 1.03,
      "grad_norm": 8.734308242797852,
      "learning_rate": 1.4622425629290618e-05,
      "loss": 0.6136,
      "step": 3584
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.390425682067871,
      "learning_rate": 1.4618135011441647e-05,
      "loss": 0.9535,
      "step": 3585
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.493962287902832,
      "learning_rate": 1.461384439359268e-05,
      "loss": 1.1426,
      "step": 3586
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.703402519226074,
      "learning_rate": 1.4609553775743708e-05,
      "loss": 0.9597,
      "step": 3587
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.467942237854004,
      "learning_rate": 1.4605263157894737e-05,
      "loss": 1.0131,
      "step": 3588
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.35848331451416,
      "learning_rate": 1.4600972540045767e-05,
      "loss": 0.8164,
      "step": 3589
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.839441299438477,
      "learning_rate": 1.4596681922196796e-05,
      "loss": 0.6921,
      "step": 3590
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.550750732421875,
      "learning_rate": 1.4592391304347827e-05,
      "loss": 0.8487,
      "step": 3591
    },
    {
      "epoch": 1.03,
      "grad_norm": 8.606681823730469,
      "learning_rate": 1.4588100686498855e-05,
      "loss": 0.7856,
      "step": 3592
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.010801315307617,
      "learning_rate": 1.4583810068649886e-05,
      "loss": 0.689,
      "step": 3593
    },
    {
      "epoch": 1.03,
      "grad_norm": 8.278401374816895,
      "learning_rate": 1.4579519450800915e-05,
      "loss": 0.6683,
      "step": 3594
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.394222259521484,
      "learning_rate": 1.4575228832951947e-05,
      "loss": 0.896,
      "step": 3595
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.620981216430664,
      "learning_rate": 1.4570938215102975e-05,
      "loss": 0.7403,
      "step": 3596
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.585387229919434,
      "learning_rate": 1.4566647597254006e-05,
      "loss": 0.7583,
      "step": 3597
    },
    {
      "epoch": 1.03,
      "grad_norm": 11.001717567443848,
      "learning_rate": 1.4562356979405035e-05,
      "loss": 0.9145,
      "step": 3598
    },
    {
      "epoch": 1.03,
      "grad_norm": 8.327040672302246,
      "learning_rate": 1.4558066361556065e-05,
      "loss": 0.6258,
      "step": 3599
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.499870300292969,
      "learning_rate": 1.4553775743707094e-05,
      "loss": 0.6328,
      "step": 3600
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.079899787902832,
      "learning_rate": 1.4549485125858124e-05,
      "loss": 0.6817,
      "step": 3601
    },
    {
      "epoch": 1.03,
      "grad_norm": 14.0177640914917,
      "learning_rate": 1.4545194508009153e-05,
      "loss": 0.916,
      "step": 3602
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.041084289550781,
      "learning_rate": 1.4540903890160182e-05,
      "loss": 0.9139,
      "step": 3603
    },
    {
      "epoch": 1.03,
      "grad_norm": 8.739937782287598,
      "learning_rate": 1.4536613272311214e-05,
      "loss": 0.5609,
      "step": 3604
    },
    {
      "epoch": 1.03,
      "grad_norm": 8.561568260192871,
      "learning_rate": 1.4532322654462243e-05,
      "loss": 0.6168,
      "step": 3605
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.857020378112793,
      "learning_rate": 1.4528032036613273e-05,
      "loss": 0.733,
      "step": 3606
    },
    {
      "epoch": 1.03,
      "grad_norm": 11.430785179138184,
      "learning_rate": 1.4523741418764302e-05,
      "loss": 0.9336,
      "step": 3607
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.511173248291016,
      "learning_rate": 1.4519450800915332e-05,
      "loss": 1.0044,
      "step": 3608
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.712701797485352,
      "learning_rate": 1.4515160183066361e-05,
      "loss": 1.0262,
      "step": 3609
    },
    {
      "epoch": 1.03,
      "grad_norm": 11.696800231933594,
      "learning_rate": 1.4510869565217392e-05,
      "loss": 1.197,
      "step": 3610
    },
    {
      "epoch": 1.03,
      "grad_norm": 8.272732734680176,
      "learning_rate": 1.450657894736842e-05,
      "loss": 0.7798,
      "step": 3611
    },
    {
      "epoch": 1.03,
      "grad_norm": 12.144145965576172,
      "learning_rate": 1.4502288329519453e-05,
      "loss": 0.6809,
      "step": 3612
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.280263900756836,
      "learning_rate": 1.4497997711670481e-05,
      "loss": 0.9479,
      "step": 3613
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.201637268066406,
      "learning_rate": 1.4493707093821512e-05,
      "loss": 0.6101,
      "step": 3614
    },
    {
      "epoch": 1.03,
      "grad_norm": 7.813424110412598,
      "learning_rate": 1.448941647597254e-05,
      "loss": 0.7455,
      "step": 3615
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.607821464538574,
      "learning_rate": 1.448512585812357e-05,
      "loss": 0.7986,
      "step": 3616
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.124610900878906,
      "learning_rate": 1.44808352402746e-05,
      "loss": 0.7356,
      "step": 3617
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.535299301147461,
      "learning_rate": 1.4476544622425629e-05,
      "loss": 0.8917,
      "step": 3618
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.994709014892578,
      "learning_rate": 1.4472254004576659e-05,
      "loss": 0.8014,
      "step": 3619
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.829028129577637,
      "learning_rate": 1.4467963386727688e-05,
      "loss": 0.7732,
      "step": 3620
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.724833488464355,
      "learning_rate": 1.446367276887872e-05,
      "loss": 0.8182,
      "step": 3621
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.023293495178223,
      "learning_rate": 1.4459382151029749e-05,
      "loss": 0.9182,
      "step": 3622
    },
    {
      "epoch": 1.04,
      "grad_norm": 11.00589656829834,
      "learning_rate": 1.445509153318078e-05,
      "loss": 0.9998,
      "step": 3623
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.75013256072998,
      "learning_rate": 1.4450800915331808e-05,
      "loss": 1.0815,
      "step": 3624
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.099947929382324,
      "learning_rate": 1.4446510297482838e-05,
      "loss": 0.9197,
      "step": 3625
    },
    {
      "epoch": 1.04,
      "grad_norm": 11.304966926574707,
      "learning_rate": 1.4442219679633867e-05,
      "loss": 1.0328,
      "step": 3626
    },
    {
      "epoch": 1.04,
      "grad_norm": 11.483154296875,
      "learning_rate": 1.4437929061784898e-05,
      "loss": 0.98,
      "step": 3627
    },
    {
      "epoch": 1.04,
      "grad_norm": 8.709813117980957,
      "learning_rate": 1.4433638443935926e-05,
      "loss": 0.7196,
      "step": 3628
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.48593521118164,
      "learning_rate": 1.4429347826086957e-05,
      "loss": 0.864,
      "step": 3629
    },
    {
      "epoch": 1.04,
      "grad_norm": 7.9823737144470215,
      "learning_rate": 1.4425057208237987e-05,
      "loss": 0.5869,
      "step": 3630
    },
    {
      "epoch": 1.04,
      "grad_norm": 8.290487289428711,
      "learning_rate": 1.4420766590389016e-05,
      "loss": 0.7856,
      "step": 3631
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.442862510681152,
      "learning_rate": 1.4416475972540047e-05,
      "loss": 0.8337,
      "step": 3632
    },
    {
      "epoch": 1.04,
      "grad_norm": 8.83592700958252,
      "learning_rate": 1.4412185354691075e-05,
      "loss": 0.5868,
      "step": 3633
    },
    {
      "epoch": 1.04,
      "grad_norm": 11.119686126708984,
      "learning_rate": 1.4407894736842106e-05,
      "loss": 0.8916,
      "step": 3634
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.67123031616211,
      "learning_rate": 1.4403604118993135e-05,
      "loss": 0.8316,
      "step": 3635
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.490544319152832,
      "learning_rate": 1.4399313501144165e-05,
      "loss": 0.999,
      "step": 3636
    },
    {
      "epoch": 1.04,
      "grad_norm": 11.423625946044922,
      "learning_rate": 1.4395022883295194e-05,
      "loss": 1.1596,
      "step": 3637
    },
    {
      "epoch": 1.04,
      "grad_norm": 8.861817359924316,
      "learning_rate": 1.4390732265446226e-05,
      "loss": 0.8421,
      "step": 3638
    },
    {
      "epoch": 1.04,
      "grad_norm": 8.873979568481445,
      "learning_rate": 1.4386441647597255e-05,
      "loss": 0.8049,
      "step": 3639
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.359341621398926,
      "learning_rate": 1.4382151029748285e-05,
      "loss": 0.7929,
      "step": 3640
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.25731372833252,
      "learning_rate": 1.4377860411899314e-05,
      "loss": 0.7527,
      "step": 3641
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.850419998168945,
      "learning_rate": 1.4373569794050343e-05,
      "loss": 0.8118,
      "step": 3642
    },
    {
      "epoch": 1.04,
      "grad_norm": 12.61296272277832,
      "learning_rate": 1.4369279176201373e-05,
      "loss": 0.9118,
      "step": 3643
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.669021606445312,
      "learning_rate": 1.4364988558352402e-05,
      "loss": 0.8012,
      "step": 3644
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.008260726928711,
      "learning_rate": 1.4360697940503432e-05,
      "loss": 0.9706,
      "step": 3645
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.625000953674316,
      "learning_rate": 1.4356407322654463e-05,
      "loss": 1.0906,
      "step": 3646
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.33679485321045,
      "learning_rate": 1.4352116704805493e-05,
      "loss": 0.7084,
      "step": 3647
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.466826438903809,
      "learning_rate": 1.4347826086956522e-05,
      "loss": 0.6289,
      "step": 3648
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.775632858276367,
      "learning_rate": 1.4343535469107553e-05,
      "loss": 1.0995,
      "step": 3649
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.4617280960083,
      "learning_rate": 1.4339244851258581e-05,
      "loss": 0.9055,
      "step": 3650
    },
    {
      "epoch": 1.04,
      "grad_norm": 8.862994194030762,
      "learning_rate": 1.4334954233409612e-05,
      "loss": 0.7649,
      "step": 3651
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.316967010498047,
      "learning_rate": 1.433066361556064e-05,
      "loss": 0.8565,
      "step": 3652
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.595606803894043,
      "learning_rate": 1.4326372997711671e-05,
      "loss": 0.9249,
      "step": 3653
    },
    {
      "epoch": 1.05,
      "grad_norm": 11.789463996887207,
      "learning_rate": 1.43220823798627e-05,
      "loss": 1.1813,
      "step": 3654
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.194077491760254,
      "learning_rate": 1.431779176201373e-05,
      "loss": 0.7936,
      "step": 3655
    },
    {
      "epoch": 1.05,
      "grad_norm": 9.44644832611084,
      "learning_rate": 1.431350114416476e-05,
      "loss": 0.7989,
      "step": 3656
    },
    {
      "epoch": 1.05,
      "grad_norm": 9.425652503967285,
      "learning_rate": 1.430921052631579e-05,
      "loss": 0.7062,
      "step": 3657
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.893973350524902,
      "learning_rate": 1.430491990846682e-05,
      "loss": 0.846,
      "step": 3658
    },
    {
      "epoch": 1.05,
      "grad_norm": 9.67643928527832,
      "learning_rate": 1.4300629290617849e-05,
      "loss": 0.7771,
      "step": 3659
    },
    {
      "epoch": 1.05,
      "grad_norm": 8.7727689743042,
      "learning_rate": 1.429633867276888e-05,
      "loss": 0.6081,
      "step": 3660
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.030600547790527,
      "learning_rate": 1.4292048054919908e-05,
      "loss": 0.5844,
      "step": 3661
    },
    {
      "epoch": 1.05,
      "grad_norm": 8.970498085021973,
      "learning_rate": 1.4287757437070938e-05,
      "loss": 0.8009,
      "step": 3662
    },
    {
      "epoch": 1.05,
      "grad_norm": 12.848243713378906,
      "learning_rate": 1.4283466819221967e-05,
      "loss": 0.8377,
      "step": 3663
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.326622009277344,
      "learning_rate": 1.4279176201373e-05,
      "loss": 0.7998,
      "step": 3664
    },
    {
      "epoch": 1.05,
      "grad_norm": 7.992002010345459,
      "learning_rate": 1.4274885583524028e-05,
      "loss": 0.6024,
      "step": 3665
    },
    {
      "epoch": 1.05,
      "grad_norm": 9.728657722473145,
      "learning_rate": 1.4270594965675059e-05,
      "loss": 0.8022,
      "step": 3666
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.044332504272461,
      "learning_rate": 1.4266304347826087e-05,
      "loss": 0.7803,
      "step": 3667
    },
    {
      "epoch": 1.05,
      "grad_norm": 11.209845542907715,
      "learning_rate": 1.4262013729977116e-05,
      "loss": 0.7259,
      "step": 3668
    },
    {
      "epoch": 1.05,
      "grad_norm": 11.56793212890625,
      "learning_rate": 1.4257723112128147e-05,
      "loss": 0.8163,
      "step": 3669
    },
    {
      "epoch": 1.05,
      "grad_norm": 11.417867660522461,
      "learning_rate": 1.4253432494279175e-05,
      "loss": 0.8054,
      "step": 3670
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.159390449523926,
      "learning_rate": 1.4249141876430206e-05,
      "loss": 0.9439,
      "step": 3671
    },
    {
      "epoch": 1.05,
      "grad_norm": 12.502946853637695,
      "learning_rate": 1.4244851258581236e-05,
      "loss": 1.1544,
      "step": 3672
    },
    {
      "epoch": 1.05,
      "grad_norm": 8.187594413757324,
      "learning_rate": 1.4240560640732267e-05,
      "loss": 0.6972,
      "step": 3673
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.784448623657227,
      "learning_rate": 1.4236270022883296e-05,
      "loss": 1.0062,
      "step": 3674
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.996561050415039,
      "learning_rate": 1.4231979405034326e-05,
      "loss": 0.9502,
      "step": 3675
    },
    {
      "epoch": 1.05,
      "grad_norm": 8.331461906433105,
      "learning_rate": 1.4227688787185355e-05,
      "loss": 0.9851,
      "step": 3676
    },
    {
      "epoch": 1.05,
      "grad_norm": 7.585824012756348,
      "learning_rate": 1.4223398169336385e-05,
      "loss": 0.8282,
      "step": 3677
    },
    {
      "epoch": 1.05,
      "grad_norm": 7.308041572570801,
      "learning_rate": 1.4219107551487414e-05,
      "loss": 0.6192,
      "step": 3678
    },
    {
      "epoch": 1.05,
      "grad_norm": 11.896117210388184,
      "learning_rate": 1.4214816933638444e-05,
      "loss": 1.153,
      "step": 3679
    },
    {
      "epoch": 1.05,
      "grad_norm": 11.243952751159668,
      "learning_rate": 1.4210526315789473e-05,
      "loss": 0.8675,
      "step": 3680
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.377750396728516,
      "learning_rate": 1.4206235697940504e-05,
      "loss": 0.8831,
      "step": 3681
    },
    {
      "epoch": 1.05,
      "grad_norm": 9.671457290649414,
      "learning_rate": 1.4201945080091534e-05,
      "loss": 0.7269,
      "step": 3682
    },
    {
      "epoch": 1.05,
      "grad_norm": 11.001798629760742,
      "learning_rate": 1.4197654462242563e-05,
      "loss": 0.8837,
      "step": 3683
    },
    {
      "epoch": 1.05,
      "grad_norm": 9.80832290649414,
      "learning_rate": 1.4193363844393593e-05,
      "loss": 0.8081,
      "step": 3684
    },
    {
      "epoch": 1.05,
      "grad_norm": 9.946518898010254,
      "learning_rate": 1.4189073226544622e-05,
      "loss": 0.8903,
      "step": 3685
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.565032005310059,
      "learning_rate": 1.4184782608695653e-05,
      "loss": 0.9812,
      "step": 3686
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.964146614074707,
      "learning_rate": 1.4180491990846681e-05,
      "loss": 0.9207,
      "step": 3687
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.63786792755127,
      "learning_rate": 1.4176201372997712e-05,
      "loss": 0.9114,
      "step": 3688
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.37463665008545,
      "learning_rate": 1.4171910755148742e-05,
      "loss": 0.8679,
      "step": 3689
    },
    {
      "epoch": 1.06,
      "grad_norm": 11.379569053649902,
      "learning_rate": 1.4167620137299773e-05,
      "loss": 0.923,
      "step": 3690
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.972661018371582,
      "learning_rate": 1.4163329519450802e-05,
      "loss": 0.7566,
      "step": 3691
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.997492790222168,
      "learning_rate": 1.4159038901601832e-05,
      "loss": 0.9684,
      "step": 3692
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.21652603149414,
      "learning_rate": 1.415474828375286e-05,
      "loss": 0.723,
      "step": 3693
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.660791397094727,
      "learning_rate": 1.415045766590389e-05,
      "loss": 0.8044,
      "step": 3694
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.309823036193848,
      "learning_rate": 1.414616704805492e-05,
      "loss": 0.8231,
      "step": 3695
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.975504875183105,
      "learning_rate": 1.4141876430205949e-05,
      "loss": 0.8977,
      "step": 3696
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.991817474365234,
      "learning_rate": 1.413758581235698e-05,
      "loss": 0.6957,
      "step": 3697
    },
    {
      "epoch": 1.06,
      "grad_norm": 7.990804672241211,
      "learning_rate": 1.413329519450801e-05,
      "loss": 0.6785,
      "step": 3698
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.592016220092773,
      "learning_rate": 1.412900457665904e-05,
      "loss": 1.0659,
      "step": 3699
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.965288162231445,
      "learning_rate": 1.4124713958810069e-05,
      "loss": 0.8429,
      "step": 3700
    },
    {
      "epoch": 1.06,
      "grad_norm": 8.703361511230469,
      "learning_rate": 1.41204233409611e-05,
      "loss": 0.8845,
      "step": 3701
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.193435668945312,
      "learning_rate": 1.4116132723112128e-05,
      "loss": 0.8447,
      "step": 3702
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.333001136779785,
      "learning_rate": 1.4111842105263159e-05,
      "loss": 0.8867,
      "step": 3703
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.7531156539917,
      "learning_rate": 1.4107551487414187e-05,
      "loss": 1.0541,
      "step": 3704
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.385746955871582,
      "learning_rate": 1.4103260869565218e-05,
      "loss": 0.852,
      "step": 3705
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.889044761657715,
      "learning_rate": 1.4098970251716248e-05,
      "loss": 0.9843,
      "step": 3706
    },
    {
      "epoch": 1.06,
      "grad_norm": 8.673348426818848,
      "learning_rate": 1.4094679633867277e-05,
      "loss": 0.8038,
      "step": 3707
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.381312370300293,
      "learning_rate": 1.4090389016018308e-05,
      "loss": 0.7471,
      "step": 3708
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.749486923217773,
      "learning_rate": 1.4086098398169336e-05,
      "loss": 0.7541,
      "step": 3709
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.783663749694824,
      "learning_rate": 1.4081807780320367e-05,
      "loss": 0.8046,
      "step": 3710
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.428963661193848,
      "learning_rate": 1.4077517162471396e-05,
      "loss": 0.8263,
      "step": 3711
    },
    {
      "epoch": 1.06,
      "grad_norm": 8.74447250366211,
      "learning_rate": 1.4073226544622426e-05,
      "loss": 0.7661,
      "step": 3712
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.322054862976074,
      "learning_rate": 1.4068935926773455e-05,
      "loss": 0.9435,
      "step": 3713
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.705693244934082,
      "learning_rate": 1.4064645308924485e-05,
      "loss": 0.8276,
      "step": 3714
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.739495277404785,
      "learning_rate": 1.4060354691075516e-05,
      "loss": 0.5656,
      "step": 3715
    },
    {
      "epoch": 1.06,
      "grad_norm": 11.19046401977539,
      "learning_rate": 1.4056064073226546e-05,
      "loss": 0.8381,
      "step": 3716
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.727036476135254,
      "learning_rate": 1.4051773455377575e-05,
      "loss": 0.8366,
      "step": 3717
    },
    {
      "epoch": 1.06,
      "grad_norm": 7.8415207862854,
      "learning_rate": 1.4047482837528605e-05,
      "loss": 0.7503,
      "step": 3718
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.288435935974121,
      "learning_rate": 1.4043192219679634e-05,
      "loss": 0.7301,
      "step": 3719
    },
    {
      "epoch": 1.06,
      "grad_norm": 8.485249519348145,
      "learning_rate": 1.4038901601830663e-05,
      "loss": 0.8689,
      "step": 3720
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.322175025939941,
      "learning_rate": 1.4034610983981693e-05,
      "loss": 0.9206,
      "step": 3721
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.204500198364258,
      "learning_rate": 1.4030320366132722e-05,
      "loss": 0.6163,
      "step": 3722
    },
    {
      "epoch": 1.06,
      "grad_norm": 12.585383415222168,
      "learning_rate": 1.4026029748283753e-05,
      "loss": 0.6983,
      "step": 3723
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.98762321472168,
      "learning_rate": 1.4021739130434783e-05,
      "loss": 0.6787,
      "step": 3724
    },
    {
      "epoch": 1.07,
      "grad_norm": 15.406034469604492,
      "learning_rate": 1.4017448512585814e-05,
      "loss": 0.947,
      "step": 3725
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.489848136901855,
      "learning_rate": 1.4013157894736842e-05,
      "loss": 0.7061,
      "step": 3726
    },
    {
      "epoch": 1.07,
      "grad_norm": 8.680780410766602,
      "learning_rate": 1.4008867276887873e-05,
      "loss": 0.6413,
      "step": 3727
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.098986625671387,
      "learning_rate": 1.4004576659038902e-05,
      "loss": 0.8731,
      "step": 3728
    },
    {
      "epoch": 1.07,
      "grad_norm": 9.972673416137695,
      "learning_rate": 1.4000286041189932e-05,
      "loss": 0.6916,
      "step": 3729
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.532546997070312,
      "learning_rate": 1.399599542334096e-05,
      "loss": 0.8237,
      "step": 3730
    },
    {
      "epoch": 1.07,
      "grad_norm": 9.868332862854004,
      "learning_rate": 1.3991704805491991e-05,
      "loss": 0.5668,
      "step": 3731
    },
    {
      "epoch": 1.07,
      "grad_norm": 9.763664245605469,
      "learning_rate": 1.3987414187643022e-05,
      "loss": 0.7665,
      "step": 3732
    },
    {
      "epoch": 1.07,
      "grad_norm": 8.648012161254883,
      "learning_rate": 1.398312356979405e-05,
      "loss": 0.526,
      "step": 3733
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.330743789672852,
      "learning_rate": 1.3978832951945081e-05,
      "loss": 0.838,
      "step": 3734
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.847214698791504,
      "learning_rate": 1.397454233409611e-05,
      "loss": 0.8665,
      "step": 3735
    },
    {
      "epoch": 1.07,
      "grad_norm": 12.673944473266602,
      "learning_rate": 1.397025171624714e-05,
      "loss": 0.9717,
      "step": 3736
    },
    {
      "epoch": 1.07,
      "grad_norm": 12.03902530670166,
      "learning_rate": 1.3965961098398169e-05,
      "loss": 1.05,
      "step": 3737
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.373015403747559,
      "learning_rate": 1.39616704805492e-05,
      "loss": 0.7497,
      "step": 3738
    },
    {
      "epoch": 1.07,
      "grad_norm": 8.553824424743652,
      "learning_rate": 1.3957379862700228e-05,
      "loss": 0.6724,
      "step": 3739
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.366500854492188,
      "learning_rate": 1.3953089244851259e-05,
      "loss": 0.8859,
      "step": 3740
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.80486011505127,
      "learning_rate": 1.3948798627002289e-05,
      "loss": 1.0033,
      "step": 3741
    },
    {
      "epoch": 1.07,
      "grad_norm": 11.4472017288208,
      "learning_rate": 1.394450800915332e-05,
      "loss": 0.9502,
      "step": 3742
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.950733184814453,
      "learning_rate": 1.3940217391304348e-05,
      "loss": 0.8685,
      "step": 3743
    },
    {
      "epoch": 1.07,
      "grad_norm": 11.05859661102295,
      "learning_rate": 1.3935926773455379e-05,
      "loss": 0.868,
      "step": 3744
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.811240196228027,
      "learning_rate": 1.3931636155606408e-05,
      "loss": 0.9819,
      "step": 3745
    },
    {
      "epoch": 1.07,
      "grad_norm": 9.162971496582031,
      "learning_rate": 1.3927345537757436e-05,
      "loss": 0.643,
      "step": 3746
    },
    {
      "epoch": 1.07,
      "grad_norm": 9.291269302368164,
      "learning_rate": 1.3923054919908467e-05,
      "loss": 0.8225,
      "step": 3747
    },
    {
      "epoch": 1.07,
      "grad_norm": 12.17906379699707,
      "learning_rate": 1.3918764302059496e-05,
      "loss": 1.1213,
      "step": 3748
    },
    {
      "epoch": 1.07,
      "grad_norm": 11.068907737731934,
      "learning_rate": 1.3914473684210528e-05,
      "loss": 1.0983,
      "step": 3749
    },
    {
      "epoch": 1.07,
      "grad_norm": 11.328256607055664,
      "learning_rate": 1.3910183066361556e-05,
      "loss": 1.0155,
      "step": 3750
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.168045997619629,
      "learning_rate": 1.3905892448512587e-05,
      "loss": 0.8069,
      "step": 3751
    },
    {
      "epoch": 1.07,
      "grad_norm": 9.096185684204102,
      "learning_rate": 1.3901601830663616e-05,
      "loss": 0.7577,
      "step": 3752
    },
    {
      "epoch": 1.07,
      "grad_norm": 7.600979328155518,
      "learning_rate": 1.3897311212814646e-05,
      "loss": 0.5877,
      "step": 3753
    },
    {
      "epoch": 1.07,
      "grad_norm": 9.016931533813477,
      "learning_rate": 1.3893020594965675e-05,
      "loss": 0.6699,
      "step": 3754
    },
    {
      "epoch": 1.07,
      "grad_norm": 7.853875160217285,
      "learning_rate": 1.3888729977116705e-05,
      "loss": 0.6426,
      "step": 3755
    },
    {
      "epoch": 1.07,
      "grad_norm": 9.79909896850586,
      "learning_rate": 1.3884439359267734e-05,
      "loss": 0.8809,
      "step": 3756
    },
    {
      "epoch": 1.07,
      "grad_norm": 9.951688766479492,
      "learning_rate": 1.3880148741418765e-05,
      "loss": 0.9389,
      "step": 3757
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.226069450378418,
      "learning_rate": 1.3875858123569795e-05,
      "loss": 1.003,
      "step": 3758
    },
    {
      "epoch": 1.08,
      "grad_norm": 7.558326244354248,
      "learning_rate": 1.3871567505720824e-05,
      "loss": 0.787,
      "step": 3759
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.79404067993164,
      "learning_rate": 1.3867276887871854e-05,
      "loss": 0.8167,
      "step": 3760
    },
    {
      "epoch": 1.08,
      "grad_norm": 10.235540390014648,
      "learning_rate": 1.3862986270022883e-05,
      "loss": 0.886,
      "step": 3761
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.964478492736816,
      "learning_rate": 1.3858695652173914e-05,
      "loss": 0.7499,
      "step": 3762
    },
    {
      "epoch": 1.08,
      "grad_norm": 10.020101547241211,
      "learning_rate": 1.3854405034324942e-05,
      "loss": 0.6242,
      "step": 3763
    },
    {
      "epoch": 1.08,
      "grad_norm": 12.146797180175781,
      "learning_rate": 1.3850114416475973e-05,
      "loss": 0.9343,
      "step": 3764
    },
    {
      "epoch": 1.08,
      "grad_norm": 8.966434478759766,
      "learning_rate": 1.3845823798627002e-05,
      "loss": 0.9949,
      "step": 3765
    },
    {
      "epoch": 1.08,
      "grad_norm": 8.525348663330078,
      "learning_rate": 1.3841533180778034e-05,
      "loss": 0.5102,
      "step": 3766
    },
    {
      "epoch": 1.08,
      "grad_norm": 10.42138957977295,
      "learning_rate": 1.3837242562929062e-05,
      "loss": 0.8386,
      "step": 3767
    },
    {
      "epoch": 1.08,
      "grad_norm": 12.72672176361084,
      "learning_rate": 1.3832951945080093e-05,
      "loss": 1.0639,
      "step": 3768
    },
    {
      "epoch": 1.08,
      "grad_norm": 11.651678085327148,
      "learning_rate": 1.3828661327231122e-05,
      "loss": 0.8836,
      "step": 3769
    },
    {
      "epoch": 1.08,
      "grad_norm": 10.607648849487305,
      "learning_rate": 1.3824370709382152e-05,
      "loss": 0.9617,
      "step": 3770
    },
    {
      "epoch": 1.08,
      "grad_norm": 10.19911003112793,
      "learning_rate": 1.3820080091533181e-05,
      "loss": 0.7797,
      "step": 3771
    },
    {
      "epoch": 1.08,
      "grad_norm": 8.973006248474121,
      "learning_rate": 1.3815789473684211e-05,
      "loss": 0.7331,
      "step": 3772
    },
    {
      "epoch": 1.08,
      "grad_norm": 10.524520874023438,
      "learning_rate": 1.381149885583524e-05,
      "loss": 0.834,
      "step": 3773
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.342079162597656,
      "learning_rate": 1.3807208237986269e-05,
      "loss": 0.6822,
      "step": 3774
    },
    {
      "epoch": 1.08,
      "grad_norm": 10.43412971496582,
      "learning_rate": 1.3802917620137301e-05,
      "loss": 0.8916,
      "step": 3775
    },
    {
      "epoch": 1.08,
      "grad_norm": 10.923917770385742,
      "learning_rate": 1.379862700228833e-05,
      "loss": 1.0817,
      "step": 3776
    },
    {
      "epoch": 1.08,
      "grad_norm": 10.417581558227539,
      "learning_rate": 1.379433638443936e-05,
      "loss": 0.6987,
      "step": 3777
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.958463668823242,
      "learning_rate": 1.3790045766590389e-05,
      "loss": 0.8642,
      "step": 3778
    },
    {
      "epoch": 1.08,
      "grad_norm": 10.143387794494629,
      "learning_rate": 1.378575514874142e-05,
      "loss": 0.6766,
      "step": 3779
    },
    {
      "epoch": 1.08,
      "grad_norm": 8.822731971740723,
      "learning_rate": 1.3781464530892448e-05,
      "loss": 0.7229,
      "step": 3780
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.64212417602539,
      "learning_rate": 1.3777173913043479e-05,
      "loss": 0.7974,
      "step": 3781
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.49821949005127,
      "learning_rate": 1.3772883295194508e-05,
      "loss": 0.5882,
      "step": 3782
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.604161262512207,
      "learning_rate": 1.3768592677345538e-05,
      "loss": 0.7547,
      "step": 3783
    },
    {
      "epoch": 1.08,
      "grad_norm": 10.2879638671875,
      "learning_rate": 1.3764302059496568e-05,
      "loss": 0.9607,
      "step": 3784
    },
    {
      "epoch": 1.08,
      "grad_norm": 8.067282676696777,
      "learning_rate": 1.3760011441647599e-05,
      "loss": 0.797,
      "step": 3785
    },
    {
      "epoch": 1.08,
      "grad_norm": 11.269867897033691,
      "learning_rate": 1.3755720823798628e-05,
      "loss": 0.9492,
      "step": 3786
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.921648979187012,
      "learning_rate": 1.3751430205949656e-05,
      "loss": 0.6434,
      "step": 3787
    },
    {
      "epoch": 1.08,
      "grad_norm": 8.82890510559082,
      "learning_rate": 1.3747139588100687e-05,
      "loss": 0.8586,
      "step": 3788
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.924644470214844,
      "learning_rate": 1.3742848970251716e-05,
      "loss": 0.7732,
      "step": 3789
    },
    {
      "epoch": 1.08,
      "grad_norm": 8.59892463684082,
      "learning_rate": 1.3738558352402746e-05,
      "loss": 0.7436,
      "step": 3790
    },
    {
      "epoch": 1.08,
      "grad_norm": 10.175561904907227,
      "learning_rate": 1.3734267734553775e-05,
      "loss": 0.8205,
      "step": 3791
    },
    {
      "epoch": 1.08,
      "grad_norm": 8.727319717407227,
      "learning_rate": 1.3729977116704807e-05,
      "loss": 0.6489,
      "step": 3792
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.501527786254883,
      "learning_rate": 1.3725686498855836e-05,
      "loss": 0.7864,
      "step": 3793
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.407328605651855,
      "learning_rate": 1.3721395881006866e-05,
      "loss": 0.707,
      "step": 3794
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.251099586486816,
      "learning_rate": 1.3717105263157895e-05,
      "loss": 0.7515,
      "step": 3795
    },
    {
      "epoch": 1.09,
      "grad_norm": 13.685705184936523,
      "learning_rate": 1.3712814645308926e-05,
      "loss": 1.0767,
      "step": 3796
    },
    {
      "epoch": 1.09,
      "grad_norm": 10.53074836730957,
      "learning_rate": 1.3708524027459954e-05,
      "loss": 0.8211,
      "step": 3797
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.404327392578125,
      "learning_rate": 1.3704233409610985e-05,
      "loss": 0.7364,
      "step": 3798
    },
    {
      "epoch": 1.09,
      "grad_norm": 8.201828956604004,
      "learning_rate": 1.3699942791762014e-05,
      "loss": 0.8571,
      "step": 3799
    },
    {
      "epoch": 1.09,
      "grad_norm": 11.381919860839844,
      "learning_rate": 1.3695652173913042e-05,
      "loss": 1.0725,
      "step": 3800
    },
    {
      "epoch": 1.09,
      "grad_norm": 11.355937957763672,
      "learning_rate": 1.3691361556064074e-05,
      "loss": 0.769,
      "step": 3801
    },
    {
      "epoch": 1.09,
      "grad_norm": 10.03324031829834,
      "learning_rate": 1.3687070938215103e-05,
      "loss": 0.7663,
      "step": 3802
    },
    {
      "epoch": 1.09,
      "grad_norm": 8.717913627624512,
      "learning_rate": 1.3682780320366134e-05,
      "loss": 0.6451,
      "step": 3803
    },
    {
      "epoch": 1.09,
      "grad_norm": 13.435998916625977,
      "learning_rate": 1.3678489702517162e-05,
      "loss": 0.9202,
      "step": 3804
    },
    {
      "epoch": 1.09,
      "grad_norm": 13.894830703735352,
      "learning_rate": 1.3674199084668193e-05,
      "loss": 0.9243,
      "step": 3805
    },
    {
      "epoch": 1.09,
      "grad_norm": 11.40660285949707,
      "learning_rate": 1.3669908466819222e-05,
      "loss": 1.0794,
      "step": 3806
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.202834129333496,
      "learning_rate": 1.3665617848970252e-05,
      "loss": 0.6797,
      "step": 3807
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.211841583251953,
      "learning_rate": 1.3661327231121281e-05,
      "loss": 0.7329,
      "step": 3808
    },
    {
      "epoch": 1.09,
      "grad_norm": 11.23617172241211,
      "learning_rate": 1.3657036613272313e-05,
      "loss": 0.8237,
      "step": 3809
    },
    {
      "epoch": 1.09,
      "grad_norm": 11.655969619750977,
      "learning_rate": 1.3652745995423342e-05,
      "loss": 0.9963,
      "step": 3810
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.310397148132324,
      "learning_rate": 1.3648455377574372e-05,
      "loss": 0.7312,
      "step": 3811
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.105588912963867,
      "learning_rate": 1.3644164759725401e-05,
      "loss": 0.6932,
      "step": 3812
    },
    {
      "epoch": 1.09,
      "grad_norm": 8.658004760742188,
      "learning_rate": 1.363987414187643e-05,
      "loss": 0.8755,
      "step": 3813
    },
    {
      "epoch": 1.09,
      "grad_norm": 11.68152904510498,
      "learning_rate": 1.363558352402746e-05,
      "loss": 1.0816,
      "step": 3814
    },
    {
      "epoch": 1.09,
      "grad_norm": 13.390514373779297,
      "learning_rate": 1.3631292906178489e-05,
      "loss": 1.0396,
      "step": 3815
    },
    {
      "epoch": 1.09,
      "grad_norm": 10.660036087036133,
      "learning_rate": 1.362700228832952e-05,
      "loss": 0.8675,
      "step": 3816
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.080789566040039,
      "learning_rate": 1.3622711670480548e-05,
      "loss": 0.7688,
      "step": 3817
    },
    {
      "epoch": 1.09,
      "grad_norm": 14.014163970947266,
      "learning_rate": 1.361842105263158e-05,
      "loss": 0.9729,
      "step": 3818
    },
    {
      "epoch": 1.09,
      "grad_norm": 10.439164161682129,
      "learning_rate": 1.361413043478261e-05,
      "loss": 0.8642,
      "step": 3819
    },
    {
      "epoch": 1.09,
      "grad_norm": 8.424198150634766,
      "learning_rate": 1.360983981693364e-05,
      "loss": 0.8619,
      "step": 3820
    },
    {
      "epoch": 1.09,
      "grad_norm": 11.352283477783203,
      "learning_rate": 1.3605549199084668e-05,
      "loss": 0.728,
      "step": 3821
    },
    {
      "epoch": 1.09,
      "grad_norm": 10.121688842773438,
      "learning_rate": 1.3601258581235699e-05,
      "loss": 0.7932,
      "step": 3822
    },
    {
      "epoch": 1.09,
      "grad_norm": 8.25421142578125,
      "learning_rate": 1.3596967963386728e-05,
      "loss": 0.7211,
      "step": 3823
    },
    {
      "epoch": 1.09,
      "grad_norm": 8.372735977172852,
      "learning_rate": 1.3592677345537758e-05,
      "loss": 0.8354,
      "step": 3824
    },
    {
      "epoch": 1.09,
      "grad_norm": 8.384273529052734,
      "learning_rate": 1.3588386727688787e-05,
      "loss": 0.7215,
      "step": 3825
    },
    {
      "epoch": 1.09,
      "grad_norm": 12.669160842895508,
      "learning_rate": 1.3584096109839817e-05,
      "loss": 0.9718,
      "step": 3826
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.039535522460938,
      "learning_rate": 1.3579805491990848e-05,
      "loss": 0.8118,
      "step": 3827
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.365995407104492,
      "learning_rate": 1.3575514874141877e-05,
      "loss": 0.7145,
      "step": 3828
    },
    {
      "epoch": 1.1,
      "grad_norm": 11.246047973632812,
      "learning_rate": 1.3571224256292907e-05,
      "loss": 0.9481,
      "step": 3829
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.856829643249512,
      "learning_rate": 1.3566933638443936e-05,
      "loss": 0.9542,
      "step": 3830
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.257083892822266,
      "learning_rate": 1.3562643020594966e-05,
      "loss": 0.9543,
      "step": 3831
    },
    {
      "epoch": 1.1,
      "grad_norm": 8.790023803710938,
      "learning_rate": 1.3558352402745995e-05,
      "loss": 0.6408,
      "step": 3832
    },
    {
      "epoch": 1.1,
      "grad_norm": 11.7524995803833,
      "learning_rate": 1.3554061784897025e-05,
      "loss": 1.1864,
      "step": 3833
    },
    {
      "epoch": 1.1,
      "grad_norm": 11.273727416992188,
      "learning_rate": 1.3549771167048054e-05,
      "loss": 1.1412,
      "step": 3834
    },
    {
      "epoch": 1.1,
      "grad_norm": 9.125393867492676,
      "learning_rate": 1.3545480549199086e-05,
      "loss": 0.9648,
      "step": 3835
    },
    {
      "epoch": 1.1,
      "grad_norm": 9.272759437561035,
      "learning_rate": 1.3541189931350115e-05,
      "loss": 0.6963,
      "step": 3836
    },
    {
      "epoch": 1.1,
      "grad_norm": 8.003986358642578,
      "learning_rate": 1.3536899313501146e-05,
      "loss": 0.8362,
      "step": 3837
    },
    {
      "epoch": 1.1,
      "grad_norm": 7.971385955810547,
      "learning_rate": 1.3532608695652174e-05,
      "loss": 0.6797,
      "step": 3838
    },
    {
      "epoch": 1.1,
      "grad_norm": 8.665493965148926,
      "learning_rate": 1.3528318077803203e-05,
      "loss": 0.8865,
      "step": 3839
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.29703426361084,
      "learning_rate": 1.3524027459954234e-05,
      "loss": 1.0045,
      "step": 3840
    },
    {
      "epoch": 1.1,
      "grad_norm": 8.791654586791992,
      "learning_rate": 1.3519736842105262e-05,
      "loss": 0.8541,
      "step": 3841
    },
    {
      "epoch": 1.1,
      "grad_norm": 9.257181167602539,
      "learning_rate": 1.3515446224256293e-05,
      "loss": 0.8603,
      "step": 3842
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.332367897033691,
      "learning_rate": 1.3511155606407322e-05,
      "loss": 0.8712,
      "step": 3843
    },
    {
      "epoch": 1.1,
      "grad_norm": 11.061684608459473,
      "learning_rate": 1.3506864988558354e-05,
      "loss": 1.1324,
      "step": 3844
    },
    {
      "epoch": 1.1,
      "grad_norm": 9.152527809143066,
      "learning_rate": 1.3502574370709383e-05,
      "loss": 0.7468,
      "step": 3845
    },
    {
      "epoch": 1.1,
      "grad_norm": 11.406342506408691,
      "learning_rate": 1.3498283752860413e-05,
      "loss": 0.8362,
      "step": 3846
    },
    {
      "epoch": 1.1,
      "grad_norm": 8.86299991607666,
      "learning_rate": 1.3493993135011442e-05,
      "loss": 0.591,
      "step": 3847
    },
    {
      "epoch": 1.1,
      "grad_norm": 9.53094482421875,
      "learning_rate": 1.3489702517162472e-05,
      "loss": 0.7772,
      "step": 3848
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.275636672973633,
      "learning_rate": 1.3485411899313501e-05,
      "loss": 0.8751,
      "step": 3849
    },
    {
      "epoch": 1.1,
      "grad_norm": 11.548718452453613,
      "learning_rate": 1.3481121281464531e-05,
      "loss": 0.709,
      "step": 3850
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.843782424926758,
      "learning_rate": 1.347683066361556e-05,
      "loss": 0.8233,
      "step": 3851
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.66096305847168,
      "learning_rate": 1.347254004576659e-05,
      "loss": 0.8643,
      "step": 3852
    },
    {
      "epoch": 1.1,
      "grad_norm": 7.803994178771973,
      "learning_rate": 1.3468249427917621e-05,
      "loss": 0.6668,
      "step": 3853
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.722987174987793,
      "learning_rate": 1.346395881006865e-05,
      "loss": 0.8912,
      "step": 3854
    },
    {
      "epoch": 1.1,
      "grad_norm": 8.521932601928711,
      "learning_rate": 1.345966819221968e-05,
      "loss": 0.8668,
      "step": 3855
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.58981990814209,
      "learning_rate": 1.345537757437071e-05,
      "loss": 0.8814,
      "step": 3856
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.837814331054688,
      "learning_rate": 1.345108695652174e-05,
      "loss": 0.9011,
      "step": 3857
    },
    {
      "epoch": 1.1,
      "grad_norm": 9.734378814697266,
      "learning_rate": 1.3446796338672768e-05,
      "loss": 0.8212,
      "step": 3858
    },
    {
      "epoch": 1.1,
      "grad_norm": 9.41157341003418,
      "learning_rate": 1.3442505720823799e-05,
      "loss": 0.7862,
      "step": 3859
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.327248573303223,
      "learning_rate": 1.3438215102974828e-05,
      "loss": 0.8898,
      "step": 3860
    },
    {
      "epoch": 1.1,
      "grad_norm": 11.293153762817383,
      "learning_rate": 1.343392448512586e-05,
      "loss": 0.8931,
      "step": 3861
    },
    {
      "epoch": 1.1,
      "grad_norm": 8.565530776977539,
      "learning_rate": 1.3429633867276889e-05,
      "loss": 0.7467,
      "step": 3862
    },
    {
      "epoch": 1.1,
      "grad_norm": 9.503556251525879,
      "learning_rate": 1.3425343249427919e-05,
      "loss": 0.7474,
      "step": 3863
    },
    {
      "epoch": 1.11,
      "grad_norm": 11.677658081054688,
      "learning_rate": 1.3421052631578948e-05,
      "loss": 0.9569,
      "step": 3864
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.089754104614258,
      "learning_rate": 1.3416762013729977e-05,
      "loss": 0.8821,
      "step": 3865
    },
    {
      "epoch": 1.11,
      "grad_norm": 9.392486572265625,
      "learning_rate": 1.3412471395881007e-05,
      "loss": 0.6777,
      "step": 3866
    },
    {
      "epoch": 1.11,
      "grad_norm": 9.124673843383789,
      "learning_rate": 1.3408180778032036e-05,
      "loss": 0.7258,
      "step": 3867
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.929582595825195,
      "learning_rate": 1.3403890160183066e-05,
      "loss": 1.1245,
      "step": 3868
    },
    {
      "epoch": 1.11,
      "grad_norm": 13.617674827575684,
      "learning_rate": 1.3399599542334097e-05,
      "loss": 1.2321,
      "step": 3869
    },
    {
      "epoch": 1.11,
      "grad_norm": 6.998832702636719,
      "learning_rate": 1.3395308924485127e-05,
      "loss": 0.5102,
      "step": 3870
    },
    {
      "epoch": 1.11,
      "grad_norm": 8.23576545715332,
      "learning_rate": 1.3391018306636156e-05,
      "loss": 0.6787,
      "step": 3871
    },
    {
      "epoch": 1.11,
      "grad_norm": 9.127097129821777,
      "learning_rate": 1.3386727688787186e-05,
      "loss": 0.8956,
      "step": 3872
    },
    {
      "epoch": 1.11,
      "grad_norm": 8.985982894897461,
      "learning_rate": 1.3382437070938215e-05,
      "loss": 0.8104,
      "step": 3873
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.850921630859375,
      "learning_rate": 1.3378146453089246e-05,
      "loss": 0.8172,
      "step": 3874
    },
    {
      "epoch": 1.11,
      "grad_norm": 11.072077751159668,
      "learning_rate": 1.3373855835240274e-05,
      "loss": 0.8077,
      "step": 3875
    },
    {
      "epoch": 1.11,
      "grad_norm": 9.985342025756836,
      "learning_rate": 1.3369565217391305e-05,
      "loss": 0.7859,
      "step": 3876
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.39843463897705,
      "learning_rate": 1.3365274599542334e-05,
      "loss": 0.9875,
      "step": 3877
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.44593620300293,
      "learning_rate": 1.3360983981693364e-05,
      "loss": 0.7658,
      "step": 3878
    },
    {
      "epoch": 1.11,
      "grad_norm": 7.363400936126709,
      "learning_rate": 1.3356693363844395e-05,
      "loss": 0.7387,
      "step": 3879
    },
    {
      "epoch": 1.11,
      "grad_norm": 8.5333251953125,
      "learning_rate": 1.3352402745995423e-05,
      "loss": 0.6563,
      "step": 3880
    },
    {
      "epoch": 1.11,
      "grad_norm": 11.564074516296387,
      "learning_rate": 1.3348112128146454e-05,
      "loss": 0.7479,
      "step": 3881
    },
    {
      "epoch": 1.11,
      "grad_norm": 11.394442558288574,
      "learning_rate": 1.3343821510297483e-05,
      "loss": 0.8078,
      "step": 3882
    },
    {
      "epoch": 1.11,
      "grad_norm": 11.571203231811523,
      "learning_rate": 1.3339530892448513e-05,
      "loss": 0.9504,
      "step": 3883
    },
    {
      "epoch": 1.11,
      "grad_norm": 9.84980297088623,
      "learning_rate": 1.3335240274599542e-05,
      "loss": 0.6538,
      "step": 3884
    },
    {
      "epoch": 1.11,
      "grad_norm": 9.279945373535156,
      "learning_rate": 1.3330949656750572e-05,
      "loss": 0.8315,
      "step": 3885
    },
    {
      "epoch": 1.11,
      "grad_norm": 8.108189582824707,
      "learning_rate": 1.3326659038901603e-05,
      "loss": 0.6845,
      "step": 3886
    },
    {
      "epoch": 1.11,
      "grad_norm": 11.106013298034668,
      "learning_rate": 1.3322368421052633e-05,
      "loss": 0.9725,
      "step": 3887
    },
    {
      "epoch": 1.11,
      "grad_norm": 9.482634544372559,
      "learning_rate": 1.3318077803203662e-05,
      "loss": 0.7943,
      "step": 3888
    },
    {
      "epoch": 1.11,
      "grad_norm": 8.352679252624512,
      "learning_rate": 1.3313787185354692e-05,
      "loss": 0.7034,
      "step": 3889
    },
    {
      "epoch": 1.11,
      "grad_norm": 9.596867561340332,
      "learning_rate": 1.3309496567505721e-05,
      "loss": 0.7767,
      "step": 3890
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.758633613586426,
      "learning_rate": 1.330520594965675e-05,
      "loss": 1.1519,
      "step": 3891
    },
    {
      "epoch": 1.11,
      "grad_norm": 8.963069915771484,
      "learning_rate": 1.330091533180778e-05,
      "loss": 0.7092,
      "step": 3892
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.518741607666016,
      "learning_rate": 1.329662471395881e-05,
      "loss": 1.0363,
      "step": 3893
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.884881973266602,
      "learning_rate": 1.329233409610984e-05,
      "loss": 0.9321,
      "step": 3894
    },
    {
      "epoch": 1.11,
      "grad_norm": 8.422544479370117,
      "learning_rate": 1.328804347826087e-05,
      "loss": 0.5192,
      "step": 3895
    },
    {
      "epoch": 1.11,
      "grad_norm": 11.745040893554688,
      "learning_rate": 1.32837528604119e-05,
      "loss": 1.0652,
      "step": 3896
    },
    {
      "epoch": 1.11,
      "grad_norm": 8.32690143585205,
      "learning_rate": 1.327946224256293e-05,
      "loss": 0.7271,
      "step": 3897
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.290936470031738,
      "learning_rate": 1.327517162471396e-05,
      "loss": 0.7431,
      "step": 3898
    },
    {
      "epoch": 1.12,
      "grad_norm": 11.083048820495605,
      "learning_rate": 1.3270881006864989e-05,
      "loss": 0.831,
      "step": 3899
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.073370933532715,
      "learning_rate": 1.3266590389016019e-05,
      "loss": 0.874,
      "step": 3900
    },
    {
      "epoch": 1.12,
      "grad_norm": 12.049382209777832,
      "learning_rate": 1.3262299771167048e-05,
      "loss": 0.8641,
      "step": 3901
    },
    {
      "epoch": 1.12,
      "grad_norm": 11.350683212280273,
      "learning_rate": 1.3258009153318078e-05,
      "loss": 0.5894,
      "step": 3902
    },
    {
      "epoch": 1.12,
      "grad_norm": 8.89463996887207,
      "learning_rate": 1.3253718535469107e-05,
      "loss": 0.699,
      "step": 3903
    },
    {
      "epoch": 1.12,
      "grad_norm": 8.856133460998535,
      "learning_rate": 1.3249427917620137e-05,
      "loss": 0.7376,
      "step": 3904
    },
    {
      "epoch": 1.12,
      "grad_norm": 8.090983390808105,
      "learning_rate": 1.3245137299771168e-05,
      "loss": 0.4932,
      "step": 3905
    },
    {
      "epoch": 1.12,
      "grad_norm": 12.358712196350098,
      "learning_rate": 1.3240846681922197e-05,
      "loss": 0.8241,
      "step": 3906
    },
    {
      "epoch": 1.12,
      "grad_norm": 8.918922424316406,
      "learning_rate": 1.3236556064073227e-05,
      "loss": 0.661,
      "step": 3907
    },
    {
      "epoch": 1.12,
      "grad_norm": 13.80139446258545,
      "learning_rate": 1.3232265446224256e-05,
      "loss": 1.0576,
      "step": 3908
    },
    {
      "epoch": 1.12,
      "grad_norm": 8.336238861083984,
      "learning_rate": 1.3227974828375286e-05,
      "loss": 0.6813,
      "step": 3909
    },
    {
      "epoch": 1.12,
      "grad_norm": 8.881771087646484,
      "learning_rate": 1.3223684210526315e-05,
      "loss": 0.5574,
      "step": 3910
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.341936111450195,
      "learning_rate": 1.3219393592677346e-05,
      "loss": 0.779,
      "step": 3911
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.47202205657959,
      "learning_rate": 1.3215102974828376e-05,
      "loss": 0.7365,
      "step": 3912
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.373587608337402,
      "learning_rate": 1.3210812356979407e-05,
      "loss": 0.9122,
      "step": 3913
    },
    {
      "epoch": 1.12,
      "grad_norm": 11.215059280395508,
      "learning_rate": 1.3206521739130435e-05,
      "loss": 0.8947,
      "step": 3914
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.4517822265625,
      "learning_rate": 1.3202231121281466e-05,
      "loss": 0.9194,
      "step": 3915
    },
    {
      "epoch": 1.12,
      "grad_norm": 13.125057220458984,
      "learning_rate": 1.3197940503432495e-05,
      "loss": 1.114,
      "step": 3916
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.323071479797363,
      "learning_rate": 1.3193649885583523e-05,
      "loss": 0.9372,
      "step": 3917
    },
    {
      "epoch": 1.12,
      "grad_norm": 9.739777565002441,
      "learning_rate": 1.3189359267734554e-05,
      "loss": 0.9486,
      "step": 3918
    },
    {
      "epoch": 1.12,
      "grad_norm": 9.97629451751709,
      "learning_rate": 1.3185068649885583e-05,
      "loss": 0.7643,
      "step": 3919
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.741100311279297,
      "learning_rate": 1.3180778032036613e-05,
      "loss": 0.8237,
      "step": 3920
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.449416160583496,
      "learning_rate": 1.3176487414187643e-05,
      "loss": 1.0419,
      "step": 3921
    },
    {
      "epoch": 1.12,
      "grad_norm": 9.64289665222168,
      "learning_rate": 1.3172196796338674e-05,
      "loss": 0.7974,
      "step": 3922
    },
    {
      "epoch": 1.12,
      "grad_norm": 8.52686595916748,
      "learning_rate": 1.3167906178489703e-05,
      "loss": 0.791,
      "step": 3923
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.28368091583252,
      "learning_rate": 1.3163615560640733e-05,
      "loss": 0.7411,
      "step": 3924
    },
    {
      "epoch": 1.12,
      "grad_norm": 8.14837646484375,
      "learning_rate": 1.3159324942791762e-05,
      "loss": 0.805,
      "step": 3925
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.000300407409668,
      "learning_rate": 1.3155034324942792e-05,
      "loss": 0.7476,
      "step": 3926
    },
    {
      "epoch": 1.12,
      "grad_norm": 11.533475875854492,
      "learning_rate": 1.3150743707093821e-05,
      "loss": 0.8103,
      "step": 3927
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.855892181396484,
      "learning_rate": 1.3146453089244852e-05,
      "loss": 0.7256,
      "step": 3928
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.99836254119873,
      "learning_rate": 1.3142162471395882e-05,
      "loss": 1.0138,
      "step": 3929
    },
    {
      "epoch": 1.12,
      "grad_norm": 9.347951889038086,
      "learning_rate": 1.3137871853546911e-05,
      "loss": 0.7863,
      "step": 3930
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.57369613647461,
      "learning_rate": 1.3133581235697941e-05,
      "loss": 0.7666,
      "step": 3931
    },
    {
      "epoch": 1.12,
      "grad_norm": 9.703213691711426,
      "learning_rate": 1.312929061784897e-05,
      "loss": 0.822,
      "step": 3932
    },
    {
      "epoch": 1.12,
      "grad_norm": 8.332062721252441,
      "learning_rate": 1.3125e-05,
      "loss": 0.5574,
      "step": 3933
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.98759651184082,
      "learning_rate": 1.312070938215103e-05,
      "loss": 0.7092,
      "step": 3934
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.835576057434082,
      "learning_rate": 1.311641876430206e-05,
      "loss": 0.7832,
      "step": 3935
    },
    {
      "epoch": 1.13,
      "grad_norm": 10.170145988464355,
      "learning_rate": 1.3112128146453089e-05,
      "loss": 0.6981,
      "step": 3936
    },
    {
      "epoch": 1.13,
      "grad_norm": 10.585335731506348,
      "learning_rate": 1.3107837528604119e-05,
      "loss": 0.8255,
      "step": 3937
    },
    {
      "epoch": 1.13,
      "grad_norm": 8.866064071655273,
      "learning_rate": 1.310354691075515e-05,
      "loss": 0.7015,
      "step": 3938
    },
    {
      "epoch": 1.13,
      "grad_norm": 11.070013999938965,
      "learning_rate": 1.309925629290618e-05,
      "loss": 0.899,
      "step": 3939
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.384458541870117,
      "learning_rate": 1.3094965675057209e-05,
      "loss": 0.4722,
      "step": 3940
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.862875938415527,
      "learning_rate": 1.309067505720824e-05,
      "loss": 0.8842,
      "step": 3941
    },
    {
      "epoch": 1.13,
      "grad_norm": 10.934288024902344,
      "learning_rate": 1.3086384439359268e-05,
      "loss": 0.7596,
      "step": 3942
    },
    {
      "epoch": 1.13,
      "grad_norm": 10.560894012451172,
      "learning_rate": 1.3082093821510297e-05,
      "loss": 0.6473,
      "step": 3943
    },
    {
      "epoch": 1.13,
      "grad_norm": 10.554158210754395,
      "learning_rate": 1.3077803203661327e-05,
      "loss": 0.8742,
      "step": 3944
    },
    {
      "epoch": 1.13,
      "grad_norm": 10.599578857421875,
      "learning_rate": 1.3073512585812356e-05,
      "loss": 0.8191,
      "step": 3945
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.034960746765137,
      "learning_rate": 1.3069221967963388e-05,
      "loss": 0.7124,
      "step": 3946
    },
    {
      "epoch": 1.13,
      "grad_norm": 11.03193473815918,
      "learning_rate": 1.3064931350114417e-05,
      "loss": 0.6531,
      "step": 3947
    },
    {
      "epoch": 1.13,
      "grad_norm": 12.117925643920898,
      "learning_rate": 1.3060640732265447e-05,
      "loss": 1.0116,
      "step": 3948
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.007270812988281,
      "learning_rate": 1.3056350114416476e-05,
      "loss": 0.7321,
      "step": 3949
    },
    {
      "epoch": 1.13,
      "grad_norm": 10.332066535949707,
      "learning_rate": 1.3052059496567507e-05,
      "loss": 1.1249,
      "step": 3950
    },
    {
      "epoch": 1.13,
      "grad_norm": 8.80511474609375,
      "learning_rate": 1.3047768878718535e-05,
      "loss": 0.6779,
      "step": 3951
    },
    {
      "epoch": 1.13,
      "grad_norm": 10.973099708557129,
      "learning_rate": 1.3043478260869566e-05,
      "loss": 0.6649,
      "step": 3952
    },
    {
      "epoch": 1.13,
      "grad_norm": 8.528388977050781,
      "learning_rate": 1.3039187643020595e-05,
      "loss": 0.7054,
      "step": 3953
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.05822467803955,
      "learning_rate": 1.3034897025171625e-05,
      "loss": 0.7759,
      "step": 3954
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.46092414855957,
      "learning_rate": 1.3030606407322655e-05,
      "loss": 0.6742,
      "step": 3955
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.287985801696777,
      "learning_rate": 1.3026315789473684e-05,
      "loss": 0.7852,
      "step": 3956
    },
    {
      "epoch": 1.13,
      "grad_norm": 10.732928276062012,
      "learning_rate": 1.3022025171624715e-05,
      "loss": 0.8557,
      "step": 3957
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.83497142791748,
      "learning_rate": 1.3017734553775743e-05,
      "loss": 0.6626,
      "step": 3958
    },
    {
      "epoch": 1.13,
      "grad_norm": 8.530620574951172,
      "learning_rate": 1.3013443935926774e-05,
      "loss": 0.6577,
      "step": 3959
    },
    {
      "epoch": 1.13,
      "grad_norm": 8.171707153320312,
      "learning_rate": 1.3009153318077803e-05,
      "loss": 0.6615,
      "step": 3960
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.138986587524414,
      "learning_rate": 1.3004862700228833e-05,
      "loss": 0.6291,
      "step": 3961
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.378610610961914,
      "learning_rate": 1.3000572082379862e-05,
      "loss": 0.8419,
      "step": 3962
    },
    {
      "epoch": 1.13,
      "grad_norm": 10.207708358764648,
      "learning_rate": 1.2996281464530894e-05,
      "loss": 0.7725,
      "step": 3963
    },
    {
      "epoch": 1.13,
      "grad_norm": 12.150755882263184,
      "learning_rate": 1.2991990846681923e-05,
      "loss": 1.2711,
      "step": 3964
    },
    {
      "epoch": 1.13,
      "grad_norm": 11.433258056640625,
      "learning_rate": 1.2987700228832953e-05,
      "loss": 0.8698,
      "step": 3965
    },
    {
      "epoch": 1.13,
      "grad_norm": 12.114084243774414,
      "learning_rate": 1.2983409610983982e-05,
      "loss": 0.9742,
      "step": 3966
    },
    {
      "epoch": 1.13,
      "grad_norm": 8.968756675720215,
      "learning_rate": 1.2979118993135013e-05,
      "loss": 0.6515,
      "step": 3967
    },
    {
      "epoch": 1.14,
      "grad_norm": 9.046331405639648,
      "learning_rate": 1.2974828375286041e-05,
      "loss": 0.7982,
      "step": 3968
    },
    {
      "epoch": 1.14,
      "grad_norm": 11.625608444213867,
      "learning_rate": 1.2970537757437072e-05,
      "loss": 1.0168,
      "step": 3969
    },
    {
      "epoch": 1.14,
      "grad_norm": 8.300956726074219,
      "learning_rate": 1.29662471395881e-05,
      "loss": 0.7453,
      "step": 3970
    },
    {
      "epoch": 1.14,
      "grad_norm": 7.171285629272461,
      "learning_rate": 1.296195652173913e-05,
      "loss": 0.5139,
      "step": 3971
    },
    {
      "epoch": 1.14,
      "grad_norm": 7.901523590087891,
      "learning_rate": 1.2957665903890161e-05,
      "loss": 0.6808,
      "step": 3972
    },
    {
      "epoch": 1.14,
      "grad_norm": 10.088178634643555,
      "learning_rate": 1.295337528604119e-05,
      "loss": 1.0606,
      "step": 3973
    },
    {
      "epoch": 1.14,
      "grad_norm": 10.532071113586426,
      "learning_rate": 1.294908466819222e-05,
      "loss": 0.6841,
      "step": 3974
    },
    {
      "epoch": 1.14,
      "grad_norm": 10.62282943725586,
      "learning_rate": 1.294479405034325e-05,
      "loss": 0.9119,
      "step": 3975
    },
    {
      "epoch": 1.14,
      "grad_norm": 9.230091094970703,
      "learning_rate": 1.294050343249428e-05,
      "loss": 0.5679,
      "step": 3976
    },
    {
      "epoch": 1.14,
      "grad_norm": 10.999259948730469,
      "learning_rate": 1.2936212814645309e-05,
      "loss": 0.8994,
      "step": 3977
    },
    {
      "epoch": 1.14,
      "grad_norm": 8.278515815734863,
      "learning_rate": 1.2931922196796339e-05,
      "loss": 0.5793,
      "step": 3978
    },
    {
      "epoch": 1.14,
      "grad_norm": 13.591418266296387,
      "learning_rate": 1.2927631578947368e-05,
      "loss": 0.8064,
      "step": 3979
    },
    {
      "epoch": 1.14,
      "grad_norm": 10.123757362365723,
      "learning_rate": 1.2923340961098398e-05,
      "loss": 0.8161,
      "step": 3980
    },
    {
      "epoch": 1.14,
      "grad_norm": 11.282858848571777,
      "learning_rate": 1.2919050343249429e-05,
      "loss": 0.8874,
      "step": 3981
    },
    {
      "epoch": 1.14,
      "grad_norm": 11.874940872192383,
      "learning_rate": 1.291475972540046e-05,
      "loss": 0.6845,
      "step": 3982
    },
    {
      "epoch": 1.14,
      "grad_norm": 9.55794906616211,
      "learning_rate": 1.2910469107551488e-05,
      "loss": 0.767,
      "step": 3983
    },
    {
      "epoch": 1.14,
      "grad_norm": 8.992429733276367,
      "learning_rate": 1.2906178489702517e-05,
      "loss": 0.5362,
      "step": 3984
    },
    {
      "epoch": 1.14,
      "grad_norm": 9.76063060760498,
      "learning_rate": 1.2901887871853547e-05,
      "loss": 0.6032,
      "step": 3985
    },
    {
      "epoch": 1.14,
      "grad_norm": 11.053279876708984,
      "learning_rate": 1.2897597254004576e-05,
      "loss": 0.8658,
      "step": 3986
    },
    {
      "epoch": 1.14,
      "grad_norm": 9.731647491455078,
      "learning_rate": 1.2893306636155607e-05,
      "loss": 0.6615,
      "step": 3987
    },
    {
      "epoch": 1.14,
      "grad_norm": 9.753975868225098,
      "learning_rate": 1.2889016018306635e-05,
      "loss": 0.702,
      "step": 3988
    },
    {
      "epoch": 1.14,
      "grad_norm": 8.342461585998535,
      "learning_rate": 1.2884725400457667e-05,
      "loss": 0.7509,
      "step": 3989
    },
    {
      "epoch": 1.14,
      "grad_norm": 10.290438652038574,
      "learning_rate": 1.2880434782608696e-05,
      "loss": 0.7519,
      "step": 3990
    },
    {
      "epoch": 1.14,
      "grad_norm": 9.696769714355469,
      "learning_rate": 1.2876144164759727e-05,
      "loss": 0.8955,
      "step": 3991
    },
    {
      "epoch": 1.14,
      "grad_norm": 10.913823127746582,
      "learning_rate": 1.2871853546910755e-05,
      "loss": 0.909,
      "step": 3992
    },
    {
      "epoch": 1.14,
      "grad_norm": 9.27153205871582,
      "learning_rate": 1.2867562929061786e-05,
      "loss": 0.7404,
      "step": 3993
    },
    {
      "epoch": 1.14,
      "grad_norm": 9.592728614807129,
      "learning_rate": 1.2863272311212815e-05,
      "loss": 1.0155,
      "step": 3994
    },
    {
      "epoch": 1.14,
      "grad_norm": 10.578533172607422,
      "learning_rate": 1.2858981693363845e-05,
      "loss": 0.7963,
      "step": 3995
    },
    {
      "epoch": 1.14,
      "grad_norm": 9.958884239196777,
      "learning_rate": 1.2854691075514874e-05,
      "loss": 0.9383,
      "step": 3996
    },
    {
      "epoch": 1.14,
      "grad_norm": 12.885809898376465,
      "learning_rate": 1.2850400457665903e-05,
      "loss": 0.9006,
      "step": 3997
    },
    {
      "epoch": 1.14,
      "grad_norm": 10.383514404296875,
      "learning_rate": 1.2846109839816935e-05,
      "loss": 0.8054,
      "step": 3998
    },
    {
      "epoch": 1.14,
      "grad_norm": 9.011198997497559,
      "learning_rate": 1.2841819221967964e-05,
      "loss": 0.7757,
      "step": 3999
    },
    {
      "epoch": 1.14,
      "grad_norm": 10.90754222869873,
      "learning_rate": 1.2837528604118994e-05,
      "loss": 0.667,
      "step": 4000
    },
    {
      "epoch": 1.14,
      "grad_norm": 8.191333770751953,
      "learning_rate": 1.2833237986270023e-05,
      "loss": 0.7656,
      "step": 4001
    },
    {
      "epoch": 1.14,
      "grad_norm": 11.077473640441895,
      "learning_rate": 1.2828947368421053e-05,
      "loss": 0.7835,
      "step": 4002
    },
    {
      "epoch": 1.15,
      "grad_norm": 9.449070930480957,
      "learning_rate": 1.2824656750572082e-05,
      "loss": 0.7441,
      "step": 4003
    },
    {
      "epoch": 1.15,
      "grad_norm": 11.44480037689209,
      "learning_rate": 1.2820366132723113e-05,
      "loss": 0.7525,
      "step": 4004
    },
    {
      "epoch": 1.15,
      "grad_norm": 9.595183372497559,
      "learning_rate": 1.2816075514874141e-05,
      "loss": 0.7875,
      "step": 4005
    },
    {
      "epoch": 1.15,
      "grad_norm": 9.679591178894043,
      "learning_rate": 1.2811784897025173e-05,
      "loss": 1.0367,
      "step": 4006
    },
    {
      "epoch": 1.15,
      "grad_norm": 11.832993507385254,
      "learning_rate": 1.2807494279176202e-05,
      "loss": 0.7952,
      "step": 4007
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.265448570251465,
      "learning_rate": 1.2803203661327233e-05,
      "loss": 0.8162,
      "step": 4008
    },
    {
      "epoch": 1.15,
      "grad_norm": 9.365416526794434,
      "learning_rate": 1.2798913043478261e-05,
      "loss": 0.8788,
      "step": 4009
    },
    {
      "epoch": 1.15,
      "grad_norm": 8.634854316711426,
      "learning_rate": 1.279462242562929e-05,
      "loss": 0.688,
      "step": 4010
    },
    {
      "epoch": 1.15,
      "grad_norm": 9.54733657836914,
      "learning_rate": 1.279033180778032e-05,
      "loss": 0.729,
      "step": 4011
    },
    {
      "epoch": 1.15,
      "grad_norm": 9.37033462524414,
      "learning_rate": 1.278604118993135e-05,
      "loss": 0.7131,
      "step": 4012
    },
    {
      "epoch": 1.15,
      "grad_norm": 11.617125511169434,
      "learning_rate": 1.278175057208238e-05,
      "loss": 0.951,
      "step": 4013
    },
    {
      "epoch": 1.15,
      "grad_norm": 11.0932035446167,
      "learning_rate": 1.2777459954233409e-05,
      "loss": 0.7694,
      "step": 4014
    },
    {
      "epoch": 1.15,
      "grad_norm": 11.520350456237793,
      "learning_rate": 1.277316933638444e-05,
      "loss": 1.0484,
      "step": 4015
    },
    {
      "epoch": 1.15,
      "grad_norm": 11.6951322555542,
      "learning_rate": 1.276887871853547e-05,
      "loss": 0.8313,
      "step": 4016
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.18323802947998,
      "learning_rate": 1.27645881006865e-05,
      "loss": 0.7908,
      "step": 4017
    },
    {
      "epoch": 1.15,
      "grad_norm": 15.108692169189453,
      "learning_rate": 1.2760297482837529e-05,
      "loss": 1.0901,
      "step": 4018
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.572484016418457,
      "learning_rate": 1.275600686498856e-05,
      "loss": 0.8613,
      "step": 4019
    },
    {
      "epoch": 1.15,
      "grad_norm": 9.657601356506348,
      "learning_rate": 1.2751716247139588e-05,
      "loss": 0.9282,
      "step": 4020
    },
    {
      "epoch": 1.15,
      "grad_norm": 11.38335132598877,
      "learning_rate": 1.2747425629290619e-05,
      "loss": 0.7404,
      "step": 4021
    },
    {
      "epoch": 1.15,
      "grad_norm": 8.653599739074707,
      "learning_rate": 1.2743135011441647e-05,
      "loss": 0.6312,
      "step": 4022
    },
    {
      "epoch": 1.15,
      "grad_norm": 11.255392074584961,
      "learning_rate": 1.2738844393592678e-05,
      "loss": 1.0732,
      "step": 4023
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.099267959594727,
      "learning_rate": 1.2734553775743708e-05,
      "loss": 0.9457,
      "step": 4024
    },
    {
      "epoch": 1.15,
      "grad_norm": 9.576580047607422,
      "learning_rate": 1.2730263157894737e-05,
      "loss": 0.8892,
      "step": 4025
    },
    {
      "epoch": 1.15,
      "grad_norm": 11.603577613830566,
      "learning_rate": 1.2725972540045767e-05,
      "loss": 0.7765,
      "step": 4026
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.528868675231934,
      "learning_rate": 1.2721681922196796e-05,
      "loss": 0.9207,
      "step": 4027
    },
    {
      "epoch": 1.15,
      "grad_norm": 12.098111152648926,
      "learning_rate": 1.2717391304347827e-05,
      "loss": 1.0071,
      "step": 4028
    },
    {
      "epoch": 1.15,
      "grad_norm": 13.240303039550781,
      "learning_rate": 1.2713100686498855e-05,
      "loss": 0.7886,
      "step": 4029
    },
    {
      "epoch": 1.15,
      "grad_norm": 9.420174598693848,
      "learning_rate": 1.2708810068649886e-05,
      "loss": 0.7186,
      "step": 4030
    },
    {
      "epoch": 1.15,
      "grad_norm": 11.674076080322266,
      "learning_rate": 1.2704519450800915e-05,
      "loss": 0.8748,
      "step": 4031
    },
    {
      "epoch": 1.15,
      "grad_norm": 11.141865730285645,
      "learning_rate": 1.2700228832951947e-05,
      "loss": 0.8274,
      "step": 4032
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.610071182250977,
      "learning_rate": 1.2695938215102976e-05,
      "loss": 0.8465,
      "step": 4033
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.964436531066895,
      "learning_rate": 1.2691647597254006e-05,
      "loss": 0.7164,
      "step": 4034
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.76241397857666,
      "learning_rate": 1.2687356979405035e-05,
      "loss": 0.8646,
      "step": 4035
    },
    {
      "epoch": 1.15,
      "grad_norm": 12.814397811889648,
      "learning_rate": 1.2683066361556064e-05,
      "loss": 0.9439,
      "step": 4036
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.622892379760742,
      "learning_rate": 1.2678775743707094e-05,
      "loss": 1.0646,
      "step": 4037
    },
    {
      "epoch": 1.16,
      "grad_norm": 12.848695755004883,
      "learning_rate": 1.2674485125858123e-05,
      "loss": 1.0784,
      "step": 4038
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.831666946411133,
      "learning_rate": 1.2670194508009153e-05,
      "loss": 0.7179,
      "step": 4039
    },
    {
      "epoch": 1.16,
      "grad_norm": 8.572142601013184,
      "learning_rate": 1.2665903890160182e-05,
      "loss": 0.5997,
      "step": 4040
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.928911209106445,
      "learning_rate": 1.2661613272311214e-05,
      "loss": 1.0008,
      "step": 4041
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.064866065979004,
      "learning_rate": 1.2657322654462243e-05,
      "loss": 0.5705,
      "step": 4042
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.2470121383667,
      "learning_rate": 1.2653032036613273e-05,
      "loss": 1.0596,
      "step": 4043
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.076916694641113,
      "learning_rate": 1.2648741418764302e-05,
      "loss": 0.7603,
      "step": 4044
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.133118629455566,
      "learning_rate": 1.2644450800915333e-05,
      "loss": 0.9168,
      "step": 4045
    },
    {
      "epoch": 1.16,
      "grad_norm": 11.268158912658691,
      "learning_rate": 1.2640160183066361e-05,
      "loss": 0.8666,
      "step": 4046
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.203883171081543,
      "learning_rate": 1.2635869565217392e-05,
      "loss": 0.6609,
      "step": 4047
    },
    {
      "epoch": 1.16,
      "grad_norm": 12.575746536254883,
      "learning_rate": 1.263157894736842e-05,
      "loss": 0.7837,
      "step": 4048
    },
    {
      "epoch": 1.16,
      "grad_norm": 11.600099563598633,
      "learning_rate": 1.2627288329519451e-05,
      "loss": 0.9264,
      "step": 4049
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.319470405578613,
      "learning_rate": 1.2622997711670482e-05,
      "loss": 0.602,
      "step": 4050
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.774476051330566,
      "learning_rate": 1.261870709382151e-05,
      "loss": 0.7665,
      "step": 4051
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.500500679016113,
      "learning_rate": 1.261441647597254e-05,
      "loss": 0.9282,
      "step": 4052
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.612297058105469,
      "learning_rate": 1.261012585812357e-05,
      "loss": 0.8067,
      "step": 4053
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.459089279174805,
      "learning_rate": 1.26058352402746e-05,
      "loss": 0.6766,
      "step": 4054
    },
    {
      "epoch": 1.16,
      "grad_norm": 12.293682098388672,
      "learning_rate": 1.2601544622425629e-05,
      "loss": 0.8211,
      "step": 4055
    },
    {
      "epoch": 1.16,
      "grad_norm": 8.434786796569824,
      "learning_rate": 1.259725400457666e-05,
      "loss": 0.6012,
      "step": 4056
    },
    {
      "epoch": 1.16,
      "grad_norm": 11.257843017578125,
      "learning_rate": 1.2592963386727688e-05,
      "loss": 0.7561,
      "step": 4057
    },
    {
      "epoch": 1.16,
      "grad_norm": 11.848149299621582,
      "learning_rate": 1.258867276887872e-05,
      "loss": 0.8905,
      "step": 4058
    },
    {
      "epoch": 1.16,
      "grad_norm": 8.863398551940918,
      "learning_rate": 1.2584382151029749e-05,
      "loss": 0.8206,
      "step": 4059
    },
    {
      "epoch": 1.16,
      "grad_norm": 7.729502201080322,
      "learning_rate": 1.258009153318078e-05,
      "loss": 0.6038,
      "step": 4060
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.019165992736816,
      "learning_rate": 1.2575800915331808e-05,
      "loss": 0.7282,
      "step": 4061
    },
    {
      "epoch": 1.16,
      "grad_norm": 8.591156959533691,
      "learning_rate": 1.2571510297482837e-05,
      "loss": 0.7666,
      "step": 4062
    },
    {
      "epoch": 1.16,
      "grad_norm": 8.982990264892578,
      "learning_rate": 1.2567219679633867e-05,
      "loss": 0.8294,
      "step": 4063
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.198776245117188,
      "learning_rate": 1.2562929061784896e-05,
      "loss": 0.8153,
      "step": 4064
    },
    {
      "epoch": 1.16,
      "grad_norm": 7.7439422607421875,
      "learning_rate": 1.2558638443935927e-05,
      "loss": 0.65,
      "step": 4065
    },
    {
      "epoch": 1.16,
      "grad_norm": 8.913726806640625,
      "learning_rate": 1.2554347826086957e-05,
      "loss": 0.6944,
      "step": 4066
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.340245246887207,
      "learning_rate": 1.2550057208237988e-05,
      "loss": 0.7589,
      "step": 4067
    },
    {
      "epoch": 1.16,
      "grad_norm": 11.276741027832031,
      "learning_rate": 1.2545766590389016e-05,
      "loss": 0.862,
      "step": 4068
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.607254028320312,
      "learning_rate": 1.2541475972540047e-05,
      "loss": 0.9057,
      "step": 4069
    },
    {
      "epoch": 1.16,
      "grad_norm": 12.084604263305664,
      "learning_rate": 1.2537185354691076e-05,
      "loss": 0.7923,
      "step": 4070
    },
    {
      "epoch": 1.16,
      "grad_norm": 12.961553573608398,
      "learning_rate": 1.2532894736842106e-05,
      "loss": 0.8188,
      "step": 4071
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.144914627075195,
      "learning_rate": 1.2528604118993135e-05,
      "loss": 0.5782,
      "step": 4072
    },
    {
      "epoch": 1.17,
      "grad_norm": 8.926443099975586,
      "learning_rate": 1.2524313501144165e-05,
      "loss": 0.8521,
      "step": 4073
    },
    {
      "epoch": 1.17,
      "grad_norm": 7.8484344482421875,
      "learning_rate": 1.2520022883295194e-05,
      "loss": 0.5207,
      "step": 4074
    },
    {
      "epoch": 1.17,
      "grad_norm": 12.259339332580566,
      "learning_rate": 1.2515732265446224e-05,
      "loss": 1.2073,
      "step": 4075
    },
    {
      "epoch": 1.17,
      "grad_norm": 10.956076622009277,
      "learning_rate": 1.2511441647597255e-05,
      "loss": 0.9287,
      "step": 4076
    },
    {
      "epoch": 1.17,
      "grad_norm": 11.137741088867188,
      "learning_rate": 1.2507151029748284e-05,
      "loss": 0.9011,
      "step": 4077
    },
    {
      "epoch": 1.17,
      "grad_norm": 11.437335014343262,
      "learning_rate": 1.2502860411899314e-05,
      "loss": 0.7987,
      "step": 4078
    },
    {
      "epoch": 1.17,
      "grad_norm": 9.757527351379395,
      "learning_rate": 1.2498569794050343e-05,
      "loss": 0.835,
      "step": 4079
    },
    {
      "epoch": 1.17,
      "grad_norm": 10.70904541015625,
      "learning_rate": 1.2494279176201373e-05,
      "loss": 0.9434,
      "step": 4080
    },
    {
      "epoch": 1.17,
      "grad_norm": 9.849149703979492,
      "learning_rate": 1.2489988558352402e-05,
      "loss": 0.8466,
      "step": 4081
    },
    {
      "epoch": 1.17,
      "grad_norm": 9.947101593017578,
      "learning_rate": 1.2485697940503433e-05,
      "loss": 0.8026,
      "step": 4082
    },
    {
      "epoch": 1.17,
      "grad_norm": 9.381686210632324,
      "learning_rate": 1.2481407322654463e-05,
      "loss": 0.5783,
      "step": 4083
    },
    {
      "epoch": 1.17,
      "grad_norm": 12.77807331085205,
      "learning_rate": 1.2477116704805494e-05,
      "loss": 0.7914,
      "step": 4084
    },
    {
      "epoch": 1.17,
      "grad_norm": 13.345623970031738,
      "learning_rate": 1.2472826086956522e-05,
      "loss": 0.7986,
      "step": 4085
    },
    {
      "epoch": 1.17,
      "grad_norm": 9.868369102478027,
      "learning_rate": 1.2468535469107553e-05,
      "loss": 0.825,
      "step": 4086
    },
    {
      "epoch": 1.17,
      "grad_norm": 9.543618202209473,
      "learning_rate": 1.2464244851258582e-05,
      "loss": 0.794,
      "step": 4087
    },
    {
      "epoch": 1.17,
      "grad_norm": 8.939046859741211,
      "learning_rate": 1.245995423340961e-05,
      "loss": 0.6565,
      "step": 4088
    },
    {
      "epoch": 1.17,
      "grad_norm": 9.811325073242188,
      "learning_rate": 1.245566361556064e-05,
      "loss": 0.7062,
      "step": 4089
    },
    {
      "epoch": 1.17,
      "grad_norm": 10.338138580322266,
      "learning_rate": 1.245137299771167e-05,
      "loss": 0.773,
      "step": 4090
    },
    {
      "epoch": 1.17,
      "grad_norm": 10.663650512695312,
      "learning_rate": 1.24470823798627e-05,
      "loss": 0.7764,
      "step": 4091
    },
    {
      "epoch": 1.17,
      "grad_norm": 10.51228141784668,
      "learning_rate": 1.244279176201373e-05,
      "loss": 0.7541,
      "step": 4092
    },
    {
      "epoch": 1.17,
      "grad_norm": 11.164430618286133,
      "learning_rate": 1.2438501144164761e-05,
      "loss": 0.8271,
      "step": 4093
    },
    {
      "epoch": 1.17,
      "grad_norm": 7.714865207672119,
      "learning_rate": 1.243421052631579e-05,
      "loss": 0.6045,
      "step": 4094
    },
    {
      "epoch": 1.17,
      "grad_norm": 10.589447021484375,
      "learning_rate": 1.242991990846682e-05,
      "loss": 0.7164,
      "step": 4095
    },
    {
      "epoch": 1.17,
      "grad_norm": 11.344781875610352,
      "learning_rate": 1.2425629290617849e-05,
      "loss": 0.7694,
      "step": 4096
    },
    {
      "epoch": 1.17,
      "grad_norm": 13.170841217041016,
      "learning_rate": 1.242133867276888e-05,
      "loss": 0.9305,
      "step": 4097
    },
    {
      "epoch": 1.17,
      "grad_norm": 9.428088188171387,
      "learning_rate": 1.2417048054919908e-05,
      "loss": 0.8694,
      "step": 4098
    },
    {
      "epoch": 1.17,
      "grad_norm": 11.26711654663086,
      "learning_rate": 1.2412757437070939e-05,
      "loss": 0.8744,
      "step": 4099
    },
    {
      "epoch": 1.17,
      "grad_norm": 11.067695617675781,
      "learning_rate": 1.2408466819221967e-05,
      "loss": 0.8843,
      "step": 4100
    },
    {
      "epoch": 1.17,
      "grad_norm": 14.268261909484863,
      "learning_rate": 1.2404176201372998e-05,
      "loss": 0.8601,
      "step": 4101
    },
    {
      "epoch": 1.17,
      "grad_norm": 9.639678001403809,
      "learning_rate": 1.2399885583524028e-05,
      "loss": 0.7101,
      "step": 4102
    },
    {
      "epoch": 1.17,
      "grad_norm": 8.908231735229492,
      "learning_rate": 1.2395594965675057e-05,
      "loss": 0.7455,
      "step": 4103
    },
    {
      "epoch": 1.17,
      "grad_norm": 11.500310897827148,
      "learning_rate": 1.2391304347826088e-05,
      "loss": 0.9985,
      "step": 4104
    },
    {
      "epoch": 1.17,
      "grad_norm": 10.697529792785645,
      "learning_rate": 1.2387013729977116e-05,
      "loss": 0.6407,
      "step": 4105
    },
    {
      "epoch": 1.17,
      "grad_norm": 7.432031154632568,
      "learning_rate": 1.2382723112128147e-05,
      "loss": 0.6257,
      "step": 4106
    },
    {
      "epoch": 1.17,
      "grad_norm": 11.1350679397583,
      "learning_rate": 1.2378432494279176e-05,
      "loss": 0.7523,
      "step": 4107
    },
    {
      "epoch": 1.18,
      "grad_norm": 10.981945037841797,
      "learning_rate": 1.2374141876430206e-05,
      "loss": 1.1006,
      "step": 4108
    },
    {
      "epoch": 1.18,
      "grad_norm": 8.67753791809082,
      "learning_rate": 1.2369851258581236e-05,
      "loss": 0.573,
      "step": 4109
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.949068069458008,
      "learning_rate": 1.2365560640732267e-05,
      "loss": 0.9602,
      "step": 4110
    },
    {
      "epoch": 1.18,
      "grad_norm": 12.82435417175293,
      "learning_rate": 1.2361270022883296e-05,
      "loss": 1.0123,
      "step": 4111
    },
    {
      "epoch": 1.18,
      "grad_norm": 10.570842742919922,
      "learning_rate": 1.2356979405034326e-05,
      "loss": 0.735,
      "step": 4112
    },
    {
      "epoch": 1.18,
      "grad_norm": 10.396074295043945,
      "learning_rate": 1.2352688787185355e-05,
      "loss": 0.7812,
      "step": 4113
    },
    {
      "epoch": 1.18,
      "grad_norm": 10.221151351928711,
      "learning_rate": 1.2348398169336384e-05,
      "loss": 0.732,
      "step": 4114
    },
    {
      "epoch": 1.18,
      "grad_norm": 10.070449829101562,
      "learning_rate": 1.2344107551487414e-05,
      "loss": 0.8596,
      "step": 4115
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.374528884887695,
      "learning_rate": 1.2339816933638443e-05,
      "loss": 0.8476,
      "step": 4116
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.961380958557129,
      "learning_rate": 1.2335526315789473e-05,
      "loss": 0.8881,
      "step": 4117
    },
    {
      "epoch": 1.18,
      "grad_norm": 8.791508674621582,
      "learning_rate": 1.2331235697940504e-05,
      "loss": 0.7018,
      "step": 4118
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.854459762573242,
      "learning_rate": 1.2326945080091534e-05,
      "loss": 0.7528,
      "step": 4119
    },
    {
      "epoch": 1.18,
      "grad_norm": 11.351456642150879,
      "learning_rate": 1.2322654462242563e-05,
      "loss": 0.5635,
      "step": 4120
    },
    {
      "epoch": 1.18,
      "grad_norm": 7.799431324005127,
      "learning_rate": 1.2318363844393594e-05,
      "loss": 0.5938,
      "step": 4121
    },
    {
      "epoch": 1.18,
      "grad_norm": 12.781096458435059,
      "learning_rate": 1.2314073226544622e-05,
      "loss": 0.9506,
      "step": 4122
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.58182430267334,
      "learning_rate": 1.2309782608695653e-05,
      "loss": 0.756,
      "step": 4123
    },
    {
      "epoch": 1.18,
      "grad_norm": 10.1422119140625,
      "learning_rate": 1.2305491990846682e-05,
      "loss": 1.0917,
      "step": 4124
    },
    {
      "epoch": 1.18,
      "grad_norm": 8.120306015014648,
      "learning_rate": 1.2301201372997712e-05,
      "loss": 0.8838,
      "step": 4125
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.112457275390625,
      "learning_rate": 1.2296910755148742e-05,
      "loss": 0.805,
      "step": 4126
    },
    {
      "epoch": 1.18,
      "grad_norm": 10.73107624053955,
      "learning_rate": 1.2292620137299771e-05,
      "loss": 1.0341,
      "step": 4127
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.700551986694336,
      "learning_rate": 1.2288329519450802e-05,
      "loss": 0.9188,
      "step": 4128
    },
    {
      "epoch": 1.18,
      "grad_norm": 15.175979614257812,
      "learning_rate": 1.228403890160183e-05,
      "loss": 1.0461,
      "step": 4129
    },
    {
      "epoch": 1.18,
      "grad_norm": 8.576842308044434,
      "learning_rate": 1.2279748283752861e-05,
      "loss": 0.8233,
      "step": 4130
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.577347755432129,
      "learning_rate": 1.227545766590389e-05,
      "loss": 0.7652,
      "step": 4131
    },
    {
      "epoch": 1.18,
      "grad_norm": 12.247037887573242,
      "learning_rate": 1.227116704805492e-05,
      "loss": 0.9427,
      "step": 4132
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.787284851074219,
      "learning_rate": 1.2266876430205949e-05,
      "loss": 0.8011,
      "step": 4133
    },
    {
      "epoch": 1.18,
      "grad_norm": 10.73867416381836,
      "learning_rate": 1.226258581235698e-05,
      "loss": 0.5968,
      "step": 4134
    },
    {
      "epoch": 1.18,
      "grad_norm": 8.938671112060547,
      "learning_rate": 1.225829519450801e-05,
      "loss": 0.8767,
      "step": 4135
    },
    {
      "epoch": 1.18,
      "grad_norm": 10.706457138061523,
      "learning_rate": 1.225400457665904e-05,
      "loss": 0.8142,
      "step": 4136
    },
    {
      "epoch": 1.18,
      "grad_norm": 11.404316902160645,
      "learning_rate": 1.2249713958810069e-05,
      "loss": 0.8691,
      "step": 4137
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.415704727172852,
      "learning_rate": 1.22454233409611e-05,
      "loss": 0.7746,
      "step": 4138
    },
    {
      "epoch": 1.18,
      "grad_norm": 10.434395790100098,
      "learning_rate": 1.2241132723112128e-05,
      "loss": 0.9429,
      "step": 4139
    },
    {
      "epoch": 1.18,
      "grad_norm": 8.272794723510742,
      "learning_rate": 1.2236842105263159e-05,
      "loss": 0.646,
      "step": 4140
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.81921100616455,
      "learning_rate": 1.2232551487414188e-05,
      "loss": 0.8917,
      "step": 4141
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.18488597869873,
      "learning_rate": 1.2228260869565216e-05,
      "loss": 0.759,
      "step": 4142
    },
    {
      "epoch": 1.19,
      "grad_norm": 9.562060356140137,
      "learning_rate": 1.2223970251716248e-05,
      "loss": 0.8422,
      "step": 4143
    },
    {
      "epoch": 1.19,
      "grad_norm": 7.701141834259033,
      "learning_rate": 1.2219679633867277e-05,
      "loss": 0.5721,
      "step": 4144
    },
    {
      "epoch": 1.19,
      "grad_norm": 11.071125030517578,
      "learning_rate": 1.2215389016018308e-05,
      "loss": 0.6635,
      "step": 4145
    },
    {
      "epoch": 1.19,
      "grad_norm": 12.54422664642334,
      "learning_rate": 1.2211098398169336e-05,
      "loss": 0.8234,
      "step": 4146
    },
    {
      "epoch": 1.19,
      "grad_norm": 11.236760139465332,
      "learning_rate": 1.2206807780320367e-05,
      "loss": 0.8189,
      "step": 4147
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.954216003417969,
      "learning_rate": 1.2202517162471396e-05,
      "loss": 0.8914,
      "step": 4148
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.158648490905762,
      "learning_rate": 1.2198226544622426e-05,
      "loss": 0.7184,
      "step": 4149
    },
    {
      "epoch": 1.19,
      "grad_norm": 11.025486946105957,
      "learning_rate": 1.2193935926773455e-05,
      "loss": 0.7505,
      "step": 4150
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.360795974731445,
      "learning_rate": 1.2189645308924485e-05,
      "loss": 0.7453,
      "step": 4151
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.180240631103516,
      "learning_rate": 1.2185354691075516e-05,
      "loss": 0.8649,
      "step": 4152
    },
    {
      "epoch": 1.19,
      "grad_norm": 11.01577091217041,
      "learning_rate": 1.2181064073226545e-05,
      "loss": 0.9037,
      "step": 4153
    },
    {
      "epoch": 1.19,
      "grad_norm": 9.540740013122559,
      "learning_rate": 1.2176773455377575e-05,
      "loss": 0.7046,
      "step": 4154
    },
    {
      "epoch": 1.19,
      "grad_norm": 9.411883354187012,
      "learning_rate": 1.2172482837528604e-05,
      "loss": 0.8421,
      "step": 4155
    },
    {
      "epoch": 1.19,
      "grad_norm": 12.0674467086792,
      "learning_rate": 1.2168192219679634e-05,
      "loss": 0.6131,
      "step": 4156
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.145977020263672,
      "learning_rate": 1.2163901601830663e-05,
      "loss": 0.932,
      "step": 4157
    },
    {
      "epoch": 1.19,
      "grad_norm": 7.649431228637695,
      "learning_rate": 1.2159610983981694e-05,
      "loss": 0.5755,
      "step": 4158
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.664711952209473,
      "learning_rate": 1.2155320366132722e-05,
      "loss": 0.8075,
      "step": 4159
    },
    {
      "epoch": 1.19,
      "grad_norm": 11.5469970703125,
      "learning_rate": 1.2151029748283753e-05,
      "loss": 0.8425,
      "step": 4160
    },
    {
      "epoch": 1.19,
      "grad_norm": 9.75073528289795,
      "learning_rate": 1.2146739130434783e-05,
      "loss": 0.6543,
      "step": 4161
    },
    {
      "epoch": 1.19,
      "grad_norm": 9.785978317260742,
      "learning_rate": 1.2142448512585814e-05,
      "loss": 0.6521,
      "step": 4162
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.929224014282227,
      "learning_rate": 1.2138157894736842e-05,
      "loss": 0.9284,
      "step": 4163
    },
    {
      "epoch": 1.19,
      "grad_norm": 7.091762542724609,
      "learning_rate": 1.2133867276887873e-05,
      "loss": 0.4471,
      "step": 4164
    },
    {
      "epoch": 1.19,
      "grad_norm": 11.026457786560059,
      "learning_rate": 1.2129576659038902e-05,
      "loss": 0.8445,
      "step": 4165
    },
    {
      "epoch": 1.19,
      "grad_norm": 13.157074928283691,
      "learning_rate": 1.2125286041189932e-05,
      "loss": 0.8575,
      "step": 4166
    },
    {
      "epoch": 1.19,
      "grad_norm": 9.50782299041748,
      "learning_rate": 1.2120995423340961e-05,
      "loss": 0.614,
      "step": 4167
    },
    {
      "epoch": 1.19,
      "grad_norm": 12.477818489074707,
      "learning_rate": 1.211670480549199e-05,
      "loss": 0.8306,
      "step": 4168
    },
    {
      "epoch": 1.19,
      "grad_norm": 12.342202186584473,
      "learning_rate": 1.2112414187643022e-05,
      "loss": 0.6937,
      "step": 4169
    },
    {
      "epoch": 1.19,
      "grad_norm": 8.733111381530762,
      "learning_rate": 1.210812356979405e-05,
      "loss": 0.6343,
      "step": 4170
    },
    {
      "epoch": 1.19,
      "grad_norm": 11.11361312866211,
      "learning_rate": 1.2103832951945081e-05,
      "loss": 0.9674,
      "step": 4171
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.384431838989258,
      "learning_rate": 1.209954233409611e-05,
      "loss": 0.7703,
      "step": 4172
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.025187492370605,
      "learning_rate": 1.209525171624714e-05,
      "loss": 0.7246,
      "step": 4173
    },
    {
      "epoch": 1.19,
      "grad_norm": 9.30639934539795,
      "learning_rate": 1.2090961098398169e-05,
      "loss": 0.713,
      "step": 4174
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.196172714233398,
      "learning_rate": 1.20866704805492e-05,
      "loss": 0.8763,
      "step": 4175
    },
    {
      "epoch": 1.19,
      "grad_norm": 12.444104194641113,
      "learning_rate": 1.2082379862700228e-05,
      "loss": 0.971,
      "step": 4176
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.587760925292969,
      "learning_rate": 1.2078089244851259e-05,
      "loss": 0.7235,
      "step": 4177
    },
    {
      "epoch": 1.2,
      "grad_norm": 7.7796478271484375,
      "learning_rate": 1.207379862700229e-05,
      "loss": 0.6905,
      "step": 4178
    },
    {
      "epoch": 1.2,
      "grad_norm": 10.688580513000488,
      "learning_rate": 1.206950800915332e-05,
      "loss": 1.1034,
      "step": 4179
    },
    {
      "epoch": 1.2,
      "grad_norm": 9.650777816772461,
      "learning_rate": 1.2065217391304348e-05,
      "loss": 0.7908,
      "step": 4180
    },
    {
      "epoch": 1.2,
      "grad_norm": 9.831267356872559,
      "learning_rate": 1.2060926773455377e-05,
      "loss": 0.7006,
      "step": 4181
    },
    {
      "epoch": 1.2,
      "grad_norm": 11.56423282623291,
      "learning_rate": 1.2056636155606408e-05,
      "loss": 0.7525,
      "step": 4182
    },
    {
      "epoch": 1.2,
      "grad_norm": 9.822474479675293,
      "learning_rate": 1.2052345537757436e-05,
      "loss": 1.1696,
      "step": 4183
    },
    {
      "epoch": 1.2,
      "grad_norm": 8.155306816101074,
      "learning_rate": 1.2048054919908467e-05,
      "loss": 0.6272,
      "step": 4184
    },
    {
      "epoch": 1.2,
      "grad_norm": 9.39324951171875,
      "learning_rate": 1.2043764302059496e-05,
      "loss": 0.7688,
      "step": 4185
    },
    {
      "epoch": 1.2,
      "grad_norm": 11.637955665588379,
      "learning_rate": 1.2039473684210528e-05,
      "loss": 1.0111,
      "step": 4186
    },
    {
      "epoch": 1.2,
      "grad_norm": 13.19721794128418,
      "learning_rate": 1.2035183066361557e-05,
      "loss": 0.9972,
      "step": 4187
    },
    {
      "epoch": 1.2,
      "grad_norm": 9.880034446716309,
      "learning_rate": 1.2030892448512587e-05,
      "loss": 0.7787,
      "step": 4188
    },
    {
      "epoch": 1.2,
      "grad_norm": 9.659402847290039,
      "learning_rate": 1.2026601830663616e-05,
      "loss": 0.71,
      "step": 4189
    },
    {
      "epoch": 1.2,
      "grad_norm": 7.872838973999023,
      "learning_rate": 1.2022311212814646e-05,
      "loss": 0.7305,
      "step": 4190
    },
    {
      "epoch": 1.2,
      "grad_norm": 8.332547187805176,
      "learning_rate": 1.2018020594965675e-05,
      "loss": 1.1115,
      "step": 4191
    },
    {
      "epoch": 1.2,
      "grad_norm": 8.256744384765625,
      "learning_rate": 1.2013729977116706e-05,
      "loss": 0.902,
      "step": 4192
    },
    {
      "epoch": 1.2,
      "grad_norm": 8.371477127075195,
      "learning_rate": 1.2009439359267734e-05,
      "loss": 0.875,
      "step": 4193
    },
    {
      "epoch": 1.2,
      "grad_norm": 10.548712730407715,
      "learning_rate": 1.2005148741418763e-05,
      "loss": 0.8353,
      "step": 4194
    },
    {
      "epoch": 1.2,
      "grad_norm": 11.617545127868652,
      "learning_rate": 1.2000858123569795e-05,
      "loss": 1.303,
      "step": 4195
    },
    {
      "epoch": 1.2,
      "grad_norm": 9.448934555053711,
      "learning_rate": 1.1996567505720824e-05,
      "loss": 0.8609,
      "step": 4196
    },
    {
      "epoch": 1.2,
      "grad_norm": 9.004947662353516,
      "learning_rate": 1.1992276887871854e-05,
      "loss": 0.7202,
      "step": 4197
    },
    {
      "epoch": 1.2,
      "grad_norm": 8.329684257507324,
      "learning_rate": 1.1987986270022883e-05,
      "loss": 0.863,
      "step": 4198
    },
    {
      "epoch": 1.2,
      "grad_norm": 10.899600982666016,
      "learning_rate": 1.1983695652173914e-05,
      "loss": 0.9036,
      "step": 4199
    },
    {
      "epoch": 1.2,
      "grad_norm": 7.325217247009277,
      "learning_rate": 1.1979405034324942e-05,
      "loss": 0.601,
      "step": 4200
    },
    {
      "epoch": 1.2,
      "grad_norm": 9.30616283416748,
      "learning_rate": 1.1975114416475973e-05,
      "loss": 0.9284,
      "step": 4201
    },
    {
      "epoch": 1.2,
      "grad_norm": 10.972339630126953,
      "learning_rate": 1.1970823798627002e-05,
      "loss": 0.9204,
      "step": 4202
    },
    {
      "epoch": 1.2,
      "grad_norm": 11.110115051269531,
      "learning_rate": 1.1966533180778034e-05,
      "loss": 0.6842,
      "step": 4203
    },
    {
      "epoch": 1.2,
      "grad_norm": 12.855428695678711,
      "learning_rate": 1.1962242562929063e-05,
      "loss": 0.8472,
      "step": 4204
    },
    {
      "epoch": 1.2,
      "grad_norm": 8.137256622314453,
      "learning_rate": 1.1957951945080093e-05,
      "loss": 0.5689,
      "step": 4205
    },
    {
      "epoch": 1.2,
      "grad_norm": 9.706517219543457,
      "learning_rate": 1.1953661327231122e-05,
      "loss": 0.7754,
      "step": 4206
    },
    {
      "epoch": 1.2,
      "grad_norm": 9.710511207580566,
      "learning_rate": 1.194937070938215e-05,
      "loss": 0.9112,
      "step": 4207
    },
    {
      "epoch": 1.2,
      "grad_norm": 11.372818946838379,
      "learning_rate": 1.1945080091533181e-05,
      "loss": 0.9699,
      "step": 4208
    },
    {
      "epoch": 1.2,
      "grad_norm": 10.132120132446289,
      "learning_rate": 1.194078947368421e-05,
      "loss": 0.641,
      "step": 4209
    },
    {
      "epoch": 1.2,
      "grad_norm": 10.642168998718262,
      "learning_rate": 1.193649885583524e-05,
      "loss": 0.776,
      "step": 4210
    },
    {
      "epoch": 1.2,
      "grad_norm": 12.64185905456543,
      "learning_rate": 1.1932208237986269e-05,
      "loss": 0.7297,
      "step": 4211
    },
    {
      "epoch": 1.2,
      "grad_norm": 13.01512622833252,
      "learning_rate": 1.1927917620137301e-05,
      "loss": 1.3426,
      "step": 4212
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.130009651184082,
      "learning_rate": 1.192362700228833e-05,
      "loss": 0.6614,
      "step": 4213
    },
    {
      "epoch": 1.21,
      "grad_norm": 8.871742248535156,
      "learning_rate": 1.191933638443936e-05,
      "loss": 0.5409,
      "step": 4214
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.790398597717285,
      "learning_rate": 1.191504576659039e-05,
      "loss": 0.7948,
      "step": 4215
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.162729263305664,
      "learning_rate": 1.191075514874142e-05,
      "loss": 0.7729,
      "step": 4216
    },
    {
      "epoch": 1.21,
      "grad_norm": 7.943366050720215,
      "learning_rate": 1.1906464530892448e-05,
      "loss": 0.7385,
      "step": 4217
    },
    {
      "epoch": 1.21,
      "grad_norm": 9.41102409362793,
      "learning_rate": 1.1902173913043479e-05,
      "loss": 0.6734,
      "step": 4218
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.491739273071289,
      "learning_rate": 1.1897883295194508e-05,
      "loss": 0.6161,
      "step": 4219
    },
    {
      "epoch": 1.21,
      "grad_norm": 11.373291015625,
      "learning_rate": 1.1893592677345536e-05,
      "loss": 0.8434,
      "step": 4220
    },
    {
      "epoch": 1.21,
      "grad_norm": 8.466850280761719,
      "learning_rate": 1.1889302059496569e-05,
      "loss": 0.7244,
      "step": 4221
    },
    {
      "epoch": 1.21,
      "grad_norm": 13.56641960144043,
      "learning_rate": 1.1885011441647597e-05,
      "loss": 0.8512,
      "step": 4222
    },
    {
      "epoch": 1.21,
      "grad_norm": 12.589375495910645,
      "learning_rate": 1.1880720823798628e-05,
      "loss": 0.9807,
      "step": 4223
    },
    {
      "epoch": 1.21,
      "grad_norm": 9.531851768493652,
      "learning_rate": 1.1876430205949657e-05,
      "loss": 0.6452,
      "step": 4224
    },
    {
      "epoch": 1.21,
      "grad_norm": 11.992056846618652,
      "learning_rate": 1.1872139588100687e-05,
      "loss": 0.8185,
      "step": 4225
    },
    {
      "epoch": 1.21,
      "grad_norm": 9.233867645263672,
      "learning_rate": 1.1867848970251716e-05,
      "loss": 0.802,
      "step": 4226
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.06538200378418,
      "learning_rate": 1.1863558352402746e-05,
      "loss": 0.6751,
      "step": 4227
    },
    {
      "epoch": 1.21,
      "grad_norm": 11.67239761352539,
      "learning_rate": 1.1859267734553775e-05,
      "loss": 0.7702,
      "step": 4228
    },
    {
      "epoch": 1.21,
      "grad_norm": 11.357415199279785,
      "learning_rate": 1.1854977116704807e-05,
      "loss": 0.8018,
      "step": 4229
    },
    {
      "epoch": 1.21,
      "grad_norm": 9.972132682800293,
      "learning_rate": 1.1850686498855836e-05,
      "loss": 0.6667,
      "step": 4230
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.696549415588379,
      "learning_rate": 1.1846395881006866e-05,
      "loss": 0.8272,
      "step": 4231
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.538824081420898,
      "learning_rate": 1.1842105263157895e-05,
      "loss": 0.7871,
      "step": 4232
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.483213424682617,
      "learning_rate": 1.1837814645308924e-05,
      "loss": 0.6551,
      "step": 4233
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.283119201660156,
      "learning_rate": 1.1833524027459954e-05,
      "loss": 0.7317,
      "step": 4234
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.398222923278809,
      "learning_rate": 1.1829233409610983e-05,
      "loss": 0.7929,
      "step": 4235
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.281462669372559,
      "learning_rate": 1.1824942791762014e-05,
      "loss": 0.7468,
      "step": 4236
    },
    {
      "epoch": 1.21,
      "grad_norm": 8.925246238708496,
      "learning_rate": 1.1820652173913042e-05,
      "loss": 0.5831,
      "step": 4237
    },
    {
      "epoch": 1.21,
      "grad_norm": 9.566448211669922,
      "learning_rate": 1.1816361556064075e-05,
      "loss": 0.7537,
      "step": 4238
    },
    {
      "epoch": 1.21,
      "grad_norm": 13.305020332336426,
      "learning_rate": 1.1812070938215103e-05,
      "loss": 0.7581,
      "step": 4239
    },
    {
      "epoch": 1.21,
      "grad_norm": 11.126168251037598,
      "learning_rate": 1.1807780320366134e-05,
      "loss": 0.5523,
      "step": 4240
    },
    {
      "epoch": 1.21,
      "grad_norm": 8.93124771118164,
      "learning_rate": 1.1803489702517163e-05,
      "loss": 0.5531,
      "step": 4241
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.155959129333496,
      "learning_rate": 1.1799199084668193e-05,
      "loss": 0.9744,
      "step": 4242
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.217401504516602,
      "learning_rate": 1.1794908466819222e-05,
      "loss": 0.9487,
      "step": 4243
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.667862892150879,
      "learning_rate": 1.1790617848970252e-05,
      "loss": 0.7737,
      "step": 4244
    },
    {
      "epoch": 1.21,
      "grad_norm": 11.557846069335938,
      "learning_rate": 1.1786327231121281e-05,
      "loss": 1.1461,
      "step": 4245
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.171220779418945,
      "learning_rate": 1.1782036613272312e-05,
      "loss": 0.8439,
      "step": 4246
    },
    {
      "epoch": 1.21,
      "grad_norm": 12.268500328063965,
      "learning_rate": 1.1777745995423342e-05,
      "loss": 0.8422,
      "step": 4247
    },
    {
      "epoch": 1.22,
      "grad_norm": 8.701354026794434,
      "learning_rate": 1.177345537757437e-05,
      "loss": 0.6734,
      "step": 4248
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.68353271484375,
      "learning_rate": 1.1769164759725401e-05,
      "loss": 0.6657,
      "step": 4249
    },
    {
      "epoch": 1.22,
      "grad_norm": 12.258493423461914,
      "learning_rate": 1.176487414187643e-05,
      "loss": 1.1132,
      "step": 4250
    },
    {
      "epoch": 1.22,
      "grad_norm": 12.400599479675293,
      "learning_rate": 1.176058352402746e-05,
      "loss": 1.0322,
      "step": 4251
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.623896598815918,
      "learning_rate": 1.175629290617849e-05,
      "loss": 0.8161,
      "step": 4252
    },
    {
      "epoch": 1.22,
      "grad_norm": 9.813971519470215,
      "learning_rate": 1.175200228832952e-05,
      "loss": 0.6358,
      "step": 4253
    },
    {
      "epoch": 1.22,
      "grad_norm": 8.067031860351562,
      "learning_rate": 1.1747711670480548e-05,
      "loss": 0.8396,
      "step": 4254
    },
    {
      "epoch": 1.22,
      "grad_norm": 9.504332542419434,
      "learning_rate": 1.174342105263158e-05,
      "loss": 0.8028,
      "step": 4255
    },
    {
      "epoch": 1.22,
      "grad_norm": 9.022904396057129,
      "learning_rate": 1.173913043478261e-05,
      "loss": 0.814,
      "step": 4256
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.11242961883545,
      "learning_rate": 1.173483981693364e-05,
      "loss": 0.6788,
      "step": 4257
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.316167831420898,
      "learning_rate": 1.1730549199084669e-05,
      "loss": 0.7584,
      "step": 4258
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.539634704589844,
      "learning_rate": 1.1726258581235697e-05,
      "loss": 0.6962,
      "step": 4259
    },
    {
      "epoch": 1.22,
      "grad_norm": 6.796972751617432,
      "learning_rate": 1.1721967963386728e-05,
      "loss": 0.5701,
      "step": 4260
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.749701499938965,
      "learning_rate": 1.1717677345537757e-05,
      "loss": 0.946,
      "step": 4261
    },
    {
      "epoch": 1.22,
      "grad_norm": 8.778101921081543,
      "learning_rate": 1.1713386727688787e-05,
      "loss": 0.8784,
      "step": 4262
    },
    {
      "epoch": 1.22,
      "grad_norm": 9.367880821228027,
      "learning_rate": 1.1709096109839818e-05,
      "loss": 0.8687,
      "step": 4263
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.653621673583984,
      "learning_rate": 1.1704805491990848e-05,
      "loss": 1.0333,
      "step": 4264
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.247994422912598,
      "learning_rate": 1.1700514874141877e-05,
      "loss": 0.8605,
      "step": 4265
    },
    {
      "epoch": 1.22,
      "grad_norm": 9.010552406311035,
      "learning_rate": 1.1696224256292907e-05,
      "loss": 0.8025,
      "step": 4266
    },
    {
      "epoch": 1.22,
      "grad_norm": 8.36340618133545,
      "learning_rate": 1.1691933638443936e-05,
      "loss": 0.6301,
      "step": 4267
    },
    {
      "epoch": 1.22,
      "grad_norm": 9.572992324829102,
      "learning_rate": 1.1687643020594966e-05,
      "loss": 0.6753,
      "step": 4268
    },
    {
      "epoch": 1.22,
      "grad_norm": 9.083409309387207,
      "learning_rate": 1.1683352402745995e-05,
      "loss": 0.7244,
      "step": 4269
    },
    {
      "epoch": 1.22,
      "grad_norm": 11.072820663452148,
      "learning_rate": 1.1679061784897026e-05,
      "loss": 0.6236,
      "step": 4270
    },
    {
      "epoch": 1.22,
      "grad_norm": 12.434370994567871,
      "learning_rate": 1.1674771167048054e-05,
      "loss": 0.9262,
      "step": 4271
    },
    {
      "epoch": 1.22,
      "grad_norm": 13.024633407592773,
      "learning_rate": 1.1670480549199085e-05,
      "loss": 0.748,
      "step": 4272
    },
    {
      "epoch": 1.22,
      "grad_norm": 7.6458659172058105,
      "learning_rate": 1.1666189931350115e-05,
      "loss": 0.7682,
      "step": 4273
    },
    {
      "epoch": 1.22,
      "grad_norm": 9.794417381286621,
      "learning_rate": 1.1661899313501144e-05,
      "loss": 0.7265,
      "step": 4274
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.18636703491211,
      "learning_rate": 1.1657608695652175e-05,
      "loss": 0.7181,
      "step": 4275
    },
    {
      "epoch": 1.22,
      "grad_norm": 11.913507461547852,
      "learning_rate": 1.1653318077803203e-05,
      "loss": 0.7299,
      "step": 4276
    },
    {
      "epoch": 1.22,
      "grad_norm": 9.873882293701172,
      "learning_rate": 1.1649027459954234e-05,
      "loss": 0.8151,
      "step": 4277
    },
    {
      "epoch": 1.22,
      "grad_norm": 11.573633193969727,
      "learning_rate": 1.1644736842105263e-05,
      "loss": 1.019,
      "step": 4278
    },
    {
      "epoch": 1.22,
      "grad_norm": 9.217798233032227,
      "learning_rate": 1.1640446224256293e-05,
      "loss": 0.6498,
      "step": 4279
    },
    {
      "epoch": 1.22,
      "grad_norm": 9.954983711242676,
      "learning_rate": 1.1636155606407322e-05,
      "loss": 0.7433,
      "step": 4280
    },
    {
      "epoch": 1.22,
      "grad_norm": 11.062023162841797,
      "learning_rate": 1.1631864988558354e-05,
      "loss": 0.9304,
      "step": 4281
    },
    {
      "epoch": 1.22,
      "grad_norm": 9.869750022888184,
      "learning_rate": 1.1627574370709383e-05,
      "loss": 0.7542,
      "step": 4282
    },
    {
      "epoch": 1.23,
      "grad_norm": 11.585716247558594,
      "learning_rate": 1.1623283752860413e-05,
      "loss": 0.8442,
      "step": 4283
    },
    {
      "epoch": 1.23,
      "grad_norm": 8.067532539367676,
      "learning_rate": 1.1618993135011442e-05,
      "loss": 0.6195,
      "step": 4284
    },
    {
      "epoch": 1.23,
      "grad_norm": 12.927903175354004,
      "learning_rate": 1.161470251716247e-05,
      "loss": 0.8351,
      "step": 4285
    },
    {
      "epoch": 1.23,
      "grad_norm": 9.903480529785156,
      "learning_rate": 1.1610411899313501e-05,
      "loss": 0.7832,
      "step": 4286
    },
    {
      "epoch": 1.23,
      "grad_norm": 9.87097454071045,
      "learning_rate": 1.160612128146453e-05,
      "loss": 1.1323,
      "step": 4287
    },
    {
      "epoch": 1.23,
      "grad_norm": 9.815967559814453,
      "learning_rate": 1.160183066361556e-05,
      "loss": 0.784,
      "step": 4288
    },
    {
      "epoch": 1.23,
      "grad_norm": 9.69139575958252,
      "learning_rate": 1.1597540045766591e-05,
      "loss": 0.7694,
      "step": 4289
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.718193054199219,
      "learning_rate": 1.1593249427917621e-05,
      "loss": 0.9341,
      "step": 4290
    },
    {
      "epoch": 1.23,
      "grad_norm": 8.258038520812988,
      "learning_rate": 1.158895881006865e-05,
      "loss": 0.6007,
      "step": 4291
    },
    {
      "epoch": 1.23,
      "grad_norm": 7.3089823722839355,
      "learning_rate": 1.158466819221968e-05,
      "loss": 0.7345,
      "step": 4292
    },
    {
      "epoch": 1.23,
      "grad_norm": 7.901584625244141,
      "learning_rate": 1.158037757437071e-05,
      "loss": 0.7332,
      "step": 4293
    },
    {
      "epoch": 1.23,
      "grad_norm": 7.5747785568237305,
      "learning_rate": 1.157608695652174e-05,
      "loss": 0.6251,
      "step": 4294
    },
    {
      "epoch": 1.23,
      "grad_norm": 8.1260347366333,
      "learning_rate": 1.1571796338672769e-05,
      "loss": 0.7314,
      "step": 4295
    },
    {
      "epoch": 1.23,
      "grad_norm": 8.920598030090332,
      "learning_rate": 1.1567505720823799e-05,
      "loss": 0.6244,
      "step": 4296
    },
    {
      "epoch": 1.23,
      "grad_norm": 9.813675880432129,
      "learning_rate": 1.1563215102974828e-05,
      "loss": 0.8088,
      "step": 4297
    },
    {
      "epoch": 1.23,
      "grad_norm": 12.650906562805176,
      "learning_rate": 1.1558924485125858e-05,
      "loss": 0.9075,
      "step": 4298
    },
    {
      "epoch": 1.23,
      "grad_norm": 12.57913589477539,
      "learning_rate": 1.1554633867276889e-05,
      "loss": 0.9788,
      "step": 4299
    },
    {
      "epoch": 1.23,
      "grad_norm": 11.513996124267578,
      "learning_rate": 1.1550343249427917e-05,
      "loss": 0.9212,
      "step": 4300
    },
    {
      "epoch": 1.23,
      "grad_norm": 8.296212196350098,
      "learning_rate": 1.1546052631578948e-05,
      "loss": 0.7903,
      "step": 4301
    },
    {
      "epoch": 1.23,
      "grad_norm": 11.32043743133545,
      "learning_rate": 1.1541762013729977e-05,
      "loss": 0.8166,
      "step": 4302
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.928311347961426,
      "learning_rate": 1.1537471395881007e-05,
      "loss": 1.1405,
      "step": 4303
    },
    {
      "epoch": 1.23,
      "grad_norm": 9.672231674194336,
      "learning_rate": 1.1533180778032036e-05,
      "loss": 0.8346,
      "step": 4304
    },
    {
      "epoch": 1.23,
      "grad_norm": 11.749388694763184,
      "learning_rate": 1.1528890160183066e-05,
      "loss": 0.8113,
      "step": 4305
    },
    {
      "epoch": 1.23,
      "grad_norm": 9.33349609375,
      "learning_rate": 1.1524599542334097e-05,
      "loss": 0.7047,
      "step": 4306
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.378589630126953,
      "learning_rate": 1.1520308924485127e-05,
      "loss": 0.9374,
      "step": 4307
    },
    {
      "epoch": 1.23,
      "grad_norm": 12.265445709228516,
      "learning_rate": 1.1516018306636156e-05,
      "loss": 0.7387,
      "step": 4308
    },
    {
      "epoch": 1.23,
      "grad_norm": 9.64606761932373,
      "learning_rate": 1.1511727688787187e-05,
      "loss": 0.9858,
      "step": 4309
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.293717384338379,
      "learning_rate": 1.1507437070938215e-05,
      "loss": 1.4012,
      "step": 4310
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.404987335205078,
      "learning_rate": 1.1503146453089244e-05,
      "loss": 1.0246,
      "step": 4311
    },
    {
      "epoch": 1.23,
      "grad_norm": 6.542437553405762,
      "learning_rate": 1.1498855835240275e-05,
      "loss": 0.6179,
      "step": 4312
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.147354125976562,
      "learning_rate": 1.1494565217391303e-05,
      "loss": 0.8771,
      "step": 4313
    },
    {
      "epoch": 1.23,
      "grad_norm": 9.722268104553223,
      "learning_rate": 1.1490274599542334e-05,
      "loss": 0.8938,
      "step": 4314
    },
    {
      "epoch": 1.23,
      "grad_norm": 8.905896186828613,
      "learning_rate": 1.1485983981693364e-05,
      "loss": 0.6041,
      "step": 4315
    },
    {
      "epoch": 1.23,
      "grad_norm": 11.898765563964844,
      "learning_rate": 1.1481693363844395e-05,
      "loss": 0.9426,
      "step": 4316
    },
    {
      "epoch": 1.23,
      "grad_norm": 13.410457611083984,
      "learning_rate": 1.1477402745995423e-05,
      "loss": 1.0473,
      "step": 4317
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.72824478149414,
      "learning_rate": 1.1473112128146454e-05,
      "loss": 0.8856,
      "step": 4318
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.007899284362793,
      "learning_rate": 1.1468821510297483e-05,
      "loss": 0.7087,
      "step": 4319
    },
    {
      "epoch": 1.24,
      "grad_norm": 12.402856826782227,
      "learning_rate": 1.1464530892448513e-05,
      "loss": 0.7871,
      "step": 4320
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.839303970336914,
      "learning_rate": 1.1460240274599542e-05,
      "loss": 0.9303,
      "step": 4321
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.47752857208252,
      "learning_rate": 1.1455949656750572e-05,
      "loss": 0.9723,
      "step": 4322
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.317313194274902,
      "learning_rate": 1.1451659038901603e-05,
      "loss": 0.7985,
      "step": 4323
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.262114524841309,
      "learning_rate": 1.1447368421052632e-05,
      "loss": 0.8299,
      "step": 4324
    },
    {
      "epoch": 1.24,
      "grad_norm": 7.015544414520264,
      "learning_rate": 1.1443077803203662e-05,
      "loss": 0.5288,
      "step": 4325
    },
    {
      "epoch": 1.24,
      "grad_norm": 8.641091346740723,
      "learning_rate": 1.1438787185354691e-05,
      "loss": 0.7868,
      "step": 4326
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.731534957885742,
      "learning_rate": 1.1434496567505721e-05,
      "loss": 0.8198,
      "step": 4327
    },
    {
      "epoch": 1.24,
      "grad_norm": 12.458267211914062,
      "learning_rate": 1.143020594965675e-05,
      "loss": 0.9479,
      "step": 4328
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.34188175201416,
      "learning_rate": 1.142591533180778e-05,
      "loss": 0.8897,
      "step": 4329
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.190241813659668,
      "learning_rate": 1.142162471395881e-05,
      "loss": 0.5841,
      "step": 4330
    },
    {
      "epoch": 1.24,
      "grad_norm": 14.47514820098877,
      "learning_rate": 1.141733409610984e-05,
      "loss": 1.1723,
      "step": 4331
    },
    {
      "epoch": 1.24,
      "grad_norm": 11.83283519744873,
      "learning_rate": 1.141304347826087e-05,
      "loss": 0.8721,
      "step": 4332
    },
    {
      "epoch": 1.24,
      "grad_norm": 11.338597297668457,
      "learning_rate": 1.14087528604119e-05,
      "loss": 0.6735,
      "step": 4333
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.52934741973877,
      "learning_rate": 1.140446224256293e-05,
      "loss": 0.7374,
      "step": 4334
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.10802936553955,
      "learning_rate": 1.140017162471396e-05,
      "loss": 0.73,
      "step": 4335
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.148085594177246,
      "learning_rate": 1.1395881006864989e-05,
      "loss": 0.5268,
      "step": 4336
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.599064826965332,
      "learning_rate": 1.139159038901602e-05,
      "loss": 0.8209,
      "step": 4337
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.743919372558594,
      "learning_rate": 1.1387299771167048e-05,
      "loss": 0.7522,
      "step": 4338
    },
    {
      "epoch": 1.24,
      "grad_norm": 11.263919830322266,
      "learning_rate": 1.1383009153318077e-05,
      "loss": 0.6177,
      "step": 4339
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.132713317871094,
      "learning_rate": 1.1378718535469107e-05,
      "loss": 0.8611,
      "step": 4340
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.401273727416992,
      "learning_rate": 1.1374427917620138e-05,
      "loss": 0.7318,
      "step": 4341
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.272201538085938,
      "learning_rate": 1.1370137299771168e-05,
      "loss": 0.8656,
      "step": 4342
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.658712387084961,
      "learning_rate": 1.1365846681922197e-05,
      "loss": 0.7594,
      "step": 4343
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.106707572937012,
      "learning_rate": 1.1361556064073227e-05,
      "loss": 0.7132,
      "step": 4344
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.241507530212402,
      "learning_rate": 1.1357265446224256e-05,
      "loss": 0.588,
      "step": 4345
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.850719451904297,
      "learning_rate": 1.1352974828375287e-05,
      "loss": 0.6053,
      "step": 4346
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.792867660522461,
      "learning_rate": 1.1348684210526315e-05,
      "loss": 0.8155,
      "step": 4347
    },
    {
      "epoch": 1.24,
      "grad_norm": 7.734493255615234,
      "learning_rate": 1.1344393592677346e-05,
      "loss": 0.6452,
      "step": 4348
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.958784103393555,
      "learning_rate": 1.1340102974828376e-05,
      "loss": 0.734,
      "step": 4349
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.508255958557129,
      "learning_rate": 1.1335812356979407e-05,
      "loss": 0.8112,
      "step": 4350
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.391280174255371,
      "learning_rate": 1.1331521739130435e-05,
      "loss": 0.6283,
      "step": 4351
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.672481536865234,
      "learning_rate": 1.1327231121281464e-05,
      "loss": 0.8086,
      "step": 4352
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.470000267028809,
      "learning_rate": 1.1322940503432495e-05,
      "loss": 0.7252,
      "step": 4353
    },
    {
      "epoch": 1.25,
      "grad_norm": 11.01935863494873,
      "learning_rate": 1.1318649885583523e-05,
      "loss": 0.9348,
      "step": 4354
    },
    {
      "epoch": 1.25,
      "grad_norm": 9.3633451461792,
      "learning_rate": 1.1314359267734554e-05,
      "loss": 0.722,
      "step": 4355
    },
    {
      "epoch": 1.25,
      "grad_norm": 13.133733749389648,
      "learning_rate": 1.1310068649885583e-05,
      "loss": 0.6562,
      "step": 4356
    },
    {
      "epoch": 1.25,
      "grad_norm": 8.972723007202148,
      "learning_rate": 1.1305778032036613e-05,
      "loss": 0.8306,
      "step": 4357
    },
    {
      "epoch": 1.25,
      "grad_norm": 8.94922161102295,
      "learning_rate": 1.1301487414187644e-05,
      "loss": 0.9429,
      "step": 4358
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.160902976989746,
      "learning_rate": 1.1297196796338674e-05,
      "loss": 0.7064,
      "step": 4359
    },
    {
      "epoch": 1.25,
      "grad_norm": 11.051900863647461,
      "learning_rate": 1.1292906178489703e-05,
      "loss": 0.7791,
      "step": 4360
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.453458786010742,
      "learning_rate": 1.1288615560640733e-05,
      "loss": 0.6881,
      "step": 4361
    },
    {
      "epoch": 1.25,
      "grad_norm": 9.616094589233398,
      "learning_rate": 1.1284324942791762e-05,
      "loss": 0.7485,
      "step": 4362
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.842020034790039,
      "learning_rate": 1.1280034324942793e-05,
      "loss": 0.8852,
      "step": 4363
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.710325241088867,
      "learning_rate": 1.1275743707093821e-05,
      "loss": 1.0494,
      "step": 4364
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.418954849243164,
      "learning_rate": 1.127145308924485e-05,
      "loss": 0.7913,
      "step": 4365
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.215678215026855,
      "learning_rate": 1.1267162471395882e-05,
      "loss": 0.8389,
      "step": 4366
    },
    {
      "epoch": 1.25,
      "grad_norm": 11.755205154418945,
      "learning_rate": 1.1262871853546911e-05,
      "loss": 0.6641,
      "step": 4367
    },
    {
      "epoch": 1.25,
      "grad_norm": 8.971964836120605,
      "learning_rate": 1.1258581235697941e-05,
      "loss": 0.707,
      "step": 4368
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.31049919128418,
      "learning_rate": 1.125429061784897e-05,
      "loss": 0.6869,
      "step": 4369
    },
    {
      "epoch": 1.25,
      "grad_norm": 9.721508979797363,
      "learning_rate": 1.125e-05,
      "loss": 0.9134,
      "step": 4370
    },
    {
      "epoch": 1.25,
      "grad_norm": 14.146642684936523,
      "learning_rate": 1.124570938215103e-05,
      "loss": 0.8786,
      "step": 4371
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.55115795135498,
      "learning_rate": 1.124141876430206e-05,
      "loss": 0.8725,
      "step": 4372
    },
    {
      "epoch": 1.25,
      "grad_norm": 9.048384666442871,
      "learning_rate": 1.1237128146453089e-05,
      "loss": 0.7171,
      "step": 4373
    },
    {
      "epoch": 1.25,
      "grad_norm": 9.084287643432617,
      "learning_rate": 1.123283752860412e-05,
      "loss": 0.6334,
      "step": 4374
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.687078475952148,
      "learning_rate": 1.122854691075515e-05,
      "loss": 0.798,
      "step": 4375
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.394928932189941,
      "learning_rate": 1.122425629290618e-05,
      "loss": 0.823,
      "step": 4376
    },
    {
      "epoch": 1.25,
      "grad_norm": 11.122708320617676,
      "learning_rate": 1.1219965675057209e-05,
      "loss": 0.6799,
      "step": 4377
    },
    {
      "epoch": 1.25,
      "grad_norm": 11.894579887390137,
      "learning_rate": 1.1215675057208238e-05,
      "loss": 0.7922,
      "step": 4378
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.686917304992676,
      "learning_rate": 1.1211384439359268e-05,
      "loss": 0.9087,
      "step": 4379
    },
    {
      "epoch": 1.25,
      "grad_norm": 11.511585235595703,
      "learning_rate": 1.1207093821510297e-05,
      "loss": 0.919,
      "step": 4380
    },
    {
      "epoch": 1.25,
      "grad_norm": 6.411844253540039,
      "learning_rate": 1.1202803203661327e-05,
      "loss": 0.449,
      "step": 4381
    },
    {
      "epoch": 1.25,
      "grad_norm": 9.558112144470215,
      "learning_rate": 1.1198512585812356e-05,
      "loss": 0.717,
      "step": 4382
    },
    {
      "epoch": 1.25,
      "grad_norm": 9.505937576293945,
      "learning_rate": 1.1194221967963388e-05,
      "loss": 0.7318,
      "step": 4383
    },
    {
      "epoch": 1.25,
      "grad_norm": 8.222380638122559,
      "learning_rate": 1.1189931350114417e-05,
      "loss": 0.7322,
      "step": 4384
    },
    {
      "epoch": 1.25,
      "grad_norm": 12.940840721130371,
      "learning_rate": 1.1185640732265447e-05,
      "loss": 0.8371,
      "step": 4385
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.579774856567383,
      "learning_rate": 1.1181350114416476e-05,
      "loss": 0.8169,
      "step": 4386
    },
    {
      "epoch": 1.25,
      "grad_norm": 8.442181587219238,
      "learning_rate": 1.1177059496567507e-05,
      "loss": 0.6415,
      "step": 4387
    },
    {
      "epoch": 1.26,
      "grad_norm": 11.065545082092285,
      "learning_rate": 1.1172768878718535e-05,
      "loss": 0.8782,
      "step": 4388
    },
    {
      "epoch": 1.26,
      "grad_norm": 9.941202163696289,
      "learning_rate": 1.1168478260869566e-05,
      "loss": 0.5738,
      "step": 4389
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.161408424377441,
      "learning_rate": 1.1164187643020595e-05,
      "loss": 0.6929,
      "step": 4390
    },
    {
      "epoch": 1.26,
      "grad_norm": 8.900789260864258,
      "learning_rate": 1.1159897025171623e-05,
      "loss": 0.5595,
      "step": 4391
    },
    {
      "epoch": 1.26,
      "grad_norm": 9.961406707763672,
      "learning_rate": 1.1155606407322656e-05,
      "loss": 0.4831,
      "step": 4392
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.347709655761719,
      "learning_rate": 1.1151315789473684e-05,
      "loss": 0.7459,
      "step": 4393
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.460041046142578,
      "learning_rate": 1.1147025171624715e-05,
      "loss": 0.599,
      "step": 4394
    },
    {
      "epoch": 1.26,
      "grad_norm": 12.151421546936035,
      "learning_rate": 1.1142734553775744e-05,
      "loss": 0.9976,
      "step": 4395
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.167964935302734,
      "learning_rate": 1.1138443935926774e-05,
      "loss": 0.6846,
      "step": 4396
    },
    {
      "epoch": 1.26,
      "grad_norm": 11.677679061889648,
      "learning_rate": 1.1134153318077803e-05,
      "loss": 1.1699,
      "step": 4397
    },
    {
      "epoch": 1.26,
      "grad_norm": 11.125264167785645,
      "learning_rate": 1.1129862700228833e-05,
      "loss": 0.7343,
      "step": 4398
    },
    {
      "epoch": 1.26,
      "grad_norm": 8.717630386352539,
      "learning_rate": 1.1125572082379862e-05,
      "loss": 0.5695,
      "step": 4399
    },
    {
      "epoch": 1.26,
      "grad_norm": 11.196426391601562,
      "learning_rate": 1.1121281464530894e-05,
      "loss": 0.963,
      "step": 4400
    },
    {
      "epoch": 1.26,
      "grad_norm": 9.22913646697998,
      "learning_rate": 1.1116990846681923e-05,
      "loss": 0.6797,
      "step": 4401
    },
    {
      "epoch": 1.26,
      "grad_norm": 12.953048706054688,
      "learning_rate": 1.1112700228832953e-05,
      "loss": 0.9139,
      "step": 4402
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.623172760009766,
      "learning_rate": 1.1108409610983982e-05,
      "loss": 1.0378,
      "step": 4403
    },
    {
      "epoch": 1.26,
      "grad_norm": 7.46474552154541,
      "learning_rate": 1.1104118993135011e-05,
      "loss": 0.5835,
      "step": 4404
    },
    {
      "epoch": 1.26,
      "grad_norm": 9.413641929626465,
      "learning_rate": 1.1099828375286041e-05,
      "loss": 0.634,
      "step": 4405
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.446019172668457,
      "learning_rate": 1.109553775743707e-05,
      "loss": 0.7857,
      "step": 4406
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.171029090881348,
      "learning_rate": 1.10912471395881e-05,
      "loss": 0.6332,
      "step": 4407
    },
    {
      "epoch": 1.26,
      "grad_norm": 9.081442832946777,
      "learning_rate": 1.108695652173913e-05,
      "loss": 0.6324,
      "step": 4408
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.221150398254395,
      "learning_rate": 1.1082665903890162e-05,
      "loss": 0.6282,
      "step": 4409
    },
    {
      "epoch": 1.26,
      "grad_norm": 9.756031036376953,
      "learning_rate": 1.107837528604119e-05,
      "loss": 0.8952,
      "step": 4410
    },
    {
      "epoch": 1.26,
      "grad_norm": 9.257505416870117,
      "learning_rate": 1.1074084668192221e-05,
      "loss": 0.6357,
      "step": 4411
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.139008522033691,
      "learning_rate": 1.106979405034325e-05,
      "loss": 0.855,
      "step": 4412
    },
    {
      "epoch": 1.26,
      "grad_norm": 13.11406421661377,
      "learning_rate": 1.106550343249428e-05,
      "loss": 1.1138,
      "step": 4413
    },
    {
      "epoch": 1.26,
      "grad_norm": 9.236861228942871,
      "learning_rate": 1.1061212814645309e-05,
      "loss": 0.5939,
      "step": 4414
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.148466110229492,
      "learning_rate": 1.105692219679634e-05,
      "loss": 0.9314,
      "step": 4415
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.256353378295898,
      "learning_rate": 1.1052631578947368e-05,
      "loss": 0.8695,
      "step": 4416
    },
    {
      "epoch": 1.26,
      "grad_norm": 9.436090469360352,
      "learning_rate": 1.1048340961098397e-05,
      "loss": 0.8043,
      "step": 4417
    },
    {
      "epoch": 1.26,
      "grad_norm": 9.323995590209961,
      "learning_rate": 1.1044050343249429e-05,
      "loss": 0.8311,
      "step": 4418
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.637639045715332,
      "learning_rate": 1.1039759725400458e-05,
      "loss": 0.961,
      "step": 4419
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.672082901000977,
      "learning_rate": 1.1035469107551488e-05,
      "loss": 0.8238,
      "step": 4420
    },
    {
      "epoch": 1.26,
      "grad_norm": 9.179261207580566,
      "learning_rate": 1.1031178489702517e-05,
      "loss": 0.8912,
      "step": 4421
    },
    {
      "epoch": 1.26,
      "grad_norm": 9.817282676696777,
      "learning_rate": 1.1026887871853547e-05,
      "loss": 0.8843,
      "step": 4422
    },
    {
      "epoch": 1.27,
      "grad_norm": 8.61851692199707,
      "learning_rate": 1.1022597254004576e-05,
      "loss": 0.7468,
      "step": 4423
    },
    {
      "epoch": 1.27,
      "grad_norm": 7.733014106750488,
      "learning_rate": 1.1018306636155607e-05,
      "loss": 0.6755,
      "step": 4424
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.36716079711914,
      "learning_rate": 1.1014016018306635e-05,
      "loss": 0.8969,
      "step": 4425
    },
    {
      "epoch": 1.27,
      "grad_norm": 10.766349792480469,
      "learning_rate": 1.1009725400457668e-05,
      "loss": 0.9567,
      "step": 4426
    },
    {
      "epoch": 1.27,
      "grad_norm": 11.445676803588867,
      "learning_rate": 1.1005434782608696e-05,
      "loss": 0.9534,
      "step": 4427
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.643470764160156,
      "learning_rate": 1.1001144164759727e-05,
      "loss": 0.7908,
      "step": 4428
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.62116813659668,
      "learning_rate": 1.0996853546910756e-05,
      "loss": 0.6876,
      "step": 4429
    },
    {
      "epoch": 1.27,
      "grad_norm": 11.832209587097168,
      "learning_rate": 1.0992562929061784e-05,
      "loss": 0.9371,
      "step": 4430
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.72209358215332,
      "learning_rate": 1.0988272311212815e-05,
      "loss": 0.6439,
      "step": 4431
    },
    {
      "epoch": 1.27,
      "grad_norm": 7.92345666885376,
      "learning_rate": 1.0983981693363844e-05,
      "loss": 0.5952,
      "step": 4432
    },
    {
      "epoch": 1.27,
      "grad_norm": 8.540083885192871,
      "learning_rate": 1.0979691075514874e-05,
      "loss": 0.7791,
      "step": 4433
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.119351387023926,
      "learning_rate": 1.0975400457665903e-05,
      "loss": 0.7931,
      "step": 4434
    },
    {
      "epoch": 1.27,
      "grad_norm": 11.509077072143555,
      "learning_rate": 1.0971109839816935e-05,
      "loss": 0.9935,
      "step": 4435
    },
    {
      "epoch": 1.27,
      "grad_norm": 11.747727394104004,
      "learning_rate": 1.0966819221967964e-05,
      "loss": 0.647,
      "step": 4436
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.080245018005371,
      "learning_rate": 1.0962528604118994e-05,
      "loss": 0.5636,
      "step": 4437
    },
    {
      "epoch": 1.27,
      "grad_norm": 8.393102645874023,
      "learning_rate": 1.0958237986270023e-05,
      "loss": 0.698,
      "step": 4438
    },
    {
      "epoch": 1.27,
      "grad_norm": 10.01490592956543,
      "learning_rate": 1.0953947368421053e-05,
      "loss": 0.4403,
      "step": 4439
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.193498611450195,
      "learning_rate": 1.0949656750572082e-05,
      "loss": 0.8632,
      "step": 4440
    },
    {
      "epoch": 1.27,
      "grad_norm": 10.101186752319336,
      "learning_rate": 1.0945366132723113e-05,
      "loss": 0.7359,
      "step": 4441
    },
    {
      "epoch": 1.27,
      "grad_norm": 11.3381929397583,
      "learning_rate": 1.0941075514874141e-05,
      "loss": 0.9855,
      "step": 4442
    },
    {
      "epoch": 1.27,
      "grad_norm": 10.720212936401367,
      "learning_rate": 1.0936784897025172e-05,
      "loss": 0.8431,
      "step": 4443
    },
    {
      "epoch": 1.27,
      "grad_norm": 10.518932342529297,
      "learning_rate": 1.0932494279176202e-05,
      "loss": 1.0291,
      "step": 4444
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.96078109741211,
      "learning_rate": 1.0928203661327231e-05,
      "loss": 0.7174,
      "step": 4445
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.446786880493164,
      "learning_rate": 1.0923913043478262e-05,
      "loss": 0.9162,
      "step": 4446
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.563207626342773,
      "learning_rate": 1.091962242562929e-05,
      "loss": 0.7722,
      "step": 4447
    },
    {
      "epoch": 1.27,
      "grad_norm": 10.334098815917969,
      "learning_rate": 1.091533180778032e-05,
      "loss": 0.8665,
      "step": 4448
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.159623146057129,
      "learning_rate": 1.091104118993135e-05,
      "loss": 0.6491,
      "step": 4449
    },
    {
      "epoch": 1.27,
      "grad_norm": 8.697135925292969,
      "learning_rate": 1.090675057208238e-05,
      "loss": 0.7193,
      "step": 4450
    },
    {
      "epoch": 1.27,
      "grad_norm": 10.513727188110352,
      "learning_rate": 1.0902459954233409e-05,
      "loss": 0.7611,
      "step": 4451
    },
    {
      "epoch": 1.27,
      "grad_norm": 8.544404983520508,
      "learning_rate": 1.0898169336384441e-05,
      "loss": 0.9147,
      "step": 4452
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.933435440063477,
      "learning_rate": 1.089387871853547e-05,
      "loss": 0.9955,
      "step": 4453
    },
    {
      "epoch": 1.27,
      "grad_norm": 11.88568115234375,
      "learning_rate": 1.08895881006865e-05,
      "loss": 0.8911,
      "step": 4454
    },
    {
      "epoch": 1.27,
      "grad_norm": 11.408987998962402,
      "learning_rate": 1.0885297482837529e-05,
      "loss": 1.0424,
      "step": 4455
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.65211296081543,
      "learning_rate": 1.0881006864988558e-05,
      "loss": 0.684,
      "step": 4456
    },
    {
      "epoch": 1.27,
      "grad_norm": 10.292887687683105,
      "learning_rate": 1.0876716247139588e-05,
      "loss": 0.623,
      "step": 4457
    },
    {
      "epoch": 1.28,
      "grad_norm": 8.282696723937988,
      "learning_rate": 1.0872425629290617e-05,
      "loss": 0.6805,
      "step": 4458
    },
    {
      "epoch": 1.28,
      "grad_norm": 11.467270851135254,
      "learning_rate": 1.0868135011441647e-05,
      "loss": 0.7225,
      "step": 4459
    },
    {
      "epoch": 1.28,
      "grad_norm": 13.024991035461426,
      "learning_rate": 1.0863844393592678e-05,
      "loss": 0.7548,
      "step": 4460
    },
    {
      "epoch": 1.28,
      "grad_norm": 9.389389991760254,
      "learning_rate": 1.0859553775743708e-05,
      "loss": 0.7311,
      "step": 4461
    },
    {
      "epoch": 1.28,
      "grad_norm": 10.110858917236328,
      "learning_rate": 1.0855263157894737e-05,
      "loss": 0.5706,
      "step": 4462
    },
    {
      "epoch": 1.28,
      "grad_norm": 8.631179809570312,
      "learning_rate": 1.0850972540045768e-05,
      "loss": 0.6806,
      "step": 4463
    },
    {
      "epoch": 1.28,
      "grad_norm": 10.76997184753418,
      "learning_rate": 1.0846681922196796e-05,
      "loss": 1.02,
      "step": 4464
    },
    {
      "epoch": 1.28,
      "grad_norm": 11.195799827575684,
      "learning_rate": 1.0842391304347827e-05,
      "loss": 0.7281,
      "step": 4465
    },
    {
      "epoch": 1.28,
      "grad_norm": 11.992496490478516,
      "learning_rate": 1.0838100686498856e-05,
      "loss": 0.938,
      "step": 4466
    },
    {
      "epoch": 1.28,
      "grad_norm": 10.878107070922852,
      "learning_rate": 1.0833810068649886e-05,
      "loss": 0.8651,
      "step": 4467
    },
    {
      "epoch": 1.28,
      "grad_norm": 11.575766563415527,
      "learning_rate": 1.0829519450800915e-05,
      "loss": 0.7973,
      "step": 4468
    },
    {
      "epoch": 1.28,
      "grad_norm": 8.012828826904297,
      "learning_rate": 1.0825228832951945e-05,
      "loss": 0.7017,
      "step": 4469
    },
    {
      "epoch": 1.28,
      "grad_norm": 9.665970802307129,
      "learning_rate": 1.0820938215102976e-05,
      "loss": 0.7088,
      "step": 4470
    },
    {
      "epoch": 1.28,
      "grad_norm": 9.927847862243652,
      "learning_rate": 1.0816647597254005e-05,
      "loss": 0.8645,
      "step": 4471
    },
    {
      "epoch": 1.28,
      "grad_norm": 9.595250129699707,
      "learning_rate": 1.0812356979405035e-05,
      "loss": 0.8021,
      "step": 4472
    },
    {
      "epoch": 1.28,
      "grad_norm": 11.20498275756836,
      "learning_rate": 1.0808066361556064e-05,
      "loss": 0.8367,
      "step": 4473
    },
    {
      "epoch": 1.28,
      "grad_norm": 10.3494234085083,
      "learning_rate": 1.0803775743707094e-05,
      "loss": 0.7279,
      "step": 4474
    },
    {
      "epoch": 1.28,
      "grad_norm": 8.78724193572998,
      "learning_rate": 1.0799485125858123e-05,
      "loss": 0.7025,
      "step": 4475
    },
    {
      "epoch": 1.28,
      "grad_norm": 9.069414138793945,
      "learning_rate": 1.0795194508009153e-05,
      "loss": 0.6999,
      "step": 4476
    },
    {
      "epoch": 1.28,
      "grad_norm": 12.110283851623535,
      "learning_rate": 1.0790903890160182e-05,
      "loss": 0.7431,
      "step": 4477
    },
    {
      "epoch": 1.28,
      "grad_norm": 10.447746276855469,
      "learning_rate": 1.0786613272311214e-05,
      "loss": 0.8934,
      "step": 4478
    },
    {
      "epoch": 1.28,
      "grad_norm": 9.537714958190918,
      "learning_rate": 1.0782322654462243e-05,
      "loss": 0.7105,
      "step": 4479
    },
    {
      "epoch": 1.28,
      "grad_norm": 10.018208503723145,
      "learning_rate": 1.0778032036613274e-05,
      "loss": 0.8458,
      "step": 4480
    },
    {
      "epoch": 1.28,
      "grad_norm": 10.869131088256836,
      "learning_rate": 1.0773741418764302e-05,
      "loss": 0.6133,
      "step": 4481
    },
    {
      "epoch": 1.28,
      "grad_norm": 10.848339080810547,
      "learning_rate": 1.0769450800915331e-05,
      "loss": 0.8584,
      "step": 4482
    },
    {
      "epoch": 1.28,
      "grad_norm": 12.082961082458496,
      "learning_rate": 1.0765160183066362e-05,
      "loss": 0.9047,
      "step": 4483
    },
    {
      "epoch": 1.28,
      "grad_norm": 10.2700834274292,
      "learning_rate": 1.076086956521739e-05,
      "loss": 0.4975,
      "step": 4484
    },
    {
      "epoch": 1.28,
      "grad_norm": 10.529144287109375,
      "learning_rate": 1.075657894736842e-05,
      "loss": 0.9198,
      "step": 4485
    },
    {
      "epoch": 1.28,
      "grad_norm": 11.378095626831055,
      "learning_rate": 1.0752288329519451e-05,
      "loss": 0.7958,
      "step": 4486
    },
    {
      "epoch": 1.28,
      "grad_norm": 10.02692985534668,
      "learning_rate": 1.0747997711670482e-05,
      "loss": 0.9134,
      "step": 4487
    },
    {
      "epoch": 1.28,
      "grad_norm": 8.611089706420898,
      "learning_rate": 1.074370709382151e-05,
      "loss": 0.7818,
      "step": 4488
    },
    {
      "epoch": 1.28,
      "grad_norm": 8.577508926391602,
      "learning_rate": 1.0739416475972541e-05,
      "loss": 0.5102,
      "step": 4489
    },
    {
      "epoch": 1.28,
      "grad_norm": 10.593514442443848,
      "learning_rate": 1.073512585812357e-05,
      "loss": 0.7386,
      "step": 4490
    },
    {
      "epoch": 1.28,
      "grad_norm": 8.785354614257812,
      "learning_rate": 1.07308352402746e-05,
      "loss": 0.4635,
      "step": 4491
    },
    {
      "epoch": 1.28,
      "grad_norm": 11.214665412902832,
      "learning_rate": 1.0726544622425629e-05,
      "loss": 1.0071,
      "step": 4492
    },
    {
      "epoch": 1.29,
      "grad_norm": 10.643991470336914,
      "learning_rate": 1.072225400457666e-05,
      "loss": 0.7866,
      "step": 4493
    },
    {
      "epoch": 1.29,
      "grad_norm": 11.196982383728027,
      "learning_rate": 1.0717963386727688e-05,
      "loss": 0.9896,
      "step": 4494
    },
    {
      "epoch": 1.29,
      "grad_norm": 9.391378402709961,
      "learning_rate": 1.0713672768878719e-05,
      "loss": 0.7679,
      "step": 4495
    },
    {
      "epoch": 1.29,
      "grad_norm": 8.35649585723877,
      "learning_rate": 1.0709382151029749e-05,
      "loss": 0.5818,
      "step": 4496
    },
    {
      "epoch": 1.29,
      "grad_norm": 12.581069946289062,
      "learning_rate": 1.0705091533180778e-05,
      "loss": 0.7659,
      "step": 4497
    },
    {
      "epoch": 1.29,
      "grad_norm": 8.290343284606934,
      "learning_rate": 1.0700800915331808e-05,
      "loss": 0.7201,
      "step": 4498
    },
    {
      "epoch": 1.29,
      "grad_norm": 10.348004341125488,
      "learning_rate": 1.0696510297482837e-05,
      "loss": 0.811,
      "step": 4499
    },
    {
      "epoch": 1.29,
      "grad_norm": 12.528030395507812,
      "learning_rate": 1.0692219679633868e-05,
      "loss": 0.9127,
      "step": 4500
    },
    {
      "epoch": 1.29,
      "grad_norm": 10.638726234436035,
      "learning_rate": 1.0687929061784896e-05,
      "loss": 0.5826,
      "step": 4501
    },
    {
      "epoch": 1.29,
      "grad_norm": 11.5818452835083,
      "learning_rate": 1.0683638443935927e-05,
      "loss": 0.792,
      "step": 4502
    },
    {
      "epoch": 1.29,
      "grad_norm": 12.89651107788086,
      "learning_rate": 1.0679347826086957e-05,
      "loss": 0.7994,
      "step": 4503
    },
    {
      "epoch": 1.29,
      "grad_norm": 11.800990104675293,
      "learning_rate": 1.0675057208237988e-05,
      "loss": 1.1297,
      "step": 4504
    },
    {
      "epoch": 1.29,
      "grad_norm": 11.796900749206543,
      "learning_rate": 1.0670766590389017e-05,
      "loss": 0.8113,
      "step": 4505
    },
    {
      "epoch": 1.29,
      "grad_norm": 12.544075012207031,
      "learning_rate": 1.0666475972540047e-05,
      "loss": 0.8221,
      "step": 4506
    },
    {
      "epoch": 1.29,
      "grad_norm": 9.659513473510742,
      "learning_rate": 1.0662185354691076e-05,
      "loss": 0.7253,
      "step": 4507
    },
    {
      "epoch": 1.29,
      "grad_norm": 12.628800392150879,
      "learning_rate": 1.0657894736842106e-05,
      "loss": 0.9197,
      "step": 4508
    },
    {
      "epoch": 1.29,
      "grad_norm": 9.06775951385498,
      "learning_rate": 1.0653604118993135e-05,
      "loss": 0.7415,
      "step": 4509
    },
    {
      "epoch": 1.29,
      "grad_norm": 8.520203590393066,
      "learning_rate": 1.0649313501144164e-05,
      "loss": 0.7182,
      "step": 4510
    },
    {
      "epoch": 1.29,
      "grad_norm": 8.279836654663086,
      "learning_rate": 1.0645022883295194e-05,
      "loss": 0.5338,
      "step": 4511
    },
    {
      "epoch": 1.29,
      "grad_norm": 10.691319465637207,
      "learning_rate": 1.0640732265446225e-05,
      "loss": 0.6707,
      "step": 4512
    },
    {
      "epoch": 1.29,
      "grad_norm": 11.865562438964844,
      "learning_rate": 1.0636441647597255e-05,
      "loss": 0.9687,
      "step": 4513
    },
    {
      "epoch": 1.29,
      "grad_norm": 8.937037467956543,
      "learning_rate": 1.0632151029748284e-05,
      "loss": 0.6894,
      "step": 4514
    },
    {
      "epoch": 1.29,
      "grad_norm": 12.400564193725586,
      "learning_rate": 1.0627860411899314e-05,
      "loss": 1.2345,
      "step": 4515
    },
    {
      "epoch": 1.29,
      "grad_norm": 8.49405288696289,
      "learning_rate": 1.0623569794050343e-05,
      "loss": 0.4766,
      "step": 4516
    },
    {
      "epoch": 1.29,
      "grad_norm": 8.88058853149414,
      "learning_rate": 1.0619279176201374e-05,
      "loss": 0.6505,
      "step": 4517
    },
    {
      "epoch": 1.29,
      "grad_norm": 8.913090705871582,
      "learning_rate": 1.0614988558352402e-05,
      "loss": 0.6876,
      "step": 4518
    },
    {
      "epoch": 1.29,
      "grad_norm": 10.349099159240723,
      "learning_rate": 1.0610697940503433e-05,
      "loss": 0.7209,
      "step": 4519
    },
    {
      "epoch": 1.29,
      "grad_norm": 11.539715766906738,
      "learning_rate": 1.0606407322654463e-05,
      "loss": 0.9389,
      "step": 4520
    },
    {
      "epoch": 1.29,
      "grad_norm": 9.029921531677246,
      "learning_rate": 1.0602116704805492e-05,
      "loss": 0.6983,
      "step": 4521
    },
    {
      "epoch": 1.29,
      "grad_norm": 10.217121124267578,
      "learning_rate": 1.0597826086956523e-05,
      "loss": 1.0282,
      "step": 4522
    },
    {
      "epoch": 1.29,
      "grad_norm": 9.556905746459961,
      "learning_rate": 1.0593535469107551e-05,
      "loss": 0.6441,
      "step": 4523
    },
    {
      "epoch": 1.29,
      "grad_norm": 13.417732238769531,
      "learning_rate": 1.0589244851258582e-05,
      "loss": 0.9225,
      "step": 4524
    },
    {
      "epoch": 1.29,
      "grad_norm": 10.074112892150879,
      "learning_rate": 1.058495423340961e-05,
      "loss": 0.7527,
      "step": 4525
    },
    {
      "epoch": 1.29,
      "grad_norm": 10.604214668273926,
      "learning_rate": 1.0580663615560641e-05,
      "loss": 0.7732,
      "step": 4526
    },
    {
      "epoch": 1.29,
      "grad_norm": 10.65685749053955,
      "learning_rate": 1.057637299771167e-05,
      "loss": 1.0403,
      "step": 4527
    },
    {
      "epoch": 1.3,
      "grad_norm": 8.894769668579102,
      "learning_rate": 1.05720823798627e-05,
      "loss": 0.6774,
      "step": 4528
    },
    {
      "epoch": 1.3,
      "grad_norm": 9.923067092895508,
      "learning_rate": 1.056779176201373e-05,
      "loss": 0.5607,
      "step": 4529
    },
    {
      "epoch": 1.3,
      "grad_norm": 8.149468421936035,
      "learning_rate": 1.0563501144164761e-05,
      "loss": 0.5877,
      "step": 4530
    },
    {
      "epoch": 1.3,
      "grad_norm": 10.948150634765625,
      "learning_rate": 1.055921052631579e-05,
      "loss": 1.0175,
      "step": 4531
    },
    {
      "epoch": 1.3,
      "grad_norm": 12.211588859558105,
      "learning_rate": 1.055491990846682e-05,
      "loss": 0.8954,
      "step": 4532
    },
    {
      "epoch": 1.3,
      "grad_norm": 12.116756439208984,
      "learning_rate": 1.0550629290617849e-05,
      "loss": 0.9302,
      "step": 4533
    },
    {
      "epoch": 1.3,
      "grad_norm": 9.606791496276855,
      "learning_rate": 1.054633867276888e-05,
      "loss": 0.7397,
      "step": 4534
    },
    {
      "epoch": 1.3,
      "grad_norm": 8.567057609558105,
      "learning_rate": 1.0542048054919908e-05,
      "loss": 0.5171,
      "step": 4535
    },
    {
      "epoch": 1.3,
      "grad_norm": 10.134928703308105,
      "learning_rate": 1.0537757437070937e-05,
      "loss": 0.7642,
      "step": 4536
    },
    {
      "epoch": 1.3,
      "grad_norm": 12.260796546936035,
      "learning_rate": 1.0533466819221968e-05,
      "loss": 0.7348,
      "step": 4537
    },
    {
      "epoch": 1.3,
      "grad_norm": 8.267999649047852,
      "learning_rate": 1.0529176201372998e-05,
      "loss": 0.6745,
      "step": 4538
    },
    {
      "epoch": 1.3,
      "grad_norm": 10.089160919189453,
      "learning_rate": 1.0524885583524028e-05,
      "loss": 0.7294,
      "step": 4539
    },
    {
      "epoch": 1.3,
      "grad_norm": 11.858624458312988,
      "learning_rate": 1.0520594965675057e-05,
      "loss": 1.0909,
      "step": 4540
    },
    {
      "epoch": 1.3,
      "grad_norm": 9.252303123474121,
      "learning_rate": 1.0516304347826088e-05,
      "loss": 0.5499,
      "step": 4541
    },
    {
      "epoch": 1.3,
      "grad_norm": 10.532364845275879,
      "learning_rate": 1.0512013729977116e-05,
      "loss": 0.716,
      "step": 4542
    },
    {
      "epoch": 1.3,
      "grad_norm": 10.907944679260254,
      "learning_rate": 1.0507723112128147e-05,
      "loss": 0.8436,
      "step": 4543
    },
    {
      "epoch": 1.3,
      "grad_norm": 9.34953498840332,
      "learning_rate": 1.0503432494279176e-05,
      "loss": 0.9408,
      "step": 4544
    },
    {
      "epoch": 1.3,
      "grad_norm": 8.764162063598633,
      "learning_rate": 1.0499141876430206e-05,
      "loss": 0.7812,
      "step": 4545
    },
    {
      "epoch": 1.3,
      "grad_norm": 10.903592109680176,
      "learning_rate": 1.0494851258581237e-05,
      "loss": 0.9772,
      "step": 4546
    },
    {
      "epoch": 1.3,
      "grad_norm": 9.145940780639648,
      "learning_rate": 1.0490560640732267e-05,
      "loss": 0.5603,
      "step": 4547
    },
    {
      "epoch": 1.3,
      "grad_norm": 8.859952926635742,
      "learning_rate": 1.0486270022883296e-05,
      "loss": 0.493,
      "step": 4548
    },
    {
      "epoch": 1.3,
      "grad_norm": 10.92986011505127,
      "learning_rate": 1.0481979405034325e-05,
      "loss": 0.8482,
      "step": 4549
    },
    {
      "epoch": 1.3,
      "grad_norm": 8.599848747253418,
      "learning_rate": 1.0477688787185355e-05,
      "loss": 0.7271,
      "step": 4550
    },
    {
      "epoch": 1.3,
      "grad_norm": 9.906673431396484,
      "learning_rate": 1.0473398169336384e-05,
      "loss": 0.7742,
      "step": 4551
    },
    {
      "epoch": 1.3,
      "grad_norm": 12.310548782348633,
      "learning_rate": 1.0469107551487414e-05,
      "loss": 0.6319,
      "step": 4552
    },
    {
      "epoch": 1.3,
      "grad_norm": 11.670842170715332,
      "learning_rate": 1.0464816933638443e-05,
      "loss": 0.8078,
      "step": 4553
    },
    {
      "epoch": 1.3,
      "grad_norm": 11.398588180541992,
      "learning_rate": 1.0460526315789474e-05,
      "loss": 0.9609,
      "step": 4554
    },
    {
      "epoch": 1.3,
      "grad_norm": 11.271353721618652,
      "learning_rate": 1.0456235697940504e-05,
      "loss": 0.8449,
      "step": 4555
    },
    {
      "epoch": 1.3,
      "grad_norm": 10.082358360290527,
      "learning_rate": 1.0451945080091534e-05,
      "loss": 0.7383,
      "step": 4556
    },
    {
      "epoch": 1.3,
      "grad_norm": 9.931964874267578,
      "learning_rate": 1.0447654462242563e-05,
      "loss": 0.6965,
      "step": 4557
    },
    {
      "epoch": 1.3,
      "grad_norm": 10.41787052154541,
      "learning_rate": 1.0443363844393594e-05,
      "loss": 0.934,
      "step": 4558
    },
    {
      "epoch": 1.3,
      "grad_norm": 12.855107307434082,
      "learning_rate": 1.0439073226544622e-05,
      "loss": 0.8262,
      "step": 4559
    },
    {
      "epoch": 1.3,
      "grad_norm": 11.924995422363281,
      "learning_rate": 1.0434782608695653e-05,
      "loss": 1.0188,
      "step": 4560
    },
    {
      "epoch": 1.3,
      "grad_norm": 10.832833290100098,
      "learning_rate": 1.0430491990846682e-05,
      "loss": 0.8195,
      "step": 4561
    },
    {
      "epoch": 1.3,
      "grad_norm": 7.587615489959717,
      "learning_rate": 1.042620137299771e-05,
      "loss": 0.539,
      "step": 4562
    },
    {
      "epoch": 1.31,
      "grad_norm": 8.638379096984863,
      "learning_rate": 1.0421910755148743e-05,
      "loss": 0.574,
      "step": 4563
    },
    {
      "epoch": 1.31,
      "grad_norm": 8.017379760742188,
      "learning_rate": 1.0417620137299771e-05,
      "loss": 0.685,
      "step": 4564
    },
    {
      "epoch": 1.31,
      "grad_norm": 9.898755073547363,
      "learning_rate": 1.0413329519450802e-05,
      "loss": 0.8604,
      "step": 4565
    },
    {
      "epoch": 1.31,
      "grad_norm": 7.718481063842773,
      "learning_rate": 1.040903890160183e-05,
      "loss": 0.7614,
      "step": 4566
    },
    {
      "epoch": 1.31,
      "grad_norm": 10.322260856628418,
      "learning_rate": 1.0404748283752861e-05,
      "loss": 0.7968,
      "step": 4567
    },
    {
      "epoch": 1.31,
      "grad_norm": 13.542582511901855,
      "learning_rate": 1.040045766590389e-05,
      "loss": 0.8681,
      "step": 4568
    },
    {
      "epoch": 1.31,
      "grad_norm": 9.584424018859863,
      "learning_rate": 1.039616704805492e-05,
      "loss": 0.9568,
      "step": 4569
    },
    {
      "epoch": 1.31,
      "grad_norm": 12.441803932189941,
      "learning_rate": 1.0391876430205949e-05,
      "loss": 0.9867,
      "step": 4570
    },
    {
      "epoch": 1.31,
      "grad_norm": 8.160896301269531,
      "learning_rate": 1.038758581235698e-05,
      "loss": 0.569,
      "step": 4571
    },
    {
      "epoch": 1.31,
      "grad_norm": 12.349885940551758,
      "learning_rate": 1.038329519450801e-05,
      "loss": 0.8415,
      "step": 4572
    },
    {
      "epoch": 1.31,
      "grad_norm": 10.131543159484863,
      "learning_rate": 1.037900457665904e-05,
      "loss": 0.8312,
      "step": 4573
    },
    {
      "epoch": 1.31,
      "grad_norm": 10.335249900817871,
      "learning_rate": 1.037471395881007e-05,
      "loss": 0.916,
      "step": 4574
    },
    {
      "epoch": 1.31,
      "grad_norm": 9.307873725891113,
      "learning_rate": 1.0370423340961098e-05,
      "loss": 0.5624,
      "step": 4575
    },
    {
      "epoch": 1.31,
      "grad_norm": 10.228752136230469,
      "learning_rate": 1.0366132723112128e-05,
      "loss": 0.7126,
      "step": 4576
    },
    {
      "epoch": 1.31,
      "grad_norm": 12.114232063293457,
      "learning_rate": 1.0361842105263157e-05,
      "loss": 0.8773,
      "step": 4577
    },
    {
      "epoch": 1.31,
      "grad_norm": 10.18767261505127,
      "learning_rate": 1.0357551487414188e-05,
      "loss": 0.8004,
      "step": 4578
    },
    {
      "epoch": 1.31,
      "grad_norm": 10.70250129699707,
      "learning_rate": 1.0353260869565216e-05,
      "loss": 0.8739,
      "step": 4579
    },
    {
      "epoch": 1.31,
      "grad_norm": 10.009703636169434,
      "learning_rate": 1.0348970251716249e-05,
      "loss": 0.8867,
      "step": 4580
    },
    {
      "epoch": 1.31,
      "grad_norm": 8.726730346679688,
      "learning_rate": 1.0344679633867277e-05,
      "loss": 0.6418,
      "step": 4581
    },
    {
      "epoch": 1.31,
      "grad_norm": 10.187596321105957,
      "learning_rate": 1.0340389016018308e-05,
      "loss": 0.6098,
      "step": 4582
    },
    {
      "epoch": 1.31,
      "grad_norm": 11.032061576843262,
      "learning_rate": 1.0336098398169337e-05,
      "loss": 0.84,
      "step": 4583
    },
    {
      "epoch": 1.31,
      "grad_norm": 9.076355934143066,
      "learning_rate": 1.0331807780320367e-05,
      "loss": 0.9079,
      "step": 4584
    },
    {
      "epoch": 1.31,
      "grad_norm": 9.244355201721191,
      "learning_rate": 1.0327517162471396e-05,
      "loss": 0.6617,
      "step": 4585
    },
    {
      "epoch": 1.31,
      "grad_norm": 8.106156349182129,
      "learning_rate": 1.0323226544622426e-05,
      "loss": 0.5986,
      "step": 4586
    },
    {
      "epoch": 1.31,
      "grad_norm": 10.858281135559082,
      "learning_rate": 1.0318935926773455e-05,
      "loss": 0.9367,
      "step": 4587
    },
    {
      "epoch": 1.31,
      "grad_norm": 9.763096809387207,
      "learning_rate": 1.0314645308924484e-05,
      "loss": 0.8339,
      "step": 4588
    },
    {
      "epoch": 1.31,
      "grad_norm": 9.7815580368042,
      "learning_rate": 1.0310354691075516e-05,
      "loss": 0.808,
      "step": 4589
    },
    {
      "epoch": 1.31,
      "grad_norm": 7.393942832946777,
      "learning_rate": 1.0306064073226545e-05,
      "loss": 0.4532,
      "step": 4590
    },
    {
      "epoch": 1.31,
      "grad_norm": 11.11086368560791,
      "learning_rate": 1.0301773455377575e-05,
      "loss": 0.7224,
      "step": 4591
    },
    {
      "epoch": 1.31,
      "grad_norm": 9.721107482910156,
      "learning_rate": 1.0297482837528604e-05,
      "loss": 0.8394,
      "step": 4592
    },
    {
      "epoch": 1.31,
      "grad_norm": 10.71251392364502,
      "learning_rate": 1.0293192219679634e-05,
      "loss": 0.9271,
      "step": 4593
    },
    {
      "epoch": 1.31,
      "grad_norm": 7.635794639587402,
      "learning_rate": 1.0288901601830663e-05,
      "loss": 0.597,
      "step": 4594
    },
    {
      "epoch": 1.31,
      "grad_norm": 10.38532829284668,
      "learning_rate": 1.0284610983981694e-05,
      "loss": 0.8118,
      "step": 4595
    },
    {
      "epoch": 1.31,
      "grad_norm": 10.115407943725586,
      "learning_rate": 1.0280320366132722e-05,
      "loss": 0.7534,
      "step": 4596
    },
    {
      "epoch": 1.31,
      "grad_norm": 10.365483283996582,
      "learning_rate": 1.0276029748283753e-05,
      "loss": 1.1437,
      "step": 4597
    },
    {
      "epoch": 1.32,
      "grad_norm": 10.19611644744873,
      "learning_rate": 1.0271739130434783e-05,
      "loss": 0.7595,
      "step": 4598
    },
    {
      "epoch": 1.32,
      "grad_norm": 11.781787872314453,
      "learning_rate": 1.0267448512585814e-05,
      "loss": 0.8448,
      "step": 4599
    },
    {
      "epoch": 1.32,
      "grad_norm": 10.059295654296875,
      "learning_rate": 1.0263157894736843e-05,
      "loss": 0.6639,
      "step": 4600
    },
    {
      "epoch": 1.32,
      "grad_norm": 10.477483749389648,
      "learning_rate": 1.0258867276887871e-05,
      "loss": 0.7552,
      "step": 4601
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.580055236816406,
      "learning_rate": 1.0254576659038902e-05,
      "loss": 0.9293,
      "step": 4602
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.448568344116211,
      "learning_rate": 1.025028604118993e-05,
      "loss": 0.716,
      "step": 4603
    },
    {
      "epoch": 1.32,
      "grad_norm": 13.001675605773926,
      "learning_rate": 1.0245995423340961e-05,
      "loss": 0.9883,
      "step": 4604
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.869972229003906,
      "learning_rate": 1.024170480549199e-05,
      "loss": 0.6913,
      "step": 4605
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.498185157775879,
      "learning_rate": 1.0237414187643022e-05,
      "loss": 0.7251,
      "step": 4606
    },
    {
      "epoch": 1.32,
      "grad_norm": 12.074101448059082,
      "learning_rate": 1.023312356979405e-05,
      "loss": 0.6859,
      "step": 4607
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.54678726196289,
      "learning_rate": 1.0228832951945081e-05,
      "loss": 0.7507,
      "step": 4608
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.271499633789062,
      "learning_rate": 1.022454233409611e-05,
      "loss": 0.6925,
      "step": 4609
    },
    {
      "epoch": 1.32,
      "grad_norm": 12.213305473327637,
      "learning_rate": 1.022025171624714e-05,
      "loss": 0.965,
      "step": 4610
    },
    {
      "epoch": 1.32,
      "grad_norm": 11.851221084594727,
      "learning_rate": 1.021596109839817e-05,
      "loss": 0.6818,
      "step": 4611
    },
    {
      "epoch": 1.32,
      "grad_norm": 10.92638111114502,
      "learning_rate": 1.02116704805492e-05,
      "loss": 0.8467,
      "step": 4612
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.772537231445312,
      "learning_rate": 1.0207379862700228e-05,
      "loss": 0.6508,
      "step": 4613
    },
    {
      "epoch": 1.32,
      "grad_norm": 10.279985427856445,
      "learning_rate": 1.0203089244851257e-05,
      "loss": 1.0279,
      "step": 4614
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.189204216003418,
      "learning_rate": 1.019879862700229e-05,
      "loss": 0.8367,
      "step": 4615
    },
    {
      "epoch": 1.32,
      "grad_norm": 10.385782241821289,
      "learning_rate": 1.0194508009153318e-05,
      "loss": 0.9493,
      "step": 4616
    },
    {
      "epoch": 1.32,
      "grad_norm": 10.314265251159668,
      "learning_rate": 1.0190217391304349e-05,
      "loss": 0.7412,
      "step": 4617
    },
    {
      "epoch": 1.32,
      "grad_norm": 11.677253723144531,
      "learning_rate": 1.0185926773455377e-05,
      "loss": 0.6281,
      "step": 4618
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.144380569458008,
      "learning_rate": 1.0181636155606408e-05,
      "loss": 0.7098,
      "step": 4619
    },
    {
      "epoch": 1.32,
      "grad_norm": 12.974599838256836,
      "learning_rate": 1.0177345537757437e-05,
      "loss": 1.1372,
      "step": 4620
    },
    {
      "epoch": 1.32,
      "grad_norm": 11.553279876708984,
      "learning_rate": 1.0173054919908467e-05,
      "loss": 0.9577,
      "step": 4621
    },
    {
      "epoch": 1.32,
      "grad_norm": 10.324438095092773,
      "learning_rate": 1.0168764302059496e-05,
      "loss": 0.6501,
      "step": 4622
    },
    {
      "epoch": 1.32,
      "grad_norm": 13.246345520019531,
      "learning_rate": 1.0164473684210528e-05,
      "loss": 1.1245,
      "step": 4623
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.062743186950684,
      "learning_rate": 1.0160183066361557e-05,
      "loss": 0.6808,
      "step": 4624
    },
    {
      "epoch": 1.32,
      "grad_norm": 10.063706398010254,
      "learning_rate": 1.0155892448512587e-05,
      "loss": 0.8209,
      "step": 4625
    },
    {
      "epoch": 1.32,
      "grad_norm": 8.378332138061523,
      "learning_rate": 1.0151601830663616e-05,
      "loss": 0.4858,
      "step": 4626
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.984187126159668,
      "learning_rate": 1.0147311212814645e-05,
      "loss": 0.7595,
      "step": 4627
    },
    {
      "epoch": 1.32,
      "grad_norm": 12.664240837097168,
      "learning_rate": 1.0143020594965675e-05,
      "loss": 0.8782,
      "step": 4628
    },
    {
      "epoch": 1.32,
      "grad_norm": 8.787510871887207,
      "learning_rate": 1.0138729977116704e-05,
      "loss": 0.7938,
      "step": 4629
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.451238632202148,
      "learning_rate": 1.0134439359267734e-05,
      "loss": 0.7193,
      "step": 4630
    },
    {
      "epoch": 1.32,
      "grad_norm": 10.58932876586914,
      "learning_rate": 1.0130148741418763e-05,
      "loss": 0.6507,
      "step": 4631
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.03365707397461,
      "learning_rate": 1.0125858123569795e-05,
      "loss": 0.7437,
      "step": 4632
    },
    {
      "epoch": 1.33,
      "grad_norm": 8.374075889587402,
      "learning_rate": 1.0121567505720824e-05,
      "loss": 0.6675,
      "step": 4633
    },
    {
      "epoch": 1.33,
      "grad_norm": 8.497538566589355,
      "learning_rate": 1.0117276887871855e-05,
      "loss": 0.6729,
      "step": 4634
    },
    {
      "epoch": 1.33,
      "grad_norm": 9.651091575622559,
      "learning_rate": 1.0112986270022883e-05,
      "loss": 0.6992,
      "step": 4635
    },
    {
      "epoch": 1.33,
      "grad_norm": 10.816292762756348,
      "learning_rate": 1.0108695652173914e-05,
      "loss": 0.8129,
      "step": 4636
    },
    {
      "epoch": 1.33,
      "grad_norm": 9.455986022949219,
      "learning_rate": 1.0104405034324943e-05,
      "loss": 0.7505,
      "step": 4637
    },
    {
      "epoch": 1.33,
      "grad_norm": 9.267904281616211,
      "learning_rate": 1.0100114416475973e-05,
      "loss": 0.6653,
      "step": 4638
    },
    {
      "epoch": 1.33,
      "grad_norm": 10.088541984558105,
      "learning_rate": 1.0095823798627002e-05,
      "loss": 0.5034,
      "step": 4639
    },
    {
      "epoch": 1.33,
      "grad_norm": 9.960984230041504,
      "learning_rate": 1.0091533180778032e-05,
      "loss": 0.634,
      "step": 4640
    },
    {
      "epoch": 1.33,
      "grad_norm": 10.747187614440918,
      "learning_rate": 1.0087242562929063e-05,
      "loss": 0.791,
      "step": 4641
    },
    {
      "epoch": 1.33,
      "grad_norm": 8.487506866455078,
      "learning_rate": 1.0082951945080092e-05,
      "loss": 0.7181,
      "step": 4642
    },
    {
      "epoch": 1.33,
      "grad_norm": 8.82575798034668,
      "learning_rate": 1.0078661327231122e-05,
      "loss": 0.7306,
      "step": 4643
    },
    {
      "epoch": 1.33,
      "grad_norm": 9.52674674987793,
      "learning_rate": 1.007437070938215e-05,
      "loss": 0.6141,
      "step": 4644
    },
    {
      "epoch": 1.33,
      "grad_norm": 11.346635818481445,
      "learning_rate": 1.0070080091533181e-05,
      "loss": 0.7195,
      "step": 4645
    },
    {
      "epoch": 1.33,
      "grad_norm": 9.532054901123047,
      "learning_rate": 1.006578947368421e-05,
      "loss": 0.5254,
      "step": 4646
    },
    {
      "epoch": 1.33,
      "grad_norm": 9.944281578063965,
      "learning_rate": 1.006149885583524e-05,
      "loss": 0.8458,
      "step": 4647
    },
    {
      "epoch": 1.33,
      "grad_norm": 10.513945579528809,
      "learning_rate": 1.005720823798627e-05,
      "loss": 0.8692,
      "step": 4648
    },
    {
      "epoch": 1.33,
      "grad_norm": 10.14871883392334,
      "learning_rate": 1.0052917620137301e-05,
      "loss": 0.6585,
      "step": 4649
    },
    {
      "epoch": 1.33,
      "grad_norm": 9.397774696350098,
      "learning_rate": 1.004862700228833e-05,
      "loss": 0.5747,
      "step": 4650
    },
    {
      "epoch": 1.33,
      "grad_norm": 10.431659698486328,
      "learning_rate": 1.004433638443936e-05,
      "loss": 0.7969,
      "step": 4651
    },
    {
      "epoch": 1.33,
      "grad_norm": 11.2437105178833,
      "learning_rate": 1.004004576659039e-05,
      "loss": 0.6863,
      "step": 4652
    },
    {
      "epoch": 1.33,
      "grad_norm": 11.774871826171875,
      "learning_rate": 1.0035755148741418e-05,
      "loss": 1.0087,
      "step": 4653
    },
    {
      "epoch": 1.33,
      "grad_norm": 12.120447158813477,
      "learning_rate": 1.0031464530892449e-05,
      "loss": 1.0396,
      "step": 4654
    },
    {
      "epoch": 1.33,
      "grad_norm": 9.248188018798828,
      "learning_rate": 1.0027173913043477e-05,
      "loss": 0.6938,
      "step": 4655
    },
    {
      "epoch": 1.33,
      "grad_norm": 8.619288444519043,
      "learning_rate": 1.0022883295194508e-05,
      "loss": 0.6464,
      "step": 4656
    },
    {
      "epoch": 1.33,
      "grad_norm": 11.303775787353516,
      "learning_rate": 1.0018592677345537e-05,
      "loss": 0.7225,
      "step": 4657
    },
    {
      "epoch": 1.33,
      "grad_norm": 11.534278869628906,
      "learning_rate": 1.0014302059496569e-05,
      "loss": 0.7771,
      "step": 4658
    },
    {
      "epoch": 1.33,
      "grad_norm": 10.095863342285156,
      "learning_rate": 1.0010011441647598e-05,
      "loss": 0.6307,
      "step": 4659
    },
    {
      "epoch": 1.33,
      "grad_norm": 9.884389877319336,
      "learning_rate": 1.0005720823798628e-05,
      "loss": 0.593,
      "step": 4660
    },
    {
      "epoch": 1.33,
      "grad_norm": 10.065033912658691,
      "learning_rate": 1.0001430205949657e-05,
      "loss": 0.6567,
      "step": 4661
    },
    {
      "epoch": 1.33,
      "grad_norm": 9.432774543762207,
      "learning_rate": 9.997139588100687e-06,
      "loss": 0.575,
      "step": 4662
    },
    {
      "epoch": 1.33,
      "grad_norm": 10.539703369140625,
      "learning_rate": 9.992848970251716e-06,
      "loss": 0.556,
      "step": 4663
    },
    {
      "epoch": 1.33,
      "grad_norm": 8.618727684020996,
      "learning_rate": 9.988558352402746e-06,
      "loss": 0.5203,
      "step": 4664
    },
    {
      "epoch": 1.33,
      "grad_norm": 8.832174301147461,
      "learning_rate": 9.984267734553775e-06,
      "loss": 0.7284,
      "step": 4665
    },
    {
      "epoch": 1.33,
      "grad_norm": 10.143733978271484,
      "learning_rate": 9.979977116704806e-06,
      "loss": 0.6738,
      "step": 4666
    },
    {
      "epoch": 1.33,
      "grad_norm": 9.429113388061523,
      "learning_rate": 9.975686498855836e-06,
      "loss": 0.5794,
      "step": 4667
    },
    {
      "epoch": 1.34,
      "grad_norm": 8.873308181762695,
      "learning_rate": 9.971395881006865e-06,
      "loss": 0.5068,
      "step": 4668
    },
    {
      "epoch": 1.34,
      "grad_norm": 9.968809127807617,
      "learning_rate": 9.967105263157895e-06,
      "loss": 0.6062,
      "step": 4669
    },
    {
      "epoch": 1.34,
      "grad_norm": 10.752883911132812,
      "learning_rate": 9.962814645308924e-06,
      "loss": 0.6434,
      "step": 4670
    },
    {
      "epoch": 1.34,
      "grad_norm": 8.815367698669434,
      "learning_rate": 9.958524027459955e-06,
      "loss": 0.6924,
      "step": 4671
    },
    {
      "epoch": 1.34,
      "grad_norm": 9.490862846374512,
      "learning_rate": 9.954233409610983e-06,
      "loss": 0.9181,
      "step": 4672
    },
    {
      "epoch": 1.34,
      "grad_norm": 10.75500202178955,
      "learning_rate": 9.949942791762014e-06,
      "loss": 0.9343,
      "step": 4673
    },
    {
      "epoch": 1.34,
      "grad_norm": 12.002522468566895,
      "learning_rate": 9.945652173913043e-06,
      "loss": 0.8571,
      "step": 4674
    },
    {
      "epoch": 1.34,
      "grad_norm": 9.769218444824219,
      "learning_rate": 9.941361556064075e-06,
      "loss": 0.7145,
      "step": 4675
    },
    {
      "epoch": 1.34,
      "grad_norm": 8.448454856872559,
      "learning_rate": 9.937070938215104e-06,
      "loss": 0.6893,
      "step": 4676
    },
    {
      "epoch": 1.34,
      "grad_norm": 11.79710865020752,
      "learning_rate": 9.932780320366134e-06,
      "loss": 0.8783,
      "step": 4677
    },
    {
      "epoch": 1.34,
      "grad_norm": 11.979585647583008,
      "learning_rate": 9.928489702517163e-06,
      "loss": 0.7278,
      "step": 4678
    },
    {
      "epoch": 1.34,
      "grad_norm": 8.608783721923828,
      "learning_rate": 9.924199084668192e-06,
      "loss": 0.567,
      "step": 4679
    },
    {
      "epoch": 1.34,
      "grad_norm": 8.807727813720703,
      "learning_rate": 9.919908466819222e-06,
      "loss": 0.5357,
      "step": 4680
    },
    {
      "epoch": 1.34,
      "grad_norm": 6.912734031677246,
      "learning_rate": 9.91561784897025e-06,
      "loss": 0.4204,
      "step": 4681
    },
    {
      "epoch": 1.34,
      "grad_norm": 10.059368133544922,
      "learning_rate": 9.911327231121281e-06,
      "loss": 0.7899,
      "step": 4682
    },
    {
      "epoch": 1.34,
      "grad_norm": 11.73801040649414,
      "learning_rate": 9.907036613272312e-06,
      "loss": 0.7755,
      "step": 4683
    },
    {
      "epoch": 1.34,
      "grad_norm": 9.313863754272461,
      "learning_rate": 9.902745995423342e-06,
      "loss": 0.7283,
      "step": 4684
    },
    {
      "epoch": 1.34,
      "grad_norm": 10.158074378967285,
      "learning_rate": 9.898455377574371e-06,
      "loss": 0.6462,
      "step": 4685
    },
    {
      "epoch": 1.34,
      "grad_norm": 10.189724922180176,
      "learning_rate": 9.894164759725401e-06,
      "loss": 0.9259,
      "step": 4686
    },
    {
      "epoch": 1.34,
      "grad_norm": 11.322792053222656,
      "learning_rate": 9.88987414187643e-06,
      "loss": 0.5923,
      "step": 4687
    },
    {
      "epoch": 1.34,
      "grad_norm": 8.962874412536621,
      "learning_rate": 9.88558352402746e-06,
      "loss": 0.7261,
      "step": 4688
    },
    {
      "epoch": 1.34,
      "grad_norm": 10.788043975830078,
      "learning_rate": 9.88129290617849e-06,
      "loss": 0.8296,
      "step": 4689
    },
    {
      "epoch": 1.34,
      "grad_norm": 13.454601287841797,
      "learning_rate": 9.87700228832952e-06,
      "loss": 0.7895,
      "step": 4690
    },
    {
      "epoch": 1.34,
      "grad_norm": 10.123590469360352,
      "learning_rate": 9.872711670480549e-06,
      "loss": 0.6016,
      "step": 4691
    },
    {
      "epoch": 1.34,
      "grad_norm": 7.896024227142334,
      "learning_rate": 9.868421052631579e-06,
      "loss": 0.6095,
      "step": 4692
    },
    {
      "epoch": 1.34,
      "grad_norm": 10.664581298828125,
      "learning_rate": 9.86413043478261e-06,
      "loss": 0.7457,
      "step": 4693
    },
    {
      "epoch": 1.34,
      "grad_norm": 11.129467964172363,
      "learning_rate": 9.859839816933638e-06,
      "loss": 0.7181,
      "step": 4694
    },
    {
      "epoch": 1.34,
      "grad_norm": 11.753436088562012,
      "learning_rate": 9.855549199084669e-06,
      "loss": 0.8247,
      "step": 4695
    },
    {
      "epoch": 1.34,
      "grad_norm": 8.845498085021973,
      "learning_rate": 9.851258581235698e-06,
      "loss": 0.5541,
      "step": 4696
    },
    {
      "epoch": 1.34,
      "grad_norm": 12.261682510375977,
      "learning_rate": 9.846967963386728e-06,
      "loss": 0.9145,
      "step": 4697
    },
    {
      "epoch": 1.34,
      "grad_norm": 11.292670249938965,
      "learning_rate": 9.842677345537757e-06,
      "loss": 0.757,
      "step": 4698
    },
    {
      "epoch": 1.34,
      "grad_norm": 13.406408309936523,
      "learning_rate": 9.838386727688787e-06,
      "loss": 1.1041,
      "step": 4699
    },
    {
      "epoch": 1.34,
      "grad_norm": 10.857704162597656,
      "learning_rate": 9.834096109839818e-06,
      "loss": 0.669,
      "step": 4700
    },
    {
      "epoch": 1.34,
      "grad_norm": 11.81068229675293,
      "learning_rate": 9.829805491990848e-06,
      "loss": 0.9058,
      "step": 4701
    },
    {
      "epoch": 1.34,
      "grad_norm": 9.678082466125488,
      "learning_rate": 9.825514874141877e-06,
      "loss": 0.7138,
      "step": 4702
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.629570007324219,
      "learning_rate": 9.821224256292907e-06,
      "loss": 0.7891,
      "step": 4703
    },
    {
      "epoch": 1.35,
      "grad_norm": 8.208858489990234,
      "learning_rate": 9.816933638443936e-06,
      "loss": 0.7526,
      "step": 4704
    },
    {
      "epoch": 1.35,
      "grad_norm": 11.03001880645752,
      "learning_rate": 9.812643020594967e-06,
      "loss": 0.9222,
      "step": 4705
    },
    {
      "epoch": 1.35,
      "grad_norm": 13.611908912658691,
      "learning_rate": 9.808352402745995e-06,
      "loss": 0.8673,
      "step": 4706
    },
    {
      "epoch": 1.35,
      "grad_norm": 11.656730651855469,
      "learning_rate": 9.804061784897024e-06,
      "loss": 0.6069,
      "step": 4707
    },
    {
      "epoch": 1.35,
      "grad_norm": 10.210548400878906,
      "learning_rate": 9.799771167048055e-06,
      "loss": 0.7342,
      "step": 4708
    },
    {
      "epoch": 1.35,
      "grad_norm": 10.285316467285156,
      "learning_rate": 9.795480549199085e-06,
      "loss": 0.9595,
      "step": 4709
    },
    {
      "epoch": 1.35,
      "grad_norm": 10.383859634399414,
      "learning_rate": 9.791189931350116e-06,
      "loss": 0.6653,
      "step": 4710
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.29322338104248,
      "learning_rate": 9.786899313501144e-06,
      "loss": 0.6616,
      "step": 4711
    },
    {
      "epoch": 1.35,
      "grad_norm": 11.92971134185791,
      "learning_rate": 9.782608695652175e-06,
      "loss": 0.9191,
      "step": 4712
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.166658401489258,
      "learning_rate": 9.778318077803204e-06,
      "loss": 0.6697,
      "step": 4713
    },
    {
      "epoch": 1.35,
      "grad_norm": 10.312832832336426,
      "learning_rate": 9.774027459954234e-06,
      "loss": 0.8248,
      "step": 4714
    },
    {
      "epoch": 1.35,
      "grad_norm": 12.214032173156738,
      "learning_rate": 9.769736842105263e-06,
      "loss": 0.7471,
      "step": 4715
    },
    {
      "epoch": 1.35,
      "grad_norm": 11.253389358520508,
      "learning_rate": 9.765446224256293e-06,
      "loss": 0.7021,
      "step": 4716
    },
    {
      "epoch": 1.35,
      "grad_norm": 11.42955493927002,
      "learning_rate": 9.761155606407322e-06,
      "loss": 0.7276,
      "step": 4717
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.202191352844238,
      "learning_rate": 9.756864988558354e-06,
      "loss": 0.6752,
      "step": 4718
    },
    {
      "epoch": 1.35,
      "grad_norm": 10.934903144836426,
      "learning_rate": 9.752574370709383e-06,
      "loss": 0.7681,
      "step": 4719
    },
    {
      "epoch": 1.35,
      "grad_norm": 10.316263198852539,
      "learning_rate": 9.748283752860412e-06,
      "loss": 0.7321,
      "step": 4720
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.91502857208252,
      "learning_rate": 9.743993135011442e-06,
      "loss": 0.8845,
      "step": 4721
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.166141510009766,
      "learning_rate": 9.739702517162471e-06,
      "loss": 0.7623,
      "step": 4722
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.540265083312988,
      "learning_rate": 9.735411899313501e-06,
      "loss": 0.8254,
      "step": 4723
    },
    {
      "epoch": 1.35,
      "grad_norm": 10.553126335144043,
      "learning_rate": 9.73112128146453e-06,
      "loss": 0.6477,
      "step": 4724
    },
    {
      "epoch": 1.35,
      "grad_norm": 10.808560371398926,
      "learning_rate": 9.72683066361556e-06,
      "loss": 0.8561,
      "step": 4725
    },
    {
      "epoch": 1.35,
      "grad_norm": 8.343021392822266,
      "learning_rate": 9.722540045766591e-06,
      "loss": 0.7708,
      "step": 4726
    },
    {
      "epoch": 1.35,
      "grad_norm": 8.422737121582031,
      "learning_rate": 9.718249427917622e-06,
      "loss": 0.5716,
      "step": 4727
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.480222702026367,
      "learning_rate": 9.71395881006865e-06,
      "loss": 0.7948,
      "step": 4728
    },
    {
      "epoch": 1.35,
      "grad_norm": 12.601561546325684,
      "learning_rate": 9.70966819221968e-06,
      "loss": 0.9539,
      "step": 4729
    },
    {
      "epoch": 1.35,
      "grad_norm": 8.475797653198242,
      "learning_rate": 9.70537757437071e-06,
      "loss": 0.597,
      "step": 4730
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.0745267868042,
      "learning_rate": 9.70108695652174e-06,
      "loss": 0.6777,
      "step": 4731
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.286697387695312,
      "learning_rate": 9.696796338672769e-06,
      "loss": 0.6527,
      "step": 4732
    },
    {
      "epoch": 1.35,
      "grad_norm": 10.54956340789795,
      "learning_rate": 9.692505720823798e-06,
      "loss": 0.6245,
      "step": 4733
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.258973121643066,
      "learning_rate": 9.688215102974828e-06,
      "loss": 0.6062,
      "step": 4734
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.817070007324219,
      "learning_rate": 9.683924485125858e-06,
      "loss": 0.6548,
      "step": 4735
    },
    {
      "epoch": 1.35,
      "grad_norm": 7.8161301612854,
      "learning_rate": 9.679633867276889e-06,
      "loss": 0.6321,
      "step": 4736
    },
    {
      "epoch": 1.35,
      "grad_norm": 8.645702362060547,
      "learning_rate": 9.675343249427918e-06,
      "loss": 0.5972,
      "step": 4737
    },
    {
      "epoch": 1.36,
      "grad_norm": 10.85690689086914,
      "learning_rate": 9.671052631578948e-06,
      "loss": 0.7775,
      "step": 4738
    },
    {
      "epoch": 1.36,
      "grad_norm": 10.943389892578125,
      "learning_rate": 9.666762013729977e-06,
      "loss": 0.8891,
      "step": 4739
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.511855125427246,
      "learning_rate": 9.662471395881007e-06,
      "loss": 0.6053,
      "step": 4740
    },
    {
      "epoch": 1.36,
      "grad_norm": 11.409897804260254,
      "learning_rate": 9.658180778032036e-06,
      "loss": 0.8391,
      "step": 4741
    },
    {
      "epoch": 1.36,
      "grad_norm": 10.824045181274414,
      "learning_rate": 9.653890160183067e-06,
      "loss": 0.7154,
      "step": 4742
    },
    {
      "epoch": 1.36,
      "grad_norm": 11.163288116455078,
      "learning_rate": 9.649599542334097e-06,
      "loss": 0.9376,
      "step": 4743
    },
    {
      "epoch": 1.36,
      "grad_norm": 12.635465621948242,
      "learning_rate": 9.645308924485128e-06,
      "loss": 0.8153,
      "step": 4744
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.781368255615234,
      "learning_rate": 9.641018306636156e-06,
      "loss": 0.7844,
      "step": 4745
    },
    {
      "epoch": 1.36,
      "grad_norm": 10.899678230285645,
      "learning_rate": 9.636727688787185e-06,
      "loss": 0.8468,
      "step": 4746
    },
    {
      "epoch": 1.36,
      "grad_norm": 10.143167495727539,
      "learning_rate": 9.632437070938216e-06,
      "loss": 0.7934,
      "step": 4747
    },
    {
      "epoch": 1.36,
      "grad_norm": 10.395912170410156,
      "learning_rate": 9.628146453089244e-06,
      "loss": 0.6913,
      "step": 4748
    },
    {
      "epoch": 1.36,
      "grad_norm": 11.43149471282959,
      "learning_rate": 9.623855835240275e-06,
      "loss": 0.8784,
      "step": 4749
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.007231712341309,
      "learning_rate": 9.619565217391304e-06,
      "loss": 0.671,
      "step": 4750
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.453091621398926,
      "learning_rate": 9.615274599542334e-06,
      "loss": 0.7037,
      "step": 4751
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.27692985534668,
      "learning_rate": 9.610983981693364e-06,
      "loss": 0.6657,
      "step": 4752
    },
    {
      "epoch": 1.36,
      "grad_norm": 10.366813659667969,
      "learning_rate": 9.606693363844395e-06,
      "loss": 0.5964,
      "step": 4753
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.110732078552246,
      "learning_rate": 9.602402745995424e-06,
      "loss": 0.6431,
      "step": 4754
    },
    {
      "epoch": 1.36,
      "grad_norm": 8.982587814331055,
      "learning_rate": 9.598112128146454e-06,
      "loss": 0.5971,
      "step": 4755
    },
    {
      "epoch": 1.36,
      "grad_norm": 10.255237579345703,
      "learning_rate": 9.593821510297483e-06,
      "loss": 0.794,
      "step": 4756
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.470309257507324,
      "learning_rate": 9.589530892448513e-06,
      "loss": 0.6336,
      "step": 4757
    },
    {
      "epoch": 1.36,
      "grad_norm": 10.870800018310547,
      "learning_rate": 9.585240274599542e-06,
      "loss": 1.029,
      "step": 4758
    },
    {
      "epoch": 1.36,
      "grad_norm": 10.282632827758789,
      "learning_rate": 9.580949656750571e-06,
      "loss": 1.0672,
      "step": 4759
    },
    {
      "epoch": 1.36,
      "grad_norm": 10.640077590942383,
      "learning_rate": 9.576659038901603e-06,
      "loss": 0.6764,
      "step": 4760
    },
    {
      "epoch": 1.36,
      "grad_norm": 8.993135452270508,
      "learning_rate": 9.572368421052632e-06,
      "loss": 0.6048,
      "step": 4761
    },
    {
      "epoch": 1.36,
      "grad_norm": 13.737961769104004,
      "learning_rate": 9.568077803203662e-06,
      "loss": 0.9287,
      "step": 4762
    },
    {
      "epoch": 1.36,
      "grad_norm": 11.905489921569824,
      "learning_rate": 9.563787185354691e-06,
      "loss": 0.7215,
      "step": 4763
    },
    {
      "epoch": 1.36,
      "grad_norm": 8.13138484954834,
      "learning_rate": 9.559496567505722e-06,
      "loss": 0.8252,
      "step": 4764
    },
    {
      "epoch": 1.36,
      "grad_norm": 11.043505668640137,
      "learning_rate": 9.55520594965675e-06,
      "loss": 0.8947,
      "step": 4765
    },
    {
      "epoch": 1.36,
      "grad_norm": 10.608139038085938,
      "learning_rate": 9.55091533180778e-06,
      "loss": 0.5847,
      "step": 4766
    },
    {
      "epoch": 1.36,
      "grad_norm": 8.512394905090332,
      "learning_rate": 9.54662471395881e-06,
      "loss": 0.7247,
      "step": 4767
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.464174270629883,
      "learning_rate": 9.54233409610984e-06,
      "loss": 0.6077,
      "step": 4768
    },
    {
      "epoch": 1.36,
      "grad_norm": 14.590190887451172,
      "learning_rate": 9.53804347826087e-06,
      "loss": 0.8994,
      "step": 4769
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.813478469848633,
      "learning_rate": 9.533752860411901e-06,
      "loss": 0.7307,
      "step": 4770
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.709108352661133,
      "learning_rate": 9.52946224256293e-06,
      "loss": 0.7144,
      "step": 4771
    },
    {
      "epoch": 1.36,
      "grad_norm": 11.180831909179688,
      "learning_rate": 9.525171624713958e-06,
      "loss": 0.9568,
      "step": 4772
    },
    {
      "epoch": 1.37,
      "grad_norm": 8.907102584838867,
      "learning_rate": 9.520881006864989e-06,
      "loss": 0.6992,
      "step": 4773
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.115985870361328,
      "learning_rate": 9.516590389016018e-06,
      "loss": 0.8351,
      "step": 4774
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.451964378356934,
      "learning_rate": 9.512299771167048e-06,
      "loss": 1.1227,
      "step": 4775
    },
    {
      "epoch": 1.37,
      "grad_norm": 9.790129661560059,
      "learning_rate": 9.508009153318077e-06,
      "loss": 0.5994,
      "step": 4776
    },
    {
      "epoch": 1.37,
      "grad_norm": 9.342123985290527,
      "learning_rate": 9.503718535469107e-06,
      "loss": 0.5986,
      "step": 4777
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.826268196105957,
      "learning_rate": 9.499427917620138e-06,
      "loss": 0.6164,
      "step": 4778
    },
    {
      "epoch": 1.37,
      "grad_norm": 12.784324645996094,
      "learning_rate": 9.495137299771168e-06,
      "loss": 0.7825,
      "step": 4779
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.142254829406738,
      "learning_rate": 9.490846681922197e-06,
      "loss": 0.7858,
      "step": 4780
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.85872745513916,
      "learning_rate": 9.486556064073227e-06,
      "loss": 0.8585,
      "step": 4781
    },
    {
      "epoch": 1.37,
      "grad_norm": 12.147515296936035,
      "learning_rate": 9.482265446224256e-06,
      "loss": 0.932,
      "step": 4782
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.450340270996094,
      "learning_rate": 9.477974828375287e-06,
      "loss": 0.7762,
      "step": 4783
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.950103759765625,
      "learning_rate": 9.473684210526315e-06,
      "loss": 0.5991,
      "step": 4784
    },
    {
      "epoch": 1.37,
      "grad_norm": 11.839818954467773,
      "learning_rate": 9.469393592677344e-06,
      "loss": 0.7055,
      "step": 4785
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.563735008239746,
      "learning_rate": 9.465102974828376e-06,
      "loss": 0.8403,
      "step": 4786
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.265862464904785,
      "learning_rate": 9.460812356979405e-06,
      "loss": 0.6591,
      "step": 4787
    },
    {
      "epoch": 1.37,
      "grad_norm": 9.833613395690918,
      "learning_rate": 9.456521739130436e-06,
      "loss": 0.7097,
      "step": 4788
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.166386604309082,
      "learning_rate": 9.452231121281464e-06,
      "loss": 0.7283,
      "step": 4789
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.88394832611084,
      "learning_rate": 9.447940503432495e-06,
      "loss": 0.67,
      "step": 4790
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.023024559020996,
      "learning_rate": 9.443649885583524e-06,
      "loss": 0.7822,
      "step": 4791
    },
    {
      "epoch": 1.37,
      "grad_norm": 11.198458671569824,
      "learning_rate": 9.439359267734554e-06,
      "loss": 0.7167,
      "step": 4792
    },
    {
      "epoch": 1.37,
      "grad_norm": 8.456743240356445,
      "learning_rate": 9.435068649885583e-06,
      "loss": 0.7078,
      "step": 4793
    },
    {
      "epoch": 1.37,
      "grad_norm": 8.988679885864258,
      "learning_rate": 9.430778032036613e-06,
      "loss": 0.5257,
      "step": 4794
    },
    {
      "epoch": 1.37,
      "grad_norm": 8.99492073059082,
      "learning_rate": 9.426487414187644e-06,
      "loss": 0.7596,
      "step": 4795
    },
    {
      "epoch": 1.37,
      "grad_norm": 11.329377174377441,
      "learning_rate": 9.422196796338674e-06,
      "loss": 0.741,
      "step": 4796
    },
    {
      "epoch": 1.37,
      "grad_norm": 9.76819896697998,
      "learning_rate": 9.417906178489703e-06,
      "loss": 0.661,
      "step": 4797
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.368531227111816,
      "learning_rate": 9.413615560640732e-06,
      "loss": 0.6498,
      "step": 4798
    },
    {
      "epoch": 1.37,
      "grad_norm": 13.120408058166504,
      "learning_rate": 9.409324942791762e-06,
      "loss": 0.7884,
      "step": 4799
    },
    {
      "epoch": 1.37,
      "grad_norm": 11.418519973754883,
      "learning_rate": 9.405034324942791e-06,
      "loss": 0.73,
      "step": 4800
    },
    {
      "epoch": 1.37,
      "grad_norm": 8.889922142028809,
      "learning_rate": 9.400743707093821e-06,
      "loss": 0.6517,
      "step": 4801
    },
    {
      "epoch": 1.37,
      "grad_norm": 8.791597366333008,
      "learning_rate": 9.39645308924485e-06,
      "loss": 0.5533,
      "step": 4802
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.5639009475708,
      "learning_rate": 9.392162471395882e-06,
      "loss": 0.8333,
      "step": 4803
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.931742668151855,
      "learning_rate": 9.387871853546911e-06,
      "loss": 0.6539,
      "step": 4804
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.472341537475586,
      "learning_rate": 9.383581235697942e-06,
      "loss": 0.762,
      "step": 4805
    },
    {
      "epoch": 1.37,
      "grad_norm": 12.587716102600098,
      "learning_rate": 9.37929061784897e-06,
      "loss": 0.7897,
      "step": 4806
    },
    {
      "epoch": 1.38,
      "grad_norm": 11.926453590393066,
      "learning_rate": 9.375000000000001e-06,
      "loss": 0.9278,
      "step": 4807
    },
    {
      "epoch": 1.38,
      "grad_norm": 13.291482925415039,
      "learning_rate": 9.37070938215103e-06,
      "loss": 0.8511,
      "step": 4808
    },
    {
      "epoch": 1.38,
      "grad_norm": 10.288946151733398,
      "learning_rate": 9.36641876430206e-06,
      "loss": 0.929,
      "step": 4809
    },
    {
      "epoch": 1.38,
      "grad_norm": 10.623122215270996,
      "learning_rate": 9.362128146453089e-06,
      "loss": 0.6458,
      "step": 4810
    },
    {
      "epoch": 1.38,
      "grad_norm": 10.468912124633789,
      "learning_rate": 9.357837528604118e-06,
      "loss": 0.6873,
      "step": 4811
    },
    {
      "epoch": 1.38,
      "grad_norm": 11.56838607788086,
      "learning_rate": 9.35354691075515e-06,
      "loss": 0.7875,
      "step": 4812
    },
    {
      "epoch": 1.38,
      "grad_norm": 11.725969314575195,
      "learning_rate": 9.349256292906179e-06,
      "loss": 0.8911,
      "step": 4813
    },
    {
      "epoch": 1.38,
      "grad_norm": 9.419934272766113,
      "learning_rate": 9.344965675057209e-06,
      "loss": 0.7614,
      "step": 4814
    },
    {
      "epoch": 1.38,
      "grad_norm": 11.268380165100098,
      "learning_rate": 9.340675057208238e-06,
      "loss": 0.7354,
      "step": 4815
    },
    {
      "epoch": 1.38,
      "grad_norm": 10.358317375183105,
      "learning_rate": 9.336384439359268e-06,
      "loss": 0.9548,
      "step": 4816
    },
    {
      "epoch": 1.38,
      "grad_norm": 8.72009563446045,
      "learning_rate": 9.332093821510297e-06,
      "loss": 0.582,
      "step": 4817
    },
    {
      "epoch": 1.38,
      "grad_norm": 10.13125991821289,
      "learning_rate": 9.327803203661327e-06,
      "loss": 0.6966,
      "step": 4818
    },
    {
      "epoch": 1.38,
      "grad_norm": 9.3549222946167,
      "learning_rate": 9.323512585812356e-06,
      "loss": 0.6896,
      "step": 4819
    },
    {
      "epoch": 1.38,
      "grad_norm": 10.324894905090332,
      "learning_rate": 9.319221967963388e-06,
      "loss": 0.7683,
      "step": 4820
    },
    {
      "epoch": 1.38,
      "grad_norm": 8.656771659851074,
      "learning_rate": 9.314931350114417e-06,
      "loss": 0.5893,
      "step": 4821
    },
    {
      "epoch": 1.38,
      "grad_norm": 11.520030975341797,
      "learning_rate": 9.310640732265448e-06,
      "loss": 1.0155,
      "step": 4822
    },
    {
      "epoch": 1.38,
      "grad_norm": 10.97884464263916,
      "learning_rate": 9.306350114416476e-06,
      "loss": 0.7444,
      "step": 4823
    },
    {
      "epoch": 1.38,
      "grad_norm": 13.074646949768066,
      "learning_rate": 9.302059496567505e-06,
      "loss": 0.9445,
      "step": 4824
    },
    {
      "epoch": 1.38,
      "grad_norm": 12.072041511535645,
      "learning_rate": 9.297768878718536e-06,
      "loss": 0.9269,
      "step": 4825
    },
    {
      "epoch": 1.38,
      "grad_norm": 8.852041244506836,
      "learning_rate": 9.293478260869564e-06,
      "loss": 0.585,
      "step": 4826
    },
    {
      "epoch": 1.38,
      "grad_norm": 10.590184211730957,
      "learning_rate": 9.289187643020595e-06,
      "loss": 0.7778,
      "step": 4827
    },
    {
      "epoch": 1.38,
      "grad_norm": 11.690543174743652,
      "learning_rate": 9.284897025171624e-06,
      "loss": 0.878,
      "step": 4828
    },
    {
      "epoch": 1.38,
      "grad_norm": 10.135167121887207,
      "learning_rate": 9.280606407322656e-06,
      "loss": 0.9343,
      "step": 4829
    },
    {
      "epoch": 1.38,
      "grad_norm": 8.094252586364746,
      "learning_rate": 9.276315789473685e-06,
      "loss": 0.6903,
      "step": 4830
    },
    {
      "epoch": 1.38,
      "grad_norm": 10.056294441223145,
      "learning_rate": 9.272025171624715e-06,
      "loss": 0.8418,
      "step": 4831
    },
    {
      "epoch": 1.38,
      "grad_norm": 11.848834037780762,
      "learning_rate": 9.267734553775744e-06,
      "loss": 0.6992,
      "step": 4832
    },
    {
      "epoch": 1.38,
      "grad_norm": 10.331584930419922,
      "learning_rate": 9.263443935926774e-06,
      "loss": 0.9414,
      "step": 4833
    },
    {
      "epoch": 1.38,
      "grad_norm": 9.208252906799316,
      "learning_rate": 9.259153318077803e-06,
      "loss": 0.5844,
      "step": 4834
    },
    {
      "epoch": 1.38,
      "grad_norm": 9.118937492370605,
      "learning_rate": 9.254862700228833e-06,
      "loss": 0.9003,
      "step": 4835
    },
    {
      "epoch": 1.38,
      "grad_norm": 9.917899131774902,
      "learning_rate": 9.250572082379862e-06,
      "loss": 0.6907,
      "step": 4836
    },
    {
      "epoch": 1.38,
      "grad_norm": 11.160566329956055,
      "learning_rate": 9.246281464530893e-06,
      "loss": 0.7855,
      "step": 4837
    },
    {
      "epoch": 1.38,
      "grad_norm": 9.2199125289917,
      "learning_rate": 9.241990846681923e-06,
      "loss": 0.7515,
      "step": 4838
    },
    {
      "epoch": 1.38,
      "grad_norm": 10.523056983947754,
      "learning_rate": 9.237700228832952e-06,
      "loss": 0.862,
      "step": 4839
    },
    {
      "epoch": 1.38,
      "grad_norm": 8.743433952331543,
      "learning_rate": 9.233409610983982e-06,
      "loss": 0.5145,
      "step": 4840
    },
    {
      "epoch": 1.38,
      "grad_norm": 10.881135940551758,
      "learning_rate": 9.229118993135011e-06,
      "loss": 0.8192,
      "step": 4841
    },
    {
      "epoch": 1.39,
      "grad_norm": 7.766360282897949,
      "learning_rate": 9.224828375286042e-06,
      "loss": 0.528,
      "step": 4842
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.022210121154785,
      "learning_rate": 9.22053775743707e-06,
      "loss": 0.7403,
      "step": 4843
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.161805152893066,
      "learning_rate": 9.216247139588101e-06,
      "loss": 0.8515,
      "step": 4844
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.7399263381958,
      "learning_rate": 9.21195652173913e-06,
      "loss": 0.7869,
      "step": 4845
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.652128219604492,
      "learning_rate": 9.207665903890162e-06,
      "loss": 0.6249,
      "step": 4846
    },
    {
      "epoch": 1.39,
      "grad_norm": 13.797903060913086,
      "learning_rate": 9.20337528604119e-06,
      "loss": 0.8755,
      "step": 4847
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.382393836975098,
      "learning_rate": 9.199084668192221e-06,
      "loss": 1.107,
      "step": 4848
    },
    {
      "epoch": 1.39,
      "grad_norm": 11.453968048095703,
      "learning_rate": 9.19479405034325e-06,
      "loss": 0.9609,
      "step": 4849
    },
    {
      "epoch": 1.39,
      "grad_norm": 9.114245414733887,
      "learning_rate": 9.190503432494279e-06,
      "loss": 0.5484,
      "step": 4850
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.182549476623535,
      "learning_rate": 9.186212814645309e-06,
      "loss": 0.6344,
      "step": 4851
    },
    {
      "epoch": 1.39,
      "grad_norm": 12.673012733459473,
      "learning_rate": 9.181922196796338e-06,
      "loss": 0.9958,
      "step": 4852
    },
    {
      "epoch": 1.39,
      "grad_norm": 9.4663667678833,
      "learning_rate": 9.177631578947368e-06,
      "loss": 0.7675,
      "step": 4853
    },
    {
      "epoch": 1.39,
      "grad_norm": 9.954926490783691,
      "learning_rate": 9.173340961098397e-06,
      "loss": 0.7907,
      "step": 4854
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.192892074584961,
      "learning_rate": 9.16905034324943e-06,
      "loss": 0.6709,
      "step": 4855
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.08601188659668,
      "learning_rate": 9.164759725400458e-06,
      "loss": 0.6938,
      "step": 4856
    },
    {
      "epoch": 1.39,
      "grad_norm": 11.993782997131348,
      "learning_rate": 9.160469107551488e-06,
      "loss": 0.9492,
      "step": 4857
    },
    {
      "epoch": 1.39,
      "grad_norm": 8.177780151367188,
      "learning_rate": 9.156178489702517e-06,
      "loss": 0.512,
      "step": 4858
    },
    {
      "epoch": 1.39,
      "grad_norm": 9.7926025390625,
      "learning_rate": 9.151887871853548e-06,
      "loss": 0.6743,
      "step": 4859
    },
    {
      "epoch": 1.39,
      "grad_norm": 13.589574813842773,
      "learning_rate": 9.147597254004576e-06,
      "loss": 0.7877,
      "step": 4860
    },
    {
      "epoch": 1.39,
      "grad_norm": 8.5984525680542,
      "learning_rate": 9.143306636155607e-06,
      "loss": 0.6091,
      "step": 4861
    },
    {
      "epoch": 1.39,
      "grad_norm": 9.036418914794922,
      "learning_rate": 9.139016018306636e-06,
      "loss": 0.7091,
      "step": 4862
    },
    {
      "epoch": 1.39,
      "grad_norm": 9.252519607543945,
      "learning_rate": 9.134725400457666e-06,
      "loss": 0.7672,
      "step": 4863
    },
    {
      "epoch": 1.39,
      "grad_norm": 11.83324146270752,
      "learning_rate": 9.130434782608697e-06,
      "loss": 1.0583,
      "step": 4864
    },
    {
      "epoch": 1.39,
      "grad_norm": 9.944408416748047,
      "learning_rate": 9.126144164759725e-06,
      "loss": 0.9297,
      "step": 4865
    },
    {
      "epoch": 1.39,
      "grad_norm": 9.997650146484375,
      "learning_rate": 9.121853546910756e-06,
      "loss": 1.0245,
      "step": 4866
    },
    {
      "epoch": 1.39,
      "grad_norm": 8.75586223602295,
      "learning_rate": 9.117562929061785e-06,
      "loss": 0.5483,
      "step": 4867
    },
    {
      "epoch": 1.39,
      "grad_norm": 11.224337577819824,
      "learning_rate": 9.113272311212815e-06,
      "loss": 0.6652,
      "step": 4868
    },
    {
      "epoch": 1.39,
      "grad_norm": 7.3334479331970215,
      "learning_rate": 9.108981693363844e-06,
      "loss": 0.634,
      "step": 4869
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.251147270202637,
      "learning_rate": 9.104691075514874e-06,
      "loss": 0.5847,
      "step": 4870
    },
    {
      "epoch": 1.39,
      "grad_norm": 9.901888847351074,
      "learning_rate": 9.100400457665903e-06,
      "loss": 0.6054,
      "step": 4871
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.24634075164795,
      "learning_rate": 9.096109839816935e-06,
      "loss": 0.8071,
      "step": 4872
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.110356330871582,
      "learning_rate": 9.091819221967964e-06,
      "loss": 0.7466,
      "step": 4873
    },
    {
      "epoch": 1.39,
      "grad_norm": 9.721631050109863,
      "learning_rate": 9.087528604118994e-06,
      "loss": 0.6514,
      "step": 4874
    },
    {
      "epoch": 1.39,
      "grad_norm": 7.648428916931152,
      "learning_rate": 9.083237986270023e-06,
      "loss": 0.5112,
      "step": 4875
    },
    {
      "epoch": 1.39,
      "grad_norm": 11.041352272033691,
      "learning_rate": 9.078947368421054e-06,
      "loss": 1.0079,
      "step": 4876
    },
    {
      "epoch": 1.4,
      "grad_norm": 7.543818950653076,
      "learning_rate": 9.074656750572082e-06,
      "loss": 0.5,
      "step": 4877
    },
    {
      "epoch": 1.4,
      "grad_norm": 7.59186315536499,
      "learning_rate": 9.070366132723111e-06,
      "loss": 0.6879,
      "step": 4878
    },
    {
      "epoch": 1.4,
      "grad_norm": 12.302193641662598,
      "learning_rate": 9.066075514874142e-06,
      "loss": 0.8182,
      "step": 4879
    },
    {
      "epoch": 1.4,
      "grad_norm": 7.864506244659424,
      "learning_rate": 9.061784897025172e-06,
      "loss": 0.7759,
      "step": 4880
    },
    {
      "epoch": 1.4,
      "grad_norm": 11.609857559204102,
      "learning_rate": 9.057494279176203e-06,
      "loss": 0.5229,
      "step": 4881
    },
    {
      "epoch": 1.4,
      "grad_norm": 7.691769123077393,
      "learning_rate": 9.053203661327231e-06,
      "loss": 0.6559,
      "step": 4882
    },
    {
      "epoch": 1.4,
      "grad_norm": 9.675695419311523,
      "learning_rate": 9.048913043478262e-06,
      "loss": 0.4562,
      "step": 4883
    },
    {
      "epoch": 1.4,
      "grad_norm": 10.433341026306152,
      "learning_rate": 9.04462242562929e-06,
      "loss": 0.8334,
      "step": 4884
    },
    {
      "epoch": 1.4,
      "grad_norm": 11.0515775680542,
      "learning_rate": 9.040331807780321e-06,
      "loss": 0.8456,
      "step": 4885
    },
    {
      "epoch": 1.4,
      "grad_norm": 10.080611228942871,
      "learning_rate": 9.03604118993135e-06,
      "loss": 0.718,
      "step": 4886
    },
    {
      "epoch": 1.4,
      "grad_norm": 11.835367202758789,
      "learning_rate": 9.03175057208238e-06,
      "loss": 0.9327,
      "step": 4887
    },
    {
      "epoch": 1.4,
      "grad_norm": 10.179398536682129,
      "learning_rate": 9.027459954233409e-06,
      "loss": 0.7596,
      "step": 4888
    },
    {
      "epoch": 1.4,
      "grad_norm": 10.481108665466309,
      "learning_rate": 9.02316933638444e-06,
      "loss": 0.7859,
      "step": 4889
    },
    {
      "epoch": 1.4,
      "grad_norm": 9.284229278564453,
      "learning_rate": 9.01887871853547e-06,
      "loss": 0.745,
      "step": 4890
    },
    {
      "epoch": 1.4,
      "grad_norm": 11.10659408569336,
      "learning_rate": 9.014588100686499e-06,
      "loss": 0.8434,
      "step": 4891
    },
    {
      "epoch": 1.4,
      "grad_norm": 9.546857833862305,
      "learning_rate": 9.010297482837529e-06,
      "loss": 0.8202,
      "step": 4892
    },
    {
      "epoch": 1.4,
      "grad_norm": 12.39665699005127,
      "learning_rate": 9.006006864988558e-06,
      "loss": 0.8723,
      "step": 4893
    },
    {
      "epoch": 1.4,
      "grad_norm": 12.502572059631348,
      "learning_rate": 9.001716247139588e-06,
      "loss": 1.0049,
      "step": 4894
    },
    {
      "epoch": 1.4,
      "grad_norm": 8.685144424438477,
      "learning_rate": 8.997425629290617e-06,
      "loss": 0.6399,
      "step": 4895
    },
    {
      "epoch": 1.4,
      "grad_norm": 12.4518404006958,
      "learning_rate": 8.993135011441648e-06,
      "loss": 0.9697,
      "step": 4896
    },
    {
      "epoch": 1.4,
      "grad_norm": 8.57666015625,
      "learning_rate": 8.988844393592678e-06,
      "loss": 0.5471,
      "step": 4897
    },
    {
      "epoch": 1.4,
      "grad_norm": 8.991373062133789,
      "learning_rate": 8.984553775743709e-06,
      "loss": 0.6502,
      "step": 4898
    },
    {
      "epoch": 1.4,
      "grad_norm": 8.511014938354492,
      "learning_rate": 8.980263157894737e-06,
      "loss": 0.5548,
      "step": 4899
    },
    {
      "epoch": 1.4,
      "grad_norm": 9.343938827514648,
      "learning_rate": 8.975972540045768e-06,
      "loss": 0.8212,
      "step": 4900
    },
    {
      "epoch": 1.4,
      "grad_norm": 12.516120910644531,
      "learning_rate": 8.971681922196797e-06,
      "loss": 0.8418,
      "step": 4901
    },
    {
      "epoch": 1.4,
      "grad_norm": 9.186525344848633,
      "learning_rate": 8.967391304347827e-06,
      "loss": 0.6415,
      "step": 4902
    },
    {
      "epoch": 1.4,
      "grad_norm": 9.080911636352539,
      "learning_rate": 8.963100686498856e-06,
      "loss": 0.7878,
      "step": 4903
    },
    {
      "epoch": 1.4,
      "grad_norm": 9.257999420166016,
      "learning_rate": 8.958810068649885e-06,
      "loss": 0.702,
      "step": 4904
    },
    {
      "epoch": 1.4,
      "grad_norm": 11.845600128173828,
      "learning_rate": 8.954519450800915e-06,
      "loss": 0.7782,
      "step": 4905
    },
    {
      "epoch": 1.4,
      "grad_norm": 12.40800666809082,
      "learning_rate": 8.950228832951945e-06,
      "loss": 0.9072,
      "step": 4906
    },
    {
      "epoch": 1.4,
      "grad_norm": 10.087108612060547,
      "learning_rate": 8.945938215102976e-06,
      "loss": 0.9709,
      "step": 4907
    },
    {
      "epoch": 1.4,
      "grad_norm": 10.10132884979248,
      "learning_rate": 8.941647597254005e-06,
      "loss": 1.0056,
      "step": 4908
    },
    {
      "epoch": 1.4,
      "grad_norm": 12.176788330078125,
      "learning_rate": 8.937356979405035e-06,
      "loss": 0.7682,
      "step": 4909
    },
    {
      "epoch": 1.4,
      "grad_norm": 7.847817420959473,
      "learning_rate": 8.933066361556064e-06,
      "loss": 0.5401,
      "step": 4910
    },
    {
      "epoch": 1.4,
      "grad_norm": 6.8500895500183105,
      "learning_rate": 8.928775743707094e-06,
      "loss": 0.4209,
      "step": 4911
    },
    {
      "epoch": 1.41,
      "grad_norm": 11.234733581542969,
      "learning_rate": 8.924485125858123e-06,
      "loss": 0.6664,
      "step": 4912
    },
    {
      "epoch": 1.41,
      "grad_norm": 9.983650207519531,
      "learning_rate": 8.920194508009154e-06,
      "loss": 0.5712,
      "step": 4913
    },
    {
      "epoch": 1.41,
      "grad_norm": 11.488990783691406,
      "learning_rate": 8.915903890160182e-06,
      "loss": 0.9202,
      "step": 4914
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.480536460876465,
      "learning_rate": 8.911613272311215e-06,
      "loss": 0.5609,
      "step": 4915
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.097199440002441,
      "learning_rate": 8.907322654462243e-06,
      "loss": 0.6498,
      "step": 4916
    },
    {
      "epoch": 1.41,
      "grad_norm": 12.0875825881958,
      "learning_rate": 8.903032036613272e-06,
      "loss": 0.7561,
      "step": 4917
    },
    {
      "epoch": 1.41,
      "grad_norm": 9.851820945739746,
      "learning_rate": 8.898741418764303e-06,
      "loss": 0.609,
      "step": 4918
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.984374046325684,
      "learning_rate": 8.894450800915331e-06,
      "loss": 0.6938,
      "step": 4919
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.526418685913086,
      "learning_rate": 8.890160183066362e-06,
      "loss": 0.9217,
      "step": 4920
    },
    {
      "epoch": 1.41,
      "grad_norm": 11.407646179199219,
      "learning_rate": 8.88586956521739e-06,
      "loss": 0.7912,
      "step": 4921
    },
    {
      "epoch": 1.41,
      "grad_norm": 13.315333366394043,
      "learning_rate": 8.881578947368421e-06,
      "loss": 0.7204,
      "step": 4922
    },
    {
      "epoch": 1.41,
      "grad_norm": 11.823431015014648,
      "learning_rate": 8.877288329519451e-06,
      "loss": 0.8202,
      "step": 4923
    },
    {
      "epoch": 1.41,
      "grad_norm": 12.348017692565918,
      "learning_rate": 8.872997711670482e-06,
      "loss": 0.8588,
      "step": 4924
    },
    {
      "epoch": 1.41,
      "grad_norm": 9.585533142089844,
      "learning_rate": 8.86870709382151e-06,
      "loss": 0.7297,
      "step": 4925
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.81173324584961,
      "learning_rate": 8.864416475972541e-06,
      "loss": 0.7518,
      "step": 4926
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.336341857910156,
      "learning_rate": 8.86012585812357e-06,
      "loss": 0.7924,
      "step": 4927
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.460423469543457,
      "learning_rate": 8.8558352402746e-06,
      "loss": 0.7537,
      "step": 4928
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.195754051208496,
      "learning_rate": 8.851544622425629e-06,
      "loss": 0.7238,
      "step": 4929
    },
    {
      "epoch": 1.41,
      "grad_norm": 11.2444429397583,
      "learning_rate": 8.847254004576658e-06,
      "loss": 0.7621,
      "step": 4930
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.065338134765625,
      "learning_rate": 8.842963386727688e-06,
      "loss": 0.5536,
      "step": 4931
    },
    {
      "epoch": 1.41,
      "grad_norm": 11.432260513305664,
      "learning_rate": 8.838672768878719e-06,
      "loss": 0.8951,
      "step": 4932
    },
    {
      "epoch": 1.41,
      "grad_norm": 12.106633186340332,
      "learning_rate": 8.83438215102975e-06,
      "loss": 1.0926,
      "step": 4933
    },
    {
      "epoch": 1.41,
      "grad_norm": 9.704801559448242,
      "learning_rate": 8.830091533180778e-06,
      "loss": 0.8252,
      "step": 4934
    },
    {
      "epoch": 1.41,
      "grad_norm": 8.296677589416504,
      "learning_rate": 8.825800915331809e-06,
      "loss": 0.4684,
      "step": 4935
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.15492057800293,
      "learning_rate": 8.821510297482837e-06,
      "loss": 0.6786,
      "step": 4936
    },
    {
      "epoch": 1.41,
      "grad_norm": 8.982447624206543,
      "learning_rate": 8.817219679633868e-06,
      "loss": 0.6677,
      "step": 4937
    },
    {
      "epoch": 1.41,
      "grad_norm": 8.942950248718262,
      "learning_rate": 8.812929061784897e-06,
      "loss": 0.6398,
      "step": 4938
    },
    {
      "epoch": 1.41,
      "grad_norm": 8.501655578613281,
      "learning_rate": 8.808638443935927e-06,
      "loss": 0.6015,
      "step": 4939
    },
    {
      "epoch": 1.41,
      "grad_norm": 9.412524223327637,
      "learning_rate": 8.804347826086957e-06,
      "loss": 0.8239,
      "step": 4940
    },
    {
      "epoch": 1.41,
      "grad_norm": 8.406737327575684,
      "learning_rate": 8.800057208237988e-06,
      "loss": 0.5314,
      "step": 4941
    },
    {
      "epoch": 1.41,
      "grad_norm": 9.219732284545898,
      "learning_rate": 8.795766590389017e-06,
      "loss": 0.6431,
      "step": 4942
    },
    {
      "epoch": 1.41,
      "grad_norm": 8.9622220993042,
      "learning_rate": 8.791475972540045e-06,
      "loss": 0.6074,
      "step": 4943
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.373034477233887,
      "learning_rate": 8.787185354691076e-06,
      "loss": 0.7059,
      "step": 4944
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.128436088562012,
      "learning_rate": 8.782894736842105e-06,
      "loss": 0.6946,
      "step": 4945
    },
    {
      "epoch": 1.41,
      "grad_norm": 13.03666877746582,
      "learning_rate": 8.778604118993135e-06,
      "loss": 0.7183,
      "step": 4946
    },
    {
      "epoch": 1.42,
      "grad_norm": 10.098986625671387,
      "learning_rate": 8.774313501144164e-06,
      "loss": 0.7109,
      "step": 4947
    },
    {
      "epoch": 1.42,
      "grad_norm": 10.848021507263184,
      "learning_rate": 8.770022883295194e-06,
      "loss": 0.6452,
      "step": 4948
    },
    {
      "epoch": 1.42,
      "grad_norm": 8.350137710571289,
      "learning_rate": 8.765732265446225e-06,
      "loss": 0.6083,
      "step": 4949
    },
    {
      "epoch": 1.42,
      "grad_norm": 9.368001937866211,
      "learning_rate": 8.761441647597255e-06,
      "loss": 0.7013,
      "step": 4950
    },
    {
      "epoch": 1.42,
      "grad_norm": 11.08501148223877,
      "learning_rate": 8.757151029748284e-06,
      "loss": 0.7593,
      "step": 4951
    },
    {
      "epoch": 1.42,
      "grad_norm": 12.226153373718262,
      "learning_rate": 8.752860411899315e-06,
      "loss": 0.6093,
      "step": 4952
    },
    {
      "epoch": 1.42,
      "grad_norm": 10.081727981567383,
      "learning_rate": 8.748569794050343e-06,
      "loss": 0.798,
      "step": 4953
    },
    {
      "epoch": 1.42,
      "grad_norm": 7.829829692840576,
      "learning_rate": 8.744279176201374e-06,
      "loss": 0.4565,
      "step": 4954
    },
    {
      "epoch": 1.42,
      "grad_norm": 11.475992202758789,
      "learning_rate": 8.739988558352403e-06,
      "loss": 0.6535,
      "step": 4955
    },
    {
      "epoch": 1.42,
      "grad_norm": 11.957167625427246,
      "learning_rate": 8.735697940503431e-06,
      "loss": 0.9409,
      "step": 4956
    },
    {
      "epoch": 1.42,
      "grad_norm": 8.489669799804688,
      "learning_rate": 8.731407322654463e-06,
      "loss": 0.7229,
      "step": 4957
    },
    {
      "epoch": 1.42,
      "grad_norm": 12.282538414001465,
      "learning_rate": 8.727116704805492e-06,
      "loss": 0.7623,
      "step": 4958
    },
    {
      "epoch": 1.42,
      "grad_norm": 8.948768615722656,
      "learning_rate": 8.722826086956523e-06,
      "loss": 0.6361,
      "step": 4959
    },
    {
      "epoch": 1.42,
      "grad_norm": 10.072561264038086,
      "learning_rate": 8.718535469107551e-06,
      "loss": 0.7876,
      "step": 4960
    },
    {
      "epoch": 1.42,
      "grad_norm": 9.630600929260254,
      "learning_rate": 8.714244851258582e-06,
      "loss": 0.5635,
      "step": 4961
    },
    {
      "epoch": 1.42,
      "grad_norm": 9.96124267578125,
      "learning_rate": 8.70995423340961e-06,
      "loss": 0.8819,
      "step": 4962
    },
    {
      "epoch": 1.42,
      "grad_norm": 10.096244812011719,
      "learning_rate": 8.705663615560641e-06,
      "loss": 0.8045,
      "step": 4963
    },
    {
      "epoch": 1.42,
      "grad_norm": 8.456628799438477,
      "learning_rate": 8.70137299771167e-06,
      "loss": 0.546,
      "step": 4964
    },
    {
      "epoch": 1.42,
      "grad_norm": 11.215423583984375,
      "learning_rate": 8.6970823798627e-06,
      "loss": 0.7391,
      "step": 4965
    },
    {
      "epoch": 1.42,
      "grad_norm": 10.309953689575195,
      "learning_rate": 8.69279176201373e-06,
      "loss": 0.7174,
      "step": 4966
    },
    {
      "epoch": 1.42,
      "grad_norm": 12.189631462097168,
      "learning_rate": 8.688501144164761e-06,
      "loss": 1.013,
      "step": 4967
    },
    {
      "epoch": 1.42,
      "grad_norm": 10.816985130310059,
      "learning_rate": 8.68421052631579e-06,
      "loss": 0.8628,
      "step": 4968
    },
    {
      "epoch": 1.42,
      "grad_norm": 9.234234809875488,
      "learning_rate": 8.679919908466819e-06,
      "loss": 0.8081,
      "step": 4969
    },
    {
      "epoch": 1.42,
      "grad_norm": 11.34066390991211,
      "learning_rate": 8.67562929061785e-06,
      "loss": 0.8229,
      "step": 4970
    },
    {
      "epoch": 1.42,
      "grad_norm": 10.21949291229248,
      "learning_rate": 8.671338672768878e-06,
      "loss": 0.5068,
      "step": 4971
    },
    {
      "epoch": 1.42,
      "grad_norm": 10.102566719055176,
      "learning_rate": 8.667048054919909e-06,
      "loss": 0.7015,
      "step": 4972
    },
    {
      "epoch": 1.42,
      "grad_norm": 8.445645332336426,
      "learning_rate": 8.662757437070937e-06,
      "loss": 0.6165,
      "step": 4973
    },
    {
      "epoch": 1.42,
      "grad_norm": 8.391802787780762,
      "learning_rate": 8.658466819221968e-06,
      "loss": 0.6852,
      "step": 4974
    },
    {
      "epoch": 1.42,
      "grad_norm": 10.527009963989258,
      "learning_rate": 8.654176201372998e-06,
      "loss": 0.6214,
      "step": 4975
    },
    {
      "epoch": 1.42,
      "grad_norm": 9.231841087341309,
      "learning_rate": 8.649885583524029e-06,
      "loss": 0.7137,
      "step": 4976
    },
    {
      "epoch": 1.42,
      "grad_norm": 9.761481285095215,
      "learning_rate": 8.645594965675057e-06,
      "loss": 0.8636,
      "step": 4977
    },
    {
      "epoch": 1.42,
      "grad_norm": 8.89303970336914,
      "learning_rate": 8.641304347826088e-06,
      "loss": 0.5767,
      "step": 4978
    },
    {
      "epoch": 1.42,
      "grad_norm": 12.108701705932617,
      "learning_rate": 8.637013729977117e-06,
      "loss": 0.6581,
      "step": 4979
    },
    {
      "epoch": 1.42,
      "grad_norm": 9.690510749816895,
      "learning_rate": 8.632723112128147e-06,
      "loss": 0.6257,
      "step": 4980
    },
    {
      "epoch": 1.42,
      "grad_norm": 11.89633560180664,
      "learning_rate": 8.628432494279176e-06,
      "loss": 0.8569,
      "step": 4981
    },
    {
      "epoch": 1.43,
      "grad_norm": 11.14169979095459,
      "learning_rate": 8.624141876430205e-06,
      "loss": 0.7619,
      "step": 4982
    },
    {
      "epoch": 1.43,
      "grad_norm": 10.043008804321289,
      "learning_rate": 8.619851258581237e-06,
      "loss": 0.6162,
      "step": 4983
    },
    {
      "epoch": 1.43,
      "grad_norm": 8.896636009216309,
      "learning_rate": 8.615560640732266e-06,
      "loss": 0.5833,
      "step": 4984
    },
    {
      "epoch": 1.43,
      "grad_norm": 9.469889640808105,
      "learning_rate": 8.611270022883296e-06,
      "loss": 0.8342,
      "step": 4985
    },
    {
      "epoch": 1.43,
      "grad_norm": 11.050342559814453,
      "learning_rate": 8.606979405034325e-06,
      "loss": 0.6791,
      "step": 4986
    },
    {
      "epoch": 1.43,
      "grad_norm": 11.315376281738281,
      "learning_rate": 8.602688787185355e-06,
      "loss": 0.7415,
      "step": 4987
    },
    {
      "epoch": 1.43,
      "grad_norm": 11.068825721740723,
      "learning_rate": 8.598398169336384e-06,
      "loss": 0.6817,
      "step": 4988
    },
    {
      "epoch": 1.43,
      "grad_norm": 8.905261993408203,
      "learning_rate": 8.594107551487415e-06,
      "loss": 0.4884,
      "step": 4989
    },
    {
      "epoch": 1.43,
      "grad_norm": 11.311631202697754,
      "learning_rate": 8.589816933638443e-06,
      "loss": 0.8217,
      "step": 4990
    },
    {
      "epoch": 1.43,
      "grad_norm": 10.067758560180664,
      "learning_rate": 8.585526315789474e-06,
      "loss": 0.7261,
      "step": 4991
    },
    {
      "epoch": 1.43,
      "grad_norm": 11.24259090423584,
      "learning_rate": 8.581235697940504e-06,
      "loss": 0.7978,
      "step": 4992
    },
    {
      "epoch": 1.43,
      "grad_norm": 10.914911270141602,
      "learning_rate": 8.576945080091535e-06,
      "loss": 0.8629,
      "step": 4993
    },
    {
      "epoch": 1.43,
      "grad_norm": 10.073575019836426,
      "learning_rate": 8.572654462242563e-06,
      "loss": 0.7815,
      "step": 4994
    },
    {
      "epoch": 1.43,
      "grad_norm": 9.616877555847168,
      "learning_rate": 8.568363844393592e-06,
      "loss": 0.639,
      "step": 4995
    },
    {
      "epoch": 1.43,
      "grad_norm": 11.003323554992676,
      "learning_rate": 8.564073226544623e-06,
      "loss": 0.5851,
      "step": 4996
    },
    {
      "epoch": 1.43,
      "grad_norm": 10.080148696899414,
      "learning_rate": 8.559782608695651e-06,
      "loss": 0.6707,
      "step": 4997
    },
    {
      "epoch": 1.43,
      "grad_norm": 8.979016304016113,
      "learning_rate": 8.555491990846682e-06,
      "loss": 0.7034,
      "step": 4998
    },
    {
      "epoch": 1.43,
      "grad_norm": 12.307140350341797,
      "learning_rate": 8.55120137299771e-06,
      "loss": 0.9493,
      "step": 4999
    },
    {
      "epoch": 1.43,
      "grad_norm": 10.767085075378418,
      "learning_rate": 8.546910755148743e-06,
      "loss": 0.6093,
      "step": 5000
    },
    {
      "epoch": 1.43,
      "grad_norm": 10.213473320007324,
      "learning_rate": 8.542620137299772e-06,
      "loss": 0.6934,
      "step": 5001
    },
    {
      "epoch": 1.43,
      "grad_norm": 9.938896179199219,
      "learning_rate": 8.538329519450802e-06,
      "loss": 0.7864,
      "step": 5002
    },
    {
      "epoch": 1.43,
      "grad_norm": 9.872451782226562,
      "learning_rate": 8.53403890160183e-06,
      "loss": 0.7673,
      "step": 5003
    },
    {
      "epoch": 1.43,
      "grad_norm": 10.641152381896973,
      "learning_rate": 8.529748283752861e-06,
      "loss": 1.0319,
      "step": 5004
    },
    {
      "epoch": 1.43,
      "grad_norm": 10.038233757019043,
      "learning_rate": 8.52545766590389e-06,
      "loss": 0.6311,
      "step": 5005
    },
    {
      "epoch": 1.43,
      "grad_norm": 9.689718246459961,
      "learning_rate": 8.52116704805492e-06,
      "loss": 0.8566,
      "step": 5006
    },
    {
      "epoch": 1.43,
      "grad_norm": 9.882425308227539,
      "learning_rate": 8.51687643020595e-06,
      "loss": 0.5638,
      "step": 5007
    },
    {
      "epoch": 1.43,
      "grad_norm": 9.150846481323242,
      "learning_rate": 8.512585812356978e-06,
      "loss": 0.5935,
      "step": 5008
    },
    {
      "epoch": 1.43,
      "grad_norm": 10.776080131530762,
      "learning_rate": 8.50829519450801e-06,
      "loss": 0.8544,
      "step": 5009
    },
    {
      "epoch": 1.43,
      "grad_norm": 10.461644172668457,
      "learning_rate": 8.504004576659039e-06,
      "loss": 0.832,
      "step": 5010
    },
    {
      "epoch": 1.43,
      "grad_norm": 11.013406753540039,
      "learning_rate": 8.49971395881007e-06,
      "loss": 0.7076,
      "step": 5011
    },
    {
      "epoch": 1.43,
      "grad_norm": 8.372725486755371,
      "learning_rate": 8.495423340961098e-06,
      "loss": 0.6376,
      "step": 5012
    },
    {
      "epoch": 1.43,
      "grad_norm": 13.2879638671875,
      "learning_rate": 8.491132723112129e-06,
      "loss": 0.7582,
      "step": 5013
    },
    {
      "epoch": 1.43,
      "grad_norm": 11.984237670898438,
      "learning_rate": 8.486842105263157e-06,
      "loss": 0.8337,
      "step": 5014
    },
    {
      "epoch": 1.43,
      "grad_norm": 8.97717571258545,
      "learning_rate": 8.482551487414188e-06,
      "loss": 0.7792,
      "step": 5015
    },
    {
      "epoch": 1.43,
      "grad_norm": 11.499009132385254,
      "learning_rate": 8.478260869565217e-06,
      "loss": 0.8281,
      "step": 5016
    },
    {
      "epoch": 1.44,
      "grad_norm": 8.801410675048828,
      "learning_rate": 8.473970251716249e-06,
      "loss": 0.7705,
      "step": 5017
    },
    {
      "epoch": 1.44,
      "grad_norm": 9.285280227661133,
      "learning_rate": 8.469679633867278e-06,
      "loss": 0.557,
      "step": 5018
    },
    {
      "epoch": 1.44,
      "grad_norm": 8.304474830627441,
      "learning_rate": 8.465389016018308e-06,
      "loss": 0.5849,
      "step": 5019
    },
    {
      "epoch": 1.44,
      "grad_norm": 9.440895080566406,
      "learning_rate": 8.461098398169337e-06,
      "loss": 0.5966,
      "step": 5020
    },
    {
      "epoch": 1.44,
      "grad_norm": 16.106412887573242,
      "learning_rate": 8.456807780320366e-06,
      "loss": 0.8179,
      "step": 5021
    },
    {
      "epoch": 1.44,
      "grad_norm": 11.817154884338379,
      "learning_rate": 8.452517162471396e-06,
      "loss": 0.7659,
      "step": 5022
    },
    {
      "epoch": 1.44,
      "grad_norm": 11.135847091674805,
      "learning_rate": 8.448226544622425e-06,
      "loss": 0.762,
      "step": 5023
    },
    {
      "epoch": 1.44,
      "grad_norm": 11.009657859802246,
      "learning_rate": 8.443935926773455e-06,
      "loss": 0.8869,
      "step": 5024
    },
    {
      "epoch": 1.44,
      "grad_norm": 9.861078262329102,
      "learning_rate": 8.439645308924484e-06,
      "loss": 0.6355,
      "step": 5025
    },
    {
      "epoch": 1.44,
      "grad_norm": 8.525294303894043,
      "learning_rate": 8.435354691075516e-06,
      "loss": 0.6068,
      "step": 5026
    },
    {
      "epoch": 1.44,
      "grad_norm": 15.591021537780762,
      "learning_rate": 8.431064073226545e-06,
      "loss": 1.1226,
      "step": 5027
    },
    {
      "epoch": 1.44,
      "grad_norm": 8.61467456817627,
      "learning_rate": 8.426773455377575e-06,
      "loss": 0.4673,
      "step": 5028
    },
    {
      "epoch": 1.44,
      "grad_norm": 11.177366256713867,
      "learning_rate": 8.422482837528604e-06,
      "loss": 0.7149,
      "step": 5029
    },
    {
      "epoch": 1.44,
      "grad_norm": 10.839001655578613,
      "learning_rate": 8.418192219679635e-06,
      "loss": 0.7335,
      "step": 5030
    },
    {
      "epoch": 1.44,
      "grad_norm": 10.47923755645752,
      "learning_rate": 8.413901601830663e-06,
      "loss": 0.77,
      "step": 5031
    },
    {
      "epoch": 1.44,
      "grad_norm": 10.473682403564453,
      "learning_rate": 8.409610983981694e-06,
      "loss": 0.5729,
      "step": 5032
    },
    {
      "epoch": 1.44,
      "grad_norm": 13.123830795288086,
      "learning_rate": 8.405320366132723e-06,
      "loss": 0.6412,
      "step": 5033
    },
    {
      "epoch": 1.44,
      "grad_norm": 13.672239303588867,
      "learning_rate": 8.401029748283751e-06,
      "loss": 0.8529,
      "step": 5034
    },
    {
      "epoch": 1.44,
      "grad_norm": 9.248321533203125,
      "learning_rate": 8.396739130434784e-06,
      "loss": 0.5481,
      "step": 5035
    },
    {
      "epoch": 1.44,
      "grad_norm": 11.924641609191895,
      "learning_rate": 8.392448512585812e-06,
      "loss": 0.9527,
      "step": 5036
    },
    {
      "epoch": 1.44,
      "grad_norm": 8.101284980773926,
      "learning_rate": 8.388157894736843e-06,
      "loss": 0.603,
      "step": 5037
    },
    {
      "epoch": 1.44,
      "grad_norm": 9.7366361618042,
      "learning_rate": 8.383867276887872e-06,
      "loss": 0.677,
      "step": 5038
    },
    {
      "epoch": 1.44,
      "grad_norm": 7.6654229164123535,
      "learning_rate": 8.379576659038902e-06,
      "loss": 0.68,
      "step": 5039
    },
    {
      "epoch": 1.44,
      "grad_norm": 12.373343467712402,
      "learning_rate": 8.37528604118993e-06,
      "loss": 0.9479,
      "step": 5040
    },
    {
      "epoch": 1.44,
      "grad_norm": 10.893401145935059,
      "learning_rate": 8.370995423340961e-06,
      "loss": 0.8434,
      "step": 5041
    },
    {
      "epoch": 1.44,
      "grad_norm": 10.966062545776367,
      "learning_rate": 8.36670480549199e-06,
      "loss": 0.5505,
      "step": 5042
    },
    {
      "epoch": 1.44,
      "grad_norm": 11.359344482421875,
      "learning_rate": 8.362414187643022e-06,
      "loss": 0.6423,
      "step": 5043
    },
    {
      "epoch": 1.44,
      "grad_norm": 10.55596923828125,
      "learning_rate": 8.358123569794051e-06,
      "loss": 0.7672,
      "step": 5044
    },
    {
      "epoch": 1.44,
      "grad_norm": 11.82313346862793,
      "learning_rate": 8.353832951945081e-06,
      "loss": 0.7701,
      "step": 5045
    },
    {
      "epoch": 1.44,
      "grad_norm": 9.508740425109863,
      "learning_rate": 8.34954233409611e-06,
      "loss": 0.6773,
      "step": 5046
    },
    {
      "epoch": 1.44,
      "grad_norm": 11.016886711120605,
      "learning_rate": 8.345251716247139e-06,
      "loss": 0.8646,
      "step": 5047
    },
    {
      "epoch": 1.44,
      "grad_norm": 10.397418022155762,
      "learning_rate": 8.34096109839817e-06,
      "loss": 0.684,
      "step": 5048
    },
    {
      "epoch": 1.44,
      "grad_norm": 10.650434494018555,
      "learning_rate": 8.336670480549198e-06,
      "loss": 0.8008,
      "step": 5049
    },
    {
      "epoch": 1.44,
      "grad_norm": 9.770722389221191,
      "learning_rate": 8.332379862700229e-06,
      "loss": 0.7797,
      "step": 5050
    },
    {
      "epoch": 1.44,
      "grad_norm": 9.314210891723633,
      "learning_rate": 8.328089244851257e-06,
      "loss": 0.7631,
      "step": 5051
    },
    {
      "epoch": 1.45,
      "grad_norm": 11.673943519592285,
      "learning_rate": 8.32379862700229e-06,
      "loss": 0.8589,
      "step": 5052
    },
    {
      "epoch": 1.45,
      "grad_norm": 12.262770652770996,
      "learning_rate": 8.319508009153318e-06,
      "loss": 0.8789,
      "step": 5053
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.737245559692383,
      "learning_rate": 8.315217391304349e-06,
      "loss": 0.8933,
      "step": 5054
    },
    {
      "epoch": 1.45,
      "grad_norm": 12.311223030090332,
      "learning_rate": 8.310926773455378e-06,
      "loss": 0.5966,
      "step": 5055
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.886199951171875,
      "learning_rate": 8.306636155606408e-06,
      "loss": 0.7343,
      "step": 5056
    },
    {
      "epoch": 1.45,
      "grad_norm": 12.238615036010742,
      "learning_rate": 8.302345537757437e-06,
      "loss": 0.8853,
      "step": 5057
    },
    {
      "epoch": 1.45,
      "grad_norm": 9.038883209228516,
      "learning_rate": 8.298054919908467e-06,
      "loss": 0.7911,
      "step": 5058
    },
    {
      "epoch": 1.45,
      "grad_norm": 7.761969089508057,
      "learning_rate": 8.293764302059496e-06,
      "loss": 0.6182,
      "step": 5059
    },
    {
      "epoch": 1.45,
      "grad_norm": 9.099241256713867,
      "learning_rate": 8.289473684210526e-06,
      "loss": 0.4603,
      "step": 5060
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.834635734558105,
      "learning_rate": 8.285183066361557e-06,
      "loss": 0.7656,
      "step": 5061
    },
    {
      "epoch": 1.45,
      "grad_norm": 13.24924373626709,
      "learning_rate": 8.280892448512586e-06,
      "loss": 0.786,
      "step": 5062
    },
    {
      "epoch": 1.45,
      "grad_norm": 13.369250297546387,
      "learning_rate": 8.276601830663616e-06,
      "loss": 1.0903,
      "step": 5063
    },
    {
      "epoch": 1.45,
      "grad_norm": 13.29235553741455,
      "learning_rate": 8.272311212814645e-06,
      "loss": 0.793,
      "step": 5064
    },
    {
      "epoch": 1.45,
      "grad_norm": 9.594003677368164,
      "learning_rate": 8.268020594965675e-06,
      "loss": 0.5489,
      "step": 5065
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.180651664733887,
      "learning_rate": 8.263729977116704e-06,
      "loss": 0.5971,
      "step": 5066
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.12555980682373,
      "learning_rate": 8.259439359267735e-06,
      "loss": 0.6838,
      "step": 5067
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.411796569824219,
      "learning_rate": 8.255148741418763e-06,
      "loss": 0.726,
      "step": 5068
    },
    {
      "epoch": 1.45,
      "grad_norm": 11.773266792297363,
      "learning_rate": 8.250858123569796e-06,
      "loss": 1.0189,
      "step": 5069
    },
    {
      "epoch": 1.45,
      "grad_norm": 9.530159950256348,
      "learning_rate": 8.246567505720824e-06,
      "loss": 0.6832,
      "step": 5070
    },
    {
      "epoch": 1.45,
      "grad_norm": 9.325148582458496,
      "learning_rate": 8.242276887871855e-06,
      "loss": 0.7125,
      "step": 5071
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.458324432373047,
      "learning_rate": 8.237986270022884e-06,
      "loss": 0.6311,
      "step": 5072
    },
    {
      "epoch": 1.45,
      "grad_norm": 8.303791999816895,
      "learning_rate": 8.233695652173914e-06,
      "loss": 0.6966,
      "step": 5073
    },
    {
      "epoch": 1.45,
      "grad_norm": 12.064472198486328,
      "learning_rate": 8.229405034324943e-06,
      "loss": 0.7607,
      "step": 5074
    },
    {
      "epoch": 1.45,
      "grad_norm": 8.434898376464844,
      "learning_rate": 8.225114416475972e-06,
      "loss": 0.6423,
      "step": 5075
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.684408187866211,
      "learning_rate": 8.220823798627002e-06,
      "loss": 0.7354,
      "step": 5076
    },
    {
      "epoch": 1.45,
      "grad_norm": 9.874177932739258,
      "learning_rate": 8.216533180778032e-06,
      "loss": 0.6514,
      "step": 5077
    },
    {
      "epoch": 1.45,
      "grad_norm": 12.246417045593262,
      "learning_rate": 8.212242562929063e-06,
      "loss": 0.8711,
      "step": 5078
    },
    {
      "epoch": 1.45,
      "grad_norm": 12.1536283493042,
      "learning_rate": 8.207951945080092e-06,
      "loss": 0.9474,
      "step": 5079
    },
    {
      "epoch": 1.45,
      "grad_norm": 6.930532932281494,
      "learning_rate": 8.203661327231122e-06,
      "loss": 0.3914,
      "step": 5080
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.783819198608398,
      "learning_rate": 8.199370709382151e-06,
      "loss": 0.8458,
      "step": 5081
    },
    {
      "epoch": 1.45,
      "grad_norm": 8.25786304473877,
      "learning_rate": 8.195080091533181e-06,
      "loss": 0.5712,
      "step": 5082
    },
    {
      "epoch": 1.45,
      "grad_norm": 9.214256286621094,
      "learning_rate": 8.19078947368421e-06,
      "loss": 0.7327,
      "step": 5083
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.542569160461426,
      "learning_rate": 8.18649885583524e-06,
      "loss": 0.7113,
      "step": 5084
    },
    {
      "epoch": 1.45,
      "grad_norm": 9.573670387268066,
      "learning_rate": 8.18220823798627e-06,
      "loss": 0.6155,
      "step": 5085
    },
    {
      "epoch": 1.45,
      "grad_norm": 9.836844444274902,
      "learning_rate": 8.177917620137302e-06,
      "loss": 0.6262,
      "step": 5086
    },
    {
      "epoch": 1.46,
      "grad_norm": 11.307541847229004,
      "learning_rate": 8.17362700228833e-06,
      "loss": 0.8254,
      "step": 5087
    },
    {
      "epoch": 1.46,
      "grad_norm": 6.5441107749938965,
      "learning_rate": 8.169336384439359e-06,
      "loss": 0.3907,
      "step": 5088
    },
    {
      "epoch": 1.46,
      "grad_norm": 9.281217575073242,
      "learning_rate": 8.16504576659039e-06,
      "loss": 0.5852,
      "step": 5089
    },
    {
      "epoch": 1.46,
      "grad_norm": 12.437320709228516,
      "learning_rate": 8.160755148741418e-06,
      "loss": 0.7752,
      "step": 5090
    },
    {
      "epoch": 1.46,
      "grad_norm": 9.643728256225586,
      "learning_rate": 8.156464530892449e-06,
      "loss": 0.5487,
      "step": 5091
    },
    {
      "epoch": 1.46,
      "grad_norm": 10.50843620300293,
      "learning_rate": 8.152173913043478e-06,
      "loss": 0.5361,
      "step": 5092
    },
    {
      "epoch": 1.46,
      "grad_norm": 9.972748756408691,
      "learning_rate": 8.147883295194508e-06,
      "loss": 0.6879,
      "step": 5093
    },
    {
      "epoch": 1.46,
      "grad_norm": 10.782821655273438,
      "learning_rate": 8.143592677345537e-06,
      "loss": 0.6245,
      "step": 5094
    },
    {
      "epoch": 1.46,
      "grad_norm": 7.436230182647705,
      "learning_rate": 8.139302059496569e-06,
      "loss": 0.4944,
      "step": 5095
    },
    {
      "epoch": 1.46,
      "grad_norm": 10.277985572814941,
      "learning_rate": 8.135011441647598e-06,
      "loss": 0.7533,
      "step": 5096
    },
    {
      "epoch": 1.46,
      "grad_norm": 9.366806030273438,
      "learning_rate": 8.130720823798628e-06,
      "loss": 0.666,
      "step": 5097
    },
    {
      "epoch": 1.46,
      "grad_norm": 13.114858627319336,
      "learning_rate": 8.126430205949657e-06,
      "loss": 0.7336,
      "step": 5098
    },
    {
      "epoch": 1.46,
      "grad_norm": 12.340123176574707,
      "learning_rate": 8.122139588100687e-06,
      "loss": 0.7865,
      "step": 5099
    },
    {
      "epoch": 1.46,
      "grad_norm": 10.428196907043457,
      "learning_rate": 8.117848970251716e-06,
      "loss": 0.59,
      "step": 5100
    },
    {
      "epoch": 1.46,
      "grad_norm": 16.39594841003418,
      "learning_rate": 8.113558352402745e-06,
      "loss": 0.8344,
      "step": 5101
    },
    {
      "epoch": 1.46,
      "grad_norm": 9.768933296203613,
      "learning_rate": 8.109267734553775e-06,
      "loss": 0.7913,
      "step": 5102
    },
    {
      "epoch": 1.46,
      "grad_norm": 11.725458145141602,
      "learning_rate": 8.104977116704806e-06,
      "loss": 0.585,
      "step": 5103
    },
    {
      "epoch": 1.46,
      "grad_norm": 12.467772483825684,
      "learning_rate": 8.100686498855836e-06,
      "loss": 0.8799,
      "step": 5104
    },
    {
      "epoch": 1.46,
      "grad_norm": 10.282450675964355,
      "learning_rate": 8.096395881006865e-06,
      "loss": 0.7503,
      "step": 5105
    },
    {
      "epoch": 1.46,
      "grad_norm": 12.790483474731445,
      "learning_rate": 8.092105263157896e-06,
      "loss": 1.0285,
      "step": 5106
    },
    {
      "epoch": 1.46,
      "grad_norm": 9.784635543823242,
      "learning_rate": 8.087814645308924e-06,
      "loss": 0.5898,
      "step": 5107
    },
    {
      "epoch": 1.46,
      "grad_norm": 10.346814155578613,
      "learning_rate": 8.083524027459955e-06,
      "loss": 0.7253,
      "step": 5108
    },
    {
      "epoch": 1.46,
      "grad_norm": 12.714737892150879,
      "learning_rate": 8.079233409610984e-06,
      "loss": 0.8393,
      "step": 5109
    },
    {
      "epoch": 1.46,
      "grad_norm": 11.371176719665527,
      "learning_rate": 8.074942791762014e-06,
      "loss": 0.8245,
      "step": 5110
    },
    {
      "epoch": 1.46,
      "grad_norm": 10.148053169250488,
      "learning_rate": 8.070652173913043e-06,
      "loss": 0.6768,
      "step": 5111
    },
    {
      "epoch": 1.46,
      "grad_norm": 11.980246543884277,
      "learning_rate": 8.066361556064075e-06,
      "loss": 0.8608,
      "step": 5112
    },
    {
      "epoch": 1.46,
      "grad_norm": 9.12360954284668,
      "learning_rate": 8.062070938215104e-06,
      "loss": 0.7646,
      "step": 5113
    },
    {
      "epoch": 1.46,
      "grad_norm": 11.374464988708496,
      "learning_rate": 8.057780320366132e-06,
      "loss": 0.8281,
      "step": 5114
    },
    {
      "epoch": 1.46,
      "grad_norm": 12.122367858886719,
      "learning_rate": 8.053489702517163e-06,
      "loss": 0.737,
      "step": 5115
    },
    {
      "epoch": 1.46,
      "grad_norm": 9.022582054138184,
      "learning_rate": 8.049199084668192e-06,
      "loss": 0.7077,
      "step": 5116
    },
    {
      "epoch": 1.46,
      "grad_norm": 7.594146251678467,
      "learning_rate": 8.044908466819222e-06,
      "loss": 0.5956,
      "step": 5117
    },
    {
      "epoch": 1.46,
      "grad_norm": 10.771967887878418,
      "learning_rate": 8.040617848970251e-06,
      "loss": 0.8169,
      "step": 5118
    },
    {
      "epoch": 1.46,
      "grad_norm": 12.040303230285645,
      "learning_rate": 8.036327231121281e-06,
      "loss": 0.7885,
      "step": 5119
    },
    {
      "epoch": 1.46,
      "grad_norm": 11.240238189697266,
      "learning_rate": 8.032036613272312e-06,
      "loss": 0.8766,
      "step": 5120
    },
    {
      "epoch": 1.46,
      "grad_norm": 9.21816635131836,
      "learning_rate": 8.027745995423342e-06,
      "loss": 0.6309,
      "step": 5121
    },
    {
      "epoch": 1.47,
      "grad_norm": 9.237491607666016,
      "learning_rate": 8.023455377574371e-06,
      "loss": 0.6262,
      "step": 5122
    },
    {
      "epoch": 1.47,
      "grad_norm": 12.814543724060059,
      "learning_rate": 8.019164759725402e-06,
      "loss": 0.862,
      "step": 5123
    },
    {
      "epoch": 1.47,
      "grad_norm": 10.67367172241211,
      "learning_rate": 8.01487414187643e-06,
      "loss": 1.0106,
      "step": 5124
    },
    {
      "epoch": 1.47,
      "grad_norm": 9.877201080322266,
      "learning_rate": 8.01058352402746e-06,
      "loss": 0.7026,
      "step": 5125
    },
    {
      "epoch": 1.47,
      "grad_norm": 11.837648391723633,
      "learning_rate": 8.00629290617849e-06,
      "loss": 0.9552,
      "step": 5126
    },
    {
      "epoch": 1.47,
      "grad_norm": 12.015484809875488,
      "learning_rate": 8.002002288329518e-06,
      "loss": 0.6225,
      "step": 5127
    },
    {
      "epoch": 1.47,
      "grad_norm": 8.69874095916748,
      "learning_rate": 7.997711670480549e-06,
      "loss": 0.6455,
      "step": 5128
    },
    {
      "epoch": 1.47,
      "grad_norm": 10.47046184539795,
      "learning_rate": 7.99342105263158e-06,
      "loss": 0.7297,
      "step": 5129
    },
    {
      "epoch": 1.47,
      "grad_norm": 10.689979553222656,
      "learning_rate": 7.98913043478261e-06,
      "loss": 0.8392,
      "step": 5130
    },
    {
      "epoch": 1.47,
      "grad_norm": 9.730806350708008,
      "learning_rate": 7.984839816933638e-06,
      "loss": 0.4623,
      "step": 5131
    },
    {
      "epoch": 1.47,
      "grad_norm": 12.895252227783203,
      "learning_rate": 7.980549199084669e-06,
      "loss": 0.9518,
      "step": 5132
    },
    {
      "epoch": 1.47,
      "grad_norm": 9.70818042755127,
      "learning_rate": 7.976258581235698e-06,
      "loss": 0.8398,
      "step": 5133
    },
    {
      "epoch": 1.47,
      "grad_norm": 10.01799488067627,
      "learning_rate": 7.971967963386728e-06,
      "loss": 0.8637,
      "step": 5134
    },
    {
      "epoch": 1.47,
      "grad_norm": 10.447866439819336,
      "learning_rate": 7.967677345537757e-06,
      "loss": 0.6831,
      "step": 5135
    },
    {
      "epoch": 1.47,
      "grad_norm": 11.744251251220703,
      "learning_rate": 7.963386727688787e-06,
      "loss": 0.9694,
      "step": 5136
    },
    {
      "epoch": 1.47,
      "grad_norm": 11.017406463623047,
      "learning_rate": 7.959096109839818e-06,
      "loss": 0.5433,
      "step": 5137
    },
    {
      "epoch": 1.47,
      "grad_norm": 8.464219093322754,
      "learning_rate": 7.954805491990848e-06,
      "loss": 0.5909,
      "step": 5138
    },
    {
      "epoch": 1.47,
      "grad_norm": 11.348611831665039,
      "learning_rate": 7.950514874141877e-06,
      "loss": 0.5935,
      "step": 5139
    },
    {
      "epoch": 1.47,
      "grad_norm": 10.337952613830566,
      "learning_rate": 7.946224256292906e-06,
      "loss": 0.5608,
      "step": 5140
    },
    {
      "epoch": 1.47,
      "grad_norm": 14.805388450622559,
      "learning_rate": 7.941933638443936e-06,
      "loss": 0.7561,
      "step": 5141
    },
    {
      "epoch": 1.47,
      "grad_norm": 9.155672073364258,
      "learning_rate": 7.937643020594965e-06,
      "loss": 0.5781,
      "step": 5142
    },
    {
      "epoch": 1.47,
      "grad_norm": 9.741296768188477,
      "learning_rate": 7.933352402745996e-06,
      "loss": 0.6299,
      "step": 5143
    },
    {
      "epoch": 1.47,
      "grad_norm": 10.216283798217773,
      "learning_rate": 7.929061784897024e-06,
      "loss": 0.8591,
      "step": 5144
    },
    {
      "epoch": 1.47,
      "grad_norm": 9.9946928024292,
      "learning_rate": 7.924771167048055e-06,
      "loss": 0.6312,
      "step": 5145
    },
    {
      "epoch": 1.47,
      "grad_norm": 9.108634948730469,
      "learning_rate": 7.920480549199085e-06,
      "loss": 0.602,
      "step": 5146
    },
    {
      "epoch": 1.47,
      "grad_norm": 12.92469596862793,
      "learning_rate": 7.916189931350116e-06,
      "loss": 0.6813,
      "step": 5147
    },
    {
      "epoch": 1.47,
      "grad_norm": 9.77869701385498,
      "learning_rate": 7.911899313501144e-06,
      "loss": 0.7651,
      "step": 5148
    },
    {
      "epoch": 1.47,
      "grad_norm": 11.956109046936035,
      "learning_rate": 7.907608695652175e-06,
      "loss": 0.7607,
      "step": 5149
    },
    {
      "epoch": 1.47,
      "grad_norm": 10.98892879486084,
      "learning_rate": 7.903318077803204e-06,
      "loss": 0.5443,
      "step": 5150
    },
    {
      "epoch": 1.47,
      "grad_norm": 10.866950988769531,
      "learning_rate": 7.899027459954234e-06,
      "loss": 0.6851,
      "step": 5151
    },
    {
      "epoch": 1.47,
      "grad_norm": 13.141809463500977,
      "learning_rate": 7.894736842105263e-06,
      "loss": 0.8302,
      "step": 5152
    },
    {
      "epoch": 1.47,
      "grad_norm": 10.317466735839844,
      "learning_rate": 7.890446224256292e-06,
      "loss": 0.7601,
      "step": 5153
    },
    {
      "epoch": 1.47,
      "grad_norm": 10.443330764770508,
      "learning_rate": 7.886155606407322e-06,
      "loss": 0.8592,
      "step": 5154
    },
    {
      "epoch": 1.47,
      "grad_norm": 10.033818244934082,
      "learning_rate": 7.881864988558353e-06,
      "loss": 0.891,
      "step": 5155
    },
    {
      "epoch": 1.47,
      "grad_norm": 12.123933792114258,
      "learning_rate": 7.877574370709383e-06,
      "loss": 1.1241,
      "step": 5156
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.95120620727539,
      "learning_rate": 7.873283752860412e-06,
      "loss": 0.5987,
      "step": 5157
    },
    {
      "epoch": 1.48,
      "grad_norm": 9.693603515625,
      "learning_rate": 7.868993135011442e-06,
      "loss": 0.4539,
      "step": 5158
    },
    {
      "epoch": 1.48,
      "grad_norm": 9.845906257629395,
      "learning_rate": 7.864702517162471e-06,
      "loss": 0.7684,
      "step": 5159
    },
    {
      "epoch": 1.48,
      "grad_norm": 9.617039680480957,
      "learning_rate": 7.860411899313502e-06,
      "loss": 0.9473,
      "step": 5160
    },
    {
      "epoch": 1.48,
      "grad_norm": 11.552712440490723,
      "learning_rate": 7.85612128146453e-06,
      "loss": 0.8603,
      "step": 5161
    },
    {
      "epoch": 1.48,
      "grad_norm": 9.689935684204102,
      "learning_rate": 7.85183066361556e-06,
      "loss": 0.5713,
      "step": 5162
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.119417190551758,
      "learning_rate": 7.847540045766591e-06,
      "loss": 0.5921,
      "step": 5163
    },
    {
      "epoch": 1.48,
      "grad_norm": 9.15572738647461,
      "learning_rate": 7.843249427917622e-06,
      "loss": 0.5496,
      "step": 5164
    },
    {
      "epoch": 1.48,
      "grad_norm": 15.126708984375,
      "learning_rate": 7.83895881006865e-06,
      "loss": 1.1661,
      "step": 5165
    },
    {
      "epoch": 1.48,
      "grad_norm": 12.176010131835938,
      "learning_rate": 7.83466819221968e-06,
      "loss": 0.8296,
      "step": 5166
    },
    {
      "epoch": 1.48,
      "grad_norm": 8.255621910095215,
      "learning_rate": 7.83037757437071e-06,
      "loss": 0.5731,
      "step": 5167
    },
    {
      "epoch": 1.48,
      "grad_norm": 11.428735733032227,
      "learning_rate": 7.826086956521738e-06,
      "loss": 0.7782,
      "step": 5168
    },
    {
      "epoch": 1.48,
      "grad_norm": 8.800541877746582,
      "learning_rate": 7.821796338672769e-06,
      "loss": 0.6463,
      "step": 5169
    },
    {
      "epoch": 1.48,
      "grad_norm": 12.368282318115234,
      "learning_rate": 7.817505720823798e-06,
      "loss": 0.6011,
      "step": 5170
    },
    {
      "epoch": 1.48,
      "grad_norm": 9.917195320129395,
      "learning_rate": 7.813215102974828e-06,
      "loss": 0.7234,
      "step": 5171
    },
    {
      "epoch": 1.48,
      "grad_norm": 9.662066459655762,
      "learning_rate": 7.808924485125859e-06,
      "loss": 0.7493,
      "step": 5172
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.145095825195312,
      "learning_rate": 7.804633867276889e-06,
      "loss": 0.8596,
      "step": 5173
    },
    {
      "epoch": 1.48,
      "grad_norm": 11.008843421936035,
      "learning_rate": 7.800343249427918e-06,
      "loss": 0.6049,
      "step": 5174
    },
    {
      "epoch": 1.48,
      "grad_norm": 11.558208465576172,
      "learning_rate": 7.796052631578948e-06,
      "loss": 0.7109,
      "step": 5175
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.150371551513672,
      "learning_rate": 7.791762013729977e-06,
      "loss": 0.8938,
      "step": 5176
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.215923309326172,
      "learning_rate": 7.787471395881008e-06,
      "loss": 0.5018,
      "step": 5177
    },
    {
      "epoch": 1.48,
      "grad_norm": 8.31438159942627,
      "learning_rate": 7.783180778032036e-06,
      "loss": 0.4783,
      "step": 5178
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.255351066589355,
      "learning_rate": 7.778890160183065e-06,
      "loss": 0.6412,
      "step": 5179
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.33067512512207,
      "learning_rate": 7.774599542334097e-06,
      "loss": 0.7935,
      "step": 5180
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.240641593933105,
      "learning_rate": 7.770308924485126e-06,
      "loss": 0.8504,
      "step": 5181
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.01538372039795,
      "learning_rate": 7.766018306636156e-06,
      "loss": 0.771,
      "step": 5182
    },
    {
      "epoch": 1.48,
      "grad_norm": 8.917338371276855,
      "learning_rate": 7.761727688787185e-06,
      "loss": 0.7849,
      "step": 5183
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.52344036102295,
      "learning_rate": 7.757437070938216e-06,
      "loss": 0.9558,
      "step": 5184
    },
    {
      "epoch": 1.48,
      "grad_norm": 8.942107200622559,
      "learning_rate": 7.753146453089244e-06,
      "loss": 0.7475,
      "step": 5185
    },
    {
      "epoch": 1.48,
      "grad_norm": 9.920065879821777,
      "learning_rate": 7.748855835240275e-06,
      "loss": 0.5781,
      "step": 5186
    },
    {
      "epoch": 1.48,
      "grad_norm": 9.799785614013672,
      "learning_rate": 7.744565217391304e-06,
      "loss": 0.7346,
      "step": 5187
    },
    {
      "epoch": 1.48,
      "grad_norm": 11.251009941101074,
      "learning_rate": 7.740274599542334e-06,
      "loss": 0.682,
      "step": 5188
    },
    {
      "epoch": 1.48,
      "grad_norm": 11.464249610900879,
      "learning_rate": 7.735983981693365e-06,
      "loss": 0.649,
      "step": 5189
    },
    {
      "epoch": 1.48,
      "grad_norm": 9.01133918762207,
      "learning_rate": 7.731693363844395e-06,
      "loss": 0.4474,
      "step": 5190
    },
    {
      "epoch": 1.48,
      "grad_norm": 9.98456859588623,
      "learning_rate": 7.727402745995424e-06,
      "loss": 0.5277,
      "step": 5191
    },
    {
      "epoch": 1.49,
      "grad_norm": 13.69474983215332,
      "learning_rate": 7.723112128146453e-06,
      "loss": 0.7136,
      "step": 5192
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.09440803527832,
      "learning_rate": 7.718821510297483e-06,
      "loss": 0.601,
      "step": 5193
    },
    {
      "epoch": 1.49,
      "grad_norm": 11.429454803466797,
      "learning_rate": 7.714530892448512e-06,
      "loss": 0.7693,
      "step": 5194
    },
    {
      "epoch": 1.49,
      "grad_norm": 12.033621788024902,
      "learning_rate": 7.710240274599542e-06,
      "loss": 0.905,
      "step": 5195
    },
    {
      "epoch": 1.49,
      "grad_norm": 8.883831024169922,
      "learning_rate": 7.705949656750571e-06,
      "loss": 0.7298,
      "step": 5196
    },
    {
      "epoch": 1.49,
      "grad_norm": 12.658049583435059,
      "learning_rate": 7.701659038901603e-06,
      "loss": 0.9474,
      "step": 5197
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.020696640014648,
      "learning_rate": 7.697368421052632e-06,
      "loss": 0.7189,
      "step": 5198
    },
    {
      "epoch": 1.49,
      "grad_norm": 9.961935043334961,
      "learning_rate": 7.693077803203662e-06,
      "loss": 0.6584,
      "step": 5199
    },
    {
      "epoch": 1.49,
      "grad_norm": 8.704461097717285,
      "learning_rate": 7.688787185354691e-06,
      "loss": 0.5271,
      "step": 5200
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.55123233795166,
      "learning_rate": 7.684496567505722e-06,
      "loss": 0.6096,
      "step": 5201
    },
    {
      "epoch": 1.49,
      "grad_norm": 13.534534454345703,
      "learning_rate": 7.68020594965675e-06,
      "loss": 0.7524,
      "step": 5202
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.84149169921875,
      "learning_rate": 7.675915331807781e-06,
      "loss": 0.7198,
      "step": 5203
    },
    {
      "epoch": 1.49,
      "grad_norm": 12.128814697265625,
      "learning_rate": 7.67162471395881e-06,
      "loss": 0.7491,
      "step": 5204
    },
    {
      "epoch": 1.49,
      "grad_norm": 12.675615310668945,
      "learning_rate": 7.667334096109838e-06,
      "loss": 1.0454,
      "step": 5205
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.708645820617676,
      "learning_rate": 7.66304347826087e-06,
      "loss": 0.7547,
      "step": 5206
    },
    {
      "epoch": 1.49,
      "grad_norm": 9.514225959777832,
      "learning_rate": 7.6587528604119e-06,
      "loss": 0.6761,
      "step": 5207
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.71254825592041,
      "learning_rate": 7.65446224256293e-06,
      "loss": 0.7642,
      "step": 5208
    },
    {
      "epoch": 1.49,
      "grad_norm": 8.257457733154297,
      "learning_rate": 7.650171624713959e-06,
      "loss": 0.6241,
      "step": 5209
    },
    {
      "epoch": 1.49,
      "grad_norm": 11.02060604095459,
      "learning_rate": 7.645881006864989e-06,
      "loss": 0.7695,
      "step": 5210
    },
    {
      "epoch": 1.49,
      "grad_norm": 7.717381477355957,
      "learning_rate": 7.641590389016018e-06,
      "loss": 0.6437,
      "step": 5211
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.31637954711914,
      "learning_rate": 7.637299771167048e-06,
      "loss": 0.9296,
      "step": 5212
    },
    {
      "epoch": 1.49,
      "grad_norm": 14.598742485046387,
      "learning_rate": 7.633009153318077e-06,
      "loss": 0.8934,
      "step": 5213
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.441706657409668,
      "learning_rate": 7.628718535469107e-06,
      "loss": 0.8108,
      "step": 5214
    },
    {
      "epoch": 1.49,
      "grad_norm": 19.70823860168457,
      "learning_rate": 7.624427917620138e-06,
      "loss": 0.7806,
      "step": 5215
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.273853302001953,
      "learning_rate": 7.620137299771168e-06,
      "loss": 0.6576,
      "step": 5216
    },
    {
      "epoch": 1.49,
      "grad_norm": 8.733698844909668,
      "learning_rate": 7.615846681922197e-06,
      "loss": 0.5898,
      "step": 5217
    },
    {
      "epoch": 1.49,
      "grad_norm": 12.016242027282715,
      "learning_rate": 7.611556064073227e-06,
      "loss": 0.6687,
      "step": 5218
    },
    {
      "epoch": 1.49,
      "grad_norm": 8.932428359985352,
      "learning_rate": 7.6072654462242564e-06,
      "loss": 0.8076,
      "step": 5219
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.967382431030273,
      "learning_rate": 7.602974828375286e-06,
      "loss": 0.8116,
      "step": 5220
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.108834266662598,
      "learning_rate": 7.598684210526316e-06,
      "loss": 0.5624,
      "step": 5221
    },
    {
      "epoch": 1.49,
      "grad_norm": 12.46695327758789,
      "learning_rate": 7.594393592677345e-06,
      "loss": 0.9826,
      "step": 5222
    },
    {
      "epoch": 1.49,
      "grad_norm": 12.058609962463379,
      "learning_rate": 7.590102974828377e-06,
      "loss": 0.6397,
      "step": 5223
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.250239372253418,
      "learning_rate": 7.585812356979406e-06,
      "loss": 0.5533,
      "step": 5224
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.57835578918457,
      "learning_rate": 7.581521739130435e-06,
      "loss": 0.5701,
      "step": 5225
    },
    {
      "epoch": 1.49,
      "grad_norm": 11.8040189743042,
      "learning_rate": 7.577231121281465e-06,
      "loss": 0.8717,
      "step": 5226
    },
    {
      "epoch": 1.5,
      "grad_norm": 9.846546173095703,
      "learning_rate": 7.572940503432494e-06,
      "loss": 0.7043,
      "step": 5227
    },
    {
      "epoch": 1.5,
      "grad_norm": 9.991266250610352,
      "learning_rate": 7.568649885583524e-06,
      "loss": 0.7032,
      "step": 5228
    },
    {
      "epoch": 1.5,
      "grad_norm": 10.166804313659668,
      "learning_rate": 7.5643592677345534e-06,
      "loss": 0.8109,
      "step": 5229
    },
    {
      "epoch": 1.5,
      "grad_norm": 9.002490997314453,
      "learning_rate": 7.560068649885583e-06,
      "loss": 0.4385,
      "step": 5230
    },
    {
      "epoch": 1.5,
      "grad_norm": 11.902188301086426,
      "learning_rate": 7.555778032036613e-06,
      "loss": 0.7558,
      "step": 5231
    },
    {
      "epoch": 1.5,
      "grad_norm": 12.483781814575195,
      "learning_rate": 7.551487414187644e-06,
      "loss": 0.6895,
      "step": 5232
    },
    {
      "epoch": 1.5,
      "grad_norm": 9.413661003112793,
      "learning_rate": 7.547196796338674e-06,
      "loss": 0.6075,
      "step": 5233
    },
    {
      "epoch": 1.5,
      "grad_norm": 9.486357688903809,
      "learning_rate": 7.542906178489703e-06,
      "loss": 0.5532,
      "step": 5234
    },
    {
      "epoch": 1.5,
      "grad_norm": 9.101693153381348,
      "learning_rate": 7.538615560640733e-06,
      "loss": 0.6248,
      "step": 5235
    },
    {
      "epoch": 1.5,
      "grad_norm": 11.137267112731934,
      "learning_rate": 7.5343249427917624e-06,
      "loss": 0.8225,
      "step": 5236
    },
    {
      "epoch": 1.5,
      "grad_norm": 11.395636558532715,
      "learning_rate": 7.530034324942792e-06,
      "loss": 0.6358,
      "step": 5237
    },
    {
      "epoch": 1.5,
      "grad_norm": 10.854764938354492,
      "learning_rate": 7.525743707093821e-06,
      "loss": 0.6775,
      "step": 5238
    },
    {
      "epoch": 1.5,
      "grad_norm": 14.450782775878906,
      "learning_rate": 7.52145308924485e-06,
      "loss": 0.865,
      "step": 5239
    },
    {
      "epoch": 1.5,
      "grad_norm": 10.17805290222168,
      "learning_rate": 7.517162471395882e-06,
      "loss": 0.5995,
      "step": 5240
    },
    {
      "epoch": 1.5,
      "grad_norm": 12.272109985351562,
      "learning_rate": 7.512871853546911e-06,
      "loss": 0.9173,
      "step": 5241
    },
    {
      "epoch": 1.5,
      "grad_norm": 11.942349433898926,
      "learning_rate": 7.508581235697941e-06,
      "loss": 0.8059,
      "step": 5242
    },
    {
      "epoch": 1.5,
      "grad_norm": 10.477256774902344,
      "learning_rate": 7.5042906178489706e-06,
      "loss": 0.6005,
      "step": 5243
    },
    {
      "epoch": 1.5,
      "grad_norm": 10.549176216125488,
      "learning_rate": 7.5e-06,
      "loss": 0.8263,
      "step": 5244
    },
    {
      "epoch": 1.5,
      "grad_norm": 9.667335510253906,
      "learning_rate": 7.49570938215103e-06,
      "loss": 0.6895,
      "step": 5245
    },
    {
      "epoch": 1.5,
      "grad_norm": 8.928720474243164,
      "learning_rate": 7.491418764302059e-06,
      "loss": 0.647,
      "step": 5246
    },
    {
      "epoch": 1.5,
      "grad_norm": 10.364776611328125,
      "learning_rate": 7.48712814645309e-06,
      "loss": 0.6653,
      "step": 5247
    },
    {
      "epoch": 1.5,
      "grad_norm": 13.162821769714355,
      "learning_rate": 7.4828375286041195e-06,
      "loss": 1.0826,
      "step": 5248
    },
    {
      "epoch": 1.5,
      "grad_norm": 8.7410249710083,
      "learning_rate": 7.478546910755149e-06,
      "loss": 0.5761,
      "step": 5249
    },
    {
      "epoch": 1.5,
      "grad_norm": 9.24797534942627,
      "learning_rate": 7.474256292906179e-06,
      "loss": 0.4566,
      "step": 5250
    },
    {
      "epoch": 1.5,
      "grad_norm": 9.684334754943848,
      "learning_rate": 7.469965675057208e-06,
      "loss": 0.6107,
      "step": 5251
    },
    {
      "epoch": 1.5,
      "grad_norm": 12.080406188964844,
      "learning_rate": 7.465675057208238e-06,
      "loss": 0.9,
      "step": 5252
    },
    {
      "epoch": 1.5,
      "grad_norm": 13.190532684326172,
      "learning_rate": 7.4613844393592676e-06,
      "loss": 0.9134,
      "step": 5253
    },
    {
      "epoch": 1.5,
      "grad_norm": 11.988898277282715,
      "learning_rate": 7.457093821510297e-06,
      "loss": 0.8086,
      "step": 5254
    },
    {
      "epoch": 1.5,
      "grad_norm": 12.422467231750488,
      "learning_rate": 7.452803203661328e-06,
      "loss": 0.809,
      "step": 5255
    },
    {
      "epoch": 1.5,
      "grad_norm": 13.21649169921875,
      "learning_rate": 7.448512585812357e-06,
      "loss": 0.9793,
      "step": 5256
    },
    {
      "epoch": 1.5,
      "grad_norm": 11.757281303405762,
      "learning_rate": 7.444221967963387e-06,
      "loss": 0.8449,
      "step": 5257
    },
    {
      "epoch": 1.5,
      "grad_norm": 10.680924415588379,
      "learning_rate": 7.4399313501144165e-06,
      "loss": 0.7947,
      "step": 5258
    },
    {
      "epoch": 1.5,
      "grad_norm": 10.09373664855957,
      "learning_rate": 7.435640732265447e-06,
      "loss": 0.7494,
      "step": 5259
    },
    {
      "epoch": 1.5,
      "grad_norm": 11.1739501953125,
      "learning_rate": 7.4313501144164766e-06,
      "loss": 0.682,
      "step": 5260
    },
    {
      "epoch": 1.5,
      "grad_norm": 10.319697380065918,
      "learning_rate": 7.427059496567506e-06,
      "loss": 0.8318,
      "step": 5261
    },
    {
      "epoch": 1.51,
      "grad_norm": 9.602401733398438,
      "learning_rate": 7.422768878718536e-06,
      "loss": 0.5827,
      "step": 5262
    },
    {
      "epoch": 1.51,
      "grad_norm": 8.633907318115234,
      "learning_rate": 7.418478260869565e-06,
      "loss": 0.5389,
      "step": 5263
    },
    {
      "epoch": 1.51,
      "grad_norm": 8.840450286865234,
      "learning_rate": 7.414187643020595e-06,
      "loss": 0.7338,
      "step": 5264
    },
    {
      "epoch": 1.51,
      "grad_norm": 9.036559104919434,
      "learning_rate": 7.409897025171625e-06,
      "loss": 0.6257,
      "step": 5265
    },
    {
      "epoch": 1.51,
      "grad_norm": 12.07451057434082,
      "learning_rate": 7.405606407322654e-06,
      "loss": 0.6456,
      "step": 5266
    },
    {
      "epoch": 1.51,
      "grad_norm": 9.394609451293945,
      "learning_rate": 7.401315789473684e-06,
      "loss": 0.7133,
      "step": 5267
    },
    {
      "epoch": 1.51,
      "grad_norm": 9.017657279968262,
      "learning_rate": 7.397025171624714e-06,
      "loss": 0.532,
      "step": 5268
    },
    {
      "epoch": 1.51,
      "grad_norm": 11.069826126098633,
      "learning_rate": 7.392734553775744e-06,
      "loss": 0.4487,
      "step": 5269
    },
    {
      "epoch": 1.51,
      "grad_norm": 11.50070571899414,
      "learning_rate": 7.3884439359267736e-06,
      "loss": 0.7551,
      "step": 5270
    },
    {
      "epoch": 1.51,
      "grad_norm": 8.076889038085938,
      "learning_rate": 7.384153318077803e-06,
      "loss": 0.52,
      "step": 5271
    },
    {
      "epoch": 1.51,
      "grad_norm": 9.858548164367676,
      "learning_rate": 7.379862700228834e-06,
      "loss": 0.7357,
      "step": 5272
    },
    {
      "epoch": 1.51,
      "grad_norm": 11.235546112060547,
      "learning_rate": 7.375572082379863e-06,
      "loss": 0.6642,
      "step": 5273
    },
    {
      "epoch": 1.51,
      "grad_norm": 12.630517959594727,
      "learning_rate": 7.371281464530893e-06,
      "loss": 0.8561,
      "step": 5274
    },
    {
      "epoch": 1.51,
      "grad_norm": 11.6703462600708,
      "learning_rate": 7.3669908466819225e-06,
      "loss": 0.9949,
      "step": 5275
    },
    {
      "epoch": 1.51,
      "grad_norm": 13.284300804138184,
      "learning_rate": 7.362700228832952e-06,
      "loss": 0.6989,
      "step": 5276
    },
    {
      "epoch": 1.51,
      "grad_norm": 10.491209030151367,
      "learning_rate": 7.358409610983982e-06,
      "loss": 0.7176,
      "step": 5277
    },
    {
      "epoch": 1.51,
      "grad_norm": 8.57515811920166,
      "learning_rate": 7.354118993135011e-06,
      "loss": 0.5642,
      "step": 5278
    },
    {
      "epoch": 1.51,
      "grad_norm": 11.295157432556152,
      "learning_rate": 7.349828375286041e-06,
      "loss": 0.5904,
      "step": 5279
    },
    {
      "epoch": 1.51,
      "grad_norm": 10.756020545959473,
      "learning_rate": 7.3455377574370706e-06,
      "loss": 0.5272,
      "step": 5280
    },
    {
      "epoch": 1.51,
      "grad_norm": 14.983441352844238,
      "learning_rate": 7.341247139588101e-06,
      "loss": 0.6953,
      "step": 5281
    },
    {
      "epoch": 1.51,
      "grad_norm": 10.309775352478027,
      "learning_rate": 7.336956521739131e-06,
      "loss": 0.6887,
      "step": 5282
    },
    {
      "epoch": 1.51,
      "grad_norm": 12.075322151184082,
      "learning_rate": 7.33266590389016e-06,
      "loss": 0.9896,
      "step": 5283
    },
    {
      "epoch": 1.51,
      "grad_norm": 11.666068077087402,
      "learning_rate": 7.32837528604119e-06,
      "loss": 0.7826,
      "step": 5284
    },
    {
      "epoch": 1.51,
      "grad_norm": 13.928191184997559,
      "learning_rate": 7.32408466819222e-06,
      "loss": 0.883,
      "step": 5285
    },
    {
      "epoch": 1.51,
      "grad_norm": 11.600682258605957,
      "learning_rate": 7.31979405034325e-06,
      "loss": 0.666,
      "step": 5286
    },
    {
      "epoch": 1.51,
      "grad_norm": 10.639303207397461,
      "learning_rate": 7.3155034324942796e-06,
      "loss": 0.7518,
      "step": 5287
    },
    {
      "epoch": 1.51,
      "grad_norm": 13.749682426452637,
      "learning_rate": 7.311212814645309e-06,
      "loss": 0.9626,
      "step": 5288
    },
    {
      "epoch": 1.51,
      "grad_norm": 8.073677062988281,
      "learning_rate": 7.30692219679634e-06,
      "loss": 0.59,
      "step": 5289
    },
    {
      "epoch": 1.51,
      "grad_norm": 10.90993881225586,
      "learning_rate": 7.302631578947368e-06,
      "loss": 0.6807,
      "step": 5290
    },
    {
      "epoch": 1.51,
      "grad_norm": 10.742498397827148,
      "learning_rate": 7.298340961098398e-06,
      "loss": 0.5879,
      "step": 5291
    },
    {
      "epoch": 1.51,
      "grad_norm": 10.239691734313965,
      "learning_rate": 7.294050343249428e-06,
      "loss": 0.7894,
      "step": 5292
    },
    {
      "epoch": 1.51,
      "grad_norm": 10.508283615112305,
      "learning_rate": 7.289759725400457e-06,
      "loss": 0.7049,
      "step": 5293
    },
    {
      "epoch": 1.51,
      "grad_norm": 10.425931930541992,
      "learning_rate": 7.285469107551488e-06,
      "loss": 0.7857,
      "step": 5294
    },
    {
      "epoch": 1.51,
      "grad_norm": 12.55759334564209,
      "learning_rate": 7.281178489702517e-06,
      "loss": 0.8908,
      "step": 5295
    },
    {
      "epoch": 1.51,
      "grad_norm": 9.504537582397461,
      "learning_rate": 7.276887871853547e-06,
      "loss": 0.6734,
      "step": 5296
    },
    {
      "epoch": 1.52,
      "grad_norm": 9.230053901672363,
      "learning_rate": 7.2725972540045766e-06,
      "loss": 0.7552,
      "step": 5297
    },
    {
      "epoch": 1.52,
      "grad_norm": 9.277558326721191,
      "learning_rate": 7.268306636155607e-06,
      "loss": 0.5683,
      "step": 5298
    },
    {
      "epoch": 1.52,
      "grad_norm": 9.663713455200195,
      "learning_rate": 7.264016018306637e-06,
      "loss": 0.8314,
      "step": 5299
    },
    {
      "epoch": 1.52,
      "grad_norm": 11.926158905029297,
      "learning_rate": 7.259725400457666e-06,
      "loss": 0.9163,
      "step": 5300
    },
    {
      "epoch": 1.52,
      "grad_norm": 10.892145156860352,
      "learning_rate": 7.255434782608696e-06,
      "loss": 0.5672,
      "step": 5301
    },
    {
      "epoch": 1.52,
      "grad_norm": 12.315850257873535,
      "learning_rate": 7.251144164759726e-06,
      "loss": 0.8399,
      "step": 5302
    },
    {
      "epoch": 1.52,
      "grad_norm": 11.595917701721191,
      "learning_rate": 7.246853546910756e-06,
      "loss": 0.8012,
      "step": 5303
    },
    {
      "epoch": 1.52,
      "grad_norm": 10.428421974182129,
      "learning_rate": 7.242562929061785e-06,
      "loss": 0.7003,
      "step": 5304
    },
    {
      "epoch": 1.52,
      "grad_norm": 10.854645729064941,
      "learning_rate": 7.238272311212814e-06,
      "loss": 0.8191,
      "step": 5305
    },
    {
      "epoch": 1.52,
      "grad_norm": 12.198843955993652,
      "learning_rate": 7.233981693363844e-06,
      "loss": 0.8671,
      "step": 5306
    },
    {
      "epoch": 1.52,
      "grad_norm": 9.188078880310059,
      "learning_rate": 7.229691075514874e-06,
      "loss": 0.6152,
      "step": 5307
    },
    {
      "epoch": 1.52,
      "grad_norm": 9.002267837524414,
      "learning_rate": 7.225400457665904e-06,
      "loss": 0.7284,
      "step": 5308
    },
    {
      "epoch": 1.52,
      "grad_norm": 9.587257385253906,
      "learning_rate": 7.221109839816934e-06,
      "loss": 0.8177,
      "step": 5309
    },
    {
      "epoch": 1.52,
      "grad_norm": 11.889808654785156,
      "learning_rate": 7.216819221967963e-06,
      "loss": 0.7482,
      "step": 5310
    },
    {
      "epoch": 1.52,
      "grad_norm": 11.84115982055664,
      "learning_rate": 7.212528604118994e-06,
      "loss": 0.7478,
      "step": 5311
    },
    {
      "epoch": 1.52,
      "grad_norm": 11.276606559753418,
      "learning_rate": 7.208237986270023e-06,
      "loss": 0.7999,
      "step": 5312
    },
    {
      "epoch": 1.52,
      "grad_norm": 9.208023071289062,
      "learning_rate": 7.203947368421053e-06,
      "loss": 0.6271,
      "step": 5313
    },
    {
      "epoch": 1.52,
      "grad_norm": 8.481173515319824,
      "learning_rate": 7.1996567505720826e-06,
      "loss": 0.5259,
      "step": 5314
    },
    {
      "epoch": 1.52,
      "grad_norm": 10.612635612487793,
      "learning_rate": 7.195366132723113e-06,
      "loss": 0.8156,
      "step": 5315
    },
    {
      "epoch": 1.52,
      "grad_norm": 9.746124267578125,
      "learning_rate": 7.191075514874143e-06,
      "loss": 0.7464,
      "step": 5316
    },
    {
      "epoch": 1.52,
      "grad_norm": 11.440072059631348,
      "learning_rate": 7.186784897025171e-06,
      "loss": 0.5033,
      "step": 5317
    },
    {
      "epoch": 1.52,
      "grad_norm": 10.415870666503906,
      "learning_rate": 7.182494279176201e-06,
      "loss": 0.6788,
      "step": 5318
    },
    {
      "epoch": 1.52,
      "grad_norm": 9.202872276306152,
      "learning_rate": 7.1782036613272315e-06,
      "loss": 0.7637,
      "step": 5319
    },
    {
      "epoch": 1.52,
      "grad_norm": 10.398765563964844,
      "learning_rate": 7.173913043478261e-06,
      "loss": 0.8702,
      "step": 5320
    },
    {
      "epoch": 1.52,
      "grad_norm": 11.210211753845215,
      "learning_rate": 7.169622425629291e-06,
      "loss": 1.0157,
      "step": 5321
    },
    {
      "epoch": 1.52,
      "grad_norm": 8.577627182006836,
      "learning_rate": 7.16533180778032e-06,
      "loss": 0.5621,
      "step": 5322
    },
    {
      "epoch": 1.52,
      "grad_norm": 9.694450378417969,
      "learning_rate": 7.16104118993135e-06,
      "loss": 0.8201,
      "step": 5323
    },
    {
      "epoch": 1.52,
      "grad_norm": 9.710101127624512,
      "learning_rate": 7.15675057208238e-06,
      "loss": 0.591,
      "step": 5324
    },
    {
      "epoch": 1.52,
      "grad_norm": 9.863205909729004,
      "learning_rate": 7.15245995423341e-06,
      "loss": 0.6081,
      "step": 5325
    },
    {
      "epoch": 1.52,
      "grad_norm": 10.321282386779785,
      "learning_rate": 7.14816933638444e-06,
      "loss": 0.7331,
      "step": 5326
    },
    {
      "epoch": 1.52,
      "grad_norm": 7.0798258781433105,
      "learning_rate": 7.143878718535469e-06,
      "loss": 0.4342,
      "step": 5327
    },
    {
      "epoch": 1.52,
      "grad_norm": 10.438071250915527,
      "learning_rate": 7.1395881006865e-06,
      "loss": 0.8411,
      "step": 5328
    },
    {
      "epoch": 1.52,
      "grad_norm": 8.400115013122559,
      "learning_rate": 7.135297482837529e-06,
      "loss": 0.6325,
      "step": 5329
    },
    {
      "epoch": 1.52,
      "grad_norm": 11.314515113830566,
      "learning_rate": 7.131006864988558e-06,
      "loss": 0.516,
      "step": 5330
    },
    {
      "epoch": 1.52,
      "grad_norm": 10.420086860656738,
      "learning_rate": 7.126716247139588e-06,
      "loss": 0.8869,
      "step": 5331
    },
    {
      "epoch": 1.53,
      "grad_norm": 10.788599014282227,
      "learning_rate": 7.122425629290618e-06,
      "loss": 0.6434,
      "step": 5332
    },
    {
      "epoch": 1.53,
      "grad_norm": 10.751967430114746,
      "learning_rate": 7.118135011441648e-06,
      "loss": 0.8512,
      "step": 5333
    },
    {
      "epoch": 1.53,
      "grad_norm": 9.534699440002441,
      "learning_rate": 7.113844393592677e-06,
      "loss": 0.718,
      "step": 5334
    },
    {
      "epoch": 1.53,
      "grad_norm": 9.03892707824707,
      "learning_rate": 7.109553775743707e-06,
      "loss": 0.6529,
      "step": 5335
    },
    {
      "epoch": 1.53,
      "grad_norm": 10.13648796081543,
      "learning_rate": 7.105263157894737e-06,
      "loss": 0.6326,
      "step": 5336
    },
    {
      "epoch": 1.53,
      "grad_norm": 9.899974822998047,
      "learning_rate": 7.100972540045767e-06,
      "loss": 0.5475,
      "step": 5337
    },
    {
      "epoch": 1.53,
      "grad_norm": 11.557777404785156,
      "learning_rate": 7.096681922196797e-06,
      "loss": 0.8745,
      "step": 5338
    },
    {
      "epoch": 1.53,
      "grad_norm": 11.886506080627441,
      "learning_rate": 7.092391304347826e-06,
      "loss": 0.9558,
      "step": 5339
    },
    {
      "epoch": 1.53,
      "grad_norm": 7.03980016708374,
      "learning_rate": 7.088100686498856e-06,
      "loss": 0.4913,
      "step": 5340
    },
    {
      "epoch": 1.53,
      "grad_norm": 10.639065742492676,
      "learning_rate": 7.083810068649886e-06,
      "loss": 0.5013,
      "step": 5341
    },
    {
      "epoch": 1.53,
      "grad_norm": 16.474300384521484,
      "learning_rate": 7.079519450800916e-06,
      "loss": 0.9308,
      "step": 5342
    },
    {
      "epoch": 1.53,
      "grad_norm": 10.744767189025879,
      "learning_rate": 7.075228832951945e-06,
      "loss": 0.7818,
      "step": 5343
    },
    {
      "epoch": 1.53,
      "grad_norm": 11.738082885742188,
      "learning_rate": 7.070938215102974e-06,
      "loss": 0.7811,
      "step": 5344
    },
    {
      "epoch": 1.53,
      "grad_norm": 11.322881698608398,
      "learning_rate": 7.066647597254005e-06,
      "loss": 0.7869,
      "step": 5345
    },
    {
      "epoch": 1.53,
      "grad_norm": 11.683786392211914,
      "learning_rate": 7.0623569794050345e-06,
      "loss": 0.938,
      "step": 5346
    },
    {
      "epoch": 1.53,
      "grad_norm": 9.391820907592773,
      "learning_rate": 7.058066361556064e-06,
      "loss": 0.5965,
      "step": 5347
    },
    {
      "epoch": 1.53,
      "grad_norm": 13.156702041625977,
      "learning_rate": 7.053775743707094e-06,
      "loss": 0.8922,
      "step": 5348
    },
    {
      "epoch": 1.53,
      "grad_norm": 13.145834922790527,
      "learning_rate": 7.049485125858124e-06,
      "loss": 0.8078,
      "step": 5349
    },
    {
      "epoch": 1.53,
      "grad_norm": 8.215470314025879,
      "learning_rate": 7.045194508009154e-06,
      "loss": 0.4871,
      "step": 5350
    },
    {
      "epoch": 1.53,
      "grad_norm": 11.24267578125,
      "learning_rate": 7.040903890160183e-06,
      "loss": 0.8555,
      "step": 5351
    },
    {
      "epoch": 1.53,
      "grad_norm": 10.632845878601074,
      "learning_rate": 7.036613272311213e-06,
      "loss": 0.5971,
      "step": 5352
    },
    {
      "epoch": 1.53,
      "grad_norm": 13.724236488342285,
      "learning_rate": 7.032322654462243e-06,
      "loss": 0.7376,
      "step": 5353
    },
    {
      "epoch": 1.53,
      "grad_norm": 10.956292152404785,
      "learning_rate": 7.028032036613273e-06,
      "loss": 0.6456,
      "step": 5354
    },
    {
      "epoch": 1.53,
      "grad_norm": 8.754426002502441,
      "learning_rate": 7.023741418764303e-06,
      "loss": 0.5247,
      "step": 5355
    },
    {
      "epoch": 1.53,
      "grad_norm": 9.3694486618042,
      "learning_rate": 7.0194508009153315e-06,
      "loss": 0.6017,
      "step": 5356
    },
    {
      "epoch": 1.53,
      "grad_norm": 9.64429759979248,
      "learning_rate": 7.015160183066361e-06,
      "loss": 0.5578,
      "step": 5357
    },
    {
      "epoch": 1.53,
      "grad_norm": 11.855950355529785,
      "learning_rate": 7.0108695652173915e-06,
      "loss": 0.8204,
      "step": 5358
    },
    {
      "epoch": 1.53,
      "grad_norm": 10.41483211517334,
      "learning_rate": 7.006578947368421e-06,
      "loss": 0.5388,
      "step": 5359
    },
    {
      "epoch": 1.53,
      "grad_norm": 10.524611473083496,
      "learning_rate": 7.002288329519451e-06,
      "loss": 0.5432,
      "step": 5360
    },
    {
      "epoch": 1.53,
      "grad_norm": 13.025598526000977,
      "learning_rate": 6.99799771167048e-06,
      "loss": 0.961,
      "step": 5361
    },
    {
      "epoch": 1.53,
      "grad_norm": 9.626395225524902,
      "learning_rate": 6.993707093821511e-06,
      "loss": 0.6729,
      "step": 5362
    },
    {
      "epoch": 1.53,
      "grad_norm": 11.571722984313965,
      "learning_rate": 6.9894164759725405e-06,
      "loss": 0.7831,
      "step": 5363
    },
    {
      "epoch": 1.53,
      "grad_norm": 13.247154235839844,
      "learning_rate": 6.98512585812357e-06,
      "loss": 0.7105,
      "step": 5364
    },
    {
      "epoch": 1.53,
      "grad_norm": 9.18348503112793,
      "learning_rate": 6.9808352402746e-06,
      "loss": 0.8202,
      "step": 5365
    },
    {
      "epoch": 1.53,
      "grad_norm": 10.731383323669434,
      "learning_rate": 6.976544622425629e-06,
      "loss": 0.711,
      "step": 5366
    },
    {
      "epoch": 1.54,
      "grad_norm": 10.938879013061523,
      "learning_rate": 6.97225400457666e-06,
      "loss": 0.7683,
      "step": 5367
    },
    {
      "epoch": 1.54,
      "grad_norm": 9.101683616638184,
      "learning_rate": 6.967963386727689e-06,
      "loss": 0.5983,
      "step": 5368
    },
    {
      "epoch": 1.54,
      "grad_norm": 8.889518737792969,
      "learning_rate": 6.963672768878718e-06,
      "loss": 0.636,
      "step": 5369
    },
    {
      "epoch": 1.54,
      "grad_norm": 10.405555725097656,
      "learning_rate": 6.959382151029748e-06,
      "loss": 0.6939,
      "step": 5370
    },
    {
      "epoch": 1.54,
      "grad_norm": 10.569013595581055,
      "learning_rate": 6.955091533180778e-06,
      "loss": 0.8429,
      "step": 5371
    },
    {
      "epoch": 1.54,
      "grad_norm": 9.115777969360352,
      "learning_rate": 6.950800915331808e-06,
      "loss": 0.5183,
      "step": 5372
    },
    {
      "epoch": 1.54,
      "grad_norm": 9.988287925720215,
      "learning_rate": 6.9465102974828375e-06,
      "loss": 0.7661,
      "step": 5373
    },
    {
      "epoch": 1.54,
      "grad_norm": 11.403606414794922,
      "learning_rate": 6.942219679633867e-06,
      "loss": 0.8205,
      "step": 5374
    },
    {
      "epoch": 1.54,
      "grad_norm": 12.239830017089844,
      "learning_rate": 6.9379290617848975e-06,
      "loss": 1.0035,
      "step": 5375
    },
    {
      "epoch": 1.54,
      "grad_norm": 9.13248348236084,
      "learning_rate": 6.933638443935927e-06,
      "loss": 0.6676,
      "step": 5376
    },
    {
      "epoch": 1.54,
      "grad_norm": 11.136905670166016,
      "learning_rate": 6.929347826086957e-06,
      "loss": 0.8766,
      "step": 5377
    },
    {
      "epoch": 1.54,
      "grad_norm": 7.92160701751709,
      "learning_rate": 6.925057208237986e-06,
      "loss": 0.6367,
      "step": 5378
    },
    {
      "epoch": 1.54,
      "grad_norm": 10.290059089660645,
      "learning_rate": 6.920766590389017e-06,
      "loss": 0.6464,
      "step": 5379
    },
    {
      "epoch": 1.54,
      "grad_norm": 10.359493255615234,
      "learning_rate": 6.9164759725400465e-06,
      "loss": 0.7502,
      "step": 5380
    },
    {
      "epoch": 1.54,
      "grad_norm": 9.81421184539795,
      "learning_rate": 6.912185354691076e-06,
      "loss": 0.8105,
      "step": 5381
    },
    {
      "epoch": 1.54,
      "grad_norm": 10.404888153076172,
      "learning_rate": 6.907894736842106e-06,
      "loss": 0.796,
      "step": 5382
    },
    {
      "epoch": 1.54,
      "grad_norm": 9.707366943359375,
      "learning_rate": 6.9036041189931345e-06,
      "loss": 0.7516,
      "step": 5383
    },
    {
      "epoch": 1.54,
      "grad_norm": 10.13158893585205,
      "learning_rate": 6.899313501144165e-06,
      "loss": 0.802,
      "step": 5384
    },
    {
      "epoch": 1.54,
      "grad_norm": 11.407342910766602,
      "learning_rate": 6.8950228832951945e-06,
      "loss": 0.6751,
      "step": 5385
    },
    {
      "epoch": 1.54,
      "grad_norm": 13.483102798461914,
      "learning_rate": 6.890732265446224e-06,
      "loss": 0.8375,
      "step": 5386
    },
    {
      "epoch": 1.54,
      "grad_norm": 11.197330474853516,
      "learning_rate": 6.886441647597254e-06,
      "loss": 0.662,
      "step": 5387
    },
    {
      "epoch": 1.54,
      "grad_norm": 11.573650360107422,
      "learning_rate": 6.882151029748284e-06,
      "loss": 0.85,
      "step": 5388
    },
    {
      "epoch": 1.54,
      "grad_norm": 13.032238006591797,
      "learning_rate": 6.877860411899314e-06,
      "loss": 1.016,
      "step": 5389
    },
    {
      "epoch": 1.54,
      "grad_norm": 10.803162574768066,
      "learning_rate": 6.8735697940503434e-06,
      "loss": 0.56,
      "step": 5390
    },
    {
      "epoch": 1.54,
      "grad_norm": 10.913394927978516,
      "learning_rate": 6.869279176201373e-06,
      "loss": 0.6293,
      "step": 5391
    },
    {
      "epoch": 1.54,
      "grad_norm": 10.91557502746582,
      "learning_rate": 6.8649885583524035e-06,
      "loss": 0.8244,
      "step": 5392
    },
    {
      "epoch": 1.54,
      "grad_norm": 10.184675216674805,
      "learning_rate": 6.860697940503433e-06,
      "loss": 0.6208,
      "step": 5393
    },
    {
      "epoch": 1.54,
      "grad_norm": 10.222221374511719,
      "learning_rate": 6.856407322654463e-06,
      "loss": 0.6482,
      "step": 5394
    },
    {
      "epoch": 1.54,
      "grad_norm": 8.793624877929688,
      "learning_rate": 6.852116704805492e-06,
      "loss": 0.5475,
      "step": 5395
    },
    {
      "epoch": 1.54,
      "grad_norm": 11.794410705566406,
      "learning_rate": 6.847826086956521e-06,
      "loss": 0.5631,
      "step": 5396
    },
    {
      "epoch": 1.54,
      "grad_norm": 13.063423156738281,
      "learning_rate": 6.843535469107552e-06,
      "loss": 0.5843,
      "step": 5397
    },
    {
      "epoch": 1.54,
      "grad_norm": 12.135796546936035,
      "learning_rate": 6.839244851258581e-06,
      "loss": 0.6722,
      "step": 5398
    },
    {
      "epoch": 1.54,
      "grad_norm": 11.323772430419922,
      "learning_rate": 6.834954233409611e-06,
      "loss": 0.5577,
      "step": 5399
    },
    {
      "epoch": 1.54,
      "grad_norm": 11.501399993896484,
      "learning_rate": 6.8306636155606404e-06,
      "loss": 0.8055,
      "step": 5400
    },
    {
      "epoch": 1.54,
      "grad_norm": 9.352534294128418,
      "learning_rate": 6.826372997711671e-06,
      "loss": 0.7704,
      "step": 5401
    },
    {
      "epoch": 1.55,
      "grad_norm": 11.826374053955078,
      "learning_rate": 6.8220823798627005e-06,
      "loss": 0.6599,
      "step": 5402
    },
    {
      "epoch": 1.55,
      "grad_norm": 9.877493858337402,
      "learning_rate": 6.81779176201373e-06,
      "loss": 0.76,
      "step": 5403
    },
    {
      "epoch": 1.55,
      "grad_norm": 11.737428665161133,
      "learning_rate": 6.81350114416476e-06,
      "loss": 0.7593,
      "step": 5404
    },
    {
      "epoch": 1.55,
      "grad_norm": 10.759040832519531,
      "learning_rate": 6.80921052631579e-06,
      "loss": 0.9183,
      "step": 5405
    },
    {
      "epoch": 1.55,
      "grad_norm": 10.864048957824707,
      "learning_rate": 6.80491990846682e-06,
      "loss": 0.7596,
      "step": 5406
    },
    {
      "epoch": 1.55,
      "grad_norm": 9.24080753326416,
      "learning_rate": 6.8006292906178494e-06,
      "loss": 0.6632,
      "step": 5407
    },
    {
      "epoch": 1.55,
      "grad_norm": 11.46875,
      "learning_rate": 6.796338672768879e-06,
      "loss": 0.6237,
      "step": 5408
    },
    {
      "epoch": 1.55,
      "grad_norm": 11.83765697479248,
      "learning_rate": 6.792048054919909e-06,
      "loss": 0.7213,
      "step": 5409
    },
    {
      "epoch": 1.55,
      "grad_norm": 12.011991500854492,
      "learning_rate": 6.787757437070938e-06,
      "loss": 0.9464,
      "step": 5410
    },
    {
      "epoch": 1.55,
      "grad_norm": 10.703045845031738,
      "learning_rate": 6.783466819221968e-06,
      "loss": 0.7118,
      "step": 5411
    },
    {
      "epoch": 1.55,
      "grad_norm": 10.701443672180176,
      "learning_rate": 6.7791762013729975e-06,
      "loss": 0.5642,
      "step": 5412
    },
    {
      "epoch": 1.55,
      "grad_norm": 10.329233169555664,
      "learning_rate": 6.774885583524027e-06,
      "loss": 0.6452,
      "step": 5413
    },
    {
      "epoch": 1.55,
      "grad_norm": 8.791563034057617,
      "learning_rate": 6.770594965675058e-06,
      "loss": 0.5704,
      "step": 5414
    },
    {
      "epoch": 1.55,
      "grad_norm": 9.633209228515625,
      "learning_rate": 6.766304347826087e-06,
      "loss": 0.8627,
      "step": 5415
    },
    {
      "epoch": 1.55,
      "grad_norm": 9.805323600769043,
      "learning_rate": 6.762013729977117e-06,
      "loss": 0.6023,
      "step": 5416
    },
    {
      "epoch": 1.55,
      "grad_norm": 10.737568855285645,
      "learning_rate": 6.7577231121281464e-06,
      "loss": 0.708,
      "step": 5417
    },
    {
      "epoch": 1.55,
      "grad_norm": 8.404878616333008,
      "learning_rate": 6.753432494279177e-06,
      "loss": 0.6025,
      "step": 5418
    },
    {
      "epoch": 1.55,
      "grad_norm": 10.053535461425781,
      "learning_rate": 6.7491418764302065e-06,
      "loss": 0.8781,
      "step": 5419
    },
    {
      "epoch": 1.55,
      "grad_norm": 13.944951057434082,
      "learning_rate": 6.744851258581236e-06,
      "loss": 1.1078,
      "step": 5420
    },
    {
      "epoch": 1.55,
      "grad_norm": 9.490021705627441,
      "learning_rate": 6.740560640732266e-06,
      "loss": 0.6951,
      "step": 5421
    },
    {
      "epoch": 1.55,
      "grad_norm": 9.664155960083008,
      "learning_rate": 6.736270022883295e-06,
      "loss": 0.8254,
      "step": 5422
    },
    {
      "epoch": 1.55,
      "grad_norm": 11.093666076660156,
      "learning_rate": 6.731979405034325e-06,
      "loss": 0.6202,
      "step": 5423
    },
    {
      "epoch": 1.55,
      "grad_norm": 11.165013313293457,
      "learning_rate": 6.727688787185355e-06,
      "loss": 0.7902,
      "step": 5424
    },
    {
      "epoch": 1.55,
      "grad_norm": 10.131797790527344,
      "learning_rate": 6.723398169336384e-06,
      "loss": 0.596,
      "step": 5425
    },
    {
      "epoch": 1.55,
      "grad_norm": 9.271760940551758,
      "learning_rate": 6.719107551487414e-06,
      "loss": 0.4633,
      "step": 5426
    },
    {
      "epoch": 1.55,
      "grad_norm": 11.775052070617676,
      "learning_rate": 6.714816933638444e-06,
      "loss": 0.8464,
      "step": 5427
    },
    {
      "epoch": 1.55,
      "grad_norm": 8.292208671569824,
      "learning_rate": 6.710526315789474e-06,
      "loss": 0.5077,
      "step": 5428
    },
    {
      "epoch": 1.55,
      "grad_norm": 10.974455833435059,
      "learning_rate": 6.7062356979405035e-06,
      "loss": 0.721,
      "step": 5429
    },
    {
      "epoch": 1.55,
      "grad_norm": 10.512555122375488,
      "learning_rate": 6.701945080091533e-06,
      "loss": 0.961,
      "step": 5430
    },
    {
      "epoch": 1.55,
      "grad_norm": 15.095166206359863,
      "learning_rate": 6.697654462242564e-06,
      "loss": 0.8586,
      "step": 5431
    },
    {
      "epoch": 1.55,
      "grad_norm": 11.317481994628906,
      "learning_rate": 6.693363844393593e-06,
      "loss": 0.9296,
      "step": 5432
    },
    {
      "epoch": 1.55,
      "grad_norm": 10.495819091796875,
      "learning_rate": 6.689073226544623e-06,
      "loss": 0.8478,
      "step": 5433
    },
    {
      "epoch": 1.55,
      "grad_norm": 11.36196231842041,
      "learning_rate": 6.6847826086956524e-06,
      "loss": 0.9135,
      "step": 5434
    },
    {
      "epoch": 1.55,
      "grad_norm": 11.978899002075195,
      "learning_rate": 6.680491990846682e-06,
      "loss": 0.8818,
      "step": 5435
    },
    {
      "epoch": 1.55,
      "grad_norm": 10.200531005859375,
      "learning_rate": 6.676201372997712e-06,
      "loss": 0.8606,
      "step": 5436
    },
    {
      "epoch": 1.56,
      "grad_norm": 11.20178508758545,
      "learning_rate": 6.671910755148741e-06,
      "loss": 0.5133,
      "step": 5437
    },
    {
      "epoch": 1.56,
      "grad_norm": 10.661775588989258,
      "learning_rate": 6.667620137299771e-06,
      "loss": 0.6154,
      "step": 5438
    },
    {
      "epoch": 1.56,
      "grad_norm": 10.742033958435059,
      "learning_rate": 6.663329519450801e-06,
      "loss": 0.4946,
      "step": 5439
    },
    {
      "epoch": 1.56,
      "grad_norm": 11.584891319274902,
      "learning_rate": 6.659038901601831e-06,
      "loss": 0.8595,
      "step": 5440
    },
    {
      "epoch": 1.56,
      "grad_norm": 9.102738380432129,
      "learning_rate": 6.654748283752861e-06,
      "loss": 0.7664,
      "step": 5441
    },
    {
      "epoch": 1.56,
      "grad_norm": 10.778501510620117,
      "learning_rate": 6.65045766590389e-06,
      "loss": 0.6236,
      "step": 5442
    },
    {
      "epoch": 1.56,
      "grad_norm": 9.565971374511719,
      "learning_rate": 6.64616704805492e-06,
      "loss": 0.6284,
      "step": 5443
    },
    {
      "epoch": 1.56,
      "grad_norm": 14.338422775268555,
      "learning_rate": 6.64187643020595e-06,
      "loss": 1.364,
      "step": 5444
    },
    {
      "epoch": 1.56,
      "grad_norm": 9.952033996582031,
      "learning_rate": 6.63758581235698e-06,
      "loss": 0.6922,
      "step": 5445
    },
    {
      "epoch": 1.56,
      "grad_norm": 11.656608581542969,
      "learning_rate": 6.6332951945080095e-06,
      "loss": 0.8932,
      "step": 5446
    },
    {
      "epoch": 1.56,
      "grad_norm": 10.877176284790039,
      "learning_rate": 6.629004576659039e-06,
      "loss": 0.7369,
      "step": 5447
    },
    {
      "epoch": 1.56,
      "grad_norm": 9.53266429901123,
      "learning_rate": 6.624713958810069e-06,
      "loss": 0.7248,
      "step": 5448
    },
    {
      "epoch": 1.56,
      "grad_norm": 10.719442367553711,
      "learning_rate": 6.620423340961098e-06,
      "loss": 0.5127,
      "step": 5449
    },
    {
      "epoch": 1.56,
      "grad_norm": 9.9187593460083,
      "learning_rate": 6.616132723112128e-06,
      "loss": 0.7927,
      "step": 5450
    },
    {
      "epoch": 1.56,
      "grad_norm": 10.836196899414062,
      "learning_rate": 6.611842105263158e-06,
      "loss": 0.9459,
      "step": 5451
    },
    {
      "epoch": 1.56,
      "grad_norm": 9.903067588806152,
      "learning_rate": 6.607551487414188e-06,
      "loss": 0.7539,
      "step": 5452
    },
    {
      "epoch": 1.56,
      "grad_norm": 9.817303657531738,
      "learning_rate": 6.603260869565218e-06,
      "loss": 0.5441,
      "step": 5453
    },
    {
      "epoch": 1.56,
      "grad_norm": 8.64751148223877,
      "learning_rate": 6.598970251716247e-06,
      "loss": 0.5629,
      "step": 5454
    },
    {
      "epoch": 1.56,
      "grad_norm": 8.386674880981445,
      "learning_rate": 6.594679633867277e-06,
      "loss": 0.5778,
      "step": 5455
    },
    {
      "epoch": 1.56,
      "grad_norm": 8.778495788574219,
      "learning_rate": 6.5903890160183065e-06,
      "loss": 0.7115,
      "step": 5456
    },
    {
      "epoch": 1.56,
      "grad_norm": 9.286686897277832,
      "learning_rate": 6.586098398169337e-06,
      "loss": 0.6961,
      "step": 5457
    },
    {
      "epoch": 1.56,
      "grad_norm": 9.776361465454102,
      "learning_rate": 6.581807780320367e-06,
      "loss": 0.7327,
      "step": 5458
    },
    {
      "epoch": 1.56,
      "grad_norm": 13.709680557250977,
      "learning_rate": 6.577517162471396e-06,
      "loss": 0.9498,
      "step": 5459
    },
    {
      "epoch": 1.56,
      "grad_norm": 11.222743034362793,
      "learning_rate": 6.573226544622426e-06,
      "loss": 0.5232,
      "step": 5460
    },
    {
      "epoch": 1.56,
      "grad_norm": 12.961259841918945,
      "learning_rate": 6.5689359267734554e-06,
      "loss": 0.7338,
      "step": 5461
    },
    {
      "epoch": 1.56,
      "grad_norm": 9.301464080810547,
      "learning_rate": 6.564645308924485e-06,
      "loss": 0.7528,
      "step": 5462
    },
    {
      "epoch": 1.56,
      "grad_norm": 19.588850021362305,
      "learning_rate": 6.560354691075515e-06,
      "loss": 0.807,
      "step": 5463
    },
    {
      "epoch": 1.56,
      "grad_norm": 10.187501907348633,
      "learning_rate": 6.556064073226544e-06,
      "loss": 1.0132,
      "step": 5464
    },
    {
      "epoch": 1.56,
      "grad_norm": 10.580513954162598,
      "learning_rate": 6.551773455377575e-06,
      "loss": 0.5971,
      "step": 5465
    },
    {
      "epoch": 1.56,
      "grad_norm": 11.745843887329102,
      "learning_rate": 6.547482837528604e-06,
      "loss": 0.9573,
      "step": 5466
    },
    {
      "epoch": 1.56,
      "grad_norm": 8.349631309509277,
      "learning_rate": 6.543192219679634e-06,
      "loss": 0.4772,
      "step": 5467
    },
    {
      "epoch": 1.56,
      "grad_norm": 9.672638893127441,
      "learning_rate": 6.538901601830664e-06,
      "loss": 0.6586,
      "step": 5468
    },
    {
      "epoch": 1.56,
      "grad_norm": 10.535762786865234,
      "learning_rate": 6.534610983981694e-06,
      "loss": 0.7865,
      "step": 5469
    },
    {
      "epoch": 1.56,
      "grad_norm": 10.90261459350586,
      "learning_rate": 6.530320366132724e-06,
      "loss": 0.5622,
      "step": 5470
    },
    {
      "epoch": 1.56,
      "grad_norm": 10.270242691040039,
      "learning_rate": 6.526029748283753e-06,
      "loss": 0.7503,
      "step": 5471
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.051499366760254,
      "learning_rate": 6.521739130434783e-06,
      "loss": 0.8852,
      "step": 5472
    },
    {
      "epoch": 1.57,
      "grad_norm": 15.139245986938477,
      "learning_rate": 6.5174485125858125e-06,
      "loss": 0.7186,
      "step": 5473
    },
    {
      "epoch": 1.57,
      "grad_norm": 12.09343433380127,
      "learning_rate": 6.513157894736842e-06,
      "loss": 0.8723,
      "step": 5474
    },
    {
      "epoch": 1.57,
      "grad_norm": 8.4949312210083,
      "learning_rate": 6.508867276887872e-06,
      "loss": 0.5647,
      "step": 5475
    },
    {
      "epoch": 1.57,
      "grad_norm": 12.268529891967773,
      "learning_rate": 6.504576659038901e-06,
      "loss": 0.8052,
      "step": 5476
    },
    {
      "epoch": 1.57,
      "grad_norm": 12.441484451293945,
      "learning_rate": 6.500286041189931e-06,
      "loss": 0.5815,
      "step": 5477
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.255128860473633,
      "learning_rate": 6.495995423340961e-06,
      "loss": 0.6331,
      "step": 5478
    },
    {
      "epoch": 1.57,
      "grad_norm": 11.862914085388184,
      "learning_rate": 6.491704805491991e-06,
      "loss": 0.6901,
      "step": 5479
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.020846366882324,
      "learning_rate": 6.487414187643021e-06,
      "loss": 0.7216,
      "step": 5480
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.554515838623047,
      "learning_rate": 6.48312356979405e-06,
      "loss": 0.7237,
      "step": 5481
    },
    {
      "epoch": 1.57,
      "grad_norm": 12.486430168151855,
      "learning_rate": 6.478832951945081e-06,
      "loss": 0.6598,
      "step": 5482
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.92392349243164,
      "learning_rate": 6.47454233409611e-06,
      "loss": 0.7801,
      "step": 5483
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.194019317626953,
      "learning_rate": 6.47025171624714e-06,
      "loss": 0.6593,
      "step": 5484
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.077391624450684,
      "learning_rate": 6.4659610983981696e-06,
      "loss": 0.6658,
      "step": 5485
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.685776710510254,
      "learning_rate": 6.461670480549199e-06,
      "loss": 0.6464,
      "step": 5486
    },
    {
      "epoch": 1.57,
      "grad_norm": 9.115002632141113,
      "learning_rate": 6.45737986270023e-06,
      "loss": 0.7477,
      "step": 5487
    },
    {
      "epoch": 1.57,
      "grad_norm": 8.218844413757324,
      "learning_rate": 6.453089244851258e-06,
      "loss": 0.6765,
      "step": 5488
    },
    {
      "epoch": 1.57,
      "grad_norm": 12.361523628234863,
      "learning_rate": 6.448798627002288e-06,
      "loss": 0.749,
      "step": 5489
    },
    {
      "epoch": 1.57,
      "grad_norm": 8.853835105895996,
      "learning_rate": 6.444508009153318e-06,
      "loss": 0.6066,
      "step": 5490
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.084883689880371,
      "learning_rate": 6.440217391304348e-06,
      "loss": 0.7014,
      "step": 5491
    },
    {
      "epoch": 1.57,
      "grad_norm": 12.231773376464844,
      "learning_rate": 6.435926773455378e-06,
      "loss": 0.9233,
      "step": 5492
    },
    {
      "epoch": 1.57,
      "grad_norm": 12.060643196105957,
      "learning_rate": 6.431636155606407e-06,
      "loss": 0.8505,
      "step": 5493
    },
    {
      "epoch": 1.57,
      "grad_norm": 12.02497386932373,
      "learning_rate": 6.427345537757437e-06,
      "loss": 0.6959,
      "step": 5494
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.105432510375977,
      "learning_rate": 6.423054919908467e-06,
      "loss": 0.6364,
      "step": 5495
    },
    {
      "epoch": 1.57,
      "grad_norm": 11.441240310668945,
      "learning_rate": 6.418764302059497e-06,
      "loss": 0.7262,
      "step": 5496
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.952935218811035,
      "learning_rate": 6.414473684210527e-06,
      "loss": 0.6929,
      "step": 5497
    },
    {
      "epoch": 1.57,
      "grad_norm": 11.399803161621094,
      "learning_rate": 6.410183066361556e-06,
      "loss": 0.857,
      "step": 5498
    },
    {
      "epoch": 1.57,
      "grad_norm": 16.05520248413086,
      "learning_rate": 6.405892448512587e-06,
      "loss": 0.8272,
      "step": 5499
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.909257888793945,
      "learning_rate": 6.401601830663616e-06,
      "loss": 0.8576,
      "step": 5500
    },
    {
      "epoch": 1.57,
      "grad_norm": 9.228020668029785,
      "learning_rate": 6.397311212814645e-06,
      "loss": 0.4105,
      "step": 5501
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.026965141296387,
      "learning_rate": 6.393020594965675e-06,
      "loss": 0.5532,
      "step": 5502
    },
    {
      "epoch": 1.57,
      "grad_norm": 12.771260261535645,
      "learning_rate": 6.388729977116704e-06,
      "loss": 0.7664,
      "step": 5503
    },
    {
      "epoch": 1.57,
      "grad_norm": 11.88595962524414,
      "learning_rate": 6.384439359267735e-06,
      "loss": 0.8299,
      "step": 5504
    },
    {
      "epoch": 1.57,
      "grad_norm": 9.56600570678711,
      "learning_rate": 6.380148741418764e-06,
      "loss": 0.7204,
      "step": 5505
    },
    {
      "epoch": 1.57,
      "grad_norm": 9.454607009887695,
      "learning_rate": 6.375858123569794e-06,
      "loss": 0.5083,
      "step": 5506
    },
    {
      "epoch": 1.58,
      "grad_norm": 11.280576705932617,
      "learning_rate": 6.371567505720824e-06,
      "loss": 0.7152,
      "step": 5507
    },
    {
      "epoch": 1.58,
      "grad_norm": 10.759710311889648,
      "learning_rate": 6.367276887871854e-06,
      "loss": 0.8317,
      "step": 5508
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.474405288696289,
      "learning_rate": 6.362986270022884e-06,
      "loss": 0.7682,
      "step": 5509
    },
    {
      "epoch": 1.58,
      "grad_norm": 11.654435157775879,
      "learning_rate": 6.358695652173913e-06,
      "loss": 0.7855,
      "step": 5510
    },
    {
      "epoch": 1.58,
      "grad_norm": 11.794870376586914,
      "learning_rate": 6.354405034324943e-06,
      "loss": 0.7901,
      "step": 5511
    },
    {
      "epoch": 1.58,
      "grad_norm": 11.223348617553711,
      "learning_rate": 6.350114416475973e-06,
      "loss": 0.6573,
      "step": 5512
    },
    {
      "epoch": 1.58,
      "grad_norm": 7.961227893829346,
      "learning_rate": 6.345823798627003e-06,
      "loss": 0.5521,
      "step": 5513
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.095694541931152,
      "learning_rate": 6.341533180778032e-06,
      "loss": 0.7054,
      "step": 5514
    },
    {
      "epoch": 1.58,
      "grad_norm": 10.895547866821289,
      "learning_rate": 6.337242562929061e-06,
      "loss": 0.9474,
      "step": 5515
    },
    {
      "epoch": 1.58,
      "grad_norm": 12.356182098388672,
      "learning_rate": 6.332951945080091e-06,
      "loss": 0.7122,
      "step": 5516
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.5460786819458,
      "learning_rate": 6.3286613272311215e-06,
      "loss": 0.6152,
      "step": 5517
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.811149597167969,
      "learning_rate": 6.324370709382151e-06,
      "loss": 0.7623,
      "step": 5518
    },
    {
      "epoch": 1.58,
      "grad_norm": 8.240185737609863,
      "learning_rate": 6.320080091533181e-06,
      "loss": 0.6884,
      "step": 5519
    },
    {
      "epoch": 1.58,
      "grad_norm": 11.257939338684082,
      "learning_rate": 6.31578947368421e-06,
      "loss": 0.481,
      "step": 5520
    },
    {
      "epoch": 1.58,
      "grad_norm": 10.784560203552246,
      "learning_rate": 6.311498855835241e-06,
      "loss": 0.5507,
      "step": 5521
    },
    {
      "epoch": 1.58,
      "grad_norm": 11.940464973449707,
      "learning_rate": 6.30720823798627e-06,
      "loss": 0.7215,
      "step": 5522
    },
    {
      "epoch": 1.58,
      "grad_norm": 11.188111305236816,
      "learning_rate": 6.3029176201373e-06,
      "loss": 0.5997,
      "step": 5523
    },
    {
      "epoch": 1.58,
      "grad_norm": 7.8785223960876465,
      "learning_rate": 6.29862700228833e-06,
      "loss": 0.592,
      "step": 5524
    },
    {
      "epoch": 1.58,
      "grad_norm": 11.687405586242676,
      "learning_rate": 6.29433638443936e-06,
      "loss": 0.679,
      "step": 5525
    },
    {
      "epoch": 1.58,
      "grad_norm": 12.611087799072266,
      "learning_rate": 6.29004576659039e-06,
      "loss": 1.0554,
      "step": 5526
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.295668601989746,
      "learning_rate": 6.2857551487414185e-06,
      "loss": 0.4928,
      "step": 5527
    },
    {
      "epoch": 1.58,
      "grad_norm": 11.037130355834961,
      "learning_rate": 6.281464530892448e-06,
      "loss": 0.6082,
      "step": 5528
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.137391090393066,
      "learning_rate": 6.2771739130434786e-06,
      "loss": 0.5593,
      "step": 5529
    },
    {
      "epoch": 1.58,
      "grad_norm": 12.662400245666504,
      "learning_rate": 6.272883295194508e-06,
      "loss": 0.8509,
      "step": 5530
    },
    {
      "epoch": 1.58,
      "grad_norm": 11.213057518005371,
      "learning_rate": 6.268592677345538e-06,
      "loss": 0.7416,
      "step": 5531
    },
    {
      "epoch": 1.58,
      "grad_norm": 10.6245698928833,
      "learning_rate": 6.264302059496567e-06,
      "loss": 0.5605,
      "step": 5532
    },
    {
      "epoch": 1.58,
      "grad_norm": 8.723930358886719,
      "learning_rate": 6.260011441647597e-06,
      "loss": 0.5844,
      "step": 5533
    },
    {
      "epoch": 1.58,
      "grad_norm": 10.928689002990723,
      "learning_rate": 6.2557208237986275e-06,
      "loss": 0.6429,
      "step": 5534
    },
    {
      "epoch": 1.58,
      "grad_norm": 11.049684524536133,
      "learning_rate": 6.251430205949657e-06,
      "loss": 0.8551,
      "step": 5535
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.772027969360352,
      "learning_rate": 6.247139588100687e-06,
      "loss": 0.7677,
      "step": 5536
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.763443946838379,
      "learning_rate": 6.242848970251716e-06,
      "loss": 0.7714,
      "step": 5537
    },
    {
      "epoch": 1.58,
      "grad_norm": 10.62363338470459,
      "learning_rate": 6.238558352402747e-06,
      "loss": 0.5815,
      "step": 5538
    },
    {
      "epoch": 1.58,
      "grad_norm": 8.02758502960205,
      "learning_rate": 6.234267734553776e-06,
      "loss": 0.5845,
      "step": 5539
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.214056015014648,
      "learning_rate": 6.229977116704805e-06,
      "loss": 0.6295,
      "step": 5540
    },
    {
      "epoch": 1.58,
      "grad_norm": 11.316672325134277,
      "learning_rate": 6.225686498855835e-06,
      "loss": 0.6987,
      "step": 5541
    },
    {
      "epoch": 1.59,
      "grad_norm": 10.765800476074219,
      "learning_rate": 6.221395881006865e-06,
      "loss": 0.6417,
      "step": 5542
    },
    {
      "epoch": 1.59,
      "grad_norm": 9.062914848327637,
      "learning_rate": 6.217105263157895e-06,
      "loss": 0.6407,
      "step": 5543
    },
    {
      "epoch": 1.59,
      "grad_norm": 9.839356422424316,
      "learning_rate": 6.2128146453089245e-06,
      "loss": 0.5741,
      "step": 5544
    },
    {
      "epoch": 1.59,
      "grad_norm": 10.004151344299316,
      "learning_rate": 6.208524027459954e-06,
      "loss": 0.5263,
      "step": 5545
    },
    {
      "epoch": 1.59,
      "grad_norm": 9.841792106628418,
      "learning_rate": 6.204233409610984e-06,
      "loss": 0.624,
      "step": 5546
    },
    {
      "epoch": 1.59,
      "grad_norm": 13.122528076171875,
      "learning_rate": 6.199942791762014e-06,
      "loss": 1.1056,
      "step": 5547
    },
    {
      "epoch": 1.59,
      "grad_norm": 9.106522560119629,
      "learning_rate": 6.195652173913044e-06,
      "loss": 0.6107,
      "step": 5548
    },
    {
      "epoch": 1.59,
      "grad_norm": 9.179657936096191,
      "learning_rate": 6.191361556064073e-06,
      "loss": 0.6582,
      "step": 5549
    },
    {
      "epoch": 1.59,
      "grad_norm": 11.946647644042969,
      "learning_rate": 6.187070938215103e-06,
      "loss": 0.7773,
      "step": 5550
    },
    {
      "epoch": 1.59,
      "grad_norm": 11.384345054626465,
      "learning_rate": 6.1827803203661335e-06,
      "loss": 0.86,
      "step": 5551
    },
    {
      "epoch": 1.59,
      "grad_norm": 10.477992057800293,
      "learning_rate": 6.178489702517163e-06,
      "loss": 0.8906,
      "step": 5552
    },
    {
      "epoch": 1.59,
      "grad_norm": 14.003066062927246,
      "learning_rate": 6.174199084668192e-06,
      "loss": 0.8674,
      "step": 5553
    },
    {
      "epoch": 1.59,
      "grad_norm": 12.793807983398438,
      "learning_rate": 6.1699084668192215e-06,
      "loss": 0.7947,
      "step": 5554
    },
    {
      "epoch": 1.59,
      "grad_norm": 10.329636573791504,
      "learning_rate": 6.165617848970252e-06,
      "loss": 0.9208,
      "step": 5555
    },
    {
      "epoch": 1.59,
      "grad_norm": 10.548229217529297,
      "learning_rate": 6.1613272311212815e-06,
      "loss": 0.6526,
      "step": 5556
    },
    {
      "epoch": 1.59,
      "grad_norm": 9.881653785705566,
      "learning_rate": 6.157036613272311e-06,
      "loss": 0.7158,
      "step": 5557
    },
    {
      "epoch": 1.59,
      "grad_norm": 10.707265853881836,
      "learning_rate": 6.152745995423341e-06,
      "loss": 0.6928,
      "step": 5558
    },
    {
      "epoch": 1.59,
      "grad_norm": 9.42878532409668,
      "learning_rate": 6.148455377574371e-06,
      "loss": 0.6004,
      "step": 5559
    },
    {
      "epoch": 1.59,
      "grad_norm": 11.764908790588379,
      "learning_rate": 6.144164759725401e-06,
      "loss": 0.8836,
      "step": 5560
    },
    {
      "epoch": 1.59,
      "grad_norm": 10.387475967407227,
      "learning_rate": 6.1398741418764305e-06,
      "loss": 0.7904,
      "step": 5561
    },
    {
      "epoch": 1.59,
      "grad_norm": 11.080256462097168,
      "learning_rate": 6.13558352402746e-06,
      "loss": 0.6863,
      "step": 5562
    },
    {
      "epoch": 1.59,
      "grad_norm": 10.54133129119873,
      "learning_rate": 6.13129290617849e-06,
      "loss": 0.8626,
      "step": 5563
    },
    {
      "epoch": 1.59,
      "grad_norm": 9.514528274536133,
      "learning_rate": 6.12700228832952e-06,
      "loss": 0.6721,
      "step": 5564
    },
    {
      "epoch": 1.59,
      "grad_norm": 12.685568809509277,
      "learning_rate": 6.12271167048055e-06,
      "loss": 0.9852,
      "step": 5565
    },
    {
      "epoch": 1.59,
      "grad_norm": 10.1242036819458,
      "learning_rate": 6.118421052631579e-06,
      "loss": 0.7792,
      "step": 5566
    },
    {
      "epoch": 1.59,
      "grad_norm": 14.765839576721191,
      "learning_rate": 6.114130434782608e-06,
      "loss": 0.8978,
      "step": 5567
    },
    {
      "epoch": 1.59,
      "grad_norm": 9.789997100830078,
      "learning_rate": 6.109839816933639e-06,
      "loss": 0.6368,
      "step": 5568
    },
    {
      "epoch": 1.59,
      "grad_norm": 8.29202938079834,
      "learning_rate": 6.105549199084668e-06,
      "loss": 0.455,
      "step": 5569
    },
    {
      "epoch": 1.59,
      "grad_norm": 10.819106101989746,
      "learning_rate": 6.101258581235698e-06,
      "loss": 0.6887,
      "step": 5570
    },
    {
      "epoch": 1.59,
      "grad_norm": 10.95679759979248,
      "learning_rate": 6.0969679633867275e-06,
      "loss": 0.792,
      "step": 5571
    },
    {
      "epoch": 1.59,
      "grad_norm": 11.622454643249512,
      "learning_rate": 6.092677345537758e-06,
      "loss": 0.6522,
      "step": 5572
    },
    {
      "epoch": 1.59,
      "grad_norm": 9.316247940063477,
      "learning_rate": 6.0883867276887875e-06,
      "loss": 0.5935,
      "step": 5573
    },
    {
      "epoch": 1.59,
      "grad_norm": 11.921667098999023,
      "learning_rate": 6.084096109839817e-06,
      "loss": 0.7116,
      "step": 5574
    },
    {
      "epoch": 1.59,
      "grad_norm": 11.171442031860352,
      "learning_rate": 6.079805491990847e-06,
      "loss": 0.8473,
      "step": 5575
    },
    {
      "epoch": 1.59,
      "grad_norm": 17.87783432006836,
      "learning_rate": 6.075514874141876e-06,
      "loss": 0.8636,
      "step": 5576
    },
    {
      "epoch": 1.6,
      "grad_norm": 9.572556495666504,
      "learning_rate": 6.071224256292907e-06,
      "loss": 0.6621,
      "step": 5577
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.472742080688477,
      "learning_rate": 6.0669336384439365e-06,
      "loss": 0.7571,
      "step": 5578
    },
    {
      "epoch": 1.6,
      "grad_norm": 9.450839042663574,
      "learning_rate": 6.062643020594966e-06,
      "loss": 0.5779,
      "step": 5579
    },
    {
      "epoch": 1.6,
      "grad_norm": 12.35782527923584,
      "learning_rate": 6.058352402745995e-06,
      "loss": 0.7953,
      "step": 5580
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.768641471862793,
      "learning_rate": 6.054061784897025e-06,
      "loss": 0.8999,
      "step": 5581
    },
    {
      "epoch": 1.6,
      "grad_norm": 11.072747230529785,
      "learning_rate": 6.049771167048055e-06,
      "loss": 0.5263,
      "step": 5582
    },
    {
      "epoch": 1.6,
      "grad_norm": 8.882772445678711,
      "learning_rate": 6.0454805491990845e-06,
      "loss": 0.6575,
      "step": 5583
    },
    {
      "epoch": 1.6,
      "grad_norm": 11.697036743164062,
      "learning_rate": 6.041189931350114e-06,
      "loss": 0.9594,
      "step": 5584
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.463035583496094,
      "learning_rate": 6.036899313501145e-06,
      "loss": 0.6768,
      "step": 5585
    },
    {
      "epoch": 1.6,
      "grad_norm": 11.004870414733887,
      "learning_rate": 6.032608695652174e-06,
      "loss": 0.5053,
      "step": 5586
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.203104972839355,
      "learning_rate": 6.028318077803204e-06,
      "loss": 0.5423,
      "step": 5587
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.33894157409668,
      "learning_rate": 6.0240274599542335e-06,
      "loss": 0.6638,
      "step": 5588
    },
    {
      "epoch": 1.6,
      "grad_norm": 9.002959251403809,
      "learning_rate": 6.019736842105264e-06,
      "loss": 0.7116,
      "step": 5589
    },
    {
      "epoch": 1.6,
      "grad_norm": 9.3934907913208,
      "learning_rate": 6.0154462242562935e-06,
      "loss": 0.6622,
      "step": 5590
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.514039993286133,
      "learning_rate": 6.011155606407323e-06,
      "loss": 0.7518,
      "step": 5591
    },
    {
      "epoch": 1.6,
      "grad_norm": 9.802041053771973,
      "learning_rate": 6.006864988558353e-06,
      "loss": 0.9095,
      "step": 5592
    },
    {
      "epoch": 1.6,
      "grad_norm": 9.759675979614258,
      "learning_rate": 6.0025743707093815e-06,
      "loss": 0.7053,
      "step": 5593
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.36616039276123,
      "learning_rate": 5.998283752860412e-06,
      "loss": 0.707,
      "step": 5594
    },
    {
      "epoch": 1.6,
      "grad_norm": 9.417503356933594,
      "learning_rate": 5.993993135011442e-06,
      "loss": 0.7954,
      "step": 5595
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.612123489379883,
      "learning_rate": 5.989702517162471e-06,
      "loss": 0.5458,
      "step": 5596
    },
    {
      "epoch": 1.6,
      "grad_norm": 11.943852424621582,
      "learning_rate": 5.985411899313501e-06,
      "loss": 0.782,
      "step": 5597
    },
    {
      "epoch": 1.6,
      "grad_norm": 9.845441818237305,
      "learning_rate": 5.981121281464531e-06,
      "loss": 0.6633,
      "step": 5598
    },
    {
      "epoch": 1.6,
      "grad_norm": 9.53563117980957,
      "learning_rate": 5.976830663615561e-06,
      "loss": 0.7448,
      "step": 5599
    },
    {
      "epoch": 1.6,
      "grad_norm": 13.612221717834473,
      "learning_rate": 5.9725400457665905e-06,
      "loss": 0.9435,
      "step": 5600
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.986820220947266,
      "learning_rate": 5.96824942791762e-06,
      "loss": 0.593,
      "step": 5601
    },
    {
      "epoch": 1.6,
      "grad_norm": 12.078536033630371,
      "learning_rate": 5.963958810068651e-06,
      "loss": 0.7624,
      "step": 5602
    },
    {
      "epoch": 1.6,
      "grad_norm": 9.601693153381348,
      "learning_rate": 5.95966819221968e-06,
      "loss": 0.5876,
      "step": 5603
    },
    {
      "epoch": 1.6,
      "grad_norm": 11.859810829162598,
      "learning_rate": 5.95537757437071e-06,
      "loss": 0.5815,
      "step": 5604
    },
    {
      "epoch": 1.6,
      "grad_norm": 8.395353317260742,
      "learning_rate": 5.9510869565217395e-06,
      "loss": 0.5982,
      "step": 5605
    },
    {
      "epoch": 1.6,
      "grad_norm": 11.383386611938477,
      "learning_rate": 5.946796338672768e-06,
      "loss": 0.8622,
      "step": 5606
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.532876968383789,
      "learning_rate": 5.942505720823799e-06,
      "loss": 0.6901,
      "step": 5607
    },
    {
      "epoch": 1.6,
      "grad_norm": 12.721446990966797,
      "learning_rate": 5.938215102974828e-06,
      "loss": 0.5373,
      "step": 5608
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.015861511230469,
      "learning_rate": 5.933924485125858e-06,
      "loss": 0.6831,
      "step": 5609
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.524345397949219,
      "learning_rate": 5.9296338672768875e-06,
      "loss": 0.5984,
      "step": 5610
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.858034133911133,
      "learning_rate": 5.925343249427918e-06,
      "loss": 0.608,
      "step": 5611
    },
    {
      "epoch": 1.61,
      "grad_norm": 8.153266906738281,
      "learning_rate": 5.921052631578948e-06,
      "loss": 0.5376,
      "step": 5612
    },
    {
      "epoch": 1.61,
      "grad_norm": 7.881692886352539,
      "learning_rate": 5.916762013729977e-06,
      "loss": 0.6893,
      "step": 5613
    },
    {
      "epoch": 1.61,
      "grad_norm": 14.807575225830078,
      "learning_rate": 5.912471395881007e-06,
      "loss": 0.9757,
      "step": 5614
    },
    {
      "epoch": 1.61,
      "grad_norm": 12.7268705368042,
      "learning_rate": 5.908180778032037e-06,
      "loss": 0.9282,
      "step": 5615
    },
    {
      "epoch": 1.61,
      "grad_norm": 8.834338188171387,
      "learning_rate": 5.903890160183067e-06,
      "loss": 0.6952,
      "step": 5616
    },
    {
      "epoch": 1.61,
      "grad_norm": 11.114319801330566,
      "learning_rate": 5.8995995423340965e-06,
      "loss": 0.9001,
      "step": 5617
    },
    {
      "epoch": 1.61,
      "grad_norm": 10.926636695861816,
      "learning_rate": 5.895308924485126e-06,
      "loss": 0.4992,
      "step": 5618
    },
    {
      "epoch": 1.61,
      "grad_norm": 12.872359275817871,
      "learning_rate": 5.891018306636156e-06,
      "loss": 0.9402,
      "step": 5619
    },
    {
      "epoch": 1.61,
      "grad_norm": 12.638120651245117,
      "learning_rate": 5.886727688787185e-06,
      "loss": 0.8616,
      "step": 5620
    },
    {
      "epoch": 1.61,
      "grad_norm": 11.012920379638672,
      "learning_rate": 5.882437070938215e-06,
      "loss": 0.83,
      "step": 5621
    },
    {
      "epoch": 1.61,
      "grad_norm": 12.366905212402344,
      "learning_rate": 5.878146453089245e-06,
      "loss": 0.7689,
      "step": 5622
    },
    {
      "epoch": 1.61,
      "grad_norm": 12.612177848815918,
      "learning_rate": 5.873855835240274e-06,
      "loss": 0.937,
      "step": 5623
    },
    {
      "epoch": 1.61,
      "grad_norm": 7.9943718910217285,
      "learning_rate": 5.869565217391305e-06,
      "loss": 0.5223,
      "step": 5624
    },
    {
      "epoch": 1.61,
      "grad_norm": 10.88665771484375,
      "learning_rate": 5.865274599542334e-06,
      "loss": 0.5805,
      "step": 5625
    },
    {
      "epoch": 1.61,
      "grad_norm": 8.64239501953125,
      "learning_rate": 5.860983981693364e-06,
      "loss": 0.5296,
      "step": 5626
    },
    {
      "epoch": 1.61,
      "grad_norm": 9.842996597290039,
      "learning_rate": 5.8566933638443935e-06,
      "loss": 0.5918,
      "step": 5627
    },
    {
      "epoch": 1.61,
      "grad_norm": 8.364131927490234,
      "learning_rate": 5.852402745995424e-06,
      "loss": 0.5218,
      "step": 5628
    },
    {
      "epoch": 1.61,
      "grad_norm": 12.647870063781738,
      "learning_rate": 5.848112128146454e-06,
      "loss": 0.7731,
      "step": 5629
    },
    {
      "epoch": 1.61,
      "grad_norm": 11.33095645904541,
      "learning_rate": 5.843821510297483e-06,
      "loss": 0.8719,
      "step": 5630
    },
    {
      "epoch": 1.61,
      "grad_norm": 9.10512924194336,
      "learning_rate": 5.839530892448513e-06,
      "loss": 0.4263,
      "step": 5631
    },
    {
      "epoch": 1.61,
      "grad_norm": 8.400649070739746,
      "learning_rate": 5.8352402745995424e-06,
      "loss": 0.6073,
      "step": 5632
    },
    {
      "epoch": 1.61,
      "grad_norm": 11.631400108337402,
      "learning_rate": 5.830949656750572e-06,
      "loss": 0.7929,
      "step": 5633
    },
    {
      "epoch": 1.61,
      "grad_norm": 11.168889045715332,
      "learning_rate": 5.826659038901602e-06,
      "loss": 0.7255,
      "step": 5634
    },
    {
      "epoch": 1.61,
      "grad_norm": 11.418717384338379,
      "learning_rate": 5.822368421052631e-06,
      "loss": 0.5485,
      "step": 5635
    },
    {
      "epoch": 1.61,
      "grad_norm": 14.530598640441895,
      "learning_rate": 5.818077803203661e-06,
      "loss": 0.7852,
      "step": 5636
    },
    {
      "epoch": 1.61,
      "grad_norm": 11.886549949645996,
      "learning_rate": 5.813787185354691e-06,
      "loss": 0.7901,
      "step": 5637
    },
    {
      "epoch": 1.61,
      "grad_norm": 14.043811798095703,
      "learning_rate": 5.809496567505721e-06,
      "loss": 1.0616,
      "step": 5638
    },
    {
      "epoch": 1.61,
      "grad_norm": 9.342390060424805,
      "learning_rate": 5.805205949656751e-06,
      "loss": 0.6468,
      "step": 5639
    },
    {
      "epoch": 1.61,
      "grad_norm": 9.720380783081055,
      "learning_rate": 5.80091533180778e-06,
      "loss": 0.6197,
      "step": 5640
    },
    {
      "epoch": 1.61,
      "grad_norm": 11.170440673828125,
      "learning_rate": 5.796624713958811e-06,
      "loss": 0.7108,
      "step": 5641
    },
    {
      "epoch": 1.61,
      "grad_norm": 9.893743515014648,
      "learning_rate": 5.79233409610984e-06,
      "loss": 0.7752,
      "step": 5642
    },
    {
      "epoch": 1.61,
      "grad_norm": 12.05979061126709,
      "learning_rate": 5.78804347826087e-06,
      "loss": 0.53,
      "step": 5643
    },
    {
      "epoch": 1.61,
      "grad_norm": 9.423600196838379,
      "learning_rate": 5.7837528604118995e-06,
      "loss": 0.7921,
      "step": 5644
    },
    {
      "epoch": 1.61,
      "grad_norm": 12.006379127502441,
      "learning_rate": 5.779462242562929e-06,
      "loss": 0.7143,
      "step": 5645
    },
    {
      "epoch": 1.61,
      "grad_norm": 13.299084663391113,
      "learning_rate": 5.775171624713959e-06,
      "loss": 0.7702,
      "step": 5646
    },
    {
      "epoch": 1.62,
      "grad_norm": 10.986786842346191,
      "learning_rate": 5.770881006864988e-06,
      "loss": 0.6752,
      "step": 5647
    },
    {
      "epoch": 1.62,
      "grad_norm": 10.160089492797852,
      "learning_rate": 5.766590389016018e-06,
      "loss": 0.6384,
      "step": 5648
    },
    {
      "epoch": 1.62,
      "grad_norm": 10.388351440429688,
      "learning_rate": 5.7622997711670484e-06,
      "loss": 0.8869,
      "step": 5649
    },
    {
      "epoch": 1.62,
      "grad_norm": 11.499975204467773,
      "learning_rate": 5.758009153318078e-06,
      "loss": 0.7715,
      "step": 5650
    },
    {
      "epoch": 1.62,
      "grad_norm": 11.53027629852295,
      "learning_rate": 5.753718535469108e-06,
      "loss": 0.6374,
      "step": 5651
    },
    {
      "epoch": 1.62,
      "grad_norm": 9.61728286743164,
      "learning_rate": 5.749427917620137e-06,
      "loss": 0.4762,
      "step": 5652
    },
    {
      "epoch": 1.62,
      "grad_norm": 10.64439582824707,
      "learning_rate": 5.745137299771167e-06,
      "loss": 0.874,
      "step": 5653
    },
    {
      "epoch": 1.62,
      "grad_norm": 10.544897079467773,
      "learning_rate": 5.740846681922197e-06,
      "loss": 0.7644,
      "step": 5654
    },
    {
      "epoch": 1.62,
      "grad_norm": 11.613075256347656,
      "learning_rate": 5.736556064073227e-06,
      "loss": 0.5905,
      "step": 5655
    },
    {
      "epoch": 1.62,
      "grad_norm": 13.937162399291992,
      "learning_rate": 5.732265446224257e-06,
      "loss": 0.8385,
      "step": 5656
    },
    {
      "epoch": 1.62,
      "grad_norm": 11.547974586486816,
      "learning_rate": 5.727974828375286e-06,
      "loss": 0.788,
      "step": 5657
    },
    {
      "epoch": 1.62,
      "grad_norm": 10.030085563659668,
      "learning_rate": 5.723684210526316e-06,
      "loss": 0.6463,
      "step": 5658
    },
    {
      "epoch": 1.62,
      "grad_norm": 11.306310653686523,
      "learning_rate": 5.7193935926773454e-06,
      "loss": 0.7094,
      "step": 5659
    },
    {
      "epoch": 1.62,
      "grad_norm": 12.70578670501709,
      "learning_rate": 5.715102974828375e-06,
      "loss": 0.7308,
      "step": 5660
    },
    {
      "epoch": 1.62,
      "grad_norm": 10.362273216247559,
      "learning_rate": 5.710812356979405e-06,
      "loss": 0.6183,
      "step": 5661
    },
    {
      "epoch": 1.62,
      "grad_norm": 11.30966854095459,
      "learning_rate": 5.706521739130435e-06,
      "loss": 0.561,
      "step": 5662
    },
    {
      "epoch": 1.62,
      "grad_norm": 9.392801284790039,
      "learning_rate": 5.702231121281465e-06,
      "loss": 0.6767,
      "step": 5663
    },
    {
      "epoch": 1.62,
      "grad_norm": 9.160632133483887,
      "learning_rate": 5.697940503432494e-06,
      "loss": 0.7882,
      "step": 5664
    },
    {
      "epoch": 1.62,
      "grad_norm": 9.824238777160645,
      "learning_rate": 5.693649885583524e-06,
      "loss": 0.5345,
      "step": 5665
    },
    {
      "epoch": 1.62,
      "grad_norm": 9.859192848205566,
      "learning_rate": 5.689359267734554e-06,
      "loss": 0.5978,
      "step": 5666
    },
    {
      "epoch": 1.62,
      "grad_norm": 10.57384967803955,
      "learning_rate": 5.685068649885584e-06,
      "loss": 0.6203,
      "step": 5667
    },
    {
      "epoch": 1.62,
      "grad_norm": 8.914021492004395,
      "learning_rate": 5.680778032036614e-06,
      "loss": 0.6637,
      "step": 5668
    },
    {
      "epoch": 1.62,
      "grad_norm": 9.800065994262695,
      "learning_rate": 5.676487414187643e-06,
      "loss": 0.6351,
      "step": 5669
    },
    {
      "epoch": 1.62,
      "grad_norm": 10.23118782043457,
      "learning_rate": 5.672196796338673e-06,
      "loss": 0.678,
      "step": 5670
    },
    {
      "epoch": 1.62,
      "grad_norm": 8.032459259033203,
      "learning_rate": 5.667906178489703e-06,
      "loss": 0.5446,
      "step": 5671
    },
    {
      "epoch": 1.62,
      "grad_norm": 10.34920883178711,
      "learning_rate": 5.663615560640732e-06,
      "loss": 0.8721,
      "step": 5672
    },
    {
      "epoch": 1.62,
      "grad_norm": 10.661069869995117,
      "learning_rate": 5.659324942791762e-06,
      "loss": 0.9113,
      "step": 5673
    },
    {
      "epoch": 1.62,
      "grad_norm": 10.502165794372559,
      "learning_rate": 5.655034324942791e-06,
      "loss": 0.9143,
      "step": 5674
    },
    {
      "epoch": 1.62,
      "grad_norm": 10.990413665771484,
      "learning_rate": 5.650743707093822e-06,
      "loss": 0.8208,
      "step": 5675
    },
    {
      "epoch": 1.62,
      "grad_norm": 9.408350944519043,
      "learning_rate": 5.6464530892448514e-06,
      "loss": 0.6148,
      "step": 5676
    },
    {
      "epoch": 1.62,
      "grad_norm": 10.625914573669434,
      "learning_rate": 5.642162471395881e-06,
      "loss": 0.6522,
      "step": 5677
    },
    {
      "epoch": 1.62,
      "grad_norm": 9.551958084106445,
      "learning_rate": 5.637871853546911e-06,
      "loss": 0.6935,
      "step": 5678
    },
    {
      "epoch": 1.62,
      "grad_norm": 13.339814186096191,
      "learning_rate": 5.633581235697941e-06,
      "loss": 0.5454,
      "step": 5679
    },
    {
      "epoch": 1.62,
      "grad_norm": 11.620394706726074,
      "learning_rate": 5.629290617848971e-06,
      "loss": 0.5537,
      "step": 5680
    },
    {
      "epoch": 1.62,
      "grad_norm": 8.314926147460938,
      "learning_rate": 5.625e-06,
      "loss": 0.6268,
      "step": 5681
    },
    {
      "epoch": 1.63,
      "grad_norm": 8.93075180053711,
      "learning_rate": 5.62070938215103e-06,
      "loss": 0.6914,
      "step": 5682
    },
    {
      "epoch": 1.63,
      "grad_norm": 10.87057113647461,
      "learning_rate": 5.61641876430206e-06,
      "loss": 0.6301,
      "step": 5683
    },
    {
      "epoch": 1.63,
      "grad_norm": 10.709811210632324,
      "learning_rate": 5.61212814645309e-06,
      "loss": 0.5426,
      "step": 5684
    },
    {
      "epoch": 1.63,
      "grad_norm": 10.641079902648926,
      "learning_rate": 5.607837528604119e-06,
      "loss": 0.9742,
      "step": 5685
    },
    {
      "epoch": 1.63,
      "grad_norm": 12.158310890197754,
      "learning_rate": 5.6035469107551484e-06,
      "loss": 0.7012,
      "step": 5686
    },
    {
      "epoch": 1.63,
      "grad_norm": 12.088842391967773,
      "learning_rate": 5.599256292906178e-06,
      "loss": 0.9431,
      "step": 5687
    },
    {
      "epoch": 1.63,
      "grad_norm": 9.152239799499512,
      "learning_rate": 5.5949656750572085e-06,
      "loss": 0.6064,
      "step": 5688
    },
    {
      "epoch": 1.63,
      "grad_norm": 8.571259498596191,
      "learning_rate": 5.590675057208238e-06,
      "loss": 0.5959,
      "step": 5689
    },
    {
      "epoch": 1.63,
      "grad_norm": 12.594555854797363,
      "learning_rate": 5.586384439359268e-06,
      "loss": 0.8432,
      "step": 5690
    },
    {
      "epoch": 1.63,
      "grad_norm": 7.2420654296875,
      "learning_rate": 5.582093821510297e-06,
      "loss": 0.5177,
      "step": 5691
    },
    {
      "epoch": 1.63,
      "grad_norm": 9.37486457824707,
      "learning_rate": 5.577803203661328e-06,
      "loss": 0.7414,
      "step": 5692
    },
    {
      "epoch": 1.63,
      "grad_norm": 11.70945930480957,
      "learning_rate": 5.5735125858123574e-06,
      "loss": 0.6451,
      "step": 5693
    },
    {
      "epoch": 1.63,
      "grad_norm": 9.084478378295898,
      "learning_rate": 5.569221967963387e-06,
      "loss": 0.5428,
      "step": 5694
    },
    {
      "epoch": 1.63,
      "grad_norm": 10.824153900146484,
      "learning_rate": 5.564931350114417e-06,
      "loss": 0.7393,
      "step": 5695
    },
    {
      "epoch": 1.63,
      "grad_norm": 11.970434188842773,
      "learning_rate": 5.560640732265447e-06,
      "loss": 0.8567,
      "step": 5696
    },
    {
      "epoch": 1.63,
      "grad_norm": 16.701438903808594,
      "learning_rate": 5.556350114416477e-06,
      "loss": 0.7511,
      "step": 5697
    },
    {
      "epoch": 1.63,
      "grad_norm": 11.472614288330078,
      "learning_rate": 5.5520594965675055e-06,
      "loss": 1.0356,
      "step": 5698
    },
    {
      "epoch": 1.63,
      "grad_norm": 8.926335334777832,
      "learning_rate": 5.547768878718535e-06,
      "loss": 0.6318,
      "step": 5699
    },
    {
      "epoch": 1.63,
      "grad_norm": 9.65076732635498,
      "learning_rate": 5.543478260869565e-06,
      "loss": 0.6234,
      "step": 5700
    },
    {
      "epoch": 1.63,
      "grad_norm": 9.989914894104004,
      "learning_rate": 5.539187643020595e-06,
      "loss": 0.5944,
      "step": 5701
    },
    {
      "epoch": 1.63,
      "grad_norm": 13.871895790100098,
      "learning_rate": 5.534897025171625e-06,
      "loss": 0.7599,
      "step": 5702
    },
    {
      "epoch": 1.63,
      "grad_norm": 9.64797592163086,
      "learning_rate": 5.5306064073226544e-06,
      "loss": 0.5855,
      "step": 5703
    },
    {
      "epoch": 1.63,
      "grad_norm": 12.163445472717285,
      "learning_rate": 5.526315789473684e-06,
      "loss": 0.7724,
      "step": 5704
    },
    {
      "epoch": 1.63,
      "grad_norm": 8.04688835144043,
      "learning_rate": 5.5220251716247145e-06,
      "loss": 0.6141,
      "step": 5705
    },
    {
      "epoch": 1.63,
      "grad_norm": 8.863974571228027,
      "learning_rate": 5.517734553775744e-06,
      "loss": 0.7573,
      "step": 5706
    },
    {
      "epoch": 1.63,
      "grad_norm": 11.304844856262207,
      "learning_rate": 5.513443935926774e-06,
      "loss": 0.9782,
      "step": 5707
    },
    {
      "epoch": 1.63,
      "grad_norm": 9.068216323852539,
      "learning_rate": 5.509153318077803e-06,
      "loss": 0.7531,
      "step": 5708
    },
    {
      "epoch": 1.63,
      "grad_norm": 10.172918319702148,
      "learning_rate": 5.504862700228834e-06,
      "loss": 0.6192,
      "step": 5709
    },
    {
      "epoch": 1.63,
      "grad_norm": 12.894253730773926,
      "learning_rate": 5.500572082379863e-06,
      "loss": 0.9341,
      "step": 5710
    },
    {
      "epoch": 1.63,
      "grad_norm": 13.358455657958984,
      "learning_rate": 5.496281464530892e-06,
      "loss": 0.8522,
      "step": 5711
    },
    {
      "epoch": 1.63,
      "grad_norm": 10.213536262512207,
      "learning_rate": 5.491990846681922e-06,
      "loss": 0.6582,
      "step": 5712
    },
    {
      "epoch": 1.63,
      "grad_norm": 12.736071586608887,
      "learning_rate": 5.487700228832951e-06,
      "loss": 0.6622,
      "step": 5713
    },
    {
      "epoch": 1.63,
      "grad_norm": 10.614320755004883,
      "learning_rate": 5.483409610983982e-06,
      "loss": 0.7729,
      "step": 5714
    },
    {
      "epoch": 1.63,
      "grad_norm": 10.915679931640625,
      "learning_rate": 5.4791189931350115e-06,
      "loss": 0.5835,
      "step": 5715
    },
    {
      "epoch": 1.64,
      "grad_norm": 10.88956356048584,
      "learning_rate": 5.474828375286041e-06,
      "loss": 0.7071,
      "step": 5716
    },
    {
      "epoch": 1.64,
      "grad_norm": 8.267797470092773,
      "learning_rate": 5.470537757437071e-06,
      "loss": 0.5224,
      "step": 5717
    },
    {
      "epoch": 1.64,
      "grad_norm": 8.929670333862305,
      "learning_rate": 5.466247139588101e-06,
      "loss": 0.4938,
      "step": 5718
    },
    {
      "epoch": 1.64,
      "grad_norm": 9.463593482971191,
      "learning_rate": 5.461956521739131e-06,
      "loss": 0.6707,
      "step": 5719
    },
    {
      "epoch": 1.64,
      "grad_norm": 7.468658924102783,
      "learning_rate": 5.45766590389016e-06,
      "loss": 0.5084,
      "step": 5720
    },
    {
      "epoch": 1.64,
      "grad_norm": 9.18384838104248,
      "learning_rate": 5.45337528604119e-06,
      "loss": 0.4628,
      "step": 5721
    },
    {
      "epoch": 1.64,
      "grad_norm": 11.797301292419434,
      "learning_rate": 5.4490846681922205e-06,
      "loss": 0.7877,
      "step": 5722
    },
    {
      "epoch": 1.64,
      "grad_norm": 10.957298278808594,
      "learning_rate": 5.44479405034325e-06,
      "loss": 0.9386,
      "step": 5723
    },
    {
      "epoch": 1.64,
      "grad_norm": 11.238849639892578,
      "learning_rate": 5.440503432494279e-06,
      "loss": 0.5708,
      "step": 5724
    },
    {
      "epoch": 1.64,
      "grad_norm": 11.021093368530273,
      "learning_rate": 5.4362128146453085e-06,
      "loss": 0.6834,
      "step": 5725
    },
    {
      "epoch": 1.64,
      "grad_norm": 9.856903076171875,
      "learning_rate": 5.431922196796339e-06,
      "loss": 0.5821,
      "step": 5726
    },
    {
      "epoch": 1.64,
      "grad_norm": 11.011984825134277,
      "learning_rate": 5.4276315789473686e-06,
      "loss": 0.7455,
      "step": 5727
    },
    {
      "epoch": 1.64,
      "grad_norm": 9.175439834594727,
      "learning_rate": 5.423340961098398e-06,
      "loss": 0.6383,
      "step": 5728
    },
    {
      "epoch": 1.64,
      "grad_norm": 11.727838516235352,
      "learning_rate": 5.419050343249428e-06,
      "loss": 0.6758,
      "step": 5729
    },
    {
      "epoch": 1.64,
      "grad_norm": 10.31438159942627,
      "learning_rate": 5.414759725400457e-06,
      "loss": 0.7227,
      "step": 5730
    },
    {
      "epoch": 1.64,
      "grad_norm": 10.410078048706055,
      "learning_rate": 5.410469107551488e-06,
      "loss": 0.6846,
      "step": 5731
    },
    {
      "epoch": 1.64,
      "grad_norm": 10.888920783996582,
      "learning_rate": 5.4061784897025175e-06,
      "loss": 0.6959,
      "step": 5732
    },
    {
      "epoch": 1.64,
      "grad_norm": 11.906867980957031,
      "learning_rate": 5.401887871853547e-06,
      "loss": 0.8296,
      "step": 5733
    },
    {
      "epoch": 1.64,
      "grad_norm": 8.357414245605469,
      "learning_rate": 5.397597254004577e-06,
      "loss": 0.4034,
      "step": 5734
    },
    {
      "epoch": 1.64,
      "grad_norm": 11.187140464782715,
      "learning_rate": 5.393306636155607e-06,
      "loss": 0.716,
      "step": 5735
    },
    {
      "epoch": 1.64,
      "grad_norm": 7.655837059020996,
      "learning_rate": 5.389016018306637e-06,
      "loss": 0.7081,
      "step": 5736
    },
    {
      "epoch": 1.64,
      "grad_norm": 10.159165382385254,
      "learning_rate": 5.3847254004576656e-06,
      "loss": 0.6526,
      "step": 5737
    },
    {
      "epoch": 1.64,
      "grad_norm": 9.793972969055176,
      "learning_rate": 5.380434782608695e-06,
      "loss": 0.7996,
      "step": 5738
    },
    {
      "epoch": 1.64,
      "grad_norm": 11.443796157836914,
      "learning_rate": 5.376144164759726e-06,
      "loss": 0.6855,
      "step": 5739
    },
    {
      "epoch": 1.64,
      "grad_norm": 9.832064628601074,
      "learning_rate": 5.371853546910755e-06,
      "loss": 0.7701,
      "step": 5740
    },
    {
      "epoch": 1.64,
      "grad_norm": 13.623041152954102,
      "learning_rate": 5.367562929061785e-06,
      "loss": 1.0157,
      "step": 5741
    },
    {
      "epoch": 1.64,
      "grad_norm": 10.20119571685791,
      "learning_rate": 5.3632723112128145e-06,
      "loss": 0.628,
      "step": 5742
    },
    {
      "epoch": 1.64,
      "grad_norm": 9.81611442565918,
      "learning_rate": 5.358981693363844e-06,
      "loss": 0.5421,
      "step": 5743
    },
    {
      "epoch": 1.64,
      "grad_norm": 8.950174331665039,
      "learning_rate": 5.3546910755148746e-06,
      "loss": 0.5966,
      "step": 5744
    },
    {
      "epoch": 1.64,
      "grad_norm": 9.24271011352539,
      "learning_rate": 5.350400457665904e-06,
      "loss": 0.6289,
      "step": 5745
    },
    {
      "epoch": 1.64,
      "grad_norm": 11.46676254272461,
      "learning_rate": 5.346109839816934e-06,
      "loss": 0.5403,
      "step": 5746
    },
    {
      "epoch": 1.64,
      "grad_norm": 8.326277732849121,
      "learning_rate": 5.341819221967963e-06,
      "loss": 0.4409,
      "step": 5747
    },
    {
      "epoch": 1.64,
      "grad_norm": 11.634745597839355,
      "learning_rate": 5.337528604118994e-06,
      "loss": 0.6893,
      "step": 5748
    },
    {
      "epoch": 1.64,
      "grad_norm": 8.160537719726562,
      "learning_rate": 5.3332379862700235e-06,
      "loss": 0.6457,
      "step": 5749
    },
    {
      "epoch": 1.64,
      "grad_norm": 12.94668197631836,
      "learning_rate": 5.328947368421053e-06,
      "loss": 0.5905,
      "step": 5750
    },
    {
      "epoch": 1.65,
      "grad_norm": 11.275633811950684,
      "learning_rate": 5.324656750572082e-06,
      "loss": 0.7235,
      "step": 5751
    },
    {
      "epoch": 1.65,
      "grad_norm": 12.722701072692871,
      "learning_rate": 5.320366132723112e-06,
      "loss": 0.6949,
      "step": 5752
    },
    {
      "epoch": 1.65,
      "grad_norm": 11.349908828735352,
      "learning_rate": 5.316075514874142e-06,
      "loss": 0.5982,
      "step": 5753
    },
    {
      "epoch": 1.65,
      "grad_norm": 9.948938369750977,
      "learning_rate": 5.3117848970251716e-06,
      "loss": 0.7043,
      "step": 5754
    },
    {
      "epoch": 1.65,
      "grad_norm": 10.155777931213379,
      "learning_rate": 5.307494279176201e-06,
      "loss": 0.6262,
      "step": 5755
    },
    {
      "epoch": 1.65,
      "grad_norm": 12.634535789489746,
      "learning_rate": 5.303203661327232e-06,
      "loss": 0.8843,
      "step": 5756
    },
    {
      "epoch": 1.65,
      "grad_norm": 12.780898094177246,
      "learning_rate": 5.298913043478261e-06,
      "loss": 0.864,
      "step": 5757
    },
    {
      "epoch": 1.65,
      "grad_norm": 9.863545417785645,
      "learning_rate": 5.294622425629291e-06,
      "loss": 0.6517,
      "step": 5758
    },
    {
      "epoch": 1.65,
      "grad_norm": 10.639693260192871,
      "learning_rate": 5.2903318077803205e-06,
      "loss": 0.6594,
      "step": 5759
    },
    {
      "epoch": 1.65,
      "grad_norm": 10.80581283569336,
      "learning_rate": 5.28604118993135e-06,
      "loss": 0.6584,
      "step": 5760
    },
    {
      "epoch": 1.65,
      "grad_norm": 9.108868598937988,
      "learning_rate": 5.2817505720823806e-06,
      "loss": 0.5799,
      "step": 5761
    },
    {
      "epoch": 1.65,
      "grad_norm": 11.019648551940918,
      "learning_rate": 5.27745995423341e-06,
      "loss": 0.7017,
      "step": 5762
    },
    {
      "epoch": 1.65,
      "grad_norm": 9.318804740905762,
      "learning_rate": 5.27316933638444e-06,
      "loss": 0.6385,
      "step": 5763
    },
    {
      "epoch": 1.65,
      "grad_norm": 10.86406421661377,
      "learning_rate": 5.2688787185354686e-06,
      "loss": 0.5914,
      "step": 5764
    },
    {
      "epoch": 1.65,
      "grad_norm": 10.691460609436035,
      "learning_rate": 5.264588100686499e-06,
      "loss": 0.845,
      "step": 5765
    },
    {
      "epoch": 1.65,
      "grad_norm": 10.869109153747559,
      "learning_rate": 5.260297482837529e-06,
      "loss": 0.9187,
      "step": 5766
    },
    {
      "epoch": 1.65,
      "grad_norm": 8.999061584472656,
      "learning_rate": 5.256006864988558e-06,
      "loss": 0.82,
      "step": 5767
    },
    {
      "epoch": 1.65,
      "grad_norm": 9.673580169677734,
      "learning_rate": 5.251716247139588e-06,
      "loss": 0.851,
      "step": 5768
    },
    {
      "epoch": 1.65,
      "grad_norm": 10.771028518676758,
      "learning_rate": 5.247425629290618e-06,
      "loss": 0.7403,
      "step": 5769
    },
    {
      "epoch": 1.65,
      "grad_norm": 9.66604232788086,
      "learning_rate": 5.243135011441648e-06,
      "loss": 0.8523,
      "step": 5770
    },
    {
      "epoch": 1.65,
      "grad_norm": 9.262810707092285,
      "learning_rate": 5.2388443935926776e-06,
      "loss": 0.5099,
      "step": 5771
    },
    {
      "epoch": 1.65,
      "grad_norm": 9.349327087402344,
      "learning_rate": 5.234553775743707e-06,
      "loss": 0.6424,
      "step": 5772
    },
    {
      "epoch": 1.65,
      "grad_norm": 10.648466110229492,
      "learning_rate": 5.230263157894737e-06,
      "loss": 0.8119,
      "step": 5773
    },
    {
      "epoch": 1.65,
      "grad_norm": 10.655963897705078,
      "learning_rate": 5.225972540045767e-06,
      "loss": 0.6532,
      "step": 5774
    },
    {
      "epoch": 1.65,
      "grad_norm": 9.538254737854004,
      "learning_rate": 5.221681922196797e-06,
      "loss": 0.6325,
      "step": 5775
    },
    {
      "epoch": 1.65,
      "grad_norm": 11.388775825500488,
      "learning_rate": 5.2173913043478265e-06,
      "loss": 0.8202,
      "step": 5776
    },
    {
      "epoch": 1.65,
      "grad_norm": 8.295978546142578,
      "learning_rate": 5.213100686498855e-06,
      "loss": 0.6304,
      "step": 5777
    },
    {
      "epoch": 1.65,
      "grad_norm": 9.392183303833008,
      "learning_rate": 5.208810068649886e-06,
      "loss": 0.9357,
      "step": 5778
    },
    {
      "epoch": 1.65,
      "grad_norm": 12.010884284973145,
      "learning_rate": 5.204519450800915e-06,
      "loss": 0.8542,
      "step": 5779
    },
    {
      "epoch": 1.65,
      "grad_norm": 9.261788368225098,
      "learning_rate": 5.200228832951945e-06,
      "loss": 0.709,
      "step": 5780
    },
    {
      "epoch": 1.65,
      "grad_norm": 11.460758209228516,
      "learning_rate": 5.1959382151029746e-06,
      "loss": 0.7145,
      "step": 5781
    },
    {
      "epoch": 1.65,
      "grad_norm": 11.312665939331055,
      "learning_rate": 5.191647597254005e-06,
      "loss": 0.7533,
      "step": 5782
    },
    {
      "epoch": 1.65,
      "grad_norm": 10.184603691101074,
      "learning_rate": 5.187356979405035e-06,
      "loss": 0.6701,
      "step": 5783
    },
    {
      "epoch": 1.65,
      "grad_norm": 10.924507141113281,
      "learning_rate": 5.183066361556064e-06,
      "loss": 0.5879,
      "step": 5784
    },
    {
      "epoch": 1.65,
      "grad_norm": 11.757448196411133,
      "learning_rate": 5.178775743707094e-06,
      "loss": 0.9316,
      "step": 5785
    },
    {
      "epoch": 1.66,
      "grad_norm": 9.38831901550293,
      "learning_rate": 5.174485125858124e-06,
      "loss": 0.5362,
      "step": 5786
    },
    {
      "epoch": 1.66,
      "grad_norm": 10.152300834655762,
      "learning_rate": 5.170194508009154e-06,
      "loss": 0.5182,
      "step": 5787
    },
    {
      "epoch": 1.66,
      "grad_norm": 10.15249252319336,
      "learning_rate": 5.1659038901601836e-06,
      "loss": 0.6635,
      "step": 5788
    },
    {
      "epoch": 1.66,
      "grad_norm": 10.151712417602539,
      "learning_rate": 5.161613272311213e-06,
      "loss": 0.9518,
      "step": 5789
    },
    {
      "epoch": 1.66,
      "grad_norm": 12.894401550292969,
      "learning_rate": 5.157322654462242e-06,
      "loss": 0.8927,
      "step": 5790
    },
    {
      "epoch": 1.66,
      "grad_norm": 10.738083839416504,
      "learning_rate": 5.153032036613272e-06,
      "loss": 0.701,
      "step": 5791
    },
    {
      "epoch": 1.66,
      "grad_norm": 12.905317306518555,
      "learning_rate": 5.148741418764302e-06,
      "loss": 1.1686,
      "step": 5792
    },
    {
      "epoch": 1.66,
      "grad_norm": 9.421422004699707,
      "learning_rate": 5.144450800915332e-06,
      "loss": 0.8037,
      "step": 5793
    },
    {
      "epoch": 1.66,
      "grad_norm": 10.57293701171875,
      "learning_rate": 5.140160183066361e-06,
      "loss": 0.8017,
      "step": 5794
    },
    {
      "epoch": 1.66,
      "grad_norm": 9.552532196044922,
      "learning_rate": 5.135869565217392e-06,
      "loss": 0.7185,
      "step": 5795
    },
    {
      "epoch": 1.66,
      "grad_norm": 9.750856399536133,
      "learning_rate": 5.131578947368421e-06,
      "loss": 0.7244,
      "step": 5796
    },
    {
      "epoch": 1.66,
      "grad_norm": 12.330699920654297,
      "learning_rate": 5.127288329519451e-06,
      "loss": 0.6662,
      "step": 5797
    },
    {
      "epoch": 1.66,
      "grad_norm": 10.388833045959473,
      "learning_rate": 5.1229977116704805e-06,
      "loss": 0.885,
      "step": 5798
    },
    {
      "epoch": 1.66,
      "grad_norm": 10.175448417663574,
      "learning_rate": 5.118707093821511e-06,
      "loss": 0.6561,
      "step": 5799
    },
    {
      "epoch": 1.66,
      "grad_norm": 9.187387466430664,
      "learning_rate": 5.114416475972541e-06,
      "loss": 0.646,
      "step": 5800
    },
    {
      "epoch": 1.66,
      "grad_norm": 8.872142791748047,
      "learning_rate": 5.11012585812357e-06,
      "loss": 0.7365,
      "step": 5801
    },
    {
      "epoch": 1.66,
      "grad_norm": 8.377429008483887,
      "learning_rate": 5.1058352402746e-06,
      "loss": 0.5354,
      "step": 5802
    },
    {
      "epoch": 1.66,
      "grad_norm": 7.479408264160156,
      "learning_rate": 5.101544622425629e-06,
      "loss": 0.5004,
      "step": 5803
    },
    {
      "epoch": 1.66,
      "grad_norm": 11.968086242675781,
      "learning_rate": 5.097254004576659e-06,
      "loss": 0.8115,
      "step": 5804
    },
    {
      "epoch": 1.66,
      "grad_norm": 8.832465171813965,
      "learning_rate": 5.092963386727689e-06,
      "loss": 0.5674,
      "step": 5805
    },
    {
      "epoch": 1.66,
      "grad_norm": 10.427408218383789,
      "learning_rate": 5.088672768878718e-06,
      "loss": 0.6168,
      "step": 5806
    },
    {
      "epoch": 1.66,
      "grad_norm": 10.2507963180542,
      "learning_rate": 5.084382151029748e-06,
      "loss": 0.6153,
      "step": 5807
    },
    {
      "epoch": 1.66,
      "grad_norm": 8.970211029052734,
      "learning_rate": 5.080091533180778e-06,
      "loss": 0.6329,
      "step": 5808
    },
    {
      "epoch": 1.66,
      "grad_norm": 12.040018081665039,
      "learning_rate": 5.075800915331808e-06,
      "loss": 0.8813,
      "step": 5809
    },
    {
      "epoch": 1.66,
      "grad_norm": 12.211027145385742,
      "learning_rate": 5.071510297482838e-06,
      "loss": 0.6981,
      "step": 5810
    },
    {
      "epoch": 1.66,
      "grad_norm": 12.553165435791016,
      "learning_rate": 5.067219679633867e-06,
      "loss": 0.6293,
      "step": 5811
    },
    {
      "epoch": 1.66,
      "grad_norm": 9.366776466369629,
      "learning_rate": 5.062929061784898e-06,
      "loss": 0.6127,
      "step": 5812
    },
    {
      "epoch": 1.66,
      "grad_norm": 10.16841983795166,
      "learning_rate": 5.058638443935927e-06,
      "loss": 0.6729,
      "step": 5813
    },
    {
      "epoch": 1.66,
      "grad_norm": 8.510993003845215,
      "learning_rate": 5.054347826086957e-06,
      "loss": 0.7226,
      "step": 5814
    },
    {
      "epoch": 1.66,
      "grad_norm": 12.433296203613281,
      "learning_rate": 5.0500572082379865e-06,
      "loss": 0.9685,
      "step": 5815
    },
    {
      "epoch": 1.66,
      "grad_norm": 11.985654830932617,
      "learning_rate": 5.045766590389016e-06,
      "loss": 0.7879,
      "step": 5816
    },
    {
      "epoch": 1.66,
      "grad_norm": 10.938732147216797,
      "learning_rate": 5.041475972540046e-06,
      "loss": 0.72,
      "step": 5817
    },
    {
      "epoch": 1.66,
      "grad_norm": 9.93817138671875,
      "learning_rate": 5.037185354691075e-06,
      "loss": 0.7446,
      "step": 5818
    },
    {
      "epoch": 1.66,
      "grad_norm": 11.764874458312988,
      "learning_rate": 5.032894736842105e-06,
      "loss": 0.7629,
      "step": 5819
    },
    {
      "epoch": 1.66,
      "grad_norm": 12.215780258178711,
      "learning_rate": 5.028604118993135e-06,
      "loss": 0.6836,
      "step": 5820
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.806153297424316,
      "learning_rate": 5.024313501144165e-06,
      "loss": 0.6898,
      "step": 5821
    },
    {
      "epoch": 1.67,
      "grad_norm": 7.726294040679932,
      "learning_rate": 5.020022883295195e-06,
      "loss": 0.4908,
      "step": 5822
    },
    {
      "epoch": 1.67,
      "grad_norm": 11.043843269348145,
      "learning_rate": 5.015732265446224e-06,
      "loss": 0.6357,
      "step": 5823
    },
    {
      "epoch": 1.67,
      "grad_norm": 9.391953468322754,
      "learning_rate": 5.011441647597254e-06,
      "loss": 0.6654,
      "step": 5824
    },
    {
      "epoch": 1.67,
      "grad_norm": 9.231965065002441,
      "learning_rate": 5.007151029748284e-06,
      "loss": 0.5421,
      "step": 5825
    },
    {
      "epoch": 1.67,
      "grad_norm": 8.77448844909668,
      "learning_rate": 5.002860411899314e-06,
      "loss": 0.5603,
      "step": 5826
    },
    {
      "epoch": 1.67,
      "grad_norm": 9.12179946899414,
      "learning_rate": 4.998569794050344e-06,
      "loss": 0.6626,
      "step": 5827
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.378664016723633,
      "learning_rate": 4.994279176201373e-06,
      "loss": 0.681,
      "step": 5828
    },
    {
      "epoch": 1.67,
      "grad_norm": 8.726974487304688,
      "learning_rate": 4.989988558352403e-06,
      "loss": 0.4317,
      "step": 5829
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.938841819763184,
      "learning_rate": 4.9856979405034325e-06,
      "loss": 0.9883,
      "step": 5830
    },
    {
      "epoch": 1.67,
      "grad_norm": 11.10641860961914,
      "learning_rate": 4.981407322654462e-06,
      "loss": 0.9498,
      "step": 5831
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.560961723327637,
      "learning_rate": 4.977116704805492e-06,
      "loss": 0.8008,
      "step": 5832
    },
    {
      "epoch": 1.67,
      "grad_norm": 11.211451530456543,
      "learning_rate": 4.972826086956521e-06,
      "loss": 0.5645,
      "step": 5833
    },
    {
      "epoch": 1.67,
      "grad_norm": 9.279936790466309,
      "learning_rate": 4.968535469107552e-06,
      "loss": 0.6464,
      "step": 5834
    },
    {
      "epoch": 1.67,
      "grad_norm": 13.09455394744873,
      "learning_rate": 4.964244851258581e-06,
      "loss": 0.8324,
      "step": 5835
    },
    {
      "epoch": 1.67,
      "grad_norm": 12.164727210998535,
      "learning_rate": 4.959954233409611e-06,
      "loss": 0.6953,
      "step": 5836
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.946770668029785,
      "learning_rate": 4.955663615560641e-06,
      "loss": 0.5145,
      "step": 5837
    },
    {
      "epoch": 1.67,
      "grad_norm": 11.676422119140625,
      "learning_rate": 4.951372997711671e-06,
      "loss": 0.7639,
      "step": 5838
    },
    {
      "epoch": 1.67,
      "grad_norm": 9.145784378051758,
      "learning_rate": 4.947082379862701e-06,
      "loss": 0.6324,
      "step": 5839
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.441269874572754,
      "learning_rate": 4.94279176201373e-06,
      "loss": 0.5452,
      "step": 5840
    },
    {
      "epoch": 1.67,
      "grad_norm": 9.933587074279785,
      "learning_rate": 4.93850114416476e-06,
      "loss": 0.6534,
      "step": 5841
    },
    {
      "epoch": 1.67,
      "grad_norm": 14.15119743347168,
      "learning_rate": 4.9342105263157895e-06,
      "loss": 0.829,
      "step": 5842
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.811209678649902,
      "learning_rate": 4.929919908466819e-06,
      "loss": 0.5488,
      "step": 5843
    },
    {
      "epoch": 1.67,
      "grad_norm": 12.233796119689941,
      "learning_rate": 4.925629290617849e-06,
      "loss": 0.8341,
      "step": 5844
    },
    {
      "epoch": 1.67,
      "grad_norm": 14.163995742797852,
      "learning_rate": 4.921338672768878e-06,
      "loss": 0.8841,
      "step": 5845
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.326066970825195,
      "learning_rate": 4.917048054919909e-06,
      "loss": 0.8647,
      "step": 5846
    },
    {
      "epoch": 1.67,
      "grad_norm": 11.209264755249023,
      "learning_rate": 4.9127574370709385e-06,
      "loss": 0.6987,
      "step": 5847
    },
    {
      "epoch": 1.67,
      "grad_norm": 11.691764831542969,
      "learning_rate": 4.908466819221968e-06,
      "loss": 0.7641,
      "step": 5848
    },
    {
      "epoch": 1.67,
      "grad_norm": 11.83598804473877,
      "learning_rate": 4.904176201372998e-06,
      "loss": 0.5928,
      "step": 5849
    },
    {
      "epoch": 1.67,
      "grad_norm": 9.098348617553711,
      "learning_rate": 4.899885583524027e-06,
      "loss": 0.5905,
      "step": 5850
    },
    {
      "epoch": 1.67,
      "grad_norm": 9.325675010681152,
      "learning_rate": 4.895594965675058e-06,
      "loss": 0.6716,
      "step": 5851
    },
    {
      "epoch": 1.67,
      "grad_norm": 9.184008598327637,
      "learning_rate": 4.891304347826087e-06,
      "loss": 0.4886,
      "step": 5852
    },
    {
      "epoch": 1.67,
      "grad_norm": 11.029022216796875,
      "learning_rate": 4.887013729977117e-06,
      "loss": 0.7218,
      "step": 5853
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.732136726379395,
      "learning_rate": 4.882723112128147e-06,
      "loss": 0.6508,
      "step": 5854
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.072098731994629,
      "learning_rate": 4.878432494279177e-06,
      "loss": 0.6364,
      "step": 5855
    },
    {
      "epoch": 1.68,
      "grad_norm": 7.013856410980225,
      "learning_rate": 4.874141876430206e-06,
      "loss": 0.4318,
      "step": 5856
    },
    {
      "epoch": 1.68,
      "grad_norm": 12.017395973205566,
      "learning_rate": 4.8698512585812354e-06,
      "loss": 0.7872,
      "step": 5857
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.142374992370605,
      "learning_rate": 4.865560640732265e-06,
      "loss": 0.5402,
      "step": 5858
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.501347541809082,
      "learning_rate": 4.8612700228832955e-06,
      "loss": 0.6895,
      "step": 5859
    },
    {
      "epoch": 1.68,
      "grad_norm": 12.26644515991211,
      "learning_rate": 4.856979405034325e-06,
      "loss": 0.8957,
      "step": 5860
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.183009147644043,
      "learning_rate": 4.852688787185355e-06,
      "loss": 0.6938,
      "step": 5861
    },
    {
      "epoch": 1.68,
      "grad_norm": 13.985980033874512,
      "learning_rate": 4.848398169336384e-06,
      "loss": 0.8486,
      "step": 5862
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.614011764526367,
      "learning_rate": 4.844107551487414e-06,
      "loss": 0.7999,
      "step": 5863
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.026040077209473,
      "learning_rate": 4.8398169336384444e-06,
      "loss": 0.7847,
      "step": 5864
    },
    {
      "epoch": 1.68,
      "grad_norm": 11.509770393371582,
      "learning_rate": 4.835526315789474e-06,
      "loss": 0.7551,
      "step": 5865
    },
    {
      "epoch": 1.68,
      "grad_norm": 8.606500625610352,
      "learning_rate": 4.831235697940504e-06,
      "loss": 0.6628,
      "step": 5866
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.944982528686523,
      "learning_rate": 4.826945080091533e-06,
      "loss": 0.7259,
      "step": 5867
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.284320831298828,
      "learning_rate": 4.822654462242564e-06,
      "loss": 0.6398,
      "step": 5868
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.699448585510254,
      "learning_rate": 4.8183638443935925e-06,
      "loss": 0.6661,
      "step": 5869
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.851250648498535,
      "learning_rate": 4.814073226544622e-06,
      "loss": 0.5445,
      "step": 5870
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.146505355834961,
      "learning_rate": 4.809782608695652e-06,
      "loss": 0.5736,
      "step": 5871
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.085628509521484,
      "learning_rate": 4.805491990846682e-06,
      "loss": 0.7359,
      "step": 5872
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.133440017700195,
      "learning_rate": 4.801201372997712e-06,
      "loss": 0.6137,
      "step": 5873
    },
    {
      "epoch": 1.68,
      "grad_norm": 8.663152694702148,
      "learning_rate": 4.7969107551487414e-06,
      "loss": 0.7389,
      "step": 5874
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.566811561584473,
      "learning_rate": 4.792620137299771e-06,
      "loss": 0.7826,
      "step": 5875
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.916440963745117,
      "learning_rate": 4.7883295194508015e-06,
      "loss": 0.5214,
      "step": 5876
    },
    {
      "epoch": 1.68,
      "grad_norm": 11.84753704071045,
      "learning_rate": 4.784038901601831e-06,
      "loss": 0.7269,
      "step": 5877
    },
    {
      "epoch": 1.68,
      "grad_norm": 11.298389434814453,
      "learning_rate": 4.779748283752861e-06,
      "loss": 0.7548,
      "step": 5878
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.253491401672363,
      "learning_rate": 4.77545766590389e-06,
      "loss": 0.7006,
      "step": 5879
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.172586441040039,
      "learning_rate": 4.77116704805492e-06,
      "loss": 0.7238,
      "step": 5880
    },
    {
      "epoch": 1.68,
      "grad_norm": 11.111824035644531,
      "learning_rate": 4.7668764302059504e-06,
      "loss": 0.6726,
      "step": 5881
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.476541519165039,
      "learning_rate": 4.762585812356979e-06,
      "loss": 0.3624,
      "step": 5882
    },
    {
      "epoch": 1.68,
      "grad_norm": 8.951101303100586,
      "learning_rate": 4.758295194508009e-06,
      "loss": 0.6643,
      "step": 5883
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.434000015258789,
      "learning_rate": 4.7540045766590384e-06,
      "loss": 0.5429,
      "step": 5884
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.742546081542969,
      "learning_rate": 4.749713958810069e-06,
      "loss": 0.6604,
      "step": 5885
    },
    {
      "epoch": 1.68,
      "grad_norm": 8.147375106811523,
      "learning_rate": 4.7454233409610985e-06,
      "loss": 0.6377,
      "step": 5886
    },
    {
      "epoch": 1.68,
      "grad_norm": 11.91827392578125,
      "learning_rate": 4.741132723112128e-06,
      "loss": 0.7549,
      "step": 5887
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.47176742553711,
      "learning_rate": 4.736842105263158e-06,
      "loss": 0.5754,
      "step": 5888
    },
    {
      "epoch": 1.68,
      "grad_norm": 11.063567161560059,
      "learning_rate": 4.732551487414188e-06,
      "loss": 0.7022,
      "step": 5889
    },
    {
      "epoch": 1.68,
      "grad_norm": 8.001396179199219,
      "learning_rate": 4.728260869565218e-06,
      "loss": 0.646,
      "step": 5890
    },
    {
      "epoch": 1.69,
      "grad_norm": 10.043627738952637,
      "learning_rate": 4.7239702517162474e-06,
      "loss": 0.943,
      "step": 5891
    },
    {
      "epoch": 1.69,
      "grad_norm": 9.53686237335205,
      "learning_rate": 4.719679633867277e-06,
      "loss": 0.5179,
      "step": 5892
    },
    {
      "epoch": 1.69,
      "grad_norm": 9.575545310974121,
      "learning_rate": 4.715389016018307e-06,
      "loss": 0.6462,
      "step": 5893
    },
    {
      "epoch": 1.69,
      "grad_norm": 13.83825397491455,
      "learning_rate": 4.711098398169337e-06,
      "loss": 0.736,
      "step": 5894
    },
    {
      "epoch": 1.69,
      "grad_norm": 10.548954963684082,
      "learning_rate": 4.706807780320366e-06,
      "loss": 0.6661,
      "step": 5895
    },
    {
      "epoch": 1.69,
      "grad_norm": 9.620893478393555,
      "learning_rate": 4.7025171624713955e-06,
      "loss": 0.6117,
      "step": 5896
    },
    {
      "epoch": 1.69,
      "grad_norm": 9.700591087341309,
      "learning_rate": 4.698226544622425e-06,
      "loss": 0.6078,
      "step": 5897
    },
    {
      "epoch": 1.69,
      "grad_norm": 11.308130264282227,
      "learning_rate": 4.693935926773456e-06,
      "loss": 0.6275,
      "step": 5898
    },
    {
      "epoch": 1.69,
      "grad_norm": 10.247930526733398,
      "learning_rate": 4.689645308924485e-06,
      "loss": 0.8355,
      "step": 5899
    },
    {
      "epoch": 1.69,
      "grad_norm": 11.837489128112793,
      "learning_rate": 4.685354691075515e-06,
      "loss": 0.7833,
      "step": 5900
    },
    {
      "epoch": 1.69,
      "grad_norm": 11.13876724243164,
      "learning_rate": 4.6810640732265444e-06,
      "loss": 0.5052,
      "step": 5901
    },
    {
      "epoch": 1.69,
      "grad_norm": 11.996820449829102,
      "learning_rate": 4.676773455377575e-06,
      "loss": 0.7055,
      "step": 5902
    },
    {
      "epoch": 1.69,
      "grad_norm": 13.874919891357422,
      "learning_rate": 4.6724828375286045e-06,
      "loss": 0.7694,
      "step": 5903
    },
    {
      "epoch": 1.69,
      "grad_norm": 13.17946720123291,
      "learning_rate": 4.668192219679634e-06,
      "loss": 0.8438,
      "step": 5904
    },
    {
      "epoch": 1.69,
      "grad_norm": 12.827901840209961,
      "learning_rate": 4.663901601830664e-06,
      "loss": 0.7192,
      "step": 5905
    },
    {
      "epoch": 1.69,
      "grad_norm": 12.541082382202148,
      "learning_rate": 4.659610983981694e-06,
      "loss": 0.8147,
      "step": 5906
    },
    {
      "epoch": 1.69,
      "grad_norm": 8.027711868286133,
      "learning_rate": 4.655320366132724e-06,
      "loss": 0.568,
      "step": 5907
    },
    {
      "epoch": 1.69,
      "grad_norm": 10.079437255859375,
      "learning_rate": 4.651029748283753e-06,
      "loss": 0.632,
      "step": 5908
    },
    {
      "epoch": 1.69,
      "grad_norm": 9.83368968963623,
      "learning_rate": 4.646739130434782e-06,
      "loss": 0.5738,
      "step": 5909
    },
    {
      "epoch": 1.69,
      "grad_norm": 14.067094802856445,
      "learning_rate": 4.642448512585812e-06,
      "loss": 0.954,
      "step": 5910
    },
    {
      "epoch": 1.69,
      "grad_norm": 6.759878635406494,
      "learning_rate": 4.638157894736842e-06,
      "loss": 0.5277,
      "step": 5911
    },
    {
      "epoch": 1.69,
      "grad_norm": 8.464375495910645,
      "learning_rate": 4.633867276887872e-06,
      "loss": 0.5855,
      "step": 5912
    },
    {
      "epoch": 1.69,
      "grad_norm": 11.092123031616211,
      "learning_rate": 4.6295766590389015e-06,
      "loss": 0.7731,
      "step": 5913
    },
    {
      "epoch": 1.69,
      "grad_norm": 11.064888954162598,
      "learning_rate": 4.625286041189931e-06,
      "loss": 0.9506,
      "step": 5914
    },
    {
      "epoch": 1.69,
      "grad_norm": 9.002968788146973,
      "learning_rate": 4.620995423340962e-06,
      "loss": 0.5487,
      "step": 5915
    },
    {
      "epoch": 1.69,
      "grad_norm": 12.4066162109375,
      "learning_rate": 4.616704805491991e-06,
      "loss": 0.9604,
      "step": 5916
    },
    {
      "epoch": 1.69,
      "grad_norm": 8.316583633422852,
      "learning_rate": 4.612414187643021e-06,
      "loss": 0.583,
      "step": 5917
    },
    {
      "epoch": 1.69,
      "grad_norm": 12.206084251403809,
      "learning_rate": 4.6081235697940504e-06,
      "loss": 0.7596,
      "step": 5918
    },
    {
      "epoch": 1.69,
      "grad_norm": 9.902289390563965,
      "learning_rate": 4.603832951945081e-06,
      "loss": 0.7456,
      "step": 5919
    },
    {
      "epoch": 1.69,
      "grad_norm": 14.221868515014648,
      "learning_rate": 4.5995423340961105e-06,
      "loss": 0.778,
      "step": 5920
    },
    {
      "epoch": 1.69,
      "grad_norm": 13.835212707519531,
      "learning_rate": 4.595251716247139e-06,
      "loss": 0.5654,
      "step": 5921
    },
    {
      "epoch": 1.69,
      "grad_norm": 10.850911140441895,
      "learning_rate": 4.590961098398169e-06,
      "loss": 0.5314,
      "step": 5922
    },
    {
      "epoch": 1.69,
      "grad_norm": 12.394707679748535,
      "learning_rate": 4.5866704805491985e-06,
      "loss": 0.5975,
      "step": 5923
    },
    {
      "epoch": 1.69,
      "grad_norm": 10.528264045715332,
      "learning_rate": 4.582379862700229e-06,
      "loss": 0.4408,
      "step": 5924
    },
    {
      "epoch": 1.69,
      "grad_norm": 13.076751708984375,
      "learning_rate": 4.578089244851259e-06,
      "loss": 0.9047,
      "step": 5925
    },
    {
      "epoch": 1.7,
      "grad_norm": 11.11559009552002,
      "learning_rate": 4.573798627002288e-06,
      "loss": 0.6955,
      "step": 5926
    },
    {
      "epoch": 1.7,
      "grad_norm": 11.401023864746094,
      "learning_rate": 4.569508009153318e-06,
      "loss": 0.4923,
      "step": 5927
    },
    {
      "epoch": 1.7,
      "grad_norm": 11.230006217956543,
      "learning_rate": 4.565217391304348e-06,
      "loss": 0.7525,
      "step": 5928
    },
    {
      "epoch": 1.7,
      "grad_norm": 9.576884269714355,
      "learning_rate": 4.560926773455378e-06,
      "loss": 0.4744,
      "step": 5929
    },
    {
      "epoch": 1.7,
      "grad_norm": 13.80195140838623,
      "learning_rate": 4.5566361556064075e-06,
      "loss": 0.8665,
      "step": 5930
    },
    {
      "epoch": 1.7,
      "grad_norm": 8.915005683898926,
      "learning_rate": 4.552345537757437e-06,
      "loss": 0.6348,
      "step": 5931
    },
    {
      "epoch": 1.7,
      "grad_norm": 11.691276550292969,
      "learning_rate": 4.548054919908468e-06,
      "loss": 0.6432,
      "step": 5932
    },
    {
      "epoch": 1.7,
      "grad_norm": 12.980295181274414,
      "learning_rate": 4.543764302059497e-06,
      "loss": 0.9116,
      "step": 5933
    },
    {
      "epoch": 1.7,
      "grad_norm": 9.581231117248535,
      "learning_rate": 4.539473684210527e-06,
      "loss": 0.6034,
      "step": 5934
    },
    {
      "epoch": 1.7,
      "grad_norm": 8.7820405960083,
      "learning_rate": 4.535183066361556e-06,
      "loss": 0.6066,
      "step": 5935
    },
    {
      "epoch": 1.7,
      "grad_norm": 11.12899398803711,
      "learning_rate": 4.530892448512586e-06,
      "loss": 0.6293,
      "step": 5936
    },
    {
      "epoch": 1.7,
      "grad_norm": 8.969178199768066,
      "learning_rate": 4.526601830663616e-06,
      "loss": 0.5033,
      "step": 5937
    },
    {
      "epoch": 1.7,
      "grad_norm": 11.642123222351074,
      "learning_rate": 4.522311212814645e-06,
      "loss": 1.0285,
      "step": 5938
    },
    {
      "epoch": 1.7,
      "grad_norm": 10.58486557006836,
      "learning_rate": 4.518020594965675e-06,
      "loss": 0.8515,
      "step": 5939
    },
    {
      "epoch": 1.7,
      "grad_norm": 13.439935684204102,
      "learning_rate": 4.5137299771167045e-06,
      "loss": 0.9082,
      "step": 5940
    },
    {
      "epoch": 1.7,
      "grad_norm": 11.394495964050293,
      "learning_rate": 4.509439359267735e-06,
      "loss": 0.8086,
      "step": 5941
    },
    {
      "epoch": 1.7,
      "grad_norm": 8.398945808410645,
      "learning_rate": 4.5051487414187646e-06,
      "loss": 0.8295,
      "step": 5942
    },
    {
      "epoch": 1.7,
      "grad_norm": 8.517271995544434,
      "learning_rate": 4.500858123569794e-06,
      "loss": 0.7168,
      "step": 5943
    },
    {
      "epoch": 1.7,
      "grad_norm": 9.0183744430542,
      "learning_rate": 4.496567505720824e-06,
      "loss": 0.6702,
      "step": 5944
    },
    {
      "epoch": 1.7,
      "grad_norm": 10.215263366699219,
      "learning_rate": 4.492276887871854e-06,
      "loss": 0.6134,
      "step": 5945
    },
    {
      "epoch": 1.7,
      "grad_norm": 12.507948875427246,
      "learning_rate": 4.487986270022884e-06,
      "loss": 0.7977,
      "step": 5946
    },
    {
      "epoch": 1.7,
      "grad_norm": 9.09069538116455,
      "learning_rate": 4.4836956521739135e-06,
      "loss": 0.629,
      "step": 5947
    },
    {
      "epoch": 1.7,
      "grad_norm": 10.303925514221191,
      "learning_rate": 4.479405034324942e-06,
      "loss": 0.6987,
      "step": 5948
    },
    {
      "epoch": 1.7,
      "grad_norm": 9.196348190307617,
      "learning_rate": 4.475114416475973e-06,
      "loss": 0.6685,
      "step": 5949
    },
    {
      "epoch": 1.7,
      "grad_norm": 11.39057445526123,
      "learning_rate": 4.470823798627002e-06,
      "loss": 0.6844,
      "step": 5950
    },
    {
      "epoch": 1.7,
      "grad_norm": 11.751300811767578,
      "learning_rate": 4.466533180778032e-06,
      "loss": 0.6349,
      "step": 5951
    },
    {
      "epoch": 1.7,
      "grad_norm": 10.53183650970459,
      "learning_rate": 4.4622425629290616e-06,
      "loss": 0.7825,
      "step": 5952
    },
    {
      "epoch": 1.7,
      "grad_norm": 9.162747383117676,
      "learning_rate": 4.457951945080091e-06,
      "loss": 0.6018,
      "step": 5953
    },
    {
      "epoch": 1.7,
      "grad_norm": 9.915216445922852,
      "learning_rate": 4.453661327231122e-06,
      "loss": 0.5699,
      "step": 5954
    },
    {
      "epoch": 1.7,
      "grad_norm": 8.251471519470215,
      "learning_rate": 4.449370709382151e-06,
      "loss": 0.5714,
      "step": 5955
    },
    {
      "epoch": 1.7,
      "grad_norm": 11.356422424316406,
      "learning_rate": 4.445080091533181e-06,
      "loss": 0.9157,
      "step": 5956
    },
    {
      "epoch": 1.7,
      "grad_norm": 7.7874836921691895,
      "learning_rate": 4.4407894736842105e-06,
      "loss": 0.5033,
      "step": 5957
    },
    {
      "epoch": 1.7,
      "grad_norm": 9.747831344604492,
      "learning_rate": 4.436498855835241e-06,
      "loss": 0.7735,
      "step": 5958
    },
    {
      "epoch": 1.7,
      "grad_norm": 9.569536209106445,
      "learning_rate": 4.4322082379862706e-06,
      "loss": 0.4177,
      "step": 5959
    },
    {
      "epoch": 1.7,
      "grad_norm": 14.832538604736328,
      "learning_rate": 4.4279176201373e-06,
      "loss": 0.9468,
      "step": 5960
    },
    {
      "epoch": 1.71,
      "grad_norm": 11.02159595489502,
      "learning_rate": 4.423627002288329e-06,
      "loss": 0.7396,
      "step": 5961
    },
    {
      "epoch": 1.71,
      "grad_norm": 8.493553161621094,
      "learning_rate": 4.419336384439359e-06,
      "loss": 0.5594,
      "step": 5962
    },
    {
      "epoch": 1.71,
      "grad_norm": 9.55902099609375,
      "learning_rate": 4.415045766590389e-06,
      "loss": 0.6982,
      "step": 5963
    },
    {
      "epoch": 1.71,
      "grad_norm": 10.69473934173584,
      "learning_rate": 4.410755148741419e-06,
      "loss": 0.7358,
      "step": 5964
    },
    {
      "epoch": 1.71,
      "grad_norm": 10.646027565002441,
      "learning_rate": 4.406464530892448e-06,
      "loss": 0.6932,
      "step": 5965
    },
    {
      "epoch": 1.71,
      "grad_norm": 10.105545043945312,
      "learning_rate": 4.402173913043479e-06,
      "loss": 0.5455,
      "step": 5966
    },
    {
      "epoch": 1.71,
      "grad_norm": 10.997332572937012,
      "learning_rate": 4.397883295194508e-06,
      "loss": 0.6765,
      "step": 5967
    },
    {
      "epoch": 1.71,
      "grad_norm": 11.80391788482666,
      "learning_rate": 4.393592677345538e-06,
      "loss": 0.9207,
      "step": 5968
    },
    {
      "epoch": 1.71,
      "grad_norm": 9.650821685791016,
      "learning_rate": 4.3893020594965676e-06,
      "loss": 0.8558,
      "step": 5969
    },
    {
      "epoch": 1.71,
      "grad_norm": 10.389674186706543,
      "learning_rate": 4.385011441647597e-06,
      "loss": 0.6678,
      "step": 5970
    },
    {
      "epoch": 1.71,
      "grad_norm": 12.466364860534668,
      "learning_rate": 4.380720823798628e-06,
      "loss": 0.8399,
      "step": 5971
    },
    {
      "epoch": 1.71,
      "grad_norm": 10.856403350830078,
      "learning_rate": 4.376430205949657e-06,
      "loss": 0.737,
      "step": 5972
    },
    {
      "epoch": 1.71,
      "grad_norm": 9.363508224487305,
      "learning_rate": 4.372139588100687e-06,
      "loss": 0.4673,
      "step": 5973
    },
    {
      "epoch": 1.71,
      "grad_norm": 11.127074241638184,
      "learning_rate": 4.367848970251716e-06,
      "loss": 1.125,
      "step": 5974
    },
    {
      "epoch": 1.71,
      "grad_norm": 9.285826683044434,
      "learning_rate": 4.363558352402746e-06,
      "loss": 0.5807,
      "step": 5975
    },
    {
      "epoch": 1.71,
      "grad_norm": 10.394722938537598,
      "learning_rate": 4.359267734553776e-06,
      "loss": 0.7673,
      "step": 5976
    },
    {
      "epoch": 1.71,
      "grad_norm": 13.34463119506836,
      "learning_rate": 4.354977116704805e-06,
      "loss": 0.5349,
      "step": 5977
    },
    {
      "epoch": 1.71,
      "grad_norm": 8.6209716796875,
      "learning_rate": 4.350686498855835e-06,
      "loss": 0.5501,
      "step": 5978
    },
    {
      "epoch": 1.71,
      "grad_norm": 13.577221870422363,
      "learning_rate": 4.346395881006865e-06,
      "loss": 0.8501,
      "step": 5979
    },
    {
      "epoch": 1.71,
      "grad_norm": 10.219799995422363,
      "learning_rate": 4.342105263157895e-06,
      "loss": 0.7215,
      "step": 5980
    },
    {
      "epoch": 1.71,
      "grad_norm": 13.079976081848145,
      "learning_rate": 4.337814645308925e-06,
      "loss": 0.7579,
      "step": 5981
    },
    {
      "epoch": 1.71,
      "grad_norm": 11.152042388916016,
      "learning_rate": 4.333524027459954e-06,
      "loss": 0.6373,
      "step": 5982
    },
    {
      "epoch": 1.71,
      "grad_norm": 11.564728736877441,
      "learning_rate": 4.329233409610984e-06,
      "loss": 0.7111,
      "step": 5983
    },
    {
      "epoch": 1.71,
      "grad_norm": 10.477561950683594,
      "learning_rate": 4.324942791762014e-06,
      "loss": 0.7411,
      "step": 5984
    },
    {
      "epoch": 1.71,
      "grad_norm": 10.553227424621582,
      "learning_rate": 4.320652173913044e-06,
      "loss": 0.8103,
      "step": 5985
    },
    {
      "epoch": 1.71,
      "grad_norm": 10.41597843170166,
      "learning_rate": 4.3163615560640736e-06,
      "loss": 0.6798,
      "step": 5986
    },
    {
      "epoch": 1.71,
      "grad_norm": 8.1062650680542,
      "learning_rate": 4.312070938215102e-06,
      "loss": 0.6382,
      "step": 5987
    },
    {
      "epoch": 1.71,
      "grad_norm": 8.41385269165039,
      "learning_rate": 4.307780320366133e-06,
      "loss": 0.5842,
      "step": 5988
    },
    {
      "epoch": 1.71,
      "grad_norm": 11.125850677490234,
      "learning_rate": 4.303489702517162e-06,
      "loss": 0.6911,
      "step": 5989
    },
    {
      "epoch": 1.71,
      "grad_norm": 10.93230152130127,
      "learning_rate": 4.299199084668192e-06,
      "loss": 0.5723,
      "step": 5990
    },
    {
      "epoch": 1.71,
      "grad_norm": 8.094341278076172,
      "learning_rate": 4.294908466819222e-06,
      "loss": 0.6396,
      "step": 5991
    },
    {
      "epoch": 1.71,
      "grad_norm": 13.452651977539062,
      "learning_rate": 4.290617848970252e-06,
      "loss": 0.7873,
      "step": 5992
    },
    {
      "epoch": 1.71,
      "grad_norm": 10.352450370788574,
      "learning_rate": 4.286327231121282e-06,
      "loss": 0.5889,
      "step": 5993
    },
    {
      "epoch": 1.71,
      "grad_norm": 9.265152931213379,
      "learning_rate": 4.282036613272311e-06,
      "loss": 0.7391,
      "step": 5994
    },
    {
      "epoch": 1.71,
      "grad_norm": 10.378124237060547,
      "learning_rate": 4.277745995423341e-06,
      "loss": 0.7037,
      "step": 5995
    },
    {
      "epoch": 1.72,
      "grad_norm": 11.603124618530273,
      "learning_rate": 4.273455377574371e-06,
      "loss": 0.73,
      "step": 5996
    },
    {
      "epoch": 1.72,
      "grad_norm": 12.37160587310791,
      "learning_rate": 4.269164759725401e-06,
      "loss": 0.5829,
      "step": 5997
    },
    {
      "epoch": 1.72,
      "grad_norm": 9.701993942260742,
      "learning_rate": 4.264874141876431e-06,
      "loss": 0.7356,
      "step": 5998
    },
    {
      "epoch": 1.72,
      "grad_norm": 10.731022834777832,
      "learning_rate": 4.26058352402746e-06,
      "loss": 0.7306,
      "step": 5999
    },
    {
      "epoch": 1.72,
      "grad_norm": 11.346711158752441,
      "learning_rate": 4.256292906178489e-06,
      "loss": 0.7553,
      "step": 6000
    },
    {
      "epoch": 1.72,
      "grad_norm": 10.900642395019531,
      "learning_rate": 4.2520022883295195e-06,
      "loss": 0.5546,
      "step": 6001
    },
    {
      "epoch": 1.72,
      "grad_norm": 13.566597938537598,
      "learning_rate": 4.247711670480549e-06,
      "loss": 0.8884,
      "step": 6002
    },
    {
      "epoch": 1.72,
      "grad_norm": 14.55006217956543,
      "learning_rate": 4.243421052631579e-06,
      "loss": 0.8857,
      "step": 6003
    },
    {
      "epoch": 1.72,
      "grad_norm": 12.602134704589844,
      "learning_rate": 4.239130434782608e-06,
      "loss": 0.8392,
      "step": 6004
    },
    {
      "epoch": 1.72,
      "grad_norm": 13.23162841796875,
      "learning_rate": 4.234839816933639e-06,
      "loss": 0.6262,
      "step": 6005
    },
    {
      "epoch": 1.72,
      "grad_norm": 11.039629936218262,
      "learning_rate": 4.230549199084668e-06,
      "loss": 0.6168,
      "step": 6006
    },
    {
      "epoch": 1.72,
      "grad_norm": 8.362481117248535,
      "learning_rate": 4.226258581235698e-06,
      "loss": 0.6102,
      "step": 6007
    },
    {
      "epoch": 1.72,
      "grad_norm": 7.873233318328857,
      "learning_rate": 4.221967963386728e-06,
      "loss": 0.6211,
      "step": 6008
    },
    {
      "epoch": 1.72,
      "grad_norm": 9.244438171386719,
      "learning_rate": 4.217677345537758e-06,
      "loss": 0.5464,
      "step": 6009
    },
    {
      "epoch": 1.72,
      "grad_norm": 11.47413444519043,
      "learning_rate": 4.213386727688788e-06,
      "loss": 0.7942,
      "step": 6010
    },
    {
      "epoch": 1.72,
      "grad_norm": 12.936746597290039,
      "learning_rate": 4.209096109839817e-06,
      "loss": 0.7059,
      "step": 6011
    },
    {
      "epoch": 1.72,
      "grad_norm": 11.094941139221191,
      "learning_rate": 4.204805491990847e-06,
      "loss": 0.4952,
      "step": 6012
    },
    {
      "epoch": 1.72,
      "grad_norm": 7.792799472808838,
      "learning_rate": 4.200514874141876e-06,
      "loss": 0.7102,
      "step": 6013
    },
    {
      "epoch": 1.72,
      "grad_norm": 11.127025604248047,
      "learning_rate": 4.196224256292906e-06,
      "loss": 0.78,
      "step": 6014
    },
    {
      "epoch": 1.72,
      "grad_norm": 9.540066719055176,
      "learning_rate": 4.191933638443936e-06,
      "loss": 0.7395,
      "step": 6015
    },
    {
      "epoch": 1.72,
      "grad_norm": 10.446444511413574,
      "learning_rate": 4.187643020594965e-06,
      "loss": 0.5893,
      "step": 6016
    },
    {
      "epoch": 1.72,
      "grad_norm": 10.989967346191406,
      "learning_rate": 4.183352402745995e-06,
      "loss": 0.58,
      "step": 6017
    },
    {
      "epoch": 1.72,
      "grad_norm": 10.281612396240234,
      "learning_rate": 4.1790617848970255e-06,
      "loss": 0.7098,
      "step": 6018
    },
    {
      "epoch": 1.72,
      "grad_norm": 11.980088233947754,
      "learning_rate": 4.174771167048055e-06,
      "loss": 0.5153,
      "step": 6019
    },
    {
      "epoch": 1.72,
      "grad_norm": 9.9744234085083,
      "learning_rate": 4.170480549199085e-06,
      "loss": 0.5124,
      "step": 6020
    },
    {
      "epoch": 1.72,
      "grad_norm": 9.606974601745605,
      "learning_rate": 4.166189931350114e-06,
      "loss": 0.6363,
      "step": 6021
    },
    {
      "epoch": 1.72,
      "grad_norm": 8.587697982788086,
      "learning_rate": 4.161899313501145e-06,
      "loss": 0.5235,
      "step": 6022
    },
    {
      "epoch": 1.72,
      "grad_norm": 10.671334266662598,
      "learning_rate": 4.157608695652174e-06,
      "loss": 0.7783,
      "step": 6023
    },
    {
      "epoch": 1.72,
      "grad_norm": 11.862448692321777,
      "learning_rate": 4.153318077803204e-06,
      "loss": 0.9054,
      "step": 6024
    },
    {
      "epoch": 1.72,
      "grad_norm": 9.895891189575195,
      "learning_rate": 4.149027459954234e-06,
      "loss": 0.5098,
      "step": 6025
    },
    {
      "epoch": 1.72,
      "grad_norm": 8.942290306091309,
      "learning_rate": 4.144736842105263e-06,
      "loss": 0.7021,
      "step": 6026
    },
    {
      "epoch": 1.72,
      "grad_norm": 10.103818893432617,
      "learning_rate": 4.140446224256293e-06,
      "loss": 0.5956,
      "step": 6027
    },
    {
      "epoch": 1.72,
      "grad_norm": 11.1176118850708,
      "learning_rate": 4.1361556064073225e-06,
      "loss": 0.8294,
      "step": 6028
    },
    {
      "epoch": 1.72,
      "grad_norm": 8.09962272644043,
      "learning_rate": 4.131864988558352e-06,
      "loss": 0.602,
      "step": 6029
    },
    {
      "epoch": 1.72,
      "grad_norm": 13.282578468322754,
      "learning_rate": 4.127574370709382e-06,
      "loss": 0.674,
      "step": 6030
    },
    {
      "epoch": 1.73,
      "grad_norm": 8.672886848449707,
      "learning_rate": 4.123283752860412e-06,
      "loss": 0.5272,
      "step": 6031
    },
    {
      "epoch": 1.73,
      "grad_norm": 11.450350761413574,
      "learning_rate": 4.118993135011442e-06,
      "loss": 0.9202,
      "step": 6032
    },
    {
      "epoch": 1.73,
      "grad_norm": 10.484188079833984,
      "learning_rate": 4.114702517162471e-06,
      "loss": 0.7453,
      "step": 6033
    },
    {
      "epoch": 1.73,
      "grad_norm": 11.441675186157227,
      "learning_rate": 4.110411899313501e-06,
      "loss": 0.716,
      "step": 6034
    },
    {
      "epoch": 1.73,
      "grad_norm": 9.349542617797852,
      "learning_rate": 4.1061212814645315e-06,
      "loss": 0.4711,
      "step": 6035
    },
    {
      "epoch": 1.73,
      "grad_norm": 7.599945068359375,
      "learning_rate": 4.101830663615561e-06,
      "loss": 0.5928,
      "step": 6036
    },
    {
      "epoch": 1.73,
      "grad_norm": 9.424768447875977,
      "learning_rate": 4.097540045766591e-06,
      "loss": 0.6022,
      "step": 6037
    },
    {
      "epoch": 1.73,
      "grad_norm": 16.08043670654297,
      "learning_rate": 4.09324942791762e-06,
      "loss": 0.7528,
      "step": 6038
    },
    {
      "epoch": 1.73,
      "grad_norm": 11.866475105285645,
      "learning_rate": 4.088958810068651e-06,
      "loss": 0.7812,
      "step": 6039
    },
    {
      "epoch": 1.73,
      "grad_norm": 7.43369197845459,
      "learning_rate": 4.0846681922196795e-06,
      "loss": 0.565,
      "step": 6040
    },
    {
      "epoch": 1.73,
      "grad_norm": 7.482633590698242,
      "learning_rate": 4.080377574370709e-06,
      "loss": 0.4834,
      "step": 6041
    },
    {
      "epoch": 1.73,
      "grad_norm": 12.829164505004883,
      "learning_rate": 4.076086956521739e-06,
      "loss": 0.6553,
      "step": 6042
    },
    {
      "epoch": 1.73,
      "grad_norm": 10.891663551330566,
      "learning_rate": 4.071796338672768e-06,
      "loss": 0.7566,
      "step": 6043
    },
    {
      "epoch": 1.73,
      "grad_norm": 10.123371124267578,
      "learning_rate": 4.067505720823799e-06,
      "loss": 0.7981,
      "step": 6044
    },
    {
      "epoch": 1.73,
      "grad_norm": 10.865095138549805,
      "learning_rate": 4.0632151029748285e-06,
      "loss": 0.9449,
      "step": 6045
    },
    {
      "epoch": 1.73,
      "grad_norm": 9.813224792480469,
      "learning_rate": 4.058924485125858e-06,
      "loss": 0.6667,
      "step": 6046
    },
    {
      "epoch": 1.73,
      "grad_norm": 10.570638656616211,
      "learning_rate": 4.054633867276888e-06,
      "loss": 0.7178,
      "step": 6047
    },
    {
      "epoch": 1.73,
      "grad_norm": 13.028220176696777,
      "learning_rate": 4.050343249427918e-06,
      "loss": 0.6327,
      "step": 6048
    },
    {
      "epoch": 1.73,
      "grad_norm": 9.264501571655273,
      "learning_rate": 4.046052631578948e-06,
      "loss": 0.6126,
      "step": 6049
    },
    {
      "epoch": 1.73,
      "grad_norm": 11.550825119018555,
      "learning_rate": 4.041762013729977e-06,
      "loss": 0.6428,
      "step": 6050
    },
    {
      "epoch": 1.73,
      "grad_norm": 10.864901542663574,
      "learning_rate": 4.037471395881007e-06,
      "loss": 0.5461,
      "step": 6051
    },
    {
      "epoch": 1.73,
      "grad_norm": 8.91755485534668,
      "learning_rate": 4.0331807780320375e-06,
      "loss": 0.3539,
      "step": 6052
    },
    {
      "epoch": 1.73,
      "grad_norm": 8.895212173461914,
      "learning_rate": 4.028890160183066e-06,
      "loss": 0.6364,
      "step": 6053
    },
    {
      "epoch": 1.73,
      "grad_norm": 10.656241416931152,
      "learning_rate": 4.024599542334096e-06,
      "loss": 0.8206,
      "step": 6054
    },
    {
      "epoch": 1.73,
      "grad_norm": 10.321805000305176,
      "learning_rate": 4.0203089244851255e-06,
      "loss": 0.6592,
      "step": 6055
    },
    {
      "epoch": 1.73,
      "grad_norm": 8.235122680664062,
      "learning_rate": 4.016018306636156e-06,
      "loss": 0.5673,
      "step": 6056
    },
    {
      "epoch": 1.73,
      "grad_norm": 10.012003898620605,
      "learning_rate": 4.0117276887871855e-06,
      "loss": 0.7307,
      "step": 6057
    },
    {
      "epoch": 1.73,
      "grad_norm": 13.317085266113281,
      "learning_rate": 4.007437070938215e-06,
      "loss": 0.8539,
      "step": 6058
    },
    {
      "epoch": 1.73,
      "grad_norm": 11.325658798217773,
      "learning_rate": 4.003146453089245e-06,
      "loss": 0.5153,
      "step": 6059
    },
    {
      "epoch": 1.73,
      "grad_norm": 11.494180679321289,
      "learning_rate": 3.998855835240274e-06,
      "loss": 0.863,
      "step": 6060
    },
    {
      "epoch": 1.73,
      "grad_norm": 11.76332950592041,
      "learning_rate": 3.994565217391305e-06,
      "loss": 0.9077,
      "step": 6061
    },
    {
      "epoch": 1.73,
      "grad_norm": 8.529455184936523,
      "learning_rate": 3.9902745995423345e-06,
      "loss": 0.5149,
      "step": 6062
    },
    {
      "epoch": 1.73,
      "grad_norm": 8.060349464416504,
      "learning_rate": 3.985983981693364e-06,
      "loss": 0.5088,
      "step": 6063
    },
    {
      "epoch": 1.73,
      "grad_norm": 9.382649421691895,
      "learning_rate": 3.981693363844394e-06,
      "loss": 0.5403,
      "step": 6064
    },
    {
      "epoch": 1.73,
      "grad_norm": 10.23032283782959,
      "learning_rate": 3.977402745995424e-06,
      "loss": 0.7574,
      "step": 6065
    },
    {
      "epoch": 1.74,
      "grad_norm": 11.511918067932129,
      "learning_rate": 3.973112128146453e-06,
      "loss": 0.7237,
      "step": 6066
    },
    {
      "epoch": 1.74,
      "grad_norm": 9.262679100036621,
      "learning_rate": 3.9688215102974825e-06,
      "loss": 0.7081,
      "step": 6067
    },
    {
      "epoch": 1.74,
      "grad_norm": 10.52244758605957,
      "learning_rate": 3.964530892448512e-06,
      "loss": 0.8418,
      "step": 6068
    },
    {
      "epoch": 1.74,
      "grad_norm": 8.32150936126709,
      "learning_rate": 3.960240274599543e-06,
      "loss": 0.5113,
      "step": 6069
    },
    {
      "epoch": 1.74,
      "grad_norm": 10.85783576965332,
      "learning_rate": 3.955949656750572e-06,
      "loss": 0.5971,
      "step": 6070
    },
    {
      "epoch": 1.74,
      "grad_norm": 12.290858268737793,
      "learning_rate": 3.951659038901602e-06,
      "loss": 0.7818,
      "step": 6071
    },
    {
      "epoch": 1.74,
      "grad_norm": 10.787084579467773,
      "learning_rate": 3.9473684210526315e-06,
      "loss": 0.5963,
      "step": 6072
    },
    {
      "epoch": 1.74,
      "grad_norm": 9.790975570678711,
      "learning_rate": 3.943077803203661e-06,
      "loss": 0.7175,
      "step": 6073
    },
    {
      "epoch": 1.74,
      "grad_norm": 10.893580436706543,
      "learning_rate": 3.9387871853546915e-06,
      "loss": 0.8964,
      "step": 6074
    },
    {
      "epoch": 1.74,
      "grad_norm": 11.846661567687988,
      "learning_rate": 3.934496567505721e-06,
      "loss": 0.6185,
      "step": 6075
    },
    {
      "epoch": 1.74,
      "grad_norm": 10.663214683532715,
      "learning_rate": 3.930205949656751e-06,
      "loss": 0.5727,
      "step": 6076
    },
    {
      "epoch": 1.74,
      "grad_norm": 11.027164459228516,
      "learning_rate": 3.92591533180778e-06,
      "loss": 0.7161,
      "step": 6077
    },
    {
      "epoch": 1.74,
      "grad_norm": 7.9825592041015625,
      "learning_rate": 3.921624713958811e-06,
      "loss": 0.3647,
      "step": 6078
    },
    {
      "epoch": 1.74,
      "grad_norm": 8.15426254272461,
      "learning_rate": 3.91733409610984e-06,
      "loss": 0.4139,
      "step": 6079
    },
    {
      "epoch": 1.74,
      "grad_norm": 8.603719711303711,
      "learning_rate": 3.913043478260869e-06,
      "loss": 0.6229,
      "step": 6080
    },
    {
      "epoch": 1.74,
      "grad_norm": 9.28713607788086,
      "learning_rate": 3.908752860411899e-06,
      "loss": 0.5782,
      "step": 6081
    },
    {
      "epoch": 1.74,
      "grad_norm": 9.771931648254395,
      "learning_rate": 3.904462242562929e-06,
      "loss": 0.5709,
      "step": 6082
    },
    {
      "epoch": 1.74,
      "grad_norm": 10.751740455627441,
      "learning_rate": 3.900171624713959e-06,
      "loss": 0.6636,
      "step": 6083
    },
    {
      "epoch": 1.74,
      "grad_norm": 11.382894515991211,
      "learning_rate": 3.8958810068649885e-06,
      "loss": 0.5822,
      "step": 6084
    },
    {
      "epoch": 1.74,
      "grad_norm": 13.858905792236328,
      "learning_rate": 3.891590389016018e-06,
      "loss": 0.9624,
      "step": 6085
    },
    {
      "epoch": 1.74,
      "grad_norm": 10.064650535583496,
      "learning_rate": 3.887299771167049e-06,
      "loss": 0.6448,
      "step": 6086
    },
    {
      "epoch": 1.74,
      "grad_norm": 8.4977445602417,
      "learning_rate": 3.883009153318078e-06,
      "loss": 0.466,
      "step": 6087
    },
    {
      "epoch": 1.74,
      "grad_norm": 13.306116104125977,
      "learning_rate": 3.878718535469108e-06,
      "loss": 0.7704,
      "step": 6088
    },
    {
      "epoch": 1.74,
      "grad_norm": 9.628791809082031,
      "learning_rate": 3.8744279176201375e-06,
      "loss": 0.4445,
      "step": 6089
    },
    {
      "epoch": 1.74,
      "grad_norm": 9.335134506225586,
      "learning_rate": 3.870137299771167e-06,
      "loss": 0.6053,
      "step": 6090
    },
    {
      "epoch": 1.74,
      "grad_norm": 13.88830852508545,
      "learning_rate": 3.8658466819221975e-06,
      "loss": 0.7347,
      "step": 6091
    },
    {
      "epoch": 1.74,
      "grad_norm": 8.299972534179688,
      "learning_rate": 3.861556064073226e-06,
      "loss": 0.3696,
      "step": 6092
    },
    {
      "epoch": 1.74,
      "grad_norm": 12.010845184326172,
      "learning_rate": 3.857265446224256e-06,
      "loss": 0.7255,
      "step": 6093
    },
    {
      "epoch": 1.74,
      "grad_norm": 12.07990837097168,
      "learning_rate": 3.8529748283752855e-06,
      "loss": 0.6944,
      "step": 6094
    },
    {
      "epoch": 1.74,
      "grad_norm": 10.700096130371094,
      "learning_rate": 3.848684210526316e-06,
      "loss": 0.764,
      "step": 6095
    },
    {
      "epoch": 1.74,
      "grad_norm": 9.16130542755127,
      "learning_rate": 3.844393592677346e-06,
      "loss": 0.536,
      "step": 6096
    },
    {
      "epoch": 1.74,
      "grad_norm": 10.576375007629395,
      "learning_rate": 3.840102974828375e-06,
      "loss": 0.6338,
      "step": 6097
    },
    {
      "epoch": 1.74,
      "grad_norm": 12.437286376953125,
      "learning_rate": 3.835812356979405e-06,
      "loss": 0.8559,
      "step": 6098
    },
    {
      "epoch": 1.74,
      "grad_norm": 10.215106964111328,
      "learning_rate": 3.831521739130435e-06,
      "loss": 0.7449,
      "step": 6099
    },
    {
      "epoch": 1.74,
      "grad_norm": 10.478534698486328,
      "learning_rate": 3.827231121281465e-06,
      "loss": 0.6246,
      "step": 6100
    },
    {
      "epoch": 1.75,
      "grad_norm": 12.23443603515625,
      "learning_rate": 3.8229405034324945e-06,
      "loss": 0.582,
      "step": 6101
    },
    {
      "epoch": 1.75,
      "grad_norm": 11.197013854980469,
      "learning_rate": 3.818649885583524e-06,
      "loss": 0.6055,
      "step": 6102
    },
    {
      "epoch": 1.75,
      "grad_norm": 9.7366361618042,
      "learning_rate": 3.8143592677345533e-06,
      "loss": 0.7919,
      "step": 6103
    },
    {
      "epoch": 1.75,
      "grad_norm": 10.818952560424805,
      "learning_rate": 3.810068649885584e-06,
      "loss": 0.7185,
      "step": 6104
    },
    {
      "epoch": 1.75,
      "grad_norm": 13.971001625061035,
      "learning_rate": 3.8057780320366134e-06,
      "loss": 1.0051,
      "step": 6105
    },
    {
      "epoch": 1.75,
      "grad_norm": 9.492419242858887,
      "learning_rate": 3.801487414187643e-06,
      "loss": 0.5068,
      "step": 6106
    },
    {
      "epoch": 1.75,
      "grad_norm": 9.137680053710938,
      "learning_rate": 3.7971967963386726e-06,
      "loss": 0.5911,
      "step": 6107
    },
    {
      "epoch": 1.75,
      "grad_norm": 10.991887092590332,
      "learning_rate": 3.792906178489703e-06,
      "loss": 0.6473,
      "step": 6108
    },
    {
      "epoch": 1.75,
      "grad_norm": 8.027948379516602,
      "learning_rate": 3.7886155606407323e-06,
      "loss": 0.5314,
      "step": 6109
    },
    {
      "epoch": 1.75,
      "grad_norm": 11.149654388427734,
      "learning_rate": 3.784324942791762e-06,
      "loss": 0.6914,
      "step": 6110
    },
    {
      "epoch": 1.75,
      "grad_norm": 9.29753589630127,
      "learning_rate": 3.7800343249427915e-06,
      "loss": 0.4906,
      "step": 6111
    },
    {
      "epoch": 1.75,
      "grad_norm": 10.15299129486084,
      "learning_rate": 3.775743707093822e-06,
      "loss": 0.5583,
      "step": 6112
    },
    {
      "epoch": 1.75,
      "grad_norm": 10.863149642944336,
      "learning_rate": 3.7714530892448516e-06,
      "loss": 0.7952,
      "step": 6113
    },
    {
      "epoch": 1.75,
      "grad_norm": 13.928953170776367,
      "learning_rate": 3.7671624713958812e-06,
      "loss": 0.6771,
      "step": 6114
    },
    {
      "epoch": 1.75,
      "grad_norm": 12.386027336120605,
      "learning_rate": 3.7628718535469104e-06,
      "loss": 0.8803,
      "step": 6115
    },
    {
      "epoch": 1.75,
      "grad_norm": 9.897882461547852,
      "learning_rate": 3.758581235697941e-06,
      "loss": 0.7667,
      "step": 6116
    },
    {
      "epoch": 1.75,
      "grad_norm": 10.644085884094238,
      "learning_rate": 3.7542906178489705e-06,
      "loss": 0.7295,
      "step": 6117
    },
    {
      "epoch": 1.75,
      "grad_norm": 8.818511009216309,
      "learning_rate": 3.75e-06,
      "loss": 0.6042,
      "step": 6118
    },
    {
      "epoch": 1.75,
      "grad_norm": 9.307840347290039,
      "learning_rate": 3.7457093821510297e-06,
      "loss": 0.6812,
      "step": 6119
    },
    {
      "epoch": 1.75,
      "grad_norm": 13.10869026184082,
      "learning_rate": 3.7414187643020597e-06,
      "loss": 0.6901,
      "step": 6120
    },
    {
      "epoch": 1.75,
      "grad_norm": 12.035422325134277,
      "learning_rate": 3.7371281464530894e-06,
      "loss": 0.6088,
      "step": 6121
    },
    {
      "epoch": 1.75,
      "grad_norm": 9.764453887939453,
      "learning_rate": 3.732837528604119e-06,
      "loss": 0.7152,
      "step": 6122
    },
    {
      "epoch": 1.75,
      "grad_norm": 12.01958179473877,
      "learning_rate": 3.7285469107551486e-06,
      "loss": 0.7195,
      "step": 6123
    },
    {
      "epoch": 1.75,
      "grad_norm": 11.711091041564941,
      "learning_rate": 3.7242562929061786e-06,
      "loss": 0.7594,
      "step": 6124
    },
    {
      "epoch": 1.75,
      "grad_norm": 12.01634407043457,
      "learning_rate": 3.7199656750572082e-06,
      "loss": 0.5769,
      "step": 6125
    },
    {
      "epoch": 1.75,
      "grad_norm": 10.122314453125,
      "learning_rate": 3.7156750572082383e-06,
      "loss": 0.6724,
      "step": 6126
    },
    {
      "epoch": 1.75,
      "grad_norm": 11.244257926940918,
      "learning_rate": 3.711384439359268e-06,
      "loss": 0.5739,
      "step": 6127
    },
    {
      "epoch": 1.75,
      "grad_norm": 11.775614738464355,
      "learning_rate": 3.7070938215102975e-06,
      "loss": 0.6301,
      "step": 6128
    },
    {
      "epoch": 1.75,
      "grad_norm": 9.514596939086914,
      "learning_rate": 3.702803203661327e-06,
      "loss": 0.6,
      "step": 6129
    },
    {
      "epoch": 1.75,
      "grad_norm": 10.512495040893555,
      "learning_rate": 3.698512585812357e-06,
      "loss": 0.7625,
      "step": 6130
    },
    {
      "epoch": 1.75,
      "grad_norm": 9.708354949951172,
      "learning_rate": 3.6942219679633868e-06,
      "loss": 0.6674,
      "step": 6131
    },
    {
      "epoch": 1.75,
      "grad_norm": 11.710132598876953,
      "learning_rate": 3.689931350114417e-06,
      "loss": 0.7033,
      "step": 6132
    },
    {
      "epoch": 1.75,
      "grad_norm": 15.638635635375977,
      "learning_rate": 3.6856407322654464e-06,
      "loss": 0.7225,
      "step": 6133
    },
    {
      "epoch": 1.75,
      "grad_norm": 10.383947372436523,
      "learning_rate": 3.681350114416476e-06,
      "loss": 0.4643,
      "step": 6134
    },
    {
      "epoch": 1.75,
      "grad_norm": 12.879046440124512,
      "learning_rate": 3.6770594965675057e-06,
      "loss": 0.6736,
      "step": 6135
    },
    {
      "epoch": 1.76,
      "grad_norm": 10.509659767150879,
      "learning_rate": 3.6727688787185353e-06,
      "loss": 0.532,
      "step": 6136
    },
    {
      "epoch": 1.76,
      "grad_norm": 8.082070350646973,
      "learning_rate": 3.6684782608695653e-06,
      "loss": 0.5275,
      "step": 6137
    },
    {
      "epoch": 1.76,
      "grad_norm": 11.578171730041504,
      "learning_rate": 3.664187643020595e-06,
      "loss": 0.7137,
      "step": 6138
    },
    {
      "epoch": 1.76,
      "grad_norm": 10.211572647094727,
      "learning_rate": 3.659897025171625e-06,
      "loss": 0.5826,
      "step": 6139
    },
    {
      "epoch": 1.76,
      "grad_norm": 9.431300163269043,
      "learning_rate": 3.6556064073226546e-06,
      "loss": 0.5547,
      "step": 6140
    },
    {
      "epoch": 1.76,
      "grad_norm": 11.96328353881836,
      "learning_rate": 3.651315789473684e-06,
      "loss": 0.7601,
      "step": 6141
    },
    {
      "epoch": 1.76,
      "grad_norm": 11.621810913085938,
      "learning_rate": 3.647025171624714e-06,
      "loss": 0.7721,
      "step": 6142
    },
    {
      "epoch": 1.76,
      "grad_norm": 11.324017524719238,
      "learning_rate": 3.642734553775744e-06,
      "loss": 0.7112,
      "step": 6143
    },
    {
      "epoch": 1.76,
      "grad_norm": 10.572038650512695,
      "learning_rate": 3.6384439359267735e-06,
      "loss": 0.7062,
      "step": 6144
    },
    {
      "epoch": 1.76,
      "grad_norm": 12.122496604919434,
      "learning_rate": 3.6341533180778035e-06,
      "loss": 0.6158,
      "step": 6145
    },
    {
      "epoch": 1.76,
      "grad_norm": 13.798746109008789,
      "learning_rate": 3.629862700228833e-06,
      "loss": 0.9658,
      "step": 6146
    },
    {
      "epoch": 1.76,
      "grad_norm": 12.25180435180664,
      "learning_rate": 3.625572082379863e-06,
      "loss": 0.6471,
      "step": 6147
    },
    {
      "epoch": 1.76,
      "grad_norm": 8.908062934875488,
      "learning_rate": 3.6212814645308924e-06,
      "loss": 0.5552,
      "step": 6148
    },
    {
      "epoch": 1.76,
      "grad_norm": 11.949036598205566,
      "learning_rate": 3.616990846681922e-06,
      "loss": 0.6593,
      "step": 6149
    },
    {
      "epoch": 1.76,
      "grad_norm": 9.419204711914062,
      "learning_rate": 3.612700228832952e-06,
      "loss": 0.6012,
      "step": 6150
    },
    {
      "epoch": 1.76,
      "grad_norm": 10.818522453308105,
      "learning_rate": 3.6084096109839816e-06,
      "loss": 0.7202,
      "step": 6151
    },
    {
      "epoch": 1.76,
      "grad_norm": 10.00659465789795,
      "learning_rate": 3.6041189931350117e-06,
      "loss": 0.4724,
      "step": 6152
    },
    {
      "epoch": 1.76,
      "grad_norm": 9.23212718963623,
      "learning_rate": 3.5998283752860413e-06,
      "loss": 0.7177,
      "step": 6153
    },
    {
      "epoch": 1.76,
      "grad_norm": 10.0756196975708,
      "learning_rate": 3.5955377574370713e-06,
      "loss": 0.714,
      "step": 6154
    },
    {
      "epoch": 1.76,
      "grad_norm": 10.584141731262207,
      "learning_rate": 3.5912471395881005e-06,
      "loss": 0.4958,
      "step": 6155
    },
    {
      "epoch": 1.76,
      "grad_norm": 9.411657333374023,
      "learning_rate": 3.5869565217391305e-06,
      "loss": 0.578,
      "step": 6156
    },
    {
      "epoch": 1.76,
      "grad_norm": 10.629715919494629,
      "learning_rate": 3.58266590389016e-06,
      "loss": 0.5276,
      "step": 6157
    },
    {
      "epoch": 1.76,
      "grad_norm": 10.674521446228027,
      "learning_rate": 3.57837528604119e-06,
      "loss": 0.6469,
      "step": 6158
    },
    {
      "epoch": 1.76,
      "grad_norm": 8.006171226501465,
      "learning_rate": 3.57408466819222e-06,
      "loss": 0.4589,
      "step": 6159
    },
    {
      "epoch": 1.76,
      "grad_norm": 10.794471740722656,
      "learning_rate": 3.56979405034325e-06,
      "loss": 0.7086,
      "step": 6160
    },
    {
      "epoch": 1.76,
      "grad_norm": 9.56449031829834,
      "learning_rate": 3.565503432494279e-06,
      "loss": 0.6166,
      "step": 6161
    },
    {
      "epoch": 1.76,
      "grad_norm": 11.872832298278809,
      "learning_rate": 3.561212814645309e-06,
      "loss": 0.5607,
      "step": 6162
    },
    {
      "epoch": 1.76,
      "grad_norm": 11.3371000289917,
      "learning_rate": 3.5569221967963387e-06,
      "loss": 0.6424,
      "step": 6163
    },
    {
      "epoch": 1.76,
      "grad_norm": 9.539777755737305,
      "learning_rate": 3.5526315789473683e-06,
      "loss": 0.5611,
      "step": 6164
    },
    {
      "epoch": 1.76,
      "grad_norm": 10.240815162658691,
      "learning_rate": 3.5483409610983983e-06,
      "loss": 0.7481,
      "step": 6165
    },
    {
      "epoch": 1.76,
      "grad_norm": 11.382087707519531,
      "learning_rate": 3.544050343249428e-06,
      "loss": 0.6187,
      "step": 6166
    },
    {
      "epoch": 1.76,
      "grad_norm": 10.351128578186035,
      "learning_rate": 3.539759725400458e-06,
      "loss": 0.6702,
      "step": 6167
    },
    {
      "epoch": 1.76,
      "grad_norm": 8.965996742248535,
      "learning_rate": 3.535469107551487e-06,
      "loss": 0.7342,
      "step": 6168
    },
    {
      "epoch": 1.76,
      "grad_norm": 12.034710884094238,
      "learning_rate": 3.5311784897025172e-06,
      "loss": 0.7361,
      "step": 6169
    },
    {
      "epoch": 1.76,
      "grad_norm": 11.237125396728516,
      "learning_rate": 3.526887871853547e-06,
      "loss": 0.7888,
      "step": 6170
    },
    {
      "epoch": 1.77,
      "grad_norm": 11.04775333404541,
      "learning_rate": 3.522597254004577e-06,
      "loss": 0.7907,
      "step": 6171
    },
    {
      "epoch": 1.77,
      "grad_norm": 11.389314651489258,
      "learning_rate": 3.5183066361556065e-06,
      "loss": 0.7913,
      "step": 6172
    },
    {
      "epoch": 1.77,
      "grad_norm": 10.503778457641602,
      "learning_rate": 3.5140160183066365e-06,
      "loss": 0.5925,
      "step": 6173
    },
    {
      "epoch": 1.77,
      "grad_norm": 11.094098091125488,
      "learning_rate": 3.5097254004576657e-06,
      "loss": 0.6376,
      "step": 6174
    },
    {
      "epoch": 1.77,
      "grad_norm": 9.925131797790527,
      "learning_rate": 3.5054347826086958e-06,
      "loss": 0.4822,
      "step": 6175
    },
    {
      "epoch": 1.77,
      "grad_norm": 9.414146423339844,
      "learning_rate": 3.5011441647597254e-06,
      "loss": 0.5317,
      "step": 6176
    },
    {
      "epoch": 1.77,
      "grad_norm": 13.774288177490234,
      "learning_rate": 3.4968535469107554e-06,
      "loss": 0.861,
      "step": 6177
    },
    {
      "epoch": 1.77,
      "grad_norm": 10.283080101013184,
      "learning_rate": 3.492562929061785e-06,
      "loss": 0.6522,
      "step": 6178
    },
    {
      "epoch": 1.77,
      "grad_norm": 10.067567825317383,
      "learning_rate": 3.4882723112128147e-06,
      "loss": 0.5531,
      "step": 6179
    },
    {
      "epoch": 1.77,
      "grad_norm": 12.848753929138184,
      "learning_rate": 3.4839816933638447e-06,
      "loss": 0.7347,
      "step": 6180
    },
    {
      "epoch": 1.77,
      "grad_norm": 9.71597671508789,
      "learning_rate": 3.479691075514874e-06,
      "loss": 0.5949,
      "step": 6181
    },
    {
      "epoch": 1.77,
      "grad_norm": 13.107844352722168,
      "learning_rate": 3.475400457665904e-06,
      "loss": 0.5409,
      "step": 6182
    },
    {
      "epoch": 1.77,
      "grad_norm": 7.763223648071289,
      "learning_rate": 3.4711098398169335e-06,
      "loss": 0.564,
      "step": 6183
    },
    {
      "epoch": 1.77,
      "grad_norm": 12.058732032775879,
      "learning_rate": 3.4668192219679636e-06,
      "loss": 0.7095,
      "step": 6184
    },
    {
      "epoch": 1.77,
      "grad_norm": 10.082940101623535,
      "learning_rate": 3.462528604118993e-06,
      "loss": 0.586,
      "step": 6185
    },
    {
      "epoch": 1.77,
      "grad_norm": 11.730283737182617,
      "learning_rate": 3.4582379862700232e-06,
      "loss": 0.6136,
      "step": 6186
    },
    {
      "epoch": 1.77,
      "grad_norm": 7.0045928955078125,
      "learning_rate": 3.453947368421053e-06,
      "loss": 0.5827,
      "step": 6187
    },
    {
      "epoch": 1.77,
      "grad_norm": 15.763656616210938,
      "learning_rate": 3.4496567505720825e-06,
      "loss": 0.8763,
      "step": 6188
    },
    {
      "epoch": 1.77,
      "grad_norm": 9.833060264587402,
      "learning_rate": 3.445366132723112e-06,
      "loss": 0.8132,
      "step": 6189
    },
    {
      "epoch": 1.77,
      "grad_norm": 10.767908096313477,
      "learning_rate": 3.441075514874142e-06,
      "loss": 0.7116,
      "step": 6190
    },
    {
      "epoch": 1.77,
      "grad_norm": 9.660247802734375,
      "learning_rate": 3.4367848970251717e-06,
      "loss": 0.6864,
      "step": 6191
    },
    {
      "epoch": 1.77,
      "grad_norm": 9.962508201599121,
      "learning_rate": 3.4324942791762018e-06,
      "loss": 0.7094,
      "step": 6192
    },
    {
      "epoch": 1.77,
      "grad_norm": 9.406431198120117,
      "learning_rate": 3.4282036613272314e-06,
      "loss": 0.4943,
      "step": 6193
    },
    {
      "epoch": 1.77,
      "grad_norm": 13.485053062438965,
      "learning_rate": 3.4239130434782606e-06,
      "loss": 1.0711,
      "step": 6194
    },
    {
      "epoch": 1.77,
      "grad_norm": 11.053961753845215,
      "learning_rate": 3.4196224256292906e-06,
      "loss": 0.6344,
      "step": 6195
    },
    {
      "epoch": 1.77,
      "grad_norm": 9.669668197631836,
      "learning_rate": 3.4153318077803202e-06,
      "loss": 0.5333,
      "step": 6196
    },
    {
      "epoch": 1.77,
      "grad_norm": 9.433085441589355,
      "learning_rate": 3.4110411899313503e-06,
      "loss": 0.6433,
      "step": 6197
    },
    {
      "epoch": 1.77,
      "grad_norm": 8.06063175201416,
      "learning_rate": 3.40675057208238e-06,
      "loss": 0.4725,
      "step": 6198
    },
    {
      "epoch": 1.77,
      "grad_norm": 11.81443977355957,
      "learning_rate": 3.40245995423341e-06,
      "loss": 0.8544,
      "step": 6199
    },
    {
      "epoch": 1.77,
      "grad_norm": 8.285453796386719,
      "learning_rate": 3.3981693363844395e-06,
      "loss": 0.5734,
      "step": 6200
    },
    {
      "epoch": 1.77,
      "grad_norm": 10.867969512939453,
      "learning_rate": 3.393878718535469e-06,
      "loss": 0.7148,
      "step": 6201
    },
    {
      "epoch": 1.77,
      "grad_norm": 10.558439254760742,
      "learning_rate": 3.3895881006864988e-06,
      "loss": 0.8534,
      "step": 6202
    },
    {
      "epoch": 1.77,
      "grad_norm": 9.229147911071777,
      "learning_rate": 3.385297482837529e-06,
      "loss": 0.6502,
      "step": 6203
    },
    {
      "epoch": 1.77,
      "grad_norm": 10.644229888916016,
      "learning_rate": 3.3810068649885584e-06,
      "loss": 0.8017,
      "step": 6204
    },
    {
      "epoch": 1.77,
      "grad_norm": 10.682366371154785,
      "learning_rate": 3.3767162471395885e-06,
      "loss": 0.6322,
      "step": 6205
    },
    {
      "epoch": 1.78,
      "grad_norm": 13.435039520263672,
      "learning_rate": 3.372425629290618e-06,
      "loss": 0.6487,
      "step": 6206
    },
    {
      "epoch": 1.78,
      "grad_norm": 11.46129035949707,
      "learning_rate": 3.3681350114416477e-06,
      "loss": 0.607,
      "step": 6207
    },
    {
      "epoch": 1.78,
      "grad_norm": 10.996201515197754,
      "learning_rate": 3.3638443935926773e-06,
      "loss": 0.5365,
      "step": 6208
    },
    {
      "epoch": 1.78,
      "grad_norm": 11.303156852722168,
      "learning_rate": 3.359553775743707e-06,
      "loss": 0.6161,
      "step": 6209
    },
    {
      "epoch": 1.78,
      "grad_norm": 10.77161693572998,
      "learning_rate": 3.355263157894737e-06,
      "loss": 0.5759,
      "step": 6210
    },
    {
      "epoch": 1.78,
      "grad_norm": 10.97195816040039,
      "learning_rate": 3.3509725400457666e-06,
      "loss": 0.5865,
      "step": 6211
    },
    {
      "epoch": 1.78,
      "grad_norm": 12.53463077545166,
      "learning_rate": 3.3466819221967966e-06,
      "loss": 0.7462,
      "step": 6212
    },
    {
      "epoch": 1.78,
      "grad_norm": 10.610153198242188,
      "learning_rate": 3.3423913043478262e-06,
      "loss": 0.65,
      "step": 6213
    },
    {
      "epoch": 1.78,
      "grad_norm": 10.48507022857666,
      "learning_rate": 3.338100686498856e-06,
      "loss": 0.6284,
      "step": 6214
    },
    {
      "epoch": 1.78,
      "grad_norm": 10.51944351196289,
      "learning_rate": 3.3338100686498854e-06,
      "loss": 0.7416,
      "step": 6215
    },
    {
      "epoch": 1.78,
      "grad_norm": 12.190573692321777,
      "learning_rate": 3.3295194508009155e-06,
      "loss": 0.6409,
      "step": 6216
    },
    {
      "epoch": 1.78,
      "grad_norm": 9.088427543640137,
      "learning_rate": 3.325228832951945e-06,
      "loss": 0.6849,
      "step": 6217
    },
    {
      "epoch": 1.78,
      "grad_norm": 7.932781219482422,
      "learning_rate": 3.320938215102975e-06,
      "loss": 0.5349,
      "step": 6218
    },
    {
      "epoch": 1.78,
      "grad_norm": 10.231762886047363,
      "learning_rate": 3.3166475972540048e-06,
      "loss": 0.6828,
      "step": 6219
    },
    {
      "epoch": 1.78,
      "grad_norm": 11.04877758026123,
      "learning_rate": 3.3123569794050344e-06,
      "loss": 0.6749,
      "step": 6220
    },
    {
      "epoch": 1.78,
      "grad_norm": 6.348419189453125,
      "learning_rate": 3.308066361556064e-06,
      "loss": 0.3623,
      "step": 6221
    },
    {
      "epoch": 1.78,
      "grad_norm": 9.415176391601562,
      "learning_rate": 3.303775743707094e-06,
      "loss": 0.5816,
      "step": 6222
    },
    {
      "epoch": 1.78,
      "grad_norm": 9.620346069335938,
      "learning_rate": 3.2994851258581236e-06,
      "loss": 0.5649,
      "step": 6223
    },
    {
      "epoch": 1.78,
      "grad_norm": 11.188264846801758,
      "learning_rate": 3.2951945080091533e-06,
      "loss": 0.7468,
      "step": 6224
    },
    {
      "epoch": 1.78,
      "grad_norm": 11.869451522827148,
      "learning_rate": 3.2909038901601833e-06,
      "loss": 0.9742,
      "step": 6225
    },
    {
      "epoch": 1.78,
      "grad_norm": 12.979801177978516,
      "learning_rate": 3.286613272311213e-06,
      "loss": 0.6048,
      "step": 6226
    },
    {
      "epoch": 1.78,
      "grad_norm": 12.555614471435547,
      "learning_rate": 3.2823226544622425e-06,
      "loss": 0.7426,
      "step": 6227
    },
    {
      "epoch": 1.78,
      "grad_norm": 9.1586275100708,
      "learning_rate": 3.278032036613272e-06,
      "loss": 0.5555,
      "step": 6228
    },
    {
      "epoch": 1.78,
      "grad_norm": 10.315823554992676,
      "learning_rate": 3.273741418764302e-06,
      "loss": 0.5908,
      "step": 6229
    },
    {
      "epoch": 1.78,
      "grad_norm": 8.977620124816895,
      "learning_rate": 3.269450800915332e-06,
      "loss": 0.638,
      "step": 6230
    },
    {
      "epoch": 1.78,
      "grad_norm": 12.179141998291016,
      "learning_rate": 3.265160183066362e-06,
      "loss": 0.6091,
      "step": 6231
    },
    {
      "epoch": 1.78,
      "grad_norm": 10.679584503173828,
      "learning_rate": 3.2608695652173914e-06,
      "loss": 0.546,
      "step": 6232
    },
    {
      "epoch": 1.78,
      "grad_norm": 8.82325267791748,
      "learning_rate": 3.256578947368421e-06,
      "loss": 0.7694,
      "step": 6233
    },
    {
      "epoch": 1.78,
      "grad_norm": 7.902207374572754,
      "learning_rate": 3.2522883295194507e-06,
      "loss": 0.6916,
      "step": 6234
    },
    {
      "epoch": 1.78,
      "grad_norm": 9.770059585571289,
      "learning_rate": 3.2479977116704807e-06,
      "loss": 0.6605,
      "step": 6235
    },
    {
      "epoch": 1.78,
      "grad_norm": 9.278421401977539,
      "learning_rate": 3.2437070938215103e-06,
      "loss": 0.5572,
      "step": 6236
    },
    {
      "epoch": 1.78,
      "grad_norm": 10.935799598693848,
      "learning_rate": 3.2394164759725404e-06,
      "loss": 0.6734,
      "step": 6237
    },
    {
      "epoch": 1.78,
      "grad_norm": 13.630748748779297,
      "learning_rate": 3.23512585812357e-06,
      "loss": 0.744,
      "step": 6238
    },
    {
      "epoch": 1.78,
      "grad_norm": 10.820865631103516,
      "learning_rate": 3.2308352402745996e-06,
      "loss": 0.6496,
      "step": 6239
    },
    {
      "epoch": 1.78,
      "grad_norm": 12.565092086791992,
      "learning_rate": 3.226544622425629e-06,
      "loss": 0.8417,
      "step": 6240
    },
    {
      "epoch": 1.79,
      "grad_norm": 10.59786319732666,
      "learning_rate": 3.222254004576659e-06,
      "loss": 0.6452,
      "step": 6241
    },
    {
      "epoch": 1.79,
      "grad_norm": 11.53398323059082,
      "learning_rate": 3.217963386727689e-06,
      "loss": 0.6539,
      "step": 6242
    },
    {
      "epoch": 1.79,
      "grad_norm": 9.137362480163574,
      "learning_rate": 3.2136727688787185e-06,
      "loss": 0.6683,
      "step": 6243
    },
    {
      "epoch": 1.79,
      "grad_norm": 9.848520278930664,
      "learning_rate": 3.2093821510297485e-06,
      "loss": 0.7281,
      "step": 6244
    },
    {
      "epoch": 1.79,
      "grad_norm": 12.271465301513672,
      "learning_rate": 3.205091533180778e-06,
      "loss": 0.5905,
      "step": 6245
    },
    {
      "epoch": 1.79,
      "grad_norm": 8.549403190612793,
      "learning_rate": 3.200800915331808e-06,
      "loss": 0.7741,
      "step": 6246
    },
    {
      "epoch": 1.79,
      "grad_norm": 10.232514381408691,
      "learning_rate": 3.1965102974828374e-06,
      "loss": 0.5319,
      "step": 6247
    },
    {
      "epoch": 1.79,
      "grad_norm": 8.45359992980957,
      "learning_rate": 3.1922196796338674e-06,
      "loss": 0.6075,
      "step": 6248
    },
    {
      "epoch": 1.79,
      "grad_norm": 9.884468078613281,
      "learning_rate": 3.187929061784897e-06,
      "loss": 0.6882,
      "step": 6249
    },
    {
      "epoch": 1.79,
      "grad_norm": 11.395963668823242,
      "learning_rate": 3.183638443935927e-06,
      "loss": 0.5281,
      "step": 6250
    },
    {
      "epoch": 1.79,
      "grad_norm": 10.683157920837402,
      "learning_rate": 3.1793478260869567e-06,
      "loss": 0.7669,
      "step": 6251
    },
    {
      "epoch": 1.79,
      "grad_norm": 10.354310989379883,
      "learning_rate": 3.1750572082379867e-06,
      "loss": 0.6731,
      "step": 6252
    },
    {
      "epoch": 1.79,
      "grad_norm": 11.60689640045166,
      "learning_rate": 3.170766590389016e-06,
      "loss": 0.6892,
      "step": 6253
    },
    {
      "epoch": 1.79,
      "grad_norm": 10.987874984741211,
      "learning_rate": 3.1664759725400455e-06,
      "loss": 0.6374,
      "step": 6254
    },
    {
      "epoch": 1.79,
      "grad_norm": 10.786839485168457,
      "learning_rate": 3.1621853546910755e-06,
      "loss": 0.5039,
      "step": 6255
    },
    {
      "epoch": 1.79,
      "grad_norm": 13.27176284790039,
      "learning_rate": 3.157894736842105e-06,
      "loss": 0.6787,
      "step": 6256
    },
    {
      "epoch": 1.79,
      "grad_norm": 11.333096504211426,
      "learning_rate": 3.153604118993135e-06,
      "loss": 0.664,
      "step": 6257
    },
    {
      "epoch": 1.79,
      "grad_norm": 10.735307693481445,
      "learning_rate": 3.149313501144165e-06,
      "loss": 0.7826,
      "step": 6258
    },
    {
      "epoch": 1.79,
      "grad_norm": 10.006080627441406,
      "learning_rate": 3.145022883295195e-06,
      "loss": 0.5373,
      "step": 6259
    },
    {
      "epoch": 1.79,
      "grad_norm": 10.427276611328125,
      "learning_rate": 3.140732265446224e-06,
      "loss": 0.9635,
      "step": 6260
    },
    {
      "epoch": 1.79,
      "grad_norm": 11.013106346130371,
      "learning_rate": 3.136441647597254e-06,
      "loss": 0.8264,
      "step": 6261
    },
    {
      "epoch": 1.79,
      "grad_norm": 9.57888126373291,
      "learning_rate": 3.1321510297482837e-06,
      "loss": 0.6285,
      "step": 6262
    },
    {
      "epoch": 1.79,
      "grad_norm": 11.295920372009277,
      "learning_rate": 3.1278604118993137e-06,
      "loss": 0.9006,
      "step": 6263
    },
    {
      "epoch": 1.79,
      "grad_norm": 11.340927124023438,
      "learning_rate": 3.1235697940503434e-06,
      "loss": 0.6735,
      "step": 6264
    },
    {
      "epoch": 1.79,
      "grad_norm": 8.415022850036621,
      "learning_rate": 3.1192791762013734e-06,
      "loss": 0.6388,
      "step": 6265
    },
    {
      "epoch": 1.79,
      "grad_norm": 11.195077896118164,
      "learning_rate": 3.1149885583524026e-06,
      "loss": 0.6975,
      "step": 6266
    },
    {
      "epoch": 1.79,
      "grad_norm": 9.73503589630127,
      "learning_rate": 3.1106979405034326e-06,
      "loss": 0.7999,
      "step": 6267
    },
    {
      "epoch": 1.79,
      "grad_norm": 13.010418891906738,
      "learning_rate": 3.1064073226544622e-06,
      "loss": 0.7029,
      "step": 6268
    },
    {
      "epoch": 1.79,
      "grad_norm": 10.655797958374023,
      "learning_rate": 3.102116704805492e-06,
      "loss": 0.6374,
      "step": 6269
    },
    {
      "epoch": 1.79,
      "grad_norm": 6.9743852615356445,
      "learning_rate": 3.097826086956522e-06,
      "loss": 0.5332,
      "step": 6270
    },
    {
      "epoch": 1.79,
      "grad_norm": 8.947936058044434,
      "learning_rate": 3.0935354691075515e-06,
      "loss": 0.5842,
      "step": 6271
    },
    {
      "epoch": 1.79,
      "grad_norm": 10.933826446533203,
      "learning_rate": 3.0892448512585815e-06,
      "loss": 0.9045,
      "step": 6272
    },
    {
      "epoch": 1.79,
      "grad_norm": 9.997200965881348,
      "learning_rate": 3.0849542334096107e-06,
      "loss": 0.7061,
      "step": 6273
    },
    {
      "epoch": 1.79,
      "grad_norm": 8.821471214294434,
      "learning_rate": 3.0806636155606408e-06,
      "loss": 0.6473,
      "step": 6274
    },
    {
      "epoch": 1.79,
      "grad_norm": 9.868239402770996,
      "learning_rate": 3.0763729977116704e-06,
      "loss": 0.6916,
      "step": 6275
    },
    {
      "epoch": 1.8,
      "grad_norm": 13.37822151184082,
      "learning_rate": 3.0720823798627004e-06,
      "loss": 0.7249,
      "step": 6276
    },
    {
      "epoch": 1.8,
      "grad_norm": 11.691786766052246,
      "learning_rate": 3.06779176201373e-06,
      "loss": 0.6795,
      "step": 6277
    },
    {
      "epoch": 1.8,
      "grad_norm": 12.730964660644531,
      "learning_rate": 3.06350114416476e-06,
      "loss": 1.0718,
      "step": 6278
    },
    {
      "epoch": 1.8,
      "grad_norm": 13.505888938903809,
      "learning_rate": 3.0592105263157897e-06,
      "loss": 0.8676,
      "step": 6279
    },
    {
      "epoch": 1.8,
      "grad_norm": 11.337920188903809,
      "learning_rate": 3.0549199084668193e-06,
      "loss": 0.7076,
      "step": 6280
    },
    {
      "epoch": 1.8,
      "grad_norm": 12.364757537841797,
      "learning_rate": 3.050629290617849e-06,
      "loss": 0.9709,
      "step": 6281
    },
    {
      "epoch": 1.8,
      "grad_norm": 10.987433433532715,
      "learning_rate": 3.046338672768879e-06,
      "loss": 0.3746,
      "step": 6282
    },
    {
      "epoch": 1.8,
      "grad_norm": 10.072982788085938,
      "learning_rate": 3.0420480549199086e-06,
      "loss": 0.6641,
      "step": 6283
    },
    {
      "epoch": 1.8,
      "grad_norm": 12.776458740234375,
      "learning_rate": 3.037757437070938e-06,
      "loss": 0.6595,
      "step": 6284
    },
    {
      "epoch": 1.8,
      "grad_norm": 11.782681465148926,
      "learning_rate": 3.0334668192219682e-06,
      "loss": 0.7771,
      "step": 6285
    },
    {
      "epoch": 1.8,
      "grad_norm": 10.204319953918457,
      "learning_rate": 3.0291762013729974e-06,
      "loss": 0.6043,
      "step": 6286
    },
    {
      "epoch": 1.8,
      "grad_norm": 10.723920822143555,
      "learning_rate": 3.0248855835240275e-06,
      "loss": 0.8629,
      "step": 6287
    },
    {
      "epoch": 1.8,
      "grad_norm": 10.14091682434082,
      "learning_rate": 3.020594965675057e-06,
      "loss": 0.5764,
      "step": 6288
    },
    {
      "epoch": 1.8,
      "grad_norm": 9.678997039794922,
      "learning_rate": 3.016304347826087e-06,
      "loss": 0.5826,
      "step": 6289
    },
    {
      "epoch": 1.8,
      "grad_norm": 8.583478927612305,
      "learning_rate": 3.0120137299771167e-06,
      "loss": 0.5579,
      "step": 6290
    },
    {
      "epoch": 1.8,
      "grad_norm": 9.08668041229248,
      "learning_rate": 3.0077231121281468e-06,
      "loss": 0.5925,
      "step": 6291
    },
    {
      "epoch": 1.8,
      "grad_norm": 9.817574501037598,
      "learning_rate": 3.0034324942791764e-06,
      "loss": 0.6855,
      "step": 6292
    },
    {
      "epoch": 1.8,
      "grad_norm": 13.133573532104492,
      "learning_rate": 2.999141876430206e-06,
      "loss": 0.8879,
      "step": 6293
    },
    {
      "epoch": 1.8,
      "grad_norm": 10.072456359863281,
      "learning_rate": 2.9948512585812356e-06,
      "loss": 0.8635,
      "step": 6294
    },
    {
      "epoch": 1.8,
      "grad_norm": 10.36079216003418,
      "learning_rate": 2.9905606407322657e-06,
      "loss": 0.7124,
      "step": 6295
    },
    {
      "epoch": 1.8,
      "grad_norm": 9.779273986816406,
      "learning_rate": 2.9862700228832953e-06,
      "loss": 0.5912,
      "step": 6296
    },
    {
      "epoch": 1.8,
      "grad_norm": 11.917943954467773,
      "learning_rate": 2.9819794050343253e-06,
      "loss": 0.9621,
      "step": 6297
    },
    {
      "epoch": 1.8,
      "grad_norm": 10.256078720092773,
      "learning_rate": 2.977688787185355e-06,
      "loss": 0.6369,
      "step": 6298
    },
    {
      "epoch": 1.8,
      "grad_norm": 8.891809463500977,
      "learning_rate": 2.973398169336384e-06,
      "loss": 0.6131,
      "step": 6299
    },
    {
      "epoch": 1.8,
      "grad_norm": 10.462176322937012,
      "learning_rate": 2.969107551487414e-06,
      "loss": 0.9024,
      "step": 6300
    },
    {
      "epoch": 1.8,
      "grad_norm": 11.854138374328613,
      "learning_rate": 2.9648169336384438e-06,
      "loss": 0.8871,
      "step": 6301
    },
    {
      "epoch": 1.8,
      "grad_norm": 11.050225257873535,
      "learning_rate": 2.960526315789474e-06,
      "loss": 0.6207,
      "step": 6302
    },
    {
      "epoch": 1.8,
      "grad_norm": 13.017751693725586,
      "learning_rate": 2.9562356979405034e-06,
      "loss": 0.5799,
      "step": 6303
    },
    {
      "epoch": 1.8,
      "grad_norm": 10.159361839294434,
      "learning_rate": 2.9519450800915335e-06,
      "loss": 0.6183,
      "step": 6304
    },
    {
      "epoch": 1.8,
      "grad_norm": 9.185395240783691,
      "learning_rate": 2.947654462242563e-06,
      "loss": 0.5525,
      "step": 6305
    },
    {
      "epoch": 1.8,
      "grad_norm": 10.041814804077148,
      "learning_rate": 2.9433638443935927e-06,
      "loss": 0.6716,
      "step": 6306
    },
    {
      "epoch": 1.8,
      "grad_norm": 11.238035202026367,
      "learning_rate": 2.9390732265446223e-06,
      "loss": 0.7286,
      "step": 6307
    },
    {
      "epoch": 1.8,
      "grad_norm": 7.606348991394043,
      "learning_rate": 2.9347826086956523e-06,
      "loss": 0.5237,
      "step": 6308
    },
    {
      "epoch": 1.8,
      "grad_norm": 11.537421226501465,
      "learning_rate": 2.930491990846682e-06,
      "loss": 0.7842,
      "step": 6309
    },
    {
      "epoch": 1.8,
      "grad_norm": 9.478804588317871,
      "learning_rate": 2.926201372997712e-06,
      "loss": 0.6767,
      "step": 6310
    },
    {
      "epoch": 1.81,
      "grad_norm": 9.909363746643066,
      "learning_rate": 2.9219107551487416e-06,
      "loss": 0.7462,
      "step": 6311
    },
    {
      "epoch": 1.81,
      "grad_norm": 11.550567626953125,
      "learning_rate": 2.9176201372997712e-06,
      "loss": 0.6006,
      "step": 6312
    },
    {
      "epoch": 1.81,
      "grad_norm": 9.454132080078125,
      "learning_rate": 2.913329519450801e-06,
      "loss": 0.684,
      "step": 6313
    },
    {
      "epoch": 1.81,
      "grad_norm": 8.273576736450195,
      "learning_rate": 2.9090389016018305e-06,
      "loss": 0.6919,
      "step": 6314
    },
    {
      "epoch": 1.81,
      "grad_norm": 9.448694229125977,
      "learning_rate": 2.9047482837528605e-06,
      "loss": 0.5262,
      "step": 6315
    },
    {
      "epoch": 1.81,
      "grad_norm": 10.833744049072266,
      "learning_rate": 2.90045766590389e-06,
      "loss": 0.7616,
      "step": 6316
    },
    {
      "epoch": 1.81,
      "grad_norm": 11.065820693969727,
      "learning_rate": 2.89616704805492e-06,
      "loss": 0.4999,
      "step": 6317
    },
    {
      "epoch": 1.81,
      "grad_norm": 9.91887092590332,
      "learning_rate": 2.8918764302059498e-06,
      "loss": 0.4809,
      "step": 6318
    },
    {
      "epoch": 1.81,
      "grad_norm": 11.010311126708984,
      "learning_rate": 2.8875858123569794e-06,
      "loss": 0.8754,
      "step": 6319
    },
    {
      "epoch": 1.81,
      "grad_norm": 12.345414161682129,
      "learning_rate": 2.883295194508009e-06,
      "loss": 0.7151,
      "step": 6320
    },
    {
      "epoch": 1.81,
      "grad_norm": 12.316413879394531,
      "learning_rate": 2.879004576659039e-06,
      "loss": 0.6672,
      "step": 6321
    },
    {
      "epoch": 1.81,
      "grad_norm": 8.642687797546387,
      "learning_rate": 2.8747139588100686e-06,
      "loss": 0.4366,
      "step": 6322
    },
    {
      "epoch": 1.81,
      "grad_norm": 10.080451011657715,
      "learning_rate": 2.8704233409610987e-06,
      "loss": 0.6632,
      "step": 6323
    },
    {
      "epoch": 1.81,
      "grad_norm": 11.197017669677734,
      "learning_rate": 2.8661327231121283e-06,
      "loss": 0.9502,
      "step": 6324
    },
    {
      "epoch": 1.81,
      "grad_norm": 11.442328453063965,
      "learning_rate": 2.861842105263158e-06,
      "loss": 0.6707,
      "step": 6325
    },
    {
      "epoch": 1.81,
      "grad_norm": 11.40806770324707,
      "learning_rate": 2.8575514874141875e-06,
      "loss": 0.6401,
      "step": 6326
    },
    {
      "epoch": 1.81,
      "grad_norm": 10.922065734863281,
      "learning_rate": 2.8532608695652176e-06,
      "loss": 0.8657,
      "step": 6327
    },
    {
      "epoch": 1.81,
      "grad_norm": 9.807641983032227,
      "learning_rate": 2.848970251716247e-06,
      "loss": 0.7005,
      "step": 6328
    },
    {
      "epoch": 1.81,
      "grad_norm": 8.661300659179688,
      "learning_rate": 2.844679633867277e-06,
      "loss": 0.5264,
      "step": 6329
    },
    {
      "epoch": 1.81,
      "grad_norm": 8.730284690856934,
      "learning_rate": 2.840389016018307e-06,
      "loss": 0.4594,
      "step": 6330
    },
    {
      "epoch": 1.81,
      "grad_norm": 9.94897747039795,
      "learning_rate": 2.8360983981693364e-06,
      "loss": 0.6994,
      "step": 6331
    },
    {
      "epoch": 1.81,
      "grad_norm": 10.989338874816895,
      "learning_rate": 2.831807780320366e-06,
      "loss": 0.6297,
      "step": 6332
    },
    {
      "epoch": 1.81,
      "grad_norm": 12.295429229736328,
      "learning_rate": 2.8275171624713957e-06,
      "loss": 0.8819,
      "step": 6333
    },
    {
      "epoch": 1.81,
      "grad_norm": 9.570334434509277,
      "learning_rate": 2.8232265446224257e-06,
      "loss": 0.4966,
      "step": 6334
    },
    {
      "epoch": 1.81,
      "grad_norm": 11.286737442016602,
      "learning_rate": 2.8189359267734553e-06,
      "loss": 0.9316,
      "step": 6335
    },
    {
      "epoch": 1.81,
      "grad_norm": 8.949596405029297,
      "learning_rate": 2.8146453089244854e-06,
      "loss": 0.4447,
      "step": 6336
    },
    {
      "epoch": 1.81,
      "grad_norm": 10.223958015441895,
      "learning_rate": 2.810354691075515e-06,
      "loss": 0.6091,
      "step": 6337
    },
    {
      "epoch": 1.81,
      "grad_norm": 10.857589721679688,
      "learning_rate": 2.806064073226545e-06,
      "loss": 0.6198,
      "step": 6338
    },
    {
      "epoch": 1.81,
      "grad_norm": 11.978499412536621,
      "learning_rate": 2.8017734553775742e-06,
      "loss": 0.6739,
      "step": 6339
    },
    {
      "epoch": 1.81,
      "grad_norm": 8.985401153564453,
      "learning_rate": 2.7974828375286043e-06,
      "loss": 0.6556,
      "step": 6340
    },
    {
      "epoch": 1.81,
      "grad_norm": 9.769513130187988,
      "learning_rate": 2.793192219679634e-06,
      "loss": 0.595,
      "step": 6341
    },
    {
      "epoch": 1.81,
      "grad_norm": 10.520943641662598,
      "learning_rate": 2.788901601830664e-06,
      "loss": 0.6485,
      "step": 6342
    },
    {
      "epoch": 1.81,
      "grad_norm": 13.118606567382812,
      "learning_rate": 2.7846109839816935e-06,
      "loss": 0.8612,
      "step": 6343
    },
    {
      "epoch": 1.81,
      "grad_norm": 10.005794525146484,
      "learning_rate": 2.7803203661327236e-06,
      "loss": 0.7682,
      "step": 6344
    },
    {
      "epoch": 1.81,
      "grad_norm": 8.515554428100586,
      "learning_rate": 2.7760297482837528e-06,
      "loss": 0.5026,
      "step": 6345
    },
    {
      "epoch": 1.82,
      "grad_norm": 14.833863258361816,
      "learning_rate": 2.7717391304347824e-06,
      "loss": 0.9351,
      "step": 6346
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.056303024291992,
      "learning_rate": 2.7674485125858124e-06,
      "loss": 0.5222,
      "step": 6347
    },
    {
      "epoch": 1.82,
      "grad_norm": 9.482291221618652,
      "learning_rate": 2.763157894736842e-06,
      "loss": 0.6263,
      "step": 6348
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.735074043273926,
      "learning_rate": 2.758867276887872e-06,
      "loss": 0.7135,
      "step": 6349
    },
    {
      "epoch": 1.82,
      "grad_norm": 9.922060012817383,
      "learning_rate": 2.7545766590389017e-06,
      "loss": 0.6701,
      "step": 6350
    },
    {
      "epoch": 1.82,
      "grad_norm": 7.189872741699219,
      "learning_rate": 2.7502860411899317e-06,
      "loss": 0.3701,
      "step": 6351
    },
    {
      "epoch": 1.82,
      "grad_norm": 10.476938247680664,
      "learning_rate": 2.745995423340961e-06,
      "loss": 0.7437,
      "step": 6352
    },
    {
      "epoch": 1.82,
      "grad_norm": 10.8806791305542,
      "learning_rate": 2.741704805491991e-06,
      "loss": 0.5855,
      "step": 6353
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.708989143371582,
      "learning_rate": 2.7374141876430206e-06,
      "loss": 0.737,
      "step": 6354
    },
    {
      "epoch": 1.82,
      "grad_norm": 8.859798431396484,
      "learning_rate": 2.7331235697940506e-06,
      "loss": 0.467,
      "step": 6355
    },
    {
      "epoch": 1.82,
      "grad_norm": 9.817736625671387,
      "learning_rate": 2.72883295194508e-06,
      "loss": 0.7523,
      "step": 6356
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.412315368652344,
      "learning_rate": 2.7245423340961102e-06,
      "loss": 0.7811,
      "step": 6357
    },
    {
      "epoch": 1.82,
      "grad_norm": 10.915645599365234,
      "learning_rate": 2.7202517162471394e-06,
      "loss": 0.8479,
      "step": 6358
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.466460227966309,
      "learning_rate": 2.7159610983981695e-06,
      "loss": 0.5644,
      "step": 6359
    },
    {
      "epoch": 1.82,
      "grad_norm": 8.003796577453613,
      "learning_rate": 2.711670480549199e-06,
      "loss": 0.5993,
      "step": 6360
    },
    {
      "epoch": 1.82,
      "grad_norm": 12.29226016998291,
      "learning_rate": 2.7073798627002287e-06,
      "loss": 0.6913,
      "step": 6361
    },
    {
      "epoch": 1.82,
      "grad_norm": 9.387224197387695,
      "learning_rate": 2.7030892448512587e-06,
      "loss": 0.5956,
      "step": 6362
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.623571395874023,
      "learning_rate": 2.6987986270022884e-06,
      "loss": 0.7545,
      "step": 6363
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.96871280670166,
      "learning_rate": 2.6945080091533184e-06,
      "loss": 0.553,
      "step": 6364
    },
    {
      "epoch": 1.82,
      "grad_norm": 12.344478607177734,
      "learning_rate": 2.6902173913043476e-06,
      "loss": 0.7015,
      "step": 6365
    },
    {
      "epoch": 1.82,
      "grad_norm": 10.739533424377441,
      "learning_rate": 2.6859267734553776e-06,
      "loss": 0.4474,
      "step": 6366
    },
    {
      "epoch": 1.82,
      "grad_norm": 8.55423641204834,
      "learning_rate": 2.6816361556064072e-06,
      "loss": 0.6801,
      "step": 6367
    },
    {
      "epoch": 1.82,
      "grad_norm": 8.852259635925293,
      "learning_rate": 2.6773455377574373e-06,
      "loss": 0.5669,
      "step": 6368
    },
    {
      "epoch": 1.82,
      "grad_norm": 9.556403160095215,
      "learning_rate": 2.673054919908467e-06,
      "loss": 0.678,
      "step": 6369
    },
    {
      "epoch": 1.82,
      "grad_norm": 12.659199714660645,
      "learning_rate": 2.668764302059497e-06,
      "loss": 0.7636,
      "step": 6370
    },
    {
      "epoch": 1.82,
      "grad_norm": 10.23427677154541,
      "learning_rate": 2.6644736842105266e-06,
      "loss": 0.684,
      "step": 6371
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.79377269744873,
      "learning_rate": 2.660183066361556e-06,
      "loss": 0.6206,
      "step": 6372
    },
    {
      "epoch": 1.82,
      "grad_norm": 10.17258358001709,
      "learning_rate": 2.6558924485125858e-06,
      "loss": 0.712,
      "step": 6373
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.21706771850586,
      "learning_rate": 2.651601830663616e-06,
      "loss": 0.8673,
      "step": 6374
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.416669845581055,
      "learning_rate": 2.6473112128146454e-06,
      "loss": 0.8331,
      "step": 6375
    },
    {
      "epoch": 1.82,
      "grad_norm": 9.43038558959961,
      "learning_rate": 2.643020594965675e-06,
      "loss": 0.5575,
      "step": 6376
    },
    {
      "epoch": 1.82,
      "grad_norm": 7.555305004119873,
      "learning_rate": 2.638729977116705e-06,
      "loss": 0.5514,
      "step": 6377
    },
    {
      "epoch": 1.82,
      "grad_norm": 8.861245155334473,
      "learning_rate": 2.6344393592677343e-06,
      "loss": 0.7121,
      "step": 6378
    },
    {
      "epoch": 1.82,
      "grad_norm": 14.446701049804688,
      "learning_rate": 2.6301487414187643e-06,
      "loss": 0.662,
      "step": 6379
    },
    {
      "epoch": 1.82,
      "grad_norm": 10.05135726928711,
      "learning_rate": 2.625858123569794e-06,
      "loss": 0.7382,
      "step": 6380
    },
    {
      "epoch": 1.83,
      "grad_norm": 11.255631446838379,
      "learning_rate": 2.621567505720824e-06,
      "loss": 0.655,
      "step": 6381
    },
    {
      "epoch": 1.83,
      "grad_norm": 11.895508766174316,
      "learning_rate": 2.6172768878718536e-06,
      "loss": 0.7342,
      "step": 6382
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.853370666503906,
      "learning_rate": 2.6129862700228836e-06,
      "loss": 0.5712,
      "step": 6383
    },
    {
      "epoch": 1.83,
      "grad_norm": 12.39453411102295,
      "learning_rate": 2.6086956521739132e-06,
      "loss": 0.9292,
      "step": 6384
    },
    {
      "epoch": 1.83,
      "grad_norm": 8.208925247192383,
      "learning_rate": 2.604405034324943e-06,
      "loss": 0.578,
      "step": 6385
    },
    {
      "epoch": 1.83,
      "grad_norm": 11.307624816894531,
      "learning_rate": 2.6001144164759725e-06,
      "loss": 0.7251,
      "step": 6386
    },
    {
      "epoch": 1.83,
      "grad_norm": 8.692691802978516,
      "learning_rate": 2.5958237986270025e-06,
      "loss": 0.5645,
      "step": 6387
    },
    {
      "epoch": 1.83,
      "grad_norm": 11.707426071166992,
      "learning_rate": 2.591533180778032e-06,
      "loss": 0.5614,
      "step": 6388
    },
    {
      "epoch": 1.83,
      "grad_norm": 11.854446411132812,
      "learning_rate": 2.587242562929062e-06,
      "loss": 0.801,
      "step": 6389
    },
    {
      "epoch": 1.83,
      "grad_norm": 7.33786678314209,
      "learning_rate": 2.5829519450800918e-06,
      "loss": 0.4136,
      "step": 6390
    },
    {
      "epoch": 1.83,
      "grad_norm": 11.048632621765137,
      "learning_rate": 2.578661327231121e-06,
      "loss": 0.7862,
      "step": 6391
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.102787017822266,
      "learning_rate": 2.574370709382151e-06,
      "loss": 0.6187,
      "step": 6392
    },
    {
      "epoch": 1.83,
      "grad_norm": 18.396352767944336,
      "learning_rate": 2.5700800915331806e-06,
      "loss": 0.682,
      "step": 6393
    },
    {
      "epoch": 1.83,
      "grad_norm": 13.416193008422852,
      "learning_rate": 2.5657894736842107e-06,
      "loss": 0.7955,
      "step": 6394
    },
    {
      "epoch": 1.83,
      "grad_norm": 7.945309162139893,
      "learning_rate": 2.5614988558352403e-06,
      "loss": 0.7932,
      "step": 6395
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.73609733581543,
      "learning_rate": 2.5572082379862703e-06,
      "loss": 0.6372,
      "step": 6396
    },
    {
      "epoch": 1.83,
      "grad_norm": 9.39498519897461,
      "learning_rate": 2.5529176201373e-06,
      "loss": 0.5106,
      "step": 6397
    },
    {
      "epoch": 1.83,
      "grad_norm": 9.900101661682129,
      "learning_rate": 2.5486270022883295e-06,
      "loss": 0.6601,
      "step": 6398
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.181139945983887,
      "learning_rate": 2.544336384439359e-06,
      "loss": 0.7377,
      "step": 6399
    },
    {
      "epoch": 1.83,
      "grad_norm": 8.726524353027344,
      "learning_rate": 2.540045766590389e-06,
      "loss": 0.4742,
      "step": 6400
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.729772567749023,
      "learning_rate": 2.535755148741419e-06,
      "loss": 0.4859,
      "step": 6401
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.562342643737793,
      "learning_rate": 2.531464530892449e-06,
      "loss": 0.7937,
      "step": 6402
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.791984558105469,
      "learning_rate": 2.5271739130434785e-06,
      "loss": 0.6454,
      "step": 6403
    },
    {
      "epoch": 1.83,
      "grad_norm": 13.546317100524902,
      "learning_rate": 2.522883295194508e-06,
      "loss": 0.8172,
      "step": 6404
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.978838920593262,
      "learning_rate": 2.5185926773455377e-06,
      "loss": 0.5623,
      "step": 6405
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.70091724395752,
      "learning_rate": 2.5143020594965673e-06,
      "loss": 0.7327,
      "step": 6406
    },
    {
      "epoch": 1.83,
      "grad_norm": 9.919500350952148,
      "learning_rate": 2.5100114416475973e-06,
      "loss": 0.6551,
      "step": 6407
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.135588645935059,
      "learning_rate": 2.505720823798627e-06,
      "loss": 0.6837,
      "step": 6408
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.382122039794922,
      "learning_rate": 2.501430205949657e-06,
      "loss": 0.6848,
      "step": 6409
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.038745880126953,
      "learning_rate": 2.4971395881006866e-06,
      "loss": 0.5203,
      "step": 6410
    },
    {
      "epoch": 1.83,
      "grad_norm": 11.14279842376709,
      "learning_rate": 2.4928489702517162e-06,
      "loss": 0.5948,
      "step": 6411
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.874357223510742,
      "learning_rate": 2.488558352402746e-06,
      "loss": 0.7269,
      "step": 6412
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.609356880187988,
      "learning_rate": 2.484267734553776e-06,
      "loss": 0.5865,
      "step": 6413
    },
    {
      "epoch": 1.83,
      "grad_norm": 9.98708724975586,
      "learning_rate": 2.4799771167048055e-06,
      "loss": 0.7918,
      "step": 6414
    },
    {
      "epoch": 1.83,
      "grad_norm": 11.888640403747559,
      "learning_rate": 2.4756864988558355e-06,
      "loss": 0.6676,
      "step": 6415
    },
    {
      "epoch": 1.84,
      "grad_norm": 10.499030113220215,
      "learning_rate": 2.471395881006865e-06,
      "loss": 0.5232,
      "step": 6416
    },
    {
      "epoch": 1.84,
      "grad_norm": 9.32913589477539,
      "learning_rate": 2.4671052631578948e-06,
      "loss": 0.5939,
      "step": 6417
    },
    {
      "epoch": 1.84,
      "grad_norm": 10.895303726196289,
      "learning_rate": 2.4628146453089244e-06,
      "loss": 0.7183,
      "step": 6418
    },
    {
      "epoch": 1.84,
      "grad_norm": 9.079419136047363,
      "learning_rate": 2.4585240274599544e-06,
      "loss": 0.7644,
      "step": 6419
    },
    {
      "epoch": 1.84,
      "grad_norm": 10.472522735595703,
      "learning_rate": 2.454233409610984e-06,
      "loss": 0.6486,
      "step": 6420
    },
    {
      "epoch": 1.84,
      "grad_norm": 8.136340141296387,
      "learning_rate": 2.4499427917620136e-06,
      "loss": 0.5381,
      "step": 6421
    },
    {
      "epoch": 1.84,
      "grad_norm": 10.92373275756836,
      "learning_rate": 2.4456521739130437e-06,
      "loss": 0.6147,
      "step": 6422
    },
    {
      "epoch": 1.84,
      "grad_norm": 12.690459251403809,
      "learning_rate": 2.4413615560640733e-06,
      "loss": 0.569,
      "step": 6423
    },
    {
      "epoch": 1.84,
      "grad_norm": 11.713183403015137,
      "learning_rate": 2.437070938215103e-06,
      "loss": 0.7141,
      "step": 6424
    },
    {
      "epoch": 1.84,
      "grad_norm": 10.668099403381348,
      "learning_rate": 2.4327803203661325e-06,
      "loss": 0.7687,
      "step": 6425
    },
    {
      "epoch": 1.84,
      "grad_norm": 14.403892517089844,
      "learning_rate": 2.4284897025171626e-06,
      "loss": 0.7468,
      "step": 6426
    },
    {
      "epoch": 1.84,
      "grad_norm": 7.870250225067139,
      "learning_rate": 2.424199084668192e-06,
      "loss": 0.6532,
      "step": 6427
    },
    {
      "epoch": 1.84,
      "grad_norm": 12.25476360321045,
      "learning_rate": 2.4199084668192222e-06,
      "loss": 0.6964,
      "step": 6428
    },
    {
      "epoch": 1.84,
      "grad_norm": 12.443902969360352,
      "learning_rate": 2.415617848970252e-06,
      "loss": 0.9736,
      "step": 6429
    },
    {
      "epoch": 1.84,
      "grad_norm": 13.772923469543457,
      "learning_rate": 2.411327231121282e-06,
      "loss": 0.7917,
      "step": 6430
    },
    {
      "epoch": 1.84,
      "grad_norm": 12.89171028137207,
      "learning_rate": 2.407036613272311e-06,
      "loss": 0.7432,
      "step": 6431
    },
    {
      "epoch": 1.84,
      "grad_norm": 10.942740440368652,
      "learning_rate": 2.402745995423341e-06,
      "loss": 0.6137,
      "step": 6432
    },
    {
      "epoch": 1.84,
      "grad_norm": 9.382735252380371,
      "learning_rate": 2.3984553775743707e-06,
      "loss": 0.6103,
      "step": 6433
    },
    {
      "epoch": 1.84,
      "grad_norm": 11.818822860717773,
      "learning_rate": 2.3941647597254008e-06,
      "loss": 0.7706,
      "step": 6434
    },
    {
      "epoch": 1.84,
      "grad_norm": 12.318153381347656,
      "learning_rate": 2.3898741418764304e-06,
      "loss": 0.5996,
      "step": 6435
    },
    {
      "epoch": 1.84,
      "grad_norm": 11.190994262695312,
      "learning_rate": 2.38558352402746e-06,
      "loss": 0.7502,
      "step": 6436
    },
    {
      "epoch": 1.84,
      "grad_norm": 10.103891372680664,
      "learning_rate": 2.3812929061784896e-06,
      "loss": 0.6892,
      "step": 6437
    },
    {
      "epoch": 1.84,
      "grad_norm": 13.064241409301758,
      "learning_rate": 2.3770022883295192e-06,
      "loss": 0.7656,
      "step": 6438
    },
    {
      "epoch": 1.84,
      "grad_norm": 15.39616870880127,
      "learning_rate": 2.3727116704805493e-06,
      "loss": 0.8093,
      "step": 6439
    },
    {
      "epoch": 1.84,
      "grad_norm": 14.783260345458984,
      "learning_rate": 2.368421052631579e-06,
      "loss": 0.7323,
      "step": 6440
    },
    {
      "epoch": 1.84,
      "grad_norm": 9.266852378845215,
      "learning_rate": 2.364130434782609e-06,
      "loss": 0.658,
      "step": 6441
    },
    {
      "epoch": 1.84,
      "grad_norm": 10.6998929977417,
      "learning_rate": 2.3598398169336385e-06,
      "loss": 0.7976,
      "step": 6442
    },
    {
      "epoch": 1.84,
      "grad_norm": 13.172496795654297,
      "learning_rate": 2.3555491990846686e-06,
      "loss": 0.6606,
      "step": 6443
    },
    {
      "epoch": 1.84,
      "grad_norm": 12.803412437438965,
      "learning_rate": 2.3512585812356978e-06,
      "loss": 0.45,
      "step": 6444
    },
    {
      "epoch": 1.84,
      "grad_norm": 7.840516567230225,
      "learning_rate": 2.346967963386728e-06,
      "loss": 0.581,
      "step": 6445
    },
    {
      "epoch": 1.84,
      "grad_norm": 10.895773887634277,
      "learning_rate": 2.3426773455377574e-06,
      "loss": 0.7987,
      "step": 6446
    },
    {
      "epoch": 1.84,
      "grad_norm": 9.917009353637695,
      "learning_rate": 2.3383867276887874e-06,
      "loss": 0.583,
      "step": 6447
    },
    {
      "epoch": 1.84,
      "grad_norm": 9.434247970581055,
      "learning_rate": 2.334096109839817e-06,
      "loss": 0.5384,
      "step": 6448
    },
    {
      "epoch": 1.84,
      "grad_norm": 12.754484176635742,
      "learning_rate": 2.329805491990847e-06,
      "loss": 0.715,
      "step": 6449
    },
    {
      "epoch": 1.84,
      "grad_norm": 10.924982070922852,
      "learning_rate": 2.3255148741418763e-06,
      "loss": 0.7895,
      "step": 6450
    },
    {
      "epoch": 1.85,
      "grad_norm": 8.170269012451172,
      "learning_rate": 2.321224256292906e-06,
      "loss": 0.5094,
      "step": 6451
    },
    {
      "epoch": 1.85,
      "grad_norm": 7.4521331787109375,
      "learning_rate": 2.316933638443936e-06,
      "loss": 0.4345,
      "step": 6452
    },
    {
      "epoch": 1.85,
      "grad_norm": 10.049761772155762,
      "learning_rate": 2.3126430205949656e-06,
      "loss": 0.6747,
      "step": 6453
    },
    {
      "epoch": 1.85,
      "grad_norm": 10.886943817138672,
      "learning_rate": 2.3083524027459956e-06,
      "loss": 0.5234,
      "step": 6454
    },
    {
      "epoch": 1.85,
      "grad_norm": 11.293790817260742,
      "learning_rate": 2.3040617848970252e-06,
      "loss": 0.6757,
      "step": 6455
    },
    {
      "epoch": 1.85,
      "grad_norm": 9.645736694335938,
      "learning_rate": 2.2997711670480553e-06,
      "loss": 0.8438,
      "step": 6456
    },
    {
      "epoch": 1.85,
      "grad_norm": 14.119002342224121,
      "learning_rate": 2.2954805491990844e-06,
      "loss": 0.7814,
      "step": 6457
    },
    {
      "epoch": 1.85,
      "grad_norm": 9.948843955993652,
      "learning_rate": 2.2911899313501145e-06,
      "loss": 0.5564,
      "step": 6458
    },
    {
      "epoch": 1.85,
      "grad_norm": 9.257054328918457,
      "learning_rate": 2.286899313501144e-06,
      "loss": 0.7186,
      "step": 6459
    },
    {
      "epoch": 1.85,
      "grad_norm": 10.969335556030273,
      "learning_rate": 2.282608695652174e-06,
      "loss": 0.8478,
      "step": 6460
    },
    {
      "epoch": 1.85,
      "grad_norm": 9.461034774780273,
      "learning_rate": 2.2783180778032038e-06,
      "loss": 0.7713,
      "step": 6461
    },
    {
      "epoch": 1.85,
      "grad_norm": 13.996525764465332,
      "learning_rate": 2.274027459954234e-06,
      "loss": 0.9145,
      "step": 6462
    },
    {
      "epoch": 1.85,
      "grad_norm": 12.406022071838379,
      "learning_rate": 2.2697368421052634e-06,
      "loss": 0.7289,
      "step": 6463
    },
    {
      "epoch": 1.85,
      "grad_norm": 7.784307956695557,
      "learning_rate": 2.265446224256293e-06,
      "loss": 0.605,
      "step": 6464
    },
    {
      "epoch": 1.85,
      "grad_norm": 10.131551742553711,
      "learning_rate": 2.2611556064073226e-06,
      "loss": 0.621,
      "step": 6465
    },
    {
      "epoch": 1.85,
      "grad_norm": 10.075953483581543,
      "learning_rate": 2.2568649885583522e-06,
      "loss": 0.6698,
      "step": 6466
    },
    {
      "epoch": 1.85,
      "grad_norm": 11.9581880569458,
      "learning_rate": 2.2525743707093823e-06,
      "loss": 0.8044,
      "step": 6467
    },
    {
      "epoch": 1.85,
      "grad_norm": 10.870621681213379,
      "learning_rate": 2.248283752860412e-06,
      "loss": 0.581,
      "step": 6468
    },
    {
      "epoch": 1.85,
      "grad_norm": 9.622464179992676,
      "learning_rate": 2.243993135011442e-06,
      "loss": 0.656,
      "step": 6469
    },
    {
      "epoch": 1.85,
      "grad_norm": 9.188612937927246,
      "learning_rate": 2.239702517162471e-06,
      "loss": 0.7667,
      "step": 6470
    },
    {
      "epoch": 1.85,
      "grad_norm": 11.247570037841797,
      "learning_rate": 2.235411899313501e-06,
      "loss": 0.7746,
      "step": 6471
    },
    {
      "epoch": 1.85,
      "grad_norm": 13.013686180114746,
      "learning_rate": 2.2311212814645308e-06,
      "loss": 0.8872,
      "step": 6472
    },
    {
      "epoch": 1.85,
      "grad_norm": 9.735376358032227,
      "learning_rate": 2.226830663615561e-06,
      "loss": 0.754,
      "step": 6473
    },
    {
      "epoch": 1.85,
      "grad_norm": 10.814042091369629,
      "learning_rate": 2.2225400457665904e-06,
      "loss": 0.8551,
      "step": 6474
    },
    {
      "epoch": 1.85,
      "grad_norm": 9.536465644836426,
      "learning_rate": 2.2182494279176205e-06,
      "loss": 0.7245,
      "step": 6475
    },
    {
      "epoch": 1.85,
      "grad_norm": 9.123568534851074,
      "learning_rate": 2.21395881006865e-06,
      "loss": 0.6765,
      "step": 6476
    },
    {
      "epoch": 1.85,
      "grad_norm": 10.892561912536621,
      "learning_rate": 2.2096681922196797e-06,
      "loss": 0.6025,
      "step": 6477
    },
    {
      "epoch": 1.85,
      "grad_norm": 8.635826110839844,
      "learning_rate": 2.2053775743707093e-06,
      "loss": 0.7246,
      "step": 6478
    },
    {
      "epoch": 1.85,
      "grad_norm": 12.406291961669922,
      "learning_rate": 2.2010869565217394e-06,
      "loss": 0.6637,
      "step": 6479
    },
    {
      "epoch": 1.85,
      "grad_norm": 10.550618171691895,
      "learning_rate": 2.196796338672769e-06,
      "loss": 0.6013,
      "step": 6480
    },
    {
      "epoch": 1.85,
      "grad_norm": 7.916990280151367,
      "learning_rate": 2.1925057208237986e-06,
      "loss": 0.5096,
      "step": 6481
    },
    {
      "epoch": 1.85,
      "grad_norm": 8.452898025512695,
      "learning_rate": 2.1882151029748286e-06,
      "loss": 0.6016,
      "step": 6482
    },
    {
      "epoch": 1.85,
      "grad_norm": 8.159934997558594,
      "learning_rate": 2.183924485125858e-06,
      "loss": 0.7398,
      "step": 6483
    },
    {
      "epoch": 1.85,
      "grad_norm": 8.472947120666504,
      "learning_rate": 2.179633867276888e-06,
      "loss": 0.4693,
      "step": 6484
    },
    {
      "epoch": 1.85,
      "grad_norm": 10.124168395996094,
      "learning_rate": 2.1753432494279175e-06,
      "loss": 0.5472,
      "step": 6485
    },
    {
      "epoch": 1.86,
      "grad_norm": 7.216623783111572,
      "learning_rate": 2.1710526315789475e-06,
      "loss": 0.5199,
      "step": 6486
    },
    {
      "epoch": 1.86,
      "grad_norm": 11.942208290100098,
      "learning_rate": 2.166762013729977e-06,
      "loss": 1.0072,
      "step": 6487
    },
    {
      "epoch": 1.86,
      "grad_norm": 9.318800926208496,
      "learning_rate": 2.162471395881007e-06,
      "loss": 0.6784,
      "step": 6488
    },
    {
      "epoch": 1.86,
      "grad_norm": 10.21501350402832,
      "learning_rate": 2.1581807780320368e-06,
      "loss": 0.5391,
      "step": 6489
    },
    {
      "epoch": 1.86,
      "grad_norm": 12.667245864868164,
      "learning_rate": 2.1538901601830664e-06,
      "loss": 0.6718,
      "step": 6490
    },
    {
      "epoch": 1.86,
      "grad_norm": 9.556787490844727,
      "learning_rate": 2.149599542334096e-06,
      "loss": 0.6574,
      "step": 6491
    },
    {
      "epoch": 1.86,
      "grad_norm": 12.611247062683105,
      "learning_rate": 2.145308924485126e-06,
      "loss": 0.821,
      "step": 6492
    },
    {
      "epoch": 1.86,
      "grad_norm": 8.203670501708984,
      "learning_rate": 2.1410183066361557e-06,
      "loss": 0.5282,
      "step": 6493
    },
    {
      "epoch": 1.86,
      "grad_norm": 11.463093757629395,
      "learning_rate": 2.1367276887871857e-06,
      "loss": 0.5527,
      "step": 6494
    },
    {
      "epoch": 1.86,
      "grad_norm": 8.550508499145508,
      "learning_rate": 2.1324370709382153e-06,
      "loss": 0.5446,
      "step": 6495
    },
    {
      "epoch": 1.86,
      "grad_norm": 9.182392120361328,
      "learning_rate": 2.1281464530892445e-06,
      "loss": 0.676,
      "step": 6496
    },
    {
      "epoch": 1.86,
      "grad_norm": 10.834150314331055,
      "learning_rate": 2.1238558352402745e-06,
      "loss": 0.743,
      "step": 6497
    },
    {
      "epoch": 1.86,
      "grad_norm": 9.196577072143555,
      "learning_rate": 2.119565217391304e-06,
      "loss": 0.7373,
      "step": 6498
    },
    {
      "epoch": 1.86,
      "grad_norm": 7.15338134765625,
      "learning_rate": 2.115274599542334e-06,
      "loss": 0.4727,
      "step": 6499
    },
    {
      "epoch": 1.86,
      "grad_norm": 10.727302551269531,
      "learning_rate": 2.110983981693364e-06,
      "loss": 0.7907,
      "step": 6500
    },
    {
      "epoch": 1.86,
      "grad_norm": 11.069611549377441,
      "learning_rate": 2.106693363844394e-06,
      "loss": 0.5606,
      "step": 6501
    },
    {
      "epoch": 1.86,
      "grad_norm": 12.03882884979248,
      "learning_rate": 2.1024027459954235e-06,
      "loss": 0.8557,
      "step": 6502
    },
    {
      "epoch": 1.86,
      "grad_norm": 9.859286308288574,
      "learning_rate": 2.098112128146453e-06,
      "loss": 0.6137,
      "step": 6503
    },
    {
      "epoch": 1.86,
      "grad_norm": 11.920726776123047,
      "learning_rate": 2.0938215102974827e-06,
      "loss": 0.652,
      "step": 6504
    },
    {
      "epoch": 1.86,
      "grad_norm": 13.011934280395508,
      "learning_rate": 2.0895308924485127e-06,
      "loss": 0.7809,
      "step": 6505
    },
    {
      "epoch": 1.86,
      "grad_norm": 10.034687995910645,
      "learning_rate": 2.0852402745995424e-06,
      "loss": 0.6016,
      "step": 6506
    },
    {
      "epoch": 1.86,
      "grad_norm": 12.651317596435547,
      "learning_rate": 2.0809496567505724e-06,
      "loss": 1.0099,
      "step": 6507
    },
    {
      "epoch": 1.86,
      "grad_norm": 11.004575729370117,
      "learning_rate": 2.076659038901602e-06,
      "loss": 0.621,
      "step": 6508
    },
    {
      "epoch": 1.86,
      "grad_norm": 11.537083625793457,
      "learning_rate": 2.0723684210526316e-06,
      "loss": 0.5622,
      "step": 6509
    },
    {
      "epoch": 1.86,
      "grad_norm": 8.899933815002441,
      "learning_rate": 2.0680778032036612e-06,
      "loss": 0.5059,
      "step": 6510
    },
    {
      "epoch": 1.86,
      "grad_norm": 10.634422302246094,
      "learning_rate": 2.063787185354691e-06,
      "loss": 0.9044,
      "step": 6511
    },
    {
      "epoch": 1.86,
      "grad_norm": 10.67432689666748,
      "learning_rate": 2.059496567505721e-06,
      "loss": 0.5265,
      "step": 6512
    },
    {
      "epoch": 1.86,
      "grad_norm": 9.965004920959473,
      "learning_rate": 2.0552059496567505e-06,
      "loss": 0.7052,
      "step": 6513
    },
    {
      "epoch": 1.86,
      "grad_norm": 8.582775115966797,
      "learning_rate": 2.0509153318077805e-06,
      "loss": 0.6821,
      "step": 6514
    },
    {
      "epoch": 1.86,
      "grad_norm": 10.898089408874512,
      "learning_rate": 2.04662471395881e-06,
      "loss": 0.6636,
      "step": 6515
    },
    {
      "epoch": 1.86,
      "grad_norm": 9.397239685058594,
      "learning_rate": 2.0423340961098398e-06,
      "loss": 0.6129,
      "step": 6516
    },
    {
      "epoch": 1.86,
      "grad_norm": 10.674189567565918,
      "learning_rate": 2.0380434782608694e-06,
      "loss": 0.6215,
      "step": 6517
    },
    {
      "epoch": 1.86,
      "grad_norm": 8.830999374389648,
      "learning_rate": 2.0337528604118994e-06,
      "loss": 0.4642,
      "step": 6518
    },
    {
      "epoch": 1.86,
      "grad_norm": 10.3596773147583,
      "learning_rate": 2.029462242562929e-06,
      "loss": 0.8204,
      "step": 6519
    },
    {
      "epoch": 1.86,
      "grad_norm": 11.077269554138184,
      "learning_rate": 2.025171624713959e-06,
      "loss": 0.6943,
      "step": 6520
    },
    {
      "epoch": 1.87,
      "grad_norm": 8.061532974243164,
      "learning_rate": 2.0208810068649887e-06,
      "loss": 0.5054,
      "step": 6521
    },
    {
      "epoch": 1.87,
      "grad_norm": 10.03957462310791,
      "learning_rate": 2.0165903890160187e-06,
      "loss": 0.7679,
      "step": 6522
    },
    {
      "epoch": 1.87,
      "grad_norm": 7.75835657119751,
      "learning_rate": 2.012299771167048e-06,
      "loss": 0.4455,
      "step": 6523
    },
    {
      "epoch": 1.87,
      "grad_norm": 9.567909240722656,
      "learning_rate": 2.008009153318078e-06,
      "loss": 0.6535,
      "step": 6524
    },
    {
      "epoch": 1.87,
      "grad_norm": 10.833852767944336,
      "learning_rate": 2.0037185354691076e-06,
      "loss": 0.7393,
      "step": 6525
    },
    {
      "epoch": 1.87,
      "grad_norm": 8.219368934631348,
      "learning_rate": 1.999427917620137e-06,
      "loss": 0.5447,
      "step": 6526
    },
    {
      "epoch": 1.87,
      "grad_norm": 11.239940643310547,
      "learning_rate": 1.9951372997711672e-06,
      "loss": 0.8011,
      "step": 6527
    },
    {
      "epoch": 1.87,
      "grad_norm": 8.554726600646973,
      "learning_rate": 1.990846681922197e-06,
      "loss": 0.5626,
      "step": 6528
    },
    {
      "epoch": 1.87,
      "grad_norm": 8.939287185668945,
      "learning_rate": 1.9865560640732265e-06,
      "loss": 0.4421,
      "step": 6529
    },
    {
      "epoch": 1.87,
      "grad_norm": 7.468092441558838,
      "learning_rate": 1.982265446224256e-06,
      "loss": 0.523,
      "step": 6530
    },
    {
      "epoch": 1.87,
      "grad_norm": 12.858736038208008,
      "learning_rate": 1.977974828375286e-06,
      "loss": 0.9184,
      "step": 6531
    },
    {
      "epoch": 1.87,
      "grad_norm": 9.084821701049805,
      "learning_rate": 1.9736842105263157e-06,
      "loss": 0.6705,
      "step": 6532
    },
    {
      "epoch": 1.87,
      "grad_norm": 11.919846534729004,
      "learning_rate": 1.9693935926773458e-06,
      "loss": 0.8067,
      "step": 6533
    },
    {
      "epoch": 1.87,
      "grad_norm": 8.74809741973877,
      "learning_rate": 1.9651029748283754e-06,
      "loss": 0.5937,
      "step": 6534
    },
    {
      "epoch": 1.87,
      "grad_norm": 9.585488319396973,
      "learning_rate": 1.9608123569794054e-06,
      "loss": 0.5007,
      "step": 6535
    },
    {
      "epoch": 1.87,
      "grad_norm": 10.856274604797363,
      "learning_rate": 1.9565217391304346e-06,
      "loss": 0.6984,
      "step": 6536
    },
    {
      "epoch": 1.87,
      "grad_norm": 9.430891036987305,
      "learning_rate": 1.9522311212814646e-06,
      "loss": 0.4644,
      "step": 6537
    },
    {
      "epoch": 1.87,
      "grad_norm": 12.390891075134277,
      "learning_rate": 1.9479405034324943e-06,
      "loss": 0.7448,
      "step": 6538
    },
    {
      "epoch": 1.87,
      "grad_norm": 11.068594932556152,
      "learning_rate": 1.9436498855835243e-06,
      "loss": 0.8033,
      "step": 6539
    },
    {
      "epoch": 1.87,
      "grad_norm": 10.79172134399414,
      "learning_rate": 1.939359267734554e-06,
      "loss": 0.7613,
      "step": 6540
    },
    {
      "epoch": 1.87,
      "grad_norm": 12.126513481140137,
      "learning_rate": 1.9350686498855835e-06,
      "loss": 0.9048,
      "step": 6541
    },
    {
      "epoch": 1.87,
      "grad_norm": 9.405591011047363,
      "learning_rate": 1.930778032036613e-06,
      "loss": 0.4953,
      "step": 6542
    },
    {
      "epoch": 1.87,
      "grad_norm": 12.395812034606934,
      "learning_rate": 1.9264874141876428e-06,
      "loss": 0.9121,
      "step": 6543
    },
    {
      "epoch": 1.87,
      "grad_norm": 12.50573444366455,
      "learning_rate": 1.922196796338673e-06,
      "loss": 0.714,
      "step": 6544
    },
    {
      "epoch": 1.87,
      "grad_norm": 9.9764986038208,
      "learning_rate": 1.9179061784897024e-06,
      "loss": 0.5217,
      "step": 6545
    },
    {
      "epoch": 1.87,
      "grad_norm": 12.422931671142578,
      "learning_rate": 1.9136155606407325e-06,
      "loss": 0.7425,
      "step": 6546
    },
    {
      "epoch": 1.87,
      "grad_norm": 11.33634090423584,
      "learning_rate": 1.909324942791762e-06,
      "loss": 0.7039,
      "step": 6547
    },
    {
      "epoch": 1.87,
      "grad_norm": 8.913440704345703,
      "learning_rate": 1.905034324942792e-06,
      "loss": 0.5332,
      "step": 6548
    },
    {
      "epoch": 1.87,
      "grad_norm": 10.53049373626709,
      "learning_rate": 1.9007437070938215e-06,
      "loss": 0.6939,
      "step": 6549
    },
    {
      "epoch": 1.87,
      "grad_norm": 9.380077362060547,
      "learning_rate": 1.8964530892448515e-06,
      "loss": 0.5322,
      "step": 6550
    },
    {
      "epoch": 1.87,
      "grad_norm": 8.71501350402832,
      "learning_rate": 1.892162471395881e-06,
      "loss": 0.6613,
      "step": 6551
    },
    {
      "epoch": 1.87,
      "grad_norm": 10.52595043182373,
      "learning_rate": 1.887871853546911e-06,
      "loss": 0.8388,
      "step": 6552
    },
    {
      "epoch": 1.87,
      "grad_norm": 18.139190673828125,
      "learning_rate": 1.8835812356979406e-06,
      "loss": 0.7646,
      "step": 6553
    },
    {
      "epoch": 1.87,
      "grad_norm": 14.185227394104004,
      "learning_rate": 1.8792906178489704e-06,
      "loss": 0.8541,
      "step": 6554
    },
    {
      "epoch": 1.88,
      "grad_norm": 11.55553913116455,
      "learning_rate": 1.875e-06,
      "loss": 0.7206,
      "step": 6555
    },
    {
      "epoch": 1.88,
      "grad_norm": 10.506942749023438,
      "learning_rate": 1.8707093821510299e-06,
      "loss": 0.5802,
      "step": 6556
    },
    {
      "epoch": 1.88,
      "grad_norm": 12.027273178100586,
      "learning_rate": 1.8664187643020595e-06,
      "loss": 0.6196,
      "step": 6557
    },
    {
      "epoch": 1.88,
      "grad_norm": 8.904149055480957,
      "learning_rate": 1.8621281464530893e-06,
      "loss": 0.6362,
      "step": 6558
    },
    {
      "epoch": 1.88,
      "grad_norm": 9.9290132522583,
      "learning_rate": 1.8578375286041191e-06,
      "loss": 0.6015,
      "step": 6559
    },
    {
      "epoch": 1.88,
      "grad_norm": 10.648885726928711,
      "learning_rate": 1.8535469107551488e-06,
      "loss": 0.6246,
      "step": 6560
    },
    {
      "epoch": 1.88,
      "grad_norm": 12.099038124084473,
      "learning_rate": 1.8492562929061786e-06,
      "loss": 0.6298,
      "step": 6561
    },
    {
      "epoch": 1.88,
      "grad_norm": 10.756821632385254,
      "learning_rate": 1.8449656750572084e-06,
      "loss": 0.67,
      "step": 6562
    },
    {
      "epoch": 1.88,
      "grad_norm": 10.574438095092773,
      "learning_rate": 1.840675057208238e-06,
      "loss": 0.4984,
      "step": 6563
    },
    {
      "epoch": 1.88,
      "grad_norm": 13.0107421875,
      "learning_rate": 1.8363844393592676e-06,
      "loss": 0.7365,
      "step": 6564
    },
    {
      "epoch": 1.88,
      "grad_norm": 10.021994590759277,
      "learning_rate": 1.8320938215102975e-06,
      "loss": 0.7611,
      "step": 6565
    },
    {
      "epoch": 1.88,
      "grad_norm": 10.102065086364746,
      "learning_rate": 1.8278032036613273e-06,
      "loss": 0.5019,
      "step": 6566
    },
    {
      "epoch": 1.88,
      "grad_norm": 12.003802299499512,
      "learning_rate": 1.823512585812357e-06,
      "loss": 0.8283,
      "step": 6567
    },
    {
      "epoch": 1.88,
      "grad_norm": 11.736248016357422,
      "learning_rate": 1.8192219679633867e-06,
      "loss": 0.7633,
      "step": 6568
    },
    {
      "epoch": 1.88,
      "grad_norm": 9.837981224060059,
      "learning_rate": 1.8149313501144166e-06,
      "loss": 0.761,
      "step": 6569
    },
    {
      "epoch": 1.88,
      "grad_norm": 13.176420211791992,
      "learning_rate": 1.8106407322654462e-06,
      "loss": 0.818,
      "step": 6570
    },
    {
      "epoch": 1.88,
      "grad_norm": 10.442317008972168,
      "learning_rate": 1.806350114416476e-06,
      "loss": 0.4333,
      "step": 6571
    },
    {
      "epoch": 1.88,
      "grad_norm": 12.715324401855469,
      "learning_rate": 1.8020594965675058e-06,
      "loss": 0.8797,
      "step": 6572
    },
    {
      "epoch": 1.88,
      "grad_norm": 12.06436538696289,
      "learning_rate": 1.7977688787185357e-06,
      "loss": 0.9243,
      "step": 6573
    },
    {
      "epoch": 1.88,
      "grad_norm": 11.596464157104492,
      "learning_rate": 1.7934782608695653e-06,
      "loss": 0.6724,
      "step": 6574
    },
    {
      "epoch": 1.88,
      "grad_norm": 9.028141021728516,
      "learning_rate": 1.789187643020595e-06,
      "loss": 0.6722,
      "step": 6575
    },
    {
      "epoch": 1.88,
      "grad_norm": 12.241581916809082,
      "learning_rate": 1.784897025171625e-06,
      "loss": 0.4524,
      "step": 6576
    },
    {
      "epoch": 1.88,
      "grad_norm": 12.716862678527832,
      "learning_rate": 1.7806064073226545e-06,
      "loss": 0.9328,
      "step": 6577
    },
    {
      "epoch": 1.88,
      "grad_norm": 9.879105567932129,
      "learning_rate": 1.7763157894736842e-06,
      "loss": 0.5696,
      "step": 6578
    },
    {
      "epoch": 1.88,
      "grad_norm": 9.3067626953125,
      "learning_rate": 1.772025171624714e-06,
      "loss": 0.6508,
      "step": 6579
    },
    {
      "epoch": 1.88,
      "grad_norm": 11.868824005126953,
      "learning_rate": 1.7677345537757436e-06,
      "loss": 0.7221,
      "step": 6580
    },
    {
      "epoch": 1.88,
      "grad_norm": 11.533590316772461,
      "learning_rate": 1.7634439359267734e-06,
      "loss": 0.658,
      "step": 6581
    },
    {
      "epoch": 1.88,
      "grad_norm": 9.161964416503906,
      "learning_rate": 1.7591533180778033e-06,
      "loss": 0.4541,
      "step": 6582
    },
    {
      "epoch": 1.88,
      "grad_norm": 11.080299377441406,
      "learning_rate": 1.7548627002288329e-06,
      "loss": 0.5937,
      "step": 6583
    },
    {
      "epoch": 1.88,
      "grad_norm": 11.026849746704102,
      "learning_rate": 1.7505720823798627e-06,
      "loss": 0.8369,
      "step": 6584
    },
    {
      "epoch": 1.88,
      "grad_norm": 13.56468391418457,
      "learning_rate": 1.7462814645308925e-06,
      "loss": 0.6955,
      "step": 6585
    },
    {
      "epoch": 1.88,
      "grad_norm": 11.296615600585938,
      "learning_rate": 1.7419908466819223e-06,
      "loss": 0.5353,
      "step": 6586
    },
    {
      "epoch": 1.88,
      "grad_norm": 12.954611778259277,
      "learning_rate": 1.737700228832952e-06,
      "loss": 0.961,
      "step": 6587
    },
    {
      "epoch": 1.88,
      "grad_norm": 14.633906364440918,
      "learning_rate": 1.7334096109839818e-06,
      "loss": 0.8205,
      "step": 6588
    },
    {
      "epoch": 1.88,
      "grad_norm": 8.348833084106445,
      "learning_rate": 1.7291189931350116e-06,
      "loss": 0.5662,
      "step": 6589
    },
    {
      "epoch": 1.89,
      "grad_norm": 11.917252540588379,
      "learning_rate": 1.7248283752860412e-06,
      "loss": 0.7883,
      "step": 6590
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.018460273742676,
      "learning_rate": 1.720537757437071e-06,
      "loss": 0.5761,
      "step": 6591
    },
    {
      "epoch": 1.89,
      "grad_norm": 9.583654403686523,
      "learning_rate": 1.7162471395881009e-06,
      "loss": 0.653,
      "step": 6592
    },
    {
      "epoch": 1.89,
      "grad_norm": 8.523519515991211,
      "learning_rate": 1.7119565217391303e-06,
      "loss": 0.599,
      "step": 6593
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.856172561645508,
      "learning_rate": 1.7076659038901601e-06,
      "loss": 0.6536,
      "step": 6594
    },
    {
      "epoch": 1.89,
      "grad_norm": 13.230097770690918,
      "learning_rate": 1.70337528604119e-06,
      "loss": 0.8298,
      "step": 6595
    },
    {
      "epoch": 1.89,
      "grad_norm": 8.523895263671875,
      "learning_rate": 1.6990846681922198e-06,
      "loss": 0.4115,
      "step": 6596
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.178942680358887,
      "learning_rate": 1.6947940503432494e-06,
      "loss": 0.4652,
      "step": 6597
    },
    {
      "epoch": 1.89,
      "grad_norm": 9.806584358215332,
      "learning_rate": 1.6905034324942792e-06,
      "loss": 0.5653,
      "step": 6598
    },
    {
      "epoch": 1.89,
      "grad_norm": 9.303384780883789,
      "learning_rate": 1.686212814645309e-06,
      "loss": 0.515,
      "step": 6599
    },
    {
      "epoch": 1.89,
      "grad_norm": 8.282113075256348,
      "learning_rate": 1.6819221967963386e-06,
      "loss": 0.6797,
      "step": 6600
    },
    {
      "epoch": 1.89,
      "grad_norm": 11.019519805908203,
      "learning_rate": 1.6776315789473685e-06,
      "loss": 0.7665,
      "step": 6601
    },
    {
      "epoch": 1.89,
      "grad_norm": 9.183219909667969,
      "learning_rate": 1.6733409610983983e-06,
      "loss": 0.4708,
      "step": 6602
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.683391571044922,
      "learning_rate": 1.669050343249428e-06,
      "loss": 0.6266,
      "step": 6603
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.17486572265625,
      "learning_rate": 1.6647597254004577e-06,
      "loss": 0.5132,
      "step": 6604
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.503311157226562,
      "learning_rate": 1.6604691075514876e-06,
      "loss": 0.5902,
      "step": 6605
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.002145767211914,
      "learning_rate": 1.6561784897025172e-06,
      "loss": 0.8209,
      "step": 6606
    },
    {
      "epoch": 1.89,
      "grad_norm": 9.178977012634277,
      "learning_rate": 1.651887871853547e-06,
      "loss": 0.5495,
      "step": 6607
    },
    {
      "epoch": 1.89,
      "grad_norm": 8.63107681274414,
      "learning_rate": 1.6475972540045766e-06,
      "loss": 0.5714,
      "step": 6608
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.814050674438477,
      "learning_rate": 1.6433066361556065e-06,
      "loss": 0.6961,
      "step": 6609
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.427483558654785,
      "learning_rate": 1.639016018306636e-06,
      "loss": 0.6196,
      "step": 6610
    },
    {
      "epoch": 1.89,
      "grad_norm": 9.812458038330078,
      "learning_rate": 1.634725400457666e-06,
      "loss": 0.5838,
      "step": 6611
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.628884315490723,
      "learning_rate": 1.6304347826086957e-06,
      "loss": 0.8683,
      "step": 6612
    },
    {
      "epoch": 1.89,
      "grad_norm": 9.876553535461426,
      "learning_rate": 1.6261441647597253e-06,
      "loss": 0.6857,
      "step": 6613
    },
    {
      "epoch": 1.89,
      "grad_norm": 8.790066719055176,
      "learning_rate": 1.6218535469107552e-06,
      "loss": 0.7048,
      "step": 6614
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.74987506866455,
      "learning_rate": 1.617562929061785e-06,
      "loss": 0.6891,
      "step": 6615
    },
    {
      "epoch": 1.89,
      "grad_norm": 9.740676879882812,
      "learning_rate": 1.6132723112128146e-06,
      "loss": 0.6121,
      "step": 6616
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.547184944152832,
      "learning_rate": 1.6089816933638444e-06,
      "loss": 0.7235,
      "step": 6617
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.935553550720215,
      "learning_rate": 1.6046910755148743e-06,
      "loss": 0.6277,
      "step": 6618
    },
    {
      "epoch": 1.89,
      "grad_norm": 12.542131423950195,
      "learning_rate": 1.600400457665904e-06,
      "loss": 0.6723,
      "step": 6619
    },
    {
      "epoch": 1.89,
      "grad_norm": 12.385489463806152,
      "learning_rate": 1.5961098398169337e-06,
      "loss": 0.8717,
      "step": 6620
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.140175819396973,
      "learning_rate": 1.5918192219679635e-06,
      "loss": 0.5769,
      "step": 6621
    },
    {
      "epoch": 1.89,
      "grad_norm": 9.585509300231934,
      "learning_rate": 1.5875286041189934e-06,
      "loss": 0.4811,
      "step": 6622
    },
    {
      "epoch": 1.89,
      "grad_norm": 9.377813339233398,
      "learning_rate": 1.5832379862700228e-06,
      "loss": 0.5785,
      "step": 6623
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.11171817779541,
      "learning_rate": 1.5789473684210526e-06,
      "loss": 0.5008,
      "step": 6624
    },
    {
      "epoch": 1.9,
      "grad_norm": 9.451621055603027,
      "learning_rate": 1.5746567505720824e-06,
      "loss": 0.6607,
      "step": 6625
    },
    {
      "epoch": 1.9,
      "grad_norm": 9.769519805908203,
      "learning_rate": 1.570366132723112e-06,
      "loss": 0.7866,
      "step": 6626
    },
    {
      "epoch": 1.9,
      "grad_norm": 10.366878509521484,
      "learning_rate": 1.5660755148741419e-06,
      "loss": 0.6435,
      "step": 6627
    },
    {
      "epoch": 1.9,
      "grad_norm": 9.071807861328125,
      "learning_rate": 1.5617848970251717e-06,
      "loss": 0.5372,
      "step": 6628
    },
    {
      "epoch": 1.9,
      "grad_norm": 10.883258819580078,
      "learning_rate": 1.5574942791762013e-06,
      "loss": 0.6525,
      "step": 6629
    },
    {
      "epoch": 1.9,
      "grad_norm": 9.583721160888672,
      "learning_rate": 1.5532036613272311e-06,
      "loss": 0.6439,
      "step": 6630
    },
    {
      "epoch": 1.9,
      "grad_norm": 15.856637954711914,
      "learning_rate": 1.548913043478261e-06,
      "loss": 0.9104,
      "step": 6631
    },
    {
      "epoch": 1.9,
      "grad_norm": 12.541048049926758,
      "learning_rate": 1.5446224256292908e-06,
      "loss": 0.8598,
      "step": 6632
    },
    {
      "epoch": 1.9,
      "grad_norm": 9.865492820739746,
      "learning_rate": 1.5403318077803204e-06,
      "loss": 0.6173,
      "step": 6633
    },
    {
      "epoch": 1.9,
      "grad_norm": 9.361900329589844,
      "learning_rate": 1.5360411899313502e-06,
      "loss": 0.6299,
      "step": 6634
    },
    {
      "epoch": 1.9,
      "grad_norm": 12.05883502960205,
      "learning_rate": 1.53175057208238e-06,
      "loss": 0.6691,
      "step": 6635
    },
    {
      "epoch": 1.9,
      "grad_norm": 12.530620574951172,
      "learning_rate": 1.5274599542334097e-06,
      "loss": 0.8825,
      "step": 6636
    },
    {
      "epoch": 1.9,
      "grad_norm": 10.966458320617676,
      "learning_rate": 1.5231693363844395e-06,
      "loss": 0.5808,
      "step": 6637
    },
    {
      "epoch": 1.9,
      "grad_norm": 10.090479850769043,
      "learning_rate": 1.518878718535469e-06,
      "loss": 0.5668,
      "step": 6638
    },
    {
      "epoch": 1.9,
      "grad_norm": 11.146265029907227,
      "learning_rate": 1.5145881006864987e-06,
      "loss": 0.6695,
      "step": 6639
    },
    {
      "epoch": 1.9,
      "grad_norm": 9.504698753356934,
      "learning_rate": 1.5102974828375285e-06,
      "loss": 0.6011,
      "step": 6640
    },
    {
      "epoch": 1.9,
      "grad_norm": 9.853711128234863,
      "learning_rate": 1.5060068649885584e-06,
      "loss": 0.6912,
      "step": 6641
    },
    {
      "epoch": 1.9,
      "grad_norm": 12.366127967834473,
      "learning_rate": 1.5017162471395882e-06,
      "loss": 0.6353,
      "step": 6642
    },
    {
      "epoch": 1.9,
      "grad_norm": 10.047423362731934,
      "learning_rate": 1.4974256292906178e-06,
      "loss": 0.5828,
      "step": 6643
    },
    {
      "epoch": 1.9,
      "grad_norm": 10.754246711730957,
      "learning_rate": 1.4931350114416476e-06,
      "loss": 0.807,
      "step": 6644
    },
    {
      "epoch": 1.9,
      "grad_norm": 10.070907592773438,
      "learning_rate": 1.4888443935926775e-06,
      "loss": 0.5376,
      "step": 6645
    },
    {
      "epoch": 1.9,
      "grad_norm": 10.664957046508789,
      "learning_rate": 1.484553775743707e-06,
      "loss": 0.5351,
      "step": 6646
    },
    {
      "epoch": 1.9,
      "grad_norm": 9.667845726013184,
      "learning_rate": 1.480263157894737e-06,
      "loss": 0.5598,
      "step": 6647
    },
    {
      "epoch": 1.9,
      "grad_norm": 9.291950225830078,
      "learning_rate": 1.4759725400457667e-06,
      "loss": 0.5589,
      "step": 6648
    },
    {
      "epoch": 1.9,
      "grad_norm": 11.45644760131836,
      "learning_rate": 1.4716819221967963e-06,
      "loss": 0.9333,
      "step": 6649
    },
    {
      "epoch": 1.9,
      "grad_norm": 11.348915100097656,
      "learning_rate": 1.4673913043478262e-06,
      "loss": 0.7001,
      "step": 6650
    },
    {
      "epoch": 1.9,
      "grad_norm": 12.178325653076172,
      "learning_rate": 1.463100686498856e-06,
      "loss": 0.9092,
      "step": 6651
    },
    {
      "epoch": 1.9,
      "grad_norm": 10.490787506103516,
      "learning_rate": 1.4588100686498856e-06,
      "loss": 0.5223,
      "step": 6652
    },
    {
      "epoch": 1.9,
      "grad_norm": 8.094230651855469,
      "learning_rate": 1.4545194508009152e-06,
      "loss": 0.543,
      "step": 6653
    },
    {
      "epoch": 1.9,
      "grad_norm": 10.10330581665039,
      "learning_rate": 1.450228832951945e-06,
      "loss": 0.6563,
      "step": 6654
    },
    {
      "epoch": 1.9,
      "grad_norm": 11.853269577026367,
      "learning_rate": 1.4459382151029749e-06,
      "loss": 0.7333,
      "step": 6655
    },
    {
      "epoch": 1.9,
      "grad_norm": 11.071447372436523,
      "learning_rate": 1.4416475972540045e-06,
      "loss": 0.7413,
      "step": 6656
    },
    {
      "epoch": 1.9,
      "grad_norm": 11.710362434387207,
      "learning_rate": 1.4373569794050343e-06,
      "loss": 0.55,
      "step": 6657
    },
    {
      "epoch": 1.9,
      "grad_norm": 11.374383926391602,
      "learning_rate": 1.4330663615560641e-06,
      "loss": 0.7556,
      "step": 6658
    },
    {
      "epoch": 1.9,
      "grad_norm": 11.201468467712402,
      "learning_rate": 1.4287757437070938e-06,
      "loss": 0.6124,
      "step": 6659
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.033202171325684,
      "learning_rate": 1.4244851258581236e-06,
      "loss": 0.5549,
      "step": 6660
    },
    {
      "epoch": 1.91,
      "grad_norm": 8.656717300415039,
      "learning_rate": 1.4201945080091534e-06,
      "loss": 0.4417,
      "step": 6661
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.967692375183105,
      "learning_rate": 1.415903890160183e-06,
      "loss": 0.8864,
      "step": 6662
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.697224617004395,
      "learning_rate": 1.4116132723112129e-06,
      "loss": 0.6277,
      "step": 6663
    },
    {
      "epoch": 1.91,
      "grad_norm": 11.160284042358398,
      "learning_rate": 1.4073226544622427e-06,
      "loss": 0.6296,
      "step": 6664
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.617480278015137,
      "learning_rate": 1.4030320366132725e-06,
      "loss": 0.6303,
      "step": 6665
    },
    {
      "epoch": 1.91,
      "grad_norm": 11.47805118560791,
      "learning_rate": 1.3987414187643021e-06,
      "loss": 0.588,
      "step": 6666
    },
    {
      "epoch": 1.91,
      "grad_norm": 9.959935188293457,
      "learning_rate": 1.394450800915332e-06,
      "loss": 0.759,
      "step": 6667
    },
    {
      "epoch": 1.91,
      "grad_norm": 9.430790901184082,
      "learning_rate": 1.3901601830663618e-06,
      "loss": 0.6728,
      "step": 6668
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.529243469238281,
      "learning_rate": 1.3858695652173912e-06,
      "loss": 0.494,
      "step": 6669
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.530109405517578,
      "learning_rate": 1.381578947368421e-06,
      "loss": 0.8574,
      "step": 6670
    },
    {
      "epoch": 1.91,
      "grad_norm": 8.783020973205566,
      "learning_rate": 1.3772883295194508e-06,
      "loss": 0.6715,
      "step": 6671
    },
    {
      "epoch": 1.91,
      "grad_norm": 9.161794662475586,
      "learning_rate": 1.3729977116704805e-06,
      "loss": 0.6489,
      "step": 6672
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.282054901123047,
      "learning_rate": 1.3687070938215103e-06,
      "loss": 0.5719,
      "step": 6673
    },
    {
      "epoch": 1.91,
      "grad_norm": 9.306244850158691,
      "learning_rate": 1.36441647597254e-06,
      "loss": 0.7034,
      "step": 6674
    },
    {
      "epoch": 1.91,
      "grad_norm": 12.614920616149902,
      "learning_rate": 1.3601258581235697e-06,
      "loss": 0.6356,
      "step": 6675
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.753173828125,
      "learning_rate": 1.3558352402745995e-06,
      "loss": 0.7029,
      "step": 6676
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.913387298583984,
      "learning_rate": 1.3515446224256294e-06,
      "loss": 0.6301,
      "step": 6677
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.129559516906738,
      "learning_rate": 1.3472540045766592e-06,
      "loss": 0.7076,
      "step": 6678
    },
    {
      "epoch": 1.91,
      "grad_norm": 7.9117608070373535,
      "learning_rate": 1.3429633867276888e-06,
      "loss": 0.427,
      "step": 6679
    },
    {
      "epoch": 1.91,
      "grad_norm": 11.85035228729248,
      "learning_rate": 1.3386727688787186e-06,
      "loss": 0.6972,
      "step": 6680
    },
    {
      "epoch": 1.91,
      "grad_norm": 8.53018569946289,
      "learning_rate": 1.3343821510297485e-06,
      "loss": 0.5078,
      "step": 6681
    },
    {
      "epoch": 1.91,
      "grad_norm": 8.861907958984375,
      "learning_rate": 1.330091533180778e-06,
      "loss": 0.6506,
      "step": 6682
    },
    {
      "epoch": 1.91,
      "grad_norm": 9.757502555847168,
      "learning_rate": 1.325800915331808e-06,
      "loss": 0.6035,
      "step": 6683
    },
    {
      "epoch": 1.91,
      "grad_norm": 9.364039421081543,
      "learning_rate": 1.3215102974828375e-06,
      "loss": 0.5463,
      "step": 6684
    },
    {
      "epoch": 1.91,
      "grad_norm": 9.148898124694824,
      "learning_rate": 1.3172196796338671e-06,
      "loss": 0.538,
      "step": 6685
    },
    {
      "epoch": 1.91,
      "grad_norm": 11.64253044128418,
      "learning_rate": 1.312929061784897e-06,
      "loss": 0.6017,
      "step": 6686
    },
    {
      "epoch": 1.91,
      "grad_norm": 8.908527374267578,
      "learning_rate": 1.3086384439359268e-06,
      "loss": 0.4138,
      "step": 6687
    },
    {
      "epoch": 1.91,
      "grad_norm": 9.46848201751709,
      "learning_rate": 1.3043478260869566e-06,
      "loss": 0.7496,
      "step": 6688
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.075117111206055,
      "learning_rate": 1.3000572082379862e-06,
      "loss": 0.8197,
      "step": 6689
    },
    {
      "epoch": 1.91,
      "grad_norm": 11.487811088562012,
      "learning_rate": 1.295766590389016e-06,
      "loss": 0.7324,
      "step": 6690
    },
    {
      "epoch": 1.91,
      "grad_norm": 8.055896759033203,
      "learning_rate": 1.2914759725400459e-06,
      "loss": 0.3772,
      "step": 6691
    },
    {
      "epoch": 1.91,
      "grad_norm": 9.442093849182129,
      "learning_rate": 1.2871853546910755e-06,
      "loss": 0.4689,
      "step": 6692
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.520005226135254,
      "learning_rate": 1.2828947368421053e-06,
      "loss": 0.6256,
      "step": 6693
    },
    {
      "epoch": 1.91,
      "grad_norm": 11.794526100158691,
      "learning_rate": 1.2786041189931352e-06,
      "loss": 0.636,
      "step": 6694
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.728378295898438,
      "learning_rate": 1.2743135011441648e-06,
      "loss": 0.6543,
      "step": 6695
    },
    {
      "epoch": 1.92,
      "grad_norm": 10.640963554382324,
      "learning_rate": 1.2700228832951946e-06,
      "loss": 0.6148,
      "step": 6696
    },
    {
      "epoch": 1.92,
      "grad_norm": 11.699413299560547,
      "learning_rate": 1.2657322654462244e-06,
      "loss": 0.766,
      "step": 6697
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.714559555053711,
      "learning_rate": 1.261441647597254e-06,
      "loss": 0.5721,
      "step": 6698
    },
    {
      "epoch": 1.92,
      "grad_norm": 8.230494499206543,
      "learning_rate": 1.2571510297482837e-06,
      "loss": 0.5532,
      "step": 6699
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.15118408203125,
      "learning_rate": 1.2528604118993135e-06,
      "loss": 0.5442,
      "step": 6700
    },
    {
      "epoch": 1.92,
      "grad_norm": 10.485254287719727,
      "learning_rate": 1.2485697940503433e-06,
      "loss": 0.6268,
      "step": 6701
    },
    {
      "epoch": 1.92,
      "grad_norm": 10.303457260131836,
      "learning_rate": 1.244279176201373e-06,
      "loss": 0.7238,
      "step": 6702
    },
    {
      "epoch": 1.92,
      "grad_norm": 10.692706108093262,
      "learning_rate": 1.2399885583524027e-06,
      "loss": 0.4998,
      "step": 6703
    },
    {
      "epoch": 1.92,
      "grad_norm": 14.482782363891602,
      "learning_rate": 1.2356979405034326e-06,
      "loss": 0.9019,
      "step": 6704
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.645691871643066,
      "learning_rate": 1.2314073226544622e-06,
      "loss": 0.707,
      "step": 6705
    },
    {
      "epoch": 1.92,
      "grad_norm": 10.623063087463379,
      "learning_rate": 1.227116704805492e-06,
      "loss": 0.843,
      "step": 6706
    },
    {
      "epoch": 1.92,
      "grad_norm": 10.0907621383667,
      "learning_rate": 1.2228260869565218e-06,
      "loss": 0.8514,
      "step": 6707
    },
    {
      "epoch": 1.92,
      "grad_norm": 11.34632396697998,
      "learning_rate": 1.2185354691075515e-06,
      "loss": 0.853,
      "step": 6708
    },
    {
      "epoch": 1.92,
      "grad_norm": 8.570047378540039,
      "learning_rate": 1.2142448512585813e-06,
      "loss": 0.4797,
      "step": 6709
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.260835647583008,
      "learning_rate": 1.2099542334096111e-06,
      "loss": 0.6946,
      "step": 6710
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.483461380004883,
      "learning_rate": 1.205663615560641e-06,
      "loss": 0.3813,
      "step": 6711
    },
    {
      "epoch": 1.92,
      "grad_norm": 11.182538032531738,
      "learning_rate": 1.2013729977116706e-06,
      "loss": 0.6669,
      "step": 6712
    },
    {
      "epoch": 1.92,
      "grad_norm": 8.788228034973145,
      "learning_rate": 1.1970823798627004e-06,
      "loss": 0.3281,
      "step": 6713
    },
    {
      "epoch": 1.92,
      "grad_norm": 11.311713218688965,
      "learning_rate": 1.19279176201373e-06,
      "loss": 0.5773,
      "step": 6714
    },
    {
      "epoch": 1.92,
      "grad_norm": 12.638751029968262,
      "learning_rate": 1.1885011441647596e-06,
      "loss": 1.0074,
      "step": 6715
    },
    {
      "epoch": 1.92,
      "grad_norm": 10.119768142700195,
      "learning_rate": 1.1842105263157894e-06,
      "loss": 0.6024,
      "step": 6716
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.68479061126709,
      "learning_rate": 1.1799199084668193e-06,
      "loss": 0.6385,
      "step": 6717
    },
    {
      "epoch": 1.92,
      "grad_norm": 8.239176750183105,
      "learning_rate": 1.1756292906178489e-06,
      "loss": 0.524,
      "step": 6718
    },
    {
      "epoch": 1.92,
      "grad_norm": 10.428190231323242,
      "learning_rate": 1.1713386727688787e-06,
      "loss": 0.6015,
      "step": 6719
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.850186347961426,
      "learning_rate": 1.1670480549199085e-06,
      "loss": 0.7612,
      "step": 6720
    },
    {
      "epoch": 1.92,
      "grad_norm": 12.244690895080566,
      "learning_rate": 1.1627574370709381e-06,
      "loss": 0.7536,
      "step": 6721
    },
    {
      "epoch": 1.92,
      "grad_norm": 10.301718711853027,
      "learning_rate": 1.158466819221968e-06,
      "loss": 0.6864,
      "step": 6722
    },
    {
      "epoch": 1.92,
      "grad_norm": 11.460555076599121,
      "learning_rate": 1.1541762013729978e-06,
      "loss": 0.6202,
      "step": 6723
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.115154266357422,
      "learning_rate": 1.1498855835240276e-06,
      "loss": 0.6418,
      "step": 6724
    },
    {
      "epoch": 1.92,
      "grad_norm": 11.595759391784668,
      "learning_rate": 1.1455949656750572e-06,
      "loss": 0.624,
      "step": 6725
    },
    {
      "epoch": 1.92,
      "grad_norm": 12.14228343963623,
      "learning_rate": 1.141304347826087e-06,
      "loss": 0.7796,
      "step": 6726
    },
    {
      "epoch": 1.92,
      "grad_norm": 10.284402847290039,
      "learning_rate": 1.137013729977117e-06,
      "loss": 0.5669,
      "step": 6727
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.583091735839844,
      "learning_rate": 1.1327231121281465e-06,
      "loss": 0.5784,
      "step": 6728
    },
    {
      "epoch": 1.92,
      "grad_norm": 11.306744575500488,
      "learning_rate": 1.1284324942791761e-06,
      "loss": 0.7993,
      "step": 6729
    },
    {
      "epoch": 1.93,
      "grad_norm": 19.395160675048828,
      "learning_rate": 1.124141876430206e-06,
      "loss": 0.7529,
      "step": 6730
    },
    {
      "epoch": 1.93,
      "grad_norm": 7.858729362487793,
      "learning_rate": 1.1198512585812356e-06,
      "loss": 0.6799,
      "step": 6731
    },
    {
      "epoch": 1.93,
      "grad_norm": 10.490175247192383,
      "learning_rate": 1.1155606407322654e-06,
      "loss": 0.8072,
      "step": 6732
    },
    {
      "epoch": 1.93,
      "grad_norm": 10.685630798339844,
      "learning_rate": 1.1112700228832952e-06,
      "loss": 0.5765,
      "step": 6733
    },
    {
      "epoch": 1.93,
      "grad_norm": 11.851250648498535,
      "learning_rate": 1.106979405034325e-06,
      "loss": 0.5958,
      "step": 6734
    },
    {
      "epoch": 1.93,
      "grad_norm": 10.589179039001465,
      "learning_rate": 1.1026887871853547e-06,
      "loss": 0.7273,
      "step": 6735
    },
    {
      "epoch": 1.93,
      "grad_norm": 8.347055435180664,
      "learning_rate": 1.0983981693363845e-06,
      "loss": 0.6655,
      "step": 6736
    },
    {
      "epoch": 1.93,
      "grad_norm": 10.162137985229492,
      "learning_rate": 1.0941075514874143e-06,
      "loss": 0.7468,
      "step": 6737
    },
    {
      "epoch": 1.93,
      "grad_norm": 11.438544273376465,
      "learning_rate": 1.089816933638444e-06,
      "loss": 0.7288,
      "step": 6738
    },
    {
      "epoch": 1.93,
      "grad_norm": 8.79272747039795,
      "learning_rate": 1.0855263157894738e-06,
      "loss": 0.6498,
      "step": 6739
    },
    {
      "epoch": 1.93,
      "grad_norm": 10.488850593566895,
      "learning_rate": 1.0812356979405036e-06,
      "loss": 0.6367,
      "step": 6740
    },
    {
      "epoch": 1.93,
      "grad_norm": 8.263250350952148,
      "learning_rate": 1.0769450800915332e-06,
      "loss": 0.5896,
      "step": 6741
    },
    {
      "epoch": 1.93,
      "grad_norm": 11.923199653625488,
      "learning_rate": 1.072654462242563e-06,
      "loss": 0.788,
      "step": 6742
    },
    {
      "epoch": 1.93,
      "grad_norm": 9.139735221862793,
      "learning_rate": 1.0683638443935929e-06,
      "loss": 0.5462,
      "step": 6743
    },
    {
      "epoch": 1.93,
      "grad_norm": 10.757696151733398,
      "learning_rate": 1.0640732265446223e-06,
      "loss": 0.6614,
      "step": 6744
    },
    {
      "epoch": 1.93,
      "grad_norm": 8.452191352844238,
      "learning_rate": 1.059782608695652e-06,
      "loss": 0.5517,
      "step": 6745
    },
    {
      "epoch": 1.93,
      "grad_norm": 11.318562507629395,
      "learning_rate": 1.055491990846682e-06,
      "loss": 0.6681,
      "step": 6746
    },
    {
      "epoch": 1.93,
      "grad_norm": 11.440497398376465,
      "learning_rate": 1.0512013729977117e-06,
      "loss": 0.7208,
      "step": 6747
    },
    {
      "epoch": 1.93,
      "grad_norm": 11.99815845489502,
      "learning_rate": 1.0469107551487413e-06,
      "loss": 0.6689,
      "step": 6748
    },
    {
      "epoch": 1.93,
      "grad_norm": 10.977968215942383,
      "learning_rate": 1.0426201372997712e-06,
      "loss": 0.5295,
      "step": 6749
    },
    {
      "epoch": 1.93,
      "grad_norm": 12.640535354614258,
      "learning_rate": 1.038329519450801e-06,
      "loss": 0.8584,
      "step": 6750
    },
    {
      "epoch": 1.93,
      "grad_norm": 44.93844985961914,
      "learning_rate": 1.0340389016018306e-06,
      "loss": 0.5963,
      "step": 6751
    },
    {
      "epoch": 1.93,
      "grad_norm": 11.811992645263672,
      "learning_rate": 1.0297482837528604e-06,
      "loss": 0.6537,
      "step": 6752
    },
    {
      "epoch": 1.93,
      "grad_norm": 9.109289169311523,
      "learning_rate": 1.0254576659038903e-06,
      "loss": 0.6972,
      "step": 6753
    },
    {
      "epoch": 1.93,
      "grad_norm": 12.60461711883545,
      "learning_rate": 1.0211670480549199e-06,
      "loss": 0.8511,
      "step": 6754
    },
    {
      "epoch": 1.93,
      "grad_norm": 10.249258995056152,
      "learning_rate": 1.0168764302059497e-06,
      "loss": 0.9507,
      "step": 6755
    },
    {
      "epoch": 1.93,
      "grad_norm": 11.223470687866211,
      "learning_rate": 1.0125858123569795e-06,
      "loss": 0.7589,
      "step": 6756
    },
    {
      "epoch": 1.93,
      "grad_norm": 11.984776496887207,
      "learning_rate": 1.0082951945080094e-06,
      "loss": 0.6934,
      "step": 6757
    },
    {
      "epoch": 1.93,
      "grad_norm": 12.150588035583496,
      "learning_rate": 1.004004576659039e-06,
      "loss": 0.6201,
      "step": 6758
    },
    {
      "epoch": 1.93,
      "grad_norm": 10.411704063415527,
      "learning_rate": 9.997139588100686e-07,
      "loss": 0.61,
      "step": 6759
    },
    {
      "epoch": 1.93,
      "grad_norm": 10.817537307739258,
      "learning_rate": 9.954233409610984e-07,
      "loss": 0.7173,
      "step": 6760
    },
    {
      "epoch": 1.93,
      "grad_norm": 10.725568771362305,
      "learning_rate": 9.91132723112128e-07,
      "loss": 0.6655,
      "step": 6761
    },
    {
      "epoch": 1.93,
      "grad_norm": 9.771780014038086,
      "learning_rate": 9.868421052631579e-07,
      "loss": 0.7,
      "step": 6762
    },
    {
      "epoch": 1.93,
      "grad_norm": 9.953685760498047,
      "learning_rate": 9.825514874141877e-07,
      "loss": 0.641,
      "step": 6763
    },
    {
      "epoch": 1.93,
      "grad_norm": 11.242582321166992,
      "learning_rate": 9.782608695652173e-07,
      "loss": 0.5187,
      "step": 6764
    },
    {
      "epoch": 1.94,
      "grad_norm": 11.4192533493042,
      "learning_rate": 9.739702517162471e-07,
      "loss": 0.5498,
      "step": 6765
    },
    {
      "epoch": 1.94,
      "grad_norm": 9.859990119934082,
      "learning_rate": 9.69679633867277e-07,
      "loss": 0.7167,
      "step": 6766
    },
    {
      "epoch": 1.94,
      "grad_norm": 15.310134887695312,
      "learning_rate": 9.653890160183066e-07,
      "loss": 1.1055,
      "step": 6767
    },
    {
      "epoch": 1.94,
      "grad_norm": 9.646717071533203,
      "learning_rate": 9.610983981693364e-07,
      "loss": 0.4438,
      "step": 6768
    },
    {
      "epoch": 1.94,
      "grad_norm": 10.532123565673828,
      "learning_rate": 9.568077803203662e-07,
      "loss": 0.6696,
      "step": 6769
    },
    {
      "epoch": 1.94,
      "grad_norm": 15.500516891479492,
      "learning_rate": 9.52517162471396e-07,
      "loss": 0.9405,
      "step": 6770
    },
    {
      "epoch": 1.94,
      "grad_norm": 11.07739543914795,
      "learning_rate": 9.482265446224258e-07,
      "loss": 0.8091,
      "step": 6771
    },
    {
      "epoch": 1.94,
      "grad_norm": 10.713896751403809,
      "learning_rate": 9.439359267734555e-07,
      "loss": 0.6927,
      "step": 6772
    },
    {
      "epoch": 1.94,
      "grad_norm": 10.74335765838623,
      "learning_rate": 9.396453089244852e-07,
      "loss": 0.6765,
      "step": 6773
    },
    {
      "epoch": 1.94,
      "grad_norm": 10.386445999145508,
      "learning_rate": 9.353546910755149e-07,
      "loss": 0.6199,
      "step": 6774
    },
    {
      "epoch": 1.94,
      "grad_norm": 9.495280265808105,
      "learning_rate": 9.310640732265447e-07,
      "loss": 0.4989,
      "step": 6775
    },
    {
      "epoch": 1.94,
      "grad_norm": 8.543712615966797,
      "learning_rate": 9.267734553775744e-07,
      "loss": 0.6049,
      "step": 6776
    },
    {
      "epoch": 1.94,
      "grad_norm": 12.661754608154297,
      "learning_rate": 9.224828375286042e-07,
      "loss": 0.8616,
      "step": 6777
    },
    {
      "epoch": 1.94,
      "grad_norm": 9.792861938476562,
      "learning_rate": 9.181922196796338e-07,
      "loss": 0.592,
      "step": 6778
    },
    {
      "epoch": 1.94,
      "grad_norm": 10.058728218078613,
      "learning_rate": 9.139016018306636e-07,
      "loss": 0.5957,
      "step": 6779
    },
    {
      "epoch": 1.94,
      "grad_norm": 11.858846664428711,
      "learning_rate": 9.096109839816934e-07,
      "loss": 0.8671,
      "step": 6780
    },
    {
      "epoch": 1.94,
      "grad_norm": 6.873499870300293,
      "learning_rate": 9.053203661327231e-07,
      "loss": 0.3545,
      "step": 6781
    },
    {
      "epoch": 1.94,
      "grad_norm": 10.451911926269531,
      "learning_rate": 9.010297482837529e-07,
      "loss": 0.6086,
      "step": 6782
    },
    {
      "epoch": 1.94,
      "grad_norm": 10.603643417358398,
      "learning_rate": 8.967391304347826e-07,
      "loss": 0.6285,
      "step": 6783
    },
    {
      "epoch": 1.94,
      "grad_norm": 11.095579147338867,
      "learning_rate": 8.924485125858125e-07,
      "loss": 0.6823,
      "step": 6784
    },
    {
      "epoch": 1.94,
      "grad_norm": 9.788519859313965,
      "learning_rate": 8.881578947368421e-07,
      "loss": 0.7352,
      "step": 6785
    },
    {
      "epoch": 1.94,
      "grad_norm": 11.36791706085205,
      "learning_rate": 8.838672768878718e-07,
      "loss": 0.7521,
      "step": 6786
    },
    {
      "epoch": 1.94,
      "grad_norm": 10.876045227050781,
      "learning_rate": 8.795766590389016e-07,
      "loss": 0.6124,
      "step": 6787
    },
    {
      "epoch": 1.94,
      "grad_norm": 14.375767707824707,
      "learning_rate": 8.752860411899313e-07,
      "loss": 0.8835,
      "step": 6788
    },
    {
      "epoch": 1.94,
      "grad_norm": 8.53582763671875,
      "learning_rate": 8.709954233409612e-07,
      "loss": 0.673,
      "step": 6789
    },
    {
      "epoch": 1.94,
      "grad_norm": 8.364534378051758,
      "learning_rate": 8.667048054919909e-07,
      "loss": 0.5763,
      "step": 6790
    },
    {
      "epoch": 1.94,
      "grad_norm": 16.54009246826172,
      "learning_rate": 8.624141876430206e-07,
      "loss": 0.6385,
      "step": 6791
    },
    {
      "epoch": 1.94,
      "grad_norm": 9.698732376098633,
      "learning_rate": 8.581235697940504e-07,
      "loss": 0.6774,
      "step": 6792
    },
    {
      "epoch": 1.94,
      "grad_norm": 10.97348690032959,
      "learning_rate": 8.538329519450801e-07,
      "loss": 0.6796,
      "step": 6793
    },
    {
      "epoch": 1.94,
      "grad_norm": 9.030004501342773,
      "learning_rate": 8.495423340961099e-07,
      "loss": 0.6908,
      "step": 6794
    },
    {
      "epoch": 1.94,
      "grad_norm": 11.053754806518555,
      "learning_rate": 8.452517162471396e-07,
      "loss": 0.9486,
      "step": 6795
    },
    {
      "epoch": 1.94,
      "grad_norm": 14.873150825500488,
      "learning_rate": 8.409610983981693e-07,
      "loss": 0.8398,
      "step": 6796
    },
    {
      "epoch": 1.94,
      "grad_norm": 11.530095100402832,
      "learning_rate": 8.366704805491992e-07,
      "loss": 0.6613,
      "step": 6797
    },
    {
      "epoch": 1.94,
      "grad_norm": 12.763251304626465,
      "learning_rate": 8.323798627002289e-07,
      "loss": 0.8406,
      "step": 6798
    },
    {
      "epoch": 1.94,
      "grad_norm": 11.035303115844727,
      "learning_rate": 8.280892448512586e-07,
      "loss": 0.579,
      "step": 6799
    },
    {
      "epoch": 1.95,
      "grad_norm": 7.569090366363525,
      "learning_rate": 8.237986270022883e-07,
      "loss": 0.3526,
      "step": 6800
    },
    {
      "epoch": 1.95,
      "grad_norm": 9.218209266662598,
      "learning_rate": 8.19508009153318e-07,
      "loss": 0.6879,
      "step": 6801
    },
    {
      "epoch": 1.95,
      "grad_norm": 12.47700023651123,
      "learning_rate": 8.152173913043479e-07,
      "loss": 0.8903,
      "step": 6802
    },
    {
      "epoch": 1.95,
      "grad_norm": 10.173751831054688,
      "learning_rate": 8.109267734553776e-07,
      "loss": 0.6248,
      "step": 6803
    },
    {
      "epoch": 1.95,
      "grad_norm": 11.127490997314453,
      "learning_rate": 8.066361556064073e-07,
      "loss": 0.84,
      "step": 6804
    },
    {
      "epoch": 1.95,
      "grad_norm": 9.903820991516113,
      "learning_rate": 8.023455377574371e-07,
      "loss": 0.5848,
      "step": 6805
    },
    {
      "epoch": 1.95,
      "grad_norm": 11.57451343536377,
      "learning_rate": 7.980549199084668e-07,
      "loss": 0.7,
      "step": 6806
    },
    {
      "epoch": 1.95,
      "grad_norm": 11.393851280212402,
      "learning_rate": 7.937643020594967e-07,
      "loss": 0.8057,
      "step": 6807
    },
    {
      "epoch": 1.95,
      "grad_norm": 8.562248229980469,
      "learning_rate": 7.894736842105263e-07,
      "loss": 0.5488,
      "step": 6808
    },
    {
      "epoch": 1.95,
      "grad_norm": 13.203279495239258,
      "learning_rate": 7.85183066361556e-07,
      "loss": 0.9368,
      "step": 6809
    },
    {
      "epoch": 1.95,
      "grad_norm": 9.013036727905273,
      "learning_rate": 7.808924485125858e-07,
      "loss": 0.6234,
      "step": 6810
    },
    {
      "epoch": 1.95,
      "grad_norm": 10.600117683410645,
      "learning_rate": 7.766018306636156e-07,
      "loss": 0.8523,
      "step": 6811
    },
    {
      "epoch": 1.95,
      "grad_norm": 9.700066566467285,
      "learning_rate": 7.723112128146454e-07,
      "loss": 0.6776,
      "step": 6812
    },
    {
      "epoch": 1.95,
      "grad_norm": 10.107597351074219,
      "learning_rate": 7.680205949656751e-07,
      "loss": 0.6422,
      "step": 6813
    },
    {
      "epoch": 1.95,
      "grad_norm": 9.581838607788086,
      "learning_rate": 7.637299771167048e-07,
      "loss": 0.609,
      "step": 6814
    },
    {
      "epoch": 1.95,
      "grad_norm": 9.373744010925293,
      "learning_rate": 7.594393592677345e-07,
      "loss": 0.5848,
      "step": 6815
    },
    {
      "epoch": 1.95,
      "grad_norm": 8.906309127807617,
      "learning_rate": 7.551487414187643e-07,
      "loss": 0.7027,
      "step": 6816
    },
    {
      "epoch": 1.95,
      "grad_norm": 9.41042709350586,
      "learning_rate": 7.508581235697941e-07,
      "loss": 0.5145,
      "step": 6817
    },
    {
      "epoch": 1.95,
      "grad_norm": 9.03809642791748,
      "learning_rate": 7.465675057208238e-07,
      "loss": 0.5115,
      "step": 6818
    },
    {
      "epoch": 1.95,
      "grad_norm": 10.127779006958008,
      "learning_rate": 7.422768878718535e-07,
      "loss": 0.6582,
      "step": 6819
    },
    {
      "epoch": 1.95,
      "grad_norm": 11.648848533630371,
      "learning_rate": 7.379862700228834e-07,
      "loss": 0.7816,
      "step": 6820
    },
    {
      "epoch": 1.95,
      "grad_norm": 11.141276359558105,
      "learning_rate": 7.336956521739131e-07,
      "loss": 0.6873,
      "step": 6821
    },
    {
      "epoch": 1.95,
      "grad_norm": 12.081658363342285,
      "learning_rate": 7.294050343249428e-07,
      "loss": 0.7483,
      "step": 6822
    },
    {
      "epoch": 1.95,
      "grad_norm": 8.0576810836792,
      "learning_rate": 7.251144164759725e-07,
      "loss": 0.69,
      "step": 6823
    },
    {
      "epoch": 1.95,
      "grad_norm": 12.894620895385742,
      "learning_rate": 7.208237986270022e-07,
      "loss": 0.8952,
      "step": 6824
    },
    {
      "epoch": 1.95,
      "grad_norm": 9.190361976623535,
      "learning_rate": 7.165331807780321e-07,
      "loss": 0.5655,
      "step": 6825
    },
    {
      "epoch": 1.95,
      "grad_norm": 10.179404258728027,
      "learning_rate": 7.122425629290618e-07,
      "loss": 0.6606,
      "step": 6826
    },
    {
      "epoch": 1.95,
      "grad_norm": 9.90598201751709,
      "learning_rate": 7.079519450800915e-07,
      "loss": 0.8001,
      "step": 6827
    },
    {
      "epoch": 1.95,
      "grad_norm": 9.589546203613281,
      "learning_rate": 7.036613272311213e-07,
      "loss": 0.5828,
      "step": 6828
    },
    {
      "epoch": 1.95,
      "grad_norm": 9.023877143859863,
      "learning_rate": 6.993707093821511e-07,
      "loss": 0.5469,
      "step": 6829
    },
    {
      "epoch": 1.95,
      "grad_norm": 10.734456062316895,
      "learning_rate": 6.950800915331809e-07,
      "loss": 0.6977,
      "step": 6830
    },
    {
      "epoch": 1.95,
      "grad_norm": 10.52493667602539,
      "learning_rate": 6.907894736842105e-07,
      "loss": 0.7481,
      "step": 6831
    },
    {
      "epoch": 1.95,
      "grad_norm": 10.465811729431152,
      "learning_rate": 6.864988558352402e-07,
      "loss": 0.5465,
      "step": 6832
    },
    {
      "epoch": 1.95,
      "grad_norm": 9.976741790771484,
      "learning_rate": 6.8220823798627e-07,
      "loss": 0.8391,
      "step": 6833
    },
    {
      "epoch": 1.95,
      "grad_norm": 11.496618270874023,
      "learning_rate": 6.779176201372998e-07,
      "loss": 0.6632,
      "step": 6834
    },
    {
      "epoch": 1.96,
      "grad_norm": 7.427152633666992,
      "learning_rate": 6.736270022883296e-07,
      "loss": 0.4168,
      "step": 6835
    },
    {
      "epoch": 1.96,
      "grad_norm": 9.783970832824707,
      "learning_rate": 6.693363844393593e-07,
      "loss": 0.7624,
      "step": 6836
    },
    {
      "epoch": 1.96,
      "grad_norm": 9.616044044494629,
      "learning_rate": 6.65045766590389e-07,
      "loss": 0.5893,
      "step": 6837
    },
    {
      "epoch": 1.96,
      "grad_norm": 9.780241012573242,
      "learning_rate": 6.607551487414188e-07,
      "loss": 0.4578,
      "step": 6838
    },
    {
      "epoch": 1.96,
      "grad_norm": 9.057365417480469,
      "learning_rate": 6.564645308924485e-07,
      "loss": 0.5626,
      "step": 6839
    },
    {
      "epoch": 1.96,
      "grad_norm": 10.086860656738281,
      "learning_rate": 6.521739130434783e-07,
      "loss": 0.6514,
      "step": 6840
    },
    {
      "epoch": 1.96,
      "grad_norm": 10.18759822845459,
      "learning_rate": 6.47883295194508e-07,
      "loss": 0.5814,
      "step": 6841
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.207710266113281,
      "learning_rate": 6.435926773455378e-07,
      "loss": 0.5791,
      "step": 6842
    },
    {
      "epoch": 1.96,
      "grad_norm": 10.417292594909668,
      "learning_rate": 6.393020594965676e-07,
      "loss": 0.6957,
      "step": 6843
    },
    {
      "epoch": 1.96,
      "grad_norm": 10.373567581176758,
      "learning_rate": 6.350114416475973e-07,
      "loss": 0.6666,
      "step": 6844
    },
    {
      "epoch": 1.96,
      "grad_norm": 9.936010360717773,
      "learning_rate": 6.30720823798627e-07,
      "loss": 0.6347,
      "step": 6845
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.712944984436035,
      "learning_rate": 6.264302059496567e-07,
      "loss": 0.8248,
      "step": 6846
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.301887512207031,
      "learning_rate": 6.221395881006865e-07,
      "loss": 0.7441,
      "step": 6847
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.015470504760742,
      "learning_rate": 6.178489702517163e-07,
      "loss": 0.7018,
      "step": 6848
    },
    {
      "epoch": 1.96,
      "grad_norm": 10.960253715515137,
      "learning_rate": 6.13558352402746e-07,
      "loss": 0.5765,
      "step": 6849
    },
    {
      "epoch": 1.96,
      "grad_norm": 10.221090316772461,
      "learning_rate": 6.092677345537757e-07,
      "loss": 0.638,
      "step": 6850
    },
    {
      "epoch": 1.96,
      "grad_norm": 8.547921180725098,
      "learning_rate": 6.049771167048056e-07,
      "loss": 0.536,
      "step": 6851
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.917266845703125,
      "learning_rate": 6.006864988558353e-07,
      "loss": 0.699,
      "step": 6852
    },
    {
      "epoch": 1.96,
      "grad_norm": 10.198384284973145,
      "learning_rate": 5.96395881006865e-07,
      "loss": 0.6332,
      "step": 6853
    },
    {
      "epoch": 1.96,
      "grad_norm": 8.864692687988281,
      "learning_rate": 5.921052631578947e-07,
      "loss": 0.5656,
      "step": 6854
    },
    {
      "epoch": 1.96,
      "grad_norm": 12.42785358428955,
      "learning_rate": 5.878146453089244e-07,
      "loss": 0.6115,
      "step": 6855
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.457018852233887,
      "learning_rate": 5.835240274599543e-07,
      "loss": 0.4691,
      "step": 6856
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.069910049438477,
      "learning_rate": 5.79233409610984e-07,
      "loss": 0.5925,
      "step": 6857
    },
    {
      "epoch": 1.96,
      "grad_norm": 10.123159408569336,
      "learning_rate": 5.749427917620138e-07,
      "loss": 0.6451,
      "step": 6858
    },
    {
      "epoch": 1.96,
      "grad_norm": 9.3661527633667,
      "learning_rate": 5.706521739130435e-07,
      "loss": 0.7237,
      "step": 6859
    },
    {
      "epoch": 1.96,
      "grad_norm": 8.912443161010742,
      "learning_rate": 5.663615560640733e-07,
      "loss": 0.6893,
      "step": 6860
    },
    {
      "epoch": 1.96,
      "grad_norm": 8.892996788024902,
      "learning_rate": 5.62070938215103e-07,
      "loss": 0.6673,
      "step": 6861
    },
    {
      "epoch": 1.96,
      "grad_norm": 8.603639602661133,
      "learning_rate": 5.577803203661327e-07,
      "loss": 0.3919,
      "step": 6862
    },
    {
      "epoch": 1.96,
      "grad_norm": 9.034506797790527,
      "learning_rate": 5.534897025171625e-07,
      "loss": 0.7098,
      "step": 6863
    },
    {
      "epoch": 1.96,
      "grad_norm": 9.981705665588379,
      "learning_rate": 5.491990846681922e-07,
      "loss": 0.7897,
      "step": 6864
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.018440246582031,
      "learning_rate": 5.44908466819222e-07,
      "loss": 0.8389,
      "step": 6865
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.106205940246582,
      "learning_rate": 5.406178489702518e-07,
      "loss": 0.8287,
      "step": 6866
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.349292755126953,
      "learning_rate": 5.363272311212815e-07,
      "loss": 0.7209,
      "step": 6867
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.872751235961914,
      "learning_rate": 5.320366132723111e-07,
      "loss": 0.6976,
      "step": 6868
    },
    {
      "epoch": 1.96,
      "grad_norm": 12.316161155700684,
      "learning_rate": 5.27745995423341e-07,
      "loss": 0.5499,
      "step": 6869
    },
    {
      "epoch": 1.97,
      "grad_norm": 8.931777000427246,
      "learning_rate": 5.234553775743707e-07,
      "loss": 0.6789,
      "step": 6870
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.08340835571289,
      "learning_rate": 5.191647597254005e-07,
      "loss": 0.7906,
      "step": 6871
    },
    {
      "epoch": 1.97,
      "grad_norm": 12.119122505187988,
      "learning_rate": 5.148741418764302e-07,
      "loss": 0.8094,
      "step": 6872
    },
    {
      "epoch": 1.97,
      "grad_norm": 14.020956039428711,
      "learning_rate": 5.105835240274599e-07,
      "loss": 0.746,
      "step": 6873
    },
    {
      "epoch": 1.97,
      "grad_norm": 8.066704750061035,
      "learning_rate": 5.062929061784898e-07,
      "loss": 0.5912,
      "step": 6874
    },
    {
      "epoch": 1.97,
      "grad_norm": 8.146242141723633,
      "learning_rate": 5.020022883295195e-07,
      "loss": 0.399,
      "step": 6875
    },
    {
      "epoch": 1.97,
      "grad_norm": 9.605037689208984,
      "learning_rate": 4.977116704805492e-07,
      "loss": 0.5589,
      "step": 6876
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.58935260772705,
      "learning_rate": 4.934210526315789e-07,
      "loss": 0.6743,
      "step": 6877
    },
    {
      "epoch": 1.97,
      "grad_norm": 12.243191719055176,
      "learning_rate": 4.891304347826087e-07,
      "loss": 0.6987,
      "step": 6878
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.33033561706543,
      "learning_rate": 4.848398169336385e-07,
      "loss": 0.5374,
      "step": 6879
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.57116413116455,
      "learning_rate": 4.805491990846682e-07,
      "loss": 0.8575,
      "step": 6880
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.175572395324707,
      "learning_rate": 4.76258581235698e-07,
      "loss": 0.626,
      "step": 6881
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.430510520935059,
      "learning_rate": 4.7196796338672775e-07,
      "loss": 0.6818,
      "step": 6882
    },
    {
      "epoch": 1.97,
      "grad_norm": 11.097426414489746,
      "learning_rate": 4.6767734553775747e-07,
      "loss": 0.8006,
      "step": 6883
    },
    {
      "epoch": 1.97,
      "grad_norm": 9.713412284851074,
      "learning_rate": 4.633867276887872e-07,
      "loss": 0.6842,
      "step": 6884
    },
    {
      "epoch": 1.97,
      "grad_norm": 14.023730278015137,
      "learning_rate": 4.590961098398169e-07,
      "loss": 0.7831,
      "step": 6885
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.488276481628418,
      "learning_rate": 4.548054919908467e-07,
      "loss": 0.5933,
      "step": 6886
    },
    {
      "epoch": 1.97,
      "grad_norm": 8.53012466430664,
      "learning_rate": 4.5051487414187646e-07,
      "loss": 0.4922,
      "step": 6887
    },
    {
      "epoch": 1.97,
      "grad_norm": 11.886279106140137,
      "learning_rate": 4.4622425629290623e-07,
      "loss": 0.5628,
      "step": 6888
    },
    {
      "epoch": 1.97,
      "grad_norm": 9.568771362304688,
      "learning_rate": 4.419336384439359e-07,
      "loss": 0.5714,
      "step": 6889
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.269181251525879,
      "learning_rate": 4.3764302059496567e-07,
      "loss": 0.586,
      "step": 6890
    },
    {
      "epoch": 1.97,
      "grad_norm": 11.323017120361328,
      "learning_rate": 4.3335240274599545e-07,
      "loss": 0.7212,
      "step": 6891
    },
    {
      "epoch": 1.97,
      "grad_norm": 11.437891006469727,
      "learning_rate": 4.290617848970252e-07,
      "loss": 0.6863,
      "step": 6892
    },
    {
      "epoch": 1.97,
      "grad_norm": 8.55978012084961,
      "learning_rate": 4.2477116704805494e-07,
      "loss": 0.5286,
      "step": 6893
    },
    {
      "epoch": 1.97,
      "grad_norm": 8.444185256958008,
      "learning_rate": 4.2048054919908466e-07,
      "loss": 0.4725,
      "step": 6894
    },
    {
      "epoch": 1.97,
      "grad_norm": 11.7064847946167,
      "learning_rate": 4.1618993135011444e-07,
      "loss": 0.6294,
      "step": 6895
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.098320007324219,
      "learning_rate": 4.1189931350114416e-07,
      "loss": 0.8641,
      "step": 6896
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.265768051147461,
      "learning_rate": 4.0760869565217393e-07,
      "loss": 0.5418,
      "step": 6897
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.396321296691895,
      "learning_rate": 4.0331807780320365e-07,
      "loss": 0.7486,
      "step": 6898
    },
    {
      "epoch": 1.97,
      "grad_norm": 9.897319793701172,
      "learning_rate": 3.990274599542334e-07,
      "loss": 0.6535,
      "step": 6899
    },
    {
      "epoch": 1.97,
      "grad_norm": 9.139906883239746,
      "learning_rate": 3.9473684210526315e-07,
      "loss": 0.5608,
      "step": 6900
    },
    {
      "epoch": 1.97,
      "grad_norm": 11.586169242858887,
      "learning_rate": 3.904462242562929e-07,
      "loss": 0.8048,
      "step": 6901
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.12833309173584,
      "learning_rate": 3.861556064073227e-07,
      "loss": 0.6463,
      "step": 6902
    },
    {
      "epoch": 1.97,
      "grad_norm": 11.549627304077148,
      "learning_rate": 3.818649885583524e-07,
      "loss": 0.5828,
      "step": 6903
    },
    {
      "epoch": 1.97,
      "grad_norm": 9.387853622436523,
      "learning_rate": 3.7757437070938213e-07,
      "loss": 0.5356,
      "step": 6904
    },
    {
      "epoch": 1.98,
      "grad_norm": 10.681439399719238,
      "learning_rate": 3.732837528604119e-07,
      "loss": 0.5921,
      "step": 6905
    },
    {
      "epoch": 1.98,
      "grad_norm": 10.00284481048584,
      "learning_rate": 3.689931350114417e-07,
      "loss": 0.6901,
      "step": 6906
    },
    {
      "epoch": 1.98,
      "grad_norm": 10.689186096191406,
      "learning_rate": 3.647025171624714e-07,
      "loss": 0.6819,
      "step": 6907
    },
    {
      "epoch": 1.98,
      "grad_norm": 7.789943695068359,
      "learning_rate": 3.604118993135011e-07,
      "loss": 0.5717,
      "step": 6908
    },
    {
      "epoch": 1.98,
      "grad_norm": 12.032934188842773,
      "learning_rate": 3.561212814645309e-07,
      "loss": 0.864,
      "step": 6909
    },
    {
      "epoch": 1.98,
      "grad_norm": 9.302270889282227,
      "learning_rate": 3.5183066361556067e-07,
      "loss": 0.5825,
      "step": 6910
    },
    {
      "epoch": 1.98,
      "grad_norm": 5.9612908363342285,
      "learning_rate": 3.4754004576659044e-07,
      "loss": 0.3212,
      "step": 6911
    },
    {
      "epoch": 1.98,
      "grad_norm": 11.921589851379395,
      "learning_rate": 3.432494279176201e-07,
      "loss": 0.7346,
      "step": 6912
    },
    {
      "epoch": 1.98,
      "grad_norm": 10.079961776733398,
      "learning_rate": 3.389588100686499e-07,
      "loss": 0.782,
      "step": 6913
    },
    {
      "epoch": 1.98,
      "grad_norm": 10.618403434753418,
      "learning_rate": 3.3466819221967966e-07,
      "loss": 0.7526,
      "step": 6914
    },
    {
      "epoch": 1.98,
      "grad_norm": 11.24497127532959,
      "learning_rate": 3.303775743707094e-07,
      "loss": 0.7186,
      "step": 6915
    },
    {
      "epoch": 1.98,
      "grad_norm": 9.965060234069824,
      "learning_rate": 3.2608695652173915e-07,
      "loss": 0.6016,
      "step": 6916
    },
    {
      "epoch": 1.98,
      "grad_norm": 10.062822341918945,
      "learning_rate": 3.217963386727689e-07,
      "loss": 0.5654,
      "step": 6917
    },
    {
      "epoch": 1.98,
      "grad_norm": 11.358135223388672,
      "learning_rate": 3.1750572082379865e-07,
      "loss": 0.7088,
      "step": 6918
    },
    {
      "epoch": 1.98,
      "grad_norm": 9.111773490905762,
      "learning_rate": 3.1321510297482837e-07,
      "loss": 0.7644,
      "step": 6919
    },
    {
      "epoch": 1.98,
      "grad_norm": 9.192373275756836,
      "learning_rate": 3.0892448512585814e-07,
      "loss": 0.6394,
      "step": 6920
    },
    {
      "epoch": 1.98,
      "grad_norm": 12.135351181030273,
      "learning_rate": 3.0463386727688786e-07,
      "loss": 0.8185,
      "step": 6921
    },
    {
      "epoch": 1.98,
      "grad_norm": 11.259346008300781,
      "learning_rate": 3.0034324942791764e-07,
      "loss": 0.6089,
      "step": 6922
    },
    {
      "epoch": 1.98,
      "grad_norm": 7.142961502075195,
      "learning_rate": 2.9605263157894736e-07,
      "loss": 0.4448,
      "step": 6923
    },
    {
      "epoch": 1.98,
      "grad_norm": 8.596514701843262,
      "learning_rate": 2.9176201372997713e-07,
      "loss": 0.603,
      "step": 6924
    },
    {
      "epoch": 1.98,
      "grad_norm": 10.275640487670898,
      "learning_rate": 2.874713958810069e-07,
      "loss": 0.9107,
      "step": 6925
    },
    {
      "epoch": 1.98,
      "grad_norm": 12.552618980407715,
      "learning_rate": 2.8318077803203663e-07,
      "loss": 0.7594,
      "step": 6926
    },
    {
      "epoch": 1.98,
      "grad_norm": 7.591946601867676,
      "learning_rate": 2.7889016018306635e-07,
      "loss": 0.5472,
      "step": 6927
    },
    {
      "epoch": 1.98,
      "grad_norm": 9.330859184265137,
      "learning_rate": 2.745995423340961e-07,
      "loss": 0.6062,
      "step": 6928
    },
    {
      "epoch": 1.98,
      "grad_norm": 11.907567024230957,
      "learning_rate": 2.703089244851259e-07,
      "loss": 0.6068,
      "step": 6929
    },
    {
      "epoch": 1.98,
      "grad_norm": 10.675895690917969,
      "learning_rate": 2.6601830663615556e-07,
      "loss": 0.683,
      "step": 6930
    },
    {
      "epoch": 1.98,
      "grad_norm": 10.48024845123291,
      "learning_rate": 2.6172768878718534e-07,
      "loss": 0.4274,
      "step": 6931
    },
    {
      "epoch": 1.98,
      "grad_norm": 8.749180793762207,
      "learning_rate": 2.574370709382151e-07,
      "loss": 0.622,
      "step": 6932
    },
    {
      "epoch": 1.98,
      "grad_norm": 10.966500282287598,
      "learning_rate": 2.531464530892449e-07,
      "loss": 0.6093,
      "step": 6933
    },
    {
      "epoch": 1.98,
      "grad_norm": 8.568903923034668,
      "learning_rate": 2.488558352402746e-07,
      "loss": 0.6865,
      "step": 6934
    },
    {
      "epoch": 1.98,
      "grad_norm": 12.321759223937988,
      "learning_rate": 2.445652173913043e-07,
      "loss": 0.86,
      "step": 6935
    },
    {
      "epoch": 1.98,
      "grad_norm": 11.141681671142578,
      "learning_rate": 2.402745995423341e-07,
      "loss": 0.7193,
      "step": 6936
    },
    {
      "epoch": 1.98,
      "grad_norm": 8.963074684143066,
      "learning_rate": 2.3598398169336387e-07,
      "loss": 0.4704,
      "step": 6937
    },
    {
      "epoch": 1.98,
      "grad_norm": 10.483250617980957,
      "learning_rate": 2.316933638443936e-07,
      "loss": 0.6919,
      "step": 6938
    },
    {
      "epoch": 1.98,
      "grad_norm": 11.217721939086914,
      "learning_rate": 2.2740274599542334e-07,
      "loss": 0.6015,
      "step": 6939
    },
    {
      "epoch": 1.99,
      "grad_norm": 9.580672264099121,
      "learning_rate": 2.2311212814645312e-07,
      "loss": 0.5452,
      "step": 6940
    },
    {
      "epoch": 1.99,
      "grad_norm": 8.676939964294434,
      "learning_rate": 2.1882151029748284e-07,
      "loss": 0.4453,
      "step": 6941
    },
    {
      "epoch": 1.99,
      "grad_norm": 9.502946853637695,
      "learning_rate": 2.145308924485126e-07,
      "loss": 0.5431,
      "step": 6942
    },
    {
      "epoch": 1.99,
      "grad_norm": 7.5875773429870605,
      "learning_rate": 2.1024027459954233e-07,
      "loss": 0.5096,
      "step": 6943
    },
    {
      "epoch": 1.99,
      "grad_norm": 9.894153594970703,
      "learning_rate": 2.0594965675057208e-07,
      "loss": 0.6333,
      "step": 6944
    },
    {
      "epoch": 1.99,
      "grad_norm": 11.48495101928711,
      "learning_rate": 2.0165903890160183e-07,
      "loss": 0.6755,
      "step": 6945
    },
    {
      "epoch": 1.99,
      "grad_norm": 9.783747673034668,
      "learning_rate": 1.9736842105263157e-07,
      "loss": 0.6055,
      "step": 6946
    },
    {
      "epoch": 1.99,
      "grad_norm": 10.02064323425293,
      "learning_rate": 1.9307780320366135e-07,
      "loss": 0.4871,
      "step": 6947
    },
    {
      "epoch": 1.99,
      "grad_norm": 12.362141609191895,
      "learning_rate": 1.8878718535469107e-07,
      "loss": 0.6577,
      "step": 6948
    },
    {
      "epoch": 1.99,
      "grad_norm": 13.172623634338379,
      "learning_rate": 1.8449656750572084e-07,
      "loss": 0.6607,
      "step": 6949
    },
    {
      "epoch": 1.99,
      "grad_norm": 8.153654098510742,
      "learning_rate": 1.8020594965675056e-07,
      "loss": 0.5096,
      "step": 6950
    },
    {
      "epoch": 1.99,
      "grad_norm": 9.872180938720703,
      "learning_rate": 1.7591533180778034e-07,
      "loss": 0.5401,
      "step": 6951
    },
    {
      "epoch": 1.99,
      "grad_norm": 8.815258026123047,
      "learning_rate": 1.7162471395881006e-07,
      "loss": 0.5247,
      "step": 6952
    },
    {
      "epoch": 1.99,
      "grad_norm": 7.608972549438477,
      "learning_rate": 1.6733409610983983e-07,
      "loss": 0.4153,
      "step": 6953
    },
    {
      "epoch": 1.99,
      "grad_norm": 10.867547035217285,
      "learning_rate": 1.6304347826086958e-07,
      "loss": 0.7146,
      "step": 6954
    },
    {
      "epoch": 1.99,
      "grad_norm": 8.751923561096191,
      "learning_rate": 1.5875286041189932e-07,
      "loss": 0.4785,
      "step": 6955
    },
    {
      "epoch": 1.99,
      "grad_norm": 12.871894836425781,
      "learning_rate": 1.5446224256292907e-07,
      "loss": 0.7724,
      "step": 6956
    },
    {
      "epoch": 1.99,
      "grad_norm": 12.138284683227539,
      "learning_rate": 1.5017162471395882e-07,
      "loss": 0.687,
      "step": 6957
    },
    {
      "epoch": 1.99,
      "grad_norm": 12.697810173034668,
      "learning_rate": 1.4588100686498857e-07,
      "loss": 0.7714,
      "step": 6958
    },
    {
      "epoch": 1.99,
      "grad_norm": 10.738144874572754,
      "learning_rate": 1.4159038901601831e-07,
      "loss": 0.7474,
      "step": 6959
    },
    {
      "epoch": 1.99,
      "grad_norm": 12.630233764648438,
      "learning_rate": 1.3729977116704806e-07,
      "loss": 0.7768,
      "step": 6960
    },
    {
      "epoch": 1.99,
      "grad_norm": 10.851924896240234,
      "learning_rate": 1.3300915331807778e-07,
      "loss": 0.6323,
      "step": 6961
    },
    {
      "epoch": 1.99,
      "grad_norm": 11.448753356933594,
      "learning_rate": 1.2871853546910756e-07,
      "loss": 0.5869,
      "step": 6962
    },
    {
      "epoch": 1.99,
      "grad_norm": 10.769323348999023,
      "learning_rate": 1.244279176201373e-07,
      "loss": 0.6573,
      "step": 6963
    },
    {
      "epoch": 1.99,
      "grad_norm": 12.800044059753418,
      "learning_rate": 1.2013729977116705e-07,
      "loss": 0.6271,
      "step": 6964
    },
    {
      "epoch": 1.99,
      "grad_norm": 11.841652870178223,
      "learning_rate": 1.158466819221968e-07,
      "loss": 0.7032,
      "step": 6965
    },
    {
      "epoch": 1.99,
      "grad_norm": 11.031814575195312,
      "learning_rate": 1.1155606407322656e-07,
      "loss": 0.6932,
      "step": 6966
    },
    {
      "epoch": 1.99,
      "grad_norm": 10.714957237243652,
      "learning_rate": 1.072654462242563e-07,
      "loss": 0.6247,
      "step": 6967
    },
    {
      "epoch": 1.99,
      "grad_norm": 12.333422660827637,
      "learning_rate": 1.0297482837528604e-07,
      "loss": 0.8033,
      "step": 6968
    },
    {
      "epoch": 1.99,
      "grad_norm": 10.396112442016602,
      "learning_rate": 9.868421052631579e-08,
      "loss": 0.6173,
      "step": 6969
    },
    {
      "epoch": 1.99,
      "grad_norm": 10.016500473022461,
      "learning_rate": 9.439359267734553e-08,
      "loss": 0.5538,
      "step": 6970
    },
    {
      "epoch": 1.99,
      "grad_norm": 10.614705085754395,
      "learning_rate": 9.010297482837528e-08,
      "loss": 0.717,
      "step": 6971
    },
    {
      "epoch": 1.99,
      "grad_norm": 7.811626434326172,
      "learning_rate": 8.581235697940503e-08,
      "loss": 0.5426,
      "step": 6972
    },
    {
      "epoch": 1.99,
      "grad_norm": 9.46379566192627,
      "learning_rate": 8.152173913043479e-08,
      "loss": 0.57,
      "step": 6973
    },
    {
      "epoch": 1.99,
      "grad_norm": 10.381284713745117,
      "learning_rate": 7.723112128146454e-08,
      "loss": 0.6539,
      "step": 6974
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.76805591583252,
      "learning_rate": 7.294050343249428e-08,
      "loss": 0.7282,
      "step": 6975
    },
    {
      "epoch": 2.0,
      "grad_norm": 9.52723217010498,
      "learning_rate": 6.864988558352403e-08,
      "loss": 0.6539,
      "step": 6976
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.24339485168457,
      "learning_rate": 6.435926773455378e-08,
      "loss": 0.6859,
      "step": 6977
    },
    {
      "epoch": 2.0,
      "grad_norm": 11.314688682556152,
      "learning_rate": 6.006864988558353e-08,
      "loss": 0.6341,
      "step": 6978
    },
    {
      "epoch": 2.0,
      "grad_norm": 11.507288932800293,
      "learning_rate": 5.577803203661328e-08,
      "loss": 0.8784,
      "step": 6979
    },
    {
      "epoch": 2.0,
      "grad_norm": 9.23391342163086,
      "learning_rate": 5.148741418764302e-08,
      "loss": 0.6483,
      "step": 6980
    },
    {
      "epoch": 2.0,
      "grad_norm": 13.069464683532715,
      "learning_rate": 4.719679633867277e-08,
      "loss": 0.7351,
      "step": 6981
    },
    {
      "epoch": 2.0,
      "grad_norm": 12.40121078491211,
      "learning_rate": 4.2906178489702514e-08,
      "loss": 0.7474,
      "step": 6982
    },
    {
      "epoch": 2.0,
      "grad_norm": 11.74376392364502,
      "learning_rate": 3.861556064073227e-08,
      "loss": 0.7727,
      "step": 6983
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.733787536621094,
      "learning_rate": 3.4324942791762015e-08,
      "loss": 0.6788,
      "step": 6984
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.416940689086914,
      "learning_rate": 3.003432494279176e-08,
      "loss": 0.5102,
      "step": 6985
    },
    {
      "epoch": 2.0,
      "grad_norm": 9.378046989440918,
      "learning_rate": 2.574370709382151e-08,
      "loss": 0.5839,
      "step": 6986
    },
    {
      "epoch": 2.0,
      "grad_norm": 13.186521530151367,
      "learning_rate": 2.1453089244851257e-08,
      "loss": 0.5744,
      "step": 6987
    },
    {
      "epoch": 2.0,
      "grad_norm": 12.878933906555176,
      "learning_rate": 1.7162471395881008e-08,
      "loss": 0.8448,
      "step": 6988
    },
    {
      "epoch": 2.0,
      "grad_norm": 9.82333755493164,
      "learning_rate": 1.2871853546910755e-08,
      "loss": 0.451,
      "step": 6989
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.506118774414062,
      "learning_rate": 8.581235697940504e-09,
      "loss": 0.6907,
      "step": 6990
    },
    {
      "epoch": 2.0,
      "grad_norm": 9.47684383392334,
      "learning_rate": 4.290617848970252e-09,
      "loss": 0.6083,
      "step": 6991
    },
    {
      "epoch": 2.0,
      "grad_norm": 14.73945140838623,
      "learning_rate": 0.0,
      "loss": 0.8721,
      "step": 6992
    },
    {
      "epoch": 2.0,
      "step": 6992,
      "total_flos": 1.315371792860375e+17,
      "train_loss": 1.0098489519694578,
      "train_runtime": 2148.6486,
      "train_samples_per_second": 312.383,
      "train_steps_per_second": 3.254
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 6992,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 1.315371792860375e+17,
  "train_batch_size": 96,
  "trial_name": null,
  "trial_params": null
}