{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9995664614584236,
  "eval_steps": 500,
  "global_step": 1441,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0006936616665221538,
      "grad_norm": 43.466719125161504,
      "learning_rate": 0.0,
      "loss": 2.7095,
      "step": 1
    },
    {
      "epoch": 0.0013873233330443076,
      "grad_norm": 6.96466711942635,
      "learning_rate": 2.5595802480981545e-06,
      "loss": 2.0266,
      "step": 2
    },
    {
      "epoch": 0.0020809849995664614,
      "grad_norm": 5.728387691489433,
      "learning_rate": 4.056838710822129e-06,
      "loss": 2.0082,
      "step": 3
    },
    {
      "epoch": 0.0027746466660886152,
      "grad_norm": 4.686811256036109,
      "learning_rate": 5.119160496196309e-06,
      "loss": 1.9589,
      "step": 4
    },
    {
      "epoch": 0.003468308332610769,
      "grad_norm": 3.3529589095838874,
      "learning_rate": 5.943161289177871e-06,
      "loss": 1.9014,
      "step": 5
    },
    {
      "epoch": 0.004161969999132923,
      "grad_norm": 3.5493873116106522,
      "learning_rate": 6.616418958920285e-06,
      "loss": 1.8478,
      "step": 6
    },
    {
      "epoch": 0.004855631665655077,
      "grad_norm": 3.2248318636630753,
      "learning_rate": 7.185650207899778e-06,
      "loss": 1.8631,
      "step": 7
    },
    {
      "epoch": 0.0055492933321772304,
      "grad_norm": 2.5559675536274558,
      "learning_rate": 7.678740744294464e-06,
      "loss": 1.8771,
      "step": 8
    },
    {
      "epoch": 0.006242954998699385,
      "grad_norm": 2.852602324925445,
      "learning_rate": 8.113677421644258e-06,
      "loss": 1.8517,
      "step": 9
    },
    {
      "epoch": 0.006936616665221538,
      "grad_norm": 2.8475514523205687,
      "learning_rate": 8.502741537276027e-06,
      "loss": 1.8444,
      "step": 10
    },
    {
      "epoch": 0.007630278331743692,
      "grad_norm": 2.4445595331823116,
      "learning_rate": 8.854692840710254e-06,
      "loss": 1.8173,
      "step": 11
    },
    {
      "epoch": 0.008323939998265846,
      "grad_norm": 2.6236339242092606,
      "learning_rate": 9.175999207018439e-06,
      "loss": 1.797,
      "step": 12
    },
    {
      "epoch": 0.009017601664787999,
      "grad_norm": 2.7480536817929178,
      "learning_rate": 9.471572411831843e-06,
      "loss": 1.7525,
      "step": 13
    },
    {
      "epoch": 0.009711263331310154,
      "grad_norm": 2.8195069251274267,
      "learning_rate": 9.745230455997932e-06,
      "loss": 1.7421,
      "step": 14
    },
    {
      "epoch": 0.010404924997832308,
      "grad_norm": 2.574781912417627,
      "learning_rate": 9.999999999999999e-06,
      "loss": 1.7563,
      "step": 15
    },
    {
      "epoch": 0.011098586664354461,
      "grad_norm": 3.018787395539229,
      "learning_rate": 1e-05,
      "loss": 1.7235,
      "step": 16
    },
    {
      "epoch": 0.011792248330876614,
      "grad_norm": 2.7321515111278085,
      "learning_rate": 1e-05,
      "loss": 1.7372,
      "step": 17
    },
    {
      "epoch": 0.01248590999739877,
      "grad_norm": 2.535876566743008,
      "learning_rate": 1e-05,
      "loss": 1.7513,
      "step": 18
    },
    {
      "epoch": 0.013179571663920923,
      "grad_norm": 2.503169241059693,
      "learning_rate": 1e-05,
      "loss": 1.7598,
      "step": 19
    },
    {
      "epoch": 0.013873233330443076,
      "grad_norm": 2.5731073684349752,
      "learning_rate": 1e-05,
      "loss": 1.7947,
      "step": 20
    },
    {
      "epoch": 0.01456689499696523,
      "grad_norm": 2.5424918030748245,
      "learning_rate": 1e-05,
      "loss": 1.7304,
      "step": 21
    },
    {
      "epoch": 0.015260556663487385,
      "grad_norm": 2.280575429076331,
      "learning_rate": 1e-05,
      "loss": 1.6995,
      "step": 22
    },
    {
      "epoch": 0.015954218330009536,
      "grad_norm": 2.493960760059663,
      "learning_rate": 1e-05,
      "loss": 1.6744,
      "step": 23
    },
    {
      "epoch": 0.01664787999653169,
      "grad_norm": 2.238029250091602,
      "learning_rate": 1e-05,
      "loss": 1.7026,
      "step": 24
    },
    {
      "epoch": 0.017341541663053846,
      "grad_norm": 2.2414764681865362,
      "learning_rate": 1e-05,
      "loss": 1.6437,
      "step": 25
    },
    {
      "epoch": 0.018035203329575998,
      "grad_norm": 2.3762425127817135,
      "learning_rate": 1e-05,
      "loss": 1.7258,
      "step": 26
    },
    {
      "epoch": 0.018728864996098153,
      "grad_norm": 2.471220653637295,
      "learning_rate": 1e-05,
      "loss": 1.7413,
      "step": 27
    },
    {
      "epoch": 0.01942252666262031,
      "grad_norm": 2.236115513685021,
      "learning_rate": 1e-05,
      "loss": 1.7788,
      "step": 28
    },
    {
      "epoch": 0.02011618832914246,
      "grad_norm": 2.1158621654066128,
      "learning_rate": 1e-05,
      "loss": 1.6955,
      "step": 29
    },
    {
      "epoch": 0.020809849995664615,
      "grad_norm": 2.438210038803027,
      "learning_rate": 1e-05,
      "loss": 1.7435,
      "step": 30
    },
    {
      "epoch": 0.021503511662186767,
      "grad_norm": 2.2049092103165484,
      "learning_rate": 1e-05,
      "loss": 1.7103,
      "step": 31
    },
    {
      "epoch": 0.022197173328708922,
      "grad_norm": 1.992834120231878,
      "learning_rate": 1e-05,
      "loss": 1.6941,
      "step": 32
    },
    {
      "epoch": 0.022890834995231077,
      "grad_norm": 2.3021272374064767,
      "learning_rate": 1e-05,
      "loss": 1.667,
      "step": 33
    },
    {
      "epoch": 0.02358449666175323,
      "grad_norm": 2.1119253371859426,
      "learning_rate": 1e-05,
      "loss": 1.6849,
      "step": 34
    },
    {
      "epoch": 0.024278158328275384,
      "grad_norm": 2.3243314568193445,
      "learning_rate": 1e-05,
      "loss": 1.6466,
      "step": 35
    },
    {
      "epoch": 0.02497181999479754,
      "grad_norm": 2.2531985009302176,
      "learning_rate": 1e-05,
      "loss": 1.6459,
      "step": 36
    },
    {
      "epoch": 0.02566548166131969,
      "grad_norm": 2.1844804086893777,
      "learning_rate": 1e-05,
      "loss": 1.6794,
      "step": 37
    },
    {
      "epoch": 0.026359143327841845,
      "grad_norm": 2.2530966998426245,
      "learning_rate": 1e-05,
      "loss": 1.6601,
      "step": 38
    },
    {
      "epoch": 0.027052804994364,
      "grad_norm": 2.1421735569008686,
      "learning_rate": 1e-05,
      "loss": 1.6335,
      "step": 39
    },
    {
      "epoch": 0.027746466660886152,
      "grad_norm": 2.181725062044003,
      "learning_rate": 1e-05,
      "loss": 1.6319,
      "step": 40
    },
    {
      "epoch": 0.028440128327408307,
      "grad_norm": 2.193692520161033,
      "learning_rate": 1e-05,
      "loss": 1.6363,
      "step": 41
    },
    {
      "epoch": 0.02913378999393046,
      "grad_norm": 2.1086393378123818,
      "learning_rate": 1e-05,
      "loss": 1.6735,
      "step": 42
    },
    {
      "epoch": 0.029827451660452614,
      "grad_norm": 2.0131860874790175,
      "learning_rate": 1e-05,
      "loss": 1.5967,
      "step": 43
    },
    {
      "epoch": 0.03052111332697477,
      "grad_norm": 2.1758701969638,
      "learning_rate": 1e-05,
      "loss": 1.667,
      "step": 44
    },
    {
      "epoch": 0.03121477499349692,
      "grad_norm": 2.10881747915129,
      "learning_rate": 1e-05,
      "loss": 1.6002,
      "step": 45
    },
    {
      "epoch": 0.03190843666001907,
      "grad_norm": 2.1929235643249707,
      "learning_rate": 1e-05,
      "loss": 1.6109,
      "step": 46
    },
    {
      "epoch": 0.03260209832654123,
      "grad_norm": 2.2382019559100503,
      "learning_rate": 1e-05,
      "loss": 1.6309,
      "step": 47
    },
    {
      "epoch": 0.03329575999306338,
      "grad_norm": 2.204542258967164,
      "learning_rate": 1e-05,
      "loss": 1.6123,
      "step": 48
    },
    {
      "epoch": 0.03398942165958554,
      "grad_norm": 2.1983829138121194,
      "learning_rate": 1e-05,
      "loss": 1.6371,
      "step": 49
    },
    {
      "epoch": 0.03468308332610769,
      "grad_norm": 2.010616851796152,
      "learning_rate": 1e-05,
      "loss": 1.648,
      "step": 50
    },
    {
      "epoch": 0.03537674499262985,
      "grad_norm": 2.109337350234141,
      "learning_rate": 1e-05,
      "loss": 1.5939,
      "step": 51
    },
    {
      "epoch": 0.036070406659151996,
      "grad_norm": 2.5304099594684657,
      "learning_rate": 1e-05,
      "loss": 1.5806,
      "step": 52
    },
    {
      "epoch": 0.03676406832567415,
      "grad_norm": 2.159949455039585,
      "learning_rate": 1e-05,
      "loss": 1.6898,
      "step": 53
    },
    {
      "epoch": 0.037457729992196306,
      "grad_norm": 2.3028114991190227,
      "learning_rate": 1e-05,
      "loss": 1.6102,
      "step": 54
    },
    {
      "epoch": 0.03815139165871846,
      "grad_norm": 2.137637210948916,
      "learning_rate": 1e-05,
      "loss": 1.5976,
      "step": 55
    },
    {
      "epoch": 0.03884505332524062,
      "grad_norm": 2.283279122337232,
      "learning_rate": 1e-05,
      "loss": 1.6182,
      "step": 56
    },
    {
      "epoch": 0.039538714991762765,
      "grad_norm": 2.352555898736503,
      "learning_rate": 1e-05,
      "loss": 1.5764,
      "step": 57
    },
    {
      "epoch": 0.04023237665828492,
      "grad_norm": 2.2640721232017578,
      "learning_rate": 1e-05,
      "loss": 1.606,
      "step": 58
    },
    {
      "epoch": 0.040926038324807075,
      "grad_norm": 2.2046947158269052,
      "learning_rate": 1e-05,
      "loss": 1.6621,
      "step": 59
    },
    {
      "epoch": 0.04161969999132923,
      "grad_norm": 2.2356515653560955,
      "learning_rate": 1e-05,
      "loss": 1.5783,
      "step": 60
    },
    {
      "epoch": 0.042313361657851385,
      "grad_norm": 2.0914853090761656,
      "learning_rate": 1e-05,
      "loss": 1.6386,
      "step": 61
    },
    {
      "epoch": 0.04300702332437353,
      "grad_norm": 2.0219313193549335,
      "learning_rate": 1e-05,
      "loss": 1.5409,
      "step": 62
    },
    {
      "epoch": 0.04370068499089569,
      "grad_norm": 2.0579915512686675,
      "learning_rate": 1e-05,
      "loss": 1.6024,
      "step": 63
    },
    {
      "epoch": 0.044394346657417844,
      "grad_norm": 2.1350362922236563,
      "learning_rate": 1e-05,
      "loss": 1.5979,
      "step": 64
    },
    {
      "epoch": 0.04508800832394,
      "grad_norm": 2.1574547660781493,
      "learning_rate": 1e-05,
      "loss": 1.5968,
      "step": 65
    },
    {
      "epoch": 0.045781669990462154,
      "grad_norm": 2.032927708489895,
      "learning_rate": 1e-05,
      "loss": 1.623,
      "step": 66
    },
    {
      "epoch": 0.04647533165698431,
      "grad_norm": 2.310697906396082,
      "learning_rate": 1e-05,
      "loss": 1.58,
      "step": 67
    },
    {
      "epoch": 0.04716899332350646,
      "grad_norm": 2.082166726223734,
      "learning_rate": 1e-05,
      "loss": 1.5855,
      "step": 68
    },
    {
      "epoch": 0.04786265499002861,
      "grad_norm": 2.1805698151273205,
      "learning_rate": 1e-05,
      "loss": 1.5406,
      "step": 69
    },
    {
      "epoch": 0.04855631665655077,
      "grad_norm": 2.1517290677451366,
      "learning_rate": 1e-05,
      "loss": 1.5794,
      "step": 70
    },
    {
      "epoch": 0.04924997832307292,
      "grad_norm": 2.2324944455303286,
      "learning_rate": 1e-05,
      "loss": 1.5897,
      "step": 71
    },
    {
      "epoch": 0.04994363998959508,
      "grad_norm": 2.117363841822313,
      "learning_rate": 1e-05,
      "loss": 1.6344,
      "step": 72
    },
    {
      "epoch": 0.050637301656117226,
      "grad_norm": 2.0840589709094055,
      "learning_rate": 1e-05,
      "loss": 1.5593,
      "step": 73
    },
    {
      "epoch": 0.05133096332263938,
      "grad_norm": 2.139635042522081,
      "learning_rate": 1e-05,
      "loss": 1.5604,
      "step": 74
    },
    {
      "epoch": 0.052024624989161536,
      "grad_norm": 1.9773852269088779,
      "learning_rate": 1e-05,
      "loss": 1.5132,
      "step": 75
    },
    {
      "epoch": 0.05271828665568369,
      "grad_norm": 2.2497342140467227,
      "learning_rate": 1e-05,
      "loss": 1.5689,
      "step": 76
    },
    {
      "epoch": 0.053411948322205846,
      "grad_norm": 2.1828363324950515,
      "learning_rate": 1e-05,
      "loss": 1.5775,
      "step": 77
    },
    {
      "epoch": 0.054105609988728,
      "grad_norm": 2.397270079677964,
      "learning_rate": 1e-05,
      "loss": 1.5876,
      "step": 78
    },
    {
      "epoch": 0.05479927165525015,
      "grad_norm": 2.132037167524842,
      "learning_rate": 1e-05,
      "loss": 1.5388,
      "step": 79
    },
    {
      "epoch": 0.055492933321772304,
      "grad_norm": 2.105011522745257,
      "learning_rate": 1e-05,
      "loss": 1.5878,
      "step": 80
    },
    {
      "epoch": 0.05618659498829446,
      "grad_norm": 2.331199900780172,
      "learning_rate": 1e-05,
      "loss": 1.6095,
      "step": 81
    },
    {
      "epoch": 0.056880256654816615,
      "grad_norm": 2.0679915620822915,
      "learning_rate": 1e-05,
      "loss": 1.5489,
      "step": 82
    },
    {
      "epoch": 0.05757391832133877,
      "grad_norm": 2.1615822894325154,
      "learning_rate": 1e-05,
      "loss": 1.6244,
      "step": 83
    },
    {
      "epoch": 0.05826757998786092,
      "grad_norm": 1.979313705736586,
      "learning_rate": 1e-05,
      "loss": 1.5881,
      "step": 84
    },
    {
      "epoch": 0.05896124165438307,
      "grad_norm": 2.1629527862643063,
      "learning_rate": 1e-05,
      "loss": 1.5775,
      "step": 85
    },
    {
      "epoch": 0.05965490332090523,
      "grad_norm": 2.3681176161050956,
      "learning_rate": 1e-05,
      "loss": 1.5714,
      "step": 86
    },
    {
      "epoch": 0.06034856498742738,
      "grad_norm": 1.9456880551179383,
      "learning_rate": 1e-05,
      "loss": 1.5516,
      "step": 87
    },
    {
      "epoch": 0.06104222665394954,
      "grad_norm": 2.1286899349051343,
      "learning_rate": 1e-05,
      "loss": 1.5368,
      "step": 88
    },
    {
      "epoch": 0.06173588832047169,
      "grad_norm": 2.124774583311733,
      "learning_rate": 1e-05,
      "loss": 1.5244,
      "step": 89
    },
    {
      "epoch": 0.06242954998699384,
      "grad_norm": 2.255844207300473,
      "learning_rate": 1e-05,
      "loss": 1.6468,
      "step": 90
    },
    {
      "epoch": 0.063123211653516,
      "grad_norm": 2.253221543014536,
      "learning_rate": 1e-05,
      "loss": 1.5459,
      "step": 91
    },
    {
      "epoch": 0.06381687332003814,
      "grad_norm": 2.161784717901113,
      "learning_rate": 1e-05,
      "loss": 1.5295,
      "step": 92
    },
    {
      "epoch": 0.0645105349865603,
      "grad_norm": 2.100171711959203,
      "learning_rate": 1e-05,
      "loss": 1.5432,
      "step": 93
    },
    {
      "epoch": 0.06520419665308246,
      "grad_norm": 2.0472421305517776,
      "learning_rate": 1e-05,
      "loss": 1.5722,
      "step": 94
    },
    {
      "epoch": 0.06589785831960461,
      "grad_norm": 2.174127269142891,
      "learning_rate": 1e-05,
      "loss": 1.5301,
      "step": 95
    },
    {
      "epoch": 0.06659151998612677,
      "grad_norm": 2.302981491215993,
      "learning_rate": 1e-05,
      "loss": 1.5495,
      "step": 96
    },
    {
      "epoch": 0.06728518165264892,
      "grad_norm": 2.093167418727749,
      "learning_rate": 1e-05,
      "loss": 1.5604,
      "step": 97
    },
    {
      "epoch": 0.06797884331917108,
      "grad_norm": 2.0951541746401263,
      "learning_rate": 1e-05,
      "loss": 1.5769,
      "step": 98
    },
    {
      "epoch": 0.06867250498569323,
      "grad_norm": 1.9576369412503245,
      "learning_rate": 1e-05,
      "loss": 1.6003,
      "step": 99
    },
    {
      "epoch": 0.06936616665221539,
      "grad_norm": 2.3419839237311266,
      "learning_rate": 1e-05,
      "loss": 1.4925,
      "step": 100
    },
    {
      "epoch": 0.07005982831873754,
      "grad_norm": 2.0633427358605774,
      "learning_rate": 1e-05,
      "loss": 1.5698,
      "step": 101
    },
    {
      "epoch": 0.0707534899852597,
      "grad_norm": 2.1113080337620795,
      "learning_rate": 1e-05,
      "loss": 1.593,
      "step": 102
    },
    {
      "epoch": 0.07144715165178184,
      "grad_norm": 2.200709143893939,
      "learning_rate": 1e-05,
      "loss": 1.5487,
      "step": 103
    },
    {
      "epoch": 0.07214081331830399,
      "grad_norm": 2.115437173154835,
      "learning_rate": 1e-05,
      "loss": 1.5184,
      "step": 104
    },
    {
      "epoch": 0.07283447498482615,
      "grad_norm": 1.9592926067021565,
      "learning_rate": 1e-05,
      "loss": 1.5824,
      "step": 105
    },
    {
      "epoch": 0.0735281366513483,
      "grad_norm": 2.0541106318447366,
      "learning_rate": 1e-05,
      "loss": 1.5265,
      "step": 106
    },
    {
      "epoch": 0.07422179831787046,
      "grad_norm": 1.9381187865210794,
      "learning_rate": 1e-05,
      "loss": 1.5628,
      "step": 107
    },
    {
      "epoch": 0.07491545998439261,
      "grad_norm": 2.1035504116541084,
      "learning_rate": 1e-05,
      "loss": 1.6001,
      "step": 108
    },
    {
      "epoch": 0.07560912165091477,
      "grad_norm": 2.143235938125612,
      "learning_rate": 1e-05,
      "loss": 1.5153,
      "step": 109
    },
    {
      "epoch": 0.07630278331743692,
      "grad_norm": 2.0039392778273357,
      "learning_rate": 1e-05,
      "loss": 1.5603,
      "step": 110
    },
    {
      "epoch": 0.07699644498395908,
      "grad_norm": 2.0158397924903233,
      "learning_rate": 1e-05,
      "loss": 1.5216,
      "step": 111
    },
    {
      "epoch": 0.07769010665048123,
      "grad_norm": 1.894029714001099,
      "learning_rate": 1e-05,
      "loss": 1.5587,
      "step": 112
    },
    {
      "epoch": 0.07838376831700339,
      "grad_norm": 2.432706928463119,
      "learning_rate": 1e-05,
      "loss": 1.5707,
      "step": 113
    },
    {
      "epoch": 0.07907742998352553,
      "grad_norm": 2.036785190887291,
      "learning_rate": 1e-05,
      "loss": 1.4926,
      "step": 114
    },
    {
      "epoch": 0.07977109165004768,
      "grad_norm": 1.9400189433970951,
      "learning_rate": 1e-05,
      "loss": 1.6043,
      "step": 115
    },
    {
      "epoch": 0.08046475331656984,
      "grad_norm": 1.919202983721404,
      "learning_rate": 1e-05,
      "loss": 1.537,
      "step": 116
    },
    {
      "epoch": 0.081158414983092,
      "grad_norm": 2.0977172125022707,
      "learning_rate": 1e-05,
      "loss": 1.5141,
      "step": 117
    },
    {
      "epoch": 0.08185207664961415,
      "grad_norm": 2.06407983728045,
      "learning_rate": 1e-05,
      "loss": 1.5672,
      "step": 118
    },
    {
      "epoch": 0.0825457383161363,
      "grad_norm": 2.078141165316271,
      "learning_rate": 1e-05,
      "loss": 1.5411,
      "step": 119
    },
    {
      "epoch": 0.08323939998265846,
      "grad_norm": 2.128515674818184,
      "learning_rate": 1e-05,
      "loss": 1.574,
      "step": 120
    },
    {
      "epoch": 0.08393306164918062,
      "grad_norm": 2.204639065305811,
      "learning_rate": 1e-05,
      "loss": 1.5411,
      "step": 121
    },
    {
      "epoch": 0.08462672331570277,
      "grad_norm": 2.028431294158661,
      "learning_rate": 1e-05,
      "loss": 1.4471,
      "step": 122
    },
    {
      "epoch": 0.08532038498222493,
      "grad_norm": 1.9634386333120701,
      "learning_rate": 1e-05,
      "loss": 1.5498,
      "step": 123
    },
    {
      "epoch": 0.08601404664874707,
      "grad_norm": 2.029215017631285,
      "learning_rate": 1e-05,
      "loss": 1.5568,
      "step": 124
    },
    {
      "epoch": 0.08670770831526922,
      "grad_norm": 2.0663413738174397,
      "learning_rate": 1e-05,
      "loss": 1.5426,
      "step": 125
    },
    {
      "epoch": 0.08740136998179138,
      "grad_norm": 1.9619014687764207,
      "learning_rate": 1e-05,
      "loss": 1.5133,
      "step": 126
    },
    {
      "epoch": 0.08809503164831353,
      "grad_norm": 1.9162219051787666,
      "learning_rate": 1e-05,
      "loss": 1.5464,
      "step": 127
    },
    {
      "epoch": 0.08878869331483569,
      "grad_norm": 2.261542371230024,
      "learning_rate": 1e-05,
      "loss": 1.4646,
      "step": 128
    },
    {
      "epoch": 0.08948235498135784,
      "grad_norm": 1.9715291734543514,
      "learning_rate": 1e-05,
      "loss": 1.5132,
      "step": 129
    },
    {
      "epoch": 0.09017601664788,
      "grad_norm": 2.0203777348766834,
      "learning_rate": 1e-05,
      "loss": 1.5354,
      "step": 130
    },
    {
      "epoch": 0.09086967831440215,
      "grad_norm": 1.9138585711617677,
      "learning_rate": 1e-05,
      "loss": 1.5407,
      "step": 131
    },
    {
      "epoch": 0.09156333998092431,
      "grad_norm": 2.0173322994738596,
      "learning_rate": 1e-05,
      "loss": 1.5439,
      "step": 132
    },
    {
      "epoch": 0.09225700164744646,
      "grad_norm": 2.107269356193457,
      "learning_rate": 1e-05,
      "loss": 1.5155,
      "step": 133
    },
    {
      "epoch": 0.09295066331396862,
      "grad_norm": 1.8484907386064835,
      "learning_rate": 1e-05,
      "loss": 1.5765,
      "step": 134
    },
    {
      "epoch": 0.09364432498049076,
      "grad_norm": 1.9245702524265067,
      "learning_rate": 1e-05,
      "loss": 1.5584,
      "step": 135
    },
    {
      "epoch": 0.09433798664701291,
      "grad_norm": 2.01491446609071,
      "learning_rate": 1e-05,
      "loss": 1.5813,
      "step": 136
    },
    {
      "epoch": 0.09503164831353507,
      "grad_norm": 2.139603451563103,
      "learning_rate": 1e-05,
      "loss": 1.5156,
      "step": 137
    },
    {
      "epoch": 0.09572530998005722,
      "grad_norm": 2.02926570002898,
      "learning_rate": 1e-05,
      "loss": 1.5331,
      "step": 138
    },
    {
      "epoch": 0.09641897164657938,
      "grad_norm": 2.0788419828866314,
      "learning_rate": 1e-05,
      "loss": 1.5504,
      "step": 139
    },
    {
      "epoch": 0.09711263331310153,
      "grad_norm": 2.1919839551775016,
      "learning_rate": 1e-05,
      "loss": 1.519,
      "step": 140
    },
    {
      "epoch": 0.09780629497962369,
      "grad_norm": 1.905930415266768,
      "learning_rate": 1e-05,
      "loss": 1.5084,
      "step": 141
    },
    {
      "epoch": 0.09849995664614584,
      "grad_norm": 2.107652533544824,
      "learning_rate": 1e-05,
      "loss": 1.5004,
      "step": 142
    },
    {
      "epoch": 0.099193618312668,
      "grad_norm": 1.9142412206590709,
      "learning_rate": 1e-05,
      "loss": 1.5488,
      "step": 143
    },
    {
      "epoch": 0.09988727997919015,
      "grad_norm": 1.8910378379533608,
      "learning_rate": 1e-05,
      "loss": 1.4912,
      "step": 144
    },
    {
      "epoch": 0.10058094164571231,
      "grad_norm": 2.0531944494577385,
      "learning_rate": 1e-05,
      "loss": 1.5011,
      "step": 145
    },
    {
      "epoch": 0.10127460331223445,
      "grad_norm": 1.9561470527096887,
      "learning_rate": 1e-05,
      "loss": 1.483,
      "step": 146
    },
    {
      "epoch": 0.1019682649787566,
      "grad_norm": 2.0182745726837186,
      "learning_rate": 1e-05,
      "loss": 1.4969,
      "step": 147
    },
    {
      "epoch": 0.10266192664527876,
      "grad_norm": 2.1655704406766305,
      "learning_rate": 1e-05,
      "loss": 1.5151,
      "step": 148
    },
    {
      "epoch": 0.10335558831180092,
      "grad_norm": 2.067383515526746,
      "learning_rate": 1e-05,
      "loss": 1.4932,
      "step": 149
    },
    {
      "epoch": 0.10404924997832307,
      "grad_norm": 2.16565372834465,
      "learning_rate": 1e-05,
      "loss": 1.5211,
      "step": 150
    },
    {
      "epoch": 0.10474291164484523,
      "grad_norm": 2.067199972931285,
      "learning_rate": 1e-05,
      "loss": 1.5182,
      "step": 151
    },
    {
      "epoch": 0.10543657331136738,
      "grad_norm": 2.021560871683723,
      "learning_rate": 1e-05,
      "loss": 1.5216,
      "step": 152
    },
    {
      "epoch": 0.10613023497788954,
      "grad_norm": 2.0837036440044914,
      "learning_rate": 1e-05,
      "loss": 1.4789,
      "step": 153
    },
    {
      "epoch": 0.10682389664441169,
      "grad_norm": 2.105326579649833,
      "learning_rate": 1e-05,
      "loss": 1.5532,
      "step": 154
    },
    {
      "epoch": 0.10751755831093385,
      "grad_norm": 1.9531358624572481,
      "learning_rate": 1e-05,
      "loss": 1.5709,
      "step": 155
    },
    {
      "epoch": 0.108211219977456,
      "grad_norm": 2.064713519765923,
      "learning_rate": 1e-05,
      "loss": 1.5378,
      "step": 156
    },
    {
      "epoch": 0.10890488164397814,
      "grad_norm": 2.1618499671142226,
      "learning_rate": 1e-05,
      "loss": 1.5176,
      "step": 157
    },
    {
      "epoch": 0.1095985433105003,
      "grad_norm": 1.9086208868139072,
      "learning_rate": 1e-05,
      "loss": 1.5546,
      "step": 158
    },
    {
      "epoch": 0.11029220497702245,
      "grad_norm": 2.02664550739396,
      "learning_rate": 1e-05,
      "loss": 1.4997,
      "step": 159
    },
    {
      "epoch": 0.11098586664354461,
      "grad_norm": 1.999544656281972,
      "learning_rate": 1e-05,
      "loss": 1.5003,
      "step": 160
    },
    {
      "epoch": 0.11167952831006676,
      "grad_norm": 2.053756926485644,
      "learning_rate": 1e-05,
      "loss": 1.5292,
      "step": 161
    },
    {
      "epoch": 0.11237318997658892,
      "grad_norm": 2.094168834382519,
      "learning_rate": 1e-05,
      "loss": 1.4676,
      "step": 162
    },
    {
      "epoch": 0.11306685164311107,
      "grad_norm": 2.002754028656303,
      "learning_rate": 1e-05,
      "loss": 1.5105,
      "step": 163
    },
    {
      "epoch": 0.11376051330963323,
      "grad_norm": 1.9434708295091558,
      "learning_rate": 1e-05,
      "loss": 1.5053,
      "step": 164
    },
    {
      "epoch": 0.11445417497615538,
      "grad_norm": 2.0330157669650815,
      "learning_rate": 1e-05,
      "loss": 1.4888,
      "step": 165
    },
    {
      "epoch": 0.11514783664267754,
      "grad_norm": 1.9289370322291217,
      "learning_rate": 1e-05,
      "loss": 1.5277,
      "step": 166
    },
    {
      "epoch": 0.1158414983091997,
      "grad_norm": 1.959995924753835,
      "learning_rate": 1e-05,
      "loss": 1.4825,
      "step": 167
    },
    {
      "epoch": 0.11653515997572184,
      "grad_norm": 1.7991848893944669,
      "learning_rate": 1e-05,
      "loss": 1.5005,
      "step": 168
    },
    {
      "epoch": 0.11722882164224399,
      "grad_norm": 1.8734306548761093,
      "learning_rate": 1e-05,
      "loss": 1.515,
      "step": 169
    },
    {
      "epoch": 0.11792248330876615,
      "grad_norm": 2.0015044635043218,
      "learning_rate": 1e-05,
      "loss": 1.5252,
      "step": 170
    },
    {
      "epoch": 0.1186161449752883,
      "grad_norm": 2.2665936004262273,
      "learning_rate": 1e-05,
      "loss": 1.4838,
      "step": 171
    },
    {
      "epoch": 0.11930980664181046,
      "grad_norm": 1.9544216259159037,
      "learning_rate": 1e-05,
      "loss": 1.4925,
      "step": 172
    },
    {
      "epoch": 0.12000346830833261,
      "grad_norm": 2.1154532935532715,
      "learning_rate": 1e-05,
      "loss": 1.5146,
      "step": 173
    },
    {
      "epoch": 0.12069712997485477,
      "grad_norm": 2.1021548296617927,
      "learning_rate": 1e-05,
      "loss": 1.4915,
      "step": 174
    },
    {
      "epoch": 0.12139079164137692,
      "grad_norm": 2.0834121176679634,
      "learning_rate": 1e-05,
      "loss": 1.5075,
      "step": 175
    },
    {
      "epoch": 0.12208445330789908,
      "grad_norm": 2.0016972202996213,
      "learning_rate": 1e-05,
      "loss": 1.5255,
      "step": 176
    },
    {
      "epoch": 0.12277811497442123,
      "grad_norm": 2.0612598236678523,
      "learning_rate": 1e-05,
      "loss": 1.5149,
      "step": 177
    },
    {
      "epoch": 0.12347177664094337,
      "grad_norm": 2.0229183358200484,
      "learning_rate": 1e-05,
      "loss": 1.5249,
      "step": 178
    },
    {
      "epoch": 0.12416543830746553,
      "grad_norm": 2.263550407359551,
      "learning_rate": 1e-05,
      "loss": 1.5244,
      "step": 179
    },
    {
      "epoch": 0.12485909997398768,
      "grad_norm": 2.08456086134308,
      "learning_rate": 1e-05,
      "loss": 1.4981,
      "step": 180
    },
    {
      "epoch": 0.12555276164050985,
      "grad_norm": 2.0774621427372386,
      "learning_rate": 1e-05,
      "loss": 1.5143,
      "step": 181
    },
    {
      "epoch": 0.126246423307032,
      "grad_norm": 2.1010285933058626,
      "learning_rate": 1e-05,
      "loss": 1.4504,
      "step": 182
    },
    {
      "epoch": 0.12694008497355416,
      "grad_norm": 2.0102176748558405,
      "learning_rate": 1e-05,
      "loss": 1.4146,
      "step": 183
    },
    {
      "epoch": 0.1276337466400763,
      "grad_norm": 2.095717278113951,
      "learning_rate": 1e-05,
      "loss": 1.5395,
      "step": 184
    },
    {
      "epoch": 0.12832740830659844,
      "grad_norm": 2.193298827450061,
      "learning_rate": 1e-05,
      "loss": 1.4769,
      "step": 185
    },
    {
      "epoch": 0.1290210699731206,
      "grad_norm": 1.9388355574681662,
      "learning_rate": 1e-05,
      "loss": 1.5375,
      "step": 186
    },
    {
      "epoch": 0.12971473163964276,
      "grad_norm": 2.0877632629967913,
      "learning_rate": 1e-05,
      "loss": 1.5471,
      "step": 187
    },
    {
      "epoch": 0.1304083933061649,
      "grad_norm": 2.22367106492369,
      "learning_rate": 1e-05,
      "loss": 1.4583,
      "step": 188
    },
    {
      "epoch": 0.13110205497268707,
      "grad_norm": 1.9943605205254191,
      "learning_rate": 1e-05,
      "loss": 1.5008,
      "step": 189
    },
    {
      "epoch": 0.13179571663920922,
      "grad_norm": 2.096604510058919,
      "learning_rate": 1e-05,
      "loss": 1.5449,
      "step": 190
    },
    {
      "epoch": 0.13248937830573138,
      "grad_norm": 1.9264619546423505,
      "learning_rate": 1e-05,
      "loss": 1.5068,
      "step": 191
    },
    {
      "epoch": 0.13318303997225353,
      "grad_norm": 2.250824707812072,
      "learning_rate": 1e-05,
      "loss": 1.4505,
      "step": 192
    },
    {
      "epoch": 0.13387670163877569,
      "grad_norm": 2.166944357294215,
      "learning_rate": 1e-05,
      "loss": 1.5341,
      "step": 193
    },
    {
      "epoch": 0.13457036330529784,
      "grad_norm": 2.0250424027409673,
      "learning_rate": 1e-05,
      "loss": 1.4852,
      "step": 194
    },
    {
      "epoch": 0.13526402497182,
      "grad_norm": 2.122488575543949,
      "learning_rate": 1e-05,
      "loss": 1.4973,
      "step": 195
    },
    {
      "epoch": 0.13595768663834215,
      "grad_norm": 2.1098754788199647,
      "learning_rate": 1e-05,
      "loss": 1.5069,
      "step": 196
    },
    {
      "epoch": 0.1366513483048643,
      "grad_norm": 2.088689323771004,
      "learning_rate": 1e-05,
      "loss": 1.4936,
      "step": 197
    },
    {
      "epoch": 0.13734500997138646,
      "grad_norm": 1.88466378266683,
      "learning_rate": 1e-05,
      "loss": 1.5068,
      "step": 198
    },
    {
      "epoch": 0.13803867163790862,
      "grad_norm": 2.0160265524905845,
      "learning_rate": 1e-05,
      "loss": 1.5368,
      "step": 199
    },
    {
      "epoch": 0.13873233330443077,
      "grad_norm": 1.8858982582224784,
      "learning_rate": 1e-05,
      "loss": 1.4676,
      "step": 200
    },
    {
      "epoch": 0.13942599497095293,
      "grad_norm": 1.8632802765837246,
      "learning_rate": 1e-05,
      "loss": 1.5151,
      "step": 201
    },
    {
      "epoch": 0.14011965663747508,
      "grad_norm": 1.9258461198592782,
      "learning_rate": 1e-05,
      "loss": 1.501,
      "step": 202
    },
    {
      "epoch": 0.14081331830399724,
      "grad_norm": 1.9872609586669983,
      "learning_rate": 1e-05,
      "loss": 1.4688,
      "step": 203
    },
    {
      "epoch": 0.1415069799705194,
      "grad_norm": 1.8678557794614834,
      "learning_rate": 1e-05,
      "loss": 1.5344,
      "step": 204
    },
    {
      "epoch": 0.14220064163704152,
      "grad_norm": 1.8914203459451417,
      "learning_rate": 1e-05,
      "loss": 1.4892,
      "step": 205
    },
    {
      "epoch": 0.14289430330356367,
      "grad_norm": 1.9911065628954558,
      "learning_rate": 1e-05,
      "loss": 1.5092,
      "step": 206
    },
    {
      "epoch": 0.14358796497008583,
      "grad_norm": 2.0485687976317966,
      "learning_rate": 1e-05,
      "loss": 1.5167,
      "step": 207
    },
    {
      "epoch": 0.14428162663660798,
      "grad_norm": 1.9496266310348234,
      "learning_rate": 1e-05,
      "loss": 1.4602,
      "step": 208
    },
    {
      "epoch": 0.14497528830313014,
      "grad_norm": 2.2871818269739754,
      "learning_rate": 1e-05,
      "loss": 1.4685,
      "step": 209
    },
    {
      "epoch": 0.1456689499696523,
      "grad_norm": 1.984619214022057,
      "learning_rate": 1e-05,
      "loss": 1.4409,
      "step": 210
    },
    {
      "epoch": 0.14636261163617445,
      "grad_norm": 1.9696667123704634,
      "learning_rate": 1e-05,
      "loss": 1.5176,
      "step": 211
    },
    {
      "epoch": 0.1470562733026966,
      "grad_norm": 1.933399558736106,
      "learning_rate": 1e-05,
      "loss": 1.5327,
      "step": 212
    },
    {
      "epoch": 0.14774993496921876,
      "grad_norm": 2.1861585612655885,
      "learning_rate": 1e-05,
      "loss": 1.4685,
      "step": 213
    },
    {
      "epoch": 0.14844359663574092,
      "grad_norm": 2.0747570404469804,
      "learning_rate": 1e-05,
      "loss": 1.4925,
      "step": 214
    },
    {
      "epoch": 0.14913725830226307,
      "grad_norm": 2.0747837540555096,
      "learning_rate": 1e-05,
      "loss": 1.5019,
      "step": 215
    },
    {
      "epoch": 0.14983091996878523,
      "grad_norm": 2.0326660466120297,
      "learning_rate": 1e-05,
      "loss": 1.4597,
      "step": 216
    },
    {
      "epoch": 0.15052458163530738,
      "grad_norm": 1.7823779998232254,
      "learning_rate": 1e-05,
      "loss": 1.4539,
      "step": 217
    },
    {
      "epoch": 0.15121824330182954,
      "grad_norm": 2.157188291473536,
      "learning_rate": 1e-05,
      "loss": 1.5397,
      "step": 218
    },
    {
      "epoch": 0.1519119049683517,
      "grad_norm": 2.0183890562420905,
      "learning_rate": 1e-05,
      "loss": 1.4753,
      "step": 219
    },
    {
      "epoch": 0.15260556663487385,
      "grad_norm": 1.9531135223586058,
      "learning_rate": 1e-05,
      "loss": 1.4315,
      "step": 220
    },
    {
      "epoch": 0.153299228301396,
      "grad_norm": 2.038347402060851,
      "learning_rate": 1e-05,
      "loss": 1.4246,
      "step": 221
    },
    {
      "epoch": 0.15399288996791816,
      "grad_norm": 2.0488940765067967,
      "learning_rate": 1e-05,
      "loss": 1.4461,
      "step": 222
    },
    {
      "epoch": 0.1546865516344403,
      "grad_norm": 2.231661501840028,
      "learning_rate": 1e-05,
      "loss": 1.472,
      "step": 223
    },
    {
      "epoch": 0.15538021330096247,
      "grad_norm": 1.9308119943251083,
      "learning_rate": 1e-05,
      "loss": 1.5128,
      "step": 224
    },
    {
      "epoch": 0.15607387496748462,
      "grad_norm": 2.229149512340474,
      "learning_rate": 1e-05,
      "loss": 1.4479,
      "step": 225
    },
    {
      "epoch": 0.15676753663400678,
      "grad_norm": 2.0928578049018483,
      "learning_rate": 1e-05,
      "loss": 1.5009,
      "step": 226
    },
    {
      "epoch": 0.1574611983005289,
      "grad_norm": 1.9663752374718868,
      "learning_rate": 1e-05,
      "loss": 1.4818,
      "step": 227
    },
    {
      "epoch": 0.15815485996705106,
      "grad_norm": 2.1034138941034786,
      "learning_rate": 1e-05,
      "loss": 1.4859,
      "step": 228
    },
    {
      "epoch": 0.15884852163357321,
      "grad_norm": 2.3560738411841626,
      "learning_rate": 1e-05,
      "loss": 1.4779,
      "step": 229
    },
    {
      "epoch": 0.15954218330009537,
      "grad_norm": 2.4332643261654403,
      "learning_rate": 1e-05,
      "loss": 1.4809,
      "step": 230
    },
    {
      "epoch": 0.16023584496661752,
      "grad_norm": 2.0278596830202757,
      "learning_rate": 1e-05,
      "loss": 1.4559,
      "step": 231
    },
    {
      "epoch": 0.16092950663313968,
      "grad_norm": 2.3341820259604984,
      "learning_rate": 1e-05,
      "loss": 1.479,
      "step": 232
    },
    {
      "epoch": 0.16162316829966183,
      "grad_norm": 1.9872750396163408,
      "learning_rate": 1e-05,
      "loss": 1.4715,
      "step": 233
    },
    {
      "epoch": 0.162316829966184,
      "grad_norm": 1.9999402472599845,
      "learning_rate": 1e-05,
      "loss": 1.5459,
      "step": 234
    },
    {
      "epoch": 0.16301049163270614,
      "grad_norm": 2.128103039977492,
      "learning_rate": 1e-05,
      "loss": 1.5096,
      "step": 235
    },
    {
      "epoch": 0.1637041532992283,
      "grad_norm": 1.9768244155184505,
      "learning_rate": 1e-05,
      "loss": 1.5127,
      "step": 236
    },
    {
      "epoch": 0.16439781496575046,
      "grad_norm": 1.788660125975532,
      "learning_rate": 1e-05,
      "loss": 1.4742,
      "step": 237
    },
    {
      "epoch": 0.1650914766322726,
      "grad_norm": 2.091066870692497,
      "learning_rate": 1e-05,
      "loss": 1.4978,
      "step": 238
    },
    {
      "epoch": 0.16578513829879477,
      "grad_norm": 1.9576004965176508,
      "learning_rate": 1e-05,
      "loss": 1.45,
      "step": 239
    },
    {
      "epoch": 0.16647879996531692,
      "grad_norm": 2.0698620965680528,
      "learning_rate": 1e-05,
      "loss": 1.4502,
      "step": 240
    },
    {
      "epoch": 0.16717246163183908,
      "grad_norm": 2.07030748449516,
      "learning_rate": 1e-05,
      "loss": 1.5028,
      "step": 241
    },
    {
      "epoch": 0.16786612329836123,
      "grad_norm": 2.0470194343255455,
      "learning_rate": 1e-05,
      "loss": 1.4714,
      "step": 242
    },
    {
      "epoch": 0.16855978496488339,
      "grad_norm": 2.0084641094309794,
      "learning_rate": 1e-05,
      "loss": 1.4819,
      "step": 243
    },
    {
      "epoch": 0.16925344663140554,
      "grad_norm": 2.0146863278209106,
      "learning_rate": 1e-05,
      "loss": 1.4604,
      "step": 244
    },
    {
      "epoch": 0.1699471082979277,
      "grad_norm": 2.068474055525701,
      "learning_rate": 1e-05,
      "loss": 1.4494,
      "step": 245
    },
    {
      "epoch": 0.17064076996444985,
      "grad_norm": 1.9451803439502662,
      "learning_rate": 1e-05,
      "loss": 1.4424,
      "step": 246
    },
    {
      "epoch": 0.171334431630972,
      "grad_norm": 2.0197781186907835,
      "learning_rate": 1e-05,
      "loss": 1.4614,
      "step": 247
    },
    {
      "epoch": 0.17202809329749413,
      "grad_norm": 1.9032139101620693,
      "learning_rate": 1e-05,
      "loss": 1.4731,
      "step": 248
    },
    {
      "epoch": 0.1727217549640163,
      "grad_norm": 2.115774791694279,
      "learning_rate": 1e-05,
      "loss": 1.416,
      "step": 249
    },
    {
      "epoch": 0.17341541663053844,
      "grad_norm": 2.0399470329500047,
      "learning_rate": 1e-05,
      "loss": 1.4481,
      "step": 250
    },
    {
      "epoch": 0.1741090782970606,
      "grad_norm": 2.2047084098275027,
      "learning_rate": 1e-05,
      "loss": 1.4822,
      "step": 251
    },
    {
      "epoch": 0.17480273996358275,
      "grad_norm": 2.0604716275395374,
      "learning_rate": 1e-05,
      "loss": 1.4698,
      "step": 252
    },
    {
      "epoch": 0.1754964016301049,
      "grad_norm": 2.0292160669190107,
      "learning_rate": 1e-05,
      "loss": 1.4292,
      "step": 253
    },
    {
      "epoch": 0.17619006329662706,
      "grad_norm": 1.8814965372884114,
      "learning_rate": 1e-05,
      "loss": 1.448,
      "step": 254
    },
    {
      "epoch": 0.17688372496314922,
      "grad_norm": 2.1057756496393805,
      "learning_rate": 1e-05,
      "loss": 1.4335,
      "step": 255
    },
    {
      "epoch": 0.17757738662967137,
      "grad_norm": 2.3132846447554227,
      "learning_rate": 1e-05,
      "loss": 1.4856,
      "step": 256
    },
    {
      "epoch": 0.17827104829619353,
      "grad_norm": 2.0736038589358734,
      "learning_rate": 1e-05,
      "loss": 1.4483,
      "step": 257
    },
    {
      "epoch": 0.17896470996271568,
      "grad_norm": 2.1910317153749985,
      "learning_rate": 1e-05,
      "loss": 1.4896,
      "step": 258
    },
    {
      "epoch": 0.17965837162923784,
      "grad_norm": 2.054248882735396,
      "learning_rate": 1e-05,
      "loss": 1.4202,
      "step": 259
    },
    {
      "epoch": 0.18035203329576,
      "grad_norm": 1.9735586840820536,
      "learning_rate": 1e-05,
      "loss": 1.492,
      "step": 260
    },
    {
      "epoch": 0.18104569496228215,
      "grad_norm": 2.0486128891661357,
      "learning_rate": 1e-05,
      "loss": 1.4842,
      "step": 261
    },
    {
      "epoch": 0.1817393566288043,
      "grad_norm": 2.0487679336561535,
      "learning_rate": 1e-05,
      "loss": 1.3914,
      "step": 262
    },
    {
      "epoch": 0.18243301829532646,
      "grad_norm": 2.122214099406227,
      "learning_rate": 1e-05,
      "loss": 1.5043,
      "step": 263
    },
    {
      "epoch": 0.18312667996184862,
      "grad_norm": 1.9819401665450367,
      "learning_rate": 1e-05,
      "loss": 1.461,
      "step": 264
    },
    {
      "epoch": 0.18382034162837077,
      "grad_norm": 1.8894079827925012,
      "learning_rate": 1e-05,
      "loss": 1.5248,
      "step": 265
    },
    {
      "epoch": 0.18451400329489293,
      "grad_norm": 1.9721789820827966,
      "learning_rate": 1e-05,
      "loss": 1.4289,
      "step": 266
    },
    {
      "epoch": 0.18520766496141508,
      "grad_norm": 1.9579977527698131,
      "learning_rate": 1e-05,
      "loss": 1.4313,
      "step": 267
    },
    {
      "epoch": 0.18590132662793724,
      "grad_norm": 2.0444647256611885,
      "learning_rate": 1e-05,
      "loss": 1.4645,
      "step": 268
    },
    {
      "epoch": 0.1865949882944594,
      "grad_norm": 2.1214502305742275,
      "learning_rate": 1e-05,
      "loss": 1.4712,
      "step": 269
    },
    {
      "epoch": 0.18728864996098152,
      "grad_norm": 1.9384086787313417,
      "learning_rate": 1e-05,
      "loss": 1.4641,
      "step": 270
    },
    {
      "epoch": 0.18798231162750367,
      "grad_norm": 1.9903323950267982,
      "learning_rate": 1e-05,
      "loss": 1.4438,
      "step": 271
    },
    {
      "epoch": 0.18867597329402583,
      "grad_norm": 2.2552584575632264,
      "learning_rate": 1e-05,
      "loss": 1.4954,
      "step": 272
    },
    {
      "epoch": 0.18936963496054798,
      "grad_norm": 2.0400461455788075,
      "learning_rate": 1e-05,
      "loss": 1.4574,
      "step": 273
    },
    {
      "epoch": 0.19006329662707014,
      "grad_norm": 1.985049295968063,
      "learning_rate": 1e-05,
      "loss": 1.4714,
      "step": 274
    },
    {
      "epoch": 0.1907569582935923,
      "grad_norm": 2.0084753961869173,
      "learning_rate": 1e-05,
      "loss": 1.4557,
      "step": 275
    },
    {
      "epoch": 0.19145061996011445,
      "grad_norm": 2.0212614192536473,
      "learning_rate": 1e-05,
      "loss": 1.4841,
      "step": 276
    },
    {
      "epoch": 0.1921442816266366,
      "grad_norm": 2.1484850741149035,
      "learning_rate": 1e-05,
      "loss": 1.5103,
      "step": 277
    },
    {
      "epoch": 0.19283794329315876,
      "grad_norm": 1.9196012631959583,
      "learning_rate": 1e-05,
      "loss": 1.4313,
      "step": 278
    },
    {
      "epoch": 0.19353160495968091,
      "grad_norm": 1.77676382629001,
      "learning_rate": 1e-05,
      "loss": 1.4468,
      "step": 279
    },
    {
      "epoch": 0.19422526662620307,
      "grad_norm": 1.9938658500301698,
      "learning_rate": 1e-05,
      "loss": 1.4,
      "step": 280
    },
    {
      "epoch": 0.19491892829272522,
      "grad_norm": 2.00675288394433,
      "learning_rate": 1e-05,
      "loss": 1.393,
      "step": 281
    },
    {
      "epoch": 0.19561258995924738,
      "grad_norm": 1.9133179166179877,
      "learning_rate": 1e-05,
      "loss": 1.4856,
      "step": 282
    },
    {
      "epoch": 0.19630625162576953,
      "grad_norm": 2.059048555946398,
      "learning_rate": 1e-05,
      "loss": 1.4583,
      "step": 283
    },
    {
      "epoch": 0.1969999132922917,
      "grad_norm": 1.9633303202331707,
      "learning_rate": 1e-05,
      "loss": 1.4713,
      "step": 284
    },
    {
      "epoch": 0.19769357495881384,
      "grad_norm": 2.062438630313021,
      "learning_rate": 1e-05,
      "loss": 1.4695,
      "step": 285
    },
    {
      "epoch": 0.198387236625336,
      "grad_norm": 2.0881034716487017,
      "learning_rate": 1e-05,
      "loss": 1.4088,
      "step": 286
    },
    {
      "epoch": 0.19908089829185815,
      "grad_norm": 2.1703550326161416,
      "learning_rate": 1e-05,
      "loss": 1.5003,
      "step": 287
    },
    {
      "epoch": 0.1997745599583803,
      "grad_norm": 2.074897325155734,
      "learning_rate": 1e-05,
      "loss": 1.4233,
      "step": 288
    },
    {
      "epoch": 0.20046822162490247,
      "grad_norm": 2.1455316048865667,
      "learning_rate": 1e-05,
      "loss": 1.499,
      "step": 289
    },
    {
      "epoch": 0.20116188329142462,
      "grad_norm": 2.101826626794352,
      "learning_rate": 1e-05,
      "loss": 1.423,
      "step": 290
    },
    {
      "epoch": 0.20185554495794678,
      "grad_norm": 1.9433232261964963,
      "learning_rate": 1e-05,
      "loss": 1.4258,
      "step": 291
    },
    {
      "epoch": 0.2025492066244689,
      "grad_norm": 2.0443499477866878,
      "learning_rate": 1e-05,
      "loss": 1.4755,
      "step": 292
    },
    {
      "epoch": 0.20324286829099106,
      "grad_norm": 2.0522507133871795,
      "learning_rate": 1e-05,
      "loss": 1.4472,
      "step": 293
    },
    {
      "epoch": 0.2039365299575132,
      "grad_norm": 2.138230646253678,
      "learning_rate": 1e-05,
      "loss": 1.49,
      "step": 294
    },
    {
      "epoch": 0.20463019162403537,
      "grad_norm": 1.8226561062473423,
      "learning_rate": 1e-05,
      "loss": 1.4556,
      "step": 295
    },
    {
      "epoch": 0.20532385329055752,
      "grad_norm": 2.038783861493903,
      "learning_rate": 1e-05,
      "loss": 1.4549,
      "step": 296
    },
    {
      "epoch": 0.20601751495707968,
      "grad_norm": 2.0515757165248054,
      "learning_rate": 1e-05,
      "loss": 1.4662,
      "step": 297
    },
    {
      "epoch": 0.20671117662360183,
      "grad_norm": 2.0235486233300564,
      "learning_rate": 1e-05,
      "loss": 1.4422,
      "step": 298
    },
    {
      "epoch": 0.207404838290124,
      "grad_norm": 1.9481327456484323,
      "learning_rate": 1e-05,
      "loss": 1.4871,
      "step": 299
    },
    {
      "epoch": 0.20809849995664614,
      "grad_norm": 2.1490389171940962,
      "learning_rate": 1e-05,
      "loss": 1.4438,
      "step": 300
    },
    {
      "epoch": 0.2087921616231683,
      "grad_norm": 1.8636272783701093,
      "learning_rate": 1e-05,
      "loss": 1.3833,
      "step": 301
    },
    {
      "epoch": 0.20948582328969045,
      "grad_norm": 2.0132879418102525,
      "learning_rate": 1e-05,
      "loss": 1.5079,
      "step": 302
    },
    {
      "epoch": 0.2101794849562126,
      "grad_norm": 1.982488095685725,
      "learning_rate": 1e-05,
      "loss": 1.4497,
      "step": 303
    },
    {
      "epoch": 0.21087314662273476,
      "grad_norm": 1.8238922950218728,
      "learning_rate": 1e-05,
      "loss": 1.4535,
      "step": 304
    },
    {
      "epoch": 0.21156680828925692,
      "grad_norm": 1.93657599411593,
      "learning_rate": 1e-05,
      "loss": 1.413,
      "step": 305
    },
    {
      "epoch": 0.21226046995577907,
      "grad_norm": 1.8702247940122898,
      "learning_rate": 1e-05,
      "loss": 1.4419,
      "step": 306
    },
    {
      "epoch": 0.21295413162230123,
      "grad_norm": 1.888122691714043,
      "learning_rate": 1e-05,
      "loss": 1.4674,
      "step": 307
    },
    {
      "epoch": 0.21364779328882338,
      "grad_norm": 1.937294530390563,
      "learning_rate": 1e-05,
      "loss": 1.5099,
      "step": 308
    },
    {
      "epoch": 0.21434145495534554,
      "grad_norm": 1.7678460926752717,
      "learning_rate": 1e-05,
      "loss": 1.4653,
      "step": 309
    },
    {
      "epoch": 0.2150351166218677,
      "grad_norm": 1.9151083060534004,
      "learning_rate": 1e-05,
      "loss": 1.4605,
      "step": 310
    },
    {
      "epoch": 0.21572877828838985,
      "grad_norm": 1.7787361825304597,
      "learning_rate": 1e-05,
      "loss": 1.4693,
      "step": 311
    },
    {
      "epoch": 0.216422439954912,
      "grad_norm": 1.9120724944581113,
      "learning_rate": 1e-05,
      "loss": 1.4881,
      "step": 312
    },
    {
      "epoch": 0.21711610162143413,
      "grad_norm": 1.779761445996506,
      "learning_rate": 1e-05,
      "loss": 1.4091,
      "step": 313
    },
    {
      "epoch": 0.2178097632879563,
      "grad_norm": 2.0485121226808527,
      "learning_rate": 1e-05,
      "loss": 1.434,
      "step": 314
    },
    {
      "epoch": 0.21850342495447844,
      "grad_norm": 1.8510715617371953,
      "learning_rate": 1e-05,
      "loss": 1.4502,
      "step": 315
    },
    {
      "epoch": 0.2191970866210006,
      "grad_norm": 2.0639411090008717,
      "learning_rate": 1e-05,
      "loss": 1.4663,
      "step": 316
    },
    {
      "epoch": 0.21989074828752275,
      "grad_norm": 1.8671225612100188,
      "learning_rate": 1e-05,
      "loss": 1.4679,
      "step": 317
    },
    {
      "epoch": 0.2205844099540449,
      "grad_norm": 2.089367239670572,
      "learning_rate": 1e-05,
      "loss": 1.4195,
      "step": 318
    },
    {
      "epoch": 0.22127807162056706,
      "grad_norm": 1.989736808079142,
      "learning_rate": 1e-05,
      "loss": 1.4397,
      "step": 319
    },
    {
      "epoch": 0.22197173328708922,
      "grad_norm": 1.9101605077692294,
      "learning_rate": 1e-05,
      "loss": 1.3985,
      "step": 320
    },
    {
      "epoch": 0.22266539495361137,
      "grad_norm": 1.9384620738328688,
      "learning_rate": 1e-05,
      "loss": 1.4223,
      "step": 321
    },
    {
      "epoch": 0.22335905662013353,
      "grad_norm": 2.1584155527633198,
      "learning_rate": 1e-05,
      "loss": 1.4836,
      "step": 322
    },
    {
      "epoch": 0.22405271828665568,
      "grad_norm": 2.0224370906370694,
      "learning_rate": 1e-05,
      "loss": 1.4454,
      "step": 323
    },
    {
      "epoch": 0.22474637995317784,
      "grad_norm": 2.0788397833125765,
      "learning_rate": 1e-05,
      "loss": 1.5232,
      "step": 324
    },
    {
      "epoch": 0.2254400416197,
      "grad_norm": 2.0341494419793427,
      "learning_rate": 1e-05,
      "loss": 1.4617,
      "step": 325
    },
    {
      "epoch": 0.22613370328622215,
      "grad_norm": 2.008192362429119,
      "learning_rate": 1e-05,
      "loss": 1.4375,
      "step": 326
    },
    {
      "epoch": 0.2268273649527443,
      "grad_norm": 1.7720050916534584,
      "learning_rate": 1e-05,
      "loss": 1.4517,
      "step": 327
    },
    {
      "epoch": 0.22752102661926646,
      "grad_norm": 1.9703464005250477,
      "learning_rate": 1e-05,
      "loss": 1.4201,
      "step": 328
    },
    {
      "epoch": 0.2282146882857886,
      "grad_norm": 1.8046683048934846,
      "learning_rate": 1e-05,
      "loss": 1.4441,
      "step": 329
    },
    {
      "epoch": 0.22890834995231077,
      "grad_norm": 1.996448453239083,
      "learning_rate": 1e-05,
      "loss": 1.411,
      "step": 330
    },
    {
      "epoch": 0.22960201161883292,
      "grad_norm": 2.202209909202205,
      "learning_rate": 1e-05,
      "loss": 1.4443,
      "step": 331
    },
    {
      "epoch": 0.23029567328535508,
      "grad_norm": 1.9288611183096158,
      "learning_rate": 1e-05,
      "loss": 1.4261,
      "step": 332
    },
    {
      "epoch": 0.23098933495187723,
      "grad_norm": 1.9931802186149232,
      "learning_rate": 1e-05,
      "loss": 1.4228,
      "step": 333
    },
    {
      "epoch": 0.2316829966183994,
      "grad_norm": 1.9300657348149677,
      "learning_rate": 1e-05,
      "loss": 1.4791,
      "step": 334
    },
    {
      "epoch": 0.23237665828492152,
      "grad_norm": 1.952656400463476,
      "learning_rate": 1e-05,
      "loss": 1.4682,
      "step": 335
    },
    {
      "epoch": 0.23307031995144367,
      "grad_norm": 1.920902486845839,
      "learning_rate": 1e-05,
      "loss": 1.4286,
      "step": 336
    },
    {
      "epoch": 0.23376398161796583,
      "grad_norm": 1.9689037515231558,
      "learning_rate": 1e-05,
      "loss": 1.439,
      "step": 337
    },
    {
      "epoch": 0.23445764328448798,
      "grad_norm": 1.928911178735969,
      "learning_rate": 1e-05,
      "loss": 1.4464,
      "step": 338
    },
    {
      "epoch": 0.23515130495101014,
      "grad_norm": 2.1288190614130134,
      "learning_rate": 1e-05,
      "loss": 1.4386,
      "step": 339
    },
    {
      "epoch": 0.2358449666175323,
      "grad_norm": 2.076072187290826,
      "learning_rate": 1e-05,
      "loss": 1.4928,
      "step": 340
    },
    {
      "epoch": 0.23653862828405445,
      "grad_norm": 1.8705047682268778,
      "learning_rate": 1e-05,
      "loss": 1.4536,
      "step": 341
    },
    {
      "epoch": 0.2372322899505766,
      "grad_norm": 1.9565604386223752,
      "learning_rate": 1e-05,
      "loss": 1.4783,
      "step": 342
    },
    {
      "epoch": 0.23792595161709876,
      "grad_norm": 2.028892379215102,
      "learning_rate": 1e-05,
      "loss": 1.4272,
      "step": 343
    },
    {
      "epoch": 0.2386196132836209,
      "grad_norm": 1.8650321597118786,
      "learning_rate": 1e-05,
      "loss": 1.4742,
      "step": 344
    },
    {
      "epoch": 0.23931327495014307,
      "grad_norm": 2.1612306768116083,
      "learning_rate": 1e-05,
      "loss": 1.4405,
      "step": 345
    },
    {
      "epoch": 0.24000693661666522,
      "grad_norm": 2.0151472915975748,
      "learning_rate": 1e-05,
      "loss": 1.4246,
      "step": 346
    },
    {
      "epoch": 0.24070059828318738,
      "grad_norm": 2.1329263531073472,
      "learning_rate": 1e-05,
      "loss": 1.4483,
      "step": 347
    },
    {
      "epoch": 0.24139425994970953,
      "grad_norm": 1.8503952674647635,
      "learning_rate": 1e-05,
      "loss": 1.3852,
      "step": 348
    },
    {
      "epoch": 0.2420879216162317,
      "grad_norm": 1.9232118601702997,
      "learning_rate": 1e-05,
      "loss": 1.4704,
      "step": 349
    },
    {
      "epoch": 0.24278158328275384,
      "grad_norm": 1.8955902858682858,
      "learning_rate": 1e-05,
      "loss": 1.4472,
      "step": 350
    },
    {
      "epoch": 0.243475244949276,
      "grad_norm": 2.0209412893438503,
      "learning_rate": 1e-05,
      "loss": 1.453,
      "step": 351
    },
    {
      "epoch": 0.24416890661579815,
      "grad_norm": 2.0049406945887736,
      "learning_rate": 1e-05,
      "loss": 1.4429,
      "step": 352
    },
    {
      "epoch": 0.2448625682823203,
      "grad_norm": 1.8970813615917181,
      "learning_rate": 1e-05,
      "loss": 1.4145,
      "step": 353
    },
    {
      "epoch": 0.24555622994884246,
      "grad_norm": 1.998898929474657,
      "learning_rate": 1e-05,
      "loss": 1.3887,
      "step": 354
    },
    {
      "epoch": 0.24624989161536462,
      "grad_norm": 2.0390603685379944,
      "learning_rate": 1e-05,
      "loss": 1.477,
      "step": 355
    },
    {
      "epoch": 0.24694355328188675,
      "grad_norm": 1.9473625966599237,
      "learning_rate": 1e-05,
      "loss": 1.422,
      "step": 356
    },
    {
      "epoch": 0.2476372149484089,
      "grad_norm": 1.8621154509930413,
      "learning_rate": 1e-05,
      "loss": 1.4551,
      "step": 357
    },
    {
      "epoch": 0.24833087661493106,
      "grad_norm": 2.079525202308318,
      "learning_rate": 1e-05,
      "loss": 1.496,
      "step": 358
    },
    {
      "epoch": 0.2490245382814532,
      "grad_norm": 2.0606010137774162,
      "learning_rate": 1e-05,
      "loss": 1.3982,
      "step": 359
    },
    {
      "epoch": 0.24971819994797537,
      "grad_norm": 2.4565497491600015,
      "learning_rate": 1e-05,
      "loss": 1.4833,
      "step": 360
    },
    {
      "epoch": 0.2504118616144975,
      "grad_norm": 1.9825019010439706,
      "learning_rate": 1e-05,
      "loss": 1.4271,
      "step": 361
    },
    {
      "epoch": 0.2511055232810197,
      "grad_norm": 1.9086652448767583,
      "learning_rate": 1e-05,
      "loss": 1.4125,
      "step": 362
    },
    {
      "epoch": 0.25179918494754183,
      "grad_norm": 2.0999318583687625,
      "learning_rate": 1e-05,
      "loss": 1.4235,
      "step": 363
    },
    {
      "epoch": 0.252492846614064,
      "grad_norm": 1.985190014569069,
      "learning_rate": 1e-05,
      "loss": 1.4501,
      "step": 364
    },
    {
      "epoch": 0.25318650828058614,
      "grad_norm": 1.910068963847788,
      "learning_rate": 1e-05,
      "loss": 1.4658,
      "step": 365
    },
    {
      "epoch": 0.2538801699471083,
      "grad_norm": 2.1692090204353764,
      "learning_rate": 1e-05,
      "loss": 1.4418,
      "step": 366
    },
    {
      "epoch": 0.25457383161363045,
      "grad_norm": 1.9609194594229094,
      "learning_rate": 1e-05,
      "loss": 1.4975,
      "step": 367
    },
    {
      "epoch": 0.2552674932801526,
      "grad_norm": 1.9921583707202606,
      "learning_rate": 1e-05,
      "loss": 1.46,
      "step": 368
    },
    {
      "epoch": 0.25596115494667476,
      "grad_norm": 1.946762435726228,
      "learning_rate": 1e-05,
      "loss": 1.4346,
      "step": 369
    },
    {
      "epoch": 0.2566548166131969,
      "grad_norm": 1.9131156156498506,
      "learning_rate": 1e-05,
      "loss": 1.4145,
      "step": 370
    },
    {
      "epoch": 0.2573484782797191,
      "grad_norm": 2.0667230542895885,
      "learning_rate": 1e-05,
      "loss": 1.4428,
      "step": 371
    },
    {
      "epoch": 0.2580421399462412,
      "grad_norm": 1.8834877477820704,
      "learning_rate": 1e-05,
      "loss": 1.4198,
      "step": 372
    },
    {
      "epoch": 0.2587358016127634,
      "grad_norm": 2.192277081033706,
      "learning_rate": 1e-05,
      "loss": 1.4732,
      "step": 373
    },
    {
      "epoch": 0.2594294632792855,
      "grad_norm": 2.0923048016266406,
      "learning_rate": 1e-05,
      "loss": 1.4748,
      "step": 374
    },
    {
      "epoch": 0.2601231249458077,
      "grad_norm": 2.086280677890939,
      "learning_rate": 1e-05,
      "loss": 1.4405,
      "step": 375
    },
    {
      "epoch": 0.2608167866123298,
      "grad_norm": 1.905489769808509,
      "learning_rate": 1e-05,
      "loss": 1.4445,
      "step": 376
    },
    {
      "epoch": 0.261510448278852,
      "grad_norm": 2.0450091978264466,
      "learning_rate": 1e-05,
      "loss": 1.4186,
      "step": 377
    },
    {
      "epoch": 0.26220410994537413,
      "grad_norm": 2.133929617755548,
      "learning_rate": 1e-05,
      "loss": 1.4572,
      "step": 378
    },
    {
      "epoch": 0.2628977716118963,
      "grad_norm": 2.040274726964717,
      "learning_rate": 1e-05,
      "loss": 1.4238,
      "step": 379
    },
    {
      "epoch": 0.26359143327841844,
      "grad_norm": 2.031802523503596,
      "learning_rate": 1e-05,
      "loss": 1.485,
      "step": 380
    },
    {
      "epoch": 0.2642850949449406,
      "grad_norm": 1.8546387757797609,
      "learning_rate": 1e-05,
      "loss": 1.4579,
      "step": 381
    },
    {
      "epoch": 0.26497875661146275,
      "grad_norm": 2.0419344295588893,
      "learning_rate": 1e-05,
      "loss": 1.4663,
      "step": 382
    },
    {
      "epoch": 0.26567241827798493,
      "grad_norm": 1.9135093773488887,
      "learning_rate": 1e-05,
      "loss": 1.3924,
      "step": 383
    },
    {
      "epoch": 0.26636607994450706,
      "grad_norm": 1.9016139415831943,
      "learning_rate": 1e-05,
      "loss": 1.4308,
      "step": 384
    },
    {
      "epoch": 0.26705974161102924,
      "grad_norm": 1.9082693062876364,
      "learning_rate": 1e-05,
      "loss": 1.3594,
      "step": 385
    },
    {
      "epoch": 0.26775340327755137,
      "grad_norm": 1.9587042651997673,
      "learning_rate": 1e-05,
      "loss": 1.4879,
      "step": 386
    },
    {
      "epoch": 0.26844706494407355,
      "grad_norm": 1.8806230175431726,
      "learning_rate": 1e-05,
      "loss": 1.4101,
      "step": 387
    },
    {
      "epoch": 0.2691407266105957,
      "grad_norm": 1.7730356951796558,
      "learning_rate": 1e-05,
      "loss": 1.4196,
      "step": 388
    },
    {
      "epoch": 0.2698343882771178,
      "grad_norm": 2.0320726341653255,
      "learning_rate": 1e-05,
      "loss": 1.4207,
      "step": 389
    },
    {
      "epoch": 0.27052804994364,
      "grad_norm": 1.9758337171264837,
      "learning_rate": 1e-05,
      "loss": 1.3527,
      "step": 390
    },
    {
      "epoch": 0.2712217116101621,
      "grad_norm": 1.8958218154268665,
      "learning_rate": 1e-05,
      "loss": 1.4802,
      "step": 391
    },
    {
      "epoch": 0.2719153732766843,
      "grad_norm": 2.0207609598652,
      "learning_rate": 1e-05,
      "loss": 1.4304,
      "step": 392
    },
    {
      "epoch": 0.27260903494320643,
      "grad_norm": 1.9475691858544313,
      "learning_rate": 1e-05,
      "loss": 1.4018,
      "step": 393
    },
    {
      "epoch": 0.2733026966097286,
      "grad_norm": 1.9389041388735107,
      "learning_rate": 1e-05,
      "loss": 1.4343,
      "step": 394
    },
    {
      "epoch": 0.27399635827625074,
      "grad_norm": 1.905336185660673,
      "learning_rate": 1e-05,
      "loss": 1.4283,
      "step": 395
    },
    {
      "epoch": 0.2746900199427729,
      "grad_norm": 1.729440202579088,
      "learning_rate": 1e-05,
      "loss": 1.4066,
      "step": 396
    },
    {
      "epoch": 0.27538368160929505,
      "grad_norm": 1.7537140656073078,
      "learning_rate": 1e-05,
      "loss": 1.3722,
      "step": 397
    },
    {
      "epoch": 0.27607734327581723,
      "grad_norm": 1.8075317974729948,
      "learning_rate": 1e-05,
      "loss": 1.3875,
      "step": 398
    },
    {
      "epoch": 0.27677100494233936,
      "grad_norm": 1.8284022154266728,
      "learning_rate": 1e-05,
      "loss": 1.401,
      "step": 399
    },
    {
      "epoch": 0.27746466660886154,
      "grad_norm": 2.0107348857071563,
      "learning_rate": 1e-05,
      "loss": 1.4018,
      "step": 400
    },
    {
      "epoch": 0.27815832827538367,
      "grad_norm": 2.117911173802508,
      "learning_rate": 1e-05,
      "loss": 1.4517,
      "step": 401
    },
    {
      "epoch": 0.27885198994190585,
      "grad_norm": 1.8033673093925715,
      "learning_rate": 1e-05,
      "loss": 1.3847,
      "step": 402
    },
    {
      "epoch": 0.279545651608428,
      "grad_norm": 1.7714377429685437,
      "learning_rate": 1e-05,
      "loss": 1.4638,
      "step": 403
    },
    {
      "epoch": 0.28023931327495016,
      "grad_norm": 1.8489883317533833,
      "learning_rate": 1e-05,
      "loss": 1.4183,
      "step": 404
    },
    {
      "epoch": 0.2809329749414723,
      "grad_norm": 2.0970756827183625,
      "learning_rate": 1e-05,
      "loss": 1.4531,
      "step": 405
    },
    {
      "epoch": 0.2816266366079945,
      "grad_norm": 2.0300925713263247,
      "learning_rate": 1e-05,
      "loss": 1.4129,
      "step": 406
    },
    {
      "epoch": 0.2823202982745166,
      "grad_norm": 2.0694538236815365,
      "learning_rate": 1e-05,
      "loss": 1.4346,
      "step": 407
    },
    {
      "epoch": 0.2830139599410388,
      "grad_norm": 2.125152836002329,
      "learning_rate": 1e-05,
      "loss": 1.4636,
      "step": 408
    },
    {
      "epoch": 0.2837076216075609,
      "grad_norm": 1.783820266730962,
      "learning_rate": 1e-05,
      "loss": 1.4327,
      "step": 409
    },
    {
      "epoch": 0.28440128327408304,
      "grad_norm": 2.0534367028415943,
      "learning_rate": 1e-05,
      "loss": 1.4691,
      "step": 410
    },
    {
      "epoch": 0.2850949449406052,
      "grad_norm": 1.876558701718411,
      "learning_rate": 1e-05,
      "loss": 1.383,
      "step": 411
    },
    {
      "epoch": 0.28578860660712735,
      "grad_norm": 2.0521158454990855,
      "learning_rate": 1e-05,
      "loss": 1.4213,
      "step": 412
    },
    {
      "epoch": 0.28648226827364953,
      "grad_norm": 2.0220971535707877,
      "learning_rate": 1e-05,
      "loss": 1.4658,
      "step": 413
    },
    {
      "epoch": 0.28717592994017166,
      "grad_norm": 2.038148661267878,
      "learning_rate": 1e-05,
      "loss": 1.4389,
      "step": 414
    },
    {
      "epoch": 0.28786959160669384,
      "grad_norm": 1.9405516561969085,
      "learning_rate": 1e-05,
      "loss": 1.4077,
      "step": 415
    },
    {
      "epoch": 0.28856325327321597,
      "grad_norm": 1.7499414272985196,
      "learning_rate": 1e-05,
      "loss": 1.4215,
      "step": 416
    },
    {
      "epoch": 0.28925691493973815,
      "grad_norm": 1.8287074775541738,
      "learning_rate": 1e-05,
      "loss": 1.4734,
      "step": 417
    },
    {
      "epoch": 0.2899505766062603,
      "grad_norm": 2.029489052381792,
      "learning_rate": 1e-05,
      "loss": 1.3857,
      "step": 418
    },
    {
      "epoch": 0.29064423827278246,
      "grad_norm": 1.9688645010655113,
      "learning_rate": 1e-05,
      "loss": 1.377,
      "step": 419
    },
    {
      "epoch": 0.2913378999393046,
      "grad_norm": 1.925094261068192,
      "learning_rate": 1e-05,
      "loss": 1.4246,
      "step": 420
    },
    {
      "epoch": 0.2920315616058268,
      "grad_norm": 2.117190679742464,
      "learning_rate": 1e-05,
      "loss": 1.4595,
      "step": 421
    },
    {
      "epoch": 0.2927252232723489,
      "grad_norm": 2.076005784675454,
      "learning_rate": 1e-05,
      "loss": 1.4242,
      "step": 422
    },
    {
      "epoch": 0.2934188849388711,
      "grad_norm": 1.9173678536303644,
      "learning_rate": 1e-05,
      "loss": 1.4209,
      "step": 423
    },
    {
      "epoch": 0.2941125466053932,
      "grad_norm": 1.9453593461528418,
      "learning_rate": 1e-05,
      "loss": 1.4067,
      "step": 424
    },
    {
      "epoch": 0.2948062082719154,
      "grad_norm": 2.067561860466064,
      "learning_rate": 1e-05,
      "loss": 1.4637,
      "step": 425
    },
    {
      "epoch": 0.2954998699384375,
      "grad_norm": 1.899343422880033,
      "learning_rate": 1e-05,
      "loss": 1.4355,
      "step": 426
    },
    {
      "epoch": 0.2961935316049597,
      "grad_norm": 1.9847242096071747,
      "learning_rate": 1e-05,
      "loss": 1.4292,
      "step": 427
    },
    {
      "epoch": 0.29688719327148183,
      "grad_norm": 1.854907399156512,
      "learning_rate": 1e-05,
      "loss": 1.443,
      "step": 428
    },
    {
      "epoch": 0.297580854938004,
      "grad_norm": 1.955263267566865,
      "learning_rate": 1e-05,
      "loss": 1.3928,
      "step": 429
    },
    {
      "epoch": 0.29827451660452614,
      "grad_norm": 2.053105452037698,
      "learning_rate": 1e-05,
      "loss": 1.378,
      "step": 430
    },
    {
      "epoch": 0.2989681782710483,
      "grad_norm": 1.9109805438113896,
      "learning_rate": 1e-05,
      "loss": 1.3645,
      "step": 431
    },
    {
      "epoch": 0.29966183993757045,
      "grad_norm": 2.043933400997838,
      "learning_rate": 1e-05,
      "loss": 1.435,
      "step": 432
    },
    {
      "epoch": 0.3003555016040926,
      "grad_norm": 2.1971134552732785,
      "learning_rate": 1e-05,
      "loss": 1.4391,
      "step": 433
    },
    {
      "epoch": 0.30104916327061476,
      "grad_norm": 1.9910023447928422,
      "learning_rate": 1e-05,
      "loss": 1.4084,
      "step": 434
    },
    {
      "epoch": 0.3017428249371369,
      "grad_norm": 2.0432125795374607,
      "learning_rate": 1e-05,
      "loss": 1.4423,
      "step": 435
    },
    {
      "epoch": 0.30243648660365907,
      "grad_norm": 1.9609905480132668,
      "learning_rate": 1e-05,
      "loss": 1.4511,
      "step": 436
    },
    {
      "epoch": 0.3031301482701812,
      "grad_norm": 1.8755027815135972,
      "learning_rate": 1e-05,
      "loss": 1.3937,
      "step": 437
    },
    {
      "epoch": 0.3038238099367034,
      "grad_norm": 1.9069584207191346,
      "learning_rate": 1e-05,
      "loss": 1.38,
      "step": 438
    },
    {
      "epoch": 0.3045174716032255,
      "grad_norm": 2.174136818483425,
      "learning_rate": 1e-05,
      "loss": 1.4535,
      "step": 439
    },
    {
      "epoch": 0.3052111332697477,
      "grad_norm": 1.8297620514780848,
      "learning_rate": 1e-05,
      "loss": 1.4285,
      "step": 440
    },
    {
      "epoch": 0.3059047949362698,
      "grad_norm": 1.8949261520180545,
      "learning_rate": 1e-05,
      "loss": 1.4336,
      "step": 441
    },
    {
      "epoch": 0.306598456602792,
      "grad_norm": 1.9549110409659416,
      "learning_rate": 1e-05,
      "loss": 1.373,
      "step": 442
    },
    {
      "epoch": 0.30729211826931413,
      "grad_norm": 1.9504761362373537,
      "learning_rate": 1e-05,
      "loss": 1.3915,
      "step": 443
    },
    {
      "epoch": 0.3079857799358363,
      "grad_norm": 1.8994399368824695,
      "learning_rate": 1e-05,
      "loss": 1.3887,
      "step": 444
    },
    {
      "epoch": 0.30867944160235844,
      "grad_norm": 1.8796327717601744,
      "learning_rate": 1e-05,
      "loss": 1.4075,
      "step": 445
    },
    {
      "epoch": 0.3093731032688806,
      "grad_norm": 1.9681787692068275,
      "learning_rate": 1e-05,
      "loss": 1.4476,
      "step": 446
    },
    {
      "epoch": 0.31006676493540275,
      "grad_norm": 2.0373492235623285,
      "learning_rate": 1e-05,
      "loss": 1.3909,
      "step": 447
    },
    {
      "epoch": 0.31076042660192493,
      "grad_norm": 1.8891756577788732,
      "learning_rate": 1e-05,
      "loss": 1.4804,
      "step": 448
    },
    {
      "epoch": 0.31145408826844706,
      "grad_norm": 1.833019021556368,
      "learning_rate": 1e-05,
      "loss": 1.4207,
      "step": 449
    },
    {
      "epoch": 0.31214774993496924,
      "grad_norm": 1.836006077142845,
      "learning_rate": 1e-05,
      "loss": 1.3933,
      "step": 450
    },
    {
      "epoch": 0.31284141160149137,
      "grad_norm": 1.8756183495188887,
      "learning_rate": 1e-05,
      "loss": 1.4081,
      "step": 451
    },
    {
      "epoch": 0.31353507326801355,
      "grad_norm": 1.8832247399343314,
      "learning_rate": 1e-05,
      "loss": 1.3978,
      "step": 452
    },
    {
      "epoch": 0.3142287349345357,
      "grad_norm": 1.9411565103780286,
      "learning_rate": 1e-05,
      "loss": 1.4346,
      "step": 453
    },
    {
      "epoch": 0.3149223966010578,
      "grad_norm": 1.9226021145167342,
      "learning_rate": 1e-05,
      "loss": 1.4192,
      "step": 454
    },
    {
      "epoch": 0.31561605826758,
      "grad_norm": 1.864849847975618,
      "learning_rate": 1e-05,
      "loss": 1.4023,
      "step": 455
    },
    {
      "epoch": 0.3163097199341021,
      "grad_norm": 1.841130556375779,
      "learning_rate": 1e-05,
      "loss": 1.3843,
      "step": 456
    },
    {
      "epoch": 0.3170033816006243,
      "grad_norm": 1.9910619133587744,
      "learning_rate": 1e-05,
      "loss": 1.4196,
      "step": 457
    },
    {
      "epoch": 0.31769704326714643,
      "grad_norm": 1.871071007237103,
      "learning_rate": 1e-05,
      "loss": 1.4398,
      "step": 458
    },
    {
      "epoch": 0.3183907049336686,
      "grad_norm": 1.9525027065397538,
      "learning_rate": 1e-05,
      "loss": 1.4221,
      "step": 459
    },
    {
      "epoch": 0.31908436660019074,
      "grad_norm": 1.875959289372641,
      "learning_rate": 1e-05,
      "loss": 1.4568,
      "step": 460
    },
    {
      "epoch": 0.3197780282667129,
      "grad_norm": 1.896488018271282,
      "learning_rate": 1e-05,
      "loss": 1.4443,
      "step": 461
    },
    {
      "epoch": 0.32047168993323505,
      "grad_norm": 1.819017143897324,
      "learning_rate": 1e-05,
      "loss": 1.3804,
      "step": 462
    },
    {
      "epoch": 0.32116535159975723,
      "grad_norm": 1.904368280434257,
      "learning_rate": 1e-05,
      "loss": 1.3803,
      "step": 463
    },
    {
      "epoch": 0.32185901326627936,
      "grad_norm": 1.9630080846645839,
      "learning_rate": 1e-05,
      "loss": 1.4099,
      "step": 464
    },
    {
      "epoch": 0.32255267493280154,
      "grad_norm": 1.8933783131596658,
      "learning_rate": 1e-05,
      "loss": 1.4426,
      "step": 465
    },
    {
      "epoch": 0.32324633659932367,
      "grad_norm": 1.9919230915473398,
      "learning_rate": 1e-05,
      "loss": 1.4099,
      "step": 466
    },
    {
      "epoch": 0.32393999826584585,
      "grad_norm": 2.0057958567267606,
      "learning_rate": 1e-05,
      "loss": 1.4409,
      "step": 467
    },
    {
      "epoch": 0.324633659932368,
      "grad_norm": 2.0237082745853088,
      "learning_rate": 1e-05,
      "loss": 1.4331,
      "step": 468
    },
    {
      "epoch": 0.32532732159889016,
      "grad_norm": 1.9072494572387,
      "learning_rate": 1e-05,
      "loss": 1.4298,
      "step": 469
    },
    {
      "epoch": 0.3260209832654123,
      "grad_norm": 2.1235561783712718,
      "learning_rate": 1e-05,
      "loss": 1.4125,
      "step": 470
    },
    {
      "epoch": 0.3267146449319345,
      "grad_norm": 1.844374731285956,
      "learning_rate": 1e-05,
      "loss": 1.3672,
      "step": 471
    },
    {
      "epoch": 0.3274083065984566,
      "grad_norm": 2.1204902498109326,
      "learning_rate": 1e-05,
      "loss": 1.3909,
      "step": 472
    },
    {
      "epoch": 0.3281019682649788,
      "grad_norm": 2.084912862868831,
      "learning_rate": 1e-05,
      "loss": 1.4521,
      "step": 473
    },
    {
      "epoch": 0.3287956299315009,
      "grad_norm": 1.9363382366089963,
      "learning_rate": 1e-05,
      "loss": 1.4166,
      "step": 474
    },
    {
      "epoch": 0.32948929159802304,
      "grad_norm": 2.043516332073307,
      "learning_rate": 1e-05,
      "loss": 1.4221,
      "step": 475
    },
    {
      "epoch": 0.3301829532645452,
      "grad_norm": 2.125873099222709,
      "learning_rate": 1e-05,
      "loss": 1.4894,
      "step": 476
    },
    {
      "epoch": 0.33087661493106735,
      "grad_norm": 2.042031475330511,
      "learning_rate": 1e-05,
      "loss": 1.4025,
      "step": 477
    },
    {
      "epoch": 0.33157027659758953,
      "grad_norm": 1.7849937086119454,
      "learning_rate": 1e-05,
      "loss": 1.4132,
      "step": 478
    },
    {
      "epoch": 0.33226393826411166,
      "grad_norm": 1.7931574592397888,
      "learning_rate": 1e-05,
      "loss": 1.4273,
      "step": 479
    },
    {
      "epoch": 0.33295759993063384,
      "grad_norm": 2.826745637249205,
      "learning_rate": 1e-05,
      "loss": 1.3552,
      "step": 480
    },
    {
      "epoch": 0.33365126159715597,
      "grad_norm": 1.9644473691572903,
      "learning_rate": 1e-05,
      "loss": 1.3728,
      "step": 481
    },
    {
      "epoch": 0.33434492326367815,
      "grad_norm": 1.8815785092027932,
      "learning_rate": 1e-05,
      "loss": 1.3836,
      "step": 482
    },
    {
      "epoch": 0.3350385849302003,
      "grad_norm": 2.1707103250407265,
      "learning_rate": 1e-05,
      "loss": 1.3872,
      "step": 483
    },
    {
      "epoch": 0.33573224659672246,
      "grad_norm": 1.805124910060426,
      "learning_rate": 1e-05,
      "loss": 1.4269,
      "step": 484
    },
    {
      "epoch": 0.3364259082632446,
      "grad_norm": 2.0566289424785165,
      "learning_rate": 1e-05,
      "loss": 1.4353,
      "step": 485
    },
    {
      "epoch": 0.33711956992976677,
      "grad_norm": 1.9392776766197952,
      "learning_rate": 1e-05,
      "loss": 1.42,
      "step": 486
    },
    {
      "epoch": 0.3378132315962889,
      "grad_norm": 1.7990589474342267,
      "learning_rate": 1e-05,
      "loss": 1.4232,
      "step": 487
    },
    {
      "epoch": 0.3385068932628111,
      "grad_norm": 2.042243634171937,
      "learning_rate": 1e-05,
      "loss": 1.4094,
      "step": 488
    },
    {
      "epoch": 0.3392005549293332,
      "grad_norm": 1.9990437442788238,
      "learning_rate": 1e-05,
      "loss": 1.4349,
      "step": 489
    },
    {
      "epoch": 0.3398942165958554,
      "grad_norm": 2.1051212689157777,
      "learning_rate": 1e-05,
      "loss": 1.3838,
      "step": 490
    },
    {
      "epoch": 0.3405878782623775,
      "grad_norm": 2.1229600713115238,
      "learning_rate": 1e-05,
      "loss": 1.3403,
      "step": 491
    },
    {
      "epoch": 0.3412815399288997,
      "grad_norm": 1.8779948313961718,
      "learning_rate": 1e-05,
      "loss": 1.3938,
      "step": 492
    },
    {
      "epoch": 0.34197520159542183,
      "grad_norm": 1.9140202797636157,
      "learning_rate": 1e-05,
      "loss": 1.3917,
      "step": 493
    },
    {
      "epoch": 0.342668863261944,
      "grad_norm": 1.974739574014709,
      "learning_rate": 1e-05,
      "loss": 1.4054,
      "step": 494
    },
    {
      "epoch": 0.34336252492846614,
      "grad_norm": 1.9199035763474215,
      "learning_rate": 1e-05,
      "loss": 1.3922,
      "step": 495
    },
    {
      "epoch": 0.34405618659498827,
      "grad_norm": 1.9177941872465927,
      "learning_rate": 1e-05,
      "loss": 1.4168,
      "step": 496
    },
    {
      "epoch": 0.34474984826151045,
      "grad_norm": 2.00337000954214,
      "learning_rate": 1e-05,
      "loss": 1.3949,
      "step": 497
    },
    {
      "epoch": 0.3454435099280326,
      "grad_norm": 1.998945535857092,
      "learning_rate": 1e-05,
      "loss": 1.3929,
      "step": 498
    },
    {
      "epoch": 0.34613717159455476,
      "grad_norm": 2.0182449229484636,
      "learning_rate": 1e-05,
      "loss": 1.422,
      "step": 499
    },
    {
      "epoch": 0.3468308332610769,
      "grad_norm": 1.9361126702422482,
      "learning_rate": 1e-05,
      "loss": 1.3964,
      "step": 500
    },
    {
      "epoch": 0.34752449492759907,
      "grad_norm": 2.0248490486946316,
      "learning_rate": 1e-05,
      "loss": 1.4293,
      "step": 501
    },
    {
      "epoch": 0.3482181565941212,
      "grad_norm": 1.8982822396985153,
      "learning_rate": 1e-05,
      "loss": 1.3778,
      "step": 502
    },
    {
      "epoch": 0.3489118182606434,
      "grad_norm": 2.059259100223746,
      "learning_rate": 1e-05,
      "loss": 1.4509,
      "step": 503
    },
    {
      "epoch": 0.3496054799271655,
      "grad_norm": 2.00329460831796,
      "learning_rate": 1e-05,
      "loss": 1.3878,
      "step": 504
    },
    {
      "epoch": 0.3502991415936877,
      "grad_norm": 1.9719718567850673,
      "learning_rate": 1e-05,
      "loss": 1.3829,
      "step": 505
    },
    {
      "epoch": 0.3509928032602098,
      "grad_norm": 1.9623427987164561,
      "learning_rate": 1e-05,
      "loss": 1.4036,
      "step": 506
    },
    {
      "epoch": 0.351686464926732,
      "grad_norm": 2.022806123443883,
      "learning_rate": 1e-05,
      "loss": 1.4068,
      "step": 507
    },
    {
      "epoch": 0.35238012659325413,
      "grad_norm": 2.00015503601285,
      "learning_rate": 1e-05,
      "loss": 1.4248,
      "step": 508
    },
    {
      "epoch": 0.3530737882597763,
      "grad_norm": 1.904741684713311,
      "learning_rate": 1e-05,
      "loss": 1.3519,
      "step": 509
    },
    {
      "epoch": 0.35376744992629844,
      "grad_norm": 1.9455166972699935,
      "learning_rate": 1e-05,
      "loss": 1.4152,
      "step": 510
    },
    {
      "epoch": 0.3544611115928206,
      "grad_norm": 1.9409989480916887,
      "learning_rate": 1e-05,
      "loss": 1.3916,
      "step": 511
    },
    {
      "epoch": 0.35515477325934275,
      "grad_norm": 1.8707742492583141,
      "learning_rate": 1e-05,
      "loss": 1.4362,
      "step": 512
    },
    {
      "epoch": 0.35584843492586493,
      "grad_norm": 1.8314588678913148,
      "learning_rate": 1e-05,
      "loss": 1.4279,
      "step": 513
    },
    {
      "epoch": 0.35654209659238706,
      "grad_norm": 1.913107053098294,
      "learning_rate": 1e-05,
      "loss": 1.4686,
      "step": 514
    },
    {
      "epoch": 0.35723575825890924,
      "grad_norm": 1.8781008328794606,
      "learning_rate": 1e-05,
      "loss": 1.4257,
      "step": 515
    },
    {
      "epoch": 0.35792941992543137,
      "grad_norm": 1.858897618474902,
      "learning_rate": 1e-05,
      "loss": 1.3299,
      "step": 516
    },
    {
      "epoch": 0.35862308159195355,
      "grad_norm": 1.7902966802183116,
      "learning_rate": 1e-05,
      "loss": 1.3806,
      "step": 517
    },
    {
      "epoch": 0.3593167432584757,
      "grad_norm": 1.9406550842391148,
      "learning_rate": 1e-05,
      "loss": 1.4069,
      "step": 518
    },
    {
      "epoch": 0.3600104049249978,
      "grad_norm": 2.032374189763469,
      "learning_rate": 1e-05,
      "loss": 1.3893,
      "step": 519
    },
    {
      "epoch": 0.36070406659152,
      "grad_norm": 2.0622944109637946,
      "learning_rate": 1e-05,
      "loss": 1.4044,
      "step": 520
    },
    {
      "epoch": 0.3613977282580421,
      "grad_norm": 1.7694621919787776,
      "learning_rate": 1e-05,
      "loss": 1.3781,
      "step": 521
    },
    {
      "epoch": 0.3620913899245643,
      "grad_norm": 1.9876645276332312,
      "learning_rate": 1e-05,
      "loss": 1.3658,
      "step": 522
    },
    {
      "epoch": 0.3627850515910864,
      "grad_norm": 1.8286541300136883,
      "learning_rate": 1e-05,
      "loss": 1.3954,
      "step": 523
    },
    {
      "epoch": 0.3634787132576086,
      "grad_norm": 2.1598158167597785,
      "learning_rate": 1e-05,
      "loss": 1.4233,
      "step": 524
    },
    {
      "epoch": 0.36417237492413074,
      "grad_norm": 2.0125775815004308,
      "learning_rate": 1e-05,
      "loss": 1.425,
      "step": 525
    },
    {
      "epoch": 0.3648660365906529,
      "grad_norm": 1.8220999673991007,
      "learning_rate": 1e-05,
      "loss": 1.4689,
      "step": 526
    },
    {
      "epoch": 0.36555969825717505,
      "grad_norm": 1.9067444062678136,
      "learning_rate": 1e-05,
      "loss": 1.4184,
      "step": 527
    },
    {
      "epoch": 0.36625335992369723,
      "grad_norm": 1.8244010981958079,
      "learning_rate": 1e-05,
      "loss": 1.4307,
      "step": 528
    },
    {
      "epoch": 0.36694702159021936,
      "grad_norm": 1.9491392457779768,
      "learning_rate": 1e-05,
      "loss": 1.4046,
      "step": 529
    },
    {
      "epoch": 0.36764068325674154,
      "grad_norm": 1.8922109612389026,
      "learning_rate": 1e-05,
      "loss": 1.4312,
      "step": 530
    },
    {
      "epoch": 0.36833434492326367,
      "grad_norm": 1.8416689512552131,
      "learning_rate": 1e-05,
      "loss": 1.3785,
      "step": 531
    },
    {
      "epoch": 0.36902800658978585,
      "grad_norm": 1.8783492413280818,
      "learning_rate": 1e-05,
      "loss": 1.4109,
      "step": 532
    },
    {
      "epoch": 0.369721668256308,
      "grad_norm": 1.849885383573315,
      "learning_rate": 1e-05,
      "loss": 1.378,
      "step": 533
    },
    {
      "epoch": 0.37041532992283016,
      "grad_norm": 2.071142129836207,
      "learning_rate": 1e-05,
      "loss": 1.3897,
      "step": 534
    },
    {
      "epoch": 0.3711089915893523,
      "grad_norm": 1.955558194146127,
      "learning_rate": 1e-05,
      "loss": 1.3781,
      "step": 535
    },
    {
      "epoch": 0.37180265325587447,
      "grad_norm": 1.8539973789480575,
      "learning_rate": 1e-05,
      "loss": 1.3616,
      "step": 536
    },
    {
      "epoch": 0.3724963149223966,
      "grad_norm": 1.9837890849438713,
      "learning_rate": 1e-05,
      "loss": 1.3713,
      "step": 537
    },
    {
      "epoch": 0.3731899765889188,
      "grad_norm": 1.9070579057354713,
      "learning_rate": 1e-05,
      "loss": 1.3479,
      "step": 538
    },
    {
      "epoch": 0.3738836382554409,
      "grad_norm": 1.9425946819156692,
      "learning_rate": 1e-05,
      "loss": 1.376,
      "step": 539
    },
    {
      "epoch": 0.37457729992196304,
      "grad_norm": 1.8771696216581655,
      "learning_rate": 1e-05,
      "loss": 1.3827,
      "step": 540
    },
    {
      "epoch": 0.3752709615884852,
      "grad_norm": 1.9186911492371799,
      "learning_rate": 1e-05,
      "loss": 1.376,
      "step": 541
    },
    {
      "epoch": 0.37596462325500735,
      "grad_norm": 1.7294504716898789,
      "learning_rate": 1e-05,
      "loss": 1.3706,
      "step": 542
    },
    {
      "epoch": 0.37665828492152953,
      "grad_norm": 1.7712039048530457,
      "learning_rate": 1e-05,
      "loss": 1.4438,
      "step": 543
    },
    {
      "epoch": 0.37735194658805166,
      "grad_norm": 1.83847300072347,
      "learning_rate": 1e-05,
      "loss": 1.4161,
      "step": 544
    },
    {
      "epoch": 0.37804560825457384,
      "grad_norm": 1.825146012889819,
      "learning_rate": 1e-05,
      "loss": 1.4118,
      "step": 545
    },
    {
      "epoch": 0.37873926992109597,
      "grad_norm": 1.7989485574696917,
      "learning_rate": 1e-05,
      "loss": 1.3689,
      "step": 546
    },
    {
      "epoch": 0.37943293158761815,
      "grad_norm": 2.1484169058996816,
      "learning_rate": 1e-05,
      "loss": 1.3477,
      "step": 547
    },
    {
      "epoch": 0.3801265932541403,
      "grad_norm": 2.0781089621700777,
      "learning_rate": 1e-05,
      "loss": 1.4372,
      "step": 548
    },
    {
      "epoch": 0.38082025492066246,
      "grad_norm": 1.8367820389064522,
      "learning_rate": 1e-05,
      "loss": 1.4556,
      "step": 549
    },
    {
      "epoch": 0.3815139165871846,
      "grad_norm": 1.8748672667110173,
      "learning_rate": 1e-05,
      "loss": 1.4059,
      "step": 550
    },
    {
      "epoch": 0.38220757825370677,
      "grad_norm": 1.8956025199638242,
      "learning_rate": 1e-05,
      "loss": 1.4102,
      "step": 551
    },
    {
      "epoch": 0.3829012399202289,
      "grad_norm": 1.9601967006624061,
      "learning_rate": 1e-05,
      "loss": 1.3879,
      "step": 552
    },
    {
      "epoch": 0.3835949015867511,
      "grad_norm": 2.0406526524881707,
      "learning_rate": 1e-05,
      "loss": 1.3885,
      "step": 553
    },
    {
      "epoch": 0.3842885632532732,
      "grad_norm": 2.091444436822791,
      "learning_rate": 1e-05,
      "loss": 1.3999,
      "step": 554
    },
    {
      "epoch": 0.3849822249197954,
      "grad_norm": 1.7714596927341815,
      "learning_rate": 1e-05,
      "loss": 1.3452,
      "step": 555
    },
    {
      "epoch": 0.3856758865863175,
      "grad_norm": 1.825942444423705,
      "learning_rate": 1e-05,
      "loss": 1.4142,
      "step": 556
    },
    {
      "epoch": 0.3863695482528397,
      "grad_norm": 1.871025102921204,
      "learning_rate": 1e-05,
      "loss": 1.3421,
      "step": 557
    },
    {
      "epoch": 0.38706320991936183,
      "grad_norm": 1.9540951936439066,
      "learning_rate": 1e-05,
      "loss": 1.384,
      "step": 558
    },
    {
      "epoch": 0.387756871585884,
      "grad_norm": 1.8147562505586048,
      "learning_rate": 1e-05,
      "loss": 1.3794,
      "step": 559
    },
    {
      "epoch": 0.38845053325240614,
      "grad_norm": 1.8138847080198641,
      "learning_rate": 1e-05,
      "loss": 1.4251,
      "step": 560
    },
    {
      "epoch": 0.38914419491892827,
      "grad_norm": 1.8462886417351692,
      "learning_rate": 1e-05,
      "loss": 1.398,
      "step": 561
    },
    {
      "epoch": 0.38983785658545045,
      "grad_norm": 2.076297894089803,
      "learning_rate": 1e-05,
      "loss": 1.3529,
      "step": 562
    },
    {
      "epoch": 0.3905315182519726,
      "grad_norm": 1.8009871240759965,
      "learning_rate": 1e-05,
      "loss": 1.3739,
      "step": 563
    },
    {
      "epoch": 0.39122517991849476,
      "grad_norm": 1.872935136959733,
      "learning_rate": 1e-05,
      "loss": 1.3842,
      "step": 564
    },
    {
      "epoch": 0.3919188415850169,
      "grad_norm": 1.8957537468315788,
      "learning_rate": 1e-05,
      "loss": 1.3633,
      "step": 565
    },
    {
      "epoch": 0.39261250325153907,
      "grad_norm": 1.991998761167588,
      "learning_rate": 1e-05,
      "loss": 1.3742,
      "step": 566
    },
    {
      "epoch": 0.3933061649180612,
      "grad_norm": 1.7016217238658489,
      "learning_rate": 1e-05,
      "loss": 1.4253,
      "step": 567
    },
    {
      "epoch": 0.3939998265845834,
      "grad_norm": 1.8800648961699629,
      "learning_rate": 1e-05,
      "loss": 1.4451,
      "step": 568
    },
    {
      "epoch": 0.3946934882511055,
      "grad_norm": 2.0019042611698774,
      "learning_rate": 1e-05,
      "loss": 1.3997,
      "step": 569
    },
    {
      "epoch": 0.3953871499176277,
      "grad_norm": 2.215323946324701,
      "learning_rate": 1e-05,
      "loss": 1.3577,
      "step": 570
    },
    {
      "epoch": 0.3960808115841498,
      "grad_norm": 2.0407346613187016,
      "learning_rate": 1e-05,
      "loss": 1.3808,
      "step": 571
    },
    {
      "epoch": 0.396774473250672,
      "grad_norm": 1.874235605467884,
      "learning_rate": 1e-05,
      "loss": 1.3883,
      "step": 572
    },
    {
      "epoch": 0.3974681349171941,
      "grad_norm": 1.9076714486170196,
      "learning_rate": 1e-05,
      "loss": 1.3785,
      "step": 573
    },
    {
      "epoch": 0.3981617965837163,
      "grad_norm": 1.9438453292084767,
      "learning_rate": 1e-05,
      "loss": 1.3888,
      "step": 574
    },
    {
      "epoch": 0.39885545825023844,
      "grad_norm": 1.9010804186629797,
      "learning_rate": 1e-05,
      "loss": 1.4044,
      "step": 575
    },
    {
      "epoch": 0.3995491199167606,
      "grad_norm": 2.1173409005527213,
      "learning_rate": 1e-05,
      "loss": 1.4043,
      "step": 576
    },
    {
      "epoch": 0.40024278158328275,
      "grad_norm": 1.924471913084561,
      "learning_rate": 1e-05,
      "loss": 1.3636,
      "step": 577
    },
    {
      "epoch": 0.40093644324980493,
      "grad_norm": 2.129867365540973,
      "learning_rate": 1e-05,
      "loss": 1.4308,
      "step": 578
    },
    {
      "epoch": 0.40163010491632706,
      "grad_norm": 1.9302914340159718,
      "learning_rate": 1e-05,
      "loss": 1.3937,
      "step": 579
    },
    {
      "epoch": 0.40232376658284924,
      "grad_norm": 1.978021194477141,
      "learning_rate": 1e-05,
      "loss": 1.4076,
      "step": 580
    },
    {
      "epoch": 0.40301742824937137,
      "grad_norm": 1.9801917793272694,
      "learning_rate": 1e-05,
      "loss": 1.3756,
      "step": 581
    },
    {
      "epoch": 0.40371108991589355,
      "grad_norm": 1.8510216673051632,
      "learning_rate": 1e-05,
      "loss": 1.3643,
      "step": 582
    },
    {
      "epoch": 0.4044047515824157,
      "grad_norm": 2.183109714099149,
      "learning_rate": 1e-05,
      "loss": 1.4537,
      "step": 583
    },
    {
      "epoch": 0.4050984132489378,
      "grad_norm": 2.349408814867446,
      "learning_rate": 1e-05,
      "loss": 1.4192,
      "step": 584
    },
    {
      "epoch": 0.40579207491546,
      "grad_norm": 2.2006000600981106,
      "learning_rate": 1e-05,
      "loss": 1.4079,
      "step": 585
    },
    {
      "epoch": 0.4064857365819821,
      "grad_norm": 1.8981822287744043,
      "learning_rate": 1e-05,
      "loss": 1.4069,
      "step": 586
    },
    {
      "epoch": 0.4071793982485043,
      "grad_norm": 1.760320500086242,
      "learning_rate": 1e-05,
      "loss": 1.3738,
      "step": 587
    },
    {
      "epoch": 0.4078730599150264,
      "grad_norm": 1.8312063491211514,
      "learning_rate": 1e-05,
      "loss": 1.3887,
      "step": 588
    },
    {
      "epoch": 0.4085667215815486,
      "grad_norm": 1.760204313519952,
      "learning_rate": 1e-05,
      "loss": 1.3332,
      "step": 589
    },
    {
      "epoch": 0.40926038324807074,
      "grad_norm": 1.7343834044934423,
      "learning_rate": 1e-05,
      "loss": 1.3435,
      "step": 590
    },
    {
      "epoch": 0.4099540449145929,
      "grad_norm": 2.2088224043427687,
      "learning_rate": 1e-05,
      "loss": 1.4162,
      "step": 591
    },
    {
      "epoch": 0.41064770658111505,
      "grad_norm": 1.8200743654853602,
      "learning_rate": 1e-05,
      "loss": 1.3931,
      "step": 592
    },
    {
      "epoch": 0.41134136824763723,
      "grad_norm": 1.8435586030279356,
      "learning_rate": 1e-05,
      "loss": 1.4093,
      "step": 593
    },
    {
      "epoch": 0.41203502991415936,
      "grad_norm": 1.916392749662975,
      "learning_rate": 1e-05,
      "loss": 1.3894,
      "step": 594
    },
    {
      "epoch": 0.41272869158068154,
      "grad_norm": 1.9161401196518564,
      "learning_rate": 1e-05,
      "loss": 1.3726,
      "step": 595
    },
    {
      "epoch": 0.41342235324720367,
      "grad_norm": 1.8756544580848145,
      "learning_rate": 1e-05,
      "loss": 1.3697,
      "step": 596
    },
    {
      "epoch": 0.41411601491372585,
      "grad_norm": 1.898787166818739,
      "learning_rate": 1e-05,
      "loss": 1.3941,
      "step": 597
    },
    {
      "epoch": 0.414809676580248,
      "grad_norm": 1.8044701136277606,
      "learning_rate": 1e-05,
      "loss": 1.3999,
      "step": 598
    },
    {
      "epoch": 0.41550333824677016,
      "grad_norm": 1.8310418953275842,
      "learning_rate": 1e-05,
      "loss": 1.3379,
      "step": 599
    },
    {
      "epoch": 0.4161969999132923,
      "grad_norm": 1.9227996432956476,
      "learning_rate": 1e-05,
      "loss": 1.362,
      "step": 600
    },
    {
      "epoch": 0.41689066157981447,
      "grad_norm": 1.8826360241413953,
      "learning_rate": 1e-05,
      "loss": 1.413,
      "step": 601
    },
    {
      "epoch": 0.4175843232463366,
      "grad_norm": 1.6984816355671049,
      "learning_rate": 1e-05,
      "loss": 1.3703,
      "step": 602
    },
    {
      "epoch": 0.4182779849128588,
      "grad_norm": 1.8568350133313958,
      "learning_rate": 1e-05,
      "loss": 1.3909,
      "step": 603
    },
    {
      "epoch": 0.4189716465793809,
      "grad_norm": 1.917415585499782,
      "learning_rate": 1e-05,
      "loss": 1.3698,
      "step": 604
    },
    {
      "epoch": 0.41966530824590303,
      "grad_norm": 1.8859603577321011,
      "learning_rate": 1e-05,
      "loss": 1.4034,
      "step": 605
    },
    {
      "epoch": 0.4203589699124252,
      "grad_norm": 1.8889156912270977,
      "learning_rate": 1e-05,
      "loss": 1.4125,
      "step": 606
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 1.911197022952282,
      "learning_rate": 1e-05,
      "loss": 1.3655,
      "step": 607
    },
    {
      "epoch": 0.42174629324546953,
      "grad_norm": 1.9782065265119402,
      "learning_rate": 1e-05,
      "loss": 1.4201,
      "step": 608
    },
    {
      "epoch": 0.42243995491199166,
      "grad_norm": 1.8747816354205493,
      "learning_rate": 1e-05,
      "loss": 1.3664,
      "step": 609
    },
    {
      "epoch": 0.42313361657851384,
      "grad_norm": 1.9012852889769853,
      "learning_rate": 1e-05,
      "loss": 1.3259,
      "step": 610
    },
    {
      "epoch": 0.42382727824503597,
      "grad_norm": 1.8979118524613148,
      "learning_rate": 1e-05,
      "loss": 1.3971,
      "step": 611
    },
    {
      "epoch": 0.42452093991155815,
      "grad_norm": 1.8520667588064368,
      "learning_rate": 1e-05,
      "loss": 1.4135,
      "step": 612
    },
    {
      "epoch": 0.4252146015780803,
      "grad_norm": 1.8342529385215973,
      "learning_rate": 1e-05,
      "loss": 1.3863,
      "step": 613
    },
    {
      "epoch": 0.42590826324460246,
      "grad_norm": 1.8295048253836184,
      "learning_rate": 1e-05,
      "loss": 1.3087,
      "step": 614
    },
    {
      "epoch": 0.4266019249111246,
      "grad_norm": 1.7695063050348687,
      "learning_rate": 1e-05,
      "loss": 1.4162,
      "step": 615
    },
    {
      "epoch": 0.42729558657764677,
      "grad_norm": 2.082777123696616,
      "learning_rate": 1e-05,
      "loss": 1.4278,
      "step": 616
    },
    {
      "epoch": 0.4279892482441689,
      "grad_norm": 1.7051067845957866,
      "learning_rate": 1e-05,
      "loss": 1.4095,
      "step": 617
    },
    {
      "epoch": 0.4286829099106911,
      "grad_norm": 1.8695270037035383,
      "learning_rate": 1e-05,
      "loss": 1.4503,
      "step": 618
    },
    {
      "epoch": 0.4293765715772132,
      "grad_norm": 1.9281893067578681,
      "learning_rate": 1e-05,
      "loss": 1.38,
      "step": 619
    },
    {
      "epoch": 0.4300702332437354,
      "grad_norm": 1.777014355958319,
      "learning_rate": 1e-05,
      "loss": 1.3984,
      "step": 620
    },
    {
      "epoch": 0.4307638949102575,
      "grad_norm": 1.842917760375874,
      "learning_rate": 1e-05,
      "loss": 1.3853,
      "step": 621
    },
    {
      "epoch": 0.4314575565767797,
      "grad_norm": 1.8267047634185034,
      "learning_rate": 1e-05,
      "loss": 1.3524,
      "step": 622
    },
    {
      "epoch": 0.4321512182433018,
      "grad_norm": 2.0773027372115394,
      "learning_rate": 1e-05,
      "loss": 1.3901,
      "step": 623
    },
    {
      "epoch": 0.432844879909824,
      "grad_norm": 1.8770974629057278,
      "learning_rate": 1e-05,
      "loss": 1.376,
      "step": 624
    },
    {
      "epoch": 0.43353854157634614,
      "grad_norm": 1.8429345410006643,
      "learning_rate": 1e-05,
      "loss": 1.3877,
      "step": 625
    },
    {
      "epoch": 0.43423220324286826,
      "grad_norm": 1.7713210328211393,
      "learning_rate": 1e-05,
      "loss": 1.3976,
      "step": 626
    },
    {
      "epoch": 0.43492586490939045,
      "grad_norm": 1.8286568674310186,
      "learning_rate": 1e-05,
      "loss": 1.3465,
      "step": 627
    },
    {
      "epoch": 0.4356195265759126,
      "grad_norm": 1.8690248103614588,
      "learning_rate": 1e-05,
      "loss": 1.3826,
      "step": 628
    },
    {
      "epoch": 0.43631318824243476,
      "grad_norm": 2.025467675895459,
      "learning_rate": 1e-05,
      "loss": 1.3671,
      "step": 629
    },
    {
      "epoch": 0.4370068499089569,
      "grad_norm": 1.8903327597767532,
      "learning_rate": 1e-05,
      "loss": 1.3548,
      "step": 630
    },
    {
      "epoch": 0.43770051157547907,
      "grad_norm": 1.7825184036448585,
      "learning_rate": 1e-05,
      "loss": 1.3978,
      "step": 631
    },
    {
      "epoch": 0.4383941732420012,
      "grad_norm": 1.9191594921321666,
      "learning_rate": 1e-05,
      "loss": 1.4214,
      "step": 632
    },
    {
      "epoch": 0.4390878349085234,
      "grad_norm": 1.8710466935982961,
      "learning_rate": 1e-05,
      "loss": 1.4685,
      "step": 633
    },
    {
      "epoch": 0.4397814965750455,
      "grad_norm": 1.9252588373442325,
      "learning_rate": 1e-05,
      "loss": 1.3529,
      "step": 634
    },
    {
      "epoch": 0.4404751582415677,
      "grad_norm": 1.8307810199559515,
      "learning_rate": 1e-05,
      "loss": 1.379,
      "step": 635
    },
    {
      "epoch": 0.4411688199080898,
      "grad_norm": 1.9458457674285412,
      "learning_rate": 1e-05,
      "loss": 1.3734,
      "step": 636
    },
    {
      "epoch": 0.441862481574612,
      "grad_norm": 1.8137288582058262,
      "learning_rate": 1e-05,
      "loss": 1.3309,
      "step": 637
    },
    {
      "epoch": 0.4425561432411341,
      "grad_norm": 1.972371773276373,
      "learning_rate": 1e-05,
      "loss": 1.4303,
      "step": 638
    },
    {
      "epoch": 0.4432498049076563,
      "grad_norm": 1.9163578588020878,
      "learning_rate": 1e-05,
      "loss": 1.3191,
      "step": 639
    },
    {
      "epoch": 0.44394346657417844,
      "grad_norm": 1.8861727336930172,
      "learning_rate": 1e-05,
      "loss": 1.374,
      "step": 640
    },
    {
      "epoch": 0.4446371282407006,
      "grad_norm": 1.7493814291306995,
      "learning_rate": 1e-05,
      "loss": 1.3992,
      "step": 641
    },
    {
      "epoch": 0.44533078990722275,
      "grad_norm": 1.8289477836679608,
      "learning_rate": 1e-05,
      "loss": 1.3472,
      "step": 642
    },
    {
      "epoch": 0.44602445157374493,
      "grad_norm": 1.9323084827259125,
      "learning_rate": 1e-05,
      "loss": 1.3966,
      "step": 643
    },
    {
      "epoch": 0.44671811324026706,
      "grad_norm": 1.866327691734681,
      "learning_rate": 1e-05,
      "loss": 1.3793,
      "step": 644
    },
    {
      "epoch": 0.44741177490678924,
      "grad_norm": 1.8217937158784923,
      "learning_rate": 1e-05,
      "loss": 1.4244,
      "step": 645
    },
    {
      "epoch": 0.44810543657331137,
      "grad_norm": 1.8364372691697852,
      "learning_rate": 1e-05,
      "loss": 1.3755,
      "step": 646
    },
    {
      "epoch": 0.4487990982398335,
      "grad_norm": 2.013774748739019,
      "learning_rate": 1e-05,
      "loss": 1.3807,
      "step": 647
    },
    {
      "epoch": 0.4494927599063557,
      "grad_norm": 1.805702981754065,
      "learning_rate": 1e-05,
      "loss": 1.3942,
      "step": 648
    },
    {
      "epoch": 0.4501864215728778,
      "grad_norm": 1.9948282199322953,
      "learning_rate": 1e-05,
      "loss": 1.4366,
      "step": 649
    },
    {
      "epoch": 0.4508800832394,
      "grad_norm": 1.978073570947312,
      "learning_rate": 1e-05,
      "loss": 1.4254,
      "step": 650
    },
    {
      "epoch": 0.4515737449059221,
      "grad_norm": 1.9232595674423625,
      "learning_rate": 1e-05,
      "loss": 1.404,
      "step": 651
    },
    {
      "epoch": 0.4522674065724443,
      "grad_norm": 1.855166028756208,
      "learning_rate": 1e-05,
      "loss": 1.3791,
      "step": 652
    },
    {
      "epoch": 0.4529610682389664,
      "grad_norm": 1.7786637265688048,
      "learning_rate": 1e-05,
      "loss": 1.3936,
      "step": 653
    },
    {
      "epoch": 0.4536547299054886,
      "grad_norm": 1.6650125213330889,
      "learning_rate": 1e-05,
      "loss": 1.38,
      "step": 654
    },
    {
      "epoch": 0.45434839157201073,
      "grad_norm": 1.9212024581799898,
      "learning_rate": 1e-05,
      "loss": 1.3951,
      "step": 655
    },
    {
      "epoch": 0.4550420532385329,
      "grad_norm": 1.855795867809702,
      "learning_rate": 1e-05,
      "loss": 1.3563,
      "step": 656
    },
    {
      "epoch": 0.45573571490505504,
      "grad_norm": 1.8769724131475898,
      "learning_rate": 1e-05,
      "loss": 1.3974,
      "step": 657
    },
    {
      "epoch": 0.4564293765715772,
      "grad_norm": 1.7926551468562113,
      "learning_rate": 1e-05,
      "loss": 1.3974,
      "step": 658
    },
    {
      "epoch": 0.45712303823809936,
      "grad_norm": 1.9178863763832097,
      "learning_rate": 1e-05,
      "loss": 1.4146,
      "step": 659
    },
    {
      "epoch": 0.45781669990462154,
      "grad_norm": 1.9353453542577745,
      "learning_rate": 1e-05,
      "loss": 1.4187,
      "step": 660
    },
    {
      "epoch": 0.45851036157114367,
      "grad_norm": 1.952403141808128,
      "learning_rate": 1e-05,
      "loss": 1.4008,
      "step": 661
    },
    {
      "epoch": 0.45920402323766585,
      "grad_norm": 1.757904596732749,
      "learning_rate": 1e-05,
      "loss": 1.4264,
      "step": 662
    },
    {
      "epoch": 0.459897684904188,
      "grad_norm": 1.9395774430498347,
      "learning_rate": 1e-05,
      "loss": 1.3671,
      "step": 663
    },
    {
      "epoch": 0.46059134657071016,
      "grad_norm": 1.8890203107447723,
      "learning_rate": 1e-05,
      "loss": 1.3715,
      "step": 664
    },
    {
      "epoch": 0.4612850082372323,
      "grad_norm": 1.8286704559781526,
      "learning_rate": 1e-05,
      "loss": 1.3941,
      "step": 665
    },
    {
      "epoch": 0.46197866990375447,
      "grad_norm": 1.810001367913328,
      "learning_rate": 1e-05,
      "loss": 1.3883,
      "step": 666
    },
    {
      "epoch": 0.4626723315702766,
      "grad_norm": 1.9152160013729407,
      "learning_rate": 1e-05,
      "loss": 1.3813,
      "step": 667
    },
    {
      "epoch": 0.4633659932367988,
      "grad_norm": 2.0094067199071546,
      "learning_rate": 1e-05,
      "loss": 1.3176,
      "step": 668
    },
    {
      "epoch": 0.4640596549033209,
      "grad_norm": 1.8734070756332013,
      "learning_rate": 1e-05,
      "loss": 1.4233,
      "step": 669
    },
    {
      "epoch": 0.46475331656984303,
      "grad_norm": 1.7612782018456719,
      "learning_rate": 1e-05,
      "loss": 1.3438,
      "step": 670
    },
    {
      "epoch": 0.4654469782363652,
      "grad_norm": 1.8683597363973128,
      "learning_rate": 1e-05,
      "loss": 1.355,
      "step": 671
    },
    {
      "epoch": 0.46614063990288734,
      "grad_norm": 2.0012042847679914,
      "learning_rate": 1e-05,
      "loss": 1.4169,
      "step": 672
    },
    {
      "epoch": 0.4668343015694095,
      "grad_norm": 1.9416276311234992,
      "learning_rate": 1e-05,
      "loss": 1.3742,
      "step": 673
    },
    {
      "epoch": 0.46752796323593165,
      "grad_norm": 1.8922567854833168,
      "learning_rate": 1e-05,
      "loss": 1.3779,
      "step": 674
    },
    {
      "epoch": 0.46822162490245384,
      "grad_norm": 1.7793620481094168,
      "learning_rate": 1e-05,
      "loss": 1.3915,
      "step": 675
    },
    {
      "epoch": 0.46891528656897596,
      "grad_norm": 2.167051856102186,
      "learning_rate": 1e-05,
      "loss": 1.3989,
      "step": 676
    },
    {
      "epoch": 0.46960894823549815,
      "grad_norm": 1.871191468625491,
      "learning_rate": 1e-05,
      "loss": 1.3547,
      "step": 677
    },
    {
      "epoch": 0.4703026099020203,
      "grad_norm": 1.7304516085856814,
      "learning_rate": 1e-05,
      "loss": 1.3808,
      "step": 678
    },
    {
      "epoch": 0.47099627156854246,
      "grad_norm": 1.745983949797038,
      "learning_rate": 1e-05,
      "loss": 1.3425,
      "step": 679
    },
    {
      "epoch": 0.4716899332350646,
      "grad_norm": 1.8779422010813842,
      "learning_rate": 1e-05,
      "loss": 1.3592,
      "step": 680
    },
    {
      "epoch": 0.47238359490158677,
      "grad_norm": 1.9934611197767778,
      "learning_rate": 1e-05,
      "loss": 1.3538,
      "step": 681
    },
    {
      "epoch": 0.4730772565681089,
      "grad_norm": 1.9623233741400323,
      "learning_rate": 1e-05,
      "loss": 1.346,
      "step": 682
    },
    {
      "epoch": 0.4737709182346311,
      "grad_norm": 1.796679910714133,
      "learning_rate": 1e-05,
      "loss": 1.355,
      "step": 683
    },
    {
      "epoch": 0.4744645799011532,
      "grad_norm": 1.7722903042578,
      "learning_rate": 1e-05,
      "loss": 1.4071,
      "step": 684
    },
    {
      "epoch": 0.4751582415676754,
      "grad_norm": 1.8052196535090335,
      "learning_rate": 1e-05,
      "loss": 1.3814,
      "step": 685
    },
    {
      "epoch": 0.4758519032341975,
      "grad_norm": 1.7948553541752041,
      "learning_rate": 1e-05,
      "loss": 1.3729,
      "step": 686
    },
    {
      "epoch": 0.4765455649007197,
      "grad_norm": 2.007593236556849,
      "learning_rate": 1e-05,
      "loss": 1.3673,
      "step": 687
    },
    {
      "epoch": 0.4772392265672418,
      "grad_norm": 1.9797683213286288,
      "learning_rate": 1e-05,
      "loss": 1.3629,
      "step": 688
    },
    {
      "epoch": 0.477932888233764,
      "grad_norm": 1.9548220267121557,
      "learning_rate": 1e-05,
      "loss": 1.383,
      "step": 689
    },
    {
      "epoch": 0.47862654990028614,
      "grad_norm": 1.929009442664775,
      "learning_rate": 1e-05,
      "loss": 1.3921,
      "step": 690
    },
    {
      "epoch": 0.47932021156680826,
      "grad_norm": 1.8421018984754722,
      "learning_rate": 1e-05,
      "loss": 1.3672,
      "step": 691
    },
    {
      "epoch": 0.48001387323333045,
      "grad_norm": 1.8234045076467948,
      "learning_rate": 1e-05,
      "loss": 1.398,
      "step": 692
    },
    {
      "epoch": 0.4807075348998526,
      "grad_norm": 2.066405568328085,
      "learning_rate": 1e-05,
      "loss": 1.3348,
      "step": 693
    },
    {
      "epoch": 0.48140119656637476,
      "grad_norm": 1.9963663982661528,
      "learning_rate": 1e-05,
      "loss": 1.3662,
      "step": 694
    },
    {
      "epoch": 0.4820948582328969,
      "grad_norm": 1.883319503377311,
      "learning_rate": 1e-05,
      "loss": 1.3835,
      "step": 695
    },
    {
      "epoch": 0.48278851989941907,
      "grad_norm": 1.56186845290527,
      "learning_rate": 1e-05,
      "loss": 1.3386,
      "step": 696
    },
    {
      "epoch": 0.4834821815659412,
      "grad_norm": 1.8687876803046082,
      "learning_rate": 1e-05,
      "loss": 1.3501,
      "step": 697
    },
    {
      "epoch": 0.4841758432324634,
      "grad_norm": 2.0432055137936036,
      "learning_rate": 1e-05,
      "loss": 1.366,
      "step": 698
    },
    {
      "epoch": 0.4848695048989855,
      "grad_norm": 1.7989863129663937,
      "learning_rate": 1e-05,
      "loss": 1.387,
      "step": 699
    },
    {
      "epoch": 0.4855631665655077,
      "grad_norm": 1.9153016999718706,
      "learning_rate": 1e-05,
      "loss": 1.4391,
      "step": 700
    },
    {
      "epoch": 0.4862568282320298,
      "grad_norm": 1.8256041578604432,
      "learning_rate": 1e-05,
      "loss": 1.385,
      "step": 701
    },
    {
      "epoch": 0.486950489898552,
      "grad_norm": 1.8391212949021127,
      "learning_rate": 1e-05,
      "loss": 1.388,
      "step": 702
    },
    {
      "epoch": 0.4876441515650741,
      "grad_norm": 1.9978982769762854,
      "learning_rate": 1e-05,
      "loss": 1.3841,
      "step": 703
    },
    {
      "epoch": 0.4883378132315963,
      "grad_norm": 1.9247875553761613,
      "learning_rate": 1e-05,
      "loss": 1.3514,
      "step": 704
    },
    {
      "epoch": 0.48903147489811843,
      "grad_norm": 1.8114868285875125,
      "learning_rate": 1e-05,
      "loss": 1.3659,
      "step": 705
    },
    {
      "epoch": 0.4897251365646406,
      "grad_norm": 1.8506801571462923,
      "learning_rate": 1e-05,
      "loss": 1.4349,
      "step": 706
    },
    {
      "epoch": 0.49041879823116274,
      "grad_norm": 1.8480926190821445,
      "learning_rate": 1e-05,
      "loss": 1.3768,
      "step": 707
    },
    {
      "epoch": 0.4911124598976849,
      "grad_norm": 1.9547558985555982,
      "learning_rate": 1e-05,
      "loss": 1.3767,
      "step": 708
    },
    {
      "epoch": 0.49180612156420705,
      "grad_norm": 1.7241605811559706,
      "learning_rate": 1e-05,
      "loss": 1.4277,
      "step": 709
    },
    {
      "epoch": 0.49249978323072924,
      "grad_norm": 1.9597827919247104,
      "learning_rate": 1e-05,
      "loss": 1.2981,
      "step": 710
    },
    {
      "epoch": 0.49319344489725137,
      "grad_norm": 2.241085697740784,
      "learning_rate": 1e-05,
      "loss": 1.3738,
      "step": 711
    },
    {
      "epoch": 0.4938871065637735,
      "grad_norm": 2.1107944914928023,
      "learning_rate": 1e-05,
      "loss": 1.3788,
      "step": 712
    },
    {
      "epoch": 0.4945807682302957,
      "grad_norm": 1.9819856716802244,
      "learning_rate": 1e-05,
      "loss": 1.3805,
      "step": 713
    },
    {
      "epoch": 0.4952744298968178,
      "grad_norm": 1.7818814238988587,
      "learning_rate": 1e-05,
      "loss": 1.3864,
      "step": 714
    },
    {
      "epoch": 0.49596809156334,
      "grad_norm": 1.8028425771808247,
      "learning_rate": 1e-05,
      "loss": 1.4047,
      "step": 715
    },
    {
      "epoch": 0.4966617532298621,
      "grad_norm": 1.810766896499592,
      "learning_rate": 1e-05,
      "loss": 1.3899,
      "step": 716
    },
    {
      "epoch": 0.4973554148963843,
      "grad_norm": 1.8166160127003885,
      "learning_rate": 1e-05,
      "loss": 1.3603,
      "step": 717
    },
    {
      "epoch": 0.4980490765629064,
      "grad_norm": 2.13361803754948,
      "learning_rate": 1e-05,
      "loss": 1.4022,
      "step": 718
    },
    {
      "epoch": 0.4987427382294286,
      "grad_norm": 2.016997065796454,
      "learning_rate": 1e-05,
      "loss": 1.4034,
      "step": 719
    },
    {
      "epoch": 0.49943639989595073,
      "grad_norm": 2.0748032742705793,
      "learning_rate": 1e-05,
      "loss": 1.4105,
      "step": 720
    },
    {
      "epoch": 0.5001300615624729,
      "grad_norm": 1.8523948307863285,
      "learning_rate": 1e-05,
      "loss": 1.3697,
      "step": 721
    },
    {
      "epoch": 0.500823723228995,
      "grad_norm": 1.8122879610484914,
      "learning_rate": 1e-05,
      "loss": 1.404,
      "step": 722
    },
    {
      "epoch": 0.5015173848955172,
      "grad_norm": 1.8832267152617044,
      "learning_rate": 1e-05,
      "loss": 1.412,
      "step": 723
    },
    {
      "epoch": 0.5022110465620394,
      "grad_norm": 1.8257618111117269,
      "learning_rate": 1e-05,
      "loss": 1.4016,
      "step": 724
    },
    {
      "epoch": 0.5029047082285615,
      "grad_norm": 1.7942051170674276,
      "learning_rate": 1e-05,
      "loss": 1.3686,
      "step": 725
    },
    {
      "epoch": 0.5035983698950837,
      "grad_norm": 1.97042982772308,
      "learning_rate": 1e-05,
      "loss": 1.3575,
      "step": 726
    },
    {
      "epoch": 0.5042920315616058,
      "grad_norm": 2.053740555710897,
      "learning_rate": 1e-05,
      "loss": 1.379,
      "step": 727
    },
    {
      "epoch": 0.504985693228128,
      "grad_norm": 1.977833812394873,
      "learning_rate": 1e-05,
      "loss": 1.3174,
      "step": 728
    },
    {
      "epoch": 0.5056793548946501,
      "grad_norm": 1.9322288272504544,
      "learning_rate": 1e-05,
      "loss": 1.4038,
      "step": 729
    },
    {
      "epoch": 0.5063730165611723,
      "grad_norm": 2.0400420303295275,
      "learning_rate": 1e-05,
      "loss": 1.3953,
      "step": 730
    },
    {
      "epoch": 0.5070666782276945,
      "grad_norm": 1.976799409278783,
      "learning_rate": 1e-05,
      "loss": 1.3895,
      "step": 731
    },
    {
      "epoch": 0.5077603398942167,
      "grad_norm": 1.9306400432028785,
      "learning_rate": 1e-05,
      "loss": 1.4228,
      "step": 732
    },
    {
      "epoch": 0.5084540015607387,
      "grad_norm": 1.7347175465813025,
      "learning_rate": 1e-05,
      "loss": 1.3704,
      "step": 733
    },
    {
      "epoch": 0.5091476632272609,
      "grad_norm": 1.7180782433090167,
      "learning_rate": 1e-05,
      "loss": 1.3547,
      "step": 734
    },
    {
      "epoch": 0.5098413248937831,
      "grad_norm": 1.852662073693016,
      "learning_rate": 1e-05,
      "loss": 1.4228,
      "step": 735
    },
    {
      "epoch": 0.5105349865603052,
      "grad_norm": 1.8075356657945834,
      "learning_rate": 1e-05,
      "loss": 1.3887,
      "step": 736
    },
    {
      "epoch": 0.5112286482268273,
      "grad_norm": 1.7367287190868546,
      "learning_rate": 1e-05,
      "loss": 1.3752,
      "step": 737
    },
    {
      "epoch": 0.5119223098933495,
      "grad_norm": 1.8286379353529887,
      "learning_rate": 1e-05,
      "loss": 1.3963,
      "step": 738
    },
    {
      "epoch": 0.5126159715598717,
      "grad_norm": 1.7825873791247597,
      "learning_rate": 1e-05,
      "loss": 1.397,
      "step": 739
    },
    {
      "epoch": 0.5133096332263938,
      "grad_norm": 2.0430860649068396,
      "learning_rate": 1e-05,
      "loss": 1.368,
      "step": 740
    },
    {
      "epoch": 0.514003294892916,
      "grad_norm": 1.82138787437088,
      "learning_rate": 1e-05,
      "loss": 1.3888,
      "step": 741
    },
    {
      "epoch": 0.5146969565594381,
      "grad_norm": 1.9949710877946953,
      "learning_rate": 1e-05,
      "loss": 1.3221,
      "step": 742
    },
    {
      "epoch": 0.5153906182259603,
      "grad_norm": 1.7999499915500219,
      "learning_rate": 1e-05,
      "loss": 1.3214,
      "step": 743
    },
    {
      "epoch": 0.5160842798924824,
      "grad_norm": 1.944101461453592,
      "learning_rate": 1e-05,
      "loss": 1.3641,
      "step": 744
    },
    {
      "epoch": 0.5167779415590046,
      "grad_norm": 1.9484142602052452,
      "learning_rate": 1e-05,
      "loss": 1.4183,
      "step": 745
    },
    {
      "epoch": 0.5174716032255268,
      "grad_norm": 1.9261361736136446,
      "learning_rate": 1e-05,
      "loss": 1.39,
      "step": 746
    },
    {
      "epoch": 0.518165264892049,
      "grad_norm": 1.873564918154059,
      "learning_rate": 1e-05,
      "loss": 1.3511,
      "step": 747
    },
    {
      "epoch": 0.518858926558571,
      "grad_norm": 1.975733457802349,
      "learning_rate": 1e-05,
      "loss": 1.3786,
      "step": 748
    },
    {
      "epoch": 0.5195525882250932,
      "grad_norm": 1.9396312909048352,
      "learning_rate": 1e-05,
      "loss": 1.3161,
      "step": 749
    },
    {
      "epoch": 0.5202462498916154,
      "grad_norm": 1.736368973000078,
      "learning_rate": 1e-05,
      "loss": 1.3704,
      "step": 750
    },
    {
      "epoch": 0.5209399115581376,
      "grad_norm": 1.7911382730465684,
      "learning_rate": 1e-05,
      "loss": 1.3379,
      "step": 751
    },
    {
      "epoch": 0.5216335732246596,
      "grad_norm": 1.8590253300206483,
      "learning_rate": 1e-05,
      "loss": 1.3531,
      "step": 752
    },
    {
      "epoch": 0.5223272348911818,
      "grad_norm": 1.9765690927121422,
      "learning_rate": 1e-05,
      "loss": 1.3728,
      "step": 753
    },
    {
      "epoch": 0.523020896557704,
      "grad_norm": 2.0025745088354148,
      "learning_rate": 1e-05,
      "loss": 1.3939,
      "step": 754
    },
    {
      "epoch": 0.5237145582242261,
      "grad_norm": 1.9192921819712603,
      "learning_rate": 1e-05,
      "loss": 1.3914,
      "step": 755
    },
    {
      "epoch": 0.5244082198907483,
      "grad_norm": 1.8600023632893699,
      "learning_rate": 1e-05,
      "loss": 1.3499,
      "step": 756
    },
    {
      "epoch": 0.5251018815572704,
      "grad_norm": 1.9355189855324415,
      "learning_rate": 1e-05,
      "loss": 1.3471,
      "step": 757
    },
    {
      "epoch": 0.5257955432237926,
      "grad_norm": 2.0191814963451495,
      "learning_rate": 1e-05,
      "loss": 1.4244,
      "step": 758
    },
    {
      "epoch": 0.5264892048903147,
      "grad_norm": 1.883726645686763,
      "learning_rate": 1e-05,
      "loss": 1.3619,
      "step": 759
    },
    {
      "epoch": 0.5271828665568369,
      "grad_norm": 1.7947607621987598,
      "learning_rate": 1e-05,
      "loss": 1.3404,
      "step": 760
    },
    {
      "epoch": 0.5278765282233591,
      "grad_norm": 1.8889256123493143,
      "learning_rate": 1e-05,
      "loss": 1.4012,
      "step": 761
    },
    {
      "epoch": 0.5285701898898812,
      "grad_norm": 1.9230513498083945,
      "learning_rate": 1e-05,
      "loss": 1.3475,
      "step": 762
    },
    {
      "epoch": 0.5292638515564033,
      "grad_norm": 1.8067019812009846,
      "learning_rate": 1e-05,
      "loss": 1.3911,
      "step": 763
    },
    {
      "epoch": 0.5299575132229255,
      "grad_norm": 1.8380178891570427,
      "learning_rate": 1e-05,
      "loss": 1.3614,
      "step": 764
    },
    {
      "epoch": 0.5306511748894477,
      "grad_norm": 1.8235814437026021,
      "learning_rate": 1e-05,
      "loss": 1.3556,
      "step": 765
    },
    {
      "epoch": 0.5313448365559699,
      "grad_norm": 1.830617817421823,
      "learning_rate": 1e-05,
      "loss": 1.3844,
      "step": 766
    },
    {
      "epoch": 0.5320384982224919,
      "grad_norm": 1.9168498268351752,
      "learning_rate": 1e-05,
      "loss": 1.3669,
      "step": 767
    },
    {
      "epoch": 0.5327321598890141,
      "grad_norm": 1.945357782268639,
      "learning_rate": 1e-05,
      "loss": 1.3595,
      "step": 768
    },
    {
      "epoch": 0.5334258215555363,
      "grad_norm": 2.237537090117557,
      "learning_rate": 1e-05,
      "loss": 1.3683,
      "step": 769
    },
    {
      "epoch": 0.5341194832220585,
      "grad_norm": 1.8612240335487669,
      "learning_rate": 1e-05,
      "loss": 1.3937,
      "step": 770
    },
    {
      "epoch": 0.5348131448885806,
      "grad_norm": 2.0804814565765266,
      "learning_rate": 1e-05,
      "loss": 1.3706,
      "step": 771
    },
    {
      "epoch": 0.5355068065551027,
      "grad_norm": 1.9951415692701584,
      "learning_rate": 1e-05,
      "loss": 1.3696,
      "step": 772
    },
    {
      "epoch": 0.5362004682216249,
      "grad_norm": 1.7248104660180126,
      "learning_rate": 1e-05,
      "loss": 1.3747,
      "step": 773
    },
    {
      "epoch": 0.5368941298881471,
      "grad_norm": 1.7711620914058126,
      "learning_rate": 1e-05,
      "loss": 1.3776,
      "step": 774
    },
    {
      "epoch": 0.5375877915546692,
      "grad_norm": 1.7502844737194139,
      "learning_rate": 1e-05,
      "loss": 1.3719,
      "step": 775
    },
    {
      "epoch": 0.5382814532211914,
      "grad_norm": 1.975837716180595,
      "learning_rate": 1e-05,
      "loss": 1.3392,
      "step": 776
    },
    {
      "epoch": 0.5389751148877135,
      "grad_norm": 2.079903617254406,
      "learning_rate": 1e-05,
      "loss": 1.3789,
      "step": 777
    },
    {
      "epoch": 0.5396687765542356,
      "grad_norm": 1.8776752694265728,
      "learning_rate": 1e-05,
      "loss": 1.3968,
      "step": 778
    },
    {
      "epoch": 0.5403624382207578,
      "grad_norm": 1.7524287384136423,
      "learning_rate": 1e-05,
      "loss": 1.3538,
      "step": 779
    },
    {
      "epoch": 0.54105609988728,
      "grad_norm": 2.034871602096054,
      "learning_rate": 1e-05,
      "loss": 1.3546,
      "step": 780
    },
    {
      "epoch": 0.5417497615538022,
      "grad_norm": 2.0262231787193117,
      "learning_rate": 1e-05,
      "loss": 1.351,
      "step": 781
    },
    {
      "epoch": 0.5424434232203242,
      "grad_norm": 1.8800790132697096,
      "learning_rate": 1e-05,
      "loss": 1.4063,
      "step": 782
    },
    {
      "epoch": 0.5431370848868464,
      "grad_norm": 1.8397602468821888,
      "learning_rate": 1e-05,
      "loss": 1.3284,
      "step": 783
    },
    {
      "epoch": 0.5438307465533686,
      "grad_norm": 1.7812165456068076,
      "learning_rate": 1e-05,
      "loss": 1.3116,
      "step": 784
    },
    {
      "epoch": 0.5445244082198908,
      "grad_norm": 1.7794765249417972,
      "learning_rate": 1e-05,
      "loss": 1.3796,
      "step": 785
    },
    {
      "epoch": 0.5452180698864129,
      "grad_norm": 1.9715381587594996,
      "learning_rate": 1e-05,
      "loss": 1.3771,
      "step": 786
    },
    {
      "epoch": 0.545911731552935,
      "grad_norm": 1.7931337412184942,
      "learning_rate": 1e-05,
      "loss": 1.4108,
      "step": 787
    },
    {
      "epoch": 0.5466053932194572,
      "grad_norm": 1.8048639309386612,
      "learning_rate": 1e-05,
      "loss": 1.3689,
      "step": 788
    },
    {
      "epoch": 0.5472990548859794,
      "grad_norm": 1.8622884245084281,
      "learning_rate": 1e-05,
      "loss": 1.3513,
      "step": 789
    },
    {
      "epoch": 0.5479927165525015,
      "grad_norm": 1.816207257972127,
      "learning_rate": 1e-05,
      "loss": 1.3331,
      "step": 790
    },
    {
      "epoch": 0.5486863782190237,
      "grad_norm": 1.7508374499546198,
      "learning_rate": 1e-05,
      "loss": 1.3594,
      "step": 791
    },
    {
      "epoch": 0.5493800398855458,
      "grad_norm": 1.7086134745690817,
      "learning_rate": 1e-05,
      "loss": 1.343,
      "step": 792
    },
    {
      "epoch": 0.550073701552068,
      "grad_norm": 1.8826776634822224,
      "learning_rate": 1e-05,
      "loss": 1.4312,
      "step": 793
    },
    {
      "epoch": 0.5507673632185901,
      "grad_norm": 1.85161413953065,
      "learning_rate": 1e-05,
      "loss": 1.403,
      "step": 794
    },
    {
      "epoch": 0.5514610248851123,
      "grad_norm": 2.169026140246512,
      "learning_rate": 1e-05,
      "loss": 1.3514,
      "step": 795
    },
    {
      "epoch": 0.5521546865516345,
      "grad_norm": 1.8974393651417674,
      "learning_rate": 1e-05,
      "loss": 1.376,
      "step": 796
    },
    {
      "epoch": 0.5528483482181566,
      "grad_norm": 1.7818922230701455,
      "learning_rate": 1e-05,
      "loss": 1.3365,
      "step": 797
    },
    {
      "epoch": 0.5535420098846787,
      "grad_norm": 1.8268621071764144,
      "learning_rate": 1e-05,
      "loss": 1.3267,
      "step": 798
    },
    {
      "epoch": 0.5542356715512009,
      "grad_norm": 1.741645703166344,
      "learning_rate": 1e-05,
      "loss": 1.3401,
      "step": 799
    },
    {
      "epoch": 0.5549293332177231,
      "grad_norm": 1.7684014664169725,
      "learning_rate": 1e-05,
      "loss": 1.3453,
      "step": 800
    },
    {
      "epoch": 0.5556229948842452,
      "grad_norm": 1.872082247187476,
      "learning_rate": 1e-05,
      "loss": 1.3849,
      "step": 801
    },
    {
      "epoch": 0.5563166565507673,
      "grad_norm": 1.843965544279036,
      "learning_rate": 1e-05,
      "loss": 1.3515,
      "step": 802
    },
    {
      "epoch": 0.5570103182172895,
      "grad_norm": 1.7503097471047913,
      "learning_rate": 1e-05,
      "loss": 1.3772,
      "step": 803
    },
    {
      "epoch": 0.5577039798838117,
      "grad_norm": 2.152027612117084,
      "learning_rate": 1e-05,
      "loss": 1.3369,
      "step": 804
    },
    {
      "epoch": 0.5583976415503338,
      "grad_norm": 1.9678830918650134,
      "learning_rate": 1e-05,
      "loss": 1.3363,
      "step": 805
    },
    {
      "epoch": 0.559091303216856,
      "grad_norm": 1.7813746189463364,
      "learning_rate": 1e-05,
      "loss": 1.3268,
      "step": 806
    },
    {
      "epoch": 0.5597849648833781,
      "grad_norm": 1.835262281512248,
      "learning_rate": 1e-05,
      "loss": 1.3155,
      "step": 807
    },
    {
      "epoch": 0.5604786265499003,
      "grad_norm": 1.778985993964162,
      "learning_rate": 1e-05,
      "loss": 1.3913,
      "step": 808
    },
    {
      "epoch": 0.5611722882164224,
      "grad_norm": 2.07464881345254,
      "learning_rate": 1e-05,
      "loss": 1.35,
      "step": 809
    },
    {
      "epoch": 0.5618659498829446,
      "grad_norm": 2.186610515953725,
      "learning_rate": 1e-05,
      "loss": 1.3954,
      "step": 810
    },
    {
      "epoch": 0.5625596115494668,
      "grad_norm": 2.129937538785471,
      "learning_rate": 1e-05,
      "loss": 1.3644,
      "step": 811
    },
    {
      "epoch": 0.563253273215989,
      "grad_norm": 2.113811860287663,
      "learning_rate": 1e-05,
      "loss": 1.3397,
      "step": 812
    },
    {
      "epoch": 0.563946934882511,
      "grad_norm": 2.0550884583810487,
      "learning_rate": 1e-05,
      "loss": 1.3316,
      "step": 813
    },
    {
      "epoch": 0.5646405965490332,
      "grad_norm": 1.85558729084722,
      "learning_rate": 1e-05,
      "loss": 1.3586,
      "step": 814
    },
    {
      "epoch": 0.5653342582155554,
      "grad_norm": 1.7940205357205186,
      "learning_rate": 1e-05,
      "loss": 1.3328,
      "step": 815
    },
    {
      "epoch": 0.5660279198820776,
      "grad_norm": 1.8867105415827423,
      "learning_rate": 1e-05,
      "loss": 1.4099,
      "step": 816
    },
    {
      "epoch": 0.5667215815485996,
      "grad_norm": 1.9561860862534044,
      "learning_rate": 1e-05,
      "loss": 1.4491,
      "step": 817
    },
    {
      "epoch": 0.5674152432151218,
      "grad_norm": 2.0876714983856313,
      "learning_rate": 1e-05,
      "loss": 1.364,
      "step": 818
    },
    {
      "epoch": 0.568108904881644,
      "grad_norm": 1.99399680820796,
      "learning_rate": 1e-05,
      "loss": 1.3033,
      "step": 819
    },
    {
      "epoch": 0.5688025665481661,
      "grad_norm": 1.976493240480479,
      "learning_rate": 1e-05,
      "loss": 1.333,
      "step": 820
    },
    {
      "epoch": 0.5694962282146883,
      "grad_norm": 1.95662043100926,
      "learning_rate": 1e-05,
      "loss": 1.3136,
      "step": 821
    },
    {
      "epoch": 0.5701898898812104,
      "grad_norm": 1.6497733053601713,
      "learning_rate": 1e-05,
      "loss": 1.3861,
      "step": 822
    },
    {
      "epoch": 0.5708835515477326,
      "grad_norm": 1.818935374314111,
      "learning_rate": 1e-05,
      "loss": 1.3857,
      "step": 823
    },
    {
      "epoch": 0.5715772132142547,
      "grad_norm": 1.747967079872631,
      "learning_rate": 1e-05,
      "loss": 1.349,
      "step": 824
    },
    {
      "epoch": 0.5722708748807769,
      "grad_norm": 1.8908128266598878,
      "learning_rate": 1e-05,
      "loss": 1.3616,
      "step": 825
    },
    {
      "epoch": 0.5729645365472991,
      "grad_norm": 1.8720455934356435,
      "learning_rate": 1e-05,
      "loss": 1.3238,
      "step": 826
    },
    {
      "epoch": 0.5736581982138212,
      "grad_norm": 1.7833794579975666,
      "learning_rate": 1e-05,
      "loss": 1.3956,
      "step": 827
    },
    {
      "epoch": 0.5743518598803433,
      "grad_norm": 2.1032994082238203,
      "learning_rate": 1e-05,
      "loss": 1.3575,
      "step": 828
    },
    {
      "epoch": 0.5750455215468655,
      "grad_norm": 2.0134774144141487,
      "learning_rate": 1e-05,
      "loss": 1.3464,
      "step": 829
    },
    {
      "epoch": 0.5757391832133877,
      "grad_norm": 1.8711603789528202,
      "learning_rate": 1e-05,
      "loss": 1.4115,
      "step": 830
    },
    {
      "epoch": 0.5764328448799099,
      "grad_norm": 1.7819151783946192,
      "learning_rate": 1e-05,
      "loss": 1.3615,
      "step": 831
    },
    {
      "epoch": 0.5771265065464319,
      "grad_norm": 1.9690584032628007,
      "learning_rate": 1e-05,
      "loss": 1.3901,
      "step": 832
    },
    {
      "epoch": 0.5778201682129541,
      "grad_norm": 1.9642954115887026,
      "learning_rate": 1e-05,
      "loss": 1.3518,
      "step": 833
    },
    {
      "epoch": 0.5785138298794763,
      "grad_norm": 1.744497422769626,
      "learning_rate": 1e-05,
      "loss": 1.3469,
      "step": 834
    },
    {
      "epoch": 0.5792074915459985,
      "grad_norm": 2.0755337336634767,
      "learning_rate": 1e-05,
      "loss": 1.3823,
      "step": 835
    },
    {
      "epoch": 0.5799011532125206,
      "grad_norm": 1.8340533083393944,
      "learning_rate": 1e-05,
      "loss": 1.374,
      "step": 836
    },
    {
      "epoch": 0.5805948148790427,
      "grad_norm": 1.7560881859736863,
      "learning_rate": 1e-05,
      "loss": 1.4051,
      "step": 837
    },
    {
      "epoch": 0.5812884765455649,
      "grad_norm": 1.7470644935895128,
      "learning_rate": 1e-05,
      "loss": 1.3751,
      "step": 838
    },
    {
      "epoch": 0.5819821382120871,
      "grad_norm": 1.6498957999043187,
      "learning_rate": 1e-05,
      "loss": 1.3425,
      "step": 839
    },
    {
      "epoch": 0.5826757998786092,
      "grad_norm": 1.708284581084529,
      "learning_rate": 1e-05,
      "loss": 1.3772,
      "step": 840
    },
    {
      "epoch": 0.5833694615451314,
      "grad_norm": 1.7341358375022322,
      "learning_rate": 1e-05,
      "loss": 1.4271,
      "step": 841
    },
    {
      "epoch": 0.5840631232116535,
      "grad_norm": 1.96090088265045,
      "learning_rate": 1e-05,
      "loss": 1.3811,
      "step": 842
    },
    {
      "epoch": 0.5847567848781756,
      "grad_norm": 1.8309156039915615,
      "learning_rate": 1e-05,
      "loss": 1.3518,
      "step": 843
    },
    {
      "epoch": 0.5854504465446978,
      "grad_norm": 1.766087750027872,
      "learning_rate": 1e-05,
      "loss": 1.2791,
      "step": 844
    },
    {
      "epoch": 0.58614410821122,
      "grad_norm": 1.840859949308765,
      "learning_rate": 1e-05,
      "loss": 1.3364,
      "step": 845
    },
    {
      "epoch": 0.5868377698777422,
      "grad_norm": 1.8811612648079954,
      "learning_rate": 1e-05,
      "loss": 1.3866,
      "step": 846
    },
    {
      "epoch": 0.5875314315442642,
      "grad_norm": 1.7882571243218808,
      "learning_rate": 1e-05,
      "loss": 1.3818,
      "step": 847
    },
    {
      "epoch": 0.5882250932107864,
      "grad_norm": 1.7751021276239416,
      "learning_rate": 1e-05,
      "loss": 1.3877,
      "step": 848
    },
    {
      "epoch": 0.5889187548773086,
      "grad_norm": 1.863598117479808,
      "learning_rate": 1e-05,
      "loss": 1.3283,
      "step": 849
    },
    {
      "epoch": 0.5896124165438308,
      "grad_norm": 1.7934939156618281,
      "learning_rate": 1e-05,
      "loss": 1.3909,
      "step": 850
    },
    {
      "epoch": 0.5903060782103529,
      "grad_norm": 1.8701501570430885,
      "learning_rate": 1e-05,
      "loss": 1.3911,
      "step": 851
    },
    {
      "epoch": 0.590999739876875,
      "grad_norm": 1.8075702135316054,
      "learning_rate": 1e-05,
      "loss": 1.3267,
      "step": 852
    },
    {
      "epoch": 0.5916934015433972,
      "grad_norm": 1.7844322891888966,
      "learning_rate": 1e-05,
      "loss": 1.3749,
      "step": 853
    },
    {
      "epoch": 0.5923870632099194,
      "grad_norm": 1.6364137503955962,
      "learning_rate": 1e-05,
      "loss": 1.3903,
      "step": 854
    },
    {
      "epoch": 0.5930807248764415,
      "grad_norm": 1.845018525439836,
      "learning_rate": 1e-05,
      "loss": 1.3638,
      "step": 855
    },
    {
      "epoch": 0.5937743865429637,
      "grad_norm": 1.6438686492795926,
      "learning_rate": 1e-05,
      "loss": 1.3356,
      "step": 856
    },
    {
      "epoch": 0.5944680482094858,
      "grad_norm": 1.8461296262824984,
      "learning_rate": 1e-05,
      "loss": 1.318,
      "step": 857
    },
    {
      "epoch": 0.595161709876008,
      "grad_norm": 1.7523581116541502,
      "learning_rate": 1e-05,
      "loss": 1.3278,
      "step": 858
    },
    {
      "epoch": 0.5958553715425301,
      "grad_norm": 1.7821675206089143,
      "learning_rate": 1e-05,
      "loss": 1.3257,
      "step": 859
    },
    {
      "epoch": 0.5965490332090523,
      "grad_norm": 1.8952855386403753,
      "learning_rate": 1e-05,
      "loss": 1.352,
      "step": 860
    },
    {
      "epoch": 0.5972426948755745,
      "grad_norm": 1.9061495251459173,
      "learning_rate": 1e-05,
      "loss": 1.361,
      "step": 861
    },
    {
      "epoch": 0.5979363565420966,
      "grad_norm": 1.807002944323855,
      "learning_rate": 1e-05,
      "loss": 1.3498,
      "step": 862
    },
    {
      "epoch": 0.5986300182086187,
      "grad_norm": 1.8427398989259318,
      "learning_rate": 1e-05,
      "loss": 1.3442,
      "step": 863
    },
    {
      "epoch": 0.5993236798751409,
      "grad_norm": 2.159582705739885,
      "learning_rate": 1e-05,
      "loss": 1.4137,
      "step": 864
    },
    {
      "epoch": 0.6000173415416631,
      "grad_norm": 1.8186698531077894,
      "learning_rate": 1e-05,
      "loss": 1.3743,
      "step": 865
    },
    {
      "epoch": 0.6007110032081852,
      "grad_norm": 1.7754924011854336,
      "learning_rate": 1e-05,
      "loss": 1.3758,
      "step": 866
    },
    {
      "epoch": 0.6014046648747073,
      "grad_norm": 1.7146434615480912,
      "learning_rate": 1e-05,
      "loss": 1.3655,
      "step": 867
    },
    {
      "epoch": 0.6020983265412295,
      "grad_norm": 1.6864187498892416,
      "learning_rate": 1e-05,
      "loss": 1.3842,
      "step": 868
    },
    {
      "epoch": 0.6027919882077517,
      "grad_norm": 1.7625940182372917,
      "learning_rate": 1e-05,
      "loss": 1.3563,
      "step": 869
    },
    {
      "epoch": 0.6034856498742738,
      "grad_norm": 1.8385401005573274,
      "learning_rate": 1e-05,
      "loss": 1.317,
      "step": 870
    },
    {
      "epoch": 0.604179311540796,
      "grad_norm": 1.8249865396470937,
      "learning_rate": 1e-05,
      "loss": 1.3666,
      "step": 871
    },
    {
      "epoch": 0.6048729732073181,
      "grad_norm": 1.774024094473835,
      "learning_rate": 1e-05,
      "loss": 1.3677,
      "step": 872
    },
    {
      "epoch": 0.6055666348738403,
      "grad_norm": 1.9300061255791403,
      "learning_rate": 1e-05,
      "loss": 1.3668,
      "step": 873
    },
    {
      "epoch": 0.6062602965403624,
      "grad_norm": 1.8538055069719046,
      "learning_rate": 1e-05,
      "loss": 1.3639,
      "step": 874
    },
    {
      "epoch": 0.6069539582068846,
      "grad_norm": 1.8234801513872547,
      "learning_rate": 1e-05,
      "loss": 1.3183,
      "step": 875
    },
    {
      "epoch": 0.6076476198734068,
      "grad_norm": 1.8887104952070137,
      "learning_rate": 1e-05,
      "loss": 1.3798,
      "step": 876
    },
    {
      "epoch": 0.608341281539929,
      "grad_norm": 1.967685873156897,
      "learning_rate": 1e-05,
      "loss": 1.3733,
      "step": 877
    },
    {
      "epoch": 0.609034943206451,
      "grad_norm": 1.8930452243619467,
      "learning_rate": 1e-05,
      "loss": 1.3627,
      "step": 878
    },
    {
      "epoch": 0.6097286048729732,
      "grad_norm": 1.9917808148339706,
      "learning_rate": 1e-05,
      "loss": 1.3637,
      "step": 879
    },
    {
      "epoch": 0.6104222665394954,
      "grad_norm": 1.7953716355637714,
      "learning_rate": 1e-05,
      "loss": 1.3552,
      "step": 880
    },
    {
      "epoch": 0.6111159282060176,
      "grad_norm": 2.054090380165633,
      "learning_rate": 1e-05,
      "loss": 1.3309,
      "step": 881
    },
    {
      "epoch": 0.6118095898725396,
      "grad_norm": 1.774125479388038,
      "learning_rate": 1e-05,
      "loss": 1.3346,
      "step": 882
    },
    {
      "epoch": 0.6125032515390618,
      "grad_norm": 1.788110427730304,
      "learning_rate": 1e-05,
      "loss": 1.3907,
      "step": 883
    },
    {
      "epoch": 0.613196913205584,
      "grad_norm": 1.9117105836931287,
      "learning_rate": 1e-05,
      "loss": 1.3662,
      "step": 884
    },
    {
      "epoch": 0.6138905748721061,
      "grad_norm": 1.9761438291924842,
      "learning_rate": 1e-05,
      "loss": 1.3487,
      "step": 885
    },
    {
      "epoch": 0.6145842365386283,
      "grad_norm": 1.7939497044726074,
      "learning_rate": 1e-05,
      "loss": 1.3617,
      "step": 886
    },
    {
      "epoch": 0.6152778982051504,
      "grad_norm": 1.8137308254099254,
      "learning_rate": 1e-05,
      "loss": 1.4017,
      "step": 887
    },
    {
      "epoch": 0.6159715598716726,
      "grad_norm": 1.9358335601206476,
      "learning_rate": 1e-05,
      "loss": 1.3221,
      "step": 888
    },
    {
      "epoch": 0.6166652215381947,
      "grad_norm": 1.8023857684374447,
      "learning_rate": 1e-05,
      "loss": 1.3645,
      "step": 889
    },
    {
      "epoch": 0.6173588832047169,
      "grad_norm": 1.7858631263519855,
      "learning_rate": 1e-05,
      "loss": 1.3915,
      "step": 890
    },
    {
      "epoch": 0.6180525448712391,
      "grad_norm": 1.8447027974665438,
      "learning_rate": 1e-05,
      "loss": 1.3331,
      "step": 891
    },
    {
      "epoch": 0.6187462065377612,
      "grad_norm": 1.8973843442097094,
      "learning_rate": 1e-05,
      "loss": 1.3053,
      "step": 892
    },
    {
      "epoch": 0.6194398682042833,
      "grad_norm": 1.7623453790808643,
      "learning_rate": 1e-05,
      "loss": 1.2926,
      "step": 893
    },
    {
      "epoch": 0.6201335298708055,
      "grad_norm": 1.8017523463560905,
      "learning_rate": 1e-05,
      "loss": 1.3647,
      "step": 894
    },
    {
      "epoch": 0.6208271915373277,
      "grad_norm": 1.7450544611778176,
      "learning_rate": 1e-05,
      "loss": 1.3658,
      "step": 895
    },
    {
      "epoch": 0.6215208532038499,
      "grad_norm": 1.823024100294463,
      "learning_rate": 1e-05,
      "loss": 1.3508,
      "step": 896
    },
    {
      "epoch": 0.6222145148703719,
      "grad_norm": 1.8524166096635302,
      "learning_rate": 1e-05,
      "loss": 1.3446,
      "step": 897
    },
    {
      "epoch": 0.6229081765368941,
      "grad_norm": 1.7203042262118677,
      "learning_rate": 1e-05,
      "loss": 1.3838,
      "step": 898
    },
    {
      "epoch": 0.6236018382034163,
      "grad_norm": 1.7896042667571013,
      "learning_rate": 1e-05,
      "loss": 1.3401,
      "step": 899
    },
    {
      "epoch": 0.6242954998699385,
      "grad_norm": 1.8566869792112495,
      "learning_rate": 1e-05,
      "loss": 1.3856,
      "step": 900
    },
    {
      "epoch": 0.6249891615364606,
      "grad_norm": 1.7236631672011284,
      "learning_rate": 1e-05,
      "loss": 1.3455,
      "step": 901
    },
    {
      "epoch": 0.6256828232029827,
      "grad_norm": 1.9857281332079058,
      "learning_rate": 1e-05,
      "loss": 1.3347,
      "step": 902
    },
    {
      "epoch": 0.6263764848695049,
      "grad_norm": 1.916490049064551,
      "learning_rate": 1e-05,
      "loss": 1.2964,
      "step": 903
    },
    {
      "epoch": 0.6270701465360271,
      "grad_norm": 1.9713572657543152,
      "learning_rate": 1e-05,
      "loss": 1.3302,
      "step": 904
    },
    {
      "epoch": 0.6277638082025492,
      "grad_norm": 1.7610441348646735,
      "learning_rate": 1e-05,
      "loss": 1.3344,
      "step": 905
    },
    {
      "epoch": 0.6284574698690714,
      "grad_norm": 1.799617843727853,
      "learning_rate": 1e-05,
      "loss": 1.3284,
      "step": 906
    },
    {
      "epoch": 0.6291511315355935,
      "grad_norm": 1.7832143262655586,
      "learning_rate": 1e-05,
      "loss": 1.3356,
      "step": 907
    },
    {
      "epoch": 0.6298447932021156,
      "grad_norm": 1.7971828284576865,
      "learning_rate": 1e-05,
      "loss": 1.3207,
      "step": 908
    },
    {
      "epoch": 0.6305384548686378,
      "grad_norm": 1.7536423641131738,
      "learning_rate": 1e-05,
      "loss": 1.3499,
      "step": 909
    },
    {
      "epoch": 0.63123211653516,
      "grad_norm": 1.7401477745871217,
      "learning_rate": 1e-05,
      "loss": 1.3398,
      "step": 910
    },
    {
      "epoch": 0.6319257782016822,
      "grad_norm": 1.7218362910164169,
      "learning_rate": 1e-05,
      "loss": 1.4193,
      "step": 911
    },
    {
      "epoch": 0.6326194398682042,
      "grad_norm": 1.9808528889101304,
      "learning_rate": 1e-05,
      "loss": 1.392,
      "step": 912
    },
    {
      "epoch": 0.6333131015347264,
      "grad_norm": 1.9064408882431807,
      "learning_rate": 1e-05,
      "loss": 1.3615,
      "step": 913
    },
    {
      "epoch": 0.6340067632012486,
      "grad_norm": 1.7248366549906144,
      "learning_rate": 1e-05,
      "loss": 1.3231,
      "step": 914
    },
    {
      "epoch": 0.6347004248677708,
      "grad_norm": 1.795395048611617,
      "learning_rate": 1e-05,
      "loss": 1.3851,
      "step": 915
    },
    {
      "epoch": 0.6353940865342929,
      "grad_norm": 1.7344884424253888,
      "learning_rate": 1e-05,
      "loss": 1.3528,
      "step": 916
    },
    {
      "epoch": 0.636087748200815,
      "grad_norm": 1.700061571947052,
      "learning_rate": 1e-05,
      "loss": 1.3348,
      "step": 917
    },
    {
      "epoch": 0.6367814098673372,
      "grad_norm": 1.8646566885856952,
      "learning_rate": 1e-05,
      "loss": 1.357,
      "step": 918
    },
    {
      "epoch": 0.6374750715338594,
      "grad_norm": 1.8652265942281396,
      "learning_rate": 1e-05,
      "loss": 1.3343,
      "step": 919
    },
    {
      "epoch": 0.6381687332003815,
      "grad_norm": 1.7925656802981118,
      "learning_rate": 1e-05,
      "loss": 1.3308,
      "step": 920
    },
    {
      "epoch": 0.6388623948669037,
      "grad_norm": 1.7896253293176538,
      "learning_rate": 1e-05,
      "loss": 1.3813,
      "step": 921
    },
    {
      "epoch": 0.6395560565334258,
      "grad_norm": 1.6534489942501098,
      "learning_rate": 1e-05,
      "loss": 1.3233,
      "step": 922
    },
    {
      "epoch": 0.640249718199948,
      "grad_norm": 1.8033026812852484,
      "learning_rate": 1e-05,
      "loss": 1.3074,
      "step": 923
    },
    {
      "epoch": 0.6409433798664701,
      "grad_norm": 1.822451647210804,
      "learning_rate": 1e-05,
      "loss": 1.3217,
      "step": 924
    },
    {
      "epoch": 0.6416370415329923,
      "grad_norm": 1.7821886850533442,
      "learning_rate": 1e-05,
      "loss": 1.3531,
      "step": 925
    },
    {
      "epoch": 0.6423307031995145,
      "grad_norm": 1.8446137766083273,
      "learning_rate": 1e-05,
      "loss": 1.3552,
      "step": 926
    },
    {
      "epoch": 0.6430243648660365,
      "grad_norm": 1.7365041612964318,
      "learning_rate": 1e-05,
      "loss": 1.3472,
      "step": 927
    },
    {
      "epoch": 0.6437180265325587,
      "grad_norm": 1.9116647427342783,
      "learning_rate": 1e-05,
      "loss": 1.3356,
      "step": 928
    },
    {
      "epoch": 0.6444116881990809,
      "grad_norm": 1.6818488752250975,
      "learning_rate": 1e-05,
      "loss": 1.35,
      "step": 929
    },
    {
      "epoch": 0.6451053498656031,
      "grad_norm": 1.8566305164008303,
      "learning_rate": 1e-05,
      "loss": 1.3643,
      "step": 930
    },
    {
      "epoch": 0.6457990115321252,
      "grad_norm": 1.7420953544687154,
      "learning_rate": 1e-05,
      "loss": 1.3148,
      "step": 931
    },
    {
      "epoch": 0.6464926731986473,
      "grad_norm": 1.885744413844102,
      "learning_rate": 1e-05,
      "loss": 1.3492,
      "step": 932
    },
    {
      "epoch": 0.6471863348651695,
      "grad_norm": 1.7944270298154161,
      "learning_rate": 1e-05,
      "loss": 1.3513,
      "step": 933
    },
    {
      "epoch": 0.6478799965316917,
      "grad_norm": 1.7535218523742484,
      "learning_rate": 1e-05,
      "loss": 1.4101,
      "step": 934
    },
    {
      "epoch": 0.6485736581982138,
      "grad_norm": 1.7229852034518358,
      "learning_rate": 1e-05,
      "loss": 1.3437,
      "step": 935
    },
    {
      "epoch": 0.649267319864736,
      "grad_norm": 1.896304422647214,
      "learning_rate": 1e-05,
      "loss": 1.3156,
      "step": 936
    },
    {
      "epoch": 0.6499609815312581,
      "grad_norm": 1.8055244846850502,
      "learning_rate": 1e-05,
      "loss": 1.3691,
      "step": 937
    },
    {
      "epoch": 0.6506546431977803,
      "grad_norm": 1.684941036557295,
      "learning_rate": 1e-05,
      "loss": 1.3579,
      "step": 938
    },
    {
      "epoch": 0.6513483048643024,
      "grad_norm": 1.8888571415510795,
      "learning_rate": 1e-05,
      "loss": 1.3632,
      "step": 939
    },
    {
      "epoch": 0.6520419665308246,
      "grad_norm": 1.8160274262290288,
      "learning_rate": 1e-05,
      "loss": 1.3099,
      "step": 940
    },
    {
      "epoch": 0.6527356281973468,
      "grad_norm": 1.755049632438486,
      "learning_rate": 1e-05,
      "loss": 1.3519,
      "step": 941
    },
    {
      "epoch": 0.653429289863869,
      "grad_norm": 1.89712944315266,
      "learning_rate": 1e-05,
      "loss": 1.3494,
      "step": 942
    },
    {
      "epoch": 0.654122951530391,
      "grad_norm": 1.765188366032801,
      "learning_rate": 1e-05,
      "loss": 1.3856,
      "step": 943
    },
    {
      "epoch": 0.6548166131969132,
      "grad_norm": 1.9497117765562002,
      "learning_rate": 1e-05,
      "loss": 1.3665,
      "step": 944
    },
    {
      "epoch": 0.6555102748634354,
      "grad_norm": 1.799101172711031,
      "learning_rate": 1e-05,
      "loss": 1.3445,
      "step": 945
    },
    {
      "epoch": 0.6562039365299576,
      "grad_norm": 1.712616408878392,
      "learning_rate": 1e-05,
      "loss": 1.3491,
      "step": 946
    },
    {
      "epoch": 0.6568975981964796,
      "grad_norm": 1.7946208261432808,
      "learning_rate": 1e-05,
      "loss": 1.3731,
      "step": 947
    },
    {
      "epoch": 0.6575912598630018,
      "grad_norm": 1.7262699314904466,
      "learning_rate": 1e-05,
      "loss": 1.3585,
      "step": 948
    },
    {
      "epoch": 0.658284921529524,
      "grad_norm": 1.9628450564778277,
      "learning_rate": 1e-05,
      "loss": 1.301,
      "step": 949
    },
    {
      "epoch": 0.6589785831960461,
      "grad_norm": 1.8061202922829884,
      "learning_rate": 1e-05,
      "loss": 1.3783,
      "step": 950
    },
    {
      "epoch": 0.6596722448625683,
      "grad_norm": 1.7121738551781767,
      "learning_rate": 1e-05,
      "loss": 1.357,
      "step": 951
    },
    {
      "epoch": 0.6603659065290904,
      "grad_norm": 1.8850562141039617,
      "learning_rate": 1e-05,
      "loss": 1.2819,
      "step": 952
    },
    {
      "epoch": 0.6610595681956126,
      "grad_norm": 1.867193802881424,
      "learning_rate": 1e-05,
      "loss": 1.3156,
      "step": 953
    },
    {
      "epoch": 0.6617532298621347,
      "grad_norm": 1.7443179431377005,
      "learning_rate": 1e-05,
      "loss": 1.2751,
      "step": 954
    },
    {
      "epoch": 0.6624468915286569,
      "grad_norm": 1.8733131607506688,
      "learning_rate": 1e-05,
      "loss": 1.3534,
      "step": 955
    },
    {
      "epoch": 0.6631405531951791,
      "grad_norm": 1.9784306105729255,
      "learning_rate": 1e-05,
      "loss": 1.2742,
      "step": 956
    },
    {
      "epoch": 0.6638342148617012,
      "grad_norm": 1.8959702823237385,
      "learning_rate": 1e-05,
      "loss": 1.4094,
      "step": 957
    },
    {
      "epoch": 0.6645278765282233,
      "grad_norm": 1.664080974193892,
      "learning_rate": 1e-05,
      "loss": 1.3658,
      "step": 958
    },
    {
      "epoch": 0.6652215381947455,
      "grad_norm": 1.782008443874851,
      "learning_rate": 1e-05,
      "loss": 1.347,
      "step": 959
    },
    {
      "epoch": 0.6659151998612677,
      "grad_norm": 1.8460350587229146,
      "learning_rate": 1e-05,
      "loss": 1.3639,
      "step": 960
    },
    {
      "epoch": 0.6666088615277899,
      "grad_norm": 1.6425972064330443,
      "learning_rate": 1e-05,
      "loss": 1.3693,
      "step": 961
    },
    {
      "epoch": 0.6673025231943119,
      "grad_norm": 1.8565662897573758,
      "learning_rate": 1e-05,
      "loss": 1.3636,
      "step": 962
    },
    {
      "epoch": 0.6679961848608341,
      "grad_norm": 1.7302118661778385,
      "learning_rate": 1e-05,
      "loss": 1.3566,
      "step": 963
    },
    {
      "epoch": 0.6686898465273563,
      "grad_norm": 1.9616014623863918,
      "learning_rate": 1e-05,
      "loss": 1.374,
      "step": 964
    },
    {
      "epoch": 0.6693835081938785,
      "grad_norm": 1.84734239559959,
      "learning_rate": 1e-05,
      "loss": 1.2604,
      "step": 965
    },
    {
      "epoch": 0.6700771698604006,
      "grad_norm": 1.6679061078794732,
      "learning_rate": 1e-05,
      "loss": 1.344,
      "step": 966
    },
    {
      "epoch": 0.6707708315269227,
      "grad_norm": 1.9249982922421873,
      "learning_rate": 1e-05,
      "loss": 1.3773,
      "step": 967
    },
    {
      "epoch": 0.6714644931934449,
      "grad_norm": 1.9130662887594385,
      "learning_rate": 1e-05,
      "loss": 1.3389,
      "step": 968
    },
    {
      "epoch": 0.6721581548599671,
      "grad_norm": 2.0490408445129553,
      "learning_rate": 1e-05,
      "loss": 1.3271,
      "step": 969
    },
    {
      "epoch": 0.6728518165264892,
      "grad_norm": 1.7966765927847634,
      "learning_rate": 1e-05,
      "loss": 1.3082,
      "step": 970
    },
    {
      "epoch": 0.6735454781930114,
      "grad_norm": 1.8365578777762348,
      "learning_rate": 1e-05,
      "loss": 1.3813,
      "step": 971
    },
    {
      "epoch": 0.6742391398595335,
      "grad_norm": 1.7856135676874048,
      "learning_rate": 1e-05,
      "loss": 1.3791,
      "step": 972
    },
    {
      "epoch": 0.6749328015260556,
      "grad_norm": 1.7189842050358264,
      "learning_rate": 1e-05,
      "loss": 1.3859,
      "step": 973
    },
    {
      "epoch": 0.6756264631925778,
      "grad_norm": 1.697736604178591,
      "learning_rate": 1e-05,
      "loss": 1.3531,
      "step": 974
    },
    {
      "epoch": 0.6763201248591,
      "grad_norm": 1.85446433118358,
      "learning_rate": 1e-05,
      "loss": 1.3377,
      "step": 975
    },
    {
      "epoch": 0.6770137865256222,
      "grad_norm": 1.817734345502182,
      "learning_rate": 1e-05,
      "loss": 1.3316,
      "step": 976
    },
    {
      "epoch": 0.6777074481921442,
      "grad_norm": 1.6362921639652548,
      "learning_rate": 1e-05,
      "loss": 1.2879,
      "step": 977
    },
    {
      "epoch": 0.6784011098586664,
      "grad_norm": 1.7236748146694982,
      "learning_rate": 1e-05,
      "loss": 1.3023,
      "step": 978
    },
    {
      "epoch": 0.6790947715251886,
      "grad_norm": 1.874808634354032,
      "learning_rate": 1e-05,
      "loss": 1.3291,
      "step": 979
    },
    {
      "epoch": 0.6797884331917108,
      "grad_norm": 2.143605571091092,
      "learning_rate": 1e-05,
      "loss": 1.3305,
      "step": 980
    },
    {
      "epoch": 0.6804820948582329,
      "grad_norm": 1.7140845838212821,
      "learning_rate": 1e-05,
      "loss": 1.2762,
      "step": 981
    },
    {
      "epoch": 0.681175756524755,
      "grad_norm": 1.7627892247793258,
      "learning_rate": 1e-05,
      "loss": 1.3793,
      "step": 982
    },
    {
      "epoch": 0.6818694181912772,
      "grad_norm": 1.9327806260640557,
      "learning_rate": 1e-05,
      "loss": 1.3314,
      "step": 983
    },
    {
      "epoch": 0.6825630798577994,
      "grad_norm": 1.9570006600139125,
      "learning_rate": 1e-05,
      "loss": 1.3092,
      "step": 984
    },
    {
      "epoch": 0.6832567415243215,
      "grad_norm": 1.857815246395867,
      "learning_rate": 1e-05,
      "loss": 1.328,
      "step": 985
    },
    {
      "epoch": 0.6839504031908437,
      "grad_norm": 1.6521408115836034,
      "learning_rate": 1e-05,
      "loss": 1.3649,
      "step": 986
    },
    {
      "epoch": 0.6846440648573658,
      "grad_norm": 1.6883217532733774,
      "learning_rate": 1e-05,
      "loss": 1.3742,
      "step": 987
    },
    {
      "epoch": 0.685337726523888,
      "grad_norm": 1.7657705434647315,
      "learning_rate": 1e-05,
      "loss": 1.4009,
      "step": 988
    },
    {
      "epoch": 0.6860313881904101,
      "grad_norm": 1.8388470770976078,
      "learning_rate": 1e-05,
      "loss": 1.3385,
      "step": 989
    },
    {
      "epoch": 0.6867250498569323,
      "grad_norm": 1.6446394362553027,
      "learning_rate": 1e-05,
      "loss": 1.3747,
      "step": 990
    },
    {
      "epoch": 0.6874187115234545,
      "grad_norm": 1.76332053954708,
      "learning_rate": 1e-05,
      "loss": 1.3744,
      "step": 991
    },
    {
      "epoch": 0.6881123731899765,
      "grad_norm": 1.8551504155963352,
      "learning_rate": 1e-05,
      "loss": 1.3461,
      "step": 992
    },
    {
      "epoch": 0.6888060348564987,
      "grad_norm": 1.8242720423216203,
      "learning_rate": 1e-05,
      "loss": 1.3949,
      "step": 993
    },
    {
      "epoch": 0.6894996965230209,
      "grad_norm": 1.8728688560334699,
      "learning_rate": 1e-05,
      "loss": 1.4202,
      "step": 994
    },
    {
      "epoch": 0.6901933581895431,
      "grad_norm": 1.8128080132317514,
      "learning_rate": 1e-05,
      "loss": 1.3528,
      "step": 995
    },
    {
      "epoch": 0.6908870198560652,
      "grad_norm": 1.7706897683233593,
      "learning_rate": 1e-05,
      "loss": 1.3772,
      "step": 996
    },
    {
      "epoch": 0.6915806815225873,
      "grad_norm": 1.867842253838222,
      "learning_rate": 1e-05,
      "loss": 1.3735,
      "step": 997
    },
    {
      "epoch": 0.6922743431891095,
      "grad_norm": 1.908519372133083,
      "learning_rate": 1e-05,
      "loss": 1.3093,
      "step": 998
    },
    {
      "epoch": 0.6929680048556317,
      "grad_norm": 1.7680738442803956,
      "learning_rate": 1e-05,
      "loss": 1.3205,
      "step": 999
    },
    {
      "epoch": 0.6936616665221538,
      "grad_norm": 1.752432098173259,
      "learning_rate": 1e-05,
      "loss": 1.3451,
      "step": 1000
    },
    {
      "epoch": 0.694355328188676,
      "grad_norm": 1.7866959543838525,
      "learning_rate": 1e-05,
      "loss": 1.3288,
      "step": 1001
    },
    {
      "epoch": 0.6950489898551981,
      "grad_norm": 1.6320555911640122,
      "learning_rate": 1e-05,
      "loss": 1.3157,
      "step": 1002
    },
    {
      "epoch": 0.6957426515217203,
      "grad_norm": 1.7766379486245896,
      "learning_rate": 1e-05,
      "loss": 1.3559,
      "step": 1003
    },
    {
      "epoch": 0.6964363131882424,
      "grad_norm": 1.7801227267982318,
      "learning_rate": 1e-05,
      "loss": 1.389,
      "step": 1004
    },
    {
      "epoch": 0.6971299748547646,
      "grad_norm": 1.7763939485733111,
      "learning_rate": 1e-05,
      "loss": 1.4214,
      "step": 1005
    },
    {
      "epoch": 0.6978236365212868,
      "grad_norm": 1.7466154961438336,
      "learning_rate": 1e-05,
      "loss": 1.3974,
      "step": 1006
    },
    {
      "epoch": 0.6985172981878089,
      "grad_norm": 1.8018054751465553,
      "learning_rate": 1e-05,
      "loss": 1.3146,
      "step": 1007
    },
    {
      "epoch": 0.699210959854331,
      "grad_norm": 1.7015362135443022,
      "learning_rate": 1e-05,
      "loss": 1.3521,
      "step": 1008
    },
    {
      "epoch": 0.6999046215208532,
      "grad_norm": 1.8044732359887248,
      "learning_rate": 1e-05,
      "loss": 1.3298,
      "step": 1009
    },
    {
      "epoch": 0.7005982831873754,
      "grad_norm": 1.7248926110752036,
      "learning_rate": 1e-05,
      "loss": 1.311,
      "step": 1010
    },
    {
      "epoch": 0.7012919448538976,
      "grad_norm": 1.7408526444267358,
      "learning_rate": 1e-05,
      "loss": 1.3346,
      "step": 1011
    },
    {
      "epoch": 0.7019856065204196,
      "grad_norm": 1.756432069962424,
      "learning_rate": 1e-05,
      "loss": 1.3608,
      "step": 1012
    },
    {
      "epoch": 0.7026792681869418,
      "grad_norm": 1.9004070384049725,
      "learning_rate": 1e-05,
      "loss": 1.3061,
      "step": 1013
    },
    {
      "epoch": 0.703372929853464,
      "grad_norm": 1.7674772646104595,
      "learning_rate": 1e-05,
      "loss": 1.3365,
      "step": 1014
    },
    {
      "epoch": 0.7040665915199861,
      "grad_norm": 1.815286596926447,
      "learning_rate": 1e-05,
      "loss": 1.3112,
      "step": 1015
    },
    {
      "epoch": 0.7047602531865083,
      "grad_norm": 1.9200058514873535,
      "learning_rate": 1e-05,
      "loss": 1.3702,
      "step": 1016
    },
    {
      "epoch": 0.7054539148530304,
      "grad_norm": 1.7499367861528972,
      "learning_rate": 1e-05,
      "loss": 1.3707,
      "step": 1017
    },
    {
      "epoch": 0.7061475765195526,
      "grad_norm": 1.925251587075512,
      "learning_rate": 1e-05,
      "loss": 1.3208,
      "step": 1018
    },
    {
      "epoch": 0.7068412381860747,
      "grad_norm": 1.7154198796482498,
      "learning_rate": 1e-05,
      "loss": 1.3336,
      "step": 1019
    },
    {
      "epoch": 0.7075348998525969,
      "grad_norm": 1.837360393002266,
      "learning_rate": 1e-05,
      "loss": 1.3328,
      "step": 1020
    },
    {
      "epoch": 0.7082285615191191,
      "grad_norm": 1.6211349139215232,
      "learning_rate": 1e-05,
      "loss": 1.3284,
      "step": 1021
    },
    {
      "epoch": 0.7089222231856412,
      "grad_norm": 1.866016563395064,
      "learning_rate": 1e-05,
      "loss": 1.3198,
      "step": 1022
    },
    {
      "epoch": 0.7096158848521633,
      "grad_norm": 1.6839566806665383,
      "learning_rate": 1e-05,
      "loss": 1.3651,
      "step": 1023
    },
    {
      "epoch": 0.7103095465186855,
      "grad_norm": 1.7159632620855965,
      "learning_rate": 1e-05,
      "loss": 1.3268,
      "step": 1024
    },
    {
      "epoch": 0.7110032081852077,
      "grad_norm": 1.806422188485046,
      "learning_rate": 1e-05,
      "loss": 1.2901,
      "step": 1025
    },
    {
      "epoch": 0.7116968698517299,
      "grad_norm": 1.7329697047767731,
      "learning_rate": 1e-05,
      "loss": 1.278,
      "step": 1026
    },
    {
      "epoch": 0.7123905315182519,
      "grad_norm": 1.7358387057504157,
      "learning_rate": 1e-05,
      "loss": 1.3959,
      "step": 1027
    },
    {
      "epoch": 0.7130841931847741,
      "grad_norm": 1.7843805164975317,
      "learning_rate": 1e-05,
      "loss": 1.3158,
      "step": 1028
    },
    {
      "epoch": 0.7137778548512963,
      "grad_norm": 1.8034487032951743,
      "learning_rate": 1e-05,
      "loss": 1.3239,
      "step": 1029
    },
    {
      "epoch": 0.7144715165178185,
      "grad_norm": 1.7806257674138806,
      "learning_rate": 1e-05,
      "loss": 1.348,
      "step": 1030
    },
    {
      "epoch": 0.7151651781843406,
      "grad_norm": 1.8562808148693768,
      "learning_rate": 1e-05,
      "loss": 1.3675,
      "step": 1031
    },
    {
      "epoch": 0.7158588398508627,
      "grad_norm": 1.9354803733254098,
      "learning_rate": 1e-05,
      "loss": 1.3331,
      "step": 1032
    },
    {
      "epoch": 0.7165525015173849,
      "grad_norm": 1.7581325487197559,
      "learning_rate": 1e-05,
      "loss": 1.3468,
      "step": 1033
    },
    {
      "epoch": 0.7172461631839071,
      "grad_norm": 1.802989571777425,
      "learning_rate": 1e-05,
      "loss": 1.3429,
      "step": 1034
    },
    {
      "epoch": 0.7179398248504292,
      "grad_norm": 1.7427548981299272,
      "learning_rate": 1e-05,
      "loss": 1.3555,
      "step": 1035
    },
    {
      "epoch": 0.7186334865169514,
      "grad_norm": 1.9113368723599877,
      "learning_rate": 1e-05,
      "loss": 1.3458,
      "step": 1036
    },
    {
      "epoch": 0.7193271481834735,
      "grad_norm": 1.746074482528233,
      "learning_rate": 1e-05,
      "loss": 1.256,
      "step": 1037
    },
    {
      "epoch": 0.7200208098499956,
      "grad_norm": 1.8354380987152568,
      "learning_rate": 1e-05,
      "loss": 1.336,
      "step": 1038
    },
    {
      "epoch": 0.7207144715165178,
      "grad_norm": 1.945273024066637,
      "learning_rate": 1e-05,
      "loss": 1.3985,
      "step": 1039
    },
    {
      "epoch": 0.72140813318304,
      "grad_norm": 1.843452934712379,
      "learning_rate": 1e-05,
      "loss": 1.3193,
      "step": 1040
    },
    {
      "epoch": 0.7221017948495622,
      "grad_norm": 1.9504440192269308,
      "learning_rate": 1e-05,
      "loss": 1.3296,
      "step": 1041
    },
    {
      "epoch": 0.7227954565160842,
      "grad_norm": 1.8706540071171702,
      "learning_rate": 1e-05,
      "loss": 1.3403,
      "step": 1042
    },
    {
      "epoch": 0.7234891181826064,
      "grad_norm": 1.9488283001981537,
      "learning_rate": 1e-05,
      "loss": 1.3245,
      "step": 1043
    },
    {
      "epoch": 0.7241827798491286,
      "grad_norm": 1.6969025009984984,
      "learning_rate": 1e-05,
      "loss": 1.3041,
      "step": 1044
    },
    {
      "epoch": 0.7248764415156508,
      "grad_norm": 2.0072497667921443,
      "learning_rate": 1e-05,
      "loss": 1.3447,
      "step": 1045
    },
    {
      "epoch": 0.7255701031821729,
      "grad_norm": 1.8741053070227567,
      "learning_rate": 1e-05,
      "loss": 1.3814,
      "step": 1046
    },
    {
      "epoch": 0.726263764848695,
      "grad_norm": 1.8874635982516978,
      "learning_rate": 1e-05,
      "loss": 1.328,
      "step": 1047
    },
    {
      "epoch": 0.7269574265152172,
      "grad_norm": 1.8011724514294678,
      "learning_rate": 1e-05,
      "loss": 1.3165,
      "step": 1048
    },
    {
      "epoch": 0.7276510881817394,
      "grad_norm": 1.7615037862154284,
      "learning_rate": 1e-05,
      "loss": 1.3405,
      "step": 1049
    },
    {
      "epoch": 0.7283447498482615,
      "grad_norm": 1.8936590686337285,
      "learning_rate": 1e-05,
      "loss": 1.3304,
      "step": 1050
    },
    {
      "epoch": 0.7290384115147837,
      "grad_norm": 1.7561903582794731,
      "learning_rate": 1e-05,
      "loss": 1.3135,
      "step": 1051
    },
    {
      "epoch": 0.7297320731813058,
      "grad_norm": 1.9910837825376402,
      "learning_rate": 1e-05,
      "loss": 1.3328,
      "step": 1052
    },
    {
      "epoch": 0.730425734847828,
      "grad_norm": 1.7932757018881984,
      "learning_rate": 1e-05,
      "loss": 1.3244,
      "step": 1053
    },
    {
      "epoch": 0.7311193965143501,
      "grad_norm": 1.8115170033234536,
      "learning_rate": 1e-05,
      "loss": 1.3256,
      "step": 1054
    },
    {
      "epoch": 0.7318130581808723,
      "grad_norm": 1.846025824397134,
      "learning_rate": 1e-05,
      "loss": 1.2897,
      "step": 1055
    },
    {
      "epoch": 0.7325067198473945,
      "grad_norm": 1.8145673138145253,
      "learning_rate": 1e-05,
      "loss": 1.3333,
      "step": 1056
    },
    {
      "epoch": 0.7332003815139165,
      "grad_norm": 1.6430330564527496,
      "learning_rate": 1e-05,
      "loss": 1.3569,
      "step": 1057
    },
    {
      "epoch": 0.7338940431804387,
      "grad_norm": 1.7140027949989802,
      "learning_rate": 1e-05,
      "loss": 1.3096,
      "step": 1058
    },
    {
      "epoch": 0.7345877048469609,
      "grad_norm": 1.9676800557132401,
      "learning_rate": 1e-05,
      "loss": 1.3327,
      "step": 1059
    },
    {
      "epoch": 0.7352813665134831,
      "grad_norm": 1.76220089185509,
      "learning_rate": 1e-05,
      "loss": 1.3837,
      "step": 1060
    },
    {
      "epoch": 0.7359750281800052,
      "grad_norm": 1.8650314152543974,
      "learning_rate": 1e-05,
      "loss": 1.31,
      "step": 1061
    },
    {
      "epoch": 0.7366686898465273,
      "grad_norm": 2.1105903583633054,
      "learning_rate": 1e-05,
      "loss": 1.3189,
      "step": 1062
    },
    {
      "epoch": 0.7373623515130495,
      "grad_norm": 1.9330298633028662,
      "learning_rate": 1e-05,
      "loss": 1.3116,
      "step": 1063
    },
    {
      "epoch": 0.7380560131795717,
      "grad_norm": 1.8697841477985335,
      "learning_rate": 1e-05,
      "loss": 1.3059,
      "step": 1064
    },
    {
      "epoch": 0.7387496748460938,
      "grad_norm": 1.8167855532087722,
      "learning_rate": 1e-05,
      "loss": 1.3725,
      "step": 1065
    },
    {
      "epoch": 0.739443336512616,
      "grad_norm": 1.9237163410520044,
      "learning_rate": 1e-05,
      "loss": 1.3216,
      "step": 1066
    },
    {
      "epoch": 0.7401369981791381,
      "grad_norm": 1.6945312309384704,
      "learning_rate": 1e-05,
      "loss": 1.3205,
      "step": 1067
    },
    {
      "epoch": 0.7408306598456603,
      "grad_norm": 1.8403636035002944,
      "learning_rate": 1e-05,
      "loss": 1.323,
      "step": 1068
    },
    {
      "epoch": 0.7415243215121824,
      "grad_norm": 1.8463721460630584,
      "learning_rate": 1e-05,
      "loss": 1.3215,
      "step": 1069
    },
    {
      "epoch": 0.7422179831787046,
      "grad_norm": 1.9058089022783489,
      "learning_rate": 1e-05,
      "loss": 1.3117,
      "step": 1070
    },
    {
      "epoch": 0.7429116448452268,
      "grad_norm": 2.1180556320179935,
      "learning_rate": 1e-05,
      "loss": 1.3762,
      "step": 1071
    },
    {
      "epoch": 0.7436053065117489,
      "grad_norm": 1.8334895855944495,
      "learning_rate": 1e-05,
      "loss": 1.3443,
      "step": 1072
    },
    {
      "epoch": 0.744298968178271,
      "grad_norm": 2.1232912269113573,
      "learning_rate": 1e-05,
      "loss": 1.3149,
      "step": 1073
    },
    {
      "epoch": 0.7449926298447932,
      "grad_norm": 1.933335743238272,
      "learning_rate": 1e-05,
      "loss": 1.3036,
      "step": 1074
    },
    {
      "epoch": 0.7456862915113154,
      "grad_norm": 1.7161884340751359,
      "learning_rate": 1e-05,
      "loss": 1.337,
      "step": 1075
    },
    {
      "epoch": 0.7463799531778376,
      "grad_norm": 1.8821904641272937,
      "learning_rate": 1e-05,
      "loss": 1.3484,
      "step": 1076
    },
    {
      "epoch": 0.7470736148443596,
      "grad_norm": 1.8420209136091366,
      "learning_rate": 1e-05,
      "loss": 1.3255,
      "step": 1077
    },
    {
      "epoch": 0.7477672765108818,
      "grad_norm": 1.6844962997358441,
      "learning_rate": 1e-05,
      "loss": 1.3763,
      "step": 1078
    },
    {
      "epoch": 0.748460938177404,
      "grad_norm": 1.6853533211643357,
      "learning_rate": 1e-05,
      "loss": 1.334,
      "step": 1079
    },
    {
      "epoch": 0.7491545998439261,
      "grad_norm": 1.7019512963652867,
      "learning_rate": 1e-05,
      "loss": 1.2932,
      "step": 1080
    },
    {
      "epoch": 0.7498482615104483,
      "grad_norm": 1.811962898514401,
      "learning_rate": 1e-05,
      "loss": 1.3587,
      "step": 1081
    },
    {
      "epoch": 0.7505419231769704,
      "grad_norm": 1.7755093149292678,
      "learning_rate": 1e-05,
      "loss": 1.2892,
      "step": 1082
    },
    {
      "epoch": 0.7512355848434926,
      "grad_norm": 1.8629164456603942,
      "learning_rate": 1e-05,
      "loss": 1.3037,
      "step": 1083
    },
    {
      "epoch": 0.7519292465100147,
      "grad_norm": 1.7886710463953541,
      "learning_rate": 1e-05,
      "loss": 1.2803,
      "step": 1084
    },
    {
      "epoch": 0.7526229081765369,
      "grad_norm": 1.695860963172146,
      "learning_rate": 1e-05,
      "loss": 1.4153,
      "step": 1085
    },
    {
      "epoch": 0.7533165698430591,
      "grad_norm": 1.9667102987846012,
      "learning_rate": 1e-05,
      "loss": 1.3436,
      "step": 1086
    },
    {
      "epoch": 0.7540102315095812,
      "grad_norm": 1.7749867238254802,
      "learning_rate": 1e-05,
      "loss": 1.3543,
      "step": 1087
    },
    {
      "epoch": 0.7547038931761033,
      "grad_norm": 1.8601842356859748,
      "learning_rate": 1e-05,
      "loss": 1.3261,
      "step": 1088
    },
    {
      "epoch": 0.7553975548426255,
      "grad_norm": 1.7933714883604228,
      "learning_rate": 1e-05,
      "loss": 1.375,
      "step": 1089
    },
    {
      "epoch": 0.7560912165091477,
      "grad_norm": 1.9108510451623624,
      "learning_rate": 1e-05,
      "loss": 1.3306,
      "step": 1090
    },
    {
      "epoch": 0.7567848781756699,
      "grad_norm": 1.7031275623765503,
      "learning_rate": 1e-05,
      "loss": 1.3776,
      "step": 1091
    },
    {
      "epoch": 0.7574785398421919,
      "grad_norm": 1.723756241244783,
      "learning_rate": 1e-05,
      "loss": 1.3069,
      "step": 1092
    },
    {
      "epoch": 0.7581722015087141,
      "grad_norm": 1.790196818419589,
      "learning_rate": 1e-05,
      "loss": 1.2887,
      "step": 1093
    },
    {
      "epoch": 0.7588658631752363,
      "grad_norm": 1.6853767916962972,
      "learning_rate": 1e-05,
      "loss": 1.3328,
      "step": 1094
    },
    {
      "epoch": 0.7595595248417585,
      "grad_norm": 1.8283907475641092,
      "learning_rate": 1e-05,
      "loss": 1.3394,
      "step": 1095
    },
    {
      "epoch": 0.7602531865082806,
      "grad_norm": 1.7686661299844872,
      "learning_rate": 1e-05,
      "loss": 1.3213,
      "step": 1096
    },
    {
      "epoch": 0.7609468481748027,
      "grad_norm": 2.0255015885014678,
      "learning_rate": 1e-05,
      "loss": 1.3553,
      "step": 1097
    },
    {
      "epoch": 0.7616405098413249,
      "grad_norm": 1.887116897314805,
      "learning_rate": 1e-05,
      "loss": 1.2652,
      "step": 1098
    },
    {
      "epoch": 0.7623341715078471,
      "grad_norm": 2.0827192956718816,
      "learning_rate": 1e-05,
      "loss": 1.2931,
      "step": 1099
    },
    {
      "epoch": 0.7630278331743692,
      "grad_norm": 1.6696475938133672,
      "learning_rate": 1e-05,
      "loss": 1.344,
      "step": 1100
    },
    {
      "epoch": 0.7637214948408914,
      "grad_norm": 1.8696319148514944,
      "learning_rate": 1e-05,
      "loss": 1.2875,
      "step": 1101
    },
    {
      "epoch": 0.7644151565074135,
      "grad_norm": 1.9161318005356125,
      "learning_rate": 1e-05,
      "loss": 1.3039,
      "step": 1102
    },
    {
      "epoch": 0.7651088181739356,
      "grad_norm": 1.7357326775365562,
      "learning_rate": 1e-05,
      "loss": 1.327,
      "step": 1103
    },
    {
      "epoch": 0.7658024798404578,
      "grad_norm": 1.9374495795431432,
      "learning_rate": 1e-05,
      "loss": 1.3118,
      "step": 1104
    },
    {
      "epoch": 0.76649614150698,
      "grad_norm": 1.8970035705740447,
      "learning_rate": 1e-05,
      "loss": 1.3518,
      "step": 1105
    },
    {
      "epoch": 0.7671898031735022,
      "grad_norm": 1.7310701480408488,
      "learning_rate": 1e-05,
      "loss": 1.3011,
      "step": 1106
    },
    {
      "epoch": 0.7678834648400242,
      "grad_norm": 1.6347319795022848,
      "learning_rate": 1e-05,
      "loss": 1.3885,
      "step": 1107
    },
    {
      "epoch": 0.7685771265065464,
      "grad_norm": 1.7389954204005627,
      "learning_rate": 1e-05,
      "loss": 1.3407,
      "step": 1108
    },
    {
      "epoch": 0.7692707881730686,
      "grad_norm": 1.8545763638618038,
      "learning_rate": 1e-05,
      "loss": 1.3475,
      "step": 1109
    },
    {
      "epoch": 0.7699644498395908,
      "grad_norm": 1.830290361513446,
      "learning_rate": 1e-05,
      "loss": 1.3372,
      "step": 1110
    },
    {
      "epoch": 0.7706581115061129,
      "grad_norm": 1.7554211037949135,
      "learning_rate": 1e-05,
      "loss": 1.3382,
      "step": 1111
    },
    {
      "epoch": 0.771351773172635,
      "grad_norm": 1.791182051330342,
      "learning_rate": 1e-05,
      "loss": 1.2865,
      "step": 1112
    },
    {
      "epoch": 0.7720454348391572,
      "grad_norm": 1.7944697955298015,
      "learning_rate": 1e-05,
      "loss": 1.3521,
      "step": 1113
    },
    {
      "epoch": 0.7727390965056794,
      "grad_norm": 1.7280244820641575,
      "learning_rate": 1e-05,
      "loss": 1.2855,
      "step": 1114
    },
    {
      "epoch": 0.7734327581722015,
      "grad_norm": 1.646134246431486,
      "learning_rate": 1e-05,
      "loss": 1.3443,
      "step": 1115
    },
    {
      "epoch": 0.7741264198387237,
      "grad_norm": 1.9614831482482202,
      "learning_rate": 1e-05,
      "loss": 1.3511,
      "step": 1116
    },
    {
      "epoch": 0.7748200815052458,
      "grad_norm": 1.8931116828060321,
      "learning_rate": 1e-05,
      "loss": 1.3014,
      "step": 1117
    },
    {
      "epoch": 0.775513743171768,
      "grad_norm": 1.766409312274434,
      "learning_rate": 1e-05,
      "loss": 1.3368,
      "step": 1118
    },
    {
      "epoch": 0.7762074048382901,
      "grad_norm": 1.7681164259480207,
      "learning_rate": 1e-05,
      "loss": 1.3277,
      "step": 1119
    },
    {
      "epoch": 0.7769010665048123,
      "grad_norm": 1.8795478880896714,
      "learning_rate": 1e-05,
      "loss": 1.358,
      "step": 1120
    },
    {
      "epoch": 0.7775947281713345,
      "grad_norm": 1.7552042041280342,
      "learning_rate": 1e-05,
      "loss": 1.338,
      "step": 1121
    },
    {
      "epoch": 0.7782883898378565,
      "grad_norm": 1.6825666179566667,
      "learning_rate": 1e-05,
      "loss": 1.3468,
      "step": 1122
    },
    {
      "epoch": 0.7789820515043787,
      "grad_norm": 2.069237558777561,
      "learning_rate": 1e-05,
      "loss": 1.3285,
      "step": 1123
    },
    {
      "epoch": 0.7796757131709009,
      "grad_norm": 1.8945277365950668,
      "learning_rate": 1e-05,
      "loss": 1.3457,
      "step": 1124
    },
    {
      "epoch": 0.7803693748374231,
      "grad_norm": 1.804604515643157,
      "learning_rate": 1e-05,
      "loss": 1.3198,
      "step": 1125
    },
    {
      "epoch": 0.7810630365039452,
      "grad_norm": 1.6687410426653992,
      "learning_rate": 1e-05,
      "loss": 1.3571,
      "step": 1126
    },
    {
      "epoch": 0.7817566981704673,
      "grad_norm": 1.8408255487647456,
      "learning_rate": 1e-05,
      "loss": 1.348,
      "step": 1127
    },
    {
      "epoch": 0.7824503598369895,
      "grad_norm": 1.7473823693404393,
      "learning_rate": 1e-05,
      "loss": 1.3449,
      "step": 1128
    },
    {
      "epoch": 0.7831440215035117,
      "grad_norm": 1.762367934431706,
      "learning_rate": 1e-05,
      "loss": 1.3296,
      "step": 1129
    },
    {
      "epoch": 0.7838376831700338,
      "grad_norm": 1.8065045373977573,
      "learning_rate": 1e-05,
      "loss": 1.3088,
      "step": 1130
    },
    {
      "epoch": 0.784531344836556,
      "grad_norm": 1.7702834178449094,
      "learning_rate": 1e-05,
      "loss": 1.3251,
      "step": 1131
    },
    {
      "epoch": 0.7852250065030781,
      "grad_norm": 1.684651956430982,
      "learning_rate": 1e-05,
      "loss": 1.3221,
      "step": 1132
    },
    {
      "epoch": 0.7859186681696003,
      "grad_norm": 1.718401813004295,
      "learning_rate": 1e-05,
      "loss": 1.3583,
      "step": 1133
    },
    {
      "epoch": 0.7866123298361224,
      "grad_norm": 1.7593076318939966,
      "learning_rate": 1e-05,
      "loss": 1.3481,
      "step": 1134
    },
    {
      "epoch": 0.7873059915026446,
      "grad_norm": 1.9066726353857328,
      "learning_rate": 1e-05,
      "loss": 1.298,
      "step": 1135
    },
    {
      "epoch": 0.7879996531691668,
      "grad_norm": 1.666639956360219,
      "learning_rate": 1e-05,
      "loss": 1.3321,
      "step": 1136
    },
    {
      "epoch": 0.7886933148356889,
      "grad_norm": 1.722161549449429,
      "learning_rate": 1e-05,
      "loss": 1.3081,
      "step": 1137
    },
    {
      "epoch": 0.789386976502211,
      "grad_norm": 1.8556050317918311,
      "learning_rate": 1e-05,
      "loss": 1.3155,
      "step": 1138
    },
    {
      "epoch": 0.7900806381687332,
      "grad_norm": 1.906163646274444,
      "learning_rate": 1e-05,
      "loss": 1.3127,
      "step": 1139
    },
    {
      "epoch": 0.7907742998352554,
      "grad_norm": 1.91777805112317,
      "learning_rate": 1e-05,
      "loss": 1.3267,
      "step": 1140
    },
    {
      "epoch": 0.7914679615017776,
      "grad_norm": 1.758622480343078,
      "learning_rate": 1e-05,
      "loss": 1.3604,
      "step": 1141
    },
    {
      "epoch": 0.7921616231682996,
      "grad_norm": 1.818304243791579,
      "learning_rate": 1e-05,
      "loss": 1.3744,
      "step": 1142
    },
    {
      "epoch": 0.7928552848348218,
      "grad_norm": 1.7096779670968696,
      "learning_rate": 1e-05,
      "loss": 1.2967,
      "step": 1143
    },
    {
      "epoch": 0.793548946501344,
      "grad_norm": 1.8288634798052434,
      "learning_rate": 1e-05,
      "loss": 1.3836,
      "step": 1144
    },
    {
      "epoch": 0.7942426081678661,
      "grad_norm": 1.837569087516797,
      "learning_rate": 1e-05,
      "loss": 1.3306,
      "step": 1145
    },
    {
      "epoch": 0.7949362698343883,
      "grad_norm": 1.757663683708377,
      "learning_rate": 1e-05,
      "loss": 1.3289,
      "step": 1146
    },
    {
      "epoch": 0.7956299315009104,
      "grad_norm": 1.5733908350037635,
      "learning_rate": 1e-05,
      "loss": 1.3293,
      "step": 1147
    },
    {
      "epoch": 0.7963235931674326,
      "grad_norm": 1.7006097263550368,
      "learning_rate": 1e-05,
      "loss": 1.3544,
      "step": 1148
    },
    {
      "epoch": 0.7970172548339547,
      "grad_norm": 1.852767855910649,
      "learning_rate": 1e-05,
      "loss": 1.367,
      "step": 1149
    },
    {
      "epoch": 0.7977109165004769,
      "grad_norm": 1.8236237376393378,
      "learning_rate": 1e-05,
      "loss": 1.3298,
      "step": 1150
    },
    {
      "epoch": 0.7984045781669991,
      "grad_norm": 1.7542840584318944,
      "learning_rate": 1e-05,
      "loss": 1.3305,
      "step": 1151
    },
    {
      "epoch": 0.7990982398335212,
      "grad_norm": 1.7363578108911824,
      "learning_rate": 1e-05,
      "loss": 1.312,
      "step": 1152
    },
    {
      "epoch": 0.7997919015000433,
      "grad_norm": 1.7447796524158012,
      "learning_rate": 1e-05,
      "loss": 1.3146,
      "step": 1153
    },
    {
      "epoch": 0.8004855631665655,
      "grad_norm": 1.7998354010324689,
      "learning_rate": 1e-05,
      "loss": 1.2651,
      "step": 1154
    },
    {
      "epoch": 0.8011792248330877,
      "grad_norm": 1.72370764035216,
      "learning_rate": 1e-05,
      "loss": 1.3313,
      "step": 1155
    },
    {
      "epoch": 0.8018728864996099,
      "grad_norm": 1.9027302235993584,
      "learning_rate": 1e-05,
      "loss": 1.3275,
      "step": 1156
    },
    {
      "epoch": 0.8025665481661319,
      "grad_norm": 1.767361577842723,
      "learning_rate": 1e-05,
      "loss": 1.281,
      "step": 1157
    },
    {
      "epoch": 0.8032602098326541,
      "grad_norm": 1.6957352595013901,
      "learning_rate": 1e-05,
      "loss": 1.339,
      "step": 1158
    },
    {
      "epoch": 0.8039538714991763,
      "grad_norm": 1.8811200117431954,
      "learning_rate": 1e-05,
      "loss": 1.3694,
      "step": 1159
    },
    {
      "epoch": 0.8046475331656985,
      "grad_norm": 1.8732110913975033,
      "learning_rate": 1e-05,
      "loss": 1.3828,
      "step": 1160
    },
    {
      "epoch": 0.8053411948322206,
      "grad_norm": 1.806929633499466,
      "learning_rate": 1e-05,
      "loss": 1.293,
      "step": 1161
    },
    {
      "epoch": 0.8060348564987427,
      "grad_norm": 1.9234019762346322,
      "learning_rate": 1e-05,
      "loss": 1.356,
      "step": 1162
    },
    {
      "epoch": 0.8067285181652649,
      "grad_norm": 1.8612299857621706,
      "learning_rate": 1e-05,
      "loss": 1.2944,
      "step": 1163
    },
    {
      "epoch": 0.8074221798317871,
      "grad_norm": 1.8949079530850956,
      "learning_rate": 1e-05,
      "loss": 1.2759,
      "step": 1164
    },
    {
      "epoch": 0.8081158414983092,
      "grad_norm": 1.835747159847785,
      "learning_rate": 1e-05,
      "loss": 1.2959,
      "step": 1165
    },
    {
      "epoch": 0.8088095031648314,
      "grad_norm": 1.6994420208999754,
      "learning_rate": 1e-05,
      "loss": 1.3546,
      "step": 1166
    },
    {
      "epoch": 0.8095031648313535,
      "grad_norm": 1.7058565978885374,
      "learning_rate": 1e-05,
      "loss": 1.3016,
      "step": 1167
    },
    {
      "epoch": 0.8101968264978756,
      "grad_norm": 1.7202328196552465,
      "learning_rate": 1e-05,
      "loss": 1.3244,
      "step": 1168
    },
    {
      "epoch": 0.8108904881643978,
      "grad_norm": 1.5969257126998517,
      "learning_rate": 1e-05,
      "loss": 1.2941,
      "step": 1169
    },
    {
      "epoch": 0.81158414983092,
      "grad_norm": 2.033122314469678,
      "learning_rate": 1e-05,
      "loss": 1.3112,
      "step": 1170
    },
    {
      "epoch": 0.8122778114974422,
      "grad_norm": 1.818607135695885,
      "learning_rate": 1e-05,
      "loss": 1.3372,
      "step": 1171
    },
    {
      "epoch": 0.8129714731639642,
      "grad_norm": 1.7332347851944176,
      "learning_rate": 1e-05,
      "loss": 1.2963,
      "step": 1172
    },
    {
      "epoch": 0.8136651348304864,
      "grad_norm": 1.79912386221675,
      "learning_rate": 1e-05,
      "loss": 1.3192,
      "step": 1173
    },
    {
      "epoch": 0.8143587964970086,
      "grad_norm": 1.759247644611021,
      "learning_rate": 1e-05,
      "loss": 1.3087,
      "step": 1174
    },
    {
      "epoch": 0.8150524581635308,
      "grad_norm": 1.81196721719679,
      "learning_rate": 1e-05,
      "loss": 1.3347,
      "step": 1175
    },
    {
      "epoch": 0.8157461198300529,
      "grad_norm": 1.8329350171829653,
      "learning_rate": 1e-05,
      "loss": 1.347,
      "step": 1176
    },
    {
      "epoch": 0.816439781496575,
      "grad_norm": 2.097683322497983,
      "learning_rate": 1e-05,
      "loss": 1.3187,
      "step": 1177
    },
    {
      "epoch": 0.8171334431630972,
      "grad_norm": 1.682730633916439,
      "learning_rate": 1e-05,
      "loss": 1.326,
      "step": 1178
    },
    {
      "epoch": 0.8178271048296194,
      "grad_norm": 1.7967406525079352,
      "learning_rate": 1e-05,
      "loss": 1.3229,
      "step": 1179
    },
    {
      "epoch": 0.8185207664961415,
      "grad_norm": 1.7155776890591734,
      "learning_rate": 1e-05,
      "loss": 1.3143,
      "step": 1180
    },
    {
      "epoch": 0.8192144281626637,
      "grad_norm": 1.731948891871397,
      "learning_rate": 1e-05,
      "loss": 1.3044,
      "step": 1181
    },
    {
      "epoch": 0.8199080898291858,
      "grad_norm": 1.6258267218121343,
      "learning_rate": 1e-05,
      "loss": 1.2966,
      "step": 1182
    },
    {
      "epoch": 0.820601751495708,
      "grad_norm": 1.8947923997247118,
      "learning_rate": 1e-05,
      "loss": 1.25,
      "step": 1183
    },
    {
      "epoch": 0.8212954131622301,
      "grad_norm": 1.9268221611589265,
      "learning_rate": 1e-05,
      "loss": 1.363,
      "step": 1184
    },
    {
      "epoch": 0.8219890748287523,
      "grad_norm": 1.6151381049081766,
      "learning_rate": 1e-05,
      "loss": 1.3323,
      "step": 1185
    },
    {
      "epoch": 0.8226827364952745,
      "grad_norm": 1.6754745238527986,
      "learning_rate": 1e-05,
      "loss": 1.2736,
      "step": 1186
    },
    {
      "epoch": 0.8233763981617965,
      "grad_norm": 1.6566597340456615,
      "learning_rate": 1e-05,
      "loss": 1.3292,
      "step": 1187
    },
    {
      "epoch": 0.8240700598283187,
      "grad_norm": 1.7868911126447826,
      "learning_rate": 1e-05,
      "loss": 1.3146,
      "step": 1188
    },
    {
      "epoch": 0.8247637214948409,
      "grad_norm": 1.7395269678979228,
      "learning_rate": 1e-05,
      "loss": 1.3172,
      "step": 1189
    },
    {
      "epoch": 0.8254573831613631,
      "grad_norm": 1.846976443991522,
      "learning_rate": 1e-05,
      "loss": 1.2845,
      "step": 1190
    },
    {
      "epoch": 0.8261510448278852,
      "grad_norm": 1.9263142545199925,
      "learning_rate": 1e-05,
      "loss": 1.3218,
      "step": 1191
    },
    {
      "epoch": 0.8268447064944073,
      "grad_norm": 1.6753572265693735,
      "learning_rate": 1e-05,
      "loss": 1.3347,
      "step": 1192
    },
    {
      "epoch": 0.8275383681609295,
      "grad_norm": 1.8402922863638769,
      "learning_rate": 1e-05,
      "loss": 1.3389,
      "step": 1193
    },
    {
      "epoch": 0.8282320298274517,
      "grad_norm": 1.7003805951507205,
      "learning_rate": 1e-05,
      "loss": 1.3016,
      "step": 1194
    },
    {
      "epoch": 0.8289256914939738,
      "grad_norm": 1.7016887764707231,
      "learning_rate": 1e-05,
      "loss": 1.3312,
      "step": 1195
    },
    {
      "epoch": 0.829619353160496,
      "grad_norm": 1.7450143183218212,
      "learning_rate": 1e-05,
      "loss": 1.303,
      "step": 1196
    },
    {
      "epoch": 0.8303130148270181,
      "grad_norm": 1.7676264585702774,
      "learning_rate": 1e-05,
      "loss": 1.317,
      "step": 1197
    },
    {
      "epoch": 0.8310066764935403,
      "grad_norm": 2.0267151482602177,
      "learning_rate": 1e-05,
      "loss": 1.2687,
      "step": 1198
    },
    {
      "epoch": 0.8317003381600624,
      "grad_norm": 1.6648828369964448,
      "learning_rate": 1e-05,
      "loss": 1.3436,
      "step": 1199
    },
    {
      "epoch": 0.8323939998265846,
      "grad_norm": 1.7437721633894316,
      "learning_rate": 1e-05,
      "loss": 1.2719,
      "step": 1200
    },
    {
      "epoch": 0.8330876614931068,
      "grad_norm": 1.76080267027574,
      "learning_rate": 1e-05,
      "loss": 1.3364,
      "step": 1201
    },
    {
      "epoch": 0.8337813231596289,
      "grad_norm": 1.6023127093441858,
      "learning_rate": 1e-05,
      "loss": 1.3567,
      "step": 1202
    },
    {
      "epoch": 0.834474984826151,
      "grad_norm": 1.807984298617728,
      "learning_rate": 1e-05,
      "loss": 1.3243,
      "step": 1203
    },
    {
      "epoch": 0.8351686464926732,
      "grad_norm": 1.7507073872441543,
      "learning_rate": 1e-05,
      "loss": 1.2748,
      "step": 1204
    },
    {
      "epoch": 0.8358623081591954,
      "grad_norm": 1.7340226078649954,
      "learning_rate": 1e-05,
      "loss": 1.3236,
      "step": 1205
    },
    {
      "epoch": 0.8365559698257176,
      "grad_norm": 1.8128370813992278,
      "learning_rate": 1e-05,
      "loss": 1.3398,
      "step": 1206
    },
    {
      "epoch": 0.8372496314922396,
      "grad_norm": 1.8385313973464554,
      "learning_rate": 1e-05,
      "loss": 1.3244,
      "step": 1207
    },
    {
      "epoch": 0.8379432931587618,
      "grad_norm": 1.8202487072236857,
      "learning_rate": 1e-05,
      "loss": 1.2841,
      "step": 1208
    },
    {
      "epoch": 0.838636954825284,
      "grad_norm": 1.8226882804903315,
      "learning_rate": 1e-05,
      "loss": 1.3251,
      "step": 1209
    },
    {
      "epoch": 0.8393306164918061,
      "grad_norm": 1.7796315284012894,
      "learning_rate": 1e-05,
      "loss": 1.3503,
      "step": 1210
    },
    {
      "epoch": 0.8400242781583283,
      "grad_norm": 1.770507934678552,
      "learning_rate": 1e-05,
      "loss": 1.2873,
      "step": 1211
    },
    {
      "epoch": 0.8407179398248504,
      "grad_norm": 1.7983739916772261,
      "learning_rate": 1e-05,
      "loss": 1.3309,
      "step": 1212
    },
    {
      "epoch": 0.8414116014913726,
      "grad_norm": 1.8581265442393458,
      "learning_rate": 1e-05,
      "loss": 1.2976,
      "step": 1213
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 1.9709414587672909,
      "learning_rate": 1e-05,
      "loss": 1.3244,
      "step": 1214
    },
    {
      "epoch": 0.8427989248244169,
      "grad_norm": 1.934195079670126,
      "learning_rate": 1e-05,
      "loss": 1.3197,
      "step": 1215
    },
    {
      "epoch": 0.8434925864909391,
      "grad_norm": 1.5905446332751805,
      "learning_rate": 1e-05,
      "loss": 1.3144,
      "step": 1216
    },
    {
      "epoch": 0.8441862481574612,
      "grad_norm": 1.8012514681125382,
      "learning_rate": 1e-05,
      "loss": 1.2984,
      "step": 1217
    },
    {
      "epoch": 0.8448799098239833,
      "grad_norm": 1.8322906369289444,
      "learning_rate": 1e-05,
      "loss": 1.3365,
      "step": 1218
    },
    {
      "epoch": 0.8455735714905055,
      "grad_norm": 1.7098942463292028,
      "learning_rate": 1e-05,
      "loss": 1.2994,
      "step": 1219
    },
    {
      "epoch": 0.8462672331570277,
      "grad_norm": 1.806262096500155,
      "learning_rate": 1e-05,
      "loss": 1.3135,
      "step": 1220
    },
    {
      "epoch": 0.8469608948235499,
      "grad_norm": 1.8088243417346301,
      "learning_rate": 1e-05,
      "loss": 1.2548,
      "step": 1221
    },
    {
      "epoch": 0.8476545564900719,
      "grad_norm": 1.7615794277621235,
      "learning_rate": 1e-05,
      "loss": 1.369,
      "step": 1222
    },
    {
      "epoch": 0.8483482181565941,
      "grad_norm": 1.7446487987023735,
      "learning_rate": 1e-05,
      "loss": 1.3417,
      "step": 1223
    },
    {
      "epoch": 0.8490418798231163,
      "grad_norm": 1.9292547174186359,
      "learning_rate": 1e-05,
      "loss": 1.2619,
      "step": 1224
    },
    {
      "epoch": 0.8497355414896385,
      "grad_norm": 1.7385855689942564,
      "learning_rate": 1e-05,
      "loss": 1.3367,
      "step": 1225
    },
    {
      "epoch": 0.8504292031561606,
      "grad_norm": 1.7893857799911939,
      "learning_rate": 1e-05,
      "loss": 1.3231,
      "step": 1226
    },
    {
      "epoch": 0.8511228648226827,
      "grad_norm": 1.7904188745955463,
      "learning_rate": 1e-05,
      "loss": 1.273,
      "step": 1227
    },
    {
      "epoch": 0.8518165264892049,
      "grad_norm": 1.7311471203311328,
      "learning_rate": 1e-05,
      "loss": 1.3036,
      "step": 1228
    },
    {
      "epoch": 0.852510188155727,
      "grad_norm": 1.8405281776102504,
      "learning_rate": 1e-05,
      "loss": 1.3266,
      "step": 1229
    },
    {
      "epoch": 0.8532038498222492,
      "grad_norm": 1.7910111933959387,
      "learning_rate": 1e-05,
      "loss": 1.332,
      "step": 1230
    },
    {
      "epoch": 0.8538975114887714,
      "grad_norm": 1.692055737327334,
      "learning_rate": 1e-05,
      "loss": 1.2594,
      "step": 1231
    },
    {
      "epoch": 0.8545911731552935,
      "grad_norm": 1.715888470034707,
      "learning_rate": 1e-05,
      "loss": 1.2974,
      "step": 1232
    },
    {
      "epoch": 0.8552848348218156,
      "grad_norm": 1.6920575665727629,
      "learning_rate": 1e-05,
      "loss": 1.3081,
      "step": 1233
    },
    {
      "epoch": 0.8559784964883378,
      "grad_norm": 1.8820212499017275,
      "learning_rate": 1e-05,
      "loss": 1.2579,
      "step": 1234
    },
    {
      "epoch": 0.85667215815486,
      "grad_norm": 1.7590796982626669,
      "learning_rate": 1e-05,
      "loss": 1.346,
      "step": 1235
    },
    {
      "epoch": 0.8573658198213822,
      "grad_norm": 1.7175897993182907,
      "learning_rate": 1e-05,
      "loss": 1.2952,
      "step": 1236
    },
    {
      "epoch": 0.8580594814879042,
      "grad_norm": 1.8872407973916834,
      "learning_rate": 1e-05,
      "loss": 1.2678,
      "step": 1237
    },
    {
      "epoch": 0.8587531431544264,
      "grad_norm": 1.820685981576064,
      "learning_rate": 1e-05,
      "loss": 1.3161,
      "step": 1238
    },
    {
      "epoch": 0.8594468048209486,
      "grad_norm": 1.7106563905509156,
      "learning_rate": 1e-05,
      "loss": 1.2805,
      "step": 1239
    },
    {
      "epoch": 0.8601404664874708,
      "grad_norm": 1.835121012618051,
      "learning_rate": 1e-05,
      "loss": 1.3411,
      "step": 1240
    },
    {
      "epoch": 0.8608341281539929,
      "grad_norm": 1.759648204282831,
      "learning_rate": 1e-05,
      "loss": 1.3169,
      "step": 1241
    },
    {
      "epoch": 0.861527789820515,
      "grad_norm": 1.8108439058590786,
      "learning_rate": 1e-05,
      "loss": 1.3365,
      "step": 1242
    },
    {
      "epoch": 0.8622214514870372,
      "grad_norm": 1.8264838456429158,
      "learning_rate": 1e-05,
      "loss": 1.3122,
      "step": 1243
    },
    {
      "epoch": 0.8629151131535594,
      "grad_norm": 1.8561243343760983,
      "learning_rate": 1e-05,
      "loss": 1.3126,
      "step": 1244
    },
    {
      "epoch": 0.8636087748200815,
      "grad_norm": 1.7744496574339597,
      "learning_rate": 1e-05,
      "loss": 1.3419,
      "step": 1245
    },
    {
      "epoch": 0.8643024364866037,
      "grad_norm": 1.7102771712690799,
      "learning_rate": 1e-05,
      "loss": 1.3913,
      "step": 1246
    },
    {
      "epoch": 0.8649960981531258,
      "grad_norm": 1.637011563504676,
      "learning_rate": 1e-05,
      "loss": 1.3294,
      "step": 1247
    },
    {
      "epoch": 0.865689759819648,
      "grad_norm": 1.7378852441605857,
      "learning_rate": 1e-05,
      "loss": 1.3604,
      "step": 1248
    },
    {
      "epoch": 0.8663834214861701,
      "grad_norm": 1.892145137680064,
      "learning_rate": 1e-05,
      "loss": 1.3839,
      "step": 1249
    },
    {
      "epoch": 0.8670770831526923,
      "grad_norm": 1.7114640198993971,
      "learning_rate": 1e-05,
      "loss": 1.3164,
      "step": 1250
    },
    {
      "epoch": 0.8677707448192145,
      "grad_norm": 1.857472545380348,
      "learning_rate": 1e-05,
      "loss": 1.3409,
      "step": 1251
    },
    {
      "epoch": 0.8684644064857365,
      "grad_norm": 1.6636749531242045,
      "learning_rate": 1e-05,
      "loss": 1.3103,
      "step": 1252
    },
    {
      "epoch": 0.8691580681522587,
      "grad_norm": 1.853753890533686,
      "learning_rate": 1e-05,
      "loss": 1.3372,
      "step": 1253
    },
    {
      "epoch": 0.8698517298187809,
      "grad_norm": 1.7274626500933439,
      "learning_rate": 1e-05,
      "loss": 1.3198,
      "step": 1254
    },
    {
      "epoch": 0.8705453914853031,
      "grad_norm": 1.7681696772117461,
      "learning_rate": 1e-05,
      "loss": 1.3451,
      "step": 1255
    },
    {
      "epoch": 0.8712390531518251,
      "grad_norm": 1.7074991789276823,
      "learning_rate": 1e-05,
      "loss": 1.306,
      "step": 1256
    },
    {
      "epoch": 0.8719327148183473,
      "grad_norm": 1.8022859354848255,
      "learning_rate": 1e-05,
      "loss": 1.307,
      "step": 1257
    },
    {
      "epoch": 0.8726263764848695,
      "grad_norm": 1.722826958334196,
      "learning_rate": 1e-05,
      "loss": 1.348,
      "step": 1258
    },
    {
      "epoch": 0.8733200381513917,
      "grad_norm": 1.6278820667115972,
      "learning_rate": 1e-05,
      "loss": 1.3156,
      "step": 1259
    },
    {
      "epoch": 0.8740136998179138,
      "grad_norm": 1.8615627558866032,
      "learning_rate": 1e-05,
      "loss": 1.3642,
      "step": 1260
    },
    {
      "epoch": 0.874707361484436,
      "grad_norm": 1.6977360433618884,
      "learning_rate": 1e-05,
      "loss": 1.3661,
      "step": 1261
    },
    {
      "epoch": 0.8754010231509581,
      "grad_norm": 1.7119935578912018,
      "learning_rate": 1e-05,
      "loss": 1.343,
      "step": 1262
    },
    {
      "epoch": 0.8760946848174803,
      "grad_norm": 1.8828273441618772,
      "learning_rate": 1e-05,
      "loss": 1.3286,
      "step": 1263
    },
    {
      "epoch": 0.8767883464840024,
      "grad_norm": 1.7991745556104566,
      "learning_rate": 1e-05,
      "loss": 1.2877,
      "step": 1264
    },
    {
      "epoch": 0.8774820081505246,
      "grad_norm": 1.7692125953469466,
      "learning_rate": 1e-05,
      "loss": 1.3132,
      "step": 1265
    },
    {
      "epoch": 0.8781756698170468,
      "grad_norm": 1.763539004235656,
      "learning_rate": 1e-05,
      "loss": 1.2516,
      "step": 1266
    },
    {
      "epoch": 0.8788693314835689,
      "grad_norm": 1.8097604071898914,
      "learning_rate": 1e-05,
      "loss": 1.322,
      "step": 1267
    },
    {
      "epoch": 0.879562993150091,
      "grad_norm": 1.8504952339835505,
      "learning_rate": 1e-05,
      "loss": 1.3257,
      "step": 1268
    },
    {
      "epoch": 0.8802566548166132,
      "grad_norm": 1.8228681067260823,
      "learning_rate": 1e-05,
      "loss": 1.3207,
      "step": 1269
    },
    {
      "epoch": 0.8809503164831354,
      "grad_norm": 1.8569141776272553,
      "learning_rate": 1e-05,
      "loss": 1.3415,
      "step": 1270
    },
    {
      "epoch": 0.8816439781496576,
      "grad_norm": 1.8653037577154865,
      "learning_rate": 1e-05,
      "loss": 1.2882,
      "step": 1271
    },
    {
      "epoch": 0.8823376398161796,
      "grad_norm": 1.7842091027351248,
      "learning_rate": 1e-05,
      "loss": 1.3011,
      "step": 1272
    },
    {
      "epoch": 0.8830313014827018,
      "grad_norm": 1.865231093038103,
      "learning_rate": 1e-05,
      "loss": 1.3421,
      "step": 1273
    },
    {
      "epoch": 0.883724963149224,
      "grad_norm": 1.8445484018508556,
      "learning_rate": 1e-05,
      "loss": 1.2994,
      "step": 1274
    },
    {
      "epoch": 0.8844186248157461,
      "grad_norm": 2.0112180595207585,
      "learning_rate": 1e-05,
      "loss": 1.3491,
      "step": 1275
    },
    {
      "epoch": 0.8851122864822683,
      "grad_norm": 1.9313320335876165,
      "learning_rate": 1e-05,
      "loss": 1.2972,
      "step": 1276
    },
    {
      "epoch": 0.8858059481487904,
      "grad_norm": 1.766314463801755,
      "learning_rate": 1e-05,
      "loss": 1.3551,
      "step": 1277
    },
    {
      "epoch": 0.8864996098153126,
      "grad_norm": 1.6939200547787165,
      "learning_rate": 1e-05,
      "loss": 1.3346,
      "step": 1278
    },
    {
      "epoch": 0.8871932714818347,
      "grad_norm": 1.6553779025228499,
      "learning_rate": 1e-05,
      "loss": 1.3331,
      "step": 1279
    },
    {
      "epoch": 0.8878869331483569,
      "grad_norm": 1.672522216377223,
      "learning_rate": 1e-05,
      "loss": 1.3143,
      "step": 1280
    },
    {
      "epoch": 0.888580594814879,
      "grad_norm": 1.6477930929287763,
      "learning_rate": 1e-05,
      "loss": 1.2655,
      "step": 1281
    },
    {
      "epoch": 0.8892742564814012,
      "grad_norm": 1.9082634726577539,
      "learning_rate": 1e-05,
      "loss": 1.3101,
      "step": 1282
    },
    {
      "epoch": 0.8899679181479233,
      "grad_norm": 1.6619545725705775,
      "learning_rate": 1e-05,
      "loss": 1.3134,
      "step": 1283
    },
    {
      "epoch": 0.8906615798144455,
      "grad_norm": 1.8186223504046093,
      "learning_rate": 1e-05,
      "loss": 1.3356,
      "step": 1284
    },
    {
      "epoch": 0.8913552414809677,
      "grad_norm": 1.6755642327768199,
      "learning_rate": 1e-05,
      "loss": 1.3005,
      "step": 1285
    },
    {
      "epoch": 0.8920489031474899,
      "grad_norm": 1.7174416264813488,
      "learning_rate": 1e-05,
      "loss": 1.3347,
      "step": 1286
    },
    {
      "epoch": 0.8927425648140119,
      "grad_norm": 1.7186710255881854,
      "learning_rate": 1e-05,
      "loss": 1.2908,
      "step": 1287
    },
    {
      "epoch": 0.8934362264805341,
      "grad_norm": 1.6346319252027368,
      "learning_rate": 1e-05,
      "loss": 1.2927,
      "step": 1288
    },
    {
      "epoch": 0.8941298881470563,
      "grad_norm": 1.8474495055475482,
      "learning_rate": 1e-05,
      "loss": 1.3255,
      "step": 1289
    },
    {
      "epoch": 0.8948235498135785,
      "grad_norm": 1.8788390378390694,
      "learning_rate": 1e-05,
      "loss": 1.3385,
      "step": 1290
    },
    {
      "epoch": 0.8955172114801006,
      "grad_norm": 1.629566356590922,
      "learning_rate": 1e-05,
      "loss": 1.3482,
      "step": 1291
    },
    {
      "epoch": 0.8962108731466227,
      "grad_norm": 1.7412289424540435,
      "learning_rate": 1e-05,
      "loss": 1.267,
      "step": 1292
    },
    {
      "epoch": 0.8969045348131449,
      "grad_norm": 1.5904673607353297,
      "learning_rate": 1e-05,
      "loss": 1.3107,
      "step": 1293
    },
    {
      "epoch": 0.897598196479667,
      "grad_norm": 1.7668320875825854,
      "learning_rate": 1e-05,
      "loss": 1.2968,
      "step": 1294
    },
    {
      "epoch": 0.8982918581461892,
      "grad_norm": 1.6434271863603802,
      "learning_rate": 1e-05,
      "loss": 1.3531,
      "step": 1295
    },
    {
      "epoch": 0.8989855198127114,
      "grad_norm": 1.6446049917777663,
      "learning_rate": 1e-05,
      "loss": 1.3067,
      "step": 1296
    },
    {
      "epoch": 0.8996791814792335,
      "grad_norm": 1.7114850493139724,
      "learning_rate": 1e-05,
      "loss": 1.2526,
      "step": 1297
    },
    {
      "epoch": 0.9003728431457556,
      "grad_norm": 1.7703812278053084,
      "learning_rate": 1e-05,
      "loss": 1.3284,
      "step": 1298
    },
    {
      "epoch": 0.9010665048122778,
      "grad_norm": 1.783598858952647,
      "learning_rate": 1e-05,
      "loss": 1.3449,
      "step": 1299
    },
    {
      "epoch": 0.9017601664788,
      "grad_norm": 1.7594643222023651,
      "learning_rate": 1e-05,
      "loss": 1.2865,
      "step": 1300
    },
    {
      "epoch": 0.9024538281453222,
      "grad_norm": 1.7979500686841217,
      "learning_rate": 1e-05,
      "loss": 1.2826,
      "step": 1301
    },
    {
      "epoch": 0.9031474898118442,
      "grad_norm": 1.7054221381249888,
      "learning_rate": 1e-05,
      "loss": 1.3525,
      "step": 1302
    },
    {
      "epoch": 0.9038411514783664,
      "grad_norm": 1.7423889164937596,
      "learning_rate": 1e-05,
      "loss": 1.3044,
      "step": 1303
    },
    {
      "epoch": 0.9045348131448886,
      "grad_norm": 1.8080520927021586,
      "learning_rate": 1e-05,
      "loss": 1.331,
      "step": 1304
    },
    {
      "epoch": 0.9052284748114108,
      "grad_norm": 1.9903383637954946,
      "learning_rate": 1e-05,
      "loss": 1.3213,
      "step": 1305
    },
    {
      "epoch": 0.9059221364779328,
      "grad_norm": 1.622554431537696,
      "learning_rate": 1e-05,
      "loss": 1.3307,
      "step": 1306
    },
    {
      "epoch": 0.906615798144455,
      "grad_norm": 1.891392187663753,
      "learning_rate": 1e-05,
      "loss": 1.3484,
      "step": 1307
    },
    {
      "epoch": 0.9073094598109772,
      "grad_norm": 1.5686215128270367,
      "learning_rate": 1e-05,
      "loss": 1.3101,
      "step": 1308
    },
    {
      "epoch": 0.9080031214774994,
      "grad_norm": 1.7879215692958745,
      "learning_rate": 1e-05,
      "loss": 1.3499,
      "step": 1309
    },
    {
      "epoch": 0.9086967831440215,
      "grad_norm": 1.7583743340224038,
      "learning_rate": 1e-05,
      "loss": 1.2757,
      "step": 1310
    },
    {
      "epoch": 0.9093904448105437,
      "grad_norm": 1.7396925828139032,
      "learning_rate": 1e-05,
      "loss": 1.2925,
      "step": 1311
    },
    {
      "epoch": 0.9100841064770658,
      "grad_norm": 1.7958710900831494,
      "learning_rate": 1e-05,
      "loss": 1.3058,
      "step": 1312
    },
    {
      "epoch": 0.910777768143588,
      "grad_norm": 1.764507973805378,
      "learning_rate": 1e-05,
      "loss": 1.297,
      "step": 1313
    },
    {
      "epoch": 0.9114714298101101,
      "grad_norm": 1.8550793576750044,
      "learning_rate": 1e-05,
      "loss": 1.3146,
      "step": 1314
    },
    {
      "epoch": 0.9121650914766323,
      "grad_norm": 1.6803054339754033,
      "learning_rate": 1e-05,
      "loss": 1.2966,
      "step": 1315
    },
    {
      "epoch": 0.9128587531431545,
      "grad_norm": 1.885929787117982,
      "learning_rate": 1e-05,
      "loss": 1.3201,
      "step": 1316
    },
    {
      "epoch": 0.9135524148096765,
      "grad_norm": 1.6131928750808537,
      "learning_rate": 1e-05,
      "loss": 1.2899,
      "step": 1317
    },
    {
      "epoch": 0.9142460764761987,
      "grad_norm": 1.8283426626966652,
      "learning_rate": 1e-05,
      "loss": 1.3149,
      "step": 1318
    },
    {
      "epoch": 0.9149397381427209,
      "grad_norm": 1.7895417190819627,
      "learning_rate": 1e-05,
      "loss": 1.3135,
      "step": 1319
    },
    {
      "epoch": 0.9156333998092431,
      "grad_norm": 1.6655898144804122,
      "learning_rate": 1e-05,
      "loss": 1.3506,
      "step": 1320
    },
    {
      "epoch": 0.9163270614757651,
      "grad_norm": 1.789395903422411,
      "learning_rate": 1e-05,
      "loss": 1.2944,
      "step": 1321
    },
    {
      "epoch": 0.9170207231422873,
      "grad_norm": 1.5554671892032146,
      "learning_rate": 1e-05,
      "loss": 1.3035,
      "step": 1322
    },
    {
      "epoch": 0.9177143848088095,
      "grad_norm": 1.7462183316659554,
      "learning_rate": 1e-05,
      "loss": 1.2871,
      "step": 1323
    },
    {
      "epoch": 0.9184080464753317,
      "grad_norm": 1.6407955316695355,
      "learning_rate": 1e-05,
      "loss": 1.2842,
      "step": 1324
    },
    {
      "epoch": 0.9191017081418538,
      "grad_norm": 1.8054786932821483,
      "learning_rate": 1e-05,
      "loss": 1.2801,
      "step": 1325
    },
    {
      "epoch": 0.919795369808376,
      "grad_norm": 1.7774882584688303,
      "learning_rate": 1e-05,
      "loss": 1.3111,
      "step": 1326
    },
    {
      "epoch": 0.9204890314748981,
      "grad_norm": 1.7487308168423374,
      "learning_rate": 1e-05,
      "loss": 1.3086,
      "step": 1327
    },
    {
      "epoch": 0.9211826931414203,
      "grad_norm": 1.830348412513128,
      "learning_rate": 1e-05,
      "loss": 1.2398,
      "step": 1328
    },
    {
      "epoch": 0.9218763548079424,
      "grad_norm": 1.8745249129153505,
      "learning_rate": 1e-05,
      "loss": 1.3174,
      "step": 1329
    },
    {
      "epoch": 0.9225700164744646,
      "grad_norm": 1.6997745347291242,
      "learning_rate": 1e-05,
      "loss": 1.2795,
      "step": 1330
    },
    {
      "epoch": 0.9232636781409868,
      "grad_norm": 1.6461004692498915,
      "learning_rate": 1e-05,
      "loss": 1.369,
      "step": 1331
    },
    {
      "epoch": 0.9239573398075089,
      "grad_norm": 1.867846237084916,
      "learning_rate": 1e-05,
      "loss": 1.3492,
      "step": 1332
    },
    {
      "epoch": 0.924651001474031,
      "grad_norm": 1.7338725633875975,
      "learning_rate": 1e-05,
      "loss": 1.3425,
      "step": 1333
    },
    {
      "epoch": 0.9253446631405532,
      "grad_norm": 1.7058365907688622,
      "learning_rate": 1e-05,
      "loss": 1.3289,
      "step": 1334
    },
    {
      "epoch": 0.9260383248070754,
      "grad_norm": 1.6812689204725293,
      "learning_rate": 1e-05,
      "loss": 1.3166,
      "step": 1335
    },
    {
      "epoch": 0.9267319864735976,
      "grad_norm": 1.7854594207936143,
      "learning_rate": 1e-05,
      "loss": 1.3357,
      "step": 1336
    },
    {
      "epoch": 0.9274256481401196,
      "grad_norm": 1.8632837097505413,
      "learning_rate": 1e-05,
      "loss": 1.2385,
      "step": 1337
    },
    {
      "epoch": 0.9281193098066418,
      "grad_norm": 1.838541328017032,
      "learning_rate": 1e-05,
      "loss": 1.2703,
      "step": 1338
    },
    {
      "epoch": 0.928812971473164,
      "grad_norm": 1.8476109239446323,
      "learning_rate": 1e-05,
      "loss": 1.2934,
      "step": 1339
    },
    {
      "epoch": 0.9295066331396861,
      "grad_norm": 1.735630365745877,
      "learning_rate": 1e-05,
      "loss": 1.2978,
      "step": 1340
    },
    {
      "epoch": 0.9302002948062082,
      "grad_norm": 1.6182767450065856,
      "learning_rate": 1e-05,
      "loss": 1.2999,
      "step": 1341
    },
    {
      "epoch": 0.9308939564727304,
      "grad_norm": 1.7314432313872,
      "learning_rate": 1e-05,
      "loss": 1.291,
      "step": 1342
    },
    {
      "epoch": 0.9315876181392526,
      "grad_norm": 1.9298812546673758,
      "learning_rate": 1e-05,
      "loss": 1.3352,
      "step": 1343
    },
    {
      "epoch": 0.9322812798057747,
      "grad_norm": 1.890305436960625,
      "learning_rate": 1e-05,
      "loss": 1.3603,
      "step": 1344
    },
    {
      "epoch": 0.9329749414722969,
      "grad_norm": 1.7012847869156689,
      "learning_rate": 1e-05,
      "loss": 1.2705,
      "step": 1345
    },
    {
      "epoch": 0.933668603138819,
      "grad_norm": 1.7624548099246455,
      "learning_rate": 1e-05,
      "loss": 1.3229,
      "step": 1346
    },
    {
      "epoch": 0.9343622648053412,
      "grad_norm": 1.8174643785961295,
      "learning_rate": 1e-05,
      "loss": 1.2613,
      "step": 1347
    },
    {
      "epoch": 0.9350559264718633,
      "grad_norm": 1.766018762161775,
      "learning_rate": 1e-05,
      "loss": 1.3907,
      "step": 1348
    },
    {
      "epoch": 0.9357495881383855,
      "grad_norm": 1.8138067691928326,
      "learning_rate": 1e-05,
      "loss": 1.2878,
      "step": 1349
    },
    {
      "epoch": 0.9364432498049077,
      "grad_norm": 1.8378030129613956,
      "learning_rate": 1e-05,
      "loss": 1.3192,
      "step": 1350
    },
    {
      "epoch": 0.9371369114714299,
      "grad_norm": 2.012247327683597,
      "learning_rate": 1e-05,
      "loss": 1.3096,
      "step": 1351
    },
    {
      "epoch": 0.9378305731379519,
      "grad_norm": 1.704402821997915,
      "learning_rate": 1e-05,
      "loss": 1.3304,
      "step": 1352
    },
    {
      "epoch": 0.9385242348044741,
      "grad_norm": 1.7885693471410014,
      "learning_rate": 1e-05,
      "loss": 1.3065,
      "step": 1353
    },
    {
      "epoch": 0.9392178964709963,
      "grad_norm": 1.6661434949017349,
      "learning_rate": 1e-05,
      "loss": 1.3307,
      "step": 1354
    },
    {
      "epoch": 0.9399115581375185,
      "grad_norm": 1.7962358503874996,
      "learning_rate": 1e-05,
      "loss": 1.3129,
      "step": 1355
    },
    {
      "epoch": 0.9406052198040405,
      "grad_norm": 1.718189556083174,
      "learning_rate": 1e-05,
      "loss": 1.2994,
      "step": 1356
    },
    {
      "epoch": 0.9412988814705627,
      "grad_norm": 1.9092876605935403,
      "learning_rate": 1e-05,
      "loss": 1.3859,
      "step": 1357
    },
    {
      "epoch": 0.9419925431370849,
      "grad_norm": 1.8331634920945954,
      "learning_rate": 1e-05,
      "loss": 1.2668,
      "step": 1358
    },
    {
      "epoch": 0.942686204803607,
      "grad_norm": 1.7795600411027266,
      "learning_rate": 1e-05,
      "loss": 1.3158,
      "step": 1359
    },
    {
      "epoch": 0.9433798664701292,
      "grad_norm": 1.7718647937619816,
      "learning_rate": 1e-05,
      "loss": 1.2766,
      "step": 1360
    },
    {
      "epoch": 0.9440735281366514,
      "grad_norm": 1.7833385647199957,
      "learning_rate": 1e-05,
      "loss": 1.3188,
      "step": 1361
    },
    {
      "epoch": 0.9447671898031735,
      "grad_norm": 1.7094537156535095,
      "learning_rate": 1e-05,
      "loss": 1.2916,
      "step": 1362
    },
    {
      "epoch": 0.9454608514696956,
      "grad_norm": 1.7218575319930665,
      "learning_rate": 1e-05,
      "loss": 1.363,
      "step": 1363
    },
    {
      "epoch": 0.9461545131362178,
      "grad_norm": 1.8291690131158997,
      "learning_rate": 1e-05,
      "loss": 1.3067,
      "step": 1364
    },
    {
      "epoch": 0.94684817480274,
      "grad_norm": 1.5981127580277035,
      "learning_rate": 1e-05,
      "loss": 1.316,
      "step": 1365
    },
    {
      "epoch": 0.9475418364692622,
      "grad_norm": 1.7097133890389782,
      "learning_rate": 1e-05,
      "loss": 1.31,
      "step": 1366
    },
    {
      "epoch": 0.9482354981357842,
      "grad_norm": 1.7177869042959357,
      "learning_rate": 1e-05,
      "loss": 1.3261,
      "step": 1367
    },
    {
      "epoch": 0.9489291598023064,
      "grad_norm": 1.7391018887085676,
      "learning_rate": 1e-05,
      "loss": 1.3476,
      "step": 1368
    },
    {
      "epoch": 0.9496228214688286,
      "grad_norm": 1.835174110490517,
      "learning_rate": 1e-05,
      "loss": 1.3198,
      "step": 1369
    },
    {
      "epoch": 0.9503164831353508,
      "grad_norm": 1.8268669014129535,
      "learning_rate": 1e-05,
      "loss": 1.3338,
      "step": 1370
    },
    {
      "epoch": 0.9510101448018728,
      "grad_norm": 1.6581115729537674,
      "learning_rate": 1e-05,
      "loss": 1.2378,
      "step": 1371
    },
    {
      "epoch": 0.951703806468395,
      "grad_norm": 1.809305155351,
      "learning_rate": 1e-05,
      "loss": 1.3432,
      "step": 1372
    },
    {
      "epoch": 0.9523974681349172,
      "grad_norm": 1.7869170912103947,
      "learning_rate": 1e-05,
      "loss": 1.3173,
      "step": 1373
    },
    {
      "epoch": 0.9530911298014394,
      "grad_norm": 1.5946853342076313,
      "learning_rate": 1e-05,
      "loss": 1.3202,
      "step": 1374
    },
    {
      "epoch": 0.9537847914679615,
      "grad_norm": 1.7449681923450435,
      "learning_rate": 1e-05,
      "loss": 1.3368,
      "step": 1375
    },
    {
      "epoch": 0.9544784531344837,
      "grad_norm": 1.7867785161300096,
      "learning_rate": 1e-05,
      "loss": 1.2523,
      "step": 1376
    },
    {
      "epoch": 0.9551721148010058,
      "grad_norm": 1.7493793708109926,
      "learning_rate": 1e-05,
      "loss": 1.3143,
      "step": 1377
    },
    {
      "epoch": 0.955865776467528,
      "grad_norm": 1.6708212386781847,
      "learning_rate": 1e-05,
      "loss": 1.3197,
      "step": 1378
    },
    {
      "epoch": 0.9565594381340501,
      "grad_norm": 1.7624923123146528,
      "learning_rate": 1e-05,
      "loss": 1.2976,
      "step": 1379
    },
    {
      "epoch": 0.9572530998005723,
      "grad_norm": 1.8444740226606935,
      "learning_rate": 1e-05,
      "loss": 1.3251,
      "step": 1380
    },
    {
      "epoch": 0.9579467614670945,
      "grad_norm": 1.5953334218837774,
      "learning_rate": 1e-05,
      "loss": 1.2664,
      "step": 1381
    },
    {
      "epoch": 0.9586404231336165,
      "grad_norm": 1.9455556499425986,
      "learning_rate": 1e-05,
      "loss": 1.2883,
      "step": 1382
    },
    {
      "epoch": 0.9593340848001387,
      "grad_norm": 1.7447825684583413,
      "learning_rate": 1e-05,
      "loss": 1.3101,
      "step": 1383
    },
    {
      "epoch": 0.9600277464666609,
      "grad_norm": 1.7815515049059591,
      "learning_rate": 1e-05,
      "loss": 1.2931,
      "step": 1384
    },
    {
      "epoch": 0.9607214081331831,
      "grad_norm": 1.8473821407663016,
      "learning_rate": 1e-05,
      "loss": 1.2916,
      "step": 1385
    },
    {
      "epoch": 0.9614150697997051,
      "grad_norm": 1.8287625516543624,
      "learning_rate": 1e-05,
      "loss": 1.3302,
      "step": 1386
    },
    {
      "epoch": 0.9621087314662273,
      "grad_norm": 1.9131178436318876,
      "learning_rate": 1e-05,
      "loss": 1.292,
      "step": 1387
    },
    {
      "epoch": 0.9628023931327495,
      "grad_norm": 1.6731211620160438,
      "learning_rate": 1e-05,
      "loss": 1.2662,
      "step": 1388
    },
    {
      "epoch": 0.9634960547992717,
      "grad_norm": 1.8100000986534872,
      "learning_rate": 1e-05,
      "loss": 1.3462,
      "step": 1389
    },
    {
      "epoch": 0.9641897164657938,
      "grad_norm": 1.915036526264764,
      "learning_rate": 1e-05,
      "loss": 1.2819,
      "step": 1390
    },
    {
      "epoch": 0.964883378132316,
      "grad_norm": 1.9069098292199362,
      "learning_rate": 1e-05,
      "loss": 1.3002,
      "step": 1391
    },
    {
      "epoch": 0.9655770397988381,
      "grad_norm": 1.7930073332304715,
      "learning_rate": 1e-05,
      "loss": 1.2816,
      "step": 1392
    },
    {
      "epoch": 0.9662707014653603,
      "grad_norm": 1.8530639734903125,
      "learning_rate": 1e-05,
      "loss": 1.3521,
      "step": 1393
    },
    {
      "epoch": 0.9669643631318824,
      "grad_norm": 1.7315815210793186,
      "learning_rate": 1e-05,
      "loss": 1.299,
      "step": 1394
    },
    {
      "epoch": 0.9676580247984046,
      "grad_norm": 1.8034499383080111,
      "learning_rate": 1e-05,
      "loss": 1.2712,
      "step": 1395
    },
    {
      "epoch": 0.9683516864649268,
      "grad_norm": 1.8019607332463998,
      "learning_rate": 1e-05,
      "loss": 1.3431,
      "step": 1396
    },
    {
      "epoch": 0.9690453481314489,
      "grad_norm": 1.7026349253270283,
      "learning_rate": 1e-05,
      "loss": 1.3306,
      "step": 1397
    },
    {
      "epoch": 0.969739009797971,
      "grad_norm": 1.8029112479233105,
      "learning_rate": 1e-05,
      "loss": 1.2791,
      "step": 1398
    },
    {
      "epoch": 0.9704326714644932,
      "grad_norm": 1.6640138276094931,
      "learning_rate": 1e-05,
      "loss": 1.3551,
      "step": 1399
    },
    {
      "epoch": 0.9711263331310154,
      "grad_norm": 1.8773693641133087,
      "learning_rate": 1e-05,
      "loss": 1.3214,
      "step": 1400
    },
    {
      "epoch": 0.9718199947975376,
      "grad_norm": 1.6189625859175014,
      "learning_rate": 1e-05,
      "loss": 1.3045,
      "step": 1401
    },
    {
      "epoch": 0.9725136564640596,
      "grad_norm": 1.8655984733394448,
      "learning_rate": 1e-05,
      "loss": 1.295,
      "step": 1402
    },
    {
      "epoch": 0.9732073181305818,
      "grad_norm": 1.6495459791796045,
      "learning_rate": 1e-05,
      "loss": 1.3209,
      "step": 1403
    },
    {
      "epoch": 0.973900979797104,
      "grad_norm": 1.7262446380853802,
      "learning_rate": 1e-05,
      "loss": 1.3179,
      "step": 1404
    },
    {
      "epoch": 0.9745946414636261,
      "grad_norm": 1.8782015233926619,
      "learning_rate": 1e-05,
      "loss": 1.2991,
      "step": 1405
    },
    {
      "epoch": 0.9752883031301482,
      "grad_norm": 1.6294972087965263,
      "learning_rate": 1e-05,
      "loss": 1.3022,
      "step": 1406
    },
    {
      "epoch": 0.9759819647966704,
      "grad_norm": 1.79967261492696,
      "learning_rate": 1e-05,
      "loss": 1.3019,
      "step": 1407
    },
    {
      "epoch": 0.9766756264631926,
      "grad_norm": 1.8301257668132722,
      "learning_rate": 1e-05,
      "loss": 1.2952,
      "step": 1408
    },
    {
      "epoch": 0.9773692881297147,
      "grad_norm": 1.7858886287008595,
      "learning_rate": 1e-05,
      "loss": 1.3455,
      "step": 1409
    },
    {
      "epoch": 0.9780629497962369,
      "grad_norm": 1.645627337737642,
      "learning_rate": 1e-05,
      "loss": 1.2788,
      "step": 1410
    },
    {
      "epoch": 0.978756611462759,
      "grad_norm": 1.8493218045117754,
      "learning_rate": 1e-05,
      "loss": 1.2946,
      "step": 1411
    },
    {
      "epoch": 0.9794502731292812,
      "grad_norm": 1.9503637295771707,
      "learning_rate": 1e-05,
      "loss": 1.2623,
      "step": 1412
    },
    {
      "epoch": 0.9801439347958033,
      "grad_norm": 1.854587002167059,
      "learning_rate": 1e-05,
      "loss": 1.2966,
      "step": 1413
    },
    {
      "epoch": 0.9808375964623255,
      "grad_norm": 1.7795692734633988,
      "learning_rate": 1e-05,
      "loss": 1.3538,
      "step": 1414
    },
    {
      "epoch": 0.9815312581288477,
      "grad_norm": 1.8026744879009458,
      "learning_rate": 1e-05,
      "loss": 1.3534,
      "step": 1415
    },
    {
      "epoch": 0.9822249197953699,
      "grad_norm": 1.7902484770921805,
      "learning_rate": 1e-05,
      "loss": 1.3155,
      "step": 1416
    },
    {
      "epoch": 0.9829185814618919,
      "grad_norm": 1.7212800748245163,
      "learning_rate": 1e-05,
      "loss": 1.2941,
      "step": 1417
    },
    {
      "epoch": 0.9836122431284141,
      "grad_norm": 1.801323328800465,
      "learning_rate": 1e-05,
      "loss": 1.3823,
      "step": 1418
    },
    {
      "epoch": 0.9843059047949363,
      "grad_norm": 1.7021881473197502,
      "learning_rate": 1e-05,
      "loss": 1.3371,
      "step": 1419
    },
    {
      "epoch": 0.9849995664614585,
      "grad_norm": 1.8046118812459044,
      "learning_rate": 1e-05,
      "loss": 1.3433,
      "step": 1420
    },
    {
      "epoch": 0.9856932281279805,
      "grad_norm": 1.6685271100241381,
      "learning_rate": 1e-05,
      "loss": 1.2897,
      "step": 1421
    },
    {
      "epoch": 0.9863868897945027,
      "grad_norm": 1.6454889789915157,
      "learning_rate": 1e-05,
      "loss": 1.3029,
      "step": 1422
    },
    {
      "epoch": 0.9870805514610249,
      "grad_norm": 1.8167751934930396,
      "learning_rate": 1e-05,
      "loss": 1.278,
      "step": 1423
    },
    {
      "epoch": 0.987774213127547,
      "grad_norm": 1.7280323726187787,
      "learning_rate": 1e-05,
      "loss": 1.3137,
      "step": 1424
    },
    {
      "epoch": 0.9884678747940692,
      "grad_norm": 1.670222942611059,
      "learning_rate": 1e-05,
      "loss": 1.3248,
      "step": 1425
    },
    {
      "epoch": 0.9891615364605914,
      "grad_norm": 1.799342131947682,
      "learning_rate": 1e-05,
      "loss": 1.3215,
      "step": 1426
    },
    {
      "epoch": 0.9898551981271135,
      "grad_norm": 1.8123264473720575,
      "learning_rate": 1e-05,
      "loss": 1.3548,
      "step": 1427
    },
    {
      "epoch": 0.9905488597936356,
      "grad_norm": 1.7747475128663022,
      "learning_rate": 1e-05,
      "loss": 1.2765,
      "step": 1428
    },
    {
      "epoch": 0.9912425214601578,
      "grad_norm": 1.8041680654218195,
      "learning_rate": 1e-05,
      "loss": 1.31,
      "step": 1429
    },
    {
      "epoch": 0.99193618312668,
      "grad_norm": 1.7105917151107914,
      "learning_rate": 1e-05,
      "loss": 1.2958,
      "step": 1430
    },
    {
      "epoch": 0.9926298447932022,
      "grad_norm": 1.6626340057182631,
      "learning_rate": 1e-05,
      "loss": 1.3484,
      "step": 1431
    },
    {
      "epoch": 0.9933235064597242,
      "grad_norm": 1.6610746706958375,
      "learning_rate": 1e-05,
      "loss": 1.3256,
      "step": 1432
    },
    {
      "epoch": 0.9940171681262464,
      "grad_norm": 1.7615692816320323,
      "learning_rate": 1e-05,
      "loss": 1.3303,
      "step": 1433
    },
    {
      "epoch": 0.9947108297927686,
      "grad_norm": 1.7533916584851055,
      "learning_rate": 1e-05,
      "loss": 1.2955,
      "step": 1434
    },
    {
      "epoch": 0.9954044914592908,
      "grad_norm": 1.7232538432339657,
      "learning_rate": 1e-05,
      "loss": 1.3177,
      "step": 1435
    },
    {
      "epoch": 0.9960981531258128,
      "grad_norm": 1.7441612385283174,
      "learning_rate": 1e-05,
      "loss": 1.3013,
      "step": 1436
    },
    {
      "epoch": 0.996791814792335,
      "grad_norm": 1.8597160448645227,
      "learning_rate": 1e-05,
      "loss": 1.2923,
      "step": 1437
    },
    {
      "epoch": 0.9974854764588572,
      "grad_norm": 1.742932722373845,
      "learning_rate": 1e-05,
      "loss": 1.3469,
      "step": 1438
    },
    {
      "epoch": 0.9981791381253794,
      "grad_norm": 1.8961285251801105,
      "learning_rate": 1e-05,
      "loss": 1.3101,
      "step": 1439
    },
    {
      "epoch": 0.9988727997919015,
      "grad_norm": 1.6967076327288442,
      "learning_rate": 1e-05,
      "loss": 1.3157,
      "step": 1440
    },
    {
      "epoch": 0.9995664614584236,
      "grad_norm": 1.669534255209671,
      "learning_rate": 1e-05,
      "loss": 1.3203,
      "step": 1441
    },
    {
      "epoch": 0.9995664614584236,
      "step": 1441,
      "total_flos": 2332396250726400.0,
      "train_loss": 1.4015557253881927,
      "train_runtime": 204310.3582,
      "train_samples_per_second": 0.903,
      "train_steps_per_second": 0.007
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 1441,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 10,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2332396250726400.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}