{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.463215258855586,
  "eval_steps": 500,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.010899182561307902,
      "grad_norm": 6.173911094665527,
      "learning_rate": 4e-05,
      "loss": 14.7136,
      "step": 1
    },
    {
      "epoch": 0.021798365122615803,
      "grad_norm": 6.3364973068237305,
      "learning_rate": 8e-05,
      "loss": 14.8041,
      "step": 2
    },
    {
      "epoch": 0.0326975476839237,
      "grad_norm": 6.231718063354492,
      "learning_rate": 0.00012,
      "loss": 14.7035,
      "step": 3
    },
    {
      "epoch": 0.043596730245231606,
      "grad_norm": 6.516734600067139,
      "learning_rate": 0.00016,
      "loss": 14.6141,
      "step": 4
    },
    {
      "epoch": 0.05449591280653951,
      "grad_norm": NaN,
      "learning_rate": 0.00016,
      "loss": 14.4407,
      "step": 5
    },
    {
      "epoch": 0.0653950953678474,
      "grad_norm": 7.817420482635498,
      "learning_rate": 0.0002,
      "loss": 14.5375,
      "step": 6
    },
    {
      "epoch": 0.07629427792915532,
      "grad_norm": 9.571914672851562,
      "learning_rate": 0.0001995959595959596,
      "loss": 14.3553,
      "step": 7
    },
    {
      "epoch": 0.08719346049046321,
      "grad_norm": 11.314117431640625,
      "learning_rate": 0.0001991919191919192,
      "loss": 14.1913,
      "step": 8
    },
    {
      "epoch": 0.09809264305177112,
      "grad_norm": 12.428942680358887,
      "learning_rate": 0.00019878787878787878,
      "loss": 14.0594,
      "step": 9
    },
    {
      "epoch": 0.10899182561307902,
      "grad_norm": 13.171401977539062,
      "learning_rate": 0.00019838383838383837,
      "loss": 13.8783,
      "step": 10
    },
    {
      "epoch": 0.11989100817438691,
      "grad_norm": 14.13656997680664,
      "learning_rate": 0.000197979797979798,
      "loss": 13.7696,
      "step": 11
    },
    {
      "epoch": 0.1307901907356948,
      "grad_norm": 14.418411254882812,
      "learning_rate": 0.0001975757575757576,
      "loss": 13.6174,
      "step": 12
    },
    {
      "epoch": 0.14168937329700274,
      "grad_norm": 14.280936241149902,
      "learning_rate": 0.0001971717171717172,
      "loss": 13.4457,
      "step": 13
    },
    {
      "epoch": 0.15258855585831063,
      "grad_norm": 15.03721809387207,
      "learning_rate": 0.00019676767676767677,
      "loss": 13.3558,
      "step": 14
    },
    {
      "epoch": 0.16348773841961853,
      "grad_norm": 15.251821517944336,
      "learning_rate": 0.00019636363636363636,
      "loss": 13.2064,
      "step": 15
    },
    {
      "epoch": 0.17438692098092642,
      "grad_norm": 15.709274291992188,
      "learning_rate": 0.00019595959595959596,
      "loss": 13.0549,
      "step": 16
    },
    {
      "epoch": 0.18528610354223432,
      "grad_norm": 15.98558521270752,
      "learning_rate": 0.00019555555555555556,
      "loss": 12.8944,
      "step": 17
    },
    {
      "epoch": 0.19618528610354224,
      "grad_norm": 16.447908401489258,
      "learning_rate": 0.00019515151515151516,
      "loss": 12.8388,
      "step": 18
    },
    {
      "epoch": 0.20708446866485014,
      "grad_norm": 17.313697814941406,
      "learning_rate": 0.00019474747474747476,
      "loss": 12.6693,
      "step": 19
    },
    {
      "epoch": 0.21798365122615804,
      "grad_norm": 18.5076961517334,
      "learning_rate": 0.00019434343434343435,
      "loss": 12.5437,
      "step": 20
    },
    {
      "epoch": 0.22888283378746593,
      "grad_norm": 18.852705001831055,
      "learning_rate": 0.00019393939393939395,
      "loss": 12.3889,
      "step": 21
    },
    {
      "epoch": 0.23978201634877383,
      "grad_norm": 16.87480926513672,
      "learning_rate": 0.00019353535353535355,
      "loss": 12.2707,
      "step": 22
    },
    {
      "epoch": 0.2506811989100817,
      "grad_norm": 23.646879196166992,
      "learning_rate": 0.00019313131313131315,
      "loss": 11.918,
      "step": 23
    },
    {
      "epoch": 0.2615803814713896,
      "grad_norm": 18.714590072631836,
      "learning_rate": 0.00019272727272727274,
      "loss": 11.6272,
      "step": 24
    },
    {
      "epoch": 0.2724795640326976,
      "grad_norm": 35.362369537353516,
      "learning_rate": 0.00019232323232323232,
      "loss": 11.5307,
      "step": 25
    },
    {
      "epoch": 0.28337874659400547,
      "grad_norm": 22.232742309570312,
      "learning_rate": 0.00019191919191919191,
      "loss": 11.0798,
      "step": 26
    },
    {
      "epoch": 0.29427792915531337,
      "grad_norm": 22.175138473510742,
      "learning_rate": 0.0001915151515151515,
      "loss": 10.6351,
      "step": 27
    },
    {
      "epoch": 0.30517711171662126,
      "grad_norm": 26.256942749023438,
      "learning_rate": 0.00019111111111111114,
      "loss": 10.4008,
      "step": 28
    },
    {
      "epoch": 0.31607629427792916,
      "grad_norm": 24.899234771728516,
      "learning_rate": 0.00019070707070707073,
      "loss": 10.0295,
      "step": 29
    },
    {
      "epoch": 0.32697547683923706,
      "grad_norm": 15.740798950195312,
      "learning_rate": 0.0001903030303030303,
      "loss": 9.7172,
      "step": 30
    },
    {
      "epoch": 0.33787465940054495,
      "grad_norm": 27.80441665649414,
      "learning_rate": 0.0001898989898989899,
      "loss": 9.0562,
      "step": 31
    },
    {
      "epoch": 0.34877384196185285,
      "grad_norm": 45.844871520996094,
      "learning_rate": 0.0001894949494949495,
      "loss": 8.8467,
      "step": 32
    },
    {
      "epoch": 0.35967302452316074,
      "grad_norm": 6.738473892211914,
      "learning_rate": 0.0001890909090909091,
      "loss": 7.9078,
      "step": 33
    },
    {
      "epoch": 0.37057220708446864,
      "grad_norm": 3.9533674716949463,
      "learning_rate": 0.0001886868686868687,
      "loss": 7.4944,
      "step": 34
    },
    {
      "epoch": 0.3814713896457766,
      "grad_norm": 8.950474739074707,
      "learning_rate": 0.0001882828282828283,
      "loss": 7.1246,
      "step": 35
    },
    {
      "epoch": 0.3923705722070845,
      "grad_norm": 4.175976276397705,
      "learning_rate": 0.0001878787878787879,
      "loss": 6.8332,
      "step": 36
    },
    {
      "epoch": 0.4032697547683924,
      "grad_norm": 1.5284682512283325,
      "learning_rate": 0.0001874747474747475,
      "loss": 6.7993,
      "step": 37
    },
    {
      "epoch": 0.4141689373297003,
      "grad_norm": 1.4985376596450806,
      "learning_rate": 0.0001870707070707071,
      "loss": 6.7921,
      "step": 38
    },
    {
      "epoch": 0.4250681198910082,
      "grad_norm": 1.0437488555908203,
      "learning_rate": 0.0001866666666666667,
      "loss": 6.5349,
      "step": 39
    },
    {
      "epoch": 0.4359673024523161,
      "grad_norm": 1.4849172830581665,
      "learning_rate": 0.00018626262626262628,
      "loss": 6.7318,
      "step": 40
    },
    {
      "epoch": 0.44686648501362397,
      "grad_norm": 6.274806022644043,
      "learning_rate": 0.00018585858585858586,
      "loss": 6.6821,
      "step": 41
    },
    {
      "epoch": 0.45776566757493187,
      "grad_norm": 6.835639953613281,
      "learning_rate": 0.00018545454545454545,
      "loss": 6.4709,
      "step": 42
    },
    {
      "epoch": 0.46866485013623976,
      "grad_norm": 2.275062322616577,
      "learning_rate": 0.00018505050505050505,
      "loss": 6.4287,
      "step": 43
    },
    {
      "epoch": 0.47956403269754766,
      "grad_norm": 1.1775331497192383,
      "learning_rate": 0.00018464646464646465,
      "loss": 6.4365,
      "step": 44
    },
    {
      "epoch": 0.4904632152588556,
      "grad_norm": 1.0454274415969849,
      "learning_rate": 0.00018424242424242427,
      "loss": 6.4323,
      "step": 45
    },
    {
      "epoch": 0.5013623978201635,
      "grad_norm": 2.2385048866271973,
      "learning_rate": 0.00018383838383838384,
      "loss": 6.6629,
      "step": 46
    },
    {
      "epoch": 0.5122615803814714,
      "grad_norm": 0.8004893660545349,
      "learning_rate": 0.00018343434343434344,
      "loss": 6.6393,
      "step": 47
    },
    {
      "epoch": 0.5231607629427792,
      "grad_norm": 0.7514417767524719,
      "learning_rate": 0.00018303030303030304,
      "loss": 6.3857,
      "step": 48
    },
    {
      "epoch": 0.5340599455040872,
      "grad_norm": 1.1417890787124634,
      "learning_rate": 0.00018262626262626264,
      "loss": 6.5261,
      "step": 49
    },
    {
      "epoch": 0.5449591280653951,
      "grad_norm": 2.334674119949341,
      "learning_rate": 0.00018222222222222224,
      "loss": 6.45,
      "step": 50
    },
    {
      "epoch": 0.555858310626703,
      "grad_norm": 1.50180983543396,
      "learning_rate": 0.00018181818181818183,
      "loss": 6.3055,
      "step": 51
    },
    {
      "epoch": 0.5667574931880109,
      "grad_norm": 1.9359692335128784,
      "learning_rate": 0.0001814141414141414,
      "loss": 6.2799,
      "step": 52
    },
    {
      "epoch": 0.5776566757493188,
      "grad_norm": 2.2008886337280273,
      "learning_rate": 0.00018101010101010103,
      "loss": 6.2713,
      "step": 53
    },
    {
      "epoch": 0.5885558583106267,
      "grad_norm": 3.091736316680908,
      "learning_rate": 0.00018060606060606063,
      "loss": 6.5858,
      "step": 54
    },
    {
      "epoch": 0.5994550408719346,
      "grad_norm": 2.224456310272217,
      "learning_rate": 0.00018020202020202023,
      "loss": 6.2544,
      "step": 55
    },
    {
      "epoch": 0.6103542234332425,
      "grad_norm": 2.97121524810791,
      "learning_rate": 0.0001797979797979798,
      "loss": 6.2433,
      "step": 56
    },
    {
      "epoch": 0.6212534059945504,
      "grad_norm": 1.7177040576934814,
      "learning_rate": 0.0001793939393939394,
      "loss": 6.2307,
      "step": 57
    },
    {
      "epoch": 0.6321525885558583,
      "grad_norm": 0.4635346233844757,
      "learning_rate": 0.000178989898989899,
      "loss": 6.2002,
      "step": 58
    },
    {
      "epoch": 0.6430517711171662,
      "grad_norm": 0.9738274216651917,
      "learning_rate": 0.0001785858585858586,
      "loss": 6.2621,
      "step": 59
    },
    {
      "epoch": 0.6539509536784741,
      "grad_norm": 0.9515509605407715,
      "learning_rate": 0.0001781818181818182,
      "loss": 6.2735,
      "step": 60
    },
    {
      "epoch": 0.6648501362397821,
      "grad_norm": 0.38928309082984924,
      "learning_rate": 0.00017777777777777779,
      "loss": 6.2145,
      "step": 61
    },
    {
      "epoch": 0.6757493188010899,
      "grad_norm": 0.4064483046531677,
      "learning_rate": 0.00017737373737373738,
      "loss": 6.2032,
      "step": 62
    },
    {
      "epoch": 0.6866485013623979,
      "grad_norm": 0.5218994617462158,
      "learning_rate": 0.00017696969696969698,
      "loss": 6.2611,
      "step": 63
    },
    {
      "epoch": 0.6975476839237057,
      "grad_norm": 0.5422987341880798,
      "learning_rate": 0.00017656565656565658,
      "loss": 6.2076,
      "step": 64
    },
    {
      "epoch": 0.7084468664850136,
      "grad_norm": 0.38721132278442383,
      "learning_rate": 0.00017616161616161618,
      "loss": 6.2255,
      "step": 65
    },
    {
      "epoch": 0.7193460490463215,
      "grad_norm": 1.0110316276550293,
      "learning_rate": 0.00017575757575757578,
      "loss": 6.2988,
      "step": 66
    },
    {
      "epoch": 0.7302452316076294,
      "grad_norm": 0.33188217878341675,
      "learning_rate": 0.00017535353535353535,
      "loss": 6.1976,
      "step": 67
    },
    {
      "epoch": 0.7411444141689373,
      "grad_norm": 0.34784093499183655,
      "learning_rate": 0.00017494949494949494,
      "loss": 6.3306,
      "step": 68
    },
    {
      "epoch": 0.7520435967302452,
      "grad_norm": 0.1645226925611496,
      "learning_rate": 0.00017454545454545454,
      "loss": 6.1857,
      "step": 69
    },
    {
      "epoch": 0.7629427792915532,
      "grad_norm": 0.19698309898376465,
      "learning_rate": 0.00017414141414141414,
      "loss": 6.1652,
      "step": 70
    },
    {
      "epoch": 0.773841961852861,
      "grad_norm": 0.1376776397228241,
      "learning_rate": 0.00017373737373737377,
      "loss": 6.1559,
      "step": 71
    },
    {
      "epoch": 0.784741144414169,
      "grad_norm": 0.18504256010055542,
      "learning_rate": 0.00017333333333333334,
      "loss": 6.1365,
      "step": 72
    },
    {
      "epoch": 0.7956403269754768,
      "grad_norm": 0.21234065294265747,
      "learning_rate": 0.00017292929292929293,
      "loss": 6.1698,
      "step": 73
    },
    {
      "epoch": 0.8065395095367848,
      "grad_norm": 0.20268601179122925,
      "learning_rate": 0.00017252525252525253,
      "loss": 6.1746,
      "step": 74
    },
    {
      "epoch": 0.8174386920980926,
      "grad_norm": 0.23631519079208374,
      "learning_rate": 0.00017212121212121213,
      "loss": 6.1354,
      "step": 75
    },
    {
      "epoch": 0.8283378746594006,
      "grad_norm": 0.24041754007339478,
      "learning_rate": 0.00017171717171717173,
      "loss": 6.1546,
      "step": 76
    },
    {
      "epoch": 0.8392370572207084,
      "grad_norm": 0.1868537813425064,
      "learning_rate": 0.00017131313131313133,
      "loss": 6.151,
      "step": 77
    },
    {
      "epoch": 0.8501362397820164,
      "grad_norm": 0.29994234442710876,
      "learning_rate": 0.0001709090909090909,
      "loss": 6.1612,
      "step": 78
    },
    {
      "epoch": 0.8610354223433242,
      "grad_norm": 0.5775282979011536,
      "learning_rate": 0.00017050505050505052,
      "loss": 6.3783,
      "step": 79
    },
    {
      "epoch": 0.8719346049046321,
      "grad_norm": 1.5403401851654053,
      "learning_rate": 0.00017010101010101012,
      "loss": 6.2874,
      "step": 80
    },
    {
      "epoch": 0.8828337874659401,
      "grad_norm": 0.389077752828598,
      "learning_rate": 0.00016969696969696972,
      "loss": 6.179,
      "step": 81
    },
    {
      "epoch": 0.8937329700272479,
      "grad_norm": 0.38043642044067383,
      "learning_rate": 0.00016929292929292932,
      "loss": 6.2538,
      "step": 82
    },
    {
      "epoch": 0.9046321525885559,
      "grad_norm": 0.6061437726020813,
      "learning_rate": 0.00016888888888888889,
      "loss": 6.4058,
      "step": 83
    },
    {
      "epoch": 0.9155313351498637,
      "grad_norm": 0.24083693325519562,
      "learning_rate": 0.00016848484848484848,
      "loss": 6.2007,
      "step": 84
    },
    {
      "epoch": 0.9264305177111717,
      "grad_norm": 0.2530721426010132,
      "learning_rate": 0.00016808080808080808,
      "loss": 6.1851,
      "step": 85
    },
    {
      "epoch": 0.9373297002724795,
      "grad_norm": 0.3793930411338806,
      "learning_rate": 0.00016767676767676768,
      "loss": 6.1921,
      "step": 86
    },
    {
      "epoch": 0.9482288828337875,
      "grad_norm": 0.255757600069046,
      "learning_rate": 0.00016727272727272728,
      "loss": 6.2641,
      "step": 87
    },
    {
      "epoch": 0.9591280653950953,
      "grad_norm": 0.16223767399787903,
      "learning_rate": 0.00016686868686868688,
      "loss": 6.1847,
      "step": 88
    },
    {
      "epoch": 0.9700272479564033,
      "grad_norm": 0.14015568792819977,
      "learning_rate": 0.00016646464646464647,
      "loss": 6.1638,
      "step": 89
    },
    {
      "epoch": 0.9809264305177112,
      "grad_norm": 0.6141758561134338,
      "learning_rate": 0.00016606060606060607,
      "loss": 6.2167,
      "step": 90
    },
    {
      "epoch": 0.9918256130790191,
      "grad_norm": 0.4424732029438019,
      "learning_rate": 0.00016565656565656567,
      "loss": 6.3013,
      "step": 91
    },
    {
      "epoch": 1.005449591280654,
      "grad_norm": 0.405213862657547,
      "learning_rate": 0.00016525252525252527,
      "loss": 9.2423,
      "step": 92
    },
    {
      "epoch": 1.0163487738419619,
      "grad_norm": 0.8009450435638428,
      "learning_rate": 0.00016484848484848487,
      "loss": 6.1953,
      "step": 93
    },
    {
      "epoch": 1.0272479564032697,
      "grad_norm": 0.2201612889766693,
      "learning_rate": 0.00016444444444444444,
      "loss": 6.153,
      "step": 94
    },
    {
      "epoch": 1.0381471389645776,
      "grad_norm": 0.4194308817386627,
      "learning_rate": 0.00016404040404040403,
      "loss": 6.1968,
      "step": 95
    },
    {
      "epoch": 1.0490463215258856,
      "grad_norm": 0.6941304206848145,
      "learning_rate": 0.00016363636363636366,
      "loss": 6.9588,
      "step": 96
    },
    {
      "epoch": 1.0599455040871935,
      "grad_norm": 0.5405890941619873,
      "learning_rate": 0.00016323232323232326,
      "loss": 6.5007,
      "step": 97
    },
    {
      "epoch": 1.0708446866485013,
      "grad_norm": 0.6613783240318298,
      "learning_rate": 0.00016282828282828283,
      "loss": 6.1685,
      "step": 98
    },
    {
      "epoch": 1.0817438692098094,
      "grad_norm": 0.49628227949142456,
      "learning_rate": 0.00016242424242424243,
      "loss": 6.2215,
      "step": 99
    },
    {
      "epoch": 1.0926430517711172,
      "grad_norm": 0.2801539897918701,
      "learning_rate": 0.00016202020202020202,
      "loss": 6.1567,
      "step": 100
    },
    {
      "epoch": 1.103542234332425,
      "grad_norm": 0.2642604410648346,
      "learning_rate": 0.00016161616161616162,
      "loss": 6.1611,
      "step": 101
    },
    {
      "epoch": 1.1144414168937329,
      "grad_norm": 0.3278007209300995,
      "learning_rate": 0.00016121212121212122,
      "loss": 6.2144,
      "step": 102
    },
    {
      "epoch": 1.125340599455041,
      "grad_norm": 0.28854313492774963,
      "learning_rate": 0.00016080808080808082,
      "loss": 6.1837,
      "step": 103
    },
    {
      "epoch": 1.1362397820163488,
      "grad_norm": 0.1120542660355568,
      "learning_rate": 0.0001604040404040404,
      "loss": 6.1733,
      "step": 104
    },
    {
      "epoch": 1.1471389645776566,
      "grad_norm": 0.1502349078655243,
      "learning_rate": 0.00016,
      "loss": 6.1956,
      "step": 105
    },
    {
      "epoch": 1.1580381471389645,
      "grad_norm": 0.1573847085237503,
      "learning_rate": 0.0001595959595959596,
      "loss": 6.2238,
      "step": 106
    },
    {
      "epoch": 1.1689373297002725,
      "grad_norm": 0.24283528327941895,
      "learning_rate": 0.0001591919191919192,
      "loss": 6.1679,
      "step": 107
    },
    {
      "epoch": 1.1798365122615804,
      "grad_norm": 0.24594663083553314,
      "learning_rate": 0.0001587878787878788,
      "loss": 6.2147,
      "step": 108
    },
    {
      "epoch": 1.1907356948228882,
      "grad_norm": 0.3162429928779602,
      "learning_rate": 0.00015838383838383838,
      "loss": 6.239,
      "step": 109
    },
    {
      "epoch": 1.2016348773841963,
      "grad_norm": 0.24397960305213928,
      "learning_rate": 0.00015797979797979798,
      "loss": 6.4416,
      "step": 110
    },
    {
      "epoch": 1.2125340599455041,
      "grad_norm": 0.1755504310131073,
      "learning_rate": 0.00015757575757575757,
      "loss": 6.1848,
      "step": 111
    },
    {
      "epoch": 1.223433242506812,
      "grad_norm": 0.1865660697221756,
      "learning_rate": 0.00015717171717171717,
      "loss": 6.1807,
      "step": 112
    },
    {
      "epoch": 1.2343324250681198,
      "grad_norm": 0.5681672692298889,
      "learning_rate": 0.0001567676767676768,
      "loss": 7.1042,
      "step": 113
    },
    {
      "epoch": 1.2452316076294279,
      "grad_norm": 0.13271750509738922,
      "learning_rate": 0.00015636363636363637,
      "loss": 6.2829,
      "step": 114
    },
    {
      "epoch": 1.2561307901907357,
      "grad_norm": 0.21046936511993408,
      "learning_rate": 0.00015595959595959597,
      "loss": 6.1624,
      "step": 115
    },
    {
      "epoch": 1.2670299727520435,
      "grad_norm": 0.6139978170394897,
      "learning_rate": 0.00015555555555555556,
      "loss": 6.1552,
      "step": 116
    },
    {
      "epoch": 1.2779291553133514,
      "grad_norm": 0.47311800718307495,
      "learning_rate": 0.00015515151515151516,
      "loss": 6.1831,
      "step": 117
    },
    {
      "epoch": 1.2888283378746594,
      "grad_norm": 0.24481365084648132,
      "learning_rate": 0.00015474747474747476,
      "loss": 6.1404,
      "step": 118
    },
    {
      "epoch": 1.2997275204359673,
      "grad_norm": 0.2416616678237915,
      "learning_rate": 0.00015434343434343436,
      "loss": 6.1316,
      "step": 119
    },
    {
      "epoch": 1.3106267029972751,
      "grad_norm": 0.29993197321891785,
      "learning_rate": 0.00015393939393939393,
      "loss": 6.1391,
      "step": 120
    },
    {
      "epoch": 1.3215258855585832,
      "grad_norm": 0.19761894643306732,
      "learning_rate": 0.00015353535353535353,
      "loss": 6.178,
      "step": 121
    },
    {
      "epoch": 1.332425068119891,
      "grad_norm": 0.12814165651798248,
      "learning_rate": 0.00015313131313131315,
      "loss": 6.1696,
      "step": 122
    },
    {
      "epoch": 1.3433242506811989,
      "grad_norm": 0.7292453050613403,
      "learning_rate": 0.00015272727272727275,
      "loss": 6.206,
      "step": 123
    },
    {
      "epoch": 1.354223433242507,
      "grad_norm": 0.34973058104515076,
      "learning_rate": 0.00015232323232323235,
      "loss": 6.4659,
      "step": 124
    },
    {
      "epoch": 1.3651226158038148,
      "grad_norm": 0.22113573551177979,
      "learning_rate": 0.00015191919191919192,
      "loss": 6.2153,
      "step": 125
    },
    {
      "epoch": 1.3760217983651226,
      "grad_norm": 0.21217972040176392,
      "learning_rate": 0.00015151515151515152,
      "loss": 6.2934,
      "step": 126
    },
    {
      "epoch": 1.3869209809264305,
      "grad_norm": 0.20437318086624146,
      "learning_rate": 0.0001511111111111111,
      "loss": 6.2159,
      "step": 127
    },
    {
      "epoch": 1.3978201634877383,
      "grad_norm": 0.15652823448181152,
      "learning_rate": 0.0001507070707070707,
      "loss": 6.2017,
      "step": 128
    },
    {
      "epoch": 1.4087193460490464,
      "grad_norm": 0.13777923583984375,
      "learning_rate": 0.0001503030303030303,
      "loss": 6.2012,
      "step": 129
    },
    {
      "epoch": 1.4196185286103542,
      "grad_norm": 0.19372999668121338,
      "learning_rate": 0.0001498989898989899,
      "loss": 6.1927,
      "step": 130
    },
    {
      "epoch": 1.430517711171662,
      "grad_norm": 0.13652633130550385,
      "learning_rate": 0.0001494949494949495,
      "loss": 6.1932,
      "step": 131
    },
    {
      "epoch": 1.44141689373297,
      "grad_norm": 0.38320279121398926,
      "learning_rate": 0.0001490909090909091,
      "loss": 6.315,
      "step": 132
    },
    {
      "epoch": 1.452316076294278,
      "grad_norm": 0.1715274155139923,
      "learning_rate": 0.0001486868686868687,
      "loss": 6.2499,
      "step": 133
    },
    {
      "epoch": 1.4632152588555858,
      "grad_norm": 0.1598396748304367,
      "learning_rate": 0.0001482828282828283,
      "loss": 6.1881,
      "step": 134
    },
    {
      "epoch": 1.4741144414168939,
      "grad_norm": 0.15926583111286163,
      "learning_rate": 0.0001478787878787879,
      "loss": 6.2575,
      "step": 135
    },
    {
      "epoch": 1.4850136239782017,
      "grad_norm": 0.33166489005088806,
      "learning_rate": 0.00014747474747474747,
      "loss": 6.1834,
      "step": 136
    },
    {
      "epoch": 1.4959128065395095,
      "grad_norm": 0.3521062731742859,
      "learning_rate": 0.00014707070707070706,
      "loss": 6.1724,
      "step": 137
    },
    {
      "epoch": 1.5068119891008176,
      "grad_norm": 0.30115848779678345,
      "learning_rate": 0.00014666666666666666,
      "loss": 6.151,
      "step": 138
    },
    {
      "epoch": 1.5177111716621252,
      "grad_norm": 0.12073715031147003,
      "learning_rate": 0.0001462626262626263,
      "loss": 6.1674,
      "step": 139
    },
    {
      "epoch": 1.5286103542234333,
      "grad_norm": 0.47121262550354004,
      "learning_rate": 0.00014585858585858586,
      "loss": 6.3034,
      "step": 140
    },
    {
      "epoch": 1.5395095367847411,
      "grad_norm": 0.17591841518878937,
      "learning_rate": 0.00014545454545454546,
      "loss": 6.1417,
      "step": 141
    },
    {
      "epoch": 1.550408719346049,
      "grad_norm": 0.13197541236877441,
      "learning_rate": 0.00014505050505050505,
      "loss": 6.1599,
      "step": 142
    },
    {
      "epoch": 1.561307901907357,
      "grad_norm": 0.12980419397354126,
      "learning_rate": 0.00014464646464646465,
      "loss": 6.1491,
      "step": 143
    },
    {
      "epoch": 1.5722070844686649,
      "grad_norm": 0.2677476704120636,
      "learning_rate": 0.00014424242424242425,
      "loss": 6.1894,
      "step": 144
    },
    {
      "epoch": 1.5831062670299727,
      "grad_norm": 0.640482485294342,
      "learning_rate": 0.00014383838383838385,
      "loss": 6.5929,
      "step": 145
    },
    {
      "epoch": 1.5940054495912808,
      "grad_norm": 0.11611925810575485,
      "learning_rate": 0.00014343434343434342,
      "loss": 6.1625,
      "step": 146
    },
    {
      "epoch": 1.6049046321525886,
      "grad_norm": 0.31824928522109985,
      "learning_rate": 0.00014303030303030304,
      "loss": 6.1576,
      "step": 147
    },
    {
      "epoch": 1.6158038147138964,
      "grad_norm": 0.3396945893764496,
      "learning_rate": 0.00014262626262626264,
      "loss": 6.1749,
      "step": 148
    },
    {
      "epoch": 1.6267029972752045,
      "grad_norm": 0.36278408765792847,
      "learning_rate": 0.00014222222222222224,
      "loss": 6.2145,
      "step": 149
    },
    {
      "epoch": 1.6376021798365121,
      "grad_norm": 0.2582668364048004,
      "learning_rate": 0.00014181818181818184,
      "loss": 6.1878,
      "step": 150
    },
    {
      "epoch": 1.6485013623978202,
      "grad_norm": 0.12749724090099335,
      "learning_rate": 0.0001414141414141414,
      "loss": 6.1796,
      "step": 151
    },
    {
      "epoch": 1.659400544959128,
      "grad_norm": 0.27475911378860474,
      "learning_rate": 0.000141010101010101,
      "loss": 6.1946,
      "step": 152
    },
    {
      "epoch": 1.6702997275204359,
      "grad_norm": 0.14458522200584412,
      "learning_rate": 0.0001406060606060606,
      "loss": 6.3475,
      "step": 153
    },
    {
      "epoch": 1.681198910081744,
      "grad_norm": 0.20999635756015778,
      "learning_rate": 0.0001402020202020202,
      "loss": 6.19,
      "step": 154
    },
    {
      "epoch": 1.6920980926430518,
      "grad_norm": 0.16225522756576538,
      "learning_rate": 0.0001397979797979798,
      "loss": 6.3598,
      "step": 155
    },
    {
      "epoch": 1.7029972752043596,
      "grad_norm": 0.11324774473905563,
      "learning_rate": 0.0001393939393939394,
      "loss": 6.1729,
      "step": 156
    },
    {
      "epoch": 1.7138964577656677,
      "grad_norm": 0.16397441923618317,
      "learning_rate": 0.000138989898989899,
      "loss": 6.2364,
      "step": 157
    },
    {
      "epoch": 1.7247956403269755,
      "grad_norm": 0.31427308917045593,
      "learning_rate": 0.0001385858585858586,
      "loss": 6.6519,
      "step": 158
    },
    {
      "epoch": 1.7356948228882834,
      "grad_norm": 0.1903781294822693,
      "learning_rate": 0.0001381818181818182,
      "loss": 6.1424,
      "step": 159
    },
    {
      "epoch": 1.7465940054495914,
      "grad_norm": 0.1074923425912857,
      "learning_rate": 0.0001377777777777778,
      "loss": 6.162,
      "step": 160
    },
    {
      "epoch": 1.757493188010899,
      "grad_norm": 0.2343117892742157,
      "learning_rate": 0.0001373737373737374,
      "loss": 6.3699,
      "step": 161
    },
    {
      "epoch": 1.768392370572207,
      "grad_norm": 0.22847014665603638,
      "learning_rate": 0.00013696969696969696,
      "loss": 6.158,
      "step": 162
    },
    {
      "epoch": 1.779291553133515,
      "grad_norm": 0.17746838927268982,
      "learning_rate": 0.00013656565656565656,
      "loss": 6.1443,
      "step": 163
    },
    {
      "epoch": 1.7901907356948228,
      "grad_norm": 0.19379661977291107,
      "learning_rate": 0.00013616161616161618,
      "loss": 6.288,
      "step": 164
    },
    {
      "epoch": 1.8010899182561309,
      "grad_norm": 0.2187827080488205,
      "learning_rate": 0.00013575757575757578,
      "loss": 6.1312,
      "step": 165
    },
    {
      "epoch": 1.8119891008174387,
      "grad_norm": 0.2696266174316406,
      "learning_rate": 0.00013535353535353538,
      "loss": 6.1643,
      "step": 166
    },
    {
      "epoch": 1.8228882833787465,
      "grad_norm": 0.4230067729949951,
      "learning_rate": 0.00013494949494949495,
      "loss": 6.1366,
      "step": 167
    },
    {
      "epoch": 1.8337874659400546,
      "grad_norm": 0.2398137003183365,
      "learning_rate": 0.00013454545454545455,
      "loss": 6.1562,
      "step": 168
    },
    {
      "epoch": 1.8446866485013624,
      "grad_norm": 0.16632282733917236,
      "learning_rate": 0.00013414141414141414,
      "loss": 6.186,
      "step": 169
    },
    {
      "epoch": 1.8555858310626703,
      "grad_norm": 0.48356759548187256,
      "learning_rate": 0.00013373737373737374,
      "loss": 6.3463,
      "step": 170
    },
    {
      "epoch": 1.8664850136239783,
      "grad_norm": 0.16733959317207336,
      "learning_rate": 0.00013333333333333334,
      "loss": 6.177,
      "step": 171
    },
    {
      "epoch": 1.877384196185286,
      "grad_norm": 0.24873673915863037,
      "learning_rate": 0.00013292929292929294,
      "loss": 6.2373,
      "step": 172
    },
    {
      "epoch": 1.888283378746594,
      "grad_norm": 0.23225297033786774,
      "learning_rate": 0.00013252525252525254,
      "loss": 6.1933,
      "step": 173
    },
    {
      "epoch": 1.8991825613079019,
      "grad_norm": 0.473389595746994,
      "learning_rate": 0.00013212121212121213,
      "loss": 6.2052,
      "step": 174
    },
    {
      "epoch": 1.9100817438692097,
      "grad_norm": 0.3113359212875366,
      "learning_rate": 0.00013171717171717173,
      "loss": 6.3197,
      "step": 175
    },
    {
      "epoch": 1.9209809264305178,
      "grad_norm": 0.1677085906267166,
      "learning_rate": 0.00013131313131313133,
      "loss": 6.3585,
      "step": 176
    },
    {
      "epoch": 1.9318801089918256,
      "grad_norm": 0.1711427867412567,
      "learning_rate": 0.00013090909090909093,
      "loss": 6.2056,
      "step": 177
    },
    {
      "epoch": 1.9427792915531334,
      "grad_norm": 0.12511757016181946,
      "learning_rate": 0.0001305050505050505,
      "loss": 6.1898,
      "step": 178
    },
    {
      "epoch": 1.9536784741144415,
      "grad_norm": 0.15003643929958344,
      "learning_rate": 0.0001301010101010101,
      "loss": 6.18,
      "step": 179
    },
    {
      "epoch": 1.9645776566757494,
      "grad_norm": 0.18387942016124725,
      "learning_rate": 0.0001296969696969697,
      "loss": 6.1497,
      "step": 180
    },
    {
      "epoch": 1.9754768392370572,
      "grad_norm": 0.09031075984239578,
      "learning_rate": 0.00012929292929292932,
      "loss": 6.1753,
      "step": 181
    },
    {
      "epoch": 1.9863760217983653,
      "grad_norm": 0.10767246037721634,
      "learning_rate": 0.00012888888888888892,
      "loss": 6.1599,
      "step": 182
    },
    {
      "epoch": 1.9972752043596729,
      "grad_norm": 0.45141157507896423,
      "learning_rate": 0.0001284848484848485,
      "loss": 9.3651,
      "step": 183
    },
    {
      "epoch": 2.010899182561308,
      "grad_norm": 0.09891144186258316,
      "learning_rate": 0.00012808080808080809,
      "loss": 6.1438,
      "step": 184
    },
    {
      "epoch": 2.0217983651226157,
      "grad_norm": 0.13178829848766327,
      "learning_rate": 0.00012767676767676768,
      "loss": 6.1458,
      "step": 185
    },
    {
      "epoch": 2.0326975476839237,
      "grad_norm": 0.08824258297681808,
      "learning_rate": 0.00012727272727272728,
      "loss": 6.143,
      "step": 186
    },
    {
      "epoch": 2.043596730245232,
      "grad_norm": 0.13943664729595184,
      "learning_rate": 0.00012686868686868688,
      "loss": 6.1356,
      "step": 187
    },
    {
      "epoch": 2.0544959128065394,
      "grad_norm": 0.14288043975830078,
      "learning_rate": 0.00012646464646464645,
      "loss": 6.1573,
      "step": 188
    },
    {
      "epoch": 2.0653950953678475,
      "grad_norm": 0.2474461942911148,
      "learning_rate": 0.00012606060606060605,
      "loss": 6.2321,
      "step": 189
    },
    {
      "epoch": 2.076294277929155,
      "grad_norm": 0.13096056878566742,
      "learning_rate": 0.00012565656565656567,
      "loss": 6.2665,
      "step": 190
    },
    {
      "epoch": 2.087193460490463,
      "grad_norm": 0.21291935443878174,
      "learning_rate": 0.00012525252525252527,
      "loss": 6.1686,
      "step": 191
    },
    {
      "epoch": 2.0980926430517712,
      "grad_norm": 0.13552981615066528,
      "learning_rate": 0.00012484848484848487,
      "loss": 6.178,
      "step": 192
    },
    {
      "epoch": 2.108991825613079,
      "grad_norm": 0.09302832186222076,
      "learning_rate": 0.00012444444444444444,
      "loss": 6.1602,
      "step": 193
    },
    {
      "epoch": 2.119891008174387,
      "grad_norm": 0.5419220924377441,
      "learning_rate": 0.00012404040404040404,
      "loss": 6.2916,
      "step": 194
    },
    {
      "epoch": 2.130790190735695,
      "grad_norm": 0.12538427114486694,
      "learning_rate": 0.00012363636363636364,
      "loss": 6.1507,
      "step": 195
    },
    {
      "epoch": 2.1416893732970026,
      "grad_norm": 0.5663388967514038,
      "learning_rate": 0.00012323232323232323,
      "loss": 6.154,
      "step": 196
    },
    {
      "epoch": 2.1525885558583107,
      "grad_norm": 0.16043990850448608,
      "learning_rate": 0.00012282828282828283,
      "loss": 6.1679,
      "step": 197
    },
    {
      "epoch": 2.1634877384196187,
      "grad_norm": 0.140010267496109,
      "learning_rate": 0.00012242424242424243,
      "loss": 6.188,
      "step": 198
    },
    {
      "epoch": 2.1743869209809263,
      "grad_norm": 0.08459494262933731,
      "learning_rate": 0.00012202020202020204,
      "loss": 6.1681,
      "step": 199
    },
    {
      "epoch": 2.1852861035422344,
      "grad_norm": 0.40000060200691223,
      "learning_rate": 0.00012161616161616162,
      "loss": 6.8772,
      "step": 200
    },
    {
      "epoch": 2.1961852861035425,
      "grad_norm": 0.18648391962051392,
      "learning_rate": 0.00012121212121212122,
      "loss": 6.1619,
      "step": 201
    },
    {
      "epoch": 2.20708446866485,
      "grad_norm": 0.5561273097991943,
      "learning_rate": 0.00012080808080808082,
      "loss": 6.2139,
      "step": 202
    },
    {
      "epoch": 2.217983651226158,
      "grad_norm": 0.5014886856079102,
      "learning_rate": 0.0001204040404040404,
      "loss": 6.6783,
      "step": 203
    },
    {
      "epoch": 2.2288828337874658,
      "grad_norm": 0.2286769449710846,
      "learning_rate": 0.00012,
      "loss": 6.2043,
      "step": 204
    },
    {
      "epoch": 2.239782016348774,
      "grad_norm": 0.33960846066474915,
      "learning_rate": 0.0001195959595959596,
      "loss": 6.157,
      "step": 205
    },
    {
      "epoch": 2.250681198910082,
      "grad_norm": 0.21916227042675018,
      "learning_rate": 0.00011919191919191919,
      "loss": 6.5121,
      "step": 206
    },
    {
      "epoch": 2.2615803814713895,
      "grad_norm": 0.40727755427360535,
      "learning_rate": 0.0001187878787878788,
      "loss": 6.1969,
      "step": 207
    },
    {
      "epoch": 2.2724795640326976,
      "grad_norm": 0.20227251946926117,
      "learning_rate": 0.0001183838383838384,
      "loss": 6.3042,
      "step": 208
    },
    {
      "epoch": 2.2833787465940056,
      "grad_norm": 0.262510746717453,
      "learning_rate": 0.00011797979797979799,
      "loss": 6.3083,
      "step": 209
    },
    {
      "epoch": 2.2942779291553133,
      "grad_norm": 0.3285979926586151,
      "learning_rate": 0.00011757575757575758,
      "loss": 6.1488,
      "step": 210
    },
    {
      "epoch": 2.3051771117166213,
      "grad_norm": 0.5452210903167725,
      "learning_rate": 0.00011717171717171717,
      "loss": 6.5011,
      "step": 211
    },
    {
      "epoch": 2.316076294277929,
      "grad_norm": 0.14714978635311127,
      "learning_rate": 0.00011676767676767677,
      "loss": 6.1809,
      "step": 212
    },
    {
      "epoch": 2.326975476839237,
      "grad_norm": 0.31165456771850586,
      "learning_rate": 0.00011636363636363636,
      "loss": 6.2561,
      "step": 213
    },
    {
      "epoch": 2.337874659400545,
      "grad_norm": 0.18160106241703033,
      "learning_rate": 0.00011595959595959596,
      "loss": 6.1561,
      "step": 214
    },
    {
      "epoch": 2.3487738419618527,
      "grad_norm": 0.35501718521118164,
      "learning_rate": 0.00011555555555555555,
      "loss": 6.1351,
      "step": 215
    },
    {
      "epoch": 2.3596730245231607,
      "grad_norm": 0.18472979962825775,
      "learning_rate": 0.00011515151515151516,
      "loss": 6.1727,
      "step": 216
    },
    {
      "epoch": 2.370572207084469,
      "grad_norm": 0.17979513108730316,
      "learning_rate": 0.00011474747474747476,
      "loss": 6.1407,
      "step": 217
    },
    {
      "epoch": 2.3814713896457764,
      "grad_norm": 0.1735665202140808,
      "learning_rate": 0.00011434343434343435,
      "loss": 6.1773,
      "step": 218
    },
    {
      "epoch": 2.3923705722070845,
      "grad_norm": 0.1667218953371048,
      "learning_rate": 0.00011393939393939394,
      "loss": 6.1469,
      "step": 219
    },
    {
      "epoch": 2.4032697547683926,
      "grad_norm": 0.12015088647603989,
      "learning_rate": 0.00011353535353535354,
      "loss": 6.1646,
      "step": 220
    },
    {
      "epoch": 2.4141689373297,
      "grad_norm": 0.12716157734394073,
      "learning_rate": 0.00011313131313131313,
      "loss": 6.1674,
      "step": 221
    },
    {
      "epoch": 2.4250681198910082,
      "grad_norm": 0.1907472461462021,
      "learning_rate": 0.00011272727272727272,
      "loss": 6.1589,
      "step": 222
    },
    {
      "epoch": 2.4359673024523163,
      "grad_norm": 0.25412097573280334,
      "learning_rate": 0.00011232323232323232,
      "loss": 6.2231,
      "step": 223
    },
    {
      "epoch": 2.446866485013624,
      "grad_norm": 0.13590645790100098,
      "learning_rate": 0.00011191919191919193,
      "loss": 6.1549,
      "step": 224
    },
    {
      "epoch": 2.457765667574932,
      "grad_norm": 0.16243083775043488,
      "learning_rate": 0.00011151515151515153,
      "loss": 6.1532,
      "step": 225
    },
    {
      "epoch": 2.4686648501362396,
      "grad_norm": 0.3192502558231354,
      "learning_rate": 0.00011111111111111112,
      "loss": 6.1976,
      "step": 226
    },
    {
      "epoch": 2.4795640326975477,
      "grad_norm": 0.11255048960447311,
      "learning_rate": 0.00011070707070707071,
      "loss": 6.1756,
      "step": 227
    },
    {
      "epoch": 2.4904632152588557,
      "grad_norm": 0.08019052445888519,
      "learning_rate": 0.00011030303030303031,
      "loss": 6.1651,
      "step": 228
    },
    {
      "epoch": 2.5013623978201633,
      "grad_norm": 0.11216670274734497,
      "learning_rate": 0.0001098989898989899,
      "loss": 6.174,
      "step": 229
    },
    {
      "epoch": 2.5122615803814714,
      "grad_norm": 0.180108904838562,
      "learning_rate": 0.0001094949494949495,
      "loss": 6.1628,
      "step": 230
    },
    {
      "epoch": 2.523160762942779,
      "grad_norm": 0.16970708966255188,
      "learning_rate": 0.00010909090909090909,
      "loss": 6.1851,
      "step": 231
    },
    {
      "epoch": 2.534059945504087,
      "grad_norm": 0.12635812163352966,
      "learning_rate": 0.00010868686868686868,
      "loss": 6.1696,
      "step": 232
    },
    {
      "epoch": 2.544959128065395,
      "grad_norm": 0.5308592915534973,
      "learning_rate": 0.0001082828282828283,
      "loss": 7.1268,
      "step": 233
    },
    {
      "epoch": 2.5558583106267028,
      "grad_norm": 0.16270631551742554,
      "learning_rate": 0.00010787878787878789,
      "loss": 6.1787,
      "step": 234
    },
    {
      "epoch": 2.566757493188011,
      "grad_norm": 0.8060057759284973,
      "learning_rate": 0.00010747474747474748,
      "loss": 6.4044,
      "step": 235
    },
    {
      "epoch": 2.577656675749319,
      "grad_norm": 0.44751596450805664,
      "learning_rate": 0.00010707070707070708,
      "loss": 6.1638,
      "step": 236
    },
    {
      "epoch": 2.5885558583106265,
      "grad_norm": 0.299060583114624,
      "learning_rate": 0.00010666666666666667,
      "loss": 6.1527,
      "step": 237
    },
    {
      "epoch": 2.5994550408719346,
      "grad_norm": 0.29560768604278564,
      "learning_rate": 0.00010626262626262626,
      "loss": 6.1387,
      "step": 238
    },
    {
      "epoch": 2.6103542234332426,
      "grad_norm": 0.31047800183296204,
      "learning_rate": 0.00010585858585858586,
      "loss": 6.1753,
      "step": 239
    },
    {
      "epoch": 2.6212534059945503,
      "grad_norm": 0.23160408437252045,
      "learning_rate": 0.00010545454545454545,
      "loss": 6.384,
      "step": 240
    },
    {
      "epoch": 2.6321525885558583,
      "grad_norm": 0.2616265118122101,
      "learning_rate": 0.00010505050505050507,
      "loss": 6.3272,
      "step": 241
    },
    {
      "epoch": 2.6430517711171664,
      "grad_norm": 0.17835183441638947,
      "learning_rate": 0.00010464646464646466,
      "loss": 6.1332,
      "step": 242
    },
    {
      "epoch": 2.653950953678474,
      "grad_norm": 0.3156978487968445,
      "learning_rate": 0.00010424242424242425,
      "loss": 6.1321,
      "step": 243
    },
    {
      "epoch": 2.664850136239782,
      "grad_norm": 0.1692386120557785,
      "learning_rate": 0.00010383838383838385,
      "loss": 6.1671,
      "step": 244
    },
    {
      "epoch": 2.67574931880109,
      "grad_norm": 0.2413683384656906,
      "learning_rate": 0.00010343434343434344,
      "loss": 6.3664,
      "step": 245
    },
    {
      "epoch": 2.6866485013623977,
      "grad_norm": 0.12921543419361115,
      "learning_rate": 0.00010303030303030303,
      "loss": 6.1585,
      "step": 246
    },
    {
      "epoch": 2.697547683923706,
      "grad_norm": 0.15248817205429077,
      "learning_rate": 0.00010262626262626263,
      "loss": 6.1404,
      "step": 247
    },
    {
      "epoch": 2.708446866485014,
      "grad_norm": 0.20268793404102325,
      "learning_rate": 0.00010222222222222222,
      "loss": 6.2903,
      "step": 248
    },
    {
      "epoch": 2.7193460490463215,
      "grad_norm": 0.1451648324728012,
      "learning_rate": 0.00010181818181818181,
      "loss": 6.167,
      "step": 249
    },
    {
      "epoch": 2.7302452316076296,
      "grad_norm": 0.14130540192127228,
      "learning_rate": 0.00010141414141414143,
      "loss": 6.1824,
      "step": 250
    },
    {
      "epoch": 2.741144414168937,
      "grad_norm": 0.16221533715724945,
      "learning_rate": 0.00010101010101010102,
      "loss": 6.1688,
      "step": 251
    },
    {
      "epoch": 2.7520435967302452,
      "grad_norm": 0.2507758140563965,
      "learning_rate": 0.00010060606060606062,
      "loss": 6.179,
      "step": 252
    },
    {
      "epoch": 2.7629427792915533,
      "grad_norm": 0.13289335370063782,
      "learning_rate": 0.0001002020202020202,
      "loss": 6.1705,
      "step": 253
    },
    {
      "epoch": 2.773841961852861,
      "grad_norm": 0.13900728523731232,
      "learning_rate": 9.97979797979798e-05,
      "loss": 6.2319,
      "step": 254
    },
    {
      "epoch": 2.784741144414169,
      "grad_norm": 0.1445474475622177,
      "learning_rate": 9.939393939393939e-05,
      "loss": 6.1917,
      "step": 255
    },
    {
      "epoch": 2.7956403269754766,
      "grad_norm": 0.26752519607543945,
      "learning_rate": 9.8989898989899e-05,
      "loss": 6.1691,
      "step": 256
    },
    {
      "epoch": 2.8065395095367847,
      "grad_norm": 0.37636271119117737,
      "learning_rate": 9.85858585858586e-05,
      "loss": 6.1943,
      "step": 257
    },
    {
      "epoch": 2.8174386920980927,
      "grad_norm": 0.11944714188575745,
      "learning_rate": 9.818181818181818e-05,
      "loss": 6.1893,
      "step": 258
    },
    {
      "epoch": 2.8283378746594003,
      "grad_norm": 0.11981720477342606,
      "learning_rate": 9.777777777777778e-05,
      "loss": 6.1768,
      "step": 259
    },
    {
      "epoch": 2.8392370572207084,
      "grad_norm": 0.20344823598861694,
      "learning_rate": 9.737373737373738e-05,
      "loss": 6.205,
      "step": 260
    },
    {
      "epoch": 2.8501362397820165,
      "grad_norm": 0.15309491753578186,
      "learning_rate": 9.696969696969698e-05,
      "loss": 6.1888,
      "step": 261
    },
    {
      "epoch": 2.861035422343324,
      "grad_norm": 0.22223514318466187,
      "learning_rate": 9.656565656565657e-05,
      "loss": 6.2169,
      "step": 262
    },
    {
      "epoch": 2.871934604904632,
      "grad_norm": 0.4575447142124176,
      "learning_rate": 9.616161616161616e-05,
      "loss": 6.1841,
      "step": 263
    },
    {
      "epoch": 2.88283378746594,
      "grad_norm": 0.18480806052684784,
      "learning_rate": 9.575757575757576e-05,
      "loss": 6.3447,
      "step": 264
    },
    {
      "epoch": 2.893732970027248,
      "grad_norm": 0.44275179505348206,
      "learning_rate": 9.535353535353537e-05,
      "loss": 6.1697,
      "step": 265
    },
    {
      "epoch": 2.904632152588556,
      "grad_norm": 0.5216860771179199,
      "learning_rate": 9.494949494949495e-05,
      "loss": 6.1836,
      "step": 266
    },
    {
      "epoch": 2.915531335149864,
      "grad_norm": 0.25538700819015503,
      "learning_rate": 9.454545454545455e-05,
      "loss": 6.1733,
      "step": 267
    },
    {
      "epoch": 2.9264305177111716,
      "grad_norm": 0.23962710797786713,
      "learning_rate": 9.414141414141415e-05,
      "loss": 6.1606,
      "step": 268
    },
    {
      "epoch": 2.9373297002724796,
      "grad_norm": 0.33051827549934387,
      "learning_rate": 9.373737373737375e-05,
      "loss": 6.2506,
      "step": 269
    },
    {
      "epoch": 2.9482288828337877,
      "grad_norm": 0.3523387610912323,
      "learning_rate": 9.333333333333334e-05,
      "loss": 6.1995,
      "step": 270
    },
    {
      "epoch": 2.9591280653950953,
      "grad_norm": 0.4528197646141052,
      "learning_rate": 9.292929292929293e-05,
      "loss": 6.5313,
      "step": 271
    },
    {
      "epoch": 2.9700272479564034,
      "grad_norm": 0.15994518995285034,
      "learning_rate": 9.252525252525253e-05,
      "loss": 6.1704,
      "step": 272
    },
    {
      "epoch": 2.9809264305177114,
      "grad_norm": 0.4959411323070526,
      "learning_rate": 9.212121212121214e-05,
      "loss": 6.1982,
      "step": 273
    },
    {
      "epoch": 2.991825613079019,
      "grad_norm": 0.36732834577560425,
      "learning_rate": 9.171717171717172e-05,
      "loss": 6.2399,
      "step": 274
    },
    {
      "epoch": 3.005449591280654,
      "grad_norm": 0.41449519991874695,
      "learning_rate": 9.131313131313132e-05,
      "loss": 9.239,
      "step": 275
    },
    {
      "epoch": 3.016348773841962,
      "grad_norm": 0.49275797605514526,
      "learning_rate": 9.090909090909092e-05,
      "loss": 6.1771,
      "step": 276
    },
    {
      "epoch": 3.02724795640327,
      "grad_norm": 0.14379490911960602,
      "learning_rate": 9.050505050505052e-05,
      "loss": 6.209,
      "step": 277
    },
    {
      "epoch": 3.0381471389645776,
      "grad_norm": 0.1562594473361969,
      "learning_rate": 9.010101010101011e-05,
      "loss": 6.1669,
      "step": 278
    },
    {
      "epoch": 3.0490463215258856,
      "grad_norm": 0.22215306758880615,
      "learning_rate": 8.96969696969697e-05,
      "loss": 6.534,
      "step": 279
    },
    {
      "epoch": 3.0599455040871932,
      "grad_norm": 0.23347879946231842,
      "learning_rate": 8.92929292929293e-05,
      "loss": 6.154,
      "step": 280
    },
    {
      "epoch": 3.0708446866485013,
      "grad_norm": 0.1128433346748352,
      "learning_rate": 8.888888888888889e-05,
      "loss": 6.1795,
      "step": 281
    },
    {
      "epoch": 3.0817438692098094,
      "grad_norm": 0.12231786549091339,
      "learning_rate": 8.848484848484849e-05,
      "loss": 6.22,
      "step": 282
    },
    {
      "epoch": 3.092643051771117,
      "grad_norm": 0.12375138700008392,
      "learning_rate": 8.808080808080809e-05,
      "loss": 6.3243,
      "step": 283
    },
    {
      "epoch": 3.103542234332425,
      "grad_norm": 0.14191976189613342,
      "learning_rate": 8.767676767676767e-05,
      "loss": 6.25,
      "step": 284
    },
    {
      "epoch": 3.114441416893733,
      "grad_norm": 0.17911039292812347,
      "learning_rate": 8.727272727272727e-05,
      "loss": 6.1752,
      "step": 285
    },
    {
      "epoch": 3.1253405994550407,
      "grad_norm": 0.13937656581401825,
      "learning_rate": 8.686868686868688e-05,
      "loss": 6.1562,
      "step": 286
    },
    {
      "epoch": 3.136239782016349,
      "grad_norm": 0.19013279676437378,
      "learning_rate": 8.646464646464647e-05,
      "loss": 6.1372,
      "step": 287
    },
    {
      "epoch": 3.147138964577657,
      "grad_norm": 0.16956225037574768,
      "learning_rate": 8.606060606060606e-05,
      "loss": 6.1445,
      "step": 288
    },
    {
      "epoch": 3.1580381471389645,
      "grad_norm": 0.08755031228065491,
      "learning_rate": 8.565656565656566e-05,
      "loss": 6.1511,
      "step": 289
    },
    {
      "epoch": 3.1689373297002725,
      "grad_norm": 0.1031273603439331,
      "learning_rate": 8.525252525252526e-05,
      "loss": 6.1424,
      "step": 290
    },
    {
      "epoch": 3.1798365122615806,
      "grad_norm": 0.14533282816410065,
      "learning_rate": 8.484848484848486e-05,
      "loss": 6.1505,
      "step": 291
    },
    {
      "epoch": 3.190735694822888,
      "grad_norm": 0.37659046053886414,
      "learning_rate": 8.444444444444444e-05,
      "loss": 6.6586,
      "step": 292
    },
    {
      "epoch": 3.2016348773841963,
      "grad_norm": 0.10601754486560822,
      "learning_rate": 8.404040404040404e-05,
      "loss": 6.143,
      "step": 293
    },
    {
      "epoch": 3.212534059945504,
      "grad_norm": 0.5452451109886169,
      "learning_rate": 8.363636363636364e-05,
      "loss": 7.1831,
      "step": 294
    },
    {
      "epoch": 3.223433242506812,
      "grad_norm": 0.22060728073120117,
      "learning_rate": 8.323232323232324e-05,
      "loss": 6.1595,
      "step": 295
    },
    {
      "epoch": 3.23433242506812,
      "grad_norm": 0.1550436168909073,
      "learning_rate": 8.282828282828283e-05,
      "loss": 6.2545,
      "step": 296
    },
    {
      "epoch": 3.2452316076294276,
      "grad_norm": 0.22410456836223602,
      "learning_rate": 8.242424242424243e-05,
      "loss": 6.1355,
      "step": 297
    },
    {
      "epoch": 3.2561307901907357,
      "grad_norm": 0.21547560393810272,
      "learning_rate": 8.202020202020202e-05,
      "loss": 6.1469,
      "step": 298
    },
    {
      "epoch": 3.2670299727520438,
      "grad_norm": 0.14056307077407837,
      "learning_rate": 8.161616161616163e-05,
      "loss": 6.4813,
      "step": 299
    },
    {
      "epoch": 3.2779291553133514,
      "grad_norm": 0.14949879050254822,
      "learning_rate": 8.121212121212121e-05,
      "loss": 6.1359,
      "step": 300
    },
    {
      "epoch": 3.2888283378746594,
      "grad_norm": 0.1382267326116562,
      "learning_rate": 8.080808080808081e-05,
      "loss": 6.2013,
      "step": 301
    },
    {
      "epoch": 3.299727520435967,
      "grad_norm": 0.19791223108768463,
      "learning_rate": 8.040404040404041e-05,
      "loss": 6.1286,
      "step": 302
    },
    {
      "epoch": 3.310626702997275,
      "grad_norm": 0.13955998420715332,
      "learning_rate": 8e-05,
      "loss": 6.127,
      "step": 303
    },
    {
      "epoch": 3.321525885558583,
      "grad_norm": 0.14550985395908356,
      "learning_rate": 7.95959595959596e-05,
      "loss": 6.1668,
      "step": 304
    },
    {
      "epoch": 3.332425068119891,
      "grad_norm": 0.1612524390220642,
      "learning_rate": 7.919191919191919e-05,
      "loss": 6.2293,
      "step": 305
    },
    {
      "epoch": 3.343324250681199,
      "grad_norm": 0.10381856560707092,
      "learning_rate": 7.878787878787879e-05,
      "loss": 6.1429,
      "step": 306
    },
    {
      "epoch": 3.354223433242507,
      "grad_norm": 0.10719522833824158,
      "learning_rate": 7.83838383838384e-05,
      "loss": 6.1476,
      "step": 307
    },
    {
      "epoch": 3.3651226158038146,
      "grad_norm": 0.18627916276454926,
      "learning_rate": 7.797979797979798e-05,
      "loss": 6.1572,
      "step": 308
    },
    {
      "epoch": 3.3760217983651226,
      "grad_norm": 0.10271579772233963,
      "learning_rate": 7.757575757575758e-05,
      "loss": 6.1598,
      "step": 309
    },
    {
      "epoch": 3.3869209809264307,
      "grad_norm": 0.19747163355350494,
      "learning_rate": 7.717171717171718e-05,
      "loss": 6.1679,
      "step": 310
    },
    {
      "epoch": 3.3978201634877383,
      "grad_norm": 0.2724840044975281,
      "learning_rate": 7.676767676767676e-05,
      "loss": 6.2574,
      "step": 311
    },
    {
      "epoch": 3.4087193460490464,
      "grad_norm": 0.15871340036392212,
      "learning_rate": 7.636363636363637e-05,
      "loss": 6.3006,
      "step": 312
    },
    {
      "epoch": 3.4196185286103544,
      "grad_norm": 0.19969478249549866,
      "learning_rate": 7.595959595959596e-05,
      "loss": 6.1685,
      "step": 313
    },
    {
      "epoch": 3.430517711171662,
      "grad_norm": 0.1830209493637085,
      "learning_rate": 7.555555555555556e-05,
      "loss": 6.2126,
      "step": 314
    },
    {
      "epoch": 3.44141689373297,
      "grad_norm": 0.12285900115966797,
      "learning_rate": 7.515151515151515e-05,
      "loss": 6.1792,
      "step": 315
    },
    {
      "epoch": 3.452316076294278,
      "grad_norm": 0.17864428460597992,
      "learning_rate": 7.474747474747475e-05,
      "loss": 6.3188,
      "step": 316
    },
    {
      "epoch": 3.463215258855586,
      "grad_norm": 0.30143600702285767,
      "learning_rate": 7.434343434343435e-05,
      "loss": 6.2182,
      "step": 317
    },
    {
      "epoch": 3.474114441416894,
      "grad_norm": 0.215902641415596,
      "learning_rate": 7.393939393939395e-05,
      "loss": 6.1728,
      "step": 318
    },
    {
      "epoch": 3.4850136239782015,
      "grad_norm": 0.2436136156320572,
      "learning_rate": 7.353535353535353e-05,
      "loss": 6.2123,
      "step": 319
    },
    {
      "epoch": 3.4959128065395095,
      "grad_norm": 0.24740266799926758,
      "learning_rate": 7.313131313131314e-05,
      "loss": 6.1889,
      "step": 320
    },
    {
      "epoch": 3.5068119891008176,
      "grad_norm": 0.21778513491153717,
      "learning_rate": 7.272727272727273e-05,
      "loss": 6.1706,
      "step": 321
    },
    {
      "epoch": 3.517711171662125,
      "grad_norm": 0.22260497510433197,
      "learning_rate": 7.232323232323233e-05,
      "loss": 6.1825,
      "step": 322
    },
    {
      "epoch": 3.5286103542234333,
      "grad_norm": 0.13955512642860413,
      "learning_rate": 7.191919191919192e-05,
      "loss": 6.1847,
      "step": 323
    },
    {
      "epoch": 3.539509536784741,
      "grad_norm": 0.34359583258628845,
      "learning_rate": 7.151515151515152e-05,
      "loss": 6.4338,
      "step": 324
    },
    {
      "epoch": 3.550408719346049,
      "grad_norm": 0.10277872532606125,
      "learning_rate": 7.111111111111112e-05,
      "loss": 6.1621,
      "step": 325
    },
    {
      "epoch": 3.561307901907357,
      "grad_norm": 0.19088654220104218,
      "learning_rate": 7.07070707070707e-05,
      "loss": 6.1867,
      "step": 326
    },
    {
      "epoch": 3.5722070844686646,
      "grad_norm": 0.0887422040104866,
      "learning_rate": 7.03030303030303e-05,
      "loss": 6.141,
      "step": 327
    },
    {
      "epoch": 3.5831062670299727,
      "grad_norm": 0.11330597847700119,
      "learning_rate": 6.98989898989899e-05,
      "loss": 6.1588,
      "step": 328
    },
    {
      "epoch": 3.5940054495912808,
      "grad_norm": 0.09927017986774445,
      "learning_rate": 6.94949494949495e-05,
      "loss": 6.1231,
      "step": 329
    },
    {
      "epoch": 3.6049046321525884,
      "grad_norm": 0.10564055293798447,
      "learning_rate": 6.90909090909091e-05,
      "loss": 6.1484,
      "step": 330
    },
    {
      "epoch": 3.6158038147138964,
      "grad_norm": 0.3282771706581116,
      "learning_rate": 6.86868686868687e-05,
      "loss": 6.1763,
      "step": 331
    },
    {
      "epoch": 3.6267029972752045,
      "grad_norm": 0.43109074234962463,
      "learning_rate": 6.828282828282828e-05,
      "loss": 6.2524,
      "step": 332
    },
    {
      "epoch": 3.637602179836512,
      "grad_norm": 0.13901947438716888,
      "learning_rate": 6.787878787878789e-05,
      "loss": 6.1351,
      "step": 333
    },
    {
      "epoch": 3.64850136239782,
      "grad_norm": 0.11033597588539124,
      "learning_rate": 6.747474747474747e-05,
      "loss": 6.1543,
      "step": 334
    },
    {
      "epoch": 3.6594005449591283,
      "grad_norm": 0.12363231182098389,
      "learning_rate": 6.707070707070707e-05,
      "loss": 6.1509,
      "step": 335
    },
    {
      "epoch": 3.670299727520436,
      "grad_norm": 0.18957512080669403,
      "learning_rate": 6.666666666666667e-05,
      "loss": 6.1652,
      "step": 336
    },
    {
      "epoch": 3.681198910081744,
      "grad_norm": 0.23333989083766937,
      "learning_rate": 6.626262626262627e-05,
      "loss": 6.1927,
      "step": 337
    },
    {
      "epoch": 3.692098092643052,
      "grad_norm": 0.3563447892665863,
      "learning_rate": 6.585858585858587e-05,
      "loss": 6.1793,
      "step": 338
    },
    {
      "epoch": 3.7029972752043596,
      "grad_norm": 0.12097752094268799,
      "learning_rate": 6.545454545454546e-05,
      "loss": 6.1671,
      "step": 339
    },
    {
      "epoch": 3.7138964577656677,
      "grad_norm": 0.14230136573314667,
      "learning_rate": 6.505050505050505e-05,
      "loss": 6.1716,
      "step": 340
    },
    {
      "epoch": 3.7247956403269757,
      "grad_norm": 0.2817299962043762,
      "learning_rate": 6.464646464646466e-05,
      "loss": 6.1803,
      "step": 341
    },
    {
      "epoch": 3.7356948228882834,
      "grad_norm": 0.20600463449954987,
      "learning_rate": 6.424242424242424e-05,
      "loss": 6.3618,
      "step": 342
    },
    {
      "epoch": 3.7465940054495914,
      "grad_norm": 0.22449436783790588,
      "learning_rate": 6.383838383838384e-05,
      "loss": 6.1726,
      "step": 343
    },
    {
      "epoch": 3.757493188010899,
      "grad_norm": 0.12208205461502075,
      "learning_rate": 6.343434343434344e-05,
      "loss": 6.1956,
      "step": 344
    },
    {
      "epoch": 3.768392370572207,
      "grad_norm": 0.10846205800771713,
      "learning_rate": 6.303030303030302e-05,
      "loss": 6.1562,
      "step": 345
    },
    {
      "epoch": 3.7792915531335147,
      "grad_norm": 0.2708032727241516,
      "learning_rate": 6.262626262626264e-05,
      "loss": 6.2429,
      "step": 346
    },
    {
      "epoch": 3.790190735694823,
      "grad_norm": 0.42959117889404297,
      "learning_rate": 6.222222222222222e-05,
      "loss": 6.2097,
      "step": 347
    },
    {
      "epoch": 3.801089918256131,
      "grad_norm": 0.22440572082996368,
      "learning_rate": 6.181818181818182e-05,
      "loss": 6.3554,
      "step": 348
    },
    {
      "epoch": 3.8119891008174385,
      "grad_norm": 0.2479640245437622,
      "learning_rate": 6.141414141414142e-05,
      "loss": 6.1556,
      "step": 349
    },
    {
      "epoch": 3.8228882833787465,
      "grad_norm": 0.49595651030540466,
      "learning_rate": 6.101010101010102e-05,
      "loss": 7.1389,
      "step": 350
    },
    {
      "epoch": 3.8337874659400546,
      "grad_norm": 0.18350739777088165,
      "learning_rate": 6.060606060606061e-05,
      "loss": 6.1918,
      "step": 351
    },
    {
      "epoch": 3.844686648501362,
      "grad_norm": 0.2975408136844635,
      "learning_rate": 6.02020202020202e-05,
      "loss": 6.1625,
      "step": 352
    },
    {
      "epoch": 3.8555858310626703,
      "grad_norm": 0.33965060114860535,
      "learning_rate": 5.97979797979798e-05,
      "loss": 6.1698,
      "step": 353
    },
    {
      "epoch": 3.8664850136239783,
      "grad_norm": 0.1836046725511551,
      "learning_rate": 5.93939393939394e-05,
      "loss": 6.1667,
      "step": 354
    },
    {
      "epoch": 3.877384196185286,
      "grad_norm": 0.19976381957530975,
      "learning_rate": 5.8989898989898996e-05,
      "loss": 6.1561,
      "step": 355
    },
    {
      "epoch": 3.888283378746594,
      "grad_norm": 0.19982588291168213,
      "learning_rate": 5.858585858585859e-05,
      "loss": 6.277,
      "step": 356
    },
    {
      "epoch": 3.899182561307902,
      "grad_norm": 0.11651279032230377,
      "learning_rate": 5.818181818181818e-05,
      "loss": 6.1483,
      "step": 357
    },
    {
      "epoch": 3.9100817438692097,
      "grad_norm": 0.27458274364471436,
      "learning_rate": 5.7777777777777776e-05,
      "loss": 6.1632,
      "step": 358
    },
    {
      "epoch": 3.9209809264305178,
      "grad_norm": 0.0916130468249321,
      "learning_rate": 5.737373737373738e-05,
      "loss": 6.1351,
      "step": 359
    },
    {
      "epoch": 3.931880108991826,
      "grad_norm": 0.23624251782894135,
      "learning_rate": 5.696969696969697e-05,
      "loss": 6.3717,
      "step": 360
    },
    {
      "epoch": 3.9427792915531334,
      "grad_norm": 0.14670449495315552,
      "learning_rate": 5.6565656565656563e-05,
      "loss": 6.1572,
      "step": 361
    },
    {
      "epoch": 3.9536784741144415,
      "grad_norm": 0.12288593500852585,
      "learning_rate": 5.616161616161616e-05,
      "loss": 6.1509,
      "step": 362
    },
    {
      "epoch": 3.9645776566757496,
      "grad_norm": 0.1172018051147461,
      "learning_rate": 5.5757575757575766e-05,
      "loss": 6.2154,
      "step": 363
    },
    {
      "epoch": 3.975476839237057,
      "grad_norm": 0.16450701653957367,
      "learning_rate": 5.535353535353536e-05,
      "loss": 6.1417,
      "step": 364
    },
    {
      "epoch": 3.9863760217983653,
      "grad_norm": 0.17729254066944122,
      "learning_rate": 5.494949494949495e-05,
      "loss": 6.237,
      "step": 365
    },
    {
      "epoch": 3.997275204359673,
      "grad_norm": 0.1622830182313919,
      "learning_rate": 5.4545454545454546e-05,
      "loss": 9.2183,
      "step": 366
    },
    {
      "epoch": 4.010899182561308,
      "grad_norm": 0.25812089443206787,
      "learning_rate": 5.414141414141415e-05,
      "loss": 6.6206,
      "step": 367
    },
    {
      "epoch": 4.021798365122616,
      "grad_norm": 0.10862333327531815,
      "learning_rate": 5.373737373737374e-05,
      "loss": 6.1384,
      "step": 368
    },
    {
      "epoch": 4.032697547683924,
      "grad_norm": 0.23395223915576935,
      "learning_rate": 5.333333333333333e-05,
      "loss": 6.1301,
      "step": 369
    },
    {
      "epoch": 4.043596730245231,
      "grad_norm": 0.17671817541122437,
      "learning_rate": 5.292929292929293e-05,
      "loss": 6.1281,
      "step": 370
    },
    {
      "epoch": 4.05449591280654,
      "grad_norm": 0.18898046016693115,
      "learning_rate": 5.2525252525252536e-05,
      "loss": 6.1806,
      "step": 371
    },
    {
      "epoch": 4.0653950953678475,
      "grad_norm": 0.16608364880084991,
      "learning_rate": 5.212121212121213e-05,
      "loss": 6.1582,
      "step": 372
    },
    {
      "epoch": 4.076294277929155,
      "grad_norm": 0.15272347629070282,
      "learning_rate": 5.171717171717172e-05,
      "loss": 6.1441,
      "step": 373
    },
    {
      "epoch": 4.087193460490464,
      "grad_norm": 0.14940135180950165,
      "learning_rate": 5.1313131313131316e-05,
      "loss": 6.1524,
      "step": 374
    },
    {
      "epoch": 4.098092643051771,
      "grad_norm": 0.12995606660842896,
      "learning_rate": 5.090909090909091e-05,
      "loss": 6.1294,
      "step": 375
    },
    {
      "epoch": 4.108991825613079,
      "grad_norm": 0.09847455471754074,
      "learning_rate": 5.050505050505051e-05,
      "loss": 6.1608,
      "step": 376
    },
    {
      "epoch": 4.1198910081743865,
      "grad_norm": 0.12721924483776093,
      "learning_rate": 5.01010101010101e-05,
      "loss": 6.1577,
      "step": 377
    },
    {
      "epoch": 4.130790190735695,
      "grad_norm": 0.16987353563308716,
      "learning_rate": 4.9696969696969694e-05,
      "loss": 6.1492,
      "step": 378
    },
    {
      "epoch": 4.141689373297003,
      "grad_norm": 0.17095069587230682,
      "learning_rate": 4.92929292929293e-05,
      "loss": 6.1358,
      "step": 379
    },
    {
      "epoch": 4.15258855585831,
      "grad_norm": 0.12413057684898376,
      "learning_rate": 4.888888888888889e-05,
      "loss": 6.1557,
      "step": 380
    },
    {
      "epoch": 4.163487738419619,
      "grad_norm": 0.09834663569927216,
      "learning_rate": 4.848484848484849e-05,
      "loss": 6.1509,
      "step": 381
    },
    {
      "epoch": 4.174386920980926,
      "grad_norm": 0.15776708722114563,
      "learning_rate": 4.808080808080808e-05,
      "loss": 6.1801,
      "step": 382
    },
    {
      "epoch": 4.185286103542234,
      "grad_norm": 0.14047148823738098,
      "learning_rate": 4.7676767676767684e-05,
      "loss": 6.17,
      "step": 383
    },
    {
      "epoch": 4.1961852861035425,
      "grad_norm": 0.4055856764316559,
      "learning_rate": 4.7272727272727275e-05,
      "loss": 6.2462,
      "step": 384
    },
    {
      "epoch": 4.20708446866485,
      "grad_norm": 0.12411222606897354,
      "learning_rate": 4.686868686868687e-05,
      "loss": 6.1751,
      "step": 385
    },
    {
      "epoch": 4.217983651226158,
      "grad_norm": 0.1417582631111145,
      "learning_rate": 4.6464646464646464e-05,
      "loss": 6.207,
      "step": 386
    },
    {
      "epoch": 4.228882833787466,
      "grad_norm": 0.1599162071943283,
      "learning_rate": 4.606060606060607e-05,
      "loss": 6.1669,
      "step": 387
    },
    {
      "epoch": 4.239782016348774,
      "grad_norm": 0.11022761464118958,
      "learning_rate": 4.565656565656566e-05,
      "loss": 6.1717,
      "step": 388
    },
    {
      "epoch": 4.2506811989100814,
      "grad_norm": 0.2099771499633789,
      "learning_rate": 4.525252525252526e-05,
      "loss": 6.1848,
      "step": 389
    },
    {
      "epoch": 4.26158038147139,
      "grad_norm": 0.38737088441848755,
      "learning_rate": 4.484848484848485e-05,
      "loss": 6.1843,
      "step": 390
    },
    {
      "epoch": 4.272479564032698,
      "grad_norm": 0.19913600385189056,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 6.181,
      "step": 391
    },
    {
      "epoch": 4.283378746594005,
      "grad_norm": 0.18423688411712646,
      "learning_rate": 4.4040404040404044e-05,
      "loss": 6.1798,
      "step": 392
    },
    {
      "epoch": 4.294277929155314,
      "grad_norm": 0.17174512147903442,
      "learning_rate": 4.3636363636363636e-05,
      "loss": 6.2697,
      "step": 393
    },
    {
      "epoch": 4.305177111716621,
      "grad_norm": 0.1219327375292778,
      "learning_rate": 4.3232323232323234e-05,
      "loss": 6.1861,
      "step": 394
    },
    {
      "epoch": 4.316076294277929,
      "grad_norm": 0.5196271538734436,
      "learning_rate": 4.282828282828283e-05,
      "loss": 6.5272,
      "step": 395
    },
    {
      "epoch": 4.3269754768392374,
      "grad_norm": 0.17318467795848846,
      "learning_rate": 4.242424242424243e-05,
      "loss": 6.1996,
      "step": 396
    },
    {
      "epoch": 4.337874659400545,
      "grad_norm": 0.40889430046081543,
      "learning_rate": 4.202020202020202e-05,
      "loss": 6.908,
      "step": 397
    },
    {
      "epoch": 4.348773841961853,
      "grad_norm": 0.18795502185821533,
      "learning_rate": 4.161616161616162e-05,
      "loss": 6.2444,
      "step": 398
    },
    {
      "epoch": 4.359673024523161,
      "grad_norm": 0.1756502389907837,
      "learning_rate": 4.1212121212121216e-05,
      "loss": 6.1558,
      "step": 399
    },
    {
      "epoch": 4.370572207084469,
      "grad_norm": 0.20153924822807312,
      "learning_rate": 4.0808080808080814e-05,
      "loss": 6.1704,
      "step": 400
    },
    {
      "epoch": 4.381471389645776,
      "grad_norm": 0.14148367941379547,
      "learning_rate": 4.0404040404040405e-05,
      "loss": 6.1617,
      "step": 401
    },
    {
      "epoch": 4.392370572207085,
      "grad_norm": 0.16998320817947388,
      "learning_rate": 4e-05,
      "loss": 6.1476,
      "step": 402
    },
    {
      "epoch": 4.4032697547683926,
      "grad_norm": 0.23891212046146393,
      "learning_rate": 3.9595959595959594e-05,
      "loss": 6.3065,
      "step": 403
    },
    {
      "epoch": 4.4141689373297,
      "grad_norm": 0.17936837673187256,
      "learning_rate": 3.91919191919192e-05,
      "loss": 6.141,
      "step": 404
    },
    {
      "epoch": 4.425068119891008,
      "grad_norm": 0.13204631209373474,
      "learning_rate": 3.878787878787879e-05,
      "loss": 6.1693,
      "step": 405
    },
    {
      "epoch": 4.435967302452316,
      "grad_norm": 0.23752065002918243,
      "learning_rate": 3.838383838383838e-05,
      "loss": 6.1486,
      "step": 406
    },
    {
      "epoch": 4.446866485013624,
      "grad_norm": 0.13474120199680328,
      "learning_rate": 3.797979797979798e-05,
      "loss": 6.1781,
      "step": 407
    },
    {
      "epoch": 4.4577656675749315,
      "grad_norm": 0.639192521572113,
      "learning_rate": 3.757575757575758e-05,
      "loss": 7.1072,
      "step": 408
    },
    {
      "epoch": 4.46866485013624,
      "grad_norm": 0.31699126958847046,
      "learning_rate": 3.7171717171717175e-05,
      "loss": 6.1528,
      "step": 409
    },
    {
      "epoch": 4.479564032697548,
      "grad_norm": 0.14006993174552917,
      "learning_rate": 3.6767676767676766e-05,
      "loss": 6.1284,
      "step": 410
    },
    {
      "epoch": 4.490463215258855,
      "grad_norm": 0.18349085748195648,
      "learning_rate": 3.6363636363636364e-05,
      "loss": 6.4731,
      "step": 411
    },
    {
      "epoch": 4.501362397820164,
      "grad_norm": 0.16099974513053894,
      "learning_rate": 3.595959595959596e-05,
      "loss": 6.1471,
      "step": 412
    },
    {
      "epoch": 4.512261580381471,
      "grad_norm": 0.17679904401302338,
      "learning_rate": 3.555555555555556e-05,
      "loss": 6.1535,
      "step": 413
    },
    {
      "epoch": 4.523160762942779,
      "grad_norm": 0.16737085580825806,
      "learning_rate": 3.515151515151515e-05,
      "loss": 6.1319,
      "step": 414
    },
    {
      "epoch": 4.5340599455040875,
      "grad_norm": 0.11516544967889786,
      "learning_rate": 3.474747474747475e-05,
      "loss": 6.1876,
      "step": 415
    },
    {
      "epoch": 4.544959128065395,
      "grad_norm": 0.13226546347141266,
      "learning_rate": 3.434343434343435e-05,
      "loss": 6.1415,
      "step": 416
    },
    {
      "epoch": 4.555858310626703,
      "grad_norm": 0.5673292875289917,
      "learning_rate": 3.3939393939393945e-05,
      "loss": 6.1813,
      "step": 417
    },
    {
      "epoch": 4.566757493188011,
      "grad_norm": 0.19020290672779083,
      "learning_rate": 3.3535353535353536e-05,
      "loss": 6.1366,
      "step": 418
    },
    {
      "epoch": 4.577656675749319,
      "grad_norm": 0.1316557228565216,
      "learning_rate": 3.3131313131313134e-05,
      "loss": 6.1256,
      "step": 419
    },
    {
      "epoch": 4.5885558583106265,
      "grad_norm": 0.18218891322612762,
      "learning_rate": 3.272727272727273e-05,
      "loss": 6.1428,
      "step": 420
    },
    {
      "epoch": 4.599455040871934,
      "grad_norm": 0.16241154074668884,
      "learning_rate": 3.232323232323233e-05,
      "loss": 6.3538,
      "step": 421
    },
    {
      "epoch": 4.610354223433243,
      "grad_norm": 0.24633854627609253,
      "learning_rate": 3.191919191919192e-05,
      "loss": 6.1339,
      "step": 422
    },
    {
      "epoch": 4.62125340599455,
      "grad_norm": 0.13929039239883423,
      "learning_rate": 3.151515151515151e-05,
      "loss": 6.1458,
      "step": 423
    },
    {
      "epoch": 4.632152588555858,
      "grad_norm": 0.18816447257995605,
      "learning_rate": 3.111111111111111e-05,
      "loss": 6.2432,
      "step": 424
    },
    {
      "epoch": 4.643051771117166,
      "grad_norm": 0.20620077848434448,
      "learning_rate": 3.070707070707071e-05,
      "loss": 6.156,
      "step": 425
    },
    {
      "epoch": 4.653950953678474,
      "grad_norm": 0.20312055945396423,
      "learning_rate": 3.0303030303030306e-05,
      "loss": 6.1537,
      "step": 426
    },
    {
      "epoch": 4.664850136239782,
      "grad_norm": 0.12666474282741547,
      "learning_rate": 2.98989898989899e-05,
      "loss": 6.2204,
      "step": 427
    },
    {
      "epoch": 4.67574931880109,
      "grad_norm": 0.41980618238449097,
      "learning_rate": 2.9494949494949498e-05,
      "loss": 6.5506,
      "step": 428
    },
    {
      "epoch": 4.686648501362398,
      "grad_norm": 0.24480164051055908,
      "learning_rate": 2.909090909090909e-05,
      "loss": 6.1571,
      "step": 429
    },
    {
      "epoch": 4.697547683923705,
      "grad_norm": 0.2902083098888397,
      "learning_rate": 2.868686868686869e-05,
      "loss": 6.1515,
      "step": 430
    },
    {
      "epoch": 4.708446866485014,
      "grad_norm": 0.11313210427761078,
      "learning_rate": 2.8282828282828282e-05,
      "loss": 6.203,
      "step": 431
    },
    {
      "epoch": 4.7193460490463215,
      "grad_norm": 0.14203494787216187,
      "learning_rate": 2.7878787878787883e-05,
      "loss": 6.1618,
      "step": 432
    },
    {
      "epoch": 4.730245231607629,
      "grad_norm": 0.13528816401958466,
      "learning_rate": 2.7474747474747474e-05,
      "loss": 6.3833,
      "step": 433
    },
    {
      "epoch": 4.741144414168938,
      "grad_norm": 0.11331304907798767,
      "learning_rate": 2.7070707070707075e-05,
      "loss": 6.1472,
      "step": 434
    },
    {
      "epoch": 4.752043596730245,
      "grad_norm": 0.11952386051416397,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 6.1477,
      "step": 435
    },
    {
      "epoch": 4.762942779291553,
      "grad_norm": 0.22013674676418304,
      "learning_rate": 2.6262626262626268e-05,
      "loss": 6.1598,
      "step": 436
    },
    {
      "epoch": 4.773841961852861,
      "grad_norm": 0.10620447248220444,
      "learning_rate": 2.585858585858586e-05,
      "loss": 6.2546,
      "step": 437
    },
    {
      "epoch": 4.784741144414169,
      "grad_norm": 0.1305541694164276,
      "learning_rate": 2.5454545454545454e-05,
      "loss": 6.2223,
      "step": 438
    },
    {
      "epoch": 4.795640326975477,
      "grad_norm": 0.14448855817317963,
      "learning_rate": 2.505050505050505e-05,
      "loss": 6.1602,
      "step": 439
    },
    {
      "epoch": 4.806539509536785,
      "grad_norm": 0.1863837093114853,
      "learning_rate": 2.464646464646465e-05,
      "loss": 6.145,
      "step": 440
    },
    {
      "epoch": 4.817438692098093,
      "grad_norm": 0.47183990478515625,
      "learning_rate": 2.4242424242424244e-05,
      "loss": 6.3824,
      "step": 441
    },
    {
      "epoch": 4.8283378746594,
      "grad_norm": 0.18477211892604828,
      "learning_rate": 2.3838383838383842e-05,
      "loss": 6.1644,
      "step": 442
    },
    {
      "epoch": 4.839237057220709,
      "grad_norm": 0.36106404662132263,
      "learning_rate": 2.3434343434343436e-05,
      "loss": 6.1992,
      "step": 443
    },
    {
      "epoch": 4.8501362397820165,
      "grad_norm": 0.11133147776126862,
      "learning_rate": 2.3030303030303034e-05,
      "loss": 6.1298,
      "step": 444
    },
    {
      "epoch": 4.861035422343324,
      "grad_norm": 0.4021960198879242,
      "learning_rate": 2.262626262626263e-05,
      "loss": 6.2763,
      "step": 445
    },
    {
      "epoch": 4.871934604904633,
      "grad_norm": 0.14010433852672577,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 6.1838,
      "step": 446
    },
    {
      "epoch": 4.88283378746594,
      "grad_norm": 0.45312055945396423,
      "learning_rate": 2.1818181818181818e-05,
      "loss": 6.66,
      "step": 447
    },
    {
      "epoch": 4.893732970027248,
      "grad_norm": 0.1600738912820816,
      "learning_rate": 2.1414141414141416e-05,
      "loss": 6.1646,
      "step": 448
    },
    {
      "epoch": 4.904632152588556,
      "grad_norm": 0.11380810290575027,
      "learning_rate": 2.101010101010101e-05,
      "loss": 6.1415,
      "step": 449
    },
    {
      "epoch": 4.915531335149864,
      "grad_norm": 0.14316502213478088,
      "learning_rate": 2.0606060606060608e-05,
      "loss": 6.1984,
      "step": 450
    },
    {
      "epoch": 4.926430517711172,
      "grad_norm": 0.2457677721977234,
      "learning_rate": 2.0202020202020203e-05,
      "loss": 6.1522,
      "step": 451
    },
    {
      "epoch": 4.937329700272479,
      "grad_norm": 0.15555907785892487,
      "learning_rate": 1.9797979797979797e-05,
      "loss": 6.1647,
      "step": 452
    },
    {
      "epoch": 4.948228882833788,
      "grad_norm": 0.12678752839565277,
      "learning_rate": 1.9393939393939395e-05,
      "loss": 6.1495,
      "step": 453
    },
    {
      "epoch": 4.959128065395095,
      "grad_norm": 0.1315479725599289,
      "learning_rate": 1.898989898989899e-05,
      "loss": 6.1946,
      "step": 454
    },
    {
      "epoch": 4.970027247956403,
      "grad_norm": 0.22341807186603546,
      "learning_rate": 1.8585858585858588e-05,
      "loss": 6.14,
      "step": 455
    },
    {
      "epoch": 4.9809264305177114,
      "grad_norm": 0.397432804107666,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 6.1562,
      "step": 456
    },
    {
      "epoch": 4.991825613079019,
      "grad_norm": 0.16656984388828278,
      "learning_rate": 1.777777777777778e-05,
      "loss": 6.1628,
      "step": 457
    },
    {
      "epoch": 5.005449591280654,
      "grad_norm": 0.47680985927581787,
      "learning_rate": 1.7373737373737375e-05,
      "loss": 9.2237,
      "step": 458
    },
    {
      "epoch": 5.016348773841962,
      "grad_norm": 0.13037891685962677,
      "learning_rate": 1.6969696969696972e-05,
      "loss": 6.1438,
      "step": 459
    },
    {
      "epoch": 5.0272479564032695,
      "grad_norm": 0.12159968167543411,
      "learning_rate": 1.6565656565656567e-05,
      "loss": 6.1595,
      "step": 460
    },
    {
      "epoch": 5.038147138964578,
      "grad_norm": 0.11064036190509796,
      "learning_rate": 1.6161616161616165e-05,
      "loss": 6.1593,
      "step": 461
    },
    {
      "epoch": 5.049046321525886,
      "grad_norm": 0.1583862155675888,
      "learning_rate": 1.5757575757575756e-05,
      "loss": 6.1594,
      "step": 462
    },
    {
      "epoch": 5.059945504087193,
      "grad_norm": 0.2619314193725586,
      "learning_rate": 1.5353535353535354e-05,
      "loss": 6.3912,
      "step": 463
    },
    {
      "epoch": 5.070844686648502,
      "grad_norm": 0.10678429901599884,
      "learning_rate": 1.494949494949495e-05,
      "loss": 6.1448,
      "step": 464
    },
    {
      "epoch": 5.081743869209809,
      "grad_norm": 0.24632836878299713,
      "learning_rate": 1.4545454545454545e-05,
      "loss": 6.1786,
      "step": 465
    },
    {
      "epoch": 5.092643051771117,
      "grad_norm": 0.24659158289432526,
      "learning_rate": 1.4141414141414141e-05,
      "loss": 6.2495,
      "step": 466
    },
    {
      "epoch": 5.1035422343324255,
      "grad_norm": 0.10519186407327652,
      "learning_rate": 1.3737373737373737e-05,
      "loss": 6.1399,
      "step": 467
    },
    {
      "epoch": 5.114441416893733,
      "grad_norm": 0.30278512835502625,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 6.1993,
      "step": 468
    },
    {
      "epoch": 5.125340599455041,
      "grad_norm": 0.41121378540992737,
      "learning_rate": 1.292929292929293e-05,
      "loss": 6.3494,
      "step": 469
    },
    {
      "epoch": 5.136239782016348,
      "grad_norm": 0.2749239504337311,
      "learning_rate": 1.2525252525252526e-05,
      "loss": 6.1745,
      "step": 470
    },
    {
      "epoch": 5.147138964577657,
      "grad_norm": 0.3106517791748047,
      "learning_rate": 1.2121212121212122e-05,
      "loss": 6.2632,
      "step": 471
    },
    {
      "epoch": 5.1580381471389645,
      "grad_norm": 0.19723156094551086,
      "learning_rate": 1.1717171717171718e-05,
      "loss": 6.1638,
      "step": 472
    },
    {
      "epoch": 5.168937329700272,
      "grad_norm": 0.38709595799446106,
      "learning_rate": 1.1313131313131314e-05,
      "loss": 6.3415,
      "step": 473
    },
    {
      "epoch": 5.179836512261581,
      "grad_norm": 0.1426631212234497,
      "learning_rate": 1.0909090909090909e-05,
      "loss": 6.3154,
      "step": 474
    },
    {
      "epoch": 5.190735694822888,
      "grad_norm": 0.23347443342208862,
      "learning_rate": 1.0505050505050505e-05,
      "loss": 6.1717,
      "step": 475
    },
    {
      "epoch": 5.201634877384196,
      "grad_norm": 0.09223368018865585,
      "learning_rate": 1.0101010101010101e-05,
      "loss": 6.1606,
      "step": 476
    },
    {
      "epoch": 5.212534059945504,
      "grad_norm": 0.1321372389793396,
      "learning_rate": 9.696969696969698e-06,
      "loss": 6.2569,
      "step": 477
    },
    {
      "epoch": 5.223433242506812,
      "grad_norm": 0.14801448583602905,
      "learning_rate": 9.292929292929294e-06,
      "loss": 6.1669,
      "step": 478
    },
    {
      "epoch": 5.23433242506812,
      "grad_norm": 0.20771123468875885,
      "learning_rate": 8.88888888888889e-06,
      "loss": 6.173,
      "step": 479
    },
    {
      "epoch": 5.245231607629428,
      "grad_norm": 0.12525993585586548,
      "learning_rate": 8.484848484848486e-06,
      "loss": 6.171,
      "step": 480
    },
    {
      "epoch": 5.256130790190736,
      "grad_norm": 0.12884493172168732,
      "learning_rate": 8.080808080808082e-06,
      "loss": 6.1318,
      "step": 481
    },
    {
      "epoch": 5.267029972752043,
      "grad_norm": 0.1414600908756256,
      "learning_rate": 7.676767676767677e-06,
      "loss": 6.1838,
      "step": 482
    },
    {
      "epoch": 5.277929155313352,
      "grad_norm": 0.25564873218536377,
      "learning_rate": 7.272727272727272e-06,
      "loss": 6.1501,
      "step": 483
    },
    {
      "epoch": 5.2888283378746594,
      "grad_norm": 0.3402186334133148,
      "learning_rate": 6.8686868686868685e-06,
      "loss": 6.855,
      "step": 484
    },
    {
      "epoch": 5.299727520435967,
      "grad_norm": 0.21217402815818787,
      "learning_rate": 6.464646464646465e-06,
      "loss": 6.1712,
      "step": 485
    },
    {
      "epoch": 5.310626702997276,
      "grad_norm": 0.20217765867710114,
      "learning_rate": 6.060606060606061e-06,
      "loss": 6.1735,
      "step": 486
    },
    {
      "epoch": 5.321525885558583,
      "grad_norm": 0.14054284989833832,
      "learning_rate": 5.656565656565657e-06,
      "loss": 6.1637,
      "step": 487
    },
    {
      "epoch": 5.332425068119891,
      "grad_norm": 0.22656294703483582,
      "learning_rate": 5.2525252525252526e-06,
      "loss": 6.1563,
      "step": 488
    },
    {
      "epoch": 5.343324250681199,
      "grad_norm": 0.10825368016958237,
      "learning_rate": 4.848484848484849e-06,
      "loss": 6.202,
      "step": 489
    },
    {
      "epoch": 5.354223433242507,
      "grad_norm": 0.17495879530906677,
      "learning_rate": 4.444444444444445e-06,
      "loss": 6.2115,
      "step": 490
    },
    {
      "epoch": 5.3651226158038146,
      "grad_norm": 0.10381829738616943,
      "learning_rate": 4.040404040404041e-06,
      "loss": 6.1697,
      "step": 491
    },
    {
      "epoch": 5.376021798365123,
      "grad_norm": 0.1631442904472351,
      "learning_rate": 3.636363636363636e-06,
      "loss": 6.1498,
      "step": 492
    },
    {
      "epoch": 5.386920980926431,
      "grad_norm": 0.13070239126682281,
      "learning_rate": 3.2323232323232324e-06,
      "loss": 6.1829,
      "step": 493
    },
    {
      "epoch": 5.397820163487738,
      "grad_norm": 0.13861213624477386,
      "learning_rate": 2.8282828282828286e-06,
      "loss": 6.1574,
      "step": 494
    },
    {
      "epoch": 5.408719346049046,
      "grad_norm": 0.24785035848617554,
      "learning_rate": 2.4242424242424244e-06,
      "loss": 6.5561,
      "step": 495
    },
    {
      "epoch": 5.419618528610354,
      "grad_norm": 0.12529440224170685,
      "learning_rate": 2.0202020202020206e-06,
      "loss": 6.168,
      "step": 496
    },
    {
      "epoch": 5.430517711171662,
      "grad_norm": 0.14261850714683533,
      "learning_rate": 1.6161616161616162e-06,
      "loss": 6.1755,
      "step": 497
    },
    {
      "epoch": 5.44141689373297,
      "grad_norm": 0.3609119951725006,
      "learning_rate": 1.2121212121212122e-06,
      "loss": 6.1591,
      "step": 498
    },
    {
      "epoch": 5.452316076294278,
      "grad_norm": 0.19529031217098236,
      "learning_rate": 8.080808080808081e-07,
      "loss": 6.2801,
      "step": 499
    },
    {
      "epoch": 5.463215258855586,
      "grad_norm": 0.1783263236284256,
      "learning_rate": 4.0404040404040405e-07,
      "loss": 6.1468,
      "step": 500
    }
  ],
  "logging_steps": 1,
  "max_steps": 500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.2096532119552e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}