{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9904153354632586,
  "eval_steps": 500,
  "global_step": 312,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.009584664536741214,
      "grad_norm": 6.805100440979004,
      "learning_rate": 3.125e-07,
      "loss": 0.8451,
      "step": 1
    },
    {
      "epoch": 0.019169329073482427,
      "grad_norm": 6.944859504699707,
      "learning_rate": 6.25e-07,
      "loss": 0.8823,
      "step": 2
    },
    {
      "epoch": 0.02875399361022364,
      "grad_norm": 6.678200721740723,
      "learning_rate": 9.375000000000001e-07,
      "loss": 0.8476,
      "step": 3
    },
    {
      "epoch": 0.038338658146964855,
      "grad_norm": 6.952979564666748,
      "learning_rate": 1.25e-06,
      "loss": 0.912,
      "step": 4
    },
    {
      "epoch": 0.04792332268370607,
      "grad_norm": 6.973270416259766,
      "learning_rate": 1.5625e-06,
      "loss": 0.8781,
      "step": 5
    },
    {
      "epoch": 0.05750798722044728,
      "grad_norm": 6.345473289489746,
      "learning_rate": 1.8750000000000003e-06,
      "loss": 0.8242,
      "step": 6
    },
    {
      "epoch": 0.0670926517571885,
      "grad_norm": 5.5535969734191895,
      "learning_rate": 2.1875000000000002e-06,
      "loss": 0.8251,
      "step": 7
    },
    {
      "epoch": 0.07667731629392971,
      "grad_norm": 5.133858680725098,
      "learning_rate": 2.5e-06,
      "loss": 0.8043,
      "step": 8
    },
    {
      "epoch": 0.08626198083067092,
      "grad_norm": 3.2064390182495117,
      "learning_rate": 2.8125e-06,
      "loss": 0.7508,
      "step": 9
    },
    {
      "epoch": 0.09584664536741214,
      "grad_norm": 3.2898361682891846,
      "learning_rate": 3.125e-06,
      "loss": 0.7751,
      "step": 10
    },
    {
      "epoch": 0.10543130990415335,
      "grad_norm": 2.8389291763305664,
      "learning_rate": 3.4375e-06,
      "loss": 0.7334,
      "step": 11
    },
    {
      "epoch": 0.11501597444089456,
      "grad_norm": 2.9657039642333984,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.6988,
      "step": 12
    },
    {
      "epoch": 0.12460063897763578,
      "grad_norm": 3.3944308757781982,
      "learning_rate": 4.0625000000000005e-06,
      "loss": 0.6876,
      "step": 13
    },
    {
      "epoch": 0.134185303514377,
      "grad_norm": 3.525303840637207,
      "learning_rate": 4.3750000000000005e-06,
      "loss": 0.6934,
      "step": 14
    },
    {
      "epoch": 0.14376996805111822,
      "grad_norm": 3.2581071853637695,
      "learning_rate": 4.6875000000000004e-06,
      "loss": 0.6867,
      "step": 15
    },
    {
      "epoch": 0.15335463258785942,
      "grad_norm": 2.7474966049194336,
      "learning_rate": 5e-06,
      "loss": 0.617,
      "step": 16
    },
    {
      "epoch": 0.16293929712460065,
      "grad_norm": 2.480701208114624,
      "learning_rate": 5.3125e-06,
      "loss": 0.6209,
      "step": 17
    },
    {
      "epoch": 0.17252396166134185,
      "grad_norm": 2.3733997344970703,
      "learning_rate": 5.625e-06,
      "loss": 0.6337,
      "step": 18
    },
    {
      "epoch": 0.18210862619808307,
      "grad_norm": 2.1485886573791504,
      "learning_rate": 5.9375e-06,
      "loss": 0.6037,
      "step": 19
    },
    {
      "epoch": 0.19169329073482427,
      "grad_norm": 1.8110982179641724,
      "learning_rate": 6.25e-06,
      "loss": 0.6138,
      "step": 20
    },
    {
      "epoch": 0.2012779552715655,
      "grad_norm": 1.5806992053985596,
      "learning_rate": 6.5625e-06,
      "loss": 0.5795,
      "step": 21
    },
    {
      "epoch": 0.2108626198083067,
      "grad_norm": 1.5841341018676758,
      "learning_rate": 6.875e-06,
      "loss": 0.6092,
      "step": 22
    },
    {
      "epoch": 0.22044728434504793,
      "grad_norm": 1.566663146018982,
      "learning_rate": 7.1875e-06,
      "loss": 0.572,
      "step": 23
    },
    {
      "epoch": 0.23003194888178913,
      "grad_norm": 1.5222035646438599,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.5908,
      "step": 24
    },
    {
      "epoch": 0.23961661341853036,
      "grad_norm": 1.2543559074401855,
      "learning_rate": 7.8125e-06,
      "loss": 0.5891,
      "step": 25
    },
    {
      "epoch": 0.24920127795527156,
      "grad_norm": 1.1692081689834595,
      "learning_rate": 8.125000000000001e-06,
      "loss": 0.5471,
      "step": 26
    },
    {
      "epoch": 0.25878594249201275,
      "grad_norm": 1.3132680654525757,
      "learning_rate": 8.4375e-06,
      "loss": 0.5494,
      "step": 27
    },
    {
      "epoch": 0.268370607028754,
      "grad_norm": 1.0260133743286133,
      "learning_rate": 8.750000000000001e-06,
      "loss": 0.5169,
      "step": 28
    },
    {
      "epoch": 0.2779552715654952,
      "grad_norm": 0.9326399564743042,
      "learning_rate": 9.0625e-06,
      "loss": 0.5331,
      "step": 29
    },
    {
      "epoch": 0.28753993610223644,
      "grad_norm": 1.008641242980957,
      "learning_rate": 9.375000000000001e-06,
      "loss": 0.5607,
      "step": 30
    },
    {
      "epoch": 0.2971246006389776,
      "grad_norm": 1.1734907627105713,
      "learning_rate": 9.6875e-06,
      "loss": 0.5451,
      "step": 31
    },
    {
      "epoch": 0.30670926517571884,
      "grad_norm": 0.8239313960075378,
      "learning_rate": 1e-05,
      "loss": 0.5468,
      "step": 32
    },
    {
      "epoch": 0.31629392971246006,
      "grad_norm": 0.8218770623207092,
      "learning_rate": 9.999685283773504e-06,
      "loss": 0.5069,
      "step": 33
    },
    {
      "epoch": 0.3258785942492013,
      "grad_norm": 0.7858874797821045,
      "learning_rate": 9.998741174712534e-06,
      "loss": 0.518,
      "step": 34
    },
    {
      "epoch": 0.3354632587859425,
      "grad_norm": 0.8673418760299683,
      "learning_rate": 9.997167791667668e-06,
      "loss": 0.539,
      "step": 35
    },
    {
      "epoch": 0.3450479233226837,
      "grad_norm": 0.8987308740615845,
      "learning_rate": 9.994965332706574e-06,
      "loss": 0.5339,
      "step": 36
    },
    {
      "epoch": 0.3546325878594249,
      "grad_norm": 0.6194221377372742,
      "learning_rate": 9.992134075089085e-06,
      "loss": 0.4994,
      "step": 37
    },
    {
      "epoch": 0.36421725239616615,
      "grad_norm": 0.7056483030319214,
      "learning_rate": 9.98867437523228e-06,
      "loss": 0.5016,
      "step": 38
    },
    {
      "epoch": 0.3738019169329074,
      "grad_norm": 0.7977483868598938,
      "learning_rate": 9.984586668665641e-06,
      "loss": 0.5228,
      "step": 39
    },
    {
      "epoch": 0.38338658146964855,
      "grad_norm": 0.8258603811264038,
      "learning_rate": 9.979871469976197e-06,
      "loss": 0.5016,
      "step": 40
    },
    {
      "epoch": 0.3929712460063898,
      "grad_norm": 0.7627078294754028,
      "learning_rate": 9.974529372743762e-06,
      "loss": 0.5001,
      "step": 41
    },
    {
      "epoch": 0.402555910543131,
      "grad_norm": 0.7237797379493713,
      "learning_rate": 9.968561049466214e-06,
      "loss": 0.5044,
      "step": 42
    },
    {
      "epoch": 0.41214057507987223,
      "grad_norm": 0.6851755380630493,
      "learning_rate": 9.961967251474823e-06,
      "loss": 0.522,
      "step": 43
    },
    {
      "epoch": 0.4217252396166134,
      "grad_norm": 0.7777822613716125,
      "learning_rate": 9.954748808839675e-06,
      "loss": 0.4991,
      "step": 44
    },
    {
      "epoch": 0.43130990415335463,
      "grad_norm": 0.7029432654380798,
      "learning_rate": 9.946906630265184e-06,
      "loss": 0.5415,
      "step": 45
    },
    {
      "epoch": 0.44089456869009586,
      "grad_norm": 0.6500408053398132,
      "learning_rate": 9.938441702975689e-06,
      "loss": 0.4806,
      "step": 46
    },
    {
      "epoch": 0.4504792332268371,
      "grad_norm": 0.7370374202728271,
      "learning_rate": 9.92935509259118e-06,
      "loss": 0.5094,
      "step": 47
    },
    {
      "epoch": 0.46006389776357826,
      "grad_norm": 0.6582643389701843,
      "learning_rate": 9.91964794299315e-06,
      "loss": 0.5158,
      "step": 48
    },
    {
      "epoch": 0.4696485623003195,
      "grad_norm": 0.6737906336784363,
      "learning_rate": 9.909321476180594e-06,
      "loss": 0.4944,
      "step": 49
    },
    {
      "epoch": 0.4792332268370607,
      "grad_norm": 0.6686286926269531,
      "learning_rate": 9.898376992116179e-06,
      "loss": 0.4889,
      "step": 50
    },
    {
      "epoch": 0.48881789137380194,
      "grad_norm": 0.689642071723938,
      "learning_rate": 9.886815868562596e-06,
      "loss": 0.518,
      "step": 51
    },
    {
      "epoch": 0.4984025559105431,
      "grad_norm": 0.6608878970146179,
      "learning_rate": 9.874639560909118e-06,
      "loss": 0.499,
      "step": 52
    },
    {
      "epoch": 0.5079872204472844,
      "grad_norm": 0.7523769736289978,
      "learning_rate": 9.861849601988384e-06,
      "loss": 0.5226,
      "step": 53
    },
    {
      "epoch": 0.5175718849840255,
      "grad_norm": 0.6801315546035767,
      "learning_rate": 9.848447601883436e-06,
      "loss": 0.4806,
      "step": 54
    },
    {
      "epoch": 0.5271565495207667,
      "grad_norm": 0.6715684533119202,
      "learning_rate": 9.834435247725032e-06,
      "loss": 0.4882,
      "step": 55
    },
    {
      "epoch": 0.536741214057508,
      "grad_norm": 0.6812077760696411,
      "learning_rate": 9.819814303479268e-06,
      "loss": 0.4752,
      "step": 56
    },
    {
      "epoch": 0.5463258785942492,
      "grad_norm": 0.7278083562850952,
      "learning_rate": 9.804586609725499e-06,
      "loss": 0.4914,
      "step": 57
    },
    {
      "epoch": 0.5559105431309904,
      "grad_norm": 0.7285177707672119,
      "learning_rate": 9.788754083424654e-06,
      "loss": 0.4659,
      "step": 58
    },
    {
      "epoch": 0.5654952076677316,
      "grad_norm": 0.6427510976791382,
      "learning_rate": 9.772318717677905e-06,
      "loss": 0.4961,
      "step": 59
    },
    {
      "epoch": 0.5750798722044729,
      "grad_norm": 0.7198605537414551,
      "learning_rate": 9.755282581475769e-06,
      "loss": 0.4894,
      "step": 60
    },
    {
      "epoch": 0.5846645367412141,
      "grad_norm": 0.7601820230484009,
      "learning_rate": 9.737647819437645e-06,
      "loss": 0.4906,
      "step": 61
    },
    {
      "epoch": 0.5942492012779552,
      "grad_norm": 0.6605767607688904,
      "learning_rate": 9.719416651541839e-06,
      "loss": 0.4854,
      "step": 62
    },
    {
      "epoch": 0.6038338658146964,
      "grad_norm": 0.7106564044952393,
      "learning_rate": 9.700591372846096e-06,
      "loss": 0.4793,
      "step": 63
    },
    {
      "epoch": 0.6134185303514377,
      "grad_norm": 0.684380829334259,
      "learning_rate": 9.681174353198687e-06,
      "loss": 0.4841,
      "step": 64
    },
    {
      "epoch": 0.6230031948881789,
      "grad_norm": 0.6601747274398804,
      "learning_rate": 9.661168036940071e-06,
      "loss": 0.4647,
      "step": 65
    },
    {
      "epoch": 0.6325878594249201,
      "grad_norm": 0.6643561124801636,
      "learning_rate": 9.640574942595195e-06,
      "loss": 0.5038,
      "step": 66
    },
    {
      "epoch": 0.6421725239616614,
      "grad_norm": 0.665458619594574,
      "learning_rate": 9.619397662556434e-06,
      "loss": 0.4716,
      "step": 67
    },
    {
      "epoch": 0.6517571884984026,
      "grad_norm": 0.6601731181144714,
      "learning_rate": 9.597638862757255e-06,
      "loss": 0.4892,
      "step": 68
    },
    {
      "epoch": 0.6613418530351438,
      "grad_norm": 0.7274516224861145,
      "learning_rate": 9.5753012823366e-06,
      "loss": 0.4876,
      "step": 69
    },
    {
      "epoch": 0.670926517571885,
      "grad_norm": 0.6386899948120117,
      "learning_rate": 9.552387733294081e-06,
      "loss": 0.4596,
      "step": 70
    },
    {
      "epoch": 0.6805111821086262,
      "grad_norm": 0.6390331983566284,
      "learning_rate": 9.528901100135971e-06,
      "loss": 0.4667,
      "step": 71
    },
    {
      "epoch": 0.6900958466453674,
      "grad_norm": 0.6199338436126709,
      "learning_rate": 9.504844339512096e-06,
      "loss": 0.4902,
      "step": 72
    },
    {
      "epoch": 0.6996805111821086,
      "grad_norm": 0.6297160387039185,
      "learning_rate": 9.480220479843627e-06,
      "loss": 0.4674,
      "step": 73
    },
    {
      "epoch": 0.7092651757188498,
      "grad_norm": 0.5854744911193848,
      "learning_rate": 9.45503262094184e-06,
      "loss": 0.4738,
      "step": 74
    },
    {
      "epoch": 0.7188498402555911,
      "grad_norm": 0.6352077722549438,
      "learning_rate": 9.4292839336179e-06,
      "loss": 0.4696,
      "step": 75
    },
    {
      "epoch": 0.7284345047923323,
      "grad_norm": 0.6318387389183044,
      "learning_rate": 9.40297765928369e-06,
      "loss": 0.4745,
      "step": 76
    },
    {
      "epoch": 0.7380191693290735,
      "grad_norm": 0.7550281286239624,
      "learning_rate": 9.376117109543769e-06,
      "loss": 0.4694,
      "step": 77
    },
    {
      "epoch": 0.7476038338658147,
      "grad_norm": 0.7698407173156738,
      "learning_rate": 9.348705665778479e-06,
      "loss": 0.4698,
      "step": 78
    },
    {
      "epoch": 0.7571884984025559,
      "grad_norm": 0.6444739103317261,
      "learning_rate": 9.320746778718274e-06,
      "loss": 0.4352,
      "step": 79
    },
    {
      "epoch": 0.7667731629392971,
      "grad_norm": 0.6695520281791687,
      "learning_rate": 9.292243968009332e-06,
      "loss": 0.4778,
      "step": 80
    },
    {
      "epoch": 0.7763578274760383,
      "grad_norm": 0.6107879877090454,
      "learning_rate": 9.263200821770462e-06,
      "loss": 0.4693,
      "step": 81
    },
    {
      "epoch": 0.7859424920127795,
      "grad_norm": 0.583711564540863,
      "learning_rate": 9.233620996141421e-06,
      "loss": 0.467,
      "step": 82
    },
    {
      "epoch": 0.7955271565495208,
      "grad_norm": 0.6243283152580261,
      "learning_rate": 9.203508214822652e-06,
      "loss": 0.4936,
      "step": 83
    },
    {
      "epoch": 0.805111821086262,
      "grad_norm": 0.6546980738639832,
      "learning_rate": 9.172866268606514e-06,
      "loss": 0.4662,
      "step": 84
    },
    {
      "epoch": 0.8146964856230032,
      "grad_norm": 0.6571932435035706,
      "learning_rate": 9.141699014900084e-06,
      "loss": 0.5032,
      "step": 85
    },
    {
      "epoch": 0.8242811501597445,
      "grad_norm": 0.6166942119598389,
      "learning_rate": 9.110010377239552e-06,
      "loss": 0.481,
      "step": 86
    },
    {
      "epoch": 0.8338658146964856,
      "grad_norm": 0.6507984399795532,
      "learning_rate": 9.077804344796302e-06,
      "loss": 0.4515,
      "step": 87
    },
    {
      "epoch": 0.8434504792332268,
      "grad_norm": 0.6561005711555481,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.4935,
      "step": 88
    },
    {
      "epoch": 0.853035143769968,
      "grad_norm": 0.6434285044670105,
      "learning_rate": 9.011856377401891e-06,
      "loss": 0.4832,
      "step": 89
    },
    {
      "epoch": 0.8626198083067093,
      "grad_norm": 0.6693336367607117,
      "learning_rate": 8.978122744408905e-06,
      "loss": 0.4561,
      "step": 90
    },
    {
      "epoch": 0.8722044728434505,
      "grad_norm": 0.6963056325912476,
      "learning_rate": 8.943888319504456e-06,
      "loss": 0.4893,
      "step": 91
    },
    {
      "epoch": 0.8817891373801917,
      "grad_norm": 0.6495336890220642,
      "learning_rate": 8.90915741234015e-06,
      "loss": 0.4851,
      "step": 92
    },
    {
      "epoch": 0.8913738019169329,
      "grad_norm": 0.6429077386856079,
      "learning_rate": 8.873934395068006e-06,
      "loss": 0.4797,
      "step": 93
    },
    {
      "epoch": 0.9009584664536742,
      "grad_norm": 0.7217315435409546,
      "learning_rate": 8.838223701790057e-06,
      "loss": 0.4641,
      "step": 94
    },
    {
      "epoch": 0.9105431309904153,
      "grad_norm": 0.6442226767539978,
      "learning_rate": 8.802029828000157e-06,
      "loss": 0.4663,
      "step": 95
    },
    {
      "epoch": 0.9201277955271565,
      "grad_norm": 0.6835136413574219,
      "learning_rate": 8.765357330018056e-06,
      "loss": 0.5088,
      "step": 96
    },
    {
      "epoch": 0.9297124600638977,
      "grad_norm": 0.672391951084137,
      "learning_rate": 8.728210824415829e-06,
      "loss": 0.4482,
      "step": 97
    },
    {
      "epoch": 0.939297124600639,
      "grad_norm": 0.7429690361022949,
      "learning_rate": 8.690594987436705e-06,
      "loss": 0.4659,
      "step": 98
    },
    {
      "epoch": 0.9488817891373802,
      "grad_norm": 0.6244953274726868,
      "learning_rate": 8.652514554406388e-06,
      "loss": 0.4798,
      "step": 99
    },
    {
      "epoch": 0.9584664536741214,
      "grad_norm": 0.606186032295227,
      "learning_rate": 8.613974319136959e-06,
      "loss": 0.4733,
      "step": 100
    },
    {
      "epoch": 0.9680511182108626,
      "grad_norm": 0.6192579865455627,
      "learning_rate": 8.574979133323378e-06,
      "loss": 0.4735,
      "step": 101
    },
    {
      "epoch": 0.9776357827476039,
      "grad_norm": 0.6590493321418762,
      "learning_rate": 8.535533905932739e-06,
      "loss": 0.4616,
      "step": 102
    },
    {
      "epoch": 0.987220447284345,
      "grad_norm": 0.7287415862083435,
      "learning_rate": 8.495643602586287e-06,
      "loss": 0.4889,
      "step": 103
    },
    {
      "epoch": 0.9968051118210862,
      "grad_norm": 0.5435300469398499,
      "learning_rate": 8.455313244934324e-06,
      "loss": 0.4721,
      "step": 104
    },
    {
      "epoch": 1.0063897763578276,
      "grad_norm": 0.6497277021408081,
      "learning_rate": 8.414547910024035e-06,
      "loss": 0.3994,
      "step": 105
    },
    {
      "epoch": 1.0159744408945688,
      "grad_norm": 0.6478316783905029,
      "learning_rate": 8.373352729660373e-06,
      "loss": 0.404,
      "step": 106
    },
    {
      "epoch": 1.0255591054313098,
      "grad_norm": 0.5634056925773621,
      "learning_rate": 8.331732889760021e-06,
      "loss": 0.4008,
      "step": 107
    },
    {
      "epoch": 1.035143769968051,
      "grad_norm": 0.6008002161979675,
      "learning_rate": 8.289693629698564e-06,
      "loss": 0.4069,
      "step": 108
    },
    {
      "epoch": 1.0447284345047922,
      "grad_norm": 0.5883151292800903,
      "learning_rate": 8.247240241650918e-06,
      "loss": 0.3912,
      "step": 109
    },
    {
      "epoch": 1.0543130990415335,
      "grad_norm": 0.6554713249206543,
      "learning_rate": 8.204378069925121e-06,
      "loss": 0.4295,
      "step": 110
    },
    {
      "epoch": 1.0638977635782747,
      "grad_norm": 0.5875764489173889,
      "learning_rate": 8.16111251028955e-06,
      "loss": 0.4055,
      "step": 111
    },
    {
      "epoch": 1.073482428115016,
      "grad_norm": 0.5865110754966736,
      "learning_rate": 8.117449009293668e-06,
      "loss": 0.3787,
      "step": 112
    },
    {
      "epoch": 1.0830670926517572,
      "grad_norm": 0.7027278542518616,
      "learning_rate": 8.073393063582386e-06,
      "loss": 0.4001,
      "step": 113
    },
    {
      "epoch": 1.0926517571884984,
      "grad_norm": 0.5848774313926697,
      "learning_rate": 8.0289502192041e-06,
      "loss": 0.4105,
      "step": 114
    },
    {
      "epoch": 1.1022364217252396,
      "grad_norm": 0.6438045501708984,
      "learning_rate": 7.984126070912519e-06,
      "loss": 0.3861,
      "step": 115
    },
    {
      "epoch": 1.1118210862619808,
      "grad_norm": 0.8243338465690613,
      "learning_rate": 7.938926261462366e-06,
      "loss": 0.4029,
      "step": 116
    },
    {
      "epoch": 1.121405750798722,
      "grad_norm": 0.6301456093788147,
      "learning_rate": 7.89335648089903e-06,
      "loss": 0.422,
      "step": 117
    },
    {
      "epoch": 1.1309904153354633,
      "grad_norm": 0.575528621673584,
      "learning_rate": 7.84742246584226e-06,
      "loss": 0.4258,
      "step": 118
    },
    {
      "epoch": 1.1405750798722045,
      "grad_norm": 0.7289822101593018,
      "learning_rate": 7.801129998764014e-06,
      "loss": 0.3868,
      "step": 119
    },
    {
      "epoch": 1.1501597444089458,
      "grad_norm": 0.6722568273544312,
      "learning_rate": 7.754484907260513e-06,
      "loss": 0.4185,
      "step": 120
    },
    {
      "epoch": 1.159744408945687,
      "grad_norm": 0.6317142844200134,
      "learning_rate": 7.70749306331863e-06,
      "loss": 0.4026,
      "step": 121
    },
    {
      "epoch": 1.1693290734824282,
      "grad_norm": 0.6734579205513,
      "learning_rate": 7.660160382576683e-06,
      "loss": 0.413,
      "step": 122
    },
    {
      "epoch": 1.1789137380191694,
      "grad_norm": 0.6377459764480591,
      "learning_rate": 7.612492823579744e-06,
      "loss": 0.4055,
      "step": 123
    },
    {
      "epoch": 1.1884984025559104,
      "grad_norm": 0.6760308146476746,
      "learning_rate": 7.564496387029532e-06,
      "loss": 0.3767,
      "step": 124
    },
    {
      "epoch": 1.1980830670926517,
      "grad_norm": 0.598756730556488,
      "learning_rate": 7.516177115029002e-06,
      "loss": 0.4244,
      "step": 125
    },
    {
      "epoch": 1.207667731629393,
      "grad_norm": 0.5769426226615906,
      "learning_rate": 7.467541090321735e-06,
      "loss": 0.4159,
      "step": 126
    },
    {
      "epoch": 1.2172523961661341,
      "grad_norm": 0.603039026260376,
      "learning_rate": 7.4185944355261996e-06,
      "loss": 0.4345,
      "step": 127
    },
    {
      "epoch": 1.2268370607028753,
      "grad_norm": 0.6537725925445557,
      "learning_rate": 7.369343312364994e-06,
      "loss": 0.397,
      "step": 128
    },
    {
      "epoch": 1.2364217252396166,
      "grad_norm": 0.593711793422699,
      "learning_rate": 7.319793920889171e-06,
      "loss": 0.4171,
      "step": 129
    },
    {
      "epoch": 1.2460063897763578,
      "grad_norm": 0.5596345663070679,
      "learning_rate": 7.269952498697734e-06,
      "loss": 0.4066,
      "step": 130
    },
    {
      "epoch": 1.255591054313099,
      "grad_norm": 0.56202232837677,
      "learning_rate": 7.219825320152411e-06,
      "loss": 0.378,
      "step": 131
    },
    {
      "epoch": 1.2651757188498403,
      "grad_norm": 0.5652729272842407,
      "learning_rate": 7.169418695587791e-06,
      "loss": 0.4094,
      "step": 132
    },
    {
      "epoch": 1.2747603833865815,
      "grad_norm": 0.665550172328949,
      "learning_rate": 7.118738970516944e-06,
      "loss": 0.4206,
      "step": 133
    },
    {
      "epoch": 1.2843450479233227,
      "grad_norm": 0.5853424668312073,
      "learning_rate": 7.067792524832604e-06,
      "loss": 0.3874,
      "step": 134
    },
    {
      "epoch": 1.293929712460064,
      "grad_norm": 0.6019495725631714,
      "learning_rate": 7.016585772004026e-06,
      "loss": 0.4303,
      "step": 135
    },
    {
      "epoch": 1.3035143769968052,
      "grad_norm": 0.6686872243881226,
      "learning_rate": 6.965125158269619e-06,
      "loss": 0.3832,
      "step": 136
    },
    {
      "epoch": 1.3130990415335464,
      "grad_norm": 0.6023411750793457,
      "learning_rate": 6.913417161825449e-06,
      "loss": 0.3898,
      "step": 137
    },
    {
      "epoch": 1.3226837060702876,
      "grad_norm": 0.5375120639801025,
      "learning_rate": 6.8614682920097265e-06,
      "loss": 0.4135,
      "step": 138
    },
    {
      "epoch": 1.3322683706070286,
      "grad_norm": 0.5534108877182007,
      "learning_rate": 6.809285088483361e-06,
      "loss": 0.3891,
      "step": 139
    },
    {
      "epoch": 1.34185303514377,
      "grad_norm": 0.5958161950111389,
      "learning_rate": 6.7568741204067145e-06,
      "loss": 0.425,
      "step": 140
    },
    {
      "epoch": 1.351437699680511,
      "grad_norm": 0.5792331099510193,
      "learning_rate": 6.704241985612625e-06,
      "loss": 0.3977,
      "step": 141
    },
    {
      "epoch": 1.3610223642172525,
      "grad_norm": 0.5817360877990723,
      "learning_rate": 6.651395309775837e-06,
      "loss": 0.3949,
      "step": 142
    },
    {
      "epoch": 1.3706070287539935,
      "grad_norm": 0.6270766854286194,
      "learning_rate": 6.598340745578908e-06,
      "loss": 0.3939,
      "step": 143
    },
    {
      "epoch": 1.3801916932907348,
      "grad_norm": 0.6821984648704529,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.3845,
      "step": 144
    },
    {
      "epoch": 1.389776357827476,
      "grad_norm": 0.6132534146308899,
      "learning_rate": 6.491634692845781e-06,
      "loss": 0.399,
      "step": 145
    },
    {
      "epoch": 1.3993610223642172,
      "grad_norm": 0.5472438931465149,
      "learning_rate": 6.437996637160086e-06,
      "loss": 0.3786,
      "step": 146
    },
    {
      "epoch": 1.4089456869009584,
      "grad_norm": 0.5496889352798462,
      "learning_rate": 6.384177557124247e-06,
      "loss": 0.3863,
      "step": 147
    },
    {
      "epoch": 1.4185303514376997,
      "grad_norm": 0.6177486777305603,
      "learning_rate": 6.330184227833376e-06,
      "loss": 0.3792,
      "step": 148
    },
    {
      "epoch": 1.428115015974441,
      "grad_norm": 0.5737434029579163,
      "learning_rate": 6.276023446318214e-06,
      "loss": 0.3948,
      "step": 149
    },
    {
      "epoch": 1.4376996805111821,
      "grad_norm": 0.5723017454147339,
      "learning_rate": 6.2217020306894705e-06,
      "loss": 0.4137,
      "step": 150
    },
    {
      "epoch": 1.4472843450479234,
      "grad_norm": 0.6154125332832336,
      "learning_rate": 6.1672268192795285e-06,
      "loss": 0.4154,
      "step": 151
    },
    {
      "epoch": 1.4568690095846646,
      "grad_norm": 0.56871497631073,
      "learning_rate": 6.112604669781572e-06,
      "loss": 0.4095,
      "step": 152
    },
    {
      "epoch": 1.4664536741214058,
      "grad_norm": 0.5961901545524597,
      "learning_rate": 6.057842458386315e-06,
      "loss": 0.402,
      "step": 153
    },
    {
      "epoch": 1.476038338658147,
      "grad_norm": 0.601497232913971,
      "learning_rate": 6.002947078916365e-06,
      "loss": 0.4026,
      "step": 154
    },
    {
      "epoch": 1.4856230031948883,
      "grad_norm": 0.6027733683586121,
      "learning_rate": 5.947925441958393e-06,
      "loss": 0.4125,
      "step": 155
    },
    {
      "epoch": 1.4952076677316293,
      "grad_norm": 0.5610053539276123,
      "learning_rate": 5.892784473993184e-06,
      "loss": 0.3937,
      "step": 156
    },
    {
      "epoch": 1.5047923322683707,
      "grad_norm": 0.5642369985580444,
      "learning_rate": 5.837531116523683e-06,
      "loss": 0.3939,
      "step": 157
    },
    {
      "epoch": 1.5143769968051117,
      "grad_norm": 0.6150848269462585,
      "learning_rate": 5.782172325201155e-06,
      "loss": 0.4254,
      "step": 158
    },
    {
      "epoch": 1.5239616613418532,
      "grad_norm": 0.5497562885284424,
      "learning_rate": 5.726715068949564e-06,
      "loss": 0.4142,
      "step": 159
    },
    {
      "epoch": 1.5335463258785942,
      "grad_norm": 0.5342154502868652,
      "learning_rate": 5.671166329088278e-06,
      "loss": 0.387,
      "step": 160
    },
    {
      "epoch": 1.5431309904153354,
      "grad_norm": 0.551790714263916,
      "learning_rate": 5.615533098453215e-06,
      "loss": 0.4116,
      "step": 161
    },
    {
      "epoch": 1.5527156549520766,
      "grad_norm": 0.5560970306396484,
      "learning_rate": 5.559822380516539e-06,
      "loss": 0.4043,
      "step": 162
    },
    {
      "epoch": 1.5623003194888179,
      "grad_norm": 0.7323923110961914,
      "learning_rate": 5.504041188505022e-06,
      "loss": 0.3892,
      "step": 163
    },
    {
      "epoch": 1.571884984025559,
      "grad_norm": 0.5998652577400208,
      "learning_rate": 5.448196544517168e-06,
      "loss": 0.4068,
      "step": 164
    },
    {
      "epoch": 1.5814696485623003,
      "grad_norm": 0.586337685585022,
      "learning_rate": 5.392295478639226e-06,
      "loss": 0.4059,
      "step": 165
    },
    {
      "epoch": 1.5910543130990416,
      "grad_norm": 0.5749455094337463,
      "learning_rate": 5.336345028060199e-06,
      "loss": 0.3849,
      "step": 166
    },
    {
      "epoch": 1.6006389776357828,
      "grad_norm": 0.5518559217453003,
      "learning_rate": 5.2803522361859596e-06,
      "loss": 0.3861,
      "step": 167
    },
    {
      "epoch": 1.610223642172524,
      "grad_norm": 0.6018110513687134,
      "learning_rate": 5.224324151752575e-06,
      "loss": 0.3985,
      "step": 168
    },
    {
      "epoch": 1.619808306709265,
      "grad_norm": 0.5411421656608582,
      "learning_rate": 5.168267827938971e-06,
      "loss": 0.4005,
      "step": 169
    },
    {
      "epoch": 1.6293929712460065,
      "grad_norm": 0.5747963190078735,
      "learning_rate": 5.112190321479026e-06,
      "loss": 0.4332,
      "step": 170
    },
    {
      "epoch": 1.6389776357827475,
      "grad_norm": 0.6073194146156311,
      "learning_rate": 5.05609869177323e-06,
      "loss": 0.4096,
      "step": 171
    },
    {
      "epoch": 1.648562300319489,
      "grad_norm": 0.5901669263839722,
      "learning_rate": 5e-06,
      "loss": 0.4068,
      "step": 172
    },
    {
      "epoch": 1.65814696485623,
      "grad_norm": 0.6021949648857117,
      "learning_rate": 4.943901308226771e-06,
      "loss": 0.394,
      "step": 173
    },
    {
      "epoch": 1.6677316293929714,
      "grad_norm": 0.5671015977859497,
      "learning_rate": 4.887809678520976e-06,
      "loss": 0.4074,
      "step": 174
    },
    {
      "epoch": 1.6773162939297124,
      "grad_norm": 0.5579338669776917,
      "learning_rate": 4.831732172061032e-06,
      "loss": 0.394,
      "step": 175
    },
    {
      "epoch": 1.6869009584664538,
      "grad_norm": 0.6558188796043396,
      "learning_rate": 4.775675848247427e-06,
      "loss": 0.4043,
      "step": 176
    },
    {
      "epoch": 1.6964856230031948,
      "grad_norm": 0.5909088850021362,
      "learning_rate": 4.719647763814041e-06,
      "loss": 0.3948,
      "step": 177
    },
    {
      "epoch": 1.706070287539936,
      "grad_norm": 0.5313936471939087,
      "learning_rate": 4.663654971939802e-06,
      "loss": 0.4152,
      "step": 178
    },
    {
      "epoch": 1.7156549520766773,
      "grad_norm": 0.5276142358779907,
      "learning_rate": 4.6077045213607765e-06,
      "loss": 0.4234,
      "step": 179
    },
    {
      "epoch": 1.7252396166134185,
      "grad_norm": 0.5799081921577454,
      "learning_rate": 4.551803455482833e-06,
      "loss": 0.3971,
      "step": 180
    },
    {
      "epoch": 1.7348242811501597,
      "grad_norm": 0.5779193043708801,
      "learning_rate": 4.4959588114949785e-06,
      "loss": 0.397,
      "step": 181
    },
    {
      "epoch": 1.744408945686901,
      "grad_norm": 0.5751073360443115,
      "learning_rate": 4.4401776194834615e-06,
      "loss": 0.3825,
      "step": 182
    },
    {
      "epoch": 1.7539936102236422,
      "grad_norm": 0.5396866798400879,
      "learning_rate": 4.384466901546786e-06,
      "loss": 0.413,
      "step": 183
    },
    {
      "epoch": 1.7635782747603834,
      "grad_norm": 0.550475001335144,
      "learning_rate": 4.3288336709117246e-06,
      "loss": 0.3708,
      "step": 184
    },
    {
      "epoch": 1.7731629392971247,
      "grad_norm": 0.5338464975357056,
      "learning_rate": 4.273284931050438e-06,
      "loss": 0.4228,
      "step": 185
    },
    {
      "epoch": 1.7827476038338657,
      "grad_norm": 0.5433242321014404,
      "learning_rate": 4.217827674798845e-06,
      "loss": 0.3989,
      "step": 186
    },
    {
      "epoch": 1.792332268370607,
      "grad_norm": 0.5268736481666565,
      "learning_rate": 4.162468883476319e-06,
      "loss": 0.4248,
      "step": 187
    },
    {
      "epoch": 1.8019169329073481,
      "grad_norm": 0.5313707590103149,
      "learning_rate": 4.107215526006818e-06,
      "loss": 0.409,
      "step": 188
    },
    {
      "epoch": 1.8115015974440896,
      "grad_norm": 0.560043215751648,
      "learning_rate": 4.052074558041608e-06,
      "loss": 0.3908,
      "step": 189
    },
    {
      "epoch": 1.8210862619808306,
      "grad_norm": 0.5267096757888794,
      "learning_rate": 3.997052921083637e-06,
      "loss": 0.3714,
      "step": 190
    },
    {
      "epoch": 1.830670926517572,
      "grad_norm": 0.5708348155021667,
      "learning_rate": 3.9421575416136866e-06,
      "loss": 0.3997,
      "step": 191
    },
    {
      "epoch": 1.840255591054313,
      "grad_norm": 0.5195730328559875,
      "learning_rate": 3.887395330218429e-06,
      "loss": 0.3997,
      "step": 192
    },
    {
      "epoch": 1.8498402555910545,
      "grad_norm": 0.5325154662132263,
      "learning_rate": 3.832773180720475e-06,
      "loss": 0.4064,
      "step": 193
    },
    {
      "epoch": 1.8594249201277955,
      "grad_norm": 0.551060676574707,
      "learning_rate": 3.778297969310529e-06,
      "loss": 0.3995,
      "step": 194
    },
    {
      "epoch": 1.8690095846645367,
      "grad_norm": 0.547404408454895,
      "learning_rate": 3.723976553681787e-06,
      "loss": 0.4069,
      "step": 195
    },
    {
      "epoch": 1.878594249201278,
      "grad_norm": 0.4963686764240265,
      "learning_rate": 3.669815772166625e-06,
      "loss": 0.3995,
      "step": 196
    },
    {
      "epoch": 1.8881789137380192,
      "grad_norm": 0.5169672966003418,
      "learning_rate": 3.6158224428757538e-06,
      "loss": 0.3937,
      "step": 197
    },
    {
      "epoch": 1.8977635782747604,
      "grad_norm": 0.5645796656608582,
      "learning_rate": 3.562003362839914e-06,
      "loss": 0.4011,
      "step": 198
    },
    {
      "epoch": 1.9073482428115016,
      "grad_norm": 0.5136457681655884,
      "learning_rate": 3.50836530715422e-06,
      "loss": 0.3994,
      "step": 199
    },
    {
      "epoch": 1.9169329073482428,
      "grad_norm": 0.527316689491272,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 0.3978,
      "step": 200
    },
    {
      "epoch": 1.926517571884984,
      "grad_norm": 0.5034081935882568,
      "learning_rate": 3.4016592544210937e-06,
      "loss": 0.4142,
      "step": 201
    },
    {
      "epoch": 1.9361022364217253,
      "grad_norm": 0.5851829051971436,
      "learning_rate": 3.3486046902241663e-06,
      "loss": 0.4212,
      "step": 202
    },
    {
      "epoch": 1.9456869009584663,
      "grad_norm": 0.5426456332206726,
      "learning_rate": 3.295758014387375e-06,
      "loss": 0.4047,
      "step": 203
    },
    {
      "epoch": 1.9552715654952078,
      "grad_norm": 0.49820733070373535,
      "learning_rate": 3.2431258795932863e-06,
      "loss": 0.4234,
      "step": 204
    },
    {
      "epoch": 1.9648562300319488,
      "grad_norm": 0.505520761013031,
      "learning_rate": 3.1907149115166403e-06,
      "loss": 0.4206,
      "step": 205
    },
    {
      "epoch": 1.9744408945686902,
      "grad_norm": 0.523756742477417,
      "learning_rate": 3.1385317079902743e-06,
      "loss": 0.368,
      "step": 206
    },
    {
      "epoch": 1.9840255591054312,
      "grad_norm": 0.49615344405174255,
      "learning_rate": 3.0865828381745515e-06,
      "loss": 0.3858,
      "step": 207
    },
    {
      "epoch": 1.9936102236421727,
      "grad_norm": 0.5383814573287964,
      "learning_rate": 3.0348748417303826e-06,
      "loss": 0.4036,
      "step": 208
    },
    {
      "epoch": 2.0031948881789137,
      "grad_norm": 0.5455489754676819,
      "learning_rate": 2.9834142279959754e-06,
      "loss": 0.3721,
      "step": 209
    },
    {
      "epoch": 2.012779552715655,
      "grad_norm": 0.5573099255561829,
      "learning_rate": 2.932207475167398e-06,
      "loss": 0.3451,
      "step": 210
    },
    {
      "epoch": 2.022364217252396,
      "grad_norm": 0.5531795620918274,
      "learning_rate": 2.8812610294830568e-06,
      "loss": 0.3294,
      "step": 211
    },
    {
      "epoch": 2.0319488817891376,
      "grad_norm": 0.49947431683540344,
      "learning_rate": 2.83058130441221e-06,
      "loss": 0.3616,
      "step": 212
    },
    {
      "epoch": 2.0415335463258786,
      "grad_norm": 0.4917317032814026,
      "learning_rate": 2.7801746798475905e-06,
      "loss": 0.3446,
      "step": 213
    },
    {
      "epoch": 2.0511182108626196,
      "grad_norm": 0.4955407381057739,
      "learning_rate": 2.7300475013022666e-06,
      "loss": 0.3423,
      "step": 214
    },
    {
      "epoch": 2.060702875399361,
      "grad_norm": 0.5965845584869385,
      "learning_rate": 2.6802060791108304e-06,
      "loss": 0.3489,
      "step": 215
    },
    {
      "epoch": 2.070287539936102,
      "grad_norm": 0.5219831466674805,
      "learning_rate": 2.6306566876350072e-06,
      "loss": 0.3175,
      "step": 216
    },
    {
      "epoch": 2.0798722044728435,
      "grad_norm": 0.510310709476471,
      "learning_rate": 2.5814055644738013e-06,
      "loss": 0.362,
      "step": 217
    },
    {
      "epoch": 2.0894568690095845,
      "grad_norm": 0.5322299599647522,
      "learning_rate": 2.532458909678266e-06,
      "loss": 0.3329,
      "step": 218
    },
    {
      "epoch": 2.099041533546326,
      "grad_norm": 0.49356314539909363,
      "learning_rate": 2.483822884971e-06,
      "loss": 0.3881,
      "step": 219
    },
    {
      "epoch": 2.108626198083067,
      "grad_norm": 0.49142566323280334,
      "learning_rate": 2.43550361297047e-06,
      "loss": 0.3194,
      "step": 220
    },
    {
      "epoch": 2.1182108626198084,
      "grad_norm": 0.5280053019523621,
      "learning_rate": 2.387507176420256e-06,
      "loss": 0.3397,
      "step": 221
    },
    {
      "epoch": 2.1277955271565494,
      "grad_norm": 0.5028843283653259,
      "learning_rate": 2.339839617423318e-06,
      "loss": 0.3779,
      "step": 222
    },
    {
      "epoch": 2.137380191693291,
      "grad_norm": 0.4654610753059387,
      "learning_rate": 2.2925069366813718e-06,
      "loss": 0.3173,
      "step": 223
    },
    {
      "epoch": 2.146964856230032,
      "grad_norm": 0.5069674253463745,
      "learning_rate": 2.245515092739488e-06,
      "loss": 0.3477,
      "step": 224
    },
    {
      "epoch": 2.1565495207667733,
      "grad_norm": 0.49440547823905945,
      "learning_rate": 2.1988700012359865e-06,
      "loss": 0.3351,
      "step": 225
    },
    {
      "epoch": 2.1661341853035143,
      "grad_norm": 0.48839282989501953,
      "learning_rate": 2.1525775341577404e-06,
      "loss": 0.3481,
      "step": 226
    },
    {
      "epoch": 2.1757188498402558,
      "grad_norm": 0.5006738305091858,
      "learning_rate": 2.1066435191009717e-06,
      "loss": 0.3302,
      "step": 227
    },
    {
      "epoch": 2.1853035143769968,
      "grad_norm": 0.5038663744926453,
      "learning_rate": 2.061073738537635e-06,
      "loss": 0.3726,
      "step": 228
    },
    {
      "epoch": 2.194888178913738,
      "grad_norm": 0.4842613637447357,
      "learning_rate": 2.0158739290874822e-06,
      "loss": 0.3676,
      "step": 229
    },
    {
      "epoch": 2.2044728434504792,
      "grad_norm": 0.46521812677383423,
      "learning_rate": 1.971049780795901e-06,
      "loss": 0.3372,
      "step": 230
    },
    {
      "epoch": 2.2140575079872207,
      "grad_norm": 0.4897405505180359,
      "learning_rate": 1.9266069364176144e-06,
      "loss": 0.3273,
      "step": 231
    },
    {
      "epoch": 2.2236421725239617,
      "grad_norm": 0.4693354666233063,
      "learning_rate": 1.8825509907063328e-06,
      "loss": 0.3431,
      "step": 232
    },
    {
      "epoch": 2.2332268370607027,
      "grad_norm": 0.4614076018333435,
      "learning_rate": 1.838887489710452e-06,
      "loss": 0.3179,
      "step": 233
    },
    {
      "epoch": 2.242811501597444,
      "grad_norm": 0.4754553735256195,
      "learning_rate": 1.7956219300748796e-06,
      "loss": 0.3279,
      "step": 234
    },
    {
      "epoch": 2.252396166134185,
      "grad_norm": 0.5363587737083435,
      "learning_rate": 1.7527597583490825e-06,
      "loss": 0.3275,
      "step": 235
    },
    {
      "epoch": 2.2619808306709266,
      "grad_norm": 0.47843068838119507,
      "learning_rate": 1.7103063703014372e-06,
      "loss": 0.3547,
      "step": 236
    },
    {
      "epoch": 2.2715654952076676,
      "grad_norm": 0.5068241357803345,
      "learning_rate": 1.6682671102399806e-06,
      "loss": 0.3509,
      "step": 237
    },
    {
      "epoch": 2.281150159744409,
      "grad_norm": 0.5344985723495483,
      "learning_rate": 1.6266472703396286e-06,
      "loss": 0.3679,
      "step": 238
    },
    {
      "epoch": 2.29073482428115,
      "grad_norm": 0.5120568871498108,
      "learning_rate": 1.5854520899759656e-06,
      "loss": 0.3418,
      "step": 239
    },
    {
      "epoch": 2.3003194888178915,
      "grad_norm": 0.4886165261268616,
      "learning_rate": 1.544686755065677e-06,
      "loss": 0.344,
      "step": 240
    },
    {
      "epoch": 2.3099041533546325,
      "grad_norm": 0.4517747461795807,
      "learning_rate": 1.5043563974137132e-06,
      "loss": 0.3464,
      "step": 241
    },
    {
      "epoch": 2.319488817891374,
      "grad_norm": 0.47726428508758545,
      "learning_rate": 1.4644660940672628e-06,
      "loss": 0.3485,
      "step": 242
    },
    {
      "epoch": 2.329073482428115,
      "grad_norm": 0.5107493996620178,
      "learning_rate": 1.4250208666766235e-06,
      "loss": 0.3552,
      "step": 243
    },
    {
      "epoch": 2.3386581469648564,
      "grad_norm": 0.4823455810546875,
      "learning_rate": 1.3860256808630429e-06,
      "loss": 0.3247,
      "step": 244
    },
    {
      "epoch": 2.3482428115015974,
      "grad_norm": 0.46558883786201477,
      "learning_rate": 1.3474854455936126e-06,
      "loss": 0.3474,
      "step": 245
    },
    {
      "epoch": 2.357827476038339,
      "grad_norm": 0.46056708693504333,
      "learning_rate": 1.3094050125632973e-06,
      "loss": 0.3644,
      "step": 246
    },
    {
      "epoch": 2.36741214057508,
      "grad_norm": 0.4884093999862671,
      "learning_rate": 1.2717891755841722e-06,
      "loss": 0.3351,
      "step": 247
    },
    {
      "epoch": 2.376996805111821,
      "grad_norm": 0.4626671075820923,
      "learning_rate": 1.234642669981946e-06,
      "loss": 0.3292,
      "step": 248
    },
    {
      "epoch": 2.3865814696485623,
      "grad_norm": 0.4433649182319641,
      "learning_rate": 1.1979701719998454e-06,
      "loss": 0.3256,
      "step": 249
    },
    {
      "epoch": 2.3961661341853033,
      "grad_norm": 0.4477299749851227,
      "learning_rate": 1.1617762982099446e-06,
      "loss": 0.3307,
      "step": 250
    },
    {
      "epoch": 2.405750798722045,
      "grad_norm": 0.47525662183761597,
      "learning_rate": 1.1260656049319957e-06,
      "loss": 0.3394,
      "step": 251
    },
    {
      "epoch": 2.415335463258786,
      "grad_norm": 0.44378718733787537,
      "learning_rate": 1.0908425876598512e-06,
      "loss": 0.346,
      "step": 252
    },
    {
      "epoch": 2.4249201277955272,
      "grad_norm": 0.4790633022785187,
      "learning_rate": 1.0561116804955451e-06,
      "loss": 0.3408,
      "step": 253
    },
    {
      "epoch": 2.4345047923322682,
      "grad_norm": 0.5277284383773804,
      "learning_rate": 1.0218772555910955e-06,
      "loss": 0.3493,
      "step": 254
    },
    {
      "epoch": 2.4440894568690097,
      "grad_norm": 0.4934959411621094,
      "learning_rate": 9.881436225981107e-07,
      "loss": 0.3551,
      "step": 255
    },
    {
      "epoch": 2.4536741214057507,
      "grad_norm": 0.4702779948711395,
      "learning_rate": 9.549150281252633e-07,
      "loss": 0.3277,
      "step": 256
    },
    {
      "epoch": 2.463258785942492,
      "grad_norm": 0.4510821998119354,
      "learning_rate": 9.221956552036992e-07,
      "loss": 0.3291,
      "step": 257
    },
    {
      "epoch": 2.472843450479233,
      "grad_norm": 0.46118900179862976,
      "learning_rate": 8.899896227604509e-07,
      "loss": 0.3445,
      "step": 258
    },
    {
      "epoch": 2.4824281150159746,
      "grad_norm": 0.47450578212738037,
      "learning_rate": 8.58300985099918e-07,
      "loss": 0.3537,
      "step": 259
    },
    {
      "epoch": 2.4920127795527156,
      "grad_norm": 0.46595561504364014,
      "learning_rate": 8.271337313934869e-07,
      "loss": 0.3294,
      "step": 260
    },
    {
      "epoch": 2.501597444089457,
      "grad_norm": 0.4666687250137329,
      "learning_rate": 7.964917851773496e-07,
      "loss": 0.3487,
      "step": 261
    },
    {
      "epoch": 2.511182108626198,
      "grad_norm": 0.4660246670246124,
      "learning_rate": 7.663790038585794e-07,
      "loss": 0.3438,
      "step": 262
    },
    {
      "epoch": 2.520766773162939,
      "grad_norm": 0.4536135494709015,
      "learning_rate": 7.367991782295392e-07,
      "loss": 0.3251,
      "step": 263
    },
    {
      "epoch": 2.5303514376996805,
      "grad_norm": 0.44076839089393616,
      "learning_rate": 7.077560319906696e-07,
      "loss": 0.3295,
      "step": 264
    },
    {
      "epoch": 2.539936102236422,
      "grad_norm": 0.4548497200012207,
      "learning_rate": 6.792532212817271e-07,
      "loss": 0.3497,
      "step": 265
    },
    {
      "epoch": 2.549520766773163,
      "grad_norm": 0.44671207666397095,
      "learning_rate": 6.512943342215234e-07,
      "loss": 0.35,
      "step": 266
    },
    {
      "epoch": 2.559105431309904,
      "grad_norm": 0.4538073241710663,
      "learning_rate": 6.238828904562316e-07,
      "loss": 0.3604,
      "step": 267
    },
    {
      "epoch": 2.5686900958466454,
      "grad_norm": 0.4668172001838684,
      "learning_rate": 5.9702234071631e-07,
      "loss": 0.3341,
      "step": 268
    },
    {
      "epoch": 2.5782747603833864,
      "grad_norm": 0.4516941010951996,
      "learning_rate": 5.707160663821009e-07,
      "loss": 0.3148,
      "step": 269
    },
    {
      "epoch": 2.587859424920128,
      "grad_norm": 0.4613952338695526,
      "learning_rate": 5.449673790581611e-07,
      "loss": 0.3597,
      "step": 270
    },
    {
      "epoch": 2.597444089456869,
      "grad_norm": 0.44554054737091064,
      "learning_rate": 5.197795201563744e-07,
      "loss": 0.3133,
      "step": 271
    },
    {
      "epoch": 2.6070287539936103,
      "grad_norm": 0.46121156215667725,
      "learning_rate": 4.951556604879049e-07,
      "loss": 0.3429,
      "step": 272
    },
    {
      "epoch": 2.6166134185303513,
      "grad_norm": 0.42949506640434265,
      "learning_rate": 4.710988998640298e-07,
      "loss": 0.3393,
      "step": 273
    },
    {
      "epoch": 2.626198083067093,
      "grad_norm": 0.44105613231658936,
      "learning_rate": 4.4761226670592074e-07,
      "loss": 0.3511,
      "step": 274
    },
    {
      "epoch": 2.635782747603834,
      "grad_norm": 0.4923551082611084,
      "learning_rate": 4.2469871766340096e-07,
      "loss": 0.3198,
      "step": 275
    },
    {
      "epoch": 2.6453674121405752,
      "grad_norm": 0.446719229221344,
      "learning_rate": 4.0236113724274716e-07,
      "loss": 0.3277,
      "step": 276
    },
    {
      "epoch": 2.6549520766773163,
      "grad_norm": 0.47062206268310547,
      "learning_rate": 3.8060233744356634e-07,
      "loss": 0.3411,
      "step": 277
    },
    {
      "epoch": 2.6645367412140573,
      "grad_norm": 0.4742945432662964,
      "learning_rate": 3.5942505740480583e-07,
      "loss": 0.3333,
      "step": 278
    },
    {
      "epoch": 2.6741214057507987,
      "grad_norm": 0.42745864391326904,
      "learning_rate": 3.3883196305992906e-07,
      "loss": 0.3182,
      "step": 279
    },
    {
      "epoch": 2.68370607028754,
      "grad_norm": 0.42992714047431946,
      "learning_rate": 3.18825646801314e-07,
      "loss": 0.3404,
      "step": 280
    },
    {
      "epoch": 2.693290734824281,
      "grad_norm": 0.47815775871276855,
      "learning_rate": 2.9940862715390483e-07,
      "loss": 0.3304,
      "step": 281
    },
    {
      "epoch": 2.702875399361022,
      "grad_norm": 0.46748435497283936,
      "learning_rate": 2.8058334845816214e-07,
      "loss": 0.3445,
      "step": 282
    },
    {
      "epoch": 2.7124600638977636,
      "grad_norm": 0.4928760528564453,
      "learning_rate": 2.6235218056235633e-07,
      "loss": 0.3276,
      "step": 283
    },
    {
      "epoch": 2.722044728434505,
      "grad_norm": 0.45993271470069885,
      "learning_rate": 2.447174185242324e-07,
      "loss": 0.3232,
      "step": 284
    },
    {
      "epoch": 2.731629392971246,
      "grad_norm": 0.4380206763744354,
      "learning_rate": 2.276812823220964e-07,
      "loss": 0.3475,
      "step": 285
    },
    {
      "epoch": 2.741214057507987,
      "grad_norm": 0.4382314383983612,
      "learning_rate": 2.1124591657534776e-07,
      "loss": 0.3249,
      "step": 286
    },
    {
      "epoch": 2.7507987220447285,
      "grad_norm": 0.47202467918395996,
      "learning_rate": 1.9541339027450256e-07,
      "loss": 0.3558,
      "step": 287
    },
    {
      "epoch": 2.7603833865814695,
      "grad_norm": 0.43734198808670044,
      "learning_rate": 1.801856965207338e-07,
      "loss": 0.3324,
      "step": 288
    },
    {
      "epoch": 2.769968051118211,
      "grad_norm": 0.4321967363357544,
      "learning_rate": 1.6556475227496816e-07,
      "loss": 0.3542,
      "step": 289
    },
    {
      "epoch": 2.779552715654952,
      "grad_norm": 0.4536594748497009,
      "learning_rate": 1.5155239811656562e-07,
      "loss": 0.3242,
      "step": 290
    },
    {
      "epoch": 2.7891373801916934,
      "grad_norm": 0.42880719900131226,
      "learning_rate": 1.3815039801161723e-07,
      "loss": 0.3688,
      "step": 291
    },
    {
      "epoch": 2.7987220447284344,
      "grad_norm": 0.44557759165763855,
      "learning_rate": 1.253604390908819e-07,
      "loss": 0.334,
      "step": 292
    },
    {
      "epoch": 2.8083067092651754,
      "grad_norm": 0.43500256538391113,
      "learning_rate": 1.1318413143740436e-07,
      "loss": 0.3411,
      "step": 293
    },
    {
      "epoch": 2.817891373801917,
      "grad_norm": 0.45296376943588257,
      "learning_rate": 1.0162300788382263e-07,
      "loss": 0.3588,
      "step": 294
    },
    {
      "epoch": 2.8274760383386583,
      "grad_norm": 0.4909366965293884,
      "learning_rate": 9.0678523819408e-08,
      "loss": 0.3162,
      "step": 295
    },
    {
      "epoch": 2.8370607028753994,
      "grad_norm": 0.46786975860595703,
      "learning_rate": 8.035205700685167e-08,
      "loss": 0.3454,
      "step": 296
    },
    {
      "epoch": 2.8466453674121404,
      "grad_norm": 0.46253442764282227,
      "learning_rate": 7.064490740882057e-08,
      "loss": 0.346,
      "step": 297
    },
    {
      "epoch": 2.856230031948882,
      "grad_norm": 0.46637165546417236,
      "learning_rate": 6.15582970243117e-08,
      "loss": 0.3368,
      "step": 298
    },
    {
      "epoch": 2.8658146964856233,
      "grad_norm": 0.41861382126808167,
      "learning_rate": 5.3093369734816824e-08,
      "loss": 0.3043,
      "step": 299
    },
    {
      "epoch": 2.8753993610223643,
      "grad_norm": 0.4365438222885132,
      "learning_rate": 4.52511911603265e-08,
      "loss": 0.3222,
      "step": 300
    },
    {
      "epoch": 2.8849840255591053,
      "grad_norm": 0.4770229160785675,
      "learning_rate": 3.8032748525179684e-08,
      "loss": 0.3288,
      "step": 301
    },
    {
      "epoch": 2.8945686900958467,
      "grad_norm": 0.4438874423503876,
      "learning_rate": 3.143895053378698e-08,
      "loss": 0.3584,
      "step": 302
    },
    {
      "epoch": 2.9041533546325877,
      "grad_norm": 0.44970646500587463,
      "learning_rate": 2.547062725623828e-08,
      "loss": 0.3581,
      "step": 303
    },
    {
      "epoch": 2.913738019169329,
      "grad_norm": 0.4476597309112549,
      "learning_rate": 2.012853002380466e-08,
      "loss": 0.339,
      "step": 304
    },
    {
      "epoch": 2.92332268370607,
      "grad_norm": 0.43612033128738403,
      "learning_rate": 1.541333133436018e-08,
      "loss": 0.3492,
      "step": 305
    },
    {
      "epoch": 2.9329073482428116,
      "grad_norm": 0.4451361298561096,
      "learning_rate": 1.132562476771959e-08,
      "loss": 0.3398,
      "step": 306
    },
    {
      "epoch": 2.9424920127795526,
      "grad_norm": 0.4753524363040924,
      "learning_rate": 7.865924910916977e-09,
      "loss": 0.3356,
      "step": 307
    },
    {
      "epoch": 2.952076677316294,
      "grad_norm": 0.43977728486061096,
      "learning_rate": 5.034667293427053e-09,
      "loss": 0.3293,
      "step": 308
    },
    {
      "epoch": 2.961661341853035,
      "grad_norm": 0.43779218196868896,
      "learning_rate": 2.8322083323334417e-09,
      "loss": 0.3244,
      "step": 309
    },
    {
      "epoch": 2.9712460063897765,
      "grad_norm": 0.4322354197502136,
      "learning_rate": 1.2588252874673469e-09,
      "loss": 0.3568,
      "step": 310
    },
    {
      "epoch": 2.9808306709265175,
      "grad_norm": 0.44659778475761414,
      "learning_rate": 3.147162264971471e-10,
      "loss": 0.3667,
      "step": 311
    },
    {
      "epoch": 2.9904153354632586,
      "grad_norm": 0.4485153555870056,
      "learning_rate": 0.0,
      "loss": 0.3413,
      "step": 312
    },
    {
      "epoch": 2.9904153354632586,
      "step": 312,
      "total_flos": 2.1907068970603315e+17,
      "train_loss": 0.0,
      "train_runtime": 6.6479,
      "train_samples_per_second": 4512.734,
      "train_steps_per_second": 46.932
    }
  ],
  "logging_steps": 1,
  "max_steps": 312,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 2.1907068970603315e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}