{
  "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": 5.816951093345819,
      "learning_rate": 6.25e-07,
      "loss": 0.8809,
      "step": 1
    },
    {
      "epoch": 0.019169329073482427,
      "grad_norm": 5.964814555570492,
      "learning_rate": 1.25e-06,
      "loss": 0.8476,
      "step": 2
    },
    {
      "epoch": 0.02875399361022364,
      "grad_norm": 5.91437668793362,
      "learning_rate": 1.8750000000000003e-06,
      "loss": 0.8771,
      "step": 3
    },
    {
      "epoch": 0.038338658146964855,
      "grad_norm": 5.7274742814752155,
      "learning_rate": 2.5e-06,
      "loss": 0.8549,
      "step": 4
    },
    {
      "epoch": 0.04792332268370607,
      "grad_norm": 5.324202754304387,
      "learning_rate": 3.125e-06,
      "loss": 0.8394,
      "step": 5
    },
    {
      "epoch": 0.05750798722044728,
      "grad_norm": 4.070433381694975,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.8099,
      "step": 6
    },
    {
      "epoch": 0.0670926517571885,
      "grad_norm": 2.4484114306405496,
      "learning_rate": 4.3750000000000005e-06,
      "loss": 0.7524,
      "step": 7
    },
    {
      "epoch": 0.07667731629392971,
      "grad_norm": 2.2317547400553917,
      "learning_rate": 5e-06,
      "loss": 0.7533,
      "step": 8
    },
    {
      "epoch": 0.08626198083067092,
      "grad_norm": 4.021700030471694,
      "learning_rate": 5.625e-06,
      "loss": 0.7731,
      "step": 9
    },
    {
      "epoch": 0.09584664536741214,
      "grad_norm": 4.242475862537413,
      "learning_rate": 6.25e-06,
      "loss": 0.7965,
      "step": 10
    },
    {
      "epoch": 0.10543130990415335,
      "grad_norm": 4.187322997258876,
      "learning_rate": 6.875e-06,
      "loss": 0.7696,
      "step": 11
    },
    {
      "epoch": 0.11501597444089456,
      "grad_norm": 3.5880075046075364,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.7105,
      "step": 12
    },
    {
      "epoch": 0.12460063897763578,
      "grad_norm": 3.5147425543441346,
      "learning_rate": 8.125000000000001e-06,
      "loss": 0.7182,
      "step": 13
    },
    {
      "epoch": 0.134185303514377,
      "grad_norm": 2.698461280108951,
      "learning_rate": 8.750000000000001e-06,
      "loss": 0.6811,
      "step": 14
    },
    {
      "epoch": 0.14376996805111822,
      "grad_norm": 1.890037499452564,
      "learning_rate": 9.375000000000001e-06,
      "loss": 0.6807,
      "step": 15
    },
    {
      "epoch": 0.15335463258785942,
      "grad_norm": 1.831321882805547,
      "learning_rate": 1e-05,
      "loss": 0.696,
      "step": 16
    },
    {
      "epoch": 0.16293929712460065,
      "grad_norm": 2.061073210204071,
      "learning_rate": 1.0625e-05,
      "loss": 0.6692,
      "step": 17
    },
    {
      "epoch": 0.17252396166134185,
      "grad_norm": 1.8734300870572917,
      "learning_rate": 1.125e-05,
      "loss": 0.6655,
      "step": 18
    },
    {
      "epoch": 0.18210862619808307,
      "grad_norm": 1.3757902351252964,
      "learning_rate": 1.1875e-05,
      "loss": 0.641,
      "step": 19
    },
    {
      "epoch": 0.19169329073482427,
      "grad_norm": 1.3308745862219742,
      "learning_rate": 1.25e-05,
      "loss": 0.6173,
      "step": 20
    },
    {
      "epoch": 0.2012779552715655,
      "grad_norm": 1.3502732169912994,
      "learning_rate": 1.3125e-05,
      "loss": 0.6143,
      "step": 21
    },
    {
      "epoch": 0.2108626198083067,
      "grad_norm": 1.179130020860865,
      "learning_rate": 1.375e-05,
      "loss": 0.6165,
      "step": 22
    },
    {
      "epoch": 0.22044728434504793,
      "grad_norm": 1.1153340921244757,
      "learning_rate": 1.4375e-05,
      "loss": 0.595,
      "step": 23
    },
    {
      "epoch": 0.23003194888178913,
      "grad_norm": 0.93782430241953,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.6081,
      "step": 24
    },
    {
      "epoch": 0.23961661341853036,
      "grad_norm": 1.0101672937842767,
      "learning_rate": 1.5625e-05,
      "loss": 0.5846,
      "step": 25
    },
    {
      "epoch": 0.24920127795527156,
      "grad_norm": 0.8994475569314913,
      "learning_rate": 1.6250000000000002e-05,
      "loss": 0.5652,
      "step": 26
    },
    {
      "epoch": 0.25878594249201275,
      "grad_norm": 0.7372445592319721,
      "learning_rate": 1.6875e-05,
      "loss": 0.569,
      "step": 27
    },
    {
      "epoch": 0.268370607028754,
      "grad_norm": 0.7790129120230198,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 0.5454,
      "step": 28
    },
    {
      "epoch": 0.2779552715654952,
      "grad_norm": 0.79836828642288,
      "learning_rate": 1.8125e-05,
      "loss": 0.5854,
      "step": 29
    },
    {
      "epoch": 0.28753993610223644,
      "grad_norm": 0.797607856474264,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 0.5628,
      "step": 30
    },
    {
      "epoch": 0.2971246006389776,
      "grad_norm": 0.8495331973366639,
      "learning_rate": 1.9375e-05,
      "loss": 0.5589,
      "step": 31
    },
    {
      "epoch": 0.30670926517571884,
      "grad_norm": 0.8939715492547688,
      "learning_rate": 2e-05,
      "loss": 0.588,
      "step": 32
    },
    {
      "epoch": 0.31629392971246006,
      "grad_norm": 0.7945325544236974,
      "learning_rate": 1.9999370567547008e-05,
      "loss": 0.5531,
      "step": 33
    },
    {
      "epoch": 0.3258785942492013,
      "grad_norm": 0.75081365184287,
      "learning_rate": 1.999748234942507e-05,
      "loss": 0.5781,
      "step": 34
    },
    {
      "epoch": 0.3354632587859425,
      "grad_norm": 0.7667797396040631,
      "learning_rate": 1.9994335583335336e-05,
      "loss": 0.5532,
      "step": 35
    },
    {
      "epoch": 0.3450479233226837,
      "grad_norm": 0.6810165168871454,
      "learning_rate": 1.9989930665413148e-05,
      "loss": 0.5592,
      "step": 36
    },
    {
      "epoch": 0.3546325878594249,
      "grad_norm": 0.8020192657808985,
      "learning_rate": 1.998426815017817e-05,
      "loss": 0.5703,
      "step": 37
    },
    {
      "epoch": 0.36421725239616615,
      "grad_norm": 0.765862886699684,
      "learning_rate": 1.997734875046456e-05,
      "loss": 0.5859,
      "step": 38
    },
    {
      "epoch": 0.3738019169329074,
      "grad_norm": 0.6206240716622619,
      "learning_rate": 1.9969173337331283e-05,
      "loss": 0.5084,
      "step": 39
    },
    {
      "epoch": 0.38338658146964855,
      "grad_norm": 0.6800428463570417,
      "learning_rate": 1.9959742939952393e-05,
      "loss": 0.5402,
      "step": 40
    },
    {
      "epoch": 0.3929712460063898,
      "grad_norm": 0.6898204949273792,
      "learning_rate": 1.9949058745487524e-05,
      "loss": 0.5576,
      "step": 41
    },
    {
      "epoch": 0.402555910543131,
      "grad_norm": 0.7038921388379388,
      "learning_rate": 1.9937122098932428e-05,
      "loss": 0.5814,
      "step": 42
    },
    {
      "epoch": 0.41214057507987223,
      "grad_norm": 0.6386669258124261,
      "learning_rate": 1.9923934502949645e-05,
      "loss": 0.5433,
      "step": 43
    },
    {
      "epoch": 0.4217252396166134,
      "grad_norm": 0.8544716563932903,
      "learning_rate": 1.990949761767935e-05,
      "loss": 0.5562,
      "step": 44
    },
    {
      "epoch": 0.43130990415335463,
      "grad_norm": 0.6590987393801254,
      "learning_rate": 1.9893813260530368e-05,
      "loss": 0.5507,
      "step": 45
    },
    {
      "epoch": 0.44089456869009586,
      "grad_norm": 0.7325945250857884,
      "learning_rate": 1.9876883405951378e-05,
      "loss": 0.5472,
      "step": 46
    },
    {
      "epoch": 0.4504792332268371,
      "grad_norm": 0.8060439236010339,
      "learning_rate": 1.985871018518236e-05,
      "loss": 0.5651,
      "step": 47
    },
    {
      "epoch": 0.46006389776357826,
      "grad_norm": 0.6140083254189492,
      "learning_rate": 1.98392958859863e-05,
      "loss": 0.5595,
      "step": 48
    },
    {
      "epoch": 0.4696485623003195,
      "grad_norm": 0.7132583174991013,
      "learning_rate": 1.9818642952361188e-05,
      "loss": 0.5526,
      "step": 49
    },
    {
      "epoch": 0.4792332268370607,
      "grad_norm": 0.59664072981028,
      "learning_rate": 1.9796753984232357e-05,
      "loss": 0.547,
      "step": 50
    },
    {
      "epoch": 0.48881789137380194,
      "grad_norm": 0.6993452708095715,
      "learning_rate": 1.9773631737125192e-05,
      "loss": 0.5361,
      "step": 51
    },
    {
      "epoch": 0.4984025559105431,
      "grad_norm": 0.6473604609615206,
      "learning_rate": 1.9749279121818235e-05,
      "loss": 0.5535,
      "step": 52
    },
    {
      "epoch": 0.5079872204472844,
      "grad_norm": 0.7074480280853533,
      "learning_rate": 1.9723699203976768e-05,
      "loss": 0.4981,
      "step": 53
    },
    {
      "epoch": 0.5175718849840255,
      "grad_norm": 0.7557340666761172,
      "learning_rate": 1.969689520376687e-05,
      "loss": 0.5379,
      "step": 54
    },
    {
      "epoch": 0.5271565495207667,
      "grad_norm": 0.636305070807052,
      "learning_rate": 1.9668870495450064e-05,
      "loss": 0.5481,
      "step": 55
    },
    {
      "epoch": 0.536741214057508,
      "grad_norm": 0.7057386270005085,
      "learning_rate": 1.9639628606958535e-05,
      "loss": 0.5118,
      "step": 56
    },
    {
      "epoch": 0.5463258785942492,
      "grad_norm": 0.5999038558802152,
      "learning_rate": 1.9609173219450998e-05,
      "loss": 0.5532,
      "step": 57
    },
    {
      "epoch": 0.5559105431309904,
      "grad_norm": 0.6446336459611116,
      "learning_rate": 1.9577508166849308e-05,
      "loss": 0.5443,
      "step": 58
    },
    {
      "epoch": 0.5654952076677316,
      "grad_norm": 0.7779704431661995,
      "learning_rate": 1.954463743535581e-05,
      "loss": 0.5282,
      "step": 59
    },
    {
      "epoch": 0.5750798722044729,
      "grad_norm": 0.6196806885678676,
      "learning_rate": 1.9510565162951538e-05,
      "loss": 0.528,
      "step": 60
    },
    {
      "epoch": 0.5846645367412141,
      "grad_norm": 0.7481597405788767,
      "learning_rate": 1.947529563887529e-05,
      "loss": 0.5258,
      "step": 61
    },
    {
      "epoch": 0.5942492012779552,
      "grad_norm": 0.5954399244612109,
      "learning_rate": 1.9438833303083677e-05,
      "loss": 0.5045,
      "step": 62
    },
    {
      "epoch": 0.6038338658146964,
      "grad_norm": 0.773965376234353,
      "learning_rate": 1.940118274569219e-05,
      "loss": 0.5025,
      "step": 63
    },
    {
      "epoch": 0.6134185303514377,
      "grad_norm": 0.5603381146719852,
      "learning_rate": 1.9362348706397374e-05,
      "loss": 0.5024,
      "step": 64
    },
    {
      "epoch": 0.6230031948881789,
      "grad_norm": 0.7685723417664032,
      "learning_rate": 1.9322336073880143e-05,
      "loss": 0.5357,
      "step": 65
    },
    {
      "epoch": 0.6325878594249201,
      "grad_norm": 0.6579285301369759,
      "learning_rate": 1.928114988519039e-05,
      "loss": 0.5126,
      "step": 66
    },
    {
      "epoch": 0.6421725239616614,
      "grad_norm": 0.7627659065606073,
      "learning_rate": 1.9238795325112867e-05,
      "loss": 0.5234,
      "step": 67
    },
    {
      "epoch": 0.6517571884984026,
      "grad_norm": 0.6307353614086453,
      "learning_rate": 1.919527772551451e-05,
      "loss": 0.5188,
      "step": 68
    },
    {
      "epoch": 0.6613418530351438,
      "grad_norm": 0.6854214793729667,
      "learning_rate": 1.91506025646732e-05,
      "loss": 0.509,
      "step": 69
    },
    {
      "epoch": 0.670926517571885,
      "grad_norm": 0.7177977977714395,
      "learning_rate": 1.9104775466588162e-05,
      "loss": 0.5363,
      "step": 70
    },
    {
      "epoch": 0.6805111821086262,
      "grad_norm": 0.5740574361421374,
      "learning_rate": 1.9057802200271943e-05,
      "loss": 0.5269,
      "step": 71
    },
    {
      "epoch": 0.6900958466453674,
      "grad_norm": 0.7278614746672593,
      "learning_rate": 1.900968867902419e-05,
      "loss": 0.5214,
      "step": 72
    },
    {
      "epoch": 0.6996805111821086,
      "grad_norm": 0.6367723242487334,
      "learning_rate": 1.8960440959687254e-05,
      "loss": 0.5559,
      "step": 73
    },
    {
      "epoch": 0.7092651757188498,
      "grad_norm": 0.6483330441964968,
      "learning_rate": 1.891006524188368e-05,
      "loss": 0.5321,
      "step": 74
    },
    {
      "epoch": 0.7188498402555911,
      "grad_norm": 0.6290627842102959,
      "learning_rate": 1.88585678672358e-05,
      "loss": 0.4939,
      "step": 75
    },
    {
      "epoch": 0.7284345047923323,
      "grad_norm": 0.6410854610772762,
      "learning_rate": 1.880595531856738e-05,
      "loss": 0.5319,
      "step": 76
    },
    {
      "epoch": 0.7380191693290735,
      "grad_norm": 0.6731639380913599,
      "learning_rate": 1.8752234219087538e-05,
      "loss": 0.5273,
      "step": 77
    },
    {
      "epoch": 0.7476038338658147,
      "grad_norm": 0.5316769493193355,
      "learning_rate": 1.8697411331556958e-05,
      "loss": 0.518,
      "step": 78
    },
    {
      "epoch": 0.7571884984025559,
      "grad_norm": 0.6062668049547616,
      "learning_rate": 1.864149355743655e-05,
      "loss": 0.5075,
      "step": 79
    },
    {
      "epoch": 0.7667731629392971,
      "grad_norm": 0.6276805556072127,
      "learning_rate": 1.8584487936018663e-05,
      "loss": 0.5058,
      "step": 80
    },
    {
      "epoch": 0.7763578274760383,
      "grad_norm": 0.5471527462104901,
      "learning_rate": 1.8526401643540924e-05,
      "loss": 0.5031,
      "step": 81
    },
    {
      "epoch": 0.7859424920127795,
      "grad_norm": 0.6123087953792351,
      "learning_rate": 1.8467241992282842e-05,
      "loss": 0.5374,
      "step": 82
    },
    {
      "epoch": 0.7955271565495208,
      "grad_norm": 0.6180809301088696,
      "learning_rate": 1.8407016429645305e-05,
      "loss": 0.5352,
      "step": 83
    },
    {
      "epoch": 0.805111821086262,
      "grad_norm": 0.5770189713402363,
      "learning_rate": 1.834573253721303e-05,
      "loss": 0.5001,
      "step": 84
    },
    {
      "epoch": 0.8146964856230032,
      "grad_norm": 0.5611559252392707,
      "learning_rate": 1.8283398029800167e-05,
      "loss": 0.4936,
      "step": 85
    },
    {
      "epoch": 0.8242811501597445,
      "grad_norm": 0.5376330084381403,
      "learning_rate": 1.8220020754479104e-05,
      "loss": 0.5128,
      "step": 86
    },
    {
      "epoch": 0.8338658146964856,
      "grad_norm": 0.5930317746662469,
      "learning_rate": 1.8155608689592604e-05,
      "loss": 0.5195,
      "step": 87
    },
    {
      "epoch": 0.8434504792332268,
      "grad_norm": 0.5582349649774669,
      "learning_rate": 1.8090169943749477e-05,
      "loss": 0.5412,
      "step": 88
    },
    {
      "epoch": 0.853035143769968,
      "grad_norm": 0.5840770892840872,
      "learning_rate": 1.8023712754803783e-05,
      "loss": 0.4931,
      "step": 89
    },
    {
      "epoch": 0.8626198083067093,
      "grad_norm": 0.5770400921032719,
      "learning_rate": 1.795624548881781e-05,
      "loss": 0.5077,
      "step": 90
    },
    {
      "epoch": 0.8722044728434505,
      "grad_norm": 0.47321773400110334,
      "learning_rate": 1.7887776639008912e-05,
      "loss": 0.5271,
      "step": 91
    },
    {
      "epoch": 0.8817891373801917,
      "grad_norm": 0.6014894804474964,
      "learning_rate": 1.78183148246803e-05,
      "loss": 0.5068,
      "step": 92
    },
    {
      "epoch": 0.8913738019169329,
      "grad_norm": 0.5268826943097094,
      "learning_rate": 1.7747868790136012e-05,
      "loss": 0.5062,
      "step": 93
    },
    {
      "epoch": 0.9009584664536742,
      "grad_norm": 0.538258599224071,
      "learning_rate": 1.7676447403580114e-05,
      "loss": 0.5192,
      "step": 94
    },
    {
      "epoch": 0.9105431309904153,
      "grad_norm": 0.5235902995089281,
      "learning_rate": 1.7604059656000313e-05,
      "loss": 0.512,
      "step": 95
    },
    {
      "epoch": 0.9201277955271565,
      "grad_norm": 0.5922248527214893,
      "learning_rate": 1.7530714660036112e-05,
      "loss": 0.5055,
      "step": 96
    },
    {
      "epoch": 0.9297124600638977,
      "grad_norm": 0.5855685386711252,
      "learning_rate": 1.7456421648831658e-05,
      "loss": 0.4984,
      "step": 97
    },
    {
      "epoch": 0.939297124600639,
      "grad_norm": 0.5458918187075826,
      "learning_rate": 1.738118997487341e-05,
      "loss": 0.5076,
      "step": 98
    },
    {
      "epoch": 0.9488817891373802,
      "grad_norm": 0.6515459002851856,
      "learning_rate": 1.7305029108812777e-05,
      "loss": 0.5229,
      "step": 99
    },
    {
      "epoch": 0.9584664536741214,
      "grad_norm": 0.5315082766100536,
      "learning_rate": 1.7227948638273918e-05,
      "loss": 0.5093,
      "step": 100
    },
    {
      "epoch": 0.9680511182108626,
      "grad_norm": 0.7875828584853896,
      "learning_rate": 1.7149958266646756e-05,
      "loss": 0.494,
      "step": 101
    },
    {
      "epoch": 0.9776357827476039,
      "grad_norm": 0.5868831184446074,
      "learning_rate": 1.7071067811865477e-05,
      "loss": 0.4865,
      "step": 102
    },
    {
      "epoch": 0.987220447284345,
      "grad_norm": 0.5730857899752412,
      "learning_rate": 1.6991287205172575e-05,
      "loss": 0.513,
      "step": 103
    },
    {
      "epoch": 0.9968051118210862,
      "grad_norm": 0.7358514287452019,
      "learning_rate": 1.691062648986865e-05,
      "loss": 0.5143,
      "step": 104
    },
    {
      "epoch": 1.0063897763578276,
      "grad_norm": 1.1699812558564358,
      "learning_rate": 1.682909582004807e-05,
      "loss": 0.8838,
      "step": 105
    },
    {
      "epoch": 1.0159744408945688,
      "grad_norm": 0.6902201583829356,
      "learning_rate": 1.6746705459320746e-05,
      "loss": 0.4084,
      "step": 106
    },
    {
      "epoch": 1.0255591054313098,
      "grad_norm": 0.6792794510906274,
      "learning_rate": 1.6663465779520042e-05,
      "loss": 0.436,
      "step": 107
    },
    {
      "epoch": 1.035143769968051,
      "grad_norm": 0.7421226478591257,
      "learning_rate": 1.657938725939713e-05,
      "loss": 0.4414,
      "step": 108
    },
    {
      "epoch": 1.0447284345047922,
      "grad_norm": 0.6881793520210084,
      "learning_rate": 1.6494480483301836e-05,
      "loss": 0.4859,
      "step": 109
    },
    {
      "epoch": 1.0543130990415335,
      "grad_norm": 0.821089177012622,
      "learning_rate": 1.6408756139850243e-05,
      "loss": 0.4273,
      "step": 110
    },
    {
      "epoch": 1.0638977635782747,
      "grad_norm": 0.6928979324587471,
      "learning_rate": 1.63222250205791e-05,
      "loss": 0.4819,
      "step": 111
    },
    {
      "epoch": 1.073482428115016,
      "grad_norm": 0.7022553739631447,
      "learning_rate": 1.6234898018587336e-05,
      "loss": 0.3695,
      "step": 112
    },
    {
      "epoch": 1.0830670926517572,
      "grad_norm": 0.725323982592671,
      "learning_rate": 1.6146786127164773e-05,
      "loss": 0.461,
      "step": 113
    },
    {
      "epoch": 1.0926517571884984,
      "grad_norm": 0.7048555706325064,
      "learning_rate": 1.60579004384082e-05,
      "loss": 0.4494,
      "step": 114
    },
    {
      "epoch": 1.1022364217252396,
      "grad_norm": 0.6640188248580515,
      "learning_rate": 1.5968252141825038e-05,
      "loss": 0.4685,
      "step": 115
    },
    {
      "epoch": 1.1118210862619808,
      "grad_norm": 0.7030304823361989,
      "learning_rate": 1.5877852522924733e-05,
      "loss": 0.4277,
      "step": 116
    },
    {
      "epoch": 1.121405750798722,
      "grad_norm": 0.7123317956342814,
      "learning_rate": 1.578671296179806e-05,
      "loss": 0.4592,
      "step": 117
    },
    {
      "epoch": 1.1309904153354633,
      "grad_norm": 0.6086700980301015,
      "learning_rate": 1.569484493168452e-05,
      "loss": 0.412,
      "step": 118
    },
    {
      "epoch": 1.1405750798722045,
      "grad_norm": 0.6390103576527882,
      "learning_rate": 1.5602259997528028e-05,
      "loss": 0.4797,
      "step": 119
    },
    {
      "epoch": 1.1501597444089458,
      "grad_norm": 0.5469865905883947,
      "learning_rate": 1.5508969814521026e-05,
      "loss": 0.3935,
      "step": 120
    },
    {
      "epoch": 1.159744408945687,
      "grad_norm": 0.5758273790847533,
      "learning_rate": 1.541498612663726e-05,
      "loss": 0.4702,
      "step": 121
    },
    {
      "epoch": 1.1693290734824282,
      "grad_norm": 0.4971600554124392,
      "learning_rate": 1.5320320765153367e-05,
      "loss": 0.416,
      "step": 122
    },
    {
      "epoch": 1.1789137380191694,
      "grad_norm": 0.6214619772407622,
      "learning_rate": 1.5224985647159489e-05,
      "loss": 0.4474,
      "step": 123
    },
    {
      "epoch": 1.1884984025559104,
      "grad_norm": 0.6547899224858453,
      "learning_rate": 1.5128992774059063e-05,
      "loss": 0.4558,
      "step": 124
    },
    {
      "epoch": 1.1980830670926517,
      "grad_norm": 0.5890574227637935,
      "learning_rate": 1.5032354230058004e-05,
      "loss": 0.47,
      "step": 125
    },
    {
      "epoch": 1.207667731629393,
      "grad_norm": 0.7361985894763436,
      "learning_rate": 1.493508218064347e-05,
      "loss": 0.4309,
      "step": 126
    },
    {
      "epoch": 1.2172523961661341,
      "grad_norm": 0.5173156701874413,
      "learning_rate": 1.4837188871052399e-05,
      "loss": 0.4814,
      "step": 127
    },
    {
      "epoch": 1.2268370607028753,
      "grad_norm": 0.6237781339666125,
      "learning_rate": 1.4738686624729987e-05,
      "loss": 0.4185,
      "step": 128
    },
    {
      "epoch": 1.2364217252396166,
      "grad_norm": 0.6273705352740975,
      "learning_rate": 1.4639587841778342e-05,
      "loss": 0.4637,
      "step": 129
    },
    {
      "epoch": 1.2460063897763578,
      "grad_norm": 0.5605628666508281,
      "learning_rate": 1.4539904997395468e-05,
      "loss": 0.4803,
      "step": 130
    },
    {
      "epoch": 1.255591054313099,
      "grad_norm": 0.597522913497888,
      "learning_rate": 1.4439650640304822e-05,
      "loss": 0.454,
      "step": 131
    },
    {
      "epoch": 1.2651757188498403,
      "grad_norm": 0.5715006951258484,
      "learning_rate": 1.4338837391175582e-05,
      "loss": 0.4441,
      "step": 132
    },
    {
      "epoch": 1.2747603833865815,
      "grad_norm": 0.5270300631154146,
      "learning_rate": 1.4237477941033888e-05,
      "loss": 0.4424,
      "step": 133
    },
    {
      "epoch": 1.2843450479233227,
      "grad_norm": 0.6111693532495731,
      "learning_rate": 1.4135585049665207e-05,
      "loss": 0.4688,
      "step": 134
    },
    {
      "epoch": 1.293929712460064,
      "grad_norm": 0.5720421473056696,
      "learning_rate": 1.4033171544008053e-05,
      "loss": 0.47,
      "step": 135
    },
    {
      "epoch": 1.3035143769968052,
      "grad_norm": 0.6142039193499456,
      "learning_rate": 1.3930250316539237e-05,
      "loss": 0.444,
      "step": 136
    },
    {
      "epoch": 1.3130990415335464,
      "grad_norm": 0.5632554602858334,
      "learning_rate": 1.3826834323650899e-05,
      "loss": 0.4447,
      "step": 137
    },
    {
      "epoch": 1.3226837060702876,
      "grad_norm": 0.5306960211119153,
      "learning_rate": 1.3722936584019453e-05,
      "loss": 0.4464,
      "step": 138
    },
    {
      "epoch": 1.3322683706070286,
      "grad_norm": 0.5026216587572911,
      "learning_rate": 1.3618570176966723e-05,
      "loss": 0.4635,
      "step": 139
    },
    {
      "epoch": 1.34185303514377,
      "grad_norm": 0.5048409273626058,
      "learning_rate": 1.3513748240813429e-05,
      "loss": 0.4719,
      "step": 140
    },
    {
      "epoch": 1.351437699680511,
      "grad_norm": 0.5143189141152029,
      "learning_rate": 1.340848397122525e-05,
      "loss": 0.423,
      "step": 141
    },
    {
      "epoch": 1.3610223642172525,
      "grad_norm": 0.4906781147146303,
      "learning_rate": 1.3302790619551673e-05,
      "loss": 0.4658,
      "step": 142
    },
    {
      "epoch": 1.3706070287539935,
      "grad_norm": 0.45394412594825295,
      "learning_rate": 1.3196681491157816e-05,
      "loss": 0.4383,
      "step": 143
    },
    {
      "epoch": 1.3801916932907348,
      "grad_norm": 0.4565249809642115,
      "learning_rate": 1.3090169943749475e-05,
      "loss": 0.4071,
      "step": 144
    },
    {
      "epoch": 1.389776357827476,
      "grad_norm": 0.5436018324080563,
      "learning_rate": 1.2983269385691562e-05,
      "loss": 0.4942,
      "step": 145
    },
    {
      "epoch": 1.3993610223642172,
      "grad_norm": 0.40894998786022985,
      "learning_rate": 1.2875993274320173e-05,
      "loss": 0.3947,
      "step": 146
    },
    {
      "epoch": 1.4089456869009584,
      "grad_norm": 0.5288187343789331,
      "learning_rate": 1.2768355114248493e-05,
      "loss": 0.4712,
      "step": 147
    },
    {
      "epoch": 1.4185303514376997,
      "grad_norm": 0.4768621544922896,
      "learning_rate": 1.2660368455666752e-05,
      "loss": 0.4382,
      "step": 148
    },
    {
      "epoch": 1.428115015974441,
      "grad_norm": 0.431738859820575,
      "learning_rate": 1.2552046892636427e-05,
      "loss": 0.4071,
      "step": 149
    },
    {
      "epoch": 1.4376996805111821,
      "grad_norm": 0.498599086228044,
      "learning_rate": 1.2443404061378941e-05,
      "loss": 0.4755,
      "step": 150
    },
    {
      "epoch": 1.4472843450479234,
      "grad_norm": 0.4988206100753027,
      "learning_rate": 1.2334453638559057e-05,
      "loss": 0.4083,
      "step": 151
    },
    {
      "epoch": 1.4568690095846646,
      "grad_norm": 0.45822495711811395,
      "learning_rate": 1.2225209339563144e-05,
      "loss": 0.485,
      "step": 152
    },
    {
      "epoch": 1.4664536741214058,
      "grad_norm": 0.5023477915233621,
      "learning_rate": 1.211568491677263e-05,
      "loss": 0.4076,
      "step": 153
    },
    {
      "epoch": 1.476038338658147,
      "grad_norm": 0.45113635635588584,
      "learning_rate": 1.200589415783273e-05,
      "loss": 0.4015,
      "step": 154
    },
    {
      "epoch": 1.4856230031948883,
      "grad_norm": 0.5334811995129992,
      "learning_rate": 1.1895850883916786e-05,
      "loss": 0.4759,
      "step": 155
    },
    {
      "epoch": 1.4952076677316293,
      "grad_norm": 0.40436266318440944,
      "learning_rate": 1.1785568947986368e-05,
      "loss": 0.419,
      "step": 156
    },
    {
      "epoch": 1.5047923322683707,
      "grad_norm": 0.4755461357936275,
      "learning_rate": 1.1675062233047365e-05,
      "loss": 0.4221,
      "step": 157
    },
    {
      "epoch": 1.5143769968051117,
      "grad_norm": 0.5300609622813354,
      "learning_rate": 1.156434465040231e-05,
      "loss": 0.4645,
      "step": 158
    },
    {
      "epoch": 1.5239616613418532,
      "grad_norm": 0.40635331262756796,
      "learning_rate": 1.1453430137899129e-05,
      "loss": 0.3713,
      "step": 159
    },
    {
      "epoch": 1.5335463258785942,
      "grad_norm": 0.4735547448374356,
      "learning_rate": 1.1342332658176556e-05,
      "loss": 0.4324,
      "step": 160
    },
    {
      "epoch": 1.5431309904153354,
      "grad_norm": 0.4495839729557387,
      "learning_rate": 1.123106619690643e-05,
      "loss": 0.4791,
      "step": 161
    },
    {
      "epoch": 1.5527156549520766,
      "grad_norm": 0.4173934080478346,
      "learning_rate": 1.1119644761033079e-05,
      "loss": 0.3956,
      "step": 162
    },
    {
      "epoch": 1.5623003194888179,
      "grad_norm": 0.5099574697612899,
      "learning_rate": 1.1008082377010045e-05,
      "loss": 0.502,
      "step": 163
    },
    {
      "epoch": 1.571884984025559,
      "grad_norm": 0.4022180449613417,
      "learning_rate": 1.0896393089034336e-05,
      "loss": 0.3678,
      "step": 164
    },
    {
      "epoch": 1.5814696485623003,
      "grad_norm": 0.46434170894732985,
      "learning_rate": 1.0784590957278452e-05,
      "loss": 0.4521,
      "step": 165
    },
    {
      "epoch": 1.5910543130990416,
      "grad_norm": 0.5222429942261165,
      "learning_rate": 1.0672690056120398e-05,
      "loss": 0.4578,
      "step": 166
    },
    {
      "epoch": 1.6006389776357828,
      "grad_norm": 0.4714562379867376,
      "learning_rate": 1.0560704472371919e-05,
      "loss": 0.4804,
      "step": 167
    },
    {
      "epoch": 1.610223642172524,
      "grad_norm": 0.4257910311693338,
      "learning_rate": 1.044864830350515e-05,
      "loss": 0.3879,
      "step": 168
    },
    {
      "epoch": 1.619808306709265,
      "grad_norm": 0.4931956302496282,
      "learning_rate": 1.0336535655877942e-05,
      "loss": 0.4756,
      "step": 169
    },
    {
      "epoch": 1.6293929712460065,
      "grad_norm": 0.4307049907622935,
      "learning_rate": 1.0224380642958052e-05,
      "loss": 0.3645,
      "step": 170
    },
    {
      "epoch": 1.6389776357827475,
      "grad_norm": 0.5837734683656293,
      "learning_rate": 1.011219738354646e-05,
      "loss": 0.4896,
      "step": 171
    },
    {
      "epoch": 1.648562300319489,
      "grad_norm": 0.4274446462676884,
      "learning_rate": 1e-05,
      "loss": 0.4297,
      "step": 172
    },
    {
      "epoch": 1.65814696485623,
      "grad_norm": 0.47050426719612987,
      "learning_rate": 9.887802616453543e-06,
      "loss": 0.4016,
      "step": 173
    },
    {
      "epoch": 1.6677316293929714,
      "grad_norm": 0.46798902119678215,
      "learning_rate": 9.775619357041952e-06,
      "loss": 0.4468,
      "step": 174
    },
    {
      "epoch": 1.6773162939297124,
      "grad_norm": 0.4231517848150463,
      "learning_rate": 9.663464344122064e-06,
      "loss": 0.4054,
      "step": 175
    },
    {
      "epoch": 1.6869009584664538,
      "grad_norm": 0.5047554859497767,
      "learning_rate": 9.551351696494854e-06,
      "loss": 0.4564,
      "step": 176
    },
    {
      "epoch": 1.6964856230031948,
      "grad_norm": 0.37022389130163336,
      "learning_rate": 9.439295527628083e-06,
      "loss": 0.4373,
      "step": 177
    },
    {
      "epoch": 1.706070287539936,
      "grad_norm": 0.5285912537102634,
      "learning_rate": 9.327309943879604e-06,
      "loss": 0.4722,
      "step": 178
    },
    {
      "epoch": 1.7156549520766773,
      "grad_norm": 0.4785755524774783,
      "learning_rate": 9.215409042721553e-06,
      "loss": 0.426,
      "step": 179
    },
    {
      "epoch": 1.7252396166134185,
      "grad_norm": 0.4506434737215472,
      "learning_rate": 9.103606910965666e-06,
      "loss": 0.4384,
      "step": 180
    },
    {
      "epoch": 1.7348242811501597,
      "grad_norm": 0.45206163197522226,
      "learning_rate": 8.991917622989957e-06,
      "loss": 0.465,
      "step": 181
    },
    {
      "epoch": 1.744408945686901,
      "grad_norm": 0.4264275243649636,
      "learning_rate": 8.880355238966923e-06,
      "loss": 0.4294,
      "step": 182
    },
    {
      "epoch": 1.7539936102236422,
      "grad_norm": 0.4305322367018383,
      "learning_rate": 8.768933803093573e-06,
      "loss": 0.439,
      "step": 183
    },
    {
      "epoch": 1.7635782747603834,
      "grad_norm": 0.444413629694292,
      "learning_rate": 8.657667341823449e-06,
      "loss": 0.4471,
      "step": 184
    },
    {
      "epoch": 1.7731629392971247,
      "grad_norm": 0.44726871944510105,
      "learning_rate": 8.546569862100876e-06,
      "loss": 0.4291,
      "step": 185
    },
    {
      "epoch": 1.7827476038338657,
      "grad_norm": 0.5141706810151837,
      "learning_rate": 8.43565534959769e-06,
      "loss": 0.4411,
      "step": 186
    },
    {
      "epoch": 1.792332268370607,
      "grad_norm": 0.4135951238897178,
      "learning_rate": 8.324937766952638e-06,
      "loss": 0.4325,
      "step": 187
    },
    {
      "epoch": 1.8019169329073481,
      "grad_norm": 0.45368873581779895,
      "learning_rate": 8.214431052013636e-06,
      "loss": 0.4228,
      "step": 188
    },
    {
      "epoch": 1.8115015974440896,
      "grad_norm": 0.491585413229281,
      "learning_rate": 8.104149116083216e-06,
      "loss": 0.4516,
      "step": 189
    },
    {
      "epoch": 1.8210862619808306,
      "grad_norm": 0.40058123705120063,
      "learning_rate": 7.994105842167274e-06,
      "loss": 0.4331,
      "step": 190
    },
    {
      "epoch": 1.830670926517572,
      "grad_norm": 0.4537379960494452,
      "learning_rate": 7.884315083227373e-06,
      "loss": 0.4685,
      "step": 191
    },
    {
      "epoch": 1.840255591054313,
      "grad_norm": 0.4212878067511151,
      "learning_rate": 7.774790660436857e-06,
      "loss": 0.4405,
      "step": 192
    },
    {
      "epoch": 1.8498402555910545,
      "grad_norm": 0.4051053145614115,
      "learning_rate": 7.66554636144095e-06,
      "loss": 0.438,
      "step": 193
    },
    {
      "epoch": 1.8594249201277955,
      "grad_norm": 0.40403658818972926,
      "learning_rate": 7.556595938621058e-06,
      "loss": 0.4153,
      "step": 194
    },
    {
      "epoch": 1.8690095846645367,
      "grad_norm": 0.39919476552418626,
      "learning_rate": 7.447953107363574e-06,
      "loss": 0.4042,
      "step": 195
    },
    {
      "epoch": 1.878594249201278,
      "grad_norm": 0.40390408443700876,
      "learning_rate": 7.33963154433325e-06,
      "loss": 0.4027,
      "step": 196
    },
    {
      "epoch": 1.8881789137380192,
      "grad_norm": 0.39441465040012735,
      "learning_rate": 7.2316448857515076e-06,
      "loss": 0.4435,
      "step": 197
    },
    {
      "epoch": 1.8977635782747604,
      "grad_norm": 0.43435324742018255,
      "learning_rate": 7.124006725679828e-06,
      "loss": 0.4618,
      "step": 198
    },
    {
      "epoch": 1.9073482428115016,
      "grad_norm": 0.4117241007254349,
      "learning_rate": 7.01673061430844e-06,
      "loss": 0.4302,
      "step": 199
    },
    {
      "epoch": 1.9169329073482428,
      "grad_norm": 0.42716318836747486,
      "learning_rate": 6.909830056250527e-06,
      "loss": 0.4766,
      "step": 200
    },
    {
      "epoch": 1.926517571884984,
      "grad_norm": 0.4284737496521204,
      "learning_rate": 6.8033185088421874e-06,
      "loss": 0.455,
      "step": 201
    },
    {
      "epoch": 1.9361022364217253,
      "grad_norm": 0.4709549975113482,
      "learning_rate": 6.697209380448333e-06,
      "loss": 0.4249,
      "step": 202
    },
    {
      "epoch": 1.9456869009584663,
      "grad_norm": 0.38457354298640695,
      "learning_rate": 6.59151602877475e-06,
      "loss": 0.4298,
      "step": 203
    },
    {
      "epoch": 1.9552715654952078,
      "grad_norm": 0.40219036129356933,
      "learning_rate": 6.486251759186573e-06,
      "loss": 0.46,
      "step": 204
    },
    {
      "epoch": 1.9648562300319488,
      "grad_norm": 0.4394898406507064,
      "learning_rate": 6.381429823033281e-06,
      "loss": 0.4507,
      "step": 205
    },
    {
      "epoch": 1.9744408945686902,
      "grad_norm": 0.428010655667797,
      "learning_rate": 6.277063415980549e-06,
      "loss": 0.4289,
      "step": 206
    },
    {
      "epoch": 1.9840255591054312,
      "grad_norm": 0.41752005507808904,
      "learning_rate": 6.173165676349103e-06,
      "loss": 0.4253,
      "step": 207
    },
    {
      "epoch": 1.9936102236421727,
      "grad_norm": 0.4446826344625602,
      "learning_rate": 6.069749683460765e-06,
      "loss": 0.4653,
      "step": 208
    },
    {
      "epoch": 2.0031948881789137,
      "grad_norm": 0.7825903799604366,
      "learning_rate": 5.966828455991951e-06,
      "loss": 0.6218,
      "step": 209
    },
    {
      "epoch": 2.012779552715655,
      "grad_norm": 0.4401481205860557,
      "learning_rate": 5.864414950334796e-06,
      "loss": 0.3884,
      "step": 210
    },
    {
      "epoch": 2.022364217252396,
      "grad_norm": 0.45142190780046537,
      "learning_rate": 5.7625220589661136e-06,
      "loss": 0.392,
      "step": 211
    },
    {
      "epoch": 2.0319488817891376,
      "grad_norm": 0.48929701776811646,
      "learning_rate": 5.66116260882442e-06,
      "loss": 0.3631,
      "step": 212
    },
    {
      "epoch": 2.0415335463258786,
      "grad_norm": 0.6210973844231107,
      "learning_rate": 5.560349359695181e-06,
      "loss": 0.4092,
      "step": 213
    },
    {
      "epoch": 2.0511182108626196,
      "grad_norm": 0.44860747421683334,
      "learning_rate": 5.460095002604533e-06,
      "loss": 0.3753,
      "step": 214
    },
    {
      "epoch": 2.060702875399361,
      "grad_norm": 0.43389253461155286,
      "learning_rate": 5.360412158221661e-06,
      "loss": 0.356,
      "step": 215
    },
    {
      "epoch": 2.070287539936102,
      "grad_norm": 0.5565953093774033,
      "learning_rate": 5.2613133752700145e-06,
      "loss": 0.4177,
      "step": 216
    },
    {
      "epoch": 2.0798722044728435,
      "grad_norm": 0.5439364657270599,
      "learning_rate": 5.1628111289476025e-06,
      "loss": 0.3691,
      "step": 217
    },
    {
      "epoch": 2.0894568690095845,
      "grad_norm": 0.4250244085573423,
      "learning_rate": 5.064917819356532e-06,
      "loss": 0.3823,
      "step": 218
    },
    {
      "epoch": 2.099041533546326,
      "grad_norm": 0.4633828235328361,
      "learning_rate": 4.967645769942e-06,
      "loss": 0.3456,
      "step": 219
    },
    {
      "epoch": 2.108626198083067,
      "grad_norm": 0.4872033142578477,
      "learning_rate": 4.87100722594094e-06,
      "loss": 0.3792,
      "step": 220
    },
    {
      "epoch": 2.1182108626198084,
      "grad_norm": 0.434441186947985,
      "learning_rate": 4.775014352840512e-06,
      "loss": 0.3891,
      "step": 221
    },
    {
      "epoch": 2.1277955271565494,
      "grad_norm": 0.4065479253193355,
      "learning_rate": 4.679679234846636e-06,
      "loss": 0.35,
      "step": 222
    },
    {
      "epoch": 2.137380191693291,
      "grad_norm": 0.4617361261302904,
      "learning_rate": 4.5850138733627435e-06,
      "loss": 0.3809,
      "step": 223
    },
    {
      "epoch": 2.146964856230032,
      "grad_norm": 0.4239909798768424,
      "learning_rate": 4.491030185478976e-06,
      "loss": 0.3707,
      "step": 224
    },
    {
      "epoch": 2.1565495207667733,
      "grad_norm": 0.4556545694731096,
      "learning_rate": 4.397740002471973e-06,
      "loss": 0.4152,
      "step": 225
    },
    {
      "epoch": 2.1661341853035143,
      "grad_norm": 0.39176510868793746,
      "learning_rate": 4.305155068315481e-06,
      "loss": 0.3372,
      "step": 226
    },
    {
      "epoch": 2.1757188498402558,
      "grad_norm": 0.4030244459032741,
      "learning_rate": 4.213287038201943e-06,
      "loss": 0.3655,
      "step": 227
    },
    {
      "epoch": 2.1853035143769968,
      "grad_norm": 0.44687274849513065,
      "learning_rate": 4.12214747707527e-06,
      "loss": 0.406,
      "step": 228
    },
    {
      "epoch": 2.194888178913738,
      "grad_norm": 0.3944935531176264,
      "learning_rate": 4.0317478581749644e-06,
      "loss": 0.372,
      "step": 229
    },
    {
      "epoch": 2.2044728434504792,
      "grad_norm": 0.40549466106843335,
      "learning_rate": 3.942099561591802e-06,
      "loss": 0.4101,
      "step": 230
    },
    {
      "epoch": 2.2140575079872207,
      "grad_norm": 0.33462197396443605,
      "learning_rate": 3.853213872835229e-06,
      "loss": 0.3337,
      "step": 231
    },
    {
      "epoch": 2.2236421725239617,
      "grad_norm": 0.4099961550027212,
      "learning_rate": 3.7651019814126656e-06,
      "loss": 0.4058,
      "step": 232
    },
    {
      "epoch": 2.2332268370607027,
      "grad_norm": 0.4303748197397045,
      "learning_rate": 3.677774979420904e-06,
      "loss": 0.3918,
      "step": 233
    },
    {
      "epoch": 2.242811501597444,
      "grad_norm": 0.35212450130725803,
      "learning_rate": 3.591243860149759e-06,
      "loss": 0.3621,
      "step": 234
    },
    {
      "epoch": 2.252396166134185,
      "grad_norm": 0.35000139197525687,
      "learning_rate": 3.505519516698165e-06,
      "loss": 0.3925,
      "step": 235
    },
    {
      "epoch": 2.2619808306709266,
      "grad_norm": 0.34903321724384345,
      "learning_rate": 3.4206127406028744e-06,
      "loss": 0.3365,
      "step": 236
    },
    {
      "epoch": 2.2715654952076676,
      "grad_norm": 0.3495761955307319,
      "learning_rate": 3.3365342204799613e-06,
      "loss": 0.3639,
      "step": 237
    },
    {
      "epoch": 2.281150159744409,
      "grad_norm": 0.3559973478644341,
      "learning_rate": 3.2532945406792573e-06,
      "loss": 0.3617,
      "step": 238
    },
    {
      "epoch": 2.29073482428115,
      "grad_norm": 0.3830718690572721,
      "learning_rate": 3.1709041799519312e-06,
      "loss": 0.3911,
      "step": 239
    },
    {
      "epoch": 2.3003194888178915,
      "grad_norm": 0.3328630776804002,
      "learning_rate": 3.089373510131354e-06,
      "loss": 0.3543,
      "step": 240
    },
    {
      "epoch": 2.3099041533546325,
      "grad_norm": 0.31154242898607376,
      "learning_rate": 3.0087127948274264e-06,
      "loss": 0.3149,
      "step": 241
    },
    {
      "epoch": 2.319488817891374,
      "grad_norm": 0.3589233869362801,
      "learning_rate": 2.9289321881345257e-06,
      "loss": 0.3974,
      "step": 242
    },
    {
      "epoch": 2.329073482428115,
      "grad_norm": 0.32834847417622853,
      "learning_rate": 2.850041733353247e-06,
      "loss": 0.3367,
      "step": 243
    },
    {
      "epoch": 2.3386581469648564,
      "grad_norm": 0.35226107708331433,
      "learning_rate": 2.7720513617260857e-06,
      "loss": 0.3803,
      "step": 244
    },
    {
      "epoch": 2.3482428115015974,
      "grad_norm": 0.32975966920662375,
      "learning_rate": 2.694970891187225e-06,
      "loss": 0.3815,
      "step": 245
    },
    {
      "epoch": 2.357827476038339,
      "grad_norm": 0.35188866709000055,
      "learning_rate": 2.6188100251265947e-06,
      "loss": 0.3935,
      "step": 246
    },
    {
      "epoch": 2.36741214057508,
      "grad_norm": 0.3180683406057308,
      "learning_rate": 2.5435783511683444e-06,
      "loss": 0.3229,
      "step": 247
    },
    {
      "epoch": 2.376996805111821,
      "grad_norm": 0.3659434112836393,
      "learning_rate": 2.469285339963892e-06,
      "loss": 0.383,
      "step": 248
    },
    {
      "epoch": 2.3865814696485623,
      "grad_norm": 0.3548003669740148,
      "learning_rate": 2.395940343999691e-06,
      "loss": 0.3828,
      "step": 249
    },
    {
      "epoch": 2.3961661341853033,
      "grad_norm": 0.3271621525630648,
      "learning_rate": 2.323552596419889e-06,
      "loss": 0.3815,
      "step": 250
    },
    {
      "epoch": 2.405750798722045,
      "grad_norm": 0.33849807601480136,
      "learning_rate": 2.2521312098639914e-06,
      "loss": 0.3819,
      "step": 251
    },
    {
      "epoch": 2.415335463258786,
      "grad_norm": 0.3558165011110794,
      "learning_rate": 2.1816851753197023e-06,
      "loss": 0.4087,
      "step": 252
    },
    {
      "epoch": 2.4249201277955272,
      "grad_norm": 0.3353146638437276,
      "learning_rate": 2.1122233609910903e-06,
      "loss": 0.3657,
      "step": 253
    },
    {
      "epoch": 2.4345047923322682,
      "grad_norm": 0.33861735298320605,
      "learning_rate": 2.043754511182191e-06,
      "loss": 0.3457,
      "step": 254
    },
    {
      "epoch": 2.4440894568690097,
      "grad_norm": 0.35349259037933695,
      "learning_rate": 1.9762872451962214e-06,
      "loss": 0.3919,
      "step": 255
    },
    {
      "epoch": 2.4536741214057507,
      "grad_norm": 0.3027836405978437,
      "learning_rate": 1.9098300562505266e-06,
      "loss": 0.3691,
      "step": 256
    },
    {
      "epoch": 2.463258785942492,
      "grad_norm": 0.3156869857843323,
      "learning_rate": 1.8443913104073984e-06,
      "loss": 0.364,
      "step": 257
    },
    {
      "epoch": 2.472843450479233,
      "grad_norm": 0.3386183902480921,
      "learning_rate": 1.7799792455209019e-06,
      "loss": 0.3308,
      "step": 258
    },
    {
      "epoch": 2.4824281150159746,
      "grad_norm": 0.3665955325247484,
      "learning_rate": 1.716601970199836e-06,
      "loss": 0.3769,
      "step": 259
    },
    {
      "epoch": 2.4920127795527156,
      "grad_norm": 0.36101096856608017,
      "learning_rate": 1.6542674627869738e-06,
      "loss": 0.4193,
      "step": 260
    },
    {
      "epoch": 2.501597444089457,
      "grad_norm": 0.3385060552291423,
      "learning_rate": 1.5929835703546992e-06,
      "loss": 0.3798,
      "step": 261
    },
    {
      "epoch": 2.511182108626198,
      "grad_norm": 0.3429361880047483,
      "learning_rate": 1.5327580077171589e-06,
      "loss": 0.3632,
      "step": 262
    },
    {
      "epoch": 2.520766773162939,
      "grad_norm": 0.33218156060244974,
      "learning_rate": 1.4735983564590784e-06,
      "loss": 0.3762,
      "step": 263
    },
    {
      "epoch": 2.5303514376996805,
      "grad_norm": 0.33211325486143817,
      "learning_rate": 1.4155120639813392e-06,
      "loss": 0.3675,
      "step": 264
    },
    {
      "epoch": 2.539936102236422,
      "grad_norm": 0.33667239529882725,
      "learning_rate": 1.3585064425634542e-06,
      "loss": 0.3627,
      "step": 265
    },
    {
      "epoch": 2.549520766773163,
      "grad_norm": 0.3520337488712381,
      "learning_rate": 1.3025886684430467e-06,
      "loss": 0.4046,
      "step": 266
    },
    {
      "epoch": 2.559105431309904,
      "grad_norm": 0.3238937013404258,
      "learning_rate": 1.2477657809124632e-06,
      "loss": 0.3789,
      "step": 267
    },
    {
      "epoch": 2.5686900958466454,
      "grad_norm": 0.3268026171586817,
      "learning_rate": 1.19404468143262e-06,
      "loss": 0.3703,
      "step": 268
    },
    {
      "epoch": 2.5782747603833864,
      "grad_norm": 0.3255582279602733,
      "learning_rate": 1.1414321327642019e-06,
      "loss": 0.3394,
      "step": 269
    },
    {
      "epoch": 2.587859424920128,
      "grad_norm": 0.31640120333334965,
      "learning_rate": 1.0899347581163222e-06,
      "loss": 0.3589,
      "step": 270
    },
    {
      "epoch": 2.597444089456869,
      "grad_norm": 0.3426513122706542,
      "learning_rate": 1.0395590403127487e-06,
      "loss": 0.3832,
      "step": 271
    },
    {
      "epoch": 2.6070287539936103,
      "grad_norm": 0.2998344813799427,
      "learning_rate": 9.903113209758098e-07,
      "loss": 0.3646,
      "step": 272
    },
    {
      "epoch": 2.6166134185303513,
      "grad_norm": 0.3271156026842227,
      "learning_rate": 9.421977997280596e-07,
      "loss": 0.3973,
      "step": 273
    },
    {
      "epoch": 2.626198083067093,
      "grad_norm": 0.3276947619903817,
      "learning_rate": 8.952245334118415e-07,
      "loss": 0.3828,
      "step": 274
    },
    {
      "epoch": 2.635782747603834,
      "grad_norm": 0.3208361600065217,
      "learning_rate": 8.493974353268019e-07,
      "loss": 0.3961,
      "step": 275
    },
    {
      "epoch": 2.6453674121405752,
      "grad_norm": 0.3187235157801297,
      "learning_rate": 8.047222744854943e-07,
      "loss": 0.3796,
      "step": 276
    },
    {
      "epoch": 2.6549520766773163,
      "grad_norm": 0.31827291746375636,
      "learning_rate": 7.612046748871327e-07,
      "loss": 0.3777,
      "step": 277
    },
    {
      "epoch": 2.6645367412140573,
      "grad_norm": 0.314592617245762,
      "learning_rate": 7.188501148096117e-07,
      "loss": 0.354,
      "step": 278
    },
    {
      "epoch": 2.6741214057507987,
      "grad_norm": 0.3411145251233779,
      "learning_rate": 6.776639261198581e-07,
      "loss": 0.4123,
      "step": 279
    },
    {
      "epoch": 2.68370607028754,
      "grad_norm": 0.30394694643551917,
      "learning_rate": 6.37651293602628e-07,
      "loss": 0.3542,
      "step": 280
    },
    {
      "epoch": 2.693290734824281,
      "grad_norm": 0.3294079402554833,
      "learning_rate": 5.988172543078097e-07,
      "loss": 0.3847,
      "step": 281
    },
    {
      "epoch": 2.702875399361022,
      "grad_norm": 0.3286750975459561,
      "learning_rate": 5.611666969163243e-07,
      "loss": 0.374,
      "step": 282
    },
    {
      "epoch": 2.7124600638977636,
      "grad_norm": 0.31556283150275327,
      "learning_rate": 5.247043611247127e-07,
      "loss": 0.3634,
      "step": 283
    },
    {
      "epoch": 2.722044728434505,
      "grad_norm": 0.28616841430584555,
      "learning_rate": 4.894348370484648e-07,
      "loss": 0.3363,
      "step": 284
    },
    {
      "epoch": 2.731629392971246,
      "grad_norm": 0.3347741210841466,
      "learning_rate": 4.553625646441928e-07,
      "loss": 0.3938,
      "step": 285
    },
    {
      "epoch": 2.741214057507987,
      "grad_norm": 0.303400322030674,
      "learning_rate": 4.224918331506955e-07,
      "loss": 0.3544,
      "step": 286
    },
    {
      "epoch": 2.7507987220447285,
      "grad_norm": 0.32472397378256507,
      "learning_rate": 3.908267805490051e-07,
      "loss": 0.3973,
      "step": 287
    },
    {
      "epoch": 2.7603833865814695,
      "grad_norm": 0.31173649557105365,
      "learning_rate": 3.603713930414676e-07,
      "loss": 0.3991,
      "step": 288
    },
    {
      "epoch": 2.769968051118211,
      "grad_norm": 0.3137833568204568,
      "learning_rate": 3.3112950454993633e-07,
      "loss": 0.3604,
      "step": 289
    },
    {
      "epoch": 2.779552715654952,
      "grad_norm": 0.30759530236879995,
      "learning_rate": 3.0310479623313125e-07,
      "loss": 0.3654,
      "step": 290
    },
    {
      "epoch": 2.7891373801916934,
      "grad_norm": 0.3181021720539179,
      "learning_rate": 2.7630079602323447e-07,
      "loss": 0.427,
      "step": 291
    },
    {
      "epoch": 2.7987220447284344,
      "grad_norm": 0.29030962682069206,
      "learning_rate": 2.507208781817638e-07,
      "loss": 0.3213,
      "step": 292
    },
    {
      "epoch": 2.8083067092651754,
      "grad_norm": 0.32615431635966197,
      "learning_rate": 2.2636826287480872e-07,
      "loss": 0.372,
      "step": 293
    },
    {
      "epoch": 2.817891373801917,
      "grad_norm": 0.29846601491141916,
      "learning_rate": 2.0324601576764525e-07,
      "loss": 0.3641,
      "step": 294
    },
    {
      "epoch": 2.8274760383386583,
      "grad_norm": 0.34873022597553716,
      "learning_rate": 1.81357047638816e-07,
      "loss": 0.4412,
      "step": 295
    },
    {
      "epoch": 2.8370607028753994,
      "grad_norm": 0.3147497934101791,
      "learning_rate": 1.6070411401370335e-07,
      "loss": 0.3375,
      "step": 296
    },
    {
      "epoch": 2.8466453674121404,
      "grad_norm": 0.3104782796474926,
      "learning_rate": 1.4128981481764115e-07,
      "loss": 0.398,
      "step": 297
    },
    {
      "epoch": 2.856230031948882,
      "grad_norm": 0.3140238921021234,
      "learning_rate": 1.231165940486234e-07,
      "loss": 0.3686,
      "step": 298
    },
    {
      "epoch": 2.8658146964856233,
      "grad_norm": 0.4288142439350452,
      "learning_rate": 1.0618673946963365e-07,
      "loss": 0.4175,
      "step": 299
    },
    {
      "epoch": 2.8753993610223643,
      "grad_norm": 0.2842824480889967,
      "learning_rate": 9.0502382320653e-08,
      "loss": 0.3317,
      "step": 300
    },
    {
      "epoch": 2.8849840255591053,
      "grad_norm": 0.3129493055661085,
      "learning_rate": 7.606549705035937e-08,
      "loss": 0.3893,
      "step": 301
    },
    {
      "epoch": 2.8945686900958467,
      "grad_norm": 0.31085989958792076,
      "learning_rate": 6.287790106757396e-08,
      "loss": 0.411,
      "step": 302
    },
    {
      "epoch": 2.9041533546325877,
      "grad_norm": 0.29788052650610436,
      "learning_rate": 5.094125451247656e-08,
      "loss": 0.3882,
      "step": 303
    },
    {
      "epoch": 2.913738019169329,
      "grad_norm": 0.330785113441627,
      "learning_rate": 4.025706004760932e-08,
      "loss": 0.4154,
      "step": 304
    },
    {
      "epoch": 2.92332268370607,
      "grad_norm": 0.29464975259883275,
      "learning_rate": 3.082666266872036e-08,
      "loss": 0.3293,
      "step": 305
    },
    {
      "epoch": 2.9329073482428116,
      "grad_norm": 0.29689954664312995,
      "learning_rate": 2.265124953543918e-08,
      "loss": 0.3585,
      "step": 306
    },
    {
      "epoch": 2.9424920127795526,
      "grad_norm": 0.3533149363735046,
      "learning_rate": 1.5731849821833955e-08,
      "loss": 0.4593,
      "step": 307
    },
    {
      "epoch": 2.952076677316294,
      "grad_norm": 0.28841598481778036,
      "learning_rate": 1.0069334586854106e-08,
      "loss": 0.3501,
      "step": 308
    },
    {
      "epoch": 2.961661341853035,
      "grad_norm": 0.3018611204905917,
      "learning_rate": 5.664416664666883e-09,
      "loss": 0.3676,
      "step": 309
    },
    {
      "epoch": 2.9712460063897765,
      "grad_norm": 0.3072862218552703,
      "learning_rate": 2.5176505749346937e-09,
      "loss": 0.3948,
      "step": 310
    },
    {
      "epoch": 2.9808306709265175,
      "grad_norm": 0.29271951419638015,
      "learning_rate": 6.294324529942942e-10,
      "loss": 0.3643,
      "step": 311
    },
    {
      "epoch": 2.9904153354632586,
      "grad_norm": 0.30292122661734633,
      "learning_rate": 0.0,
      "loss": 0.3668,
      "step": 312
    },
    {
      "epoch": 2.9904153354632586,
      "step": 312,
      "total_flos": 4.4115474684667494e+17,
      "train_loss": 0.46701200955953354,
      "train_runtime": 9975.6836,
      "train_samples_per_second": 3.006,
      "train_steps_per_second": 0.031
    }
  ],
  "logging_steps": 1.0,
  "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": 4.4115474684667494e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}