diff --git "a/last-checkpoint/trainer_state.json" "b/last-checkpoint/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/last-checkpoint/trainer_state.json"
@@ -0,0 +1,7033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.3260780957039211,
+  "eval_steps": 500,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00032607809570392107,
+      "grad_norm": 0.0851932093501091,
+      "learning_rate": 0.00019999950652018584,
+      "loss": 1.0001,
+      "step": 1
+    },
+    {
+      "epoch": 0.0006521561914078421,
+      "grad_norm": 0.09610063582658768,
+      "learning_rate": 0.0001999980260856137,
+      "loss": 0.8702,
+      "step": 2
+    },
+    {
+      "epoch": 0.0009782342871117633,
+      "grad_norm": 0.14597254991531372,
+      "learning_rate": 0.000199995558710895,
+      "loss": 1.0361,
+      "step": 3
+    },
+    {
+      "epoch": 0.0013043123828156843,
+      "grad_norm": 0.2123030424118042,
+      "learning_rate": 0.00019999210442038162,
+      "loss": 0.9474,
+      "step": 4
+    },
+    {
+      "epoch": 0.0016303904785196055,
+      "grad_norm": 0.2290404736995697,
+      "learning_rate": 0.00019998766324816607,
+      "loss": 0.859,
+      "step": 5
+    },
+    {
+      "epoch": 0.0019564685742235266,
+      "grad_norm": 0.20041389763355255,
+      "learning_rate": 0.0001999822352380809,
+      "loss": 0.8485,
+      "step": 6
+    },
+    {
+      "epoch": 0.0022825466699274476,
+      "grad_norm": 0.32096758484840393,
+      "learning_rate": 0.00019997582044369843,
+      "loss": 0.9402,
+      "step": 7
+    },
+    {
+      "epoch": 0.0026086247656313686,
+      "grad_norm": 0.16999490559101105,
+      "learning_rate": 0.00019996841892833,
+      "loss": 0.8291,
+      "step": 8
+    },
+    {
+      "epoch": 0.00293470286133529,
+      "grad_norm": 0.2701761722564697,
+      "learning_rate": 0.00019996003076502565,
+      "loss": 0.8865,
+      "step": 9
+    },
+    {
+      "epoch": 0.003260780957039211,
+      "grad_norm": 0.18380843102931976,
+      "learning_rate": 0.00019995065603657316,
+      "loss": 0.7624,
+      "step": 10
+    },
+    {
+      "epoch": 0.003586859052743132,
+      "grad_norm": 0.19820477068424225,
+      "learning_rate": 0.0001999402948354973,
+      "loss": 0.6386,
+      "step": 11
+    },
+    {
+      "epoch": 0.003912937148447053,
+      "grad_norm": 0.3971559703350067,
+      "learning_rate": 0.00019992894726405893,
+      "loss": 0.817,
+      "step": 12
+    },
+    {
+      "epoch": 0.004239015244150974,
+      "grad_norm": 0.26180964708328247,
+      "learning_rate": 0.000199916613434254,
+      "loss": 0.7921,
+      "step": 13
+    },
+    {
+      "epoch": 0.004565093339854895,
+      "grad_norm": 0.31556156277656555,
+      "learning_rate": 0.0001999032934678125,
+      "loss": 0.8024,
+      "step": 14
+    },
+    {
+      "epoch": 0.004891171435558816,
+      "grad_norm": 0.22404684126377106,
+      "learning_rate": 0.00019988898749619702,
+      "loss": 0.8024,
+      "step": 15
+    },
+    {
+      "epoch": 0.005217249531262737,
+      "grad_norm": 0.2191840410232544,
+      "learning_rate": 0.00019987369566060176,
+      "loss": 0.7378,
+      "step": 16
+    },
+    {
+      "epoch": 0.005543327626966659,
+      "grad_norm": 0.21089725196361542,
+      "learning_rate": 0.00019985741811195097,
+      "loss": 0.8019,
+      "step": 17
+    },
+    {
+      "epoch": 0.00586940572267058,
+      "grad_norm": 0.2955284118652344,
+      "learning_rate": 0.00019984015501089752,
+      "loss": 0.8887,
+      "step": 18
+    },
+    {
+      "epoch": 0.006195483818374501,
+      "grad_norm": 0.23684082925319672,
+      "learning_rate": 0.0001998219065278212,
+      "loss": 0.817,
+      "step": 19
+    },
+    {
+      "epoch": 0.006521561914078422,
+      "grad_norm": 0.2309175729751587,
+      "learning_rate": 0.00019980267284282717,
+      "loss": 0.7905,
+      "step": 20
+    },
+    {
+      "epoch": 0.006847640009782343,
+      "grad_norm": 0.22210851311683655,
+      "learning_rate": 0.00019978245414574417,
+      "loss": 0.7685,
+      "step": 21
+    },
+    {
+      "epoch": 0.007173718105486264,
+      "grad_norm": 0.18386678397655487,
+      "learning_rate": 0.00019976125063612252,
+      "loss": 0.6951,
+      "step": 22
+    },
+    {
+      "epoch": 0.007499796201190185,
+      "grad_norm": 0.2344907522201538,
+      "learning_rate": 0.00019973906252323238,
+      "loss": 0.8061,
+      "step": 23
+    },
+    {
+      "epoch": 0.007825874296894107,
+      "grad_norm": 0.23887653648853302,
+      "learning_rate": 0.0001997158900260614,
+      "loss": 0.8196,
+      "step": 24
+    },
+    {
+      "epoch": 0.008151952392598028,
+      "grad_norm": 0.21526898443698883,
+      "learning_rate": 0.0001996917333733128,
+      "loss": 0.7566,
+      "step": 25
+    },
+    {
+      "epoch": 0.008478030488301948,
+      "grad_norm": 0.28801462054252625,
+      "learning_rate": 0.00019966659280340297,
+      "loss": 0.8329,
+      "step": 26
+    },
+    {
+      "epoch": 0.00880410858400587,
+      "grad_norm": 0.2439056932926178,
+      "learning_rate": 0.00019964046856445924,
+      "loss": 0.6542,
+      "step": 27
+    },
+    {
+      "epoch": 0.00913018667970979,
+      "grad_norm": 0.33892586827278137,
+      "learning_rate": 0.00019961336091431727,
+      "loss": 0.669,
+      "step": 28
+    },
+    {
+      "epoch": 0.009456264775413711,
+      "grad_norm": 0.37088584899902344,
+      "learning_rate": 0.00019958527012051857,
+      "loss": 0.7722,
+      "step": 29
+    },
+    {
+      "epoch": 0.009782342871117632,
+      "grad_norm": 0.27336689829826355,
+      "learning_rate": 0.00019955619646030802,
+      "loss": 0.6662,
+      "step": 30
+    },
+    {
+      "epoch": 0.010108420966821553,
+      "grad_norm": 0.24305802583694458,
+      "learning_rate": 0.00019952614022063084,
+      "loss": 0.7181,
+      "step": 31
+    },
+    {
+      "epoch": 0.010434499062525474,
+      "grad_norm": 0.2589680552482605,
+      "learning_rate": 0.00019949510169813003,
+      "loss": 0.7044,
+      "step": 32
+    },
+    {
+      "epoch": 0.010760577158229395,
+      "grad_norm": 0.27365782856941223,
+      "learning_rate": 0.00019946308119914323,
+      "loss": 0.7446,
+      "step": 33
+    },
+    {
+      "epoch": 0.011086655253933318,
+      "grad_norm": 0.34046655893325806,
+      "learning_rate": 0.0001994300790396999,
+      "loss": 0.7853,
+      "step": 34
+    },
+    {
+      "epoch": 0.011412733349637239,
+      "grad_norm": 0.19071121513843536,
+      "learning_rate": 0.000199396095545518,
+      "loss": 0.5924,
+      "step": 35
+    },
+    {
+      "epoch": 0.01173881144534116,
+      "grad_norm": 0.36323902010917664,
+      "learning_rate": 0.00019936113105200085,
+      "loss": 0.8158,
+      "step": 36
+    },
+    {
+      "epoch": 0.01206488954104508,
+      "grad_norm": 0.19574415683746338,
+      "learning_rate": 0.00019932518590423394,
+      "loss": 0.5983,
+      "step": 37
+    },
+    {
+      "epoch": 0.012390967636749002,
+      "grad_norm": 0.24282293021678925,
+      "learning_rate": 0.00019928826045698136,
+      "loss": 0.8028,
+      "step": 38
+    },
+    {
+      "epoch": 0.012717045732452923,
+      "grad_norm": 0.30920177698135376,
+      "learning_rate": 0.0001992503550746824,
+      "loss": 0.7311,
+      "step": 39
+    },
+    {
+      "epoch": 0.013043123828156844,
+      "grad_norm": 0.3591444194316864,
+      "learning_rate": 0.0001992114701314478,
+      "loss": 0.7909,
+      "step": 40
+    },
+    {
+      "epoch": 0.013369201923860765,
+      "grad_norm": 0.30144479870796204,
+      "learning_rate": 0.0001991716060110563,
+      "loss": 0.6631,
+      "step": 41
+    },
+    {
+      "epoch": 0.013695280019564686,
+      "grad_norm": 0.25601276755332947,
+      "learning_rate": 0.00019913076310695068,
+      "loss": 0.7587,
+      "step": 42
+    },
+    {
+      "epoch": 0.014021358115268607,
+      "grad_norm": 0.269055038690567,
+      "learning_rate": 0.00019908894182223388,
+      "loss": 0.7726,
+      "step": 43
+    },
+    {
+      "epoch": 0.014347436210972528,
+      "grad_norm": 0.35874077677726746,
+      "learning_rate": 0.00019904614256966512,
+      "loss": 0.7229,
+      "step": 44
+    },
+    {
+      "epoch": 0.014673514306676448,
+      "grad_norm": 0.2554594576358795,
+      "learning_rate": 0.00019900236577165576,
+      "loss": 0.6733,
+      "step": 45
+    },
+    {
+      "epoch": 0.01499959240238037,
+      "grad_norm": 0.20462395250797272,
+      "learning_rate": 0.0001989576118602651,
+      "loss": 0.6097,
+      "step": 46
+    },
+    {
+      "epoch": 0.01532567049808429,
+      "grad_norm": 0.31137871742248535,
+      "learning_rate": 0.00019891188127719618,
+      "loss": 0.7929,
+      "step": 47
+    },
+    {
+      "epoch": 0.015651748593788213,
+      "grad_norm": 0.22744540870189667,
+      "learning_rate": 0.0001988651744737914,
+      "loss": 0.7205,
+      "step": 48
+    },
+    {
+      "epoch": 0.015977826689492134,
+      "grad_norm": 0.23255527019500732,
+      "learning_rate": 0.00019881749191102808,
+      "loss": 0.6804,
+      "step": 49
+    },
+    {
+      "epoch": 0.016303904785196055,
+      "grad_norm": 0.2959457337856293,
+      "learning_rate": 0.00019876883405951377,
+      "loss": 0.8133,
+      "step": 50
+    },
+    {
+      "epoch": 0.016629982880899976,
+      "grad_norm": 0.2753635346889496,
+      "learning_rate": 0.00019871920139948192,
+      "loss": 0.6824,
+      "step": 51
+    },
+    {
+      "epoch": 0.016956060976603897,
+      "grad_norm": 0.23834997415542603,
+      "learning_rate": 0.0001986685944207868,
+      "loss": 0.5754,
+      "step": 52
+    },
+    {
+      "epoch": 0.017282139072307818,
+      "grad_norm": 0.385995477437973,
+      "learning_rate": 0.0001986170136228989,
+      "loss": 0.7432,
+      "step": 53
+    },
+    {
+      "epoch": 0.01760821716801174,
+      "grad_norm": 0.26523032784461975,
+      "learning_rate": 0.00019856445951489982,
+      "loss": 0.6528,
+      "step": 54
+    },
+    {
+      "epoch": 0.01793429526371566,
+      "grad_norm": 0.25099095702171326,
+      "learning_rate": 0.0001985109326154774,
+      "loss": 0.6592,
+      "step": 55
+    },
+    {
+      "epoch": 0.01826037335941958,
+      "grad_norm": 0.2450903058052063,
+      "learning_rate": 0.00019845643345292054,
+      "loss": 0.7029,
+      "step": 56
+    },
+    {
+      "epoch": 0.018586451455123502,
+      "grad_norm": 0.21463742852210999,
+      "learning_rate": 0.00019840096256511398,
+      "loss": 0.6694,
+      "step": 57
+    },
+    {
+      "epoch": 0.018912529550827423,
+      "grad_norm": 0.2570534348487854,
+      "learning_rate": 0.00019834452049953297,
+      "loss": 0.7292,
+      "step": 58
+    },
+    {
+      "epoch": 0.019238607646531344,
+      "grad_norm": 0.263324111700058,
+      "learning_rate": 0.00019828710781323792,
+      "loss": 0.6538,
+      "step": 59
+    },
+    {
+      "epoch": 0.019564685742235265,
+      "grad_norm": 0.33475950360298157,
+      "learning_rate": 0.0001982287250728689,
+      "loss": 0.725,
+      "step": 60
+    },
+    {
+      "epoch": 0.019890763837939186,
+      "grad_norm": 0.30244898796081543,
+      "learning_rate": 0.0001981693728546399,
+      "loss": 0.7368,
+      "step": 61
+    },
+    {
+      "epoch": 0.020216841933643107,
+      "grad_norm": 0.3100570738315582,
+      "learning_rate": 0.0001981090517443334,
+      "loss": 0.7252,
+      "step": 62
+    },
+    {
+      "epoch": 0.020542920029347028,
+      "grad_norm": 0.21203063428401947,
+      "learning_rate": 0.00019804776233729444,
+      "loss": 0.6033,
+      "step": 63
+    },
+    {
+      "epoch": 0.02086899812505095,
+      "grad_norm": 0.3419407904148102,
+      "learning_rate": 0.0001979855052384247,
+      "loss": 0.7105,
+      "step": 64
+    },
+    {
+      "epoch": 0.02119507622075487,
+      "grad_norm": 0.23826122283935547,
+      "learning_rate": 0.00019792228106217658,
+      "loss": 0.6479,
+      "step": 65
+    },
+    {
+      "epoch": 0.02152115431645879,
+      "grad_norm": 0.26379406452178955,
+      "learning_rate": 0.00019785809043254722,
+      "loss": 0.6375,
+      "step": 66
+    },
+    {
+      "epoch": 0.02184723241216271,
+      "grad_norm": 0.27263978123664856,
+      "learning_rate": 0.0001977929339830722,
+      "loss": 0.6962,
+      "step": 67
+    },
+    {
+      "epoch": 0.022173310507866636,
+      "grad_norm": 0.23443619906902313,
+      "learning_rate": 0.00019772681235681936,
+      "loss": 0.6485,
+      "step": 68
+    },
+    {
+      "epoch": 0.022499388603570557,
+      "grad_norm": 0.22951410710811615,
+      "learning_rate": 0.00019765972620638248,
+      "loss": 0.6682,
+      "step": 69
+    },
+    {
+      "epoch": 0.022825466699274478,
+      "grad_norm": 0.27327480912208557,
+      "learning_rate": 0.00019759167619387476,
+      "loss": 0.6075,
+      "step": 70
+    },
+    {
+      "epoch": 0.0231515447949784,
+      "grad_norm": 0.2695823013782501,
+      "learning_rate": 0.00019752266299092236,
+      "loss": 0.6932,
+      "step": 71
+    },
+    {
+      "epoch": 0.02347762289068232,
+      "grad_norm": 0.25518912076950073,
+      "learning_rate": 0.00019745268727865774,
+      "loss": 0.7467,
+      "step": 72
+    },
+    {
+      "epoch": 0.02380370098638624,
+      "grad_norm": 0.20592662692070007,
+      "learning_rate": 0.0001973817497477129,
+      "loss": 0.6205,
+      "step": 73
+    },
+    {
+      "epoch": 0.02412977908209016,
+      "grad_norm": 0.26220276951789856,
+      "learning_rate": 0.00019730985109821266,
+      "loss": 0.6708,
+      "step": 74
+    },
+    {
+      "epoch": 0.024455857177794083,
+      "grad_norm": 0.21277664601802826,
+      "learning_rate": 0.00019723699203976766,
+      "loss": 0.6089,
+      "step": 75
+    },
+    {
+      "epoch": 0.024781935273498003,
+      "grad_norm": 0.24941107630729675,
+      "learning_rate": 0.0001971631732914674,
+      "loss": 0.653,
+      "step": 76
+    },
+    {
+      "epoch": 0.025108013369201924,
+      "grad_norm": 0.2536955177783966,
+      "learning_rate": 0.0001970883955818731,
+      "loss": 0.6835,
+      "step": 77
+    },
+    {
+      "epoch": 0.025434091464905845,
+      "grad_norm": 0.2956002354621887,
+      "learning_rate": 0.0001970126596490106,
+      "loss": 0.7187,
+      "step": 78
+    },
+    {
+      "epoch": 0.025760169560609766,
+      "grad_norm": 0.4782562553882599,
+      "learning_rate": 0.00019693596624036292,
+      "loss": 0.7792,
+      "step": 79
+    },
+    {
+      "epoch": 0.026086247656313687,
+      "grad_norm": 0.2989657521247864,
+      "learning_rate": 0.0001968583161128631,
+      "loss": 0.6652,
+      "step": 80
+    },
+    {
+      "epoch": 0.02641232575201761,
+      "grad_norm": 0.2501319348812103,
+      "learning_rate": 0.00019677971003288655,
+      "loss": 0.6437,
+      "step": 81
+    },
+    {
+      "epoch": 0.02673840384772153,
+      "grad_norm": 0.25894472002983093,
+      "learning_rate": 0.00019670014877624353,
+      "loss": 0.643,
+      "step": 82
+    },
+    {
+      "epoch": 0.02706448194342545,
+      "grad_norm": 0.26122599840164185,
+      "learning_rate": 0.00019661963312817148,
+      "loss": 0.6002,
+      "step": 83
+    },
+    {
+      "epoch": 0.02739056003912937,
+      "grad_norm": 0.30422499775886536,
+      "learning_rate": 0.0001965381638833274,
+      "loss": 0.7274,
+      "step": 84
+    },
+    {
+      "epoch": 0.027716638134833292,
+      "grad_norm": 0.22934012115001678,
+      "learning_rate": 0.00019645574184577982,
+      "loss": 0.6188,
+      "step": 85
+    },
+    {
+      "epoch": 0.028042716230537213,
+      "grad_norm": 0.2155739665031433,
+      "learning_rate": 0.000196372367829001,
+      "loss": 0.6079,
+      "step": 86
+    },
+    {
+      "epoch": 0.028368794326241134,
+      "grad_norm": 0.2555270791053772,
+      "learning_rate": 0.00019628804265585877,
+      "loss": 0.6143,
+      "step": 87
+    },
+    {
+      "epoch": 0.028694872421945055,
+      "grad_norm": 0.2770369350910187,
+      "learning_rate": 0.0001962027671586086,
+      "loss": 0.7382,
+      "step": 88
+    },
+    {
+      "epoch": 0.029020950517648976,
+      "grad_norm": 0.2100551426410675,
+      "learning_rate": 0.0001961165421788852,
+      "loss": 0.609,
+      "step": 89
+    },
+    {
+      "epoch": 0.029347028613352897,
+      "grad_norm": 0.29590874910354614,
+      "learning_rate": 0.0001960293685676943,
+      "loss": 0.7482,
+      "step": 90
+    },
+    {
+      "epoch": 0.029673106709056818,
+      "grad_norm": 0.33272069692611694,
+      "learning_rate": 0.0001959412471854043,
+      "loss": 0.7156,
+      "step": 91
+    },
+    {
+      "epoch": 0.02999918480476074,
+      "grad_norm": 0.2670280337333679,
+      "learning_rate": 0.0001958521789017376,
+      "loss": 0.7449,
+      "step": 92
+    },
+    {
+      "epoch": 0.03032526290046466,
+      "grad_norm": 0.2555425763130188,
+      "learning_rate": 0.00019576216459576222,
+      "loss": 0.7356,
+      "step": 93
+    },
+    {
+      "epoch": 0.03065134099616858,
+      "grad_norm": 0.2829780578613281,
+      "learning_rate": 0.00019567120515588308,
+      "loss": 0.6172,
+      "step": 94
+    },
+    {
+      "epoch": 0.030977419091872502,
+      "grad_norm": 0.24343392252922058,
+      "learning_rate": 0.00019557930147983302,
+      "loss": 0.575,
+      "step": 95
+    },
+    {
+      "epoch": 0.031303497187576426,
+      "grad_norm": 0.27918389439582825,
+      "learning_rate": 0.00019548645447466431,
+      "loss": 0.6214,
+      "step": 96
+    },
+    {
+      "epoch": 0.031629575283280344,
+      "grad_norm": 0.4070862829685211,
+      "learning_rate": 0.00019539266505673938,
+      "loss": 0.7078,
+      "step": 97
+    },
+    {
+      "epoch": 0.03195565337898427,
+      "grad_norm": 0.27973341941833496,
+      "learning_rate": 0.00019529793415172192,
+      "loss": 0.6759,
+      "step": 98
+    },
+    {
+      "epoch": 0.032281731474688186,
+      "grad_norm": 0.3154214322566986,
+      "learning_rate": 0.00019520226269456768,
+      "loss": 0.7701,
+      "step": 99
+    },
+    {
+      "epoch": 0.03260780957039211,
+      "grad_norm": 0.2046702355146408,
+      "learning_rate": 0.00019510565162951537,
+      "loss": 0.5448,
+      "step": 100
+    },
+    {
+      "epoch": 0.03293388766609603,
+      "grad_norm": 0.2343735247850418,
+      "learning_rate": 0.00019500810191007718,
+      "loss": 0.6245,
+      "step": 101
+    },
+    {
+      "epoch": 0.03325996576179995,
+      "grad_norm": 0.2645024061203003,
+      "learning_rate": 0.00019490961449902946,
+      "loss": 0.646,
+      "step": 102
+    },
+    {
+      "epoch": 0.03358604385750387,
+      "grad_norm": 0.28727737069129944,
+      "learning_rate": 0.0001948101903684032,
+      "loss": 0.6788,
+      "step": 103
+    },
+    {
+      "epoch": 0.033912121953207794,
+      "grad_norm": 0.2755887806415558,
+      "learning_rate": 0.00019470983049947444,
+      "loss": 0.6203,
+      "step": 104
+    },
+    {
+      "epoch": 0.03423820004891171,
+      "grad_norm": 0.2064685970544815,
+      "learning_rate": 0.00019460853588275454,
+      "loss": 0.6794,
+      "step": 105
+    },
+    {
+      "epoch": 0.034564278144615636,
+      "grad_norm": 0.2290899008512497,
+      "learning_rate": 0.00019450630751798048,
+      "loss": 0.5644,
+      "step": 106
+    },
+    {
+      "epoch": 0.03489035624031955,
+      "grad_norm": 0.26528871059417725,
+      "learning_rate": 0.000194403146414105,
+      "loss": 0.6283,
+      "step": 107
+    },
+    {
+      "epoch": 0.03521643433602348,
+      "grad_norm": 0.28655117750167847,
+      "learning_rate": 0.00019429905358928646,
+      "loss": 0.7008,
+      "step": 108
+    },
+    {
+      "epoch": 0.0355425124317274,
+      "grad_norm": 0.2962978780269623,
+      "learning_rate": 0.00019419403007087907,
+      "loss": 0.5863,
+      "step": 109
+    },
+    {
+      "epoch": 0.03586859052743132,
+      "grad_norm": 0.2612760365009308,
+      "learning_rate": 0.00019408807689542257,
+      "loss": 0.6873,
+      "step": 110
+    },
+    {
+      "epoch": 0.036194668623135244,
+      "grad_norm": 0.2672303020954132,
+      "learning_rate": 0.00019398119510863197,
+      "loss": 0.6674,
+      "step": 111
+    },
+    {
+      "epoch": 0.03652074671883916,
+      "grad_norm": 0.27370524406433105,
+      "learning_rate": 0.00019387338576538744,
+      "loss": 0.716,
+      "step": 112
+    },
+    {
+      "epoch": 0.036846824814543086,
+      "grad_norm": 0.23076298832893372,
+      "learning_rate": 0.00019376464992972356,
+      "loss": 0.6218,
+      "step": 113
+    },
+    {
+      "epoch": 0.037172902910247003,
+      "grad_norm": 0.31532204151153564,
+      "learning_rate": 0.00019365498867481923,
+      "loss": 0.7019,
+      "step": 114
+    },
+    {
+      "epoch": 0.03749898100595093,
+      "grad_norm": 0.2647252678871155,
+      "learning_rate": 0.00019354440308298675,
+      "loss": 0.7017,
+      "step": 115
+    },
+    {
+      "epoch": 0.037825059101654845,
+      "grad_norm": 0.283130407333374,
+      "learning_rate": 0.00019343289424566122,
+      "loss": 0.6234,
+      "step": 116
+    },
+    {
+      "epoch": 0.03815113719735877,
+      "grad_norm": 0.25351446866989136,
+      "learning_rate": 0.00019332046326338986,
+      "loss": 0.5254,
+      "step": 117
+    },
+    {
+      "epoch": 0.03847721529306269,
+      "grad_norm": 0.3242204785346985,
+      "learning_rate": 0.0001932071112458211,
+      "loss": 0.7043,
+      "step": 118
+    },
+    {
+      "epoch": 0.03880329338876661,
+      "grad_norm": 0.34721121191978455,
+      "learning_rate": 0.00019309283931169356,
+      "loss": 0.7116,
+      "step": 119
+    },
+    {
+      "epoch": 0.03912937148447053,
+      "grad_norm": 0.3425745964050293,
+      "learning_rate": 0.00019297764858882514,
+      "loss": 0.7402,
+      "step": 120
+    },
+    {
+      "epoch": 0.039455449580174454,
+      "grad_norm": 0.2215355932712555,
+      "learning_rate": 0.00019286154021410173,
+      "loss": 0.583,
+      "step": 121
+    },
+    {
+      "epoch": 0.03978152767587837,
+      "grad_norm": 0.2788958251476288,
+      "learning_rate": 0.00019274451533346615,
+      "loss": 0.6151,
+      "step": 122
+    },
+    {
+      "epoch": 0.040107605771582296,
+      "grad_norm": 0.22468440234661102,
+      "learning_rate": 0.00019262657510190666,
+      "loss": 0.5505,
+      "step": 123
+    },
+    {
+      "epoch": 0.04043368386728621,
+      "grad_norm": 0.23359987139701843,
+      "learning_rate": 0.0001925077206834458,
+      "loss": 0.6524,
+      "step": 124
+    },
+    {
+      "epoch": 0.04075976196299014,
+      "grad_norm": 0.2379453182220459,
+      "learning_rate": 0.0001923879532511287,
+      "loss": 0.5616,
+      "step": 125
+    },
+    {
+      "epoch": 0.041085840058694055,
+      "grad_norm": 0.24692785739898682,
+      "learning_rate": 0.0001922672739870115,
+      "loss": 0.6027,
+      "step": 126
+    },
+    {
+      "epoch": 0.04141191815439798,
+      "grad_norm": 0.2335415929555893,
+      "learning_rate": 0.00019214568408214985,
+      "loss": 0.5839,
+      "step": 127
+    },
+    {
+      "epoch": 0.0417379962501019,
+      "grad_norm": 0.34272462129592896,
+      "learning_rate": 0.00019202318473658705,
+      "loss": 0.6932,
+      "step": 128
+    },
+    {
+      "epoch": 0.04206407434580582,
+      "grad_norm": 0.23207567632198334,
+      "learning_rate": 0.00019189977715934213,
+      "loss": 0.6163,
+      "step": 129
+    },
+    {
+      "epoch": 0.04239015244150974,
+      "grad_norm": 0.18426033854484558,
+      "learning_rate": 0.00019177546256839812,
+      "loss": 0.522,
+      "step": 130
+    },
+    {
+      "epoch": 0.04271623053721366,
+      "grad_norm": 0.2792806625366211,
+      "learning_rate": 0.0001916502421906898,
+      "loss": 0.6846,
+      "step": 131
+    },
+    {
+      "epoch": 0.04304230863291758,
+      "grad_norm": 0.2858756184577942,
+      "learning_rate": 0.00019152411726209176,
+      "loss": 0.7003,
+      "step": 132
+    },
+    {
+      "epoch": 0.043368386728621505,
+      "grad_norm": 0.22560660541057587,
+      "learning_rate": 0.00019139708902740613,
+      "loss": 0.5148,
+      "step": 133
+    },
+    {
+      "epoch": 0.04369446482432542,
+      "grad_norm": 0.2962760329246521,
+      "learning_rate": 0.0001912691587403503,
+      "loss": 0.7421,
+      "step": 134
+    },
+    {
+      "epoch": 0.04402054292002935,
+      "grad_norm": 0.23213566839694977,
+      "learning_rate": 0.00019114032766354453,
+      "loss": 0.5798,
+      "step": 135
+    },
+    {
+      "epoch": 0.04434662101573327,
+      "grad_norm": 0.3144775629043579,
+      "learning_rate": 0.00019101059706849957,
+      "loss": 0.6473,
+      "step": 136
+    },
+    {
+      "epoch": 0.04467269911143719,
+      "grad_norm": 0.27858781814575195,
+      "learning_rate": 0.00019087996823560402,
+      "loss": 0.6913,
+      "step": 137
+    },
+    {
+      "epoch": 0.044998777207141114,
+      "grad_norm": 0.24840930104255676,
+      "learning_rate": 0.0001907484424541117,
+      "loss": 0.5928,
+      "step": 138
+    },
+    {
+      "epoch": 0.04532485530284503,
+      "grad_norm": 0.20698444545269012,
+      "learning_rate": 0.00019061602102212898,
+      "loss": 0.5884,
+      "step": 139
+    },
+    {
+      "epoch": 0.045650933398548955,
+      "grad_norm": 0.3082457482814789,
+      "learning_rate": 0.00019048270524660196,
+      "loss": 0.6686,
+      "step": 140
+    },
+    {
+      "epoch": 0.04597701149425287,
+      "grad_norm": 0.3713861107826233,
+      "learning_rate": 0.0001903484964433035,
+      "loss": 0.7604,
+      "step": 141
+    },
+    {
+      "epoch": 0.0463030895899568,
+      "grad_norm": 0.41255638003349304,
+      "learning_rate": 0.00019021339593682028,
+      "loss": 0.698,
+      "step": 142
+    },
+    {
+      "epoch": 0.046629167685660715,
+      "grad_norm": 0.3440454602241516,
+      "learning_rate": 0.00019007740506053983,
+      "loss": 0.5728,
+      "step": 143
+    },
+    {
+      "epoch": 0.04695524578136464,
+      "grad_norm": 0.2812284529209137,
+      "learning_rate": 0.0001899405251566371,
+      "loss": 0.6911,
+      "step": 144
+    },
+    {
+      "epoch": 0.04728132387706856,
+      "grad_norm": 0.29430341720581055,
+      "learning_rate": 0.00018980275757606157,
+      "loss": 0.682,
+      "step": 145
+    },
+    {
+      "epoch": 0.04760740197277248,
+      "grad_norm": 0.39614227414131165,
+      "learning_rate": 0.00018966410367852362,
+      "loss": 0.6881,
+      "step": 146
+    },
+    {
+      "epoch": 0.0479334800684764,
+      "grad_norm": 0.2873314917087555,
+      "learning_rate": 0.00018952456483248119,
+      "loss": 0.6422,
+      "step": 147
+    },
+    {
+      "epoch": 0.04825955816418032,
+      "grad_norm": 0.2976369857788086,
+      "learning_rate": 0.0001893841424151264,
+      "loss": 0.5745,
+      "step": 148
+    },
+    {
+      "epoch": 0.04858563625988424,
+      "grad_norm": 0.3633524477481842,
+      "learning_rate": 0.0001892428378123718,
+      "loss": 0.69,
+      "step": 149
+    },
+    {
+      "epoch": 0.048911714355588165,
+      "grad_norm": 0.275623619556427,
+      "learning_rate": 0.0001891006524188368,
+      "loss": 0.7673,
+      "step": 150
+    },
+    {
+      "epoch": 0.04923779245129208,
+      "grad_norm": 0.26629534363746643,
+      "learning_rate": 0.00018895758763783383,
+      "loss": 0.7725,
+      "step": 151
+    },
+    {
+      "epoch": 0.04956387054699601,
+      "grad_norm": 0.2281135469675064,
+      "learning_rate": 0.00018881364488135448,
+      "loss": 0.6021,
+      "step": 152
+    },
+    {
+      "epoch": 0.049889948642699924,
+      "grad_norm": 0.24734225869178772,
+      "learning_rate": 0.00018866882557005567,
+      "loss": 0.6097,
+      "step": 153
+    },
+    {
+      "epoch": 0.05021602673840385,
+      "grad_norm": 0.2616749405860901,
+      "learning_rate": 0.00018852313113324552,
+      "loss": 0.7009,
+      "step": 154
+    },
+    {
+      "epoch": 0.050542104834107766,
+      "grad_norm": 0.2619054913520813,
+      "learning_rate": 0.00018837656300886937,
+      "loss": 0.6613,
+      "step": 155
+    },
+    {
+      "epoch": 0.05086818292981169,
+      "grad_norm": 0.3681729733943939,
+      "learning_rate": 0.00018822912264349534,
+      "loss": 0.6548,
+      "step": 156
+    },
+    {
+      "epoch": 0.05119426102551561,
+      "grad_norm": 0.2739261984825134,
+      "learning_rate": 0.00018808081149230036,
+      "loss": 0.6548,
+      "step": 157
+    },
+    {
+      "epoch": 0.05152033912121953,
+      "grad_norm": 0.28752514719963074,
+      "learning_rate": 0.00018793163101905563,
+      "loss": 0.6172,
+      "step": 158
+    },
+    {
+      "epoch": 0.05184641721692345,
+      "grad_norm": 0.249834805727005,
+      "learning_rate": 0.00018778158269611218,
+      "loss": 0.6068,
+      "step": 159
+    },
+    {
+      "epoch": 0.052172495312627375,
+      "grad_norm": 0.3328039348125458,
+      "learning_rate": 0.00018763066800438636,
+      "loss": 0.6898,
+      "step": 160
+    },
+    {
+      "epoch": 0.05249857340833129,
+      "grad_norm": 0.3795255422592163,
+      "learning_rate": 0.0001874788884333453,
+      "loss": 0.6397,
+      "step": 161
+    },
+    {
+      "epoch": 0.05282465150403522,
+      "grad_norm": 0.30090928077697754,
+      "learning_rate": 0.00018732624548099204,
+      "loss": 0.6447,
+      "step": 162
+    },
+    {
+      "epoch": 0.053150729599739134,
+      "grad_norm": 0.22852292656898499,
+      "learning_rate": 0.0001871727406538509,
+      "loss": 0.5784,
+      "step": 163
+    },
+    {
+      "epoch": 0.05347680769544306,
+      "grad_norm": 0.24613617360591888,
+      "learning_rate": 0.0001870183754669526,
+      "loss": 0.6584,
+      "step": 164
+    },
+    {
+      "epoch": 0.05380288579114698,
+      "grad_norm": 0.3243715465068817,
+      "learning_rate": 0.00018686315144381913,
+      "loss": 0.6972,
+      "step": 165
+    },
+    {
+      "epoch": 0.0541289638868509,
+      "grad_norm": 0.29082372784614563,
+      "learning_rate": 0.000186707070116449,
+      "loss": 0.5996,
+      "step": 166
+    },
+    {
+      "epoch": 0.054455041982554825,
+      "grad_norm": 0.19261658191680908,
+      "learning_rate": 0.0001865501330253019,
+      "loss": 0.5629,
+      "step": 167
+    },
+    {
+      "epoch": 0.05478112007825874,
+      "grad_norm": 0.3714055120944977,
+      "learning_rate": 0.00018639234171928353,
+      "loss": 0.7309,
+      "step": 168
+    },
+    {
+      "epoch": 0.05510719817396267,
+      "grad_norm": 0.21879787743091583,
+      "learning_rate": 0.0001862336977557304,
+      "loss": 0.5907,
+      "step": 169
+    },
+    {
+      "epoch": 0.055433276269666584,
+      "grad_norm": 0.2829175293445587,
+      "learning_rate": 0.0001860742027003944,
+      "loss": 0.6705,
+      "step": 170
+    },
+    {
+      "epoch": 0.05575935436537051,
+      "grad_norm": 0.2817269563674927,
+      "learning_rate": 0.00018591385812742725,
+      "loss": 0.6662,
+      "step": 171
+    },
+    {
+      "epoch": 0.056085432461074426,
+      "grad_norm": 0.21578969061374664,
+      "learning_rate": 0.00018575266561936523,
+      "loss": 0.5573,
+      "step": 172
+    },
+    {
+      "epoch": 0.05641151055677835,
+      "grad_norm": 0.2769645154476166,
+      "learning_rate": 0.00018559062676711332,
+      "loss": 0.5742,
+      "step": 173
+    },
+    {
+      "epoch": 0.05673758865248227,
+      "grad_norm": 0.221250519156456,
+      "learning_rate": 0.0001854277431699295,
+      "loss": 0.6006,
+      "step": 174
+    },
+    {
+      "epoch": 0.05706366674818619,
+      "grad_norm": 0.391968309879303,
+      "learning_rate": 0.00018526401643540922,
+      "loss": 0.7147,
+      "step": 175
+    },
+    {
+      "epoch": 0.05738974484389011,
+      "grad_norm": 0.20950183272361755,
+      "learning_rate": 0.00018509944817946922,
+      "loss": 0.5665,
+      "step": 176
+    },
+    {
+      "epoch": 0.057715822939594034,
+      "grad_norm": 0.4676218330860138,
+      "learning_rate": 0.00018493404002633166,
+      "loss": 0.6862,
+      "step": 177
+    },
+    {
+      "epoch": 0.05804190103529795,
+      "grad_norm": 0.30444085597991943,
+      "learning_rate": 0.00018476779360850832,
+      "loss": 0.6271,
+      "step": 178
+    },
+    {
+      "epoch": 0.058367979131001876,
+      "grad_norm": 0.30755650997161865,
+      "learning_rate": 0.00018460071056678422,
+      "loss": 0.7104,
+      "step": 179
+    },
+    {
+      "epoch": 0.058694057226705794,
+      "grad_norm": 0.266366571187973,
+      "learning_rate": 0.00018443279255020152,
+      "loss": 0.7298,
+      "step": 180
+    },
+    {
+      "epoch": 0.05902013532240972,
+      "grad_norm": 0.3003389537334442,
+      "learning_rate": 0.00018426404121604323,
+      "loss": 0.7506,
+      "step": 181
+    },
+    {
+      "epoch": 0.059346213418113636,
+      "grad_norm": 0.30031818151474,
+      "learning_rate": 0.00018409445822981693,
+      "loss": 0.6368,
+      "step": 182
+    },
+    {
+      "epoch": 0.05967229151381756,
+      "grad_norm": 0.23968857526779175,
+      "learning_rate": 0.00018392404526523817,
+      "loss": 0.5467,
+      "step": 183
+    },
+    {
+      "epoch": 0.05999836960952148,
+      "grad_norm": 0.24690750241279602,
+      "learning_rate": 0.0001837528040042142,
+      "loss": 0.6422,
+      "step": 184
+    },
+    {
+      "epoch": 0.0603244477052254,
+      "grad_norm": 0.2936251759529114,
+      "learning_rate": 0.00018358073613682706,
+      "loss": 0.7603,
+      "step": 185
+    },
+    {
+      "epoch": 0.06065052580092932,
+      "grad_norm": 0.33426618576049805,
+      "learning_rate": 0.00018340784336131713,
+      "loss": 0.82,
+      "step": 186
+    },
+    {
+      "epoch": 0.060976603896633244,
+      "grad_norm": 0.29302123188972473,
+      "learning_rate": 0.00018323412738406635,
+      "loss": 0.643,
+      "step": 187
+    },
+    {
+      "epoch": 0.06130268199233716,
+      "grad_norm": 0.23355039954185486,
+      "learning_rate": 0.00018305958991958127,
+      "loss": 0.6233,
+      "step": 188
+    },
+    {
+      "epoch": 0.061628760088041086,
+      "grad_norm": 0.503938615322113,
+      "learning_rate": 0.0001828842326904762,
+      "loss": 0.717,
+      "step": 189
+    },
+    {
+      "epoch": 0.061954838183745004,
+      "grad_norm": 0.23740191757678986,
+      "learning_rate": 0.00018270805742745617,
+      "loss": 0.6435,
+      "step": 190
+    },
+    {
+      "epoch": 0.06228091627944893,
+      "grad_norm": 0.23052096366882324,
+      "learning_rate": 0.00018253106586929997,
+      "loss": 0.6395,
+      "step": 191
+    },
+    {
+      "epoch": 0.06260699437515285,
+      "grad_norm": 0.2588057219982147,
+      "learning_rate": 0.00018235325976284275,
+      "loss": 0.5851,
+      "step": 192
+    },
+    {
+      "epoch": 0.06293307247085678,
+      "grad_norm": 0.26309555768966675,
+      "learning_rate": 0.00018217464086295904,
+      "loss": 0.623,
+      "step": 193
+    },
+    {
+      "epoch": 0.06325915056656069,
+      "grad_norm": 0.30716729164123535,
+      "learning_rate": 0.00018199521093254523,
+      "loss": 0.7346,
+      "step": 194
+    },
+    {
+      "epoch": 0.06358522866226461,
+      "grad_norm": 0.2664608061313629,
+      "learning_rate": 0.00018181497174250236,
+      "loss": 0.737,
+      "step": 195
+    },
+    {
+      "epoch": 0.06391130675796854,
+      "grad_norm": 0.33862313628196716,
+      "learning_rate": 0.00018163392507171842,
+      "loss": 0.6553,
+      "step": 196
+    },
+    {
+      "epoch": 0.06423738485367246,
+      "grad_norm": 0.29159292578697205,
+      "learning_rate": 0.00018145207270705096,
+      "loss": 0.6077,
+      "step": 197
+    },
+    {
+      "epoch": 0.06456346294937637,
+      "grad_norm": 0.28656378388404846,
+      "learning_rate": 0.0001812694164433094,
+      "loss": 0.7094,
+      "step": 198
+    },
+    {
+      "epoch": 0.0648895410450803,
+      "grad_norm": 0.3513023853302002,
+      "learning_rate": 0.00018108595808323736,
+      "loss": 0.6811,
+      "step": 199
+    },
+    {
+      "epoch": 0.06521561914078422,
+      "grad_norm": 0.44914510846138,
+      "learning_rate": 0.00018090169943749476,
+      "loss": 0.5998,
+      "step": 200
+    },
+    {
+      "epoch": 0.06554169723648814,
+      "grad_norm": 0.25120270252227783,
+      "learning_rate": 0.00018071664232464002,
+      "loss": 0.5754,
+      "step": 201
+    },
+    {
+      "epoch": 0.06586777533219206,
+      "grad_norm": 0.3265678584575653,
+      "learning_rate": 0.0001805307885711122,
+      "loss": 0.6652,
+      "step": 202
+    },
+    {
+      "epoch": 0.06619385342789598,
+      "grad_norm": 0.2386000156402588,
+      "learning_rate": 0.00018034414001121278,
+      "loss": 0.6503,
+      "step": 203
+    },
+    {
+      "epoch": 0.0665199315235999,
+      "grad_norm": 0.3131890892982483,
+      "learning_rate": 0.00018015669848708767,
+      "loss": 0.6924,
+      "step": 204
+    },
+    {
+      "epoch": 0.06684600961930383,
+      "grad_norm": 0.2541196048259735,
+      "learning_rate": 0.00017996846584870908,
+      "loss": 0.7512,
+      "step": 205
+    },
+    {
+      "epoch": 0.06717208771500774,
+      "grad_norm": 0.2935178577899933,
+      "learning_rate": 0.0001797794439538571,
+      "loss": 0.6095,
+      "step": 206
+    },
+    {
+      "epoch": 0.06749816581071166,
+      "grad_norm": 0.25794175267219543,
+      "learning_rate": 0.0001795896346681016,
+      "loss": 0.5873,
+      "step": 207
+    },
+    {
+      "epoch": 0.06782424390641559,
+      "grad_norm": 0.34597495198249817,
+      "learning_rate": 0.00017939903986478355,
+      "loss": 0.6594,
+      "step": 208
+    },
+    {
+      "epoch": 0.06815032200211951,
+      "grad_norm": 0.26853707432746887,
+      "learning_rate": 0.00017920766142499672,
+      "loss": 0.6388,
+      "step": 209
+    },
+    {
+      "epoch": 0.06847640009782342,
+      "grad_norm": 0.2943996787071228,
+      "learning_rate": 0.00017901550123756906,
+      "loss": 0.6592,
+      "step": 210
+    },
+    {
+      "epoch": 0.06880247819352735,
+      "grad_norm": 0.2579059600830078,
+      "learning_rate": 0.00017882256119904403,
+      "loss": 0.7697,
+      "step": 211
+    },
+    {
+      "epoch": 0.06912855628923127,
+      "grad_norm": 0.30710938572883606,
+      "learning_rate": 0.00017862884321366188,
+      "loss": 0.6289,
+      "step": 212
+    },
+    {
+      "epoch": 0.0694546343849352,
+      "grad_norm": 0.5300430059432983,
+      "learning_rate": 0.000178434349193341,
+      "loss": 0.7128,
+      "step": 213
+    },
+    {
+      "epoch": 0.0697807124806391,
+      "grad_norm": 0.34696683287620544,
+      "learning_rate": 0.0001782390810576588,
+      "loss": 0.7006,
+      "step": 214
+    },
+    {
+      "epoch": 0.07010679057634303,
+      "grad_norm": 0.2758086621761322,
+      "learning_rate": 0.000178043040733833,
+      "loss": 0.6286,
+      "step": 215
+    },
+    {
+      "epoch": 0.07043286867204696,
+      "grad_norm": 0.3609292507171631,
+      "learning_rate": 0.00017784623015670238,
+      "loss": 0.6675,
+      "step": 216
+    },
+    {
+      "epoch": 0.07075894676775088,
+      "grad_norm": 0.2829347848892212,
+      "learning_rate": 0.00017764865126870786,
+      "loss": 0.6893,
+      "step": 217
+    },
+    {
+      "epoch": 0.0710850248634548,
+      "grad_norm": 0.5950068235397339,
+      "learning_rate": 0.00017745030601987337,
+      "loss": 0.7254,
+      "step": 218
+    },
+    {
+      "epoch": 0.07141110295915871,
+      "grad_norm": 0.30742985010147095,
+      "learning_rate": 0.00017725119636778644,
+      "loss": 0.6998,
+      "step": 219
+    },
+    {
+      "epoch": 0.07173718105486264,
+      "grad_norm": 0.4018792510032654,
+      "learning_rate": 0.00017705132427757895,
+      "loss": 0.8801,
+      "step": 220
+    },
+    {
+      "epoch": 0.07206325915056656,
+      "grad_norm": 0.3693302273750305,
+      "learning_rate": 0.00017685069172190766,
+      "loss": 0.6576,
+      "step": 221
+    },
+    {
+      "epoch": 0.07238933724627049,
+      "grad_norm": 0.23466476798057556,
+      "learning_rate": 0.00017664930068093498,
+      "loss": 0.6179,
+      "step": 222
+    },
+    {
+      "epoch": 0.0727154153419744,
+      "grad_norm": 0.4448927938938141,
+      "learning_rate": 0.00017644715314230918,
+      "loss": 0.7786,
+      "step": 223
+    },
+    {
+      "epoch": 0.07304149343767832,
+      "grad_norm": 0.25553828477859497,
+      "learning_rate": 0.0001762442511011448,
+      "loss": 0.5989,
+      "step": 224
+    },
+    {
+      "epoch": 0.07336757153338225,
+      "grad_norm": 0.253617525100708,
+      "learning_rate": 0.0001760405965600031,
+      "loss": 0.6312,
+      "step": 225
+    },
+    {
+      "epoch": 0.07369364962908617,
+      "grad_norm": 0.24342387914657593,
+      "learning_rate": 0.0001758361915288722,
+      "loss": 0.5719,
+      "step": 226
+    },
+    {
+      "epoch": 0.07401972772479008,
+      "grad_norm": 0.26518356800079346,
+      "learning_rate": 0.0001756310380251472,
+      "loss": 0.5901,
+      "step": 227
+    },
+    {
+      "epoch": 0.07434580582049401,
+      "grad_norm": 0.22742488980293274,
+      "learning_rate": 0.00017542513807361037,
+      "loss": 0.5329,
+      "step": 228
+    },
+    {
+      "epoch": 0.07467188391619793,
+      "grad_norm": 0.3179355263710022,
+      "learning_rate": 0.00017521849370641114,
+      "loss": 0.6908,
+      "step": 229
+    },
+    {
+      "epoch": 0.07499796201190186,
+      "grad_norm": 0.2648365795612335,
+      "learning_rate": 0.00017501110696304596,
+      "loss": 0.6078,
+      "step": 230
+    },
+    {
+      "epoch": 0.07532404010760577,
+      "grad_norm": 0.2643827497959137,
+      "learning_rate": 0.00017480297989033825,
+      "loss": 0.7111,
+      "step": 231
+    },
+    {
+      "epoch": 0.07565011820330969,
+      "grad_norm": 0.2998198866844177,
+      "learning_rate": 0.00017459411454241822,
+      "loss": 0.6759,
+      "step": 232
+    },
+    {
+      "epoch": 0.07597619629901362,
+      "grad_norm": 0.2904861867427826,
+      "learning_rate": 0.00017438451298070252,
+      "loss": 0.6876,
+      "step": 233
+    },
+    {
+      "epoch": 0.07630227439471754,
+      "grad_norm": 0.33539697527885437,
+      "learning_rate": 0.00017417417727387394,
+      "loss": 0.6744,
+      "step": 234
+    },
+    {
+      "epoch": 0.07662835249042145,
+      "grad_norm": 0.2617104947566986,
+      "learning_rate": 0.000173963109497861,
+      "loss": 0.5529,
+      "step": 235
+    },
+    {
+      "epoch": 0.07695443058612537,
+      "grad_norm": 0.23713330924510956,
+      "learning_rate": 0.0001737513117358174,
+      "loss": 0.606,
+      "step": 236
+    },
+    {
+      "epoch": 0.0772805086818293,
+      "grad_norm": 0.2559371888637543,
+      "learning_rate": 0.0001735387860781016,
+      "loss": 0.5811,
+      "step": 237
+    },
+    {
+      "epoch": 0.07760658677753322,
+      "grad_norm": 0.27742600440979004,
+      "learning_rate": 0.00017332553462225602,
+      "loss": 0.6415,
+      "step": 238
+    },
+    {
+      "epoch": 0.07793266487323713,
+      "grad_norm": 0.26602035760879517,
+      "learning_rate": 0.00017311155947298643,
+      "loss": 0.6722,
+      "step": 239
+    },
+    {
+      "epoch": 0.07825874296894106,
+      "grad_norm": 0.2551373839378357,
+      "learning_rate": 0.00017289686274214118,
+      "loss": 0.5842,
+      "step": 240
+    },
+    {
+      "epoch": 0.07858482106464498,
+      "grad_norm": 0.3161643147468567,
+      "learning_rate": 0.0001726814465486903,
+      "loss": 0.5979,
+      "step": 241
+    },
+    {
+      "epoch": 0.07891089916034891,
+      "grad_norm": 0.3652503788471222,
+      "learning_rate": 0.0001724653130187047,
+      "loss": 0.6652,
+      "step": 242
+    },
+    {
+      "epoch": 0.07923697725605282,
+      "grad_norm": 0.2684518098831177,
+      "learning_rate": 0.00017224846428533499,
+      "loss": 0.6001,
+      "step": 243
+    },
+    {
+      "epoch": 0.07956305535175674,
+      "grad_norm": 0.21544717252254486,
+      "learning_rate": 0.0001720309024887907,
+      "loss": 0.61,
+      "step": 244
+    },
+    {
+      "epoch": 0.07988913344746067,
+      "grad_norm": 0.2404128611087799,
+      "learning_rate": 0.00017181262977631888,
+      "loss": 0.6769,
+      "step": 245
+    },
+    {
+      "epoch": 0.08021521154316459,
+      "grad_norm": 0.21001021564006805,
+      "learning_rate": 0.00017159364830218312,
+      "loss": 0.6083,
+      "step": 246
+    },
+    {
+      "epoch": 0.08054128963886852,
+      "grad_norm": 0.36963921785354614,
+      "learning_rate": 0.00017137396022764214,
+      "loss": 0.6379,
+      "step": 247
+    },
+    {
+      "epoch": 0.08086736773457243,
+      "grad_norm": 0.259547621011734,
+      "learning_rate": 0.00017115356772092857,
+      "loss": 0.6637,
+      "step": 248
+    },
+    {
+      "epoch": 0.08119344583027635,
+      "grad_norm": 0.31841689348220825,
+      "learning_rate": 0.0001709324729572274,
+      "loss": 0.7091,
+      "step": 249
+    },
+    {
+      "epoch": 0.08151952392598028,
+      "grad_norm": 0.35306358337402344,
+      "learning_rate": 0.00017071067811865476,
+      "loss": 0.6489,
+      "step": 250
+    },
+    {
+      "epoch": 0.0818456020216842,
+      "grad_norm": 0.21508830785751343,
+      "learning_rate": 0.00017048818539423615,
+      "loss": 0.5831,
+      "step": 251
+    },
+    {
+      "epoch": 0.08217168011738811,
+      "grad_norm": 0.2714361250400543,
+      "learning_rate": 0.00017026499697988493,
+      "loss": 0.6528,
+      "step": 252
+    },
+    {
+      "epoch": 0.08249775821309203,
+      "grad_norm": 0.2910292148590088,
+      "learning_rate": 0.00017004111507838064,
+      "loss": 0.7083,
+      "step": 253
+    },
+    {
+      "epoch": 0.08282383630879596,
+      "grad_norm": 0.2184605747461319,
+      "learning_rate": 0.00016981654189934727,
+      "loss": 0.5559,
+      "step": 254
+    },
+    {
+      "epoch": 0.08314991440449988,
+      "grad_norm": 0.222558856010437,
+      "learning_rate": 0.00016959127965923142,
+      "loss": 0.6332,
+      "step": 255
+    },
+    {
+      "epoch": 0.0834759925002038,
+      "grad_norm": 0.37371590733528137,
+      "learning_rate": 0.0001693653305812805,
+      "loss": 0.5877,
+      "step": 256
+    },
+    {
+      "epoch": 0.08380207059590772,
+      "grad_norm": 0.3450126051902771,
+      "learning_rate": 0.00016913869689552064,
+      "loss": 0.7029,
+      "step": 257
+    },
+    {
+      "epoch": 0.08412814869161164,
+      "grad_norm": 0.24566428363323212,
+      "learning_rate": 0.00016891138083873487,
+      "loss": 0.5703,
+      "step": 258
+    },
+    {
+      "epoch": 0.08445422678731557,
+      "grad_norm": 0.21838568150997162,
+      "learning_rate": 0.00016868338465444085,
+      "loss": 0.5863,
+      "step": 259
+    },
+    {
+      "epoch": 0.08478030488301948,
+      "grad_norm": 0.2161954939365387,
+      "learning_rate": 0.00016845471059286887,
+      "loss": 0.5707,
+      "step": 260
+    },
+    {
+      "epoch": 0.0851063829787234,
+      "grad_norm": 0.29614922404289246,
+      "learning_rate": 0.00016822536091093965,
+      "loss": 0.6503,
+      "step": 261
+    },
+    {
+      "epoch": 0.08543246107442733,
+      "grad_norm": 0.32689064741134644,
+      "learning_rate": 0.00016799533787224192,
+      "loss": 0.6301,
+      "step": 262
+    },
+    {
+      "epoch": 0.08575853917013125,
+      "grad_norm": 0.2884279787540436,
+      "learning_rate": 0.00016776464374701025,
+      "loss": 0.5587,
+      "step": 263
+    },
+    {
+      "epoch": 0.08608461726583516,
+      "grad_norm": 0.435344398021698,
+      "learning_rate": 0.00016753328081210245,
+      "loss": 0.8083,
+      "step": 264
+    },
+    {
+      "epoch": 0.08641069536153909,
+      "grad_norm": 0.2473522573709488,
+      "learning_rate": 0.00016730125135097735,
+      "loss": 0.6562,
+      "step": 265
+    },
+    {
+      "epoch": 0.08673677345724301,
+      "grad_norm": 0.3048084080219269,
+      "learning_rate": 0.000167068557653672,
+      "loss": 0.6652,
+      "step": 266
+    },
+    {
+      "epoch": 0.08706285155294693,
+      "grad_norm": 0.3015248775482178,
+      "learning_rate": 0.0001668352020167793,
+      "loss": 0.5716,
+      "step": 267
+    },
+    {
+      "epoch": 0.08738892964865085,
+      "grad_norm": 0.26765477657318115,
+      "learning_rate": 0.00016660118674342517,
+      "loss": 0.6112,
+      "step": 268
+    },
+    {
+      "epoch": 0.08771500774435477,
+      "grad_norm": 0.2710655629634857,
+      "learning_rate": 0.00016636651414324587,
+      "loss": 0.6236,
+      "step": 269
+    },
+    {
+      "epoch": 0.0880410858400587,
+      "grad_norm": 0.3481336534023285,
+      "learning_rate": 0.00016613118653236518,
+      "loss": 0.6561,
+      "step": 270
+    },
+    {
+      "epoch": 0.08836716393576262,
+      "grad_norm": 0.39478129148483276,
+      "learning_rate": 0.0001658952062333717,
+      "loss": 0.6904,
+      "step": 271
+    },
+    {
+      "epoch": 0.08869324203146654,
+      "grad_norm": 0.24365971982479095,
+      "learning_rate": 0.00016565857557529566,
+      "loss": 0.7372,
+      "step": 272
+    },
+    {
+      "epoch": 0.08901932012717045,
+      "grad_norm": 0.2448200285434723,
+      "learning_rate": 0.00016542129689358612,
+      "loss": 0.6072,
+      "step": 273
+    },
+    {
+      "epoch": 0.08934539822287438,
+      "grad_norm": 0.23193484544754028,
+      "learning_rate": 0.0001651833725300879,
+      "loss": 0.6249,
+      "step": 274
+    },
+    {
+      "epoch": 0.0896714763185783,
+      "grad_norm": 0.3657950460910797,
+      "learning_rate": 0.00016494480483301836,
+      "loss": 0.7056,
+      "step": 275
+    },
+    {
+      "epoch": 0.08999755441428223,
+      "grad_norm": 0.22652854025363922,
+      "learning_rate": 0.00016470559615694446,
+      "loss": 0.6169,
+      "step": 276
+    },
+    {
+      "epoch": 0.09032363250998614,
+      "grad_norm": 0.23606394231319427,
+      "learning_rate": 0.00016446574886275913,
+      "loss": 0.6428,
+      "step": 277
+    },
+    {
+      "epoch": 0.09064971060569006,
+      "grad_norm": 0.21306492388248444,
+      "learning_rate": 0.00016422526531765846,
+      "loss": 0.6148,
+      "step": 278
+    },
+    {
+      "epoch": 0.09097578870139399,
+      "grad_norm": 0.2294398844242096,
+      "learning_rate": 0.00016398414789511786,
+      "loss": 0.6457,
+      "step": 279
+    },
+    {
+      "epoch": 0.09130186679709791,
+      "grad_norm": 0.3088754117488861,
+      "learning_rate": 0.000163742398974869,
+      "loss": 0.6488,
+      "step": 280
+    },
+    {
+      "epoch": 0.09162794489280182,
+      "grad_norm": 0.32271504402160645,
+      "learning_rate": 0.00016350002094287609,
+      "loss": 0.6858,
+      "step": 281
+    },
+    {
+      "epoch": 0.09195402298850575,
+      "grad_norm": 0.348728746175766,
+      "learning_rate": 0.00016325701619131246,
+      "loss": 0.671,
+      "step": 282
+    },
+    {
+      "epoch": 0.09228010108420967,
+      "grad_norm": 0.28784212470054626,
+      "learning_rate": 0.00016301338711853693,
+      "loss": 0.6244,
+      "step": 283
+    },
+    {
+      "epoch": 0.0926061791799136,
+      "grad_norm": 0.3205079734325409,
+      "learning_rate": 0.00016276913612907007,
+      "loss": 0.6944,
+      "step": 284
+    },
+    {
+      "epoch": 0.0929322572756175,
+      "grad_norm": 0.23461277782917023,
+      "learning_rate": 0.00016252426563357055,
+      "loss": 0.5726,
+      "step": 285
+    },
+    {
+      "epoch": 0.09325833537132143,
+      "grad_norm": 0.2635505795478821,
+      "learning_rate": 0.00016227877804881127,
+      "loss": 0.6212,
+      "step": 286
+    },
+    {
+      "epoch": 0.09358441346702535,
+      "grad_norm": 0.24859635531902313,
+      "learning_rate": 0.00016203267579765563,
+      "loss": 0.6139,
+      "step": 287
+    },
+    {
+      "epoch": 0.09391049156272928,
+      "grad_norm": 0.32381847500801086,
+      "learning_rate": 0.00016178596130903344,
+      "loss": 0.7499,
+      "step": 288
+    },
+    {
+      "epoch": 0.09423656965843319,
+      "grad_norm": 0.22918426990509033,
+      "learning_rate": 0.00016153863701791717,
+      "loss": 0.6352,
+      "step": 289
+    },
+    {
+      "epoch": 0.09456264775413711,
+      "grad_norm": 0.2531386911869049,
+      "learning_rate": 0.00016129070536529766,
+      "loss": 0.6116,
+      "step": 290
+    },
+    {
+      "epoch": 0.09488872584984104,
+      "grad_norm": 0.31397831439971924,
+      "learning_rate": 0.00016104216879816026,
+      "loss": 0.6613,
+      "step": 291
+    },
+    {
+      "epoch": 0.09521480394554496,
+      "grad_norm": 0.26814454793930054,
+      "learning_rate": 0.00016079302976946055,
+      "loss": 0.5927,
+      "step": 292
+    },
+    {
+      "epoch": 0.09554088204124887,
+      "grad_norm": 0.2622220516204834,
+      "learning_rate": 0.00016054329073810015,
+      "loss": 0.5552,
+      "step": 293
+    },
+    {
+      "epoch": 0.0958669601369528,
+      "grad_norm": 0.20141161978244781,
+      "learning_rate": 0.00016029295416890248,
+      "loss": 0.5583,
+      "step": 294
+    },
+    {
+      "epoch": 0.09619303823265672,
+      "grad_norm": 0.24076247215270996,
+      "learning_rate": 0.00016004202253258842,
+      "loss": 0.6295,
+      "step": 295
+    },
+    {
+      "epoch": 0.09651911632836065,
+      "grad_norm": 0.28313252329826355,
+      "learning_rate": 0.0001597904983057519,
+      "loss": 0.655,
+      "step": 296
+    },
+    {
+      "epoch": 0.09684519442406456,
+      "grad_norm": 0.3352673351764679,
+      "learning_rate": 0.00015953838397083552,
+      "loss": 0.6478,
+      "step": 297
+    },
+    {
+      "epoch": 0.09717127251976848,
+      "grad_norm": 0.2615090012550354,
+      "learning_rate": 0.00015928568201610595,
+      "loss": 0.5878,
+      "step": 298
+    },
+    {
+      "epoch": 0.0974973506154724,
+      "grad_norm": 0.22464829683303833,
+      "learning_rate": 0.00015903239493562948,
+      "loss": 0.5686,
+      "step": 299
+    },
+    {
+      "epoch": 0.09782342871117633,
+      "grad_norm": 0.24365228414535522,
+      "learning_rate": 0.00015877852522924732,
+      "loss": 0.4556,
+      "step": 300
+    },
+    {
+      "epoch": 0.09814950680688025,
+      "grad_norm": 0.22546721994876862,
+      "learning_rate": 0.00015852407540255104,
+      "loss": 0.5845,
+      "step": 301
+    },
+    {
+      "epoch": 0.09847558490258417,
+      "grad_norm": 0.3008763790130615,
+      "learning_rate": 0.00015826904796685762,
+      "loss": 0.5832,
+      "step": 302
+    },
+    {
+      "epoch": 0.09880166299828809,
+      "grad_norm": 0.37207353115081787,
+      "learning_rate": 0.00015801344543918495,
+      "loss": 0.8106,
+      "step": 303
+    },
+    {
+      "epoch": 0.09912774109399201,
+      "grad_norm": 0.27629679441452026,
+      "learning_rate": 0.00015775727034222675,
+      "loss": 0.6696,
+      "step": 304
+    },
+    {
+      "epoch": 0.09945381918969594,
+      "grad_norm": 0.32836827635765076,
+      "learning_rate": 0.00015750052520432787,
+      "loss": 0.6924,
+      "step": 305
+    },
+    {
+      "epoch": 0.09977989728539985,
+      "grad_norm": 0.2459554374217987,
+      "learning_rate": 0.0001572432125594591,
+      "loss": 0.6045,
+      "step": 306
+    },
+    {
+      "epoch": 0.10010597538110377,
+      "grad_norm": 0.2423304170370102,
+      "learning_rate": 0.00015698533494719238,
+      "loss": 0.6067,
+      "step": 307
+    },
+    {
+      "epoch": 0.1004320534768077,
+      "grad_norm": 0.33577221632003784,
+      "learning_rate": 0.00015672689491267567,
+      "loss": 0.6626,
+      "step": 308
+    },
+    {
+      "epoch": 0.10075813157251162,
+      "grad_norm": 0.2271413505077362,
+      "learning_rate": 0.00015646789500660773,
+      "loss": 0.6213,
+      "step": 309
+    },
+    {
+      "epoch": 0.10108420966821553,
+      "grad_norm": 0.3947584629058838,
+      "learning_rate": 0.00015620833778521307,
+      "loss": 0.738,
+      "step": 310
+    },
+    {
+      "epoch": 0.10141028776391946,
+      "grad_norm": 0.2767365872859955,
+      "learning_rate": 0.0001559482258102167,
+      "loss": 0.563,
+      "step": 311
+    },
+    {
+      "epoch": 0.10173636585962338,
+      "grad_norm": 0.26970767974853516,
+      "learning_rate": 0.00015568756164881882,
+      "loss": 0.7104,
+      "step": 312
+    },
+    {
+      "epoch": 0.1020624439553273,
+      "grad_norm": 0.26086539030075073,
+      "learning_rate": 0.00015542634787366942,
+      "loss": 0.6614,
+      "step": 313
+    },
+    {
+      "epoch": 0.10238852205103122,
+      "grad_norm": 0.17937611043453217,
+      "learning_rate": 0.00015516458706284303,
+      "loss": 0.5259,
+      "step": 314
+    },
+    {
+      "epoch": 0.10271460014673514,
+      "grad_norm": 0.17191629111766815,
+      "learning_rate": 0.0001549022817998132,
+      "loss": 0.5171,
+      "step": 315
+    },
+    {
+      "epoch": 0.10304067824243907,
+      "grad_norm": 0.2208365947008133,
+      "learning_rate": 0.00015463943467342693,
+      "loss": 0.5459,
+      "step": 316
+    },
+    {
+      "epoch": 0.10336675633814299,
+      "grad_norm": 0.28957873582839966,
+      "learning_rate": 0.00015437604827787927,
+      "loss": 0.6093,
+      "step": 317
+    },
+    {
+      "epoch": 0.1036928344338469,
+      "grad_norm": 0.2496960610151291,
+      "learning_rate": 0.00015411212521268758,
+      "loss": 0.6104,
+      "step": 318
+    },
+    {
+      "epoch": 0.10401891252955082,
+      "grad_norm": 0.2421676069498062,
+      "learning_rate": 0.00015384766808266602,
+      "loss": 0.6567,
+      "step": 319
+    },
+    {
+      "epoch": 0.10434499062525475,
+      "grad_norm": 0.2969132363796234,
+      "learning_rate": 0.00015358267949789966,
+      "loss": 0.6265,
+      "step": 320
+    },
+    {
+      "epoch": 0.10467106872095867,
+      "grad_norm": 0.2575230300426483,
+      "learning_rate": 0.00015331716207371888,
+      "loss": 0.6724,
+      "step": 321
+    },
+    {
+      "epoch": 0.10499714681666258,
+      "grad_norm": 0.364106148481369,
+      "learning_rate": 0.0001530511184306734,
+      "loss": 0.7101,
+      "step": 322
+    },
+    {
+      "epoch": 0.10532322491236651,
+      "grad_norm": 0.26639074087142944,
+      "learning_rate": 0.00015278455119450664,
+      "loss": 0.572,
+      "step": 323
+    },
+    {
+      "epoch": 0.10564930300807043,
+      "grad_norm": 0.24254579842090607,
+      "learning_rate": 0.0001525174629961296,
+      "loss": 0.5839,
+      "step": 324
+    },
+    {
+      "epoch": 0.10597538110377436,
+      "grad_norm": 0.2786118984222412,
+      "learning_rate": 0.0001522498564715949,
+      "loss": 0.6216,
+      "step": 325
+    },
+    {
+      "epoch": 0.10630145919947827,
+      "grad_norm": 0.2908596396446228,
+      "learning_rate": 0.00015198173426207094,
+      "loss": 0.5082,
+      "step": 326
+    },
+    {
+      "epoch": 0.10662753729518219,
+      "grad_norm": 0.26790931820869446,
+      "learning_rate": 0.00015171309901381572,
+      "loss": 0.6836,
+      "step": 327
+    },
+    {
+      "epoch": 0.10695361539088612,
+      "grad_norm": 0.25367841124534607,
+      "learning_rate": 0.00015144395337815064,
+      "loss": 0.5351,
+      "step": 328
+    },
+    {
+      "epoch": 0.10727969348659004,
+      "grad_norm": 0.23655357956886292,
+      "learning_rate": 0.00015117430001143452,
+      "loss": 0.6234,
+      "step": 329
+    },
+    {
+      "epoch": 0.10760577158229397,
+      "grad_norm": 0.29328015446662903,
+      "learning_rate": 0.00015090414157503714,
+      "loss": 0.6051,
+      "step": 330
+    },
+    {
+      "epoch": 0.10793184967799788,
+      "grad_norm": 0.2480703741312027,
+      "learning_rate": 0.00015063348073531324,
+      "loss": 0.6228,
+      "step": 331
+    },
+    {
+      "epoch": 0.1082579277737018,
+      "grad_norm": 0.21856042742729187,
+      "learning_rate": 0.0001503623201635761,
+      "loss": 0.5922,
+      "step": 332
+    },
+    {
+      "epoch": 0.10858400586940573,
+      "grad_norm": 0.27874556183815,
+      "learning_rate": 0.000150090662536071,
+      "loss": 0.648,
+      "step": 333
+    },
+    {
+      "epoch": 0.10891008396510965,
+      "grad_norm": 0.3144196569919586,
+      "learning_rate": 0.0001498185105339491,
+      "loss": 0.6663,
+      "step": 334
+    },
+    {
+      "epoch": 0.10923616206081356,
+      "grad_norm": 0.24514897167682648,
+      "learning_rate": 0.00014954586684324078,
+      "loss": 0.5493,
+      "step": 335
+    },
+    {
+      "epoch": 0.10956224015651748,
+      "grad_norm": 0.33484047651290894,
+      "learning_rate": 0.00014927273415482915,
+      "loss": 0.6464,
+      "step": 336
+    },
+    {
+      "epoch": 0.10988831825222141,
+      "grad_norm": 0.26239341497421265,
+      "learning_rate": 0.00014899911516442365,
+      "loss": 0.6216,
+      "step": 337
+    },
+    {
+      "epoch": 0.11021439634792533,
+      "grad_norm": 0.2449112832546234,
+      "learning_rate": 0.00014872501257253323,
+      "loss": 0.6243,
+      "step": 338
+    },
+    {
+      "epoch": 0.11054047444362924,
+      "grad_norm": 0.25176045298576355,
+      "learning_rate": 0.0001484504290844398,
+      "loss": 0.5852,
+      "step": 339
+    },
+    {
+      "epoch": 0.11086655253933317,
+      "grad_norm": 0.3558025360107422,
+      "learning_rate": 0.00014817536741017152,
+      "loss": 0.6305,
+      "step": 340
+    },
+    {
+      "epoch": 0.11119263063503709,
+      "grad_norm": 0.3823552131652832,
+      "learning_rate": 0.00014789983026447612,
+      "loss": 0.6243,
+      "step": 341
+    },
+    {
+      "epoch": 0.11151870873074102,
+      "grad_norm": 0.30289652943611145,
+      "learning_rate": 0.0001476238203667939,
+      "loss": 0.5996,
+      "step": 342
+    },
+    {
+      "epoch": 0.11184478682644493,
+      "grad_norm": 0.27371859550476074,
+      "learning_rate": 0.0001473473404412312,
+      "loss": 0.6582,
+      "step": 343
+    },
+    {
+      "epoch": 0.11217086492214885,
+      "grad_norm": 0.26179400086402893,
+      "learning_rate": 0.0001470703932165333,
+      "loss": 0.7194,
+      "step": 344
+    },
+    {
+      "epoch": 0.11249694301785278,
+      "grad_norm": 0.32942402362823486,
+      "learning_rate": 0.00014679298142605734,
+      "loss": 0.6949,
+      "step": 345
+    },
+    {
+      "epoch": 0.1128230211135567,
+      "grad_norm": 0.2867915630340576,
+      "learning_rate": 0.00014651510780774583,
+      "loss": 0.6307,
+      "step": 346
+    },
+    {
+      "epoch": 0.11314909920926061,
+      "grad_norm": 0.23128168284893036,
+      "learning_rate": 0.00014623677510409918,
+      "loss": 0.5856,
+      "step": 347
+    },
+    {
+      "epoch": 0.11347517730496454,
+      "grad_norm": 0.23641528189182281,
+      "learning_rate": 0.00014595798606214882,
+      "loss": 0.693,
+      "step": 348
+    },
+    {
+      "epoch": 0.11380125540066846,
+      "grad_norm": 0.23140428960323334,
+      "learning_rate": 0.00014567874343342997,
+      "loss": 0.6498,
+      "step": 349
+    },
+    {
+      "epoch": 0.11412733349637239,
+      "grad_norm": 0.3104979991912842,
+      "learning_rate": 0.00014539904997395468,
+      "loss": 0.6793,
+      "step": 350
+    },
+    {
+      "epoch": 0.1144534115920763,
+      "grad_norm": 0.28645816445350647,
+      "learning_rate": 0.00014511890844418453,
+      "loss": 0.7425,
+      "step": 351
+    },
+    {
+      "epoch": 0.11477948968778022,
+      "grad_norm": 0.31605294346809387,
+      "learning_rate": 0.00014483832160900326,
+      "loss": 0.5866,
+      "step": 352
+    },
+    {
+      "epoch": 0.11510556778348414,
+      "grad_norm": 0.27439796924591064,
+      "learning_rate": 0.00014455729223768966,
+      "loss": 0.5785,
+      "step": 353
+    },
+    {
+      "epoch": 0.11543164587918807,
+      "grad_norm": 0.3302454650402069,
+      "learning_rate": 0.0001442758231038902,
+      "loss": 0.658,
+      "step": 354
+    },
+    {
+      "epoch": 0.115757723974892,
+      "grad_norm": 0.18580487370491028,
+      "learning_rate": 0.00014399391698559152,
+      "loss": 0.439,
+      "step": 355
+    },
+    {
+      "epoch": 0.1160838020705959,
+      "grad_norm": 0.2431912124156952,
+      "learning_rate": 0.0001437115766650933,
+      "loss": 0.6042,
+      "step": 356
+    },
+    {
+      "epoch": 0.11640988016629983,
+      "grad_norm": 0.2991044223308563,
+      "learning_rate": 0.00014342880492898048,
+      "loss": 0.649,
+      "step": 357
+    },
+    {
+      "epoch": 0.11673595826200375,
+      "grad_norm": 0.31540584564208984,
+      "learning_rate": 0.0001431456045680959,
+      "loss": 0.5912,
+      "step": 358
+    },
+    {
+      "epoch": 0.11706203635770768,
+      "grad_norm": 0.2192750871181488,
+      "learning_rate": 0.00014286197837751286,
+      "loss": 0.6253,
+      "step": 359
+    },
+    {
+      "epoch": 0.11738811445341159,
+      "grad_norm": 0.23064061999320984,
+      "learning_rate": 0.00014257792915650728,
+      "loss": 0.5753,
+      "step": 360
+    },
+    {
+      "epoch": 0.11771419254911551,
+      "grad_norm": 0.29536232352256775,
+      "learning_rate": 0.00014229345970853032,
+      "loss": 0.6757,
+      "step": 361
+    },
+    {
+      "epoch": 0.11804027064481944,
+      "grad_norm": 0.30984580516815186,
+      "learning_rate": 0.00014200857284118066,
+      "loss": 0.601,
+      "step": 362
+    },
+    {
+      "epoch": 0.11836634874052336,
+      "grad_norm": 0.2990546226501465,
+      "learning_rate": 0.00014172327136617656,
+      "loss": 0.6723,
+      "step": 363
+    },
+    {
+      "epoch": 0.11869242683622727,
+      "grad_norm": 0.2658703029155731,
+      "learning_rate": 0.00014143755809932845,
+      "loss": 0.6146,
+      "step": 364
+    },
+    {
+      "epoch": 0.1190185049319312,
+      "grad_norm": 0.20527562499046326,
+      "learning_rate": 0.00014115143586051088,
+      "loss": 0.5963,
+      "step": 365
+    },
+    {
+      "epoch": 0.11934458302763512,
+      "grad_norm": 0.24419987201690674,
+      "learning_rate": 0.00014086490747363493,
+      "loss": 0.6174,
+      "step": 366
+    },
+    {
+      "epoch": 0.11967066112333904,
+      "grad_norm": 0.2944442629814148,
+      "learning_rate": 0.00014057797576662,
+      "loss": 0.6109,
+      "step": 367
+    },
+    {
+      "epoch": 0.11999673921904296,
+      "grad_norm": 0.26260483264923096,
+      "learning_rate": 0.00014029064357136628,
+      "loss": 0.681,
+      "step": 368
+    },
+    {
+      "epoch": 0.12032281731474688,
+      "grad_norm": 0.2430226057767868,
+      "learning_rate": 0.00014000291372372647,
+      "loss": 0.5597,
+      "step": 369
+    },
+    {
+      "epoch": 0.1206488954104508,
+      "grad_norm": 0.3209778666496277,
+      "learning_rate": 0.00013971478906347806,
+      "loss": 0.7486,
+      "step": 370
+    },
+    {
+      "epoch": 0.12097497350615473,
+      "grad_norm": 0.36402198672294617,
+      "learning_rate": 0.00013942627243429512,
+      "loss": 0.7214,
+      "step": 371
+    },
+    {
+      "epoch": 0.12130105160185864,
+      "grad_norm": 0.4271096885204315,
+      "learning_rate": 0.00013913736668372026,
+      "loss": 0.7041,
+      "step": 372
+    },
+    {
+      "epoch": 0.12162712969756256,
+      "grad_norm": 0.26176485419273376,
+      "learning_rate": 0.00013884807466313663,
+      "loss": 0.6308,
+      "step": 373
+    },
+    {
+      "epoch": 0.12195320779326649,
+      "grad_norm": 0.30541542172431946,
+      "learning_rate": 0.00013855839922773968,
+      "loss": 0.6901,
+      "step": 374
+    },
+    {
+      "epoch": 0.12227928588897041,
+      "grad_norm": 0.2745092511177063,
+      "learning_rate": 0.000138268343236509,
+      "loss": 0.7374,
+      "step": 375
+    },
+    {
+      "epoch": 0.12260536398467432,
+      "grad_norm": 0.23461657762527466,
+      "learning_rate": 0.00013797790955218014,
+      "loss": 0.5745,
+      "step": 376
+    },
+    {
+      "epoch": 0.12293144208037825,
+      "grad_norm": 0.2348318099975586,
+      "learning_rate": 0.00013768710104121627,
+      "loss": 0.5153,
+      "step": 377
+    },
+    {
+      "epoch": 0.12325752017608217,
+      "grad_norm": 0.22238482534885406,
+      "learning_rate": 0.00013739592057378003,
+      "loss": 0.6065,
+      "step": 378
+    },
+    {
+      "epoch": 0.1235835982717861,
+      "grad_norm": 0.34176740050315857,
+      "learning_rate": 0.0001371043710237051,
+      "loss": 0.6059,
+      "step": 379
+    },
+    {
+      "epoch": 0.12390967636749001,
+      "grad_norm": 0.33095425367355347,
+      "learning_rate": 0.00013681245526846783,
+      "loss": 0.6373,
+      "step": 380
+    },
+    {
+      "epoch": 0.12423575446319393,
+      "grad_norm": 0.28466686606407166,
+      "learning_rate": 0.0001365201761891588,
+      "loss": 0.5869,
+      "step": 381
+    },
+    {
+      "epoch": 0.12456183255889786,
+      "grad_norm": 0.37012436985969543,
+      "learning_rate": 0.00013622753667045457,
+      "loss": 0.6946,
+      "step": 382
+    },
+    {
+      "epoch": 0.12488791065460178,
+      "grad_norm": 0.2527792453765869,
+      "learning_rate": 0.00013593453960058908,
+      "loss": 0.5629,
+      "step": 383
+    },
+    {
+      "epoch": 0.1252139887503057,
+      "grad_norm": 0.3835316002368927,
+      "learning_rate": 0.00013564118787132506,
+      "loss": 0.7292,
+      "step": 384
+    },
+    {
+      "epoch": 0.12554006684600963,
+      "grad_norm": 0.22863373160362244,
+      "learning_rate": 0.00013534748437792573,
+      "loss": 0.5662,
+      "step": 385
+    },
+    {
+      "epoch": 0.12586614494171355,
+      "grad_norm": 0.2078106552362442,
+      "learning_rate": 0.0001350534320191259,
+      "loss": 0.614,
+      "step": 386
+    },
+    {
+      "epoch": 0.12619222303741745,
+      "grad_norm": 0.2273118942975998,
+      "learning_rate": 0.0001347590336971037,
+      "loss": 0.6229,
+      "step": 387
+    },
+    {
+      "epoch": 0.12651830113312137,
+      "grad_norm": 0.26143860816955566,
+      "learning_rate": 0.0001344642923174517,
+      "loss": 0.5764,
+      "step": 388
+    },
+    {
+      "epoch": 0.1268443792288253,
+      "grad_norm": 0.33049219846725464,
+      "learning_rate": 0.00013416921078914835,
+      "loss": 0.6867,
+      "step": 389
+    },
+    {
+      "epoch": 0.12717045732452922,
+      "grad_norm": 0.3913268446922302,
+      "learning_rate": 0.00013387379202452917,
+      "loss": 0.6019,
+      "step": 390
+    },
+    {
+      "epoch": 0.12749653542023315,
+      "grad_norm": 0.35554900765419006,
+      "learning_rate": 0.00013357803893925807,
+      "loss": 0.6319,
+      "step": 391
+    },
+    {
+      "epoch": 0.12782261351593707,
+      "grad_norm": 0.322226345539093,
+      "learning_rate": 0.00013328195445229868,
+      "loss": 0.6853,
+      "step": 392
+    },
+    {
+      "epoch": 0.128148691611641,
+      "grad_norm": 0.33744320273399353,
+      "learning_rate": 0.00013298554148588528,
+      "loss": 0.5545,
+      "step": 393
+    },
+    {
+      "epoch": 0.12847476970734492,
+      "grad_norm": 0.2647642493247986,
+      "learning_rate": 0.00013268880296549425,
+      "loss": 0.6286,
+      "step": 394
+    },
+    {
+      "epoch": 0.12880084780304882,
+      "grad_norm": 0.34832268953323364,
+      "learning_rate": 0.00013239174181981495,
+      "loss": 0.7018,
+      "step": 395
+    },
+    {
+      "epoch": 0.12912692589875274,
+      "grad_norm": 0.2576987147331238,
+      "learning_rate": 0.00013209436098072095,
+      "loss": 0.6339,
+      "step": 396
+    },
+    {
+      "epoch": 0.12945300399445667,
+      "grad_norm": 0.2219506800174713,
+      "learning_rate": 0.00013179666338324108,
+      "loss": 0.5791,
+      "step": 397
+    },
+    {
+      "epoch": 0.1297790820901606,
+      "grad_norm": 0.2386060357093811,
+      "learning_rate": 0.0001314986519655305,
+      "loss": 0.6095,
+      "step": 398
+    },
+    {
+      "epoch": 0.13010516018586452,
+      "grad_norm": 0.2350076287984848,
+      "learning_rate": 0.0001312003296688415,
+      "loss": 0.5839,
+      "step": 399
+    },
+    {
+      "epoch": 0.13043123828156844,
+      "grad_norm": 0.27746814489364624,
+      "learning_rate": 0.00013090169943749476,
+      "loss": 0.6154,
+      "step": 400
+    },
+    {
+      "epoch": 0.13075731637727236,
+      "grad_norm": 0.2833282947540283,
+      "learning_rate": 0.0001306027642188501,
+      "loss": 0.658,
+      "step": 401
+    },
+    {
+      "epoch": 0.1310833944729763,
+      "grad_norm": 0.444374680519104,
+      "learning_rate": 0.00013030352696327742,
+      "loss": 0.5924,
+      "step": 402
+    },
+    {
+      "epoch": 0.13140947256868019,
+      "grad_norm": 0.30098557472229004,
+      "learning_rate": 0.00013000399062412763,
+      "loss": 0.7752,
+      "step": 403
+    },
+    {
+      "epoch": 0.1317355506643841,
+      "grad_norm": 0.3449711501598358,
+      "learning_rate": 0.0001297041581577035,
+      "loss": 0.6651,
+      "step": 404
+    },
+    {
+      "epoch": 0.13206162876008803,
+      "grad_norm": 0.273874431848526,
+      "learning_rate": 0.0001294040325232304,
+      "loss": 0.5093,
+      "step": 405
+    },
+    {
+      "epoch": 0.13238770685579196,
+      "grad_norm": 0.23658856749534607,
+      "learning_rate": 0.00012910361668282719,
+      "loss": 0.6612,
+      "step": 406
+    },
+    {
+      "epoch": 0.13271378495149588,
+      "grad_norm": 0.3019971251487732,
+      "learning_rate": 0.00012880291360147693,
+      "loss": 0.6271,
+      "step": 407
+    },
+    {
+      "epoch": 0.1330398630471998,
+      "grad_norm": 0.29895782470703125,
+      "learning_rate": 0.0001285019262469976,
+      "loss": 0.641,
+      "step": 408
+    },
+    {
+      "epoch": 0.13336594114290373,
+      "grad_norm": 0.2666492164134979,
+      "learning_rate": 0.00012820065759001293,
+      "loss": 0.6078,
+      "step": 409
+    },
+    {
+      "epoch": 0.13369201923860766,
+      "grad_norm": 0.28037238121032715,
+      "learning_rate": 0.00012789911060392294,
+      "loss": 0.6904,
+      "step": 410
+    },
+    {
+      "epoch": 0.13401809733431158,
+      "grad_norm": 0.34532639384269714,
+      "learning_rate": 0.0001275972882648746,
+      "loss": 0.6884,
+      "step": 411
+    },
+    {
+      "epoch": 0.13434417543001548,
+      "grad_norm": 0.2206190675497055,
+      "learning_rate": 0.00012729519355173254,
+      "loss": 0.5713,
+      "step": 412
+    },
+    {
+      "epoch": 0.1346702535257194,
+      "grad_norm": 0.2744779586791992,
+      "learning_rate": 0.00012699282944604967,
+      "loss": 0.6322,
+      "step": 413
+    },
+    {
+      "epoch": 0.13499633162142333,
+      "grad_norm": 0.2511173486709595,
+      "learning_rate": 0.00012669019893203759,
+      "loss": 0.7319,
+      "step": 414
+    },
+    {
+      "epoch": 0.13532240971712725,
+      "grad_norm": 0.24852187931537628,
+      "learning_rate": 0.0001263873049965373,
+      "loss": 0.6683,
+      "step": 415
+    },
+    {
+      "epoch": 0.13564848781283118,
+      "grad_norm": 0.2592228651046753,
+      "learning_rate": 0.00012608415062898972,
+      "loss": 0.5839,
+      "step": 416
+    },
+    {
+      "epoch": 0.1359745659085351,
+      "grad_norm": 0.34304317831993103,
+      "learning_rate": 0.000125780738821406,
+      "loss": 0.7071,
+      "step": 417
+    },
+    {
+      "epoch": 0.13630064400423902,
+      "grad_norm": 0.3506695628166199,
+      "learning_rate": 0.00012547707256833823,
+      "loss": 0.6805,
+      "step": 418
+    },
+    {
+      "epoch": 0.13662672209994295,
+      "grad_norm": 0.2388596534729004,
+      "learning_rate": 0.00012517315486684972,
+      "loss": 0.5318,
+      "step": 419
+    },
+    {
+      "epoch": 0.13695280019564685,
+      "grad_norm": 0.22909346222877502,
+      "learning_rate": 0.0001248689887164855,
+      "loss": 0.4896,
+      "step": 420
+    },
+    {
+      "epoch": 0.13727887829135077,
+      "grad_norm": 0.2821944057941437,
+      "learning_rate": 0.00012456457711924266,
+      "loss": 0.7472,
+      "step": 421
+    },
+    {
+      "epoch": 0.1376049563870547,
+      "grad_norm": 0.3464897572994232,
+      "learning_rate": 0.00012425992307954075,
+      "loss": 0.6315,
+      "step": 422
+    },
+    {
+      "epoch": 0.13793103448275862,
+      "grad_norm": 0.513347864151001,
+      "learning_rate": 0.0001239550296041922,
+      "loss": 0.84,
+      "step": 423
+    },
+    {
+      "epoch": 0.13825711257846254,
+      "grad_norm": 0.2560180425643921,
+      "learning_rate": 0.00012364989970237248,
+      "loss": 0.647,
+      "step": 424
+    },
+    {
+      "epoch": 0.13858319067416647,
+      "grad_norm": 0.3197629749774933,
+      "learning_rate": 0.00012334453638559057,
+      "loss": 0.7108,
+      "step": 425
+    },
+    {
+      "epoch": 0.1389092687698704,
+      "grad_norm": 0.23635463416576385,
+      "learning_rate": 0.00012303894266765908,
+      "loss": 0.6198,
+      "step": 426
+    },
+    {
+      "epoch": 0.13923534686557432,
+      "grad_norm": 0.26680633425712585,
+      "learning_rate": 0.00012273312156466464,
+      "loss": 0.6366,
+      "step": 427
+    },
+    {
+      "epoch": 0.1395614249612782,
+      "grad_norm": 0.23483791947364807,
+      "learning_rate": 0.00012242707609493814,
+      "loss": 0.7021,
+      "step": 428
+    },
+    {
+      "epoch": 0.13988750305698214,
+      "grad_norm": 0.3056071996688843,
+      "learning_rate": 0.00012212080927902474,
+      "loss": 0.632,
+      "step": 429
+    },
+    {
+      "epoch": 0.14021358115268606,
+      "grad_norm": 0.24552787840366364,
+      "learning_rate": 0.00012181432413965428,
+      "loss": 0.5489,
+      "step": 430
+    },
+    {
+      "epoch": 0.14053965924839,
+      "grad_norm": 0.32417768239974976,
+      "learning_rate": 0.00012150762370171136,
+      "loss": 0.5433,
+      "step": 431
+    },
+    {
+      "epoch": 0.1408657373440939,
+      "grad_norm": 0.2855895161628723,
+      "learning_rate": 0.00012120071099220549,
+      "loss": 0.6434,
+      "step": 432
+    },
+    {
+      "epoch": 0.14119181543979784,
+      "grad_norm": 0.26909559965133667,
+      "learning_rate": 0.00012089358904024117,
+      "loss": 0.6011,
+      "step": 433
+    },
+    {
+      "epoch": 0.14151789353550176,
+      "grad_norm": 0.3203361928462982,
+      "learning_rate": 0.00012058626087698814,
+      "loss": 0.5739,
+      "step": 434
+    },
+    {
+      "epoch": 0.14184397163120568,
+      "grad_norm": 0.2709220349788666,
+      "learning_rate": 0.00012027872953565125,
+      "loss": 0.6286,
+      "step": 435
+    },
+    {
+      "epoch": 0.1421700497269096,
+      "grad_norm": 0.26857098937034607,
+      "learning_rate": 0.00011997099805144069,
+      "loss": 0.5723,
+      "step": 436
+    },
+    {
+      "epoch": 0.1424961278226135,
+      "grad_norm": 0.24100883305072784,
+      "learning_rate": 0.000119663069461542,
+      "loss": 0.5598,
+      "step": 437
+    },
+    {
+      "epoch": 0.14282220591831743,
+      "grad_norm": 0.21419620513916016,
+      "learning_rate": 0.00011935494680508606,
+      "loss": 0.5755,
+      "step": 438
+    },
+    {
+      "epoch": 0.14314828401402135,
+      "grad_norm": 0.28282371163368225,
+      "learning_rate": 0.00011904663312311901,
+      "loss": 0.6764,
+      "step": 439
+    },
+    {
+      "epoch": 0.14347436210972528,
+      "grad_norm": 0.18891361355781555,
+      "learning_rate": 0.00011873813145857249,
+      "loss": 0.5354,
+      "step": 440
+    },
+    {
+      "epoch": 0.1438004402054292,
+      "grad_norm": 0.245451420545578,
+      "learning_rate": 0.00011842944485623335,
+      "loss": 0.619,
+      "step": 441
+    },
+    {
+      "epoch": 0.14412651830113313,
+      "grad_norm": 0.2450558841228485,
+      "learning_rate": 0.00011812057636271374,
+      "loss": 0.7004,
+      "step": 442
+    },
+    {
+      "epoch": 0.14445259639683705,
+      "grad_norm": 0.27985724806785583,
+      "learning_rate": 0.000117811529026421,
+      "loss": 0.628,
+      "step": 443
+    },
+    {
+      "epoch": 0.14477867449254098,
+      "grad_norm": 0.2647560238838196,
+      "learning_rate": 0.00011750230589752762,
+      "loss": 0.6445,
+      "step": 444
+    },
+    {
+      "epoch": 0.14510475258824487,
+      "grad_norm": 0.28567782044410706,
+      "learning_rate": 0.00011719291002794096,
+      "loss": 0.5969,
+      "step": 445
+    },
+    {
+      "epoch": 0.1454308306839488,
+      "grad_norm": 0.26054808497428894,
+      "learning_rate": 0.00011688334447127338,
+      "loss": 0.6482,
+      "step": 446
+    },
+    {
+      "epoch": 0.14575690877965272,
+      "grad_norm": 0.2926684617996216,
+      "learning_rate": 0.00011657361228281199,
+      "loss": 0.5566,
+      "step": 447
+    },
+    {
+      "epoch": 0.14608298687535665,
+      "grad_norm": 0.24087432026863098,
+      "learning_rate": 0.00011626371651948838,
+      "loss": 0.651,
+      "step": 448
+    },
+    {
+      "epoch": 0.14640906497106057,
+      "grad_norm": 0.1882346272468567,
+      "learning_rate": 0.00011595366023984864,
+      "loss": 0.5317,
+      "step": 449
+    },
+    {
+      "epoch": 0.1467351430667645,
+      "grad_norm": 0.26859766244888306,
+      "learning_rate": 0.0001156434465040231,
+      "loss": 0.6414,
+      "step": 450
+    },
+    {
+      "epoch": 0.14706122116246842,
+      "grad_norm": 0.264344185590744,
+      "learning_rate": 0.00011533307837369607,
+      "loss": 0.5283,
+      "step": 451
+    },
+    {
+      "epoch": 0.14738729925817234,
+      "grad_norm": 0.2349400818347931,
+      "learning_rate": 0.00011502255891207572,
+      "loss": 0.5494,
+      "step": 452
+    },
+    {
+      "epoch": 0.14771337735387624,
+      "grad_norm": 0.23789772391319275,
+      "learning_rate": 0.00011471189118386375,
+      "loss": 0.6511,
+      "step": 453
+    },
+    {
+      "epoch": 0.14803945544958017,
+      "grad_norm": 0.2902562916278839,
+      "learning_rate": 0.00011440107825522521,
+      "loss": 0.5956,
+      "step": 454
+    },
+    {
+      "epoch": 0.1483655335452841,
+      "grad_norm": 0.33084043860435486,
+      "learning_rate": 0.00011409012319375827,
+      "loss": 0.7127,
+      "step": 455
+    },
+    {
+      "epoch": 0.14869161164098801,
+      "grad_norm": 0.2966511845588684,
+      "learning_rate": 0.0001137790290684638,
+      "loss": 0.6293,
+      "step": 456
+    },
+    {
+      "epoch": 0.14901768973669194,
+      "grad_norm": 0.2903098464012146,
+      "learning_rate": 0.00011346779894971527,
+      "loss": 0.5653,
+      "step": 457
+    },
+    {
+      "epoch": 0.14934376783239586,
+      "grad_norm": 0.30501604080200195,
+      "learning_rate": 0.00011315643590922827,
+      "loss": 0.6074,
+      "step": 458
+    },
+    {
+      "epoch": 0.1496698459280998,
+      "grad_norm": 0.24236589670181274,
+      "learning_rate": 0.0001128449430200303,
+      "loss": 0.5867,
+      "step": 459
+    },
+    {
+      "epoch": 0.1499959240238037,
+      "grad_norm": 0.2912566363811493,
+      "learning_rate": 0.00011253332335643043,
+      "loss": 0.6263,
+      "step": 460
+    },
+    {
+      "epoch": 0.1503220021195076,
+      "grad_norm": 0.2772620916366577,
+      "learning_rate": 0.00011222157999398895,
+      "loss": 0.6283,
+      "step": 461
+    },
+    {
+      "epoch": 0.15064808021521153,
+      "grad_norm": 0.2634068727493286,
+      "learning_rate": 0.00011190971600948699,
+      "loss": 0.6163,
+      "step": 462
+    },
+    {
+      "epoch": 0.15097415831091546,
+      "grad_norm": 0.2546554207801819,
+      "learning_rate": 0.00011159773448089614,
+      "loss": 0.6658,
+      "step": 463
+    },
+    {
+      "epoch": 0.15130023640661938,
+      "grad_norm": 0.24717013537883759,
+      "learning_rate": 0.00011128563848734816,
+      "loss": 0.6137,
+      "step": 464
+    },
+    {
+      "epoch": 0.1516263145023233,
+      "grad_norm": 0.3441718816757202,
+      "learning_rate": 0.00011097343110910452,
+      "loss": 0.6057,
+      "step": 465
+    },
+    {
+      "epoch": 0.15195239259802723,
+      "grad_norm": 0.24183110892772675,
+      "learning_rate": 0.000110661115427526,
+      "loss": 0.4939,
+      "step": 466
+    },
+    {
+      "epoch": 0.15227847069373115,
+      "grad_norm": 0.30775824189186096,
+      "learning_rate": 0.00011034869452504226,
+      "loss": 0.6185,
+      "step": 467
+    },
+    {
+      "epoch": 0.15260454878943508,
+      "grad_norm": 0.28553271293640137,
+      "learning_rate": 0.00011003617148512149,
+      "loss": 0.6158,
+      "step": 468
+    },
+    {
+      "epoch": 0.152930626885139,
+      "grad_norm": 0.2486485242843628,
+      "learning_rate": 0.00010972354939223996,
+      "loss": 0.5079,
+      "step": 469
+    },
+    {
+      "epoch": 0.1532567049808429,
+      "grad_norm": 0.30782243609428406,
+      "learning_rate": 0.00010941083133185146,
+      "loss": 0.5769,
+      "step": 470
+    },
+    {
+      "epoch": 0.15358278307654682,
+      "grad_norm": 0.3133150339126587,
+      "learning_rate": 0.00010909802039035701,
+      "loss": 0.679,
+      "step": 471
+    },
+    {
+      "epoch": 0.15390886117225075,
+      "grad_norm": 0.23745952546596527,
+      "learning_rate": 0.00010878511965507434,
+      "loss": 0.487,
+      "step": 472
+    },
+    {
+      "epoch": 0.15423493926795467,
+      "grad_norm": 0.29805904626846313,
+      "learning_rate": 0.00010847213221420736,
+      "loss": 0.7348,
+      "step": 473
+    },
+    {
+      "epoch": 0.1545610173636586,
+      "grad_norm": 0.22502878308296204,
+      "learning_rate": 0.00010815906115681578,
+      "loss": 0.5491,
+      "step": 474
+    },
+    {
+      "epoch": 0.15488709545936252,
+      "grad_norm": 0.4142134487628937,
+      "learning_rate": 0.0001078459095727845,
+      "loss": 0.644,
+      "step": 475
+    },
+    {
+      "epoch": 0.15521317355506645,
+      "grad_norm": 0.23912709951400757,
+      "learning_rate": 0.00010753268055279329,
+      "loss": 0.6337,
+      "step": 476
+    },
+    {
+      "epoch": 0.15553925165077037,
+      "grad_norm": 0.2665061354637146,
+      "learning_rate": 0.0001072193771882861,
+      "loss": 0.5108,
+      "step": 477
+    },
+    {
+      "epoch": 0.15586532974647427,
+      "grad_norm": 0.276652991771698,
+      "learning_rate": 0.00010690600257144061,
+      "loss": 0.6463,
+      "step": 478
+    },
+    {
+      "epoch": 0.1561914078421782,
+      "grad_norm": 0.2724633812904358,
+      "learning_rate": 0.0001065925597951378,
+      "loss": 0.6021,
+      "step": 479
+    },
+    {
+      "epoch": 0.15651748593788212,
+      "grad_norm": 0.2907373309135437,
+      "learning_rate": 0.00010627905195293135,
+      "loss": 0.68,
+      "step": 480
+    },
+    {
+      "epoch": 0.15684356403358604,
+      "grad_norm": 0.2480301856994629,
+      "learning_rate": 0.00010596548213901708,
+      "loss": 0.5805,
+      "step": 481
+    },
+    {
+      "epoch": 0.15716964212928997,
+      "grad_norm": 0.31199586391448975,
+      "learning_rate": 0.00010565185344820247,
+      "loss": 0.6244,
+      "step": 482
+    },
+    {
+      "epoch": 0.1574957202249939,
+      "grad_norm": 0.23156419396400452,
+      "learning_rate": 0.00010533816897587606,
+      "loss": 0.6027,
+      "step": 483
+    },
+    {
+      "epoch": 0.15782179832069781,
+      "grad_norm": 0.1928403526544571,
+      "learning_rate": 0.00010502443181797697,
+      "loss": 0.4903,
+      "step": 484
+    },
+    {
+      "epoch": 0.15814787641640174,
+      "grad_norm": 0.2800855338573456,
+      "learning_rate": 0.00010471064507096426,
+      "loss": 0.6735,
+      "step": 485
+    },
+    {
+      "epoch": 0.15847395451210564,
+      "grad_norm": 0.325988233089447,
+      "learning_rate": 0.0001043968118317865,
+      "loss": 0.6988,
+      "step": 486
+    },
+    {
+      "epoch": 0.15880003260780956,
+      "grad_norm": 0.26524752378463745,
+      "learning_rate": 0.00010408293519785101,
+      "loss": 0.6748,
+      "step": 487
+    },
+    {
+      "epoch": 0.15912611070351348,
+      "grad_norm": 0.368566632270813,
+      "learning_rate": 0.00010376901826699348,
+      "loss": 0.6704,
+      "step": 488
+    },
+    {
+      "epoch": 0.1594521887992174,
+      "grad_norm": 0.36456090211868286,
+      "learning_rate": 0.00010345506413744726,
+      "loss": 0.575,
+      "step": 489
+    },
+    {
+      "epoch": 0.15977826689492133,
+      "grad_norm": 0.2953360974788666,
+      "learning_rate": 0.00010314107590781284,
+      "loss": 0.5831,
+      "step": 490
+    },
+    {
+      "epoch": 0.16010434499062526,
+      "grad_norm": 0.22688160836696625,
+      "learning_rate": 0.00010282705667702734,
+      "loss": 0.5816,
+      "step": 491
+    },
+    {
+      "epoch": 0.16043042308632918,
+      "grad_norm": 0.34496596455574036,
+      "learning_rate": 0.00010251300954433376,
+      "loss": 0.6377,
+      "step": 492
+    },
+    {
+      "epoch": 0.1607565011820331,
+      "grad_norm": 0.249411940574646,
+      "learning_rate": 0.00010219893760925052,
+      "loss": 0.5299,
+      "step": 493
+    },
+    {
+      "epoch": 0.16108257927773703,
+      "grad_norm": 0.33367717266082764,
+      "learning_rate": 0.00010188484397154084,
+      "loss": 0.5045,
+      "step": 494
+    },
+    {
+      "epoch": 0.16140865737344093,
+      "grad_norm": 0.22329236567020416,
+      "learning_rate": 0.00010157073173118208,
+      "loss": 0.5531,
+      "step": 495
+    },
+    {
+      "epoch": 0.16173473546914485,
+      "grad_norm": 0.2595437467098236,
+      "learning_rate": 0.00010125660398833528,
+      "loss": 0.5787,
+      "step": 496
+    },
+    {
+      "epoch": 0.16206081356484878,
+      "grad_norm": 0.30315423011779785,
+      "learning_rate": 0.00010094246384331442,
+      "loss": 0.6634,
+      "step": 497
+    },
+    {
+      "epoch": 0.1623868916605527,
+      "grad_norm": 0.3824993073940277,
+      "learning_rate": 0.00010062831439655591,
+      "loss": 0.6479,
+      "step": 498
+    },
+    {
+      "epoch": 0.16271296975625663,
+      "grad_norm": 0.28295570611953735,
+      "learning_rate": 0.00010031415874858797,
+      "loss": 0.5965,
+      "step": 499
+    },
+    {
+      "epoch": 0.16303904785196055,
+      "grad_norm": 0.26669177412986755,
+      "learning_rate": 0.0001,
+      "loss": 0.5936,
+      "step": 500
+    },
+    {
+      "epoch": 0.16336512594766447,
+      "grad_norm": 0.2724348306655884,
+      "learning_rate": 9.968584125141204e-05,
+      "loss": 0.655,
+      "step": 501
+    },
+    {
+      "epoch": 0.1636912040433684,
+      "grad_norm": 0.2577010691165924,
+      "learning_rate": 9.937168560344412e-05,
+      "loss": 0.5863,
+      "step": 502
+    },
+    {
+      "epoch": 0.1640172821390723,
+      "grad_norm": 0.2395378202199936,
+      "learning_rate": 9.90575361566856e-05,
+      "loss": 0.5955,
+      "step": 503
+    },
+    {
+      "epoch": 0.16434336023477622,
+      "grad_norm": 0.2337670922279358,
+      "learning_rate": 9.874339601166473e-05,
+      "loss": 0.5532,
+      "step": 504
+    },
+    {
+      "epoch": 0.16466943833048014,
+      "grad_norm": 0.23461255431175232,
+      "learning_rate": 9.842926826881796e-05,
+      "loss": 0.4932,
+      "step": 505
+    },
+    {
+      "epoch": 0.16499551642618407,
+      "grad_norm": 0.3790539503097534,
+      "learning_rate": 9.81151560284592e-05,
+      "loss": 0.6278,
+      "step": 506
+    },
+    {
+      "epoch": 0.165321594521888,
+      "grad_norm": 0.24186423420906067,
+      "learning_rate": 9.78010623907495e-05,
+      "loss": 0.6043,
+      "step": 507
+    },
+    {
+      "epoch": 0.16564767261759192,
+      "grad_norm": 0.18181544542312622,
+      "learning_rate": 9.748699045566626e-05,
+      "loss": 0.5098,
+      "step": 508
+    },
+    {
+      "epoch": 0.16597375071329584,
+      "grad_norm": 0.26670199632644653,
+      "learning_rate": 9.717294332297268e-05,
+      "loss": 0.5501,
+      "step": 509
+    },
+    {
+      "epoch": 0.16629982880899977,
+      "grad_norm": 0.31869029998779297,
+      "learning_rate": 9.685892409218717e-05,
+      "loss": 0.6554,
+      "step": 510
+    },
+    {
+      "epoch": 0.16662590690470366,
+      "grad_norm": 0.2595357298851013,
+      "learning_rate": 9.654493586255278e-05,
+      "loss": 0.603,
+      "step": 511
+    },
+    {
+      "epoch": 0.1669519850004076,
+      "grad_norm": 0.23499582707881927,
+      "learning_rate": 9.623098173300654e-05,
+      "loss": 0.58,
+      "step": 512
+    },
+    {
+      "epoch": 0.1672780630961115,
+      "grad_norm": 0.2135399580001831,
+      "learning_rate": 9.591706480214901e-05,
+      "loss": 0.5097,
+      "step": 513
+    },
+    {
+      "epoch": 0.16760414119181544,
+      "grad_norm": 0.37453222274780273,
+      "learning_rate": 9.560318816821353e-05,
+      "loss": 0.7076,
+      "step": 514
+    },
+    {
+      "epoch": 0.16793021928751936,
+      "grad_norm": 0.23250311613082886,
+      "learning_rate": 9.528935492903575e-05,
+      "loss": 0.5374,
+      "step": 515
+    },
+    {
+      "epoch": 0.16825629738322329,
+      "grad_norm": 0.3057493567466736,
+      "learning_rate": 9.497556818202306e-05,
+      "loss": 0.7232,
+      "step": 516
+    },
+    {
+      "epoch": 0.1685823754789272,
+      "grad_norm": 0.3576625883579254,
+      "learning_rate": 9.466183102412395e-05,
+      "loss": 0.557,
+      "step": 517
+    },
+    {
+      "epoch": 0.16890845357463113,
+      "grad_norm": 0.24760064482688904,
+      "learning_rate": 9.434814655179755e-05,
+      "loss": 0.5358,
+      "step": 518
+    },
+    {
+      "epoch": 0.16923453167033506,
+      "grad_norm": 0.25141358375549316,
+      "learning_rate": 9.403451786098294e-05,
+      "loss": 0.5919,
+      "step": 519
+    },
+    {
+      "epoch": 0.16956060976603896,
+      "grad_norm": 0.31347140669822693,
+      "learning_rate": 9.372094804706867e-05,
+      "loss": 0.6379,
+      "step": 520
+    },
+    {
+      "epoch": 0.16988668786174288,
+      "grad_norm": 0.31334182620048523,
+      "learning_rate": 9.340744020486222e-05,
+      "loss": 0.6163,
+      "step": 521
+    },
+    {
+      "epoch": 0.1702127659574468,
+      "grad_norm": 0.30304384231567383,
+      "learning_rate": 9.309399742855942e-05,
+      "loss": 0.7165,
+      "step": 522
+    },
+    {
+      "epoch": 0.17053884405315073,
+      "grad_norm": 0.48817434906959534,
+      "learning_rate": 9.278062281171393e-05,
+      "loss": 0.6673,
+      "step": 523
+    },
+    {
+      "epoch": 0.17086492214885465,
+      "grad_norm": 0.3089975118637085,
+      "learning_rate": 9.246731944720675e-05,
+      "loss": 0.7226,
+      "step": 524
+    },
+    {
+      "epoch": 0.17119100024455858,
+      "grad_norm": 0.2963530123233795,
+      "learning_rate": 9.215409042721552e-05,
+      "loss": 0.5804,
+      "step": 525
+    },
+    {
+      "epoch": 0.1715170783402625,
+      "grad_norm": 0.2515937387943268,
+      "learning_rate": 9.184093884318425e-05,
+      "loss": 0.5827,
+      "step": 526
+    },
+    {
+      "epoch": 0.17184315643596643,
+      "grad_norm": 0.2320161759853363,
+      "learning_rate": 9.152786778579267e-05,
+      "loss": 0.5849,
+      "step": 527
+    },
+    {
+      "epoch": 0.17216923453167032,
+      "grad_norm": 0.4115968644618988,
+      "learning_rate": 9.121488034492569e-05,
+      "loss": 0.5939,
+      "step": 528
+    },
+    {
+      "epoch": 0.17249531262737425,
+      "grad_norm": 0.24110546708106995,
+      "learning_rate": 9.090197960964301e-05,
+      "loss": 0.5599,
+      "step": 529
+    },
+    {
+      "epoch": 0.17282139072307817,
+      "grad_norm": 0.2787460684776306,
+      "learning_rate": 9.058916866814858e-05,
+      "loss": 0.6877,
+      "step": 530
+    },
+    {
+      "epoch": 0.1731474688187821,
+      "grad_norm": 0.2140059471130371,
+      "learning_rate": 9.027645060776006e-05,
+      "loss": 0.5474,
+      "step": 531
+    },
+    {
+      "epoch": 0.17347354691448602,
+      "grad_norm": 0.20941783487796783,
+      "learning_rate": 8.99638285148785e-05,
+      "loss": 0.6098,
+      "step": 532
+    },
+    {
+      "epoch": 0.17379962501018995,
+      "grad_norm": 0.3070947527885437,
+      "learning_rate": 8.965130547495776e-05,
+      "loss": 0.7024,
+      "step": 533
+    },
+    {
+      "epoch": 0.17412570310589387,
+      "grad_norm": 0.4354031980037689,
+      "learning_rate": 8.933888457247402e-05,
+      "loss": 0.6728,
+      "step": 534
+    },
+    {
+      "epoch": 0.1744517812015978,
+      "grad_norm": 0.22803445160388947,
+      "learning_rate": 8.902656889089548e-05,
+      "loss": 0.6227,
+      "step": 535
+    },
+    {
+      "epoch": 0.1747778592973017,
+      "grad_norm": 0.2915150821208954,
+      "learning_rate": 8.871436151265184e-05,
+      "loss": 0.6151,
+      "step": 536
+    },
+    {
+      "epoch": 0.17510393739300562,
+      "grad_norm": 0.2954872250556946,
+      "learning_rate": 8.840226551910387e-05,
+      "loss": 0.6851,
+      "step": 537
+    },
+    {
+      "epoch": 0.17543001548870954,
+      "grad_norm": 0.2495146542787552,
+      "learning_rate": 8.809028399051302e-05,
+      "loss": 0.5932,
+      "step": 538
+    },
+    {
+      "epoch": 0.17575609358441346,
+      "grad_norm": 0.2785921096801758,
+      "learning_rate": 8.777842000601105e-05,
+      "loss": 0.6403,
+      "step": 539
+    },
+    {
+      "epoch": 0.1760821716801174,
+      "grad_norm": 0.258619487285614,
+      "learning_rate": 8.746667664356956e-05,
+      "loss": 0.5545,
+      "step": 540
+    },
+    {
+      "epoch": 0.1764082497758213,
+      "grad_norm": 0.2323625683784485,
+      "learning_rate": 8.715505697996971e-05,
+      "loss": 0.6349,
+      "step": 541
+    },
+    {
+      "epoch": 0.17673432787152524,
+      "grad_norm": 0.2878716289997101,
+      "learning_rate": 8.684356409077176e-05,
+      "loss": 0.6123,
+      "step": 542
+    },
+    {
+      "epoch": 0.17706040596722916,
+      "grad_norm": 0.5154581665992737,
+      "learning_rate": 8.653220105028474e-05,
+      "loss": 0.8217,
+      "step": 543
+    },
+    {
+      "epoch": 0.1773864840629331,
+      "grad_norm": 0.2401517629623413,
+      "learning_rate": 8.62209709315362e-05,
+      "loss": 0.5884,
+      "step": 544
+    },
+    {
+      "epoch": 0.17771256215863698,
+      "grad_norm": 0.3214757442474365,
+      "learning_rate": 8.590987680624174e-05,
+      "loss": 0.6488,
+      "step": 545
+    },
+    {
+      "epoch": 0.1780386402543409,
+      "grad_norm": 0.2335442304611206,
+      "learning_rate": 8.559892174477479e-05,
+      "loss": 0.5544,
+      "step": 546
+    },
+    {
+      "epoch": 0.17836471835004483,
+      "grad_norm": 0.23834139108657837,
+      "learning_rate": 8.528810881613626e-05,
+      "loss": 0.5983,
+      "step": 547
+    },
+    {
+      "epoch": 0.17869079644574876,
+      "grad_norm": 0.23438434302806854,
+      "learning_rate": 8.497744108792429e-05,
+      "loss": 0.566,
+      "step": 548
+    },
+    {
+      "epoch": 0.17901687454145268,
+      "grad_norm": 0.37170785665512085,
+      "learning_rate": 8.466692162630392e-05,
+      "loss": 0.6824,
+      "step": 549
+    },
+    {
+      "epoch": 0.1793429526371566,
+      "grad_norm": 0.263026624917984,
+      "learning_rate": 8.435655349597689e-05,
+      "loss": 0.6363,
+      "step": 550
+    },
+    {
+      "epoch": 0.17966903073286053,
+      "grad_norm": 0.3249225616455078,
+      "learning_rate": 8.404633976015134e-05,
+      "loss": 0.618,
+      "step": 551
+    },
+    {
+      "epoch": 0.17999510882856445,
+      "grad_norm": 0.26690465211868286,
+      "learning_rate": 8.373628348051165e-05,
+      "loss": 0.6547,
+      "step": 552
+    },
+    {
+      "epoch": 0.18032118692426835,
+      "grad_norm": 0.22415734827518463,
+      "learning_rate": 8.342638771718802e-05,
+      "loss": 0.5189,
+      "step": 553
+    },
+    {
+      "epoch": 0.18064726501997228,
+      "grad_norm": 0.21609823405742645,
+      "learning_rate": 8.311665552872662e-05,
+      "loss": 0.6112,
+      "step": 554
+    },
+    {
+      "epoch": 0.1809733431156762,
+      "grad_norm": 0.3650791049003601,
+      "learning_rate": 8.280708997205904e-05,
+      "loss": 0.6109,
+      "step": 555
+    },
+    {
+      "epoch": 0.18129942121138012,
+      "grad_norm": 0.2111487239599228,
+      "learning_rate": 8.249769410247239e-05,
+      "loss": 0.4548,
+      "step": 556
+    },
+    {
+      "epoch": 0.18162549930708405,
+      "grad_norm": 0.3168624937534332,
+      "learning_rate": 8.218847097357898e-05,
+      "loss": 0.6538,
+      "step": 557
+    },
+    {
+      "epoch": 0.18195157740278797,
+      "grad_norm": 0.27414387464523315,
+      "learning_rate": 8.187942363728625e-05,
+      "loss": 0.688,
+      "step": 558
+    },
+    {
+      "epoch": 0.1822776554984919,
+      "grad_norm": 0.29763197898864746,
+      "learning_rate": 8.157055514376666e-05,
+      "loss": 0.673,
+      "step": 559
+    },
+    {
+      "epoch": 0.18260373359419582,
+      "grad_norm": 0.26111218333244324,
+      "learning_rate": 8.126186854142752e-05,
+      "loss": 0.6391,
+      "step": 560
+    },
+    {
+      "epoch": 0.18292981168989972,
+      "grad_norm": 0.34893324971199036,
+      "learning_rate": 8.095336687688102e-05,
+      "loss": 0.638,
+      "step": 561
+    },
+    {
+      "epoch": 0.18325588978560364,
+      "grad_norm": 0.279060423374176,
+      "learning_rate": 8.064505319491398e-05,
+      "loss": 0.6277,
+      "step": 562
+    },
+    {
+      "epoch": 0.18358196788130757,
+      "grad_norm": 0.30398038029670715,
+      "learning_rate": 8.033693053845801e-05,
+      "loss": 0.5687,
+      "step": 563
+    },
+    {
+      "epoch": 0.1839080459770115,
+      "grad_norm": 0.2892376780509949,
+      "learning_rate": 8.002900194855932e-05,
+      "loss": 0.6962,
+      "step": 564
+    },
+    {
+      "epoch": 0.18423412407271542,
+      "grad_norm": 0.34928885102272034,
+      "learning_rate": 7.972127046434878e-05,
+      "loss": 0.747,
+      "step": 565
+    },
+    {
+      "epoch": 0.18456020216841934,
+      "grad_norm": 0.27299997210502625,
+      "learning_rate": 7.941373912301189e-05,
+      "loss": 0.6286,
+      "step": 566
+    },
+    {
+      "epoch": 0.18488628026412326,
+      "grad_norm": 0.34455394744873047,
+      "learning_rate": 7.910641095975886e-05,
+      "loss": 0.667,
+      "step": 567
+    },
+    {
+      "epoch": 0.1852123583598272,
+      "grad_norm": 0.2987124025821686,
+      "learning_rate": 7.879928900779456e-05,
+      "loss": 0.7059,
+      "step": 568
+    },
+    {
+      "epoch": 0.18553843645553109,
+      "grad_norm": 0.24651587009429932,
+      "learning_rate": 7.849237629828869e-05,
+      "loss": 0.5836,
+      "step": 569
+    },
+    {
+      "epoch": 0.185864514551235,
+      "grad_norm": 0.24579085409641266,
+      "learning_rate": 7.818567586034577e-05,
+      "loss": 0.5498,
+      "step": 570
+    },
+    {
+      "epoch": 0.18619059264693893,
+      "grad_norm": 0.31513264775276184,
+      "learning_rate": 7.787919072097531e-05,
+      "loss": 0.6475,
+      "step": 571
+    },
+    {
+      "epoch": 0.18651667074264286,
+      "grad_norm": 0.2718188464641571,
+      "learning_rate": 7.75729239050619e-05,
+      "loss": 0.6896,
+      "step": 572
+    },
+    {
+      "epoch": 0.18684274883834678,
+      "grad_norm": 0.3963709771633148,
+      "learning_rate": 7.726687843533538e-05,
+      "loss": 0.7449,
+      "step": 573
+    },
+    {
+      "epoch": 0.1871688269340507,
+      "grad_norm": 0.2776338756084442,
+      "learning_rate": 7.696105733234098e-05,
+      "loss": 0.6492,
+      "step": 574
+    },
+    {
+      "epoch": 0.18749490502975463,
+      "grad_norm": 0.20792852342128754,
+      "learning_rate": 7.66554636144095e-05,
+      "loss": 0.536,
+      "step": 575
+    },
+    {
+      "epoch": 0.18782098312545856,
+      "grad_norm": 0.29838401079177856,
+      "learning_rate": 7.635010029762756e-05,
+      "loss": 0.6181,
+      "step": 576
+    },
+    {
+      "epoch": 0.18814706122116248,
+      "grad_norm": 0.2557230293750763,
+      "learning_rate": 7.604497039580785e-05,
+      "loss": 0.5459,
+      "step": 577
+    },
+    {
+      "epoch": 0.18847313931686638,
+      "grad_norm": 0.283884733915329,
+      "learning_rate": 7.574007692045928e-05,
+      "loss": 0.5817,
+      "step": 578
+    },
+    {
+      "epoch": 0.1887992174125703,
+      "grad_norm": 0.26855558156967163,
+      "learning_rate": 7.543542288075739e-05,
+      "loss": 0.6018,
+      "step": 579
+    },
+    {
+      "epoch": 0.18912529550827423,
+      "grad_norm": 0.2804105579853058,
+      "learning_rate": 7.513101128351454e-05,
+      "loss": 0.5872,
+      "step": 580
+    },
+    {
+      "epoch": 0.18945137360397815,
+      "grad_norm": 0.40271496772766113,
+      "learning_rate": 7.48268451331503e-05,
+      "loss": 0.7282,
+      "step": 581
+    },
+    {
+      "epoch": 0.18977745169968208,
+      "grad_norm": 0.27677327394485474,
+      "learning_rate": 7.45229274316618e-05,
+      "loss": 0.7078,
+      "step": 582
+    },
+    {
+      "epoch": 0.190103529795386,
+      "grad_norm": 0.28340569138526917,
+      "learning_rate": 7.421926117859403e-05,
+      "loss": 0.5747,
+      "step": 583
+    },
+    {
+      "epoch": 0.19042960789108992,
+      "grad_norm": 0.2599877715110779,
+      "learning_rate": 7.391584937101033e-05,
+      "loss": 0.5895,
+      "step": 584
+    },
+    {
+      "epoch": 0.19075568598679385,
+      "grad_norm": 0.2972414791584015,
+      "learning_rate": 7.361269500346274e-05,
+      "loss": 0.5901,
+      "step": 585
+    },
+    {
+      "epoch": 0.19108176408249775,
+      "grad_norm": 0.2536401152610779,
+      "learning_rate": 7.330980106796246e-05,
+      "loss": 0.5877,
+      "step": 586
+    },
+    {
+      "epoch": 0.19140784217820167,
+      "grad_norm": 0.2621424198150635,
+      "learning_rate": 7.300717055395039e-05,
+      "loss": 0.7006,
+      "step": 587
+    },
+    {
+      "epoch": 0.1917339202739056,
+      "grad_norm": 0.3111814558506012,
+      "learning_rate": 7.270480644826749e-05,
+      "loss": 0.6821,
+      "step": 588
+    },
+    {
+      "epoch": 0.19205999836960952,
+      "grad_norm": 0.3220475912094116,
+      "learning_rate": 7.240271173512546e-05,
+      "loss": 0.6543,
+      "step": 589
+    },
+    {
+      "epoch": 0.19238607646531344,
+      "grad_norm": 0.3028581142425537,
+      "learning_rate": 7.210088939607708e-05,
+      "loss": 0.6336,
+      "step": 590
+    },
+    {
+      "epoch": 0.19271215456101737,
+      "grad_norm": 0.2866508364677429,
+      "learning_rate": 7.179934240998706e-05,
+      "loss": 0.6173,
+      "step": 591
+    },
+    {
+      "epoch": 0.1930382326567213,
+      "grad_norm": 0.26337945461273193,
+      "learning_rate": 7.149807375300239e-05,
+      "loss": 0.5194,
+      "step": 592
+    },
+    {
+      "epoch": 0.19336431075242522,
+      "grad_norm": 0.31201040744781494,
+      "learning_rate": 7.119708639852312e-05,
+      "loss": 0.5724,
+      "step": 593
+    },
+    {
+      "epoch": 0.1936903888481291,
+      "grad_norm": 0.2324737012386322,
+      "learning_rate": 7.089638331717284e-05,
+      "loss": 0.562,
+      "step": 594
+    },
+    {
+      "epoch": 0.19401646694383304,
+      "grad_norm": 0.3376362919807434,
+      "learning_rate": 7.059596747676962e-05,
+      "loss": 0.5773,
+      "step": 595
+    },
+    {
+      "epoch": 0.19434254503953696,
+      "grad_norm": 0.26487740874290466,
+      "learning_rate": 7.029584184229653e-05,
+      "loss": 0.5336,
+      "step": 596
+    },
+    {
+      "epoch": 0.1946686231352409,
+      "grad_norm": 0.35700923204421997,
+      "learning_rate": 6.999600937587239e-05,
+      "loss": 0.592,
+      "step": 597
+    },
+    {
+      "epoch": 0.1949947012309448,
+      "grad_norm": 0.33490103483200073,
+      "learning_rate": 6.969647303672262e-05,
+      "loss": 0.6538,
+      "step": 598
+    },
+    {
+      "epoch": 0.19532077932664874,
+      "grad_norm": 0.27960991859436035,
+      "learning_rate": 6.939723578114993e-05,
+      "loss": 0.5769,
+      "step": 599
+    },
+    {
+      "epoch": 0.19564685742235266,
+      "grad_norm": 0.2691517472267151,
+      "learning_rate": 6.909830056250527e-05,
+      "loss": 0.6472,
+      "step": 600
+    },
+    {
+      "epoch": 0.19597293551805658,
+      "grad_norm": 0.2690907120704651,
+      "learning_rate": 6.879967033115853e-05,
+      "loss": 0.6053,
+      "step": 601
+    },
+    {
+      "epoch": 0.1962990136137605,
+      "grad_norm": 0.2998523712158203,
+      "learning_rate": 6.850134803446954e-05,
+      "loss": 0.6843,
+      "step": 602
+    },
+    {
+      "epoch": 0.1966250917094644,
+      "grad_norm": 0.2649098336696625,
+      "learning_rate": 6.820333661675893e-05,
+      "loss": 0.4998,
+      "step": 603
+    },
+    {
+      "epoch": 0.19695116980516833,
+      "grad_norm": 0.2461191862821579,
+      "learning_rate": 6.790563901927907e-05,
+      "loss": 0.5161,
+      "step": 604
+    },
+    {
+      "epoch": 0.19727724790087225,
+      "grad_norm": 0.221087247133255,
+      "learning_rate": 6.760825818018508e-05,
+      "loss": 0.5993,
+      "step": 605
+    },
+    {
+      "epoch": 0.19760332599657618,
+      "grad_norm": 0.3097057640552521,
+      "learning_rate": 6.731119703450577e-05,
+      "loss": 0.6602,
+      "step": 606
+    },
+    {
+      "epoch": 0.1979294040922801,
+      "grad_norm": 0.27952730655670166,
+      "learning_rate": 6.701445851411472e-05,
+      "loss": 0.6358,
+      "step": 607
+    },
+    {
+      "epoch": 0.19825548218798403,
+      "grad_norm": 0.29174724221229553,
+      "learning_rate": 6.671804554770135e-05,
+      "loss": 0.6346,
+      "step": 608
+    },
+    {
+      "epoch": 0.19858156028368795,
+      "grad_norm": 0.3168889582157135,
+      "learning_rate": 6.642196106074194e-05,
+      "loss": 0.7173,
+      "step": 609
+    },
+    {
+      "epoch": 0.19890763837939188,
+      "grad_norm": 0.3026885390281677,
+      "learning_rate": 6.612620797547087e-05,
+      "loss": 0.6518,
+      "step": 610
+    },
+    {
+      "epoch": 0.19923371647509577,
+      "grad_norm": 0.27130576968193054,
+      "learning_rate": 6.583078921085167e-05,
+      "loss": 0.5877,
+      "step": 611
+    },
+    {
+      "epoch": 0.1995597945707997,
+      "grad_norm": 0.2264910638332367,
+      "learning_rate": 6.55357076825483e-05,
+      "loss": 0.5495,
+      "step": 612
+    },
+    {
+      "epoch": 0.19988587266650362,
+      "grad_norm": 0.2999652326107025,
+      "learning_rate": 6.52409663028963e-05,
+      "loss": 0.7243,
+      "step": 613
+    },
+    {
+      "epoch": 0.20021195076220755,
+      "grad_norm": 0.3234879970550537,
+      "learning_rate": 6.494656798087412e-05,
+      "loss": 0.5561,
+      "step": 614
+    },
+    {
+      "epoch": 0.20053802885791147,
+      "grad_norm": 0.3255922198295593,
+      "learning_rate": 6.465251562207431e-05,
+      "loss": 0.7417,
+      "step": 615
+    },
+    {
+      "epoch": 0.2008641069536154,
+      "grad_norm": 0.3037979304790497,
+      "learning_rate": 6.435881212867493e-05,
+      "loss": 0.6452,
+      "step": 616
+    },
+    {
+      "epoch": 0.20119018504931932,
+      "grad_norm": 0.3250497281551361,
+      "learning_rate": 6.406546039941094e-05,
+      "loss": 0.6424,
+      "step": 617
+    },
+    {
+      "epoch": 0.20151626314502324,
+      "grad_norm": 0.28922873735427856,
+      "learning_rate": 6.377246332954544e-05,
+      "loss": 0.5418,
+      "step": 618
+    },
+    {
+      "epoch": 0.20184234124072714,
+      "grad_norm": 0.31836992502212524,
+      "learning_rate": 6.347982381084123e-05,
+      "loss": 0.7,
+      "step": 619
+    },
+    {
+      "epoch": 0.20216841933643107,
+      "grad_norm": 0.242594376206398,
+      "learning_rate": 6.318754473153221e-05,
+      "loss": 0.5934,
+      "step": 620
+    },
+    {
+      "epoch": 0.202494497432135,
+      "grad_norm": 0.2678995132446289,
+      "learning_rate": 6.289562897629492e-05,
+      "loss": 0.6335,
+      "step": 621
+    },
+    {
+      "epoch": 0.20282057552783891,
+      "grad_norm": 0.2010084092617035,
+      "learning_rate": 6.260407942621998e-05,
+      "loss": 0.4732,
+      "step": 622
+    },
+    {
+      "epoch": 0.20314665362354284,
+      "grad_norm": 0.2803085446357727,
+      "learning_rate": 6.231289895878375e-05,
+      "loss": 0.6144,
+      "step": 623
+    },
+    {
+      "epoch": 0.20347273171924676,
+      "grad_norm": 0.29869019985198975,
+      "learning_rate": 6.20220904478199e-05,
+      "loss": 0.6179,
+      "step": 624
+    },
+    {
+      "epoch": 0.2037988098149507,
+      "grad_norm": 0.2984759509563446,
+      "learning_rate": 6.173165676349103e-05,
+      "loss": 0.6074,
+      "step": 625
+    },
+    {
+      "epoch": 0.2041248879106546,
+      "grad_norm": 0.24173347651958466,
+      "learning_rate": 6.144160077226036e-05,
+      "loss": 0.5561,
+      "step": 626
+    },
+    {
+      "epoch": 0.20445096600635854,
+      "grad_norm": 0.25193408131599426,
+      "learning_rate": 6.11519253368634e-05,
+      "loss": 0.5448,
+      "step": 627
+    },
+    {
+      "epoch": 0.20477704410206243,
+      "grad_norm": 0.29573777318000793,
+      "learning_rate": 6.086263331627976e-05,
+      "loss": 0.6412,
+      "step": 628
+    },
+    {
+      "epoch": 0.20510312219776636,
+      "grad_norm": 0.27745357155799866,
+      "learning_rate": 6.05737275657049e-05,
+      "loss": 0.6071,
+      "step": 629
+    },
+    {
+      "epoch": 0.20542920029347028,
+      "grad_norm": 0.2653226852416992,
+      "learning_rate": 6.0285210936521955e-05,
+      "loss": 0.57,
+      "step": 630
+    },
+    {
+      "epoch": 0.2057552783891742,
+      "grad_norm": 0.32329148054122925,
+      "learning_rate": 5.999708627627354e-05,
+      "loss": 0.6697,
+      "step": 631
+    },
+    {
+      "epoch": 0.20608135648487813,
+      "grad_norm": 0.2372172623872757,
+      "learning_rate": 5.9709356428633746e-05,
+      "loss": 0.5417,
+      "step": 632
+    },
+    {
+      "epoch": 0.20640743458058206,
+      "grad_norm": 0.2067628800868988,
+      "learning_rate": 5.9422024233380013e-05,
+      "loss": 0.5155,
+      "step": 633
+    },
+    {
+      "epoch": 0.20673351267628598,
+      "grad_norm": 0.27399685978889465,
+      "learning_rate": 5.913509252636511e-05,
+      "loss": 0.5758,
+      "step": 634
+    },
+    {
+      "epoch": 0.2070595907719899,
+      "grad_norm": 0.2475450485944748,
+      "learning_rate": 5.884856413948913e-05,
+      "loss": 0.5713,
+      "step": 635
+    },
+    {
+      "epoch": 0.2073856688676938,
+      "grad_norm": 0.23439644277095795,
+      "learning_rate": 5.856244190067159e-05,
+      "loss": 0.5979,
+      "step": 636
+    },
+    {
+      "epoch": 0.20771174696339773,
+      "grad_norm": 0.22557716071605682,
+      "learning_rate": 5.82767286338235e-05,
+      "loss": 0.5951,
+      "step": 637
+    },
+    {
+      "epoch": 0.20803782505910165,
+      "grad_norm": 0.3093213140964508,
+      "learning_rate": 5.799142715881938e-05,
+      "loss": 0.6018,
+      "step": 638
+    },
+    {
+      "epoch": 0.20836390315480557,
+      "grad_norm": 0.3633502125740051,
+      "learning_rate": 5.770654029146969e-05,
+      "loss": 0.7025,
+      "step": 639
+    },
+    {
+      "epoch": 0.2086899812505095,
+      "grad_norm": 0.2910383343696594,
+      "learning_rate": 5.7422070843492734e-05,
+      "loss": 0.6501,
+      "step": 640
+    },
+    {
+      "epoch": 0.20901605934621342,
+      "grad_norm": 0.36639973521232605,
+      "learning_rate": 5.713802162248718e-05,
+      "loss": 0.6685,
+      "step": 641
+    },
+    {
+      "epoch": 0.20934213744191735,
+      "grad_norm": 0.3218449652194977,
+      "learning_rate": 5.6854395431904094e-05,
+      "loss": 0.5902,
+      "step": 642
+    },
+    {
+      "epoch": 0.20966821553762127,
+      "grad_norm": 0.2687339186668396,
+      "learning_rate": 5.657119507101954e-05,
+      "loss": 0.6373,
+      "step": 643
+    },
+    {
+      "epoch": 0.20999429363332517,
+      "grad_norm": 0.25556617975234985,
+      "learning_rate": 5.6288423334906735e-05,
+      "loss": 0.5198,
+      "step": 644
+    },
+    {
+      "epoch": 0.2103203717290291,
+      "grad_norm": 0.20101326704025269,
+      "learning_rate": 5.6006083014408484e-05,
+      "loss": 0.5569,
+      "step": 645
+    },
+    {
+      "epoch": 0.21064644982473302,
+      "grad_norm": 0.27940645813941956,
+      "learning_rate": 5.572417689610987e-05,
+      "loss": 0.6211,
+      "step": 646
+    },
+    {
+      "epoch": 0.21097252792043694,
+      "grad_norm": 0.24627013504505157,
+      "learning_rate": 5.544270776231038e-05,
+      "loss": 0.5832,
+      "step": 647
+    },
+    {
+      "epoch": 0.21129860601614087,
+      "grad_norm": 0.2317035049200058,
+      "learning_rate": 5.5161678390996796e-05,
+      "loss": 0.5157,
+      "step": 648
+    },
+    {
+      "epoch": 0.2116246841118448,
+      "grad_norm": 0.36045899987220764,
+      "learning_rate": 5.488109155581549e-05,
+      "loss": 0.6291,
+      "step": 649
+    },
+    {
+      "epoch": 0.21195076220754872,
+      "grad_norm": 0.24675047397613525,
+      "learning_rate": 5.4600950026045326e-05,
+      "loss": 0.5645,
+      "step": 650
+    },
+    {
+      "epoch": 0.21227684030325264,
+      "grad_norm": 0.29204484820365906,
+      "learning_rate": 5.4321256566570036e-05,
+      "loss": 0.6878,
+      "step": 651
+    },
+    {
+      "epoch": 0.21260291839895654,
+      "grad_norm": 0.2781514823436737,
+      "learning_rate": 5.404201393785122e-05,
+      "loss": 0.6069,
+      "step": 652
+    },
+    {
+      "epoch": 0.21292899649466046,
+      "grad_norm": 0.27885934710502625,
+      "learning_rate": 5.3763224895900846e-05,
+      "loss": 0.6134,
+      "step": 653
+    },
+    {
+      "epoch": 0.21325507459036439,
+      "grad_norm": 0.3275914788246155,
+      "learning_rate": 5.348489219225416e-05,
+      "loss": 0.6795,
+      "step": 654
+    },
+    {
+      "epoch": 0.2135811526860683,
+      "grad_norm": 0.2829757034778595,
+      "learning_rate": 5.320701857394268e-05,
+      "loss": 0.6291,
+      "step": 655
+    },
+    {
+      "epoch": 0.21390723078177223,
+      "grad_norm": 0.31990575790405273,
+      "learning_rate": 5.292960678346675e-05,
+      "loss": 0.6298,
+      "step": 656
+    },
+    {
+      "epoch": 0.21423330887747616,
+      "grad_norm": 0.2904050350189209,
+      "learning_rate": 5.265265955876879e-05,
+      "loss": 0.6357,
+      "step": 657
+    },
+    {
+      "epoch": 0.21455938697318008,
+      "grad_norm": 0.2836860418319702,
+      "learning_rate": 5.237617963320608e-05,
+      "loss": 0.6394,
+      "step": 658
+    },
+    {
+      "epoch": 0.214885465068884,
+      "grad_norm": 0.2522071301937103,
+      "learning_rate": 5.210016973552391e-05,
+      "loss": 0.6214,
+      "step": 659
+    },
+    {
+      "epoch": 0.21521154316458793,
+      "grad_norm": 0.2518799901008606,
+      "learning_rate": 5.182463258982846e-05,
+      "loss": 0.6404,
+      "step": 660
+    },
+    {
+      "epoch": 0.21553762126029183,
+      "grad_norm": 0.21216027438640594,
+      "learning_rate": 5.1549570915560206e-05,
+      "loss": 0.4331,
+      "step": 661
+    },
+    {
+      "epoch": 0.21586369935599575,
+      "grad_norm": 0.34990862011909485,
+      "learning_rate": 5.127498742746675e-05,
+      "loss": 0.5828,
+      "step": 662
+    },
+    {
+      "epoch": 0.21618977745169968,
+      "grad_norm": 0.2469099760055542,
+      "learning_rate": 5.100088483557634e-05,
+      "loss": 0.5904,
+      "step": 663
+    },
+    {
+      "epoch": 0.2165158555474036,
+      "grad_norm": 0.33630579710006714,
+      "learning_rate": 5.072726584517086e-05,
+      "loss": 0.5634,
+      "step": 664
+    },
+    {
+      "epoch": 0.21684193364310753,
+      "grad_norm": 0.2696487307548523,
+      "learning_rate": 5.045413315675924e-05,
+      "loss": 0.5211,
+      "step": 665
+    },
+    {
+      "epoch": 0.21716801173881145,
+      "grad_norm": 0.30904674530029297,
+      "learning_rate": 5.018148946605092e-05,
+      "loss": 0.624,
+      "step": 666
+    },
+    {
+      "epoch": 0.21749408983451538,
+      "grad_norm": 0.3870840072631836,
+      "learning_rate": 4.990933746392899e-05,
+      "loss": 0.6745,
+      "step": 667
+    },
+    {
+      "epoch": 0.2178201679302193,
+      "grad_norm": 0.3054164946079254,
+      "learning_rate": 4.9637679836423924e-05,
+      "loss": 0.6029,
+      "step": 668
+    },
+    {
+      "epoch": 0.2181462460259232,
+      "grad_norm": 0.24271585047245026,
+      "learning_rate": 4.9366519264686725e-05,
+      "loss": 0.5774,
+      "step": 669
+    },
+    {
+      "epoch": 0.21847232412162712,
+      "grad_norm": 0.2951551377773285,
+      "learning_rate": 4.909585842496287e-05,
+      "loss": 0.6597,
+      "step": 670
+    },
+    {
+      "epoch": 0.21879840221733104,
+      "grad_norm": 0.26155775785446167,
+      "learning_rate": 4.8825699988565485e-05,
+      "loss": 0.6342,
+      "step": 671
+    },
+    {
+      "epoch": 0.21912448031303497,
+      "grad_norm": 0.2630130648612976,
+      "learning_rate": 4.8556046621849346e-05,
+      "loss": 0.5336,
+      "step": 672
+    },
+    {
+      "epoch": 0.2194505584087389,
+      "grad_norm": 0.21812640130519867,
+      "learning_rate": 4.828690098618429e-05,
+      "loss": 0.6072,
+      "step": 673
+    },
+    {
+      "epoch": 0.21977663650444282,
+      "grad_norm": 0.25095483660697937,
+      "learning_rate": 4.8018265737929044e-05,
+      "loss": 0.6364,
+      "step": 674
+    },
+    {
+      "epoch": 0.22010271460014674,
+      "grad_norm": 0.20363305509090424,
+      "learning_rate": 4.7750143528405126e-05,
+      "loss": 0.5275,
+      "step": 675
+    },
+    {
+      "epoch": 0.22042879269585067,
+      "grad_norm": 0.29310062527656555,
+      "learning_rate": 4.748253700387042e-05,
+      "loss": 0.6871,
+      "step": 676
+    },
+    {
+      "epoch": 0.22075487079155456,
+      "grad_norm": 0.24926213920116425,
+      "learning_rate": 4.721544880549337e-05,
+      "loss": 0.5561,
+      "step": 677
+    },
+    {
+      "epoch": 0.2210809488872585,
+      "grad_norm": 0.3037482798099518,
+      "learning_rate": 4.694888156932658e-05,
+      "loss": 0.6398,
+      "step": 678
+    },
+    {
+      "epoch": 0.2214070269829624,
+      "grad_norm": 0.2365943193435669,
+      "learning_rate": 4.668283792628114e-05,
+      "loss": 0.5932,
+      "step": 679
+    },
+    {
+      "epoch": 0.22173310507866634,
+      "grad_norm": 0.2507438361644745,
+      "learning_rate": 4.6417320502100316e-05,
+      "loss": 0.6336,
+      "step": 680
+    },
+    {
+      "epoch": 0.22205918317437026,
+      "grad_norm": 0.2556549310684204,
+      "learning_rate": 4.615233191733398e-05,
+      "loss": 0.5695,
+      "step": 681
+    },
+    {
+      "epoch": 0.22238526127007419,
+      "grad_norm": 0.3738943040370941,
+      "learning_rate": 4.588787478731242e-05,
+      "loss": 0.6191,
+      "step": 682
+    },
+    {
+      "epoch": 0.2227113393657781,
+      "grad_norm": 0.2213764786720276,
+      "learning_rate": 4.5623951722120736e-05,
+      "loss": 0.4365,
+      "step": 683
+    },
+    {
+      "epoch": 0.22303741746148203,
+      "grad_norm": 0.24922537803649902,
+      "learning_rate": 4.5360565326573104e-05,
+      "loss": 0.5419,
+      "step": 684
+    },
+    {
+      "epoch": 0.22336349555718596,
+      "grad_norm": 0.27778947353363037,
+      "learning_rate": 4.5097718200186814e-05,
+      "loss": 0.5728,
+      "step": 685
+    },
+    {
+      "epoch": 0.22368957365288986,
+      "grad_norm": 0.273242324590683,
+      "learning_rate": 4.483541293715698e-05,
+      "loss": 0.547,
+      "step": 686
+    },
+    {
+      "epoch": 0.22401565174859378,
+      "grad_norm": 0.325244277715683,
+      "learning_rate": 4.457365212633058e-05,
+      "loss": 0.5824,
+      "step": 687
+    },
+    {
+      "epoch": 0.2243417298442977,
+      "grad_norm": 0.24153675138950348,
+      "learning_rate": 4.431243835118124e-05,
+      "loss": 0.5492,
+      "step": 688
+    },
+    {
+      "epoch": 0.22466780794000163,
+      "grad_norm": 0.2707926332950592,
+      "learning_rate": 4.4051774189783315e-05,
+      "loss": 0.5661,
+      "step": 689
+    },
+    {
+      "epoch": 0.22499388603570555,
+      "grad_norm": 0.22814729809761047,
+      "learning_rate": 4.379166221478697e-05,
+      "loss": 0.5473,
+      "step": 690
+    },
+    {
+      "epoch": 0.22531996413140948,
+      "grad_norm": 0.3418341279029846,
+      "learning_rate": 4.3532104993392306e-05,
+      "loss": 0.6342,
+      "step": 691
+    },
+    {
+      "epoch": 0.2256460422271134,
+      "grad_norm": 0.2698429822921753,
+      "learning_rate": 4.327310508732437e-05,
+      "loss": 0.6629,
+      "step": 692
+    },
+    {
+      "epoch": 0.22597212032281733,
+      "grad_norm": 0.25924572348594666,
+      "learning_rate": 4.301466505280762e-05,
+      "loss": 0.6371,
+      "step": 693
+    },
+    {
+      "epoch": 0.22629819841852122,
+      "grad_norm": 0.24862578511238098,
+      "learning_rate": 4.2756787440540936e-05,
+      "loss": 0.6055,
+      "step": 694
+    },
+    {
+      "epoch": 0.22662427651422515,
+      "grad_norm": 0.2641250491142273,
+      "learning_rate": 4.249947479567218e-05,
+      "loss": 0.5953,
+      "step": 695
+    },
+    {
+      "epoch": 0.22695035460992907,
+      "grad_norm": 0.34082797169685364,
+      "learning_rate": 4.224272965777326e-05,
+      "loss": 0.6108,
+      "step": 696
+    },
+    {
+      "epoch": 0.227276432705633,
+      "grad_norm": 0.26527827978134155,
+      "learning_rate": 4.1986554560815096e-05,
+      "loss": 0.5238,
+      "step": 697
+    },
+    {
+      "epoch": 0.22760251080133692,
+      "grad_norm": 0.2595398426055908,
+      "learning_rate": 4.173095203314241e-05,
+      "loss": 0.5896,
+      "step": 698
+    },
+    {
+      "epoch": 0.22792858889704085,
+      "grad_norm": 0.317978173494339,
+      "learning_rate": 4.1475924597449024e-05,
+      "loss": 0.6303,
+      "step": 699
+    },
+    {
+      "epoch": 0.22825466699274477,
+      "grad_norm": 0.28211531043052673,
+      "learning_rate": 4.12214747707527e-05,
+      "loss": 0.5635,
+      "step": 700
+    },
+    {
+      "epoch": 0.2285807450884487,
+      "grad_norm": 0.2978782653808594,
+      "learning_rate": 4.096760506437057e-05,
+      "loss": 0.6719,
+      "step": 701
+    },
+    {
+      "epoch": 0.2289068231841526,
+      "grad_norm": 0.22416797280311584,
+      "learning_rate": 4.071431798389408e-05,
+      "loss": 0.6223,
+      "step": 702
+    },
+    {
+      "epoch": 0.22923290127985652,
+      "grad_norm": 0.25638195872306824,
+      "learning_rate": 4.0461616029164526e-05,
+      "loss": 0.5414,
+      "step": 703
+    },
+    {
+      "epoch": 0.22955897937556044,
+      "grad_norm": 0.3361692428588867,
+      "learning_rate": 4.020950169424815e-05,
+      "loss": 0.6814,
+      "step": 704
+    },
+    {
+      "epoch": 0.22988505747126436,
+      "grad_norm": 0.45392799377441406,
+      "learning_rate": 3.9957977467411615e-05,
+      "loss": 0.7154,
+      "step": 705
+    },
+    {
+      "epoch": 0.2302111355669683,
+      "grad_norm": 0.31325310468673706,
+      "learning_rate": 3.9707045831097555e-05,
+      "loss": 0.73,
+      "step": 706
+    },
+    {
+      "epoch": 0.2305372136626722,
+      "grad_norm": 0.31382250785827637,
+      "learning_rate": 3.945670926189987e-05,
+      "loss": 0.5842,
+      "step": 707
+    },
+    {
+      "epoch": 0.23086329175837614,
+      "grad_norm": 0.31560176610946655,
+      "learning_rate": 3.920697023053949e-05,
+      "loss": 0.6322,
+      "step": 708
+    },
+    {
+      "epoch": 0.23118936985408006,
+      "grad_norm": 0.28578540682792664,
+      "learning_rate": 3.895783120183976e-05,
+      "loss": 0.5562,
+      "step": 709
+    },
+    {
+      "epoch": 0.231515447949784,
+      "grad_norm": 0.3130005896091461,
+      "learning_rate": 3.8709294634702376e-05,
+      "loss": 0.5816,
+      "step": 710
+    },
+    {
+      "epoch": 0.23184152604548788,
+      "grad_norm": 0.21266983449459076,
+      "learning_rate": 3.846136298208285e-05,
+      "loss": 0.6178,
+      "step": 711
+    },
+    {
+      "epoch": 0.2321676041411918,
+      "grad_norm": 0.28499025106430054,
+      "learning_rate": 3.821403869096658e-05,
+      "loss": 0.5944,
+      "step": 712
+    },
+    {
+      "epoch": 0.23249368223689573,
+      "grad_norm": 0.2727053761482239,
+      "learning_rate": 3.796732420234443e-05,
+      "loss": 0.5336,
+      "step": 713
+    },
+    {
+      "epoch": 0.23281976033259966,
+      "grad_norm": 0.4271116554737091,
+      "learning_rate": 3.7721221951188765e-05,
+      "loss": 0.5785,
+      "step": 714
+    },
+    {
+      "epoch": 0.23314583842830358,
+      "grad_norm": 0.2158544808626175,
+      "learning_rate": 3.747573436642951e-05,
+      "loss": 0.5463,
+      "step": 715
+    },
+    {
+      "epoch": 0.2334719165240075,
+      "grad_norm": 0.23925550282001495,
+      "learning_rate": 3.7230863870929964e-05,
+      "loss": 0.5424,
+      "step": 716
+    },
+    {
+      "epoch": 0.23379799461971143,
+      "grad_norm": 0.2507498562335968,
+      "learning_rate": 3.698661288146311e-05,
+      "loss": 0.5701,
+      "step": 717
+    },
+    {
+      "epoch": 0.23412407271541535,
+      "grad_norm": 0.3710465133190155,
+      "learning_rate": 3.674298380868756e-05,
+      "loss": 0.5927,
+      "step": 718
+    },
+    {
+      "epoch": 0.23445015081111925,
+      "grad_norm": 0.28400731086730957,
+      "learning_rate": 3.649997905712396e-05,
+      "loss": 0.5458,
+      "step": 719
+    },
+    {
+      "epoch": 0.23477622890682318,
+      "grad_norm": 0.2662917375564575,
+      "learning_rate": 3.6257601025131026e-05,
+      "loss": 0.5832,
+      "step": 720
+    },
+    {
+      "epoch": 0.2351023070025271,
+      "grad_norm": 0.28415587544441223,
+      "learning_rate": 3.601585210488218e-05,
+      "loss": 0.5417,
+      "step": 721
+    },
+    {
+      "epoch": 0.23542838509823102,
+      "grad_norm": 0.27368491888046265,
+      "learning_rate": 3.577473468234156e-05,
+      "loss": 0.6384,
+      "step": 722
+    },
+    {
+      "epoch": 0.23575446319393495,
+      "grad_norm": 0.2925770878791809,
+      "learning_rate": 3.553425113724088e-05,
+      "loss": 0.6143,
+      "step": 723
+    },
+    {
+      "epoch": 0.23608054128963887,
+      "grad_norm": 0.3160930871963501,
+      "learning_rate": 3.52944038430556e-05,
+      "loss": 0.5924,
+      "step": 724
+    },
+    {
+      "epoch": 0.2364066193853428,
+      "grad_norm": 0.3158300518989563,
+      "learning_rate": 3.5055195166981645e-05,
+      "loss": 0.4604,
+      "step": 725
+    },
+    {
+      "epoch": 0.23673269748104672,
+      "grad_norm": 0.2738193869590759,
+      "learning_rate": 3.481662746991214e-05,
+      "loss": 0.5122,
+      "step": 726
+    },
+    {
+      "epoch": 0.23705877557675062,
+      "grad_norm": 0.39256951212882996,
+      "learning_rate": 3.4578703106413904e-05,
+      "loss": 0.6385,
+      "step": 727
+    },
+    {
+      "epoch": 0.23738485367245454,
+      "grad_norm": 0.2841635048389435,
+      "learning_rate": 3.4341424424704375e-05,
+      "loss": 0.5826,
+      "step": 728
+    },
+    {
+      "epoch": 0.23771093176815847,
+      "grad_norm": 0.33367428183555603,
+      "learning_rate": 3.4104793766628304e-05,
+      "loss": 0.5943,
+      "step": 729
+    },
+    {
+      "epoch": 0.2380370098638624,
+      "grad_norm": 0.324015349149704,
+      "learning_rate": 3.386881346763483e-05,
+      "loss": 0.5332,
+      "step": 730
+    },
+    {
+      "epoch": 0.23836308795956632,
+      "grad_norm": 0.30556005239486694,
+      "learning_rate": 3.363348585675414e-05,
+      "loss": 0.536,
+      "step": 731
+    },
+    {
+      "epoch": 0.23868916605527024,
+      "grad_norm": 0.236478790640831,
+      "learning_rate": 3.339881325657484e-05,
+      "loss": 0.4597,
+      "step": 732
+    },
+    {
+      "epoch": 0.23901524415097417,
+      "grad_norm": 0.3337570130825043,
+      "learning_rate": 3.316479798322072e-05,
+      "loss": 0.5152,
+      "step": 733
+    },
+    {
+      "epoch": 0.2393413222466781,
+      "grad_norm": 0.3730219006538391,
+      "learning_rate": 3.2931442346328004e-05,
+      "loss": 0.7172,
+      "step": 734
+    },
+    {
+      "epoch": 0.23966740034238201,
+      "grad_norm": 0.3032122254371643,
+      "learning_rate": 3.269874864902269e-05,
+      "loss": 0.6797,
+      "step": 735
+    },
+    {
+      "epoch": 0.2399934784380859,
+      "grad_norm": 0.2508755624294281,
+      "learning_rate": 3.246671918789755e-05,
+      "loss": 0.5697,
+      "step": 736
+    },
+    {
+      "epoch": 0.24031955653378984,
+      "grad_norm": 0.26841485500335693,
+      "learning_rate": 3.223535625298979e-05,
+      "loss": 0.5376,
+      "step": 737
+    },
+    {
+      "epoch": 0.24064563462949376,
+      "grad_norm": 0.21499516069889069,
+      "learning_rate": 3.200466212775808e-05,
+      "loss": 0.5711,
+      "step": 738
+    },
+    {
+      "epoch": 0.24097171272519768,
+      "grad_norm": 0.2505497336387634,
+      "learning_rate": 3.1774639089060363e-05,
+      "loss": 0.5365,
+      "step": 739
+    },
+    {
+      "epoch": 0.2412977908209016,
+      "grad_norm": 0.2722863256931305,
+      "learning_rate": 3.154528940713113e-05,
+      "loss": 0.5966,
+      "step": 740
+    },
+    {
+      "epoch": 0.24162386891660553,
+      "grad_norm": 0.26614895462989807,
+      "learning_rate": 3.1316615345559185e-05,
+      "loss": 0.6171,
+      "step": 741
+    },
+    {
+      "epoch": 0.24194994701230946,
+      "grad_norm": 0.29123929142951965,
+      "learning_rate": 3.108861916126518e-05,
+      "loss": 0.6223,
+      "step": 742
+    },
+    {
+      "epoch": 0.24227602510801338,
+      "grad_norm": 0.2488095760345459,
+      "learning_rate": 3.086130310447937e-05,
+      "loss": 0.5843,
+      "step": 743
+    },
+    {
+      "epoch": 0.24260210320371728,
+      "grad_norm": 0.4080983102321625,
+      "learning_rate": 3.063466941871952e-05,
+      "loss": 0.649,
+      "step": 744
+    },
+    {
+      "epoch": 0.2429281812994212,
+      "grad_norm": 0.278074711561203,
+      "learning_rate": 3.0408720340768572e-05,
+      "loss": 0.6112,
+      "step": 745
+    },
+    {
+      "epoch": 0.24325425939512513,
+      "grad_norm": 0.22914646565914154,
+      "learning_rate": 3.018345810065275e-05,
+      "loss": 0.5193,
+      "step": 746
+    },
+    {
+      "epoch": 0.24358033749082905,
+      "grad_norm": 0.2255176156759262,
+      "learning_rate": 2.9958884921619367e-05,
+      "loss": 0.4962,
+      "step": 747
+    },
+    {
+      "epoch": 0.24390641558653298,
+      "grad_norm": 0.24137818813323975,
+      "learning_rate": 2.9735003020115092e-05,
+      "loss": 0.4636,
+      "step": 748
+    },
+    {
+      "epoch": 0.2442324936822369,
+      "grad_norm": 0.2586713433265686,
+      "learning_rate": 2.9511814605763855e-05,
+      "loss": 0.6269,
+      "step": 749
+    },
+    {
+      "epoch": 0.24455857177794083,
+      "grad_norm": 0.29199546575546265,
+      "learning_rate": 2.9289321881345254e-05,
+      "loss": 0.6608,
+      "step": 750
+    },
+    {
+      "epoch": 0.24488464987364475,
+      "grad_norm": 0.2844295799732208,
+      "learning_rate": 2.9067527042772636e-05,
+      "loss": 0.6155,
+      "step": 751
+    },
+    {
+      "epoch": 0.24521072796934865,
+      "grad_norm": 0.25179925560951233,
+      "learning_rate": 2.8846432279071467e-05,
+      "loss": 0.529,
+      "step": 752
+    },
+    {
+      "epoch": 0.24553680606505257,
+      "grad_norm": 0.2996719181537628,
+      "learning_rate": 2.8626039772357882e-05,
+      "loss": 0.626,
+      "step": 753
+    },
+    {
+      "epoch": 0.2458628841607565,
+      "grad_norm": 0.2830248177051544,
+      "learning_rate": 2.840635169781688e-05,
+      "loss": 0.5863,
+      "step": 754
+    },
+    {
+      "epoch": 0.24618896225646042,
+      "grad_norm": 0.32690149545669556,
+      "learning_rate": 2.8187370223681132e-05,
+      "loss": 0.6031,
+      "step": 755
+    },
+    {
+      "epoch": 0.24651504035216434,
+      "grad_norm": 0.2886967658996582,
+      "learning_rate": 2.7969097511209308e-05,
+      "loss": 0.6521,
+      "step": 756
+    },
+    {
+      "epoch": 0.24684111844786827,
+      "grad_norm": 0.2926975190639496,
+      "learning_rate": 2.775153571466502e-05,
+      "loss": 0.6166,
+      "step": 757
+    },
+    {
+      "epoch": 0.2471671965435722,
+      "grad_norm": 0.25793954730033875,
+      "learning_rate": 2.753468698129533e-05,
+      "loss": 0.6065,
+      "step": 758
+    },
+    {
+      "epoch": 0.24749327463927612,
+      "grad_norm": 0.30857494473457336,
+      "learning_rate": 2.7318553451309726e-05,
+      "loss": 0.6399,
+      "step": 759
+    },
+    {
+      "epoch": 0.24781935273498001,
+      "grad_norm": 0.3076545298099518,
+      "learning_rate": 2.7103137257858868e-05,
+      "loss": 0.58,
+      "step": 760
+    },
+    {
+      "epoch": 0.24814543083068394,
+      "grad_norm": 0.28424713015556335,
+      "learning_rate": 2.688844052701359e-05,
+      "loss": 0.5929,
+      "step": 761
+    },
+    {
+      "epoch": 0.24847150892638786,
+      "grad_norm": 0.26437848806381226,
+      "learning_rate": 2.6674465377744017e-05,
+      "loss": 0.5904,
+      "step": 762
+    },
+    {
+      "epoch": 0.2487975870220918,
+      "grad_norm": 0.35582268238067627,
+      "learning_rate": 2.646121392189841e-05,
+      "loss": 0.6299,
+      "step": 763
+    },
+    {
+      "epoch": 0.2491236651177957,
+      "grad_norm": 0.3153063654899597,
+      "learning_rate": 2.624868826418262e-05,
+      "loss": 0.6415,
+      "step": 764
+    },
+    {
+      "epoch": 0.24944974321349964,
+      "grad_norm": 0.23518048226833344,
+      "learning_rate": 2.603689050213902e-05,
+      "loss": 0.5852,
+      "step": 765
+    },
+    {
+      "epoch": 0.24977582130920356,
+      "grad_norm": 0.3225111663341522,
+      "learning_rate": 2.582582272612609e-05,
+      "loss": 0.6137,
+      "step": 766
+    },
+    {
+      "epoch": 0.25010189940490746,
+      "grad_norm": 0.28972920775413513,
+      "learning_rate": 2.561548701929749e-05,
+      "loss": 0.6576,
+      "step": 767
+    },
+    {
+      "epoch": 0.2504279775006114,
+      "grad_norm": 0.2671186625957489,
+      "learning_rate": 2.540588545758179e-05,
+      "loss": 0.557,
+      "step": 768
+    },
+    {
+      "epoch": 0.2507540555963153,
+      "grad_norm": 0.29164326190948486,
+      "learning_rate": 2.5197020109661772e-05,
+      "loss": 0.5875,
+      "step": 769
+    },
+    {
+      "epoch": 0.25108013369201926,
+      "grad_norm": 0.39768028259277344,
+      "learning_rate": 2.4988893036954043e-05,
+      "loss": 0.6615,
+      "step": 770
+    },
+    {
+      "epoch": 0.25140621178772316,
+      "grad_norm": 0.44214484095573425,
+      "learning_rate": 2.4781506293588873e-05,
+      "loss": 0.5663,
+      "step": 771
+    },
+    {
+      "epoch": 0.2517322898834271,
+      "grad_norm": 0.3203497529029846,
+      "learning_rate": 2.4574861926389615e-05,
+      "loss": 0.6532,
+      "step": 772
+    },
+    {
+      "epoch": 0.252058367979131,
+      "grad_norm": 0.30829140543937683,
+      "learning_rate": 2.436896197485282e-05,
+      "loss": 0.6116,
+      "step": 773
+    },
+    {
+      "epoch": 0.2523844460748349,
+      "grad_norm": 0.3043927550315857,
+      "learning_rate": 2.4163808471127812e-05,
+      "loss": 0.6069,
+      "step": 774
+    },
+    {
+      "epoch": 0.25271052417053885,
+      "grad_norm": 0.2910175323486328,
+      "learning_rate": 2.3959403439996907e-05,
+      "loss": 0.5352,
+      "step": 775
+    },
+    {
+      "epoch": 0.25303660226624275,
+      "grad_norm": 0.29559823870658875,
+      "learning_rate": 2.37557488988552e-05,
+      "loss": 0.5325,
+      "step": 776
+    },
+    {
+      "epoch": 0.2533626803619467,
+      "grad_norm": 0.26113376021385193,
+      "learning_rate": 2.3552846857690846e-05,
+      "loss": 0.6223,
+      "step": 777
+    },
+    {
+      "epoch": 0.2536887584576506,
+      "grad_norm": 0.3582890033721924,
+      "learning_rate": 2.3350699319065026e-05,
+      "loss": 0.6703,
+      "step": 778
+    },
+    {
+      "epoch": 0.25401483655335455,
+      "grad_norm": 0.25528156757354736,
+      "learning_rate": 2.3149308278092342e-05,
+      "loss": 0.5875,
+      "step": 779
+    },
+    {
+      "epoch": 0.25434091464905845,
+      "grad_norm": 0.2476278692483902,
+      "learning_rate": 2.2948675722421086e-05,
+      "loss": 0.5797,
+      "step": 780
+    },
+    {
+      "epoch": 0.25466699274476234,
+      "grad_norm": 0.2513749897480011,
+      "learning_rate": 2.2748803632213557e-05,
+      "loss": 0.5855,
+      "step": 781
+    },
+    {
+      "epoch": 0.2549930708404663,
+      "grad_norm": 0.23542198538780212,
+      "learning_rate": 2.254969398012663e-05,
+      "loss": 0.5742,
+      "step": 782
+    },
+    {
+      "epoch": 0.2553191489361702,
+      "grad_norm": 0.3274197578430176,
+      "learning_rate": 2.235134873129213e-05,
+      "loss": 0.6306,
+      "step": 783
+    },
+    {
+      "epoch": 0.25564522703187414,
+      "grad_norm": 0.3096897602081299,
+      "learning_rate": 2.2153769843297667e-05,
+      "loss": 0.5931,
+      "step": 784
+    },
+    {
+      "epoch": 0.25597130512757804,
+      "grad_norm": 0.22587507963180542,
+      "learning_rate": 2.195695926616702e-05,
+      "loss": 0.4754,
+      "step": 785
+    },
+    {
+      "epoch": 0.256297383223282,
+      "grad_norm": 0.22682449221611023,
+      "learning_rate": 2.1760918942341192e-05,
+      "loss": 0.4945,
+      "step": 786
+    },
+    {
+      "epoch": 0.2566234613189859,
+      "grad_norm": 0.24025246500968933,
+      "learning_rate": 2.1565650806658975e-05,
+      "loss": 0.5782,
+      "step": 787
+    },
+    {
+      "epoch": 0.25694953941468984,
+      "grad_norm": 0.2858300805091858,
+      "learning_rate": 2.137115678633811e-05,
+      "loss": 0.6889,
+      "step": 788
+    },
+    {
+      "epoch": 0.25727561751039374,
+      "grad_norm": 0.24029089510440826,
+      "learning_rate": 2.1177438800956007e-05,
+      "loss": 0.6623,
+      "step": 789
+    },
+    {
+      "epoch": 0.25760169560609764,
+      "grad_norm": 0.2904897928237915,
+      "learning_rate": 2.098449876243096e-05,
+      "loss": 0.6324,
+      "step": 790
+    },
+    {
+      "epoch": 0.2579277737018016,
+      "grad_norm": 0.38068699836730957,
+      "learning_rate": 2.07923385750033e-05,
+      "loss": 0.6569,
+      "step": 791
+    },
+    {
+      "epoch": 0.2582538517975055,
+      "grad_norm": 0.37293192744255066,
+      "learning_rate": 2.0600960135216462e-05,
+      "loss": 0.7311,
+      "step": 792
+    },
+    {
+      "epoch": 0.25857992989320944,
+      "grad_norm": 0.43448659777641296,
+      "learning_rate": 2.0410365331898416e-05,
+      "loss": 0.5977,
+      "step": 793
+    },
+    {
+      "epoch": 0.25890600798891333,
+      "grad_norm": 0.3264097273349762,
+      "learning_rate": 2.0220556046142893e-05,
+      "loss": 0.6446,
+      "step": 794
+    },
+    {
+      "epoch": 0.2592320860846173,
+      "grad_norm": 0.31544405221939087,
+      "learning_rate": 2.0031534151290943e-05,
+      "loss": 0.63,
+      "step": 795
+    },
+    {
+      "epoch": 0.2595581641803212,
+      "grad_norm": 0.2896043658256531,
+      "learning_rate": 1.9843301512912327e-05,
+      "loss": 0.6207,
+      "step": 796
+    },
+    {
+      "epoch": 0.25988424227602513,
+      "grad_norm": 0.22150889039039612,
+      "learning_rate": 1.965585998878724e-05,
+      "loss": 0.5631,
+      "step": 797
+    },
+    {
+      "epoch": 0.26021032037172903,
+      "grad_norm": 0.31330549716949463,
+      "learning_rate": 1.946921142888781e-05,
+      "loss": 0.5656,
+      "step": 798
+    },
+    {
+      "epoch": 0.26053639846743293,
+      "grad_norm": 0.2872115969657898,
+      "learning_rate": 1.928335767535997e-05,
+      "loss": 0.5956,
+      "step": 799
+    },
+    {
+      "epoch": 0.2608624765631369,
+      "grad_norm": 0.23877061903476715,
+      "learning_rate": 1.9098300562505266e-05,
+      "loss": 0.5805,
+      "step": 800
+    },
+    {
+      "epoch": 0.2611885546588408,
+      "grad_norm": 0.4347909688949585,
+      "learning_rate": 1.891404191676265e-05,
+      "loss": 0.7054,
+      "step": 801
+    },
+    {
+      "epoch": 0.26151463275454473,
+      "grad_norm": 0.28407391905784607,
+      "learning_rate": 1.8730583556690605e-05,
+      "loss": 0.5604,
+      "step": 802
+    },
+    {
+      "epoch": 0.2618407108502486,
+      "grad_norm": 0.26138556003570557,
+      "learning_rate": 1.854792729294905e-05,
+      "loss": 0.6588,
+      "step": 803
+    },
+    {
+      "epoch": 0.2621667889459526,
+      "grad_norm": 0.2357640415430069,
+      "learning_rate": 1.8366074928281607e-05,
+      "loss": 0.578,
+      "step": 804
+    },
+    {
+      "epoch": 0.2624928670416565,
+      "grad_norm": 0.45062610507011414,
+      "learning_rate": 1.818502825749764e-05,
+      "loss": 0.6529,
+      "step": 805
+    },
+    {
+      "epoch": 0.26281894513736037,
+      "grad_norm": 0.2354976236820221,
+      "learning_rate": 1.8004789067454764e-05,
+      "loss": 0.5139,
+      "step": 806
+    },
+    {
+      "epoch": 0.2631450232330643,
+      "grad_norm": 0.23322217166423798,
+      "learning_rate": 1.7825359137040988e-05,
+      "loss": 0.5918,
+      "step": 807
+    },
+    {
+      "epoch": 0.2634711013287682,
+      "grad_norm": 0.2705889046192169,
+      "learning_rate": 1.7646740237157256e-05,
+      "loss": 0.5813,
+      "step": 808
+    },
+    {
+      "epoch": 0.2637971794244722,
+      "grad_norm": 0.2796480655670166,
+      "learning_rate": 1.7468934130700044e-05,
+      "loss": 0.624,
+      "step": 809
+    },
+    {
+      "epoch": 0.26412325752017607,
+      "grad_norm": 0.4603550136089325,
+      "learning_rate": 1.7291942572543807e-05,
+      "loss": 0.6995,
+      "step": 810
+    },
+    {
+      "epoch": 0.26444933561588,
+      "grad_norm": 0.30643367767333984,
+      "learning_rate": 1.7115767309523812e-05,
+      "loss": 0.6055,
+      "step": 811
+    },
+    {
+      "epoch": 0.2647754137115839,
+      "grad_norm": 0.207069531083107,
+      "learning_rate": 1.6940410080418723e-05,
+      "loss": 0.4537,
+      "step": 812
+    },
+    {
+      "epoch": 0.26510149180728787,
+      "grad_norm": 0.2721695005893707,
+      "learning_rate": 1.6765872615933677e-05,
+      "loss": 0.5403,
+      "step": 813
+    },
+    {
+      "epoch": 0.26542756990299177,
+      "grad_norm": 0.20679564774036407,
+      "learning_rate": 1.6592156638682886e-05,
+      "loss": 0.506,
+      "step": 814
+    },
+    {
+      "epoch": 0.26575364799869566,
+      "grad_norm": 0.33122187852859497,
+      "learning_rate": 1.6419263863172997e-05,
+      "loss": 0.7106,
+      "step": 815
+    },
+    {
+      "epoch": 0.2660797260943996,
+      "grad_norm": 0.2706076204776764,
+      "learning_rate": 1.6247195995785837e-05,
+      "loss": 0.6488,
+      "step": 816
+    },
+    {
+      "epoch": 0.2664058041901035,
+      "grad_norm": 0.2812751233577728,
+      "learning_rate": 1.6075954734761845e-05,
+      "loss": 0.5977,
+      "step": 817
+    },
+    {
+      "epoch": 0.26673188228580746,
+      "grad_norm": 0.2687004804611206,
+      "learning_rate": 1.5905541770183096e-05,
+      "loss": 0.5824,
+      "step": 818
+    },
+    {
+      "epoch": 0.26705796038151136,
+      "grad_norm": 0.2888942062854767,
+      "learning_rate": 1.5735958783956794e-05,
+      "loss": 0.604,
+      "step": 819
+    },
+    {
+      "epoch": 0.2673840384772153,
+      "grad_norm": 0.26342645287513733,
+      "learning_rate": 1.5567207449798515e-05,
+      "loss": 0.5679,
+      "step": 820
+    },
+    {
+      "epoch": 0.2677101165729192,
+      "grad_norm": 0.2661057412624359,
+      "learning_rate": 1.539928943321579e-05,
+      "loss": 0.5325,
+      "step": 821
+    },
+    {
+      "epoch": 0.26803619466862316,
+      "grad_norm": 0.21790291368961334,
+      "learning_rate": 1.5232206391491699e-05,
+      "loss": 0.5227,
+      "step": 822
+    },
+    {
+      "epoch": 0.26836227276432706,
+      "grad_norm": 0.2722076177597046,
+      "learning_rate": 1.5065959973668353e-05,
+      "loss": 0.5633,
+      "step": 823
+    },
+    {
+      "epoch": 0.26868835086003096,
+      "grad_norm": 0.24521185457706451,
+      "learning_rate": 1.4900551820530828e-05,
+      "loss": 0.5588,
+      "step": 824
+    },
+    {
+      "epoch": 0.2690144289557349,
+      "grad_norm": 0.3043542802333832,
+      "learning_rate": 1.4735983564590783e-05,
+      "loss": 0.6736,
+      "step": 825
+    },
+    {
+      "epoch": 0.2693405070514388,
+      "grad_norm": 0.2543700933456421,
+      "learning_rate": 1.4572256830070497e-05,
+      "loss": 0.4988,
+      "step": 826
+    },
+    {
+      "epoch": 0.26966658514714276,
+      "grad_norm": 0.30012184381484985,
+      "learning_rate": 1.4409373232886702e-05,
+      "loss": 0.5979,
+      "step": 827
+    },
+    {
+      "epoch": 0.26999266324284665,
+      "grad_norm": 0.2988223135471344,
+      "learning_rate": 1.4247334380634792e-05,
+      "loss": 0.5374,
+      "step": 828
+    },
+    {
+      "epoch": 0.2703187413385506,
+      "grad_norm": 0.29680508375167847,
+      "learning_rate": 1.4086141872572789e-05,
+      "loss": 0.6135,
+      "step": 829
+    },
+    {
+      "epoch": 0.2706448194342545,
+      "grad_norm": 0.3112308382987976,
+      "learning_rate": 1.3925797299605647e-05,
+      "loss": 0.5622,
+      "step": 830
+    },
+    {
+      "epoch": 0.2709708975299584,
+      "grad_norm": 0.23930135369300842,
+      "learning_rate": 1.3766302244269624e-05,
+      "loss": 0.6004,
+      "step": 831
+    },
+    {
+      "epoch": 0.27129697562566235,
+      "grad_norm": 0.28775033354759216,
+      "learning_rate": 1.3607658280716473e-05,
+      "loss": 0.562,
+      "step": 832
+    },
+    {
+      "epoch": 0.27162305372136625,
+      "grad_norm": 0.32726067304611206,
+      "learning_rate": 1.3449866974698122e-05,
+      "loss": 0.5781,
+      "step": 833
+    },
+    {
+      "epoch": 0.2719491318170702,
+      "grad_norm": 0.25154900550842285,
+      "learning_rate": 1.3292929883550998e-05,
+      "loss": 0.5886,
+      "step": 834
+    },
+    {
+      "epoch": 0.2722752099127741,
+      "grad_norm": 0.27102169394493103,
+      "learning_rate": 1.3136848556180892e-05,
+      "loss": 0.5878,
+      "step": 835
+    },
+    {
+      "epoch": 0.27260128800847805,
+      "grad_norm": 0.26259303092956543,
+      "learning_rate": 1.2981624533047432e-05,
+      "loss": 0.6176,
+      "step": 836
+    },
+    {
+      "epoch": 0.27292736610418195,
+      "grad_norm": 0.3033280372619629,
+      "learning_rate": 1.2827259346149122e-05,
+      "loss": 0.541,
+      "step": 837
+    },
+    {
+      "epoch": 0.2732534441998859,
+      "grad_norm": 0.32570040225982666,
+      "learning_rate": 1.2673754519008008e-05,
+      "loss": 0.6198,
+      "step": 838
+    },
+    {
+      "epoch": 0.2735795222955898,
+      "grad_norm": 0.2976042628288269,
+      "learning_rate": 1.2521111566654731e-05,
+      "loss": 0.5765,
+      "step": 839
+    },
+    {
+      "epoch": 0.2739056003912937,
+      "grad_norm": 0.29010704159736633,
+      "learning_rate": 1.2369331995613665e-05,
+      "loss": 0.5873,
+      "step": 840
+    },
+    {
+      "epoch": 0.27423167848699764,
+      "grad_norm": 0.2495676726102829,
+      "learning_rate": 1.2218417303887842e-05,
+      "loss": 0.533,
+      "step": 841
+    },
+    {
+      "epoch": 0.27455775658270154,
+      "grad_norm": 0.25514352321624756,
+      "learning_rate": 1.206836898094439e-05,
+      "loss": 0.4843,
+      "step": 842
+    },
+    {
+      "epoch": 0.2748838346784055,
+      "grad_norm": 0.3069244921207428,
+      "learning_rate": 1.191918850769964e-05,
+      "loss": 0.5408,
+      "step": 843
+    },
+    {
+      "epoch": 0.2752099127741094,
+      "grad_norm": 0.31279370188713074,
+      "learning_rate": 1.1770877356504683e-05,
+      "loss": 0.6755,
+      "step": 844
+    },
+    {
+      "epoch": 0.27553599086981334,
+      "grad_norm": 0.2172246128320694,
+      "learning_rate": 1.1623436991130654e-05,
+      "loss": 0.5187,
+      "step": 845
+    },
+    {
+      "epoch": 0.27586206896551724,
+      "grad_norm": 0.3862195611000061,
+      "learning_rate": 1.1476868866754486e-05,
+      "loss": 0.6153,
+      "step": 846
+    },
+    {
+      "epoch": 0.2761881470612212,
+      "grad_norm": 0.27538466453552246,
+      "learning_rate": 1.1331174429944347e-05,
+      "loss": 0.5874,
+      "step": 847
+    },
+    {
+      "epoch": 0.2765142251569251,
+      "grad_norm": 0.2538582980632782,
+      "learning_rate": 1.1186355118645554e-05,
+      "loss": 0.5447,
+      "step": 848
+    },
+    {
+      "epoch": 0.276840303252629,
+      "grad_norm": 0.34064552187919617,
+      "learning_rate": 1.1042412362166222e-05,
+      "loss": 0.5341,
+      "step": 849
+    },
+    {
+      "epoch": 0.27716638134833294,
+      "grad_norm": 0.239769846200943,
+      "learning_rate": 1.0899347581163221e-05,
+      "loss": 0.5149,
+      "step": 850
+    },
+    {
+      "epoch": 0.27749245944403683,
+      "grad_norm": 0.3682393431663513,
+      "learning_rate": 1.0757162187628222e-05,
+      "loss": 0.6912,
+      "step": 851
+    },
+    {
+      "epoch": 0.2778185375397408,
+      "grad_norm": 0.2674311697483063,
+      "learning_rate": 1.0615857584873623e-05,
+      "loss": 0.6053,
+      "step": 852
+    },
+    {
+      "epoch": 0.2781446156354447,
+      "grad_norm": 0.26375558972358704,
+      "learning_rate": 1.0475435167518843e-05,
+      "loss": 0.5715,
+      "step": 853
+    },
+    {
+      "epoch": 0.27847069373114863,
+      "grad_norm": 0.2378285825252533,
+      "learning_rate": 1.0335896321476413e-05,
+      "loss": 0.5853,
+      "step": 854
+    },
+    {
+      "epoch": 0.27879677182685253,
+      "grad_norm": 0.2970817983150482,
+      "learning_rate": 1.0197242423938446e-05,
+      "loss": 0.5929,
+      "step": 855
+    },
+    {
+      "epoch": 0.2791228499225564,
+      "grad_norm": 0.30819419026374817,
+      "learning_rate": 1.0059474843362892e-05,
+      "loss": 0.577,
+      "step": 856
+    },
+    {
+      "epoch": 0.2794489280182604,
+      "grad_norm": 0.2650148868560791,
+      "learning_rate": 9.922594939460194e-06,
+      "loss": 0.5986,
+      "step": 857
+    },
+    {
+      "epoch": 0.2797750061139643,
+      "grad_norm": 0.33730190992355347,
+      "learning_rate": 9.786604063179728e-06,
+      "loss": 0.596,
+      "step": 858
+    },
+    {
+      "epoch": 0.2801010842096682,
+      "grad_norm": 0.2908703088760376,
+      "learning_rate": 9.651503556696516e-06,
+      "loss": 0.6086,
+      "step": 859
+    },
+    {
+      "epoch": 0.2804271623053721,
+      "grad_norm": 0.3620714545249939,
+      "learning_rate": 9.517294753398064e-06,
+      "loss": 0.6332,
+      "step": 860
+    },
+    {
+      "epoch": 0.2807532404010761,
+      "grad_norm": 0.25597527623176575,
+      "learning_rate": 9.383978977871021e-06,
+      "loss": 0.5788,
+      "step": 861
+    },
+    {
+      "epoch": 0.28107931849678,
+      "grad_norm": 0.23144716024398804,
+      "learning_rate": 9.251557545888312e-06,
+      "loss": 0.5323,
+      "step": 862
+    },
+    {
+      "epoch": 0.2814053965924839,
+      "grad_norm": 0.2824934422969818,
+      "learning_rate": 9.120031764395987e-06,
+      "loss": 0.5951,
+      "step": 863
+    },
+    {
+      "epoch": 0.2817314746881878,
+      "grad_norm": 0.2706288695335388,
+      "learning_rate": 8.989402931500434e-06,
+      "loss": 0.5499,
+      "step": 864
+    },
+    {
+      "epoch": 0.2820575527838917,
+      "grad_norm": 0.2830828130245209,
+      "learning_rate": 8.85967233645547e-06,
+      "loss": 0.6568,
+      "step": 865
+    },
+    {
+      "epoch": 0.28238363087959567,
+      "grad_norm": 0.2941051423549652,
+      "learning_rate": 8.730841259649725e-06,
+      "loss": 0.5747,
+      "step": 866
+    },
+    {
+      "epoch": 0.28270970897529957,
+      "grad_norm": 0.20262877643108368,
+      "learning_rate": 8.602910972593892e-06,
+      "loss": 0.6067,
+      "step": 867
+    },
+    {
+      "epoch": 0.2830357870710035,
+      "grad_norm": 0.25369739532470703,
+      "learning_rate": 8.475882737908248e-06,
+      "loss": 0.4986,
+      "step": 868
+    },
+    {
+      "epoch": 0.2833618651667074,
+      "grad_norm": 0.27120092511177063,
+      "learning_rate": 8.34975780931021e-06,
+      "loss": 0.632,
+      "step": 869
+    },
+    {
+      "epoch": 0.28368794326241137,
+      "grad_norm": 0.2529854476451874,
+      "learning_rate": 8.224537431601886e-06,
+      "loss": 0.556,
+      "step": 870
+    },
+    {
+      "epoch": 0.28401402135811527,
+      "grad_norm": 0.347957968711853,
+      "learning_rate": 8.100222840657878e-06,
+      "loss": 0.5932,
+      "step": 871
+    },
+    {
+      "epoch": 0.2843400994538192,
+      "grad_norm": 0.34991616010665894,
+      "learning_rate": 7.976815263412963e-06,
+      "loss": 0.6714,
+      "step": 872
+    },
+    {
+      "epoch": 0.2846661775495231,
+      "grad_norm": 0.27129024267196655,
+      "learning_rate": 7.854315917850163e-06,
+      "loss": 0.6417,
+      "step": 873
+    },
+    {
+      "epoch": 0.284992255645227,
+      "grad_norm": 0.20815640687942505,
+      "learning_rate": 7.73272601298851e-06,
+      "loss": 0.5793,
+      "step": 874
+    },
+    {
+      "epoch": 0.28531833374093096,
+      "grad_norm": 0.27495530247688293,
+      "learning_rate": 7.612046748871327e-06,
+      "loss": 0.4999,
+      "step": 875
+    },
+    {
+      "epoch": 0.28564441183663486,
+      "grad_norm": 0.25796929001808167,
+      "learning_rate": 7.492279316554207e-06,
+      "loss": 0.5908,
+      "step": 876
+    },
+    {
+      "epoch": 0.2859704899323388,
+      "grad_norm": 0.23965142667293549,
+      "learning_rate": 7.3734248980933395e-06,
+      "loss": 0.5887,
+      "step": 877
+    },
+    {
+      "epoch": 0.2862965680280427,
+      "grad_norm": 0.23252816498279572,
+      "learning_rate": 7.255484666533874e-06,
+      "loss": 0.5622,
+      "step": 878
+    },
+    {
+      "epoch": 0.28662264612374666,
+      "grad_norm": 0.31792140007019043,
+      "learning_rate": 7.138459785898266e-06,
+      "loss": 0.5411,
+      "step": 879
+    },
+    {
+      "epoch": 0.28694872421945056,
+      "grad_norm": 0.27042651176452637,
+      "learning_rate": 7.022351411174866e-06,
+      "loss": 0.6432,
+      "step": 880
+    },
+    {
+      "epoch": 0.28727480231515445,
+      "grad_norm": 0.2737380862236023,
+      "learning_rate": 6.907160688306425e-06,
+      "loss": 0.5333,
+      "step": 881
+    },
+    {
+      "epoch": 0.2876008804108584,
+      "grad_norm": 0.49083641171455383,
+      "learning_rate": 6.7928887541789055e-06,
+      "loss": 0.7366,
+      "step": 882
+    },
+    {
+      "epoch": 0.2879269585065623,
+      "grad_norm": 0.2284003347158432,
+      "learning_rate": 6.679536736610137e-06,
+      "loss": 0.4846,
+      "step": 883
+    },
+    {
+      "epoch": 0.28825303660226625,
+      "grad_norm": 0.27481234073638916,
+      "learning_rate": 6.5671057543387985e-06,
+      "loss": 0.5543,
+      "step": 884
+    },
+    {
+      "epoch": 0.28857911469797015,
+      "grad_norm": 0.2470434457063675,
+      "learning_rate": 6.455596917013273e-06,
+      "loss": 0.5875,
+      "step": 885
+    },
+    {
+      "epoch": 0.2889051927936741,
+      "grad_norm": 0.30581676959991455,
+      "learning_rate": 6.345011325180772e-06,
+      "loss": 0.565,
+      "step": 886
+    },
+    {
+      "epoch": 0.289231270889378,
+      "grad_norm": 0.22337621450424194,
+      "learning_rate": 6.235350070276447e-06,
+      "loss": 0.5746,
+      "step": 887
+    },
+    {
+      "epoch": 0.28955734898508195,
+      "grad_norm": 0.2908400595188141,
+      "learning_rate": 6.126614234612593e-06,
+      "loss": 0.4911,
+      "step": 888
+    },
+    {
+      "epoch": 0.28988342708078585,
+      "grad_norm": 0.26192542910575867,
+      "learning_rate": 6.018804891368035e-06,
+      "loss": 0.5577,
+      "step": 889
+    },
+    {
+      "epoch": 0.29020950517648975,
+      "grad_norm": 0.21913784742355347,
+      "learning_rate": 5.911923104577455e-06,
+      "loss": 0.4661,
+      "step": 890
+    },
+    {
+      "epoch": 0.2905355832721937,
+      "grad_norm": 0.24812312424182892,
+      "learning_rate": 5.805969929120947e-06,
+      "loss": 0.5104,
+      "step": 891
+    },
+    {
+      "epoch": 0.2908616613678976,
+      "grad_norm": 0.285163015127182,
+      "learning_rate": 5.700946410713548e-06,
+      "loss": 0.5446,
+      "step": 892
+    },
+    {
+      "epoch": 0.29118773946360155,
+      "grad_norm": 0.28700485825538635,
+      "learning_rate": 5.5968535858950345e-06,
+      "loss": 0.6212,
+      "step": 893
+    },
+    {
+      "epoch": 0.29151381755930544,
+      "grad_norm": 0.311657190322876,
+      "learning_rate": 5.49369248201953e-06,
+      "loss": 0.6656,
+      "step": 894
+    },
+    {
+      "epoch": 0.2918398956550094,
+      "grad_norm": 0.26617059111595154,
+      "learning_rate": 5.39146411724547e-06,
+      "loss": 0.6205,
+      "step": 895
+    },
+    {
+      "epoch": 0.2921659737507133,
+      "grad_norm": 0.23219060897827148,
+      "learning_rate": 5.290169500525577e-06,
+      "loss": 0.5324,
+      "step": 896
+    },
+    {
+      "epoch": 0.2924920518464172,
+      "grad_norm": 0.2956946790218353,
+      "learning_rate": 5.189809631596798e-06,
+      "loss": 0.5709,
+      "step": 897
+    },
+    {
+      "epoch": 0.29281812994212114,
+      "grad_norm": 0.24315470457077026,
+      "learning_rate": 5.0903855009705514e-06,
+      "loss": 0.566,
+      "step": 898
+    },
+    {
+      "epoch": 0.29314420803782504,
+      "grad_norm": 0.2868155539035797,
+      "learning_rate": 4.991898089922819e-06,
+      "loss": 0.6156,
+      "step": 899
+    },
+    {
+      "epoch": 0.293470286133529,
+      "grad_norm": 0.29666030406951904,
+      "learning_rate": 4.8943483704846475e-06,
+      "loss": 0.5783,
+      "step": 900
+    },
+    {
+      "epoch": 0.2937963642292329,
+      "grad_norm": 0.23827630281448364,
+      "learning_rate": 4.797737305432337e-06,
+      "loss": 0.5259,
+      "step": 901
+    },
+    {
+      "epoch": 0.29412244232493684,
+      "grad_norm": 0.33247071504592896,
+      "learning_rate": 4.702065848278126e-06,
+      "loss": 0.6135,
+      "step": 902
+    },
+    {
+      "epoch": 0.29444852042064074,
+      "grad_norm": 0.24986249208450317,
+      "learning_rate": 4.607334943260655e-06,
+      "loss": 0.5976,
+      "step": 903
+    },
+    {
+      "epoch": 0.2947745985163447,
+      "grad_norm": 0.31358757615089417,
+      "learning_rate": 4.513545525335705e-06,
+      "loss": 0.6994,
+      "step": 904
+    },
+    {
+      "epoch": 0.2951006766120486,
+      "grad_norm": 0.3073498010635376,
+      "learning_rate": 4.420698520166988e-06,
+      "loss": 0.5653,
+      "step": 905
+    },
+    {
+      "epoch": 0.2954267547077525,
+      "grad_norm": 0.24950246512889862,
+      "learning_rate": 4.328794844116946e-06,
+      "loss": 0.6577,
+      "step": 906
+    },
+    {
+      "epoch": 0.29575283280345643,
+      "grad_norm": 0.3969646990299225,
+      "learning_rate": 4.237835404237778e-06,
+      "loss": 0.5671,
+      "step": 907
+    },
+    {
+      "epoch": 0.29607891089916033,
+      "grad_norm": 0.33874574303627014,
+      "learning_rate": 4.147821098262405e-06,
+      "loss": 0.7208,
+      "step": 908
+    },
+    {
+      "epoch": 0.2964049889948643,
+      "grad_norm": 0.2954644560813904,
+      "learning_rate": 4.0587528145957235e-06,
+      "loss": 0.5627,
+      "step": 909
+    },
+    {
+      "epoch": 0.2967310670905682,
+      "grad_norm": 0.26160573959350586,
+      "learning_rate": 3.970631432305694e-06,
+      "loss": 0.5891,
+      "step": 910
+    },
+    {
+      "epoch": 0.29705714518627213,
+      "grad_norm": 0.27814751863479614,
+      "learning_rate": 3.883457821114811e-06,
+      "loss": 0.6133,
+      "step": 911
+    },
+    {
+      "epoch": 0.29738322328197603,
+      "grad_norm": 0.4170272946357727,
+      "learning_rate": 3.797232841391407e-06,
+      "loss": 0.5701,
+      "step": 912
+    },
+    {
+      "epoch": 0.29770930137768,
+      "grad_norm": 0.21299175918102264,
+      "learning_rate": 3.711957344141237e-06,
+      "loss": 0.5677,
+      "step": 913
+    },
+    {
+      "epoch": 0.2980353794733839,
+      "grad_norm": 0.3248412311077118,
+      "learning_rate": 3.627632170999029e-06,
+      "loss": 0.6522,
+      "step": 914
+    },
+    {
+      "epoch": 0.2983614575690878,
+      "grad_norm": 0.3027745485305786,
+      "learning_rate": 3.5442581542201923e-06,
+      "loss": 0.5193,
+      "step": 915
+    },
+    {
+      "epoch": 0.2986875356647917,
+      "grad_norm": 0.2598162591457367,
+      "learning_rate": 3.461836116672612e-06,
+      "loss": 0.5704,
+      "step": 916
+    },
+    {
+      "epoch": 0.2990136137604956,
+      "grad_norm": 0.2748893201351166,
+      "learning_rate": 3.380366871828522e-06,
+      "loss": 0.6077,
+      "step": 917
+    },
+    {
+      "epoch": 0.2993396918561996,
+      "grad_norm": 0.2935779094696045,
+      "learning_rate": 3.2998512237565005e-06,
+      "loss": 0.6494,
+      "step": 918
+    },
+    {
+      "epoch": 0.29966576995190347,
+      "grad_norm": 0.28127992153167725,
+      "learning_rate": 3.2202899671134546e-06,
+      "loss": 0.5611,
+      "step": 919
+    },
+    {
+      "epoch": 0.2999918480476074,
+      "grad_norm": 0.27138617634773254,
+      "learning_rate": 3.1416838871368924e-06,
+      "loss": 0.5692,
+      "step": 920
+    },
+    {
+      "epoch": 0.3003179261433113,
+      "grad_norm": 0.33780840039253235,
+      "learning_rate": 3.064033759637064e-06,
+      "loss": 0.5973,
+      "step": 921
+    },
+    {
+      "epoch": 0.3006440042390152,
+      "grad_norm": 0.2962702214717865,
+      "learning_rate": 2.9873403509894203e-06,
+      "loss": 0.5803,
+      "step": 922
+    },
+    {
+      "epoch": 0.30097008233471917,
+      "grad_norm": 0.3189234435558319,
+      "learning_rate": 2.9116044181269007e-06,
+      "loss": 0.6576,
+      "step": 923
+    },
+    {
+      "epoch": 0.30129616043042307,
+      "grad_norm": 0.39585644006729126,
+      "learning_rate": 2.836826708532603e-06,
+      "loss": 0.6375,
+      "step": 924
+    },
+    {
+      "epoch": 0.301622238526127,
+      "grad_norm": 0.37866878509521484,
+      "learning_rate": 2.7630079602323442e-06,
+      "loss": 0.6671,
+      "step": 925
+    },
+    {
+      "epoch": 0.3019483166218309,
+      "grad_norm": 0.23039337992668152,
+      "learning_rate": 2.690148901787337e-06,
+      "loss": 0.4663,
+      "step": 926
+    },
+    {
+      "epoch": 0.30227439471753487,
+      "grad_norm": 0.20355091989040375,
+      "learning_rate": 2.618250252287113e-06,
+      "loss": 0.5337,
+      "step": 927
+    },
+    {
+      "epoch": 0.30260047281323876,
+      "grad_norm": 0.2754460871219635,
+      "learning_rate": 2.5473127213422763e-06,
+      "loss": 0.5808,
+      "step": 928
+    },
+    {
+      "epoch": 0.3029265509089427,
+      "grad_norm": 0.2289859503507614,
+      "learning_rate": 2.4773370090776626e-06,
+      "loss": 0.5963,
+      "step": 929
+    },
+    {
+      "epoch": 0.3032526290046466,
+      "grad_norm": 0.3919331729412079,
+      "learning_rate": 2.4083238061252567e-06,
+      "loss": 0.5978,
+      "step": 930
+    },
+    {
+      "epoch": 0.3035787071003505,
+      "grad_norm": 0.28480133414268494,
+      "learning_rate": 2.3402737936175425e-06,
+      "loss": 0.6194,
+      "step": 931
+    },
+    {
+      "epoch": 0.30390478519605446,
+      "grad_norm": 0.24756163358688354,
+      "learning_rate": 2.273187643180652e-06,
+      "loss": 0.5772,
+      "step": 932
+    },
+    {
+      "epoch": 0.30423086329175836,
+      "grad_norm": 0.3628486096858978,
+      "learning_rate": 2.2070660169278166e-06,
+      "loss": 0.5918,
+      "step": 933
+    },
+    {
+      "epoch": 0.3045569413874623,
+      "grad_norm": 0.38187867403030396,
+      "learning_rate": 2.141909567452793e-06,
+      "loss": 0.5799,
+      "step": 934
+    },
+    {
+      "epoch": 0.3048830194831662,
+      "grad_norm": 0.32940131425857544,
+      "learning_rate": 2.0777189378234143e-06,
+      "loss": 0.5906,
+      "step": 935
+    },
+    {
+      "epoch": 0.30520909757887016,
+      "grad_norm": 0.2577744126319885,
+      "learning_rate": 2.014494761575314e-06,
+      "loss": 0.5325,
+      "step": 936
+    },
+    {
+      "epoch": 0.30553517567457406,
+      "grad_norm": 0.21027593314647675,
+      "learning_rate": 1.9522376627055583e-06,
+      "loss": 0.5189,
+      "step": 937
+    },
+    {
+      "epoch": 0.305861253770278,
+      "grad_norm": 0.31960952281951904,
+      "learning_rate": 1.8909482556666024e-06,
+      "loss": 0.6651,
+      "step": 938
+    },
+    {
+      "epoch": 0.3061873318659819,
+      "grad_norm": 0.28659772872924805,
+      "learning_rate": 1.8306271453601199e-06,
+      "loss": 0.6643,
+      "step": 939
+    },
+    {
+      "epoch": 0.3065134099616858,
+      "grad_norm": 0.3308188021183014,
+      "learning_rate": 1.771274927131139e-06,
+      "loss": 0.6336,
+      "step": 940
+    },
+    {
+      "epoch": 0.30683948805738975,
+      "grad_norm": 0.23489847779273987,
+      "learning_rate": 1.712892186762083e-06,
+      "loss": 0.5784,
+      "step": 941
+    },
+    {
+      "epoch": 0.30716556615309365,
+      "grad_norm": 0.2495911717414856,
+      "learning_rate": 1.6554795004670388e-06,
+      "loss": 0.4387,
+      "step": 942
+    },
+    {
+      "epoch": 0.3074916442487976,
+      "grad_norm": 0.26292648911476135,
+      "learning_rate": 1.5990374348860305e-06,
+      "loss": 0.5927,
+      "step": 943
+    },
+    {
+      "epoch": 0.3078177223445015,
+      "grad_norm": 0.31785205006599426,
+      "learning_rate": 1.543566547079467e-06,
+      "loss": 0.6102,
+      "step": 944
+    },
+    {
+      "epoch": 0.30814380044020545,
+      "grad_norm": 0.21520088613033295,
+      "learning_rate": 1.4890673845226133e-06,
+      "loss": 0.4613,
+      "step": 945
+    },
+    {
+      "epoch": 0.30846987853590935,
+      "grad_norm": 0.23349148035049438,
+      "learning_rate": 1.4355404851001952e-06,
+      "loss": 0.4913,
+      "step": 946
+    },
+    {
+      "epoch": 0.30879595663161324,
+      "grad_norm": 0.28058943152427673,
+      "learning_rate": 1.3829863771011253e-06,
+      "loss": 0.5491,
+      "step": 947
+    },
+    {
+      "epoch": 0.3091220347273172,
+      "grad_norm": 0.29834824800491333,
+      "learning_rate": 1.3314055792131964e-06,
+      "loss": 0.6751,
+      "step": 948
+    },
+    {
+      "epoch": 0.3094481128230211,
+      "grad_norm": 0.29814356565475464,
+      "learning_rate": 1.280798600518085e-06,
+      "loss": 0.5418,
+      "step": 949
+    },
+    {
+      "epoch": 0.30977419091872505,
+      "grad_norm": 0.24206741154193878,
+      "learning_rate": 1.231165940486234e-06,
+      "loss": 0.5177,
+      "step": 950
+    },
+    {
+      "epoch": 0.31010026901442894,
+      "grad_norm": 0.24119798839092255,
+      "learning_rate": 1.1825080889719563e-06,
+      "loss": 0.6286,
+      "step": 951
+    },
+    {
+      "epoch": 0.3104263471101329,
+      "grad_norm": 0.45291998982429504,
+      "learning_rate": 1.134825526208605e-06,
+      "loss": 0.6756,
+      "step": 952
+    },
+    {
+      "epoch": 0.3107524252058368,
+      "grad_norm": 0.36271461844444275,
+      "learning_rate": 1.0881187228038215e-06,
+      "loss": 0.6432,
+      "step": 953
+    },
+    {
+      "epoch": 0.31107850330154074,
+      "grad_norm": 0.3073123097419739,
+      "learning_rate": 1.0423881397349068e-06,
+      "loss": 0.6344,
+      "step": 954
+    },
+    {
+      "epoch": 0.31140458139724464,
+      "grad_norm": 0.29428645968437195,
+      "learning_rate": 9.976342283442463e-07,
+      "loss": 0.6755,
+      "step": 955
+    },
+    {
+      "epoch": 0.31173065949294854,
+      "grad_norm": 0.2660908102989197,
+      "learning_rate": 9.538574303348813e-07,
+      "loss": 0.5873,
+      "step": 956
+    },
+    {
+      "epoch": 0.3120567375886525,
+      "grad_norm": 0.27123311161994934,
+      "learning_rate": 9.110581777661331e-07,
+      "loss": 0.6825,
+      "step": 957
+    },
+    {
+      "epoch": 0.3123828156843564,
+      "grad_norm": 0.24501000344753265,
+      "learning_rate": 8.692368930493521e-07,
+      "loss": 0.5176,
+      "step": 958
+    },
+    {
+      "epoch": 0.31270889378006034,
+      "grad_norm": 0.19210560619831085,
+      "learning_rate": 8.283939889437209e-07,
+      "loss": 0.4684,
+      "step": 959
+    },
+    {
+      "epoch": 0.31303497187576423,
+      "grad_norm": 0.2760256230831146,
+      "learning_rate": 7.885298685522235e-07,
+      "loss": 0.5789,
+      "step": 960
+    },
+    {
+      "epoch": 0.3133610499714682,
+      "grad_norm": 0.3339109420776367,
+      "learning_rate": 7.496449253176274e-07,
+      "loss": 0.6615,
+      "step": 961
+    },
+    {
+      "epoch": 0.3136871280671721,
+      "grad_norm": 0.2638777196407318,
+      "learning_rate": 7.117395430186414e-07,
+      "loss": 0.6077,
+      "step": 962
+    },
+    {
+      "epoch": 0.31401320616287604,
+      "grad_norm": 0.34010159969329834,
+      "learning_rate": 6.748140957660631e-07,
+      "loss": 0.7114,
+      "step": 963
+    },
+    {
+      "epoch": 0.31433928425857993,
+      "grad_norm": 0.26364463567733765,
+      "learning_rate": 6.388689479991605e-07,
+      "loss": 0.5443,
+      "step": 964
+    },
+    {
+      "epoch": 0.31466536235428383,
+      "grad_norm": 0.24462150037288666,
+      "learning_rate": 6.039044544820404e-07,
+      "loss": 0.5761,
+      "step": 965
+    },
+    {
+      "epoch": 0.3149914404499878,
+      "grad_norm": 0.28215262293815613,
+      "learning_rate": 5.699209603001076e-07,
+      "loss": 0.6064,
+      "step": 966
+    },
+    {
+      "epoch": 0.3153175185456917,
+      "grad_norm": 0.23975957930088043,
+      "learning_rate": 5.369188008567672e-07,
+      "loss": 0.567,
+      "step": 967
+    },
+    {
+      "epoch": 0.31564359664139563,
+      "grad_norm": 0.251577764749527,
+      "learning_rate": 5.048983018699827e-07,
+      "loss": 0.6322,
+      "step": 968
+    },
+    {
+      "epoch": 0.3159696747370995,
+      "grad_norm": 0.34745168685913086,
+      "learning_rate": 4.738597793691679e-07,
+      "loss": 0.6558,
+      "step": 969
+    },
+    {
+      "epoch": 0.3162957528328035,
+      "grad_norm": 0.2653641402721405,
+      "learning_rate": 4.438035396920004e-07,
+      "loss": 0.5111,
+      "step": 970
+    },
+    {
+      "epoch": 0.3166218309285074,
+      "grad_norm": 0.2363329529762268,
+      "learning_rate": 4.1472987948143473e-07,
+      "loss": 0.5657,
+      "step": 971
+    },
+    {
+      "epoch": 0.31694790902421127,
+      "grad_norm": 0.37044990062713623,
+      "learning_rate": 3.866390856827495e-07,
+      "loss": 0.7043,
+      "step": 972
+    },
+    {
+      "epoch": 0.3172739871199152,
+      "grad_norm": 0.2447771430015564,
+      "learning_rate": 3.595314355407609e-07,
+      "loss": 0.5734,
+      "step": 973
+    },
+    {
+      "epoch": 0.3176000652156191,
+      "grad_norm": 0.36151984333992004,
+      "learning_rate": 3.3340719659701313e-07,
+      "loss": 0.6627,
+      "step": 974
+    },
+    {
+      "epoch": 0.3179261433113231,
+      "grad_norm": 0.256767213344574,
+      "learning_rate": 3.0826662668720364e-07,
+      "loss": 0.5589,
+      "step": 975
+    },
+    {
+      "epoch": 0.31825222140702697,
+      "grad_norm": 0.2616105079650879,
+      "learning_rate": 2.841099739386066e-07,
+      "loss": 0.5758,
+      "step": 976
+    },
+    {
+      "epoch": 0.3185782995027309,
+      "grad_norm": 0.20511144399642944,
+      "learning_rate": 2.609374767676309e-07,
+      "loss": 0.5115,
+      "step": 977
+    },
+    {
+      "epoch": 0.3189043775984348,
+      "grad_norm": 0.2790910601615906,
+      "learning_rate": 2.387493638774774e-07,
+      "loss": 0.6598,
+      "step": 978
+    },
+    {
+      "epoch": 0.31923045569413877,
+      "grad_norm": 0.2062654048204422,
+      "learning_rate": 2.175458542558517e-07,
+      "loss": 0.5254,
+      "step": 979
+    },
+    {
+      "epoch": 0.31955653378984267,
+      "grad_norm": 0.3161718249320984,
+      "learning_rate": 1.973271571728441e-07,
+      "loss": 0.6436,
+      "step": 980
+    },
+    {
+      "epoch": 0.31988261188554656,
+      "grad_norm": 0.24626140296459198,
+      "learning_rate": 1.7809347217881966e-07,
+      "loss": 0.5604,
+      "step": 981
+    },
+    {
+      "epoch": 0.3202086899812505,
+      "grad_norm": 0.2532305419445038,
+      "learning_rate": 1.598449891024978e-07,
+      "loss": 0.633,
+      "step": 982
+    },
+    {
+      "epoch": 0.3205347680769544,
+      "grad_norm": 0.2982385456562042,
+      "learning_rate": 1.425818880490315e-07,
+      "loss": 0.5065,
+      "step": 983
+    },
+    {
+      "epoch": 0.32086084617265836,
+      "grad_norm": 0.3654409646987915,
+      "learning_rate": 1.2630433939825327e-07,
+      "loss": 0.699,
+      "step": 984
+    },
+    {
+      "epoch": 0.32118692426836226,
+      "grad_norm": 0.3042480945587158,
+      "learning_rate": 1.1101250380300965e-07,
+      "loss": 0.6072,
+      "step": 985
+    },
+    {
+      "epoch": 0.3215130023640662,
+      "grad_norm": 0.344954252243042,
+      "learning_rate": 9.670653218752934e-08,
+      "loss": 0.6552,
+      "step": 986
+    },
+    {
+      "epoch": 0.3218390804597701,
+      "grad_norm": 0.23715342581272125,
+      "learning_rate": 8.33865657459909e-08,
+      "loss": 0.5171,
+      "step": 987
+    },
+    {
+      "epoch": 0.32216515855547406,
+      "grad_norm": 0.22725951671600342,
+      "learning_rate": 7.105273594107953e-08,
+      "loss": 0.5189,
+      "step": 988
+    },
+    {
+      "epoch": 0.32249123665117796,
+      "grad_norm": 0.3205645680427551,
+      "learning_rate": 5.970516450271025e-08,
+      "loss": 0.6342,
+      "step": 989
+    },
+    {
+      "epoch": 0.32281731474688186,
+      "grad_norm": 0.22355954349040985,
+      "learning_rate": 4.934396342684e-08,
+      "loss": 0.56,
+      "step": 990
+    },
+    {
+      "epoch": 0.3231433928425858,
+      "grad_norm": 0.30192822217941284,
+      "learning_rate": 3.996923497434635e-08,
+      "loss": 0.6033,
+      "step": 991
+    },
+    {
+      "epoch": 0.3234694709382897,
+      "grad_norm": 0.24073319137096405,
+      "learning_rate": 3.1581071670006015e-08,
+      "loss": 0.5943,
+      "step": 992
+    },
+    {
+      "epoch": 0.32379554903399366,
+      "grad_norm": 0.24182894825935364,
+      "learning_rate": 2.417955630159563e-08,
+      "loss": 0.5463,
+      "step": 993
+    },
+    {
+      "epoch": 0.32412162712969755,
+      "grad_norm": 0.28949111700057983,
+      "learning_rate": 1.7764761919103477e-08,
+      "loss": 0.5812,
+      "step": 994
+    },
+    {
+      "epoch": 0.3244477052254015,
+      "grad_norm": 0.2450643628835678,
+      "learning_rate": 1.2336751833941229e-08,
+      "loss": 0.5409,
+      "step": 995
+    },
+    {
+      "epoch": 0.3247737833211054,
+      "grad_norm": 0.25702303647994995,
+      "learning_rate": 7.895579618388827e-09,
+      "loss": 0.5836,
+      "step": 996
+    },
+    {
+      "epoch": 0.3250998614168093,
+      "grad_norm": 0.29124775528907776,
+      "learning_rate": 4.4412891050171765e-09,
+      "loss": 0.5742,
+      "step": 997
+    },
+    {
+      "epoch": 0.32542593951251325,
+      "grad_norm": 0.25140178203582764,
+      "learning_rate": 1.973914386288467e-09,
+      "loss": 0.5419,
+      "step": 998
+    },
+    {
+      "epoch": 0.32575201760821715,
+      "grad_norm": 0.2724677622318268,
+      "learning_rate": 4.934798141786879e-10,
+      "loss": 0.5966,
+      "step": 999
+    },
+    {
+      "epoch": 0.3260780957039211,
+      "grad_norm": 0.24416160583496094,
+      "learning_rate": 0.0,
+      "loss": 0.6293,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "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": 7.53166934979117e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}