End of training

all_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 2.04,
-    "train_loss": 0.0,
-    "train_runtime": 3248433.7519,
+    "train_loss": 1.4677173125030247,
+    "train_runtime": 22959.9641,
     "train_samples": 16407928,
-    "train_samples_per_second": 0.005,
-    "train_steps_per_second": 0.0
+    "train_samples_per_second": 1454.296,
+    "train_steps_per_second": 0.089
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 2.04,
-    "train_loss": 0.0,
-    "train_runtime": 3248433.7519,
+    "train_loss": 1.4677173125030247,
+    "train_runtime": 22959.9641,
     "train_samples": 16407928,
-    "train_samples_per_second": 0.005,
-    "train_steps_per_second": 0.0
+    "train_samples_per_second": 1454.296,
+    "train_steps_per_second": 0.089
 }

trainer_state.json

@@ -9,2469 +9,2460 @@
   "log_history": [
     {
       "epoch": 0.0,
-      "learning_rate": 0.0013736056394868903,
-      "loss": 1.4723,
+      "learning_rate": 0.0012782749814122844,
+      "loss": 1.6279,
       "step": 5
     },
     {
       "epoch": 0.01,
-      "learning_rate": 0.0019425717247145285,
-      "loss": 1.568,
+      "learning_rate": 0.0018077538151554684,
+      "loss": 1.2509,
       "step": 10
     },
     {
       "epoch": 0.01,
-      "learning_rate": 0.0023791547571544325,
-      "loss": 1.2706,
+      "learning_rate": 0.0022140372138502386,
+      "loss": 1.4568,
       "step": 15
     },
     {
       "epoch": 0.02,
-      "learning_rate": 0.0027472112789737805,
-      "loss": 1.3258,
+      "learning_rate": 0.0025565499628245687,
+      "loss": 1.4557,
       "step": 20
     },
     {
       "epoch": 0.02,
-      "learning_rate": 0.003071475584169756,
-      "loss": 1.2558,
+      "learning_rate": 0.0028583097523751473,
+      "loss": 1.1046,
       "step": 25
     },
     {
       "epoch": 0.03,
-      "learning_rate": 0.0033646329245522657,
-      "loss": 0.8074,
+      "learning_rate": 0.0031311214554257476,
+      "loss": 1.3079,
       "step": 30
     },
     {
       "epoch": 0.03,
-      "learning_rate": 0.0036342189215581556,
-      "loss": 1.4713,
+      "learning_rate": 0.003381997707972616,
+      "loss": 1.1913,
       "step": 35
     },
     {
       "epoch": 0.04,
-      "learning_rate": 0.003885143449429057,
-      "loss": 1.2681,
+      "learning_rate": 0.0036155076303109367,
+      "loss": 1.415,
       "step": 40
     },
     {
       "epoch": 0.04,
-      "learning_rate": 0.004120816918460671,
-      "loss": 1.2291,
+      "learning_rate": 0.003834824944236853,
+      "loss": 1.2117,
       "step": 45
     },
     {
       "epoch": 0.05,
-      "learning_rate": 0.004343722427630694,
-      "loss": 1.306,
+      "learning_rate": 0.0040422604172722166,
+      "loss": 1.3286,
       "step": 50
     },
     {
       "epoch": 0.05,
-      "learning_rate": 0.004555734516094203,
-      "loss": 1.1953,
+      "learning_rate": 0.004239558492243069,
+      "loss": 1.2645,
       "step": 55
     },
     {
       "epoch": 0.06,
-      "learning_rate": 0.004758309514308865,
-      "loss": 1.2176,
+      "learning_rate": 0.004428074427700477,
+      "loss": 1.3262,
       "step": 60
     },
     {
       "epoch": 0.06,
-      "learning_rate": 0.0049526055654364855,
-      "loss": 1.2338,
+      "learning_rate": 0.0046088859896247685,
+      "loss": 1.3905,
       "step": 65
     },
     {
       "epoch": 0.07,
-      "learning_rate": 0.005139561687500467,
-      "loss": 1.381,
+      "learning_rate": 0.004782867026529596,
+      "loss": 1.2985,
       "step": 70
     },
     {
       "epoch": 0.07,
-      "learning_rate": 0.005319951765989316,
-      "loss": 1.5546,
+      "learning_rate": 0.004950737714883372,
+      "loss": 1.1367,
       "step": 75
     },
     {
       "epoch": 0.08,
-      "learning_rate": 0.005494422557947561,
-      "loss": 1.5126,
+      "learning_rate": 0.0051130999256491375,
+      "loss": 1.3175,
       "step": 80
     },
     {
       "epoch": 0.08,
-      "learning_rate": 0.005663521139548541,
-      "loss": 1.5132,
+      "learning_rate": 0.005270462766947299,
+      "loss": 1.4444,
       "step": 85
     },
     {
       "epoch": 0.09,
-      "learning_rate": 0.005827715174143585,
-      "loss": 1.2736,
+      "learning_rate": 0.0054232614454664055,
+      "loss": 1.2882,
       "step": 90
     },
     {
       "epoch": 0.09,
-      "learning_rate": 0.005987408170800916,
-      "loss": 1.1807,
+      "learning_rate": 0.005571871466032479,
+      "loss": 1.4907,
       "step": 95
     },
     {
       "epoch": 0.1,
-      "learning_rate": 0.006142951168339512,
-      "loss": 1.392,
+      "learning_rate": 0.0057166195047502946,
+      "loss": 1.4369,
       "step": 100
     },
     {
       "epoch": 0.1,
-      "learning_rate": 0.0062946518179668965,
-      "loss": 1.4068,
+      "learning_rate": 0.005857791861290061,
+      "loss": 1.2819,
       "step": 105
     },
     {
       "epoch": 0.11,
-      "learning_rate": 0.0064427815392316505,
-      "loss": 1.2436,
+      "learning_rate": 0.005995641118204179,
+      "loss": 1.3525,
       "step": 110
     },
     {
       "epoch": 0.11,
-      "learning_rate": 0.0065875812264513545,
-      "loss": 1.6382,
+      "learning_rate": 0.006130391451319013,
+      "loss": 1.3266,
       "step": 115
     },
     {
       "epoch": 0.12,
-      "learning_rate": 0.006729265849104531,
-      "loss": 1.5855,
+      "learning_rate": 0.006262242910851495,
+      "loss": 1.4452,
       "step": 120
     },
     {
       "epoch": 0.12,
-      "learning_rate": 0.006868028197434452,
-      "loss": 1.8619,
+      "learning_rate": 0.006391374907061422,
+      "loss": 1.5723,
       "step": 125
     },
     {
       "epoch": 0.13,
-      "learning_rate": 0.007004041959724749,
-      "loss": 1.4162,
+      "learning_rate": 0.006517949073958691,
+      "loss": 1.4178,
       "step": 130
     },
     {
       "epoch": 0.13,
-      "learning_rate": 0.007137464271463298,
-      "loss": 1.397,
+      "learning_rate": 0.0066421116415507145,
+      "loss": 1.4234,
       "step": 135
     },
     {
       "epoch": 0.14,
-      "learning_rate": 0.007268437843116311,
-      "loss": 1.376,
+      "learning_rate": 0.006763995415945232,
+      "loss": 1.5096,
       "step": 140
     },
     {
       "epoch": 0.14,
-      "learning_rate": 0.0073970927486462865,
-      "loss": 1.5437,
+      "learning_rate": 0.006883721443741945,
+      "loss": 1.3332,
       "step": 145
     },
     {
       "epoch": 0.15,
-      "learning_rate": 0.007523547938632788,
-      "loss": 1.8505,
+      "learning_rate": 0.007001400420140049,
+      "loss": 1.5342,
       "step": 150
     },
     {
       "epoch": 0.15,
-      "learning_rate": 0.0076479125281174514,
-      "loss": 1.3138,
+      "learning_rate": 0.007117133887404524,
+      "loss": 1.5825,
       "step": 155
     },
     {
       "epoch": 0.16,
-      "learning_rate": 0.007770286898858114,
-      "loss": 1.6157,
+      "learning_rate": 0.007231015260621873,
+      "loss": 1.3028,
       "step": 160
     },
     {
       "epoch": 0.16,
-      "learning_rate": 0.00789076364767037,
-      "loss": 1.4061,
+      "learning_rate": 0.007343130710225101,
+      "loss": 1.5338,
       "step": 165
     },
     {
       "epoch": 0.17,
-      "learning_rate": 0.008009428406336274,
-      "loss": 1.4829,
+      "learning_rate": 0.0074535599249993005,
+      "loss": 1.4564,
       "step": 170
     },
     {
       "epoch": 0.17,
-      "learning_rate": 0.008126360553720012,
-      "loss": 1.3897,
+      "learning_rate": 0.007562376774775252,
+      "loss": 1.319,
       "step": 175
     },
     {
       "epoch": 0.18,
-      "learning_rate": 0.008241633836921341,
-      "loss": 1.3801,
+      "learning_rate": 0.007669649888473706,
+      "loss": 1.6363,
       "step": 180
     },
     {
       "epoch": 0.18,
-      "learning_rate": 0.008355316915277182,
-      "loss": 1.6133,
+      "learning_rate": 0.007775443160352296,
+      "loss": 1.4293,
       "step": 185
     },
     {
       "epoch": 0.19,
-      "learning_rate": 0.008467473838610143,
-      "loss": 1.6826,
+      "learning_rate": 0.007879816195062792,
+      "loss": 1.4431,
       "step": 190
     },
     {
       "epoch": 0.19,
-      "learning_rate": 0.008578164469184382,
-      "loss": 1.6101,
+      "learning_rate": 0.007982824700322464,
+      "loss": 1.3983,
       "step": 195
     },
     {
       "epoch": 0.2,
-      "learning_rate": 0.008687444855261389,
-      "loss": 1.7164,
+      "learning_rate": 0.008084520834544433,
+      "loss": 1.448,
       "step": 200
     },
     {
       "epoch": 0.2,
-      "learning_rate": 0.008795367562872955,
-      "loss": 1.6603,
+      "learning_rate": 0.008184953515585021,
+      "loss": 1.6799,
       "step": 205
     },
     {
       "epoch": 0.21,
-      "learning_rate": 0.008901981971385245,
-      "loss": 1.3604,
+      "learning_rate": 0.008284168695795142,
+      "loss": 1.5823,
       "step": 210
     },
     {
       "epoch": 0.21,
-      "learning_rate": 0.009007334537569819,
-      "loss": 1.5517,
+      "learning_rate": 0.008382209607764337,
+      "loss": 1.4564,
       "step": 215
     },
     {
       "epoch": 0.22,
-      "learning_rate": 0.009111469032188405,
-      "loss": 1.6048,
+      "learning_rate": 0.008479116984486139,
+      "loss": 1.576,
       "step": 220
     },
     {
       "epoch": 0.22,
-      "learning_rate": 0.009214426752509268,
-      "loss": 1.6281,
+      "learning_rate": 0.008574929257125443,
+      "loss": 1.5307,
       "step": 225
     },
     {
       "epoch": 0.23,
-      "learning_rate": 0.009316246713681893,
-      "loss": 2.1204,
+      "learning_rate": 0.00866968273311143,
+      "loss": 1.4519,
       "step": 230
     },
     {
       "epoch": 0.23,
-      "learning_rate": 0.009416965821485117,
-      "loss": 1.6439,
+      "learning_rate": 0.008763411756896685,
+      "loss": 1.6421,
       "step": 235
     },
     {
       "epoch": 0.24,
-      "learning_rate": 0.00951661902861773,
-      "loss": 1.6982,
+      "learning_rate": 0.008856148855400954,
+      "loss": 1.321,
       "step": 240
     },
     {
       "epoch": 0.24,
-      "learning_rate": 0.009615239476408233,
-      "loss": 1.857,
+      "learning_rate": 0.008947924869885989,
+      "loss": 1.3645,
       "step": 245
     },
     {
       "epoch": 0.25,
-      "learning_rate": 0.009712858623572643,
-      "loss": 1.5612,
+      "learning_rate": 0.009038769075777341,
+      "loss": 1.6528,
       "step": 250
     },
     {
       "epoch": 0.25,
-      "learning_rate": 0.009809506363438459,
-      "loss": 1.4843,
+      "learning_rate": 0.00912870929175277,
+      "loss": 1.5315,
       "step": 255
     },
     {
       "epoch": 0.26,
-      "learning_rate": 0.009905211130872971,
-      "loss": 1.81,
+      "learning_rate": 0.009217771979249537,
+      "loss": 1.6313,
       "step": 260
     },
     {
       "epoch": 0.26,
-      "learning_rate": 0.01,
-      "loss": 1.5551,
+      "learning_rate": 0.00930598233339944,
+      "loss": 1.6657,
       "step": 265
     },
     {
       "epoch": 0.27,
-      "learning_rate": 0.009971799492385787,
-      "loss": 1.8817,
+      "learning_rate": 0.009393364366277242,
+      "loss": 1.602,
       "step": 270
     },
     {
       "epoch": 0.27,
-      "learning_rate": 0.009943598984771575,
-      "loss": 1.7807,
+      "learning_rate": 0.00947994098324202,
+      "loss": 1.5175,
       "step": 275
     },
     {
       "epoch": 0.28,
-      "learning_rate": 0.009915398477157361,
-      "loss": 1.7788,
+      "learning_rate": 0.009565734053059192,
+      "loss": 1.4379,
       "step": 280
     },
     {
       "epoch": 0.28,
-      "learning_rate": 0.009887197969543148,
-      "loss": 1.9268,
+      "learning_rate": 0.009650764472411541,
+      "loss": 1.5895,
       "step": 285
     },
     {
       "epoch": 0.29,
-      "learning_rate": 0.009858997461928934,
-      "loss": 1.767,
+      "learning_rate": 0.009735052225338362,
+      "loss": 1.9034,
       "step": 290
     },
     {
       "epoch": 0.29,
-      "learning_rate": 0.00983079695431472,
-      "loss": 2.0681,
+      "learning_rate": 0.00981861643808179,
+      "loss": 1.5616,
       "step": 295
     },
     {
       "epoch": 0.3,
-      "learning_rate": 0.009802596446700508,
-      "loss": 1.8078,
+      "learning_rate": 0.009901475429766745,
+      "loss": 1.5318,
       "step": 300
     },
     {
       "epoch": 0.3,
-      "learning_rate": 0.009774395939086295,
-      "loss": 1.5819,
+      "learning_rate": 0.009983646759294876,
+      "loss": 1.6532,
       "step": 305
     },
     {
       "epoch": 0.31,
-      "learning_rate": 0.009746195431472081,
-      "loss": 1.3608,
+      "learning_rate": 0.009976905542725173,
+      "loss": 1.6601,
       "step": 310
     },
     {
       "epoch": 0.31,
-      "learning_rate": 0.00971799492385787,
-      "loss": 1.4035,
+      "learning_rate": 0.00994803747113164,
+      "loss": 1.9683,
       "step": 315
     },
     {
       "epoch": 0.32,
-      "learning_rate": 0.009689794416243654,
-      "loss": 1.7838,
+      "learning_rate": 0.009919169399538106,
+      "loss": 1.694,
       "step": 320
     },
     {
       "epoch": 0.32,
-      "learning_rate": 0.009661593908629442,
-      "loss": 1.462,
+      "learning_rate": 0.009890301327944573,
+      "loss": 1.4985,
       "step": 325
     },
     {
       "epoch": 0.33,
-      "learning_rate": 0.009633393401015228,
-      "loss": 1.8537,
+      "learning_rate": 0.009861433256351039,
+      "loss": 1.7896,
       "step": 330
     },
     {
       "epoch": 0.33,
-      "learning_rate": 0.009605192893401015,
-      "loss": 1.6035,
+      "learning_rate": 0.009832565184757506,
+      "loss": 1.6075,
       "step": 335
     },
     {
       "epoch": 0.34,
-      "learning_rate": 0.009576992385786803,
-      "loss": 1.7534,
+      "learning_rate": 0.009803697113163973,
+      "loss": 1.7854,
       "step": 340
     },
     {
       "epoch": 0.34,
-      "learning_rate": 0.00954879187817259,
-      "loss": 1.5117,
+      "learning_rate": 0.009774829041570439,
+      "loss": 1.6752,
       "step": 345
     },
     {
       "epoch": 0.35,
-      "learning_rate": 0.009520591370558376,
-      "loss": 1.6606,
+      "learning_rate": 0.009745960969976906,
+      "loss": 1.6439,
       "step": 350
     },
     {
       "epoch": 0.35,
-      "learning_rate": 0.009492390862944162,
-      "loss": 2.2899,
+      "learning_rate": 0.009717092898383371,
+      "loss": 1.9989,
       "step": 355
     },
     {
       "epoch": 0.36,
-      "learning_rate": 0.009464190355329949,
-      "loss": 1.6254,
+      "learning_rate": 0.009688224826789838,
+      "loss": 1.8198,
       "step": 360
     },
     {
       "epoch": 0.36,
-      "learning_rate": 0.009435989847715737,
-      "loss": 1.7699,
+      "learning_rate": 0.009659356755196306,
+      "loss": 1.8466,
       "step": 365
     },
     {
       "epoch": 0.37,
-      "learning_rate": 0.009407789340101523,
-      "loss": 1.5836,
+      "learning_rate": 0.009630488683602771,
+      "loss": 1.7034,
       "step": 370
     },
     {
       "epoch": 0.37,
-      "learning_rate": 0.00937958883248731,
-      "loss": 1.9447,
+      "learning_rate": 0.009601620612009238,
+      "loss": 1.6235,
       "step": 375
     },
     {
       "epoch": 0.38,
-      "learning_rate": 0.009351388324873098,
-      "loss": 1.7334,
+      "learning_rate": 0.009572752540415704,
+      "loss": 1.5958,
       "step": 380
     },
     {
       "epoch": 0.38,
-      "learning_rate": 0.009323187817258882,
-      "loss": 1.5246,
+      "learning_rate": 0.009543884468822171,
+      "loss": 1.7292,
       "step": 385
     },
     {
       "epoch": 0.39,
-      "learning_rate": 0.00929498730964467,
-      "loss": 1.7258,
+      "learning_rate": 0.009515016397228638,
+      "loss": 1.6004,
       "step": 390
     },
     {
       "epoch": 0.39,
-      "learning_rate": 0.009266786802030457,
-      "loss": 1.6757,
+      "learning_rate": 0.009486148325635104,
+      "loss": 1.8302,
       "step": 395
     },
     {
       "epoch": 0.4,
-      "learning_rate": 0.009238586294416243,
-      "loss": 1.7912,
+      "learning_rate": 0.009457280254041571,
+      "loss": 1.6322,
       "step": 400
     },
     {
       "epoch": 0.4,
-      "learning_rate": 0.009210385786802031,
-      "loss": 1.8435,
+      "learning_rate": 0.009428412182448036,
+      "loss": 1.4802,
       "step": 405
     },
     {
       "epoch": 0.41,
-      "learning_rate": 0.009182185279187818,
-      "loss": 2.0267,
+      "learning_rate": 0.009399544110854504,
+      "loss": 1.5544,
       "step": 410
     },
     {
       "epoch": 0.41,
-      "learning_rate": 0.009153984771573604,
-      "loss": 1.6059,
+      "learning_rate": 0.009370676039260971,
+      "loss": 1.6265,
       "step": 415
     },
     {
       "epoch": 0.42,
-      "learning_rate": 0.00912578426395939,
-      "loss": 1.7753,
+      "learning_rate": 0.009341807967667436,
+      "loss": 1.7169,
       "step": 420
     },
     {
       "epoch": 0.42,
-      "learning_rate": 0.009097583756345177,
-      "loss": 1.988,
+      "learning_rate": 0.009312939896073904,
+      "loss": 1.511,
       "step": 425
     },
     {
       "epoch": 0.43,
-      "learning_rate": 0.009069383248730965,
-      "loss": 1.5201,
+      "learning_rate": 0.009284071824480369,
+      "loss": 1.63,
       "step": 430
     },
     {
       "epoch": 0.43,
-      "learning_rate": 0.009041182741116751,
-      "loss": 1.397,
+      "learning_rate": 0.009255203752886836,
+      "loss": 1.6076,
       "step": 435
     },
     {
       "epoch": 0.44,
-      "learning_rate": 0.009012982233502538,
-      "loss": 1.7266,
+      "learning_rate": 0.009226335681293303,
+      "loss": 1.5973,
       "step": 440
     },
     {
       "epoch": 0.44,
-      "learning_rate": 0.008984781725888326,
-      "loss": 1.6747,
+      "learning_rate": 0.009197467609699769,
+      "loss": 1.6757,
       "step": 445
     },
     {
       "epoch": 0.45,
-      "learning_rate": 0.008956581218274112,
-      "loss": 1.5811,
+      "learning_rate": 0.009168599538106236,
+      "loss": 1.5146,
       "step": 450
     },
     {
       "epoch": 0.45,
-      "learning_rate": 0.008928380710659899,
-      "loss": 1.4325,
+      "learning_rate": 0.009139731466512702,
+      "loss": 1.459,
       "step": 455
     },
     {
       "epoch": 0.46,
-      "learning_rate": 0.008900180203045685,
-      "loss": 1.9487,
+      "learning_rate": 0.009110863394919169,
+      "loss": 1.4292,
       "step": 460
     },
     {
       "epoch": 0.46,
-      "learning_rate": 0.008871979695431471,
-      "loss": 1.4969,
+      "learning_rate": 0.009081995323325636,
+      "loss": 1.7206,
       "step": 465
     },
     {
       "epoch": 0.47,
-      "learning_rate": 0.00884377918781726,
-      "loss": 1.6207,
+      "learning_rate": 0.009053127251732102,
+      "loss": 1.491,
       "step": 470
     },
     {
       "epoch": 0.47,
-      "learning_rate": 0.008815578680203046,
-      "loss": 1.9026,
+      "learning_rate": 0.009024259180138567,
+      "loss": 1.5426,
       "step": 475
     },
     {
       "epoch": 0.48,
-      "learning_rate": 0.008787378172588832,
-      "loss": 1.6797,
+      "learning_rate": 0.008995391108545034,
+      "loss": 1.6884,
       "step": 480
     },
     {
       "epoch": 0.48,
-      "learning_rate": 0.00875917766497462,
-      "loss": 1.5079,
+      "learning_rate": 0.008966523036951501,
+      "loss": 1.6555,
       "step": 485
     },
     {
       "epoch": 0.49,
-      "learning_rate": 0.008730977157360405,
-      "loss": 1.5912,
+      "learning_rate": 0.008937654965357967,
+      "loss": 1.6036,
       "step": 490
     },
     {
       "epoch": 0.49,
-      "learning_rate": 0.008702776649746193,
-      "loss": 1.706,
+      "learning_rate": 0.008908786893764434,
+      "loss": 1.4405,
       "step": 495
     },
     {
       "epoch": 0.5,
-      "learning_rate": 0.00867457614213198,
-      "loss": 1.6559,
+      "learning_rate": 0.008879918822170901,
+      "loss": 1.5134,
       "step": 500
     },
     {
       "epoch": 0.5,
-      "learning_rate": 0.008646375634517766,
-      "loss": 1.6109,
+      "learning_rate": 0.008851050750577369,
+      "loss": 1.6087,
       "step": 505
     },
     {
       "epoch": 0.51,
-      "learning_rate": 0.008618175126903554,
-      "loss": 1.9024,
+      "learning_rate": 0.008822182678983834,
+      "loss": 1.6898,
       "step": 510
     },
     {
       "epoch": 0.51,
-      "learning_rate": 0.00858997461928934,
-      "loss": 1.4508,
+      "learning_rate": 0.0087933146073903,
+      "loss": 1.7172,
       "step": 515
     },
     {
       "epoch": 0.52,
-      "learning_rate": 0.008561774111675127,
-      "loss": 1.7011,
+      "learning_rate": 0.008764446535796767,
+      "loss": 1.6621,
       "step": 520
     },
     {
       "epoch": 0.52,
-      "learning_rate": 0.008533573604060913,
-      "loss": 1.4233,
+      "learning_rate": 0.008735578464203234,
+      "loss": 1.4144,
       "step": 525
     },
     {
       "epoch": 0.53,
-      "learning_rate": 0.0085053730964467,
-      "loss": 1.5886,
+      "learning_rate": 0.008706710392609701,
+      "loss": 1.685,
       "step": 530
     },
     {
       "epoch": 0.53,
-      "learning_rate": 0.008477172588832488,
-      "loss": 1.6088,
+      "learning_rate": 0.008677842321016167,
+      "loss": 1.6112,
       "step": 535
     },
     {
       "epoch": 0.54,
-      "learning_rate": 0.008448972081218274,
-      "loss": 2.3507,
+      "learning_rate": 0.008648974249422632,
+      "loss": 1.5041,
       "step": 540
     },
     {
       "epoch": 0.54,
-      "learning_rate": 0.00842077157360406,
-      "loss": 1.457,
+      "learning_rate": 0.0086201061778291,
+      "loss": 1.5515,
       "step": 545
     },
     {
       "epoch": 0.55,
-      "learning_rate": 0.008392571065989849,
-      "loss": 1.7145,
+      "learning_rate": 0.008591238106235567,
+      "loss": 1.711,
       "step": 550
     },
     {
       "epoch": 0.55,
-      "learning_rate": 0.008364370558375635,
-      "loss": 1.6431,
+      "learning_rate": 0.008562370034642034,
+      "loss": 1.6166,
       "step": 555
     },
     {
       "epoch": 0.56,
-      "learning_rate": 0.008336170050761421,
-      "loss": 1.6156,
+      "learning_rate": 0.0085335019630485,
+      "loss": 1.5324,
       "step": 560
     },
     {
       "epoch": 0.56,
-      "learning_rate": 0.008307969543147208,
-      "loss": 1.4988,
+      "learning_rate": 0.008504633891454965,
+      "loss": 1.4422,
       "step": 565
     },
     {
       "epoch": 0.57,
-      "learning_rate": 0.008279769035532994,
-      "loss": 1.6655,
+      "learning_rate": 0.008475765819861432,
+      "loss": 1.6069,
       "step": 570
     },
     {
       "epoch": 0.57,
-      "learning_rate": 0.008251568527918782,
-      "loss": 1.3964,
+      "learning_rate": 0.008446897748267899,
+      "loss": 1.6231,
       "step": 575
     },
     {
       "epoch": 0.58,
-      "learning_rate": 0.008223368020304569,
-      "loss": 1.5442,
+      "learning_rate": 0.008418029676674366,
+      "loss": 1.3719,
       "step": 580
     },
     {
       "epoch": 0.58,
-      "learning_rate": 0.008195167512690355,
-      "loss": 1.576,
+      "learning_rate": 0.008389161605080832,
+      "loss": 1.6578,
       "step": 585
     },
     {
       "epoch": 0.59,
-      "learning_rate": 0.008166967005076143,
-      "loss": 1.6838,
+      "learning_rate": 0.008360293533487297,
+      "loss": 1.6525,
       "step": 590
     },
     {
       "epoch": 0.59,
-      "learning_rate": 0.008138766497461928,
-      "loss": 1.5472,
+      "learning_rate": 0.008331425461893764,
+      "loss": 1.6094,
       "step": 595
     },
     {
       "epoch": 0.6,
-      "learning_rate": 0.008110565989847716,
-      "loss": 1.6603,
+      "learning_rate": 0.008302557390300232,
+      "loss": 1.633,
       "step": 600
     },
     {
       "epoch": 0.6,
-      "learning_rate": 0.008082365482233502,
-      "loss": 1.8896,
+      "learning_rate": 0.008273689318706697,
+      "loss": 1.7865,
       "step": 605
     },
     {
       "epoch": 0.61,
-      "learning_rate": 0.008054164974619289,
-      "loss": 1.3879,
+      "learning_rate": 0.008244821247113164,
+      "loss": 1.602,
       "step": 610
     },
     {
       "epoch": 0.61,
-      "learning_rate": 0.008025964467005077,
-      "loss": 1.7506,
+      "learning_rate": 0.00821595317551963,
+      "loss": 1.6723,
       "step": 615
     },
     {
       "epoch": 0.62,
-      "learning_rate": 0.007997763959390863,
-      "loss": 1.6107,
+      "learning_rate": 0.008187085103926097,
+      "loss": 1.5495,
       "step": 620
     },
     {
       "epoch": 0.62,
-      "learning_rate": 0.00796956345177665,
-      "loss": 1.6015,
+      "learning_rate": 0.008158217032332564,
+      "loss": 1.5716,
       "step": 625
     },
     {
       "epoch": 0.63,
-      "learning_rate": 0.007941362944162436,
-      "loss": 1.4905,
+      "learning_rate": 0.00812934896073903,
+      "loss": 1.6036,
       "step": 630
     },
     {
       "epoch": 0.63,
-      "learning_rate": 0.007913162436548223,
-      "loss": 1.7908,
+      "learning_rate": 0.008100480889145497,
+      "loss": 1.4638,
       "step": 635
     },
     {
       "epoch": 0.64,
-      "learning_rate": 0.00788496192893401,
-      "loss": 1.5734,
+      "learning_rate": 0.008071612817551962,
+      "loss": 1.6912,
       "step": 640
     },
     {
       "epoch": 0.64,
-      "learning_rate": 0.007856761421319797,
-      "loss": 1.4382,
+      "learning_rate": 0.00804274474595843,
+      "loss": 1.6037,
       "step": 645
     },
     {
       "epoch": 0.65,
-      "learning_rate": 0.007828560913705583,
-      "loss": 1.5483,
+      "learning_rate": 0.008013876674364897,
+      "loss": 1.5457,
       "step": 650
     },
     {
       "epoch": 0.65,
-      "learning_rate": 0.00780036040609137,
-      "loss": 1.4697,
+      "learning_rate": 0.007985008602771362,
+      "loss": 1.4561,
       "step": 655
     },
     {
       "epoch": 0.66,
-      "learning_rate": 0.007772159898477159,
-      "loss": 1.5878,
+      "learning_rate": 0.00795614053117783,
+      "loss": 1.4871,
       "step": 660
     },
     {
       "epoch": 0.66,
-      "learning_rate": 0.007743959390862944,
-      "loss": 1.9756,
+      "learning_rate": 0.007927272459584295,
+      "loss": 1.5253,
       "step": 665
     },
     {
       "epoch": 0.67,
-      "learning_rate": 0.007715758883248732,
-      "loss": 1.6656,
+      "learning_rate": 0.007898404387990762,
+      "loss": 1.5775,
       "step": 670
     },
     {
       "epoch": 0.67,
-      "learning_rate": 0.007687558375634519,
-      "loss": 1.6744,
+      "learning_rate": 0.00786953631639723,
+      "loss": 1.724,
       "step": 675
     },
     {
       "epoch": 0.68,
-      "learning_rate": 0.007659357868020306,
-      "loss": 1.762,
+      "learning_rate": 0.007840668244803695,
+      "loss": 1.6629,
       "step": 680
     },
     {
       "epoch": 0.68,
-      "learning_rate": 0.0076311573604060925,
-      "loss": 1.6982,
+      "learning_rate": 0.007811800173210163,
+      "loss": 1.519,
       "step": 685
     },
     {
       "epoch": 0.69,
-      "learning_rate": 0.007602956852791878,
-      "loss": 1.4908,
+      "learning_rate": 0.00778293210161663,
+      "loss": 1.5684,
       "step": 690
     },
     {
       "epoch": 0.69,
-      "learning_rate": 0.007574756345177666,
-      "loss": 1.6065,
+      "learning_rate": 0.007754064030023096,
+      "loss": 1.6109,
       "step": 695
     },
     {
       "epoch": 0.7,
-      "learning_rate": 0.0075465558375634526,
-      "loss": 1.2293,
+      "learning_rate": 0.007725195958429562,
+      "loss": 1.6378,
       "step": 700
     },
     {
       "epoch": 0.7,
-      "learning_rate": 0.007518355329949239,
-      "loss": 1.6052,
+      "learning_rate": 0.007696327886836028,
+      "loss": 1.4673,
       "step": 705
     },
     {
       "epoch": 0.71,
-      "learning_rate": 0.007490154822335025,
-      "loss": 1.3118,
+      "learning_rate": 0.007667459815242496,
+      "loss": 1.3671,
       "step": 710
     },
     {
       "epoch": 0.71,
-      "learning_rate": 0.0074619543147208135,
-      "loss": 1.1742,
+      "learning_rate": 0.007638591743648962,
+      "loss": 1.6971,
       "step": 715
     },
     {
       "epoch": 0.72,
-      "learning_rate": 0.0074337538071066,
-      "loss": 1.632,
+      "learning_rate": 0.007609723672055428,
+      "loss": 1.5656,
       "step": 720
     },
     {
       "epoch": 0.72,
-      "learning_rate": 0.007405553299492385,
-      "loss": 1.6181,
+      "learning_rate": 0.007580855600461894,
+      "loss": 1.6135,
       "step": 725
     },
     {
       "epoch": 0.73,
-      "learning_rate": 0.007377352791878174,
-      "loss": 1.6726,
+      "learning_rate": 0.007551987528868362,
+      "loss": 1.6247,
       "step": 730
     },
     {
       "epoch": 0.73,
-      "learning_rate": 0.00734915228426396,
-      "loss": 1.5759,
+      "learning_rate": 0.007523119457274827,
+      "loss": 1.4691,
       "step": 735
     },
     {
       "epoch": 0.74,
-      "learning_rate": 0.007320951776649747,
-      "loss": 1.3197,
+      "learning_rate": 0.007494251385681295,
+      "loss": 1.5108,
       "step": 740
     },
     {
       "epoch": 0.74,
-      "learning_rate": 0.0072927512690355335,
-      "loss": 1.8665,
+      "learning_rate": 0.007465383314087762,
+      "loss": 1.7138,
       "step": 745
     },
     {
       "epoch": 0.75,
-      "learning_rate": 0.007264550761421322,
-      "loss": 1.1938,
+      "learning_rate": 0.007436515242494226,
+      "loss": 1.4466,
       "step": 750
     },
     {
       "epoch": 0.75,
-      "learning_rate": 0.007236350253807108,
-      "loss": 1.2782,
+      "learning_rate": 0.007407647170900693,
+      "loss": 1.6503,
       "step": 755
     },
     {
       "epoch": 0.76,
-      "learning_rate": 0.0072081497461928936,
-      "loss": 1.6276,
+      "learning_rate": 0.00737877909930716,
+      "loss": 1.5161,
       "step": 760
     },
     {
       "epoch": 0.76,
-      "learning_rate": 0.007179949238578682,
-      "loss": 1.5382,
+      "learning_rate": 0.007349911027713627,
+      "loss": 1.5638,
       "step": 765
     },
     {
       "epoch": 0.77,
-      "learning_rate": 0.007151748730964467,
-      "loss": 1.5528,
+      "learning_rate": 0.007321042956120093,
+      "loss": 1.6321,
       "step": 770
     },
     {
       "epoch": 0.77,
-      "learning_rate": 0.0071235482233502545,
-      "loss": 1.663,
+      "learning_rate": 0.007292174884526559,
+      "loss": 1.5957,
       "step": 775
     },
     {
       "epoch": 0.78,
-      "learning_rate": 0.007095347715736042,
-      "loss": 1.5098,
+      "learning_rate": 0.0072633068129330245,
+      "loss": 1.4813,
       "step": 780
     },
     {
       "epoch": 0.78,
-      "learning_rate": 0.007067147208121828,
-      "loss": 1.781,
+      "learning_rate": 0.0072344387413394926,
+      "loss": 1.6003,
       "step": 785
     },
     {
       "epoch": 0.79,
-      "learning_rate": 0.007038946700507615,
-      "loss": 1.3787,
+      "learning_rate": 0.00720557066974596,
+      "loss": 1.474,
       "step": 790
     },
     {
       "epoch": 0.79,
-      "learning_rate": 0.007010746192893401,
-      "loss": 2.1853,
+      "learning_rate": 0.007176702598152426,
+      "loss": 1.6341,
       "step": 795
     },
     {
       "epoch": 0.8,
-      "learning_rate": 0.006982545685279189,
-      "loss": 1.6639,
+      "learning_rate": 0.007147834526558892,
+      "loss": 1.6502,
       "step": 800
     },
     {
       "epoch": 0.8,
-      "learning_rate": 0.006954345177664975,
-      "loss": 1.7152,
+      "learning_rate": 0.007118966454965358,
+      "loss": 1.4565,
       "step": 805
     },
     {
       "epoch": 0.81,
-      "learning_rate": 0.006926144670050761,
-      "loss": 1.8148,
+      "learning_rate": 0.007090098383371824,
+      "loss": 1.4588,
       "step": 810
     },
     {
       "epoch": 0.81,
-      "learning_rate": 0.006897944162436549,
-      "loss": 1.6564,
+      "learning_rate": 0.007061230311778292,
+      "loss": 1.5407,
       "step": 815
     },
     {
       "epoch": 0.82,
-      "learning_rate": 0.0068697436548223354,
-      "loss": 1.3046,
+      "learning_rate": 0.007032362240184759,
+      "loss": 1.3778,
       "step": 820
     },
     {
       "epoch": 0.82,
-      "learning_rate": 0.006841543147208123,
-      "loss": 1.3462,
+      "learning_rate": 0.007003494168591224,
+      "loss": 1.5602,
       "step": 825
     },
     {
       "epoch": 0.83,
-      "learning_rate": 0.006813342639593909,
-      "loss": 1.4969,
+      "learning_rate": 0.0069746260969976905,
+      "loss": 1.6861,
       "step": 830
     },
     {
       "epoch": 0.83,
-      "learning_rate": 0.006785142131979697,
-      "loss": 1.7367,
+      "learning_rate": 0.006945758025404158,
+      "loss": 1.5511,
       "step": 835
     },
     {
       "epoch": 0.84,
-      "learning_rate": 0.006756941624365483,
-      "loss": 1.5385,
+      "learning_rate": 0.006916889953810624,
+      "loss": 1.44,
       "step": 840
     },
     {
       "epoch": 0.84,
-      "learning_rate": 0.00672874111675127,
-      "loss": 1.5918,
+      "learning_rate": 0.006888021882217091,
+      "loss": 1.4159,
       "step": 845
     },
     {
       "epoch": 0.85,
-      "learning_rate": 0.006700540609137057,
-      "loss": 1.4092,
+      "learning_rate": 0.006859153810623558,
+      "loss": 1.6642,
       "step": 850
     },
     {
       "epoch": 0.85,
-      "learning_rate": 0.006672340101522844,
-      "loss": 1.6609,
+      "learning_rate": 0.006830285739030023,
+      "loss": 1.461,
       "step": 855
     },
     {
       "epoch": 0.86,
-      "learning_rate": 0.006644139593908631,
-      "loss": 1.6529,
+      "learning_rate": 0.00680141766743649,
+      "loss": 1.6532,
       "step": 860
     },
     {
       "epoch": 0.86,
-      "learning_rate": 0.006615939086294417,
-      "loss": 2.0543,
+      "learning_rate": 0.0067725495958429575,
+      "loss": 1.5106,
       "step": 865
     },
     {
       "epoch": 0.87,
-      "learning_rate": 0.0065877385786802046,
-      "loss": 1.7507,
+      "learning_rate": 0.006743681524249424,
+      "loss": 1.564,
       "step": 870
     },
     {
       "epoch": 0.87,
-      "learning_rate": 0.00655953807106599,
-      "loss": 1.3364,
+      "learning_rate": 0.00671481345265589,
+      "loss": 1.5954,
       "step": 875
     },
     {
       "epoch": 0.88,
-      "learning_rate": 0.0065313375634517765,
-      "loss": 1.5436,
+      "learning_rate": 0.006685945381062356,
+      "loss": 1.5075,
       "step": 880
     },
     {
       "epoch": 0.88,
-      "learning_rate": 0.006503137055837565,
-      "loss": 1.5808,
+      "learning_rate": 0.006657077309468823,
+      "loss": 1.4821,
       "step": 885
     },
     {
       "epoch": 0.89,
-      "learning_rate": 0.00647493654822335,
-      "loss": 1.1435,
+      "learning_rate": 0.006628209237875289,
+      "loss": 1.6993,
       "step": 890
     },
     {
       "epoch": 0.89,
-      "learning_rate": 0.006446736040609138,
-      "loss": 1.7857,
+      "learning_rate": 0.006599341166281756,
+      "loss": 1.3872,
       "step": 895
     },
     {
       "epoch": 0.9,
-      "learning_rate": 0.006418535532994924,
-      "loss": 1.506,
+      "learning_rate": 0.006570473094688223,
+      "loss": 1.684,
       "step": 900
     },
     {
       "epoch": 0.9,
-      "learning_rate": 0.006390335025380713,
-      "loss": 1.5276,
+      "learning_rate": 0.006541605023094689,
+      "loss": 1.3419,
       "step": 905
     },
     {
       "epoch": 0.91,
-      "learning_rate": 0.006362134517766498,
-      "loss": 1.7656,
+      "learning_rate": 0.006512736951501155,
+      "loss": 1.3974,
       "step": 910
     },
     {
       "epoch": 0.91,
-      "learning_rate": 0.006333934010152285,
-      "loss": 1.2465,
+      "learning_rate": 0.006483868879907622,
+      "loss": 1.5703,
       "step": 915
     },
     {
       "epoch": 0.92,
-      "learning_rate": 0.006305733502538073,
-      "loss": 1.1414,
+      "learning_rate": 0.006455000808314089,
+      "loss": 1.5572,
       "step": 920
     },
     {
       "epoch": 0.92,
-      "learning_rate": 0.006277532994923858,
-      "loss": 1.7263,
+      "learning_rate": 0.006426132736720555,
+      "loss": 1.5941,
       "step": 925
     },
     {
       "epoch": 0.93,
-      "learning_rate": 0.0062493324873096456,
-      "loss": 1.6158,
+      "learning_rate": 0.006397264665127021,
+      "loss": 1.706,
       "step": 930
     },
     {
       "epoch": 0.93,
-      "learning_rate": 0.006221131979695433,
-      "loss": 1.3896,
+      "learning_rate": 0.006368396593533487,
+      "loss": 1.5107,
       "step": 935
     },
     {
       "epoch": 0.94,
-      "learning_rate": 0.00619293147208122,
-      "loss": 1.8046,
+      "learning_rate": 0.0063395285219399545,
+      "loss": 1.4807,
       "step": 940
     },
     {
       "epoch": 0.94,
-      "learning_rate": 0.006164730964467006,
-      "loss": 1.6806,
+      "learning_rate": 0.006310660450346422,
+      "loss": 1.3222,
       "step": 945
     },
     {
       "epoch": 0.95,
-      "learning_rate": 0.006136530456852793,
-      "loss": 1.7003,
+      "learning_rate": 0.006281792378752889,
+      "loss": 1.4944,
       "step": 950
     },
     {
       "epoch": 0.95,
-      "learning_rate": 0.006108329949238579,
-      "loss": 1.4235,
+      "learning_rate": 0.006252924307159354,
+      "loss": 1.4851,
       "step": 955
     },
     {
       "epoch": 0.96,
-      "learning_rate": 0.006080129441624366,
-      "loss": 1.5528,
+      "learning_rate": 0.006224056235565821,
+      "loss": 1.3356,
       "step": 960
     },
     {
       "epoch": 0.96,
-      "learning_rate": 0.006051928934010154,
-      "loss": 1.6251,
+      "learning_rate": 0.006195188163972287,
+      "loss": 1.5368,
       "step": 965
     },
     {
       "epoch": 0.97,
-      "learning_rate": 0.006023728426395939,
-      "loss": 1.7944,
+      "learning_rate": 0.006166320092378754,
+      "loss": 1.4879,
       "step": 970
     },
     {
       "epoch": 0.97,
-      "learning_rate": 0.0059955279187817265,
-      "loss": 1.6361,
+      "learning_rate": 0.006137452020785221,
+      "loss": 1.7248,
       "step": 975
     },
     {
       "epoch": 0.98,
-      "learning_rate": 0.005967327411167512,
-      "loss": 1.6142,
+      "learning_rate": 0.006108583949191687,
+      "loss": 1.5369,
       "step": 980
     },
     {
       "epoch": 0.98,
-      "learning_rate": 0.0059391269035533,
-      "loss": 1.6014,
+      "learning_rate": 0.006079715877598152,
+      "loss": 1.6353,
       "step": 985
     },
     {
       "epoch": 0.99,
-      "learning_rate": 0.005910926395939088,
-      "loss": 1.5756,
+      "learning_rate": 0.0060508478060046205,
+      "loss": 1.4272,
       "step": 990
     },
     {
       "epoch": 0.99,
-      "learning_rate": 0.005882725888324872,
-      "loss": 1.2877,
+      "learning_rate": 0.006021979734411086,
+      "loss": 1.4956,
       "step": 995
     },
     {
       "epoch": 1.0,
-      "learning_rate": 0.005854525380710661,
-      "loss": 1.6494,
+      "learning_rate": 0.005993111662817553,
+      "loss": 1.6992,
       "step": 1000
     },
     {
       "epoch": 1.0,
-      "learning_rate": 0.005826324873096447,
-      "loss": 1.4704,
+      "learning_rate": 0.00596424359122402,
+      "loss": 1.49,
       "step": 1005
     },
     {
       "epoch": 1.01,
-      "learning_rate": 0.005798124365482234,
-      "loss": 1.3735,
+      "learning_rate": 0.005935375519630485,
+      "loss": 1.5019,
       "step": 1010
     },
     {
       "epoch": 1.01,
-      "learning_rate": 0.005769923857868021,
-      "loss": 1.3836,
+      "learning_rate": 0.005906507448036952,
+      "loss": 1.4958,
       "step": 1015
     },
     {
       "epoch": 1.02,
-      "learning_rate": 0.005741723350253808,
-      "loss": 1.4496,
+      "learning_rate": 0.005877639376443419,
+      "loss": 1.5754,
       "step": 1020
     },
     {
       "epoch": 1.02,
-      "learning_rate": 0.005713522842639595,
-      "loss": 1.4613,
+      "learning_rate": 0.005848771304849886,
+      "loss": 1.5251,
       "step": 1025
     },
     {
       "epoch": 1.03,
-      "learning_rate": 0.00568532233502538,
-      "loss": 1.4161,
+      "learning_rate": 0.005819903233256352,
+      "loss": 1.4752,
       "step": 1030
     },
     {
       "epoch": 1.03,
-      "learning_rate": 0.0056571218274111676,
-      "loss": 1.592,
+      "learning_rate": 0.005791035161662818,
+      "loss": 1.4888,
       "step": 1035
     },
     {
       "epoch": 1.04,
-      "learning_rate": 0.005628921319796955,
-      "loss": 1.585,
+      "learning_rate": 0.005762167090069283,
+      "loss": 1.7421,
       "step": 1040
     },
     {
       "epoch": 1.04,
-      "learning_rate": 0.005600720812182741,
-      "loss": 1.2444,
+      "learning_rate": 0.005733299018475752,
+      "loss": 1.4161,
       "step": 1045
     },
     {
       "epoch": 1.05,
-      "learning_rate": 0.005572520304568528,
-      "loss": 1.3218,
+      "learning_rate": 0.005704430946882218,
+      "loss": 1.5144,
       "step": 1050
     },
     {
       "epoch": 1.05,
-      "learning_rate": 0.005544319796954316,
-      "loss": 1.4106,
+      "learning_rate": 0.005675562875288685,
+      "loss": 1.4593,
       "step": 1055
     },
     {
       "epoch": 1.06,
-      "learning_rate": 0.005516119289340102,
-      "loss": 1.6454,
+      "learning_rate": 0.00564669480369515,
+      "loss": 1.5569,
       "step": 1060
     },
     {
       "epoch": 1.06,
-      "learning_rate": 0.005487918781725888,
-      "loss": 1.4884,
+      "learning_rate": 0.0056178267321016166,
+      "loss": 1.699,
       "step": 1065
     },
     {
       "epoch": 1.07,
-      "learning_rate": 0.005459718274111677,
-      "loss": 1.387,
+      "learning_rate": 0.005588958660508083,
+      "loss": 1.382,
       "step": 1070
     },
     {
       "epoch": 1.07,
-      "learning_rate": 0.005431517766497462,
-      "loss": 1.5909,
+      "learning_rate": 0.005560090588914552,
+      "loss": 1.5344,
       "step": 1075
     },
     {
       "epoch": 1.08,
-      "learning_rate": 0.005403317258883249,
-      "loss": 1.2899,
+      "learning_rate": 0.005531222517321017,
+      "loss": 1.5147,
       "step": 1080
     },
     {
       "epoch": 1.08,
-      "learning_rate": 0.005375116751269036,
-      "loss": 1.2528,
+      "learning_rate": 0.005502354445727483,
+      "loss": 1.483,
       "step": 1085
     },
     {
       "epoch": 1.09,
-      "learning_rate": 0.005346916243654824,
-      "loss": 1.3185,
+      "learning_rate": 0.005473486374133949,
+      "loss": 1.4432,
       "step": 1090
     },
     {
       "epoch": 1.09,
-      "learning_rate": 0.00531871573604061,
-      "loss": 1.751,
+      "learning_rate": 0.005444618302540416,
+      "loss": 1.4741,
       "step": 1095
     },
     {
       "epoch": 1.1,
-      "learning_rate": 0.005290515228426396,
-      "loss": 1.2762,
+      "learning_rate": 0.005415750230946883,
+      "loss": 1.3889,
       "step": 1100
     },
     {
       "epoch": 1.1,
-      "learning_rate": 0.005262314720812184,
-      "loss": 1.4006,
+      "learning_rate": 0.00538688215935335,
+      "loss": 1.5193,
       "step": 1105
     },
     {
       "epoch": 1.11,
-      "learning_rate": 0.0052341142131979695,
-      "loss": 1.3565,
+      "learning_rate": 0.005358014087759816,
+      "loss": 1.3495,
       "step": 1110
     },
     {
       "epoch": 1.11,
-      "learning_rate": 0.005205913705583757,
-      "loss": 1.7962,
+      "learning_rate": 0.005329146016166282,
+      "loss": 1.4738,
       "step": 1115
     },
     {
       "epoch": 1.12,
-      "learning_rate": 0.005177713197969543,
-      "loss": 1.3809,
+      "learning_rate": 0.005300277944572749,
+      "loss": 1.4875,
       "step": 1120
     },
     {
       "epoch": 1.12,
-      "learning_rate": 0.00514951269035533,
-      "loss": 1.6438,
+      "learning_rate": 0.005271409872979214,
+      "loss": 1.5795,
       "step": 1125
     },
     {
       "epoch": 1.13,
-      "learning_rate": 0.005121312182741118,
-      "loss": 1.4397,
+      "learning_rate": 0.0052425418013856825,
+      "loss": 1.5649,
       "step": 1130
     },
     {
       "epoch": 1.13,
-      "learning_rate": 0.005093111675126903,
-      "loss": 1.8387,
+      "learning_rate": 0.005213673729792149,
+      "loss": 1.3385,
       "step": 1135
     },
     {
       "epoch": 1.14,
-      "learning_rate": 0.005064911167512691,
-      "loss": 1.4722,
+      "learning_rate": 0.005184805658198614,
+      "loss": 1.6503,
       "step": 1140
     },
     {
       "epoch": 1.14,
-      "learning_rate": 0.005036710659898478,
-      "loss": 1.6683,
+      "learning_rate": 0.005155937586605081,
+      "loss": 1.4296,
       "step": 1145
     },
     {
       "epoch": 1.15,
-      "learning_rate": 0.005008510152284263,
-      "loss": 1.3579,
+      "learning_rate": 0.005127069515011548,
+      "loss": 1.4508,
       "step": 1150
     },
     {
       "epoch": 1.15,
-      "learning_rate": 0.004980309644670051,
-      "loss": 1.4072,
+      "learning_rate": 0.005098201443418014,
+      "loss": 1.6112,
       "step": 1155
     },
     {
       "epoch": 1.16,
-      "learning_rate": 0.004952109137055839,
-      "loss": 1.4808,
+      "learning_rate": 0.0050693333718244814,
+      "loss": 1.4412,
       "step": 1160
     },
     {
       "epoch": 1.16,
-      "learning_rate": 0.004923908629441625,
-      "loss": 1.37,
+      "learning_rate": 0.005040465300230946,
+      "loss": 1.3971,
       "step": 1165
     },
     {
       "epoch": 1.17,
-      "learning_rate": 0.004895708121827412,
-      "loss": 1.4239,
+      "learning_rate": 0.005011597228637413,
+      "loss": 1.5099,
       "step": 1170
     },
     {
       "epoch": 1.17,
-      "learning_rate": 0.004867507614213198,
-      "loss": 1.6803,
+      "learning_rate": 0.0049827291570438805,
+      "loss": 1.5568,
       "step": 1175
     },
     {
       "epoch": 1.18,
-      "learning_rate": 0.004839307106598985,
-      "loss": 1.3899,
+      "learning_rate": 0.004953861085450347,
+      "loss": 1.5243,
       "step": 1180
     },
     {
       "epoch": 1.18,
-      "learning_rate": 0.004811106598984772,
-      "loss": 1.209,
+      "learning_rate": 0.004924993013856814,
+      "loss": 1.2718,
       "step": 1185
     },
     {
       "epoch": 1.19,
-      "learning_rate": 0.004782906091370559,
-      "loss": 1.3444,
+      "learning_rate": 0.0048961249422632795,
+      "loss": 1.6473,
       "step": 1190
     },
     {
       "epoch": 1.19,
-      "learning_rate": 0.004754705583756346,
-      "loss": 1.5994,
+      "learning_rate": 0.004867256870669747,
+      "loss": 1.5639,
       "step": 1195
     },
     {
       "epoch": 1.2,
-      "learning_rate": 0.004726505076142133,
-      "loss": 1.7138,
+      "learning_rate": 0.004838388799076213,
+      "loss": 1.3558,
       "step": 1200
     },
     {
       "epoch": 1.2,
-      "learning_rate": 0.0046983045685279196,
-      "loss": 1.6732,
+      "learning_rate": 0.004809520727482679,
+      "loss": 1.4211,
       "step": 1205
     },
     {
       "epoch": 1.21,
-      "learning_rate": 0.004670104060913706,
-      "loss": 1.6533,
+      "learning_rate": 0.004780652655889147,
+      "loss": 1.5578,
       "step": 1210
     },
     {
       "epoch": 1.21,
-      "learning_rate": 0.004641903553299492,
-      "loss": 1.4821,
+      "learning_rate": 0.004751784584295613,
+      "loss": 1.6661,
       "step": 1215
     },
     {
       "epoch": 1.22,
-      "learning_rate": 0.00461370304568528,
-      "loss": 1.9074,
+      "learning_rate": 0.0047229165127020785,
+      "loss": 1.3678,
       "step": 1220
     },
     {
       "epoch": 1.22,
-      "learning_rate": 0.004585502538071067,
-      "loss": 1.6595,
+      "learning_rate": 0.004694048441108546,
+      "loss": 1.4871,
       "step": 1225
     },
     {
       "epoch": 1.23,
-      "learning_rate": 0.004557302030456853,
-      "loss": 1.5424,
+      "learning_rate": 0.004665180369515012,
+      "loss": 1.4524,
       "step": 1230
     },
     {
       "epoch": 1.23,
-      "learning_rate": 0.0045291015228426405,
-      "loss": 1.669,
+      "learning_rate": 0.004636312297921479,
+      "loss": 1.4723,
       "step": 1235
     },
     {
       "epoch": 1.24,
-      "learning_rate": 0.004500901015228428,
-      "loss": 1.7157,
+      "learning_rate": 0.004607444226327946,
+      "loss": 1.3289,
       "step": 1240
     },
     {
       "epoch": 1.24,
-      "learning_rate": 0.004472700507614213,
-      "loss": 1.3508,
+      "learning_rate": 0.004578576154734411,
+      "loss": 1.5643,
       "step": 1245
     },
     {
       "epoch": 1.25,
-      "learning_rate": 0.0044445000000000005,
-      "loss": 1.6086,
+      "learning_rate": 0.004549708083140878,
+      "loss": 1.2773,
       "step": 1250
     },
     {
       "epoch": 1.25,
-      "learning_rate": 0.004416299492385787,
-      "loss": 2.1126,
+      "learning_rate": 0.004520840011547345,
+      "loss": 1.4601,
       "step": 1255
     },
     {
       "epoch": 1.26,
-      "learning_rate": 0.004388098984771574,
-      "loss": 1.2235,
+      "learning_rate": 0.004491971939953812,
+      "loss": 1.2864,
       "step": 1260
     },
     {
       "epoch": 1.26,
-      "learning_rate": 0.0043598984771573614,
-      "loss": 1.3403,
+      "learning_rate": 0.004463103868360278,
+      "loss": 1.3144,
       "step": 1265
     },
     {
       "epoch": 1.27,
-      "learning_rate": 0.004331697969543148,
-      "loss": 1.4274,
+      "learning_rate": 0.004434235796766744,
+      "loss": 1.4182,
       "step": 1270
     },
     {
       "epoch": 1.27,
-      "learning_rate": 0.004303497461928934,
-      "loss": 1.4858,
+      "learning_rate": 0.004405367725173211,
+      "loss": 1.5354,
       "step": 1275
     },
     {
       "epoch": 1.28,
-      "learning_rate": 0.0042752969543147215,
-      "loss": 1.4251,
+      "learning_rate": 0.004376499653579677,
+      "loss": 1.4776,
       "step": 1280
     },
     {
       "epoch": 1.28,
-      "learning_rate": 0.004247096446700508,
-      "loss": 1.2581,
+      "learning_rate": 0.0043476315819861435,
+      "loss": 1.5618,
       "step": 1285
     },
     {
       "epoch": 1.29,
-      "learning_rate": 0.004218895939086295,
-      "loss": 1.3023,
+      "learning_rate": 0.004318763510392611,
+      "loss": 1.3853,
       "step": 1290
     },
     {
       "epoch": 1.29,
-      "learning_rate": 0.0041906954314720815,
-      "loss": 1.2824,
+      "learning_rate": 0.004289895438799076,
+      "loss": 1.4081,
       "step": 1295
     },
     {
       "epoch": 1.3,
-      "learning_rate": 0.004162494923857869,
-      "loss": 1.6138,
+      "learning_rate": 0.0042610273672055434,
+      "loss": 1.2231,
       "step": 1300
     },
     {
       "epoch": 1.3,
-      "learning_rate": 0.004134294416243656,
-      "loss": 1.6641,
+      "learning_rate": 0.00423215929561201,
+      "loss": 1.4096,
       "step": 1305
     },
     {
       "epoch": 1.31,
-      "learning_rate": 0.0041060939086294415,
-      "loss": 1.2635,
+      "learning_rate": 0.004203291224018476,
+      "loss": 1.724,
       "step": 1310
     },
     {
       "epoch": 1.31,
-      "learning_rate": 0.004077893401015229,
-      "loss": 1.5835,
+      "learning_rate": 0.004174423152424943,
+      "loss": 1.2597,
       "step": 1315
     },
     {
       "epoch": 1.32,
-      "learning_rate": 0.004049692893401016,
-      "loss": 1.6182,
+      "learning_rate": 0.004145555080831409,
+      "loss": 1.3268,
       "step": 1320
     },
     {
       "epoch": 1.32,
-      "learning_rate": 0.0040214923857868025,
-      "loss": 1.2402,
+      "learning_rate": 0.004116687009237876,
+      "loss": 1.3372,
       "step": 1325
     },
     {
       "epoch": 1.33,
-      "learning_rate": 0.00399329187817259,
-      "loss": 1.6024,
+      "learning_rate": 0.004087818937644342,
+      "loss": 1.4024,
       "step": 1330
     },
     {
       "epoch": 1.33,
-      "learning_rate": 0.003965091370558376,
-      "loss": 1.3873,
+      "learning_rate": 0.004058950866050809,
+      "loss": 1.3837,
       "step": 1335
     },
     {
       "epoch": 1.34,
-      "learning_rate": 0.003936890862944163,
-      "loss": 1.5199,
+      "learning_rate": 0.004030082794457276,
+      "loss": 1.4427,
       "step": 1340
     },
     {
       "epoch": 1.34,
-      "learning_rate": 0.00390869035532995,
-      "loss": 1.3065,
+      "learning_rate": 0.004001214722863742,
+      "loss": 1.5162,
       "step": 1345
     },
     {
       "epoch": 1.35,
-      "learning_rate": 0.0038804898477157357,
-      "loss": 1.3709,
+      "learning_rate": 0.003972346651270208,
+      "loss": 1.5025,
       "step": 1350
     },
     {
       "epoch": 1.35,
-      "learning_rate": 0.003852289340101523,
-      "loss": 1.5692,
+      "learning_rate": 0.003943478579676675,
+      "loss": 1.3713,
       "step": 1355
     },
     {
       "epoch": 1.36,
-      "learning_rate": 0.0038240888324873098,
-      "loss": 1.2774,
+      "learning_rate": 0.003914610508083141,
+      "loss": 1.2259,
       "step": 1360
     },
     {
       "epoch": 1.36,
-      "learning_rate": 0.003795888324873097,
-      "loss": 1.3966,
+      "learning_rate": 0.003885742436489608,
+      "loss": 1.4999,
       "step": 1365
     },
     {
       "epoch": 1.37,
-      "learning_rate": 0.0037676878172588843,
-      "loss": 1.0405,
+      "learning_rate": 0.003856874364896074,
+      "loss": 1.3068,
       "step": 1370
     },
     {
       "epoch": 1.37,
-      "learning_rate": 0.0037394873096446707,
-      "loss": 1.6305,
+      "learning_rate": 0.003828006293302541,
+      "loss": 1.4383,
       "step": 1375
     },
     {
       "epoch": 1.38,
-      "learning_rate": 0.003711286802030457,
-      "loss": 1.3404,
+      "learning_rate": 0.003799138221709007,
+      "loss": 1.3846,
       "step": 1380
     },
     {
       "epoch": 1.38,
-      "learning_rate": 0.0036830862944162435,
-      "loss": 1.5225,
+      "learning_rate": 0.0037702701501154735,
+      "loss": 1.4892,
       "step": 1385
     },
     {
       "epoch": 1.39,
-      "learning_rate": 0.0036548857868020303,
-      "loss": 1.4402,
+      "learning_rate": 0.0037414020785219403,
+      "loss": 1.4041,
       "step": 1390
     },
     {
       "epoch": 1.39,
-      "learning_rate": 0.0036266852791878175,
-      "loss": 1.8302,
+      "learning_rate": 0.003712534006928406,
+      "loss": 1.6152,
       "step": 1395
     },
     {
       "epoch": 1.4,
-      "learning_rate": 0.003598484771573604,
-      "loss": 1.1884,
+      "learning_rate": 0.003683665935334873,
+      "loss": 1.3959,
       "step": 1400
     },
     {
       "epoch": 1.4,
-      "learning_rate": 0.003570284263959391,
-      "loss": 1.6533,
+      "learning_rate": 0.00365479786374134,
+      "loss": 1.2466,
       "step": 1405
     },
     {
       "epoch": 1.41,
-      "learning_rate": 0.0035420837563451785,
-      "loss": 1.449,
+      "learning_rate": 0.003625929792147806,
+      "loss": 1.4626,
       "step": 1410
     },
     {
       "epoch": 1.41,
-      "learning_rate": 0.003513883248730964,
-      "loss": 1.4537,
+      "learning_rate": 0.003597061720554272,
+      "loss": 1.1847,
       "step": 1415
     },
     {
       "epoch": 1.42,
-      "learning_rate": 0.0034856827411167512,
-      "loss": 1.4386,
+      "learning_rate": 0.003568193648960739,
+      "loss": 1.3083,
       "step": 1420
     },
     {
       "epoch": 1.42,
-      "learning_rate": 0.003457482233502538,
-      "loss": 1.3043,
+      "learning_rate": 0.0035393255773672055,
+      "loss": 1.3253,
       "step": 1425
     },
     {
       "epoch": 1.43,
-      "learning_rate": 0.0034292817258883244,
-      "loss": 1.5577,
+      "learning_rate": 0.0035104575057736728,
+      "loss": 1.372,
       "step": 1430
     },
     {
       "epoch": 1.43,
-      "learning_rate": 0.0034010812182741117,
-      "loss": 1.6184,
+      "learning_rate": 0.0034815894341801387,
+      "loss": 1.4528,
       "step": 1435
     },
     {
       "epoch": 1.44,
-      "learning_rate": 0.003372880710659899,
-      "loss": 1.2835,
+      "learning_rate": 0.003452721362586605,
+      "loss": 1.5115,
       "step": 1440
     },
     {
       "epoch": 1.44,
-      "learning_rate": 0.0033446802030456858,
-      "loss": 1.4659,
+      "learning_rate": 0.003423853290993072,
+      "loss": 1.3855,
       "step": 1445
     },
     {
       "epoch": 1.45,
-      "learning_rate": 0.0033164796954314717,
-      "loss": 1.1781,
+      "learning_rate": 0.0033949852193995377,
+      "loss": 1.2547,
       "step": 1450
     },
     {
       "epoch": 1.45,
-      "learning_rate": 0.0032882791878172586,
-      "loss": 1.4166,
+      "learning_rate": 0.0033661171478060045,
+      "loss": 1.4584,
       "step": 1455
     },
     {
       "epoch": 1.46,
-      "learning_rate": 0.003260078680203046,
-      "loss": 1.3821,
+      "learning_rate": 0.0033372490762124713,
+      "loss": 1.2172,
       "step": 1460
     },
     {
       "epoch": 1.46,
-      "learning_rate": 0.0032318781725888326,
-      "loss": 1.417,
+      "learning_rate": 0.0033083810046189376,
+      "loss": 1.3224,
       "step": 1465
     },
     {
       "epoch": 1.47,
-      "learning_rate": 0.0032036776649746195,
-      "loss": 1.0738,
+      "learning_rate": 0.003279512933025405,
+      "loss": 1.5014,
       "step": 1470
     },
     {
       "epoch": 1.47,
-      "learning_rate": 0.0031754771573604067,
-      "loss": 1.2626,
+      "learning_rate": 0.0032506448614318707,
+      "loss": 1.4566,
       "step": 1475
     },
     {
       "epoch": 1.48,
-      "learning_rate": 0.0031472766497461935,
-      "loss": 1.4132,
+      "learning_rate": 0.0032217767898383366,
+      "loss": 1.5488,
       "step": 1480
     },
     {
       "epoch": 1.48,
-      "learning_rate": 0.0031190761421319795,
-      "loss": 1.3976,
+      "learning_rate": 0.0031929087182448043,
+      "loss": 1.3555,
       "step": 1485
     },
     {
       "epoch": 1.49,
-      "learning_rate": 0.0030908756345177663,
-      "loss": 1.1655,
+      "learning_rate": 0.00316404064665127,
+      "loss": 1.488,
       "step": 1490
     },
     {
       "epoch": 1.49,
-      "learning_rate": 0.003062675126903553,
-      "loss": 1.6067,
+      "learning_rate": 0.0031351725750577374,
+      "loss": 1.3927,
       "step": 1495
     },
     {
       "epoch": 1.5,
-      "learning_rate": 0.0030344746192893404,
-      "loss": 1.2399,
+      "learning_rate": 0.0031063045034642033,
+      "loss": 1.1759,
       "step": 1500
     },
     {
       "epoch": 1.5,
-      "learning_rate": 0.0030062741116751272,
-      "loss": 1.3499,
+      "learning_rate": 0.00307743643187067,
+      "loss": 1.4156,
       "step": 1505
     },
     {
       "epoch": 1.51,
-      "learning_rate": 0.002978073604060914,
-      "loss": 1.2958,
+      "learning_rate": 0.0030485683602771364,
+      "loss": 1.3284,
       "step": 1510
     },
     {
       "epoch": 1.51,
-      "learning_rate": 0.0029498730964467,
-      "loss": 0.9839,
+      "learning_rate": 0.003019700288683603,
+      "loss": 1.5301,
       "step": 1515
     },
     {
       "epoch": 1.52,
-      "learning_rate": 0.002921672588832487,
-      "loss": 1.423,
+      "learning_rate": 0.0029908322170900687,
+      "loss": 1.4688,
       "step": 1520
     },
     {
       "epoch": 1.52,
-      "learning_rate": 0.002893472081218274,
-      "loss": 1.4083,
+      "learning_rate": 0.002961964145496536,
+      "loss": 1.3861,
       "step": 1525
     },
     {
       "epoch": 1.53,
-      "learning_rate": 0.002865271573604061,
-      "loss": 1.552,
+      "learning_rate": 0.0029330960739030023,
+      "loss": 1.3078,
       "step": 1530
     },
     {
       "epoch": 1.53,
-      "learning_rate": 0.002837071065989848,
-      "loss": 1.1468,
+      "learning_rate": 0.00290422800230947,
+      "loss": 1.5033,
       "step": 1535
     },
     {
       "epoch": 1.54,
-      "learning_rate": 0.002808870558375635,
-      "loss": 1.1524,
+      "learning_rate": 0.0028753599307159354,
+      "loss": 1.3179,
       "step": 1540
     },
     {
       "epoch": 1.54,
-      "learning_rate": 0.002780670050761422,
-      "loss": 1.7909,
+      "learning_rate": 0.0028464918591224013,
+      "loss": 1.3151,
       "step": 1545
     },
     {
       "epoch": 1.55,
-      "learning_rate": 0.0027524695431472078,
-      "loss": 1.2374,
+      "learning_rate": 0.0028176237875288685,
+      "loss": 1.2228,
       "step": 1550
     },
     {
       "epoch": 1.55,
-      "learning_rate": 0.0027242690355329946,
-      "loss": 1.5076,
+      "learning_rate": 0.002788755715935335,
+      "loss": 1.4193,
       "step": 1555
     },
     {
       "epoch": 1.56,
-      "learning_rate": 0.0026960685279187814,
-      "loss": 1.2157,
+      "learning_rate": 0.002759887644341802,
+      "loss": 1.5102,
       "step": 1560
     },
     {
       "epoch": 1.56,
-      "learning_rate": 0.0026678680203045687,
-      "loss": 1.4255,
+      "learning_rate": 0.002731019572748268,
+      "loss": 1.4108,
       "step": 1565
     },
     {
       "epoch": 1.57,
-      "learning_rate": 0.002639667512690356,
-      "loss": 1.1614,
+      "learning_rate": 0.0027021515011547343,
+      "loss": 1.6702,
       "step": 1570
     },
     {
       "epoch": 1.57,
-      "learning_rate": 0.0026114670050761423,
-      "loss": 1.2836,
+      "learning_rate": 0.002673283429561201,
+      "loss": 1.4893,
       "step": 1575
     },
     {
       "epoch": 1.58,
-      "learning_rate": 0.0025832664974619296,
-      "loss": 0.9739,
+      "learning_rate": 0.002644415357967667,
+      "loss": 1.5393,
       "step": 1580
     },
     {
       "epoch": 1.58,
-      "learning_rate": 0.002555065989847715,
-      "loss": 1.445,
+      "learning_rate": 0.002615547286374133,
+      "loss": 1.3174,
       "step": 1585
     },
     {
       "epoch": 1.59,
-      "learning_rate": 0.0025268654822335024,
-      "loss": 1.3724,
+      "learning_rate": 0.0025866792147806006,
+      "loss": 1.5737,
       "step": 1590
     },
     {
       "epoch": 1.59,
-      "learning_rate": 0.002498664974619289,
-      "loss": 1.1985,
+      "learning_rate": 0.002557811143187067,
+      "loss": 1.3413,
       "step": 1595
     },
     {
       "epoch": 1.6,
-      "learning_rate": 0.002470464467005076,
-      "loss": 1.5102,
+      "learning_rate": 0.002528943071593534,
+      "loss": 1.4321,
       "step": 1600
     },
     {
       "epoch": 1.6,
-      "learning_rate": 0.0024422639593908633,
-      "loss": 1.4111,
+      "learning_rate": 0.002500075,
+      "loss": 1.5292,
       "step": 1605
     },
     {
       "epoch": 1.61,
-      "learning_rate": 0.00241406345177665,
-      "loss": 1.4832,
+      "learning_rate": 0.002471206928406466,
+      "loss": 1.1747,
       "step": 1610
     },
     {
       "epoch": 1.61,
-      "learning_rate": 0.002385862944162437,
-      "loss": 1.7259,
+      "learning_rate": 0.002442338856812933,
+      "loss": 1.3613,
       "step": 1615
     },
     {
       "epoch": 1.62,
-      "learning_rate": 0.002357662436548223,
-      "loss": 1.4051,
+      "learning_rate": 0.0024134707852193995,
+      "loss": 1.4922,
       "step": 1620
     },
     {
       "epoch": 1.62,
-      "learning_rate": 0.0023294619289340097,
-      "loss": 1.1585,
+      "learning_rate": 0.0023846027136258667,
+      "loss": 1.4311,
       "step": 1625
     },
     {
       "epoch": 1.63,
-      "learning_rate": 0.002301261421319797,
-      "loss": 1.4892,
+      "learning_rate": 0.0023557346420323326,
+      "loss": 1.4672,
       "step": 1630
     },
     {
       "epoch": 1.63,
-      "learning_rate": 0.0022730609137055838,
-      "loss": 1.2785,
+      "learning_rate": 0.002326866570438799,
+      "loss": 1.3726,
       "step": 1635
     },
     {
       "epoch": 1.64,
-      "learning_rate": 0.0022448604060913706,
-      "loss": 1.4697,
+      "learning_rate": 0.0022979984988452658,
+      "loss": 1.2646,
       "step": 1640
     },
     {
       "epoch": 1.64,
-      "learning_rate": 0.002216659898477158,
-      "loss": 1.5335,
+      "learning_rate": 0.002269130427251732,
+      "loss": 1.2819,
       "step": 1645
     },
     {
       "epoch": 1.65,
-      "learning_rate": 0.0021884593908629447,
-      "loss": 1.6132,
+      "learning_rate": 0.002240262355658198,
+      "loss": 1.3917,
       "step": 1650
     },
     {
       "epoch": 1.65,
-      "learning_rate": 0.0021602588832487306,
-      "loss": 1.2333,
+      "learning_rate": 0.0022113942840646652,
+      "loss": 1.3339,
       "step": 1655
     },
     {
       "epoch": 1.66,
-      "learning_rate": 0.0021320583756345175,
-      "loss": 1.7243,
+      "learning_rate": 0.0021825262124711316,
+      "loss": 1.439,
       "step": 1660
     },
     {
       "epoch": 1.66,
-      "learning_rate": 0.0021038578680203043,
-      "loss": 1.5654,
+      "learning_rate": 0.002153658140877599,
+      "loss": 1.3474,
       "step": 1665
     },
     {
       "epoch": 1.67,
-      "learning_rate": 0.0020756573604060915,
-      "loss": 1.3233,
+      "learning_rate": 0.0021247900692840647,
+      "loss": 1.1514,
       "step": 1670
     },
     {
       "epoch": 1.67,
-      "learning_rate": 0.0020474568527918784,
-      "loss": 1.2538,
+      "learning_rate": 0.0020959219976905306,
+      "loss": 1.3472,
       "step": 1675
     },
     {
       "epoch": 1.68,
-      "learning_rate": 0.002019256345177665,
-      "loss": 1.5129,
+      "learning_rate": 0.002067053926096998,
+      "loss": 1.3059,
       "step": 1680
     },
     {
       "epoch": 1.68,
-      "learning_rate": 0.0019910558375634524,
-      "loss": 1.2342,
+      "learning_rate": 0.002038185854503464,
+      "loss": 1.4306,
       "step": 1685
     },
     {
       "epoch": 1.69,
-      "learning_rate": 0.001962855329949238,
-      "loss": 1.2484,
+      "learning_rate": 0.0020093177829099314,
+      "loss": 1.4762,
       "step": 1690
     },
     {
       "epoch": 1.69,
-      "learning_rate": 0.0019346548223350254,
-      "loss": 1.6045,
+      "learning_rate": 0.0019804497113163973,
+      "loss": 1.2309,
       "step": 1695
     },
     {
       "epoch": 1.7,
-      "learning_rate": 0.0019064543147208123,
-      "loss": 1.2544,
+      "learning_rate": 0.0019515816397228636,
+      "loss": 1.4477,
       "step": 1700
     },
     {
       "epoch": 1.7,
-      "learning_rate": 0.0018782538071065993,
-      "loss": 1.1989,
+      "learning_rate": 0.0019227135681293308,
+      "loss": 1.3584,
       "step": 1705
     },
     {
       "epoch": 1.71,
-      "learning_rate": 0.0018500532994923861,
-      "loss": 1.3844,
+      "learning_rate": 0.001893845496535797,
+      "loss": 1.2721,
       "step": 1710
     },
     {
       "epoch": 1.71,
-      "learning_rate": 0.0018218527918781732,
-      "loss": 1.2898,
+      "learning_rate": 0.0018649774249422629,
+      "loss": 1.3599,
       "step": 1715
     },
     {
       "epoch": 1.72,
-      "learning_rate": 0.0017936522842639591,
-      "loss": 1.176,
+      "learning_rate": 0.0018361093533487299,
+      "loss": 1.3039,
       "step": 1720
     },
     {
       "epoch": 1.72,
-      "learning_rate": 0.001765451776649746,
-      "loss": 1.3886,
+      "learning_rate": 0.0018072412817551962,
+      "loss": 1.4414,
       "step": 1725
     },
     {
       "epoch": 1.73,
-      "learning_rate": 0.001737251269035533,
-      "loss": 1.4818,
+      "learning_rate": 0.0017783732101616634,
+      "loss": 1.3903,
       "step": 1730
     },
     {
       "epoch": 1.73,
-      "learning_rate": 0.00170905076142132,
-      "loss": 1.4334,
+      "learning_rate": 0.0017495051385681296,
+      "loss": 1.4272,
       "step": 1735
     },
     {
       "epoch": 1.74,
-      "learning_rate": 0.0016808502538071069,
-      "loss": 1.3581,
+      "learning_rate": 0.0017206370669745957,
+      "loss": 1.3649,
       "step": 1740
     },
     {
       "epoch": 1.74,
-      "learning_rate": 0.0016526497461928939,
-      "loss": 1.287,
+      "learning_rate": 0.0016917689953810627,
+      "loss": 1.2328,
       "step": 1745
     },
     {
       "epoch": 1.75,
-      "learning_rate": 0.0016244492385786807,
-      "loss": 1.4921,
+      "learning_rate": 0.001662900923787529,
+      "loss": 1.2621,
       "step": 1750
     },
     {
       "epoch": 1.75,
-      "learning_rate": 0.0015962487309644667,
-      "loss": 1.6024,
+      "learning_rate": 0.001634032852193996,
+      "loss": 1.2631,
       "step": 1755
     },
     {
       "epoch": 1.76,
-      "learning_rate": 0.001568048223350254,
-      "loss": 1.3267,
+      "learning_rate": 0.0016051647806004622,
+      "loss": 1.3229,
       "step": 1760
     },
     {
       "epoch": 1.76,
-      "learning_rate": 0.0015398477157360405,
-      "loss": 1.3324,
+      "learning_rate": 0.0015762967090069283,
+      "loss": 1.4429,
       "step": 1765
     },
     {
       "epoch": 1.77,
-      "learning_rate": 0.0015116472081218278,
-      "loss": 1.4282,
+      "learning_rate": 0.0015474286374133955,
+      "loss": 1.3552,
       "step": 1770
     },
     {
       "epoch": 1.77,
-      "learning_rate": 0.0014834467005076146,
-      "loss": 1.4165,
+      "learning_rate": 0.0015185605658198618,
+      "loss": 1.3257,
       "step": 1775
     },
     {
       "epoch": 1.78,
-      "learning_rate": 0.0014552461928934012,
-      "loss": 1.5209,
+      "learning_rate": 0.0014896924942263275,
+      "loss": 1.1749,
       "step": 1780
     },
     {
       "epoch": 1.78,
-      "learning_rate": 0.0014270456852791885,
-      "loss": 1.5915,
+      "learning_rate": 0.0014608244226327945,
+      "loss": 1.2705,
       "step": 1785
     },
     {
       "epoch": 1.79,
-      "learning_rate": 0.001398845177664974,
-      "loss": 1.5822,
+      "learning_rate": 0.0014319563510392609,
+      "loss": 1.4362,
       "step": 1790
     },
     {
       "epoch": 1.79,
-      "learning_rate": 0.0013706446700507613,
-      "loss": 1.4356,
+      "learning_rate": 0.001403088279445728,
+      "loss": 1.1888,
       "step": 1795
     },
     {
       "epoch": 1.8,
-      "learning_rate": 0.0013424441624365483,
-      "loss": 1.3999,
+      "learning_rate": 0.0013742202078521942,
+      "loss": 1.3183,
       "step": 1800
     },
     {
       "epoch": 1.8,
-      "learning_rate": 0.0013142436548223351,
-      "loss": 1.3192,
+      "learning_rate": 0.0013453521362586604,
+      "loss": 1.4149,
       "step": 1805
     },
     {
       "epoch": 1.81,
-      "learning_rate": 0.0012860431472081222,
-      "loss": 0.9928,
+      "learning_rate": 0.0013164840646651273,
+      "loss": 1.2567,
       "step": 1810
     },
     {
       "epoch": 1.81,
-      "learning_rate": 0.0012578426395939092,
-      "loss": 1.7191,
+      "learning_rate": 0.0012876159930715937,
+      "loss": 1.251,
       "step": 1815
     },
     {
       "epoch": 1.82,
-      "learning_rate": 0.001229642131979696,
-      "loss": 1.2173,
+      "learning_rate": 0.0012587479214780596,
+      "loss": 1.1756,
       "step": 1820
     },
     {
       "epoch": 1.82,
-      "learning_rate": 0.001201441624365482,
-      "loss": 1.1303,
+      "learning_rate": 0.0012298798498845268,
+      "loss": 1.4292,
       "step": 1825
     },
     {
       "epoch": 1.83,
-      "learning_rate": 0.0011732411167512688,
-      "loss": 1.649,
+      "learning_rate": 0.001201011778290993,
+      "loss": 1.354,
       "step": 1830
     },
     {
       "epoch": 1.83,
-      "learning_rate": 0.0011450406091370558,
-      "loss": 1.5897,
+      "learning_rate": 0.0011721437066974602,
+      "loss": 1.2577,
       "step": 1835
     },
     {
       "epoch": 1.84,
-      "learning_rate": 0.0011168401015228429,
-      "loss": 1.163,
+      "learning_rate": 0.0011432756351039263,
+      "loss": 1.3255,
       "step": 1840
     },
     {
       "epoch": 1.84,
-      "learning_rate": 0.0010886395939086297,
-      "loss": 1.1745,
+      "learning_rate": 0.0011144075635103924,
+      "loss": 1.1484,
       "step": 1845
     },
     {
       "epoch": 1.85,
-      "learning_rate": 0.0010604390862944168,
-      "loss": 1.2499,
+      "learning_rate": 0.0010855394919168594,
+      "loss": 1.3225,
       "step": 1850
     },
     {
       "epoch": 1.85,
-      "learning_rate": 0.0010322385786802038,
-      "loss": 1.1123,
+      "learning_rate": 0.0010566714203233255,
+      "loss": 1.4918,
       "step": 1855
     },
     {
       "epoch": 1.86,
-      "learning_rate": 0.0010040380710659895,
-      "loss": 1.6267,
+      "learning_rate": 0.0010278033487297927,
+      "loss": 1.2976,
       "step": 1860
     },
     {
       "epoch": 1.86,
-      "learning_rate": 0.0009758375634517764,
-      "loss": 1.7138,
+      "learning_rate": 0.0009989352771362589,
+      "loss": 1.2532,
       "step": 1865
     },
     {
       "epoch": 1.87,
-      "learning_rate": 0.0009476370558375634,
-      "loss": 1.0705,
+      "learning_rate": 0.000970067205542725,
+      "loss": 1.3491,
       "step": 1870
     },
     {
       "epoch": 1.87,
-      "learning_rate": 0.0009194365482233503,
-      "loss": 1.5925,
+      "learning_rate": 0.000941199133949192,
+      "loss": 1.3206,
       "step": 1875
     },
     {
       "epoch": 1.88,
-      "learning_rate": 0.0008912360406091373,
-      "loss": 1.2733,
+      "learning_rate": 0.0009123310623556581,
+      "loss": 1.4938,
       "step": 1880
     },
     {
       "epoch": 1.88,
-      "learning_rate": 0.0008630355329949242,
-      "loss": 1.5511,
+      "learning_rate": 0.0008834629907621243,
+      "loss": 1.424,
       "step": 1885
     },
     {
       "epoch": 1.89,
-      "learning_rate": 0.0008348350253807112,
-      "loss": 1.3351,
+      "learning_rate": 0.0008545949191685913,
+      "loss": 1.3984,
       "step": 1890
     },
     {
       "epoch": 1.89,
-      "learning_rate": 0.000806634517766497,
-      "loss": 1.1094,
+      "learning_rate": 0.0008257268475750575,
+      "loss": 1.1661,
       "step": 1895
     },
     {
       "epoch": 1.9,
-      "learning_rate": 0.000778434010152284,
-      "loss": 1.1718,
+      "learning_rate": 0.0007968587759815248,
+      "loss": 1.4248,
       "step": 1900
     },
     {
       "epoch": 1.9,
-      "learning_rate": 0.0007502335025380709,
-      "loss": 1.2751,
+      "learning_rate": 0.0007679907043879908,
+      "loss": 1.357,
       "step": 1905
     },
     {
       "epoch": 1.91,
-      "learning_rate": 0.000722032994923858,
-      "loss": 0.9808,
+      "learning_rate": 0.0007391226327944569,
+      "loss": 1.269,
       "step": 1910
     },
     {
       "epoch": 1.91,
-      "learning_rate": 0.0006938324873096449,
-      "loss": 1.5196,
+      "learning_rate": 0.0007102545612009241,
+      "loss": 1.3927,
       "step": 1915
     },
     {
       "epoch": 1.92,
-      "learning_rate": 0.0006656319796954318,
-      "loss": 1.1567,
+      "learning_rate": 0.0006813864896073901,
+      "loss": 1.2803,
       "step": 1920
     },
     {
       "epoch": 1.92,
-      "learning_rate": 0.0006374314720812188,
-      "loss": 1.2295,
+      "learning_rate": 0.0006525184180138573,
+      "loss": 1.5049,
       "step": 1925
     },
     {
       "epoch": 1.93,
-      "learning_rate": 0.0006092309644670046,
-      "loss": 1.3395,
+      "learning_rate": 0.0006236503464203234,
+      "loss": 1.3776,
       "step": 1930
     },
     {
       "epoch": 1.93,
-      "learning_rate": 0.0005810304568527917,
-      "loss": 1.2176,
+      "learning_rate": 0.0005947822748267896,
+      "loss": 1.3658,
       "step": 1935
     },
     {
       "epoch": 1.94,
-      "learning_rate": 0.0005528299492385786,
-      "loss": 1.2218,
+      "learning_rate": 0.0005659142032332567,
+      "loss": 1.0963,
       "step": 1940
     },
     {
       "epoch": 1.94,
-      "learning_rate": 0.0005246294416243655,
-      "loss": 1.1824,
+      "learning_rate": 0.0005370461316397228,
+      "loss": 1.4584,
       "step": 1945
     },
     {
       "epoch": 1.95,
-      "learning_rate": 0.0004964289340101526,
-      "loss": 1.4026,
+      "learning_rate": 0.0005081780600461889,
+      "loss": 1.2742,
       "step": 1950
     },
     {
       "epoch": 1.95,
-      "learning_rate": 0.0004682284263959396,
-      "loss": 1.391,
+      "learning_rate": 0.0004793099884526561,
+      "loss": 1.1752,
       "step": 1955
     },
     {
       "epoch": 1.96,
-      "learning_rate": 0.0004400279187817254,
-      "loss": 1.5254,
+      "learning_rate": 0.0004504419168591222,
+      "loss": 1.1821,
       "step": 1960
     },
     {
       "epoch": 1.96,
-      "learning_rate": 0.0004118274111675124,
-      "loss": 1.2331,
+      "learning_rate": 0.0004215738452655894,
+      "loss": 1.3403,
       "step": 1965
     },
     {
       "epoch": 1.97,
-      "learning_rate": 0.0003836269035532993,
-      "loss": 1.2529,
+      "learning_rate": 0.00039270577367205543,
+      "loss": 1.449,
       "step": 1970
     },
     {
       "epoch": 1.97,
-      "learning_rate": 0.00035542639593908625,
-      "loss": 1.0726,
+      "learning_rate": 0.00036383770207852156,
+      "loss": 1.1671,
       "step": 1975
     },
     {
       "epoch": 1.98,
-      "learning_rate": 0.00032722588832487324,
-      "loss": 1.2203,
+      "learning_rate": 0.00033496963048498877,
+      "loss": 1.4356,
       "step": 1980
     },
     {
       "epoch": 1.98,
-      "learning_rate": 0.0002990253807106602,
-      "loss": 1.2456,
+      "learning_rate": 0.00030610155889145485,
+      "loss": 1.2352,
       "step": 1985
     },
     {
       "epoch": 1.99,
-      "learning_rate": 0.00027082487309644716,
-      "loss": 1.3003,
+      "learning_rate": 0.00027723348729792206,
+      "loss": 1.3243,
       "step": 1990
     },
     {
       "epoch": 1.99,
-      "learning_rate": 0.00024262436548223298,
-      "loss": 1.299,
+      "learning_rate": 0.00024836541570438813,
+      "loss": 1.2791,
       "step": 1995
     },
     {
       "epoch": 2.0,
-      "learning_rate": 0.00021442385786802,
-      "loss": 1.488,
+      "learning_rate": 0.0002194973441108542,
+      "loss": 1.2112,
       "step": 2000
     },
     {
       "epoch": 2.0,
-      "learning_rate": 0.00018622335025380695,
-      "loss": 937.4512,
+      "learning_rate": 0.0001906292725173214,
+      "loss": 1.1661,
       "step": 2005
     },
     {
       "epoch": 2.01,
-      "learning_rate": 0.00015802284263959394,
-      "loss": 924.4408,
+      "learning_rate": 0.0001617612009237875,
+      "loss": 1.4046,
       "step": 2010
     },
     {
       "epoch": 2.01,
-      "learning_rate": 0.00012982233502538087,
-      "loss": 903.1811,
+      "learning_rate": 0.00013289312933025357,
+      "loss": 1.2514,
       "step": 2015
     },
     {
       "epoch": 2.02,
-      "learning_rate": 0.00010162182741116783,
-      "loss": 868.8327,
+      "learning_rate": 0.00010402505773672078,
+      "loss": 1.1908,
       "step": 2020
     },
     {
       "epoch": 2.02,
-      "learning_rate": 7.342131979695479e-05,
-      "loss": 833.6856,
+      "learning_rate": 7.515698614318685e-05,
+      "loss": 1.3698,
       "step": 2025
     },
     {
       "epoch": 2.03,
-      "learning_rate": 4.522081218274065e-05,
-      "loss": 816.0007,
+      "learning_rate": 4.628891454965405e-05,
+      "loss": 1.5172,
       "step": 2030
     },
     {
       "epoch": 2.03,
-      "learning_rate": 1.70203045685276e-05,
-      "loss": 795.98,
+      "learning_rate": 1.7420842956120128e-05,
+      "loss": 1.3045,
       "step": 2035
     },
     {
       "epoch": 2.04,
       "step": 2038,
-      "total_flos": 2.2810681001063887e+23,
-      "train_loss": 16.08543400488377,
-      "train_runtime": 323733.6046,
-      "train_samples_per_second": 103.142,
-      "train_steps_per_second": 0.006
-    },
-    {
-      "epoch": 2.04,
-      "step": 2038,
-      "total_flos": 2.2810681001063887e+23,
-      "train_loss": 0.0,
-      "train_runtime": 3248433.7519,
-      "train_samples_per_second": 0.005,
-      "train_steps_per_second": 0.0
+      "total_flos": 1.1512671589937643e+23,
+      "train_loss": 1.4677173125030247,
+      "train_runtime": 22959.9641,
+      "train_samples_per_second": 1454.296,
+      "train_steps_per_second": 0.089
     }
   ],
-  "max_steps": 1,
-  "num_train_epochs": 1,
-  "start_time": 1663027920.2259,
-  "total_flos": 2.2810681001063887e+23,
+  "max_steps": 2038,
+  "num_train_epochs": 3,
+  "start_time": 1669316150.4760864,
+  "total_flos": 1.1512671589937643e+23,
   "trial_name": null,
   "trial_params": null
 }