diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6319 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9926262053318208,
+  "eval_steps": 500,
+  "global_step": 10500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0002997163925127623,
+      "loss": 2.2978,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00029943278502552465,
+      "loss": 2.2765,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.000299149177538287,
+      "loss": 2.2289,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0002988655700510493,
+      "loss": 2.3721,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00029858196256381166,
+      "loss": 2.135,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.000298298355076574,
+      "loss": 2.2899,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00029801474758933634,
+      "loss": 2.1047,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0002977311401020987,
+      "loss": 2.1712,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.000297447532614861,
+      "loss": 2.1048,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00029716392512762335,
+      "loss": 2.1801,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0002968803176403857,
+      "loss": 2.1241,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00029659671015314803,
+      "loss": 2.1783,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00029631310266591037,
+      "loss": 2.2588,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0002960294951786727,
+      "loss": 2.1962,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00029574588769143504,
+      "loss": 1.9953,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002954622802041974,
+      "loss": 2.15,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002951786727169597,
+      "loss": 2.1306,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000294895065229722,
+      "loss": 2.2464,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002946114577424844,
+      "loss": 2.2083,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00029432785025524673,
+      "loss": 2.0959,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00029404424276800907,
+      "loss": 2.2083,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002937606352807714,
+      "loss": 2.2126,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002934770277935337,
+      "loss": 2.1264,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00029319342030629603,
+      "loss": 2.1913,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002929098128190584,
+      "loss": 2.1374,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00029262620533182076,
+      "loss": 2.1884,
+      "step": 260
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00029234259784458304,
+      "loss": 2.0839,
+      "step": 270
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002920873511060692,
+      "loss": 1.9849,
+      "step": 280
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00029180374361883153,
+      "loss": 2.0729,
+      "step": 290
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002915201361315938,
+      "loss": 1.925,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00029123652864435616,
+      "loss": 2.0478,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00029095292115711855,
+      "loss": 2.1546,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002906693136698809,
+      "loss": 2.1341,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00029038570618264317,
+      "loss": 2.1653,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002901020986954055,
+      "loss": 1.9772,
+      "step": 350
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00028981849120816785,
+      "loss": 2.0511,
+      "step": 360
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00028953488372093024,
+      "loss": 2.0662,
+      "step": 370
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002892512762336926,
+      "loss": 2.1331,
+      "step": 380
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00028896766874645486,
+      "loss": 2.1118,
+      "step": 390
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002886840612592172,
+      "loss": 1.9541,
+      "step": 400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00028840045377197954,
+      "loss": 2.1632,
+      "step": 410
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002881168462847419,
+      "loss": 2.1462,
+      "step": 420
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00028783323879750427,
+      "loss": 2.1207,
+      "step": 430
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00028754963131026655,
+      "loss": 2.058,
+      "step": 440
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002872660238230289,
+      "loss": 2.0473,
+      "step": 450
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002869824163357912,
+      "loss": 2.1528,
+      "step": 460
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00028669880884855356,
+      "loss": 2.0529,
+      "step": 470
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002864152013613159,
+      "loss": 2.1924,
+      "step": 480
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00028613159387407824,
+      "loss": 2.0982,
+      "step": 490
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002858479863868406,
+      "loss": 2.0854,
+      "step": 500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002855643788996029,
+      "loss": 2.179,
+      "step": 510
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00028528077141236525,
+      "loss": 1.9784,
+      "step": 520
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002849971639251276,
+      "loss": 2.1489,
+      "step": 530
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00028471355643788993,
+      "loss": 2.1446,
+      "step": 540
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00028442994895065227,
+      "loss": 2.0817,
+      "step": 550
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002841463414634146,
+      "loss": 2.1402,
+      "step": 560
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00028386273397617694,
+      "loss": 2.027,
+      "step": 570
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002835791264889393,
+      "loss": 2.098,
+      "step": 580
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002832955190017016,
+      "loss": 2.0974,
+      "step": 590
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00028301191151446396,
+      "loss": 1.8892,
+      "step": 600
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002827283040272263,
+      "loss": 1.9368,
+      "step": 610
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00028244469653998863,
+      "loss": 2.1184,
+      "step": 620
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00028216108905275097,
+      "loss": 2.1468,
+      "step": 630
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002818774815655133,
+      "loss": 2.0261,
+      "step": 640
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00028159387407827565,
+      "loss": 2.2235,
+      "step": 650
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.000281310266591038,
+      "loss": 2.1657,
+      "step": 660
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002810266591038003,
+      "loss": 2.0709,
+      "step": 670
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00028074305161656266,
+      "loss": 2.1451,
+      "step": 680
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.000280459444129325,
+      "loss": 2.0991,
+      "step": 690
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00028017583664208734,
+      "loss": 1.983,
+      "step": 700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002798922291548497,
+      "loss": 2.0995,
+      "step": 710
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.000279608621667612,
+      "loss": 1.9696,
+      "step": 720
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00027932501418037435,
+      "loss": 2.1549,
+      "step": 730
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00027904140669313663,
+      "loss": 1.9938,
+      "step": 740
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.000278757799205899,
+      "loss": 2.1518,
+      "step": 750
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00027847419171866136,
+      "loss": 2.0179,
+      "step": 760
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002781905842314237,
+      "loss": 2.0337,
+      "step": 770
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00027790697674418604,
+      "loss": 2.1366,
+      "step": 780
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002776233692569483,
+      "loss": 2.0452,
+      "step": 790
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002773397617697107,
+      "loss": 2.177,
+      "step": 800
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00027705615428247305,
+      "loss": 2.0401,
+      "step": 810
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002767725467952354,
+      "loss": 2.06,
+      "step": 820
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00027648893930799773,
+      "loss": 2.1167,
+      "step": 830
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00027620533182076,
+      "loss": 2.0292,
+      "step": 840
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00027592172433352235,
+      "loss": 2.0566,
+      "step": 850
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00027563811684628474,
+      "loss": 2.054,
+      "step": 860
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002753545093590471,
+      "loss": 2.0844,
+      "step": 870
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00027507090187180936,
+      "loss": 2.1432,
+      "step": 880
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002747872943845717,
+      "loss": 2.0712,
+      "step": 890
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00027450368689733404,
+      "loss": 2.1214,
+      "step": 900
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00027422007941009643,
+      "loss": 1.9663,
+      "step": 910
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00027393647192285877,
+      "loss": 2.0059,
+      "step": 920
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00027368122518434487,
+      "loss": 2.0752,
+      "step": 930
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002733976176971072,
+      "loss": 2.0892,
+      "step": 940
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002731140102098695,
+      "loss": 2.0141,
+      "step": 950
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00027283040272263183,
+      "loss": 2.0029,
+      "step": 960
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00027254679523539417,
+      "loss": 2.0242,
+      "step": 970
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00027226318774815656,
+      "loss": 2.1536,
+      "step": 980
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002719795802609189,
+      "loss": 1.9372,
+      "step": 990
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002716959727736812,
+      "loss": 2.0371,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002714123652864435,
+      "loss": 2.1432,
+      "step": 1010
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00027112875779920586,
+      "loss": 2.1021,
+      "step": 1020
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00027084515031196825,
+      "loss": 2.0531,
+      "step": 1030
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002705615428247306,
+      "loss": 2.0348,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00027027793533749287,
+      "loss": 1.9867,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002699943278502552,
+      "loss": 2.1114,
+      "step": 1060
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00026971072036301755,
+      "loss": 2.0896,
+      "step": 1070
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002694271128757799,
+      "loss": 2.1385,
+      "step": 1080
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002691435053885422,
+      "loss": 2.0062,
+      "step": 1090
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00026885989790130456,
+      "loss": 2.0613,
+      "step": 1100
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002685762904140669,
+      "loss": 2.1112,
+      "step": 1110
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00026829268292682924,
+      "loss": 2.0592,
+      "step": 1120
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002680090754395916,
+      "loss": 2.0427,
+      "step": 1130
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002677254679523539,
+      "loss": 2.1737,
+      "step": 1140
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00026744186046511625,
+      "loss": 2.1267,
+      "step": 1150
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002671582529778786,
+      "loss": 2.1244,
+      "step": 1160
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002668746454906409,
+      "loss": 2.1278,
+      "step": 1170
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00026659103800340326,
+      "loss": 2.0743,
+      "step": 1180
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002663074305161656,
+      "loss": 2.1632,
+      "step": 1190
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00026602382302892794,
+      "loss": 2.0233,
+      "step": 1200
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002657402155416903,
+      "loss": 2.0631,
+      "step": 1210
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002654566080544526,
+      "loss": 2.2041,
+      "step": 1220
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00026517300056721495,
+      "loss": 2.095,
+      "step": 1230
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002648893930799773,
+      "loss": 1.9956,
+      "step": 1240
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00026460578559273963,
+      "loss": 1.9052,
+      "step": 1250
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00026432217810550197,
+      "loss": 1.8921,
+      "step": 1260
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002640385706182643,
+      "loss": 2.171,
+      "step": 1270
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00026375496313102664,
+      "loss": 1.9069,
+      "step": 1280
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.000263471355643789,
+      "loss": 1.9907,
+      "step": 1290
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002631877481565513,
+      "loss": 2.1011,
+      "step": 1300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00026290414066931366,
+      "loss": 2.1417,
+      "step": 1310
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.000262620533182076,
+      "loss": 2.039,
+      "step": 1320
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00026233692569483833,
+      "loss": 2.0006,
+      "step": 1330
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00026205331820760067,
+      "loss": 1.9599,
+      "step": 1340
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00026176971072036295,
+      "loss": 1.893,
+      "step": 1350
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002615144639818491,
+      "loss": 2.0587,
+      "step": 1360
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00026123085649461144,
+      "loss": 1.9907,
+      "step": 1370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002609472490073738,
+      "loss": 2.0395,
+      "step": 1380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002606636415201361,
+      "loss": 2.1024,
+      "step": 1390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00026038003403289846,
+      "loss": 2.0419,
+      "step": 1400
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002600964265456608,
+      "loss": 2.0421,
+      "step": 1410
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002598128190584231,
+      "loss": 1.9999,
+      "step": 1420
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00025952921157118547,
+      "loss": 2.0371,
+      "step": 1430
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002592456040839478,
+      "loss": 2.0273,
+      "step": 1440
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00025896199659671015,
+      "loss": 2.0051,
+      "step": 1450
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002586783891094725,
+      "loss": 2.0452,
+      "step": 1460
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00025839478162223477,
+      "loss": 2.0932,
+      "step": 1470
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00025811117413499716,
+      "loss": 2.0414,
+      "step": 1480
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002578275666477595,
+      "loss": 2.1954,
+      "step": 1490
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00025754395916052184,
+      "loss": 2.0392,
+      "step": 1500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002572603516732841,
+      "loss": 2.0825,
+      "step": 1510
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00025697674418604646,
+      "loss": 2.0541,
+      "step": 1520
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002566931366988088,
+      "loss": 2.0768,
+      "step": 1530
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002564095292115712,
+      "loss": 2.0877,
+      "step": 1540
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00025612592172433353,
+      "loss": 2.0012,
+      "step": 1550
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002558423142370958,
+      "loss": 2.1406,
+      "step": 1560
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00025555870674985815,
+      "loss": 1.9447,
+      "step": 1570
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002552750992626205,
+      "loss": 2.0579,
+      "step": 1580
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002549914917753829,
+      "loss": 2.0272,
+      "step": 1590
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002547078842881452,
+      "loss": 1.9909,
+      "step": 1600
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002544242768009075,
+      "loss": 1.891,
+      "step": 1610
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00025414066931366984,
+      "loss": 2.0163,
+      "step": 1620
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002538570618264322,
+      "loss": 2.1711,
+      "step": 1630
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00025357345433919457,
+      "loss": 2.0521,
+      "step": 1640
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00025328984685195685,
+      "loss": 2.0709,
+      "step": 1650
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002530062393647192,
+      "loss": 2.0268,
+      "step": 1660
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00025272263187748153,
+      "loss": 1.9734,
+      "step": 1670
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00025243902439024387,
+      "loss": 2.028,
+      "step": 1680
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002521554169030062,
+      "loss": 1.9305,
+      "step": 1690
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002519001701644923,
+      "loss": 2.1173,
+      "step": 1700
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002516165626772547,
+      "loss": 2.0794,
+      "step": 1710
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.000251332955190017,
+      "loss": 2.0245,
+      "step": 1720
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002510493477027793,
+      "loss": 1.9065,
+      "step": 1730
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00025076574021554165,
+      "loss": 1.9802,
+      "step": 1740
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.000250482132728304,
+      "loss": 2.0261,
+      "step": 1750
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00025019852524106633,
+      "loss": 2.0128,
+      "step": 1760
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00024991491775382867,
+      "loss": 2.0552,
+      "step": 1770
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.000249631310266591,
+      "loss": 2.0746,
+      "step": 1780
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00024934770277935334,
+      "loss": 2.1647,
+      "step": 1790
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002490640952921157,
+      "loss": 2.0698,
+      "step": 1800
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.000248780487804878,
+      "loss": 2.0405,
+      "step": 1810
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00024849688031764036,
+      "loss": 1.9449,
+      "step": 1820
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002482132728304027,
+      "loss": 1.9748,
+      "step": 1830
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00024792966534316503,
+      "loss": 2.0121,
+      "step": 1840
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00024764605785592737,
+      "loss": 2.0935,
+      "step": 1850
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002473624503686897,
+      "loss": 1.9214,
+      "step": 1860
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00024707884288145205,
+      "loss": 1.9404,
+      "step": 1870
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002467952353942144,
+      "loss": 2.0133,
+      "step": 1880
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002465116279069767,
+      "loss": 2.098,
+      "step": 1890
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00024622802041973906,
+      "loss": 2.0551,
+      "step": 1900
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002459444129325014,
+      "loss": 1.9603,
+      "step": 1910
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00024566080544526374,
+      "loss": 2.0956,
+      "step": 1920
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002453771979580261,
+      "loss": 1.9561,
+      "step": 1930
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002450935904707884,
+      "loss": 1.9836,
+      "step": 1940
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00024480998298355075,
+      "loss": 1.8087,
+      "step": 1950
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002445263754963131,
+      "loss": 1.919,
+      "step": 1960
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00024424276800907543,
+      "loss": 2.0155,
+      "step": 1970
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00024395916052183777,
+      "loss": 2.0345,
+      "step": 1980
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002436755530346001,
+      "loss": 2.0431,
+      "step": 1990
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00024339194554736244,
+      "loss": 2.0517,
+      "step": 2000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00024310833806012475,
+      "loss": 2.0122,
+      "step": 2010
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002428247305728871,
+      "loss": 1.8942,
+      "step": 2020
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00024254112308564943,
+      "loss": 1.8664,
+      "step": 2030
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002422575155984118,
+      "loss": 1.9414,
+      "step": 2040
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00024197390811117413,
+      "loss": 2.0272,
+      "step": 2050
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00024169030062393644,
+      "loss": 1.9806,
+      "step": 2060
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00024140669313669878,
+      "loss": 2.0221,
+      "step": 2070
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00024112308564946112,
+      "loss": 2.0129,
+      "step": 2080
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00024083947816222348,
+      "loss": 2.0592,
+      "step": 2090
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002405558706749858,
+      "loss": 1.9784,
+      "step": 2100
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00024027226318774813,
+      "loss": 1.9588,
+      "step": 2110
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00023998865570051047,
+      "loss": 2.0713,
+      "step": 2120
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002397050482132728,
+      "loss": 1.8854,
+      "step": 2130
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00023942144072603512,
+      "loss": 2.0537,
+      "step": 2140
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00023913783323879748,
+      "loss": 2.0167,
+      "step": 2150
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00023885422575155982,
+      "loss": 1.9523,
+      "step": 2160
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00023857061826432216,
+      "loss": 1.9412,
+      "step": 2170
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002382870107770845,
+      "loss": 1.9623,
+      "step": 2180
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002380034032898468,
+      "loss": 2.0483,
+      "step": 2190
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00023771979580260917,
+      "loss": 1.9854,
+      "step": 2200
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002374361883153715,
+      "loss": 1.9711,
+      "step": 2210
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00023715258082813385,
+      "loss": 1.9786,
+      "step": 2220
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002368689733408962,
+      "loss": 2.0829,
+      "step": 2230
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002365853658536585,
+      "loss": 1.8776,
+      "step": 2240
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00023630175836642086,
+      "loss": 1.9447,
+      "step": 2250
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002360181508791832,
+      "loss": 1.997,
+      "step": 2260
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00023573454339194554,
+      "loss": 2.0287,
+      "step": 2270
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00023545093590470785,
+      "loss": 2.0302,
+      "step": 2280
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002351673284174702,
+      "loss": 1.9696,
+      "step": 2290
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00023488372093023252,
+      "loss": 2.0416,
+      "step": 2300
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002346001134429949,
+      "loss": 1.8271,
+      "step": 2310
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00023431650595575723,
+      "loss": 1.9771,
+      "step": 2320
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00023403289846851954,
+      "loss": 2.0091,
+      "step": 2330
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00023374929098128188,
+      "loss": 2.1508,
+      "step": 2340
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00023346568349404421,
+      "loss": 1.9832,
+      "step": 2350
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00023318207600680658,
+      "loss": 2.0144,
+      "step": 2360
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002328984685195689,
+      "loss": 1.9328,
+      "step": 2370
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00023261486103233123,
+      "loss": 1.891,
+      "step": 2380
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00023233125354509357,
+      "loss": 2.0427,
+      "step": 2390
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002320476460578559,
+      "loss": 1.874,
+      "step": 2400
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00023176403857061827,
+      "loss": 1.9653,
+      "step": 2410
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00023148043108338058,
+      "loss": 2.0203,
+      "step": 2420
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00023119682359614292,
+      "loss": 1.8663,
+      "step": 2430
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00023091321610890526,
+      "loss": 1.8917,
+      "step": 2440
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002306296086216676,
+      "loss": 1.9734,
+      "step": 2450
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002303460011344299,
+      "loss": 1.9634,
+      "step": 2460
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00023006239364719227,
+      "loss": 1.9635,
+      "step": 2470
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002297787861599546,
+      "loss": 1.9262,
+      "step": 2480
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00022949517867271695,
+      "loss": 1.9753,
+      "step": 2490
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00022921157118547928,
+      "loss": 2.0465,
+      "step": 2500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002289279636982416,
+      "loss": 1.879,
+      "step": 2510
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00022864435621100396,
+      "loss": 2.0032,
+      "step": 2520
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002283607487237663,
+      "loss": 1.9896,
+      "step": 2530
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00022807714123652864,
+      "loss": 1.9966,
+      "step": 2540
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00022779353374929095,
+      "loss": 1.9653,
+      "step": 2550
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00022750992626205328,
+      "loss": 1.9386,
+      "step": 2560
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00022722631877481562,
+      "loss": 2.0176,
+      "step": 2570
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.000226942711287578,
+      "loss": 1.9585,
+      "step": 2580
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00022665910380034033,
+      "loss": 1.9366,
+      "step": 2590
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00022637549631310264,
+      "loss": 2.0283,
+      "step": 2600
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00022609188882586497,
+      "loss": 1.9384,
+      "step": 2610
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002258082813386273,
+      "loss": 2.1084,
+      "step": 2620
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00022552467385138968,
+      "loss": 2.0181,
+      "step": 2630
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.000225241066364152,
+      "loss": 1.8862,
+      "step": 2640
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00022495745887691433,
+      "loss": 2.0016,
+      "step": 2650
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00022467385138967666,
+      "loss": 1.9543,
+      "step": 2660
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.000224390243902439,
+      "loss": 1.9392,
+      "step": 2670
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00022410663641520137,
+      "loss": 1.9454,
+      "step": 2680
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00022382302892796368,
+      "loss": 2.0417,
+      "step": 2690
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00022353942144072602,
+      "loss": 2.0467,
+      "step": 2700
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00022325581395348835,
+      "loss": 1.8961,
+      "step": 2710
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002229722064662507,
+      "loss": 2.0104,
+      "step": 2720
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.000222688598979013,
+      "loss": 1.9292,
+      "step": 2730
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00022240499149177537,
+      "loss": 2.0145,
+      "step": 2740
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002221213840045377,
+      "loss": 2.0363,
+      "step": 2750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00022183777651730004,
+      "loss": 2.0434,
+      "step": 2760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00022155416903006238,
+      "loss": 1.9628,
+      "step": 2770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002212705615428247,
+      "loss": 2.0745,
+      "step": 2780
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00022098695405558706,
+      "loss": 2.0417,
+      "step": 2790
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002207033465683494,
+      "loss": 1.9554,
+      "step": 2800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00022041973908111173,
+      "loss": 2.0323,
+      "step": 2810
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00022013613159387404,
+      "loss": 2.0586,
+      "step": 2820
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00021985252410663638,
+      "loss": 1.8894,
+      "step": 2830
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00021956891661939875,
+      "loss": 1.8927,
+      "step": 2840
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00021928530913216108,
+      "loss": 1.8973,
+      "step": 2850
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00021900170164492342,
+      "loss": 1.9644,
+      "step": 2860
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00021871809415768573,
+      "loss": 1.9039,
+      "step": 2870
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00021843448667044807,
+      "loss": 1.9316,
+      "step": 2880
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002181508791832104,
+      "loss": 1.8483,
+      "step": 2890
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00021786727169597277,
+      "loss": 2.0534,
+      "step": 2900
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00021758366420873508,
+      "loss": 1.9126,
+      "step": 2910
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00021730005672149742,
+      "loss": 2.0031,
+      "step": 2920
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00021701644923425976,
+      "loss": 2.1399,
+      "step": 2930
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002167328417470221,
+      "loss": 2.1048,
+      "step": 2940
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00021644923425978446,
+      "loss": 1.8944,
+      "step": 2950
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00021616562677254677,
+      "loss": 2.0135,
+      "step": 2960
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002158820192853091,
+      "loss": 1.9103,
+      "step": 2970
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00021559841179807145,
+      "loss": 1.8927,
+      "step": 2980
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002153148043108338,
+      "loss": 1.9573,
+      "step": 2990
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002150311968235961,
+      "loss": 2.0191,
+      "step": 3000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00021474758933635846,
+      "loss": 1.9678,
+      "step": 3010
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002144639818491208,
+      "loss": 1.8932,
+      "step": 3020
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00021418037436188314,
+      "loss": 1.9244,
+      "step": 3030
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00021389676687464548,
+      "loss": 2.0792,
+      "step": 3040
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002136131593874078,
+      "loss": 1.8409,
+      "step": 3050
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00021332955190017015,
+      "loss": 1.9291,
+      "step": 3060
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002130459444129325,
+      "loss": 1.9294,
+      "step": 3070
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00021276233692569483,
+      "loss": 2.0805,
+      "step": 3080
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00021247872943845714,
+      "loss": 1.8575,
+      "step": 3090
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00021219512195121948,
+      "loss": 2.0922,
+      "step": 3100
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00021191151446398184,
+      "loss": 2.0071,
+      "step": 3110
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00021162790697674418,
+      "loss": 1.9723,
+      "step": 3120
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00021134429948950652,
+      "loss": 1.936,
+      "step": 3130
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00021106069200226883,
+      "loss": 2.0164,
+      "step": 3140
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00021077708451503117,
+      "loss": 2.029,
+      "step": 3150
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002104934770277935,
+      "loss": 1.9273,
+      "step": 3160
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00021020986954055587,
+      "loss": 2.0218,
+      "step": 3170
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00020992626205331818,
+      "loss": 1.9076,
+      "step": 3180
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00020964265456608052,
+      "loss": 1.9888,
+      "step": 3190
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00020935904707884286,
+      "loss": 1.9866,
+      "step": 3200
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002090754395916052,
+      "loss": 1.9614,
+      "step": 3210
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00020879183210436756,
+      "loss": 1.9859,
+      "step": 3220
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00020850822461712987,
+      "loss": 1.9241,
+      "step": 3230
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0002082246171298922,
+      "loss": 2.0371,
+      "step": 3240
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00020794100964265455,
+      "loss": 1.982,
+      "step": 3250
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00020765740215541689,
+      "loss": 2.0358,
+      "step": 3260
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00020737379466817925,
+      "loss": 1.9841,
+      "step": 3270
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00020709018718094156,
+      "loss": 1.9346,
+      "step": 3280
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0002068065796937039,
+      "loss": 1.8959,
+      "step": 3290
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00020652297220646624,
+      "loss": 1.9524,
+      "step": 3300
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00020623936471922855,
+      "loss": 1.8774,
+      "step": 3310
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00020595575723199089,
+      "loss": 2.0079,
+      "step": 3320
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00020567214974475325,
+      "loss": 1.9101,
+      "step": 3330
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0002053885422575156,
+      "loss": 1.992,
+      "step": 3340
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00020510493477027793,
+      "loss": 1.9441,
+      "step": 3350
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00020482132728304024,
+      "loss": 2.0872,
+      "step": 3360
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00020453771979580258,
+      "loss": 2.1295,
+      "step": 3370
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00020425411230856494,
+      "loss": 1.9662,
+      "step": 3380
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00020397050482132728,
+      "loss": 2.0128,
+      "step": 3390
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00020368689733408962,
+      "loss": 1.9651,
+      "step": 3400
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00020340328984685193,
+      "loss": 1.9138,
+      "step": 3410
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00020311968235961427,
+      "loss": 2.0154,
+      "step": 3420
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0002028360748723766,
+      "loss": 2.0664,
+      "step": 3430
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00020255246738513897,
+      "loss": 2.0125,
+      "step": 3440
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00020226885989790128,
+      "loss": 1.8252,
+      "step": 3450
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00020198525241066362,
+      "loss": 1.9527,
+      "step": 3460
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00020170164492342595,
+      "loss": 1.9952,
+      "step": 3470
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0002014180374361883,
+      "loss": 1.8446,
+      "step": 3480
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00020113442994895066,
+      "loss": 1.9602,
+      "step": 3490
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00020085082246171297,
+      "loss": 2.0458,
+      "step": 3500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0002005672149744753,
+      "loss": 1.9447,
+      "step": 3510
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00020028360748723764,
+      "loss": 2.0325,
+      "step": 3520
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00019999999999999998,
+      "loss": 1.9888,
+      "step": 3530
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00019971639251276235,
+      "loss": 1.9477,
+      "step": 3540
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00019943278502552466,
+      "loss": 1.9787,
+      "step": 3550
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.000199149177538287,
+      "loss": 1.8773,
+      "step": 3560
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00019886557005104933,
+      "loss": 1.9589,
+      "step": 3570
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00019858196256381165,
+      "loss": 2.0078,
+      "step": 3580
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00019829835507657398,
+      "loss": 1.9154,
+      "step": 3590
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00019801474758933635,
+      "loss": 1.9619,
+      "step": 3600
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00019773114010209869,
+      "loss": 1.9137,
+      "step": 3610
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00019744753261486102,
+      "loss": 1.9595,
+      "step": 3620
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00019716392512762333,
+      "loss": 1.953,
+      "step": 3630
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00019688031764038567,
+      "loss": 1.9825,
+      "step": 3640
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019659671015314804,
+      "loss": 1.8326,
+      "step": 3650
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019631310266591038,
+      "loss": 1.9083,
+      "step": 3660
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019602949517867271,
+      "loss": 2.0078,
+      "step": 3670
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019574588769143502,
+      "loss": 1.9444,
+      "step": 3680
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019546228020419736,
+      "loss": 1.8508,
+      "step": 3690
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001951786727169597,
+      "loss": 1.8987,
+      "step": 3700
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019489506522972207,
+      "loss": 2.0121,
+      "step": 3710
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019461145774248438,
+      "loss": 1.9966,
+      "step": 3720
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019432785025524671,
+      "loss": 1.8469,
+      "step": 3730
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019404424276800905,
+      "loss": 2.0692,
+      "step": 3740
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001937606352807714,
+      "loss": 2.0527,
+      "step": 3750
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019347702779353375,
+      "loss": 1.9751,
+      "step": 3760
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019319342030629607,
+      "loss": 1.8138,
+      "step": 3770
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001929098128190584,
+      "loss": 1.981,
+      "step": 3780
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019262620533182074,
+      "loss": 2.0095,
+      "step": 3790
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019234259784458308,
+      "loss": 2.1045,
+      "step": 3800
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019205899035734544,
+      "loss": 1.9747,
+      "step": 3810
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019177538287010776,
+      "loss": 1.8581,
+      "step": 3820
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001914917753828701,
+      "loss": 1.871,
+      "step": 3830
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019120816789563243,
+      "loss": 1.9531,
+      "step": 3840
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019092456040839474,
+      "loss": 1.8435,
+      "step": 3850
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019064095292115708,
+      "loss": 1.9877,
+      "step": 3860
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00019035734543391945,
+      "loss": 1.9131,
+      "step": 3870
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00019007373794668178,
+      "loss": 1.8909,
+      "step": 3880
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00018979013045944412,
+      "loss": 1.9501,
+      "step": 3890
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00018950652297220643,
+      "loss": 1.9938,
+      "step": 3900
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00018922291548496877,
+      "loss": 1.9519,
+      "step": 3910
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00018893930799773113,
+      "loss": 1.9293,
+      "step": 3920
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00018865570051049347,
+      "loss": 1.9221,
+      "step": 3930
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001883720930232558,
+      "loss": 1.8675,
+      "step": 3940
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00018808848553601812,
+      "loss": 1.9782,
+      "step": 3950
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00018780487804878046,
+      "loss": 1.9747,
+      "step": 3960
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00018752127056154282,
+      "loss": 1.9417,
+      "step": 3970
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00018723766307430516,
+      "loss": 2.0062,
+      "step": 3980
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00018698241633579126,
+      "loss": 2.0212,
+      "step": 3990
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001866988088485536,
+      "loss": 2.0481,
+      "step": 4000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00018641520136131594,
+      "loss": 2.0239,
+      "step": 4010
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00018613159387407825,
+      "loss": 1.8528,
+      "step": 4020
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00018584798638684059,
+      "loss": 1.9132,
+      "step": 4030
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00018556437889960292,
+      "loss": 1.9824,
+      "step": 4040
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001852807714123653,
+      "loss": 1.9436,
+      "step": 4050
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001849971639251276,
+      "loss": 1.8854,
+      "step": 4060
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00018471355643788994,
+      "loss": 2.0388,
+      "step": 4070
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00018442994895065228,
+      "loss": 1.978,
+      "step": 4080
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001841463414634146,
+      "loss": 1.9922,
+      "step": 4090
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00018386273397617698,
+      "loss": 1.8897,
+      "step": 4100
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001835791264889393,
+      "loss": 1.7802,
+      "step": 4110
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00018329551900170163,
+      "loss": 1.9593,
+      "step": 4120
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00018301191151446397,
+      "loss": 1.9404,
+      "step": 4130
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001827283040272263,
+      "loss": 1.9313,
+      "step": 4140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00018244469653998867,
+      "loss": 1.8596,
+      "step": 4150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00018216108905275098,
+      "loss": 1.9912,
+      "step": 4160
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00018187748156551332,
+      "loss": 1.8539,
+      "step": 4170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00018159387407827565,
+      "loss": 2.0064,
+      "step": 4180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00018131026659103797,
+      "loss": 1.9917,
+      "step": 4190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001810266591038003,
+      "loss": 2.0214,
+      "step": 4200
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00018074305161656267,
+      "loss": 1.7989,
+      "step": 4210
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.000180459444129325,
+      "loss": 1.9071,
+      "step": 4220
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00018017583664208734,
+      "loss": 2.0126,
+      "step": 4230
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00017989222915484966,
+      "loss": 1.9182,
+      "step": 4240
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.000179608621667612,
+      "loss": 1.9052,
+      "step": 4250
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00017932501418037436,
+      "loss": 2.038,
+      "step": 4260
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001790414066931367,
+      "loss": 1.8673,
+      "step": 4270
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00017875779920589903,
+      "loss": 2.0043,
+      "step": 4280
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00017847419171866134,
+      "loss": 1.9785,
+      "step": 4290
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00017819058423142368,
+      "loss": 1.871,
+      "step": 4300
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00017790697674418605,
+      "loss": 2.1149,
+      "step": 4310
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00017762336925694839,
+      "loss": 1.8676,
+      "step": 4320
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001773397617697107,
+      "loss": 2.0189,
+      "step": 4330
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00017705615428247303,
+      "loss": 1.8926,
+      "step": 4340
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00017677254679523537,
+      "loss": 1.9703,
+      "step": 4350
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001764889393079977,
+      "loss": 1.9466,
+      "step": 4360
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00017620533182076008,
+      "loss": 1.8895,
+      "step": 4370
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00017592172433352239,
+      "loss": 1.9457,
+      "step": 4380
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00017563811684628472,
+      "loss": 1.8517,
+      "step": 4390
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00017535450935904706,
+      "loss": 1.8771,
+      "step": 4400
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001750709018718094,
+      "loss": 2.0174,
+      "step": 4410
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00017478729438457177,
+      "loss": 1.976,
+      "step": 4420
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00017450368689733408,
+      "loss": 2.0323,
+      "step": 4430
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00017422007941009641,
+      "loss": 1.9091,
+      "step": 4440
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00017393647192285875,
+      "loss": 1.9384,
+      "step": 4450
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00017365286443562106,
+      "loss": 1.8759,
+      "step": 4460
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001733692569483834,
+      "loss": 1.9307,
+      "step": 4470
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00017308564946114577,
+      "loss": 2.025,
+      "step": 4480
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001728020419739081,
+      "loss": 1.9593,
+      "step": 4490
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00017251843448667044,
+      "loss": 2.0366,
+      "step": 4500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00017223482699943275,
+      "loss": 2.0388,
+      "step": 4510
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001719512195121951,
+      "loss": 1.9827,
+      "step": 4520
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00017166761202495746,
+      "loss": 1.9018,
+      "step": 4530
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001713840045377198,
+      "loss": 1.8422,
+      "step": 4540
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00017110039705048213,
+      "loss": 1.8816,
+      "step": 4550
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00017081678956324444,
+      "loss": 1.9177,
+      "step": 4560
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00017053318207600678,
+      "loss": 1.9647,
+      "step": 4570
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00017024957458876915,
+      "loss": 1.8367,
+      "step": 4580
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00016996596710153148,
+      "loss": 1.958,
+      "step": 4590
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001696823596142938,
+      "loss": 1.8869,
+      "step": 4600
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00016939875212705613,
+      "loss": 1.976,
+      "step": 4610
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00016911514463981847,
+      "loss": 1.9574,
+      "step": 4620
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001688315371525808,
+      "loss": 1.9362,
+      "step": 4630
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00016854792966534317,
+      "loss": 1.9701,
+      "step": 4640
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00016826432217810548,
+      "loss": 1.8939,
+      "step": 4650
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00016798071469086782,
+      "loss": 1.8127,
+      "step": 4660
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00016769710720363016,
+      "loss": 1.965,
+      "step": 4670
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001674134997163925,
+      "loss": 1.8539,
+      "step": 4680
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00016712989222915486,
+      "loss": 1.9504,
+      "step": 4690
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00016684628474191717,
+      "loss": 1.9817,
+      "step": 4700
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001665626772546795,
+      "loss": 1.9137,
+      "step": 4710
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001663074305161656,
+      "loss": 1.9258,
+      "step": 4720
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00016602382302892795,
+      "loss": 1.7631,
+      "step": 4730
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00016574021554169029,
+      "loss": 1.8479,
+      "step": 4740
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00016545660805445262,
+      "loss": 1.8883,
+      "step": 4750
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00016517300056721496,
+      "loss": 1.9619,
+      "step": 4760
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001648893930799773,
+      "loss": 1.9964,
+      "step": 4770
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00016460578559273964,
+      "loss": 1.9687,
+      "step": 4780
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00016432217810550198,
+      "loss": 2.0085,
+      "step": 4790
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00016403857061826429,
+      "loss": 1.9943,
+      "step": 4800
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00016375496313102662,
+      "loss": 1.9711,
+      "step": 4810
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.000163471355643789,
+      "loss": 1.7781,
+      "step": 4820
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016318774815655133,
+      "loss": 1.849,
+      "step": 4830
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016290414066931367,
+      "loss": 1.9066,
+      "step": 4840
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016262053318207598,
+      "loss": 1.9813,
+      "step": 4850
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016233692569483831,
+      "loss": 1.9247,
+      "step": 4860
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016205331820760068,
+      "loss": 2.0478,
+      "step": 4870
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016176971072036302,
+      "loss": 1.8815,
+      "step": 4880
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016148610323312535,
+      "loss": 1.8873,
+      "step": 4890
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016120249574588767,
+      "loss": 1.8828,
+      "step": 4900
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016091888825865,
+      "loss": 1.9985,
+      "step": 4910
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00016063528077141237,
+      "loss": 1.9237,
+      "step": 4920
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001603516732841747,
+      "loss": 2.03,
+      "step": 4930
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00016006806579693702,
+      "loss": 1.881,
+      "step": 4940
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00015978445830969936,
+      "loss": 1.9496,
+      "step": 4950
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001595008508224617,
+      "loss": 1.9751,
+      "step": 4960
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00015921724333522403,
+      "loss": 1.9493,
+      "step": 4970
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001589336358479864,
+      "loss": 1.9607,
+      "step": 4980
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001586500283607487,
+      "loss": 1.9989,
+      "step": 4990
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00015836642087351104,
+      "loss": 2.0145,
+      "step": 5000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00015808281338627338,
+      "loss": 1.8441,
+      "step": 5010
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00015779920589903572,
+      "loss": 1.9853,
+      "step": 5020
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00015751559841179806,
+      "loss": 1.8068,
+      "step": 5030
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001572319909245604,
+      "loss": 1.9993,
+      "step": 5040
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00015694838343732273,
+      "loss": 1.8017,
+      "step": 5050
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00015666477595008507,
+      "loss": 1.9598,
+      "step": 5060
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00015638116846284738,
+      "loss": 1.8583,
+      "step": 5070
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00015609756097560975,
+      "loss": 1.9907,
+      "step": 5080
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00015581395348837209,
+      "loss": 1.9518,
+      "step": 5090
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00015553034600113442,
+      "loss": 2.013,
+      "step": 5100
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00015524673851389676,
+      "loss": 1.8681,
+      "step": 5110
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00015496313102665907,
+      "loss": 1.8049,
+      "step": 5120
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001546795235394214,
+      "loss": 1.9509,
+      "step": 5130
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00015439591605218378,
+      "loss": 2.0249,
+      "step": 5140
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00015411230856494611,
+      "loss": 1.9814,
+      "step": 5150
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00015382870107770845,
+      "loss": 1.9319,
+      "step": 5160
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00015354509359047076,
+      "loss": 1.8496,
+      "step": 5170
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001532614861032331,
+      "loss": 1.9162,
+      "step": 5180
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00015297787861599547,
+      "loss": 1.9752,
+      "step": 5190
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001526942711287578,
+      "loss": 1.9887,
+      "step": 5200
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00015241066364152011,
+      "loss": 1.9524,
+      "step": 5210
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00015212705615428245,
+      "loss": 2.0026,
+      "step": 5220
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001518434486670448,
+      "loss": 1.9157,
+      "step": 5230
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00015155984117980713,
+      "loss": 2.0611,
+      "step": 5240
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001512762336925695,
+      "loss": 2.0075,
+      "step": 5250
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001509926262053318,
+      "loss": 1.9902,
+      "step": 5260
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00015070901871809414,
+      "loss": 1.9127,
+      "step": 5270
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00015042541123085648,
+      "loss": 1.998,
+      "step": 5280
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00015014180374361882,
+      "loss": 1.8734,
+      "step": 5290
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00014985819625638116,
+      "loss": 1.9243,
+      "step": 5300
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001495745887691435,
+      "loss": 1.9015,
+      "step": 5310
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00014929098128190583,
+      "loss": 1.8077,
+      "step": 5320
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00014900737379466817,
+      "loss": 1.8186,
+      "step": 5330
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001487237663074305,
+      "loss": 1.9202,
+      "step": 5340
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00014844015882019285,
+      "loss": 1.8258,
+      "step": 5350
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00014815655133295518,
+      "loss": 1.8155,
+      "step": 5360
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00014787294384571752,
+      "loss": 1.9628,
+      "step": 5370
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00014758933635847986,
+      "loss": 1.9668,
+      "step": 5380
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001473057288712422,
+      "loss": 1.956,
+      "step": 5390
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00014702212138400454,
+      "loss": 1.9612,
+      "step": 5400
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00014673851389676685,
+      "loss": 1.9822,
+      "step": 5410
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001464549064095292,
+      "loss": 1.9493,
+      "step": 5420
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00014617129892229152,
+      "loss": 1.9886,
+      "step": 5430
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00014588769143505386,
+      "loss": 1.9212,
+      "step": 5440
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00014560408394781622,
+      "loss": 1.8092,
+      "step": 5450
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00014532047646057854,
+      "loss": 1.9926,
+      "step": 5460
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001450368689733409,
+      "loss": 1.8548,
+      "step": 5470
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001447532614861032,
+      "loss": 1.8581,
+      "step": 5480
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00014446965399886555,
+      "loss": 1.9153,
+      "step": 5490
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001441860465116279,
+      "loss": 2.0769,
+      "step": 5500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00014390243902439023,
+      "loss": 1.9996,
+      "step": 5510
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00014361883153715256,
+      "loss": 1.9398,
+      "step": 5520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001433352240499149,
+      "loss": 1.9979,
+      "step": 5530
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00014305161656267724,
+      "loss": 1.804,
+      "step": 5540
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00014276800907543958,
+      "loss": 1.936,
+      "step": 5550
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00014248440158820192,
+      "loss": 1.9309,
+      "step": 5560
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00014220079410096425,
+      "loss": 1.9491,
+      "step": 5570
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001419171866137266,
+      "loss": 1.9203,
+      "step": 5580
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00014163357912648893,
+      "loss": 1.8223,
+      "step": 5590
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00014134997163925127,
+      "loss": 1.922,
+      "step": 5600
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001410663641520136,
+      "loss": 1.9175,
+      "step": 5610
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00014078275666477594,
+      "loss": 1.9536,
+      "step": 5620
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00014049914917753828,
+      "loss": 1.8396,
+      "step": 5630
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00014021554169030062,
+      "loss": 1.9098,
+      "step": 5640
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00013993193420306296,
+      "loss": 1.9491,
+      "step": 5650
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001396483267158253,
+      "loss": 1.8775,
+      "step": 5660
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013936471922858763,
+      "loss": 1.9059,
+      "step": 5670
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013908111174134994,
+      "loss": 2.0328,
+      "step": 5680
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001387975042541123,
+      "loss": 1.9487,
+      "step": 5690
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013851389676687462,
+      "loss": 1.9911,
+      "step": 5700
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013823028927963696,
+      "loss": 1.9483,
+      "step": 5710
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013794668179239932,
+      "loss": 1.8927,
+      "step": 5720
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013766307430516163,
+      "loss": 1.8362,
+      "step": 5730
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.000137379466817924,
+      "loss": 1.9886,
+      "step": 5740
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001370958593306863,
+      "loss": 2.0213,
+      "step": 5750
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013681225184344865,
+      "loss": 2.0016,
+      "step": 5760
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00013652864435621098,
+      "loss": 1.9383,
+      "step": 5770
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00013624503686897332,
+      "loss": 2.0477,
+      "step": 5780
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00013596142938173566,
+      "loss": 1.8091,
+      "step": 5790
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.000135677821894498,
+      "loss": 1.9684,
+      "step": 5800
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00013539421440726034,
+      "loss": 1.9759,
+      "step": 5810
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00013511060692002267,
+      "loss": 1.8849,
+      "step": 5820
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.000134826999432785,
+      "loss": 1.926,
+      "step": 5830
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00013454339194554735,
+      "loss": 1.972,
+      "step": 5840
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001342597844583097,
+      "loss": 1.812,
+      "step": 5850
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00013397617697107203,
+      "loss": 1.9176,
+      "step": 5860
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00013369256948383436,
+      "loss": 1.9561,
+      "step": 5870
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001334089619965967,
+      "loss": 1.982,
+      "step": 5880
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00013312535450935904,
+      "loss": 1.9534,
+      "step": 5890
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00013284174702212138,
+      "loss": 1.9424,
+      "step": 5900
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00013255813953488372,
+      "loss": 1.9401,
+      "step": 5910
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00013227453204764605,
+      "loss": 1.9386,
+      "step": 5920
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001319909245604084,
+      "loss": 1.9403,
+      "step": 5930
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00013170731707317073,
+      "loss": 2.0248,
+      "step": 5940
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00013142370958593304,
+      "loss": 1.8946,
+      "step": 5950
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001311401020986954,
+      "loss": 1.9736,
+      "step": 5960
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00013085649461145772,
+      "loss": 2.041,
+      "step": 5970
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00013057288712422005,
+      "loss": 1.8931,
+      "step": 5980
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00013028927963698242,
+      "loss": 1.9143,
+      "step": 5990
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00013000567214974473,
+      "loss": 1.8607,
+      "step": 6000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001297220646625071,
+      "loss": 2.0425,
+      "step": 6010
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001294384571752694,
+      "loss": 1.9615,
+      "step": 6020
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00012915484968803174,
+      "loss": 1.9234,
+      "step": 6030
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00012887124220079408,
+      "loss": 1.996,
+      "step": 6040
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00012858763471355642,
+      "loss": 1.9451,
+      "step": 6050
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00012830402722631876,
+      "loss": 1.9887,
+      "step": 6060
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001280204197390811,
+      "loss": 1.8884,
+      "step": 6070
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00012773681225184343,
+      "loss": 1.8801,
+      "step": 6080
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00012745320476460577,
+      "loss": 1.948,
+      "step": 6090
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001271695972773681,
+      "loss": 1.9517,
+      "step": 6100
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00012688598979013045,
+      "loss": 1.9363,
+      "step": 6110
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00012660238230289279,
+      "loss": 1.985,
+      "step": 6120
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00012631877481565512,
+      "loss": 2.0393,
+      "step": 6130
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00012603516732841746,
+      "loss": 1.8576,
+      "step": 6140
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001257515598411798,
+      "loss": 1.8394,
+      "step": 6150
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00012546795235394214,
+      "loss": 1.9693,
+      "step": 6160
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00012518434486670447,
+      "loss": 2.0355,
+      "step": 6170
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001249007373794668,
+      "loss": 1.8122,
+      "step": 6180
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00012461712989222915,
+      "loss": 1.8931,
+      "step": 6190
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001243335224049915,
+      "loss": 1.9087,
+      "step": 6200
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00012404991491775383,
+      "loss": 1.9078,
+      "step": 6210
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00012376630743051614,
+      "loss": 1.917,
+      "step": 6220
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001234826999432785,
+      "loss": 2.0256,
+      "step": 6230
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001231990924560408,
+      "loss": 2.0448,
+      "step": 6240
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00012291548496880315,
+      "loss": 1.9633,
+      "step": 6250
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00012263187748156552,
+      "loss": 2.0159,
+      "step": 6260
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00012234826999432783,
+      "loss": 1.8067,
+      "step": 6270
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00012206466250709018,
+      "loss": 1.9475,
+      "step": 6280
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00012178105501985252,
+      "loss": 1.9107,
+      "step": 6290
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00012149744753261484,
+      "loss": 1.8108,
+      "step": 6300
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00012121384004537719,
+      "loss": 2.0893,
+      "step": 6310
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00012093023255813952,
+      "loss": 1.9486,
+      "step": 6320
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00012064662507090185,
+      "loss": 1.8204,
+      "step": 6330
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001203630175836642,
+      "loss": 1.9909,
+      "step": 6340
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00012007941009642653,
+      "loss": 1.9413,
+      "step": 6350
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00011979580260918888,
+      "loss": 1.88,
+      "step": 6360
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001195121951219512,
+      "loss": 1.8708,
+      "step": 6370
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00011922858763471354,
+      "loss": 2.019,
+      "step": 6380
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00011894498014747588,
+      "loss": 1.82,
+      "step": 6390
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00011866137266023822,
+      "loss": 1.9024,
+      "step": 6400
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00011837776517300054,
+      "loss": 2.0031,
+      "step": 6410
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001180941576857629,
+      "loss": 1.9088,
+      "step": 6420
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00011781055019852523,
+      "loss": 1.8936,
+      "step": 6430
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00011752694271128757,
+      "loss": 1.7903,
+      "step": 6440
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00011724333522404991,
+      "loss": 1.9177,
+      "step": 6450
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00011695972773681223,
+      "loss": 1.9584,
+      "step": 6460
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00011667612024957459,
+      "loss": 1.9577,
+      "step": 6470
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00011639251276233691,
+      "loss": 1.9879,
+      "step": 6480
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00011610890527509925,
+      "loss": 1.8903,
+      "step": 6490
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001158252977878616,
+      "loss": 1.8622,
+      "step": 6500
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00011554169030062392,
+      "loss": 1.8421,
+      "step": 6510
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00011525808281338628,
+      "loss": 1.8843,
+      "step": 6520
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001149744753261486,
+      "loss": 1.7929,
+      "step": 6530
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00011469086783891094,
+      "loss": 1.987,
+      "step": 6540
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00011440726035167328,
+      "loss": 1.9068,
+      "step": 6550
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00011412365286443561,
+      "loss": 1.9718,
+      "step": 6560
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00011384004537719794,
+      "loss": 2.045,
+      "step": 6570
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00011355643788996029,
+      "loss": 1.9014,
+      "step": 6580
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00011327283040272261,
+      "loss": 1.7638,
+      "step": 6590
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00011298922291548497,
+      "loss": 2.0565,
+      "step": 6600
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001127056154282473,
+      "loss": 1.8918,
+      "step": 6610
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00011242200794100963,
+      "loss": 2.0033,
+      "step": 6620
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00011213840045377198,
+      "loss": 1.93,
+      "step": 6630
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001118547929665343,
+      "loss": 1.8651,
+      "step": 6640
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00011159954622802042,
+      "loss": 1.8975,
+      "step": 6650
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00011131593874078274,
+      "loss": 1.8809,
+      "step": 6660
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00011103233125354508,
+      "loss": 1.9653,
+      "step": 6670
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00011074872376630743,
+      "loss": 1.9804,
+      "step": 6680
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00011046511627906975,
+      "loss": 2.0938,
+      "step": 6690
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001101815087918321,
+      "loss": 1.9611,
+      "step": 6700
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00010989790130459443,
+      "loss": 2.0472,
+      "step": 6710
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00010961429381735677,
+      "loss": 1.8546,
+      "step": 6720
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001093306863301191,
+      "loss": 1.9998,
+      "step": 6730
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00010904707884288144,
+      "loss": 1.8905,
+      "step": 6740
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00010876347135564377,
+      "loss": 1.8993,
+      "step": 6750
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00010847986386840612,
+      "loss": 1.9437,
+      "step": 6760
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00010819625638116844,
+      "loss": 1.9391,
+      "step": 6770
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001079126488939308,
+      "loss": 1.8582,
+      "step": 6780
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00010762904140669313,
+      "loss": 1.9534,
+      "step": 6790
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00010734543391945546,
+      "loss": 1.8838,
+      "step": 6800
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00010706182643221781,
+      "loss": 1.8608,
+      "step": 6810
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00010677821894498013,
+      "loss": 2.0729,
+      "step": 6820
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00010649461145774247,
+      "loss": 1.8698,
+      "step": 6830
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00010621100397050481,
+      "loss": 2.0031,
+      "step": 6840
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00010592739648326715,
+      "loss": 1.8842,
+      "step": 6850
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001056437889960295,
+      "loss": 1.9431,
+      "step": 6860
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00010536018150879182,
+      "loss": 1.9264,
+      "step": 6870
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00010507657402155416,
+      "loss": 1.9616,
+      "step": 6880
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001047929665343165,
+      "loss": 2.0009,
+      "step": 6890
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00010450935904707884,
+      "loss": 1.9505,
+      "step": 6900
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00010422575155984116,
+      "loss": 2.0046,
+      "step": 6910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00010394214407260351,
+      "loss": 1.9535,
+      "step": 6920
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00010365853658536584,
+      "loss": 1.9317,
+      "step": 6930
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00010337492909812819,
+      "loss": 2.0639,
+      "step": 6940
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00010309132161089053,
+      "loss": 1.9932,
+      "step": 6950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00010280771412365285,
+      "loss": 1.9404,
+      "step": 6960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001025241066364152,
+      "loss": 1.8413,
+      "step": 6970
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00010224049914917753,
+      "loss": 1.9619,
+      "step": 6980
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00010195689166193986,
+      "loss": 1.9393,
+      "step": 6990
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001016732841747022,
+      "loss": 1.9883,
+      "step": 7000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00010138967668746454,
+      "loss": 1.9447,
+      "step": 7010
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00010110606920022687,
+      "loss": 2.0046,
+      "step": 7020
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00010082246171298922,
+      "loss": 1.9276,
+      "step": 7030
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010053885422575154,
+      "loss": 1.8701,
+      "step": 7040
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010025524673851389,
+      "loss": 1.9827,
+      "step": 7050
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.997163925127623e-05,
+      "loss": 2.0274,
+      "step": 7060
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.968803176403855e-05,
+      "loss": 1.9328,
+      "step": 7070
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.94044242768009e-05,
+      "loss": 1.9588,
+      "step": 7080
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.912081678956323e-05,
+      "loss": 1.8266,
+      "step": 7090
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.883720930232557e-05,
+      "loss": 2.0324,
+      "step": 7100
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.85536018150879e-05,
+      "loss": 1.9151,
+      "step": 7110
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.826999432785024e-05,
+      "loss": 1.9188,
+      "step": 7120
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.79863868406126e-05,
+      "loss": 1.8823,
+      "step": 7130
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.770277935337492e-05,
+      "loss": 2.0327,
+      "step": 7140
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.741917186613726e-05,
+      "loss": 1.8703,
+      "step": 7150
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.71355643788996e-05,
+      "loss": 1.9775,
+      "step": 7160
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.685195689166193e-05,
+      "loss": 1.9586,
+      "step": 7170
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.656834940442426e-05,
+      "loss": 1.9808,
+      "step": 7180
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.628474191718661e-05,
+      "loss": 1.9306,
+      "step": 7190
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.600113442994893e-05,
+      "loss": 1.9038,
+      "step": 7200
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.571752694271129e-05,
+      "loss": 1.9689,
+      "step": 7210
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.543391945547362e-05,
+      "loss": 1.9347,
+      "step": 7220
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.515031196823595e-05,
+      "loss": 1.9965,
+      "step": 7230
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.48667044809983e-05,
+      "loss": 1.9164,
+      "step": 7240
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.458309699376062e-05,
+      "loss": 1.9053,
+      "step": 7250
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.429948950652296e-05,
+      "loss": 1.8375,
+      "step": 7260
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.40158820192853e-05,
+      "loss": 1.9083,
+      "step": 7270
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.373227453204764e-05,
+      "loss": 1.953,
+      "step": 7280
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.344866704480999e-05,
+      "loss": 2.0066,
+      "step": 7290
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.316505955757231e-05,
+      "loss": 1.9737,
+      "step": 7300
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.288145207033464e-05,
+      "loss": 1.93,
+      "step": 7310
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.259784458309699e-05,
+      "loss": 1.9436,
+      "step": 7320
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.231423709585933e-05,
+      "loss": 1.9847,
+      "step": 7330
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.203062960862165e-05,
+      "loss": 1.9744,
+      "step": 7340
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.1747022121384e-05,
+      "loss": 1.9623,
+      "step": 7350
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.146341463414633e-05,
+      "loss": 1.83,
+      "step": 7360
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.117980714690867e-05,
+      "loss": 1.741,
+      "step": 7370
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.0896199659671e-05,
+      "loss": 1.8454,
+      "step": 7380
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.061259217243334e-05,
+      "loss": 2.004,
+      "step": 7390
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.032898468519569e-05,
+      "loss": 1.9973,
+      "step": 7400
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.004537719795802e-05,
+      "loss": 1.9251,
+      "step": 7410
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.976176971072036e-05,
+      "loss": 2.0345,
+      "step": 7420
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.94781622234827e-05,
+      "loss": 2.0359,
+      "step": 7430
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.919455473624503e-05,
+      "loss": 1.9897,
+      "step": 7440
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.891094724900736e-05,
+      "loss": 1.9629,
+      "step": 7450
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.862733976176971e-05,
+      "loss": 1.9894,
+      "step": 7460
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.834373227453203e-05,
+      "loss": 2.059,
+      "step": 7470
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.806012478729438e-05,
+      "loss": 2.0083,
+      "step": 7480
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.777651730005672e-05,
+      "loss": 1.9406,
+      "step": 7490
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.749290981281905e-05,
+      "loss": 2.0806,
+      "step": 7500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.72093023255814e-05,
+      "loss": 1.9512,
+      "step": 7510
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.692569483834372e-05,
+      "loss": 1.8797,
+      "step": 7520
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.664208735110606e-05,
+      "loss": 1.9013,
+      "step": 7530
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.63584798638684e-05,
+      "loss": 2.1102,
+      "step": 7540
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.607487237663074e-05,
+      "loss": 1.963,
+      "step": 7550
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.579126488939309e-05,
+      "loss": 1.9602,
+      "step": 7560
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 8.550765740215541e-05,
+      "loss": 1.8007,
+      "step": 7570
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 8.522404991491774e-05,
+      "loss": 1.8766,
+      "step": 7580
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 8.494044242768009e-05,
+      "loss": 1.9586,
+      "step": 7590
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 8.465683494044242e-05,
+      "loss": 1.9787,
+      "step": 7600
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 8.437322745320475e-05,
+      "loss": 1.8037,
+      "step": 7610
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 8.40896199659671e-05,
+      "loss": 1.9972,
+      "step": 7620
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 8.380601247872942e-05,
+      "loss": 1.8995,
+      "step": 7630
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 8.352240499149178e-05,
+      "loss": 1.8151,
+      "step": 7640
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 8.32387975042541e-05,
+      "loss": 1.9566,
+      "step": 7650
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 8.295519001701644e-05,
+      "loss": 1.887,
+      "step": 7660
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 8.267158252977879e-05,
+      "loss": 1.8363,
+      "step": 7670
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 8.238797504254111e-05,
+      "loss": 1.9418,
+      "step": 7680
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 8.210436755530345e-05,
+      "loss": 1.9908,
+      "step": 7690
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 8.182076006806579e-05,
+      "loss": 1.9422,
+      "step": 7700
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 8.153715258082813e-05,
+      "loss": 1.8687,
+      "step": 7710
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 8.125354509359047e-05,
+      "loss": 2.0009,
+      "step": 7720
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 8.09699376063528e-05,
+      "loss": 1.9095,
+      "step": 7730
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 8.068633011911513e-05,
+      "loss": 1.997,
+      "step": 7740
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 8.040272263187748e-05,
+      "loss": 1.8182,
+      "step": 7750
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 8.011911514463982e-05,
+      "loss": 1.9101,
+      "step": 7760
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.983550765740214e-05,
+      "loss": 2.0283,
+      "step": 7770
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 7.95519001701645e-05,
+      "loss": 1.9184,
+      "step": 7780
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 7.926829268292682e-05,
+      "loss": 1.9559,
+      "step": 7790
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 7.898468519568916e-05,
+      "loss": 1.9898,
+      "step": 7800
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 7.87010777084515e-05,
+      "loss": 1.9356,
+      "step": 7810
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 7.841747022121383e-05,
+      "loss": 1.9283,
+      "step": 7820
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 7.813386273397618e-05,
+      "loss": 1.9031,
+      "step": 7830
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 7.785025524673851e-05,
+      "loss": 1.9627,
+      "step": 7840
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 7.756664775950083e-05,
+      "loss": 1.8959,
+      "step": 7850
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 7.728304027226318e-05,
+      "loss": 2.0207,
+      "step": 7860
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 7.699943278502552e-05,
+      "loss": 1.9682,
+      "step": 7870
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 7.671582529778785e-05,
+      "loss": 2.0098,
+      "step": 7880
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 7.64322178105502e-05,
+      "loss": 2.0244,
+      "step": 7890
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 7.614861032331252e-05,
+      "loss": 1.959,
+      "step": 7900
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 7.586500283607487e-05,
+      "loss": 1.9689,
+      "step": 7910
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 7.55813953488372e-05,
+      "loss": 1.8624,
+      "step": 7920
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 7.529778786159954e-05,
+      "loss": 1.8887,
+      "step": 7930
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 7.501418037436189e-05,
+      "loss": 1.9566,
+      "step": 7940
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 7.473057288712421e-05,
+      "loss": 1.9733,
+      "step": 7950
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 7.444696539988655e-05,
+      "loss": 2.0292,
+      "step": 7960
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 7.416335791264889e-05,
+      "loss": 1.8787,
+      "step": 7970
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 7.387975042541123e-05,
+      "loss": 1.9056,
+      "step": 7980
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 7.359614293817356e-05,
+      "loss": 2.0239,
+      "step": 7990
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 7.33125354509359e-05,
+      "loss": 1.8771,
+      "step": 8000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 7.302892796369824e-05,
+      "loss": 1.9212,
+      "step": 8010
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 7.274532047646056e-05,
+      "loss": 1.9028,
+      "step": 8020
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 7.246171298922292e-05,
+      "loss": 1.9409,
+      "step": 8030
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 7.217810550198525e-05,
+      "loss": 1.9232,
+      "step": 8040
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 7.189449801474759e-05,
+      "loss": 1.9214,
+      "step": 8050
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 7.161089052750992e-05,
+      "loss": 1.9733,
+      "step": 8060
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 7.132728304027225e-05,
+      "loss": 1.8845,
+      "step": 8070
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 7.104367555303459e-05,
+      "loss": 1.9782,
+      "step": 8080
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 7.076006806579693e-05,
+      "loss": 1.8242,
+      "step": 8090
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 7.047646057855927e-05,
+      "loss": 1.9213,
+      "step": 8100
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 7.01928530913216e-05,
+      "loss": 1.9098,
+      "step": 8110
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 6.990924560408394e-05,
+      "loss": 1.9736,
+      "step": 8120
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 6.962563811684628e-05,
+      "loss": 1.883,
+      "step": 8130
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 6.934203062960862e-05,
+      "loss": 1.827,
+      "step": 8140
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 6.905842314237096e-05,
+      "loss": 1.8746,
+      "step": 8150
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 6.87748156551333e-05,
+      "loss": 1.8928,
+      "step": 8160
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 6.849120816789562e-05,
+      "loss": 1.9172,
+      "step": 8170
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 6.820760068065796e-05,
+      "loss": 2.0127,
+      "step": 8180
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 6.79239931934203e-05,
+      "loss": 1.9741,
+      "step": 8190
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 6.764038570618265e-05,
+      "loss": 2.0312,
+      "step": 8200
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 6.735677821894497e-05,
+      "loss": 1.953,
+      "step": 8210
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 6.707317073170731e-05,
+      "loss": 2.0159,
+      "step": 8220
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 6.678956324446965e-05,
+      "loss": 1.9546,
+      "step": 8230
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 6.650595575723198e-05,
+      "loss": 1.8581,
+      "step": 8240
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 6.622234826999432e-05,
+      "loss": 1.878,
+      "step": 8250
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 6.593874078275666e-05,
+      "loss": 1.9166,
+      "step": 8260
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 6.5655133295519e-05,
+      "loss": 1.9166,
+      "step": 8270
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 6.537152580828134e-05,
+      "loss": 1.9435,
+      "step": 8280
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 6.508791832104366e-05,
+      "loss": 1.8599,
+      "step": 8290
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 6.480431083380601e-05,
+      "loss": 1.8612,
+      "step": 8300
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 6.452070334656835e-05,
+      "loss": 2.0328,
+      "step": 8310
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 6.423709585933069e-05,
+      "loss": 2.0004,
+      "step": 8320
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 6.395348837209301e-05,
+      "loss": 2.007,
+      "step": 8330
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 6.366988088485535e-05,
+      "loss": 1.9076,
+      "step": 8340
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 6.338627339761769e-05,
+      "loss": 1.893,
+      "step": 8350
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 6.310266591038003e-05,
+      "loss": 2.0122,
+      "step": 8360
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 6.281905842314236e-05,
+      "loss": 1.9318,
+      "step": 8370
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 6.25354509359047e-05,
+      "loss": 1.9334,
+      "step": 8380
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 6.225184344866704e-05,
+      "loss": 1.8622,
+      "step": 8390
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 6.196823596142938e-05,
+      "loss": 1.9893,
+      "step": 8400
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 6.168462847419172e-05,
+      "loss": 1.8939,
+      "step": 8410
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 6.140102098695405e-05,
+      "loss": 1.9368,
+      "step": 8420
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 6.111741349971639e-05,
+      "loss": 1.8463,
+      "step": 8430
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 6.083380601247873e-05,
+      "loss": 1.9361,
+      "step": 8440
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 6.055019852524106e-05,
+      "loss": 1.9151,
+      "step": 8450
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 6.02665910380034e-05,
+      "loss": 1.9659,
+      "step": 8460
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.998298355076574e-05,
+      "loss": 1.9531,
+      "step": 8470
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.9699376063528075e-05,
+      "loss": 1.9611,
+      "step": 8480
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.9415768576290406e-05,
+      "loss": 1.9284,
+      "step": 8490
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.9132161089052744e-05,
+      "loss": 2.0525,
+      "step": 8500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.884855360181508e-05,
+      "loss": 1.9746,
+      "step": 8510
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.856494611457741e-05,
+      "loss": 1.7876,
+      "step": 8520
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.828133862733976e-05,
+      "loss": 1.941,
+      "step": 8530
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.7997731140102096e-05,
+      "loss": 1.925,
+      "step": 8540
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.7714123652864434e-05,
+      "loss": 1.7862,
+      "step": 8550
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.7430516165626765e-05,
+      "loss": 1.8708,
+      "step": 8560
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.71469086783891e-05,
+      "loss": 1.9129,
+      "step": 8570
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.686330119115144e-05,
+      "loss": 2.0183,
+      "step": 8580
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.657969370391378e-05,
+      "loss": 1.965,
+      "step": 8590
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.629608621667611e-05,
+      "loss": 2.0176,
+      "step": 8600
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.601247872943845e-05,
+      "loss": 1.8755,
+      "step": 8610
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.572887124220079e-05,
+      "loss": 1.975,
+      "step": 8620
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.544526375496313e-05,
+      "loss": 1.9415,
+      "step": 8630
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.516165626772546e-05,
+      "loss": 1.992,
+      "step": 8640
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.48780487804878e-05,
+      "loss": 2.0002,
+      "step": 8650
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.459444129325014e-05,
+      "loss": 1.8894,
+      "step": 8660
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.4310833806012475e-05,
+      "loss": 1.8315,
+      "step": 8670
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.4027226318774806e-05,
+      "loss": 1.9616,
+      "step": 8680
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.3743618831537144e-05,
+      "loss": 1.9351,
+      "step": 8690
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.346001134429949e-05,
+      "loss": 1.8639,
+      "step": 8700
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.317640385706183e-05,
+      "loss": 1.9152,
+      "step": 8710
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.289279636982416e-05,
+      "loss": 1.8928,
+      "step": 8720
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.2609188882586496e-05,
+      "loss": 1.9043,
+      "step": 8730
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.2325581395348834e-05,
+      "loss": 1.8675,
+      "step": 8740
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.204197390811117e-05,
+      "loss": 1.9687,
+      "step": 8750
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.17583664208735e-05,
+      "loss": 1.8306,
+      "step": 8760
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.147475893363584e-05,
+      "loss": 1.9105,
+      "step": 8770
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.119115144639818e-05,
+      "loss": 2.0177,
+      "step": 8780
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.0907543959160524e-05,
+      "loss": 1.871,
+      "step": 8790
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.0623936471922855e-05,
+      "loss": 1.8813,
+      "step": 8800
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.034032898468519e-05,
+      "loss": 1.9172,
+      "step": 8810
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.005672149744753e-05,
+      "loss": 1.8479,
+      "step": 8820
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.977311401020987e-05,
+      "loss": 2.0173,
+      "step": 8830
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.94895065229722e-05,
+      "loss": 1.8729,
+      "step": 8840
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.920589903573454e-05,
+      "loss": 1.9665,
+      "step": 8850
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.8922291548496876e-05,
+      "loss": 1.9225,
+      "step": 8860
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.863868406125922e-05,
+      "loss": 1.956,
+      "step": 8870
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.8355076574021545e-05,
+      "loss": 1.9591,
+      "step": 8880
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.807146908678389e-05,
+      "loss": 1.9469,
+      "step": 8890
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.778786159954623e-05,
+      "loss": 1.8405,
+      "step": 8900
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.750425411230856e-05,
+      "loss": 1.8911,
+      "step": 8910
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.7220646625070897e-05,
+      "loss": 1.9739,
+      "step": 8920
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.6937039137833234e-05,
+      "loss": 1.934,
+      "step": 8930
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.665343165059557e-05,
+      "loss": 1.9334,
+      "step": 8940
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.6369824163357904e-05,
+      "loss": 1.9642,
+      "step": 8950
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.608621667612024e-05,
+      "loss": 2.009,
+      "step": 8960
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.5802609188882586e-05,
+      "loss": 1.914,
+      "step": 8970
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.5519001701644924e-05,
+      "loss": 2.0272,
+      "step": 8980
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.5235394214407255e-05,
+      "loss": 1.9943,
+      "step": 8990
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.495178672716959e-05,
+      "loss": 1.8665,
+      "step": 9000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.466817923993193e-05,
+      "loss": 1.989,
+      "step": 9010
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.438457175269427e-05,
+      "loss": 1.776,
+      "step": 9020
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.41009642654566e-05,
+      "loss": 1.7324,
+      "step": 9030
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.381735677821894e-05,
+      "loss": 1.8907,
+      "step": 9040
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.3533749290981276e-05,
+      "loss": 1.8821,
+      "step": 9050
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.325014180374362e-05,
+      "loss": 1.9867,
+      "step": 9060
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.296653431650595e-05,
+      "loss": 1.869,
+      "step": 9070
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.268292682926829e-05,
+      "loss": 1.9415,
+      "step": 9080
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.239931934203063e-05,
+      "loss": 1.906,
+      "step": 9090
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.2115711854792966e-05,
+      "loss": 2.0106,
+      "step": 9100
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.18321043675553e-05,
+      "loss": 1.9134,
+      "step": 9110
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.1548496880317635e-05,
+      "loss": 1.849,
+      "step": 9120
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.126488939307997e-05,
+      "loss": 1.839,
+      "step": 9130
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.098128190584232e-05,
+      "loss": 1.8784,
+      "step": 9140
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.069767441860464e-05,
+      "loss": 2.0755,
+      "step": 9150
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.041406693136699e-05,
+      "loss": 1.909,
+      "step": 9160
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.0130459444129325e-05,
+      "loss": 1.966,
+      "step": 9170
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.984685195689166e-05,
+      "loss": 1.8651,
+      "step": 9180
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.9563244469653994e-05,
+      "loss": 2.1033,
+      "step": 9190
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.927963698241633e-05,
+      "loss": 1.9832,
+      "step": 9200
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.899602949517867e-05,
+      "loss": 1.8741,
+      "step": 9210
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.871242200794101e-05,
+      "loss": 1.8932,
+      "step": 9220
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.842881452070334e-05,
+      "loss": 1.845,
+      "step": 9230
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.814520703346568e-05,
+      "loss": 1.8516,
+      "step": 9240
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.786159954622802e-05,
+      "loss": 1.9229,
+      "step": 9250
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.757799205899035e-05,
+      "loss": 1.9784,
+      "step": 9260
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.729438457175269e-05,
+      "loss": 1.8802,
+      "step": 9270
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.701077708451503e-05,
+      "loss": 1.9196,
+      "step": 9280
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.6727169597277366e-05,
+      "loss": 1.9718,
+      "step": 9290
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.6443562110039704e-05,
+      "loss": 1.9188,
+      "step": 9300
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.6159954622802035e-05,
+      "loss": 2.0001,
+      "step": 9310
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.587634713556437e-05,
+      "loss": 1.8871,
+      "step": 9320
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.559273964832671e-05,
+      "loss": 1.909,
+      "step": 9330
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.530913216108905e-05,
+      "loss": 1.9234,
+      "step": 9340
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.502552467385139e-05,
+      "loss": 1.9902,
+      "step": 9350
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.4741917186613725e-05,
+      "loss": 1.8774,
+      "step": 9360
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.4458309699376056e-05,
+      "loss": 2.0578,
+      "step": 9370
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.41747022121384e-05,
+      "loss": 1.9046,
+      "step": 9380
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.389109472490073e-05,
+      "loss": 1.8949,
+      "step": 9390
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.360748723766307e-05,
+      "loss": 1.8696,
+      "step": 9400
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.332387975042541e-05,
+      "loss": 2.0034,
+      "step": 9410
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.3040272263187746e-05,
+      "loss": 1.9803,
+      "step": 9420
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.2756664775950084e-05,
+      "loss": 1.8447,
+      "step": 9430
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.247305728871242e-05,
+      "loss": 1.871,
+      "step": 9440
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.218944980147475e-05,
+      "loss": 1.8718,
+      "step": 9450
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.19058423142371e-05,
+      "loss": 1.7916,
+      "step": 9460
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.162223482699943e-05,
+      "loss": 1.9947,
+      "step": 9470
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.133862733976177e-05,
+      "loss": 1.981,
+      "step": 9480
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.1055019852524105e-05,
+      "loss": 1.9623,
+      "step": 9490
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.077141236528644e-05,
+      "loss": 1.8504,
+      "step": 9500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.0487804878048777e-05,
+      "loss": 1.909,
+      "step": 9510
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.0204197390811115e-05,
+      "loss": 2.0195,
+      "step": 9520
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.992058990357345e-05,
+      "loss": 1.9324,
+      "step": 9530
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.963698241633579e-05,
+      "loss": 1.9728,
+      "step": 9540
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.9353374929098125e-05,
+      "loss": 1.9669,
+      "step": 9550
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.9069767441860463e-05,
+      "loss": 1.9203,
+      "step": 9560
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.881452070334657e-05,
+      "loss": 1.9111,
+      "step": 9570
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.8530913216108903e-05,
+      "loss": 1.9052,
+      "step": 9580
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.824730572887124e-05,
+      "loss": 1.9317,
+      "step": 9590
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.7963698241633576e-05,
+      "loss": 1.9512,
+      "step": 9600
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.768009075439591e-05,
+      "loss": 1.9274,
+      "step": 9610
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.7396483267158252e-05,
+      "loss": 1.9878,
+      "step": 9620
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.7112875779920586e-05,
+      "loss": 1.8672,
+      "step": 9630
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.6829268292682924e-05,
+      "loss": 1.8437,
+      "step": 9640
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.654566080544526e-05,
+      "loss": 1.9504,
+      "step": 9650
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.62620533182076e-05,
+      "loss": 1.9867,
+      "step": 9660
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.5978445830969935e-05,
+      "loss": 1.9193,
+      "step": 9670
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.5694838343732273e-05,
+      "loss": 1.8887,
+      "step": 9680
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.5411230856494607e-05,
+      "loss": 1.8255,
+      "step": 9690
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.5127623369256945e-05,
+      "loss": 1.8051,
+      "step": 9700
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.4844015882019283e-05,
+      "loss": 1.9183,
+      "step": 9710
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.456040839478162e-05,
+      "loss": 1.9585,
+      "step": 9720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.4276800907543955e-05,
+      "loss": 1.8288,
+      "step": 9730
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.3993193420306293e-05,
+      "loss": 1.8407,
+      "step": 9740
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.370958593306863e-05,
+      "loss": 1.9804,
+      "step": 9750
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.342597844583097e-05,
+      "loss": 1.9133,
+      "step": 9760
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.3142370958593304e-05,
+      "loss": 2.0379,
+      "step": 9770
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.2858763471355642e-05,
+      "loss": 1.9466,
+      "step": 9780
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.2575155984117976e-05,
+      "loss": 2.023,
+      "step": 9790
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.2291548496880318e-05,
+      "loss": 1.8804,
+      "step": 9800
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.2007941009642652e-05,
+      "loss": 1.8823,
+      "step": 9810
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.172433352240499e-05,
+      "loss": 1.8368,
+      "step": 9820
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.1440726035167325e-05,
+      "loss": 1.8944,
+      "step": 9830
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.1157118547929666e-05,
+      "loss": 1.9715,
+      "step": 9840
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.0873511060692e-05,
+      "loss": 1.8169,
+      "step": 9850
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.058990357345434e-05,
+      "loss": 1.926,
+      "step": 9860
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.0306296086216673e-05,
+      "loss": 1.8995,
+      "step": 9870
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.0022688598979014e-05,
+      "loss": 1.964,
+      "step": 9880
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.973908111174135e-05,
+      "loss": 2.0038,
+      "step": 9890
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.9455473624503687e-05,
+      "loss": 1.8676,
+      "step": 9900
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.917186613726602e-05,
+      "loss": 1.8659,
+      "step": 9910
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.888825865002836e-05,
+      "loss": 1.8943,
+      "step": 9920
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.8604651162790697e-05,
+      "loss": 1.9027,
+      "step": 9930
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.8321043675553032e-05,
+      "loss": 1.9053,
+      "step": 9940
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.803743618831537e-05,
+      "loss": 2.1544,
+      "step": 9950
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.7753828701077708e-05,
+      "loss": 1.9002,
+      "step": 9960
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.7470221213840042e-05,
+      "loss": 1.8899,
+      "step": 9970
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.718661372660238e-05,
+      "loss": 1.7576,
+      "step": 9980
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6903006239364718e-05,
+      "loss": 2.0429,
+      "step": 9990
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.6619398752127056e-05,
+      "loss": 1.9995,
+      "step": 10000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.633579126488939e-05,
+      "loss": 1.7999,
+      "step": 10010
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.605218377765173e-05,
+      "loss": 2.0613,
+      "step": 10020
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.5768576290414066e-05,
+      "loss": 1.9488,
+      "step": 10030
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.5484968803176404e-05,
+      "loss": 1.8953,
+      "step": 10040
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.520136131593874e-05,
+      "loss": 1.9878,
+      "step": 10050
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.4917753828701077e-05,
+      "loss": 1.9123,
+      "step": 10060
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.4634146341463413e-05,
+      "loss": 1.9685,
+      "step": 10070
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.4350538854225751e-05,
+      "loss": 1.9945,
+      "step": 10080
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.4066931366988087e-05,
+      "loss": 2.0429,
+      "step": 10090
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.3783323879750425e-05,
+      "loss": 1.9244,
+      "step": 10100
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.3499716392512761e-05,
+      "loss": 1.8477,
+      "step": 10110
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.32161089052751e-05,
+      "loss": 2.0131,
+      "step": 10120
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.2932501418037436e-05,
+      "loss": 1.9577,
+      "step": 10130
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.2648893930799774e-05,
+      "loss": 1.893,
+      "step": 10140
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.236528644356211e-05,
+      "loss": 1.9509,
+      "step": 10150
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.2081678956324448e-05,
+      "loss": 1.8446,
+      "step": 10160
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.1798071469086782e-05,
+      "loss": 1.9316,
+      "step": 10170
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.1514463981849118e-05,
+      "loss": 1.9555,
+      "step": 10180
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.1230856494611456e-05,
+      "loss": 1.944,
+      "step": 10190
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0947249007373793e-05,
+      "loss": 2.0048,
+      "step": 10200
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.066364152013613e-05,
+      "loss": 1.9539,
+      "step": 10210
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.0380034032898467e-05,
+      "loss": 1.8775,
+      "step": 10220
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.0096426545660803e-05,
+      "loss": 1.8658,
+      "step": 10230
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 9.812819058423141e-06,
+      "loss": 1.8844,
+      "step": 10240
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 9.529211571185477e-06,
+      "loss": 1.8888,
+      "step": 10250
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 9.245604083947815e-06,
+      "loss": 1.93,
+      "step": 10260
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 8.961996596710151e-06,
+      "loss": 1.9094,
+      "step": 10270
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 8.67838910947249e-06,
+      "loss": 1.9298,
+      "step": 10280
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 8.394781622234826e-06,
+      "loss": 1.9718,
+      "step": 10290
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 8.111174134997164e-06,
+      "loss": 1.9241,
+      "step": 10300
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 7.8275666477595e-06,
+      "loss": 1.9156,
+      "step": 10310
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 7.543959160521837e-06,
+      "loss": 1.9464,
+      "step": 10320
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 7.260351673284174e-06,
+      "loss": 1.8982,
+      "step": 10330
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 6.976744186046511e-06,
+      "loss": 2.0217,
+      "step": 10340
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 6.693136698808848e-06,
+      "loss": 1.9342,
+      "step": 10350
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 6.409529211571185e-06,
+      "loss": 1.8258,
+      "step": 10360
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 6.125921724333522e-06,
+      "loss": 1.9249,
+      "step": 10370
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.8423142370958585e-06,
+      "loss": 1.8313,
+      "step": 10380
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.558706749858196e-06,
+      "loss": 1.917,
+      "step": 10390
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.275099262620533e-06,
+      "loss": 1.8811,
+      "step": 10400
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.99149177538287e-06,
+      "loss": 1.8812,
+      "step": 10410
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.707884288145207e-06,
+      "loss": 1.9099,
+      "step": 10420
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.424276800907543e-06,
+      "loss": 1.8592,
+      "step": 10430
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.14066931366988e-06,
+      "loss": 1.738,
+      "step": 10440
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.857061826432217e-06,
+      "loss": 1.8415,
+      "step": 10450
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.5734543391945544e-06,
+      "loss": 1.9281,
+      "step": 10460
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.2898468519568915e-06,
+      "loss": 1.9303,
+      "step": 10470
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.0062393647192286e-06,
+      "loss": 1.9314,
+      "step": 10480
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.7226318774815657e-06,
+      "loss": 1.8479,
+      "step": 10490
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.4390243902439023e-06,
+      "loss": 1.8967,
+      "step": 10500
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 10578,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "total_flos": 4.3967525098082304e+17,
+  "trial_name": null,
+  "trial_params": null
+}