diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5149 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 11.998420151840962,
+  "global_step": 8544,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.945525291828794e-05,
+      "loss": 0.8062,
+      "step": 10
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.891050583657588e-05,
+      "loss": 0.7667,
+      "step": 20
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.836575875486382e-05,
+      "loss": 0.7266,
+      "step": 30
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 7.782101167315176e-05,
+      "loss": 0.7113,
+      "step": 40
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.72762645914397e-05,
+      "loss": 0.7038,
+      "step": 50
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00011673151750972763,
+      "loss": 0.6806,
+      "step": 60
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00013618677042801555,
+      "loss": 0.6916,
+      "step": 70
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001556420233463035,
+      "loss": 0.6688,
+      "step": 80
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017509727626459142,
+      "loss": 0.6723,
+      "step": 90
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001945525291828794,
+      "loss": 0.6821,
+      "step": 100
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002140077821011673,
+      "loss": 0.6703,
+      "step": 110
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00023346303501945527,
+      "loss": 0.6714,
+      "step": 120
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002529182879377432,
+      "loss": 0.6746,
+      "step": 130
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002723735408560311,
+      "loss": 0.6678,
+      "step": 140
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002918287937743191,
+      "loss": 0.6683,
+      "step": 150
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.000311284046692607,
+      "loss": 0.6742,
+      "step": 160
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00033073929961089494,
+      "loss": 0.6661,
+      "step": 170
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00035019455252918285,
+      "loss": 0.6541,
+      "step": 180
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00036964980544747087,
+      "loss": 0.6625,
+      "step": 190
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0003891050583657588,
+      "loss": 0.6682,
+      "step": 200
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0004085603112840467,
+      "loss": 0.6692,
+      "step": 210
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0004280155642023346,
+      "loss": 0.6612,
+      "step": 220
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00044747081712062257,
+      "loss": 0.6587,
+      "step": 230
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00046692607003891054,
+      "loss": 0.6478,
+      "step": 240
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00048638132295719845,
+      "loss": 0.6618,
+      "step": 250
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0004999998383195374,
+      "loss": 0.655,
+      "step": 260
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0004999969640060184,
+      "loss": 0.6486,
+      "step": 270
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0004999904968408763,
+      "loss": 0.6403,
+      "step": 280
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0004999804369170547,
+      "loss": 0.637,
+      "step": 290
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0004999667843791306,
+      "loss": 0.6543,
+      "step": 300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0004999495394233122,
+      "loss": 0.6542,
+      "step": 310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0004999287022974372,
+      "loss": 0.6454,
+      "step": 320
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0004999042733009678,
+      "loss": 0.637,
+      "step": 330
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0004998762527849873,
+      "loss": 0.6267,
+      "step": 340
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.000499844641152195,
+      "loss": 0.6377,
+      "step": 350
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0004998094388568999,
+      "loss": 0.6257,
+      "step": 360
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0004997706464050146,
+      "loss": 0.6375,
+      "step": 370
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.000499728264354048,
+      "loss": 0.6427,
+      "step": 380
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0004996822933130973,
+      "loss": 0.6452,
+      "step": 390
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.000499632733942839,
+      "loss": 0.6368,
+      "step": 400
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0004995795869555194,
+      "loss": 0.6274,
+      "step": 410
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0004995228531149451,
+      "loss": 0.6251,
+      "step": 420
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0004994625332364707,
+      "loss": 0.6204,
+      "step": 430
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0004993986281869884,
+      "loss": 0.6459,
+      "step": 440
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0004993380490792496,
+      "loss": 0.64,
+      "step": 450
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0004992745674314514,
+      "loss": 0.6552,
+      "step": 460
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0004992006289535387,
+      "loss": 0.6474,
+      "step": 470
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0004991231090685967,
+      "loss": 0.6341,
+      "step": 480
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0004990420088907088,
+      "loss": 0.6337,
+      "step": 490
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0004989573295854126,
+      "loss": 0.6246,
+      "step": 500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0004988690723696838,
+      "loss": 0.6198,
+      "step": 510
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.000498777238511918,
+      "loss": 0.6253,
+      "step": 520
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0004986818293319129,
+      "loss": 0.6219,
+      "step": 530
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0004985828462008493,
+      "loss": 0.6228,
+      "step": 540
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0004984802905412714,
+      "loss": 0.6172,
+      "step": 550
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0004983741638270661,
+      "loss": 0.62,
+      "step": 560
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0004982644675834418,
+      "loss": 0.6169,
+      "step": 570
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0004981512033869074,
+      "loss": 0.6183,
+      "step": 580
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0004980343728652482,
+      "loss": 0.6173,
+      "step": 590
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0004979139776975036,
+      "loss": 0.6206,
+      "step": 600
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0004977900196139425,
+      "loss": 0.6043,
+      "step": 610
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0004976625003960385,
+      "loss": 0.6191,
+      "step": 620
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0004975314218764444,
+      "loss": 0.5894,
+      "step": 630
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0004973967859389658,
+      "loss": 0.6131,
+      "step": 640
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0004972585945185338,
+      "loss": 0.6016,
+      "step": 650
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.000497116849601178,
+      "loss": 0.6061,
+      "step": 660
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0004969715532239965,
+      "loss": 0.6133,
+      "step": 670
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0004968227074751282,
+      "loss": 0.6112,
+      "step": 680
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0004966703144937218,
+      "loss": 0.6024,
+      "step": 690
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0004965143764699055,
+      "loss": 0.5901,
+      "step": 700
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.000496354895644755,
+      "loss": 0.6036,
+      "step": 710
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.000496191874310262,
+      "loss": 0.5658,
+      "step": 720
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.000496025314809301,
+      "loss": 0.5662,
+      "step": 730
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0004958552195355951,
+      "loss": 0.5543,
+      "step": 740
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.0004956815909336824,
+      "loss": 0.5503,
+      "step": 750
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0004955044314988805,
+      "loss": 0.5604,
+      "step": 760
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.0004953237437772504,
+      "loss": 0.5584,
+      "step": 770
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0004951395303655605,
+      "loss": 0.5646,
+      "step": 780
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.0004949517939112487,
+      "loss": 0.5523,
+      "step": 790
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.0004947605371123845,
+      "loss": 0.5587,
+      "step": 800
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.0004945657627176309,
+      "loss": 0.5484,
+      "step": 810
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0004943674735262035,
+      "loss": 0.5641,
+      "step": 820
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0004941656723878317,
+      "loss": 0.5566,
+      "step": 830
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.0004939603622027168,
+      "loss": 0.5504,
+      "step": 840
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.0004937515459214908,
+      "loss": 0.5445,
+      "step": 850
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.0004935392265451738,
+      "loss": 0.5586,
+      "step": 860
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.0004933234071251309,
+      "loss": 0.5489,
+      "step": 870
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0004931040907630283,
+      "loss": 0.5561,
+      "step": 880
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.0004928812806107891,
+      "loss": 0.5565,
+      "step": 890
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0004926549798705473,
+      "loss": 0.5568,
+      "step": 900
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.0004924251917946023,
+      "loss": 0.5564,
+      "step": 910
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.0004921919196853721,
+      "loss": 0.5536,
+      "step": 920
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.0004919551668953457,
+      "loss": 0.5551,
+      "step": 930
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.0004917149368270351,
+      "loss": 0.5458,
+      "step": 940
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.000491471232932926,
+      "loss": 0.5517,
+      "step": 950
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.0004912240587154289,
+      "loss": 0.5452,
+      "step": 960
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.0004909734177268278,
+      "loss": 0.5442,
+      "step": 970
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.0004907193135692302,
+      "loss": 0.5438,
+      "step": 980
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.0004904617498945145,
+      "loss": 0.5458,
+      "step": 990
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.000490200730404278,
+      "loss": 0.5545,
+      "step": 1000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.0004899362588497832,
+      "loss": 0.5585,
+      "step": 1010
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.0004896683390319048,
+      "loss": 0.5505,
+      "step": 1020
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.0004893969748010741,
+      "loss": 0.5469,
+      "step": 1030
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.000489122170057224,
+      "loss": 0.5524,
+      "step": 1040
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.0004888439287497331,
+      "loss": 0.5439,
+      "step": 1050
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.0004885622548773692,
+      "loss": 0.5454,
+      "step": 1060
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00048827715248823077,
+      "loss": 0.5442,
+      "step": 1070
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00048798862567969026,
+      "loss": 0.5419,
+      "step": 1080
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00048769667859833386,
+      "loss": 0.5465,
+      "step": 1090
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00048740131543990276,
+      "loss": 0.5444,
+      "step": 1100
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00048710254044923246,
+      "loss": 0.5507,
+      "step": 1110
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00048680035792019203,
+      "loss": 0.5365,
+      "step": 1120
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.0004864947721956222,
+      "loss": 0.5475,
+      "step": 1130
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00048618578766727286,
+      "loss": 0.5472,
+      "step": 1140
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00048587340877574026,
+      "loss": 0.5449,
+      "step": 1150
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.0004855576400104029,
+      "loss": 0.5371,
+      "step": 1160
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00048523848590935727,
+      "loss": 0.5453,
+      "step": 1170
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00048491595105935225,
+      "loss": 0.5423,
+      "step": 1180
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.0004845900400957237,
+      "loss": 0.5349,
+      "step": 1190
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00048426075770232745,
+      "loss": 0.537,
+      "step": 1200
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.000483928108611472,
+      "loss": 0.5417,
+      "step": 1210
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.0004835920976038507,
+      "loss": 0.5301,
+      "step": 1220
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.000483252729508473,
+      "loss": 0.5326,
+      "step": 1230
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00048291000920259504,
+      "loss": 0.5474,
+      "step": 1240
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.0004825639416116494,
+      "loss": 0.5254,
+      "step": 1250
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00048221453170917453,
+      "loss": 0.5426,
+      "step": 1260
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.0004818617845167431,
+      "loss": 0.5358,
+      "step": 1270
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00048150570510389,
+      "loss": 0.5376,
+      "step": 1280
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00048114629858803944,
+      "loss": 0.5329,
+      "step": 1290
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.00048078357013443113,
+      "loss": 0.5261,
+      "step": 1300
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.0004804175249560466,
+      "loss": 0.5461,
+      "step": 1310
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.0004800481683135336,
+      "loss": 0.5388,
+      "step": 1320
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.00047967550551513116,
+      "loss": 0.5309,
+      "step": 1330
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.00047929954191659285,
+      "loss": 0.5364,
+      "step": 1340
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.00047892028292111007,
+      "loss": 0.5397,
+      "step": 1350
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.00047853773397923407,
+      "loss": 0.5407,
+      "step": 1360
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.00047815190058879804,
+      "loss": 0.5352,
+      "step": 1370
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.0004777627882948378,
+      "loss": 0.5279,
+      "step": 1380
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.0004773704026895121,
+      "loss": 0.5366,
+      "step": 1390
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.0004769747494120227,
+      "loss": 0.5269,
+      "step": 1400
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.0004765758341485325,
+      "loss": 0.5367,
+      "step": 1410
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.00047617366263208466,
+      "loss": 0.5274,
+      "step": 1420
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 0.0004757682406425197,
+      "loss": 0.496,
+      "step": 1430
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 0.0004753595740063925,
+      "loss": 0.4646,
+      "step": 1440
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 0.0004749476685968888,
+      "loss": 0.4699,
+      "step": 1450
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 0.00047453253033374063,
+      "loss": 0.4638,
+      "step": 1460
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 0.00047411416518314096,
+      "loss": 0.4657,
+      "step": 1470
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 0.00047369257915765874,
+      "loss": 0.4587,
+      "step": 1480
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 0.0004732677783161515,
+      "loss": 0.4669,
+      "step": 1490
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 0.00047283976876367904,
+      "loss": 0.47,
+      "step": 1500
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 0.0004724085566514152,
+      "loss": 0.4613,
+      "step": 1510
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 0.0004719741481765599,
+      "loss": 0.4698,
+      "step": 1520
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 0.00047153654958224966,
+      "loss": 0.4546,
+      "step": 1530
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 0.0004710957671574682,
+      "loss": 0.4765,
+      "step": 1540
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 0.00047065180723695576,
+      "loss": 0.4656,
+      "step": 1550
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 0.00047020467620111843,
+      "loss": 0.4728,
+      "step": 1560
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.00046975438047593604,
+      "loss": 0.464,
+      "step": 1570
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 0.00046930092653287014,
+      "loss": 0.466,
+      "step": 1580
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 0.00046884432088877074,
+      "loss": 0.4784,
+      "step": 1590
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 0.00046838457010578286,
+      "loss": 0.4693,
+      "step": 1600
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 0.0004679216807912522,
+      "loss": 0.4808,
+      "step": 1610
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 0.0004674556595976299,
+      "loss": 0.4638,
+      "step": 1620
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 0.00046698651322237743,
+      "loss": 0.4725,
+      "step": 1630
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 0.0004665142484078699,
+      "loss": 0.4658,
+      "step": 1640
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 0.00046603887194129944,
+      "loss": 0.4687,
+      "step": 1650
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 0.0004655603906545773,
+      "loss": 0.4704,
+      "step": 1660
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 0.00046507881142423613,
+      "loss": 0.468,
+      "step": 1670
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 0.00046459414117133084,
+      "loss": 0.4721,
+      "step": 1680
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 0.00046410638686133934,
+      "loss": 0.4665,
+      "step": 1690
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 0.000463615555504062,
+      "loss": 0.4717,
+      "step": 1700
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 0.00046312165415352165,
+      "loss": 0.4736,
+      "step": 1710
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 0.00046262468990786144,
+      "loss": 0.4689,
+      "step": 1720
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 0.0004621246699092433,
+      "loss": 0.4761,
+      "step": 1730
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 0.00046162160134374526,
+      "loss": 0.4667,
+      "step": 1740
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 0.00046111549144125787,
+      "loss": 0.4669,
+      "step": 1750
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 0.00046060634747538057,
+      "loss": 0.4708,
+      "step": 1760
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 0.00046009417676331723,
+      "loss": 0.4598,
+      "step": 1770
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 0.0004595789866657706,
+      "loss": 0.4739,
+      "step": 1780
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 0.00045906078458683697,
+      "loss": 0.4529,
+      "step": 1790
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 0.0004585395779738996,
+      "loss": 0.4738,
+      "step": 1800
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 0.0004580153743175214,
+      "loss": 0.4677,
+      "step": 1810
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 0.0004574881811513377,
+      "loss": 0.4626,
+      "step": 1820
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 0.00045695800605194794,
+      "loss": 0.4716,
+      "step": 1830
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 0.0004564248566388064,
+      "loss": 0.4602,
+      "step": 1840
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 0.0004558887405741131,
+      "loss": 0.4652,
+      "step": 1850
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 0.0004553496655627035,
+      "loss": 0.4686,
+      "step": 1860
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 0.0004548076393519378,
+      "loss": 0.4698,
+      "step": 1870
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 0.0004542626697315896,
+      "loss": 0.4675,
+      "step": 1880
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 0.000453714764533734,
+      "loss": 0.4771,
+      "step": 1890
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 0.0004531639316326349,
+      "loss": 0.4577,
+      "step": 1900
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 0.00045261017894463184,
+      "loss": 0.4562,
+      "step": 1910
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 0.00045205351442802666,
+      "loss": 0.4601,
+      "step": 1920
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 0.00045149394608296845,
+      "loss": 0.4699,
+      "step": 1930
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 0.00045093148195133904,
+      "loss": 0.4583,
+      "step": 1940
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 0.0004503661301166373,
+      "loss": 0.4742,
+      "step": 1950
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 0.00044979789870386284,
+      "loss": 0.4773,
+      "step": 1960
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 0.00044922679587939955,
+      "loss": 0.4689,
+      "step": 1970
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 0.00044865282985089776,
+      "loss": 0.4537,
+      "step": 1980
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 0.000448076008867157,
+      "loss": 0.4654,
+      "step": 1990
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 0.0004474963412180066,
+      "loss": 0.4701,
+      "step": 2000
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 0.0004469138352341872,
+      "loss": 0.4782,
+      "step": 2010
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 0.00044632849928723077,
+      "loss": 0.4762,
+      "step": 2020
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 0.00044574034178934024,
+      "loss": 0.4652,
+      "step": 2030
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 0.0004451493711932688,
+      "loss": 0.4604,
+      "step": 2040
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 0.0004445555959921982,
+      "loss": 0.4661,
+      "step": 2050
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 0.0004439590247196169,
+      "loss": 0.4658,
+      "step": 2060
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 0.0004433596659491972,
+      "loss": 0.4617,
+      "step": 2070
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 0.0004427575282946722,
+      "loss": 0.4715,
+      "step": 2080
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 0.00044215262040971223,
+      "loss": 0.4598,
+      "step": 2090
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 0.00044154495098779967,
+      "loss": 0.4618,
+      "step": 2100
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 0.00044093452876210504,
+      "loss": 0.4634,
+      "step": 2110
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 0.0004403213625053608,
+      "loss": 0.4609,
+      "step": 2120
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 0.0004397054610297354,
+      "loss": 0.4652,
+      "step": 2130
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 0.00043908683318670685,
+      "loss": 0.4428,
+      "step": 2140
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 0.0004384654878669353,
+      "loss": 0.3926,
+      "step": 2150
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 0.0004378414340001353,
+      "loss": 0.3898,
+      "step": 2160
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 0.00043721468055494774,
+      "loss": 0.3887,
+      "step": 2170
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 0.00043658523653881037,
+      "loss": 0.3818,
+      "step": 2180
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 0.0004359531109978289,
+      "loss": 0.3916,
+      "step": 2190
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 0.0004353183130166468,
+      "loss": 0.389,
+      "step": 2200
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 0.00043468085171831464,
+      "loss": 0.3886,
+      "step": 2210
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 0.0004340407362641591,
+      "loss": 0.3908,
+      "step": 2220
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 0.0004333979758536514,
+      "loss": 0.3886,
+      "step": 2230
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 0.0004327525797242747,
+      "loss": 0.3861,
+      "step": 2240
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 0.0004321045571513919,
+      "loss": 0.395,
+      "step": 2250
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 0.00043145391744811183,
+      "loss": 0.3967,
+      "step": 2260
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 0.0004308006699651558,
+      "loss": 0.396,
+      "step": 2270
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 0.0004301448240907228,
+      "loss": 0.39,
+      "step": 2280
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 0.000429486389250355,
+      "loss": 0.3862,
+      "step": 2290
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 0.0004288253749068021,
+      "loss": 0.3918,
+      "step": 2300
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 0.0004281617905598852,
+      "loss": 0.3885,
+      "step": 2310
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 0.0004274956457463607,
+      "loss": 0.3959,
+      "step": 2320
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 0.0004268269500397825,
+      "loss": 0.3933,
+      "step": 2330
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 0.0004261557130503653,
+      "loss": 0.4,
+      "step": 2340
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 0.00042548194442484585,
+      "loss": 0.3958,
+      "step": 2350
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 0.0004248056538463445,
+      "loss": 0.3977,
+      "step": 2360
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 0.0004241268510342261,
+      "loss": 0.3952,
+      "step": 2370
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 0.0004234455457439603,
+      "loss": 0.3894,
+      "step": 2380
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 0.00042276174776698127,
+      "loss": 0.4068,
+      "step": 2390
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 0.000422075466930547,
+      "loss": 0.3924,
+      "step": 2400
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 0.0004213867130975981,
+      "loss": 0.3979,
+      "step": 2410
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 0.00042069549616661616,
+      "loss": 0.3994,
+      "step": 2420
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 0.00042000182607148127,
+      "loss": 0.4007,
+      "step": 2430
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 0.0004193057127813295,
+      "loss": 0.4062,
+      "step": 2440
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 0.0004186071663004093,
+      "loss": 0.4013,
+      "step": 2450
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 0.00041790619666793796,
+      "loss": 0.3906,
+      "step": 2460
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 0.00041720281395795744,
+      "loss": 0.3982,
+      "step": 2470
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 0.0004164970282791893,
+      "loss": 0.3992,
+      "step": 2480
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 0.0004157888497748897,
+      "loss": 0.3982,
+      "step": 2490
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 0.0004150782886227033,
+      "loss": 0.3871,
+      "step": 2500
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 0.0004143653550345173,
+      "loss": 0.3906,
+      "step": 2510
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 0.0004136500592563144,
+      "loss": 0.387,
+      "step": 2520
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 0.0004129324115680261,
+      "loss": 0.3888,
+      "step": 2530
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 0.000412212422283384,
+      "loss": 0.402,
+      "step": 2540
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 0.00041149010174977265,
+      "loss": 0.4053,
+      "step": 2550
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 0.00041076546034808,
+      "loss": 0.3949,
+      "step": 2560
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 0.00041003850849254873,
+      "loss": 0.4027,
+      "step": 2570
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 0.00040930925663062637,
+      "loss": 0.3876,
+      "step": 2580
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 0.00040857771524281504,
+      "loss": 0.3973,
+      "step": 2590
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 0.00040784389484252106,
+      "loss": 0.397,
+      "step": 2600
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 0.00040710780597590383,
+      "loss": 0.4024,
+      "step": 2610
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 0.0004063694592217242,
+      "loss": 0.3917,
+      "step": 2620
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 0.0004056288651911922,
+      "loss": 0.3939,
+      "step": 2630
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 0.00040488603452781504,
+      "loss": 0.4034,
+      "step": 2640
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 0.0004041409779072437,
+      "loss": 0.3956,
+      "step": 2650
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 0.00040339370603712,
+      "loss": 0.4004,
+      "step": 2660
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 0.0004026442296569218,
+      "loss": 0.4044,
+      "step": 2670
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 0.00040189255953781,
+      "loss": 0.4059,
+      "step": 2680
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 0.00040113870648247234,
+      "loss": 0.3941,
+      "step": 2690
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 0.0004003826813249692,
+      "loss": 0.3991,
+      "step": 2700
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 0.00039962449493057743,
+      "loss": 0.4032,
+      "step": 2710
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 0.00039886415819563413,
+      "loss": 0.3945,
+      "step": 2720
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 0.0003981016820473803,
+      "loss": 0.4034,
+      "step": 2730
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 0.0003973370774438037,
+      "loss": 0.3968,
+      "step": 2740
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 0.0003965703553734814,
+      "loss": 0.4027,
+      "step": 2750
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 0.0003958015268554216,
+      "loss": 0.389,
+      "step": 2760
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 0.0003950306029389058,
+      "loss": 0.4024,
+      "step": 2770
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 0.00039425759470332946,
+      "loss": 0.3952,
+      "step": 2780
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 0.00039348251325804296,
+      "loss": 0.4001,
+      "step": 2790
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 0.0003927053697421923,
+      "loss": 0.3963,
+      "step": 2800
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 0.00039192617532455843,
+      "loss": 0.3985,
+      "step": 2810
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 0.0003911449412033969,
+      "loss": 0.3846,
+      "step": 2820
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 0.0003903616786062772,
+      "loss": 0.3895,
+      "step": 2830
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 0.0003895763987899215,
+      "loss": 0.4002,
+      "step": 2840
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 0.00038878911304004196,
+      "loss": 0.387,
+      "step": 2850
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 0.00038799983267117955,
+      "loss": 0.3133,
+      "step": 2860
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 0.00038720856902654127,
+      "loss": 0.3171,
+      "step": 2870
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 0.0003864153334778365,
+      "loss": 0.329,
+      "step": 2880
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 0.00038562013742511435,
+      "loss": 0.317,
+      "step": 2890
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 0.0003848229922965991,
+      "loss": 0.3092,
+      "step": 2900
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 0.0003840239095485268,
+      "loss": 0.3162,
+      "step": 2910
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 0.00038322290066497973,
+      "loss": 0.3189,
+      "step": 2920
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 0.00038241997715772203,
+      "loss": 0.3207,
+      "step": 2930
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 0.00038161515056603384,
+      "loss": 0.3137,
+      "step": 2940
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 0.0003808084324565459,
+      "loss": 0.3253,
+      "step": 2950
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 0.00037999983442307274,
+      "loss": 0.3245,
+      "step": 2960
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 0.00037918936808644657,
+      "loss": 0.3193,
+      "step": 2970
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 0.00037837704509435,
+      "loss": 0.3195,
+      "step": 2980
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 0.0003775628771211487,
+      "loss": 0.323,
+      "step": 2990
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 0.00037674687586772374,
+      "loss": 0.3229,
+      "step": 3000
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 0.0003759290530613031,
+      "loss": 0.3227,
+      "step": 3010
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 0.0003751094204552938,
+      "loss": 0.3239,
+      "step": 3020
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 0.00037428798982911185,
+      "loss": 0.329,
+      "step": 3030
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 0.0003734647729880143,
+      "loss": 0.3284,
+      "step": 3040
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 0.00037263978176292856,
+      "loss": 0.3248,
+      "step": 3050
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 0.0003718130280102829,
+      "loss": 0.3232,
+      "step": 3060
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 0.00037098452361183606,
+      "loss": 0.3311,
+      "step": 3070
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 0.0003701542804745058,
+      "loss": 0.3277,
+      "step": 3080
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 0.00036932231053019883,
+      "loss": 0.3333,
+      "step": 3090
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 0.0003684886257356385,
+      "loss": 0.3322,
+      "step": 3100
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 0.00036765323807219343,
+      "loss": 0.3282,
+      "step": 3110
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 0.0003668161595457052,
+      "loss": 0.3253,
+      "step": 3120
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 0.0003659774021863156,
+      "loss": 0.3221,
+      "step": 3130
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 0.0003651369780482939,
+      "loss": 0.3301,
+      "step": 3140
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 0.00036429489920986383,
+      "loss": 0.3222,
+      "step": 3150
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 0.00036345117777302947,
+      "loss": 0.3312,
+      "step": 3160
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 0.00036260582586340197,
+      "loss": 0.3307,
+      "step": 3170
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 0.00036175885563002473,
+      "loss": 0.3202,
+      "step": 3180
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 0.0003609102792451991,
+      "loss": 0.3254,
+      "step": 3190
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 0.0003600601089043093,
+      "loss": 0.3298,
+      "step": 3200
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 0.0003592083568256473,
+      "loss": 0.3334,
+      "step": 3210
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 0.0003583550352502371,
+      "loss": 0.3338,
+      "step": 3220
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 0.0003575001564416588,
+      "loss": 0.3247,
+      "step": 3230
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 0.0003566437326858727,
+      "loss": 0.3287,
+      "step": 3240
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 0.0003557857762910418,
+      "loss": 0.3255,
+      "step": 3250
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 0.0003549262995873562,
+      "loss": 0.3316,
+      "step": 3260
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 0.000354065314926855,
+      "loss": 0.3295,
+      "step": 3270
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 0.00035320283468324895,
+      "loss": 0.325,
+      "step": 3280
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 0.0003523388712517427,
+      "loss": 0.332,
+      "step": 3290
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 0.00035147343704885673,
+      "loss": 0.3256,
+      "step": 3300
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 0.0003506065445122488,
+      "loss": 0.3388,
+      "step": 3310
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 0.0003497382061005353,
+      "loss": 0.3316,
+      "step": 3320
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 0.000348868434293112,
+      "loss": 0.3307,
+      "step": 3330
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 0.0003479972415899749,
+      "loss": 0.3302,
+      "step": 3340
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 0.0003471246405115406,
+      "loss": 0.3213,
+      "step": 3350
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 0.00034625064359846607,
+      "loss": 0.3293,
+      "step": 3360
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 0.00034537526341146883,
+      "loss": 0.3276,
+      "step": 3370
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 0.00034449851253114615,
+      "loss": 0.3268,
+      "step": 3380
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 0.0003436204035577943,
+      "loss": 0.3293,
+      "step": 3390
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 0.00034274094911122767,
+      "loss": 0.3302,
+      "step": 3400
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 0.000341860161830597,
+      "loss": 0.3276,
+      "step": 3410
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 0.00034097805437420796,
+      "loss": 0.3329,
+      "step": 3420
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 0.0003400946394193395,
+      "loss": 0.3262,
+      "step": 3430
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 0.0003392099296620611,
+      "loss": 0.3306,
+      "step": 3440
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 0.0003383239378170507,
+      "loss": 0.3291,
+      "step": 3450
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 0.00033743667661741215,
+      "loss": 0.3299,
+      "step": 3460
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 0.0003365481588144913,
+      "loss": 0.328,
+      "step": 3470
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 0.00033565839717769396,
+      "loss": 0.324,
+      "step": 3480
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 0.0003347674044943014,
+      "loss": 0.3295,
+      "step": 3490
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 0.0003338751935692874,
+      "loss": 0.3354,
+      "step": 3500
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 0.00033298177722513347,
+      "loss": 0.3315,
+      "step": 3510
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 0.0003320871683016448,
+      "loss": 0.3361,
+      "step": 3520
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 0.0003311913796557663,
+      "loss": 0.3293,
+      "step": 3530
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 0.0003302944241613969,
+      "loss": 0.3362,
+      "step": 3540
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 0.0003293963147092052,
+      "loss": 0.3312,
+      "step": 3550
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 0.0003284970642064439,
+      "loss": 0.3275,
+      "step": 3560
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 0.00032759668557676456,
+      "loss": 0.2653,
+      "step": 3570
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 0.0003266951917600316,
+      "loss": 0.2626,
+      "step": 3580
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 0.0003257925957121363,
+      "loss": 0.2521,
+      "step": 3590
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 0.00032488891040481107,
+      "loss": 0.2522,
+      "step": 3600
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 0.00032398414882544235,
+      "loss": 0.2577,
+      "step": 3610
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 0.0003230783239768846,
+      "loss": 0.2541,
+      "step": 3620
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 0.00032217144887727314,
+      "loss": 0.2527,
+      "step": 3630
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 0.0003212635365598366,
+      "loss": 0.2556,
+      "step": 3640
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 0.0003203546000727108,
+      "loss": 0.2495,
+      "step": 3650
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 0.00031944465247874974,
+      "loss": 0.2618,
+      "step": 3660
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 0.00031853370685533926,
+      "loss": 0.2569,
+      "step": 3670
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 0.00031762177629420814,
+      "loss": 0.2578,
+      "step": 3680
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 0.0003167088739012402,
+      "loss": 0.2558,
+      "step": 3690
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 0.00031579501279628623,
+      "loss": 0.2611,
+      "step": 3700
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 0.0003148802061129751,
+      "loss": 0.2531,
+      "step": 3710
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 0.0003139644669985253,
+      "loss": 0.2634,
+      "step": 3720
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 0.00031304780861355556,
+      "loss": 0.2577,
+      "step": 3730
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 0.00031213024413189616,
+      "loss": 0.2567,
+      "step": 3740
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 0.00031121178674039954,
+      "loss": 0.2599,
+      "step": 3750
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 0.00031029244963875027,
+      "loss": 0.2716,
+      "step": 3760
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 0.00030937224603927617,
+      "loss": 0.25,
+      "step": 3770
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 0.00030845118916675773,
+      "loss": 0.2614,
+      "step": 3780
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 0.00030752929225823846,
+      "loss": 0.2653,
+      "step": 3790
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 0.0003066065685628345,
+      "loss": 0.2608,
+      "step": 3800
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 0.0003056830313415441,
+      "loss": 0.2613,
+      "step": 3810
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 0.0003047586938670574,
+      "loss": 0.2528,
+      "step": 3820
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 0.00030383356942356515,
+      "loss": 0.2631,
+      "step": 3830
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 0.00030290767130656843,
+      "loss": 0.2642,
+      "step": 3840
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 0.00030198101282268686,
+      "loss": 0.2584,
+      "step": 3850
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 0.000301053607289468,
+      "loss": 0.259,
+      "step": 3860
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 0.00030012546803519566,
+      "loss": 0.266,
+      "step": 3870
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 0.00029919660839869823,
+      "loss": 0.2623,
+      "step": 3880
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 0.0002982670417291573,
+      "loss": 0.2626,
+      "step": 3890
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 0.0002973367813859156,
+      "loss": 0.269,
+      "step": 3900
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 0.00029640584073828484,
+      "loss": 0.263,
+      "step": 3910
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 0.00029547423316535426,
+      "loss": 0.2678,
+      "step": 3920
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 0.0002945419720557973,
+      "loss": 0.2647,
+      "step": 3930
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 0.00029360907080768015,
+      "loss": 0.2673,
+      "step": 3940
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 0.00029267554282826877,
+      "loss": 0.2605,
+      "step": 3950
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 0.00029174140153383595,
+      "loss": 0.2568,
+      "step": 3960
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 0.0002908066603494691,
+      "loss": 0.2571,
+      "step": 3970
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 0.00028987133270887686,
+      "loss": 0.2694,
+      "step": 3980
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 0.00028893543205419633,
+      "loss": 0.266,
+      "step": 3990
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 0.0002879989718357994,
+      "loss": 0.2658,
+      "step": 4000
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 0.0002870619655121001,
+      "loss": 0.2636,
+      "step": 4010
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 0.0002861244265493606,
+      "loss": 0.2597,
+      "step": 4020
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 0.00028518636842149816,
+      "loss": 0.2564,
+      "step": 4030
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 0.0002842478046098912,
+      "loss": 0.2644,
+      "step": 4040
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 0.00028330874860318543,
+      "loss": 0.2626,
+      "step": 4050
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 0.00028236921389710053,
+      "loss": 0.2632,
+      "step": 4060
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 0.00028142921399423545,
+      "loss": 0.2612,
+      "step": 4070
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 0.0002804887624038751,
+      "loss": 0.268,
+      "step": 4080
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 0.00027954787264179563,
+      "loss": 0.2725,
+      "step": 4090
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 0.00027860655823007055,
+      "loss": 0.2656,
+      "step": 4100
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 0.0002776648326968762,
+      "loss": 0.2715,
+      "step": 4110
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 0.00027672270957629727,
+      "loss": 0.2714,
+      "step": 4120
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 0.00027578020240813256,
+      "loss": 0.2695,
+      "step": 4130
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 0.00027483732473770027,
+      "loss": 0.2613,
+      "step": 4140
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 0.00027389409011564314,
+      "loss": 0.2633,
+      "step": 4150
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 0.0002729505120977341,
+      "loss": 0.261,
+      "step": 4160
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 0.0002720066042446809,
+      "loss": 0.2659,
+      "step": 4170
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 0.0002710623801219319,
+      "loss": 0.2735,
+      "step": 4180
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 0.00027011785329948045,
+      "loss": 0.2644,
+      "step": 4190
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 0.00026917303735167036,
+      "loss": 0.2609,
+      "step": 4200
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 0.0002682279458570006,
+      "loss": 0.2618,
+      "step": 4210
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 0.0002672825923979301,
+      "loss": 0.2731,
+      "step": 4220
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 0.0002663369905606828,
+      "loss": 0.2723,
+      "step": 4230
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 0.00026539115393505185,
+      "loss": 0.2669,
+      "step": 4240
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 0.0002644450961142053,
+      "loss": 0.2811,
+      "step": 4250
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 0.00026349883069448946,
+      "loss": 0.2654,
+      "step": 4260
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 0.00026255237127523467,
+      "loss": 0.279,
+      "step": 4270
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 0.00026160573145855913,
+      "loss": 0.2139,
+      "step": 4280
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 0.00026065892484917363,
+      "loss": 0.1942,
+      "step": 4290
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 0.0002597119650541861,
+      "loss": 0.1964,
+      "step": 4300
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 0.00025876486568290587,
+      "loss": 0.2006,
+      "step": 4310
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 0.0002578176403466483,
+      "loss": 0.1992,
+      "step": 4320
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 0.00025687030265853926,
+      "loss": 0.1988,
+      "step": 4330
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 0.0002559228662333188,
+      "loss": 0.1933,
+      "step": 4340
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 0.00025497534468714645,
+      "loss": 0.2003,
+      "step": 4350
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 0.0002540277516374048,
+      "loss": 0.2007,
+      "step": 4360
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 0.00025308010070250415,
+      "loss": 0.1979,
+      "step": 4370
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 0.00025213240550168677,
+      "loss": 0.2061,
+      "step": 4380
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 0.0002511846796548309,
+      "loss": 0.2021,
+      "step": 4390
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 0.00025023693678225537,
+      "loss": 0.2018,
+      "step": 4400
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 0.0002492891905045237,
+      "loss": 0.2022,
+      "step": 4410
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 0.00024834145444224833,
+      "loss": 0.2026,
+      "step": 4420
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 0.00024739374221589475,
+      "loss": 0.2037,
+      "step": 4430
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 0.0002464460674455861,
+      "loss": 0.2021,
+      "step": 4440
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 0.0002454984437509071,
+      "loss": 0.2024,
+      "step": 4450
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 0.00024455088475070847,
+      "loss": 0.2073,
+      "step": 4460
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 0.0002436034040629111,
+      "loss": 0.2098,
+      "step": 4470
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 0.00024265601530431053,
+      "loss": 0.2099,
+      "step": 4480
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 0.00024170873209038103,
+      "loss": 0.2074,
+      "step": 4490
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 0.00024076156803508,
+      "loss": 0.1981,
+      "step": 4500
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 0.0002398145367506525,
+      "loss": 0.2019,
+      "step": 4510
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 0.00023886765184743527,
+      "loss": 0.2053,
+      "step": 4520
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 0.00023792092693366153,
+      "loss": 0.2016,
+      "step": 4530
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 0.00023697437561526492,
+      "loss": 0.21,
+      "step": 4540
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 0.00023602801149568456,
+      "loss": 0.2038,
+      "step": 4550
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 0.0002350818481756689,
+      "loss": 0.2019,
+      "step": 4560
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 0.00023413589925308102,
+      "loss": 0.2091,
+      "step": 4570
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 0.00023319017832270218,
+      "loss": 0.1961,
+      "step": 4580
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 0.0002322446989760376,
+      "loss": 0.2098,
+      "step": 4590
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 0.0002312994748011203,
+      "loss": 0.1934,
+      "step": 4600
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 0.00023035451938231583,
+      "loss": 0.2029,
+      "step": 4610
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 0.0002294098463001279,
+      "loss": 0.2066,
+      "step": 4620
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 0.00022846546913100194,
+      "loss": 0.2013,
+      "step": 4630
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 0.0002275214014471312,
+      "loss": 0.2117,
+      "step": 4640
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 0.00022657765681626072,
+      "loss": 0.2009,
+      "step": 4650
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 0.00022563424880149305,
+      "loss": 0.2115,
+      "step": 4660
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 0.0002246911909610928,
+      "loss": 0.2095,
+      "step": 4670
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 0.0002237484968482922,
+      "loss": 0.2046,
+      "step": 4680
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 0.00022280618001109604,
+      "loss": 0.2059,
+      "step": 4690
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 0.00022186425399208713,
+      "loss": 0.2004,
+      "step": 4700
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 0.0002209227323282315,
+      "loss": 0.2043,
+      "step": 4710
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 0.00021998162855068406,
+      "loss": 0.2098,
+      "step": 4720
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 0.00021904095618459403,
+      "loss": 0.2077,
+      "step": 4730
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 0.00021810072874891044,
+      "loss": 0.2108,
+      "step": 4740
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 0.0002171609597561882,
+      "loss": 0.2053,
+      "step": 4750
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 0.00021622166271239335,
+      "loss": 0.2111,
+      "step": 4760
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 0.00021528285111670965,
+      "loss": 0.2035,
+      "step": 4770
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 0.00021434453846134384,
+      "loss": 0.2037,
+      "step": 4780
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 0.0002134067382313324,
+      "loss": 0.2093,
+      "step": 4790
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 0.00021246946390434716,
+      "loss": 0.2059,
+      "step": 4800
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 0.0002115327289505023,
+      "loss": 0.2101,
+      "step": 4810
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 0.0002105965468321599,
+      "loss": 0.2149,
+      "step": 4820
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 0.00020966093100373707,
+      "loss": 0.2058,
+      "step": 4830
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 0.00020872589491151278,
+      "loss": 0.2094,
+      "step": 4840
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 0.00020779145199343362,
+      "loss": 0.2108,
+      "step": 4850
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 0.00020685761567892206,
+      "loss": 0.2046,
+      "step": 4860
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 0.000205924399388682,
+      "loss": 0.2103,
+      "step": 4870
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 0.00020499181653450723,
+      "loss": 0.2182,
+      "step": 4880
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 0.0002040598805190877,
+      "loss": 0.2092,
+      "step": 4890
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 0.00020312860473581737,
+      "loss": 0.2061,
+      "step": 4900
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 0.00020219800256860159,
+      "loss": 0.2053,
+      "step": 4910
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 0.00020126808739166487,
+      "loss": 0.2144,
+      "step": 4920
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 0.00020033887256935842,
+      "loss": 0.2101,
+      "step": 4930
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 0.0001994103714559684,
+      "loss": 0.2068,
+      "step": 4940
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 0.00019848259739552382,
+      "loss": 0.2126,
+      "step": 4950
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 0.00019755556372160466,
+      "loss": 0.2023,
+      "step": 4960
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 0.00019662928375715057,
+      "loss": 0.2102,
+      "step": 4970
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 0.00019570377081426903,
+      "loss": 0.2111,
+      "step": 4980
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 0.00019477903819404435,
+      "loss": 0.1718,
+      "step": 4990
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 0.00019385509918634628,
+      "loss": 0.1542,
+      "step": 5000
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 0.00019293196706963913,
+      "loss": 0.156,
+      "step": 5010
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 0.00019200965511079082,
+      "loss": 0.1501,
+      "step": 5020
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 0.00019108817656488263,
+      "loss": 0.1526,
+      "step": 5030
+    },
+    {
+      "epoch": 7.08,
+      "learning_rate": 0.0001901675446750179,
+      "loss": 0.1483,
+      "step": 5040
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 0.0001892477726721325,
+      "loss": 0.1484,
+      "step": 5050
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 0.0001883288737748043,
+      "loss": 0.1566,
+      "step": 5060
+    },
+    {
+      "epoch": 7.12,
+      "learning_rate": 0.0001874108611890632,
+      "loss": 0.1543,
+      "step": 5070
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 0.00018649374810820164,
+      "loss": 0.1477,
+      "step": 5080
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 0.00018557754771258425,
+      "loss": 0.1493,
+      "step": 5090
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 0.00018466227316945962,
+      "loss": 0.1539,
+      "step": 5100
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 0.00018374793763276966,
+      "loss": 0.1521,
+      "step": 5110
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 0.0001828345542429618,
+      "loss": 0.1579,
+      "step": 5120
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 0.00018192213612679942,
+      "loss": 0.154,
+      "step": 5130
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 0.00018101069639717338,
+      "loss": 0.1519,
+      "step": 5140
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 0.0001801002481529136,
+      "loss": 0.1517,
+      "step": 5150
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 0.00017919080447860076,
+      "loss": 0.1567,
+      "step": 5160
+    },
+    {
+      "epoch": 7.26,
+      "learning_rate": 0.00017828237844437833,
+      "loss": 0.1477,
+      "step": 5170
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 0.00017737498310576463,
+      "loss": 0.1556,
+      "step": 5180
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 0.00017646863150346537,
+      "loss": 0.1606,
+      "step": 5190
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 0.00017556333666318596,
+      "loss": 0.1551,
+      "step": 5200
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 0.0001746591115954446,
+      "loss": 0.153,
+      "step": 5210
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 0.000173755969295385,
+      "loss": 0.1571,
+      "step": 5220
+    },
+    {
+      "epoch": 7.34,
+      "learning_rate": 0.00017285392274259004,
+      "loss": 0.1488,
+      "step": 5230
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 0.00017195298490089474,
+      "loss": 0.1607,
+      "step": 5240
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 0.00017105316871820036,
+      "loss": 0.159,
+      "step": 5250
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 0.00017015448712628803,
+      "loss": 0.1607,
+      "step": 5260
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 0.00016925695304063302,
+      "loss": 0.1578,
+      "step": 5270
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 0.00016836057936021924,
+      "loss": 0.1608,
+      "step": 5280
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 0.0001674653789673536,
+      "loss": 0.157,
+      "step": 5290
+    },
+    {
+      "epoch": 7.44,
+      "learning_rate": 0.00016657136472748114,
+      "loss": 0.1572,
+      "step": 5300
+    },
+    {
+      "epoch": 7.46,
+      "learning_rate": 0.00016567854948899976,
+      "loss": 0.1577,
+      "step": 5310
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 0.00016478694608307635,
+      "loss": 0.1511,
+      "step": 5320
+    },
+    {
+      "epoch": 7.48,
+      "learning_rate": 0.0001638965673234611,
+      "loss": 0.162,
+      "step": 5330
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 0.00016300742600630478,
+      "loss": 0.1595,
+      "step": 5340
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 0.00016211953490997372,
+      "loss": 0.153,
+      "step": 5350
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 0.00016123290679486682,
+      "loss": 0.1568,
+      "step": 5360
+    },
+    {
+      "epoch": 7.54,
+      "learning_rate": 0.0001603475544032319,
+      "loss": 0.1578,
+      "step": 5370
+    },
+    {
+      "epoch": 7.56,
+      "learning_rate": 0.0001594634904589824,
+      "loss": 0.1544,
+      "step": 5380
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 0.00015858072766751513,
+      "loss": 0.1613,
+      "step": 5390
+    },
+    {
+      "epoch": 7.58,
+      "learning_rate": 0.00015769927871552697,
+      "loss": 0.1562,
+      "step": 5400
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 0.00015681915627083302,
+      "loss": 0.1614,
+      "step": 5410
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 0.00015594037298218417,
+      "loss": 0.1602,
+      "step": 5420
+    },
+    {
+      "epoch": 7.63,
+      "learning_rate": 0.0001550629414790858,
+      "loss": 0.1606,
+      "step": 5430
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 0.00015418687437161575,
+      "loss": 0.1582,
+      "step": 5440
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 0.0001533121842502435,
+      "loss": 0.1555,
+      "step": 5450
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 0.0001524388836856489,
+      "loss": 0.1528,
+      "step": 5460
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 0.0001515669852285419,
+      "loss": 0.1625,
+      "step": 5470
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 0.00015069650140948172,
+      "loss": 0.1575,
+      "step": 5480
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 0.0001498274447386971,
+      "loss": 0.1562,
+      "step": 5490
+    },
+    {
+      "epoch": 7.72,
+      "learning_rate": 0.00014895982770590644,
+      "loss": 0.163,
+      "step": 5500
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 0.00014809366278013817,
+      "loss": 0.1619,
+      "step": 5510
+    },
+    {
+      "epoch": 7.75,
+      "learning_rate": 0.00014722896240955175,
+      "loss": 0.1563,
+      "step": 5520
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 0.0001463657390212585,
+      "loss": 0.1656,
+      "step": 5530
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 0.00014550400502114346,
+      "loss": 0.1506,
+      "step": 5540
+    },
+    {
+      "epoch": 7.79,
+      "learning_rate": 0.00014464377279368654,
+      "loss": 0.1557,
+      "step": 5550
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 0.00014378505470178496,
+      "loss": 0.1571,
+      "step": 5560
+    },
+    {
+      "epoch": 7.82,
+      "learning_rate": 0.00014292786308657519,
+      "loss": 0.1585,
+      "step": 5570
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 0.00014207221026725609,
+      "loss": 0.1575,
+      "step": 5580
+    },
+    {
+      "epoch": 7.85,
+      "learning_rate": 0.00014121810854091156,
+      "loss": 0.1577,
+      "step": 5590
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 0.00014036557018233355,
+      "loss": 0.1612,
+      "step": 5600
+    },
+    {
+      "epoch": 7.88,
+      "learning_rate": 0.00013951460744384626,
+      "loss": 0.1586,
+      "step": 5610
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 0.00013866523255512961,
+      "loss": 0.1643,
+      "step": 5620
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 0.0001378174577230438,
+      "loss": 0.1597,
+      "step": 5630
+    },
+    {
+      "epoch": 7.92,
+      "learning_rate": 0.00013697129513145327,
+      "loss": 0.1632,
+      "step": 5640
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 0.00013612675694105235,
+      "loss": 0.1504,
+      "step": 5650
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 0.00013528385528919007,
+      "loss": 0.1599,
+      "step": 5660
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 0.00013444260228969595,
+      "loss": 0.1616,
+      "step": 5670
+    },
+    {
+      "epoch": 7.98,
+      "learning_rate": 0.0001336030100327054,
+      "loss": 0.1545,
+      "step": 5680
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 0.00013276509058448666,
+      "loss": 0.1621,
+      "step": 5690
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 0.00013192885598726706,
+      "loss": 0.1444,
+      "step": 5700
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 0.00013109431825905962,
+      "loss": 0.1246,
+      "step": 5710
+    },
+    {
+      "epoch": 8.03,
+      "learning_rate": 0.00013026148939349126,
+      "loss": 0.1173,
+      "step": 5720
+    },
+    {
+      "epoch": 8.05,
+      "learning_rate": 0.00012943038135962922,
+      "loss": 0.1181,
+      "step": 5730
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 0.00012860100610181005,
+      "loss": 0.1161,
+      "step": 5740
+    },
+    {
+      "epoch": 8.07,
+      "learning_rate": 0.00012777337553946744,
+      "loss": 0.1121,
+      "step": 5750
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 0.00012694750156696123,
+      "loss": 0.1208,
+      "step": 5760
+    },
+    {
+      "epoch": 8.1,
+      "learning_rate": 0.00012612339605340585,
+      "loss": 0.1186,
+      "step": 5770
+    },
+    {
+      "epoch": 8.12,
+      "learning_rate": 0.0001253010708425006,
+      "loss": 0.1184,
+      "step": 5780
+    },
+    {
+      "epoch": 8.13,
+      "learning_rate": 0.00012448053775235892,
+      "loss": 0.1155,
+      "step": 5790
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 0.00012366180857533847,
+      "loss": 0.1176,
+      "step": 5800
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 0.00012284489507787206,
+      "loss": 0.1216,
+      "step": 5810
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 0.00012202980900029814,
+      "loss": 0.1222,
+      "step": 5820
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 0.00012121656205669243,
+      "loss": 0.1117,
+      "step": 5830
+    },
+    {
+      "epoch": 8.2,
+      "learning_rate": 0.00012040516593469908,
+      "loss": 0.11,
+      "step": 5840
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 0.0001195956322953633,
+      "loss": 0.1144,
+      "step": 5850
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 0.00011878797277296338,
+      "loss": 0.1233,
+      "step": 5860
+    },
+    {
+      "epoch": 8.24,
+      "learning_rate": 0.00011798219897484366,
+      "loss": 0.1182,
+      "step": 5870
+    },
+    {
+      "epoch": 8.26,
+      "learning_rate": 0.00011717832248124732,
+      "loss": 0.1241,
+      "step": 5880
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 0.00011637635484515083,
+      "loss": 0.1184,
+      "step": 5890
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 0.00011557630759209691,
+      "loss": 0.118,
+      "step": 5900
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 0.00011477819222002935,
+      "loss": 0.1193,
+      "step": 5910
+    },
+    {
+      "epoch": 8.31,
+      "learning_rate": 0.00011398202019912801,
+      "loss": 0.1162,
+      "step": 5920
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 0.00011318780297164358,
+      "loss": 0.1151,
+      "step": 5930
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 0.00011239555195173337,
+      "loss": 0.1176,
+      "step": 5940
+    },
+    {
+      "epoch": 8.36,
+      "learning_rate": 0.00011160527852529686,
+      "loss": 0.1182,
+      "step": 5950
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 0.00011081699404981297,
+      "loss": 0.1178,
+      "step": 5960
+    },
+    {
+      "epoch": 8.38,
+      "learning_rate": 0.00011003070985417563,
+      "loss": 0.1194,
+      "step": 5970
+    },
+    {
+      "epoch": 8.4,
+      "learning_rate": 0.00010924643723853192,
+      "loss": 0.1174,
+      "step": 5980
+    },
+    {
+      "epoch": 8.41,
+      "learning_rate": 0.00010846418747411934,
+      "loss": 0.1093,
+      "step": 5990
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 0.00010768397180310386,
+      "loss": 0.119,
+      "step": 6000
+    },
+    {
+      "epoch": 8.44,
+      "learning_rate": 0.00010690580143841794,
+      "loss": 0.121,
+      "step": 6010
+    },
+    {
+      "epoch": 8.45,
+      "learning_rate": 0.0001061296875636002,
+      "loss": 0.1202,
+      "step": 6020
+    },
+    {
+      "epoch": 8.47,
+      "learning_rate": 0.00010535564133263409,
+      "loss": 0.1196,
+      "step": 6030
+    },
+    {
+      "epoch": 8.48,
+      "learning_rate": 0.00010458367386978762,
+      "loss": 0.1197,
+      "step": 6040
+    },
+    {
+      "epoch": 8.5,
+      "learning_rate": 0.00010381379626945378,
+      "loss": 0.1178,
+      "step": 6050
+    },
+    {
+      "epoch": 8.51,
+      "learning_rate": 0.00010304601959599089,
+      "loss": 0.1249,
+      "step": 6060
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 0.00010228035488356369,
+      "loss": 0.1173,
+      "step": 6070
+    },
+    {
+      "epoch": 8.54,
+      "learning_rate": 0.00010151681313598446,
+      "loss": 0.117,
+      "step": 6080
+    },
+    {
+      "epoch": 8.55,
+      "learning_rate": 0.0001007554053265554,
+      "loss": 0.1191,
+      "step": 6090
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 9.999614239791049e-05,
+      "loss": 0.1243,
+      "step": 6100
+    },
+    {
+      "epoch": 8.58,
+      "learning_rate": 9.923903526185854e-05,
+      "loss": 0.1231,
+      "step": 6110
+    },
+    {
+      "epoch": 8.59,
+      "learning_rate": 9.848409479922594e-05,
+      "loss": 0.118,
+      "step": 6120
+    },
+    {
+      "epoch": 8.61,
+      "learning_rate": 9.773133185970076e-05,
+      "loss": 0.1208,
+      "step": 6130
+    },
+    {
+      "epoch": 8.62,
+      "learning_rate": 9.698075726167669e-05,
+      "loss": 0.1193,
+      "step": 6140
+    },
+    {
+      "epoch": 8.64,
+      "learning_rate": 9.623238179209717e-05,
+      "loss": 0.1197,
+      "step": 6150
+    },
+    {
+      "epoch": 8.65,
+      "learning_rate": 9.548621620630102e-05,
+      "loss": 0.1173,
+      "step": 6160
+    },
+    {
+      "epoch": 8.66,
+      "learning_rate": 9.474227122786738e-05,
+      "loss": 0.1161,
+      "step": 6170
+    },
+    {
+      "epoch": 8.68,
+      "learning_rate": 9.40005575484619e-05,
+      "loss": 0.1175,
+      "step": 6180
+    },
+    {
+      "epoch": 8.69,
+      "learning_rate": 9.326108582768253e-05,
+      "loss": 0.112,
+      "step": 6190
+    },
+    {
+      "epoch": 8.71,
+      "learning_rate": 9.252386669290738e-05,
+      "loss": 0.1188,
+      "step": 6200
+    },
+    {
+      "epoch": 8.72,
+      "learning_rate": 9.178891073914073e-05,
+      "loss": 0.1206,
+      "step": 6210
+    },
+    {
+      "epoch": 8.73,
+      "learning_rate": 9.105622852886166e-05,
+      "loss": 0.126,
+      "step": 6220
+    },
+    {
+      "epoch": 8.75,
+      "learning_rate": 9.032583059187203e-05,
+      "loss": 0.1168,
+      "step": 6230
+    },
+    {
+      "epoch": 8.76,
+      "learning_rate": 8.95977274251448e-05,
+      "loss": 0.1157,
+      "step": 6240
+    },
+    {
+      "epoch": 8.78,
+      "learning_rate": 8.88719294926737e-05,
+      "loss": 0.1196,
+      "step": 6250
+    },
+    {
+      "epoch": 8.79,
+      "learning_rate": 8.81484472253225e-05,
+      "loss": 0.117,
+      "step": 6260
+    },
+    {
+      "epoch": 8.81,
+      "learning_rate": 8.742729102067535e-05,
+      "loss": 0.1244,
+      "step": 6270
+    },
+    {
+      "epoch": 8.82,
+      "learning_rate": 8.670847124288686e-05,
+      "loss": 0.1204,
+      "step": 6280
+    },
+    {
+      "epoch": 8.83,
+      "learning_rate": 8.599199822253384e-05,
+      "loss": 0.1245,
+      "step": 6290
+    },
+    {
+      "epoch": 8.85,
+      "learning_rate": 8.527788225646629e-05,
+      "loss": 0.1139,
+      "step": 6300
+    },
+    {
+      "epoch": 8.86,
+      "learning_rate": 8.45661336076598e-05,
+      "loss": 0.1172,
+      "step": 6310
+    },
+    {
+      "epoch": 8.88,
+      "learning_rate": 8.385676250506757e-05,
+      "loss": 0.1153,
+      "step": 6320
+    },
+    {
+      "epoch": 8.89,
+      "learning_rate": 8.314977914347397e-05,
+      "loss": 0.1202,
+      "step": 6330
+    },
+    {
+      "epoch": 8.9,
+      "learning_rate": 8.244519368334771e-05,
+      "loss": 0.1253,
+      "step": 6340
+    },
+    {
+      "epoch": 8.92,
+      "learning_rate": 8.174301625069566e-05,
+      "loss": 0.1215,
+      "step": 6350
+    },
+    {
+      "epoch": 8.93,
+      "learning_rate": 8.104325693691778e-05,
+      "loss": 0.1261,
+      "step": 6360
+    },
+    {
+      "epoch": 8.95,
+      "learning_rate": 8.034592579866174e-05,
+      "loss": 0.1234,
+      "step": 6370
+    },
+    {
+      "epoch": 8.96,
+      "learning_rate": 7.965103285767857e-05,
+      "loss": 0.1205,
+      "step": 6380
+    },
+    {
+      "epoch": 8.97,
+      "learning_rate": 7.89585881006783e-05,
+      "loss": 0.1246,
+      "step": 6390
+    },
+    {
+      "epoch": 8.99,
+      "learning_rate": 7.826860147918696e-05,
+      "loss": 0.1206,
+      "step": 6400
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 7.758108290940317e-05,
+      "loss": 0.1219,
+      "step": 6410
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 7.68960422720558e-05,
+      "loss": 0.0919,
+      "step": 6420
+    },
+    {
+      "epoch": 9.03,
+      "learning_rate": 7.621348941226169e-05,
+      "loss": 0.0909,
+      "step": 6430
+    },
+    {
+      "epoch": 9.04,
+      "learning_rate": 7.553343413938488e-05,
+      "loss": 0.0927,
+      "step": 6440
+    },
+    {
+      "epoch": 9.06,
+      "learning_rate": 7.485588622689466e-05,
+      "loss": 0.0933,
+      "step": 6450
+    },
+    {
+      "epoch": 9.07,
+      "learning_rate": 7.418085541222555e-05,
+      "loss": 0.093,
+      "step": 6460
+    },
+    {
+      "epoch": 9.09,
+      "learning_rate": 7.350835139663794e-05,
+      "loss": 0.0952,
+      "step": 6470
+    },
+    {
+      "epoch": 9.1,
+      "learning_rate": 7.283838384507755e-05,
+      "loss": 0.0919,
+      "step": 6480
+    },
+    {
+      "epoch": 9.11,
+      "learning_rate": 7.217096238603741e-05,
+      "loss": 0.0898,
+      "step": 6490
+    },
+    {
+      "epoch": 9.13,
+      "learning_rate": 7.150609661141905e-05,
+      "loss": 0.0848,
+      "step": 6500
+    },
+    {
+      "epoch": 9.14,
+      "learning_rate": 7.084379607639493e-05,
+      "loss": 0.0973,
+      "step": 6510
+    },
+    {
+      "epoch": 9.16,
+      "learning_rate": 7.018407029927071e-05,
+      "loss": 0.0967,
+      "step": 6520
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 6.952692876134886e-05,
+      "loss": 0.0934,
+      "step": 6530
+    },
+    {
+      "epoch": 9.18,
+      "learning_rate": 6.887238090679231e-05,
+      "loss": 0.0959,
+      "step": 6540
+    },
+    {
+      "epoch": 9.2,
+      "learning_rate": 6.822043614248863e-05,
+      "loss": 0.0897,
+      "step": 6550
+    },
+    {
+      "epoch": 9.21,
+      "learning_rate": 6.757110383791465e-05,
+      "loss": 0.0935,
+      "step": 6560
+    },
+    {
+      "epoch": 9.23,
+      "learning_rate": 6.692439332500227e-05,
+      "loss": 0.0958,
+      "step": 6570
+    },
+    {
+      "epoch": 9.24,
+      "learning_rate": 6.628031389800416e-05,
+      "loss": 0.0894,
+      "step": 6580
+    },
+    {
+      "epoch": 9.25,
+      "learning_rate": 6.563887481335976e-05,
+      "loss": 0.0946,
+      "step": 6590
+    },
+    {
+      "epoch": 9.27,
+      "learning_rate": 6.500008528956295e-05,
+      "loss": 0.0924,
+      "step": 6600
+    },
+    {
+      "epoch": 9.28,
+      "learning_rate": 6.436395450702911e-05,
+      "loss": 0.0975,
+      "step": 6610
+    },
+    {
+      "epoch": 9.3,
+      "learning_rate": 6.373049160796343e-05,
+      "loss": 0.0923,
+      "step": 6620
+    },
+    {
+      "epoch": 9.31,
+      "learning_rate": 6.309970569622906e-05,
+      "loss": 0.0952,
+      "step": 6630
+    },
+    {
+      "epoch": 9.32,
+      "learning_rate": 6.247160583721708e-05,
+      "loss": 0.092,
+      "step": 6640
+    },
+    {
+      "epoch": 9.34,
+      "learning_rate": 6.184620105771535e-05,
+      "loss": 0.0925,
+      "step": 6650
+    },
+    {
+      "epoch": 9.35,
+      "learning_rate": 6.122350034577945e-05,
+      "loss": 0.0889,
+      "step": 6660
+    },
+    {
+      "epoch": 9.37,
+      "learning_rate": 6.060351265060296e-05,
+      "loss": 0.0862,
+      "step": 6670
+    },
+    {
+      "epoch": 9.38,
+      "learning_rate": 5.998624688238932e-05,
+      "loss": 0.0956,
+      "step": 6680
+    },
+    {
+      "epoch": 9.39,
+      "learning_rate": 5.9371711912223655e-05,
+      "loss": 0.0906,
+      "step": 6690
+    },
+    {
+      "epoch": 9.41,
+      "learning_rate": 5.875991657194485e-05,
+      "loss": 0.0971,
+      "step": 6700
+    },
+    {
+      "epoch": 9.42,
+      "learning_rate": 5.815086965401944e-05,
+      "loss": 0.0906,
+      "step": 6710
+    },
+    {
+      "epoch": 9.44,
+      "learning_rate": 5.7544579911414404e-05,
+      "loss": 0.0903,
+      "step": 6720
+    },
+    {
+      "epoch": 9.45,
+      "learning_rate": 5.694105605747196e-05,
+      "loss": 0.0917,
+      "step": 6730
+    },
+    {
+      "epoch": 9.47,
+      "learning_rate": 5.6340306765784025e-05,
+      "loss": 0.0993,
+      "step": 6740
+    },
+    {
+      "epoch": 9.48,
+      "learning_rate": 5.5742340670067835e-05,
+      "loss": 0.0927,
+      "step": 6750
+    },
+    {
+      "epoch": 9.49,
+      "learning_rate": 5.5147166364041405e-05,
+      "loss": 0.0909,
+      "step": 6760
+    },
+    {
+      "epoch": 9.51,
+      "learning_rate": 5.4554792401300595e-05,
+      "loss": 0.087,
+      "step": 6770
+    },
+    {
+      "epoch": 9.52,
+      "learning_rate": 5.396522729519587e-05,
+      "loss": 0.0933,
+      "step": 6780
+    },
+    {
+      "epoch": 9.54,
+      "learning_rate": 5.3378479518709777e-05,
+      "loss": 0.0966,
+      "step": 6790
+    },
+    {
+      "epoch": 9.55,
+      "learning_rate": 5.279455750433557e-05,
+      "loss": 0.091,
+      "step": 6800
+    },
+    {
+      "epoch": 9.56,
+      "learning_rate": 5.221346964395585e-05,
+      "loss": 0.0943,
+      "step": 6810
+    },
+    {
+      "epoch": 9.58,
+      "learning_rate": 5.163522428872186e-05,
+      "loss": 0.0975,
+      "step": 6820
+    },
+    {
+      "epoch": 9.59,
+      "learning_rate": 5.105982974893353e-05,
+      "loss": 0.0942,
+      "step": 6830
+    },
+    {
+      "epoch": 9.61,
+      "learning_rate": 5.04872942939201e-05,
+      "loss": 0.094,
+      "step": 6840
+    },
+    {
+      "epoch": 9.62,
+      "learning_rate": 4.991762615192128e-05,
+      "loss": 0.1002,
+      "step": 6850
+    },
+    {
+      "epoch": 9.63,
+      "learning_rate": 4.9350833509969e-05,
+      "loss": 0.0922,
+      "step": 6860
+    },
+    {
+      "epoch": 9.65,
+      "learning_rate": 4.8786924513769424e-05,
+      "loss": 0.0919,
+      "step": 6870
+    },
+    {
+      "epoch": 9.66,
+      "learning_rate": 4.822590726758666e-05,
+      "loss": 0.0939,
+      "step": 6880
+    },
+    {
+      "epoch": 9.68,
+      "learning_rate": 4.7667789834125366e-05,
+      "loss": 0.0946,
+      "step": 6890
+    },
+    {
+      "epoch": 9.69,
+      "learning_rate": 4.711258023441542e-05,
+      "loss": 0.092,
+      "step": 6900
+    },
+    {
+      "epoch": 9.7,
+      "learning_rate": 4.6560286447696585e-05,
+      "loss": 0.0949,
+      "step": 6910
+    },
+    {
+      "epoch": 9.72,
+      "learning_rate": 4.6010916411303784e-05,
+      "loss": 0.0956,
+      "step": 6920
+    },
+    {
+      "epoch": 9.73,
+      "learning_rate": 4.546447802055309e-05,
+      "loss": 0.0918,
+      "step": 6930
+    },
+    {
+      "epoch": 9.75,
+      "learning_rate": 4.4920979128627884e-05,
+      "loss": 0.0885,
+      "step": 6940
+    },
+    {
+      "epoch": 9.76,
+      "learning_rate": 4.4380427546466816e-05,
+      "loss": 0.0919,
+      "step": 6950
+    },
+    {
+      "epoch": 9.77,
+      "learning_rate": 4.384283104265052e-05,
+      "loss": 0.0865,
+      "step": 6960
+    },
+    {
+      "epoch": 9.79,
+      "learning_rate": 4.3308197343290776e-05,
+      "loss": 0.0937,
+      "step": 6970
+    },
+    {
+      "epoch": 9.8,
+      "learning_rate": 4.2776534131919147e-05,
+      "loss": 0.0937,
+      "step": 6980
+    },
+    {
+      "epoch": 9.82,
+      "learning_rate": 4.2247849049376566e-05,
+      "loss": 0.0941,
+      "step": 6990
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 4.1722149693703415e-05,
+      "loss": 0.0895,
+      "step": 7000
+    },
+    {
+      "epoch": 9.84,
+      "learning_rate": 4.119944362003067e-05,
+      "loss": 0.0954,
+      "step": 7010
+    },
+    {
+      "epoch": 9.86,
+      "learning_rate": 4.067973834047106e-05,
+      "loss": 0.0909,
+      "step": 7020
+    },
+    {
+      "epoch": 9.87,
+      "learning_rate": 4.016304132401097e-05,
+      "loss": 0.0931,
+      "step": 7030
+    },
+    {
+      "epoch": 9.89,
+      "learning_rate": 3.9649359996403536e-05,
+      "loss": 0.091,
+      "step": 7040
+    },
+    {
+      "epoch": 9.9,
+      "learning_rate": 3.9138701740061564e-05,
+      "loss": 0.0925,
+      "step": 7050
+    },
+    {
+      "epoch": 9.91,
+      "learning_rate": 3.8631073893951596e-05,
+      "loss": 0.0947,
+      "step": 7060
+    },
+    {
+      "epoch": 9.93,
+      "learning_rate": 3.8126483753488263e-05,
+      "loss": 0.0961,
+      "step": 7070
+    },
+    {
+      "epoch": 9.94,
+      "learning_rate": 3.76249385704297e-05,
+      "loss": 0.0938,
+      "step": 7080
+    },
+    {
+      "epoch": 9.96,
+      "learning_rate": 3.712644555277311e-05,
+      "loss": 0.0951,
+      "step": 7090
+    },
+    {
+      "epoch": 9.97,
+      "learning_rate": 3.66310118646514e-05,
+      "loss": 0.0923,
+      "step": 7100
+    },
+    {
+      "epoch": 9.98,
+      "learning_rate": 3.613864462622979e-05,
+      "loss": 0.0985,
+      "step": 7110
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 3.564935091360405e-05,
+      "loss": 0.0941,
+      "step": 7120
+    },
+    {
+      "epoch": 10.01,
+      "learning_rate": 3.516313775869842e-05,
+      "loss": 0.0783,
+      "step": 7130
+    },
+    {
+      "epoch": 10.03,
+      "learning_rate": 3.4680012149164494e-05,
+      "loss": 0.0823,
+      "step": 7140
+    },
+    {
+      "epoch": 10.04,
+      "learning_rate": 3.419998102828134e-05,
+      "loss": 0.0768,
+      "step": 7150
+    },
+    {
+      "epoch": 10.05,
+      "learning_rate": 3.372305129485495e-05,
+      "loss": 0.0796,
+      "step": 7160
+    },
+    {
+      "epoch": 10.07,
+      "learning_rate": 3.32492298031197e-05,
+      "loss": 0.0747,
+      "step": 7170
+    },
+    {
+      "epoch": 10.08,
+      "learning_rate": 3.277852336263945e-05,
+      "loss": 0.0777,
+      "step": 7180
+    },
+    {
+      "epoch": 10.1,
+      "learning_rate": 3.231093873821009e-05,
+      "loss": 0.0756,
+      "step": 7190
+    },
+    {
+      "epoch": 10.11,
+      "learning_rate": 3.1846482649761865e-05,
+      "loss": 0.0822,
+      "step": 7200
+    },
+    {
+      "epoch": 10.13,
+      "learning_rate": 3.1385161772262997e-05,
+      "loss": 0.0743,
+      "step": 7210
+    },
+    {
+      "epoch": 10.14,
+      "learning_rate": 3.0926982735624026e-05,
+      "loss": 0.0794,
+      "step": 7220
+    },
+    {
+      "epoch": 10.15,
+      "learning_rate": 3.0471952124601936e-05,
+      "loss": 0.0796,
+      "step": 7230
+    },
+    {
+      "epoch": 10.17,
+      "learning_rate": 3.0020076478706083e-05,
+      "loss": 0.0728,
+      "step": 7240
+    },
+    {
+      "epoch": 10.18,
+      "learning_rate": 2.9571362292103942e-05,
+      "loss": 0.0788,
+      "step": 7250
+    },
+    {
+      "epoch": 10.2,
+      "learning_rate": 2.9125816013527843e-05,
+      "loss": 0.0776,
+      "step": 7260
+    },
+    {
+      "epoch": 10.21,
+      "learning_rate": 2.8683444046182118e-05,
+      "loss": 0.0802,
+      "step": 7270
+    },
+    {
+      "epoch": 10.22,
+      "learning_rate": 2.8244252747651383e-05,
+      "loss": 0.0773,
+      "step": 7280
+    },
+    {
+      "epoch": 10.24,
+      "learning_rate": 2.7808248429808973e-05,
+      "loss": 0.0791,
+      "step": 7290
+    },
+    {
+      "epoch": 10.25,
+      "learning_rate": 2.737543735872633e-05,
+      "loss": 0.0763,
+      "step": 7300
+    },
+    {
+      "epoch": 10.27,
+      "learning_rate": 2.694582575458271e-05,
+      "loss": 0.0796,
+      "step": 7310
+    },
+    {
+      "epoch": 10.28,
+      "learning_rate": 2.6519419791576156e-05,
+      "loss": 0.0743,
+      "step": 7320
+    },
+    {
+      "epoch": 10.29,
+      "learning_rate": 2.6096225597834534e-05,
+      "loss": 0.0754,
+      "step": 7330
+    },
+    {
+      "epoch": 10.31,
+      "learning_rate": 2.567624925532741e-05,
+      "loss": 0.0751,
+      "step": 7340
+    },
+    {
+      "epoch": 10.32,
+      "learning_rate": 2.5259496799778877e-05,
+      "loss": 0.0763,
+      "step": 7350
+    },
+    {
+      "epoch": 10.34,
+      "learning_rate": 2.48459742205806e-05,
+      "loss": 0.0798,
+      "step": 7360
+    },
+    {
+      "epoch": 10.35,
+      "learning_rate": 2.4435687460705874e-05,
+      "loss": 0.0753,
+      "step": 7370
+    },
+    {
+      "epoch": 10.36,
+      "learning_rate": 2.4028642416623974e-05,
+      "loss": 0.0774,
+      "step": 7380
+    },
+    {
+      "epoch": 10.38,
+      "learning_rate": 2.36248449382159e-05,
+      "loss": 0.0767,
+      "step": 7390
+    },
+    {
+      "epoch": 10.39,
+      "learning_rate": 2.322430082868973e-05,
+      "loss": 0.0754,
+      "step": 7400
+    },
+    {
+      "epoch": 10.41,
+      "learning_rate": 2.2827015844497616e-05,
+      "loss": 0.0794,
+      "step": 7410
+    },
+    {
+      "epoch": 10.42,
+      "learning_rate": 2.2432995695252866e-05,
+      "loss": 0.08,
+      "step": 7420
+    },
+    {
+      "epoch": 10.43,
+      "learning_rate": 2.2042246043648017e-05,
+      "loss": 0.0751,
+      "step": 7430
+    },
+    {
+      "epoch": 10.45,
+      "learning_rate": 2.1654772505373343e-05,
+      "loss": 0.082,
+      "step": 7440
+    },
+    {
+      "epoch": 10.46,
+      "learning_rate": 2.1270580649036114e-05,
+      "loss": 0.0787,
+      "step": 7450
+    },
+    {
+      "epoch": 10.48,
+      "learning_rate": 2.0889675996080855e-05,
+      "loss": 0.0801,
+      "step": 7460
+    },
+    {
+      "epoch": 10.49,
+      "learning_rate": 2.0512064020709543e-05,
+      "loss": 0.0805,
+      "step": 7470
+    },
+    {
+      "epoch": 10.5,
+      "learning_rate": 2.013775014980332e-05,
+      "loss": 0.0776,
+      "step": 7480
+    },
+    {
+      "epoch": 10.52,
+      "learning_rate": 1.9766739762844295e-05,
+      "loss": 0.0792,
+      "step": 7490
+    },
+    {
+      "epoch": 10.53,
+      "learning_rate": 1.939903819183833e-05,
+      "loss": 0.0776,
+      "step": 7500
+    },
+    {
+      "epoch": 10.55,
+      "learning_rate": 1.9034650721238305e-05,
+      "loss": 0.0789,
+      "step": 7510
+    },
+    {
+      "epoch": 10.56,
+      "learning_rate": 1.8673582587868276e-05,
+      "loss": 0.08,
+      "step": 7520
+    },
+    {
+      "epoch": 10.57,
+      "learning_rate": 1.8315838980848238e-05,
+      "loss": 0.0761,
+      "step": 7530
+    },
+    {
+      "epoch": 10.59,
+      "learning_rate": 1.7961425041519318e-05,
+      "loss": 0.079,
+      "step": 7540
+    },
+    {
+      "epoch": 10.6,
+      "learning_rate": 1.7610345863370237e-05,
+      "loss": 0.0766,
+      "step": 7550
+    },
+    {
+      "epoch": 10.62,
+      "learning_rate": 1.7262606491963843e-05,
+      "loss": 0.0766,
+      "step": 7560
+    },
+    {
+      "epoch": 10.63,
+      "learning_rate": 1.69182119248647e-05,
+      "loss": 0.0857,
+      "step": 7570
+    },
+    {
+      "epoch": 10.64,
+      "learning_rate": 1.6577167111567176e-05,
+      "loss": 0.0778,
+      "step": 7580
+    },
+    {
+      "epoch": 10.66,
+      "learning_rate": 1.6239476953424453e-05,
+      "loss": 0.0802,
+      "step": 7590
+    },
+    {
+      "epoch": 10.67,
+      "learning_rate": 1.5905146303578032e-05,
+      "loss": 0.0787,
+      "step": 7600
+    },
+    {
+      "epoch": 10.69,
+      "learning_rate": 1.5574179966887953e-05,
+      "loss": 0.0777,
+      "step": 7610
+    },
+    {
+      "epoch": 10.7,
+      "learning_rate": 1.5246582699863593e-05,
+      "loss": 0.0756,
+      "step": 7620
+    },
+    {
+      "epoch": 10.71,
+      "learning_rate": 1.4922359210595842e-05,
+      "loss": 0.0713,
+      "step": 7630
+    },
+    {
+      "epoch": 10.73,
+      "learning_rate": 1.4601514158688684e-05,
+      "loss": 0.08,
+      "step": 7640
+    },
+    {
+      "epoch": 10.74,
+      "learning_rate": 1.4284052155192822e-05,
+      "loss": 0.0761,
+      "step": 7650
+    },
+    {
+      "epoch": 10.76,
+      "learning_rate": 1.3969977762539138e-05,
+      "loss": 0.0794,
+      "step": 7660
+    },
+    {
+      "epoch": 10.77,
+      "learning_rate": 1.3659295494473228e-05,
+      "loss": 0.0816,
+      "step": 7670
+    },
+    {
+      "epoch": 10.79,
+      "learning_rate": 1.3352009815990552e-05,
+      "loss": 0.0767,
+      "step": 7680
+    },
+    {
+      "epoch": 10.8,
+      "learning_rate": 1.3048125143271972e-05,
+      "loss": 0.082,
+      "step": 7690
+    },
+    {
+      "epoch": 10.81,
+      "learning_rate": 1.2747645843620842e-05,
+      "loss": 0.0808,
+      "step": 7700
+    },
+    {
+      "epoch": 10.83,
+      "learning_rate": 1.245057623539969e-05,
+      "loss": 0.0705,
+      "step": 7710
+    },
+    {
+      "epoch": 10.84,
+      "learning_rate": 1.2156920587968439e-05,
+      "loss": 0.0735,
+      "step": 7720
+    },
+    {
+      "epoch": 10.86,
+      "learning_rate": 1.1866683121623062e-05,
+      "loss": 0.0768,
+      "step": 7730
+    },
+    {
+      "epoch": 10.87,
+      "learning_rate": 1.1579868007534827e-05,
+      "loss": 0.0746,
+      "step": 7740
+    },
+    {
+      "epoch": 10.88,
+      "learning_rate": 1.1296479367690348e-05,
+      "loss": 0.0797,
+      "step": 7750
+    },
+    {
+      "epoch": 10.9,
+      "learning_rate": 1.1016521274832436e-05,
+      "loss": 0.0776,
+      "step": 7760
+    },
+    {
+      "epoch": 10.91,
+      "learning_rate": 1.0739997752401531e-05,
+      "loss": 0.0773,
+      "step": 7770
+    },
+    {
+      "epoch": 10.93,
+      "learning_rate": 1.0466912774477754e-05,
+      "loss": 0.0775,
+      "step": 7780
+    },
+    {
+      "epoch": 10.94,
+      "learning_rate": 1.019727026572409e-05,
+      "loss": 0.0778,
+      "step": 7790
+    },
+    {
+      "epoch": 10.95,
+      "learning_rate": 9.931074101329651e-06,
+      "loss": 0.0759,
+      "step": 7800
+    },
+    {
+      "epoch": 10.97,
+      "learning_rate": 9.668328106954227e-06,
+      "loss": 0.0824,
+      "step": 7810
+    },
+    {
+      "epoch": 10.98,
+      "learning_rate": 9.409036058673075e-06,
+      "loss": 0.0774,
+      "step": 7820
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 9.153201682922968e-06,
+      "loss": 0.0799,
+      "step": 7830
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 8.900828656448284e-06,
+      "loss": 0.0733,
+      "step": 7840
+    },
+    {
+      "epoch": 11.02,
+      "learning_rate": 8.651920606248503e-06,
+      "loss": 0.0709,
+      "step": 7850
+    },
+    {
+      "epoch": 11.04,
+      "learning_rate": 8.406481109525771e-06,
+      "loss": 0.0715,
+      "step": 7860
+    },
+    {
+      "epoch": 11.05,
+      "learning_rate": 8.164513693633719e-06,
+      "loss": 0.0777,
+      "step": 7870
+    },
+    {
+      "epoch": 11.07,
+      "learning_rate": 7.926021836026698e-06,
+      "loss": 0.0702,
+      "step": 7880
+    },
+    {
+      "epoch": 11.08,
+      "learning_rate": 7.691008964209684e-06,
+      "loss": 0.0678,
+      "step": 7890
+    },
+    {
+      "epoch": 11.09,
+      "learning_rate": 7.459478455689339e-06,
+      "loss": 0.0668,
+      "step": 7900
+    },
+    {
+      "epoch": 11.11,
+      "learning_rate": 7.2314336379249725e-06,
+      "loss": 0.071,
+      "step": 7910
+    },
+    {
+      "epoch": 11.12,
+      "learning_rate": 7.006877788281268e-06,
+      "loss": 0.0657,
+      "step": 7920
+    },
+    {
+      "epoch": 11.14,
+      "learning_rate": 6.7858141339807445e-06,
+      "loss": 0.0726,
+      "step": 7930
+    },
+    {
+      "epoch": 11.15,
+      "learning_rate": 6.568245852057675e-06,
+      "loss": 0.0718,
+      "step": 7940
+    },
+    {
+      "epoch": 11.16,
+      "learning_rate": 6.35417606931224e-06,
+      "loss": 0.0714,
+      "step": 7950
+    },
+    {
+      "epoch": 11.18,
+      "learning_rate": 6.143607862265616e-06,
+      "loss": 0.0699,
+      "step": 7960
+    },
+    {
+      "epoch": 11.19,
+      "learning_rate": 5.936544257115928e-06,
+      "loss": 0.0698,
+      "step": 7970
+    },
+    {
+      "epoch": 11.21,
+      "learning_rate": 5.73298822969448e-06,
+      "loss": 0.0675,
+      "step": 7980
+    },
+    {
+      "epoch": 11.22,
+      "learning_rate": 5.532942705423233e-06,
+      "loss": 0.0757,
+      "step": 7990
+    },
+    {
+      "epoch": 11.23,
+      "learning_rate": 5.336410559272669e-06,
+      "loss": 0.0703,
+      "step": 8000
+    },
+    {
+      "epoch": 11.25,
+      "learning_rate": 5.14339461572047e-06,
+      "loss": 0.068,
+      "step": 8010
+    },
+    {
+      "epoch": 11.26,
+      "learning_rate": 4.953897648710848e-06,
+      "loss": 0.0719,
+      "step": 8020
+    },
+    {
+      "epoch": 11.28,
+      "learning_rate": 4.767922381614859e-06,
+      "loss": 0.0635,
+      "step": 8030
+    },
+    {
+      "epoch": 11.29,
+      "learning_rate": 4.585471487191101e-06,
+      "loss": 0.0734,
+      "step": 8040
+    },
+    {
+      "epoch": 11.3,
+      "learning_rate": 4.40654758754741e-06,
+      "loss": 0.0723,
+      "step": 8050
+    },
+    {
+      "epoch": 11.32,
+      "learning_rate": 4.23115325410306e-06,
+      "loss": 0.0723,
+      "step": 8060
+    },
+    {
+      "epoch": 11.33,
+      "learning_rate": 4.0592910075519795e-06,
+      "loss": 0.0717,
+      "step": 8070
+    },
+    {
+      "epoch": 11.35,
+      "learning_rate": 3.890963317826346e-06,
+      "loss": 0.0685,
+      "step": 8080
+    },
+    {
+      "epoch": 11.36,
+      "learning_rate": 3.7261726040611623e-06,
+      "loss": 0.0746,
+      "step": 8090
+    },
+    {
+      "epoch": 11.37,
+      "learning_rate": 3.5649212345595374e-06,
+      "loss": 0.0722,
+      "step": 8100
+    },
+    {
+      "epoch": 11.39,
+      "learning_rate": 3.40721152675863e-06,
+      "loss": 0.0729,
+      "step": 8110
+    },
+    {
+      "epoch": 11.4,
+      "learning_rate": 3.2530457471962593e-06,
+      "loss": 0.0738,
+      "step": 8120
+    },
+    {
+      "epoch": 11.42,
+      "learning_rate": 3.1024261114784013e-06,
+      "loss": 0.0728,
+      "step": 8130
+    },
+    {
+      "epoch": 11.43,
+      "learning_rate": 2.9553547842473827e-06,
+      "loss": 0.0702,
+      "step": 8140
+    },
+    {
+      "epoch": 11.45,
+      "learning_rate": 2.8118338791506546e-06,
+      "loss": 0.0735,
+      "step": 8150
+    },
+    {
+      "epoch": 11.46,
+      "learning_rate": 2.671865458810596e-06,
+      "loss": 0.075,
+      "step": 8160
+    },
+    {
+      "epoch": 11.47,
+      "learning_rate": 2.5354515347946193e-06,
+      "loss": 0.0712,
+      "step": 8170
+    },
+    {
+      "epoch": 11.49,
+      "learning_rate": 2.4025940675865833e-06,
+      "loss": 0.0754,
+      "step": 8180
+    },
+    {
+      "epoch": 11.5,
+      "learning_rate": 2.273294966558287e-06,
+      "loss": 0.0719,
+      "step": 8190
+    },
+    {
+      "epoch": 11.52,
+      "learning_rate": 2.1475560899421876e-06,
+      "loss": 0.0697,
+      "step": 8200
+    },
+    {
+      "epoch": 11.53,
+      "learning_rate": 2.025379244804837e-06,
+      "loss": 0.0702,
+      "step": 8210
+    },
+    {
+      "epoch": 11.54,
+      "learning_rate": 1.9067661870205977e-06,
+      "loss": 0.0743,
+      "step": 8220
+    },
+    {
+      "epoch": 11.56,
+      "learning_rate": 1.791718621246663e-06,
+      "loss": 0.0733,
+      "step": 8230
+    },
+    {
+      "epoch": 11.57,
+      "learning_rate": 1.6802382008984374e-06,
+      "loss": 0.0723,
+      "step": 8240
+    },
+    {
+      "epoch": 11.59,
+      "learning_rate": 1.5723265281258336e-06,
+      "loss": 0.0741,
+      "step": 8250
+    },
+    {
+      "epoch": 11.6,
+      "learning_rate": 1.46798515379018e-06,
+      "loss": 0.0709,
+      "step": 8260
+    },
+    {
+      "epoch": 11.61,
+      "learning_rate": 1.367215577441988e-06,
+      "loss": 0.0726,
+      "step": 8270
+    },
+    {
+      "epoch": 11.63,
+      "learning_rate": 1.2700192472993865e-06,
+      "loss": 0.073,
+      "step": 8280
+    },
+    {
+      "epoch": 11.64,
+      "learning_rate": 1.1763975602273325e-06,
+      "loss": 0.0671,
+      "step": 8290
+    },
+    {
+      "epoch": 11.66,
+      "learning_rate": 1.0863518617174884e-06,
+      "loss": 0.0708,
+      "step": 8300
+    },
+    {
+      "epoch": 11.67,
+      "learning_rate": 9.998834458688488e-07,
+      "loss": 0.0771,
+      "step": 8310
+    },
+    {
+      "epoch": 11.68,
+      "learning_rate": 9.169935553693388e-07,
+      "loss": 0.0728,
+      "step": 8320
+    },
+    {
+      "epoch": 11.7,
+      "learning_rate": 8.376833814776886e-07,
+      "loss": 0.0686,
+      "step": 8330
+    },
+    {
+      "epoch": 11.71,
+      "learning_rate": 7.61954064006476e-07,
+      "loss": 0.0692,
+      "step": 8340
+    },
+    {
+      "epoch": 11.73,
+      "learning_rate": 6.898066913057777e-07,
+      "loss": 0.0742,
+      "step": 8350
+    },
+    {
+      "epoch": 11.74,
+      "learning_rate": 6.212423002473489e-07,
+      "loss": 0.0728,
+      "step": 8360
+    },
+    {
+      "epoch": 11.75,
+      "learning_rate": 5.562618762099125e-07,
+      "loss": 0.0667,
+      "step": 8370
+    },
+    {
+      "epoch": 11.77,
+      "learning_rate": 4.948663530648934e-07,
+      "loss": 0.0659,
+      "step": 8380
+    },
+    {
+      "epoch": 11.78,
+      "learning_rate": 4.3705661316303936e-07,
+      "loss": 0.0727,
+      "step": 8390
+    },
+    {
+      "epoch": 11.8,
+      "learning_rate": 3.828334873217099e-07,
+      "loss": 0.0708,
+      "step": 8400
+    },
+    {
+      "epoch": 11.81,
+      "learning_rate": 3.321977548129407e-07,
+      "loss": 0.075,
+      "step": 8410
+    },
+    {
+      "epoch": 11.82,
+      "learning_rate": 2.8515014335225853e-07,
+      "loss": 0.068,
+      "step": 8420
+    },
+    {
+      "epoch": 11.84,
+      "learning_rate": 2.41691329088245e-07,
+      "loss": 0.0689,
+      "step": 8430
+    },
+    {
+      "epoch": 11.85,
+      "learning_rate": 2.0182193659273873e-07,
+      "loss": 0.0781,
+      "step": 8440
+    },
+    {
+      "epoch": 11.87,
+      "learning_rate": 1.6554253885198156e-07,
+      "loss": 0.071,
+      "step": 8450
+    },
+    {
+      "epoch": 11.88,
+      "learning_rate": 1.3285365725823616e-07,
+      "loss": 0.0711,
+      "step": 8460
+    },
+    {
+      "epoch": 11.89,
+      "learning_rate": 1.037557616024587e-07,
+      "loss": 0.0717,
+      "step": 8470
+    },
+    {
+      "epoch": 11.91,
+      "learning_rate": 7.824927006735982e-08,
+      "loss": 0.0701,
+      "step": 8480
+    },
+    {
+      "epoch": 11.92,
+      "learning_rate": 5.6334549221603856e-08,
+      "loss": 0.0737,
+      "step": 8490
+    },
+    {
+      "epoch": 11.94,
+      "learning_rate": 3.801191401436865e-08,
+      "loss": 0.0725,
+      "step": 8500
+    },
+    {
+      "epoch": 11.95,
+      "learning_rate": 2.328162777084919e-08,
+      "loss": 0.0717,
+      "step": 8510
+    },
+    {
+      "epoch": 11.96,
+      "learning_rate": 1.2143902188593847e-08,
+      "loss": 0.0729,
+      "step": 8520
+    },
+    {
+      "epoch": 11.98,
+      "learning_rate": 4.598897334284757e-09,
+      "loss": 0.0685,
+      "step": 8530
+    },
+    {
+      "epoch": 11.99,
+      "learning_rate": 6.467216416006227e-10,
+      "loss": 0.0697,
+      "step": 8540
+    },
+    {
+      "epoch": 12.0,
+      "step": 8544,
+      "total_flos": 2.7233594633118286e+19,
+      "train_loss": 0.2806222443381666,
+      "train_runtime": 216583.3872,
+      "train_samples_per_second": 5.05,
+      "train_steps_per_second": 0.039
+    }
+  ],
+  "max_steps": 8544,
+  "num_train_epochs": 12,
+  "total_flos": 2.7233594633118286e+19,
+  "trial_name": null,
+  "trial_params": null
+}