diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,8 +1,8 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.9993337774816788,
-  "global_step": 10500,
+  "epoch": 1.9986675549633577,
+  "global_step": 21000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -6306,11 +6306,6318 @@
       "learning_rate": 6.69011170872983e-05,
       "loss": 1.5428,
       "step": 10500
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.5751161575317383,
+      "eval_runtime": 4381.6742,
+      "eval_samples_per_second": 2.433,
+      "step": 10507
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.686929123834378e-05,
+      "loss": 1.3253,
+      "step": 10510
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.683746538938927e-05,
+      "loss": 1.2831,
+      "step": 10520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.680563954043474e-05,
+      "loss": 1.4592,
+      "step": 10530
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.677381369148022e-05,
+      "loss": 1.3726,
+      "step": 10540
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.67419878425257e-05,
+      "loss": 1.3083,
+      "step": 10550
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.671016199357119e-05,
+      "loss": 1.1417,
+      "step": 10560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.667833614461666e-05,
+      "loss": 1.8066,
+      "step": 10570
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.664651029566213e-05,
+      "loss": 1.3587,
+      "step": 10580
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.661468444670763e-05,
+      "loss": 1.5318,
+      "step": 10590
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.65828585977531e-05,
+      "loss": 1.3035,
+      "step": 10600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.655103274879858e-05,
+      "loss": 1.2573,
+      "step": 10610
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.651920689984405e-05,
+      "loss": 1.2519,
+      "step": 10620
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.648738105088955e-05,
+      "loss": 1.0242,
+      "step": 10630
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.645555520193502e-05,
+      "loss": 1.0862,
+      "step": 10640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.642372935298049e-05,
+      "loss": 1.3527,
+      "step": 10650
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.639190350402597e-05,
+      "loss": 1.16,
+      "step": 10660
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.636007765507145e-05,
+      "loss": 1.5568,
+      "step": 10670
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.632825180611694e-05,
+      "loss": 1.2472,
+      "step": 10680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.629642595716241e-05,
+      "loss": 1.3669,
+      "step": 10690
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.626460010820789e-05,
+      "loss": 0.9845,
+      "step": 10700
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.623277425925338e-05,
+      "loss": 1.1218,
+      "step": 10710
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.620094841029885e-05,
+      "loss": 1.4078,
+      "step": 10720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.616912256134433e-05,
+      "loss": 1.2828,
+      "step": 10730
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.613729671238981e-05,
+      "loss": 1.3693,
+      "step": 10740
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.610547086343528e-05,
+      "loss": 1.4504,
+      "step": 10750
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.607364501448077e-05,
+      "loss": 1.2069,
+      "step": 10760
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 6.604181916552624e-05,
+      "loss": 1.3613,
+      "step": 10770
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 6.600999331657173e-05,
+      "loss": 1.4412,
+      "step": 10780
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 6.59781674676172e-05,
+      "loss": 1.4046,
+      "step": 10790
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 6.594634161866267e-05,
+      "loss": 1.2768,
+      "step": 10800
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 6.591451576970816e-05,
+      "loss": 1.3946,
+      "step": 10810
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 6.588268992075364e-05,
+      "loss": 1.034,
+      "step": 10820
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 6.585086407179913e-05,
+      "loss": 1.1968,
+      "step": 10830
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 6.58190382228446e-05,
+      "loss": 1.0514,
+      "step": 10840
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 6.578721237389008e-05,
+      "loss": 1.3141,
+      "step": 10850
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 6.575538652493556e-05,
+      "loss": 1.0505,
+      "step": 10860
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 6.572356067598103e-05,
+      "loss": 1.3938,
+      "step": 10870
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 6.569173482702652e-05,
+      "loss": 1.2018,
+      "step": 10880
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 6.5659908978072e-05,
+      "loss": 1.5693,
+      "step": 10890
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 6.562808312911747e-05,
+      "loss": 1.089,
+      "step": 10900
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 6.559625728016295e-05,
+      "loss": 1.2522,
+      "step": 10910
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 6.556443143120842e-05,
+      "loss": 1.0244,
+      "step": 10920
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 6.553260558225391e-05,
+      "loss": 1.6794,
+      "step": 10930
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 6.550077973329939e-05,
+      "loss": 1.3156,
+      "step": 10940
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 6.546895388434486e-05,
+      "loss": 1.3705,
+      "step": 10950
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 6.543712803539034e-05,
+      "loss": 1.2222,
+      "step": 10960
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 6.540530218643583e-05,
+      "loss": 1.2035,
+      "step": 10970
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 6.537347633748131e-05,
+      "loss": 1.3753,
+      "step": 10980
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 6.534165048852678e-05,
+      "loss": 1.3482,
+      "step": 10990
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 6.530982463957225e-05,
+      "loss": 1.6365,
+      "step": 11000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 6.527799879061775e-05,
+      "loss": 1.4579,
+      "step": 11010
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 6.524617294166322e-05,
+      "loss": 1.389,
+      "step": 11020
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 6.52143470927087e-05,
+      "loss": 1.7417,
+      "step": 11030
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 6.518252124375417e-05,
+      "loss": 1.1974,
+      "step": 11040
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 6.515069539479966e-05,
+      "loss": 1.281,
+      "step": 11050
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 6.511886954584514e-05,
+      "loss": 1.4675,
+      "step": 11060
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 6.508704369689061e-05,
+      "loss": 1.2103,
+      "step": 11070
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 6.50552178479361e-05,
+      "loss": 1.2186,
+      "step": 11080
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 6.502339199898158e-05,
+      "loss": 1.4346,
+      "step": 11090
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 6.499156615002705e-05,
+      "loss": 1.4353,
+      "step": 11100
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 6.495974030107253e-05,
+      "loss": 0.9587,
+      "step": 11110
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 6.492791445211802e-05,
+      "loss": 1.2586,
+      "step": 11120
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 6.48960886031635e-05,
+      "loss": 1.3769,
+      "step": 11130
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 6.486426275420897e-05,
+      "loss": 1.0935,
+      "step": 11140
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 6.483243690525444e-05,
+      "loss": 1.1413,
+      "step": 11150
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 6.480061105629994e-05,
+      "loss": 1.4108,
+      "step": 11160
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 6.47687852073454e-05,
+      "loss": 1.5105,
+      "step": 11170
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 6.473695935839089e-05,
+      "loss": 1.3528,
+      "step": 11180
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 6.470513350943636e-05,
+      "loss": 1.1827,
+      "step": 11190
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 6.467330766048184e-05,
+      "loss": 1.6177,
+      "step": 11200
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 6.464148181152733e-05,
+      "loss": 1.6265,
+      "step": 11210
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 6.46096559625728e-05,
+      "loss": 1.4097,
+      "step": 11220
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 6.457783011361828e-05,
+      "loss": 1.5653,
+      "step": 11230
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 6.454600426466376e-05,
+      "loss": 1.1577,
+      "step": 11240
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 6.451417841570923e-05,
+      "loss": 1.1006,
+      "step": 11250
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 6.448235256675472e-05,
+      "loss": 1.163,
+      "step": 11260
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 6.44505267178002e-05,
+      "loss": 1.4222,
+      "step": 11270
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 6.441870086884569e-05,
+      "loss": 1.6539,
+      "step": 11280
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 6.438687501989116e-05,
+      "loss": 1.5317,
+      "step": 11290
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 6.435504917093664e-05,
+      "loss": 1.4867,
+      "step": 11300
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 6.432322332198212e-05,
+      "loss": 1.577,
+      "step": 11310
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 6.42913974730276e-05,
+      "loss": 1.3068,
+      "step": 11320
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 6.425957162407308e-05,
+      "loss": 1.3321,
+      "step": 11330
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 6.422774577511855e-05,
+      "loss": 1.2002,
+      "step": 11340
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 6.419591992616404e-05,
+      "loss": 1.3479,
+      "step": 11350
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 6.416409407720951e-05,
+      "loss": 1.4076,
+      "step": 11360
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 6.413226822825498e-05,
+      "loss": 1.4901,
+      "step": 11370
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 6.410044237930047e-05,
+      "loss": 0.9651,
+      "step": 11380
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 6.406861653034595e-05,
+      "loss": 1.462,
+      "step": 11390
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 6.403679068139144e-05,
+      "loss": 1.3459,
+      "step": 11400
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 6.40049648324369e-05,
+      "loss": 1.1993,
+      "step": 11410
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 6.397313898348239e-05,
+      "loss": 1.4077,
+      "step": 11420
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 6.394131313452787e-05,
+      "loss": 1.0938,
+      "step": 11430
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 6.390948728557334e-05,
+      "loss": 1.4993,
+      "step": 11440
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 6.387766143661883e-05,
+      "loss": 1.31,
+      "step": 11450
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 6.384583558766431e-05,
+      "loss": 1.5857,
+      "step": 11460
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 6.381400973870978e-05,
+      "loss": 1.3809,
+      "step": 11470
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 6.378218388975526e-05,
+      "loss": 1.6283,
+      "step": 11480
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 6.375035804080073e-05,
+      "loss": 1.2548,
+      "step": 11490
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 6.371853219184623e-05,
+      "loss": 1.2578,
+      "step": 11500
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 6.36867063428917e-05,
+      "loss": 1.3211,
+      "step": 11510
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 6.365488049393717e-05,
+      "loss": 1.1977,
+      "step": 11520
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 6.362305464498265e-05,
+      "loss": 1.2643,
+      "step": 11530
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 6.359122879602814e-05,
+      "loss": 1.3753,
+      "step": 11540
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 6.355940294707362e-05,
+      "loss": 1.3017,
+      "step": 11550
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 6.352757709811909e-05,
+      "loss": 1.3721,
+      "step": 11560
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 6.349575124916458e-05,
+      "loss": 1.6016,
+      "step": 11570
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 6.346392540021006e-05,
+      "loss": 1.2166,
+      "step": 11580
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 6.343209955125553e-05,
+      "loss": 1.5257,
+      "step": 11590
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 6.340027370230101e-05,
+      "loss": 1.5632,
+      "step": 11600
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 6.33684478533465e-05,
+      "loss": 1.3745,
+      "step": 11610
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 6.333662200439197e-05,
+      "loss": 1.5288,
+      "step": 11620
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 6.330479615543745e-05,
+      "loss": 1.2667,
+      "step": 11630
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 6.327297030648292e-05,
+      "loss": 1.5822,
+      "step": 11640
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 6.324114445752842e-05,
+      "loss": 1.0847,
+      "step": 11650
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 6.320931860857389e-05,
+      "loss": 1.503,
+      "step": 11660
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 6.317749275961936e-05,
+      "loss": 1.4796,
+      "step": 11670
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 6.314566691066484e-05,
+      "loss": 1.1655,
+      "step": 11680
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 6.311384106171033e-05,
+      "loss": 1.213,
+      "step": 11690
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 6.308201521275581e-05,
+      "loss": 1.5796,
+      "step": 11700
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 6.305018936380128e-05,
+      "loss": 1.1776,
+      "step": 11710
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 6.301836351484676e-05,
+      "loss": 1.2745,
+      "step": 11720
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 6.298653766589225e-05,
+      "loss": 1.5491,
+      "step": 11730
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 6.295471181693772e-05,
+      "loss": 1.3805,
+      "step": 11740
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 6.29228859679832e-05,
+      "loss": 1.4238,
+      "step": 11750
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 6.289106011902868e-05,
+      "loss": 1.3267,
+      "step": 11760
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 6.285923427007415e-05,
+      "loss": 1.2011,
+      "step": 11770
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 6.282740842111964e-05,
+      "loss": 1.2998,
+      "step": 11780
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 6.279558257216511e-05,
+      "loss": 1.0914,
+      "step": 11790
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 6.27637567232106e-05,
+      "loss": 1.4221,
+      "step": 11800
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 6.273193087425607e-05,
+      "loss": 1.4707,
+      "step": 11810
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 6.270010502530154e-05,
+      "loss": 1.2843,
+      "step": 11820
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 6.266827917634703e-05,
+      "loss": 1.5337,
+      "step": 11830
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 6.263645332739251e-05,
+      "loss": 1.3396,
+      "step": 11840
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 6.2604627478438e-05,
+      "loss": 1.0909,
+      "step": 11850
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 6.257280162948347e-05,
+      "loss": 1.4662,
+      "step": 11860
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 6.254097578052895e-05,
+      "loss": 1.1194,
+      "step": 11870
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 6.250914993157443e-05,
+      "loss": 1.2887,
+      "step": 11880
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 6.24773240826199e-05,
+      "loss": 1.4379,
+      "step": 11890
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 6.244549823366539e-05,
+      "loss": 1.2255,
+      "step": 11900
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 6.241367238471087e-05,
+      "loss": 1.5839,
+      "step": 11910
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 6.238184653575635e-05,
+      "loss": 1.4361,
+      "step": 11920
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 6.235002068680182e-05,
+      "loss": 1.1001,
+      "step": 11930
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 6.23181948378473e-05,
+      "loss": 1.5917,
+      "step": 11940
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 6.228636898889279e-05,
+      "loss": 1.1532,
+      "step": 11950
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 6.225454313993826e-05,
+      "loss": 1.3314,
+      "step": 11960
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 6.222271729098375e-05,
+      "loss": 1.1664,
+      "step": 11970
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 6.219089144202922e-05,
+      "loss": 1.1507,
+      "step": 11980
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 6.21590655930747e-05,
+      "loss": 1.3231,
+      "step": 11990
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 6.212723974412018e-05,
+      "loss": 1.401,
+      "step": 12000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 6.209541389516565e-05,
+      "loss": 1.148,
+      "step": 12010
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 6.206358804621114e-05,
+      "loss": 1.3955,
+      "step": 12020
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 6.203176219725662e-05,
+      "loss": 1.714,
+      "step": 12030
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 6.199993634830209e-05,
+      "loss": 1.3327,
+      "step": 12040
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 6.196811049934757e-05,
+      "loss": 1.3942,
+      "step": 12050
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 6.193628465039304e-05,
+      "loss": 1.2902,
+      "step": 12060
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 6.190445880143854e-05,
+      "loss": 1.2024,
+      "step": 12070
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 6.187263295248401e-05,
+      "loss": 1.6921,
+      "step": 12080
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 6.184080710352948e-05,
+      "loss": 1.5338,
+      "step": 12090
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 6.180898125457496e-05,
+      "loss": 1.3473,
+      "step": 12100
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 6.177715540562045e-05,
+      "loss": 1.1758,
+      "step": 12110
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 6.174532955666593e-05,
+      "loss": 0.9934,
+      "step": 12120
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 6.17135037077114e-05,
+      "loss": 1.0722,
+      "step": 12130
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 6.168167785875689e-05,
+      "loss": 1.5355,
+      "step": 12140
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 6.164985200980237e-05,
+      "loss": 1.1969,
+      "step": 12150
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 6.161802616084784e-05,
+      "loss": 1.2552,
+      "step": 12160
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 6.158620031189332e-05,
+      "loss": 1.4347,
+      "step": 12170
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 6.155437446293881e-05,
+      "loss": 1.2723,
+      "step": 12180
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 6.152254861398428e-05,
+      "loss": 1.3817,
+      "step": 12190
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 6.149072276502976e-05,
+      "loss": 1.4304,
+      "step": 12200
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 6.145889691607523e-05,
+      "loss": 1.3085,
+      "step": 12210
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 6.142707106712073e-05,
+      "loss": 1.4066,
+      "step": 12220
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 6.13952452181662e-05,
+      "loss": 1.2759,
+      "step": 12230
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 6.136341936921167e-05,
+      "loss": 1.3005,
+      "step": 12240
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 6.133159352025715e-05,
+      "loss": 1.4923,
+      "step": 12250
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 6.129976767130264e-05,
+      "loss": 1.2941,
+      "step": 12260
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 6.126794182234812e-05,
+      "loss": 1.188,
+      "step": 12270
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 6.123611597339359e-05,
+      "loss": 1.515,
+      "step": 12280
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 6.120429012443907e-05,
+      "loss": 1.1593,
+      "step": 12290
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 6.117246427548456e-05,
+      "loss": 1.1917,
+      "step": 12300
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 6.114063842653003e-05,
+      "loss": 1.4694,
+      "step": 12310
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 6.110881257757551e-05,
+      "loss": 1.3641,
+      "step": 12320
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 6.1076986728621e-05,
+      "loss": 1.0053,
+      "step": 12330
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 6.104516087966646e-05,
+      "loss": 1.3955,
+      "step": 12340
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 6.101333503071195e-05,
+      "loss": 1.4972,
+      "step": 12350
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 6.0981509181757424e-05,
+      "loss": 1.5457,
+      "step": 12360
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 6.094968333280291e-05,
+      "loss": 1.295,
+      "step": 12370
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 6.0917857483848385e-05,
+      "loss": 1.3848,
+      "step": 12380
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 6.088603163489386e-05,
+      "loss": 1.3321,
+      "step": 12390
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 6.0854205785939345e-05,
+      "loss": 1.5353,
+      "step": 12400
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 6.082237993698482e-05,
+      "loss": 1.2588,
+      "step": 12410
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 6.0790554088030306e-05,
+      "loss": 1.274,
+      "step": 12420
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 6.075872823907578e-05,
+      "loss": 1.1487,
+      "step": 12430
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 6.072690239012125e-05,
+      "loss": 1.2118,
+      "step": 12440
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 6.069507654116674e-05,
+      "loss": 1.1819,
+      "step": 12450
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 6.066325069221221e-05,
+      "loss": 1.4046,
+      "step": 12460
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 6.0631424843257704e-05,
+      "loss": 1.4242,
+      "step": 12470
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 6.0599598994303174e-05,
+      "loss": 1.4458,
+      "step": 12480
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 6.056777314534865e-05,
+      "loss": 1.7952,
+      "step": 12490
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 6.0535947296394134e-05,
+      "loss": 1.6093,
+      "step": 12500
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 6.050412144743961e-05,
+      "loss": 1.3165,
+      "step": 12510
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 6.0472295598485095e-05,
+      "loss": 1.1477,
+      "step": 12520
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 6.044046974953057e-05,
+      "loss": 1.3023,
+      "step": 12530
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 6.0408643900576055e-05,
+      "loss": 1.1291,
+      "step": 12540
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 6.037681805162153e-05,
+      "loss": 1.1525,
+      "step": 12550
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 6.0344992202667e-05,
+      "loss": 1.4392,
+      "step": 12560
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 6.031316635371249e-05,
+      "loss": 1.2571,
+      "step": 12570
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 6.028134050475796e-05,
+      "loss": 1.4243,
+      "step": 12580
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 6.024951465580345e-05,
+      "loss": 1.3288,
+      "step": 12590
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 6.021768880684892e-05,
+      "loss": 1.2786,
+      "step": 12600
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 6.01858629578944e-05,
+      "loss": 1.4158,
+      "step": 12610
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 6.0154037108939884e-05,
+      "loss": 1.6776,
+      "step": 12620
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 6.012221125998536e-05,
+      "loss": 1.2973,
+      "step": 12630
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 6.0090385411030844e-05,
+      "loss": 1.6589,
+      "step": 12640
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 6.005855956207632e-05,
+      "loss": 1.4143,
+      "step": 12650
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 6.00267337131218e-05,
+      "loss": 1.1845,
+      "step": 12660
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.999490786416728e-05,
+      "loss": 1.379,
+      "step": 12670
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.996308201521276e-05,
+      "loss": 1.1948,
+      "step": 12680
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.993125616625824e-05,
+      "loss": 1.4856,
+      "step": 12690
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.989943031730372e-05,
+      "loss": 1.0585,
+      "step": 12700
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.986760446834919e-05,
+      "loss": 1.0362,
+      "step": 12710
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.983577861939468e-05,
+      "loss": 1.5063,
+      "step": 12720
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.980395277044015e-05,
+      "loss": 1.3054,
+      "step": 12730
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.977212692148564e-05,
+      "loss": 1.2419,
+      "step": 12740
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.974030107253111e-05,
+      "loss": 1.3128,
+      "step": 12750
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.970847522357659e-05,
+      "loss": 1.3978,
+      "step": 12760
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.967664937462207e-05,
+      "loss": 1.3241,
+      "step": 12770
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.964482352566755e-05,
+      "loss": 1.3458,
+      "step": 12780
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.961299767671303e-05,
+      "loss": 1.1913,
+      "step": 12790
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.958117182775851e-05,
+      "loss": 1.207,
+      "step": 12800
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.9549345978803985e-05,
+      "loss": 1.8335,
+      "step": 12810
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.951752012984947e-05,
+      "loss": 1.2501,
+      "step": 12820
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.9485694280894945e-05,
+      "loss": 1.3696,
+      "step": 12830
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.945386843194043e-05,
+      "loss": 1.4096,
+      "step": 12840
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.9422042582985906e-05,
+      "loss": 1.2826,
+      "step": 12850
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.9390216734031376e-05,
+      "loss": 1.781,
+      "step": 12860
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.9358390885076866e-05,
+      "loss": 1.2511,
+      "step": 12870
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.9326565036122336e-05,
+      "loss": 1.2594,
+      "step": 12880
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.929473918716783e-05,
+      "loss": 1.2174,
+      "step": 12890
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.92629133382133e-05,
+      "loss": 1.3863,
+      "step": 12900
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.9231087489258774e-05,
+      "loss": 1.5468,
+      "step": 12910
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.919926164030426e-05,
+      "loss": 1.3738,
+      "step": 12920
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.9167435791349734e-05,
+      "loss": 1.2275,
+      "step": 12930
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.913560994239522e-05,
+      "loss": 1.2463,
+      "step": 12940
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.9103784093440695e-05,
+      "loss": 1.5187,
+      "step": 12950
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.907195824448617e-05,
+      "loss": 1.2007,
+      "step": 12960
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.9040132395531655e-05,
+      "loss": 1.5909,
+      "step": 12970
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.900830654657713e-05,
+      "loss": 1.1711,
+      "step": 12980
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.8976480697622616e-05,
+      "loss": 1.498,
+      "step": 12990
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.894465484866809e-05,
+      "loss": 1.2641,
+      "step": 13000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.891282899971356e-05,
+      "loss": 1.4552,
+      "step": 13010
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.888100315075905e-05,
+      "loss": 1.4062,
+      "step": 13020
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.884917730180452e-05,
+      "loss": 1.6072,
+      "step": 13030
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.8817351452850014e-05,
+      "loss": 1.5844,
+      "step": 13040
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.8785525603895484e-05,
+      "loss": 1.3093,
+      "step": 13050
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.875369975494096e-05,
+      "loss": 1.3364,
+      "step": 13060
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.8721873905986444e-05,
+      "loss": 1.2106,
+      "step": 13070
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.869004805703192e-05,
+      "loss": 1.4542,
+      "step": 13080
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.8658222208077405e-05,
+      "loss": 1.3747,
+      "step": 13090
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.862639635912288e-05,
+      "loss": 1.4218,
+      "step": 13100
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.859457051016836e-05,
+      "loss": 1.3271,
+      "step": 13110
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.856274466121384e-05,
+      "loss": 1.3049,
+      "step": 13120
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.853091881225932e-05,
+      "loss": 1.1618,
+      "step": 13130
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.84990929633048e-05,
+      "loss": 1.5606,
+      "step": 13140
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.846726711435028e-05,
+      "loss": 1.5192,
+      "step": 13150
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.843544126539575e-05,
+      "loss": 1.2331,
+      "step": 13160
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.840361541644124e-05,
+      "loss": 1.3765,
+      "step": 13170
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.837178956748671e-05,
+      "loss": 1.3304,
+      "step": 13180
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.83399637185322e-05,
+      "loss": 1.0399,
+      "step": 13190
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.830813786957767e-05,
+      "loss": 1.4977,
+      "step": 13200
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.827631202062316e-05,
+      "loss": 1.2552,
+      "step": 13210
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.824448617166863e-05,
+      "loss": 1.1272,
+      "step": 13220
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.821266032271411e-05,
+      "loss": 1.686,
+      "step": 13230
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.818083447375959e-05,
+      "loss": 1.3021,
+      "step": 13240
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.814900862480507e-05,
+      "loss": 1.3472,
+      "step": 13250
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.811718277585055e-05,
+      "loss": 1.1018,
+      "step": 13260
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.808535692689603e-05,
+      "loss": 1.2079,
+      "step": 13270
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.8053531077941506e-05,
+      "loss": 1.205,
+      "step": 13280
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.802170522898699e-05,
+      "loss": 1.4772,
+      "step": 13290
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.7989879380032466e-05,
+      "loss": 1.1362,
+      "step": 13300
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.795805353107795e-05,
+      "loss": 1.5549,
+      "step": 13310
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.792622768212343e-05,
+      "loss": 1.2256,
+      "step": 13320
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.78944018331689e-05,
+      "loss": 1.1792,
+      "step": 13330
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.786257598421439e-05,
+      "loss": 1.444,
+      "step": 13340
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.783075013525986e-05,
+      "loss": 1.3135,
+      "step": 13350
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.779892428630535e-05,
+      "loss": 1.2954,
+      "step": 13360
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.776709843735082e-05,
+      "loss": 1.2679,
+      "step": 13370
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.7735272588396295e-05,
+      "loss": 1.3775,
+      "step": 13380
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.770344673944178e-05,
+      "loss": 1.0079,
+      "step": 13390
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.7671620890487255e-05,
+      "loss": 1.3256,
+      "step": 13400
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.763979504153274e-05,
+      "loss": 1.2643,
+      "step": 13410
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.7607969192578216e-05,
+      "loss": 1.0838,
+      "step": 13420
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.757614334362369e-05,
+      "loss": 1.1447,
+      "step": 13430
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.7544317494669176e-05,
+      "loss": 1.0348,
+      "step": 13440
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.751249164571465e-05,
+      "loss": 1.4069,
+      "step": 13450
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.748066579676014e-05,
+      "loss": 1.5608,
+      "step": 13460
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.7448839947805614e-05,
+      "loss": 1.0844,
+      "step": 13470
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.7417014098851084e-05,
+      "loss": 1.4903,
+      "step": 13480
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.7385188249896574e-05,
+      "loss": 1.2309,
+      "step": 13490
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.7353362400942044e-05,
+      "loss": 1.4884,
+      "step": 13500
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.7321536551987535e-05,
+      "loss": 1.4128,
+      "step": 13510
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.7289710703033005e-05,
+      "loss": 1.2433,
+      "step": 13520
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.725788485407848e-05,
+      "loss": 1.4097,
+      "step": 13530
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.7226059005123965e-05,
+      "loss": 1.2903,
+      "step": 13540
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.719423315616944e-05,
+      "loss": 1.2305,
+      "step": 13550
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.7162407307214926e-05,
+      "loss": 1.2742,
+      "step": 13560
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.71305814582604e-05,
+      "loss": 1.1847,
+      "step": 13570
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.709875560930588e-05,
+      "loss": 1.1794,
+      "step": 13580
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.706692976035136e-05,
+      "loss": 1.4876,
+      "step": 13590
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.703510391139684e-05,
+      "loss": 1.3414,
+      "step": 13600
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.7003278062442324e-05,
+      "loss": 1.6806,
+      "step": 13610
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.6971452213487794e-05,
+      "loss": 1.167,
+      "step": 13620
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.693962636453327e-05,
+      "loss": 1.6015,
+      "step": 13630
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.6907800515578754e-05,
+      "loss": 1.3552,
+      "step": 13640
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.687597466662423e-05,
+      "loss": 1.4357,
+      "step": 13650
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.6844148817669715e-05,
+      "loss": 1.2554,
+      "step": 13660
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.681232296871519e-05,
+      "loss": 1.3044,
+      "step": 13670
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.678049711976067e-05,
+      "loss": 1.0156,
+      "step": 13680
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.674867127080615e-05,
+      "loss": 1.3119,
+      "step": 13690
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.671684542185163e-05,
+      "loss": 1.4552,
+      "step": 13700
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.668501957289711e-05,
+      "loss": 1.3558,
+      "step": 13710
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.665319372394259e-05,
+      "loss": 1.3309,
+      "step": 13720
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.662136787498806e-05,
+      "loss": 1.4877,
+      "step": 13730
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.658954202603355e-05,
+      "loss": 1.2664,
+      "step": 13740
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.655771617707902e-05,
+      "loss": 1.2099,
+      "step": 13750
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.652589032812451e-05,
+      "loss": 1.2328,
+      "step": 13760
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.649406447916998e-05,
+      "loss": 1.1204,
+      "step": 13770
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.646223863021546e-05,
+      "loss": 1.4109,
+      "step": 13780
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.643041278126094e-05,
+      "loss": 1.1479,
+      "step": 13790
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.639858693230642e-05,
+      "loss": 1.2953,
+      "step": 13800
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.63667610833519e-05,
+      "loss": 1.4705,
+      "step": 13810
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.633493523439738e-05,
+      "loss": 1.5887,
+      "step": 13820
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.630310938544286e-05,
+      "loss": 1.2498,
+      "step": 13830
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.627128353648834e-05,
+      "loss": 1.3203,
+      "step": 13840
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.6239457687533816e-05,
+      "loss": 1.6888,
+      "step": 13850
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.62076318385793e-05,
+      "loss": 1.2531,
+      "step": 13860
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.6175805989624776e-05,
+      "loss": 1.0861,
+      "step": 13870
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.614398014067026e-05,
+      "loss": 1.4684,
+      "step": 13880
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.611215429171574e-05,
+      "loss": 0.8508,
+      "step": 13890
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.608032844276121e-05,
+      "loss": 1.4018,
+      "step": 13900
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.60485025938067e-05,
+      "loss": 1.1384,
+      "step": 13910
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.601667674485217e-05,
+      "loss": 1.3172,
+      "step": 13920
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.598485089589766e-05,
+      "loss": 1.3609,
+      "step": 13930
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.595302504694313e-05,
+      "loss": 1.3021,
+      "step": 13940
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.5921199197988605e-05,
+      "loss": 1.4604,
+      "step": 13950
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.588937334903409e-05,
+      "loss": 1.4656,
+      "step": 13960
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.5857547500079565e-05,
+      "loss": 1.0608,
+      "step": 13970
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.582572165112505e-05,
+      "loss": 1.4819,
+      "step": 13980
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.5793895802170526e-05,
+      "loss": 1.3399,
+      "step": 13990
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.5762069953216e-05,
+      "loss": 1.1616,
+      "step": 14000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.5730244104261486e-05,
+      "loss": 1.2001,
+      "step": 14010
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.569841825530696e-05,
+      "loss": 1.5204,
+      "step": 14020
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.566659240635245e-05,
+      "loss": 1.2357,
+      "step": 14030
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.5634766557397924e-05,
+      "loss": 1.2601,
+      "step": 14040
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.5602940708443394e-05,
+      "loss": 1.4038,
+      "step": 14050
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.5571114859488884e-05,
+      "loss": 1.3067,
+      "step": 14060
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.5539289010534354e-05,
+      "loss": 1.269,
+      "step": 14070
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.5507463161579845e-05,
+      "loss": 1.3595,
+      "step": 14080
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.5475637312625315e-05,
+      "loss": 1.2674,
+      "step": 14090
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.544381146367079e-05,
+      "loss": 1.2473,
+      "step": 14100
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.5411985614716275e-05,
+      "loss": 1.5753,
+      "step": 14110
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.538015976576175e-05,
+      "loss": 1.4645,
+      "step": 14120
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.5348333916807236e-05,
+      "loss": 1.4572,
+      "step": 14130
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.531650806785271e-05,
+      "loss": 1.2548,
+      "step": 14140
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.528468221889819e-05,
+      "loss": 1.3171,
+      "step": 14150
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.525285636994367e-05,
+      "loss": 1.3876,
+      "step": 14160
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.522103052098915e-05,
+      "loss": 1.4725,
+      "step": 14170
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.5189204672034634e-05,
+      "loss": 1.2703,
+      "step": 14180
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.515737882308011e-05,
+      "loss": 1.2428,
+      "step": 14190
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.512555297412558e-05,
+      "loss": 1.5501,
+      "step": 14200
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.509372712517107e-05,
+      "loss": 1.4164,
+      "step": 14210
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.506190127621654e-05,
+      "loss": 1.4017,
+      "step": 14220
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.503007542726203e-05,
+      "loss": 1.2215,
+      "step": 14230
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.49982495783075e-05,
+      "loss": 1.1972,
+      "step": 14240
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.496642372935298e-05,
+      "loss": 1.4031,
+      "step": 14250
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.493459788039846e-05,
+      "loss": 1.1811,
+      "step": 14260
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.490277203144394e-05,
+      "loss": 1.2467,
+      "step": 14270
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.487094618248942e-05,
+      "loss": 1.4318,
+      "step": 14280
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.48391203335349e-05,
+      "loss": 1.3557,
+      "step": 14290
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.4807294484580376e-05,
+      "loss": 1.3901,
+      "step": 14300
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.477546863562586e-05,
+      "loss": 1.5302,
+      "step": 14310
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.474364278667134e-05,
+      "loss": 1.1885,
+      "step": 14320
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.471181693771682e-05,
+      "loss": 1.1578,
+      "step": 14330
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.46799910887623e-05,
+      "loss": 1.3881,
+      "step": 14340
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.464816523980777e-05,
+      "loss": 1.1401,
+      "step": 14350
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.461633939085326e-05,
+      "loss": 1.0311,
+      "step": 14360
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.458451354189873e-05,
+      "loss": 1.1228,
+      "step": 14370
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.455268769294422e-05,
+      "loss": 1.2884,
+      "step": 14380
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.452086184398969e-05,
+      "loss": 1.3827,
+      "step": 14390
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.4489035995035165e-05,
+      "loss": 1.239,
+      "step": 14400
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.445721014608065e-05,
+      "loss": 1.4542,
+      "step": 14410
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.4425384297126126e-05,
+      "loss": 1.0058,
+      "step": 14420
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.439355844817161e-05,
+      "loss": 1.3152,
+      "step": 14430
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.4361732599217086e-05,
+      "loss": 1.4563,
+      "step": 14440
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.432990675026257e-05,
+      "loss": 1.2788,
+      "step": 14450
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.429808090130805e-05,
+      "loss": 1.5673,
+      "step": 14460
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.4266255052353524e-05,
+      "loss": 1.5005,
+      "step": 14470
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.423442920339901e-05,
+      "loss": 1.1309,
+      "step": 14480
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.4202603354444484e-05,
+      "loss": 1.4121,
+      "step": 14490
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.417077750548997e-05,
+      "loss": 1.3966,
+      "step": 14500
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.4138951656535445e-05,
+      "loss": 1.1501,
+      "step": 14510
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.4107125807580915e-05,
+      "loss": 0.9098,
+      "step": 14520
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.4075299958626405e-05,
+      "loss": 1.123,
+      "step": 14530
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.4043474109671875e-05,
+      "loss": 1.2539,
+      "step": 14540
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.4011648260717366e-05,
+      "loss": 1.4045,
+      "step": 14550
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.3979822411762836e-05,
+      "loss": 1.6323,
+      "step": 14560
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.394799656280831e-05,
+      "loss": 1.3246,
+      "step": 14570
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.3916170713853796e-05,
+      "loss": 1.1879,
+      "step": 14580
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.388434486489927e-05,
+      "loss": 1.3357,
+      "step": 14590
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.385251901594476e-05,
+      "loss": 1.2795,
+      "step": 14600
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.3820693166990234e-05,
+      "loss": 1.521,
+      "step": 14610
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.378886731803571e-05,
+      "loss": 1.2611,
+      "step": 14620
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.3757041469081194e-05,
+      "loss": 1.4852,
+      "step": 14630
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.372521562012667e-05,
+      "loss": 0.9539,
+      "step": 14640
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.3693389771172155e-05,
+      "loss": 1.1787,
+      "step": 14650
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.366156392221763e-05,
+      "loss": 1.3916,
+      "step": 14660
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.36297380732631e-05,
+      "loss": 1.0306,
+      "step": 14670
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.3597912224308585e-05,
+      "loss": 1.3695,
+      "step": 14680
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.356608637535406e-05,
+      "loss": 1.2022,
+      "step": 14690
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.3534260526399546e-05,
+      "loss": 1.1744,
+      "step": 14700
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.350243467744502e-05,
+      "loss": 1.4006,
+      "step": 14710
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.34706088284905e-05,
+      "loss": 1.3549,
+      "step": 14720
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.343878297953598e-05,
+      "loss": 1.2094,
+      "step": 14730
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.340695713058146e-05,
+      "loss": 1.2815,
+      "step": 14740
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.3375131281626944e-05,
+      "loss": 1.1214,
+      "step": 14750
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.334330543267242e-05,
+      "loss": 1.5456,
+      "step": 14760
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.331147958371789e-05,
+      "loss": 1.3289,
+      "step": 14770
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.327965373476338e-05,
+      "loss": 1.5725,
+      "step": 14780
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.324782788580885e-05,
+      "loss": 1.3129,
+      "step": 14790
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.321600203685434e-05,
+      "loss": 1.1212,
+      "step": 14800
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.318417618789981e-05,
+      "loss": 1.5021,
+      "step": 14810
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.315235033894529e-05,
+      "loss": 1.2091,
+      "step": 14820
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.312052448999077e-05,
+      "loss": 1.3694,
+      "step": 14830
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.308869864103625e-05,
+      "loss": 1.2515,
+      "step": 14840
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.305687279208173e-05,
+      "loss": 1.2529,
+      "step": 14850
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.302504694312721e-05,
+      "loss": 1.2047,
+      "step": 14860
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.2993221094172686e-05,
+      "loss": 1.3786,
+      "step": 14870
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.296139524521817e-05,
+      "loss": 1.3992,
+      "step": 14880
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.292956939626365e-05,
+      "loss": 1.3026,
+      "step": 14890
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.289774354730913e-05,
+      "loss": 1.2121,
+      "step": 14900
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.286591769835461e-05,
+      "loss": 1.4507,
+      "step": 14910
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.283409184940008e-05,
+      "loss": 1.0681,
+      "step": 14920
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.280226600044557e-05,
+      "loss": 1.4707,
+      "step": 14930
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.277044015149104e-05,
+      "loss": 1.3271,
+      "step": 14940
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.273861430253653e-05,
+      "loss": 1.1235,
+      "step": 14950
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.2706788453582e-05,
+      "loss": 1.0133,
+      "step": 14960
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.2674962604627475e-05,
+      "loss": 1.2553,
+      "step": 14970
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.264313675567296e-05,
+      "loss": 1.2721,
+      "step": 14980
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.2611310906718436e-05,
+      "loss": 1.2123,
+      "step": 14990
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.257948505776392e-05,
+      "loss": 1.4009,
+      "step": 15000
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.2547659208809396e-05,
+      "loss": 1.2326,
+      "step": 15010
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.251583335985487e-05,
+      "loss": 1.3902,
+      "step": 15020
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.248400751090036e-05,
+      "loss": 1.4369,
+      "step": 15030
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.2452181661945834e-05,
+      "loss": 1.2699,
+      "step": 15040
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.242035581299132e-05,
+      "loss": 1.2985,
+      "step": 15050
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.2388529964036794e-05,
+      "loss": 1.5252,
+      "step": 15060
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.2356704115082264e-05,
+      "loss": 1.5001,
+      "step": 15070
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.2324878266127755e-05,
+      "loss": 1.4595,
+      "step": 15080
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.2293052417173225e-05,
+      "loss": 1.1815,
+      "step": 15090
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.2261226568218715e-05,
+      "loss": 1.4579,
+      "step": 15100
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.2229400719264185e-05,
+      "loss": 1.1008,
+      "step": 15110
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.2197574870309676e-05,
+      "loss": 1.2104,
+      "step": 15120
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.2165749021355146e-05,
+      "loss": 1.0013,
+      "step": 15130
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.213392317240062e-05,
+      "loss": 1.2941,
+      "step": 15140
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.2102097323446106e-05,
+      "loss": 1.2361,
+      "step": 15150
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.207027147449158e-05,
+      "loss": 1.1462,
+      "step": 15160
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.203844562553707e-05,
+      "loss": 1.2805,
+      "step": 15170
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.2006619776582544e-05,
+      "loss": 1.157,
+      "step": 15180
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.197479392762802e-05,
+      "loss": 1.2395,
+      "step": 15190
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.1942968078673504e-05,
+      "loss": 1.2716,
+      "step": 15200
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.191114222971898e-05,
+      "loss": 1.1911,
+      "step": 15210
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.1879316380764465e-05,
+      "loss": 1.2859,
+      "step": 15220
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.184749053180994e-05,
+      "loss": 0.9757,
+      "step": 15230
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.181566468285541e-05,
+      "loss": 1.1481,
+      "step": 15240
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.17838388339009e-05,
+      "loss": 1.3675,
+      "step": 15250
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.175201298494637e-05,
+      "loss": 1.1877,
+      "step": 15260
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.172018713599186e-05,
+      "loss": 1.2303,
+      "step": 15270
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.168836128703733e-05,
+      "loss": 1.6646,
+      "step": 15280
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.165653543808281e-05,
+      "loss": 1.4336,
+      "step": 15290
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.162470958912829e-05,
+      "loss": 1.612,
+      "step": 15300
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.159288374017377e-05,
+      "loss": 1.2497,
+      "step": 15310
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.1561057891219253e-05,
+      "loss": 1.2639,
+      "step": 15320
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.152923204226473e-05,
+      "loss": 1.3976,
+      "step": 15330
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.149740619331021e-05,
+      "loss": 1.5085,
+      "step": 15340
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.146558034435569e-05,
+      "loss": 1.5096,
+      "step": 15350
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.143375449540117e-05,
+      "loss": 1.6825,
+      "step": 15360
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.140192864644665e-05,
+      "loss": 1.265,
+      "step": 15370
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.137010279749213e-05,
+      "loss": 1.4147,
+      "step": 15380
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.13382769485376e-05,
+      "loss": 1.3444,
+      "step": 15390
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.130645109958309e-05,
+      "loss": 1.1047,
+      "step": 15400
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.127462525062856e-05,
+      "loss": 1.6598,
+      "step": 15410
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.124279940167405e-05,
+      "loss": 1.309,
+      "step": 15420
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.121097355271952e-05,
+      "loss": 1.148,
+      "step": 15430
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.1179147703764996e-05,
+      "loss": 1.2568,
+      "step": 15440
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.114732185481048e-05,
+      "loss": 1.3135,
+      "step": 15450
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.111549600585596e-05,
+      "loss": 1.2472,
+      "step": 15460
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.108367015690144e-05,
+      "loss": 1.3674,
+      "step": 15470
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.105184430794692e-05,
+      "loss": 1.3011,
+      "step": 15480
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.1020018458992394e-05,
+      "loss": 1.1521,
+      "step": 15490
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.098819261003788e-05,
+      "loss": 1.1709,
+      "step": 15500
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.0956366761083355e-05,
+      "loss": 1.3494,
+      "step": 15510
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.092454091212884e-05,
+      "loss": 1.2846,
+      "step": 15520
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.0892715063174315e-05,
+      "loss": 1.3508,
+      "step": 15530
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.0860889214219785e-05,
+      "loss": 1.3032,
+      "step": 15540
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.0829063365265276e-05,
+      "loss": 1.1961,
+      "step": 15550
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.0797237516310746e-05,
+      "loss": 1.4699,
+      "step": 15560
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.0765411667356236e-05,
+      "loss": 1.2985,
+      "step": 15570
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.0733585818401706e-05,
+      "loss": 1.2926,
+      "step": 15580
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.070175996944718e-05,
+      "loss": 1.3092,
+      "step": 15590
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.066993412049267e-05,
+      "loss": 1.4662,
+      "step": 15600
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.0638108271538144e-05,
+      "loss": 1.1928,
+      "step": 15610
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.060628242258363e-05,
+      "loss": 1.3266,
+      "step": 15620
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.0574456573629104e-05,
+      "loss": 1.2545,
+      "step": 15630
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.054263072467458e-05,
+      "loss": 1.3546,
+      "step": 15640
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.0510804875720065e-05,
+      "loss": 1.4485,
+      "step": 15650
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.047897902676554e-05,
+      "loss": 1.0811,
+      "step": 15660
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.0447153177811025e-05,
+      "loss": 1.256,
+      "step": 15670
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.04153273288565e-05,
+      "loss": 1.3061,
+      "step": 15680
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.038350147990197e-05,
+      "loss": 1.1861,
+      "step": 15690
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.035167563094746e-05,
+      "loss": 1.3347,
+      "step": 15700
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.031984978199293e-05,
+      "loss": 1.105,
+      "step": 15710
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.028802393303842e-05,
+      "loss": 1.2095,
+      "step": 15720
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.025619808408389e-05,
+      "loss": 1.0656,
+      "step": 15730
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.022437223512938e-05,
+      "loss": 1.291,
+      "step": 15740
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.0192546386174853e-05,
+      "loss": 1.2578,
+      "step": 15750
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.016072053722033e-05,
+      "loss": 1.2552,
+      "step": 15760
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.0128894688265814e-05,
+      "loss": 1.0318,
+      "step": 15770
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.009706883931129e-05,
+      "loss": 1.2062,
+      "step": 15780
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.0065242990356774e-05,
+      "loss": 1.1256,
+      "step": 15790
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.003341714140225e-05,
+      "loss": 1.0347,
+      "step": 15800
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.000159129244772e-05,
+      "loss": 1.1673,
+      "step": 15810
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.9969765443493205e-05,
+      "loss": 1.4957,
+      "step": 15820
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.993793959453868e-05,
+      "loss": 1.3662,
+      "step": 15830
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.9906113745584166e-05,
+      "loss": 1.2346,
+      "step": 15840
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.987428789662964e-05,
+      "loss": 1.2201,
+      "step": 15850
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.9842462047675126e-05,
+      "loss": 1.2533,
+      "step": 15860
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.98106361987206e-05,
+      "loss": 1.2212,
+      "step": 15870
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.977881034976608e-05,
+      "loss": 1.2752,
+      "step": 15880
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.9746984500811563e-05,
+      "loss": 1.3381,
+      "step": 15890
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.971515865185704e-05,
+      "loss": 1.0002,
+      "step": 15900
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.9683332802902524e-05,
+      "loss": 1.0658,
+      "step": 15910
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.9651506953948e-05,
+      "loss": 1.3124,
+      "step": 15920
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.961968110499348e-05,
+      "loss": 1.4023,
+      "step": 15930
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.9587855256038955e-05,
+      "loss": 1.426,
+      "step": 15940
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.955602940708444e-05,
+      "loss": 1.2818,
+      "step": 15950
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.9524203558129915e-05,
+      "loss": 1.5452,
+      "step": 15960
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.94923777091754e-05,
+      "loss": 1.5367,
+      "step": 15970
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.946055186022087e-05,
+      "loss": 1.2374,
+      "step": 15980
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.942872601126635e-05,
+      "loss": 1.0215,
+      "step": 15990
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.939690016231183e-05,
+      "loss": 1.4311,
+      "step": 16000
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.936507431335731e-05,
+      "loss": 1.419,
+      "step": 16010
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.933324846440279e-05,
+      "loss": 1.1318,
+      "step": 16020
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.930142261544827e-05,
+      "loss": 0.9369,
+      "step": 16030
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.926959676649375e-05,
+      "loss": 1.2709,
+      "step": 16040
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.923777091753923e-05,
+      "loss": 1.0934,
+      "step": 16050
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.920594506858471e-05,
+      "loss": 0.9907,
+      "step": 16060
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.917411921963019e-05,
+      "loss": 1.4633,
+      "step": 16070
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.9142293370675665e-05,
+      "loss": 1.0784,
+      "step": 16080
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.911046752172114e-05,
+      "loss": 1.5323,
+      "step": 16090
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.9078641672766625e-05,
+      "loss": 1.2013,
+      "step": 16100
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.90468158238121e-05,
+      "loss": 1.261,
+      "step": 16110
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.9014989974857586e-05,
+      "loss": 1.4675,
+      "step": 16120
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.8983164125903056e-05,
+      "loss": 1.0573,
+      "step": 16130
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.895133827694854e-05,
+      "loss": 1.0632,
+      "step": 16140
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.8919512427994016e-05,
+      "loss": 1.3218,
+      "step": 16150
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.88876865790395e-05,
+      "loss": 1.0788,
+      "step": 16160
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.8855860730084977e-05,
+      "loss": 1.3465,
+      "step": 16170
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.8824034881130453e-05,
+      "loss": 1.2767,
+      "step": 16180
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.879220903217594e-05,
+      "loss": 1.6768,
+      "step": 16190
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.8760383183221414e-05,
+      "loss": 1.2286,
+      "step": 16200
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.87285573342669e-05,
+      "loss": 1.4228,
+      "step": 16210
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.8696731485312374e-05,
+      "loss": 1.2702,
+      "step": 16220
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.866490563635786e-05,
+      "loss": 1.3598,
+      "step": 16230
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.863307978740333e-05,
+      "loss": 1.2043,
+      "step": 16240
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.860125393844881e-05,
+      "loss": 1.4265,
+      "step": 16250
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.856942808949429e-05,
+      "loss": 1.1054,
+      "step": 16260
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.853760224053977e-05,
+      "loss": 1.5339,
+      "step": 16270
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.850577639158525e-05,
+      "loss": 1.2976,
+      "step": 16280
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.8473950542630726e-05,
+      "loss": 1.4514,
+      "step": 16290
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.84421246936762e-05,
+      "loss": 1.2413,
+      "step": 16300
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.8410298844721687e-05,
+      "loss": 1.0662,
+      "step": 16310
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.8378472995767163e-05,
+      "loss": 1.3955,
+      "step": 16320
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.834664714681265e-05,
+      "loss": 1.208,
+      "step": 16330
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.831482129785812e-05,
+      "loss": 1.4724,
+      "step": 16340
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.82829954489036e-05,
+      "loss": 1.2841,
+      "step": 16350
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.825116959994908e-05,
+      "loss": 1.2628,
+      "step": 16360
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.821934375099456e-05,
+      "loss": 0.989,
+      "step": 16370
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.818751790204004e-05,
+      "loss": 0.9159,
+      "step": 16380
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.8155692053085515e-05,
+      "loss": 1.2558,
+      "step": 16390
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.8123866204131e-05,
+      "loss": 1.0551,
+      "step": 16400
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.8092040355176476e-05,
+      "loss": 1.2573,
+      "step": 16410
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.806021450622196e-05,
+      "loss": 1.4726,
+      "step": 16420
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.8028388657267436e-05,
+      "loss": 1.3828,
+      "step": 16430
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.799656280831291e-05,
+      "loss": 1.2052,
+      "step": 16440
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.796473695935839e-05,
+      "loss": 1.149,
+      "step": 16450
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.7932911110403873e-05,
+      "loss": 1.3104,
+      "step": 16460
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.790108526144935e-05,
+      "loss": 1.3008,
+      "step": 16470
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.7869259412494834e-05,
+      "loss": 1.314,
+      "step": 16480
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.7837433563540304e-05,
+      "loss": 1.2192,
+      "step": 16490
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.780560771458579e-05,
+      "loss": 1.6693,
+      "step": 16500
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.7773781865631265e-05,
+      "loss": 1.2332,
+      "step": 16510
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.774195601667675e-05,
+      "loss": 1.6067,
+      "step": 16520
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.7710130167722225e-05,
+      "loss": 1.1059,
+      "step": 16530
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.767830431876771e-05,
+      "loss": 1.3827,
+      "step": 16540
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.7646478469813186e-05,
+      "loss": 1.2931,
+      "step": 16550
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.761465262085866e-05,
+      "loss": 1.4552,
+      "step": 16560
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.7582826771904146e-05,
+      "loss": 1.1972,
+      "step": 16570
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.755100092294962e-05,
+      "loss": 0.8767,
+      "step": 16580
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.7519175073995107e-05,
+      "loss": 0.9475,
+      "step": 16590
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.7487349225040577e-05,
+      "loss": 1.5473,
+      "step": 16600
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.745552337608606e-05,
+      "loss": 1.269,
+      "step": 16610
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.742369752713154e-05,
+      "loss": 1.3995,
+      "step": 16620
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.739187167817702e-05,
+      "loss": 1.4818,
+      "step": 16630
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.73600458292225e-05,
+      "loss": 1.2699,
+      "step": 16640
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.7328219980267974e-05,
+      "loss": 1.3036,
+      "step": 16650
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.729639413131345e-05,
+      "loss": 1.3475,
+      "step": 16660
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.7264568282358935e-05,
+      "loss": 1.1247,
+      "step": 16670
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.723274243340441e-05,
+      "loss": 1.2363,
+      "step": 16680
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.7200916584449895e-05,
+      "loss": 1.4974,
+      "step": 16690
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.716909073549537e-05,
+      "loss": 1.3437,
+      "step": 16700
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.713726488654085e-05,
+      "loss": 1.162,
+      "step": 16710
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.710543903758633e-05,
+      "loss": 1.3693,
+      "step": 16720
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.707361318863181e-05,
+      "loss": 1.42,
+      "step": 16730
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.704178733967729e-05,
+      "loss": 1.3814,
+      "step": 16740
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.7009961490722763e-05,
+      "loss": 1.1444,
+      "step": 16750
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.697813564176825e-05,
+      "loss": 1.1105,
+      "step": 16760
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.6946309792813724e-05,
+      "loss": 1.2547,
+      "step": 16770
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.691448394385921e-05,
+      "loss": 1.2134,
+      "step": 16780
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.6882658094904684e-05,
+      "loss": 1.3874,
+      "step": 16790
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.685083224595016e-05,
+      "loss": 1.2903,
+      "step": 16800
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.681900639699564e-05,
+      "loss": 1.1891,
+      "step": 16810
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.678718054804112e-05,
+      "loss": 1.1899,
+      "step": 16820
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.67553546990866e-05,
+      "loss": 1.3783,
+      "step": 16830
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.672352885013208e-05,
+      "loss": 1.1659,
+      "step": 16840
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.669170300117756e-05,
+      "loss": 1.4276,
+      "step": 16850
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.6659877152223036e-05,
+      "loss": 1.2746,
+      "step": 16860
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.662805130326851e-05,
+      "loss": 1.222,
+      "step": 16870
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.6596225454313997e-05,
+      "loss": 1.157,
+      "step": 16880
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.6564399605359473e-05,
+      "loss": 1.3028,
+      "step": 16890
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.653257375640496e-05,
+      "loss": 1.4982,
+      "step": 16900
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.6500747907450434e-05,
+      "loss": 1.3935,
+      "step": 16910
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.646892205849591e-05,
+      "loss": 1.5659,
+      "step": 16920
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.6437096209541394e-05,
+      "loss": 1.3223,
+      "step": 16930
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.640527036058687e-05,
+      "loss": 1.1122,
+      "step": 16940
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.6373444511632355e-05,
+      "loss": 1.0243,
+      "step": 16950
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.6341618662677825e-05,
+      "loss": 1.1851,
+      "step": 16960
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.630979281372331e-05,
+      "loss": 1.321,
+      "step": 16970
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.6277966964768786e-05,
+      "loss": 1.5267,
+      "step": 16980
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.624614111581427e-05,
+      "loss": 1.0554,
+      "step": 16990
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.6214315266859746e-05,
+      "loss": 1.2804,
+      "step": 17000
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.618248941790522e-05,
+      "loss": 1.2965,
+      "step": 17010
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.61506635689507e-05,
+      "loss": 1.0722,
+      "step": 17020
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.611883771999618e-05,
+      "loss": 1.5165,
+      "step": 17030
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.608701187104166e-05,
+      "loss": 1.441,
+      "step": 17040
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.6055186022087144e-05,
+      "loss": 0.9771,
+      "step": 17050
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.602336017313262e-05,
+      "loss": 1.3015,
+      "step": 17060
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.59915343241781e-05,
+      "loss": 1.1422,
+      "step": 17070
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.595970847522358e-05,
+      "loss": 1.3455,
+      "step": 17080
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.592788262626906e-05,
+      "loss": 1.004,
+      "step": 17090
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.589605677731454e-05,
+      "loss": 1.0713,
+      "step": 17100
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.586423092836001e-05,
+      "loss": 1.0479,
+      "step": 17110
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.5832405079405495e-05,
+      "loss": 1.4771,
+      "step": 17120
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.580057923045097e-05,
+      "loss": 1.2325,
+      "step": 17130
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.5768753381496456e-05,
+      "loss": 1.262,
+      "step": 17140
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.573692753254193e-05,
+      "loss": 1.4222,
+      "step": 17150
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.5705101683587416e-05,
+      "loss": 1.2611,
+      "step": 17160
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.5673275834632887e-05,
+      "loss": 1.2773,
+      "step": 17170
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.564144998567837e-05,
+      "loss": 1.3417,
+      "step": 17180
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.560962413672385e-05,
+      "loss": 1.2499,
+      "step": 17190
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.557779828776933e-05,
+      "loss": 1.3159,
+      "step": 17200
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.554597243881481e-05,
+      "loss": 1.2028,
+      "step": 17210
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.5514146589860284e-05,
+      "loss": 1.2043,
+      "step": 17220
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.548232074090577e-05,
+      "loss": 0.967,
+      "step": 17230
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.5450494891951245e-05,
+      "loss": 1.2259,
+      "step": 17240
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.541866904299673e-05,
+      "loss": 1.1098,
+      "step": 17250
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.5386843194042205e-05,
+      "loss": 1.2959,
+      "step": 17260
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.535501734508768e-05,
+      "loss": 1.0986,
+      "step": 17270
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.532319149613316e-05,
+      "loss": 1.2249,
+      "step": 17280
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.529136564717864e-05,
+      "loss": 1.3841,
+      "step": 17290
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.525953979822412e-05,
+      "loss": 0.9871,
+      "step": 17300
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.52277139492696e-05,
+      "loss": 1.2703,
+      "step": 17310
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.5195888100315073e-05,
+      "loss": 1.224,
+      "step": 17320
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.516406225136056e-05,
+      "loss": 1.1882,
+      "step": 17330
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.5132236402406034e-05,
+      "loss": 1.0581,
+      "step": 17340
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.510041055345152e-05,
+      "loss": 1.4191,
+      "step": 17350
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.5068584704496994e-05,
+      "loss": 1.3129,
+      "step": 17360
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.503675885554247e-05,
+      "loss": 1.4355,
+      "step": 17370
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.5004933006587955e-05,
+      "loss": 1.4007,
+      "step": 17380
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.497310715763343e-05,
+      "loss": 1.235,
+      "step": 17390
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.494128130867891e-05,
+      "loss": 1.124,
+      "step": 17400
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.490945545972439e-05,
+      "loss": 1.3663,
+      "step": 17410
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.487762961076987e-05,
+      "loss": 1.2598,
+      "step": 17420
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.4845803761815346e-05,
+      "loss": 1.2365,
+      "step": 17430
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.481397791286083e-05,
+      "loss": 1.3799,
+      "step": 17440
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.4782152063906306e-05,
+      "loss": 1.1697,
+      "step": 17450
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.475032621495179e-05,
+      "loss": 1.6229,
+      "step": 17460
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.471850036599727e-05,
+      "loss": 1.3189,
+      "step": 17470
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.4686674517042744e-05,
+      "loss": 1.3392,
+      "step": 17480
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.465484866808822e-05,
+      "loss": 1.5232,
+      "step": 17490
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.4623022819133704e-05,
+      "loss": 1.3666,
+      "step": 17500
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.459119697017918e-05,
+      "loss": 1.1053,
+      "step": 17510
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.4559371121224665e-05,
+      "loss": 1.102,
+      "step": 17520
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.4527545272270135e-05,
+      "loss": 0.7984,
+      "step": 17530
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.449571942331562e-05,
+      "loss": 1.175,
+      "step": 17540
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.4463893574361095e-05,
+      "loss": 1.2059,
+      "step": 17550
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.443206772540658e-05,
+      "loss": 1.4837,
+      "step": 17560
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.4400241876452056e-05,
+      "loss": 1.2114,
+      "step": 17570
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.436841602749753e-05,
+      "loss": 1.278,
+      "step": 17580
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.4336590178543016e-05,
+      "loss": 1.2685,
+      "step": 17590
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.430476432958849e-05,
+      "loss": 1.3156,
+      "step": 17600
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.427293848063398e-05,
+      "loss": 1.1934,
+      "step": 17610
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.4241112631679454e-05,
+      "loss": 1.2918,
+      "step": 17620
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.420928678272493e-05,
+      "loss": 1.1634,
+      "step": 17630
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.417746093377041e-05,
+      "loss": 0.9987,
+      "step": 17640
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.414563508481589e-05,
+      "loss": 1.4489,
+      "step": 17650
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.411380923586137e-05,
+      "loss": 1.2199,
+      "step": 17660
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.408198338690685e-05,
+      "loss": 1.1415,
+      "step": 17670
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.405015753795232e-05,
+      "loss": 1.5274,
+      "step": 17680
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.4018331688997805e-05,
+      "loss": 1.21,
+      "step": 17690
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.398650584004328e-05,
+      "loss": 0.9351,
+      "step": 17700
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 4.3954679991088766e-05,
+      "loss": 1.3924,
+      "step": 17710
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 4.392285414213424e-05,
+      "loss": 1.0862,
+      "step": 17720
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 4.389102829317972e-05,
+      "loss": 1.2006,
+      "step": 17730
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 4.38592024442252e-05,
+      "loss": 1.3581,
+      "step": 17740
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 4.382737659527068e-05,
+      "loss": 1.091,
+      "step": 17750
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 4.3795550746316164e-05,
+      "loss": 1.537,
+      "step": 17760
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 4.376372489736164e-05,
+      "loss": 1.29,
+      "step": 17770
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 4.373189904840712e-05,
+      "loss": 1.2013,
+      "step": 17780
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 4.3700073199452594e-05,
+      "loss": 1.23,
+      "step": 17790
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 4.366824735049808e-05,
+      "loss": 1.4847,
+      "step": 17800
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 4.3636421501543555e-05,
+      "loss": 1.3086,
+      "step": 17810
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 4.360459565258904e-05,
+      "loss": 1.303,
+      "step": 17820
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 4.3572769803634515e-05,
+      "loss": 1.271,
+      "step": 17830
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 4.354094395467999e-05,
+      "loss": 1.2053,
+      "step": 17840
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 4.350911810572547e-05,
+      "loss": 1.3765,
+      "step": 17850
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 4.347729225677095e-05,
+      "loss": 1.2532,
+      "step": 17860
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 4.344546640781643e-05,
+      "loss": 1.2865,
+      "step": 17870
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 4.341364055886191e-05,
+      "loss": 1.2211,
+      "step": 17880
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 4.338181470990739e-05,
+      "loss": 1.4564,
+      "step": 17890
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 4.334998886095287e-05,
+      "loss": 1.0033,
+      "step": 17900
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 4.331816301199835e-05,
+      "loss": 0.962,
+      "step": 17910
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 4.328633716304383e-05,
+      "loss": 1.6116,
+      "step": 17920
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 4.3254511314089304e-05,
+      "loss": 1.0693,
+      "step": 17930
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 4.322268546513478e-05,
+      "loss": 1.4096,
+      "step": 17940
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 4.3190859616180265e-05,
+      "loss": 1.226,
+      "step": 17950
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 4.315903376722574e-05,
+      "loss": 1.0396,
+      "step": 17960
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 4.3127207918271225e-05,
+      "loss": 1.2501,
+      "step": 17970
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 4.30953820693167e-05,
+      "loss": 1.2864,
+      "step": 17980
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 4.306355622036218e-05,
+      "loss": 1.2269,
+      "step": 17990
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 4.3031730371407656e-05,
+      "loss": 1.116,
+      "step": 18000
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 4.299990452245314e-05,
+      "loss": 1.3539,
+      "step": 18010
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 4.2968078673498616e-05,
+      "loss": 1.5695,
+      "step": 18020
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 4.29362528245441e-05,
+      "loss": 1.3146,
+      "step": 18030
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 4.290442697558957e-05,
+      "loss": 1.1909,
+      "step": 18040
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 4.2872601126635054e-05,
+      "loss": 1.3509,
+      "step": 18050
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 4.284077527768053e-05,
+      "loss": 1.3447,
+      "step": 18060
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 4.2808949428726014e-05,
+      "loss": 1.258,
+      "step": 18070
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 4.277712357977149e-05,
+      "loss": 1.5396,
+      "step": 18080
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 4.274529773081697e-05,
+      "loss": 1.5036,
+      "step": 18090
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 4.271347188186245e-05,
+      "loss": 0.9903,
+      "step": 18100
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 4.268164603290793e-05,
+      "loss": 1.188,
+      "step": 18110
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 4.264982018395341e-05,
+      "loss": 1.4617,
+      "step": 18120
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 4.261799433499889e-05,
+      "loss": 1.5529,
+      "step": 18130
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 4.258616848604437e-05,
+      "loss": 0.9911,
+      "step": 18140
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 4.255434263708984e-05,
+      "loss": 1.2402,
+      "step": 18150
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 4.2522516788135326e-05,
+      "loss": 1.1673,
+      "step": 18160
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 4.24906909391808e-05,
+      "loss": 1.5972,
+      "step": 18170
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 4.245886509022629e-05,
+      "loss": 1.3765,
+      "step": 18180
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 4.2427039241271764e-05,
+      "loss": 1.0896,
+      "step": 18190
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 4.239521339231724e-05,
+      "loss": 1.0541,
+      "step": 18200
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 4.236338754336272e-05,
+      "loss": 1.2994,
+      "step": 18210
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 4.23315616944082e-05,
+      "loss": 1.5104,
+      "step": 18220
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 4.229973584545368e-05,
+      "loss": 1.7087,
+      "step": 18230
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 4.226790999649916e-05,
+      "loss": 1.5083,
+      "step": 18240
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 4.223608414754464e-05,
+      "loss": 1.5388,
+      "step": 18250
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 4.2204258298590115e-05,
+      "loss": 1.3103,
+      "step": 18260
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 4.21724324496356e-05,
+      "loss": 1.2792,
+      "step": 18270
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 4.2140606600681076e-05,
+      "loss": 1.1251,
+      "step": 18280
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 4.210878075172656e-05,
+      "loss": 1.2299,
+      "step": 18290
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 4.207695490277203e-05,
+      "loss": 1.2919,
+      "step": 18300
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 4.204512905381751e-05,
+      "loss": 1.0468,
+      "step": 18310
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 4.201330320486299e-05,
+      "loss": 1.4261,
+      "step": 18320
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 4.1981477355908474e-05,
+      "loss": 1.2958,
+      "step": 18330
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 4.194965150695395e-05,
+      "loss": 1.3843,
+      "step": 18340
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 4.191782565799943e-05,
+      "loss": 1.0905,
+      "step": 18350
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 4.1885999809044904e-05,
+      "loss": 1.3191,
+      "step": 18360
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 4.185417396009039e-05,
+      "loss": 1.3327,
+      "step": 18370
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 4.1822348111135865e-05,
+      "loss": 0.9465,
+      "step": 18380
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 4.179052226218135e-05,
+      "loss": 1.1875,
+      "step": 18390
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 4.1758696413226825e-05,
+      "loss": 0.936,
+      "step": 18400
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 4.17268705642723e-05,
+      "loss": 1.3193,
+      "step": 18410
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 4.1695044715317786e-05,
+      "loss": 1.1251,
+      "step": 18420
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 4.166321886636326e-05,
+      "loss": 1.0732,
+      "step": 18430
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 4.1631393017408746e-05,
+      "loss": 1.5964,
+      "step": 18440
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 4.159956716845422e-05,
+      "loss": 1.3061,
+      "step": 18450
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 4.15677413194997e-05,
+      "loss": 1.5171,
+      "step": 18460
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 4.153591547054518e-05,
+      "loss": 1.2409,
+      "step": 18470
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 4.150408962159066e-05,
+      "loss": 1.4052,
+      "step": 18480
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 4.147226377263614e-05,
+      "loss": 1.053,
+      "step": 18490
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 4.144043792368162e-05,
+      "loss": 1.3229,
+      "step": 18500
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 4.140861207472709e-05,
+      "loss": 1.4538,
+      "step": 18510
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 4.1376786225772575e-05,
+      "loss": 1.2916,
+      "step": 18520
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 4.134496037681805e-05,
+      "loss": 0.9643,
+      "step": 18530
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 4.1313134527863535e-05,
+      "loss": 1.4604,
+      "step": 18540
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 4.128130867890901e-05,
+      "loss": 1.2127,
+      "step": 18550
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 4.124948282995449e-05,
+      "loss": 1.0547,
+      "step": 18560
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 4.1217656980999966e-05,
+      "loss": 1.2186,
+      "step": 18570
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 4.118583113204545e-05,
+      "loss": 1.0003,
+      "step": 18580
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 4.1154005283090926e-05,
+      "loss": 1.2351,
+      "step": 18590
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 4.112217943413641e-05,
+      "loss": 1.4122,
+      "step": 18600
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 4.109035358518189e-05,
+      "loss": 1.438,
+      "step": 18610
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 4.1058527736227364e-05,
+      "loss": 1.196,
+      "step": 18620
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 4.102670188727285e-05,
+      "loss": 1.062,
+      "step": 18630
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 4.0994876038318324e-05,
+      "loss": 1.3738,
+      "step": 18640
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 4.096305018936381e-05,
+      "loss": 1.1569,
+      "step": 18650
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 4.093122434040928e-05,
+      "loss": 1.2354,
+      "step": 18660
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 4.089939849145476e-05,
+      "loss": 0.9465,
+      "step": 18670
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 4.086757264250024e-05,
+      "loss": 1.1662,
+      "step": 18680
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 4.083574679354572e-05,
+      "loss": 1.3165,
+      "step": 18690
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 4.08039209445912e-05,
+      "loss": 1.5689,
+      "step": 18700
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 4.0772095095636676e-05,
+      "loss": 1.2426,
+      "step": 18710
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 4.074026924668215e-05,
+      "loss": 1.2528,
+      "step": 18720
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 4.0708443397727636e-05,
+      "loss": 1.4647,
+      "step": 18730
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 4.067661754877311e-05,
+      "loss": 1.0062,
+      "step": 18740
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 4.06447916998186e-05,
+      "loss": 1.1861,
+      "step": 18750
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 4.0612965850864074e-05,
+      "loss": 1.3053,
+      "step": 18760
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 4.058114000190955e-05,
+      "loss": 1.3605,
+      "step": 18770
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 4.0549314152955034e-05,
+      "loss": 1.2673,
+      "step": 18780
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 4.051748830400051e-05,
+      "loss": 1.2602,
+      "step": 18790
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 4.0485662455045995e-05,
+      "loss": 1.3034,
+      "step": 18800
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 4.045383660609147e-05,
+      "loss": 1.2683,
+      "step": 18810
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 4.042201075713695e-05,
+      "loss": 1.3468,
+      "step": 18820
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 4.0390184908182425e-05,
+      "loss": 1.1144,
+      "step": 18830
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 4.035835905922791e-05,
+      "loss": 1.3821,
+      "step": 18840
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 4.0326533210273386e-05,
+      "loss": 1.3033,
+      "step": 18850
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 4.029470736131887e-05,
+      "loss": 1.4069,
+      "step": 18860
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 4.026288151236434e-05,
+      "loss": 1.3087,
+      "step": 18870
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 4.023105566340982e-05,
+      "loss": 1.3872,
+      "step": 18880
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 4.01992298144553e-05,
+      "loss": 1.4077,
+      "step": 18890
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 4.0167403965500784e-05,
+      "loss": 1.172,
+      "step": 18900
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 4.013557811654626e-05,
+      "loss": 1.2941,
+      "step": 18910
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 4.010375226759174e-05,
+      "loss": 1.4621,
+      "step": 18920
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 4.007192641863722e-05,
+      "loss": 1.2763,
+      "step": 18930
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 4.00401005696827e-05,
+      "loss": 1.392,
+      "step": 18940
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 4.000827472072818e-05,
+      "loss": 1.4797,
+      "step": 18950
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.997644887177366e-05,
+      "loss": 0.9922,
+      "step": 18960
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.9944623022819135e-05,
+      "loss": 1.2553,
+      "step": 18970
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.991279717386461e-05,
+      "loss": 1.1796,
+      "step": 18980
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.9880971324910096e-05,
+      "loss": 1.098,
+      "step": 18990
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.984914547595557e-05,
+      "loss": 1.3635,
+      "step": 19000
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.9817319627001056e-05,
+      "loss": 1.5246,
+      "step": 19010
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.9785493778046526e-05,
+      "loss": 1.3422,
+      "step": 19020
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.975366792909201e-05,
+      "loss": 1.1719,
+      "step": 19030
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.972184208013749e-05,
+      "loss": 1.3276,
+      "step": 19040
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.969001623118297e-05,
+      "loss": 1.4109,
+      "step": 19050
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.965819038222845e-05,
+      "loss": 1.3793,
+      "step": 19060
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.962636453327393e-05,
+      "loss": 1.2927,
+      "step": 19070
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.95945386843194e-05,
+      "loss": 1.3429,
+      "step": 19080
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.9562712835364885e-05,
+      "loss": 0.9858,
+      "step": 19090
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.953088698641036e-05,
+      "loss": 1.4387,
+      "step": 19100
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.9499061137455845e-05,
+      "loss": 1.4169,
+      "step": 19110
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.946723528850132e-05,
+      "loss": 1.1409,
+      "step": 19120
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.94354094395468e-05,
+      "loss": 1.133,
+      "step": 19130
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.940358359059228e-05,
+      "loss": 1.2668,
+      "step": 19140
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.937175774163776e-05,
+      "loss": 1.2295,
+      "step": 19150
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.933993189268324e-05,
+      "loss": 1.3709,
+      "step": 19160
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.930810604372872e-05,
+      "loss": 1.4518,
+      "step": 19170
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.92762801947742e-05,
+      "loss": 0.9845,
+      "step": 19180
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.9244454345819674e-05,
+      "loss": 1.2617,
+      "step": 19190
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.921262849686516e-05,
+      "loss": 1.1381,
+      "step": 19200
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.9180802647910634e-05,
+      "loss": 1.007,
+      "step": 19210
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.914897679895612e-05,
+      "loss": 1.2175,
+      "step": 19220
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.911715095000159e-05,
+      "loss": 1.2663,
+      "step": 19230
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.908532510104707e-05,
+      "loss": 1.1347,
+      "step": 19240
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.905349925209255e-05,
+      "loss": 1.1037,
+      "step": 19250
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.902167340313803e-05,
+      "loss": 1.2356,
+      "step": 19260
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.898984755418351e-05,
+      "loss": 1.355,
+      "step": 19270
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.8958021705228986e-05,
+      "loss": 1.2678,
+      "step": 19280
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.892619585627447e-05,
+      "loss": 1.295,
+      "step": 19290
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.8894370007319946e-05,
+      "loss": 1.1144,
+      "step": 19300
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.886254415836543e-05,
+      "loss": 1.0753,
+      "step": 19310
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.883071830941091e-05,
+      "loss": 1.2428,
+      "step": 19320
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.8798892460456384e-05,
+      "loss": 1.2343,
+      "step": 19330
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.876706661150186e-05,
+      "loss": 1.2299,
+      "step": 19340
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.8735240762547344e-05,
+      "loss": 1.2575,
+      "step": 19350
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.870341491359282e-05,
+      "loss": 1.3199,
+      "step": 19360
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.8671589064638305e-05,
+      "loss": 1.2286,
+      "step": 19370
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.863976321568378e-05,
+      "loss": 1.1311,
+      "step": 19380
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.860793736672926e-05,
+      "loss": 1.1787,
+      "step": 19390
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.8576111517774735e-05,
+      "loss": 1.3663,
+      "step": 19400
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.854428566882022e-05,
+      "loss": 1.28,
+      "step": 19410
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.8512459819865696e-05,
+      "loss": 1.488,
+      "step": 19420
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.848063397091118e-05,
+      "loss": 1.0685,
+      "step": 19430
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.8448808121956656e-05,
+      "loss": 1.1896,
+      "step": 19440
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.841698227300213e-05,
+      "loss": 1.2928,
+      "step": 19450
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.838515642404762e-05,
+      "loss": 1.1183,
+      "step": 19460
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.8353330575093094e-05,
+      "loss": 1.6208,
+      "step": 19470
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.832150472613858e-05,
+      "loss": 0.9104,
+      "step": 19480
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.828967887718405e-05,
+      "loss": 1.4046,
+      "step": 19490
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.825785302822953e-05,
+      "loss": 1.0457,
+      "step": 19500
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.822602717927501e-05,
+      "loss": 1.1939,
+      "step": 19510
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.819420133032049e-05,
+      "loss": 1.0983,
+      "step": 19520
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.816237548136597e-05,
+      "loss": 1.5828,
+      "step": 19530
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.8130549632411445e-05,
+      "loss": 1.2187,
+      "step": 19540
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.809872378345692e-05,
+      "loss": 1.1794,
+      "step": 19550
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.8066897934502406e-05,
+      "loss": 0.9364,
+      "step": 19560
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.803507208554788e-05,
+      "loss": 1.4084,
+      "step": 19570
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.8003246236593366e-05,
+      "loss": 1.4533,
+      "step": 19580
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.7971420387638836e-05,
+      "loss": 1.0989,
+      "step": 19590
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.793959453868432e-05,
+      "loss": 1.1185,
+      "step": 19600
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.79077686897298e-05,
+      "loss": 1.0785,
+      "step": 19610
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.787594284077528e-05,
+      "loss": 1.3841,
+      "step": 19620
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.784411699182076e-05,
+      "loss": 1.2914,
+      "step": 19630
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.7812291142866234e-05,
+      "loss": 1.2935,
+      "step": 19640
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.778046529391172e-05,
+      "loss": 1.0084,
+      "step": 19650
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.7748639444957195e-05,
+      "loss": 1.2559,
+      "step": 19660
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.771681359600268e-05,
+      "loss": 1.5865,
+      "step": 19670
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.7684987747048155e-05,
+      "loss": 1.2636,
+      "step": 19680
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.765316189809363e-05,
+      "loss": 1.2656,
+      "step": 19690
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.762133604913911e-05,
+      "loss": 0.9276,
+      "step": 19700
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.758951020018459e-05,
+      "loss": 1.5347,
+      "step": 19710
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.755768435123007e-05,
+      "loss": 1.053,
+      "step": 19720
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.752585850227555e-05,
+      "loss": 1.3505,
+      "step": 19730
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.749403265332103e-05,
+      "loss": 1.33,
+      "step": 19740
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.746220680436651e-05,
+      "loss": 0.9352,
+      "step": 19750
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.7430380955411984e-05,
+      "loss": 1.2561,
+      "step": 19760
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.739855510645747e-05,
+      "loss": 1.0789,
+      "step": 19770
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.7366729257502944e-05,
+      "loss": 1.2284,
+      "step": 19780
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.733490340854843e-05,
+      "loss": 1.2263,
+      "step": 19790
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.7303077559593905e-05,
+      "loss": 1.0512,
+      "step": 19800
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.727125171063938e-05,
+      "loss": 1.0489,
+      "step": 19810
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.7239425861684865e-05,
+      "loss": 1.4009,
+      "step": 19820
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.720760001273034e-05,
+      "loss": 1.1195,
+      "step": 19830
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.7175774163775826e-05,
+      "loss": 0.8015,
+      "step": 19840
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.7143948314821296e-05,
+      "loss": 1.4234,
+      "step": 19850
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.711212246586678e-05,
+      "loss": 1.3675,
+      "step": 19860
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.7080296616912256e-05,
+      "loss": 0.9495,
+      "step": 19870
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.704847076795774e-05,
+      "loss": 1.1224,
+      "step": 19880
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.701664491900322e-05,
+      "loss": 1.409,
+      "step": 19890
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.6984819070048694e-05,
+      "loss": 1.3952,
+      "step": 19900
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.695299322109417e-05,
+      "loss": 1.2911,
+      "step": 19910
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.6921167372139654e-05,
+      "loss": 1.1573,
+      "step": 19920
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.688934152318513e-05,
+      "loss": 1.2936,
+      "step": 19930
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.6857515674230615e-05,
+      "loss": 1.2558,
+      "step": 19940
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.682568982527609e-05,
+      "loss": 1.1733,
+      "step": 19950
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.679386397632157e-05,
+      "loss": 1.19,
+      "step": 19960
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.676203812736705e-05,
+      "loss": 1.022,
+      "step": 19970
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.673021227841253e-05,
+      "loss": 1.1534,
+      "step": 19980
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.669838642945801e-05,
+      "loss": 1.1945,
+      "step": 19990
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.666656058050348e-05,
+      "loss": 1.2391,
+      "step": 20000
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.6634734731548966e-05,
+      "loss": 1.133,
+      "step": 20010
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.660290888259444e-05,
+      "loss": 1.1127,
+      "step": 20020
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.657108303363993e-05,
+      "loss": 1.2856,
+      "step": 20030
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.6539257184685404e-05,
+      "loss": 1.2292,
+      "step": 20040
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.650743133573089e-05,
+      "loss": 1.5068,
+      "step": 20050
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.647560548677636e-05,
+      "loss": 1.3691,
+      "step": 20060
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.644377963782184e-05,
+      "loss": 1.2278,
+      "step": 20070
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.641195378886732e-05,
+      "loss": 1.142,
+      "step": 20080
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.63801279399128e-05,
+      "loss": 1.4245,
+      "step": 20090
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.634830209095828e-05,
+      "loss": 1.2538,
+      "step": 20100
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.6316476242003755e-05,
+      "loss": 1.0865,
+      "step": 20110
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.628465039304923e-05,
+      "loss": 1.5445,
+      "step": 20120
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.6252824544094716e-05,
+      "loss": 1.1644,
+      "step": 20130
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.622099869514019e-05,
+      "loss": 1.1228,
+      "step": 20140
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.6189172846185676e-05,
+      "loss": 1.2033,
+      "step": 20150
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.615734699723115e-05,
+      "loss": 1.2266,
+      "step": 20160
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.612552114827663e-05,
+      "loss": 1.2489,
+      "step": 20170
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.6093695299322114e-05,
+      "loss": 1.1645,
+      "step": 20180
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.606186945036759e-05,
+      "loss": 1.1382,
+      "step": 20190
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.6030043601413074e-05,
+      "loss": 1.0547,
+      "step": 20200
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.5998217752458544e-05,
+      "loss": 1.0566,
+      "step": 20210
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.596639190350403e-05,
+      "loss": 1.305,
+      "step": 20220
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.5934566054549505e-05,
+      "loss": 1.3903,
+      "step": 20230
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.590274020559499e-05,
+      "loss": 1.3047,
+      "step": 20240
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.5870914356640465e-05,
+      "loss": 1.1542,
+      "step": 20250
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.583908850768594e-05,
+      "loss": 1.1636,
+      "step": 20260
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.580726265873142e-05,
+      "loss": 0.9835,
+      "step": 20270
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.57754368097769e-05,
+      "loss": 1.1189,
+      "step": 20280
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.574361096082238e-05,
+      "loss": 1.448,
+      "step": 20290
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.571178511186786e-05,
+      "loss": 0.9613,
+      "step": 20300
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.567995926291334e-05,
+      "loss": 1.092,
+      "step": 20310
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.564813341395882e-05,
+      "loss": 1.2723,
+      "step": 20320
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.56163075650043e-05,
+      "loss": 1.37,
+      "step": 20330
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.558448171604978e-05,
+      "loss": 1.0441,
+      "step": 20340
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.555265586709526e-05,
+      "loss": 0.9713,
+      "step": 20350
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.552083001814074e-05,
+      "loss": 1.1469,
+      "step": 20360
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.5489004169186215e-05,
+      "loss": 1.1932,
+      "step": 20370
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.545717832023169e-05,
+      "loss": 1.4208,
+      "step": 20380
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.5425352471277175e-05,
+      "loss": 1.3295,
+      "step": 20390
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.539352662232265e-05,
+      "loss": 1.1444,
+      "step": 20400
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.5361700773368136e-05,
+      "loss": 1.3685,
+      "step": 20410
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.5329874924413606e-05,
+      "loss": 1.3868,
+      "step": 20420
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.529804907545909e-05,
+      "loss": 1.1486,
+      "step": 20430
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.5266223226504566e-05,
+      "loss": 1.1494,
+      "step": 20440
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.523439737755005e-05,
+      "loss": 1.3013,
+      "step": 20450
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.520257152859553e-05,
+      "loss": 1.0297,
+      "step": 20460
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.5170745679641004e-05,
+      "loss": 1.2975,
+      "step": 20470
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.513891983068649e-05,
+      "loss": 1.2635,
+      "step": 20480
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.5107093981731964e-05,
+      "loss": 1.3708,
+      "step": 20490
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.507526813277745e-05,
+      "loss": 1.0579,
+      "step": 20500
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.5043442283822925e-05,
+      "loss": 1.4045,
+      "step": 20510
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.50116164348684e-05,
+      "loss": 1.2214,
+      "step": 20520
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.497979058591388e-05,
+      "loss": 1.3855,
+      "step": 20530
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.494796473695936e-05,
+      "loss": 1.2537,
+      "step": 20540
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.491613888800484e-05,
+      "loss": 1.1688,
+      "step": 20550
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.488431303905032e-05,
+      "loss": 1.2166,
+      "step": 20560
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.485248719009579e-05,
+      "loss": 1.0007,
+      "step": 20570
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.4820661341141276e-05,
+      "loss": 1.5996,
+      "step": 20580
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.478883549218675e-05,
+      "loss": 1.2712,
+      "step": 20590
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.475700964323224e-05,
+      "loss": 1.0818,
+      "step": 20600
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.4725183794277714e-05,
+      "loss": 1.3156,
+      "step": 20610
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.469335794532319e-05,
+      "loss": 1.0122,
+      "step": 20620
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.466153209636867e-05,
+      "loss": 1.098,
+      "step": 20630
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.462970624741415e-05,
+      "loss": 1.158,
+      "step": 20640
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.459788039845963e-05,
+      "loss": 1.1666,
+      "step": 20650
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.456605454950511e-05,
+      "loss": 1.2442,
+      "step": 20660
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.453422870055059e-05,
+      "loss": 1.3601,
+      "step": 20670
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.4502402851596065e-05,
+      "loss": 1.0865,
+      "step": 20680
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.447057700264155e-05,
+      "loss": 1.121,
+      "step": 20690
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.4438751153687026e-05,
+      "loss": 1.148,
+      "step": 20700
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.440692530473251e-05,
+      "loss": 1.2275,
+      "step": 20710
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.4375099455777986e-05,
+      "loss": 0.8846,
+      "step": 20720
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.434327360682346e-05,
+      "loss": 1.3629,
+      "step": 20730
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.431144775786894e-05,
+      "loss": 1.3894,
+      "step": 20740
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.4279621908914424e-05,
+      "loss": 1.1238,
+      "step": 20750
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.42477960599599e-05,
+      "loss": 1.3383,
+      "step": 20760
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.4215970211005384e-05,
+      "loss": 1.1081,
+      "step": 20770
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.4184144362050854e-05,
+      "loss": 1.4528,
+      "step": 20780
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.415231851309634e-05,
+      "loss": 1.2063,
+      "step": 20790
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.4120492664141815e-05,
+      "loss": 0.9596,
+      "step": 20800
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.40886668151873e-05,
+      "loss": 1.4973,
+      "step": 20810
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.4056840966232775e-05,
+      "loss": 1.2717,
+      "step": 20820
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.402501511727825e-05,
+      "loss": 1.4137,
+      "step": 20830
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.3993189268323736e-05,
+      "loss": 1.1622,
+      "step": 20840
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.396136341936921e-05,
+      "loss": 0.962,
+      "step": 20850
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.3929537570414696e-05,
+      "loss": 1.1181,
+      "step": 20860
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.389771172146017e-05,
+      "loss": 1.2876,
+      "step": 20870
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.386588587250565e-05,
+      "loss": 1.5209,
+      "step": 20880
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.383406002355113e-05,
+      "loss": 1.1866,
+      "step": 20890
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.380223417459661e-05,
+      "loss": 1.1722,
+      "step": 20900
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.377040832564209e-05,
+      "loss": 1.1426,
+      "step": 20910
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.373858247668757e-05,
+      "loss": 1.0596,
+      "step": 20920
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.370675662773304e-05,
+      "loss": 1.2358,
+      "step": 20930
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.3674930778778525e-05,
+      "loss": 0.8627,
+      "step": 20940
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.3643104929824e-05,
+      "loss": 1.1639,
+      "step": 20950
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.3611279080869485e-05,
+      "loss": 1.3031,
+      "step": 20960
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.357945323191496e-05,
+      "loss": 0.9835,
+      "step": 20970
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.3547627382960446e-05,
+      "loss": 1.3002,
+      "step": 20980
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.351580153400592e-05,
+      "loss": 1.5188,
+      "step": 20990
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.34839756850514e-05,
+      "loss": 1.0337,
+      "step": 21000
     }
   ],
   "max_steps": 31521,
   "num_train_epochs": 3,
-  "total_flos": 9.955030912708116e+17,
+  "total_flos": 1.9910406818664284e+18,
   "trial_name": null,
   "trial_params": null
 }