diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6727 @@
+{
+  "best_metric": 0.23809599876403809,
+  "best_model_checkpoint": "./trained-gender/checkpoint-10000",
+  "epoch": 2.0287716709701216,
+  "eval_steps": 1000,
+  "global_step": 11000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.000199907783105865,
+      "loss": 0.681,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00019981556621172998,
+      "loss": 0.6802,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.000199723349317595,
+      "loss": 0.6835,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00019963113242345997,
+      "loss": 0.6822,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00019953891552932499,
+      "loss": 0.6896,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00019944669863518997,
+      "loss": 0.6792,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00019935448174105498,
+      "loss": 0.6828,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00019926226484691997,
+      "loss": 0.6648,
+      "step": 80
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00019917004795278495,
+      "loss": 0.6673,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00019907783105864993,
+      "loss": 0.6628,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00019898561416451495,
+      "loss": 0.6715,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00019889339727037993,
+      "loss": 0.6828,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00019880118037624494,
+      "loss": 0.6602,
+      "step": 130
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019870896348210993,
+      "loss": 0.6653,
+      "step": 140
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019861674658797494,
+      "loss": 0.6628,
+      "step": 150
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019852452969383992,
+      "loss": 0.643,
+      "step": 160
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001984323127997049,
+      "loss": 0.64,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001983400959055699,
+      "loss": 0.6225,
+      "step": 180
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001982478790114349,
+      "loss": 0.6459,
+      "step": 190
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001981556621172999,
+      "loss": 0.6327,
+      "step": 200
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001980634452231649,
+      "loss": 0.6477,
+      "step": 210
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019797122832902988,
+      "loss": 0.6269,
+      "step": 220
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001978790114348949,
+      "loss": 0.6123,
+      "step": 230
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019778679454075988,
+      "loss": 0.616,
+      "step": 240
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019769457764662486,
+      "loss": 0.6306,
+      "step": 250
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019760236075248985,
+      "loss": 0.556,
+      "step": 260
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019751014385835486,
+      "loss": 0.6079,
+      "step": 270
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019741792696421984,
+      "loss": 0.564,
+      "step": 280
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019732571007008486,
+      "loss": 0.569,
+      "step": 290
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019723349317594984,
+      "loss": 0.5822,
+      "step": 300
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019714127628181485,
+      "loss": 0.5743,
+      "step": 310
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019704905938767984,
+      "loss": 0.5493,
+      "step": 320
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019695684249354482,
+      "loss": 0.5691,
+      "step": 330
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001968646255994098,
+      "loss": 0.5501,
+      "step": 340
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019677240870527482,
+      "loss": 0.5496,
+      "step": 350
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001966801918111398,
+      "loss": 0.5383,
+      "step": 360
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001965879749170048,
+      "loss": 0.5355,
+      "step": 370
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001964957580228698,
+      "loss": 0.5576,
+      "step": 380
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001964035411287348,
+      "loss": 0.5555,
+      "step": 390
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001963113242345998,
+      "loss": 0.5702,
+      "step": 400
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019621910734046478,
+      "loss": 0.6111,
+      "step": 410
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019612689044632976,
+      "loss": 0.5336,
+      "step": 420
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019603467355219477,
+      "loss": 0.5369,
+      "step": 430
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019594245665805976,
+      "loss": 0.5608,
+      "step": 440
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019585023976392477,
+      "loss": 0.5255,
+      "step": 450
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019575802286978975,
+      "loss": 0.4979,
+      "step": 460
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00019566580597565476,
+      "loss": 0.5375,
+      "step": 470
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00019557358908151975,
+      "loss": 0.5485,
+      "step": 480
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00019548137218738473,
+      "loss": 0.5324,
+      "step": 490
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00019538915529324972,
+      "loss": 0.4639,
+      "step": 500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00019529693839911473,
+      "loss": 0.4846,
+      "step": 510
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019520472150497971,
+      "loss": 0.541,
+      "step": 520
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019511250461084473,
+      "loss": 0.5626,
+      "step": 530
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001950202877167097,
+      "loss": 0.5714,
+      "step": 540
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019492807082257472,
+      "loss": 0.5184,
+      "step": 550
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001948358539284397,
+      "loss": 0.5706,
+      "step": 560
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001947436370343047,
+      "loss": 0.5423,
+      "step": 570
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019465142014016967,
+      "loss": 0.5032,
+      "step": 580
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019455920324603469,
+      "loss": 0.4187,
+      "step": 590
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019446698635189967,
+      "loss": 0.4939,
+      "step": 600
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019437476945776468,
+      "loss": 0.4286,
+      "step": 610
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019428255256362967,
+      "loss": 0.4813,
+      "step": 620
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019419033566949468,
+      "loss": 0.5191,
+      "step": 630
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019409811877535966,
+      "loss": 0.4069,
+      "step": 640
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019400590188122465,
+      "loss": 0.4806,
+      "step": 650
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019391368498708963,
+      "loss": 0.4167,
+      "step": 660
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019382146809295462,
+      "loss": 0.4626,
+      "step": 670
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019372925119881963,
+      "loss": 0.4348,
+      "step": 680
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019363703430468464,
+      "loss": 0.5163,
+      "step": 690
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019354481741054962,
+      "loss": 0.4216,
+      "step": 700
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001934526005164146,
+      "loss": 0.4174,
+      "step": 710
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019336038362227962,
+      "loss": 0.3745,
+      "step": 720
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001932681667281446,
+      "loss": 0.4898,
+      "step": 730
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001931759498340096,
+      "loss": 0.4293,
+      "step": 740
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019308373293987457,
+      "loss": 0.4105,
+      "step": 750
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019299151604573958,
+      "loss": 0.4906,
+      "step": 760
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001928992991516046,
+      "loss": 0.4998,
+      "step": 770
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019280708225746958,
+      "loss": 0.4471,
+      "step": 780
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019271486536333456,
+      "loss": 0.4673,
+      "step": 790
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019262264846919958,
+      "loss": 0.4972,
+      "step": 800
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019253043157506456,
+      "loss": 0.5052,
+      "step": 810
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019243821468092954,
+      "loss": 0.4974,
+      "step": 820
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019234599778679453,
+      "loss": 0.4504,
+      "step": 830
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019225378089265954,
+      "loss": 0.4322,
+      "step": 840
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019216156399852455,
+      "loss": 0.5118,
+      "step": 850
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019206934710438954,
+      "loss": 0.4815,
+      "step": 860
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019197713021025452,
+      "loss": 0.4534,
+      "step": 870
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019188491331611953,
+      "loss": 0.4236,
+      "step": 880
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019179269642198452,
+      "loss": 0.4925,
+      "step": 890
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001917004795278495,
+      "loss": 0.5201,
+      "step": 900
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00019160826263371449,
+      "loss": 0.4272,
+      "step": 910
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001915160457395795,
+      "loss": 0.4909,
+      "step": 920
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001914238288454445,
+      "loss": 0.4485,
+      "step": 930
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001913316119513095,
+      "loss": 0.4188,
+      "step": 940
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00019123939505717448,
+      "loss": 0.437,
+      "step": 950
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001911471781630395,
+      "loss": 0.4208,
+      "step": 960
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00019105496126890447,
+      "loss": 0.5295,
+      "step": 970
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00019096274437476946,
+      "loss": 0.39,
+      "step": 980
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00019087052748063444,
+      "loss": 0.4634,
+      "step": 990
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00019077831058649945,
+      "loss": 0.4277,
+      "step": 1000
+    },
+    {
+      "epoch": 0.18,
+      "eval_accuracy": 0.8089282453898119,
+      "eval_loss": 0.4054337739944458,
+      "eval_runtime": 99.6522,
+      "eval_samples_per_second": 109.922,
+      "eval_steps_per_second": 13.748,
+      "step": 1000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00019068609369236446,
+      "loss": 0.4296,
+      "step": 1010
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00019059387679822945,
+      "loss": 0.4035,
+      "step": 1020
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00019050165990409443,
+      "loss": 0.3379,
+      "step": 1030
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00019040944300995944,
+      "loss": 0.451,
+      "step": 1040
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00019031722611582443,
+      "loss": 0.4277,
+      "step": 1050
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00019022500922168941,
+      "loss": 0.5039,
+      "step": 1060
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001901327923275544,
+      "loss": 0.4543,
+      "step": 1070
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001900405754334194,
+      "loss": 0.3792,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00018994835853928442,
+      "loss": 0.4495,
+      "step": 1090
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001898561416451494,
+      "loss": 0.445,
+      "step": 1100
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001897639247510144,
+      "loss": 0.4661,
+      "step": 1110
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001896717078568794,
+      "loss": 0.4948,
+      "step": 1120
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018957949096274439,
+      "loss": 0.4696,
+      "step": 1130
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018948727406860937,
+      "loss": 0.3564,
+      "step": 1140
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018939505717447435,
+      "loss": 0.3959,
+      "step": 1150
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018930284028033937,
+      "loss": 0.3899,
+      "step": 1160
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00018921062338620438,
+      "loss": 0.3641,
+      "step": 1170
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00018911840649206936,
+      "loss": 0.4144,
+      "step": 1180
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00018902618959793435,
+      "loss": 0.3547,
+      "step": 1190
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00018893397270379936,
+      "loss": 0.5267,
+      "step": 1200
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00018884175580966434,
+      "loss": 0.4542,
+      "step": 1210
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00018874953891552933,
+      "loss": 0.4595,
+      "step": 1220
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001886573220213943,
+      "loss": 0.4909,
+      "step": 1230
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00018856510512725932,
+      "loss": 0.4425,
+      "step": 1240
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00018847288823312433,
+      "loss": 0.393,
+      "step": 1250
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00018838067133898932,
+      "loss": 0.4167,
+      "step": 1260
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001882884544448543,
+      "loss": 0.3037,
+      "step": 1270
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001881962375507193,
+      "loss": 0.4309,
+      "step": 1280
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001881040206565843,
+      "loss": 0.4805,
+      "step": 1290
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00018801180376244928,
+      "loss": 0.4452,
+      "step": 1300
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00018791958686831427,
+      "loss": 0.3939,
+      "step": 1310
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00018782736997417928,
+      "loss": 0.3899,
+      "step": 1320
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001877351530800443,
+      "loss": 0.4246,
+      "step": 1330
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00018764293618590928,
+      "loss": 0.4555,
+      "step": 1340
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00018755071929177426,
+      "loss": 0.3036,
+      "step": 1350
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00018745850239763924,
+      "loss": 0.4121,
+      "step": 1360
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00018736628550350426,
+      "loss": 0.4736,
+      "step": 1370
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00018727406860936924,
+      "loss": 0.4223,
+      "step": 1380
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00018718185171523422,
+      "loss": 0.3058,
+      "step": 1390
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00018708963482109924,
+      "loss": 0.4031,
+      "step": 1400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00018699741792696425,
+      "loss": 0.3931,
+      "step": 1410
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00018690520103282923,
+      "loss": 0.5029,
+      "step": 1420
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00018681298413869422,
+      "loss": 0.5434,
+      "step": 1430
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001867207672445592,
+      "loss": 0.4353,
+      "step": 1440
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001866285503504242,
+      "loss": 0.4171,
+      "step": 1450
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001865363334562892,
+      "loss": 0.4074,
+      "step": 1460
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00018644411656215418,
+      "loss": 0.3915,
+      "step": 1470
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001863518996680192,
+      "loss": 0.3376,
+      "step": 1480
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001862596827738842,
+      "loss": 0.3333,
+      "step": 1490
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001861674658797492,
+      "loss": 0.4467,
+      "step": 1500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00018607524898561417,
+      "loss": 0.435,
+      "step": 1510
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00018598303209147916,
+      "loss": 0.3931,
+      "step": 1520
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00018589081519734417,
+      "loss": 0.3659,
+      "step": 1530
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00018579859830320915,
+      "loss": 0.4438,
+      "step": 1540
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00018570638140907414,
+      "loss": 0.3943,
+      "step": 1550
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00018561416451493915,
+      "loss": 0.4003,
+      "step": 1560
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00018552194762080416,
+      "loss": 0.3227,
+      "step": 1570
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00018542973072666915,
+      "loss": 0.4422,
+      "step": 1580
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00018533751383253413,
+      "loss": 0.3905,
+      "step": 1590
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00018524529693839911,
+      "loss": 0.3801,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00018515308004426413,
+      "loss": 0.3582,
+      "step": 1610
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001850608631501291,
+      "loss": 0.4174,
+      "step": 1620
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001849686462559941,
+      "loss": 0.3648,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001848764293618591,
+      "loss": 0.4227,
+      "step": 1640
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00018478421246772412,
+      "loss": 0.3653,
+      "step": 1650
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001846919955735891,
+      "loss": 0.3624,
+      "step": 1660
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00018459977867945409,
+      "loss": 0.3214,
+      "step": 1670
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00018450756178531907,
+      "loss": 0.3415,
+      "step": 1680
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00018441534489118408,
+      "loss": 0.3219,
+      "step": 1690
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00018432312799704907,
+      "loss": 0.3679,
+      "step": 1700
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00018423091110291405,
+      "loss": 0.3729,
+      "step": 1710
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00018413869420877906,
+      "loss": 0.4188,
+      "step": 1720
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00018404647731464407,
+      "loss": 0.3648,
+      "step": 1730
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00018395426042050906,
+      "loss": 0.4181,
+      "step": 1740
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00018386204352637404,
+      "loss": 0.315,
+      "step": 1750
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00018376982663223903,
+      "loss": 0.3348,
+      "step": 1760
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00018367760973810404,
+      "loss": 0.4496,
+      "step": 1770
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00018358539284396902,
+      "loss": 0.3835,
+      "step": 1780
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.000183493175949834,
+      "loss": 0.3148,
+      "step": 1790
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00018340095905569902,
+      "loss": 0.441,
+      "step": 1800
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00018330874216156403,
+      "loss": 0.3282,
+      "step": 1810
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00018321652526742901,
+      "loss": 0.4019,
+      "step": 1820
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.000183124308373294,
+      "loss": 0.3623,
+      "step": 1830
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00018303209147915898,
+      "loss": 0.3729,
+      "step": 1840
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.000182939874585024,
+      "loss": 0.2986,
+      "step": 1850
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00018284765769088898,
+      "loss": 0.372,
+      "step": 1860
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00018275544079675396,
+      "loss": 0.3837,
+      "step": 1870
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00018266322390261898,
+      "loss": 0.2998,
+      "step": 1880
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.000182571007008484,
+      "loss": 0.3775,
+      "step": 1890
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00018247879011434897,
+      "loss": 0.2937,
+      "step": 1900
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00018238657322021396,
+      "loss": 0.2826,
+      "step": 1910
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00018229435632607894,
+      "loss": 0.4867,
+      "step": 1920
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00018220213943194392,
+      "loss": 0.3199,
+      "step": 1930
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00018210992253780894,
+      "loss": 0.3392,
+      "step": 1940
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00018201770564367392,
+      "loss": 0.3835,
+      "step": 1950
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00018192548874953893,
+      "loss": 0.3517,
+      "step": 1960
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00018183327185540392,
+      "loss": 0.3842,
+      "step": 1970
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00018174105496126893,
+      "loss": 0.3184,
+      "step": 1980
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001816488380671339,
+      "loss": 0.3106,
+      "step": 1990
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001815566211729989,
+      "loss": 0.315,
+      "step": 2000
+    },
+    {
+      "epoch": 0.37,
+      "eval_accuracy": 0.8317509585539529,
+      "eval_loss": 0.3487056791782379,
+      "eval_runtime": 98.9474,
+      "eval_samples_per_second": 110.705,
+      "eval_steps_per_second": 13.846,
+      "step": 2000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00018146440427886388,
+      "loss": 0.4269,
+      "step": 2010
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001813721873847289,
+      "loss": 0.4103,
+      "step": 2020
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00018127997049059388,
+      "loss": 0.3383,
+      "step": 2030
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001811877535964589,
+      "loss": 0.3569,
+      "step": 2040
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00018109553670232387,
+      "loss": 0.295,
+      "step": 2050
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00018100331980818888,
+      "loss": 0.3816,
+      "step": 2060
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00018091110291405387,
+      "loss": 0.3156,
+      "step": 2070
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00018081888601991885,
+      "loss": 0.3574,
+      "step": 2080
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00018072666912578384,
+      "loss": 0.3496,
+      "step": 2090
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00018063445223164885,
+      "loss": 0.3897,
+      "step": 2100
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00018054223533751383,
+      "loss": 0.4285,
+      "step": 2110
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00018045001844337885,
+      "loss": 0.461,
+      "step": 2120
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00018035780154924383,
+      "loss": 0.3399,
+      "step": 2130
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00018026558465510884,
+      "loss": 0.3455,
+      "step": 2140
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00018017336776097383,
+      "loss": 0.3925,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001800811508668388,
+      "loss": 0.3985,
+      "step": 2160
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001799889339727038,
+      "loss": 0.3067,
+      "step": 2170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001798967170785688,
+      "loss": 0.3036,
+      "step": 2180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001798045001844338,
+      "loss": 0.3976,
+      "step": 2190
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001797122832902988,
+      "loss": 0.3375,
+      "step": 2200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00017962006639616379,
+      "loss": 0.4176,
+      "step": 2210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001795278495020288,
+      "loss": 0.3729,
+      "step": 2220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00017943563260789378,
+      "loss": 0.4237,
+      "step": 2230
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00017934341571375877,
+      "loss": 0.3436,
+      "step": 2240
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00017925119881962375,
+      "loss": 0.2412,
+      "step": 2250
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00017915898192548876,
+      "loss": 0.3455,
+      "step": 2260
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00017906676503135375,
+      "loss": 0.3741,
+      "step": 2270
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00017897454813721876,
+      "loss": 0.2836,
+      "step": 2280
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00017888233124308374,
+      "loss": 0.2703,
+      "step": 2290
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00017879011434894875,
+      "loss": 0.3872,
+      "step": 2300
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00017869789745481374,
+      "loss": 0.3514,
+      "step": 2310
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00017860568056067872,
+      "loss": 0.3548,
+      "step": 2320
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001785134636665437,
+      "loss": 0.4352,
+      "step": 2330
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00017842124677240872,
+      "loss": 0.2912,
+      "step": 2340
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001783290298782737,
+      "loss": 0.3551,
+      "step": 2350
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00017823681298413871,
+      "loss": 0.4348,
+      "step": 2360
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001781445960900037,
+      "loss": 0.4094,
+      "step": 2370
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001780523791958687,
+      "loss": 0.3368,
+      "step": 2380
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001779601623017337,
+      "loss": 0.3732,
+      "step": 2390
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00017786794540759868,
+      "loss": 0.2987,
+      "step": 2400
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00017777572851346366,
+      "loss": 0.341,
+      "step": 2410
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00017768351161932868,
+      "loss": 0.3953,
+      "step": 2420
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00017759129472519366,
+      "loss": 0.4143,
+      "step": 2430
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00017749907783105864,
+      "loss": 0.317,
+      "step": 2440
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00017740686093692366,
+      "loss": 0.2584,
+      "step": 2450
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00017731464404278867,
+      "loss": 0.3006,
+      "step": 2460
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00017722242714865365,
+      "loss": 0.4184,
+      "step": 2470
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00017713021025451864,
+      "loss": 0.3758,
+      "step": 2480
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00017703799336038362,
+      "loss": 0.4137,
+      "step": 2490
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00017694577646624863,
+      "loss": 0.3617,
+      "step": 2500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00017685355957211362,
+      "loss": 0.3775,
+      "step": 2510
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001767613426779786,
+      "loss": 0.3578,
+      "step": 2520
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001766691257838436,
+      "loss": 0.3031,
+      "step": 2530
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001765769088897086,
+      "loss": 0.335,
+      "step": 2540
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001764846919955736,
+      "loss": 0.3057,
+      "step": 2550
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001763924751014386,
+      "loss": 0.4072,
+      "step": 2560
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00017630025820730358,
+      "loss": 0.4231,
+      "step": 2570
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00017620804131316856,
+      "loss": 0.3353,
+      "step": 2580
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00017611582441903357,
+      "loss": 0.3978,
+      "step": 2590
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00017602360752489856,
+      "loss": 0.2996,
+      "step": 2600
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00017593139063076357,
+      "loss": 0.3473,
+      "step": 2610
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00017583917373662855,
+      "loss": 0.3141,
+      "step": 2620
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00017574695684249356,
+      "loss": 0.3099,
+      "step": 2630
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00017565473994835855,
+      "loss": 0.3627,
+      "step": 2640
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00017556252305422353,
+      "loss": 0.2984,
+      "step": 2650
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00017547030616008852,
+      "loss": 0.308,
+      "step": 2660
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00017537808926595353,
+      "loss": 0.3109,
+      "step": 2670
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00017528587237181851,
+      "loss": 0.3344,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00017519365547768353,
+      "loss": 0.348,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001751014385835485,
+      "loss": 0.3095,
+      "step": 2700
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00017500922168941352,
+      "loss": 0.3603,
+      "step": 2710
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001749170047952785,
+      "loss": 0.339,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001748247879011435,
+      "loss": 0.359,
+      "step": 2730
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00017473257100700847,
+      "loss": 0.339,
+      "step": 2740
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00017464035411287349,
+      "loss": 0.3681,
+      "step": 2750
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00017454813721873847,
+      "loss": 0.2514,
+      "step": 2760
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00017445592032460348,
+      "loss": 0.3092,
+      "step": 2770
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00017436370343046847,
+      "loss": 0.2572,
+      "step": 2780
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00017427148653633348,
+      "loss": 0.4704,
+      "step": 2790
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00017417926964219846,
+      "loss": 0.2768,
+      "step": 2800
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00017408705274806345,
+      "loss": 0.3077,
+      "step": 2810
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00017399483585392843,
+      "loss": 0.4038,
+      "step": 2820
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00017390261895979344,
+      "loss": 0.2926,
+      "step": 2830
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00017381040206565843,
+      "loss": 0.3433,
+      "step": 2840
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00017371818517152344,
+      "loss": 0.4099,
+      "step": 2850
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00017362596827738842,
+      "loss": 0.3401,
+      "step": 2860
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00017353375138325343,
+      "loss": 0.3805,
+      "step": 2870
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00017344153448911842,
+      "loss": 0.3354,
+      "step": 2880
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001733493175949834,
+      "loss": 0.2953,
+      "step": 2890
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001732571007008484,
+      "loss": 0.3217,
+      "step": 2900
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001731648838067134,
+      "loss": 0.289,
+      "step": 2910
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00017307266691257838,
+      "loss": 0.4819,
+      "step": 2920
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001729804500184434,
+      "loss": 0.288,
+      "step": 2930
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00017288823312430838,
+      "loss": 0.4555,
+      "step": 2940
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001727960162301734,
+      "loss": 0.327,
+      "step": 2950
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00017270379933603838,
+      "loss": 0.337,
+      "step": 2960
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00017261158244190336,
+      "loss": 0.2624,
+      "step": 2970
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00017251936554776834,
+      "loss": 0.2694,
+      "step": 2980
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00017242714865363336,
+      "loss": 0.4381,
+      "step": 2990
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00017233493175949834,
+      "loss": 0.3082,
+      "step": 3000
+    },
+    {
+      "epoch": 0.55,
+      "eval_accuracy": 0.863337593573124,
+      "eval_loss": 0.30519211292266846,
+      "eval_runtime": 99.0719,
+      "eval_samples_per_second": 110.566,
+      "eval_steps_per_second": 13.828,
+      "step": 3000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017224271486536335,
+      "loss": 0.3275,
+      "step": 3010
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017215049797122834,
+      "loss": 0.3314,
+      "step": 3020
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017205828107709335,
+      "loss": 0.289,
+      "step": 3030
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017196606418295833,
+      "loss": 0.3257,
+      "step": 3040
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017187384728882332,
+      "loss": 0.3443,
+      "step": 3050
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001717816303946883,
+      "loss": 0.3118,
+      "step": 3060
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001716894135005533,
+      "loss": 0.4607,
+      "step": 3070
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001715971966064183,
+      "loss": 0.2485,
+      "step": 3080
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001715049797122833,
+      "loss": 0.3853,
+      "step": 3090
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001714127628181483,
+      "loss": 0.3685,
+      "step": 3100
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001713205459240133,
+      "loss": 0.3942,
+      "step": 3110
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001712283290298783,
+      "loss": 0.3057,
+      "step": 3120
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00017113611213574327,
+      "loss": 0.2269,
+      "step": 3130
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00017104389524160826,
+      "loss": 0.3431,
+      "step": 3140
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00017095167834747327,
+      "loss": 0.2295,
+      "step": 3150
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00017085946145333825,
+      "loss": 0.3489,
+      "step": 3160
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00017076724455920326,
+      "loss": 0.3596,
+      "step": 3170
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00017067502766506825,
+      "loss": 0.4055,
+      "step": 3180
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00017058281077093323,
+      "loss": 0.4416,
+      "step": 3190
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00017049059387679825,
+      "loss": 0.2621,
+      "step": 3200
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00017039837698266323,
+      "loss": 0.346,
+      "step": 3210
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00017030616008852821,
+      "loss": 0.3386,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001702139431943932,
+      "loss": 0.3263,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001701217263002582,
+      "loss": 0.3014,
+      "step": 3240
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00017002950940612322,
+      "loss": 0.3152,
+      "step": 3250
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001699372925119882,
+      "loss": 0.2802,
+      "step": 3260
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001698450756178532,
+      "loss": 0.3908,
+      "step": 3270
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001697528587237182,
+      "loss": 0.2139,
+      "step": 3280
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00016966064182958319,
+      "loss": 0.3051,
+      "step": 3290
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00016956842493544817,
+      "loss": 0.3304,
+      "step": 3300
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00016947620804131316,
+      "loss": 0.4109,
+      "step": 3310
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00016938399114717817,
+      "loss": 0.298,
+      "step": 3320
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00016929177425304318,
+      "loss": 0.3299,
+      "step": 3330
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00016919955735890816,
+      "loss": 0.3102,
+      "step": 3340
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00016910734046477315,
+      "loss": 0.2846,
+      "step": 3350
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00016901512357063816,
+      "loss": 0.329,
+      "step": 3360
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00016892290667650314,
+      "loss": 0.2862,
+      "step": 3370
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00016883068978236813,
+      "loss": 0.2984,
+      "step": 3380
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001687384728882331,
+      "loss": 0.3246,
+      "step": 3390
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00016864625599409812,
+      "loss": 0.2689,
+      "step": 3400
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00016855403909996313,
+      "loss": 0.2825,
+      "step": 3410
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00016846182220582812,
+      "loss": 0.3375,
+      "step": 3420
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001683696053116931,
+      "loss": 0.3298,
+      "step": 3430
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00016827738841755812,
+      "loss": 0.334,
+      "step": 3440
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001681851715234231,
+      "loss": 0.3414,
+      "step": 3450
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00016809295462928808,
+      "loss": 0.2998,
+      "step": 3460
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00016800073773515307,
+      "loss": 0.3031,
+      "step": 3470
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00016790852084101808,
+      "loss": 0.2044,
+      "step": 3480
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001678163039468831,
+      "loss": 0.4127,
+      "step": 3490
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00016772408705274808,
+      "loss": 0.2921,
+      "step": 3500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00016763187015861306,
+      "loss": 0.3712,
+      "step": 3510
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00016753965326447807,
+      "loss": 0.342,
+      "step": 3520
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00016744743637034306,
+      "loss": 0.3056,
+      "step": 3530
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00016735521947620804,
+      "loss": 0.3342,
+      "step": 3540
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00016726300258207302,
+      "loss": 0.3005,
+      "step": 3550
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00016717078568793804,
+      "loss": 0.2978,
+      "step": 3560
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00016707856879380305,
+      "loss": 0.415,
+      "step": 3570
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00016698635189966803,
+      "loss": 0.3048,
+      "step": 3580
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00016689413500553302,
+      "loss": 0.3362,
+      "step": 3590
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00016680191811139803,
+      "loss": 0.3437,
+      "step": 3600
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.000166709701217263,
+      "loss": 0.3877,
+      "step": 3610
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.000166617484323128,
+      "loss": 0.2679,
+      "step": 3620
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00016652526742899298,
+      "loss": 0.2978,
+      "step": 3630
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.000166433050534858,
+      "loss": 0.3177,
+      "step": 3640
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.000166340833640723,
+      "loss": 0.2971,
+      "step": 3650
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.000166248616746588,
+      "loss": 0.3507,
+      "step": 3660
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00016615639985245297,
+      "loss": 0.3602,
+      "step": 3670
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00016606418295831798,
+      "loss": 0.2991,
+      "step": 3680
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00016597196606418297,
+      "loss": 0.3303,
+      "step": 3690
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00016587974917004795,
+      "loss": 0.2405,
+      "step": 3700
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00016578753227591294,
+      "loss": 0.3534,
+      "step": 3710
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00016569531538177795,
+      "loss": 0.3164,
+      "step": 3720
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00016560309848764296,
+      "loss": 0.2857,
+      "step": 3730
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00016551088159350795,
+      "loss": 0.3527,
+      "step": 3740
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00016541866469937293,
+      "loss": 0.2585,
+      "step": 3750
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00016532644780523794,
+      "loss": 0.3003,
+      "step": 3760
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00016523423091110293,
+      "loss": 0.3127,
+      "step": 3770
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001651420140169679,
+      "loss": 0.325,
+      "step": 3780
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001650497971228329,
+      "loss": 0.3222,
+      "step": 3790
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001649575802286979,
+      "loss": 0.2263,
+      "step": 3800
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00016486536333456292,
+      "loss": 0.2938,
+      "step": 3810
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001647731464404279,
+      "loss": 0.2571,
+      "step": 3820
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00016468092954629289,
+      "loss": 0.2619,
+      "step": 3830
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00016458871265215787,
+      "loss": 0.3952,
+      "step": 3840
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00016449649575802288,
+      "loss": 0.4201,
+      "step": 3850
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00016440427886388787,
+      "loss": 0.3919,
+      "step": 3860
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00016431206196975285,
+      "loss": 0.316,
+      "step": 3870
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00016421984507561786,
+      "loss": 0.3487,
+      "step": 3880
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00016412762818148287,
+      "loss": 0.3164,
+      "step": 3890
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00016403541128734786,
+      "loss": 0.311,
+      "step": 3900
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00016394319439321284,
+      "loss": 0.4413,
+      "step": 3910
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00016385097749907783,
+      "loss": 0.3052,
+      "step": 3920
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00016375876060494284,
+      "loss": 0.3408,
+      "step": 3930
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00016366654371080782,
+      "loss": 0.235,
+      "step": 3940
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001635743268166728,
+      "loss": 0.2828,
+      "step": 3950
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00016348210992253782,
+      "loss": 0.3549,
+      "step": 3960
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00016338989302840283,
+      "loss": 0.305,
+      "step": 3970
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00016329767613426782,
+      "loss": 0.22,
+      "step": 3980
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001632054592401328,
+      "loss": 0.3241,
+      "step": 3990
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00016311324234599778,
+      "loss": 0.3235,
+      "step": 4000
+    },
+    {
+      "epoch": 0.74,
+      "eval_accuracy": 0.868358590469235,
+      "eval_loss": 0.28994396328926086,
+      "eval_runtime": 99.0495,
+      "eval_samples_per_second": 110.591,
+      "eval_steps_per_second": 13.831,
+      "step": 4000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001630210254518628,
+      "loss": 0.3307,
+      "step": 4010
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00016292880855772778,
+      "loss": 0.342,
+      "step": 4020
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00016283659166359276,
+      "loss": 0.3057,
+      "step": 4030
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00016274437476945778,
+      "loss": 0.3428,
+      "step": 4040
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001626521578753228,
+      "loss": 0.3025,
+      "step": 4050
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00016255994098118777,
+      "loss": 0.3565,
+      "step": 4060
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00016246772408705276,
+      "loss": 0.2313,
+      "step": 4070
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00016237550719291774,
+      "loss": 0.2808,
+      "step": 4080
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00016228329029878275,
+      "loss": 0.2751,
+      "step": 4090
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00016219107340464774,
+      "loss": 0.3921,
+      "step": 4100
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00016209885651051272,
+      "loss": 0.2999,
+      "step": 4110
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00016200663961637773,
+      "loss": 0.2514,
+      "step": 4120
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00016191442272224274,
+      "loss": 0.2381,
+      "step": 4130
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00016182220582810773,
+      "loss": 0.3691,
+      "step": 4140
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001617299889339727,
+      "loss": 0.2547,
+      "step": 4150
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001616377720398377,
+      "loss": 0.2251,
+      "step": 4160
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001615455551457027,
+      "loss": 0.3225,
+      "step": 4170
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001614533382515677,
+      "loss": 0.403,
+      "step": 4180
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00016136112135743268,
+      "loss": 0.2988,
+      "step": 4190
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001612689044632977,
+      "loss": 0.3268,
+      "step": 4200
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001611766875691627,
+      "loss": 0.2629,
+      "step": 4210
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00016108447067502768,
+      "loss": 0.2314,
+      "step": 4220
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00016099225378089267,
+      "loss": 0.2813,
+      "step": 4230
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00016090003688675765,
+      "loss": 0.2538,
+      "step": 4240
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00016080781999262267,
+      "loss": 0.3281,
+      "step": 4250
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00016071560309848765,
+      "loss": 0.324,
+      "step": 4260
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00016062338620435263,
+      "loss": 0.2894,
+      "step": 4270
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00016053116931021765,
+      "loss": 0.3562,
+      "step": 4280
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00016043895241608266,
+      "loss": 0.2743,
+      "step": 4290
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00016034673552194764,
+      "loss": 0.4109,
+      "step": 4300
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00016025451862781263,
+      "loss": 0.3324,
+      "step": 4310
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001601623017336776,
+      "loss": 0.2997,
+      "step": 4320
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00016007008483954262,
+      "loss": 0.342,
+      "step": 4330
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001599778679454076,
+      "loss": 0.3475,
+      "step": 4340
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001598856510512726,
+      "loss": 0.2972,
+      "step": 4350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001597934341571376,
+      "loss": 0.3234,
+      "step": 4360
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001597012172630026,
+      "loss": 0.3133,
+      "step": 4370
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001596090003688676,
+      "loss": 0.3302,
+      "step": 4380
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00015951678347473258,
+      "loss": 0.3424,
+      "step": 4390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00015942456658059757,
+      "loss": 0.2909,
+      "step": 4400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00015933234968646258,
+      "loss": 0.433,
+      "step": 4410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00015924013279232756,
+      "loss": 0.2782,
+      "step": 4420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00015914791589819255,
+      "loss": 0.316,
+      "step": 4430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00015905569900405756,
+      "loss": 0.3129,
+      "step": 4440
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00015896348210992254,
+      "loss": 0.326,
+      "step": 4450
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00015887126521578755,
+      "loss": 0.2647,
+      "step": 4460
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00015877904832165254,
+      "loss": 0.316,
+      "step": 4470
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00015868683142751752,
+      "loss": 0.3266,
+      "step": 4480
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001585946145333825,
+      "loss": 0.3605,
+      "step": 4490
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00015850239763924752,
+      "loss": 0.3225,
+      "step": 4500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001584101807451125,
+      "loss": 0.3426,
+      "step": 4510
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00015831796385097752,
+      "loss": 0.322,
+      "step": 4520
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001582257469568425,
+      "loss": 0.2688,
+      "step": 4530
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001581335300627075,
+      "loss": 0.3182,
+      "step": 4540
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001580413131685725,
+      "loss": 0.2599,
+      "step": 4550
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00015794909627443748,
+      "loss": 0.2918,
+      "step": 4560
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00015785687938030246,
+      "loss": 0.3931,
+      "step": 4570
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00015776466248616748,
+      "loss": 0.3272,
+      "step": 4580
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00015767244559203246,
+      "loss": 0.2132,
+      "step": 4590
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00015758022869789747,
+      "loss": 0.2428,
+      "step": 4600
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00015748801180376246,
+      "loss": 0.4144,
+      "step": 4610
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00015739579490962747,
+      "loss": 0.2609,
+      "step": 4620
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00015730357801549245,
+      "loss": 0.284,
+      "step": 4630
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00015721136112135744,
+      "loss": 0.315,
+      "step": 4640
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00015711914422722242,
+      "loss": 0.3503,
+      "step": 4650
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00015702692733308743,
+      "loss": 0.2572,
+      "step": 4660
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00015693471043895242,
+      "loss": 0.2936,
+      "step": 4670
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00015684249354481743,
+      "loss": 0.2479,
+      "step": 4680
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001567502766506824,
+      "loss": 0.3408,
+      "step": 4690
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00015665805975654742,
+      "loss": 0.358,
+      "step": 4700
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001565658428624124,
+      "loss": 0.3405,
+      "step": 4710
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001564736259682774,
+      "loss": 0.2231,
+      "step": 4720
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00015638140907414238,
+      "loss": 0.3411,
+      "step": 4730
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001562891921800074,
+      "loss": 0.2874,
+      "step": 4740
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00015619697528587237,
+      "loss": 0.3631,
+      "step": 4750
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00015610475839173736,
+      "loss": 0.2846,
+      "step": 4760
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00015601254149760237,
+      "loss": 0.3823,
+      "step": 4770
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00015592032460346738,
+      "loss": 0.3304,
+      "step": 4780
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00015582810770933237,
+      "loss": 0.2658,
+      "step": 4790
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00015573589081519735,
+      "loss": 0.2804,
+      "step": 4800
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00015564367392106233,
+      "loss": 0.2625,
+      "step": 4810
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00015555145702692735,
+      "loss": 0.2694,
+      "step": 4820
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00015545924013279233,
+      "loss": 0.4037,
+      "step": 4830
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00015536702323865731,
+      "loss": 0.2782,
+      "step": 4840
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00015527480634452233,
+      "loss": 0.2277,
+      "step": 4850
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00015518258945038734,
+      "loss": 0.3049,
+      "step": 4860
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00015509037255625232,
+      "loss": 0.315,
+      "step": 4870
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001549981556621173,
+      "loss": 0.2892,
+      "step": 4880
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001549059387679823,
+      "loss": 0.2593,
+      "step": 4890
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001548137218738473,
+      "loss": 0.2331,
+      "step": 4900
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001547215049797123,
+      "loss": 0.3328,
+      "step": 4910
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00015462928808557727,
+      "loss": 0.2698,
+      "step": 4920
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00015453707119144228,
+      "loss": 0.2729,
+      "step": 4930
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001544448542973073,
+      "loss": 0.2428,
+      "step": 4940
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00015435263740317228,
+      "loss": 0.3144,
+      "step": 4950
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00015426042050903726,
+      "loss": 0.2702,
+      "step": 4960
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00015416820361490225,
+      "loss": 0.2847,
+      "step": 4970
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00015407598672076726,
+      "loss": 0.2186,
+      "step": 4980
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00015398376982663224,
+      "loss": 0.3029,
+      "step": 4990
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00015389155293249723,
+      "loss": 0.2505,
+      "step": 5000
+    },
+    {
+      "epoch": 0.92,
+      "eval_accuracy": 0.8784918751141135,
+      "eval_loss": 0.26932471990585327,
+      "eval_runtime": 99.0676,
+      "eval_samples_per_second": 110.571,
+      "eval_steps_per_second": 13.829,
+      "step": 5000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00015379933603836224,
+      "loss": 0.2593,
+      "step": 5010
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00015370711914422725,
+      "loss": 0.3717,
+      "step": 5020
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00015361490225009223,
+      "loss": 0.379,
+      "step": 5030
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00015352268535595722,
+      "loss": 0.2452,
+      "step": 5040
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001534304684618222,
+      "loss": 0.2741,
+      "step": 5050
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001533382515676872,
+      "loss": 0.3541,
+      "step": 5060
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001532460346735522,
+      "loss": 0.4734,
+      "step": 5070
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00015315381777941718,
+      "loss": 0.2334,
+      "step": 5080
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001530616008852822,
+      "loss": 0.2989,
+      "step": 5090
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00015296938399114718,
+      "loss": 0.3124,
+      "step": 5100
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001528771670970122,
+      "loss": 0.2227,
+      "step": 5110
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00015278495020287718,
+      "loss": 0.3233,
+      "step": 5120
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00015269273330874216,
+      "loss": 0.3275,
+      "step": 5130
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00015260051641460714,
+      "loss": 0.2369,
+      "step": 5140
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00015250829952047216,
+      "loss": 0.2122,
+      "step": 5150
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00015241608262633714,
+      "loss": 0.2965,
+      "step": 5160
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00015232386573220215,
+      "loss": 0.3045,
+      "step": 5170
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00015223164883806714,
+      "loss": 0.3304,
+      "step": 5180
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00015213943194393215,
+      "loss": 0.1698,
+      "step": 5190
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00015204721504979713,
+      "loss": 0.2182,
+      "step": 5200
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00015195499815566212,
+      "loss": 0.2228,
+      "step": 5210
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001518627812615271,
+      "loss": 0.2691,
+      "step": 5220
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001517705643673921,
+      "loss": 0.3049,
+      "step": 5230
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001516783474732571,
+      "loss": 0.268,
+      "step": 5240
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001515861305791221,
+      "loss": 0.2829,
+      "step": 5250
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001514939136849871,
+      "loss": 0.2393,
+      "step": 5260
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001514016967908521,
+      "loss": 0.2802,
+      "step": 5270
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001513094798967171,
+      "loss": 0.2809,
+      "step": 5280
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00015121726300258207,
+      "loss": 0.349,
+      "step": 5290
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00015112504610844706,
+      "loss": 0.3409,
+      "step": 5300
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00015103282921431207,
+      "loss": 0.328,
+      "step": 5310
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00015094061232017705,
+      "loss": 0.2118,
+      "step": 5320
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00015084839542604207,
+      "loss": 0.1884,
+      "step": 5330
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00015075617853190705,
+      "loss": 0.3182,
+      "step": 5340
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00015066396163777206,
+      "loss": 0.3171,
+      "step": 5350
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00015057174474363705,
+      "loss": 0.3254,
+      "step": 5360
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00015047952784950203,
+      "loss": 0.2539,
+      "step": 5370
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00015038731095536701,
+      "loss": 0.2322,
+      "step": 5380
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00015029509406123203,
+      "loss": 0.2637,
+      "step": 5390
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.000150202877167097,
+      "loss": 0.3089,
+      "step": 5400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00015011066027296202,
+      "loss": 0.2709,
+      "step": 5410
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.000150018443378827,
+      "loss": 0.3575,
+      "step": 5420
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00014992622648469202,
+      "loss": 0.1982,
+      "step": 5430
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.000149834009590557,
+      "loss": 0.1868,
+      "step": 5440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.000149741792696422,
+      "loss": 0.2507,
+      "step": 5450
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00014964957580228697,
+      "loss": 0.212,
+      "step": 5460
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00014955735890815198,
+      "loss": 0.3149,
+      "step": 5470
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00014946514201401697,
+      "loss": 0.2096,
+      "step": 5480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00014937292511988198,
+      "loss": 0.1928,
+      "step": 5490
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00014928070822574696,
+      "loss": 0.2239,
+      "step": 5500
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00014918849133161197,
+      "loss": 0.3118,
+      "step": 5510
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00014909627443747696,
+      "loss": 0.2,
+      "step": 5520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00014900405754334194,
+      "loss": 0.2386,
+      "step": 5530
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00014891184064920693,
+      "loss": 0.2723,
+      "step": 5540
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00014881962375507194,
+      "loss": 0.2094,
+      "step": 5550
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00014872740686093692,
+      "loss": 0.1656,
+      "step": 5560
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00014863518996680193,
+      "loss": 0.2381,
+      "step": 5570
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00014854297307266692,
+      "loss": 0.1589,
+      "step": 5580
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00014845075617853193,
+      "loss": 0.2366,
+      "step": 5590
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00014835853928439692,
+      "loss": 0.2766,
+      "step": 5600
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.0001482663223902619,
+      "loss": 0.2008,
+      "step": 5610
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00014817410549612688,
+      "loss": 0.2507,
+      "step": 5620
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0001480818886019919,
+      "loss": 0.2559,
+      "step": 5630
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00014798967170785688,
+      "loss": 0.2385,
+      "step": 5640
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0001478974548137219,
+      "loss": 0.2586,
+      "step": 5650
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00014780523791958688,
+      "loss": 0.3115,
+      "step": 5660
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.0001477130210254519,
+      "loss": 0.3633,
+      "step": 5670
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00014762080413131687,
+      "loss": 0.1495,
+      "step": 5680
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00014752858723718186,
+      "loss": 0.1919,
+      "step": 5690
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00014743637034304684,
+      "loss": 0.2895,
+      "step": 5700
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00014734415344891183,
+      "loss": 0.2837,
+      "step": 5710
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00014725193655477684,
+      "loss": 0.2793,
+      "step": 5720
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00014715971966064185,
+      "loss": 0.2243,
+      "step": 5730
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00014706750276650683,
+      "loss": 0.1981,
+      "step": 5740
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00014697528587237182,
+      "loss": 0.2063,
+      "step": 5750
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00014688306897823683,
+      "loss": 0.252,
+      "step": 5760
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.0001467908520841018,
+      "loss": 0.2525,
+      "step": 5770
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0001466986351899668,
+      "loss": 0.3201,
+      "step": 5780
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00014660641829583178,
+      "loss": 0.2411,
+      "step": 5790
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0001465142014016968,
+      "loss": 0.3181,
+      "step": 5800
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0001464219845075618,
+      "loss": 0.2225,
+      "step": 5810
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0001463297676134268,
+      "loss": 0.4026,
+      "step": 5820
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00014623755071929177,
+      "loss": 0.3204,
+      "step": 5830
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00014614533382515679,
+      "loss": 0.2252,
+      "step": 5840
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00014605311693102177,
+      "loss": 0.2579,
+      "step": 5850
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00014596090003688675,
+      "loss": 0.2791,
+      "step": 5860
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00014586868314275174,
+      "loss": 0.2399,
+      "step": 5870
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00014577646624861675,
+      "loss": 0.2829,
+      "step": 5880
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00014568424935448176,
+      "loss": 0.1904,
+      "step": 5890
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00014559203246034675,
+      "loss": 0.234,
+      "step": 5900
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00014549981556621173,
+      "loss": 0.243,
+      "step": 5910
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00014540759867207674,
+      "loss": 0.3356,
+      "step": 5920
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00014531538177794173,
+      "loss": 0.2302,
+      "step": 5930
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0001452231648838067,
+      "loss": 0.2237,
+      "step": 5940
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0001451309479896717,
+      "loss": 0.2933,
+      "step": 5950
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0001450387310955367,
+      "loss": 0.1666,
+      "step": 5960
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00014494651420140172,
+      "loss": 0.2424,
+      "step": 5970
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0001448542973072667,
+      "loss": 0.1896,
+      "step": 5980
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0001447620804131317,
+      "loss": 0.2417,
+      "step": 5990
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.0001446698635189967,
+      "loss": 0.2484,
+      "step": 6000
+    },
+    {
+      "epoch": 1.11,
+      "eval_accuracy": 0.8888990323169619,
+      "eval_loss": 0.25469714403152466,
+      "eval_runtime": 99.1559,
+      "eval_samples_per_second": 110.472,
+      "eval_steps_per_second": 13.817,
+      "step": 6000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00014457764662486168,
+      "loss": 0.1637,
+      "step": 6010
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00014448542973072667,
+      "loss": 0.25,
+      "step": 6020
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00014439321283659165,
+      "loss": 0.2656,
+      "step": 6030
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00014430099594245666,
+      "loss": 0.231,
+      "step": 6040
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00014420877904832167,
+      "loss": 0.2561,
+      "step": 6050
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00014411656215418666,
+      "loss": 0.2932,
+      "step": 6060
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00014402434526005164,
+      "loss": 0.2187,
+      "step": 6070
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00014393212836591665,
+      "loss": 0.2097,
+      "step": 6080
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00014383991147178164,
+      "loss": 0.2195,
+      "step": 6090
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00014374769457764662,
+      "loss": 0.2479,
+      "step": 6100
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.0001436554776835116,
+      "loss": 0.2269,
+      "step": 6110
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00014356326078937662,
+      "loss": 0.3276,
+      "step": 6120
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00014347104389524163,
+      "loss": 0.1641,
+      "step": 6130
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00014337882700110662,
+      "loss": 0.2639,
+      "step": 6140
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.0001432866101069716,
+      "loss": 0.2498,
+      "step": 6150
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.0001431943932128366,
+      "loss": 0.259,
+      "step": 6160
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.0001431021763187016,
+      "loss": 0.242,
+      "step": 6170
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00014300995942456658,
+      "loss": 0.3055,
+      "step": 6180
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00014291774253043156,
+      "loss": 0.2876,
+      "step": 6190
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00014282552563629658,
+      "loss": 0.2389,
+      "step": 6200
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0001427333087421616,
+      "loss": 0.2328,
+      "step": 6210
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00014264109184802657,
+      "loss": 0.2067,
+      "step": 6220
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00014254887495389156,
+      "loss": 0.2483,
+      "step": 6230
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00014245665805975657,
+      "loss": 0.1956,
+      "step": 6240
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00014236444116562155,
+      "loss": 0.1735,
+      "step": 6250
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00014227222427148654,
+      "loss": 0.2671,
+      "step": 6260
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00014218000737735152,
+      "loss": 0.191,
+      "step": 6270
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00014208779048321653,
+      "loss": 0.2442,
+      "step": 6280
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00014199557358908154,
+      "loss": 0.3221,
+      "step": 6290
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00014190335669494653,
+      "loss": 0.2409,
+      "step": 6300
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.0001418111398008115,
+      "loss": 0.3028,
+      "step": 6310
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0001417189229066765,
+      "loss": 0.2569,
+      "step": 6320
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0001416267060125415,
+      "loss": 0.2352,
+      "step": 6330
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0001415344891184065,
+      "loss": 0.2273,
+      "step": 6340
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00014144227222427148,
+      "loss": 0.2696,
+      "step": 6350
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0001413500553301365,
+      "loss": 0.2499,
+      "step": 6360
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0001412578384360015,
+      "loss": 0.2661,
+      "step": 6370
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00014116562154186649,
+      "loss": 0.1747,
+      "step": 6380
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00014107340464773147,
+      "loss": 0.2103,
+      "step": 6390
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00014098118775359645,
+      "loss": 0.2952,
+      "step": 6400
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00014088897085946147,
+      "loss": 0.3027,
+      "step": 6410
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00014079675396532645,
+      "loss": 0.1631,
+      "step": 6420
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00014070453707119143,
+      "loss": 0.2349,
+      "step": 6430
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00014061232017705645,
+      "loss": 0.2795,
+      "step": 6440
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00014052010328292146,
+      "loss": 0.2568,
+      "step": 6450
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00014042788638878644,
+      "loss": 0.3058,
+      "step": 6460
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00014033566949465143,
+      "loss": 0.2498,
+      "step": 6470
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0001402434526005164,
+      "loss": 0.283,
+      "step": 6480
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00014015123570638142,
+      "loss": 0.2488,
+      "step": 6490
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0001400590188122464,
+      "loss": 0.3068,
+      "step": 6500
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0001399668019181114,
+      "loss": 0.1956,
+      "step": 6510
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0001398745850239764,
+      "loss": 0.3158,
+      "step": 6520
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00013978236812984141,
+      "loss": 0.2339,
+      "step": 6530
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.0001396901512357064,
+      "loss": 0.1909,
+      "step": 6540
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00013959793434157138,
+      "loss": 0.1971,
+      "step": 6550
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00013950571744743637,
+      "loss": 0.2328,
+      "step": 6560
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00013941350055330138,
+      "loss": 0.2454,
+      "step": 6570
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00013932128365916636,
+      "loss": 0.2013,
+      "step": 6580
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00013922906676503135,
+      "loss": 0.2084,
+      "step": 6590
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00013913684987089636,
+      "loss": 0.2192,
+      "step": 6600
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00013904463297676137,
+      "loss": 0.205,
+      "step": 6610
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00013895241608262635,
+      "loss": 0.271,
+      "step": 6620
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00013886019918849134,
+      "loss": 0.2771,
+      "step": 6630
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00013876798229435632,
+      "loss": 0.2482,
+      "step": 6640
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00013867576540022134,
+      "loss": 0.2578,
+      "step": 6650
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00013858354850608632,
+      "loss": 0.23,
+      "step": 6660
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.0001384913316119513,
+      "loss": 0.1738,
+      "step": 6670
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00013839911471781632,
+      "loss": 0.2621,
+      "step": 6680
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00013830689782368133,
+      "loss": 0.2004,
+      "step": 6690
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0001382146809295463,
+      "loss": 0.1957,
+      "step": 6700
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0001381224640354113,
+      "loss": 0.2362,
+      "step": 6710
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00013803024714127628,
+      "loss": 0.2001,
+      "step": 6720
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0001379380302471413,
+      "loss": 0.2112,
+      "step": 6730
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00013784581335300628,
+      "loss": 0.1598,
+      "step": 6740
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00013775359645887126,
+      "loss": 0.2422,
+      "step": 6750
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00013766137956473627,
+      "loss": 0.2663,
+      "step": 6760
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00013756916267060128,
+      "loss": 0.1904,
+      "step": 6770
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00013747694577646627,
+      "loss": 0.2727,
+      "step": 6780
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00013738472888233125,
+      "loss": 0.2276,
+      "step": 6790
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00013729251198819624,
+      "loss": 0.2364,
+      "step": 6800
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00013720029509406125,
+      "loss": 0.2491,
+      "step": 6810
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00013710807819992623,
+      "loss": 0.2182,
+      "step": 6820
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00013701586130579122,
+      "loss": 0.1737,
+      "step": 6830
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00013692364441165623,
+      "loss": 0.2261,
+      "step": 6840
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00013683142751752124,
+      "loss": 0.3482,
+      "step": 6850
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00013673921062338622,
+      "loss": 0.2293,
+      "step": 6860
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.0001366469937292512,
+      "loss": 0.2106,
+      "step": 6870
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.0001365547768351162,
+      "loss": 0.2274,
+      "step": 6880
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.0001364625599409812,
+      "loss": 0.2122,
+      "step": 6890
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.0001363703430468462,
+      "loss": 0.2738,
+      "step": 6900
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00013627812615271117,
+      "loss": 0.2573,
+      "step": 6910
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00013618590925857619,
+      "loss": 0.2886,
+      "step": 6920
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.0001360936923644412,
+      "loss": 0.2405,
+      "step": 6930
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00013600147547030618,
+      "loss": 0.2874,
+      "step": 6940
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00013590925857617117,
+      "loss": 0.2308,
+      "step": 6950
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00013581704168203615,
+      "loss": 0.1814,
+      "step": 6960
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00013572482478790113,
+      "loss": 0.2587,
+      "step": 6970
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00013563260789376615,
+      "loss": 0.1863,
+      "step": 6980
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00013554039099963113,
+      "loss": 0.238,
+      "step": 6990
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00013544817410549614,
+      "loss": 0.1933,
+      "step": 7000
+    },
+    {
+      "epoch": 1.29,
+      "eval_accuracy": 0.8900858134014972,
+      "eval_loss": 0.25212499499320984,
+      "eval_runtime": 99.1283,
+      "eval_samples_per_second": 110.503,
+      "eval_steps_per_second": 13.82,
+      "step": 7000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00013535595721136113,
+      "loss": 0.3187,
+      "step": 7010
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00013526374031722614,
+      "loss": 0.2061,
+      "step": 7020
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00013517152342309112,
+      "loss": 0.1919,
+      "step": 7030
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.0001350793065289561,
+      "loss": 0.2051,
+      "step": 7040
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.0001349870896348211,
+      "loss": 0.1793,
+      "step": 7050
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.0001348948727406861,
+      "loss": 0.1765,
+      "step": 7060
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.0001348026558465511,
+      "loss": 0.267,
+      "step": 7070
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.0001347104389524161,
+      "loss": 0.1876,
+      "step": 7080
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00013461822205828108,
+      "loss": 0.255,
+      "step": 7090
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.0001345260051641461,
+      "loss": 0.1907,
+      "step": 7100
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00013443378827001108,
+      "loss": 0.3561,
+      "step": 7110
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00013434157137587606,
+      "loss": 0.2443,
+      "step": 7120
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00013424935448174105,
+      "loss": 0.1717,
+      "step": 7130
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00013415713758760606,
+      "loss": 0.2216,
+      "step": 7140
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00013406492069347104,
+      "loss": 0.1926,
+      "step": 7150
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00013397270379933603,
+      "loss": 0.3015,
+      "step": 7160
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00013388048690520104,
+      "loss": 0.1808,
+      "step": 7170
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00013378827001106605,
+      "loss": 0.2813,
+      "step": 7180
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00013369605311693104,
+      "loss": 0.2651,
+      "step": 7190
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00013360383622279602,
+      "loss": 0.1868,
+      "step": 7200
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.000133511619328661,
+      "loss": 0.2673,
+      "step": 7210
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00013341940243452602,
+      "loss": 0.1917,
+      "step": 7220
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.000133327185540391,
+      "loss": 0.2866,
+      "step": 7230
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00013323496864625598,
+      "loss": 0.269,
+      "step": 7240
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.000133142751752121,
+      "loss": 0.205,
+      "step": 7250
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.000133050534857986,
+      "loss": 0.2133,
+      "step": 7260
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.000132958317963851,
+      "loss": 0.2186,
+      "step": 7270
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00013286610106971598,
+      "loss": 0.2304,
+      "step": 7280
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00013277388417558096,
+      "loss": 0.189,
+      "step": 7290
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00013268166728144597,
+      "loss": 0.2829,
+      "step": 7300
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00013258945038731096,
+      "loss": 0.2703,
+      "step": 7310
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00013249723349317594,
+      "loss": 0.1833,
+      "step": 7320
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00013240501659904095,
+      "loss": 0.2662,
+      "step": 7330
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00013231279970490596,
+      "loss": 0.2284,
+      "step": 7340
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00013222058281077095,
+      "loss": 0.1519,
+      "step": 7350
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00013212836591663593,
+      "loss": 0.2817,
+      "step": 7360
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00013203614902250092,
+      "loss": 0.1787,
+      "step": 7370
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00013194393212836593,
+      "loss": 0.1595,
+      "step": 7380
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.0001318517152342309,
+      "loss": 0.2201,
+      "step": 7390
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.0001317594983400959,
+      "loss": 0.2606,
+      "step": 7400
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.0001316672814459609,
+      "loss": 0.2759,
+      "step": 7410
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00013157506455182592,
+      "loss": 0.2757,
+      "step": 7420
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.0001314828476576909,
+      "loss": 0.2708,
+      "step": 7430
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.0001313906307635559,
+      "loss": 0.1542,
+      "step": 7440
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00013129841386942087,
+      "loss": 0.2149,
+      "step": 7450
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00013120619697528589,
+      "loss": 0.1677,
+      "step": 7460
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00013111398008115087,
+      "loss": 0.2352,
+      "step": 7470
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00013102176318701585,
+      "loss": 0.2292,
+      "step": 7480
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00013092954629288087,
+      "loss": 0.2387,
+      "step": 7490
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00013083732939874588,
+      "loss": 0.2278,
+      "step": 7500
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00013074511250461086,
+      "loss": 0.1767,
+      "step": 7510
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00013065289561047585,
+      "loss": 0.1616,
+      "step": 7520
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00013056067871634083,
+      "loss": 0.2363,
+      "step": 7530
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00013046846182220584,
+      "loss": 0.1996,
+      "step": 7540
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00013037624492807083,
+      "loss": 0.2386,
+      "step": 7550
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.0001302840280339358,
+      "loss": 0.1635,
+      "step": 7560
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00013019181113980082,
+      "loss": 0.1882,
+      "step": 7570
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.0001300995942456658,
+      "loss": 0.2306,
+      "step": 7580
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00013000737735153082,
+      "loss": 0.3108,
+      "step": 7590
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.0001299151604573958,
+      "loss": 0.1904,
+      "step": 7600
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.0001298229435632608,
+      "loss": 0.3081,
+      "step": 7610
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00012973072666912577,
+      "loss": 0.1683,
+      "step": 7620
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00012963850977499078,
+      "loss": 0.2353,
+      "step": 7630
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00012954629288085577,
+      "loss": 0.1807,
+      "step": 7640
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00012945407598672078,
+      "loss": 0.1619,
+      "step": 7650
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00012936185909258576,
+      "loss": 0.1439,
+      "step": 7660
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00012926964219845077,
+      "loss": 0.2855,
+      "step": 7670
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00012917742530431576,
+      "loss": 0.2415,
+      "step": 7680
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00012908520841018074,
+      "loss": 0.112,
+      "step": 7690
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00012899299151604573,
+      "loss": 0.3052,
+      "step": 7700
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00012890077462191074,
+      "loss": 0.2934,
+      "step": 7710
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00012880855772777572,
+      "loss": 0.2106,
+      "step": 7720
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00012871634083364074,
+      "loss": 0.2402,
+      "step": 7730
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00012862412393950572,
+      "loss": 0.2989,
+      "step": 7740
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00012853190704537073,
+      "loss": 0.2562,
+      "step": 7750
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00012843969015123572,
+      "loss": 0.2612,
+      "step": 7760
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.0001283474732571007,
+      "loss": 0.2899,
+      "step": 7770
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00012825525636296568,
+      "loss": 0.1914,
+      "step": 7780
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.0001281630394688307,
+      "loss": 0.2035,
+      "step": 7790
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00012807082257469568,
+      "loss": 0.1966,
+      "step": 7800
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.0001279786056805607,
+      "loss": 0.3944,
+      "step": 7810
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00012788638878642568,
+      "loss": 0.3371,
+      "step": 7820
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.0001277941718922907,
+      "loss": 0.2058,
+      "step": 7830
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00012770195499815567,
+      "loss": 0.286,
+      "step": 7840
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00012760973810402066,
+      "loss": 0.2834,
+      "step": 7850
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00012751752120988564,
+      "loss": 0.1462,
+      "step": 7860
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00012742530431575065,
+      "loss": 0.2344,
+      "step": 7870
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00012733308742161564,
+      "loss": 0.2592,
+      "step": 7880
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00012724087052748065,
+      "loss": 0.3563,
+      "step": 7890
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00012714865363334563,
+      "loss": 0.1902,
+      "step": 7900
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00012705643673921064,
+      "loss": 0.2049,
+      "step": 7910
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00012696421984507563,
+      "loss": 0.2537,
+      "step": 7920
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.0001268720029509406,
+      "loss": 0.2205,
+      "step": 7930
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.0001267797860568056,
+      "loss": 0.1966,
+      "step": 7940
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.0001266875691626706,
+      "loss": 0.2639,
+      "step": 7950
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.0001265953522685356,
+      "loss": 0.2864,
+      "step": 7960
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.0001265031353744006,
+      "loss": 0.2161,
+      "step": 7970
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.0001264109184802656,
+      "loss": 0.2435,
+      "step": 7980
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.0001263187015861306,
+      "loss": 0.3763,
+      "step": 7990
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00012622648469199559,
+      "loss": 0.1497,
+      "step": 8000
+    },
+    {
+      "epoch": 1.48,
+      "eval_accuracy": 0.8929158298338506,
+      "eval_loss": 0.24426758289337158,
+      "eval_runtime": 99.0537,
+      "eval_samples_per_second": 110.586,
+      "eval_steps_per_second": 13.831,
+      "step": 8000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00012613426779786057,
+      "loss": 0.2817,
+      "step": 8010
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00012604205090372555,
+      "loss": 0.1709,
+      "step": 8020
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00012594983400959057,
+      "loss": 0.2355,
+      "step": 8030
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00012585761711545555,
+      "loss": 0.1934,
+      "step": 8040
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00012576540022132056,
+      "loss": 0.1798,
+      "step": 8050
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00012567318332718555,
+      "loss": 0.2171,
+      "step": 8060
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00012558096643305056,
+      "loss": 0.1634,
+      "step": 8070
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00012548874953891554,
+      "loss": 0.2284,
+      "step": 8080
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00012539653264478053,
+      "loss": 0.1464,
+      "step": 8090
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.0001253043157506455,
+      "loss": 0.1662,
+      "step": 8100
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00012521209885651052,
+      "loss": 0.2556,
+      "step": 8110
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.0001251198819623755,
+      "loss": 0.2933,
+      "step": 8120
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00012502766506824052,
+      "loss": 0.268,
+      "step": 8130
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.0001249354481741055,
+      "loss": 0.2616,
+      "step": 8140
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00012484323127997051,
+      "loss": 0.2024,
+      "step": 8150
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.0001247510143858355,
+      "loss": 0.2202,
+      "step": 8160
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00012465879749170048,
+      "loss": 0.2795,
+      "step": 8170
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00012456658059756547,
+      "loss": 0.1465,
+      "step": 8180
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00012447436370343048,
+      "loss": 0.2327,
+      "step": 8190
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00012438214680929546,
+      "loss": 0.2233,
+      "step": 8200
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00012428992991516047,
+      "loss": 0.2079,
+      "step": 8210
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00012419771302102546,
+      "loss": 0.2067,
+      "step": 8220
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00012410549612689044,
+      "loss": 0.231,
+      "step": 8230
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00012401327923275546,
+      "loss": 0.3321,
+      "step": 8240
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00012392106233862044,
+      "loss": 0.2292,
+      "step": 8250
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00012382884544448542,
+      "loss": 0.2285,
+      "step": 8260
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.0001237366285503504,
+      "loss": 0.145,
+      "step": 8270
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00012364441165621542,
+      "loss": 0.2573,
+      "step": 8280
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00012355219476208043,
+      "loss": 0.2345,
+      "step": 8290
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00012345997786794542,
+      "loss": 0.2883,
+      "step": 8300
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.0001233677609738104,
+      "loss": 0.2041,
+      "step": 8310
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.0001232755440796754,
+      "loss": 0.2447,
+      "step": 8320
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.0001231833271855404,
+      "loss": 0.1614,
+      "step": 8330
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00012309111029140538,
+      "loss": 0.1595,
+      "step": 8340
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00012299889339727036,
+      "loss": 0.1773,
+      "step": 8350
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00012290667650313538,
+      "loss": 0.2009,
+      "step": 8360
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.0001228144596090004,
+      "loss": 0.134,
+      "step": 8370
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00012272224271486537,
+      "loss": 0.2702,
+      "step": 8380
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00012263002582073036,
+      "loss": 0.1946,
+      "step": 8390
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00012253780892659537,
+      "loss": 0.2947,
+      "step": 8400
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00012244559203246035,
+      "loss": 0.3152,
+      "step": 8410
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00012235337513832534,
+      "loss": 0.2391,
+      "step": 8420
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00012226115824419032,
+      "loss": 0.2751,
+      "step": 8430
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00012216894135005533,
+      "loss": 0.204,
+      "step": 8440
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00012207672445592034,
+      "loss": 0.2294,
+      "step": 8450
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00012198450756178532,
+      "loss": 0.1751,
+      "step": 8460
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00012189229066765031,
+      "loss": 0.185,
+      "step": 8470
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00012180007377351532,
+      "loss": 0.2549,
+      "step": 8480
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00012170785687938031,
+      "loss": 0.2544,
+      "step": 8490
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.0001216156399852453,
+      "loss": 0.212,
+      "step": 8500
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00012152342309111029,
+      "loss": 0.2575,
+      "step": 8510
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.0001214312061969753,
+      "loss": 0.2937,
+      "step": 8520
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00012133898930284029,
+      "loss": 0.1906,
+      "step": 8530
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00012124677240870527,
+      "loss": 0.2207,
+      "step": 8540
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00012115455551457027,
+      "loss": 0.2573,
+      "step": 8550
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00012106233862043528,
+      "loss": 0.2415,
+      "step": 8560
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00012097012172630027,
+      "loss": 0.2171,
+      "step": 8570
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00012087790483216525,
+      "loss": 0.1182,
+      "step": 8580
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00012078568793803025,
+      "loss": 0.166,
+      "step": 8590
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00012069347104389526,
+      "loss": 0.2243,
+      "step": 8600
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00012060125414976024,
+      "loss": 0.2082,
+      "step": 8610
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00012050903725562523,
+      "loss": 0.2802,
+      "step": 8620
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00012041682036149023,
+      "loss": 0.2671,
+      "step": 8630
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00012032460346735524,
+      "loss": 0.1579,
+      "step": 8640
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00012023238657322022,
+      "loss": 0.2012,
+      "step": 8650
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.0001201401696790852,
+      "loss": 0.1878,
+      "step": 8660
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.0001200479527849502,
+      "loss": 0.2192,
+      "step": 8670
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00011995573589081522,
+      "loss": 0.2217,
+      "step": 8680
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.0001198635189966802,
+      "loss": 0.1999,
+      "step": 8690
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00011977130210254518,
+      "loss": 0.2146,
+      "step": 8700
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00011967908520841018,
+      "loss": 0.1185,
+      "step": 8710
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.0001195868683142752,
+      "loss": 0.1502,
+      "step": 8720
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00011949465142014018,
+      "loss": 0.2598,
+      "step": 8730
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00011940243452600516,
+      "loss": 0.2108,
+      "step": 8740
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00011931021763187016,
+      "loss": 0.1722,
+      "step": 8750
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00011921800073773517,
+      "loss": 0.1995,
+      "step": 8760
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00011912578384360016,
+      "loss": 0.2303,
+      "step": 8770
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00011903356694946514,
+      "loss": 0.1352,
+      "step": 8780
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00011894135005533014,
+      "loss": 0.2972,
+      "step": 8790
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00011884913316119515,
+      "loss": 0.2021,
+      "step": 8800
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00011875691626706014,
+      "loss": 0.2159,
+      "step": 8810
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00011866469937292512,
+      "loss": 0.2628,
+      "step": 8820
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00011857248247879012,
+      "loss": 0.2672,
+      "step": 8830
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.0001184802655846551,
+      "loss": 0.2086,
+      "step": 8840
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00011838804869052011,
+      "loss": 0.2027,
+      "step": 8850
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.0001182958317963851,
+      "loss": 0.2272,
+      "step": 8860
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.0001182036149022501,
+      "loss": 0.284,
+      "step": 8870
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00011811139800811508,
+      "loss": 0.2287,
+      "step": 8880
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00011801918111398009,
+      "loss": 0.2576,
+      "step": 8890
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00011792696421984508,
+      "loss": 0.1977,
+      "step": 8900
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00011783474732571007,
+      "loss": 0.2285,
+      "step": 8910
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00011774253043157506,
+      "loss": 0.2527,
+      "step": 8920
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00011765031353744007,
+      "loss": 0.1921,
+      "step": 8930
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00011755809664330505,
+      "loss": 0.1627,
+      "step": 8940
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00011746587974917005,
+      "loss": 0.1663,
+      "step": 8950
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00011737366285503504,
+      "loss": 0.2219,
+      "step": 8960
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00011728144596090005,
+      "loss": 0.2183,
+      "step": 8970
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00011718922906676503,
+      "loss": 0.1422,
+      "step": 8980
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00011709701217263003,
+      "loss": 0.3236,
+      "step": 8990
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00011700479527849502,
+      "loss": 0.326,
+      "step": 9000
+    },
+    {
+      "epoch": 1.66,
+      "eval_accuracy": 0.8958371371188607,
+      "eval_loss": 0.24060481786727905,
+      "eval_runtime": 99.1063,
+      "eval_samples_per_second": 110.528,
+      "eval_steps_per_second": 13.824,
+      "step": 9000
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00011691257838436003,
+      "loss": 0.2334,
+      "step": 9010
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00011682036149022501,
+      "loss": 0.2106,
+      "step": 9020
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00011672814459609001,
+      "loss": 0.1898,
+      "step": 9030
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.000116635927701955,
+      "loss": 0.2647,
+      "step": 9040
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00011654371080782,
+      "loss": 0.2088,
+      "step": 9050
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00011645149391368499,
+      "loss": 0.2472,
+      "step": 9060
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00011635927701954999,
+      "loss": 0.1796,
+      "step": 9070
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00011626706012541497,
+      "loss": 0.1225,
+      "step": 9080
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00011617484323127998,
+      "loss": 0.3306,
+      "step": 9090
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00011608262633714497,
+      "loss": 0.2479,
+      "step": 9100
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00011599040944300997,
+      "loss": 0.2157,
+      "step": 9110
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00011589819254887495,
+      "loss": 0.1434,
+      "step": 9120
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00011580597565473996,
+      "loss": 0.1893,
+      "step": 9130
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00011571375876060495,
+      "loss": 0.2089,
+      "step": 9140
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00011562154186646994,
+      "loss": 0.1933,
+      "step": 9150
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00011552932497233493,
+      "loss": 0.2579,
+      "step": 9160
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00011543710807819994,
+      "loss": 0.2344,
+      "step": 9170
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00011534489118406492,
+      "loss": 0.1639,
+      "step": 9180
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00011525267428992992,
+      "loss": 0.2719,
+      "step": 9190
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.0001151604573957949,
+      "loss": 0.2147,
+      "step": 9200
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00011506824050165992,
+      "loss": 0.2561,
+      "step": 9210
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.0001149760236075249,
+      "loss": 0.2279,
+      "step": 9220
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.0001148838067133899,
+      "loss": 0.2453,
+      "step": 9230
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00011479158981925489,
+      "loss": 0.2262,
+      "step": 9240
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.0001146993729251199,
+      "loss": 0.2004,
+      "step": 9250
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00011460715603098488,
+      "loss": 0.1751,
+      "step": 9260
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00011451493913684988,
+      "loss": 0.26,
+      "step": 9270
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00011442272224271486,
+      "loss": 0.1794,
+      "step": 9280
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00011433050534857987,
+      "loss": 0.2668,
+      "step": 9290
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00011423828845444486,
+      "loss": 0.2006,
+      "step": 9300
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00011414607156030986,
+      "loss": 0.2378,
+      "step": 9310
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00011405385466617484,
+      "loss": 0.1551,
+      "step": 9320
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00011396163777203985,
+      "loss": 0.1739,
+      "step": 9330
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00011386942087790484,
+      "loss": 0.1707,
+      "step": 9340
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00011377720398376984,
+      "loss": 0.2313,
+      "step": 9350
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00011368498708963482,
+      "loss": 0.1719,
+      "step": 9360
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00011359277019549983,
+      "loss": 0.2656,
+      "step": 9370
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00011350055330136482,
+      "loss": 0.1628,
+      "step": 9380
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00011340833640722981,
+      "loss": 0.2269,
+      "step": 9390
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.0001133161195130948,
+      "loss": 0.2695,
+      "step": 9400
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00011322390261895981,
+      "loss": 0.2848,
+      "step": 9410
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.0001131316857248248,
+      "loss": 0.2628,
+      "step": 9420
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00011303946883068979,
+      "loss": 0.2679,
+      "step": 9430
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00011294725193655478,
+      "loss": 0.151,
+      "step": 9440
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00011285503504241979,
+      "loss": 0.2359,
+      "step": 9450
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00011276281814828477,
+      "loss": 0.1497,
+      "step": 9460
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00011267060125414977,
+      "loss": 0.2328,
+      "step": 9470
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00011257838436001475,
+      "loss": 0.1297,
+      "step": 9480
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00011248616746587974,
+      "loss": 0.2342,
+      "step": 9490
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00011239395057174475,
+      "loss": 0.2353,
+      "step": 9500
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00011230173367760975,
+      "loss": 0.2787,
+      "step": 9510
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00011220951678347473,
+      "loss": 0.2113,
+      "step": 9520
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00011211729988933972,
+      "loss": 0.2745,
+      "step": 9530
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00011202508299520473,
+      "loss": 0.1921,
+      "step": 9540
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00011193286610106973,
+      "loss": 0.179,
+      "step": 9550
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00011184064920693471,
+      "loss": 0.1999,
+      "step": 9560
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.0001117484323127997,
+      "loss": 0.1753,
+      "step": 9570
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00011165621541866471,
+      "loss": 0.2441,
+      "step": 9580
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.0001115639985245297,
+      "loss": 0.2039,
+      "step": 9590
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00011147178163039469,
+      "loss": 0.179,
+      "step": 9600
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00011137956473625967,
+      "loss": 0.2161,
+      "step": 9610
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00011128734784212469,
+      "loss": 0.1772,
+      "step": 9620
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00011119513094798968,
+      "loss": 0.2605,
+      "step": 9630
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00011110291405385467,
+      "loss": 0.2228,
+      "step": 9640
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00011101069715971965,
+      "loss": 0.223,
+      "step": 9650
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00011091848026558466,
+      "loss": 0.2553,
+      "step": 9660
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00011082626337144966,
+      "loss": 0.2314,
+      "step": 9670
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00011073404647731465,
+      "loss": 0.2686,
+      "step": 9680
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00011064182958317963,
+      "loss": 0.2472,
+      "step": 9690
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00011054961268904464,
+      "loss": 0.2428,
+      "step": 9700
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00011045739579490964,
+      "loss": 0.136,
+      "step": 9710
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00011036517890077462,
+      "loss": 0.1727,
+      "step": 9720
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00011027296200663961,
+      "loss": 0.1694,
+      "step": 9730
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00011018074511250462,
+      "loss": 0.1562,
+      "step": 9740
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00011008852821836962,
+      "loss": 0.2279,
+      "step": 9750
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.0001099963113242346,
+      "loss": 0.1468,
+      "step": 9760
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00010990409443009959,
+      "loss": 0.2378,
+      "step": 9770
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.0001098118775359646,
+      "loss": 0.1784,
+      "step": 9780
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.0001097196606418296,
+      "loss": 0.2022,
+      "step": 9790
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00010962744374769458,
+      "loss": 0.159,
+      "step": 9800
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00010953522685355957,
+      "loss": 0.2206,
+      "step": 9810
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00010944300995942458,
+      "loss": 0.1881,
+      "step": 9820
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00010935079306528957,
+      "loss": 0.1979,
+      "step": 9830
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00010925857617115456,
+      "loss": 0.2833,
+      "step": 9840
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00010916635927701954,
+      "loss": 0.2631,
+      "step": 9850
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00010907414238288456,
+      "loss": 0.1653,
+      "step": 9860
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00010898192548874955,
+      "loss": 0.1997,
+      "step": 9870
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00010888970859461454,
+      "loss": 0.121,
+      "step": 9880
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00010879749170047952,
+      "loss": 0.2092,
+      "step": 9890
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.00010870527480634453,
+      "loss": 0.1775,
+      "step": 9900
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.00010861305791220953,
+      "loss": 0.2089,
+      "step": 9910
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.00010852084101807452,
+      "loss": 0.2334,
+      "step": 9920
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.0001084286241239395,
+      "loss": 0.1916,
+      "step": 9930
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.00010833640722980451,
+      "loss": 0.2291,
+      "step": 9940
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.00010824419033566951,
+      "loss": 0.2331,
+      "step": 9950
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.0001081519734415345,
+      "loss": 0.1836,
+      "step": 9960
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.00010805975654739948,
+      "loss": 0.2869,
+      "step": 9970
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.00010796753965326449,
+      "loss": 0.1925,
+      "step": 9980
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.00010787532275912949,
+      "loss": 0.1442,
+      "step": 9990
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.00010778310586499447,
+      "loss": 0.215,
+      "step": 10000
+    },
+    {
+      "epoch": 1.84,
+      "eval_accuracy": 0.9006755523096586,
+      "eval_loss": 0.23809599876403809,
+      "eval_runtime": 99.4319,
+      "eval_samples_per_second": 110.166,
+      "eval_steps_per_second": 13.778,
+      "step": 10000
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.00010769088897085946,
+      "loss": 0.1787,
+      "step": 10010
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.00010759867207672447,
+      "loss": 0.2511,
+      "step": 10020
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.00010750645518258947,
+      "loss": 0.2396,
+      "step": 10030
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.00010741423828845445,
+      "loss": 0.1376,
+      "step": 10040
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.00010732202139431944,
+      "loss": 0.2806,
+      "step": 10050
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.00010722980450018445,
+      "loss": 0.238,
+      "step": 10060
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.00010713758760604944,
+      "loss": 0.2295,
+      "step": 10070
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.00010704537071191443,
+      "loss": 0.2549,
+      "step": 10080
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.00010695315381777941,
+      "loss": 0.1697,
+      "step": 10090
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.00010686093692364441,
+      "loss": 0.2492,
+      "step": 10100
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.00010676872002950942,
+      "loss": 0.2332,
+      "step": 10110
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.00010667650313537441,
+      "loss": 0.2462,
+      "step": 10120
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.00010658428624123939,
+      "loss": 0.1817,
+      "step": 10130
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.00010649206934710439,
+      "loss": 0.1883,
+      "step": 10140
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.0001063998524529694,
+      "loss": 0.1145,
+      "step": 10150
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.00010630763555883439,
+      "loss": 0.1207,
+      "step": 10160
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.00010621541866469937,
+      "loss": 0.1852,
+      "step": 10170
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.00010612320177056437,
+      "loss": 0.2175,
+      "step": 10180
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.00010603098487642938,
+      "loss": 0.1844,
+      "step": 10190
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.00010593876798229436,
+      "loss": 0.2802,
+      "step": 10200
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.00010584655108815935,
+      "loss": 0.2221,
+      "step": 10210
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.00010575433419402435,
+      "loss": 0.2463,
+      "step": 10220
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.00010566211729988936,
+      "loss": 0.1879,
+      "step": 10230
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.00010556990040575434,
+      "loss": 0.1723,
+      "step": 10240
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.00010547768351161933,
+      "loss": 0.1791,
+      "step": 10250
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.00010538546661748432,
+      "loss": 0.2838,
+      "step": 10260
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.00010529324972334934,
+      "loss": 0.2493,
+      "step": 10270
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.00010520103282921432,
+      "loss": 0.1853,
+      "step": 10280
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.0001051088159350793,
+      "loss": 0.2794,
+      "step": 10290
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.0001050165990409443,
+      "loss": 0.2217,
+      "step": 10300
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.00010492438214680931,
+      "loss": 0.1742,
+      "step": 10310
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.0001048321652526743,
+      "loss": 0.2308,
+      "step": 10320
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.00010473994835853928,
+      "loss": 0.1497,
+      "step": 10330
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.00010464773146440428,
+      "loss": 0.2586,
+      "step": 10340
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.00010455551457026929,
+      "loss": 0.2479,
+      "step": 10350
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.00010446329767613428,
+      "loss": 0.1498,
+      "step": 10360
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.00010437108078199926,
+      "loss": 0.1545,
+      "step": 10370
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.00010427886388786426,
+      "loss": 0.1987,
+      "step": 10380
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.00010418664699372927,
+      "loss": 0.1942,
+      "step": 10390
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.00010409443009959426,
+      "loss": 0.1845,
+      "step": 10400
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.00010400221320545924,
+      "loss": 0.1511,
+      "step": 10410
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.00010390999631132424,
+      "loss": 0.198,
+      "step": 10420
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.00010381777941718925,
+      "loss": 0.2414,
+      "step": 10430
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.00010372556252305423,
+      "loss": 0.169,
+      "step": 10440
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.00010363334562891922,
+      "loss": 0.2181,
+      "step": 10450
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.00010354112873478422,
+      "loss": 0.1829,
+      "step": 10460
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.00010344891184064923,
+      "loss": 0.1543,
+      "step": 10470
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.00010335669494651421,
+      "loss": 0.2585,
+      "step": 10480
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.0001032644780523792,
+      "loss": 0.1768,
+      "step": 10490
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.0001031722611582442,
+      "loss": 0.2628,
+      "step": 10500
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.0001030800442641092,
+      "loss": 0.2547,
+      "step": 10510
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.00010298782736997419,
+      "loss": 0.2313,
+      "step": 10520
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.00010289561047583917,
+      "loss": 0.1581,
+      "step": 10530
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.00010280339358170417,
+      "loss": 0.1884,
+      "step": 10540
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.00010271117668756917,
+      "loss": 0.2104,
+      "step": 10550
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.00010261895979343417,
+      "loss": 0.1989,
+      "step": 10560
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.00010252674289929915,
+      "loss": 0.213,
+      "step": 10570
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.00010243452600516415,
+      "loss": 0.1792,
+      "step": 10580
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.00010234230911102915,
+      "loss": 0.1808,
+      "step": 10590
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.00010225009221689415,
+      "loss": 0.1969,
+      "step": 10600
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.00010215787532275913,
+      "loss": 0.2469,
+      "step": 10610
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.00010206565842862413,
+      "loss": 0.1471,
+      "step": 10620
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.00010197344153448913,
+      "loss": 0.2731,
+      "step": 10630
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.00010188122464035413,
+      "loss": 0.1954,
+      "step": 10640
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.00010178900774621911,
+      "loss": 0.1555,
+      "step": 10650
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.00010169679085208411,
+      "loss": 0.1597,
+      "step": 10660
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.0001016045739579491,
+      "loss": 0.1966,
+      "step": 10670
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.0001015123570638141,
+      "loss": 0.183,
+      "step": 10680
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.00010142014016967909,
+      "loss": 0.2784,
+      "step": 10690
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.00010132792327554407,
+      "loss": 0.3295,
+      "step": 10700
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.00010123570638140908,
+      "loss": 0.2518,
+      "step": 10710
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.00010114348948727408,
+      "loss": 0.2137,
+      "step": 10720
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.00010105127259313907,
+      "loss": 0.1548,
+      "step": 10730
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.00010095905569900405,
+      "loss": 0.3132,
+      "step": 10740
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.00010086683880486905,
+      "loss": 0.2445,
+      "step": 10750
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.00010077462191073406,
+      "loss": 0.2612,
+      "step": 10760
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.00010068240501659904,
+      "loss": 0.2073,
+      "step": 10770
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.00010059018812246403,
+      "loss": 0.2407,
+      "step": 10780
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.00010049797122832903,
+      "loss": 0.139,
+      "step": 10790
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.00010040575433419404,
+      "loss": 0.2008,
+      "step": 10800
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.00010031353744005902,
+      "loss": 0.3201,
+      "step": 10810
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 0.00010022132054592401,
+      "loss": 0.2554,
+      "step": 10820
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 0.000100129103651789,
+      "loss": 0.2992,
+      "step": 10830
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 0.00010003688675765402,
+      "loss": 0.2177,
+      "step": 10840
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 9.9944669863519e-05,
+      "loss": 0.1616,
+      "step": 10850
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 9.985245296938399e-05,
+      "loss": 0.1123,
+      "step": 10860
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 9.9760236075249e-05,
+      "loss": 0.1704,
+      "step": 10870
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 9.966801918111398e-05,
+      "loss": 0.1256,
+      "step": 10880
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 9.957580228697898e-05,
+      "loss": 0.1458,
+      "step": 10890
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 9.948358539284396e-05,
+      "loss": 0.2031,
+      "step": 10900
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 9.939136849870898e-05,
+      "loss": 0.204,
+      "step": 10910
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 9.929915160457396e-05,
+      "loss": 0.2048,
+      "step": 10920
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 9.920693471043896e-05,
+      "loss": 0.1439,
+      "step": 10930
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 9.911471781630394e-05,
+      "loss": 0.1381,
+      "step": 10940
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 9.902250092216895e-05,
+      "loss": 0.0731,
+      "step": 10950
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 9.893028402803394e-05,
+      "loss": 0.2045,
+      "step": 10960
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 9.883806713389894e-05,
+      "loss": 0.1455,
+      "step": 10970
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 9.874585023976392e-05,
+      "loss": 0.1463,
+      "step": 10980
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 9.865363334562893e-05,
+      "loss": 0.12,
+      "step": 10990
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 9.856141645149392e-05,
+      "loss": 0.2035,
+      "step": 11000
+    },
+    {
+      "epoch": 2.03,
+      "eval_accuracy": 0.8985758626985576,
+      "eval_loss": 0.24368144571781158,
+      "eval_runtime": 99.1817,
+      "eval_samples_per_second": 110.444,
+      "eval_steps_per_second": 13.813,
+      "step": 11000
+    },
+    {
+      "epoch": 2.03,
+      "step": 11000,
+      "total_flos": 3.7370912368449946e+18,
+      "train_loss": 0.29799577762321994,
+      "train_runtime": 2883.713,
+      "train_samples_per_second": 120.323,
+      "train_steps_per_second": 7.521
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 21688,
+  "num_train_epochs": 4,
+  "save_steps": 1000,
+  "total_flos": 3.7370912368449946e+18,
+  "trial_name": null,
+  "trial_params": null
+}