diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,13318 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.47001316036849033,
+  "eval_steps": 500,
+  "global_step": 20000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0003999999863730145,
+      "loss": 8.118,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0003999999454920598,
+      "loss": 6.3286,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0003999998773571415,
+      "loss": 6.1146,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00039999978196826884,
+      "loss": 5.8777,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0003999996593254548,
+      "loss": 5.5376,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00039999950942871623,
+      "loss": 5.2922,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0003999993322780734,
+      "loss": 5.0843,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00039999912787355056,
+      "loss": 4.8923,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0003999988962151755,
+      "loss": 4.7488,
+      "step": 90
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0003999986373029798,
+      "loss": 4.6656,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00039999835113699875,
+      "loss": 4.5166,
+      "step": 110
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00039999803771727135,
+      "loss": 4.4329,
+      "step": 120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0003999976970438403,
+      "loss": 4.408,
+      "step": 130
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00039999732911675205,
+      "loss": 4.2873,
+      "step": 140
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00039999693393605676,
+      "loss": 4.2245,
+      "step": 150
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0003999965115018082,
+      "loss": 4.2141,
+      "step": 160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0003999960618140639,
+      "loss": 4.0998,
+      "step": 170
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00039999558487288525,
+      "loss": 4.0249,
+      "step": 180
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00039999508067833727,
+      "loss": 3.9842,
+      "step": 190
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00039999454923048854,
+      "loss": 3.9469,
+      "step": 200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0003999939905294116,
+      "loss": 3.8331,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039999340457518245,
+      "loss": 3.8387,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039999279136788106,
+      "loss": 3.7919,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039999215090759095,
+      "loss": 3.7312,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039999148319439935,
+      "loss": 3.674,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039999078822839734,
+      "loss": 3.6375,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0003999900660096795,
+      "loss": 3.6078,
+      "step": 270
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0003999893165383444,
+      "loss": 3.5456,
+      "step": 280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039998853981449404,
+      "loss": 3.4769,
+      "step": 290
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039998773583823434,
+      "loss": 3.4389,
+      "step": 300
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039998690460967484,
+      "loss": 3.389,
+      "step": 310
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039998604612892876,
+      "loss": 3.4085,
+      "step": 320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039998516039611317,
+      "loss": 3.3085,
+      "step": 330
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039998424741134877,
+      "loss": 3.2635,
+      "step": 340
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0003999833071747598,
+      "loss": 3.2114,
+      "step": 350
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0003999823396864746,
+      "loss": 3.1907,
+      "step": 360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0003999813449466249,
+      "loss": 3.1599,
+      "step": 370
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0003999803229553463,
+      "loss": 3.1176,
+      "step": 380
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039997927371277804,
+      "loss": 3.0683,
+      "step": 390
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0003999781972190631,
+      "loss": 3.0275,
+      "step": 400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0003999770934743482,
+      "loss": 3.0021,
+      "step": 410
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0003999759624787837,
+      "loss": 2.9149,
+      "step": 420
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0003999748042325237,
+      "loss": 2.884,
+      "step": 430
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0003999736187357261,
+      "loss": 2.9275,
+      "step": 440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039997240598855247,
+      "loss": 2.9095,
+      "step": 450
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039997116599116796,
+      "loss": 2.9409,
+      "step": 460
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039996989874374167,
+      "loss": 2.8867,
+      "step": 470
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039996860424644623,
+      "loss": 2.849,
+      "step": 480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039996728249945806,
+      "loss": 2.8417,
+      "step": 490
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039996593350295724,
+      "loss": 2.8284,
+      "step": 500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039996455725712754,
+      "loss": 2.7558,
+      "step": 510
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039996315376215663,
+      "loss": 2.7321,
+      "step": 520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0003999617230182358,
+      "loss": 2.6914,
+      "step": 530
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039996026502555975,
+      "loss": 2.6755,
+      "step": 540
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039995877978432745,
+      "loss": 2.6632,
+      "step": 550
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039995726729474113,
+      "loss": 2.6723,
+      "step": 560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039995572755700695,
+      "loss": 2.5948,
+      "step": 570
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039995416057133477,
+      "loss": 2.6085,
+      "step": 580
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039995256633793805,
+      "loss": 2.5718,
+      "step": 590
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.000399950944857034,
+      "loss": 2.5738,
+      "step": 600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039994929612884375,
+      "loss": 2.5005,
+      "step": 610
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0003999476201535918,
+      "loss": 2.5343,
+      "step": 620
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0003999459169315066,
+      "loss": 2.547,
+      "step": 630
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0003999441864628203,
+      "loss": 2.5294,
+      "step": 640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0003999424287477686,
+      "loss": 2.518,
+      "step": 650
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039994064378659114,
+      "loss": 2.5036,
+      "step": 660
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039993883157953106,
+      "loss": 2.4465,
+      "step": 670
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0003999369921268353,
+      "loss": 2.4591,
+      "step": 680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039993512542875465,
+      "loss": 2.4806,
+      "step": 690
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0003999332314855434,
+      "loss": 2.4915,
+      "step": 700
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0003999313102974596,
+      "loss": 2.4737,
+      "step": 710
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039992936186476513,
+      "loss": 2.4002,
+      "step": 720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0003999273861877254,
+      "loss": 2.4821,
+      "step": 730
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0003999253832666098,
+      "loss": 2.445,
+      "step": 740
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0003999233531016911,
+      "loss": 2.3891,
+      "step": 750
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000399921295693246,
+      "loss": 2.3283,
+      "step": 760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039991921104155494,
+      "loss": 2.4011,
+      "step": 770
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0003999170991469019,
+      "loss": 2.4398,
+      "step": 780
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039991496000957473,
+      "loss": 2.4085,
+      "step": 790
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039991279362986483,
+      "loss": 2.3319,
+      "step": 800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039991060000806755,
+      "loss": 2.3053,
+      "step": 810
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039990837914448174,
+      "loss": 2.3027,
+      "step": 820
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039990613103941007,
+      "loss": 2.4183,
+      "step": 830
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0003999038556931588,
+      "loss": 2.3537,
+      "step": 840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0003999015531060381,
+      "loss": 2.3063,
+      "step": 850
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0003998992232783617,
+      "loss": 2.3585,
+      "step": 860
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039989686621044713,
+      "loss": 2.3435,
+      "step": 870
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039989448190261545,
+      "loss": 2.3105,
+      "step": 880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0003998920703551917,
+      "loss": 2.3027,
+      "step": 890
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0003998896315685045,
+      "loss": 2.2045,
+      "step": 900
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0003998871655428861,
+      "loss": 2.2763,
+      "step": 910
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0003998846722786726,
+      "loss": 2.2635,
+      "step": 920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039988215177620375,
+      "loss": 2.2985,
+      "step": 930
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039987960403582305,
+      "loss": 2.2659,
+      "step": 940
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039987702905787756,
+      "loss": 2.2959,
+      "step": 950
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039987442684271833,
+      "loss": 2.2554,
+      "step": 960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0003998717973906999,
+      "loss": 2.2962,
+      "step": 970
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039986914070218055,
+      "loss": 2.2713,
+      "step": 980
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039986645677752235,
+      "loss": 2.2178,
+      "step": 990
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000399863745617091,
+      "loss": 2.2294,
+      "step": 1000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000399861007221256,
+      "loss": 2.2203,
+      "step": 1010
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039985824159039044,
+      "loss": 2.2326,
+      "step": 1020
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039985544872487124,
+      "loss": 2.2135,
+      "step": 1030
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039985262862507904,
+      "loss": 2.1806,
+      "step": 1040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039984978129139803,
+      "loss": 2.2396,
+      "step": 1050
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039984690672421625,
+      "loss": 2.1817,
+      "step": 1060
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0003998440049239254,
+      "loss": 2.1996,
+      "step": 1070
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00039984107589092094,
+      "loss": 2.139,
+      "step": 1080
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0003998381196256021,
+      "loss": 2.1516,
+      "step": 1090
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00039983513612837146,
+      "loss": 2.2708,
+      "step": 1100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0003998321253996359,
+      "loss": 2.2253,
+      "step": 1110
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00039982908743980543,
+      "loss": 2.1869,
+      "step": 1120
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00039982602224929425,
+      "loss": 2.1836,
+      "step": 1130
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0003998229298285199,
+      "loss": 2.165,
+      "step": 1140
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0003998198101779038,
+      "loss": 2.1514,
+      "step": 1150
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00039981666329787116,
+      "loss": 2.2043,
+      "step": 1160
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0003998134891888507,
+      "loss": 2.2431,
+      "step": 1170
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000399810287851275,
+      "loss": 2.1645,
+      "step": 1180
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0003998070592855803,
+      "loss": 2.0972,
+      "step": 1190
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0003998038034922066,
+      "loss": 2.1676,
+      "step": 1200
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00039980052047159753,
+      "loss": 2.1515,
+      "step": 1210
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00039979721022420043,
+      "loss": 2.2242,
+      "step": 1220
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00039979387275046636,
+      "loss": 2.1498,
+      "step": 1230
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00039979050805085027,
+      "loss": 2.1727,
+      "step": 1240
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00039978711612581055,
+      "loss": 2.1848,
+      "step": 1250
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0003997836969758094,
+      "loss": 2.2028,
+      "step": 1260
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00039978025060131284,
+      "loss": 2.1458,
+      "step": 1270
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0003997767770027904,
+      "loss": 2.1835,
+      "step": 1280
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0003997732761807155,
+      "loss": 2.1333,
+      "step": 1290
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0003997697481355652,
+      "loss": 2.1516,
+      "step": 1300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0003997661928678203,
+      "loss": 2.1426,
+      "step": 1310
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00039976261037796514,
+      "loss": 2.2161,
+      "step": 1320
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000399759000666488,
+      "loss": 2.2368,
+      "step": 1330
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00039975536373388075,
+      "loss": 2.1565,
+      "step": 1340
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000399751699580639,
+      "loss": 2.083,
+      "step": 1350
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00039974800820726214,
+      "loss": 2.0986,
+      "step": 1360
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00039974428961425304,
+      "loss": 2.0592,
+      "step": 1370
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00039974054380211856,
+      "loss": 2.0629,
+      "step": 1380
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000399736770771369,
+      "loss": 2.0514,
+      "step": 1390
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00039973297052251877,
+      "loss": 2.0839,
+      "step": 1400
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00039972914305608546,
+      "loss": 2.1392,
+      "step": 1410
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00039972528837259075,
+      "loss": 2.1073,
+      "step": 1420
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0003997214064725599,
+      "loss": 2.0508,
+      "step": 1430
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0003997174973565219,
+      "loss": 2.0333,
+      "step": 1440
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0003997135610250095,
+      "loss": 2.0698,
+      "step": 1450
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000399709597478559,
+      "loss": 2.0841,
+      "step": 1460
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0003997056067177106,
+      "loss": 2.0923,
+      "step": 1470
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0003997015887430081,
+      "loss": 2.064,
+      "step": 1480
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039969754355499894,
+      "loss": 2.0102,
+      "step": 1490
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039969347115423445,
+      "loss": 2.063,
+      "step": 1500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039968937154126965,
+      "loss": 2.0544,
+      "step": 1510
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.000399685244716663,
+      "loss": 2.0805,
+      "step": 1520
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039968109068097703,
+      "loss": 2.0353,
+      "step": 1530
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003996769094347777,
+      "loss": 2.0498,
+      "step": 1540
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003996727009786349,
+      "loss": 2.0424,
+      "step": 1550
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039966846531312196,
+      "loss": 2.0201,
+      "step": 1560
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003996642024388162,
+      "loss": 2.006,
+      "step": 1570
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003996599123562985,
+      "loss": 2.0948,
+      "step": 1580
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003996555950661534,
+      "loss": 2.0324,
+      "step": 1590
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003996512505689693,
+      "loss": 2.046,
+      "step": 1600
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003996468788653382,
+      "loss": 2.0836,
+      "step": 1610
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039964247995585573,
+      "loss": 2.0208,
+      "step": 1620
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003996380538411215,
+      "loss": 2.061,
+      "step": 1630
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003996336005217386,
+      "loss": 2.0576,
+      "step": 1640
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003996291199983138,
+      "loss": 1.9996,
+      "step": 1650
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039962461227145777,
+      "loss": 1.9895,
+      "step": 1660
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003996200773417847,
+      "loss": 2.1121,
+      "step": 1670
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039961551520991263,
+      "loss": 2.1313,
+      "step": 1680
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003996109258764632,
+      "loss": 2.0451,
+      "step": 1690
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003996063093420618,
+      "loss": 2.0172,
+      "step": 1700
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039960166560733745,
+      "loss": 2.1171,
+      "step": 1710
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.000399596994672923,
+      "loss": 2.1142,
+      "step": 1720
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003995922965394551,
+      "loss": 2.0765,
+      "step": 1730
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003995875712075738,
+      "loss": 1.9818,
+      "step": 1740
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003995828186779231,
+      "loss": 2.0097,
+      "step": 1750
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039957803895115056,
+      "loss": 2.0391,
+      "step": 1760
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039957323202790754,
+      "loss": 2.0719,
+      "step": 1770
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003995683979088491,
+      "loss": 1.9692,
+      "step": 1780
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.000399563536594634,
+      "loss": 1.9804,
+      "step": 1790
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003995586480859246,
+      "loss": 2.0772,
+      "step": 1800
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039955373238338715,
+      "loss": 2.0121,
+      "step": 1810
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003995487894876915,
+      "loss": 2.1017,
+      "step": 1820
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003995438193995111,
+      "loss": 2.0021,
+      "step": 1830
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003995388221195234,
+      "loss": 2.0173,
+      "step": 1840
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003995337976484093,
+      "loss": 2.12,
+      "step": 1850
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039952874598685347,
+      "loss": 2.0556,
+      "step": 1860
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003995236671355444,
+      "loss": 2.1336,
+      "step": 1870
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003995185610951739,
+      "loss": 2.0388,
+      "step": 1880
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003995134278664382,
+      "loss": 2.0134,
+      "step": 1890
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003995082674500364,
+      "loss": 1.997,
+      "step": 1900
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039950307984667197,
+      "loss": 2.0323,
+      "step": 1910
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039949786505705167,
+      "loss": 2.0107,
+      "step": 1920
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039949262308188615,
+      "loss": 2.0288,
+      "step": 1930
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003994873539218898,
+      "loss": 1.9891,
+      "step": 1940
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003994820575777807,
+      "loss": 2.0481,
+      "step": 1950
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003994767340502804,
+      "loss": 2.0216,
+      "step": 1960
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039947138334011446,
+      "loss": 2.0167,
+      "step": 1970
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039946600544801196,
+      "loss": 1.9372,
+      "step": 1980
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039946060037470587,
+      "loss": 1.935,
+      "step": 1990
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039945516812093254,
+      "loss": 2.0051,
+      "step": 2000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003994497086874324,
+      "loss": 2.087,
+      "step": 2010
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003994442220749493,
+      "loss": 1.9893,
+      "step": 2020
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003994387082842309,
+      "loss": 1.9887,
+      "step": 2030
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003994331673160287,
+      "loss": 1.9826,
+      "step": 2040
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039942759917109755,
+      "loss": 1.9206,
+      "step": 2050
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003994220038501964,
+      "loss": 2.0448,
+      "step": 2060
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003994163813540876,
+      "loss": 1.9772,
+      "step": 2070
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003994107316835375,
+      "loss": 2.0458,
+      "step": 2080
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039940505483931573,
+      "loss": 1.9609,
+      "step": 2090
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039939935082219604,
+      "loss": 1.9851,
+      "step": 2100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039939361963295573,
+      "loss": 1.954,
+      "step": 2110
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039938786127237576,
+      "loss": 1.9591,
+      "step": 2120
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003993820757412407,
+      "loss": 1.9289,
+      "step": 2130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039937626304033914,
+      "loss": 1.9337,
+      "step": 2140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.000399370423170463,
+      "loss": 1.9348,
+      "step": 2150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003993645561324082,
+      "loss": 2.0066,
+      "step": 2160
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003993586619269742,
+      "loss": 1.9672,
+      "step": 2170
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039935274055496426,
+      "loss": 2.0074,
+      "step": 2180
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039934679201718515,
+      "loss": 2.0142,
+      "step": 2190
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039934081631444755,
+      "loss": 1.9753,
+      "step": 2200
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039933481344756587,
+      "loss": 1.9794,
+      "step": 2210
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003993287834173579,
+      "loss": 2.01,
+      "step": 2220
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003993227262246455,
+      "loss": 1.9906,
+      "step": 2230
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003993166418702541,
+      "loss": 1.9185,
+      "step": 2240
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003993105303550127,
+      "loss": 1.919,
+      "step": 2250
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003993043916797543,
+      "loss": 1.9834,
+      "step": 2260
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003992982258453152,
+      "loss": 1.9505,
+      "step": 2270
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003992920328525358,
+      "loss": 1.9617,
+      "step": 2280
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039928581270225986,
+      "loss": 1.9801,
+      "step": 2290
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003992795653953351,
+      "loss": 1.9085,
+      "step": 2300
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003992732909326128,
+      "loss": 2.01,
+      "step": 2310
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.000399266989314948,
+      "loss": 2.0069,
+      "step": 2320
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003992606605431994,
+      "loss": 1.9633,
+      "step": 2330
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003992543046182295,
+      "loss": 2.0229,
+      "step": 2340
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039924792154090433,
+      "loss": 1.9732,
+      "step": 2350
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039924151131209373,
+      "loss": 1.8956,
+      "step": 2360
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003992350739326712,
+      "loss": 1.9605,
+      "step": 2370
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003992286094035141,
+      "loss": 1.9142,
+      "step": 2380
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003992221177255031,
+      "loss": 1.9233,
+      "step": 2390
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039921559889952304,
+      "loss": 1.95,
+      "step": 2400
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003992090529264622,
+      "loss": 1.9638,
+      "step": 2410
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003992024798072125,
+      "loss": 1.9341,
+      "step": 2420
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003991958795426697,
+      "loss": 1.9212,
+      "step": 2430
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039918925213373326,
+      "loss": 1.9003,
+      "step": 2440
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003991825975813062,
+      "loss": 1.91,
+      "step": 2450
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003991759158862955,
+      "loss": 1.9405,
+      "step": 2460
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039916920704961156,
+      "loss": 1.939,
+      "step": 2470
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003991624710721686,
+      "loss": 1.8843,
+      "step": 2480
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039915570795488454,
+      "loss": 1.9161,
+      "step": 2490
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.000399148917698681,
+      "loss": 1.912,
+      "step": 2500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039914210030448326,
+      "loss": 2.029,
+      "step": 2510
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039913525577322037,
+      "loss": 1.9293,
+      "step": 2520
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039912838410582495,
+      "loss": 1.9556,
+      "step": 2530
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003991214853032335,
+      "loss": 1.9542,
+      "step": 2540
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003991145593663861,
+      "loss": 1.931,
+      "step": 2550
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039910760629622644,
+      "loss": 1.8945,
+      "step": 2560
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003991006260937022,
+      "loss": 1.9022,
+      "step": 2570
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039909361875976445,
+      "loss": 1.9475,
+      "step": 2580
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039908658429536805,
+      "loss": 1.9387,
+      "step": 2590
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003990795227014716,
+      "loss": 1.966,
+      "step": 2600
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003990724339790375,
+      "loss": 1.9118,
+      "step": 2610
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003990653181290316,
+      "loss": 1.9823,
+      "step": 2620
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039905817515242364,
+      "loss": 1.9006,
+      "step": 2630
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.000399051005050187,
+      "loss": 1.9616,
+      "step": 2640
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039904380782329867,
+      "loss": 2.0014,
+      "step": 2650
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003990365834727395,
+      "loss": 1.9463,
+      "step": 2660
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039902933199949383,
+      "loss": 1.9225,
+      "step": 2670
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039902205340454995,
+      "loss": 1.9477,
+      "step": 2680
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039901474768889967,
+      "loss": 1.8736,
+      "step": 2690
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039900741485353856,
+      "loss": 1.9268,
+      "step": 2700
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039900005489946583,
+      "loss": 1.9873,
+      "step": 2710
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003989926678276844,
+      "loss": 1.9226,
+      "step": 2720
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.000398985253639201,
+      "loss": 1.8946,
+      "step": 2730
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039897781233502586,
+      "loss": 1.9803,
+      "step": 2740
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.000398970343916173,
+      "loss": 1.9063,
+      "step": 2750
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039896284838366025,
+      "loss": 1.9195,
+      "step": 2760
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003989553257385089,
+      "loss": 1.8676,
+      "step": 2770
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003989477759817442,
+      "loss": 1.9615,
+      "step": 2780
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003989401991143948,
+      "loss": 1.9745,
+      "step": 2790
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003989325951374933,
+      "loss": 1.9975,
+      "step": 2800
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039892496405207584,
+      "loss": 1.906,
+      "step": 2810
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003989173058591823,
+      "loss": 1.8932,
+      "step": 2820
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039890962055985635,
+      "loss": 1.9158,
+      "step": 2830
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039890190815514517,
+      "loss": 1.9122,
+      "step": 2840
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039889416864609975,
+      "loss": 1.8768,
+      "step": 2850
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003988864020337748,
+      "loss": 1.9045,
+      "step": 2860
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003988786083192286,
+      "loss": 1.9719,
+      "step": 2870
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003988707875035233,
+      "loss": 1.8816,
+      "step": 2880
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039886293958772456,
+      "loss": 1.962,
+      "step": 2890
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003988550645729018,
+      "loss": 1.967,
+      "step": 2900
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039884716246012817,
+      "loss": 1.9085,
+      "step": 2910
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003988392332504805,
+      "loss": 1.8951,
+      "step": 2920
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003988312769450394,
+      "loss": 1.9603,
+      "step": 2930
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003988232935448889,
+      "loss": 1.9027,
+      "step": 2940
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039881528305111695,
+      "loss": 1.8541,
+      "step": 2950
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003988072454648152,
+      "loss": 1.9238,
+      "step": 2960
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003987991807870789,
+      "loss": 1.899,
+      "step": 2970
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039879108901900697,
+      "loss": 1.8818,
+      "step": 2980
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003987829701617022,
+      "loss": 1.8995,
+      "step": 2990
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039877482421627077,
+      "loss": 1.9103,
+      "step": 3000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003987666511838229,
+      "loss": 1.8458,
+      "step": 3010
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003987584510654722,
+      "loss": 1.9115,
+      "step": 3020
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039875022386233624,
+      "loss": 1.9156,
+      "step": 3030
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039874196957553594,
+      "loss": 1.8478,
+      "step": 3040
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039873368820619626,
+      "loss": 1.8816,
+      "step": 3050
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003987253797554457,
+      "loss": 1.9044,
+      "step": 3060
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003987170442244164,
+      "loss": 1.9477,
+      "step": 3070
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039870868161424416,
+      "loss": 1.8782,
+      "step": 3080
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003987002919260687,
+      "loss": 1.9034,
+      "step": 3090
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003986918751610333,
+      "loss": 1.851,
+      "step": 3100
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039868343132028476,
+      "loss": 1.8951,
+      "step": 3110
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003986749604049738,
+      "loss": 1.8947,
+      "step": 3120
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003986664624162548,
+      "loss": 1.8742,
+      "step": 3130
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003986579373552857,
+      "loss": 1.8696,
+      "step": 3140
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039864938522322826,
+      "loss": 1.8318,
+      "step": 3150
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039864080602124776,
+      "loss": 1.8975,
+      "step": 3160
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003986321997505135,
+      "loss": 1.8946,
+      "step": 3170
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003986235664121981,
+      "loss": 1.9007,
+      "step": 3180
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.000398614906007478,
+      "loss": 1.8543,
+      "step": 3190
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003986062185375335,
+      "loss": 1.8906,
+      "step": 3200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039859750400354836,
+      "loss": 1.8717,
+      "step": 3210
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039858876240671004,
+      "loss": 1.9091,
+      "step": 3220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039857999374820984,
+      "loss": 1.8534,
+      "step": 3230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039857119802924265,
+      "loss": 1.9424,
+      "step": 3240
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039856237525100706,
+      "loss": 1.8563,
+      "step": 3250
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039855352541470535,
+      "loss": 1.8508,
+      "step": 3260
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003985446485215434,
+      "loss": 1.7738,
+      "step": 3270
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039853574457273104,
+      "loss": 1.8507,
+      "step": 3280
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003985268135694815,
+      "loss": 1.87,
+      "step": 3290
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039851785551301176,
+      "loss": 1.8337,
+      "step": 3300
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003985088704045426,
+      "loss": 1.9019,
+      "step": 3310
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039849985824529847,
+      "loss": 1.892,
+      "step": 3320
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039849081903650733,
+      "loss": 1.8693,
+      "step": 3330
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039848175277940105,
+      "loss": 1.9654,
+      "step": 3340
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003984726594752151,
+      "loss": 1.9894,
+      "step": 3350
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003984635391251885,
+      "loss": 1.8574,
+      "step": 3360
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003984543917305642,
+      "loss": 1.8305,
+      "step": 3370
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039844521729258864,
+      "loss": 1.8939,
+      "step": 3380
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039843601581251206,
+      "loss": 1.8208,
+      "step": 3390
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003984267872915883,
+      "loss": 1.8424,
+      "step": 3400
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039841753173107505,
+      "loss": 1.8248,
+      "step": 3410
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003984082491322334,
+      "loss": 1.863,
+      "step": 3420
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003983989394963284,
+      "loss": 1.8603,
+      "step": 3430
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039838960282462867,
+      "loss": 1.9456,
+      "step": 3440
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039838023911840644,
+      "loss": 1.8169,
+      "step": 3450
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003983708483789378,
+      "loss": 1.8177,
+      "step": 3460
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039836143060750225,
+      "loss": 1.8531,
+      "step": 3470
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039835198580538336,
+      "loss": 1.9448,
+      "step": 3480
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039834251397386807,
+      "loss": 1.8522,
+      "step": 3490
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039833301511424716,
+      "loss": 1.8559,
+      "step": 3500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003983234892278149,
+      "loss": 1.864,
+      "step": 3510
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003983139363158696,
+      "loss": 1.8782,
+      "step": 3520
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039830435637971284,
+      "loss": 1.8399,
+      "step": 3530
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003982947494206501,
+      "loss": 1.8912,
+      "step": 3540
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039828511543999064,
+      "loss": 1.8414,
+      "step": 3550
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003982754544390471,
+      "loss": 1.8746,
+      "step": 3560
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039826576641913624,
+      "loss": 1.8617,
+      "step": 3570
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.000398256051381578,
+      "loss": 1.8677,
+      "step": 3580
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039824630932769643,
+      "loss": 1.8652,
+      "step": 3590
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003982365402588189,
+      "loss": 1.8284,
+      "step": 3600
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003982267441762767,
+      "loss": 1.8363,
+      "step": 3610
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003982169210814049,
+      "loss": 1.8223,
+      "step": 3620
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039820707097554183,
+      "loss": 1.8814,
+      "step": 3630
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039819719386002995,
+      "loss": 1.889,
+      "step": 3640
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003981872897362152,
+      "loss": 1.8921,
+      "step": 3650
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039817735860544714,
+      "loss": 1.8233,
+      "step": 3660
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003981674004690791,
+      "loss": 1.8574,
+      "step": 3670
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039815741532846807,
+      "loss": 1.8394,
+      "step": 3680
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003981474031849748,
+      "loss": 1.8477,
+      "step": 3690
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039813736403996355,
+      "loss": 1.8046,
+      "step": 3700
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003981272978948024,
+      "loss": 1.8568,
+      "step": 3710
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.000398117204750863,
+      "loss": 1.8511,
+      "step": 3720
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003981070846095208,
+      "loss": 1.8222,
+      "step": 3730
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039809693747215494,
+      "loss": 1.8691,
+      "step": 3740
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039808676334014806,
+      "loss": 1.8915,
+      "step": 3750
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003980765622148866,
+      "loss": 1.8411,
+      "step": 3760
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003980663340977607,
+      "loss": 1.8555,
+      "step": 3770
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039805607899016406,
+      "loss": 1.8563,
+      "step": 3780
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039804579689349436,
+      "loss": 1.8344,
+      "step": 3790
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003980354878091524,
+      "loss": 1.8375,
+      "step": 3800
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003980251517385433,
+      "loss": 1.8031,
+      "step": 3810
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003980147886830755,
+      "loss": 1.8695,
+      "step": 3820
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039800439864416106,
+      "loss": 1.8657,
+      "step": 3830
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039799398162321584,
+      "loss": 1.8441,
+      "step": 3840
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039798353762165946,
+      "loss": 1.9097,
+      "step": 3850
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039797306664091504,
+      "loss": 1.9122,
+      "step": 3860
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003979625686824095,
+      "loss": 1.8143,
+      "step": 3870
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003979520437475734,
+      "loss": 1.859,
+      "step": 3880
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003979414918378409,
+      "loss": 1.8044,
+      "step": 3890
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039793091295465004,
+      "loss": 1.8097,
+      "step": 3900
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003979203070994423,
+      "loss": 1.8495,
+      "step": 3910
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.000397909674273663,
+      "loss": 1.8872,
+      "step": 3920
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039789901447876097,
+      "loss": 1.8748,
+      "step": 3930
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039788832771618887,
+      "loss": 1.7635,
+      "step": 3940
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003978776139874031,
+      "loss": 1.8179,
+      "step": 3950
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003978668732938635,
+      "loss": 1.8709,
+      "step": 3960
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003978561056370337,
+      "loss": 1.8009,
+      "step": 3970
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039784531101838105,
+      "loss": 1.8061,
+      "step": 3980
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039783448943937646,
+      "loss": 1.8609,
+      "step": 3990
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003978236409014947,
+      "loss": 1.8439,
+      "step": 4000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039781276540621393,
+      "loss": 1.8459,
+      "step": 4010
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003978018629550164,
+      "loss": 1.8014,
+      "step": 4020
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003977909335493876,
+      "loss": 1.7949,
+      "step": 4030
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003977799771908169,
+      "loss": 1.8656,
+      "step": 4040
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003977689938807973,
+      "loss": 1.842,
+      "step": 4050
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039775798362082557,
+      "loss": 1.8599,
+      "step": 4060
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039774694641240204,
+      "loss": 1.8555,
+      "step": 4070
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003977358822570308,
+      "loss": 1.8812,
+      "step": 4080
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039772479115621944,
+      "loss": 1.8152,
+      "step": 4090
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039771367311147943,
+      "loss": 1.8578,
+      "step": 4100
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039770252812432584,
+      "loss": 1.7968,
+      "step": 4110
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003976913561962774,
+      "loss": 1.812,
+      "step": 4120
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039768015732885646,
+      "loss": 1.799,
+      "step": 4130
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039766893152358906,
+      "loss": 1.8426,
+      "step": 4140
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039765767878200493,
+      "loss": 1.8531,
+      "step": 4150
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003976463991056376,
+      "loss": 1.8041,
+      "step": 4160
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003976350924960241,
+      "loss": 1.8543,
+      "step": 4170
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039762375895470507,
+      "loss": 1.8657,
+      "step": 4180
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003976123984832251,
+      "loss": 1.8165,
+      "step": 4190
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039760101108313216,
+      "loss": 1.8312,
+      "step": 4200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003975895967559781,
+      "loss": 1.8181,
+      "step": 4210
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003975781555033183,
+      "loss": 1.8071,
+      "step": 4220
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003975666873267118,
+      "loss": 1.8136,
+      "step": 4230
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039755519222772144,
+      "loss": 1.8227,
+      "step": 4240
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003975436702079136,
+      "loss": 1.7746,
+      "step": 4250
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039753212126885843,
+      "loss": 1.856,
+      "step": 4260
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039752054541212974,
+      "loss": 1.8853,
+      "step": 4270
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003975089426393049,
+      "loss": 1.8294,
+      "step": 4280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039749731295196503,
+      "loss": 1.7912,
+      "step": 4290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.000397485656351695,
+      "loss": 1.791,
+      "step": 4300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039747397284008303,
+      "loss": 1.7691,
+      "step": 4310
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003974622624187215,
+      "loss": 1.8341,
+      "step": 4320
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.000397450525089206,
+      "loss": 1.7926,
+      "step": 4330
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.000397438760853136,
+      "loss": 1.7929,
+      "step": 4340
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003974269697121147,
+      "loss": 1.8295,
+      "step": 4350
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039741515166774876,
+      "loss": 1.7966,
+      "step": 4360
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039740330672164873,
+      "loss": 1.785,
+      "step": 4370
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039739143487542867,
+      "loss": 1.8314,
+      "step": 4380
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039737953613070635,
+      "loss": 2.1719,
+      "step": 4390
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003973676104891032,
+      "loss": 2.1545,
+      "step": 4400
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003973556579522444,
+      "loss": 1.8765,
+      "step": 4410
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039734367852175867,
+      "loss": 1.8129,
+      "step": 4420
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039733167219927834,
+      "loss": 1.8307,
+      "step": 4430
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003973196389864397,
+      "loss": 1.8195,
+      "step": 4440
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039730757888488237,
+      "loss": 1.8062,
+      "step": 4450
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003972954918962499,
+      "loss": 1.8539,
+      "step": 4460
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039728337802218925,
+      "loss": 1.815,
+      "step": 4470
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003972712372643512,
+      "loss": 1.7471,
+      "step": 4480
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003972590696243903,
+      "loss": 1.8931,
+      "step": 4490
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003972468751039645,
+      "loss": 1.8712,
+      "step": 4500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003972346537047356,
+      "loss": 1.7898,
+      "step": 4510
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.000397222405428369,
+      "loss": 1.8339,
+      "step": 4520
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039721013027653376,
+      "loss": 1.7983,
+      "step": 4530
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003971978282509026,
+      "loss": 1.7834,
+      "step": 4540
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039718549935315197,
+      "loss": 1.8152,
+      "step": 4550
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003971731435849619,
+      "loss": 1.7459,
+      "step": 4560
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039716076094801605,
+      "loss": 1.8813,
+      "step": 4570
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003971483514440019,
+      "loss": 1.8476,
+      "step": 4580
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003971359150746104,
+      "loss": 1.8125,
+      "step": 4590
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039712345184153634,
+      "loss": 1.8821,
+      "step": 4600
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003971109617464779,
+      "loss": 1.8308,
+      "step": 4610
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003970984447911374,
+      "loss": 1.8794,
+      "step": 4620
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039708590097722033,
+      "loss": 1.8233,
+      "step": 4630
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003970733303064361,
+      "loss": 1.8059,
+      "step": 4640
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039706073278049755,
+      "loss": 1.794,
+      "step": 4650
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003970481084011215,
+      "loss": 1.788,
+      "step": 4660
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039703545717002835,
+      "loss": 1.838,
+      "step": 4670
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039702277908894193,
+      "loss": 1.8613,
+      "step": 4680
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003970100741595899,
+      "loss": 1.8033,
+      "step": 4690
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039699734238370364,
+      "loss": 1.8565,
+      "step": 4700
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039698458376301803,
+      "loss": 1.8227,
+      "step": 4710
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003969717982992717,
+      "loss": 1.8421,
+      "step": 4720
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039695898599420695,
+      "loss": 1.7647,
+      "step": 4730
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003969461468495697,
+      "loss": 1.8001,
+      "step": 4740
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039693328086710955,
+      "loss": 1.8001,
+      "step": 4750
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003969203880485797,
+      "loss": 1.782,
+      "step": 4760
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.000396907468395737,
+      "loss": 1.8169,
+      "step": 4770
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003968945219103422,
+      "loss": 1.7674,
+      "step": 4780
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003968815485941594,
+      "loss": 1.8024,
+      "step": 4790
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003968685484489564,
+      "loss": 1.7552,
+      "step": 4800
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003968555214765049,
+      "loss": 1.8105,
+      "step": 4810
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039684246767857995,
+      "loss": 1.8472,
+      "step": 4820
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039682938705696045,
+      "loss": 1.8428,
+      "step": 4830
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003968162796134288,
+      "loss": 1.7878,
+      "step": 4840
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003968031453497712,
+      "loss": 1.773,
+      "step": 4850
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039678998426777756,
+      "loss": 1.7859,
+      "step": 4860
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003967767963692413,
+      "loss": 1.8525,
+      "step": 4870
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003967635816559594,
+      "loss": 1.8496,
+      "step": 4880
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003967503401297328,
+      "loss": 1.7596,
+      "step": 4890
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039673707179236576,
+      "loss": 1.8011,
+      "step": 4900
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039672377664566643,
+      "loss": 1.9062,
+      "step": 4910
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003967104546914466,
+      "loss": 1.7891,
+      "step": 4920
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039669710593152153,
+      "loss": 1.7634,
+      "step": 4930
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003966837303677103,
+      "loss": 1.8714,
+      "step": 4940
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003966703280018356,
+      "loss": 1.7708,
+      "step": 4950
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039665689883572385,
+      "loss": 1.7492,
+      "step": 4960
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039664344287120495,
+      "loss": 1.7867,
+      "step": 4970
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003966299601101125,
+      "loss": 1.7348,
+      "step": 4980
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003966164505542839,
+      "loss": 1.7343,
+      "step": 4990
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039660291420556007,
+      "loss": 1.8217,
+      "step": 5000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003965893510657856,
+      "loss": 1.7967,
+      "step": 5010
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003965757611368086,
+      "loss": 1.8114,
+      "step": 5020
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003965621444204812,
+      "loss": 1.7517,
+      "step": 5030
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003965485009186588,
+      "loss": 1.7804,
+      "step": 5040
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039653483063320066,
+      "loss": 1.8003,
+      "step": 5050
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003965211335659696,
+      "loss": 1.7153,
+      "step": 5060
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003965074097188321,
+      "loss": 1.8146,
+      "step": 5070
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003964936590936583,
+      "loss": 1.7414,
+      "step": 5080
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003964798816923221,
+      "loss": 1.7686,
+      "step": 5090
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039646607751670084,
+      "loss": 1.8036,
+      "step": 5100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003964522465686757,
+      "loss": 1.8148,
+      "step": 5110
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039643838885013123,
+      "loss": 1.7598,
+      "step": 5120
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.000396424504362956,
+      "loss": 1.7619,
+      "step": 5130
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.000396410593109042,
+      "loss": 1.8385,
+      "step": 5140
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.000396396655090285,
+      "loss": 1.8058,
+      "step": 5150
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039638269030858414,
+      "loss": 1.7824,
+      "step": 5160
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003963686987658426,
+      "loss": 1.7732,
+      "step": 5170
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003963546804639668,
+      "loss": 2.0963,
+      "step": 5180
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039634063540486714,
+      "loss": 1.9065,
+      "step": 5190
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003963265635904576,
+      "loss": 1.7381,
+      "step": 5200
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003963124650226556,
+      "loss": 1.8123,
+      "step": 5210
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003962983397033824,
+      "loss": 1.7298,
+      "step": 5220
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039628418763456295,
+      "loss": 1.7675,
+      "step": 5230
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039627000881812554,
+      "loss": 1.7765,
+      "step": 5240
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039625580325600255,
+      "loss": 1.7708,
+      "step": 5250
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003962415709501296,
+      "loss": 1.8024,
+      "step": 5260
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003962273119024462,
+      "loss": 1.8149,
+      "step": 5270
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003962130261148954,
+      "loss": 1.7905,
+      "step": 5280
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039619871358942405,
+      "loss": 1.7407,
+      "step": 5290
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039618437432798235,
+      "loss": 1.7789,
+      "step": 5300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039617000833252434,
+      "loss": 1.74,
+      "step": 5310
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003961556156050077,
+      "loss": 1.7821,
+      "step": 5320
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003961411961473937,
+      "loss": 1.8277,
+      "step": 5330
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003961267499616473,
+      "loss": 1.7875,
+      "step": 5340
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003961122770497371,
+      "loss": 1.7887,
+      "step": 5350
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003960977774136353,
+      "loss": 1.7135,
+      "step": 5360
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039608325105531774,
+      "loss": 1.7723,
+      "step": 5370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.000396068697976764,
+      "loss": 1.7617,
+      "step": 5380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003960541181799571,
+      "loss": 1.7773,
+      "step": 5390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.000396039511666884,
+      "loss": 1.7796,
+      "step": 5400
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039602487843953495,
+      "loss": 1.8129,
+      "step": 5410
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003960102184999041,
+      "loss": 1.7779,
+      "step": 5420
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003959955318499892,
+      "loss": 1.8043,
+      "step": 5430
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003959808184917915,
+      "loss": 1.8121,
+      "step": 5440
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039596607842731606,
+      "loss": 1.8458,
+      "step": 5450
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039595131165857147,
+      "loss": 1.6925,
+      "step": 5460
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039593651818757005,
+      "loss": 1.7781,
+      "step": 5470
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003959216980163276,
+      "loss": 1.7734,
+      "step": 5480
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003959068511468638,
+      "loss": 1.7421,
+      "step": 5490
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003958919775812017,
+      "loss": 1.7665,
+      "step": 5500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003958770773213682,
+      "loss": 1.8133,
+      "step": 5510
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003958621503693937,
+      "loss": 1.8489,
+      "step": 5520
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039584719672731234,
+      "loss": 1.7686,
+      "step": 5530
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039583221639716184,
+      "loss": 1.7774,
+      "step": 5540
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003958172093809836,
+      "loss": 1.7615,
+      "step": 5550
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039580217568082254,
+      "loss": 1.7265,
+      "step": 5560
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003957871152987274,
+      "loss": 1.7839,
+      "step": 5570
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003957720282367504,
+      "loss": 1.8069,
+      "step": 5580
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039575691449694744,
+      "loss": 1.7924,
+      "step": 5590
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039574177408137806,
+      "loss": 1.7417,
+      "step": 5600
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039572660699210545,
+      "loss": 1.7806,
+      "step": 5610
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003957114132311965,
+      "loss": 1.7968,
+      "step": 5620
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003956961928007216,
+      "loss": 1.7929,
+      "step": 5630
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003956809457027548,
+      "loss": 1.7942,
+      "step": 5640
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003956656719393739,
+      "loss": 1.7724,
+      "step": 5650
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039565037151266015,
+      "loss": 1.7645,
+      "step": 5660
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039563504442469865,
+      "loss": 1.825,
+      "step": 5670
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039561969067757795,
+      "loss": 1.7706,
+      "step": 5680
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039560431027339034,
+      "loss": 1.6977,
+      "step": 5690
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039558890321423165,
+      "loss": 1.821,
+      "step": 5700
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039557346950220147,
+      "loss": 1.8348,
+      "step": 5710
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039555800913940295,
+      "loss": 1.7382,
+      "step": 5720
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003955425221279428,
+      "loss": 1.8237,
+      "step": 5730
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039552700846993147,
+      "loss": 1.7565,
+      "step": 5740
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.000395511468167483,
+      "loss": 1.755,
+      "step": 5750
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00039549590122271505,
+      "loss": 1.7847,
+      "step": 5760
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.000395480307637749,
+      "loss": 1.7783,
+      "step": 5770
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003954646874147097,
+      "loss": 1.794,
+      "step": 5780
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003954490405557257,
+      "loss": 1.8117,
+      "step": 5790
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003954333670629293,
+      "loss": 1.7388,
+      "step": 5800
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00039541766693845624,
+      "loss": 1.7699,
+      "step": 5810
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00039540194018444606,
+      "loss": 1.7813,
+      "step": 5820
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00039538618680304174,
+      "loss": 1.7695,
+      "step": 5830
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00039537040679638995,
+      "loss": 1.7895,
+      "step": 5840
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003953546001666412,
+      "loss": 1.7183,
+      "step": 5850
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003953387669159494,
+      "loss": 1.7388,
+      "step": 5860
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003953229070464721,
+      "loss": 1.7812,
+      "step": 5870
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003953070205603704,
+      "loss": 1.7602,
+      "step": 5880
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003952911074598094,
+      "loss": 1.7441,
+      "step": 5890
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003952751677469575,
+      "loss": 1.7681,
+      "step": 5900
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00039525920142398667,
+      "loss": 1.7261,
+      "step": 5910
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003952432084930728,
+      "loss": 1.7352,
+      "step": 5920
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00039522718895639513,
+      "loss": 1.7927,
+      "step": 5930
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00039521114281613676,
+      "loss": 1.7422,
+      "step": 5940
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003951950700744841,
+      "loss": 1.8138,
+      "step": 5950
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003951789707336276,
+      "loss": 1.7728,
+      "step": 5960
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00039516284479576097,
+      "loss": 1.7646,
+      "step": 5970
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00039514669226308174,
+      "loss": 1.8766,
+      "step": 5980
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00039513051313779104,
+      "loss": 1.7573,
+      "step": 5990
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00039511430742209354,
+      "loss": 1.774,
+      "step": 6000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003950980751181976,
+      "loss": 1.7494,
+      "step": 6010
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003950818162283152,
+      "loss": 1.8108,
+      "step": 6020
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00039506553075466195,
+      "loss": 1.7655,
+      "step": 6030
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00039504921869945716,
+      "loss": 1.735,
+      "step": 6040
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003950328800649235,
+      "loss": 1.8083,
+      "step": 6050
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00039501651485328754,
+      "loss": 1.7709,
+      "step": 6060
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00039500012306677927,
+      "loss": 1.7238,
+      "step": 6070
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003949837047076325,
+      "loss": 1.7706,
+      "step": 6080
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003949672597780845,
+      "loss": 1.731,
+      "step": 6090
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003949507882803763,
+      "loss": 1.7088,
+      "step": 6100
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003949342902167524,
+      "loss": 1.797,
+      "step": 6110
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003949177655894609,
+      "loss": 1.7863,
+      "step": 6120
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003949012144007538,
+      "loss": 1.8108,
+      "step": 6130
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003948846366528864,
+      "loss": 1.8412,
+      "step": 6140
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00039486803234811777,
+      "loss": 1.8077,
+      "step": 6150
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003948514014887106,
+      "loss": 1.7675,
+      "step": 6160
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003948347440769311,
+      "loss": 1.8441,
+      "step": 6170
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003948180601150493,
+      "loss": 1.7729,
+      "step": 6180
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003948013496053387,
+      "loss": 1.7453,
+      "step": 6190
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003947846125500763,
+      "loss": 1.7238,
+      "step": 6200
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039476784895154293,
+      "loss": 1.7232,
+      "step": 6210
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039475105881202306,
+      "loss": 1.7615,
+      "step": 6220
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039473424213380454,
+      "loss": 1.7313,
+      "step": 6230
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003947173989191791,
+      "loss": 1.7711,
+      "step": 6240
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039470052917044184,
+      "loss": 1.754,
+      "step": 6250
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003946836328898917,
+      "loss": 1.7765,
+      "step": 6260
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003946667100798311,
+      "loss": 1.7149,
+      "step": 6270
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039464976074256606,
+      "loss": 1.7595,
+      "step": 6280
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039463278488040634,
+      "loss": 1.7473,
+      "step": 6290
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003946157824956652,
+      "loss": 1.6789,
+      "step": 6300
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039459875359065957,
+      "loss": 1.7578,
+      "step": 6310
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039458169816770987,
+      "loss": 1.7213,
+      "step": 6320
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039456461622914044,
+      "loss": 1.7212,
+      "step": 6330
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039454750777727886,
+      "loss": 1.7128,
+      "step": 6340
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039453037281445664,
+      "loss": 1.7406,
+      "step": 6350
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003945132113430086,
+      "loss": 1.8279,
+      "step": 6360
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003944960233652734,
+      "loss": 1.7301,
+      "step": 6370
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003944788088835933,
+      "loss": 1.7591,
+      "step": 6380
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003944615679003141,
+      "loss": 1.769,
+      "step": 6390
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039444430041778514,
+      "loss": 1.6983,
+      "step": 6400
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003944270064383595,
+      "loss": 1.7071,
+      "step": 6410
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039440968596439393,
+      "loss": 1.7247,
+      "step": 6420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039439233899824855,
+      "loss": 1.7116,
+      "step": 6430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039437496554228725,
+      "loss": 1.7587,
+      "step": 6440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039435756559887754,
+      "loss": 1.7289,
+      "step": 6450
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039434013917039057,
+      "loss": 1.7415,
+      "step": 6460
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003943226862592009,
+      "loss": 1.8086,
+      "step": 6470
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039430520686768695,
+      "loss": 1.8101,
+      "step": 6480
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039428770099823055,
+      "loss": 1.7399,
+      "step": 6490
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003942701686532173,
+      "loss": 1.7788,
+      "step": 6500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039425260983503625,
+      "loss": 1.7449,
+      "step": 6510
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003942350245460802,
+      "loss": 1.7242,
+      "step": 6520
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039421741278874555,
+      "loss": 1.7774,
+      "step": 6530
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003941997745654321,
+      "loss": 1.6937,
+      "step": 6540
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003941821098785435,
+      "loss": 1.7575,
+      "step": 6550
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.000394164418730487,
+      "loss": 1.7652,
+      "step": 6560
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003941467011236731,
+      "loss": 1.7747,
+      "step": 6570
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039412895706051644,
+      "loss": 1.7496,
+      "step": 6580
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003941111865434349,
+      "loss": 1.7113,
+      "step": 6590
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039409338957485006,
+      "loss": 1.7558,
+      "step": 6600
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039407556615718713,
+      "loss": 1.7222,
+      "step": 6610
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003940577162928749,
+      "loss": 1.7327,
+      "step": 6620
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039403983998434575,
+      "loss": 1.7117,
+      "step": 6630
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039402193723403576,
+      "loss": 1.7442,
+      "step": 6640
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003940040080443844,
+      "loss": 1.708,
+      "step": 6650
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039398605241783497,
+      "loss": 1.7873,
+      "step": 6660
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039396807035683426,
+      "loss": 1.7372,
+      "step": 6670
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003939500618638327,
+      "loss": 1.7293,
+      "step": 6680
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039393202694128427,
+      "loss": 1.7413,
+      "step": 6690
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003939139655916466,
+      "loss": 1.7382,
+      "step": 6700
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039389587781738097,
+      "loss": 1.7376,
+      "step": 6710
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003938777636209521,
+      "loss": 1.8539,
+      "step": 6720
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003938596230048285,
+      "loss": 2.0847,
+      "step": 6730
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039384145597148206,
+      "loss": 1.7983,
+      "step": 6740
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039382326252338857,
+      "loss": 1.7416,
+      "step": 6750
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039380504266302703,
+      "loss": 1.7959,
+      "step": 6760
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039378679639288054,
+      "loss": 1.7394,
+      "step": 6770
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039376852371543524,
+      "loss": 1.7364,
+      "step": 6780
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003937502246331813,
+      "loss": 1.7741,
+      "step": 6790
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003937318991486123,
+      "loss": 1.7287,
+      "step": 6800
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003937135472642255,
+      "loss": 1.6729,
+      "step": 6810
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003936951689825216,
+      "loss": 1.7514,
+      "step": 6820
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039367676430600507,
+      "loss": 1.7261,
+      "step": 6830
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003936583332371839,
+      "loss": 1.6996,
+      "step": 6840
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039363987577856977,
+      "loss": 1.7291,
+      "step": 6850
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003936213919326777,
+      "loss": 1.7375,
+      "step": 6860
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003936028817020267,
+      "loss": 1.773,
+      "step": 6870
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039358434508913895,
+      "loss": 1.7226,
+      "step": 6880
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039356578209654054,
+      "loss": 1.7358,
+      "step": 6890
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039354719272676104,
+      "loss": 1.7529,
+      "step": 6900
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039352857698233363,
+      "loss": 1.7016,
+      "step": 6910
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039350993486579497,
+      "loss": 1.7673,
+      "step": 6920
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039349126637968566,
+      "loss": 1.8237,
+      "step": 6930
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039347257152654936,
+      "loss": 1.7175,
+      "step": 6940
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039345385030893377,
+      "loss": 1.7287,
+      "step": 6950
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039343510272939,
+      "loss": 1.7156,
+      "step": 6960
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003934163287904729,
+      "loss": 1.6863,
+      "step": 6970
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039339752849474057,
+      "loss": 1.7854,
+      "step": 6980
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.000393378701844755,
+      "loss": 1.7371,
+      "step": 6990
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003933598488430818,
+      "loss": 1.7483,
+      "step": 7000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039334096949228994,
+      "loss": 1.7263,
+      "step": 7010
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003933220637949522,
+      "loss": 1.7644,
+      "step": 7020
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0003933031317536448,
+      "loss": 1.752,
+      "step": 7030
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15631213784217834,
+      "learning_rate": 0.00039328417337094763,
+      "loss": 1.7599,
+      "step": 7040
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.14645032584667206,
+      "learning_rate": 0.00039326518864944417,
+      "loss": 1.7166,
+      "step": 7050
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1466301679611206,
+      "learning_rate": 0.00039324617759172137,
+      "loss": 1.7425,
+      "step": 7060
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.13789162039756775,
+      "learning_rate": 0.00039322714020037,
+      "loss": 1.7232,
+      "step": 7070
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1505664587020874,
+      "learning_rate": 0.00039320807647798414,
+      "loss": 1.6992,
+      "step": 7080
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.131247878074646,
+      "learning_rate": 0.00039318898642716173,
+      "loss": 1.7142,
+      "step": 7090
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.13916432857513428,
+      "learning_rate": 0.0003931698700505041,
+      "loss": 1.7512,
+      "step": 7100
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.13484828174114227,
+      "learning_rate": 0.0003931507273506162,
+      "loss": 1.7542,
+      "step": 7110
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.14107249677181244,
+      "learning_rate": 0.00039313155833010666,
+      "loss": 1.6634,
+      "step": 7120
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.14845772087574005,
+      "learning_rate": 0.0003931123629915876,
+      "loss": 1.7984,
+      "step": 7130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1481531858444214,
+      "learning_rate": 0.0003930931413376748,
+      "loss": 1.781,
+      "step": 7140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.14172106981277466,
+      "learning_rate": 0.00039307389337098767,
+      "loss": 1.7845,
+      "step": 7150
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.14321057498455048,
+      "learning_rate": 0.000393054619094149,
+      "loss": 1.7123,
+      "step": 7160
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.14234977960586548,
+      "learning_rate": 0.0003930353185097853,
+      "loss": 1.7058,
+      "step": 7170
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.17808882892131805,
+      "learning_rate": 0.00039301599162052677,
+      "loss": 1.7017,
+      "step": 7180
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.13419701159000397,
+      "learning_rate": 0.000392996638429007,
+      "loss": 1.7307,
+      "step": 7190
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1409374326467514,
+      "learning_rate": 0.00039297725893786315,
+      "loss": 1.7113,
+      "step": 7200
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.13434813916683197,
+      "learning_rate": 0.0003929578531497362,
+      "loss": 1.6991,
+      "step": 7210
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.13483397662639618,
+      "learning_rate": 0.0003929384210672706,
+      "loss": 1.7135,
+      "step": 7220
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15328499674797058,
+      "learning_rate": 0.00039291896269311427,
+      "loss": 1.7431,
+      "step": 7230
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.14323565363883972,
+      "learning_rate": 0.0003928994780299188,
+      "loss": 1.7087,
+      "step": 7240
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.14380577206611633,
+      "learning_rate": 0.0003928799670803394,
+      "loss": 1.7823,
+      "step": 7250
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.13544294238090515,
+      "learning_rate": 0.00039286042984703484,
+      "loss": 1.7303,
+      "step": 7260
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.13900905847549438,
+      "learning_rate": 0.0003928408663326674,
+      "loss": 1.7391,
+      "step": 7270
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1786203235387802,
+      "learning_rate": 0.00039282127653990297,
+      "loss": 1.7404,
+      "step": 7280
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15546490252017975,
+      "learning_rate": 0.0003928016604714112,
+      "loss": 1.7522,
+      "step": 7290
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16079726815223694,
+      "learning_rate": 0.000392782018129865,
+      "loss": 1.7286,
+      "step": 7300
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.14534011483192444,
+      "learning_rate": 0.0003927623495179411,
+      "loss": 1.7904,
+      "step": 7310
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.14322422444820404,
+      "learning_rate": 0.0003927426546383198,
+      "loss": 1.6674,
+      "step": 7320
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.139568492770195,
+      "learning_rate": 0.00039272293349368483,
+      "loss": 1.7203,
+      "step": 7330
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.21763065457344055,
+      "learning_rate": 0.0003927031860867236,
+      "loss": 1.731,
+      "step": 7340
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.13749991357326508,
+      "learning_rate": 0.00039268341242012714,
+      "loss": 1.709,
+      "step": 7350
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.14703145623207092,
+      "learning_rate": 0.0003926636124965899,
+      "loss": 1.6717,
+      "step": 7360
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1383516639471054,
+      "learning_rate": 0.00039264378631881017,
+      "loss": 1.6542,
+      "step": 7370
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1362847089767456,
+      "learning_rate": 0.0003926239338894895,
+      "loss": 1.7087,
+      "step": 7380
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1404048204421997,
+      "learning_rate": 0.00039260405521133324,
+      "loss": 1.7072,
+      "step": 7390
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16335636377334595,
+      "learning_rate": 0.0003925841502870503,
+      "loss": 1.7198,
+      "step": 7400
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.13588540256023407,
+      "learning_rate": 0.00039256421911935306,
+      "loss": 1.6984,
+      "step": 7410
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15666501224040985,
+      "learning_rate": 0.0003925442617109575,
+      "loss": 1.7881,
+      "step": 7420
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.14240530133247375,
+      "learning_rate": 0.00039252427806458346,
+      "loss": 1.698,
+      "step": 7430
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.14285792410373688,
+      "learning_rate": 0.0003925042681829537,
+      "loss": 1.7634,
+      "step": 7440
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.17508766055107117,
+      "learning_rate": 0.0003924842320687953,
+      "loss": 1.6889,
+      "step": 7450
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.13306574523448944,
+      "learning_rate": 0.0003924641697248385,
+      "loss": 1.6307,
+      "step": 7460
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1410142034292221,
+      "learning_rate": 0.0003924440811538171,
+      "loss": 1.7104,
+      "step": 7470
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.14873316884040833,
+      "learning_rate": 0.0003924239663584686,
+      "loss": 1.7204,
+      "step": 7480
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.13496127724647522,
+      "learning_rate": 0.0003924038253415341,
+      "loss": 1.6493,
+      "step": 7490
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1559535413980484,
+      "learning_rate": 0.00039238365810575815,
+      "loss": 1.7679,
+      "step": 7500
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1314055621623993,
+      "learning_rate": 0.000392363464653889,
+      "loss": 1.6889,
+      "step": 7510
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.13722346723079681,
+      "learning_rate": 0.00039234324498867835,
+      "loss": 1.7167,
+      "step": 7520
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.17334945499897003,
+      "learning_rate": 0.0003923229991128816,
+      "loss": 1.7667,
+      "step": 7530
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.13046298921108246,
+      "learning_rate": 0.00039230272702925757,
+      "loss": 1.6632,
+      "step": 7540
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.13783209025859833,
+      "learning_rate": 0.00039228242874056875,
+      "loss": 1.7177,
+      "step": 7550
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16211003065109253,
+      "learning_rate": 0.0003922621042495812,
+      "loss": 1.7712,
+      "step": 7560
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1397620439529419,
+      "learning_rate": 0.0003922417535590646,
+      "loss": 1.6844,
+      "step": 7570
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1390129178762436,
+      "learning_rate": 0.00039222137667179195,
+      "loss": 1.6965,
+      "step": 7580
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.14877186715602875,
+      "learning_rate": 0.00039220097359054027,
+      "loss": 1.6844,
+      "step": 7590
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.13841237127780914,
+      "learning_rate": 0.0003921805443180897,
+      "loss": 1.6794,
+      "step": 7600
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.14043334126472473,
+      "learning_rate": 0.00039216008885722416,
+      "loss": 1.7271,
+      "step": 7610
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2009606659412384,
+      "learning_rate": 0.0003921396072107311,
+      "loss": 1.8196,
+      "step": 7620
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.170047327876091,
+      "learning_rate": 0.0003921190993814016,
+      "loss": 1.7412,
+      "step": 7630
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16911502182483673,
+      "learning_rate": 0.0003920985653720303,
+      "loss": 1.7761,
+      "step": 7640
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15350545942783356,
+      "learning_rate": 0.00039207800518541527,
+      "loss": 1.7545,
+      "step": 7650
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.14801712334156036,
+      "learning_rate": 0.0003920574188243583,
+      "loss": 1.6783,
+      "step": 7660
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1496005654335022,
+      "learning_rate": 0.0003920368062916647,
+      "loss": 1.6807,
+      "step": 7670
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.14550770819187164,
+      "learning_rate": 0.00039201616759014323,
+      "loss": 1.7223,
+      "step": 7680
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.14356763660907745,
+      "learning_rate": 0.00039199550272260645,
+      "loss": 1.7058,
+      "step": 7690
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.14512434601783752,
+      "learning_rate": 0.0003919748116918703,
+      "loss": 1.7302,
+      "step": 7700
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1333114504814148,
+      "learning_rate": 0.00039195409450075436,
+      "loss": 1.7544,
+      "step": 7710
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.14479930698871613,
+      "learning_rate": 0.00039193335115208177,
+      "loss": 1.6746,
+      "step": 7720
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.13208623230457306,
+      "learning_rate": 0.00039191258164867916,
+      "loss": 1.6208,
+      "step": 7730
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1354825794696808,
+      "learning_rate": 0.0003918917859933769,
+      "loss": 1.6876,
+      "step": 7740
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.13403142988681793,
+      "learning_rate": 0.0003918709641890087,
+      "loss": 1.7227,
+      "step": 7750
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.14381615817546844,
+      "learning_rate": 0.000391850116238412,
+      "loss": 1.7347,
+      "step": 7760
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.148904949426651,
+      "learning_rate": 0.0003918292421444277,
+      "loss": 1.7538,
+      "step": 7770
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.14611361920833588,
+      "learning_rate": 0.0003918083419099004,
+      "loss": 1.7352,
+      "step": 7780
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.13526111841201782,
+      "learning_rate": 0.0003917874155376781,
+      "loss": 1.7334,
+      "step": 7790
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.14207859337329865,
+      "learning_rate": 0.00039176646303061243,
+      "loss": 1.7621,
+      "step": 7800
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15307915210723877,
+      "learning_rate": 0.00039174548439155864,
+      "loss": 1.7134,
+      "step": 7810
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.14540807902812958,
+      "learning_rate": 0.0003917244796233754,
+      "loss": 1.7038,
+      "step": 7820
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.14969046413898468,
+      "learning_rate": 0.00039170344872892514,
+      "loss": 1.6982,
+      "step": 7830
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1437288224697113,
+      "learning_rate": 0.0003916823917110737,
+      "loss": 1.7362,
+      "step": 7840
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.13481366634368896,
+      "learning_rate": 0.00039166130857269037,
+      "loss": 1.757,
+      "step": 7850
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16136403381824493,
+      "learning_rate": 0.0003916401993166484,
+      "loss": 1.6612,
+      "step": 7860
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.13674139976501465,
+      "learning_rate": 0.0003916190639458242,
+      "loss": 1.6701,
+      "step": 7870
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.155400812625885,
+      "learning_rate": 0.0003915979024630978,
+      "loss": 1.7115,
+      "step": 7880
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15934404730796814,
+      "learning_rate": 0.000391576714871353,
+      "loss": 1.7014,
+      "step": 7890
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.13780388236045837,
+      "learning_rate": 0.00039155550117347704,
+      "loss": 1.6889,
+      "step": 7900
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.14378343522548676,
+      "learning_rate": 0.0003915342613723607,
+      "loss": 1.7577,
+      "step": 7910
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2770020067691803,
+      "learning_rate": 0.00039151299547089826,
+      "loss": 1.7168,
+      "step": 7920
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16923096776008606,
+      "learning_rate": 0.0003914917034719876,
+      "loss": 1.7309,
+      "step": 7930
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.14824189245700836,
+      "learning_rate": 0.0003914703853785303,
+      "loss": 1.7077,
+      "step": 7940
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.148031547665596,
+      "learning_rate": 0.0003914490411934313,
+      "loss": 1.7435,
+      "step": 7950
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.14204919338226318,
+      "learning_rate": 0.0003914276709195991,
+      "loss": 1.7234,
+      "step": 7960
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15563704073429108,
+      "learning_rate": 0.000391406274559946,
+      "loss": 1.7848,
+      "step": 7970
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.14522221684455872,
+      "learning_rate": 0.00039138485211738755,
+      "loss": 1.7095,
+      "step": 7980
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.14838719367980957,
+      "learning_rate": 0.00039136340359484297,
+      "loss": 1.726,
+      "step": 7990
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15066173672676086,
+      "learning_rate": 0.00039134192899523514,
+      "loss": 1.7538,
+      "step": 8000
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15595562756061554,
+      "learning_rate": 0.00039132042832149027,
+      "loss": 1.6987,
+      "step": 8010
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1424400955438614,
+      "learning_rate": 0.0003912989015765384,
+      "loss": 1.6906,
+      "step": 8020
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.14538030326366425,
+      "learning_rate": 0.0003912773487633129,
+      "loss": 1.6979,
+      "step": 8030
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1990845650434494,
+      "learning_rate": 0.00039125576988475075,
+      "loss": 1.7169,
+      "step": 8040
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.14819040894508362,
+      "learning_rate": 0.0003912341649437925,
+      "loss": 1.6725,
+      "step": 8050
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.23003418743610382,
+      "learning_rate": 0.0003912125339433824,
+      "loss": 1.7566,
+      "step": 8060
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1365618258714676,
+      "learning_rate": 0.0003911908768864679,
+      "loss": 1.7369,
+      "step": 8070
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1465931236743927,
+      "learning_rate": 0.0003911691937760003,
+      "loss": 1.7282,
+      "step": 8080
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.13349132239818573,
+      "learning_rate": 0.0003911474846149344,
+      "loss": 1.6716,
+      "step": 8090
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1357334852218628,
+      "learning_rate": 0.00039112574940622835,
+      "loss": 1.7131,
+      "step": 8100
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.14389128983020782,
+      "learning_rate": 0.0003911039881528441,
+      "loss": 1.7469,
+      "step": 8110
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.13784164190292358,
+      "learning_rate": 0.0003910822008577471,
+      "loss": 1.7734,
+      "step": 8120
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.13456936180591583,
+      "learning_rate": 0.00039106038752390623,
+      "loss": 1.6929,
+      "step": 8130
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.12721242010593414,
+      "learning_rate": 0.00039103854815429406,
+      "loss": 1.6928,
+      "step": 8140
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.12918342649936676,
+      "learning_rate": 0.00039101668275188655,
+      "loss": 1.7344,
+      "step": 8150
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1323053538799286,
+      "learning_rate": 0.0003909947913196633,
+      "loss": 1.6671,
+      "step": 8160
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.13387516140937805,
+      "learning_rate": 0.00039097287386060753,
+      "loss": 1.657,
+      "step": 8170
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1323385089635849,
+      "learning_rate": 0.0003909509303777059,
+      "loss": 1.6923,
+      "step": 8180
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.12993253767490387,
+      "learning_rate": 0.0003909289608739486,
+      "loss": 1.6682,
+      "step": 8190
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.14957071840763092,
+      "learning_rate": 0.0003909069653523295,
+      "loss": 1.6946,
+      "step": 8200
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1388004869222641,
+      "learning_rate": 0.0003908849438158458,
+      "loss": 1.7008,
+      "step": 8210
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1595703661441803,
+      "learning_rate": 0.00039086289626749854,
+      "loss": 1.7514,
+      "step": 8220
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1537179946899414,
+      "learning_rate": 0.0003908408227102919,
+      "loss": 1.695,
+      "step": 8230
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1576649695634842,
+      "learning_rate": 0.0003908187231472341,
+      "loss": 1.6994,
+      "step": 8240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.14125628769397736,
+      "learning_rate": 0.0003907965975813365,
+      "loss": 1.7204,
+      "step": 8250
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1359732747077942,
+      "learning_rate": 0.0003907744460156142,
+      "loss": 1.7711,
+      "step": 8260
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1417861431837082,
+      "learning_rate": 0.0003907522684530857,
+      "loss": 1.7593,
+      "step": 8270
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16784122586250305,
+      "learning_rate": 0.00039073006489677317,
+      "loss": 1.7329,
+      "step": 8280
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.14502134919166565,
+      "learning_rate": 0.00039070783534970235,
+      "loss": 1.6954,
+      "step": 8290
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13937999308109283,
+      "learning_rate": 0.00039068557981490246,
+      "loss": 1.6908,
+      "step": 8300
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1434570550918579,
+      "learning_rate": 0.00039066329829540614,
+      "loss": 1.7373,
+      "step": 8310
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13864974677562714,
+      "learning_rate": 0.00039064099079424985,
+      "loss": 1.7309,
+      "step": 8320
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.17600135505199432,
+      "learning_rate": 0.00039061865731447327,
+      "loss": 1.6646,
+      "step": 8330
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14923590421676636,
+      "learning_rate": 0.00039059629785911985,
+      "loss": 1.7094,
+      "step": 8340
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16249333322048187,
+      "learning_rate": 0.00039057391243123654,
+      "loss": 1.672,
+      "step": 8350
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15163445472717285,
+      "learning_rate": 0.0003905515010338738,
+      "loss": 1.7534,
+      "step": 8360
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13050870597362518,
+      "learning_rate": 0.00039052906367008556,
+      "loss": 1.6948,
+      "step": 8370
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14099529385566711,
+      "learning_rate": 0.0003905066003429294,
+      "loss": 1.7127,
+      "step": 8380
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13801635801792145,
+      "learning_rate": 0.00039048411105546644,
+      "loss": 1.7082,
+      "step": 8390
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13237683475017548,
+      "learning_rate": 0.00039046159581076123,
+      "loss": 1.6959,
+      "step": 8400
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.144191712141037,
+      "learning_rate": 0.00039043905461188193,
+      "loss": 1.6596,
+      "step": 8410
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1366361379623413,
+      "learning_rate": 0.0003904164874619002,
+      "loss": 1.7135,
+      "step": 8420
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13347311317920685,
+      "learning_rate": 0.0003903938943638913,
+      "loss": 1.6979,
+      "step": 8430
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1298627257347107,
+      "learning_rate": 0.000390371275320934,
+      "loss": 1.6883,
+      "step": 8440
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15101325511932373,
+      "learning_rate": 0.0003903486303361106,
+      "loss": 1.7202,
+      "step": 8450
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13881361484527588,
+      "learning_rate": 0.00039032595941250687,
+      "loss": 1.7568,
+      "step": 8460
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1474679857492447,
+      "learning_rate": 0.0003903032625532122,
+      "loss": 1.7472,
+      "step": 8470
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13569600880146027,
+      "learning_rate": 0.00039028053976131947,
+      "loss": 1.6686,
+      "step": 8480
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15183793008327484,
+      "learning_rate": 0.0003902577910399252,
+      "loss": 1.7271,
+      "step": 8490
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14885376393795013,
+      "learning_rate": 0.00039023501639212925,
+      "loss": 1.7249,
+      "step": 8500
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13697759807109833,
+      "learning_rate": 0.0003902122158210352,
+      "loss": 1.7183,
+      "step": 8510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13530884683132172,
+      "learning_rate": 0.00039018938932974996,
+      "loss": 1.7494,
+      "step": 8520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13527202606201172,
+      "learning_rate": 0.0003901665369213842,
+      "loss": 1.7127,
+      "step": 8530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13087345659732819,
+      "learning_rate": 0.000390143658599052,
+      "loss": 1.7008,
+      "step": 8540
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1484183967113495,
+      "learning_rate": 0.00039012075436587097,
+      "loss": 1.719,
+      "step": 8550
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13455341756343842,
+      "learning_rate": 0.0003900978242249623,
+      "loss": 1.6938,
+      "step": 8560
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13760384917259216,
+      "learning_rate": 0.0003900748681794506,
+      "loss": 1.685,
+      "step": 8570
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13497385382652283,
+      "learning_rate": 0.00039005188623246413,
+      "loss": 1.7293,
+      "step": 8580
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1361473649740219,
+      "learning_rate": 0.0003900288783871347,
+      "loss": 1.7187,
+      "step": 8590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13627833127975464,
+      "learning_rate": 0.00039000584464659753,
+      "loss": 1.6916,
+      "step": 8600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13532902300357819,
+      "learning_rate": 0.00038998278501399135,
+      "loss": 1.747,
+      "step": 8610
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13195374608039856,
+      "learning_rate": 0.00038995969949245853,
+      "loss": 1.6929,
+      "step": 8620
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13569962978363037,
+      "learning_rate": 0.000389936588085145,
+      "loss": 1.65,
+      "step": 8630
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15099965035915375,
+      "learning_rate": 0.00038991345079520014,
+      "loss": 1.7891,
+      "step": 8640
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13594239950180054,
+      "learning_rate": 0.0003898902876257768,
+      "loss": 1.6748,
+      "step": 8650
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14765936136245728,
+      "learning_rate": 0.0003898670985800315,
+      "loss": 1.7528,
+      "step": 8660
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13044315576553345,
+      "learning_rate": 0.00038984388366112416,
+      "loss": 1.6972,
+      "step": 8670
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13899655640125275,
+      "learning_rate": 0.0003898206428722182,
+      "loss": 1.6653,
+      "step": 8680
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13996469974517822,
+      "learning_rate": 0.0003897973762164808,
+      "loss": 1.6601,
+      "step": 8690
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14995421469211578,
+      "learning_rate": 0.0003897740836970825,
+      "loss": 1.7096,
+      "step": 8700
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14182275533676147,
+      "learning_rate": 0.00038975076531719716,
+      "loss": 1.71,
+      "step": 8710
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14055702090263367,
+      "learning_rate": 0.0003897274210800025,
+      "loss": 1.721,
+      "step": 8720
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14095278084278107,
+      "learning_rate": 0.0003897040509886797,
+      "loss": 1.6573,
+      "step": 8730
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1559416651725769,
+      "learning_rate": 0.00038968065504641327,
+      "loss": 1.6651,
+      "step": 8740
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14234113693237305,
+      "learning_rate": 0.00038965723325639154,
+      "loss": 1.7532,
+      "step": 8750
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1320653259754181,
+      "learning_rate": 0.000389633785621806,
+      "loss": 1.704,
+      "step": 8760
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14077289402484894,
+      "learning_rate": 0.000389610312145852,
+      "loss": 1.6925,
+      "step": 8770
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15022748708724976,
+      "learning_rate": 0.0003895868128317282,
+      "loss": 1.6837,
+      "step": 8780
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14192801713943481,
+      "learning_rate": 0.0003895632876826369,
+      "loss": 1.6562,
+      "step": 8790
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1271694451570511,
+      "learning_rate": 0.0003895397367017838,
+      "loss": 1.6415,
+      "step": 8800
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14491376280784607,
+      "learning_rate": 0.00038951615989237816,
+      "loss": 1.6486,
+      "step": 8810
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13827703893184662,
+      "learning_rate": 0.0003894925572576329,
+      "loss": 1.6637,
+      "step": 8820
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13990582525730133,
+      "learning_rate": 0.0003894689288007643,
+      "loss": 1.7087,
+      "step": 8830
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14701290428638458,
+      "learning_rate": 0.0003894452745249922,
+      "loss": 1.651,
+      "step": 8840
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.21782691776752472,
+      "learning_rate": 0.00038942159443353997,
+      "loss": 1.6388,
+      "step": 8850
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16319188475608826,
+      "learning_rate": 0.0003893978885296345,
+      "loss": 1.6662,
+      "step": 8860
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14546512067317963,
+      "learning_rate": 0.0003893741568165062,
+      "loss": 1.6514,
+      "step": 8870
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.145062655210495,
+      "learning_rate": 0.0003893503992973889,
+      "loss": 1.6861,
+      "step": 8880
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14725039899349213,
+      "learning_rate": 0.00038932661597552016,
+      "loss": 1.7311,
+      "step": 8890
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15224291384220123,
+      "learning_rate": 0.00038930280685414085,
+      "loss": 1.7714,
+      "step": 8900
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14499187469482422,
+      "learning_rate": 0.0003892789719364955,
+      "loss": 1.6955,
+      "step": 8910
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13085386157035828,
+      "learning_rate": 0.00038925511122583196,
+      "loss": 1.649,
+      "step": 8920
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.12321346998214722,
+      "learning_rate": 0.0003892312247254019,
+      "loss": 1.6696,
+      "step": 8930
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1457478255033493,
+      "learning_rate": 0.00038920731243846017,
+      "loss": 1.6824,
+      "step": 8940
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1479731947183609,
+      "learning_rate": 0.0003891833743682654,
+      "loss": 1.6955,
+      "step": 8950
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14455614984035492,
+      "learning_rate": 0.00038915941051807963,
+      "loss": 1.7087,
+      "step": 8960
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1325332075357437,
+      "learning_rate": 0.00038913542089116836,
+      "loss": 1.6943,
+      "step": 8970
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14660602807998657,
+      "learning_rate": 0.00038911140549080064,
+      "loss": 1.716,
+      "step": 8980
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1645248383283615,
+      "learning_rate": 0.0003890873643202491,
+      "loss": 1.7111,
+      "step": 8990
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13546110689640045,
+      "learning_rate": 0.00038906329738278976,
+      "loss": 1.754,
+      "step": 9000
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.12389056384563446,
+      "learning_rate": 0.0003890392046817023,
+      "loss": 1.6444,
+      "step": 9010
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13471394777297974,
+      "learning_rate": 0.0003890150862202698,
+      "loss": 1.6893,
+      "step": 9020
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14159949123859406,
+      "learning_rate": 0.0003889909420017788,
+      "loss": 1.685,
+      "step": 9030
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13195070624351501,
+      "learning_rate": 0.00038896677202951955,
+      "loss": 1.6555,
+      "step": 9040
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13802525401115417,
+      "learning_rate": 0.00038894257630678564,
+      "loss": 1.6851,
+      "step": 9050
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.130794957280159,
+      "learning_rate": 0.00038891835483687416,
+      "loss": 1.6961,
+      "step": 9060
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13027848303318024,
+      "learning_rate": 0.0003888941076230858,
+      "loss": 1.6602,
+      "step": 9070
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15656857192516327,
+      "learning_rate": 0.0003888698346687249,
+      "loss": 1.7005,
+      "step": 9080
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13540861010551453,
+      "learning_rate": 0.00038884553597709877,
+      "loss": 1.6786,
+      "step": 9090
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.12966085970401764,
+      "learning_rate": 0.0003888212115515189,
+      "loss": 1.7051,
+      "step": 9100
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1450546681880951,
+      "learning_rate": 0.0003887968613952999,
+      "loss": 1.7378,
+      "step": 9110
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13198581337928772,
+      "learning_rate": 0.0003887724855117599,
+      "loss": 1.7166,
+      "step": 9120
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14655737578868866,
+      "learning_rate": 0.00038874808390422055,
+      "loss": 1.7048,
+      "step": 9130
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14178557693958282,
+      "learning_rate": 0.00038872365657600727,
+      "loss": 1.7175,
+      "step": 9140
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1497267335653305,
+      "learning_rate": 0.0003886992035304485,
+      "loss": 1.6875,
+      "step": 9150
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.13873864710330963,
+      "learning_rate": 0.00038867472477087666,
+      "loss": 1.6518,
+      "step": 9160
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14022114872932434,
+      "learning_rate": 0.00038865022030062745,
+      "loss": 1.6726,
+      "step": 9170
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16763968765735626,
+      "learning_rate": 0.0003886256901230399,
+      "loss": 1.6726,
+      "step": 9180
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.13509979844093323,
+      "learning_rate": 0.00038860113424145696,
+      "loss": 1.6489,
+      "step": 9190
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14755505323410034,
+      "learning_rate": 0.00038857655265922477,
+      "loss": 1.7358,
+      "step": 9200
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1449791043996811,
+      "learning_rate": 0.00038855194537969293,
+      "loss": 1.729,
+      "step": 9210
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14036956429481506,
+      "learning_rate": 0.0003885273124062149,
+      "loss": 1.6459,
+      "step": 9220
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.13039113581180573,
+      "learning_rate": 0.00038850265374214726,
+      "loss": 1.6881,
+      "step": 9230
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.19317910075187683,
+      "learning_rate": 0.00038847796939085033,
+      "loss": 1.7277,
+      "step": 9240
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1365085244178772,
+      "learning_rate": 0.0003884532593556878,
+      "loss": 1.6708,
+      "step": 9250
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14536632597446442,
+      "learning_rate": 0.00038842852364002686,
+      "loss": 1.6428,
+      "step": 9260
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.17549118399620056,
+      "learning_rate": 0.0003884037622472383,
+      "loss": 1.7109,
+      "step": 9270
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16368693113327026,
+      "learning_rate": 0.00038837897518069627,
+      "loss": 1.659,
+      "step": 9280
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1407514065504074,
+      "learning_rate": 0.0003883541624437787,
+      "loss": 1.6635,
+      "step": 9290
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.18877118825912476,
+      "learning_rate": 0.0003883293240398666,
+      "loss": 1.7073,
+      "step": 9300
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1376110315322876,
+      "learning_rate": 0.0003883044599723448,
+      "loss": 1.6782,
+      "step": 9310
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1291997730731964,
+      "learning_rate": 0.0003882795702446015,
+      "loss": 1.6951,
+      "step": 9320
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1362183392047882,
+      "learning_rate": 0.0003882546548600284,
+      "loss": 1.6561,
+      "step": 9330
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1504468470811844,
+      "learning_rate": 0.0003882297138220207,
+      "loss": 1.7128,
+      "step": 9340
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.15345141291618347,
+      "learning_rate": 0.00038820474713397726,
+      "loss": 1.7581,
+      "step": 9350
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.15381696820259094,
+      "learning_rate": 0.0003881797547993001,
+      "loss": 1.6886,
+      "step": 9360
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14143483340740204,
+      "learning_rate": 0.00038815473682139506,
+      "loss": 1.6893,
+      "step": 9370
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.13443134725093842,
+      "learning_rate": 0.0003881296932036712,
+      "loss": 1.7155,
+      "step": 9380
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1517164260149002,
+      "learning_rate": 0.00038810462394954135,
+      "loss": 1.7083,
+      "step": 9390
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.13738253712654114,
+      "learning_rate": 0.0003880795290624216,
+      "loss": 1.7058,
+      "step": 9400
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1431797295808792,
+      "learning_rate": 0.0003880544085457316,
+      "loss": 1.6776,
+      "step": 9410
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1337369680404663,
+      "learning_rate": 0.00038802926240289463,
+      "loss": 1.6781,
+      "step": 9420
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1325959414243698,
+      "learning_rate": 0.0003880040906373373,
+      "loss": 1.7043,
+      "step": 9430
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.13038308918476105,
+      "learning_rate": 0.00038797889325248975,
+      "loss": 1.6585,
+      "step": 9440
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1321701556444168,
+      "learning_rate": 0.0003879536702517856,
+      "loss": 1.6258,
+      "step": 9450
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1335831731557846,
+      "learning_rate": 0.0003879284216386621,
+      "loss": 1.6631,
+      "step": 9460
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.13223014771938324,
+      "learning_rate": 0.0003879031474165597,
+      "loss": 1.6625,
+      "step": 9470
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.152293398976326,
+      "learning_rate": 0.00038787784758892267,
+      "loss": 1.6885,
+      "step": 9480
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.13212381303310394,
+      "learning_rate": 0.0003878525221591985,
+      "loss": 1.6607,
+      "step": 9490
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.13295498490333557,
+      "learning_rate": 0.0003878271711308384,
+      "loss": 1.7179,
+      "step": 9500
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16751062870025635,
+      "learning_rate": 0.00038780179450729677,
+      "loss": 1.7262,
+      "step": 9510
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14195337891578674,
+      "learning_rate": 0.00038777639229203186,
+      "loss": 1.6958,
+      "step": 9520
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1498609334230423,
+      "learning_rate": 0.0003877509644885052,
+      "loss": 1.7011,
+      "step": 9530
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1556314378976822,
+      "learning_rate": 0.00038772551110018177,
+      "loss": 1.7051,
+      "step": 9540
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.13929566740989685,
+      "learning_rate": 0.00038770003213053005,
+      "loss": 1.7249,
+      "step": 9550
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1451740711927414,
+      "learning_rate": 0.0003876745275830222,
+      "loss": 1.7215,
+      "step": 9560
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14072318375110626,
+      "learning_rate": 0.0003876489974611336,
+      "loss": 1.6929,
+      "step": 9570
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.15003930032253265,
+      "learning_rate": 0.0003876234417683433,
+      "loss": 1.6999,
+      "step": 9580
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13815206289291382,
+      "learning_rate": 0.0003875978605081338,
+      "loss": 1.7009,
+      "step": 9590
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.12934628129005432,
+      "learning_rate": 0.00038757225368399095,
+      "loss": 1.6663,
+      "step": 9600
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13583825528621674,
+      "learning_rate": 0.00038754662129940435,
+      "loss": 1.6905,
+      "step": 9610
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13525907695293427,
+      "learning_rate": 0.0003875209633578667,
+      "loss": 1.652,
+      "step": 9620
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.138809934258461,
+      "learning_rate": 0.00038749527986287454,
+      "loss": 1.6523,
+      "step": 9630
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.21345479786396027,
+      "learning_rate": 0.0003874695708179278,
+      "loss": 1.7747,
+      "step": 9640
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1562160700559616,
+      "learning_rate": 0.00038744383622652976,
+      "loss": 1.6557,
+      "step": 9650
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16242921352386475,
+      "learning_rate": 0.00038741807609218723,
+      "loss": 1.7411,
+      "step": 9660
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1379001885652542,
+      "learning_rate": 0.0003873922904184107,
+      "loss": 1.6967,
+      "step": 9670
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.14510823786258698,
+      "learning_rate": 0.0003873664792087138,
+      "loss": 1.7457,
+      "step": 9680
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13455097377300262,
+      "learning_rate": 0.000387340642466614,
+      "loss": 1.6778,
+      "step": 9690
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.15125523507595062,
+      "learning_rate": 0.0003873147801956319,
+      "loss": 1.6823,
+      "step": 9700
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1279924213886261,
+      "learning_rate": 0.00038728889239929184,
+      "loss": 1.6838,
+      "step": 9710
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13603992760181427,
+      "learning_rate": 0.00038726297908112155,
+      "loss": 1.6892,
+      "step": 9720
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.14827416837215424,
+      "learning_rate": 0.0003872370402446522,
+      "loss": 1.7074,
+      "step": 9730
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13511140644550323,
+      "learning_rate": 0.0003872110758934185,
+      "loss": 1.6726,
+      "step": 9740
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13389559090137482,
+      "learning_rate": 0.0003871850860309585,
+      "loss": 1.6368,
+      "step": 9750
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1559256613254547,
+      "learning_rate": 0.00038715907066081404,
+      "loss": 1.6941,
+      "step": 9760
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13027766346931458,
+      "learning_rate": 0.0003871330297865301,
+      "loss": 1.6764,
+      "step": 9770
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13140283524990082,
+      "learning_rate": 0.00038710696341165527,
+      "loss": 1.6284,
+      "step": 9780
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13621726632118225,
+      "learning_rate": 0.00038708087153974157,
+      "loss": 1.6927,
+      "step": 9790
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.22002233564853668,
+      "learning_rate": 0.0003870547541743446,
+      "loss": 1.753,
+      "step": 9800
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.163285031914711,
+      "learning_rate": 0.00038702861131902343,
+      "loss": 1.7628,
+      "step": 9810
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.14639011025428772,
+      "learning_rate": 0.00038700244297734047,
+      "loss": 1.7623,
+      "step": 9820
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1737416684627533,
+      "learning_rate": 0.0003869762491528616,
+      "loss": 1.7072,
+      "step": 9830
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1327442079782486,
+      "learning_rate": 0.00038695002984915646,
+      "loss": 1.6497,
+      "step": 9840
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.14944949746131897,
+      "learning_rate": 0.0003869237850697977,
+      "loss": 1.6907,
+      "step": 9850
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1352148801088333,
+      "learning_rate": 0.0003868975148183618,
+      "loss": 1.7065,
+      "step": 9860
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13450466096401215,
+      "learning_rate": 0.0003868712190984287,
+      "loss": 1.6629,
+      "step": 9870
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.14856620132923126,
+      "learning_rate": 0.0003868448979135816,
+      "loss": 1.7225,
+      "step": 9880
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1564270704984665,
+      "learning_rate": 0.0003868185512674073,
+      "loss": 1.7181,
+      "step": 9890
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13846179842948914,
+      "learning_rate": 0.00038679217916349606,
+      "loss": 1.6527,
+      "step": 9900
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13526107370853424,
+      "learning_rate": 0.0003867657816054417,
+      "loss": 1.65,
+      "step": 9910
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.18866793811321259,
+      "learning_rate": 0.0003867393585968412,
+      "loss": 1.7315,
+      "step": 9920
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1704215109348297,
+      "learning_rate": 0.0003867129101412954,
+      "loss": 1.6784,
+      "step": 9930
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13856296241283417,
+      "learning_rate": 0.00038668643624240837,
+      "loss": 1.7344,
+      "step": 9940
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1453276425600052,
+      "learning_rate": 0.00038665993690378765,
+      "loss": 1.6428,
+      "step": 9950
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.14118853211402893,
+      "learning_rate": 0.0003866334121290444,
+      "loss": 1.6918,
+      "step": 9960
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1492616832256317,
+      "learning_rate": 0.00038660686192179306,
+      "loss": 1.7261,
+      "step": 9970
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1425030678510666,
+      "learning_rate": 0.0003865802862856517,
+      "loss": 1.7019,
+      "step": 9980
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1411997377872467,
+      "learning_rate": 0.0003865536852242417,
+      "loss": 1.6625,
+      "step": 9990
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1392057240009308,
+      "learning_rate": 0.0003865270587411881,
+      "loss": 1.6669,
+      "step": 10000
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1435200423002243,
+      "learning_rate": 0.0003865004068401191,
+      "loss": 1.7356,
+      "step": 10010
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13941244781017303,
+      "learning_rate": 0.0003864737295246667,
+      "loss": 1.7035,
+      "step": 10020
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1439218372106552,
+      "learning_rate": 0.0003864470267984663,
+      "loss": 1.6769,
+      "step": 10030
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13170225918293,
+      "learning_rate": 0.0003864202986651564,
+      "loss": 1.6613,
+      "step": 10040
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13548241555690765,
+      "learning_rate": 0.0003863935451283794,
+      "loss": 1.6213,
+      "step": 10050
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15594840049743652,
+      "learning_rate": 0.0003863667661917811,
+      "loss": 1.7619,
+      "step": 10060
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1251944750547409,
+      "learning_rate": 0.00038633996185901044,
+      "loss": 1.6814,
+      "step": 10070
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13390634953975677,
+      "learning_rate": 0.00038631313213372025,
+      "loss": 1.6783,
+      "step": 10080
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16436685621738434,
+      "learning_rate": 0.0003862862770195665,
+      "loss": 1.6136,
+      "step": 10090
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1480519026517868,
+      "learning_rate": 0.00038625939652020867,
+      "loss": 1.685,
+      "step": 10100
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13994118571281433,
+      "learning_rate": 0.00038623249063930986,
+      "loss": 1.6469,
+      "step": 10110
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.17021489143371582,
+      "learning_rate": 0.00038620555938053655,
+      "loss": 1.6881,
+      "step": 10120
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14997856318950653,
+      "learning_rate": 0.0003861786027475586,
+      "loss": 1.6629,
+      "step": 10130
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13746066391468048,
+      "learning_rate": 0.0003861516207440495,
+      "loss": 1.69,
+      "step": 10140
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.17035061120986938,
+      "learning_rate": 0.0003861246133736859,
+      "loss": 1.6606,
+      "step": 10150
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13511179387569427,
+      "learning_rate": 0.00038609758064014817,
+      "loss": 1.6296,
+      "step": 10160
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15048129856586456,
+      "learning_rate": 0.0003860705225471201,
+      "loss": 1.704,
+      "step": 10170
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1398865431547165,
+      "learning_rate": 0.00038604343909828885,
+      "loss": 1.6259,
+      "step": 10180
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13808929920196533,
+      "learning_rate": 0.0003860163302973451,
+      "loss": 1.6376,
+      "step": 10190
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1575845628976822,
+      "learning_rate": 0.000385989196147983,
+      "loss": 1.7395,
+      "step": 10200
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13429851830005646,
+      "learning_rate": 0.0003859620366539,
+      "loss": 1.6674,
+      "step": 10210
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1420392394065857,
+      "learning_rate": 0.00038593485181879724,
+      "loss": 1.6765,
+      "step": 10220
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15411992371082306,
+      "learning_rate": 0.0003859076416463791,
+      "loss": 1.702,
+      "step": 10230
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.12618638575077057,
+      "learning_rate": 0.00038588040614035357,
+      "loss": 1.6554,
+      "step": 10240
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13572292029857635,
+      "learning_rate": 0.000385853145304432,
+      "loss": 1.6874,
+      "step": 10250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13603955507278442,
+      "learning_rate": 0.0003858258591423292,
+      "loss": 1.7074,
+      "step": 10260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1411031037569046,
+      "learning_rate": 0.00038579854765776356,
+      "loss": 1.6871,
+      "step": 10270
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1405111402273178,
+      "learning_rate": 0.00038577121085445665,
+      "loss": 1.6843,
+      "step": 10280
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14742854237556458,
+      "learning_rate": 0.00038574384873613375,
+      "loss": 1.6349,
+      "step": 10290
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.12978221476078033,
+      "learning_rate": 0.0003857164613065235,
+      "loss": 1.6206,
+      "step": 10300
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1334395706653595,
+      "learning_rate": 0.00038568904856935794,
+      "loss": 1.6445,
+      "step": 10310
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14176109433174133,
+      "learning_rate": 0.0003856616105283726,
+      "loss": 1.6516,
+      "step": 10320
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13495835661888123,
+      "learning_rate": 0.0003856341471873065,
+      "loss": 1.6288,
+      "step": 10330
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1410268247127533,
+      "learning_rate": 0.00038560665854990204,
+      "loss": 1.7328,
+      "step": 10340
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14796845614910126,
+      "learning_rate": 0.000385579144619905,
+      "loss": 1.7034,
+      "step": 10350
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13006213307380676,
+      "learning_rate": 0.00038555160540106493,
+      "loss": 1.6634,
+      "step": 10360
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.17649726569652557,
+      "learning_rate": 0.0003855240408971344,
+      "loss": 1.6931,
+      "step": 10370
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14633838832378387,
+      "learning_rate": 0.0003854964511118697,
+      "loss": 1.7006,
+      "step": 10380
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1325550377368927,
+      "learning_rate": 0.0003854688360490304,
+      "loss": 1.6795,
+      "step": 10390
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13270264863967896,
+      "learning_rate": 0.0003854411957123797,
+      "loss": 1.6789,
+      "step": 10400
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13439826667308807,
+      "learning_rate": 0.00038541353010568415,
+      "loss": 1.68,
+      "step": 10410
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14042603969573975,
+      "learning_rate": 0.0003853858392327136,
+      "loss": 1.6885,
+      "step": 10420
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13671109080314636,
+      "learning_rate": 0.0003853581230972417,
+      "loss": 1.6751,
+      "step": 10430
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14137002825737,
+      "learning_rate": 0.0003853303817030451,
+      "loss": 1.6279,
+      "step": 10440
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1310359090566635,
+      "learning_rate": 0.0003853026150539042,
+      "loss": 1.7029,
+      "step": 10450
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13415858149528503,
+      "learning_rate": 0.0003852748231536029,
+      "loss": 1.6509,
+      "step": 10460
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13460080325603485,
+      "learning_rate": 0.00038524700600592814,
+      "loss": 1.6602,
+      "step": 10470
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13028070330619812,
+      "learning_rate": 0.00038521916361467076,
+      "loss": 1.6457,
+      "step": 10480
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1308981478214264,
+      "learning_rate": 0.00038519129598362477,
+      "loss": 1.6638,
+      "step": 10490
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13529054820537567,
+      "learning_rate": 0.0003851634031165877,
+      "loss": 1.6556,
+      "step": 10500
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1281631588935852,
+      "learning_rate": 0.0003851354850173604,
+      "loss": 1.6516,
+      "step": 10510
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.21898482739925385,
+      "learning_rate": 0.00038510754168974746,
+      "loss": 1.6933,
+      "step": 10520
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1400601863861084,
+      "learning_rate": 0.00038507957313755665,
+      "loss": 1.6911,
+      "step": 10530
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13771529495716095,
+      "learning_rate": 0.0003850515793645991,
+      "loss": 1.6812,
+      "step": 10540
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14199618995189667,
+      "learning_rate": 0.00038502356037468965,
+      "loss": 1.6635,
+      "step": 10550
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13736966252326965,
+      "learning_rate": 0.00038499551617164645,
+      "loss": 1.6241,
+      "step": 10560
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.12804976105690002,
+      "learning_rate": 0.000384967446759291,
+      "loss": 1.6394,
+      "step": 10570
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14472809433937073,
+      "learning_rate": 0.00038493935214144843,
+      "loss": 1.6647,
+      "step": 10580
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.15256865322589874,
+      "learning_rate": 0.00038491123232194704,
+      "loss": 1.7305,
+      "step": 10590
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.145729199051857,
+      "learning_rate": 0.00038488308730461886,
+      "loss": 1.6832,
+      "step": 10600
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1337866336107254,
+      "learning_rate": 0.0003848549170932991,
+      "loss": 1.6071,
+      "step": 10610
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13401570916175842,
+      "learning_rate": 0.0003848267216918266,
+      "loss": 1.6091,
+      "step": 10620
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13169196248054504,
+      "learning_rate": 0.0003847985011040434,
+      "loss": 1.6604,
+      "step": 10630
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14133350551128387,
+      "learning_rate": 0.0003847702553337952,
+      "loss": 1.6919,
+      "step": 10640
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.12538287043571472,
+      "learning_rate": 0.0003847419843849312,
+      "loss": 1.6396,
+      "step": 10650
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14528688788414001,
+      "learning_rate": 0.0003847136882613036,
+      "loss": 1.6548,
+      "step": 10660
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14478085935115814,
+      "learning_rate": 0.00038468536696676856,
+      "loss": 1.6884,
+      "step": 10670
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16863298416137695,
+      "learning_rate": 0.00038465702050518517,
+      "loss": 1.7228,
+      "step": 10680
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13326987624168396,
+      "learning_rate": 0.00038462864888041644,
+      "loss": 1.6505,
+      "step": 10690
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13895738124847412,
+      "learning_rate": 0.0003846002520963284,
+      "loss": 1.6431,
+      "step": 10700
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13446184992790222,
+      "learning_rate": 0.00038457183015679075,
+      "loss": 1.6843,
+      "step": 10710
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13801562786102295,
+      "learning_rate": 0.0003845433830656765,
+      "loss": 1.6598,
+      "step": 10720
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.12989461421966553,
+      "learning_rate": 0.0003845149108268622,
+      "loss": 1.6852,
+      "step": 10730
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13179564476013184,
+      "learning_rate": 0.0003844864134442277,
+      "loss": 1.6415,
+      "step": 10740
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14234520494937897,
+      "learning_rate": 0.0003844578909216564,
+      "loss": 1.7073,
+      "step": 10750
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1515664905309677,
+      "learning_rate": 0.0003844293432630349,
+      "loss": 1.7238,
+      "step": 10760
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.17290601134300232,
+      "learning_rate": 0.0003844007704722536,
+      "loss": 1.6837,
+      "step": 10770
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1464894413948059,
+      "learning_rate": 0.00038437217255320595,
+      "loss": 1.6918,
+      "step": 10780
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13859233260154724,
+      "learning_rate": 0.00038434354950978906,
+      "loss": 1.6699,
+      "step": 10790
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1609164923429489,
+      "learning_rate": 0.00038431490134590337,
+      "loss": 1.6377,
+      "step": 10800
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.19883990287780762,
+      "learning_rate": 0.0003842862280654527,
+      "loss": 1.7499,
+      "step": 10810
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14206425845623016,
+      "learning_rate": 0.0003842575296723445,
+      "loss": 1.7072,
+      "step": 10820
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14590507745742798,
+      "learning_rate": 0.00038422880617048933,
+      "loss": 1.6986,
+      "step": 10830
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14065930247306824,
+      "learning_rate": 0.00038420005756380145,
+      "loss": 1.6698,
+      "step": 10840
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13067026436328888,
+      "learning_rate": 0.00038417128385619836,
+      "loss": 1.6227,
+      "step": 10850
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14096234738826752,
+      "learning_rate": 0.0003841424850516012,
+      "loss": 1.6479,
+      "step": 10860
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1395232230424881,
+      "learning_rate": 0.00038411366115393417,
+      "loss": 1.6281,
+      "step": 10870
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14379915595054626,
+      "learning_rate": 0.0003840848121671252,
+      "loss": 1.6553,
+      "step": 10880
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.13473458588123322,
+      "learning_rate": 0.00038405593809510546,
+      "loss": 1.7524,
+      "step": 10890
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16775494813919067,
+      "learning_rate": 0.00038402703894180985,
+      "loss": 1.7263,
+      "step": 10900
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1440572440624237,
+      "learning_rate": 0.00038399811471117614,
+      "loss": 1.6824,
+      "step": 10910
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14449696242809296,
+      "learning_rate": 0.00038396916540714605,
+      "loss": 1.6667,
+      "step": 10920
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14525967836380005,
+      "learning_rate": 0.0003839401910336644,
+      "loss": 1.6845,
+      "step": 10930
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1591772586107254,
+      "learning_rate": 0.0003839111915946795,
+      "loss": 1.6729,
+      "step": 10940
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.13817182183265686,
+      "learning_rate": 0.0003838821670941432,
+      "loss": 1.6986,
+      "step": 10950
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.15360833704471588,
+      "learning_rate": 0.00038385311753601063,
+      "loss": 1.697,
+      "step": 10960
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.13232658803462982,
+      "learning_rate": 0.00038382404292424034,
+      "loss": 1.6921,
+      "step": 10970
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.15353403985500336,
+      "learning_rate": 0.00038379494326279433,
+      "loss": 1.6952,
+      "step": 10980
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14521262049674988,
+      "learning_rate": 0.00038376581855563804,
+      "loss": 1.6929,
+      "step": 10990
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14435723423957825,
+      "learning_rate": 0.00038373666880674016,
+      "loss": 1.6781,
+      "step": 11000
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.13432404398918152,
+      "learning_rate": 0.0003837074940200731,
+      "loss": 1.6779,
+      "step": 11010
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.12679973244667053,
+      "learning_rate": 0.0003836782941996124,
+      "loss": 1.582,
+      "step": 11020
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.13105441629886627,
+      "learning_rate": 0.0003836490693493372,
+      "loss": 1.7472,
+      "step": 11030
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1506175696849823,
+      "learning_rate": 0.0003836198194732299,
+      "loss": 1.6584,
+      "step": 11040
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14088502526283264,
+      "learning_rate": 0.0003835905445752763,
+      "loss": 1.6933,
+      "step": 11050
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14554311335086823,
+      "learning_rate": 0.00038356124465946585,
+      "loss": 1.7205,
+      "step": 11060
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1857965588569641,
+      "learning_rate": 0.0003835319197297912,
+      "loss": 1.6814,
+      "step": 11070
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.167624831199646,
+      "learning_rate": 0.0003835025697902483,
+      "loss": 1.6152,
+      "step": 11080
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.15485908091068268,
+      "learning_rate": 0.00038347319484483683,
+      "loss": 1.657,
+      "step": 11090
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.15104304254055023,
+      "learning_rate": 0.00038344379489755974,
+      "loss": 1.637,
+      "step": 11100
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.132297083735466,
+      "learning_rate": 0.0003834143699524232,
+      "loss": 1.6593,
+      "step": 11110
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1328713595867157,
+      "learning_rate": 0.00038338492001343703,
+      "loss": 1.6385,
+      "step": 11120
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14161983132362366,
+      "learning_rate": 0.00038335544508461434,
+      "loss": 1.6647,
+      "step": 11130
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16857671737670898,
+      "learning_rate": 0.00038332594516997176,
+      "loss": 1.6488,
+      "step": 11140
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16283215582370758,
+      "learning_rate": 0.0003832964202735291,
+      "loss": 1.6551,
+      "step": 11150
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14631517231464386,
+      "learning_rate": 0.00038326687039930986,
+      "loss": 1.6439,
+      "step": 11160
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.13835646212100983,
+      "learning_rate": 0.00038323729555134065,
+      "loss": 1.6701,
+      "step": 11170
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.13729950785636902,
+      "learning_rate": 0.0003832076957336518,
+      "loss": 1.666,
+      "step": 11180
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.13787849247455597,
+      "learning_rate": 0.0003831780709502767,
+      "loss": 1.6191,
+      "step": 11190
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.13259431719779968,
+      "learning_rate": 0.0003831484212052524,
+      "loss": 1.6287,
+      "step": 11200
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1325533390045166,
+      "learning_rate": 0.00038311874650261927,
+      "loss": 1.6757,
+      "step": 11210
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1378818154335022,
+      "learning_rate": 0.0003830890468464211,
+      "loss": 1.6832,
+      "step": 11220
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.12810388207435608,
+      "learning_rate": 0.000383059322240705,
+      "loss": 1.6793,
+      "step": 11230
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.12740343809127808,
+      "learning_rate": 0.00038302957268952146,
+      "loss": 1.6851,
+      "step": 11240
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14042499661445618,
+      "learning_rate": 0.00038299979819692465,
+      "loss": 1.6539,
+      "step": 11250
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14357367157936096,
+      "learning_rate": 0.00038296999876697184,
+      "loss": 1.7095,
+      "step": 11260
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.13453765213489532,
+      "learning_rate": 0.0003829401744037238,
+      "loss": 1.6169,
+      "step": 11270
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.12387678772211075,
+      "learning_rate": 0.00038291032511124465,
+      "loss": 1.6301,
+      "step": 11280
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.12793013453483582,
+      "learning_rate": 0.00038288045089360196,
+      "loss": 1.6076,
+      "step": 11290
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13627608120441437,
+      "learning_rate": 0.0003828505517548667,
+      "loss": 1.642,
+      "step": 11300
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13471059501171112,
+      "learning_rate": 0.00038282062769911325,
+      "loss": 1.6657,
+      "step": 11310
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13584761321544647,
+      "learning_rate": 0.00038279067873041935,
+      "loss": 1.6755,
+      "step": 11320
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.14653141796588898,
+      "learning_rate": 0.00038276070485286607,
+      "loss": 1.6427,
+      "step": 11330
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1346314251422882,
+      "learning_rate": 0.00038273070607053805,
+      "loss": 1.6493,
+      "step": 11340
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1295168250799179,
+      "learning_rate": 0.00038270068238752315,
+      "loss": 1.6254,
+      "step": 11350
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.12681815028190613,
+      "learning_rate": 0.0003826706338079127,
+      "loss": 1.6393,
+      "step": 11360
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13591228425502777,
+      "learning_rate": 0.0003826405603358015,
+      "loss": 1.6267,
+      "step": 11370
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1302986741065979,
+      "learning_rate": 0.0003826104619752875,
+      "loss": 1.6913,
+      "step": 11380
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.14144158363342285,
+      "learning_rate": 0.0003825803387304723,
+      "loss": 1.6379,
+      "step": 11390
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16063153743743896,
+      "learning_rate": 0.00038255019060546083,
+      "loss": 1.5988,
+      "step": 11400
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1450476348400116,
+      "learning_rate": 0.00038252001760436124,
+      "loss": 1.6554,
+      "step": 11410
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.14311237633228302,
+      "learning_rate": 0.00038248981973128536,
+      "loss": 1.6349,
+      "step": 11420
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.12719784677028656,
+      "learning_rate": 0.0003824595969903481,
+      "loss": 1.6197,
+      "step": 11430
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13748832046985626,
+      "learning_rate": 0.000382429349385668,
+      "loss": 1.6987,
+      "step": 11440
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.14233890175819397,
+      "learning_rate": 0.00038239907692136685,
+      "loss": 1.6496,
+      "step": 11450
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1435275673866272,
+      "learning_rate": 0.00038236877960156994,
+      "loss": 1.6624,
+      "step": 11460
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1407058984041214,
+      "learning_rate": 0.00038233845743040575,
+      "loss": 1.6457,
+      "step": 11470
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1362675428390503,
+      "learning_rate": 0.00038230811041200646,
+      "loss": 1.6568,
+      "step": 11480
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13402876257896423,
+      "learning_rate": 0.0003822777385505073,
+      "loss": 1.6396,
+      "step": 11490
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.12936891615390778,
+      "learning_rate": 0.00038224734185004713,
+      "loss": 1.6775,
+      "step": 11500
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.14382876455783844,
+      "learning_rate": 0.0003822169203147681,
+      "loss": 1.6451,
+      "step": 11510
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.15172629058361053,
+      "learning_rate": 0.00038218647394881565,
+      "loss": 1.7103,
+      "step": 11520
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.132208451628685,
+      "learning_rate": 0.0003821560027563388,
+      "loss": 1.6221,
+      "step": 11530
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13271410763263702,
+      "learning_rate": 0.00038212550674148984,
+      "loss": 1.6939,
+      "step": 11540
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13946011662483215,
+      "learning_rate": 0.00038209498590842443,
+      "loss": 1.6767,
+      "step": 11550
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.12962143123149872,
+      "learning_rate": 0.00038206444026130164,
+      "loss": 1.6481,
+      "step": 11560
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.14416781067848206,
+      "learning_rate": 0.00038203386980428393,
+      "loss": 1.6877,
+      "step": 11570
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13405841588974,
+      "learning_rate": 0.0003820032745415372,
+      "loss": 1.6254,
+      "step": 11580
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.12885086238384247,
+      "learning_rate": 0.00038197265447723056,
+      "loss": 1.6485,
+      "step": 11590
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1339426189661026,
+      "learning_rate": 0.00038194200961553663,
+      "loss": 1.6197,
+      "step": 11600
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17396745085716248,
+      "learning_rate": 0.00038191133996063145,
+      "loss": 1.6687,
+      "step": 11610
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.14083170890808105,
+      "learning_rate": 0.0003818806455166942,
+      "loss": 1.6882,
+      "step": 11620
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.14531998336315155,
+      "learning_rate": 0.0003818499262879078,
+      "loss": 1.6539,
+      "step": 11630
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13853946328163147,
+      "learning_rate": 0.00038181918227845826,
+      "loss": 1.6537,
+      "step": 11640
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1355024129152298,
+      "learning_rate": 0.00038178841349253506,
+      "loss": 1.6609,
+      "step": 11650
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13852939009666443,
+      "learning_rate": 0.00038175761993433107,
+      "loss": 1.6367,
+      "step": 11660
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.21341007947921753,
+      "learning_rate": 0.00038172680160804257,
+      "loss": 1.6878,
+      "step": 11670
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1535138040781021,
+      "learning_rate": 0.000381695958517869,
+      "loss": 1.6729,
+      "step": 11680
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.12709595263004303,
+      "learning_rate": 0.00038166509066801357,
+      "loss": 1.6377,
+      "step": 11690
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.14343519508838654,
+      "learning_rate": 0.0003816341980626826,
+      "loss": 1.6043,
+      "step": 11700
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13254350423812866,
+      "learning_rate": 0.00038160328070608565,
+      "loss": 1.6784,
+      "step": 11710
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.12950347363948822,
+      "learning_rate": 0.00038157233860243596,
+      "loss": 1.6355,
+      "step": 11720
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14896464347839355,
+      "learning_rate": 0.00038154137175594997,
+      "loss": 1.6216,
+      "step": 11730
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.12893833220005035,
+      "learning_rate": 0.00038151038017084755,
+      "loss": 1.6339,
+      "step": 11740
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13772602379322052,
+      "learning_rate": 0.0003814793638513519,
+      "loss": 1.6551,
+      "step": 11750
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14746156334877014,
+      "learning_rate": 0.00038144832280168956,
+      "loss": 1.6918,
+      "step": 11760
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1271076649427414,
+      "learning_rate": 0.0003814172570260906,
+      "loss": 1.6094,
+      "step": 11770
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1246032565832138,
+      "learning_rate": 0.0003813861665287883,
+      "loss": 1.708,
+      "step": 11780
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13147948682308197,
+      "learning_rate": 0.00038135505131401933,
+      "loss": 1.6325,
+      "step": 11790
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.12232056260108948,
+      "learning_rate": 0.0003813239113860238,
+      "loss": 1.6363,
+      "step": 11800
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13298611342906952,
+      "learning_rate": 0.00038129274674904507,
+      "loss": 1.68,
+      "step": 11810
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.12976321578025818,
+      "learning_rate": 0.00038126155740733,
+      "loss": 1.638,
+      "step": 11820
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.12459643930196762,
+      "learning_rate": 0.00038123034336512874,
+      "loss": 1.6418,
+      "step": 11830
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1583673506975174,
+      "learning_rate": 0.0003811991046266949,
+      "loss": 1.7034,
+      "step": 11840
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14118176698684692,
+      "learning_rate": 0.0003811678411962852,
+      "loss": 1.677,
+      "step": 11850
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1380910873413086,
+      "learning_rate": 0.00038113655307816006,
+      "loss": 1.6716,
+      "step": 11860
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13520793616771698,
+      "learning_rate": 0.0003811052402765831,
+      "loss": 1.6578,
+      "step": 11870
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1493106186389923,
+      "learning_rate": 0.0003810739027958212,
+      "loss": 1.6334,
+      "step": 11880
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13702504336833954,
+      "learning_rate": 0.0003810425406401448,
+      "loss": 1.6805,
+      "step": 11890
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13339079916477203,
+      "learning_rate": 0.0003810111538138276,
+      "loss": 1.6433,
+      "step": 11900
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13484062254428864,
+      "learning_rate": 0.00038097974232114665,
+      "loss": 1.6363,
+      "step": 11910
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13990023732185364,
+      "learning_rate": 0.0003809483061663825,
+      "loss": 1.6858,
+      "step": 11920
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1360066533088684,
+      "learning_rate": 0.0003809168453538188,
+      "loss": 1.6551,
+      "step": 11930
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.15456506609916687,
+      "learning_rate": 0.0003808853598877428,
+      "loss": 1.6435,
+      "step": 11940
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1491072177886963,
+      "learning_rate": 0.00038085384977244496,
+      "loss": 1.6834,
+      "step": 11950
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14172150194644928,
+      "learning_rate": 0.0003808223150122192,
+      "loss": 1.734,
+      "step": 11960
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14219501614570618,
+      "learning_rate": 0.0003807907556113628,
+      "loss": 1.6582,
+      "step": 11970
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14249321818351746,
+      "learning_rate": 0.0003807591715741763,
+      "loss": 1.6797,
+      "step": 11980
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.12278520315885544,
+      "learning_rate": 0.0003807275629049636,
+      "loss": 1.6567,
+      "step": 11990
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.12179483473300934,
+      "learning_rate": 0.0003806959296080321,
+      "loss": 1.5853,
+      "step": 12000
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13621146976947784,
+      "learning_rate": 0.0003806642716876925,
+      "loss": 1.6556,
+      "step": 12010
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1444106101989746,
+      "learning_rate": 0.00038063258914825866,
+      "loss": 1.6513,
+      "step": 12020
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.12801142036914825,
+      "learning_rate": 0.0003806008819940481,
+      "loss": 1.6943,
+      "step": 12030
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1357094943523407,
+      "learning_rate": 0.0003805691502293815,
+      "loss": 1.6531,
+      "step": 12040
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.15934710204601288,
+      "learning_rate": 0.0003805373938585829,
+      "loss": 1.6565,
+      "step": 12050
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1452389806509018,
+      "learning_rate": 0.0003805056128859798,
+      "loss": 1.7022,
+      "step": 12060
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13581794500350952,
+      "learning_rate": 0.00038047380731590297,
+      "loss": 1.6122,
+      "step": 12070
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13787458837032318,
+      "learning_rate": 0.00038044197715268654,
+      "loss": 1.6503,
+      "step": 12080
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1335810273885727,
+      "learning_rate": 0.000380410122400668,
+      "loss": 1.6816,
+      "step": 12090
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1318478137254715,
+      "learning_rate": 0.00038037824306418825,
+      "loss": 1.6417,
+      "step": 12100
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.12821009755134583,
+      "learning_rate": 0.0003803463391475914,
+      "loss": 1.6117,
+      "step": 12110
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13755151629447937,
+      "learning_rate": 0.00038031441065522497,
+      "loss": 1.6591,
+      "step": 12120
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1282658874988556,
+      "learning_rate": 0.00038028245759144,
+      "loss": 1.6227,
+      "step": 12130
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14331896603107452,
+      "learning_rate": 0.00038025047996059053,
+      "loss": 1.6536,
+      "step": 12140
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16974402964115143,
+      "learning_rate": 0.00038021847776703436,
+      "loss": 1.6194,
+      "step": 12150
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14152204990386963,
+      "learning_rate": 0.0003801864510151322,
+      "loss": 1.6287,
+      "step": 12160
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14644131064414978,
+      "learning_rate": 0.00038015439970924856,
+      "loss": 1.6666,
+      "step": 12170
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.13402456045150757,
+      "learning_rate": 0.00038012232385375094,
+      "loss": 1.6162,
+      "step": 12180
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14258407056331635,
+      "learning_rate": 0.0003800902234530103,
+      "loss": 1.6782,
+      "step": 12190
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1407168209552765,
+      "learning_rate": 0.000380058098511401,
+      "loss": 1.6466,
+      "step": 12200
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16823522746562958,
+      "learning_rate": 0.00038002594903330066,
+      "loss": 1.6639,
+      "step": 12210
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.13747598230838776,
+      "learning_rate": 0.00037999377502309027,
+      "loss": 1.6836,
+      "step": 12220
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1358615607023239,
+      "learning_rate": 0.00037996157648515427,
+      "loss": 1.7415,
+      "step": 12230
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.134037584066391,
+      "learning_rate": 0.0003799293534238803,
+      "loss": 1.6901,
+      "step": 12240
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.13624466955661774,
+      "learning_rate": 0.0003798971058436594,
+      "loss": 1.6769,
+      "step": 12250
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.12982290983200073,
+      "learning_rate": 0.0003798648337488859,
+      "loss": 1.6238,
+      "step": 12260
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.12749430537223816,
+      "learning_rate": 0.00037983253714395755,
+      "loss": 1.622,
+      "step": 12270
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.13169650733470917,
+      "learning_rate": 0.0003798002160332754,
+      "loss": 1.5854,
+      "step": 12280
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1505732536315918,
+      "learning_rate": 0.0003797678704212438,
+      "loss": 1.6751,
+      "step": 12290
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1397910863161087,
+      "learning_rate": 0.00037973550031227064,
+      "loss": 1.6846,
+      "step": 12300
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14828398823738098,
+      "learning_rate": 0.00037970310571076675,
+      "loss": 1.7484,
+      "step": 12310
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14721670746803284,
+      "learning_rate": 0.00037967068662114674,
+      "loss": 1.6764,
+      "step": 12320
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1265481561422348,
+      "learning_rate": 0.0003796382430478283,
+      "loss": 1.6179,
+      "step": 12330
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.12669861316680908,
+      "learning_rate": 0.0003796057749952324,
+      "loss": 1.618,
+      "step": 12340
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.15905021131038666,
+      "learning_rate": 0.0003795732824677836,
+      "loss": 1.662,
+      "step": 12350
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.13711245357990265,
+      "learning_rate": 0.0003795407654699096,
+      "loss": 1.644,
+      "step": 12360
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14382456243038177,
+      "learning_rate": 0.00037950822400604144,
+      "loss": 1.6198,
+      "step": 12370
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.148932546377182,
+      "learning_rate": 0.00037947565808061365,
+      "loss": 1.643,
+      "step": 12380
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1323649138212204,
+      "learning_rate": 0.0003794430676980639,
+      "loss": 1.5892,
+      "step": 12390
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.13822904229164124,
+      "learning_rate": 0.0003794104528628332,
+      "loss": 1.6183,
+      "step": 12400
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.15499617159366608,
+      "learning_rate": 0.00037937781357936617,
+      "loss": 1.6209,
+      "step": 12410
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.134845569729805,
+      "learning_rate": 0.00037934514985211044,
+      "loss": 1.5826,
+      "step": 12420
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1421300321817398,
+      "learning_rate": 0.0003793124616855171,
+      "loss": 1.6302,
+      "step": 12430
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14278700947761536,
+      "learning_rate": 0.0003792797490840405,
+      "loss": 1.6141,
+      "step": 12440
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.135926753282547,
+      "learning_rate": 0.0003792470120521385,
+      "loss": 1.6718,
+      "step": 12450
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.13299962878227234,
+      "learning_rate": 0.00037921425059427217,
+      "loss": 1.716,
+      "step": 12460
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1322973519563675,
+      "learning_rate": 0.0003791814647149058,
+      "loss": 1.6853,
+      "step": 12470
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.13378873467445374,
+      "learning_rate": 0.0003791486544185072,
+      "loss": 1.5849,
+      "step": 12480
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14735238254070282,
+      "learning_rate": 0.0003791158197095474,
+      "loss": 1.6453,
+      "step": 12490
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.13957203924655914,
+      "learning_rate": 0.00037908296059250077,
+      "loss": 1.6303,
+      "step": 12500
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14032688736915588,
+      "learning_rate": 0.00037905007707184497,
+      "loss": 1.6073,
+      "step": 12510
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.12238329648971558,
+      "learning_rate": 0.00037901716915206117,
+      "loss": 1.6451,
+      "step": 12520
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1261938512325287,
+      "learning_rate": 0.0003789842368376336,
+      "loss": 1.5894,
+      "step": 12530
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1827414631843567,
+      "learning_rate": 0.00037895128013305,
+      "loss": 1.6792,
+      "step": 12540
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1656017303466797,
+      "learning_rate": 0.0003789182990428014,
+      "loss": 1.6189,
+      "step": 12550
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1515422910451889,
+      "learning_rate": 0.0003788852935713821,
+      "loss": 1.6907,
+      "step": 12560
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1227220669388771,
+      "learning_rate": 0.0003788522637232897,
+      "loss": 1.6704,
+      "step": 12570
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13533446192741394,
+      "learning_rate": 0.0003788192095030252,
+      "loss": 1.6518,
+      "step": 12580
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13504154980182648,
+      "learning_rate": 0.00037878613091509294,
+      "loss": 1.6018,
+      "step": 12590
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.14135916531085968,
+      "learning_rate": 0.0003787530279640005,
+      "loss": 1.6618,
+      "step": 12600
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13996636867523193,
+      "learning_rate": 0.0003787199006542588,
+      "loss": 1.6601,
+      "step": 12610
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.12917594611644745,
+      "learning_rate": 0.0003786867489903822,
+      "loss": 1.6333,
+      "step": 12620
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.136281356215477,
+      "learning_rate": 0.00037865357297688806,
+      "loss": 1.6466,
+      "step": 12630
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.12821024656295776,
+      "learning_rate": 0.00037862037261829747,
+      "loss": 1.6444,
+      "step": 12640
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13542772829532623,
+      "learning_rate": 0.00037858714791913457,
+      "loss": 1.6095,
+      "step": 12650
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13337557017803192,
+      "learning_rate": 0.0003785538988839268,
+      "loss": 1.6014,
+      "step": 12660
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1317620724439621,
+      "learning_rate": 0.0003785206255172052,
+      "loss": 1.6565,
+      "step": 12670
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1483888179063797,
+      "learning_rate": 0.0003784873278235037,
+      "loss": 1.6427,
+      "step": 12680
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.18616509437561035,
+      "learning_rate": 0.00037845400580736,
+      "loss": 1.6442,
+      "step": 12690
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.143222838640213,
+      "learning_rate": 0.0003784206594733147,
+      "loss": 1.6156,
+      "step": 12700
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.20240454375743866,
+      "learning_rate": 0.00037838728882591197,
+      "loss": 1.6448,
+      "step": 12710
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.14517773687839508,
+      "learning_rate": 0.0003783538938696991,
+      "loss": 1.7029,
+      "step": 12720
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13726265728473663,
+      "learning_rate": 0.0003783204746092271,
+      "loss": 1.6735,
+      "step": 12730
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13154472410678864,
+      "learning_rate": 0.00037828703104904977,
+      "loss": 1.6409,
+      "step": 12740
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.14166295528411865,
+      "learning_rate": 0.0003782535631937245,
+      "loss": 1.6306,
+      "step": 12750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.15384288132190704,
+      "learning_rate": 0.0003782200710478121,
+      "loss": 1.6308,
+      "step": 12760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13441230356693268,
+      "learning_rate": 0.0003781865546158763,
+      "loss": 1.626,
+      "step": 12770
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.14196231961250305,
+      "learning_rate": 0.0003781530139024846,
+      "loss": 1.6878,
+      "step": 12780
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1452767252922058,
+      "learning_rate": 0.0003781194489122074,
+      "loss": 1.6199,
+      "step": 12790
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.12843547761440277,
+      "learning_rate": 0.0003780858596496187,
+      "loss": 1.6524,
+      "step": 12800
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13439583778381348,
+      "learning_rate": 0.0003780522461192957,
+      "loss": 1.5977,
+      "step": 12810
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.12836812436580658,
+      "learning_rate": 0.00037801860832581896,
+      "loss": 1.6389,
+      "step": 12820
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.14791585505008698,
+      "learning_rate": 0.00037798494627377217,
+      "loss": 1.6415,
+      "step": 12830
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13145408034324646,
+      "learning_rate": 0.0003779512599677425,
+      "loss": 1.6438,
+      "step": 12840
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3971962630748749,
+      "learning_rate": 0.0003779175494123204,
+      "loss": 1.6774,
+      "step": 12850
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.16610465943813324,
+      "learning_rate": 0.0003778838146120997,
+      "loss": 1.7203,
+      "step": 12860
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13829699158668518,
+      "learning_rate": 0.00037785005557167725,
+      "loss": 1.6261,
+      "step": 12870
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13570912182331085,
+      "learning_rate": 0.0003778162722956535,
+      "loss": 1.619,
+      "step": 12880
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13839080929756165,
+      "learning_rate": 0.00037778246478863207,
+      "loss": 1.5813,
+      "step": 12890
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13922907412052155,
+      "learning_rate": 0.00037774863305521986,
+      "loss": 1.6481,
+      "step": 12900
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.14377102255821228,
+      "learning_rate": 0.0003777147771000272,
+      "loss": 1.682,
+      "step": 12910
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1406242698431015,
+      "learning_rate": 0.0003776808969276676,
+      "loss": 1.7073,
+      "step": 12920
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13428319990634918,
+      "learning_rate": 0.00037764699254275787,
+      "loss": 1.7087,
+      "step": 12930
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1269492506980896,
+      "learning_rate": 0.0003776130639499182,
+      "loss": 1.631,
+      "step": 12940
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1411420702934265,
+      "learning_rate": 0.00037757911115377207,
+      "loss": 1.6879,
+      "step": 12950
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13483257591724396,
+      "learning_rate": 0.0003775451341589461,
+      "loss": 1.6402,
+      "step": 12960
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.12121181190013885,
+      "learning_rate": 0.0003775111329700704,
+      "loss": 1.6591,
+      "step": 12970
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.13432370126247406,
+      "learning_rate": 0.00037747710759177835,
+      "loss": 1.6349,
+      "step": 12980
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1336212307214737,
+      "learning_rate": 0.00037744305802870655,
+      "loss": 1.6736,
+      "step": 12990
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1378161460161209,
+      "learning_rate": 0.00037740898428549494,
+      "loss": 1.6303,
+      "step": 13000
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.13662372529506683,
+      "learning_rate": 0.00037737488636678664,
+      "loss": 1.634,
+      "step": 13010
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.13213995099067688,
+      "learning_rate": 0.00037734076427722835,
+      "loss": 1.6469,
+      "step": 13020
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.13600227236747742,
+      "learning_rate": 0.00037730661802146977,
+      "loss": 1.6144,
+      "step": 13030
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.13998766243457794,
+      "learning_rate": 0.0003772724476041639,
+      "loss": 1.6659,
+      "step": 13040
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1359536051750183,
+      "learning_rate": 0.00037723825302996743,
+      "loss": 1.6886,
+      "step": 13050
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.15289464592933655,
+      "learning_rate": 0.0003772040343035398,
+      "loss": 1.6488,
+      "step": 13060
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.14165322482585907,
+      "learning_rate": 0.0003771697914295441,
+      "loss": 1.6394,
+      "step": 13070
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.13506761193275452,
+      "learning_rate": 0.00037713552441264657,
+      "loss": 1.6581,
+      "step": 13080
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1323695331811905,
+      "learning_rate": 0.00037710123325751673,
+      "loss": 1.7184,
+      "step": 13090
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.12729106843471527,
+      "learning_rate": 0.0003770669179688275,
+      "loss": 1.6154,
+      "step": 13100
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.14182405173778534,
+      "learning_rate": 0.000377032578551255,
+      "loss": 1.6521,
+      "step": 13110
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.142019122838974,
+      "learning_rate": 0.0003769982150094786,
+      "loss": 1.609,
+      "step": 13120
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.13171623647212982,
+      "learning_rate": 0.00037696382734818114,
+      "loss": 1.6386,
+      "step": 13130
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.14993920922279358,
+      "learning_rate": 0.00037692941557204853,
+      "loss": 1.586,
+      "step": 13140
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.12679728865623474,
+      "learning_rate": 0.0003768949796857701,
+      "loss": 1.6606,
+      "step": 13150
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.13762407004833221,
+      "learning_rate": 0.0003768605196940383,
+      "loss": 1.7056,
+      "step": 13160
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.17662636935710907,
+      "learning_rate": 0.0003768260356015492,
+      "loss": 1.6784,
+      "step": 13170
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.12629279494285583,
+      "learning_rate": 0.0003767915274130018,
+      "loss": 1.647,
+      "step": 13180
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1376374363899231,
+      "learning_rate": 0.00037675699513309855,
+      "loss": 1.6823,
+      "step": 13190
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1502215564250946,
+      "learning_rate": 0.00037672243876654514,
+      "loss": 1.6928,
+      "step": 13200
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.15364877879619598,
+      "learning_rate": 0.0003766878583180506,
+      "loss": 1.6604,
+      "step": 13210
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16463319957256317,
+      "learning_rate": 0.0003766532537923272,
+      "loss": 1.6847,
+      "step": 13220
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.13310016691684723,
+      "learning_rate": 0.0003766186251940904,
+      "loss": 1.6721,
+      "step": 13230
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1365755796432495,
+      "learning_rate": 0.00037658397252805917,
+      "loss": 1.6736,
+      "step": 13240
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.12797911465168,
+      "learning_rate": 0.00037654929579895564,
+      "loss": 1.6636,
+      "step": 13250
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1630721390247345,
+      "learning_rate": 0.000376514595011505,
+      "loss": 1.6586,
+      "step": 13260
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.13059785962104797,
+      "learning_rate": 0.0003764798701704362,
+      "loss": 1.5854,
+      "step": 13270
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.12410352379083633,
+      "learning_rate": 0.000376445121280481,
+      "loss": 1.6473,
+      "step": 13280
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.15035715699195862,
+      "learning_rate": 0.0003764103483463746,
+      "loss": 1.6485,
+      "step": 13290
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.13317257165908813,
+      "learning_rate": 0.0003763755513728556,
+      "loss": 1.5817,
+      "step": 13300
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16770432889461517,
+      "learning_rate": 0.0003763407303646658,
+      "loss": 1.6085,
+      "step": 13310
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.12674319744110107,
+      "learning_rate": 0.0003763058853265501,
+      "loss": 1.6021,
+      "step": 13320
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.12082844972610474,
+      "learning_rate": 0.000376271016263257,
+      "loss": 1.6156,
+      "step": 13330
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.15674911439418793,
+      "learning_rate": 0.0003762361231795381,
+      "loss": 1.6758,
+      "step": 13340
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1440170854330063,
+      "learning_rate": 0.00037620120608014806,
+      "loss": 1.5488,
+      "step": 13350
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.13002000749111176,
+      "learning_rate": 0.00037616626496984526,
+      "loss": 1.5897,
+      "step": 13360
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.12813784182071686,
+      "learning_rate": 0.00037613129985339107,
+      "loss": 1.5834,
+      "step": 13370
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.121329165995121,
+      "learning_rate": 0.0003760963107355501,
+      "loss": 1.5628,
+      "step": 13380
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.12972715497016907,
+      "learning_rate": 0.0003760612976210904,
+      "loss": 1.6213,
+      "step": 13390
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1307857781648636,
+      "learning_rate": 0.00037602626051478315,
+      "loss": 1.603,
+      "step": 13400
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1268593817949295,
+      "learning_rate": 0.00037599119942140284,
+      "loss": 1.6403,
+      "step": 13410
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12571527063846588,
+      "learning_rate": 0.00037595611434572727,
+      "loss": 1.5791,
+      "step": 13420
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.13057558238506317,
+      "learning_rate": 0.0003759210052925375,
+      "loss": 1.6253,
+      "step": 13430
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.15414628386497498,
+      "learning_rate": 0.0003758858722666178,
+      "loss": 1.6588,
+      "step": 13440
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12613502144813538,
+      "learning_rate": 0.00037585071527275574,
+      "loss": 1.6017,
+      "step": 13450
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.13810329139232635,
+      "learning_rate": 0.00037581553431574214,
+      "loss": 1.6473,
+      "step": 13460
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12223131209611893,
+      "learning_rate": 0.00037578032940037117,
+      "loss": 1.5951,
+      "step": 13470
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1235814169049263,
+      "learning_rate": 0.0003757451005314402,
+      "loss": 1.6169,
+      "step": 13480
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.20623689889907837,
+      "learning_rate": 0.0003757098477137498,
+      "loss": 1.6436,
+      "step": 13490
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.144382506608963,
+      "learning_rate": 0.0003756745709521038,
+      "loss": 1.6425,
+      "step": 13500
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.13835234940052032,
+      "learning_rate": 0.00037563927025130954,
+      "loss": 1.6612,
+      "step": 13510
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.13487409055233002,
+      "learning_rate": 0.00037560394561617734,
+      "loss": 1.7046,
+      "step": 13520
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12433511763811111,
+      "learning_rate": 0.0003755685970515209,
+      "loss": 1.6285,
+      "step": 13530
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12851616740226746,
+      "learning_rate": 0.0003755332245621571,
+      "loss": 1.5926,
+      "step": 13540
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1232232078909874,
+      "learning_rate": 0.0003754978281529063,
+      "loss": 1.6012,
+      "step": 13550
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14085985720157623,
+      "learning_rate": 0.0003754624078285918,
+      "loss": 1.6062,
+      "step": 13560
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12836585938930511,
+      "learning_rate": 0.00037542696359404044,
+      "loss": 1.681,
+      "step": 13570
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1346902847290039,
+      "learning_rate": 0.0003753914954540821,
+      "loss": 1.6875,
+      "step": 13580
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.13850773870944977,
+      "learning_rate": 0.00037535600341355007,
+      "loss": 1.6487,
+      "step": 13590
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.13894310593605042,
+      "learning_rate": 0.0003753204874772808,
+      "loss": 1.6004,
+      "step": 13600
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12756860256195068,
+      "learning_rate": 0.00037528494765011423,
+      "loss": 1.6055,
+      "step": 13610
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12638744711875916,
+      "learning_rate": 0.00037524938393689315,
+      "loss": 1.6189,
+      "step": 13620
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.13009041547775269,
+      "learning_rate": 0.0003752137963424638,
+      "loss": 1.601,
+      "step": 13630
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1364549696445465,
+      "learning_rate": 0.00037517818487167585,
+      "loss": 1.6194,
+      "step": 13640
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12034954130649567,
+      "learning_rate": 0.00037514254952938204,
+      "loss": 1.6218,
+      "step": 13650
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14631476998329163,
+      "learning_rate": 0.00037510689032043836,
+      "loss": 1.6338,
+      "step": 13660
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.13141344487667084,
+      "learning_rate": 0.00037507120724970404,
+      "loss": 1.6142,
+      "step": 13670
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14006882905960083,
+      "learning_rate": 0.00037503550032204173,
+      "loss": 1.6742,
+      "step": 13680
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.13324075937271118,
+      "learning_rate": 0.0003749997695423171,
+      "loss": 1.6363,
+      "step": 13690
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12037849426269531,
+      "learning_rate": 0.0003749640149153992,
+      "loss": 1.6478,
+      "step": 13700
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1296912431716919,
+      "learning_rate": 0.00037492823644616035,
+      "loss": 1.5908,
+      "step": 13710
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14283686876296997,
+      "learning_rate": 0.000374892434139476,
+      "loss": 1.6915,
+      "step": 13720
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.133306086063385,
+      "learning_rate": 0.000374856608000225,
+      "loss": 1.6557,
+      "step": 13730
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14094007015228271,
+      "learning_rate": 0.0003748207580332893,
+      "loss": 1.6292,
+      "step": 13740
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1342521756887436,
+      "learning_rate": 0.0003747848842435542,
+      "loss": 1.6288,
+      "step": 13750
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1393992304801941,
+      "learning_rate": 0.00037474898663590833,
+      "loss": 1.6324,
+      "step": 13760
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.13209590315818787,
+      "learning_rate": 0.0003747130652152433,
+      "loss": 1.6145,
+      "step": 13770
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2357804775238037,
+      "learning_rate": 0.00037467711998645417,
+      "loss": 1.6578,
+      "step": 13780
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14991991221904755,
+      "learning_rate": 0.0003746411509544392,
+      "loss": 1.675,
+      "step": 13790
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1279059201478958,
+      "learning_rate": 0.00037460515812409983,
+      "loss": 1.6294,
+      "step": 13800
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12366055697202682,
+      "learning_rate": 0.00037456914150034085,
+      "loss": 1.5961,
+      "step": 13810
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1304793655872345,
+      "learning_rate": 0.0003745331010880703,
+      "loss": 1.6246,
+      "step": 13820
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1325242668390274,
+      "learning_rate": 0.00037449703689219926,
+      "loss": 1.6157,
+      "step": 13830
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.15635986626148224,
+      "learning_rate": 0.0003744609489176423,
+      "loss": 1.6555,
+      "step": 13840
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.15102840960025787,
+      "learning_rate": 0.00037442483716931704,
+      "loss": 1.6652,
+      "step": 13850
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1392315775156021,
+      "learning_rate": 0.0003743887016521445,
+      "loss": 1.6667,
+      "step": 13860
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.13750429451465607,
+      "learning_rate": 0.0003743525423710489,
+      "loss": 1.68,
+      "step": 13870
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.13358047604560852,
+      "learning_rate": 0.0003743163593309575,
+      "loss": 1.6177,
+      "step": 13880
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1445418745279312,
+      "learning_rate": 0.00037428015253680117,
+      "loss": 1.6581,
+      "step": 13890
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.12553058564662933,
+      "learning_rate": 0.0003742439219935135,
+      "loss": 1.6276,
+      "step": 13900
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1535179615020752,
+      "learning_rate": 0.000374207667706032,
+      "loss": 1.615,
+      "step": 13910
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1400972157716751,
+      "learning_rate": 0.0003741713896792967,
+      "loss": 1.5924,
+      "step": 13920
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.12700070440769196,
+      "learning_rate": 0.00037413508791825145,
+      "loss": 1.614,
+      "step": 13930
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.12791931629180908,
+      "learning_rate": 0.00037409876242784293,
+      "loss": 1.6329,
+      "step": 13940
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1336628645658493,
+      "learning_rate": 0.00037406241321302124,
+      "loss": 1.6127,
+      "step": 13950
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.14038360118865967,
+      "learning_rate": 0.0003740260402787398,
+      "loss": 1.6381,
+      "step": 13960
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.13828876614570618,
+      "learning_rate": 0.0003739896436299549,
+      "loss": 1.6303,
+      "step": 13970
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.11978207528591156,
+      "learning_rate": 0.00037395322327162656,
+      "loss": 1.5761,
+      "step": 13980
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.12977859377861023,
+      "learning_rate": 0.0003739167792087176,
+      "loss": 1.6058,
+      "step": 13990
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1237422525882721,
+      "learning_rate": 0.00037388031144619435,
+      "loss": 1.6475,
+      "step": 14000
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1510535180568695,
+      "learning_rate": 0.00037384381998902623,
+      "loss": 1.6642,
+      "step": 14010
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.14058199524879456,
+      "learning_rate": 0.000373807304842186,
+      "loss": 1.6568,
+      "step": 14020
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.14129306375980377,
+      "learning_rate": 0.00037377076601064933,
+      "loss": 1.6617,
+      "step": 14030
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.12125241756439209,
+      "learning_rate": 0.00037373420349939574,
+      "loss": 1.606,
+      "step": 14040
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1290244609117508,
+      "learning_rate": 0.00037369761731340724,
+      "loss": 1.6396,
+      "step": 14050
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.12763898074626923,
+      "learning_rate": 0.00037366100745766964,
+      "loss": 1.5939,
+      "step": 14060
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.13529957830905914,
+      "learning_rate": 0.00037362437393717167,
+      "loss": 1.5901,
+      "step": 14070
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.11993245035409927,
+      "learning_rate": 0.00037358771675690544,
+      "loss": 1.6631,
+      "step": 14080
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.12410163134336472,
+      "learning_rate": 0.0003735510359218662,
+      "loss": 1.6279,
+      "step": 14090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.13325445353984833,
+      "learning_rate": 0.00037351433143705237,
+      "loss": 1.5623,
+      "step": 14100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.13742899894714355,
+      "learning_rate": 0.0003734776033074657,
+      "loss": 1.5947,
+      "step": 14110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1345348060131073,
+      "learning_rate": 0.00037344085153811113,
+      "loss": 1.6241,
+      "step": 14120
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.13709436357021332,
+      "learning_rate": 0.00037340407613399687,
+      "loss": 1.6487,
+      "step": 14130
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.14037799835205078,
+      "learning_rate": 0.0003733672771001343,
+      "loss": 1.6424,
+      "step": 14140
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.14871887862682343,
+      "learning_rate": 0.0003733304544415379,
+      "loss": 1.5902,
+      "step": 14150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1745438277721405,
+      "learning_rate": 0.00037329360816322564,
+      "loss": 1.6594,
+      "step": 14160
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.13017408549785614,
+      "learning_rate": 0.00037325673827021843,
+      "loss": 1.6437,
+      "step": 14170
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1599813997745514,
+      "learning_rate": 0.00037321984476754063,
+      "loss": 1.6244,
+      "step": 14180
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.14345701038837433,
+      "learning_rate": 0.0003731829276602196,
+      "loss": 1.629,
+      "step": 14190
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.13529981672763824,
+      "learning_rate": 0.00037314598695328615,
+      "loss": 1.5916,
+      "step": 14200
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.12083511799573898,
+      "learning_rate": 0.00037310902265177406,
+      "loss": 1.6394,
+      "step": 14210
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.18949878215789795,
+      "learning_rate": 0.0003730720347607206,
+      "loss": 1.604,
+      "step": 14220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.130552276968956,
+      "learning_rate": 0.00037303502328516597,
+      "loss": 1.6337,
+      "step": 14230
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.13656294345855713,
+      "learning_rate": 0.0003729979882301538,
+      "loss": 1.6348,
+      "step": 14240
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.13620498776435852,
+      "learning_rate": 0.0003729609296007308,
+      "loss": 1.6602,
+      "step": 14250
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.15204866230487823,
+      "learning_rate": 0.00037292384740194695,
+      "loss": 1.6101,
+      "step": 14260
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13973408937454224,
+      "learning_rate": 0.0003728867416388555,
+      "loss": 1.6176,
+      "step": 14270
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.14117884635925293,
+      "learning_rate": 0.0003728496123165127,
+      "loss": 1.6141,
+      "step": 14280
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13209649920463562,
+      "learning_rate": 0.00037281245943997834,
+      "loss": 1.6781,
+      "step": 14290
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13493242859840393,
+      "learning_rate": 0.00037277528301431516,
+      "loss": 1.6936,
+      "step": 14300
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1251293271780014,
+      "learning_rate": 0.0003727380830445891,
+      "loss": 1.6099,
+      "step": 14310
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.12920813262462616,
+      "learning_rate": 0.00037270085953586957,
+      "loss": 1.5764,
+      "step": 14320
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.11848762631416321,
+      "learning_rate": 0.00037266361249322884,
+      "loss": 1.5963,
+      "step": 14330
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.14428497850894928,
+      "learning_rate": 0.00037262634192174263,
+      "loss": 1.6077,
+      "step": 14340
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1292484998703003,
+      "learning_rate": 0.00037258904782648983,
+      "loss": 1.582,
+      "step": 14350
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13617852330207825,
+      "learning_rate": 0.00037255173021255247,
+      "loss": 1.6491,
+      "step": 14360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1400119662284851,
+      "learning_rate": 0.00037251438908501584,
+      "loss": 1.6375,
+      "step": 14370
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.22918426990509033,
+      "learning_rate": 0.0003724770244489683,
+      "loss": 1.6219,
+      "step": 14380
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13372540473937988,
+      "learning_rate": 0.0003724396363095017,
+      "loss": 1.6306,
+      "step": 14390
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1299789845943451,
+      "learning_rate": 0.0003724022246717108,
+      "loss": 1.6331,
+      "step": 14400
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1293337345123291,
+      "learning_rate": 0.0003723647895406937,
+      "loss": 1.5746,
+      "step": 14410
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13581116497516632,
+      "learning_rate": 0.00037232733092155165,
+      "loss": 1.6375,
+      "step": 14420
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1325233429670334,
+      "learning_rate": 0.0003722898488193892,
+      "loss": 1.576,
+      "step": 14430
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13030831515789032,
+      "learning_rate": 0.000372252343239314,
+      "loss": 1.6179,
+      "step": 14440
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13602511584758759,
+      "learning_rate": 0.0003722148141864369,
+      "loss": 1.6261,
+      "step": 14450
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16156691312789917,
+      "learning_rate": 0.00037217726166587203,
+      "loss": 1.561,
+      "step": 14460
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13414260745048523,
+      "learning_rate": 0.0003721396856827366,
+      "loss": 1.6126,
+      "step": 14470
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1478995531797409,
+      "learning_rate": 0.0003721020862421512,
+      "loss": 1.6407,
+      "step": 14480
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16678500175476074,
+      "learning_rate": 0.00037206446334923937,
+      "loss": 1.5854,
+      "step": 14490
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13141700625419617,
+      "learning_rate": 0.00037202681700912807,
+      "loss": 1.5906,
+      "step": 14500
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13795627653598785,
+      "learning_rate": 0.00037198914722694725,
+      "loss": 1.555,
+      "step": 14510
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1269928514957428,
+      "learning_rate": 0.0003719514540078303,
+      "loss": 1.613,
+      "step": 14520
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1316051185131073,
+      "learning_rate": 0.00037191373735691357,
+      "loss": 1.6278,
+      "step": 14530
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13085073232650757,
+      "learning_rate": 0.00037187599727933683,
+      "loss": 1.5888,
+      "step": 14540
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.12712758779525757,
+      "learning_rate": 0.0003718382337802427,
+      "loss": 1.6208,
+      "step": 14550
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13212016224861145,
+      "learning_rate": 0.00037180044686477737,
+      "loss": 1.6117,
+      "step": 14560
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1601962149143219,
+      "learning_rate": 0.00037176263653809007,
+      "loss": 1.6429,
+      "step": 14570
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13773435354232788,
+      "learning_rate": 0.0003717248028053331,
+      "loss": 1.6246,
+      "step": 14580
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1217036321759224,
+      "learning_rate": 0.0003716869456716621,
+      "loss": 1.57,
+      "step": 14590
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1296040564775467,
+      "learning_rate": 0.00037164906514223594,
+      "loss": 1.7054,
+      "step": 14600
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.12574143707752228,
+      "learning_rate": 0.00037161116122221656,
+      "loss": 1.632,
+      "step": 14610
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.14457228779792786,
+      "learning_rate": 0.00037157323391676897,
+      "loss": 1.6575,
+      "step": 14620
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13964404165744781,
+      "learning_rate": 0.00037153528323106164,
+      "loss": 1.6874,
+      "step": 14630
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13058677315711975,
+      "learning_rate": 0.00037149730917026614,
+      "loss": 1.5947,
+      "step": 14640
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.14184308052062988,
+      "learning_rate": 0.00037145931173955716,
+      "loss": 1.6341,
+      "step": 14650
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.14103791117668152,
+      "learning_rate": 0.0003714212909441126,
+      "loss": 1.6366,
+      "step": 14660
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1346432864665985,
+      "learning_rate": 0.0003713832467891135,
+      "loss": 1.6448,
+      "step": 14670
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13459618389606476,
+      "learning_rate": 0.00037134517927974424,
+      "loss": 1.6545,
+      "step": 14680
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.12230374664068222,
+      "learning_rate": 0.0003713070884211921,
+      "loss": 1.5801,
+      "step": 14690
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.12996219098567963,
+      "learning_rate": 0.00037126897421864793,
+      "loss": 1.6195,
+      "step": 14700
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1345432847738266,
+      "learning_rate": 0.0003712308366773054,
+      "loss": 1.6342,
+      "step": 14710
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.13267439603805542,
+      "learning_rate": 0.0003711926758023615,
+      "loss": 1.5958,
+      "step": 14720
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.14793846011161804,
+      "learning_rate": 0.00037115449159901654,
+      "loss": 1.5989,
+      "step": 14730
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.14753471314907074,
+      "learning_rate": 0.0003711162840724737,
+      "loss": 1.5798,
+      "step": 14740
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.13883794844150543,
+      "learning_rate": 0.00037107805322793963,
+      "loss": 1.5742,
+      "step": 14750
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.14023089408874512,
+      "learning_rate": 0.00037103979907062405,
+      "loss": 1.6243,
+      "step": 14760
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1318616420030594,
+      "learning_rate": 0.0003710015216057398,
+      "loss": 1.6287,
+      "step": 14770
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1379183828830719,
+      "learning_rate": 0.00037096322083850295,
+      "loss": 1.5711,
+      "step": 14780
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.14544400572776794,
+      "learning_rate": 0.0003709248967741328,
+      "loss": 1.6764,
+      "step": 14790
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.14848382771015167,
+      "learning_rate": 0.0003708865494178516,
+      "loss": 1.6643,
+      "step": 14800
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1400999277830124,
+      "learning_rate": 0.00037084817877488515,
+      "loss": 1.5818,
+      "step": 14810
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.14183419942855835,
+      "learning_rate": 0.0003708097848504621,
+      "loss": 1.6044,
+      "step": 14820
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.13940398395061493,
+      "learning_rate": 0.0003707713676498144,
+      "loss": 1.5932,
+      "step": 14830
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.16999736428260803,
+      "learning_rate": 0.0003707329271781771,
+      "loss": 1.5748,
+      "step": 14840
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.15817441046237946,
+      "learning_rate": 0.00037069446344078854,
+      "loss": 1.6793,
+      "step": 14850
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.14373977482318878,
+      "learning_rate": 0.0003706559764428902,
+      "loss": 1.6377,
+      "step": 14860
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1450965404510498,
+      "learning_rate": 0.0003706174661897267,
+      "loss": 1.6379,
+      "step": 14870
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1337869018316269,
+      "learning_rate": 0.0003705789326865457,
+      "loss": 1.5725,
+      "step": 14880
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.12434647232294083,
+      "learning_rate": 0.0003705403759385982,
+      "loss": 1.6538,
+      "step": 14890
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1302241086959839,
+      "learning_rate": 0.0003705017959511385,
+      "loss": 1.6041,
+      "step": 14900
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.12320354580879211,
+      "learning_rate": 0.0003704631927294237,
+      "loss": 1.6227,
+      "step": 14910
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.13008472323417664,
+      "learning_rate": 0.0003704245662787143,
+      "loss": 1.6397,
+      "step": 14920
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.12906397879123688,
+      "learning_rate": 0.00037038591660427386,
+      "loss": 1.6242,
+      "step": 14930
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.14738160371780396,
+      "learning_rate": 0.0003703472437113693,
+      "loss": 1.62,
+      "step": 14940
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.12653198838233948,
+      "learning_rate": 0.00037030854760527045,
+      "loss": 1.6273,
+      "step": 14950
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.12363409996032715,
+      "learning_rate": 0.0003702698282912506,
+      "loss": 1.604,
+      "step": 14960
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1385618895292282,
+      "learning_rate": 0.00037023108577458573,
+      "loss": 1.6037,
+      "step": 14970
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1327894628047943,
+      "learning_rate": 0.00037019232006055557,
+      "loss": 1.6604,
+      "step": 14980
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11730336397886276,
+      "learning_rate": 0.0003701535311544426,
+      "loss": 1.6179,
+      "step": 14990
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.13437233865261078,
+      "learning_rate": 0.0003701147190615325,
+      "loss": 1.6073,
+      "step": 15000
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.13813023269176483,
+      "learning_rate": 0.0003700758837871143,
+      "loss": 1.5743,
+      "step": 15010
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1432405263185501,
+      "learning_rate": 0.00037003702533647997,
+      "loss": 1.5793,
+      "step": 15020
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.15601477026939392,
+      "learning_rate": 0.00036999814371492496,
+      "loss": 1.6561,
+      "step": 15030
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.13550876080989838,
+      "learning_rate": 0.0003699592389277474,
+      "loss": 1.6301,
+      "step": 15040
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1494302898645401,
+      "learning_rate": 0.000369920310980249,
+      "loss": 1.6244,
+      "step": 15050
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1322883516550064,
+      "learning_rate": 0.0003698813598777344,
+      "loss": 1.6139,
+      "step": 15060
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1428704857826233,
+      "learning_rate": 0.00036984238562551156,
+      "loss": 1.6053,
+      "step": 15070
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.13099798560142517,
+      "learning_rate": 0.00036980338822889135,
+      "loss": 1.5871,
+      "step": 15080
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.12746968865394592,
+      "learning_rate": 0.000369764367693188,
+      "loss": 1.6247,
+      "step": 15090
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1304878443479538,
+      "learning_rate": 0.0003697253240237189,
+      "loss": 1.635,
+      "step": 15100
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.13392950594425201,
+      "learning_rate": 0.0003696862572258044,
+      "loss": 1.5906,
+      "step": 15110
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1288476288318634,
+      "learning_rate": 0.00036964716730476826,
+      "loss": 1.5855,
+      "step": 15120
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1258256584405899,
+      "learning_rate": 0.0003696080542659372,
+      "loss": 1.5825,
+      "step": 15130
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.14086346328258514,
+      "learning_rate": 0.0003695689181146411,
+      "loss": 1.6258,
+      "step": 15140
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.13240265846252441,
+      "learning_rate": 0.00036952975885621306,
+      "loss": 1.6109,
+      "step": 15150
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12934376299381256,
+      "learning_rate": 0.00036949057649598933,
+      "loss": 1.6495,
+      "step": 15160
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1496889889240265,
+      "learning_rate": 0.00036945137103930933,
+      "loss": 1.6093,
+      "step": 15170
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.14470835030078888,
+      "learning_rate": 0.0003694121424915155,
+      "loss": 1.5829,
+      "step": 15180
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.13254313170909882,
+      "learning_rate": 0.00036937289085795357,
+      "loss": 1.577,
+      "step": 15190
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.13890144228935242,
+      "learning_rate": 0.0003693336161439723,
+      "loss": 1.5437,
+      "step": 15200
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1294306218624115,
+      "learning_rate": 0.00036929431835492363,
+      "loss": 1.5835,
+      "step": 15210
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.13542836904525757,
+      "learning_rate": 0.0003692549974961627,
+      "loss": 1.6302,
+      "step": 15220
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1277715265750885,
+      "learning_rate": 0.00036921565357304784,
+      "loss": 1.5678,
+      "step": 15230
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12343065440654755,
+      "learning_rate": 0.00036917628659094036,
+      "loss": 1.6407,
+      "step": 15240
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.15068763494491577,
+      "learning_rate": 0.00036913689655520475,
+      "loss": 1.6359,
+      "step": 15250
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1351775974035263,
+      "learning_rate": 0.0003690974834712088,
+      "loss": 1.6509,
+      "step": 15260
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12874259054660797,
+      "learning_rate": 0.0003690580473443232,
+      "loss": 1.626,
+      "step": 15270
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1362953782081604,
+      "learning_rate": 0.000369018588179922,
+      "loss": 1.6038,
+      "step": 15280
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.13399767875671387,
+      "learning_rate": 0.0003689791059833822,
+      "loss": 1.643,
+      "step": 15290
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1279444843530655,
+      "learning_rate": 0.0003689396007600842,
+      "loss": 1.6213,
+      "step": 15300
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.15049542486667633,
+      "learning_rate": 0.0003689000725154112,
+      "loss": 1.6267,
+      "step": 15310
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12989181280136108,
+      "learning_rate": 0.00036886052125474975,
+      "loss": 1.6071,
+      "step": 15320
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12253989279270172,
+      "learning_rate": 0.0003688209469834895,
+      "loss": 1.5614,
+      "step": 15330
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.19118964672088623,
+      "learning_rate": 0.0003687813497070233,
+      "loss": 1.6439,
+      "step": 15340
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12991493940353394,
+      "learning_rate": 0.000368741729430747,
+      "loss": 1.6379,
+      "step": 15350
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1368943452835083,
+      "learning_rate": 0.00036870208616005966,
+      "loss": 1.6876,
+      "step": 15360
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1273137629032135,
+      "learning_rate": 0.0003686624199003635,
+      "loss": 1.6021,
+      "step": 15370
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1327163577079773,
+      "learning_rate": 0.00036862273065706375,
+      "loss": 1.584,
+      "step": 15380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1375041902065277,
+      "learning_rate": 0.00036858301843556895,
+      "loss": 1.5925,
+      "step": 15390
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12788818776607513,
+      "learning_rate": 0.00036854328324129054,
+      "loss": 1.662,
+      "step": 15400
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1442573368549347,
+      "learning_rate": 0.00036850352507964345,
+      "loss": 1.6183,
+      "step": 15410
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.13934104144573212,
+      "learning_rate": 0.0003684637439560453,
+      "loss": 1.6704,
+      "step": 15420
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12524129450321198,
+      "learning_rate": 0.0003684239398759172,
+      "loss": 1.5987,
+      "step": 15430
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12950356304645538,
+      "learning_rate": 0.00036838411284468317,
+      "loss": 1.5963,
+      "step": 15440
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12582528591156006,
+      "learning_rate": 0.00036834426286777046,
+      "loss": 1.628,
+      "step": 15450
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.13867102563381195,
+      "learning_rate": 0.0003683043899506094,
+      "loss": 1.6975,
+      "step": 15460
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.14964598417282104,
+      "learning_rate": 0.00036826449409863357,
+      "loss": 1.6353,
+      "step": 15470
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12712357938289642,
+      "learning_rate": 0.00036822457531727944,
+      "loss": 1.6047,
+      "step": 15480
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.14426526427268982,
+      "learning_rate": 0.0003681846336119868,
+      "loss": 1.63,
+      "step": 15490
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.13479648530483246,
+      "learning_rate": 0.0003681446689881985,
+      "loss": 1.6556,
+      "step": 15500
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.13704581558704376,
+      "learning_rate": 0.00036810468145136047,
+      "loss": 1.6075,
+      "step": 15510
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1371973603963852,
+      "learning_rate": 0.0003680646710069218,
+      "loss": 1.6082,
+      "step": 15520
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1427917182445526,
+      "learning_rate": 0.0003680246376603348,
+      "loss": 1.5428,
+      "step": 15530
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1293700784444809,
+      "learning_rate": 0.00036798458141705477,
+      "loss": 1.6045,
+      "step": 15540
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.12121811509132385,
+      "learning_rate": 0.0003679445022825401,
+      "loss": 1.5962,
+      "step": 15550
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13424797356128693,
+      "learning_rate": 0.00036790440026225244,
+      "loss": 1.5898,
+      "step": 15560
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.134863018989563,
+      "learning_rate": 0.00036786427536165646,
+      "loss": 1.618,
+      "step": 15570
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.14139671623706818,
+      "learning_rate": 0.00036782412758621996,
+      "loss": 1.5694,
+      "step": 15580
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13203677535057068,
+      "learning_rate": 0.00036778395694141397,
+      "loss": 1.6044,
+      "step": 15590
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13818727433681488,
+      "learning_rate": 0.00036774376343271236,
+      "loss": 1.6111,
+      "step": 15600
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1444254368543625,
+      "learning_rate": 0.00036770354706559247,
+      "loss": 1.5719,
+      "step": 15610
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13501466810703278,
+      "learning_rate": 0.00036766330784553444,
+      "loss": 1.6396,
+      "step": 15620
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13209888339042664,
+      "learning_rate": 0.00036762304577802175,
+      "loss": 1.6502,
+      "step": 15630
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13570676743984222,
+      "learning_rate": 0.0003675827608685409,
+      "loss": 1.5912,
+      "step": 15640
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.14930690824985504,
+      "learning_rate": 0.0003675424531225815,
+      "loss": 1.6024,
+      "step": 15650
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1256052553653717,
+      "learning_rate": 0.00036750212254563626,
+      "loss": 1.6353,
+      "step": 15660
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.12670467793941498,
+      "learning_rate": 0.00036746176914320103,
+      "loss": 1.5602,
+      "step": 15670
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1287185400724411,
+      "learning_rate": 0.00036742139292077474,
+      "loss": 1.5914,
+      "step": 15680
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.12172731757164001,
+      "learning_rate": 0.00036738099388385943,
+      "loss": 1.6048,
+      "step": 15690
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13237637281417847,
+      "learning_rate": 0.0003673405720379604,
+      "loss": 1.5982,
+      "step": 15700
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.14493237435817719,
+      "learning_rate": 0.00036730012738858583,
+      "loss": 1.6641,
+      "step": 15710
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.12684105336666107,
+      "learning_rate": 0.00036725965994124715,
+      "loss": 1.6057,
+      "step": 15720
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13709315657615662,
+      "learning_rate": 0.00036721916970145873,
+      "loss": 1.6742,
+      "step": 15730
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.21690481901168823,
+      "learning_rate": 0.00036717865667473836,
+      "loss": 1.6453,
+      "step": 15740
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.14305225014686584,
+      "learning_rate": 0.00036713812086660663,
+      "loss": 1.5953,
+      "step": 15750
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.15004584193229675,
+      "learning_rate": 0.00036709756228258734,
+      "loss": 1.6289,
+      "step": 15760
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13330994546413422,
+      "learning_rate": 0.0003670569809282074,
+      "loss": 1.642,
+      "step": 15770
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1308055967092514,
+      "learning_rate": 0.00036701637680899694,
+      "loss": 1.5843,
+      "step": 15780
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.135955348610878,
+      "learning_rate": 0.00036697574993048887,
+      "loss": 1.6371,
+      "step": 15790
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13670805096626282,
+      "learning_rate": 0.0003669351002982197,
+      "loss": 1.6549,
+      "step": 15800
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13485939800739288,
+      "learning_rate": 0.0003668944279177284,
+      "loss": 1.6395,
+      "step": 15810
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.12879978120326996,
+      "learning_rate": 0.0003668537327945577,
+      "loss": 1.6125,
+      "step": 15820
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1367185264825821,
+      "learning_rate": 0.0003668130149342529,
+      "loss": 1.6852,
+      "step": 15830
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13566270470619202,
+      "learning_rate": 0.0003667722743423628,
+      "loss": 1.6224,
+      "step": 15840
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.14928579330444336,
+      "learning_rate": 0.000366731511024439,
+      "loss": 1.5977,
+      "step": 15850
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.14337016642093658,
+      "learning_rate": 0.0003666907249860363,
+      "loss": 1.623,
+      "step": 15860
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.14127035439014435,
+      "learning_rate": 0.0003666499162327126,
+      "loss": 1.6455,
+      "step": 15870
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13427312672138214,
+      "learning_rate": 0.000366609084770029,
+      "loss": 1.6298,
+      "step": 15880
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.12484833598136902,
+      "learning_rate": 0.0003665682306035495,
+      "loss": 1.5874,
+      "step": 15890
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1538485586643219,
+      "learning_rate": 0.0003665273537388414,
+      "loss": 1.6411,
+      "step": 15900
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1360001266002655,
+      "learning_rate": 0.0003664864541814749,
+      "loss": 1.5923,
+      "step": 15910
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13738390803337097,
+      "learning_rate": 0.00036644553193702337,
+      "loss": 1.5919,
+      "step": 15920
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13256971538066864,
+      "learning_rate": 0.00036640458701106326,
+      "loss": 1.5939,
+      "step": 15930
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13592809438705444,
+      "learning_rate": 0.0003663636194091742,
+      "loss": 1.5941,
+      "step": 15940
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13964982330799103,
+      "learning_rate": 0.00036632262913693887,
+      "loss": 1.609,
+      "step": 15950
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.14251075685024261,
+      "learning_rate": 0.00036628161619994284,
+      "loss": 1.6015,
+      "step": 15960
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1393299102783203,
+      "learning_rate": 0.00036624058060377517,
+      "loss": 1.5993,
+      "step": 15970
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1459953784942627,
+      "learning_rate": 0.00036619952235402754,
+      "loss": 1.6411,
+      "step": 15980
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12790268659591675,
+      "learning_rate": 0.0003661584414562952,
+      "loss": 1.6101,
+      "step": 15990
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.13234680891036987,
+      "learning_rate": 0.00036611733791617595,
+      "loss": 1.6473,
+      "step": 16000
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.13906261324882507,
+      "learning_rate": 0.0003660762117392712,
+      "loss": 1.6715,
+      "step": 16010
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1301203817129135,
+      "learning_rate": 0.000366035062931185,
+      "loss": 1.618,
+      "step": 16020
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.13661043345928192,
+      "learning_rate": 0.0003659938914975249,
+      "loss": 1.5981,
+      "step": 16030
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1323031336069107,
+      "learning_rate": 0.0003659526974439013,
+      "loss": 1.6031,
+      "step": 16040
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.15984101593494415,
+      "learning_rate": 0.00036591148077592755,
+      "loss": 1.5977,
+      "step": 16050
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12953698635101318,
+      "learning_rate": 0.0003658702414992204,
+      "loss": 1.6713,
+      "step": 16060
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.16284003853797913,
+      "learning_rate": 0.00036582897961939933,
+      "loss": 1.657,
+      "step": 16070
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12465067207813263,
+      "learning_rate": 0.00036578769514208735,
+      "loss": 1.5869,
+      "step": 16080
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.14358744025230408,
+      "learning_rate": 0.00036574638807291014,
+      "loss": 1.6158,
+      "step": 16090
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.13521994650363922,
+      "learning_rate": 0.0003657050584174966,
+      "loss": 1.6355,
+      "step": 16100
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.11851070076227188,
+      "learning_rate": 0.0003656637061814787,
+      "loss": 1.6161,
+      "step": 16110
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.13625980913639069,
+      "learning_rate": 0.0003656223313704916,
+      "loss": 1.6648,
+      "step": 16120
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1362447291612625,
+      "learning_rate": 0.0003655809339901734,
+      "loss": 1.6356,
+      "step": 16130
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1277030110359192,
+      "learning_rate": 0.0003655395140461653,
+      "loss": 1.6262,
+      "step": 16140
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.15926358103752136,
+      "learning_rate": 0.00036549807154411147,
+      "loss": 1.6067,
+      "step": 16150
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.14102047681808472,
+      "learning_rate": 0.0003654566064896595,
+      "loss": 1.6199,
+      "step": 16160
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1393364667892456,
+      "learning_rate": 0.0003654151188884597,
+      "loss": 1.6645,
+      "step": 16170
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1274523288011551,
+      "learning_rate": 0.00036537360874616556,
+      "loss": 1.643,
+      "step": 16180
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12530012428760529,
+      "learning_rate": 0.00036533207606843373,
+      "loss": 1.5815,
+      "step": 16190
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12372483313083649,
+      "learning_rate": 0.00036529052086092383,
+      "loss": 1.5878,
+      "step": 16200
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1443752497434616,
+      "learning_rate": 0.00036524894312929855,
+      "loss": 1.6485,
+      "step": 16210
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.14389920234680176,
+      "learning_rate": 0.0003652073428792238,
+      "loss": 1.5924,
+      "step": 16220
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1192794144153595,
+      "learning_rate": 0.0003651657201163683,
+      "loss": 1.5825,
+      "step": 16230
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.13570775091648102,
+      "learning_rate": 0.000365124074846404,
+      "loss": 1.6108,
+      "step": 16240
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1270972192287445,
+      "learning_rate": 0.000365082407075006,
+      "loss": 1.5794,
+      "step": 16250
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12847332656383514,
+      "learning_rate": 0.00036504071680785224,
+      "loss": 1.5423,
+      "step": 16260
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12084919214248657,
+      "learning_rate": 0.0003649990040506239,
+      "loss": 1.6089,
+      "step": 16270
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12442401796579361,
+      "learning_rate": 0.0003649572688090052,
+      "loss": 1.5695,
+      "step": 16280
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12247755378484726,
+      "learning_rate": 0.00036491551108868325,
+      "loss": 1.5969,
+      "step": 16290
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.13822278380393982,
+      "learning_rate": 0.00036487373089534857,
+      "loss": 1.6128,
+      "step": 16300
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.139133483171463,
+      "learning_rate": 0.0003648319282346945,
+      "loss": 1.6805,
+      "step": 16310
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.15741145610809326,
+      "learning_rate": 0.00036479010311241735,
+      "loss": 1.6141,
+      "step": 16320
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1394251585006714,
+      "learning_rate": 0.00036474825553421675,
+      "loss": 1.5934,
+      "step": 16330
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.13420073688030243,
+      "learning_rate": 0.0003647063855057952,
+      "loss": 1.6157,
+      "step": 16340
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1311005800962448,
+      "learning_rate": 0.00036466449303285836,
+      "loss": 1.611,
+      "step": 16350
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1383652240037918,
+      "learning_rate": 0.0003646225781211149,
+      "loss": 1.6192,
+      "step": 16360
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1383495181798935,
+      "learning_rate": 0.00036458064077627653,
+      "loss": 1.6139,
+      "step": 16370
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.13181906938552856,
+      "learning_rate": 0.0003645386810040581,
+      "loss": 1.6445,
+      "step": 16380
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.12868648767471313,
+      "learning_rate": 0.0003644966988101774,
+      "loss": 1.6133,
+      "step": 16390
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.14313265681266785,
+      "learning_rate": 0.00036445469420035535,
+      "loss": 1.6993,
+      "step": 16400
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.13959407806396484,
+      "learning_rate": 0.0003644126671803159,
+      "loss": 1.6038,
+      "step": 16410
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.12419301271438599,
+      "learning_rate": 0.0003643706177557862,
+      "loss": 1.5975,
+      "step": 16420
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.14102110266685486,
+      "learning_rate": 0.0003643285459324962,
+      "loss": 1.614,
+      "step": 16430
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1222139224410057,
+      "learning_rate": 0.00036428645171617896,
+      "loss": 1.5582,
+      "step": 16440
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.133991077542305,
+      "learning_rate": 0.0003642443351125708,
+      "loss": 1.6062,
+      "step": 16450
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1295316517353058,
+      "learning_rate": 0.00036420219612741086,
+      "loss": 1.6014,
+      "step": 16460
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.12938158214092255,
+      "learning_rate": 0.00036416003476644144,
+      "loss": 1.6378,
+      "step": 16470
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.13738398253917694,
+      "learning_rate": 0.00036411785103540785,
+      "loss": 1.6551,
+      "step": 16480
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.14538368582725525,
+      "learning_rate": 0.0003640756449400584,
+      "loss": 1.6226,
+      "step": 16490
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.13002678751945496,
+      "learning_rate": 0.00036403341648614455,
+      "loss": 1.6083,
+      "step": 16500
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.13674494624137878,
+      "learning_rate": 0.0003639911656794209,
+      "loss": 1.6005,
+      "step": 16510
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1209230050444603,
+      "learning_rate": 0.00036394889252564476,
+      "loss": 1.6166,
+      "step": 16520
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1257326304912567,
+      "learning_rate": 0.0003639065970305768,
+      "loss": 1.5613,
+      "step": 16530
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.12522849440574646,
+      "learning_rate": 0.0003638642791999806,
+      "loss": 1.5945,
+      "step": 16540
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.13467837870121002,
+      "learning_rate": 0.00036382193903962277,
+      "loss": 1.6102,
+      "step": 16550
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1238376796245575,
+      "learning_rate": 0.00036377957655527303,
+      "loss": 1.6119,
+      "step": 16560
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.14226779341697693,
+      "learning_rate": 0.00036373719175270415,
+      "loss": 1.6265,
+      "step": 16570
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1349538266658783,
+      "learning_rate": 0.0003636947846376918,
+      "loss": 1.5881,
+      "step": 16580
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.13550011813640594,
+      "learning_rate": 0.0003636523552160149,
+      "loss": 1.6889,
+      "step": 16590
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1485264152288437,
+      "learning_rate": 0.0003636099034934552,
+      "loss": 1.668,
+      "step": 16600
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1284945011138916,
+      "learning_rate": 0.00036356742947579767,
+      "loss": 1.6367,
+      "step": 16610
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1161959320306778,
+      "learning_rate": 0.0003635249331688302,
+      "loss": 1.5979,
+      "step": 16620
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.13754938542842865,
+      "learning_rate": 0.00036348241457834376,
+      "loss": 1.6322,
+      "step": 16630
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.13457804918289185,
+      "learning_rate": 0.0003634398737101323,
+      "loss": 1.615,
+      "step": 16640
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.13922441005706787,
+      "learning_rate": 0.000363397310569993,
+      "loss": 1.5582,
+      "step": 16650
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1432957500219345,
+      "learning_rate": 0.0003633547251637258,
+      "loss": 1.605,
+      "step": 16660
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.12548604607582092,
+      "learning_rate": 0.0003633121174971339,
+      "loss": 1.5685,
+      "step": 16670
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.11861354857683182,
+      "learning_rate": 0.00036326948757602333,
+      "loss": 1.5585,
+      "step": 16680
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.12478451430797577,
+      "learning_rate": 0.00036322683540620333,
+      "loss": 1.6186,
+      "step": 16690
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.12997831404209137,
+      "learning_rate": 0.00036318416099348613,
+      "loss": 1.6047,
+      "step": 16700
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.15071137249469757,
+      "learning_rate": 0.00036314146434368697,
+      "loss": 1.6878,
+      "step": 16710
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1548747420310974,
+      "learning_rate": 0.000363098745462624,
+      "loss": 1.6071,
+      "step": 16720
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.12561798095703125,
+      "learning_rate": 0.00036305600435611866,
+      "loss": 1.5821,
+      "step": 16730
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1342022567987442,
+      "learning_rate": 0.0003630132410299952,
+      "loss": 1.6475,
+      "step": 16740
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1241414025425911,
+      "learning_rate": 0.00036297045549008104,
+      "loss": 1.6367,
+      "step": 16750
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.12133163213729858,
+      "learning_rate": 0.00036292764774220637,
+      "loss": 1.5694,
+      "step": 16760
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1390984058380127,
+      "learning_rate": 0.00036288481779220484,
+      "loss": 1.5986,
+      "step": 16770
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1254926323890686,
+      "learning_rate": 0.0003628419656459127,
+      "loss": 1.5701,
+      "step": 16780
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.13598591089248657,
+      "learning_rate": 0.00036279909130916956,
+      "loss": 1.6334,
+      "step": 16790
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.12620100378990173,
+      "learning_rate": 0.00036275619478781776,
+      "loss": 1.5454,
+      "step": 16800
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1288425177335739,
+      "learning_rate": 0.00036271327608770285,
+      "loss": 1.6227,
+      "step": 16810
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12438348680734634,
+      "learning_rate": 0.0003626703352146734,
+      "loss": 1.5901,
+      "step": 16820
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.14789782464504242,
+      "learning_rate": 0.00036262737217458094,
+      "loss": 1.5694,
+      "step": 16830
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1263388991355896,
+      "learning_rate": 0.00036258438697327997,
+      "loss": 1.5749,
+      "step": 16840
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12653589248657227,
+      "learning_rate": 0.00036254137961662814,
+      "loss": 1.5994,
+      "step": 16850
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13028274476528168,
+      "learning_rate": 0.000362498350110486,
+      "loss": 1.6408,
+      "step": 16860
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1280732899904251,
+      "learning_rate": 0.0003624552984607173,
+      "loss": 1.5507,
+      "step": 16870
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1309758424758911,
+      "learning_rate": 0.0003624122246731886,
+      "loss": 1.5827,
+      "step": 16880
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1449895054101944,
+      "learning_rate": 0.0003623691287537695,
+      "loss": 1.6316,
+      "step": 16890
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12817351520061493,
+      "learning_rate": 0.00036232601070833277,
+      "loss": 1.5995,
+      "step": 16900
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.14033293724060059,
+      "learning_rate": 0.00036228287054275405,
+      "loss": 1.5687,
+      "step": 16910
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13607163727283478,
+      "learning_rate": 0.000362239708262912,
+      "loss": 1.5834,
+      "step": 16920
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13933010399341583,
+      "learning_rate": 0.0003621965238746885,
+      "loss": 1.5823,
+      "step": 16930
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13570238649845123,
+      "learning_rate": 0.0003621533173839681,
+      "loss": 1.6722,
+      "step": 16940
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1349225640296936,
+      "learning_rate": 0.0003621100887966386,
+      "loss": 1.6012,
+      "step": 16950
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.14689001441001892,
+      "learning_rate": 0.00036206683811859083,
+      "loss": 1.5975,
+      "step": 16960
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13853569328784943,
+      "learning_rate": 0.00036202356535571846,
+      "loss": 1.6413,
+      "step": 16970
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.16470015048980713,
+      "learning_rate": 0.00036198027051391837,
+      "loss": 1.6172,
+      "step": 16980
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13591763377189636,
+      "learning_rate": 0.00036193695359909014,
+      "loss": 1.6524,
+      "step": 16990
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1244286447763443,
+      "learning_rate": 0.0003618936146171368,
+      "loss": 1.604,
+      "step": 17000
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13039223849773407,
+      "learning_rate": 0.000361850253573964,
+      "loss": 1.7184,
+      "step": 17010
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12885712087154388,
+      "learning_rate": 0.0003618068704754806,
+      "loss": 1.6786,
+      "step": 17020
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13418753445148468,
+      "learning_rate": 0.0003617634653275983,
+      "loss": 1.6454,
+      "step": 17030
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1279221624135971,
+      "learning_rate": 0.0003617200381362321,
+      "loss": 1.5892,
+      "step": 17040
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1341562271118164,
+      "learning_rate": 0.0003616765889072996,
+      "loss": 1.589,
+      "step": 17050
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1447325348854065,
+      "learning_rate": 0.00036163311764672183,
+      "loss": 1.6553,
+      "step": 17060
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12853869795799255,
+      "learning_rate": 0.0003615896243604225,
+      "loss": 1.6447,
+      "step": 17070
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12335439026355743,
+      "learning_rate": 0.0003615461090543284,
+      "loss": 1.5994,
+      "step": 17080
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13983649015426636,
+      "learning_rate": 0.0003615025717343695,
+      "loss": 1.5977,
+      "step": 17090
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13573993742465973,
+      "learning_rate": 0.0003614590124064785,
+      "loss": 1.5058,
+      "step": 17100
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12648895382881165,
+      "learning_rate": 0.0003614154310765913,
+      "loss": 1.6316,
+      "step": 17110
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1266384720802307,
+      "learning_rate": 0.00036137182775064654,
+      "loss": 1.5884,
+      "step": 17120
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.125233992934227,
+      "learning_rate": 0.0003613282024345863,
+      "loss": 1.5952,
+      "step": 17130
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13106931746006012,
+      "learning_rate": 0.0003612845551343552,
+      "loss": 1.6127,
+      "step": 17140
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1413438320159912,
+      "learning_rate": 0.0003612408858559012,
+      "loss": 1.5783,
+      "step": 17150
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.14359228312969208,
+      "learning_rate": 0.00036119719460517497,
+      "loss": 1.6525,
+      "step": 17160
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12513476610183716,
+      "learning_rate": 0.0003611534813881304,
+      "loss": 1.6034,
+      "step": 17170
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13809636235237122,
+      "learning_rate": 0.00036110974621072426,
+      "loss": 1.5736,
+      "step": 17180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12614567577838898,
+      "learning_rate": 0.00036106598907891625,
+      "loss": 1.5562,
+      "step": 17190
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1535719484090805,
+      "learning_rate": 0.0003610222099986693,
+      "loss": 1.6186,
+      "step": 17200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12354706227779388,
+      "learning_rate": 0.0003609784089759491,
+      "loss": 1.5717,
+      "step": 17210
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1161801815032959,
+      "learning_rate": 0.0003609345860167244,
+      "loss": 1.6135,
+      "step": 17220
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1545819640159607,
+      "learning_rate": 0.0003608907411269669,
+      "loss": 1.618,
+      "step": 17230
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.13866360485553741,
+      "learning_rate": 0.00036084687431265153,
+      "loss": 1.6237,
+      "step": 17240
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1447690725326538,
+      "learning_rate": 0.0003608029855797558,
+      "loss": 1.5795,
+      "step": 17250
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.12596924602985382,
+      "learning_rate": 0.0003607590749342605,
+      "loss": 1.555,
+      "step": 17260
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1205577403306961,
+      "learning_rate": 0.00036071514238214937,
+      "loss": 1.6004,
+      "step": 17270
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.13527412712574005,
+      "learning_rate": 0.0003606711879294091,
+      "loss": 1.5618,
+      "step": 17280
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.122543103992939,
+      "learning_rate": 0.0003606272115820292,
+      "loss": 1.5911,
+      "step": 17290
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.12264387309551239,
+      "learning_rate": 0.00036058321334600247,
+      "loss": 1.6142,
+      "step": 17300
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.12851102650165558,
+      "learning_rate": 0.00036053919322732455,
+      "loss": 1.5639,
+      "step": 17310
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.13379542529582977,
+      "learning_rate": 0.000360495151231994,
+      "loss": 1.6082,
+      "step": 17320
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1274455040693283,
+      "learning_rate": 0.00036045108736601245,
+      "loss": 1.6021,
+      "step": 17330
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.12506166100502014,
+      "learning_rate": 0.0003604070016353844,
+      "loss": 1.642,
+      "step": 17340
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.14403843879699707,
+      "learning_rate": 0.0003603628940461175,
+      "loss": 1.5917,
+      "step": 17350
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.13639913499355316,
+      "learning_rate": 0.0003603187646042222,
+      "loss": 1.6461,
+      "step": 17360
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.12027144432067871,
+      "learning_rate": 0.0003602746133157121,
+      "loss": 1.5796,
+      "step": 17370
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.13953126966953278,
+      "learning_rate": 0.0003602304401866037,
+      "loss": 1.6169,
+      "step": 17380
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.13660165667533875,
+      "learning_rate": 0.00036018624522291627,
+      "loss": 1.6167,
+      "step": 17390
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.15013617277145386,
+      "learning_rate": 0.0003601420284306725,
+      "loss": 1.5748,
+      "step": 17400
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.13552677631378174,
+      "learning_rate": 0.0003600977898158977,
+      "loss": 1.5752,
+      "step": 17410
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.14618028700351715,
+      "learning_rate": 0.0003600535293846202,
+      "loss": 1.5769,
+      "step": 17420
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.13513758778572083,
+      "learning_rate": 0.0003600092471428715,
+      "loss": 1.6294,
+      "step": 17430
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.15010881423950195,
+      "learning_rate": 0.0003599649430966858,
+      "loss": 1.6045,
+      "step": 17440
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.12532857060432434,
+      "learning_rate": 0.0003599206172521005,
+      "loss": 1.5705,
+      "step": 17450
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1190020889043808,
+      "learning_rate": 0.00035987626961515574,
+      "loss": 1.6192,
+      "step": 17460
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.11795544624328613,
+      "learning_rate": 0.000359831900191895,
+      "loss": 1.5889,
+      "step": 17470
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1490863412618637,
+      "learning_rate": 0.0003597875089883643,
+      "loss": 1.5876,
+      "step": 17480
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.20040272176265717,
+      "learning_rate": 0.0003597430960106129,
+      "loss": 1.5795,
+      "step": 17490
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.12842829525470734,
+      "learning_rate": 0.0003596986612646929,
+      "loss": 1.7063,
+      "step": 17500
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.13042472302913666,
+      "learning_rate": 0.00035965420475665954,
+      "loss": 1.5868,
+      "step": 17510
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.13298243284225464,
+      "learning_rate": 0.00035960972649257074,
+      "loss": 1.5398,
+      "step": 17520
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.13182935118675232,
+      "learning_rate": 0.0003595652264784876,
+      "loss": 1.5903,
+      "step": 17530
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1285000890493393,
+      "learning_rate": 0.0003595207047204742,
+      "loss": 1.6092,
+      "step": 17540
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.11306576430797577,
+      "learning_rate": 0.0003594761612245975,
+      "loss": 1.5463,
+      "step": 17550
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.13159814476966858,
+      "learning_rate": 0.00035943159599692733,
+      "loss": 1.586,
+      "step": 17560
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.13080434501171112,
+      "learning_rate": 0.00035938700904353666,
+      "loss": 1.5726,
+      "step": 17570
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.14132794737815857,
+      "learning_rate": 0.0003593424003705013,
+      "loss": 1.5901,
+      "step": 17580
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.13450951874256134,
+      "learning_rate": 0.00035929776998390017,
+      "loss": 1.5319,
+      "step": 17590
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.18116261065006256,
+      "learning_rate": 0.000359253117889815,
+      "loss": 1.5861,
+      "step": 17600
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1406196802854538,
+      "learning_rate": 0.0003592084440943305,
+      "loss": 1.5917,
+      "step": 17610
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.17923685908317566,
+      "learning_rate": 0.0003591637486035344,
+      "loss": 1.6051,
+      "step": 17620
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1503029763698578,
+      "learning_rate": 0.00035911903142351725,
+      "loss": 1.5905,
+      "step": 17630
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.14010436832904816,
+      "learning_rate": 0.00035907429256037276,
+      "loss": 1.6887,
+      "step": 17640
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.14434535801410675,
+      "learning_rate": 0.00035902953202019745,
+      "loss": 1.6516,
+      "step": 17650
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13706041872501373,
+      "learning_rate": 0.00035898474980909086,
+      "loss": 1.6626,
+      "step": 17660
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13165655732154846,
+      "learning_rate": 0.0003589399459331554,
+      "loss": 1.571,
+      "step": 17670
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13651034235954285,
+      "learning_rate": 0.00035889512039849654,
+      "loss": 1.5678,
+      "step": 17680
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13355237245559692,
+      "learning_rate": 0.0003588502732112226,
+      "loss": 1.6254,
+      "step": 17690
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1287340521812439,
+      "learning_rate": 0.00035880540437744496,
+      "loss": 1.5939,
+      "step": 17700
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13894182443618774,
+      "learning_rate": 0.0003587605139032778,
+      "loss": 1.6317,
+      "step": 17710
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13954025506973267,
+      "learning_rate": 0.0003587156017948385,
+      "loss": 1.6378,
+      "step": 17720
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.11821527034044266,
+      "learning_rate": 0.0003586706680582471,
+      "loss": 1.581,
+      "step": 17730
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12625131011009216,
+      "learning_rate": 0.0003586257126996267,
+      "loss": 1.5865,
+      "step": 17740
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1357714682817459,
+      "learning_rate": 0.00035858073572510334,
+      "loss": 1.6391,
+      "step": 17750
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12236014008522034,
+      "learning_rate": 0.0003585357371408061,
+      "loss": 1.6483,
+      "step": 17760
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13785655796527863,
+      "learning_rate": 0.00035849071695286697,
+      "loss": 1.5684,
+      "step": 17770
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13554228842258453,
+      "learning_rate": 0.00035844567516742077,
+      "loss": 1.6019,
+      "step": 17780
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1393551528453827,
+      "learning_rate": 0.00035840061179060536,
+      "loss": 1.6155,
+      "step": 17790
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1255672574043274,
+      "learning_rate": 0.00035835552682856155,
+      "loss": 1.5794,
+      "step": 17800
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12974193692207336,
+      "learning_rate": 0.000358310420287433,
+      "loss": 1.6027,
+      "step": 17810
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12758152186870575,
+      "learning_rate": 0.0003582652921733663,
+      "loss": 1.6585,
+      "step": 17820
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.14791662991046906,
+      "learning_rate": 0.00035822014249251125,
+      "loss": 1.57,
+      "step": 17830
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.14504961669445038,
+      "learning_rate": 0.00035817497125102026,
+      "loss": 1.6074,
+      "step": 17840
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12555773556232452,
+      "learning_rate": 0.0003581297784550489,
+      "loss": 1.5591,
+      "step": 17850
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12130322307348251,
+      "learning_rate": 0.00035808456411075544,
+      "loss": 1.628,
+      "step": 17860
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.14973516762256622,
+      "learning_rate": 0.00035803932822430135,
+      "loss": 1.6337,
+      "step": 17870
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13796919584274292,
+      "learning_rate": 0.00035799407080185086,
+      "loss": 1.5623,
+      "step": 17880
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1370391845703125,
+      "learning_rate": 0.0003579487918495713,
+      "loss": 1.6117,
+      "step": 17890
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13227681815624237,
+      "learning_rate": 0.0003579034913736326,
+      "loss": 1.5952,
+      "step": 17900
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13334843516349792,
+      "learning_rate": 0.0003578581693802081,
+      "loss": 1.5633,
+      "step": 17910
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13593466579914093,
+      "learning_rate": 0.0003578128258754737,
+      "loss": 1.5912,
+      "step": 17920
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13452664017677307,
+      "learning_rate": 0.0003577674608656083,
+      "loss": 1.5515,
+      "step": 17930
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12645120918750763,
+      "learning_rate": 0.00035772207435679393,
+      "loss": 1.6265,
+      "step": 17940
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12202449887990952,
+      "learning_rate": 0.00035767666635521533,
+      "loss": 1.566,
+      "step": 17950
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1435176134109497,
+      "learning_rate": 0.00035763123686706016,
+      "loss": 1.6506,
+      "step": 17960
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13037852942943573,
+      "learning_rate": 0.0003575857858985192,
+      "loss": 1.5795,
+      "step": 17970
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13606557250022888,
+      "learning_rate": 0.000357540313455786,
+      "loss": 1.5988,
+      "step": 17980
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.18163901567459106,
+      "learning_rate": 0.0003574948195450572,
+      "loss": 1.6009,
+      "step": 17990
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13054554164409637,
+      "learning_rate": 0.00035744930417253197,
+      "loss": 1.6055,
+      "step": 18000
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.11851833015680313,
+      "learning_rate": 0.0003574037673444129,
+      "loss": 1.6852,
+      "step": 18010
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13111279904842377,
+      "learning_rate": 0.0003573582090669053,
+      "loss": 1.6595,
+      "step": 18020
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13551484048366547,
+      "learning_rate": 0.0003573126293462173,
+      "loss": 1.6674,
+      "step": 18030
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1378137171268463,
+      "learning_rate": 0.0003572670281885601,
+      "loss": 1.6414,
+      "step": 18040
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17129214107990265,
+      "learning_rate": 0.0003572214056001477,
+      "loss": 1.6055,
+      "step": 18050
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12832503020763397,
+      "learning_rate": 0.00035717576158719716,
+      "loss": 1.597,
+      "step": 18060
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.138489231467247,
+      "learning_rate": 0.0003571300961559283,
+      "loss": 1.5725,
+      "step": 18070
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12459228187799454,
+      "learning_rate": 0.00035708440931256407,
+      "loss": 1.5779,
+      "step": 18080
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.13217143714427948,
+      "learning_rate": 0.00035703870106333006,
+      "loss": 1.5854,
+      "step": 18090
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.13522735238075256,
+      "learning_rate": 0.000356992971414455,
+      "loss": 1.5899,
+      "step": 18100
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.13625244796276093,
+      "learning_rate": 0.0003569472203721704,
+      "loss": 1.6548,
+      "step": 18110
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1573963314294815,
+      "learning_rate": 0.0003569014479427109,
+      "loss": 1.5922,
+      "step": 18120
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.13964055478572845,
+      "learning_rate": 0.00035685565413231374,
+      "loss": 1.5917,
+      "step": 18130
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.13809914886951447,
+      "learning_rate": 0.00035680983894721934,
+      "loss": 1.617,
+      "step": 18140
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1583004742860794,
+      "learning_rate": 0.00035676400239367085,
+      "loss": 1.6302,
+      "step": 18150
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1293550729751587,
+      "learning_rate": 0.0003567181444779145,
+      "loss": 1.603,
+      "step": 18160
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.13756272196769714,
+      "learning_rate": 0.0003566722652061993,
+      "loss": 1.6356,
+      "step": 18170
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.12604734301567078,
+      "learning_rate": 0.0003566263645847771,
+      "loss": 1.5885,
+      "step": 18180
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.143514946103096,
+      "learning_rate": 0.000356580442619903,
+      "loss": 1.6202,
+      "step": 18190
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.12903961539268494,
+      "learning_rate": 0.00035653449931783457,
+      "loss": 1.6126,
+      "step": 18200
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.12308374047279358,
+      "learning_rate": 0.00035648853468483267,
+      "loss": 1.5685,
+      "step": 18210
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1230534017086029,
+      "learning_rate": 0.0003564425487271607,
+      "loss": 1.5359,
+      "step": 18220
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1379854679107666,
+      "learning_rate": 0.0003563965414510853,
+      "loss": 1.5916,
+      "step": 18230
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17015758156776428,
+      "learning_rate": 0.0003563505128628759,
+      "loss": 1.6039,
+      "step": 18240
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.13459518551826477,
+      "learning_rate": 0.0003563044629688046,
+      "loss": 1.6051,
+      "step": 18250
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.13336966931819916,
+      "learning_rate": 0.00035625839177514686,
+      "loss": 1.5736,
+      "step": 18260
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.15190520882606506,
+      "learning_rate": 0.0003562122992881808,
+      "loss": 1.6056,
+      "step": 18270
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.13517926633358002,
+      "learning_rate": 0.0003561661855141872,
+      "loss": 1.6356,
+      "step": 18280
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1199948638677597,
+      "learning_rate": 0.0003561200504594501,
+      "loss": 1.6407,
+      "step": 18290
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.132409930229187,
+      "learning_rate": 0.00035607389413025637,
+      "loss": 1.6302,
+      "step": 18300
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.12950561940670013,
+      "learning_rate": 0.00035602771653289565,
+      "loss": 1.5887,
+      "step": 18310
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.15904468297958374,
+      "learning_rate": 0.0003559815176736606,
+      "loss": 1.5899,
+      "step": 18320
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.13247163593769073,
+      "learning_rate": 0.00035593529755884674,
+      "loss": 1.6145,
+      "step": 18330
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1395135521888733,
+      "learning_rate": 0.00035588905619475243,
+      "loss": 1.5837,
+      "step": 18340
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.14861518144607544,
+      "learning_rate": 0.00035584279358767906,
+      "loss": 1.5762,
+      "step": 18350
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1333700716495514,
+      "learning_rate": 0.00035579650974393073,
+      "loss": 1.5767,
+      "step": 18360
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1199478954076767,
+      "learning_rate": 0.00035575020466981457,
+      "loss": 1.5463,
+      "step": 18370
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1338033825159073,
+      "learning_rate": 0.00035570387837164063,
+      "loss": 1.5693,
+      "step": 18380
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.13631267845630646,
+      "learning_rate": 0.0003556575308557216,
+      "loss": 1.5441,
+      "step": 18390
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.12850597500801086,
+      "learning_rate": 0.0003556111621283734,
+      "loss": 1.643,
+      "step": 18400
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1382976919412613,
+      "learning_rate": 0.0003555647721959148,
+      "loss": 1.6327,
+      "step": 18410
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1332872360944748,
+      "learning_rate": 0.00035551836106466713,
+      "loss": 1.6149,
+      "step": 18420
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.13765592873096466,
+      "learning_rate": 0.0003554719287409549,
+      "loss": 1.6236,
+      "step": 18430
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1573352962732315,
+      "learning_rate": 0.0003554254752311055,
+      "loss": 1.6666,
+      "step": 18440
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1276737004518509,
+      "learning_rate": 0.00035537900054144906,
+      "loss": 1.578,
+      "step": 18450
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.12131106108427048,
+      "learning_rate": 0.00035533250467831865,
+      "loss": 1.6673,
+      "step": 18460
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1268194168806076,
+      "learning_rate": 0.0003552859876480504,
+      "loss": 1.5796,
+      "step": 18470
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.12316969782114029,
+      "learning_rate": 0.00035523944945698306,
+      "loss": 1.6052,
+      "step": 18480
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1469998061656952,
+      "learning_rate": 0.00035519289011145844,
+      "loss": 1.6342,
+      "step": 18490
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1315140575170517,
+      "learning_rate": 0.00035514630961782116,
+      "loss": 1.6061,
+      "step": 18500
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.16311466693878174,
+      "learning_rate": 0.00035509970798241867,
+      "loss": 1.6093,
+      "step": 18510
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1728275567293167,
+      "learning_rate": 0.0003550530852116015,
+      "loss": 1.6369,
+      "step": 18520
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.15360009670257568,
+      "learning_rate": 0.0003550064413117228,
+      "loss": 1.6107,
+      "step": 18530
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1581055074930191,
+      "learning_rate": 0.00035495977628913883,
+      "loss": 1.6541,
+      "step": 18540
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1480787843465805,
+      "learning_rate": 0.00035491309015020853,
+      "loss": 1.5769,
+      "step": 18550
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1461239755153656,
+      "learning_rate": 0.0003548663829012939,
+      "loss": 1.6665,
+      "step": 18560
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.14492259919643402,
+      "learning_rate": 0.0003548196545487597,
+      "loss": 1.6137,
+      "step": 18570
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.12792524695396423,
+      "learning_rate": 0.0003547729050989736,
+      "loss": 1.6377,
+      "step": 18580
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1240549087524414,
+      "learning_rate": 0.0003547261345583061,
+      "loss": 1.5857,
+      "step": 18590
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.13046002388000488,
+      "learning_rate": 0.0003546793429331307,
+      "loss": 1.5843,
+      "step": 18600
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.12912024557590485,
+      "learning_rate": 0.00035463253022982355,
+      "loss": 1.5945,
+      "step": 18610
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1412425935268402,
+      "learning_rate": 0.0003545856964547639,
+      "loss": 1.6117,
+      "step": 18620
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.13252173364162445,
+      "learning_rate": 0.00035453884161433383,
+      "loss": 1.5917,
+      "step": 18630
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1491372287273407,
+      "learning_rate": 0.00035449196571491815,
+      "loss": 1.5826,
+      "step": 18640
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.14020954072475433,
+      "learning_rate": 0.00035444506876290466,
+      "loss": 1.5947,
+      "step": 18650
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1251155287027359,
+      "learning_rate": 0.00035439815076468407,
+      "loss": 1.5975,
+      "step": 18660
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.13068650662899017,
+      "learning_rate": 0.00035435121172664976,
+      "loss": 1.5988,
+      "step": 18670
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.13086429238319397,
+      "learning_rate": 0.00035430425165519825,
+      "loss": 1.5676,
+      "step": 18680
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1352221965789795,
+      "learning_rate": 0.0003542572705567286,
+      "loss": 1.6254,
+      "step": 18690
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.13545511662960052,
+      "learning_rate": 0.00035421026843764314,
+      "loss": 1.5783,
+      "step": 18700
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.13622403144836426,
+      "learning_rate": 0.00035416324530434666,
+      "loss": 1.5584,
+      "step": 18710
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.12750837206840515,
+      "learning_rate": 0.00035411620116324705,
+      "loss": 1.5641,
+      "step": 18720
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.12281973659992218,
+      "learning_rate": 0.0003540691360207551,
+      "loss": 1.5353,
+      "step": 18730
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.12332561612129211,
+      "learning_rate": 0.0003540220498832842,
+      "loss": 1.5466,
+      "step": 18740
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1287834346294403,
+      "learning_rate": 0.00035397494275725086,
+      "loss": 1.6103,
+      "step": 18750
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.12880933284759521,
+      "learning_rate": 0.00035392781464907446,
+      "loss": 1.6153,
+      "step": 18760
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.14241217076778412,
+      "learning_rate": 0.00035388066556517697,
+      "loss": 1.6119,
+      "step": 18770
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.14779892563819885,
+      "learning_rate": 0.0003538334955119835,
+      "loss": 1.6311,
+      "step": 18780
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.129306823015213,
+      "learning_rate": 0.00035378630449592183,
+      "loss": 1.5931,
+      "step": 18790
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.14117935299873352,
+      "learning_rate": 0.0003537390925234227,
+      "loss": 1.5984,
+      "step": 18800
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.12805402278900146,
+      "learning_rate": 0.00035369185960091974,
+      "loss": 1.6055,
+      "step": 18810
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.14149801433086395,
+      "learning_rate": 0.0003536446057348493,
+      "loss": 1.5883,
+      "step": 18820
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.12275634706020355,
+      "learning_rate": 0.0003535973309316507,
+      "loss": 1.6146,
+      "step": 18830
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.12997306883335114,
+      "learning_rate": 0.000353550035197766,
+      "loss": 1.5797,
+      "step": 18840
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.13923434913158417,
+      "learning_rate": 0.0003535027185396403,
+      "loss": 1.5703,
+      "step": 18850
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1351623684167862,
+      "learning_rate": 0.0003534553809637213,
+      "loss": 1.5537,
+      "step": 18860
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.13322000205516815,
+      "learning_rate": 0.0003534080224764598,
+      "loss": 1.6483,
+      "step": 18870
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.13878951966762543,
+      "learning_rate": 0.00035336064308430927,
+      "loss": 1.6697,
+      "step": 18880
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.13363203406333923,
+      "learning_rate": 0.0003533132427937261,
+      "loss": 1.605,
+      "step": 18890
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.17968600988388062,
+      "learning_rate": 0.0003532658216111695,
+      "loss": 1.9606,
+      "step": 18900
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.15599747002124786,
+      "learning_rate": 0.00035321837954310173,
+      "loss": 1.6529,
+      "step": 18910
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1412549465894699,
+      "learning_rate": 0.0003531709165959874,
+      "loss": 1.6552,
+      "step": 18920
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.12435323745012283,
+      "learning_rate": 0.00035312343277629456,
+      "loss": 1.5852,
+      "step": 18930
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.13163238763809204,
+      "learning_rate": 0.0003530759280904936,
+      "loss": 1.6127,
+      "step": 18940
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1244221180677414,
+      "learning_rate": 0.00035302840254505813,
+      "loss": 1.6367,
+      "step": 18950
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1289689689874649,
+      "learning_rate": 0.0003529808561464644,
+      "loss": 1.6118,
+      "step": 18960
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.14280271530151367,
+      "learning_rate": 0.0003529332889011916,
+      "loss": 1.5785,
+      "step": 18970
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1336173117160797,
+      "learning_rate": 0.0003528857008157216,
+      "loss": 1.5502,
+      "step": 18980
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.12242089211940765,
+      "learning_rate": 0.0003528380918965393,
+      "loss": 1.5348,
+      "step": 18990
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.12518438696861267,
+      "learning_rate": 0.0003527904621501323,
+      "loss": 1.5828,
+      "step": 19000
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.13296419382095337,
+      "learning_rate": 0.0003527428115829912,
+      "loss": 1.5208,
+      "step": 19010
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.13181109726428986,
+      "learning_rate": 0.0003526951402016093,
+      "loss": 1.5844,
+      "step": 19020
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.13117043673992157,
+      "learning_rate": 0.0003526474480124827,
+      "loss": 1.6027,
+      "step": 19030
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3249737322330475,
+      "learning_rate": 0.00035259973502211044,
+      "loss": 1.6708,
+      "step": 19040
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.137907937169075,
+      "learning_rate": 0.00035255200123699446,
+      "loss": 1.6616,
+      "step": 19050
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.12870895862579346,
+      "learning_rate": 0.0003525042466636393,
+      "loss": 1.5573,
+      "step": 19060
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.12688829004764557,
+      "learning_rate": 0.0003524564713085526,
+      "loss": 1.5711,
+      "step": 19070
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.13367979228496552,
+      "learning_rate": 0.00035240867517824455,
+      "loss": 1.5796,
+      "step": 19080
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.12754486501216888,
+      "learning_rate": 0.00035236085827922836,
+      "loss": 1.5725,
+      "step": 19090
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.12467335164546967,
+      "learning_rate": 0.00035231302061802014,
+      "loss": 1.5939,
+      "step": 19100
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.14020243287086487,
+      "learning_rate": 0.00035226516220113863,
+      "loss": 1.5897,
+      "step": 19110
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1368483304977417,
+      "learning_rate": 0.00035221728303510554,
+      "loss": 1.5631,
+      "step": 19120
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.141549751162529,
+      "learning_rate": 0.00035216938312644533,
+      "loss": 1.6069,
+      "step": 19130
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1438661366701126,
+      "learning_rate": 0.0003521214624816853,
+      "loss": 1.6044,
+      "step": 19140
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.15190082788467407,
+      "learning_rate": 0.0003520735211073556,
+      "loss": 1.6238,
+      "step": 19150
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.12435395270586014,
+      "learning_rate": 0.0003520255590099892,
+      "loss": 1.6393,
+      "step": 19160
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.12782694399356842,
+      "learning_rate": 0.00035197757619612186,
+      "loss": 1.5998,
+      "step": 19170
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1262042224407196,
+      "learning_rate": 0.00035192957267229225,
+      "loss": 1.5822,
+      "step": 19180
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1273982673883438,
+      "learning_rate": 0.0003518815484450418,
+      "loss": 1.6225,
+      "step": 19190
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.12178373336791992,
+      "learning_rate": 0.00035183350352091465,
+      "loss": 1.565,
+      "step": 19200
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.12814302742481232,
+      "learning_rate": 0.00035178543790645797,
+      "loss": 1.6319,
+      "step": 19210
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.13452085852622986,
+      "learning_rate": 0.0003517373516082216,
+      "loss": 1.5903,
+      "step": 19220
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.12467807531356812,
+      "learning_rate": 0.0003516892446327584,
+      "loss": 1.5449,
+      "step": 19230
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.14870715141296387,
+      "learning_rate": 0.00035164111698662376,
+      "loss": 1.6069,
+      "step": 19240
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1357821226119995,
+      "learning_rate": 0.0003515929686763761,
+      "loss": 1.5752,
+      "step": 19250
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.13286936283111572,
+      "learning_rate": 0.0003515447997085765,
+      "loss": 1.5414,
+      "step": 19260
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1417759507894516,
+      "learning_rate": 0.00035149661008978893,
+      "loss": 1.669,
+      "step": 19270
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.12912607192993164,
+      "learning_rate": 0.0003514483998265804,
+      "loss": 1.6329,
+      "step": 19280
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.13168570399284363,
+      "learning_rate": 0.0003514001689255202,
+      "loss": 1.6046,
+      "step": 19290
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1332264095544815,
+      "learning_rate": 0.0003513519173931809,
+      "loss": 1.6142,
+      "step": 19300
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.14723975956439972,
+      "learning_rate": 0.0003513036452361378,
+      "loss": 1.5926,
+      "step": 19310
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.16024403274059296,
+      "learning_rate": 0.00035125535246096886,
+      "loss": 1.6164,
+      "step": 19320
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.14766928553581238,
+      "learning_rate": 0.0003512070390742549,
+      "loss": 1.6107,
+      "step": 19330
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.13881303369998932,
+      "learning_rate": 0.0003511587050825797,
+      "loss": 1.6363,
+      "step": 19340
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.15090309083461761,
+      "learning_rate": 0.00035111035049252955,
+      "loss": 1.571,
+      "step": 19350
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1362057328224182,
+      "learning_rate": 0.00035106197531069387,
+      "loss": 1.62,
+      "step": 19360
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.15984965860843658,
+      "learning_rate": 0.00035101357954366464,
+      "loss": 1.6545,
+      "step": 19370
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.12805554270744324,
+      "learning_rate": 0.00035096516319803686,
+      "loss": 1.595,
+      "step": 19380
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.14639818668365479,
+      "learning_rate": 0.00035091672628040805,
+      "loss": 1.6354,
+      "step": 19390
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.13649944961071014,
+      "learning_rate": 0.0003508682687973789,
+      "loss": 1.6099,
+      "step": 19400
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.15236856043338776,
+      "learning_rate": 0.0003508197907555525,
+      "loss": 1.6071,
+      "step": 19410
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.12846870720386505,
+      "learning_rate": 0.0003507712921615351,
+      "loss": 1.5722,
+      "step": 19420
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.13661068677902222,
+      "learning_rate": 0.00035072277302193556,
+      "loss": 1.6073,
+      "step": 19430
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1407945156097412,
+      "learning_rate": 0.0003506742333433656,
+      "loss": 1.5977,
+      "step": 19440
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.12419061362743378,
+      "learning_rate": 0.00035062567313243957,
+      "loss": 1.6379,
+      "step": 19450
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.12311229854822159,
+      "learning_rate": 0.000350577092395775,
+      "loss": 1.5852,
+      "step": 19460
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.12563548982143402,
+      "learning_rate": 0.0003505284911399917,
+      "loss": 1.6405,
+      "step": 19470
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.14141380786895752,
+      "learning_rate": 0.00035047986937171286,
+      "loss": 1.6001,
+      "step": 19480
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.13704250752925873,
+      "learning_rate": 0.0003504312270975639,
+      "loss": 1.6066,
+      "step": 19490
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.13638803362846375,
+      "learning_rate": 0.0003503825643241735,
+      "loss": 1.5896,
+      "step": 19500
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.13082872331142426,
+      "learning_rate": 0.0003503338810581727,
+      "loss": 1.5901,
+      "step": 19510
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1453154981136322,
+      "learning_rate": 0.00035028517730619575,
+      "loss": 1.6188,
+      "step": 19520
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.12285885959863663,
+      "learning_rate": 0.00035023645307487945,
+      "loss": 1.5683,
+      "step": 19530
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.14864851534366608,
+      "learning_rate": 0.00035018770837086345,
+      "loss": 1.6603,
+      "step": 19540
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.14033930003643036,
+      "learning_rate": 0.0003501389432007902,
+      "loss": 1.64,
+      "step": 19550
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.16176994144916534,
+      "learning_rate": 0.00035009015757130486,
+      "loss": 1.6354,
+      "step": 19560
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1283935308456421,
+      "learning_rate": 0.0003500413514890555,
+      "loss": 1.5697,
+      "step": 19570
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1200854480266571,
+      "learning_rate": 0.00034999252496069284,
+      "loss": 1.5617,
+      "step": 19580
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.12797559797763824,
+      "learning_rate": 0.0003499436779928706,
+      "loss": 1.6111,
+      "step": 19590
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.14817097783088684,
+      "learning_rate": 0.000349894810592245,
+      "loss": 1.6463,
+      "step": 19600
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.12815338373184204,
+      "learning_rate": 0.00034984592276547535,
+      "loss": 1.5745,
+      "step": 19610
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1392802745103836,
+      "learning_rate": 0.0003497970145192234,
+      "loss": 1.6261,
+      "step": 19620
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.14371749758720398,
+      "learning_rate": 0.00034974808586015404,
+      "loss": 1.5905,
+      "step": 19630
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.13723789155483246,
+      "learning_rate": 0.0003496991367949347,
+      "loss": 1.5603,
+      "step": 19640
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.14505890011787415,
+      "learning_rate": 0.00034965016733023557,
+      "loss": 1.6496,
+      "step": 19650
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.13015948235988617,
+      "learning_rate": 0.0003496011774727299,
+      "loss": 1.5698,
+      "step": 19660
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.13103990256786346,
+      "learning_rate": 0.0003495521672290934,
+      "loss": 1.6192,
+      "step": 19670
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.11489203572273254,
+      "learning_rate": 0.00034950313660600475,
+      "loss": 1.6,
+      "step": 19680
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.137835294008255,
+      "learning_rate": 0.0003494540856101453,
+      "loss": 1.599,
+      "step": 19690
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1199876219034195,
+      "learning_rate": 0.0003494050142481992,
+      "loss": 1.501,
+      "step": 19700
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.13219058513641357,
+      "learning_rate": 0.0003493559225268535,
+      "loss": 1.5476,
+      "step": 19710
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.13448412716388702,
+      "learning_rate": 0.00034930681045279783,
+      "loss": 1.6071,
+      "step": 19720
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.11972352862358093,
+      "learning_rate": 0.0003492576780327247,
+      "loss": 1.5649,
+      "step": 19730
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.12670572102069855,
+      "learning_rate": 0.00034920852527332944,
+      "loss": 1.5484,
+      "step": 19740
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.12329743057489395,
+      "learning_rate": 0.00034915935218131003,
+      "loss": 1.5431,
+      "step": 19750
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.14874130487442017,
+      "learning_rate": 0.0003491101587633673,
+      "loss": 1.5921,
+      "step": 19760
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.14924252033233643,
+      "learning_rate": 0.0003490609450262048,
+      "loss": 1.6466,
+      "step": 19770
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.13762734830379486,
+      "learning_rate": 0.00034901171097652894,
+      "loss": 1.602,
+      "step": 19780
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.12964162230491638,
+      "learning_rate": 0.0003489624566210488,
+      "loss": 1.5935,
+      "step": 19790
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.12984727323055267,
+      "learning_rate": 0.00034891318196647613,
+      "loss": 1.5959,
+      "step": 19800
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1572873443365097,
+      "learning_rate": 0.00034886388701952585,
+      "loss": 1.6094,
+      "step": 19810
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.13977202773094177,
+      "learning_rate": 0.0003488145717869152,
+      "loss": 1.5617,
+      "step": 19820
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.13012699782848358,
+      "learning_rate": 0.0003487652362753644,
+      "loss": 1.5929,
+      "step": 19830
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.12909632921218872,
+      "learning_rate": 0.00034871588049159645,
+      "loss": 1.6072,
+      "step": 19840
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.12324431538581848,
+      "learning_rate": 0.0003486665044423369,
+      "loss": 1.5448,
+      "step": 19850
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.13812634348869324,
+      "learning_rate": 0.00034861710813431434,
+      "loss": 1.573,
+      "step": 19860
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.13106317818164825,
+      "learning_rate": 0.00034856769157426,
+      "loss": 1.5753,
+      "step": 19870
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.14410927891731262,
+      "learning_rate": 0.0003485182547689078,
+      "loss": 1.5802,
+      "step": 19880
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.14472812414169312,
+      "learning_rate": 0.0003484687977249946,
+      "loss": 1.6134,
+      "step": 19890
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.13466855883598328,
+      "learning_rate": 0.0003484193204492598,
+      "loss": 1.6461,
+      "step": 19900
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.12836550176143646,
+      "learning_rate": 0.0003483698229484456,
+      "loss": 1.5641,
+      "step": 19910
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.13823148608207703,
+      "learning_rate": 0.00034832030522929724,
+      "loss": 1.574,
+      "step": 19920
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.14605990052223206,
+      "learning_rate": 0.0003482707672985623,
+      "loss": 1.6078,
+      "step": 19930
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.13076242804527283,
+      "learning_rate": 0.0003482212091629913,
+      "loss": 1.5609,
+      "step": 19940
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.13470326364040375,
+      "learning_rate": 0.0003481716308293377,
+      "loss": 1.6432,
+      "step": 19950
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.12680357694625854,
+      "learning_rate": 0.0003481220323043574,
+      "loss": 1.5688,
+      "step": 19960
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.20143021643161774,
+      "learning_rate": 0.00034807241359480917,
+      "loss": 1.592,
+      "step": 19970
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.14630061388015747,
+      "learning_rate": 0.00034802277470745457,
+      "loss": 1.5627,
+      "step": 19980
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.14682652056217194,
+      "learning_rate": 0.0003479731156490579,
+      "loss": 1.5707,
+      "step": 19990
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1285618245601654,
+      "learning_rate": 0.00034792343642638617,
+      "loss": 1.5892,
+      "step": 20000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 85104,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 1000,
+  "total_flos": 1.6916265566208e+19,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}