diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12478 @@
+{
+  "best_metric": 0.9988272087568413,
+  "best_model_checkpoint": "neunit-ks-kangyuan0601/checkpoint-20680",
+  "epoch": 4.999093491267299,
+  "global_step": 20680,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4506769825918764e-07,
+      "loss": 3.6087,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.901353965183753e-07,
+      "loss": 3.6067,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3520309477756283e-07,
+      "loss": 3.6018,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.802707930367505e-07,
+      "loss": 3.5874,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.253384912959381e-07,
+      "loss": 3.5679,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.704061895551257e-07,
+      "loss": 3.533,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0154738878143134e-06,
+      "loss": 3.494,
+      "step": 70
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.160541586073501e-06,
+      "loss": 3.4492,
+      "step": 80
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3056092843326885e-06,
+      "loss": 3.3843,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4506769825918762e-06,
+      "loss": 3.3302,
+      "step": 100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.5957446808510639e-06,
+      "loss": 3.234,
+      "step": 110
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.7408123791102513e-06,
+      "loss": 3.1428,
+      "step": 120
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.8858800773694392e-06,
+      "loss": 3.0476,
+      "step": 130
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.030947775628627e-06,
+      "loss": 2.9635,
+      "step": 140
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.1760154738878143e-06,
+      "loss": 2.8853,
+      "step": 150
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.321083172147002e-06,
+      "loss": 2.774,
+      "step": 160
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.4661508704061896e-06,
+      "loss": 2.6864,
+      "step": 170
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.611218568665377e-06,
+      "loss": 2.5956,
+      "step": 180
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.7562862669245645e-06,
+      "loss": 2.5463,
+      "step": 190
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.9013539651837524e-06,
+      "loss": 2.5028,
+      "step": 200
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 3.0464216634429403e-06,
+      "loss": 2.4588,
+      "step": 210
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 3.1914893617021277e-06,
+      "loss": 2.423,
+      "step": 220
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.336557059961315e-06,
+      "loss": 2.3541,
+      "step": 230
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.4816247582205027e-06,
+      "loss": 2.3408,
+      "step": 240
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.626692456479691e-06,
+      "loss": 2.2828,
+      "step": 250
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.7717601547388784e-06,
+      "loss": 2.2949,
+      "step": 260
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.916827852998066e-06,
+      "loss": 2.2277,
+      "step": 270
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.061895551257254e-06,
+      "loss": 2.2131,
+      "step": 280
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.206963249516441e-06,
+      "loss": 2.1815,
+      "step": 290
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.352030947775629e-06,
+      "loss": 2.1323,
+      "step": 300
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.497098646034816e-06,
+      "loss": 2.1039,
+      "step": 310
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.642166344294004e-06,
+      "loss": 2.0783,
+      "step": 320
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.787234042553192e-06,
+      "loss": 2.0823,
+      "step": 330
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.932301740812379e-06,
+      "loss": 2.0488,
+      "step": 340
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.077369439071567e-06,
+      "loss": 2.0152,
+      "step": 350
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.222437137330754e-06,
+      "loss": 1.9838,
+      "step": 360
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.367504835589942e-06,
+      "loss": 1.9098,
+      "step": 370
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.512572533849129e-06,
+      "loss": 1.9469,
+      "step": 380
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.657640232108317e-06,
+      "loss": 1.7973,
+      "step": 390
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.802707930367505e-06,
+      "loss": 1.8763,
+      "step": 400
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.947775628626693e-06,
+      "loss": 1.8339,
+      "step": 410
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 6.092843326885881e-06,
+      "loss": 1.7809,
+      "step": 420
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 6.237911025145068e-06,
+      "loss": 1.7618,
+      "step": 430
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 6.3829787234042555e-06,
+      "loss": 1.7279,
+      "step": 440
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 6.528046421663443e-06,
+      "loss": 1.6877,
+      "step": 450
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 6.67311411992263e-06,
+      "loss": 1.6922,
+      "step": 460
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 6.818181818181818e-06,
+      "loss": 1.7102,
+      "step": 470
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 6.963249516441005e-06,
+      "loss": 1.6482,
+      "step": 480
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 7.108317214700193e-06,
+      "loss": 1.5763,
+      "step": 490
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 7.253384912959382e-06,
+      "loss": 1.6417,
+      "step": 500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 7.398452611218569e-06,
+      "loss": 1.5569,
+      "step": 510
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 7.543520309477757e-06,
+      "loss": 1.647,
+      "step": 520
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 7.688588007736945e-06,
+      "loss": 1.5256,
+      "step": 530
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 7.833655705996132e-06,
+      "loss": 1.581,
+      "step": 540
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 7.978723404255319e-06,
+      "loss": 1.5362,
+      "step": 550
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 8.123791102514507e-06,
+      "loss": 1.4984,
+      "step": 560
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 8.268858800773694e-06,
+      "loss": 1.4771,
+      "step": 570
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 8.413926499032882e-06,
+      "loss": 1.448,
+      "step": 580
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 8.558994197292069e-06,
+      "loss": 1.3807,
+      "step": 590
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 8.704061895551257e-06,
+      "loss": 1.4097,
+      "step": 600
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 8.849129593810444e-06,
+      "loss": 1.4087,
+      "step": 610
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 8.994197292069631e-06,
+      "loss": 1.4188,
+      "step": 620
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.139264990328822e-06,
+      "loss": 1.3713,
+      "step": 630
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.284332688588009e-06,
+      "loss": 1.372,
+      "step": 640
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.429400386847196e-06,
+      "loss": 1.2908,
+      "step": 650
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.574468085106385e-06,
+      "loss": 1.2895,
+      "step": 660
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.719535783365572e-06,
+      "loss": 1.3011,
+      "step": 670
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.864603481624759e-06,
+      "loss": 1.3462,
+      "step": 680
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.0009671179883946e-05,
+      "loss": 1.2927,
+      "step": 690
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.0154738878143134e-05,
+      "loss": 1.3235,
+      "step": 700
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.0299806576402321e-05,
+      "loss": 1.377,
+      "step": 710
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.0444874274661508e-05,
+      "loss": 1.2689,
+      "step": 720
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.0589941972920697e-05,
+      "loss": 1.205,
+      "step": 730
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.0735009671179884e-05,
+      "loss": 1.2214,
+      "step": 740
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.0880077369439071e-05,
+      "loss": 1.159,
+      "step": 750
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.1025145067698258e-05,
+      "loss": 1.192,
+      "step": 760
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.1170212765957447e-05,
+      "loss": 1.2057,
+      "step": 770
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.1315280464216634e-05,
+      "loss": 1.2375,
+      "step": 780
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.1460348162475821e-05,
+      "loss": 1.1799,
+      "step": 790
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.160541586073501e-05,
+      "loss": 1.1789,
+      "step": 800
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.1750483558994198e-05,
+      "loss": 1.2571,
+      "step": 810
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.1895551257253385e-05,
+      "loss": 1.2166,
+      "step": 820
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.2040618955512574e-05,
+      "loss": 1.1872,
+      "step": 830
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.2185686653771761e-05,
+      "loss": 1.0994,
+      "step": 840
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.2330754352030948e-05,
+      "loss": 1.2466,
+      "step": 850
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.2475822050290135e-05,
+      "loss": 1.1409,
+      "step": 860
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.2620889748549324e-05,
+      "loss": 1.2037,
+      "step": 870
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.2765957446808511e-05,
+      "loss": 1.1653,
+      "step": 880
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.2911025145067698e-05,
+      "loss": 1.18,
+      "step": 890
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.3056092843326887e-05,
+      "loss": 1.0836,
+      "step": 900
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.3201160541586074e-05,
+      "loss": 1.0508,
+      "step": 910
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.334622823984526e-05,
+      "loss": 1.1853,
+      "step": 920
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.349129593810445e-05,
+      "loss": 1.1426,
+      "step": 930
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 1.109,
+      "step": 940
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.3781431334622824e-05,
+      "loss": 1.1112,
+      "step": 950
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.392649903288201e-05,
+      "loss": 1.0941,
+      "step": 960
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.40715667311412e-05,
+      "loss": 1.1127,
+      "step": 970
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.4216634429400386e-05,
+      "loss": 1.0519,
+      "step": 980
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.4361702127659575e-05,
+      "loss": 1.0951,
+      "step": 990
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.4506769825918764e-05,
+      "loss": 1.1315,
+      "step": 1000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.465183752417795e-05,
+      "loss": 1.134,
+      "step": 1010
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.4796905222437138e-05,
+      "loss": 1.2437,
+      "step": 1020
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.4941972920696325e-05,
+      "loss": 1.074,
+      "step": 1030
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.5087040618955514e-05,
+      "loss": 1.0936,
+      "step": 1040
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.5232108317214699e-05,
+      "loss": 1.0643,
+      "step": 1050
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.537717601547389e-05,
+      "loss": 1.0754,
+      "step": 1060
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.5522243713733078e-05,
+      "loss": 1.1271,
+      "step": 1070
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.5667311411992263e-05,
+      "loss": 1.1107,
+      "step": 1080
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.5812379110251452e-05,
+      "loss": 1.0862,
+      "step": 1090
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.5957446808510637e-05,
+      "loss": 1.1167,
+      "step": 1100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.6102514506769826e-05,
+      "loss": 1.0312,
+      "step": 1110
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.6247582205029015e-05,
+      "loss": 1.0372,
+      "step": 1120
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.6392649903288204e-05,
+      "loss": 1.0742,
+      "step": 1130
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.653771760154739e-05,
+      "loss": 1.0241,
+      "step": 1140
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.6682785299806578e-05,
+      "loss": 1.0362,
+      "step": 1150
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.6827852998065763e-05,
+      "loss": 1.017,
+      "step": 1160
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.6972920696324952e-05,
+      "loss": 1.0495,
+      "step": 1170
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7117988394584137e-05,
+      "loss": 0.9279,
+      "step": 1180
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.726305609284333e-05,
+      "loss": 0.9946,
+      "step": 1190
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7408123791102515e-05,
+      "loss": 1.0187,
+      "step": 1200
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7553191489361703e-05,
+      "loss": 0.9941,
+      "step": 1210
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.769825918762089e-05,
+      "loss": 1.0503,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7843326885880077e-05,
+      "loss": 0.9335,
+      "step": 1230
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7988394584139263e-05,
+      "loss": 0.995,
+      "step": 1240
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.813346228239845e-05,
+      "loss": 0.9537,
+      "step": 1250
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.8278529980657643e-05,
+      "loss": 0.9344,
+      "step": 1260
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.842359767891683e-05,
+      "loss": 0.9847,
+      "step": 1270
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.8568665377176018e-05,
+      "loss": 0.9047,
+      "step": 1280
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.8713733075435203e-05,
+      "loss": 0.9094,
+      "step": 1290
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.885880077369439e-05,
+      "loss": 0.9398,
+      "step": 1300
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9003868471953577e-05,
+      "loss": 0.8791,
+      "step": 1310
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.914893617021277e-05,
+      "loss": 0.964,
+      "step": 1320
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9294003868471954e-05,
+      "loss": 0.9511,
+      "step": 1330
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9439071566731143e-05,
+      "loss": 0.9397,
+      "step": 1340
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.958413926499033e-05,
+      "loss": 1.0304,
+      "step": 1350
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9729206963249517e-05,
+      "loss": 0.9152,
+      "step": 1360
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9874274661508702e-05,
+      "loss": 0.9074,
+      "step": 1370
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.001934235976789e-05,
+      "loss": 0.8923,
+      "step": 1380
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.016441005802708e-05,
+      "loss": 0.8856,
+      "step": 1390
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.030947775628627e-05,
+      "loss": 0.9185,
+      "step": 1400
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.0454545454545454e-05,
+      "loss": 0.8707,
+      "step": 1410
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.0599613152804643e-05,
+      "loss": 0.8677,
+      "step": 1420
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.074468085106383e-05,
+      "loss": 0.8575,
+      "step": 1430
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.0889748549323017e-05,
+      "loss": 0.8805,
+      "step": 1440
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.1034816247582205e-05,
+      "loss": 0.8959,
+      "step": 1450
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.1179883945841394e-05,
+      "loss": 0.9332,
+      "step": 1460
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.1324951644100583e-05,
+      "loss": 0.8593,
+      "step": 1470
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.1470019342359768e-05,
+      "loss": 0.912,
+      "step": 1480
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.1615087040618957e-05,
+      "loss": 0.8355,
+      "step": 1490
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.1760154738878142e-05,
+      "loss": 0.8022,
+      "step": 1500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.190522243713733e-05,
+      "loss": 0.8008,
+      "step": 1510
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.2050290135396516e-05,
+      "loss": 0.8536,
+      "step": 1520
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.219535783365571e-05,
+      "loss": 0.8341,
+      "step": 1530
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.2340425531914894e-05,
+      "loss": 0.8361,
+      "step": 1540
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.2485493230174083e-05,
+      "loss": 0.8543,
+      "step": 1550
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.2630560928433268e-05,
+      "loss": 0.8018,
+      "step": 1560
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.2775628626692457e-05,
+      "loss": 0.7866,
+      "step": 1570
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.2920696324951642e-05,
+      "loss": 0.7149,
+      "step": 1580
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.3065764023210834e-05,
+      "loss": 0.728,
+      "step": 1590
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.321083172147002e-05,
+      "loss": 0.8033,
+      "step": 1600
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.3355899419729208e-05,
+      "loss": 0.8056,
+      "step": 1610
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.3500967117988397e-05,
+      "loss": 0.7153,
+      "step": 1620
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.3646034816247582e-05,
+      "loss": 0.7266,
+      "step": 1630
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.379110251450677e-05,
+      "loss": 0.7097,
+      "step": 1640
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.3936170212765956e-05,
+      "loss": 0.6806,
+      "step": 1650
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.408123791102515e-05,
+      "loss": 0.7037,
+      "step": 1660
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.4226305609284334e-05,
+      "loss": 0.7074,
+      "step": 1670
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.4371373307543522e-05,
+      "loss": 0.7155,
+      "step": 1680
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.4516441005802708e-05,
+      "loss": 0.6867,
+      "step": 1690
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.4661508704061896e-05,
+      "loss": 0.6751,
+      "step": 1700
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.4806576402321082e-05,
+      "loss": 0.7068,
+      "step": 1710
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.495164410058027e-05,
+      "loss": 0.6911,
+      "step": 1720
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.509671179883946e-05,
+      "loss": 0.6477,
+      "step": 1730
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.5241779497098648e-05,
+      "loss": 0.6692,
+      "step": 1740
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.5386847195357833e-05,
+      "loss": 0.6552,
+      "step": 1750
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.5531914893617022e-05,
+      "loss": 0.6731,
+      "step": 1760
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.5676982591876207e-05,
+      "loss": 0.6242,
+      "step": 1770
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.5822050290135396e-05,
+      "loss": 0.6043,
+      "step": 1780
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.5967117988394585e-05,
+      "loss": 0.6808,
+      "step": 1790
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.6112185686653773e-05,
+      "loss": 0.5857,
+      "step": 1800
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.6257253384912962e-05,
+      "loss": 0.6091,
+      "step": 1810
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.6402321083172148e-05,
+      "loss": 0.572,
+      "step": 1820
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.6547388781431336e-05,
+      "loss": 0.601,
+      "step": 1830
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.669245647969052e-05,
+      "loss": 0.559,
+      "step": 1840
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.683752417794971e-05,
+      "loss": 0.6194,
+      "step": 1850
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.69825918762089e-05,
+      "loss": 0.5725,
+      "step": 1860
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.7127659574468088e-05,
+      "loss": 0.5407,
+      "step": 1870
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.7272727272727273e-05,
+      "loss": 0.6064,
+      "step": 1880
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.7417794970986462e-05,
+      "loss": 0.6159,
+      "step": 1890
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.7562862669245647e-05,
+      "loss": 0.546,
+      "step": 1900
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.7707930367504836e-05,
+      "loss": 0.5221,
+      "step": 1910
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.785299806576402e-05,
+      "loss": 0.5245,
+      "step": 1920
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.7998065764023213e-05,
+      "loss": 0.5209,
+      "step": 1930
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.81431334622824e-05,
+      "loss": 0.5455,
+      "step": 1940
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.8288201160541587e-05,
+      "loss": 0.5334,
+      "step": 1950
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.8433268858800773e-05,
+      "loss": 0.4948,
+      "step": 1960
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.857833655705996e-05,
+      "loss": 0.554,
+      "step": 1970
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.872340425531915e-05,
+      "loss": 0.5556,
+      "step": 1980
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8868471953578336e-05,
+      "loss": 0.4988,
+      "step": 1990
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.9013539651837528e-05,
+      "loss": 0.4764,
+      "step": 2000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.9158607350096713e-05,
+      "loss": 0.473,
+      "step": 2010
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.93036750483559e-05,
+      "loss": 0.486,
+      "step": 2020
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.9448742746615087e-05,
+      "loss": 0.4826,
+      "step": 2030
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.9593810444874276e-05,
+      "loss": 0.4774,
+      "step": 2040
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.973887814313346e-05,
+      "loss": 0.5261,
+      "step": 2050
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.988394584139265e-05,
+      "loss": 0.4329,
+      "step": 2060
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.9996776273372018e-05,
+      "loss": 0.47,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.9980657640232108e-05,
+      "loss": 0.4458,
+      "step": 2080
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.99645390070922e-05,
+      "loss": 0.4636,
+      "step": 2090
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.994842037395229e-05,
+      "loss": 0.451,
+      "step": 2100
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.993230174081238e-05,
+      "loss": 0.4206,
+      "step": 2110
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.991618310767247e-05,
+      "loss": 0.439,
+      "step": 2120
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.990006447453256e-05,
+      "loss": 0.3792,
+      "step": 2130
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.988394584139265e-05,
+      "loss": 0.4408,
+      "step": 2140
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.9867827208252743e-05,
+      "loss": 0.4087,
+      "step": 2150
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.9851708575112833e-05,
+      "loss": 0.4033,
+      "step": 2160
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.9835589941972922e-05,
+      "loss": 0.3655,
+      "step": 2170
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.9819471308833012e-05,
+      "loss": 0.3486,
+      "step": 2180
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.9803352675693102e-05,
+      "loss": 0.3835,
+      "step": 2190
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.9787234042553192e-05,
+      "loss": 0.409,
+      "step": 2200
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.977111540941328e-05,
+      "loss": 0.3398,
+      "step": 2210
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.975499677627337e-05,
+      "loss": 0.3848,
+      "step": 2220
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.973887814313346e-05,
+      "loss": 0.3652,
+      "step": 2230
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.972275950999355e-05,
+      "loss": 0.4004,
+      "step": 2240
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.9708252740167635e-05,
+      "loss": 0.3353,
+      "step": 2250
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.9692134107027725e-05,
+      "loss": 0.3491,
+      "step": 2260
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.9676015473887815e-05,
+      "loss": 0.332,
+      "step": 2270
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.9659896840747905e-05,
+      "loss": 0.3889,
+      "step": 2280
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.9643778207607994e-05,
+      "loss": 0.3274,
+      "step": 2290
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.9627659574468084e-05,
+      "loss": 0.3444,
+      "step": 2300
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.9611540941328174e-05,
+      "loss": 0.3087,
+      "step": 2310
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.9595422308188267e-05,
+      "loss": 0.3065,
+      "step": 2320
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.9579303675048357e-05,
+      "loss": 0.3383,
+      "step": 2330
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.9563185041908447e-05,
+      "loss": 0.3429,
+      "step": 2340
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.954706640876854e-05,
+      "loss": 0.2564,
+      "step": 2350
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.953094777562863e-05,
+      "loss": 0.3202,
+      "step": 2360
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.951482914248872e-05,
+      "loss": 0.2857,
+      "step": 2370
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.949871050934881e-05,
+      "loss": 0.3063,
+      "step": 2380
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.94825918762089e-05,
+      "loss": 0.28,
+      "step": 2390
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.946647324306899e-05,
+      "loss": 0.3035,
+      "step": 2400
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.945035460992908e-05,
+      "loss": 0.2834,
+      "step": 2410
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.9434235976789168e-05,
+      "loss": 0.2938,
+      "step": 2420
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.9418117343649258e-05,
+      "loss": 0.3124,
+      "step": 2430
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.9401998710509348e-05,
+      "loss": 0.2542,
+      "step": 2440
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.938588007736944e-05,
+      "loss": 0.25,
+      "step": 2450
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.936976144422953e-05,
+      "loss": 0.268,
+      "step": 2460
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.935364281108962e-05,
+      "loss": 0.2995,
+      "step": 2470
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.933752417794971e-05,
+      "loss": 0.2579,
+      "step": 2480
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.93214055448098e-05,
+      "loss": 0.2694,
+      "step": 2490
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.930528691166989e-05,
+      "loss": 0.2524,
+      "step": 2500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.9289168278529983e-05,
+      "loss": 0.2575,
+      "step": 2510
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.9273049645390073e-05,
+      "loss": 0.2328,
+      "step": 2520
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.9256931012250162e-05,
+      "loss": 0.2604,
+      "step": 2530
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.9240812379110252e-05,
+      "loss": 0.2066,
+      "step": 2540
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.9224693745970342e-05,
+      "loss": 0.2461,
+      "step": 2550
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.9208575112830435e-05,
+      "loss": 0.2191,
+      "step": 2560
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.9192456479690525e-05,
+      "loss": 0.231,
+      "step": 2570
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.9176337846550615e-05,
+      "loss": 0.2422,
+      "step": 2580
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.9160219213410704e-05,
+      "loss": 0.207,
+      "step": 2590
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.9144100580270794e-05,
+      "loss": 0.2196,
+      "step": 2600
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.9127981947130884e-05,
+      "loss": 0.2157,
+      "step": 2610
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.9111863313990974e-05,
+      "loss": 0.2317,
+      "step": 2620
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.9095744680851063e-05,
+      "loss": 0.1931,
+      "step": 2630
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.9079626047711153e-05,
+      "loss": 0.2088,
+      "step": 2640
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.9063507414571243e-05,
+      "loss": 0.2392,
+      "step": 2650
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.9047388781431336e-05,
+      "loss": 0.1929,
+      "step": 2660
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.9031270148291426e-05,
+      "loss": 0.1762,
+      "step": 2670
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.9015151515151516e-05,
+      "loss": 0.2501,
+      "step": 2680
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.8999032882011605e-05,
+      "loss": 0.2149,
+      "step": 2690
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.89829142488717e-05,
+      "loss": 0.2047,
+      "step": 2700
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.8966795615731788e-05,
+      "loss": 0.2281,
+      "step": 2710
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.8950676982591878e-05,
+      "loss": 0.2038,
+      "step": 2720
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.8934558349451968e-05,
+      "loss": 0.205,
+      "step": 2730
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.8918439716312058e-05,
+      "loss": 0.2023,
+      "step": 2740
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.8902321083172147e-05,
+      "loss": 0.1989,
+      "step": 2750
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.8886202450032237e-05,
+      "loss": 0.2024,
+      "step": 2760
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.887008381689233e-05,
+      "loss": 0.2065,
+      "step": 2770
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.885396518375242e-05,
+      "loss": 0.2134,
+      "step": 2780
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.883784655061251e-05,
+      "loss": 0.1969,
+      "step": 2790
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.88217279174726e-05,
+      "loss": 0.1821,
+      "step": 2800
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.880560928433269e-05,
+      "loss": 0.1396,
+      "step": 2810
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.878949065119278e-05,
+      "loss": 0.1946,
+      "step": 2820
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.877337201805287e-05,
+      "loss": 0.1846,
+      "step": 2830
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.875725338491296e-05,
+      "loss": 0.1653,
+      "step": 2840
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.874113475177305e-05,
+      "loss": 0.1818,
+      "step": 2850
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.8725016118633138e-05,
+      "loss": 0.1842,
+      "step": 2860
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.8708897485493228e-05,
+      "loss": 0.1472,
+      "step": 2870
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.869277885235332e-05,
+      "loss": 0.1975,
+      "step": 2880
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.8676660219213414e-05,
+      "loss": 0.1571,
+      "step": 2890
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.8660541586073504e-05,
+      "loss": 0.1703,
+      "step": 2900
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.8644422952933594e-05,
+      "loss": 0.1569,
+      "step": 2910
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.8628304319793684e-05,
+      "loss": 0.1929,
+      "step": 2920
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.8612185686653773e-05,
+      "loss": 0.209,
+      "step": 2930
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.8596067053513863e-05,
+      "loss": 0.1483,
+      "step": 2940
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.8579948420373953e-05,
+      "loss": 0.1768,
+      "step": 2950
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.8563829787234043e-05,
+      "loss": 0.181,
+      "step": 2960
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.8547711154094132e-05,
+      "loss": 0.1773,
+      "step": 2970
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.8531592520954225e-05,
+      "loss": 0.1563,
+      "step": 2980
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.8515473887814315e-05,
+      "loss": 0.1649,
+      "step": 2990
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.8499355254674405e-05,
+      "loss": 0.1546,
+      "step": 3000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.8483236621534495e-05,
+      "loss": 0.1981,
+      "step": 3010
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.8467117988394585e-05,
+      "loss": 0.1721,
+      "step": 3020
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.8450999355254674e-05,
+      "loss": 0.1717,
+      "step": 3030
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.8434880722114764e-05,
+      "loss": 0.1488,
+      "step": 3040
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.8418762088974854e-05,
+      "loss": 0.1311,
+      "step": 3050
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.8402643455834944e-05,
+      "loss": 0.1385,
+      "step": 3060
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.8386524822695033e-05,
+      "loss": 0.151,
+      "step": 3070
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.8370406189555127e-05,
+      "loss": 0.1748,
+      "step": 3080
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.835428755641522e-05,
+      "loss": 0.1518,
+      "step": 3090
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.833816892327531e-05,
+      "loss": 0.1402,
+      "step": 3100
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.83220502901354e-05,
+      "loss": 0.1584,
+      "step": 3110
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.830593165699549e-05,
+      "loss": 0.1291,
+      "step": 3120
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.828981302385558e-05,
+      "loss": 0.1423,
+      "step": 3130
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.827369439071567e-05,
+      "loss": 0.1565,
+      "step": 3140
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.8257575757575758e-05,
+      "loss": 0.1375,
+      "step": 3150
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.8241457124435848e-05,
+      "loss": 0.1263,
+      "step": 3160
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.8225338491295938e-05,
+      "loss": 0.1056,
+      "step": 3170
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.8209219858156028e-05,
+      "loss": 0.1055,
+      "step": 3180
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.819310122501612e-05,
+      "loss": 0.1227,
+      "step": 3190
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.817698259187621e-05,
+      "loss": 0.1281,
+      "step": 3200
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.81608639587363e-05,
+      "loss": 0.1421,
+      "step": 3210
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.814474532559639e-05,
+      "loss": 0.1471,
+      "step": 3220
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.812862669245648e-05,
+      "loss": 0.1465,
+      "step": 3230
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.811250805931657e-05,
+      "loss": 0.1352,
+      "step": 3240
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.809638942617666e-05,
+      "loss": 0.1154,
+      "step": 3250
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.808027079303675e-05,
+      "loss": 0.1401,
+      "step": 3260
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.806415215989684e-05,
+      "loss": 0.1625,
+      "step": 3270
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.8048033526756932e-05,
+      "loss": 0.1149,
+      "step": 3280
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.8031914893617022e-05,
+      "loss": 0.1368,
+      "step": 3290
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.8015796260477115e-05,
+      "loss": 0.1046,
+      "step": 3300
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.7999677627337205e-05,
+      "loss": 0.1357,
+      "step": 3310
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.7983558994197294e-05,
+      "loss": 0.1578,
+      "step": 3320
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.7967440361057384e-05,
+      "loss": 0.1278,
+      "step": 3330
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.7951321727917474e-05,
+      "loss": 0.0944,
+      "step": 3340
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.7935203094777564e-05,
+      "loss": 0.1207,
+      "step": 3350
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.7919084461637654e-05,
+      "loss": 0.1361,
+      "step": 3360
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.7902965828497743e-05,
+      "loss": 0.1216,
+      "step": 3370
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.7886847195357833e-05,
+      "loss": 0.1405,
+      "step": 3380
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.7870728562217923e-05,
+      "loss": 0.1316,
+      "step": 3390
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.7854609929078016e-05,
+      "loss": 0.1217,
+      "step": 3400
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.7838491295938106e-05,
+      "loss": 0.1337,
+      "step": 3410
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.7822372662798196e-05,
+      "loss": 0.1465,
+      "step": 3420
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.7806254029658285e-05,
+      "loss": 0.1184,
+      "step": 3430
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.7790135396518375e-05,
+      "loss": 0.1374,
+      "step": 3440
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.7774016763378465e-05,
+      "loss": 0.0939,
+      "step": 3450
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.7757898130238555e-05,
+      "loss": 0.1354,
+      "step": 3460
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.7741779497098648e-05,
+      "loss": 0.1185,
+      "step": 3470
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.7725660863958737e-05,
+      "loss": 0.1118,
+      "step": 3480
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.7709542230818827e-05,
+      "loss": 0.1025,
+      "step": 3490
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.7693423597678917e-05,
+      "loss": 0.1107,
+      "step": 3500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.767730496453901e-05,
+      "loss": 0.1112,
+      "step": 3510
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.76611863313991e-05,
+      "loss": 0.1111,
+      "step": 3520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.764506769825919e-05,
+      "loss": 0.107,
+      "step": 3530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.762894906511928e-05,
+      "loss": 0.0883,
+      "step": 3540
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.761283043197937e-05,
+      "loss": 0.1114,
+      "step": 3550
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.759671179883946e-05,
+      "loss": 0.1177,
+      "step": 3560
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.758059316569955e-05,
+      "loss": 0.0831,
+      "step": 3570
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.756447453255964e-05,
+      "loss": 0.1584,
+      "step": 3580
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.7548355899419728e-05,
+      "loss": 0.1379,
+      "step": 3590
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.7532237266279818e-05,
+      "loss": 0.1128,
+      "step": 3600
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.7516118633139908e-05,
+      "loss": 0.1372,
+      "step": 3610
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.75e-05,
+      "loss": 0.1217,
+      "step": 3620
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.748388136686009e-05,
+      "loss": 0.0952,
+      "step": 3630
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.746776273372018e-05,
+      "loss": 0.0875,
+      "step": 3640
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.745164410058027e-05,
+      "loss": 0.1204,
+      "step": 3650
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.7435525467440363e-05,
+      "loss": 0.1175,
+      "step": 3660
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.7419406834300453e-05,
+      "loss": 0.1074,
+      "step": 3670
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.7403288201160543e-05,
+      "loss": 0.1098,
+      "step": 3680
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.7387169568020633e-05,
+      "loss": 0.1184,
+      "step": 3690
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.7371050934880722e-05,
+      "loss": 0.1084,
+      "step": 3700
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.7354932301740812e-05,
+      "loss": 0.0828,
+      "step": 3710
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.7338813668600905e-05,
+      "loss": 0.1236,
+      "step": 3720
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.7322695035460995e-05,
+      "loss": 0.1141,
+      "step": 3730
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.7306576402321085e-05,
+      "loss": 0.1277,
+      "step": 3740
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.7290457769181175e-05,
+      "loss": 0.139,
+      "step": 3750
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.7274339136041264e-05,
+      "loss": 0.0894,
+      "step": 3760
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.7258220502901354e-05,
+      "loss": 0.0954,
+      "step": 3770
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.7242101869761444e-05,
+      "loss": 0.139,
+      "step": 3780
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.7225983236621534e-05,
+      "loss": 0.0857,
+      "step": 3790
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.7209864603481624e-05,
+      "loss": 0.0793,
+      "step": 3800
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.7193745970341713e-05,
+      "loss": 0.0974,
+      "step": 3810
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.7177627337201803e-05,
+      "loss": 0.1193,
+      "step": 3820
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7161508704061896e-05,
+      "loss": 0.0936,
+      "step": 3830
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7145390070921986e-05,
+      "loss": 0.0932,
+      "step": 3840
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.712927143778208e-05,
+      "loss": 0.0858,
+      "step": 3850
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.711315280464217e-05,
+      "loss": 0.1032,
+      "step": 3860
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.709703417150226e-05,
+      "loss": 0.0862,
+      "step": 3870
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.708091553836235e-05,
+      "loss": 0.1423,
+      "step": 3880
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.7064796905222438e-05,
+      "loss": 0.0898,
+      "step": 3890
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.7048678272082528e-05,
+      "loss": 0.0934,
+      "step": 3900
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.7032559638942618e-05,
+      "loss": 0.104,
+      "step": 3910
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.7016441005802707e-05,
+      "loss": 0.0779,
+      "step": 3920
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.70003223726628e-05,
+      "loss": 0.1145,
+      "step": 3930
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.698420373952289e-05,
+      "loss": 0.0845,
+      "step": 3940
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.696808510638298e-05,
+      "loss": 0.1204,
+      "step": 3950
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.695196647324307e-05,
+      "loss": 0.1277,
+      "step": 3960
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.693584784010316e-05,
+      "loss": 0.0718,
+      "step": 3970
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.691972920696325e-05,
+      "loss": 0.072,
+      "step": 3980
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.690361057382334e-05,
+      "loss": 0.0838,
+      "step": 3990
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.688749194068343e-05,
+      "loss": 0.0948,
+      "step": 4000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.687137330754352e-05,
+      "loss": 0.0873,
+      "step": 4010
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.685525467440361e-05,
+      "loss": 0.0782,
+      "step": 4020
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.6839136041263698e-05,
+      "loss": 0.111,
+      "step": 4030
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.6823017408123795e-05,
+      "loss": 0.1203,
+      "step": 4040
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.6806898774983885e-05,
+      "loss": 0.0857,
+      "step": 4050
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.6790780141843974e-05,
+      "loss": 0.1184,
+      "step": 4060
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.6774661508704064e-05,
+      "loss": 0.0804,
+      "step": 4070
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.6758542875564154e-05,
+      "loss": 0.0802,
+      "step": 4080
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.6742424242424244e-05,
+      "loss": 0.0873,
+      "step": 4090
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.6726305609284333e-05,
+      "loss": 0.1046,
+      "step": 4100
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.6710186976144423e-05,
+      "loss": 0.1051,
+      "step": 4110
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6694068343004513e-05,
+      "loss": 0.0932,
+      "step": 4120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6677949709864603e-05,
+      "loss": 0.0835,
+      "step": 4130
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.9877791753067954,
+      "eval_loss": 0.04899689182639122,
+      "eval_runtime": 69.0471,
+      "eval_samples_per_second": 852.085,
+      "eval_steps_per_second": 26.634,
+      "step": 4136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6661831076724696e-05,
+      "loss": 0.0987,
+      "step": 4140
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6645712443584786e-05,
+      "loss": 0.0993,
+      "step": 4150
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6629593810444875e-05,
+      "loss": 0.0853,
+      "step": 4160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6613475177304965e-05,
+      "loss": 0.0737,
+      "step": 4170
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6597356544165055e-05,
+      "loss": 0.0791,
+      "step": 4180
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6581237911025145e-05,
+      "loss": 0.1027,
+      "step": 4190
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6565119277885234e-05,
+      "loss": 0.07,
+      "step": 4200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6549000644745324e-05,
+      "loss": 0.0793,
+      "step": 4210
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6532882011605414e-05,
+      "loss": 0.0782,
+      "step": 4220
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6516763378465507e-05,
+      "loss": 0.0948,
+      "step": 4230
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6500644745325597e-05,
+      "loss": 0.0789,
+      "step": 4240
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.648452611218569e-05,
+      "loss": 0.0754,
+      "step": 4250
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.646840747904578e-05,
+      "loss": 0.0824,
+      "step": 4260
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.645228884590587e-05,
+      "loss": 0.0653,
+      "step": 4270
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.643617021276596e-05,
+      "loss": 0.0921,
+      "step": 4280
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.642005157962605e-05,
+      "loss": 0.0634,
+      "step": 4290
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.640393294648614e-05,
+      "loss": 0.0793,
+      "step": 4300
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.638781431334623e-05,
+      "loss": 0.0774,
+      "step": 4310
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.637169568020632e-05,
+      "loss": 0.0971,
+      "step": 4320
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.6355577047066408e-05,
+      "loss": 0.1067,
+      "step": 4330
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.6339458413926498e-05,
+      "loss": 0.1042,
+      "step": 4340
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.6323339780786588e-05,
+      "loss": 0.0839,
+      "step": 4350
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.630722114764668e-05,
+      "loss": 0.0651,
+      "step": 4360
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.629110251450677e-05,
+      "loss": 0.0914,
+      "step": 4370
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.627498388136686e-05,
+      "loss": 0.0999,
+      "step": 4380
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.625886524822695e-05,
+      "loss": 0.1092,
+      "step": 4390
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.624274661508704e-05,
+      "loss": 0.0882,
+      "step": 4400
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.622662798194713e-05,
+      "loss": 0.0824,
+      "step": 4410
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.6210509348807223e-05,
+      "loss": 0.0973,
+      "step": 4420
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.6194390715667313e-05,
+      "loss": 0.0702,
+      "step": 4430
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.6178272082527402e-05,
+      "loss": 0.0818,
+      "step": 4440
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.6162153449387492e-05,
+      "loss": 0.0758,
+      "step": 4450
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.6146034816247585e-05,
+      "loss": 0.1016,
+      "step": 4460
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.6131528046421666e-05,
+      "loss": 0.0847,
+      "step": 4470
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.6115409413281756e-05,
+      "loss": 0.084,
+      "step": 4480
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.6099290780141846e-05,
+      "loss": 0.0606,
+      "step": 4490
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.6083172147001936e-05,
+      "loss": 0.0985,
+      "step": 4500
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.6067053513862026e-05,
+      "loss": 0.0933,
+      "step": 4510
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.6050934880722115e-05,
+      "loss": 0.0718,
+      "step": 4520
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.6034816247582205e-05,
+      "loss": 0.075,
+      "step": 4530
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.6018697614442295e-05,
+      "loss": 0.081,
+      "step": 4540
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.6002578981302385e-05,
+      "loss": 0.0755,
+      "step": 4550
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.5986460348162478e-05,
+      "loss": 0.0645,
+      "step": 4560
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.5970341715022567e-05,
+      "loss": 0.0781,
+      "step": 4570
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.5954223081882657e-05,
+      "loss": 0.0895,
+      "step": 4580
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.5938104448742747e-05,
+      "loss": 0.0868,
+      "step": 4590
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.5921985815602837e-05,
+      "loss": 0.0663,
+      "step": 4600
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.5905867182462927e-05,
+      "loss": 0.0542,
+      "step": 4610
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.5889748549323016e-05,
+      "loss": 0.0644,
+      "step": 4620
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.5873629916183106e-05,
+      "loss": 0.0746,
+      "step": 4630
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.5857511283043196e-05,
+      "loss": 0.0872,
+      "step": 4640
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.584139264990329e-05,
+      "loss": 0.0922,
+      "step": 4650
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.5825274016763382e-05,
+      "loss": 0.0586,
+      "step": 4660
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.5809155383623472e-05,
+      "loss": 0.0825,
+      "step": 4670
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.5793036750483562e-05,
+      "loss": 0.0762,
+      "step": 4680
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.577691811734365e-05,
+      "loss": 0.0756,
+      "step": 4690
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.576079948420374e-05,
+      "loss": 0.0629,
+      "step": 4700
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.574468085106383e-05,
+      "loss": 0.0704,
+      "step": 4710
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.572856221792392e-05,
+      "loss": 0.0853,
+      "step": 4720
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.571244358478401e-05,
+      "loss": 0.0826,
+      "step": 4730
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.56963249516441e-05,
+      "loss": 0.0565,
+      "step": 4740
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.568020631850419e-05,
+      "loss": 0.0819,
+      "step": 4750
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.566408768536428e-05,
+      "loss": 0.0811,
+      "step": 4760
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.5647969052224373e-05,
+      "loss": 0.0883,
+      "step": 4770
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.5631850419084463e-05,
+      "loss": 0.0676,
+      "step": 4780
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.5615731785944553e-05,
+      "loss": 0.061,
+      "step": 4790
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.5599613152804642e-05,
+      "loss": 0.0552,
+      "step": 4800
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.5583494519664732e-05,
+      "loss": 0.0831,
+      "step": 4810
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.5567375886524822e-05,
+      "loss": 0.0736,
+      "step": 4820
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.555125725338491e-05,
+      "loss": 0.0854,
+      "step": 4830
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.5535138620245005e-05,
+      "loss": 0.0722,
+      "step": 4840
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.5519019987105094e-05,
+      "loss": 0.0531,
+      "step": 4850
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.5502901353965184e-05,
+      "loss": 0.089,
+      "step": 4860
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.5486782720825277e-05,
+      "loss": 0.0677,
+      "step": 4870
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.5470664087685367e-05,
+      "loss": 0.0754,
+      "step": 4880
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.5454545454545457e-05,
+      "loss": 0.0823,
+      "step": 4890
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.5438426821405547e-05,
+      "loss": 0.0837,
+      "step": 4900
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.5422308188265636e-05,
+      "loss": 0.0758,
+      "step": 4910
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.5406189555125726e-05,
+      "loss": 0.0912,
+      "step": 4920
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.5390070921985816e-05,
+      "loss": 0.0654,
+      "step": 4930
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.5373952288845906e-05,
+      "loss": 0.0663,
+      "step": 4940
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.5357833655705996e-05,
+      "loss": 0.0687,
+      "step": 4950
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.5341715022566085e-05,
+      "loss": 0.0558,
+      "step": 4960
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.5325596389426175e-05,
+      "loss": 0.0945,
+      "step": 4970
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.5309477756286268e-05,
+      "loss": 0.0789,
+      "step": 4980
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.5293359123146358e-05,
+      "loss": 0.068,
+      "step": 4990
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.5277240490006448e-05,
+      "loss": 0.0596,
+      "step": 5000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.5261121856866538e-05,
+      "loss": 0.0585,
+      "step": 5010
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.5245003223726627e-05,
+      "loss": 0.0446,
+      "step": 5020
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.522888459058672e-05,
+      "loss": 0.0515,
+      "step": 5030
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.521276595744681e-05,
+      "loss": 0.0785,
+      "step": 5040
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.51966473243069e-05,
+      "loss": 0.0382,
+      "step": 5050
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.518052869116699e-05,
+      "loss": 0.0446,
+      "step": 5060
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.516441005802708e-05,
+      "loss": 0.0681,
+      "step": 5070
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.514829142488717e-05,
+      "loss": 0.0535,
+      "step": 5080
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.5132172791747262e-05,
+      "loss": 0.0673,
+      "step": 5090
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.5116054158607352e-05,
+      "loss": 0.0647,
+      "step": 5100
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.5099935525467442e-05,
+      "loss": 0.0603,
+      "step": 5110
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.5083816892327532e-05,
+      "loss": 0.0586,
+      "step": 5120
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.506769825918762e-05,
+      "loss": 0.0628,
+      "step": 5130
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.505157962604771e-05,
+      "loss": 0.0574,
+      "step": 5140
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.50354609929078e-05,
+      "loss": 0.0778,
+      "step": 5150
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.501934235976789e-05,
+      "loss": 0.0621,
+      "step": 5160
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.500322372662798e-05,
+      "loss": 0.0552,
+      "step": 5170
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.498710509348807e-05,
+      "loss": 0.0524,
+      "step": 5180
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.4970986460348163e-05,
+      "loss": 0.0828,
+      "step": 5190
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.4954867827208253e-05,
+      "loss": 0.0666,
+      "step": 5200
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.4938749194068343e-05,
+      "loss": 0.0717,
+      "step": 5210
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.4922630560928436e-05,
+      "loss": 0.0697,
+      "step": 5220
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.4906511927788526e-05,
+      "loss": 0.0739,
+      "step": 5230
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.4890393294648616e-05,
+      "loss": 0.076,
+      "step": 5240
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.4874274661508705e-05,
+      "loss": 0.0429,
+      "step": 5250
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.4858156028368795e-05,
+      "loss": 0.0532,
+      "step": 5260
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.4842037395228885e-05,
+      "loss": 0.067,
+      "step": 5270
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.4825918762088975e-05,
+      "loss": 0.0578,
+      "step": 5280
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.4809800128949064e-05,
+      "loss": 0.0478,
+      "step": 5290
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.4793681495809158e-05,
+      "loss": 0.0626,
+      "step": 5300
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.4777562862669247e-05,
+      "loss": 0.0829,
+      "step": 5310
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.4761444229529337e-05,
+      "loss": 0.0554,
+      "step": 5320
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.4745325596389427e-05,
+      "loss": 0.0606,
+      "step": 5330
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.4729206963249517e-05,
+      "loss": 0.0536,
+      "step": 5340
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.4713088330109606e-05,
+      "loss": 0.0291,
+      "step": 5350
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.4696969696969696e-05,
+      "loss": 0.0499,
+      "step": 5360
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.4680851063829786e-05,
+      "loss": 0.0413,
+      "step": 5370
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.4664732430689876e-05,
+      "loss": 0.0436,
+      "step": 5380
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.4648613797549966e-05,
+      "loss": 0.0675,
+      "step": 5390
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.463249516441006e-05,
+      "loss": 0.065,
+      "step": 5400
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.4616376531270152e-05,
+      "loss": 0.0671,
+      "step": 5410
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.460025789813024e-05,
+      "loss": 0.0674,
+      "step": 5420
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.458413926499033e-05,
+      "loss": 0.0542,
+      "step": 5430
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.456802063185042e-05,
+      "loss": 0.0579,
+      "step": 5440
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.455190199871051e-05,
+      "loss": 0.0443,
+      "step": 5450
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.45357833655706e-05,
+      "loss": 0.0587,
+      "step": 5460
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.451966473243069e-05,
+      "loss": 0.0712,
+      "step": 5470
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.450354609929078e-05,
+      "loss": 0.0491,
+      "step": 5480
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.448742746615087e-05,
+      "loss": 0.0605,
+      "step": 5490
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.447130883301096e-05,
+      "loss": 0.0522,
+      "step": 5500
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.4455190199871053e-05,
+      "loss": 0.0435,
+      "step": 5510
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.4439071566731143e-05,
+      "loss": 0.0408,
+      "step": 5520
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.4422952933591232e-05,
+      "loss": 0.0625,
+      "step": 5530
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.4406834300451322e-05,
+      "loss": 0.068,
+      "step": 5540
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.4390715667311412e-05,
+      "loss": 0.0623,
+      "step": 5550
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.4374597034171502e-05,
+      "loss": 0.0433,
+      "step": 5560
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.435847840103159e-05,
+      "loss": 0.0493,
+      "step": 5570
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.434235976789168e-05,
+      "loss": 0.0527,
+      "step": 5580
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.432624113475177e-05,
+      "loss": 0.0664,
+      "step": 5590
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.4310122501611864e-05,
+      "loss": 0.0584,
+      "step": 5600
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.4294003868471957e-05,
+      "loss": 0.0756,
+      "step": 5610
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.4277885235332047e-05,
+      "loss": 0.0446,
+      "step": 5620
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.4261766602192137e-05,
+      "loss": 0.0413,
+      "step": 5630
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.4245647969052227e-05,
+      "loss": 0.0536,
+      "step": 5640
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.4229529335912316e-05,
+      "loss": 0.0683,
+      "step": 5650
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.4213410702772406e-05,
+      "loss": 0.0593,
+      "step": 5660
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.4197292069632496e-05,
+      "loss": 0.0553,
+      "step": 5670
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.4181173436492586e-05,
+      "loss": 0.0639,
+      "step": 5680
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.4165054803352675e-05,
+      "loss": 0.0483,
+      "step": 5690
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.4148936170212765e-05,
+      "loss": 0.0563,
+      "step": 5700
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.4132817537072855e-05,
+      "loss": 0.0787,
+      "step": 5710
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.4116698903932948e-05,
+      "loss": 0.0651,
+      "step": 5720
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.4100580270793038e-05,
+      "loss": 0.0598,
+      "step": 5730
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.4084461637653128e-05,
+      "loss": 0.0716,
+      "step": 5740
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.4068343004513217e-05,
+      "loss": 0.0439,
+      "step": 5750
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.4052224371373307e-05,
+      "loss": 0.0551,
+      "step": 5760
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.4036105738233397e-05,
+      "loss": 0.0424,
+      "step": 5770
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.4019987105093487e-05,
+      "loss": 0.0514,
+      "step": 5780
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.400386847195358e-05,
+      "loss": 0.0556,
+      "step": 5790
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.398774983881367e-05,
+      "loss": 0.0422,
+      "step": 5800
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.397163120567376e-05,
+      "loss": 0.0523,
+      "step": 5810
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.395551257253385e-05,
+      "loss": 0.0761,
+      "step": 5820
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.3939393939393942e-05,
+      "loss": 0.0588,
+      "step": 5830
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.3923275306254032e-05,
+      "loss": 0.0675,
+      "step": 5840
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.3907156673114122e-05,
+      "loss": 0.0523,
+      "step": 5850
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.389103803997421e-05,
+      "loss": 0.0463,
+      "step": 5860
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.38749194068343e-05,
+      "loss": 0.0575,
+      "step": 5870
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.385880077369439e-05,
+      "loss": 0.0521,
+      "step": 5880
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.384268214055448e-05,
+      "loss": 0.0488,
+      "step": 5890
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.382656350741457e-05,
+      "loss": 0.0422,
+      "step": 5900
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.381044487427466e-05,
+      "loss": 0.0583,
+      "step": 5910
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.379432624113475e-05,
+      "loss": 0.0596,
+      "step": 5920
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.3778207607994843e-05,
+      "loss": 0.0648,
+      "step": 5930
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.3762088974854933e-05,
+      "loss": 0.0744,
+      "step": 5940
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.3745970341715023e-05,
+      "loss": 0.0396,
+      "step": 5950
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.3729851708575113e-05,
+      "loss": 0.0444,
+      "step": 5960
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.3713733075435202e-05,
+      "loss": 0.0931,
+      "step": 5970
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.3697614442295296e-05,
+      "loss": 0.043,
+      "step": 5980
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.3681495809155385e-05,
+      "loss": 0.0577,
+      "step": 5990
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.3665377176015475e-05,
+      "loss": 0.0425,
+      "step": 6000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.3649258542875565e-05,
+      "loss": 0.0353,
+      "step": 6010
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.3633139909735655e-05,
+      "loss": 0.054,
+      "step": 6020
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.3617021276595744e-05,
+      "loss": 0.0397,
+      "step": 6030
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.3600902643455838e-05,
+      "loss": 0.0468,
+      "step": 6040
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.3584784010315927e-05,
+      "loss": 0.0393,
+      "step": 6050
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.3568665377176017e-05,
+      "loss": 0.0881,
+      "step": 6060
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.3552546744036107e-05,
+      "loss": 0.0634,
+      "step": 6070
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.3536428110896197e-05,
+      "loss": 0.044,
+      "step": 6080
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.3520309477756286e-05,
+      "loss": 0.0279,
+      "step": 6090
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.3504190844616376e-05,
+      "loss": 0.0569,
+      "step": 6100
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.3488072211476466e-05,
+      "loss": 0.0496,
+      "step": 6110
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.3471953578336556e-05,
+      "loss": 0.0448,
+      "step": 6120
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.3455834945196645e-05,
+      "loss": 0.0433,
+      "step": 6130
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.343971631205674e-05,
+      "loss": 0.0308,
+      "step": 6140
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.342359767891683e-05,
+      "loss": 0.0485,
+      "step": 6150
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.3407479045776918e-05,
+      "loss": 0.0291,
+      "step": 6160
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.339136041263701e-05,
+      "loss": 0.0359,
+      "step": 6170
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.33752417794971e-05,
+      "loss": 0.063,
+      "step": 6180
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.335912314635719e-05,
+      "loss": 0.06,
+      "step": 6190
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.334300451321728e-05,
+      "loss": 0.0502,
+      "step": 6200
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.332688588007737e-05,
+      "loss": 0.0461,
+      "step": 6210
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.331076724693746e-05,
+      "loss": 0.0338,
+      "step": 6220
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.329464861379755e-05,
+      "loss": 0.0668,
+      "step": 6230
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.327852998065764e-05,
+      "loss": 0.0593,
+      "step": 6240
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.3262411347517733e-05,
+      "loss": 0.0742,
+      "step": 6250
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.3246292714377823e-05,
+      "loss": 0.0441,
+      "step": 6260
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.3230174081237912e-05,
+      "loss": 0.0752,
+      "step": 6270
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.3214055448098002e-05,
+      "loss": 0.061,
+      "step": 6280
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.3197936814958092e-05,
+      "loss": 0.0375,
+      "step": 6290
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.318181818181818e-05,
+      "loss": 0.0301,
+      "step": 6300
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.316569954867827e-05,
+      "loss": 0.0325,
+      "step": 6310
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.314958091553836e-05,
+      "loss": 0.0405,
+      "step": 6320
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.313346228239845e-05,
+      "loss": 0.0394,
+      "step": 6330
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.311734364925854e-05,
+      "loss": 0.0512,
+      "step": 6340
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.3101225016118634e-05,
+      "loss": 0.0437,
+      "step": 6350
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.3085106382978724e-05,
+      "loss": 0.0639,
+      "step": 6360
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.3068987749838817e-05,
+      "loss": 0.041,
+      "step": 6370
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.3052869116698907e-05,
+      "loss": 0.0415,
+      "step": 6380
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.3036750483558996e-05,
+      "loss": 0.0426,
+      "step": 6390
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.3020631850419086e-05,
+      "loss": 0.0329,
+      "step": 6400
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.3004513217279176e-05,
+      "loss": 0.0431,
+      "step": 6410
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.2988394584139266e-05,
+      "loss": 0.0391,
+      "step": 6420
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.2972275950999355e-05,
+      "loss": 0.0428,
+      "step": 6430
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.2956157317859445e-05,
+      "loss": 0.0767,
+      "step": 6440
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.2940038684719535e-05,
+      "loss": 0.0498,
+      "step": 6450
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.2923920051579628e-05,
+      "loss": 0.0579,
+      "step": 6460
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.2907801418439718e-05,
+      "loss": 0.0381,
+      "step": 6470
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.2891682785299808e-05,
+      "loss": 0.0408,
+      "step": 6480
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.2875564152159897e-05,
+      "loss": 0.0449,
+      "step": 6490
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.2859445519019987e-05,
+      "loss": 0.0599,
+      "step": 6500
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.2843326885880077e-05,
+      "loss": 0.0509,
+      "step": 6510
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.2827208252740167e-05,
+      "loss": 0.0556,
+      "step": 6520
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.2811089619600256e-05,
+      "loss": 0.034,
+      "step": 6530
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.2794970986460346e-05,
+      "loss": 0.0302,
+      "step": 6540
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.2778852353320436e-05,
+      "loss": 0.0462,
+      "step": 6550
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.276273372018053e-05,
+      "loss": 0.0326,
+      "step": 6560
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.2746615087040622e-05,
+      "loss": 0.0404,
+      "step": 6570
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.2730496453900712e-05,
+      "loss": 0.0391,
+      "step": 6580
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.2714377820760802e-05,
+      "loss": 0.0343,
+      "step": 6590
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.269825918762089e-05,
+      "loss": 0.0485,
+      "step": 6600
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.268214055448098e-05,
+      "loss": 0.0592,
+      "step": 6610
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.266602192134107e-05,
+      "loss": 0.0708,
+      "step": 6620
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.264990328820116e-05,
+      "loss": 0.043,
+      "step": 6630
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.263378465506125e-05,
+      "loss": 0.0561,
+      "step": 6640
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.261766602192134e-05,
+      "loss": 0.0628,
+      "step": 6650
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.260154738878143e-05,
+      "loss": 0.0284,
+      "step": 6660
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.2585428755641523e-05,
+      "loss": 0.0496,
+      "step": 6670
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.2569310122501613e-05,
+      "loss": 0.0474,
+      "step": 6680
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.2553191489361703e-05,
+      "loss": 0.0493,
+      "step": 6690
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.2537072856221793e-05,
+      "loss": 0.0307,
+      "step": 6700
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.2520954223081882e-05,
+      "loss": 0.0338,
+      "step": 6710
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.2504835589941972e-05,
+      "loss": 0.0345,
+      "step": 6720
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.2488716956802062e-05,
+      "loss": 0.0454,
+      "step": 6730
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.247259832366215e-05,
+      "loss": 0.0501,
+      "step": 6740
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.2456479690522245e-05,
+      "loss": 0.0455,
+      "step": 6750
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.2440361057382335e-05,
+      "loss": 0.0334,
+      "step": 6760
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.2424242424242424e-05,
+      "loss": 0.042,
+      "step": 6770
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.2408123791102517e-05,
+      "loss": 0.0531,
+      "step": 6780
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.2392005157962607e-05,
+      "loss": 0.0276,
+      "step": 6790
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.2375886524822697e-05,
+      "loss": 0.0664,
+      "step": 6800
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.2359767891682787e-05,
+      "loss": 0.0352,
+      "step": 6810
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.2343649258542877e-05,
+      "loss": 0.0313,
+      "step": 6820
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.2327530625402966e-05,
+      "loss": 0.046,
+      "step": 6830
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.2311411992263056e-05,
+      "loss": 0.0448,
+      "step": 6840
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.2295293359123146e-05,
+      "loss": 0.035,
+      "step": 6850
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.2279174725983236e-05,
+      "loss": 0.0443,
+      "step": 6860
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.2263056092843325e-05,
+      "loss": 0.0477,
+      "step": 6870
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.224693745970342e-05,
+      "loss": 0.0356,
+      "step": 6880
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.2230818826563508e-05,
+      "loss": 0.0516,
+      "step": 6890
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.2214700193423598e-05,
+      "loss": 0.0483,
+      "step": 6900
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.2198581560283688e-05,
+      "loss": 0.0434,
+      "step": 6910
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.2182462927143778e-05,
+      "loss": 0.0453,
+      "step": 6920
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.2166344294003867e-05,
+      "loss": 0.0322,
+      "step": 6930
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.215022566086396e-05,
+      "loss": 0.0442,
+      "step": 6940
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.213410702772405e-05,
+      "loss": 0.0482,
+      "step": 6950
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.211798839458414e-05,
+      "loss": 0.0457,
+      "step": 6960
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.210186976144423e-05,
+      "loss": 0.0471,
+      "step": 6970
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.208575112830432e-05,
+      "loss": 0.0429,
+      "step": 6980
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.2069632495164413e-05,
+      "loss": 0.0237,
+      "step": 6990
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.2053513862024502e-05,
+      "loss": 0.0371,
+      "step": 7000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.2037395228884592e-05,
+      "loss": 0.0367,
+      "step": 7010
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.2021276595744682e-05,
+      "loss": 0.0724,
+      "step": 7020
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.2005157962604772e-05,
+      "loss": 0.0455,
+      "step": 7030
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.198903932946486e-05,
+      "loss": 0.0516,
+      "step": 7040
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.197292069632495e-05,
+      "loss": 0.0324,
+      "step": 7050
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.195680206318504e-05,
+      "loss": 0.0344,
+      "step": 7060
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.194068343004513e-05,
+      "loss": 0.0393,
+      "step": 7070
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.192456479690522e-05,
+      "loss": 0.0435,
+      "step": 7080
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.1908446163765314e-05,
+      "loss": 0.0645,
+      "step": 7090
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.1892327530625404e-05,
+      "loss": 0.0268,
+      "step": 7100
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.1877820760799485e-05,
+      "loss": 0.0497,
+      "step": 7110
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.1861702127659574e-05,
+      "loss": 0.0503,
+      "step": 7120
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.1845583494519664e-05,
+      "loss": 0.0256,
+      "step": 7130
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.1829464861379754e-05,
+      "loss": 0.043,
+      "step": 7140
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.1813346228239844e-05,
+      "loss": 0.0527,
+      "step": 7150
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.1797227595099933e-05,
+      "loss": 0.0317,
+      "step": 7160
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.1781108961960027e-05,
+      "loss": 0.0283,
+      "step": 7170
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.1764990328820116e-05,
+      "loss": 0.0457,
+      "step": 7180
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.174887169568021e-05,
+      "loss": 0.0454,
+      "step": 7190
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.17327530625403e-05,
+      "loss": 0.0475,
+      "step": 7200
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.171663442940039e-05,
+      "loss": 0.0525,
+      "step": 7210
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.170051579626048e-05,
+      "loss": 0.0386,
+      "step": 7220
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.168439716312057e-05,
+      "loss": 0.0445,
+      "step": 7230
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.166827852998066e-05,
+      "loss": 0.0363,
+      "step": 7240
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.1652159896840748e-05,
+      "loss": 0.0481,
+      "step": 7250
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.1636041263700838e-05,
+      "loss": 0.0321,
+      "step": 7260
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.1619922630560928e-05,
+      "loss": 0.0415,
+      "step": 7270
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.1603803997421017e-05,
+      "loss": 0.0507,
+      "step": 7280
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.1589297227595102e-05,
+      "loss": 0.0421,
+      "step": 7290
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.1573178594455192e-05,
+      "loss": 0.0529,
+      "step": 7300
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.155705996131528e-05,
+      "loss": 0.0565,
+      "step": 7310
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.154094132817537e-05,
+      "loss": 0.048,
+      "step": 7320
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.152482269503546e-05,
+      "loss": 0.0336,
+      "step": 7330
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.150870406189555e-05,
+      "loss": 0.0301,
+      "step": 7340
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.149258542875564e-05,
+      "loss": 0.0473,
+      "step": 7350
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.147646679561573e-05,
+      "loss": 0.0437,
+      "step": 7360
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.146034816247582e-05,
+      "loss": 0.0324,
+      "step": 7370
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.144422952933591e-05,
+      "loss": 0.0348,
+      "step": 7380
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.1428110896196003e-05,
+      "loss": 0.0645,
+      "step": 7390
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.1411992263056096e-05,
+      "loss": 0.0291,
+      "step": 7400
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.1395873629916186e-05,
+      "loss": 0.0366,
+      "step": 7410
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.1379754996776276e-05,
+      "loss": 0.0317,
+      "step": 7420
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.1363636363636365e-05,
+      "loss": 0.0511,
+      "step": 7430
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.1347517730496455e-05,
+      "loss": 0.0366,
+      "step": 7440
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.1331399097356545e-05,
+      "loss": 0.0382,
+      "step": 7450
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.1315280464216635e-05,
+      "loss": 0.0401,
+      "step": 7460
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.1299161831076725e-05,
+      "loss": 0.0479,
+      "step": 7470
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.1283043197936814e-05,
+      "loss": 0.0583,
+      "step": 7480
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.1266924564796904e-05,
+      "loss": 0.053,
+      "step": 7490
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.1250805931656997e-05,
+      "loss": 0.0345,
+      "step": 7500
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.1234687298517087e-05,
+      "loss": 0.0426,
+      "step": 7510
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.1218568665377177e-05,
+      "loss": 0.03,
+      "step": 7520
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.1202450032237267e-05,
+      "loss": 0.0357,
+      "step": 7530
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.1186331399097356e-05,
+      "loss": 0.0268,
+      "step": 7540
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.1170212765957446e-05,
+      "loss": 0.0756,
+      "step": 7550
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.1154094132817536e-05,
+      "loss": 0.0335,
+      "step": 7560
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.1137975499677626e-05,
+      "loss": 0.0396,
+      "step": 7570
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.1121856866537715e-05,
+      "loss": 0.034,
+      "step": 7580
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.110573823339781e-05,
+      "loss": 0.0345,
+      "step": 7590
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.10896196002579e-05,
+      "loss": 0.0308,
+      "step": 7600
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.107350096711799e-05,
+      "loss": 0.0317,
+      "step": 7610
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.105738233397808e-05,
+      "loss": 0.051,
+      "step": 7620
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.104126370083817e-05,
+      "loss": 0.0455,
+      "step": 7630
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.102514506769826e-05,
+      "loss": 0.0449,
+      "step": 7640
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.100902643455835e-05,
+      "loss": 0.05,
+      "step": 7650
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.099290780141844e-05,
+      "loss": 0.0281,
+      "step": 7660
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.097678916827853e-05,
+      "loss": 0.0424,
+      "step": 7670
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.096067053513862e-05,
+      "loss": 0.0367,
+      "step": 7680
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.094455190199871e-05,
+      "loss": 0.0294,
+      "step": 7690
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.09284332688588e-05,
+      "loss": 0.05,
+      "step": 7700
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.0912314635718892e-05,
+      "loss": 0.0431,
+      "step": 7710
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2.0896196002578982e-05,
+      "loss": 0.0267,
+      "step": 7720
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2.0880077369439072e-05,
+      "loss": 0.0331,
+      "step": 7730
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2.0863958736299162e-05,
+      "loss": 0.0449,
+      "step": 7740
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2.084784010315925e-05,
+      "loss": 0.0463,
+      "step": 7750
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2.083172147001934e-05,
+      "loss": 0.0376,
+      "step": 7760
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2.081560283687943e-05,
+      "loss": 0.0579,
+      "step": 7770
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2.0799484203739524e-05,
+      "loss": 0.0414,
+      "step": 7780
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2.0783365570599614e-05,
+      "loss": 0.0427,
+      "step": 7790
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 2.0767246937459704e-05,
+      "loss": 0.0294,
+      "step": 7800
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 2.0751128304319797e-05,
+      "loss": 0.0393,
+      "step": 7810
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 2.0735009671179887e-05,
+      "loss": 0.0433,
+      "step": 7820
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 2.0718891038039976e-05,
+      "loss": 0.0368,
+      "step": 7830
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 2.0702772404900066e-05,
+      "loss": 0.0271,
+      "step": 7840
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 2.0686653771760156e-05,
+      "loss": 0.0256,
+      "step": 7850
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 2.0670535138620246e-05,
+      "loss": 0.0473,
+      "step": 7860
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 2.0654416505480335e-05,
+      "loss": 0.0485,
+      "step": 7870
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 2.0638297872340425e-05,
+      "loss": 0.0541,
+      "step": 7880
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.0622179239200515e-05,
+      "loss": 0.0352,
+      "step": 7890
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.0606060606060605e-05,
+      "loss": 0.0424,
+      "step": 7900
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.0589941972920695e-05,
+      "loss": 0.0476,
+      "step": 7910
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.0573823339780788e-05,
+      "loss": 0.0515,
+      "step": 7920
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.0557704706640877e-05,
+      "loss": 0.0498,
+      "step": 7930
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.0541586073500967e-05,
+      "loss": 0.0363,
+      "step": 7940
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.0525467440361057e-05,
+      "loss": 0.0312,
+      "step": 7950
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.0509348807221147e-05,
+      "loss": 0.0287,
+      "step": 7960
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2.049323017408124e-05,
+      "loss": 0.0424,
+      "step": 7970
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2.047711154094133e-05,
+      "loss": 0.0377,
+      "step": 7980
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2.046099290780142e-05,
+      "loss": 0.0373,
+      "step": 7990
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2.044487427466151e-05,
+      "loss": 0.039,
+      "step": 8000
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2.04287556415216e-05,
+      "loss": 0.046,
+      "step": 8010
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2.0412637008381692e-05,
+      "loss": 0.0276,
+      "step": 8020
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2.0396518375241782e-05,
+      "loss": 0.0484,
+      "step": 8030
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2.038039974210187e-05,
+      "loss": 0.0347,
+      "step": 8040
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 2.036428110896196e-05,
+      "loss": 0.0337,
+      "step": 8050
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 2.034816247582205e-05,
+      "loss": 0.0349,
+      "step": 8060
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 2.033204384268214e-05,
+      "loss": 0.0368,
+      "step": 8070
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 2.031592520954223e-05,
+      "loss": 0.0361,
+      "step": 8080
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.029980657640232e-05,
+      "loss": 0.0222,
+      "step": 8090
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.028368794326241e-05,
+      "loss": 0.0517,
+      "step": 8100
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.02675693101225e-05,
+      "loss": 0.0443,
+      "step": 8110
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.025145067698259e-05,
+      "loss": 0.0139,
+      "step": 8120
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.0235332043842683e-05,
+      "loss": 0.0251,
+      "step": 8130
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.0219213410702773e-05,
+      "loss": 0.0276,
+      "step": 8140
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.0203094777562862e-05,
+      "loss": 0.0272,
+      "step": 8150
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.0186976144422956e-05,
+      "loss": 0.022,
+      "step": 8160
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.0170857511283045e-05,
+      "loss": 0.0501,
+      "step": 8170
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.0154738878143135e-05,
+      "loss": 0.0437,
+      "step": 8180
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.0138620245003225e-05,
+      "loss": 0.0404,
+      "step": 8190
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.0122501611863315e-05,
+      "loss": 0.0295,
+      "step": 8200
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.0106382978723404e-05,
+      "loss": 0.0435,
+      "step": 8210
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2.0090264345583494e-05,
+      "loss": 0.0524,
+      "step": 8220
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2.0074145712443584e-05,
+      "loss": 0.0303,
+      "step": 8230
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2.0058027079303677e-05,
+      "loss": 0.0396,
+      "step": 8240
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2.0041908446163767e-05,
+      "loss": 0.0532,
+      "step": 8250
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.0025789813023857e-05,
+      "loss": 0.0385,
+      "step": 8260
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.0009671179883946e-05,
+      "loss": 0.0534,
+      "step": 8270
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9958867321616752,
+      "eval_loss": 0.015995986759662628,
+      "eval_runtime": 68.6513,
+      "eval_samples_per_second": 856.998,
+      "eval_steps_per_second": 26.788,
+      "step": 8273
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9993552546744036e-05,
+      "loss": 0.0367,
+      "step": 8280
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.9977433913604126e-05,
+      "loss": 0.0474,
+      "step": 8290
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9961315280464216e-05,
+      "loss": 0.0274,
+      "step": 8300
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9945196647324305e-05,
+      "loss": 0.0329,
+      "step": 8310
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9929078014184395e-05,
+      "loss": 0.0307,
+      "step": 8320
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.9912959381044485e-05,
+      "loss": 0.0524,
+      "step": 8330
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.9896840747904578e-05,
+      "loss": 0.0505,
+      "step": 8340
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.988072211476467e-05,
+      "loss": 0.0221,
+      "step": 8350
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.986460348162476e-05,
+      "loss": 0.0277,
+      "step": 8360
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.984848484848485e-05,
+      "loss": 0.0333,
+      "step": 8370
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.983236621534494e-05,
+      "loss": 0.0192,
+      "step": 8380
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.981624758220503e-05,
+      "loss": 0.0556,
+      "step": 8390
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.980012894906512e-05,
+      "loss": 0.0313,
+      "step": 8400
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.978401031592521e-05,
+      "loss": 0.0454,
+      "step": 8410
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.97678916827853e-05,
+      "loss": 0.0304,
+      "step": 8420
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.975177304964539e-05,
+      "loss": 0.0304,
+      "step": 8430
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.973565441650548e-05,
+      "loss": 0.0354,
+      "step": 8440
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.9719535783365572e-05,
+      "loss": 0.0469,
+      "step": 8450
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.9703417150225662e-05,
+      "loss": 0.034,
+      "step": 8460
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.9687298517085752e-05,
+      "loss": 0.0375,
+      "step": 8470
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.967117988394584e-05,
+      "loss": 0.0215,
+      "step": 8480
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.965506125080593e-05,
+      "loss": 0.0424,
+      "step": 8490
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.963894261766602e-05,
+      "loss": 0.035,
+      "step": 8500
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.962282398452611e-05,
+      "loss": 0.0343,
+      "step": 8510
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.96067053513862e-05,
+      "loss": 0.0372,
+      "step": 8520
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.959058671824629e-05,
+      "loss": 0.0591,
+      "step": 8530
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.9574468085106384e-05,
+      "loss": 0.0336,
+      "step": 8540
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.9558349451966477e-05,
+      "loss": 0.033,
+      "step": 8550
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.9542230818826567e-05,
+      "loss": 0.0212,
+      "step": 8560
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.9526112185686656e-05,
+      "loss": 0.0464,
+      "step": 8570
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.9509993552546746e-05,
+      "loss": 0.0263,
+      "step": 8580
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.9493874919406836e-05,
+      "loss": 0.0501,
+      "step": 8590
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.9477756286266926e-05,
+      "loss": 0.0178,
+      "step": 8600
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.9461637653127015e-05,
+      "loss": 0.0153,
+      "step": 8610
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.9445519019987105e-05,
+      "loss": 0.0218,
+      "step": 8620
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.9429400386847195e-05,
+      "loss": 0.0287,
+      "step": 8630
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.9413281753707285e-05,
+      "loss": 0.0355,
+      "step": 8640
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.9397163120567374e-05,
+      "loss": 0.0263,
+      "step": 8650
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.9381044487427468e-05,
+      "loss": 0.0253,
+      "step": 8660
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.9364925854287557e-05,
+      "loss": 0.0406,
+      "step": 8670
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.9348807221147647e-05,
+      "loss": 0.022,
+      "step": 8680
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.9332688588007737e-05,
+      "loss": 0.0298,
+      "step": 8690
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.9316569954867827e-05,
+      "loss": 0.0365,
+      "step": 8700
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.9300451321727916e-05,
+      "loss": 0.0167,
+      "step": 8710
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.9284332688588006e-05,
+      "loss": 0.0305,
+      "step": 8720
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.92682140554481e-05,
+      "loss": 0.0276,
+      "step": 8730
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.925209542230819e-05,
+      "loss": 0.0465,
+      "step": 8740
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.923597678916828e-05,
+      "loss": 0.0161,
+      "step": 8750
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.9219858156028372e-05,
+      "loss": 0.0261,
+      "step": 8760
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.9203739522888462e-05,
+      "loss": 0.0503,
+      "step": 8770
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.918762088974855e-05,
+      "loss": 0.0339,
+      "step": 8780
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.917150225660864e-05,
+      "loss": 0.021,
+      "step": 8790
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.915538362346873e-05,
+      "loss": 0.0501,
+      "step": 8800
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.913926499032882e-05,
+      "loss": 0.0434,
+      "step": 8810
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.912314635718891e-05,
+      "loss": 0.0425,
+      "step": 8820
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.9107027724049e-05,
+      "loss": 0.045,
+      "step": 8830
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.909090909090909e-05,
+      "loss": 0.028,
+      "step": 8840
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.907479045776918e-05,
+      "loss": 0.0441,
+      "step": 8850
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.905867182462927e-05,
+      "loss": 0.0272,
+      "step": 8860
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.9042553191489363e-05,
+      "loss": 0.0291,
+      "step": 8870
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.9026434558349453e-05,
+      "loss": 0.0272,
+      "step": 8880
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.9010315925209542e-05,
+      "loss": 0.0369,
+      "step": 8890
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.8994197292069632e-05,
+      "loss": 0.0245,
+      "step": 8900
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.8978078658929722e-05,
+      "loss": 0.0412,
+      "step": 8910
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.8961960025789815e-05,
+      "loss": 0.0372,
+      "step": 8920
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.8945841392649905e-05,
+      "loss": 0.029,
+      "step": 8930
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.8929722759509995e-05,
+      "loss": 0.0264,
+      "step": 8940
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.8913604126370084e-05,
+      "loss": 0.0322,
+      "step": 8950
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.8897485493230174e-05,
+      "loss": 0.0238,
+      "step": 8960
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.8881366860090264e-05,
+      "loss": 0.0421,
+      "step": 8970
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.8865248226950357e-05,
+      "loss": 0.0274,
+      "step": 8980
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.8849129593810447e-05,
+      "loss": 0.0267,
+      "step": 8990
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.8833010960670537e-05,
+      "loss": 0.0461,
+      "step": 9000
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.8816892327530626e-05,
+      "loss": 0.0359,
+      "step": 9010
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.8800773694390716e-05,
+      "loss": 0.0311,
+      "step": 9020
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.8784655061250806e-05,
+      "loss": 0.0437,
+      "step": 9030
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.8768536428110896e-05,
+      "loss": 0.0164,
+      "step": 9040
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.8752417794970985e-05,
+      "loss": 0.0305,
+      "step": 9050
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.8736299161831075e-05,
+      "loss": 0.0265,
+      "step": 9060
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.8720180528691165e-05,
+      "loss": 0.0231,
+      "step": 9070
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.8704061895551258e-05,
+      "loss": 0.0137,
+      "step": 9080
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.8687943262411348e-05,
+      "loss": 0.0313,
+      "step": 9090
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.8671824629271438e-05,
+      "loss": 0.0267,
+      "step": 9100
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.865570599613153e-05,
+      "loss": 0.032,
+      "step": 9110
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.863958736299162e-05,
+      "loss": 0.051,
+      "step": 9120
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.862346872985171e-05,
+      "loss": 0.0303,
+      "step": 9130
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.86073500967118e-05,
+      "loss": 0.0442,
+      "step": 9140
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.859123146357189e-05,
+      "loss": 0.0336,
+      "step": 9150
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.857511283043198e-05,
+      "loss": 0.0291,
+      "step": 9160
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.855899419729207e-05,
+      "loss": 0.0333,
+      "step": 9170
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.854287556415216e-05,
+      "loss": 0.0261,
+      "step": 9180
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.8526756931012252e-05,
+      "loss": 0.0171,
+      "step": 9190
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.8510638297872342e-05,
+      "loss": 0.0241,
+      "step": 9200
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.8494519664732432e-05,
+      "loss": 0.0252,
+      "step": 9210
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.847840103159252e-05,
+      "loss": 0.0298,
+      "step": 9220
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.846228239845261e-05,
+      "loss": 0.0192,
+      "step": 9230
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.84461637653127e-05,
+      "loss": 0.052,
+      "step": 9240
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.843004513217279e-05,
+      "loss": 0.0309,
+      "step": 9250
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.841392649903288e-05,
+      "loss": 0.0283,
+      "step": 9260
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.839780786589297e-05,
+      "loss": 0.0141,
+      "step": 9270
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.838168923275306e-05,
+      "loss": 0.0337,
+      "step": 9280
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.8365570599613153e-05,
+      "loss": 0.036,
+      "step": 9290
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.8349451966473246e-05,
+      "loss": 0.0361,
+      "step": 9300
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.8333333333333336e-05,
+      "loss": 0.0457,
+      "step": 9310
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.8317214700193426e-05,
+      "loss": 0.0328,
+      "step": 9320
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.8301096067053516e-05,
+      "loss": 0.0395,
+      "step": 9330
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.8284977433913606e-05,
+      "loss": 0.0277,
+      "step": 9340
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.8268858800773695e-05,
+      "loss": 0.0226,
+      "step": 9350
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.8252740167633785e-05,
+      "loss": 0.0187,
+      "step": 9360
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.8236621534493875e-05,
+      "loss": 0.0234,
+      "step": 9370
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.8220502901353965e-05,
+      "loss": 0.0256,
+      "step": 9380
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.8204384268214054e-05,
+      "loss": 0.026,
+      "step": 9390
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.8188265635074148e-05,
+      "loss": 0.0292,
+      "step": 9400
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.8172147001934237e-05,
+      "loss": 0.0202,
+      "step": 9410
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.8156028368794327e-05,
+      "loss": 0.0368,
+      "step": 9420
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.8139909735654417e-05,
+      "loss": 0.0129,
+      "step": 9430
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.8123791102514507e-05,
+      "loss": 0.0163,
+      "step": 9440
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.8107672469374596e-05,
+      "loss": 0.04,
+      "step": 9450
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.8091553836234686e-05,
+      "loss": 0.0233,
+      "step": 9460
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.8075435203094776e-05,
+      "loss": 0.0334,
+      "step": 9470
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.8059316569954866e-05,
+      "loss": 0.0231,
+      "step": 9480
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.8043197936814955e-05,
+      "loss": 0.0285,
+      "step": 9490
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.8027079303675052e-05,
+      "loss": 0.0308,
+      "step": 9500
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.8010960670535142e-05,
+      "loss": 0.0488,
+      "step": 9510
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.799484203739523e-05,
+      "loss": 0.0335,
+      "step": 9520
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.797872340425532e-05,
+      "loss": 0.0325,
+      "step": 9530
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.796260477111541e-05,
+      "loss": 0.0283,
+      "step": 9540
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.79464861379755e-05,
+      "loss": 0.0336,
+      "step": 9550
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.793036750483559e-05,
+      "loss": 0.0322,
+      "step": 9560
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.791424887169568e-05,
+      "loss": 0.0279,
+      "step": 9570
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.789813023855577e-05,
+      "loss": 0.0432,
+      "step": 9580
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.788201160541586e-05,
+      "loss": 0.0333,
+      "step": 9590
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.786589297227595e-05,
+      "loss": 0.0312,
+      "step": 9600
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.7849774339136043e-05,
+      "loss": 0.0235,
+      "step": 9610
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.7833655705996133e-05,
+      "loss": 0.0305,
+      "step": 9620
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.7817537072856222e-05,
+      "loss": 0.0301,
+      "step": 9630
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.7801418439716312e-05,
+      "loss": 0.0356,
+      "step": 9640
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.7785299806576402e-05,
+      "loss": 0.0341,
+      "step": 9650
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.776918117343649e-05,
+      "loss": 0.0175,
+      "step": 9660
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.775306254029658e-05,
+      "loss": 0.0426,
+      "step": 9670
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.773694390715667e-05,
+      "loss": 0.0188,
+      "step": 9680
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.7720825274016764e-05,
+      "loss": 0.0498,
+      "step": 9690
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.7704706640876854e-05,
+      "loss": 0.0279,
+      "step": 9700
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.7688588007736947e-05,
+      "loss": 0.0188,
+      "step": 9710
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.7672469374597037e-05,
+      "loss": 0.0357,
+      "step": 9720
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.7656350741457127e-05,
+      "loss": 0.0262,
+      "step": 9730
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.7640232108317216e-05,
+      "loss": 0.052,
+      "step": 9740
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.7624113475177306e-05,
+      "loss": 0.0328,
+      "step": 9750
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.7607994842037396e-05,
+      "loss": 0.0187,
+      "step": 9760
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.7591876208897486e-05,
+      "loss": 0.0311,
+      "step": 9770
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.7575757575757576e-05,
+      "loss": 0.0173,
+      "step": 9780
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.7559638942617665e-05,
+      "loss": 0.0143,
+      "step": 9790
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.7543520309477755e-05,
+      "loss": 0.0351,
+      "step": 9800
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.7527401676337845e-05,
+      "loss": 0.0187,
+      "step": 9810
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.7511283043197938e-05,
+      "loss": 0.033,
+      "step": 9820
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.7495164410058028e-05,
+      "loss": 0.0265,
+      "step": 9830
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.7479045776918118e-05,
+      "loss": 0.0258,
+      "step": 9840
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.7462927143778207e-05,
+      "loss": 0.0162,
+      "step": 9850
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.7446808510638297e-05,
+      "loss": 0.0258,
+      "step": 9860
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.7430689877498387e-05,
+      "loss": 0.0239,
+      "step": 9870
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.741457124435848e-05,
+      "loss": 0.0365,
+      "step": 9880
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.739845261121857e-05,
+      "loss": 0.0389,
+      "step": 9890
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.738233397807866e-05,
+      "loss": 0.0199,
+      "step": 9900
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.736621534493875e-05,
+      "loss": 0.0399,
+      "step": 9910
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.735009671179884e-05,
+      "loss": 0.026,
+      "step": 9920
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.7333978078658932e-05,
+      "loss": 0.0351,
+      "step": 9930
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.7317859445519022e-05,
+      "loss": 0.036,
+      "step": 9940
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.7301740812379112e-05,
+      "loss": 0.0208,
+      "step": 9950
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.72856221792392e-05,
+      "loss": 0.0237,
+      "step": 9960
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.726950354609929e-05,
+      "loss": 0.0732,
+      "step": 9970
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.725338491295938e-05,
+      "loss": 0.0466,
+      "step": 9980
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.723726627981947e-05,
+      "loss": 0.0356,
+      "step": 9990
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.722114764667956e-05,
+      "loss": 0.0265,
+      "step": 10000
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.720502901353965e-05,
+      "loss": 0.0169,
+      "step": 10010
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.718891038039974e-05,
+      "loss": 0.0279,
+      "step": 10020
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.7172791747259833e-05,
+      "loss": 0.022,
+      "step": 10030
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.7156673114119923e-05,
+      "loss": 0.0204,
+      "step": 10040
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.7140554480980013e-05,
+      "loss": 0.0212,
+      "step": 10050
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.7124435847840103e-05,
+      "loss": 0.0319,
+      "step": 10060
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.7108317214700196e-05,
+      "loss": 0.0386,
+      "step": 10070
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7092198581560285e-05,
+      "loss": 0.015,
+      "step": 10080
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7076079948420375e-05,
+      "loss": 0.0393,
+      "step": 10090
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7059961315280465e-05,
+      "loss": 0.0146,
+      "step": 10100
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7043842682140555e-05,
+      "loss": 0.0201,
+      "step": 10110
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.7027724049000645e-05,
+      "loss": 0.0238,
+      "step": 10120
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.7011605415860734e-05,
+      "loss": 0.0181,
+      "step": 10130
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.6995486782720827e-05,
+      "loss": 0.0334,
+      "step": 10140
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.6979368149580917e-05,
+      "loss": 0.0209,
+      "step": 10150
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6963249516441007e-05,
+      "loss": 0.0205,
+      "step": 10160
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6947130883301097e-05,
+      "loss": 0.031,
+      "step": 10170
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6931012250161187e-05,
+      "loss": 0.0214,
+      "step": 10180
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6914893617021276e-05,
+      "loss": 0.0306,
+      "step": 10190
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.6898774983881366e-05,
+      "loss": 0.0321,
+      "step": 10200
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.6882656350741456e-05,
+      "loss": 0.0273,
+      "step": 10210
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.6866537717601546e-05,
+      "loss": 0.0257,
+      "step": 10220
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.6850419084461635e-05,
+      "loss": 0.0474,
+      "step": 10230
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.683430045132173e-05,
+      "loss": 0.0249,
+      "step": 10240
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.6818181818181818e-05,
+      "loss": 0.0354,
+      "step": 10250
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.680206318504191e-05,
+      "loss": 0.0219,
+      "step": 10260
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.6785944551902e-05,
+      "loss": 0.0227,
+      "step": 10270
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.676982591876209e-05,
+      "loss": 0.0189,
+      "step": 10280
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.675370728562218e-05,
+      "loss": 0.0162,
+      "step": 10290
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.673758865248227e-05,
+      "loss": 0.0188,
+      "step": 10300
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.672147001934236e-05,
+      "loss": 0.0356,
+      "step": 10310
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.670535138620245e-05,
+      "loss": 0.0376,
+      "step": 10320
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.668923275306254e-05,
+      "loss": 0.0277,
+      "step": 10330
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.667311411992263e-05,
+      "loss": 0.0133,
+      "step": 10340
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.6656995486782723e-05,
+      "loss": 0.0195,
+      "step": 10350
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.6640876853642812e-05,
+      "loss": 0.0159,
+      "step": 10360
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.6624758220502902e-05,
+      "loss": 0.0307,
+      "step": 10370
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.6608639587362992e-05,
+      "loss": 0.0277,
+      "step": 10380
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.6592520954223082e-05,
+      "loss": 0.0251,
+      "step": 10390
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.657640232108317e-05,
+      "loss": 0.0236,
+      "step": 10400
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.656028368794326e-05,
+      "loss": 0.0388,
+      "step": 10410
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.654416505480335e-05,
+      "loss": 0.0465,
+      "step": 10420
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.652804642166344e-05,
+      "loss": 0.0275,
+      "step": 10430
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.651192778852353e-05,
+      "loss": 0.0269,
+      "step": 10440
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.6495809155383627e-05,
+      "loss": 0.0351,
+      "step": 10450
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.6479690522243717e-05,
+      "loss": 0.0195,
+      "step": 10460
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.6463571889103807e-05,
+      "loss": 0.0203,
+      "step": 10470
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.6447453255963896e-05,
+      "loss": 0.0228,
+      "step": 10480
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.6431334622823986e-05,
+      "loss": 0.0227,
+      "step": 10490
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.6415215989684076e-05,
+      "loss": 0.0244,
+      "step": 10500
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.6399097356544166e-05,
+      "loss": 0.0338,
+      "step": 10510
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.6382978723404255e-05,
+      "loss": 0.0267,
+      "step": 10520
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.6366860090264345e-05,
+      "loss": 0.0216,
+      "step": 10530
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.6350741457124435e-05,
+      "loss": 0.0411,
+      "step": 10540
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.6334622823984525e-05,
+      "loss": 0.031,
+      "step": 10550
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.6318504190844618e-05,
+      "loss": 0.0143,
+      "step": 10560
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.6302385557704708e-05,
+      "loss": 0.0269,
+      "step": 10570
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.6286266924564797e-05,
+      "loss": 0.0112,
+      "step": 10580
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.6270148291424887e-05,
+      "loss": 0.019,
+      "step": 10590
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.6254029658284977e-05,
+      "loss": 0.0165,
+      "step": 10600
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.6237911025145067e-05,
+      "loss": 0.0427,
+      "step": 10610
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.6221792392005157e-05,
+      "loss": 0.0282,
+      "step": 10620
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.6205673758865246e-05,
+      "loss": 0.0354,
+      "step": 10630
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.618955512572534e-05,
+      "loss": 0.0224,
+      "step": 10640
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.617343649258543e-05,
+      "loss": 0.0238,
+      "step": 10650
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.615731785944552e-05,
+      "loss": 0.0292,
+      "step": 10660
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.6141199226305612e-05,
+      "loss": 0.0207,
+      "step": 10670
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.6125080593165702e-05,
+      "loss": 0.0233,
+      "step": 10680
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.610896196002579e-05,
+      "loss": 0.0516,
+      "step": 10690
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.609284332688588e-05,
+      "loss": 0.0271,
+      "step": 10700
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.607672469374597e-05,
+      "loss": 0.027,
+      "step": 10710
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.606060606060606e-05,
+      "loss": 0.0174,
+      "step": 10720
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.604448742746615e-05,
+      "loss": 0.015,
+      "step": 10730
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.602836879432624e-05,
+      "loss": 0.0147,
+      "step": 10740
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.601225016118633e-05,
+      "loss": 0.0311,
+      "step": 10750
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.599613152804642e-05,
+      "loss": 0.0401,
+      "step": 10760
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.5980012894906513e-05,
+      "loss": 0.0187,
+      "step": 10770
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.5963894261766603e-05,
+      "loss": 0.0252,
+      "step": 10780
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.5947775628626693e-05,
+      "loss": 0.0296,
+      "step": 10790
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.5931656995486782e-05,
+      "loss": 0.0097,
+      "step": 10800
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.5915538362346872e-05,
+      "loss": 0.028,
+      "step": 10810
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.5899419729206962e-05,
+      "loss": 0.0444,
+      "step": 10820
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.5883301096067055e-05,
+      "loss": 0.0221,
+      "step": 10830
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.5867182462927145e-05,
+      "loss": 0.0375,
+      "step": 10840
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.5851063829787235e-05,
+      "loss": 0.0178,
+      "step": 10850
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.5834945196647324e-05,
+      "loss": 0.029,
+      "step": 10860
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.5818826563507414e-05,
+      "loss": 0.0475,
+      "step": 10870
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.5802707930367507e-05,
+      "loss": 0.0211,
+      "step": 10880
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.5786589297227597e-05,
+      "loss": 0.0434,
+      "step": 10890
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.5770470664087687e-05,
+      "loss": 0.0276,
+      "step": 10900
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.5754352030947777e-05,
+      "loss": 0.021,
+      "step": 10910
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.5738233397807866e-05,
+      "loss": 0.0364,
+      "step": 10920
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.5722114764667956e-05,
+      "loss": 0.0168,
+      "step": 10930
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.5705996131528046e-05,
+      "loss": 0.0257,
+      "step": 10940
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.5689877498388136e-05,
+      "loss": 0.0425,
+      "step": 10950
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.5673758865248225e-05,
+      "loss": 0.0153,
+      "step": 10960
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.5657640232108315e-05,
+      "loss": 0.0239,
+      "step": 10970
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.564152159896841e-05,
+      "loss": 0.0181,
+      "step": 10980
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.5625402965828498e-05,
+      "loss": 0.0149,
+      "step": 10990
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.5609284332688588e-05,
+      "loss": 0.0314,
+      "step": 11000
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.5593165699548678e-05,
+      "loss": 0.0233,
+      "step": 11010
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.557704706640877e-05,
+      "loss": 0.0337,
+      "step": 11020
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.556092843326886e-05,
+      "loss": 0.0357,
+      "step": 11030
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.554480980012895e-05,
+      "loss": 0.0165,
+      "step": 11040
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.552869116698904e-05,
+      "loss": 0.0209,
+      "step": 11050
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.551257253384913e-05,
+      "loss": 0.0249,
+      "step": 11060
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.549645390070922e-05,
+      "loss": 0.0245,
+      "step": 11070
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.548033526756931e-05,
+      "loss": 0.0225,
+      "step": 11080
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.5464216634429403e-05,
+      "loss": 0.0311,
+      "step": 11090
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.5448098001289492e-05,
+      "loss": 0.024,
+      "step": 11100
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.5431979368149582e-05,
+      "loss": 0.0291,
+      "step": 11110
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.5415860735009672e-05,
+      "loss": 0.0478,
+      "step": 11120
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.539974210186976e-05,
+      "loss": 0.0286,
+      "step": 11130
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.538362346872985e-05,
+      "loss": 0.0213,
+      "step": 11140
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.536750483558994e-05,
+      "loss": 0.0165,
+      "step": 11150
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.535138620245003e-05,
+      "loss": 0.0328,
+      "step": 11160
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.533526756931012e-05,
+      "loss": 0.0265,
+      "step": 11170
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.531914893617021e-05,
+      "loss": 0.0142,
+      "step": 11180
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.5303030303030304e-05,
+      "loss": 0.0277,
+      "step": 11190
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.5286911669890393e-05,
+      "loss": 0.0357,
+      "step": 11200
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.5270793036750487e-05,
+      "loss": 0.0244,
+      "step": 11210
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.5254674403610576e-05,
+      "loss": 0.0101,
+      "step": 11220
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.5238555770470666e-05,
+      "loss": 0.0355,
+      "step": 11230
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.5222437137330756e-05,
+      "loss": 0.0099,
+      "step": 11240
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.5206318504190846e-05,
+      "loss": 0.026,
+      "step": 11250
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.5190199871050935e-05,
+      "loss": 0.0286,
+      "step": 11260
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.5174081237911027e-05,
+      "loss": 0.0382,
+      "step": 11270
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.5157962604771117e-05,
+      "loss": 0.0192,
+      "step": 11280
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.5141843971631206e-05,
+      "loss": 0.0292,
+      "step": 11290
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.5125725338491296e-05,
+      "loss": 0.0273,
+      "step": 11300
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.5109606705351386e-05,
+      "loss": 0.0316,
+      "step": 11310
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.5093488072211477e-05,
+      "loss": 0.014,
+      "step": 11320
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.5077369439071567e-05,
+      "loss": 0.0296,
+      "step": 11330
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.5061250805931657e-05,
+      "loss": 0.0192,
+      "step": 11340
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.5045132172791747e-05,
+      "loss": 0.0205,
+      "step": 11350
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.5029013539651836e-05,
+      "loss": 0.0293,
+      "step": 11360
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.5012894906511926e-05,
+      "loss": 0.0198,
+      "step": 11370
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.499677627337202e-05,
+      "loss": 0.0166,
+      "step": 11380
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.4980657640232109e-05,
+      "loss": 0.0208,
+      "step": 11390
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.4964539007092199e-05,
+      "loss": 0.0189,
+      "step": 11400
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.4948420373952289e-05,
+      "loss": 0.0341,
+      "step": 11410
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.4932301740812378e-05,
+      "loss": 0.0272,
+      "step": 11420
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.491618310767247e-05,
+      "loss": 0.0283,
+      "step": 11430
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.490006447453256e-05,
+      "loss": 0.026,
+      "step": 11440
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.4883945841392651e-05,
+      "loss": 0.0148,
+      "step": 11450
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.4867827208252741e-05,
+      "loss": 0.037,
+      "step": 11460
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.485170857511283e-05,
+      "loss": 0.0198,
+      "step": 11470
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.4835589941972922e-05,
+      "loss": 0.0361,
+      "step": 11480
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.4819471308833012e-05,
+      "loss": 0.0261,
+      "step": 11490
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.4803352675693102e-05,
+      "loss": 0.0189,
+      "step": 11500
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.4787234042553191e-05,
+      "loss": 0.0281,
+      "step": 11510
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.4771115409413281e-05,
+      "loss": 0.0327,
+      "step": 11520
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.4754996776273371e-05,
+      "loss": 0.0197,
+      "step": 11530
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.4738878143133462e-05,
+      "loss": 0.0226,
+      "step": 11540
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.4722759509993554e-05,
+      "loss": 0.0326,
+      "step": 11550
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.4706640876853644e-05,
+      "loss": 0.0103,
+      "step": 11560
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.4690522243713733e-05,
+      "loss": 0.0411,
+      "step": 11570
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.4674403610573823e-05,
+      "loss": 0.0135,
+      "step": 11580
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.4658284977433915e-05,
+      "loss": 0.0149,
+      "step": 11590
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.4642166344294004e-05,
+      "loss": 0.0163,
+      "step": 11600
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.4626047711154094e-05,
+      "loss": 0.031,
+      "step": 11610
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.4609929078014184e-05,
+      "loss": 0.0289,
+      "step": 11620
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.4593810444874274e-05,
+      "loss": 0.0407,
+      "step": 11630
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.4577691811734367e-05,
+      "loss": 0.0131,
+      "step": 11640
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.4561573178594457e-05,
+      "loss": 0.0204,
+      "step": 11650
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.4545454545454546e-05,
+      "loss": 0.0153,
+      "step": 11660
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.4529335912314636e-05,
+      "loss": 0.0171,
+      "step": 11670
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.4513217279174726e-05,
+      "loss": 0.0275,
+      "step": 11680
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.4497098646034817e-05,
+      "loss": 0.043,
+      "step": 11690
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.4480980012894907e-05,
+      "loss": 0.0359,
+      "step": 11700
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.4464861379754997e-05,
+      "loss": 0.038,
+      "step": 11710
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.4448742746615087e-05,
+      "loss": 0.0215,
+      "step": 11720
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.4432624113475176e-05,
+      "loss": 0.021,
+      "step": 11730
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.4416505480335268e-05,
+      "loss": 0.0349,
+      "step": 11740
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.440038684719536e-05,
+      "loss": 0.015,
+      "step": 11750
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.4384268214055449e-05,
+      "loss": 0.0224,
+      "step": 11760
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.4368149580915539e-05,
+      "loss": 0.0162,
+      "step": 11770
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.4352030947775629e-05,
+      "loss": 0.0303,
+      "step": 11780
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.4335912314635718e-05,
+      "loss": 0.0278,
+      "step": 11790
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.431979368149581e-05,
+      "loss": 0.038,
+      "step": 11800
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.43036750483559e-05,
+      "loss": 0.0221,
+      "step": 11810
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.428755641521599e-05,
+      "loss": 0.033,
+      "step": 11820
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.427143778207608e-05,
+      "loss": 0.0135,
+      "step": 11830
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.425531914893617e-05,
+      "loss": 0.0269,
+      "step": 11840
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.4239200515796262e-05,
+      "loss": 0.0244,
+      "step": 11850
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.4223081882656352e-05,
+      "loss": 0.0261,
+      "step": 11860
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.4206963249516442e-05,
+      "loss": 0.0313,
+      "step": 11870
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.4190844616376531e-05,
+      "loss": 0.0155,
+      "step": 11880
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.4174725983236621e-05,
+      "loss": 0.0139,
+      "step": 11890
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.4158607350096711e-05,
+      "loss": 0.0304,
+      "step": 11900
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.4142488716956802e-05,
+      "loss": 0.0105,
+      "step": 11910
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.4126370083816892e-05,
+      "loss": 0.0359,
+      "step": 11920
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.4110251450676984e-05,
+      "loss": 0.0193,
+      "step": 11930
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.4094132817537073e-05,
+      "loss": 0.0186,
+      "step": 11940
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.4078014184397163e-05,
+      "loss": 0.0188,
+      "step": 11950
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.4061895551257255e-05,
+      "loss": 0.0183,
+      "step": 11960
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.4045776918117344e-05,
+      "loss": 0.0152,
+      "step": 11970
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.4029658284977434e-05,
+      "loss": 0.0226,
+      "step": 11980
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.4013539651837524e-05,
+      "loss": 0.0339,
+      "step": 11990
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.3997421018697614e-05,
+      "loss": 0.0211,
+      "step": 12000
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.3981302385557705e-05,
+      "loss": 0.0403,
+      "step": 12010
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.3965183752417797e-05,
+      "loss": 0.0172,
+      "step": 12020
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.3949065119277886e-05,
+      "loss": 0.0352,
+      "step": 12030
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.3932946486137976e-05,
+      "loss": 0.0194,
+      "step": 12040
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.3916827852998066e-05,
+      "loss": 0.0136,
+      "step": 12050
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.3900709219858157e-05,
+      "loss": 0.0167,
+      "step": 12060
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.3884590586718247e-05,
+      "loss": 0.0266,
+      "step": 12070
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.3868471953578337e-05,
+      "loss": 0.0237,
+      "step": 12080
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.3852353320438427e-05,
+      "loss": 0.019,
+      "step": 12090
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.3836234687298516e-05,
+      "loss": 0.0233,
+      "step": 12100
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.3820116054158606e-05,
+      "loss": 0.017,
+      "step": 12110
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.38039974210187e-05,
+      "loss": 0.0079,
+      "step": 12120
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.3787878787878789e-05,
+      "loss": 0.0304,
+      "step": 12130
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.3771760154738879e-05,
+      "loss": 0.026,
+      "step": 12140
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.3755641521598969e-05,
+      "loss": 0.0217,
+      "step": 12150
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.3739522888459058e-05,
+      "loss": 0.0256,
+      "step": 12160
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.372340425531915e-05,
+      "loss": 0.0188,
+      "step": 12170
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.370728562217924e-05,
+      "loss": 0.0273,
+      "step": 12180
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.369116698903933e-05,
+      "loss": 0.0298,
+      "step": 12190
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.3675048355899419e-05,
+      "loss": 0.0157,
+      "step": 12200
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.365892972275951e-05,
+      "loss": 0.0448,
+      "step": 12210
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.3642811089619602e-05,
+      "loss": 0.0158,
+      "step": 12220
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.3626692456479692e-05,
+      "loss": 0.0217,
+      "step": 12230
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.3610573823339782e-05,
+      "loss": 0.0252,
+      "step": 12240
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.3594455190199871e-05,
+      "loss": 0.0241,
+      "step": 12250
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.3578336557059961e-05,
+      "loss": 0.0256,
+      "step": 12260
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.356221792392005e-05,
+      "loss": 0.0172,
+      "step": 12270
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.3546099290780142e-05,
+      "loss": 0.0253,
+      "step": 12280
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.3529980657640232e-05,
+      "loss": 0.0253,
+      "step": 12290
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.3513862024500322e-05,
+      "loss": 0.0141,
+      "step": 12300
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.3497743391360413e-05,
+      "loss": 0.0172,
+      "step": 12310
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.3481624758220503e-05,
+      "loss": 0.0269,
+      "step": 12320
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.3465506125080595e-05,
+      "loss": 0.0119,
+      "step": 12330
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.3449387491940684e-05,
+      "loss": 0.0234,
+      "step": 12340
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.3433268858800774e-05,
+      "loss": 0.0231,
+      "step": 12350
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.3417150225660864e-05,
+      "loss": 0.0208,
+      "step": 12360
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.3401031592520954e-05,
+      "loss": 0.0258,
+      "step": 12370
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.3384912959381045e-05,
+      "loss": 0.0128,
+      "step": 12380
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.3368794326241135e-05,
+      "loss": 0.0104,
+      "step": 12390
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.3352675693101226e-05,
+      "loss": 0.0136,
+      "step": 12400
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.3336557059961316e-05,
+      "loss": 0.0548,
+      "step": 12410
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.9976204235646055,
+      "eval_loss": 0.00988665409386158,
+      "eval_runtime": 68.5442,
+      "eval_samples_per_second": 858.336,
+      "eval_steps_per_second": 26.829,
+      "step": 12410
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.3320438426821406e-05,
+      "loss": 0.029,
+      "step": 12420
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.3304319793681497e-05,
+      "loss": 0.0179,
+      "step": 12430
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.3288201160541587e-05,
+      "loss": 0.0106,
+      "step": 12440
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.3272082527401677e-05,
+      "loss": 0.0257,
+      "step": 12450
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.3255963894261767e-05,
+      "loss": 0.0243,
+      "step": 12460
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.3239845261121856e-05,
+      "loss": 0.0234,
+      "step": 12470
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.3223726627981946e-05,
+      "loss": 0.0127,
+      "step": 12480
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.3207607994842038e-05,
+      "loss": 0.0132,
+      "step": 12490
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.3191489361702129e-05,
+      "loss": 0.0137,
+      "step": 12500
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.3175370728562219e-05,
+      "loss": 0.0135,
+      "step": 12510
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.3159252095422309e-05,
+      "loss": 0.0187,
+      "step": 12520
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.3143133462282398e-05,
+      "loss": 0.0201,
+      "step": 12530
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.312701482914249e-05,
+      "loss": 0.0088,
+      "step": 12540
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.311089619600258e-05,
+      "loss": 0.0197,
+      "step": 12550
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.309477756286267e-05,
+      "loss": 0.0152,
+      "step": 12560
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.3078658929722759e-05,
+      "loss": 0.0291,
+      "step": 12570
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.3062540296582849e-05,
+      "loss": 0.0195,
+      "step": 12580
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.3046421663442942e-05,
+      "loss": 0.0126,
+      "step": 12590
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.3030303030303032e-05,
+      "loss": 0.0216,
+      "step": 12600
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.3014184397163121e-05,
+      "loss": 0.0303,
+      "step": 12610
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.2998065764023211e-05,
+      "loss": 0.0203,
+      "step": 12620
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.2981947130883301e-05,
+      "loss": 0.0278,
+      "step": 12630
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.296582849774339e-05,
+      "loss": 0.0215,
+      "step": 12640
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.2949709864603482e-05,
+      "loss": 0.0152,
+      "step": 12650
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.2933591231463572e-05,
+      "loss": 0.0444,
+      "step": 12660
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.2917472598323662e-05,
+      "loss": 0.0292,
+      "step": 12670
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.2901353965183752e-05,
+      "loss": 0.0181,
+      "step": 12680
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.2885235332043843e-05,
+      "loss": 0.0154,
+      "step": 12690
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.2869116698903934e-05,
+      "loss": 0.0224,
+      "step": 12700
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.2852998065764024e-05,
+      "loss": 0.0313,
+      "step": 12710
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.2836879432624114e-05,
+      "loss": 0.0289,
+      "step": 12720
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.2820760799484204e-05,
+      "loss": 0.0171,
+      "step": 12730
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.2804642166344294e-05,
+      "loss": 0.0128,
+      "step": 12740
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.2788523533204385e-05,
+      "loss": 0.0171,
+      "step": 12750
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.2772404900064475e-05,
+      "loss": 0.0301,
+      "step": 12760
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.2756286266924565e-05,
+      "loss": 0.01,
+      "step": 12770
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.2740167633784656e-05,
+      "loss": 0.0251,
+      "step": 12780
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.2724049000644746e-05,
+      "loss": 0.0195,
+      "step": 12790
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.2707930367504837e-05,
+      "loss": 0.0277,
+      "step": 12800
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.2691811734364927e-05,
+      "loss": 0.0147,
+      "step": 12810
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.2675693101225017e-05,
+      "loss": 0.0337,
+      "step": 12820
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.2659574468085106e-05,
+      "loss": 0.0201,
+      "step": 12830
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.2643455834945196e-05,
+      "loss": 0.0273,
+      "step": 12840
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.2627337201805286e-05,
+      "loss": 0.0177,
+      "step": 12850
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.2611218568665377e-05,
+      "loss": 0.0363,
+      "step": 12860
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.2595099935525467e-05,
+      "loss": 0.0169,
+      "step": 12870
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.2578981302385559e-05,
+      "loss": 0.0111,
+      "step": 12880
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.2562862669245648e-05,
+      "loss": 0.027,
+      "step": 12890
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.2546744036105738e-05,
+      "loss": 0.0283,
+      "step": 12900
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.253223726627982e-05,
+      "loss": 0.0296,
+      "step": 12910
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.251611863313991e-05,
+      "loss": 0.0311,
+      "step": 12920
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.25e-05,
+      "loss": 0.0269,
+      "step": 12930
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.248388136686009e-05,
+      "loss": 0.0119,
+      "step": 12940
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.2467762733720182e-05,
+      "loss": 0.0224,
+      "step": 12950
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.2451644100580272e-05,
+      "loss": 0.0097,
+      "step": 12960
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.2435525467440361e-05,
+      "loss": 0.0363,
+      "step": 12970
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.2419406834300451e-05,
+      "loss": 0.0041,
+      "step": 12980
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.2403288201160541e-05,
+      "loss": 0.0171,
+      "step": 12990
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.238716956802063e-05,
+      "loss": 0.0167,
+      "step": 13000
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.2371050934880724e-05,
+      "loss": 0.0104,
+      "step": 13010
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.2354932301740814e-05,
+      "loss": 0.0259,
+      "step": 13020
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.2338813668600903e-05,
+      "loss": 0.0104,
+      "step": 13030
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.2322695035460993e-05,
+      "loss": 0.0111,
+      "step": 13040
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.2306576402321083e-05,
+      "loss": 0.0134,
+      "step": 13050
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.2290457769181174e-05,
+      "loss": 0.0207,
+      "step": 13060
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.2274339136041264e-05,
+      "loss": 0.0268,
+      "step": 13070
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.2258220502901354e-05,
+      "loss": 0.0188,
+      "step": 13080
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.2242101869761444e-05,
+      "loss": 0.0279,
+      "step": 13090
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.2225983236621533e-05,
+      "loss": 0.0176,
+      "step": 13100
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.2209864603481627e-05,
+      "loss": 0.0163,
+      "step": 13110
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.2193745970341716e-05,
+      "loss": 0.0188,
+      "step": 13120
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.2177627337201806e-05,
+      "loss": 0.0297,
+      "step": 13130
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.2161508704061896e-05,
+      "loss": 0.0207,
+      "step": 13140
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.2145390070921986e-05,
+      "loss": 0.0144,
+      "step": 13150
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.2129271437782077e-05,
+      "loss": 0.0519,
+      "step": 13160
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.2113152804642167e-05,
+      "loss": 0.0091,
+      "step": 13170
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.2097034171502257e-05,
+      "loss": 0.0215,
+      "step": 13180
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.2080915538362346e-05,
+      "loss": 0.0088,
+      "step": 13190
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.2064796905222436e-05,
+      "loss": 0.0113,
+      "step": 13200
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.2048678272082528e-05,
+      "loss": 0.019,
+      "step": 13210
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.2032559638942619e-05,
+      "loss": 0.016,
+      "step": 13220
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.2016441005802709e-05,
+      "loss": 0.014,
+      "step": 13230
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.2000322372662799e-05,
+      "loss": 0.02,
+      "step": 13240
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.1984203739522888e-05,
+      "loss": 0.0406,
+      "step": 13250
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.1968085106382978e-05,
+      "loss": 0.0149,
+      "step": 13260
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.195196647324307e-05,
+      "loss": 0.0265,
+      "step": 13270
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.193584784010316e-05,
+      "loss": 0.0146,
+      "step": 13280
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.1919729206963249e-05,
+      "loss": 0.0163,
+      "step": 13290
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.190361057382334e-05,
+      "loss": 0.016,
+      "step": 13300
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.188749194068343e-05,
+      "loss": 0.0197,
+      "step": 13310
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.1871373307543522e-05,
+      "loss": 0.0228,
+      "step": 13320
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.1855254674403612e-05,
+      "loss": 0.0134,
+      "step": 13330
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.1839136041263701e-05,
+      "loss": 0.0123,
+      "step": 13340
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1.1823017408123791e-05,
+      "loss": 0.0135,
+      "step": 13350
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1.180689877498388e-05,
+      "loss": 0.0346,
+      "step": 13360
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1.1790780141843972e-05,
+      "loss": 0.0312,
+      "step": 13370
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1.1774661508704062e-05,
+      "loss": 0.0201,
+      "step": 13380
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.1758542875564152e-05,
+      "loss": 0.0194,
+      "step": 13390
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.1742424242424243e-05,
+      "loss": 0.0117,
+      "step": 13400
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.1726305609284333e-05,
+      "loss": 0.0112,
+      "step": 13410
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.1710186976144423e-05,
+      "loss": 0.0239,
+      "step": 13420
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1.1694068343004514e-05,
+      "loss": 0.0151,
+      "step": 13430
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1.1677949709864604e-05,
+      "loss": 0.0343,
+      "step": 13440
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1.1661831076724694e-05,
+      "loss": 0.0141,
+      "step": 13450
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1.1645712443584784e-05,
+      "loss": 0.0142,
+      "step": 13460
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.1629593810444873e-05,
+      "loss": 0.0136,
+      "step": 13470
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.1613475177304965e-05,
+      "loss": 0.0237,
+      "step": 13480
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.1597356544165056e-05,
+      "loss": 0.0119,
+      "step": 13490
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.1581237911025146e-05,
+      "loss": 0.018,
+      "step": 13500
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.1565119277885236e-05,
+      "loss": 0.0134,
+      "step": 13510
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.1549000644745326e-05,
+      "loss": 0.049,
+      "step": 13520
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.1532882011605417e-05,
+      "loss": 0.0169,
+      "step": 13530
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.1516763378465507e-05,
+      "loss": 0.02,
+      "step": 13540
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.1500644745325597e-05,
+      "loss": 0.0185,
+      "step": 13550
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.1484526112185686e-05,
+      "loss": 0.0191,
+      "step": 13560
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.1468407479045776e-05,
+      "loss": 0.0334,
+      "step": 13570
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.1452288845905866e-05,
+      "loss": 0.031,
+      "step": 13580
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.1436170212765959e-05,
+      "loss": 0.0248,
+      "step": 13590
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.1420051579626049e-05,
+      "loss": 0.0218,
+      "step": 13600
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.1403932946486139e-05,
+      "loss": 0.032,
+      "step": 13610
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.1387814313346228e-05,
+      "loss": 0.0171,
+      "step": 13620
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.1371695680206318e-05,
+      "loss": 0.0191,
+      "step": 13630
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.135557704706641e-05,
+      "loss": 0.0333,
+      "step": 13640
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.13394584139265e-05,
+      "loss": 0.0098,
+      "step": 13650
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.1323339780786589e-05,
+      "loss": 0.0211,
+      "step": 13660
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.1307221147646679e-05,
+      "loss": 0.0147,
+      "step": 13670
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1.129110251450677e-05,
+      "loss": 0.0303,
+      "step": 13680
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1.1274983881366862e-05,
+      "loss": 0.013,
+      "step": 13690
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1.1258865248226952e-05,
+      "loss": 0.022,
+      "step": 13700
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1.1242746615087041e-05,
+      "loss": 0.008,
+      "step": 13710
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.1226627981947131e-05,
+      "loss": 0.0133,
+      "step": 13720
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.121050934880722e-05,
+      "loss": 0.0152,
+      "step": 13730
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.1194390715667312e-05,
+      "loss": 0.0178,
+      "step": 13740
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.1178272082527402e-05,
+      "loss": 0.0219,
+      "step": 13750
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1.1162153449387492e-05,
+      "loss": 0.0139,
+      "step": 13760
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1.1146034816247582e-05,
+      "loss": 0.0194,
+      "step": 13770
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1.1129916183107673e-05,
+      "loss": 0.0242,
+      "step": 13780
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1.1113797549967763e-05,
+      "loss": 0.0198,
+      "step": 13790
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.1097678916827854e-05,
+      "loss": 0.0341,
+      "step": 13800
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.1081560283687944e-05,
+      "loss": 0.0119,
+      "step": 13810
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.1065441650548034e-05,
+      "loss": 0.0223,
+      "step": 13820
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.1049323017408124e-05,
+      "loss": 0.021,
+      "step": 13830
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.1033204384268213e-05,
+      "loss": 0.0111,
+      "step": 13840
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.1017085751128305e-05,
+      "loss": 0.0134,
+      "step": 13850
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.1000967117988395e-05,
+      "loss": 0.0164,
+      "step": 13860
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.0984848484848486e-05,
+      "loss": 0.0143,
+      "step": 13870
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.0968729851708576e-05,
+      "loss": 0.0126,
+      "step": 13880
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.0952611218568666e-05,
+      "loss": 0.011,
+      "step": 13890
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.0936492585428757e-05,
+      "loss": 0.0156,
+      "step": 13900
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.0920373952288847e-05,
+      "loss": 0.0116,
+      "step": 13910
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.0904255319148937e-05,
+      "loss": 0.0125,
+      "step": 13920
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.0888136686009026e-05,
+      "loss": 0.0225,
+      "step": 13930
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.0872018052869116e-05,
+      "loss": 0.0021,
+      "step": 13940
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.0855899419729206e-05,
+      "loss": 0.0157,
+      "step": 13950
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.0839780786589297e-05,
+      "loss": 0.022,
+      "step": 13960
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.0823662153449389e-05,
+      "loss": 0.0172,
+      "step": 13970
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.0807543520309478e-05,
+      "loss": 0.0213,
+      "step": 13980
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.0791424887169568e-05,
+      "loss": 0.0199,
+      "step": 13990
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.0775306254029658e-05,
+      "loss": 0.0231,
+      "step": 14000
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1.075918762088975e-05,
+      "loss": 0.0129,
+      "step": 14010
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1.074306898774984e-05,
+      "loss": 0.0256,
+      "step": 14020
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1.0726950354609929e-05,
+      "loss": 0.0183,
+      "step": 14030
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1.0710831721470019e-05,
+      "loss": 0.0278,
+      "step": 14040
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 1.0694713088330109e-05,
+      "loss": 0.0222,
+      "step": 14050
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 1.0678594455190202e-05,
+      "loss": 0.018,
+      "step": 14060
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 1.0662475822050291e-05,
+      "loss": 0.0105,
+      "step": 14070
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 1.0646357188910381e-05,
+      "loss": 0.0141,
+      "step": 14080
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 1.0630238555770471e-05,
+      "loss": 0.01,
+      "step": 14090
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 1.061411992263056e-05,
+      "loss": 0.0062,
+      "step": 14100
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 1.0598001289490652e-05,
+      "loss": 0.0156,
+      "step": 14110
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 1.0581882656350742e-05,
+      "loss": 0.0279,
+      "step": 14120
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 1.0565764023210832e-05,
+      "loss": 0.0138,
+      "step": 14130
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 1.0549645390070922e-05,
+      "loss": 0.0139,
+      "step": 14140
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 1.0533526756931011e-05,
+      "loss": 0.0256,
+      "step": 14150
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 1.0517408123791103e-05,
+      "loss": 0.0326,
+      "step": 14160
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 1.0501289490651194e-05,
+      "loss": 0.0273,
+      "step": 14170
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 1.0485170857511284e-05,
+      "loss": 0.0222,
+      "step": 14180
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 1.0469052224371374e-05,
+      "loss": 0.0114,
+      "step": 14190
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 1.0452933591231463e-05,
+      "loss": 0.0303,
+      "step": 14200
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1.0436814958091553e-05,
+      "loss": 0.0217,
+      "step": 14210
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1.0420696324951645e-05,
+      "loss": 0.0158,
+      "step": 14220
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1.0404577691811734e-05,
+      "loss": 0.0142,
+      "step": 14230
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1.0388459058671824e-05,
+      "loss": 0.0159,
+      "step": 14240
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1.0372340425531916e-05,
+      "loss": 0.0299,
+      "step": 14250
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 1.0356221792392005e-05,
+      "loss": 0.0353,
+      "step": 14260
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 1.0340103159252097e-05,
+      "loss": 0.0119,
+      "step": 14270
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 1.0323984526112187e-05,
+      "loss": 0.0072,
+      "step": 14280
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 1.0307865892972276e-05,
+      "loss": 0.0078,
+      "step": 14290
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 1.0291747259832366e-05,
+      "loss": 0.0267,
+      "step": 14300
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 1.0275628626692456e-05,
+      "loss": 0.0123,
+      "step": 14310
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 1.0259509993552546e-05,
+      "loss": 0.0104,
+      "step": 14320
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 1.0243391360412637e-05,
+      "loss": 0.0126,
+      "step": 14330
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 1.0227272727272727e-05,
+      "loss": 0.0178,
+      "step": 14340
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 1.0211154094132818e-05,
+      "loss": 0.0307,
+      "step": 14350
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 1.0195035460992908e-05,
+      "loss": 0.014,
+      "step": 14360
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 1.0178916827852998e-05,
+      "loss": 0.0266,
+      "step": 14370
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 1.016279819471309e-05,
+      "loss": 0.0122,
+      "step": 14380
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 1.014667956157318e-05,
+      "loss": 0.0119,
+      "step": 14390
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 1.0130560928433269e-05,
+      "loss": 0.0229,
+      "step": 14400
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 1.0114442295293359e-05,
+      "loss": 0.0107,
+      "step": 14410
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 1.0098323662153449e-05,
+      "loss": 0.0168,
+      "step": 14420
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 1.008220502901354e-05,
+      "loss": 0.018,
+      "step": 14430
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 1.0066086395873631e-05,
+      "loss": 0.0247,
+      "step": 14440
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 1.0049967762733721e-05,
+      "loss": 0.0229,
+      "step": 14450
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 1.0033849129593811e-05,
+      "loss": 0.028,
+      "step": 14460
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 1.00177304964539e-05,
+      "loss": 0.0156,
+      "step": 14470
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 1.0001611863313992e-05,
+      "loss": 0.0237,
+      "step": 14480
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 9.985493230174082e-06,
+      "loss": 0.0127,
+      "step": 14490
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 9.969374597034172e-06,
+      "loss": 0.0325,
+      "step": 14500
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 9.953255963894261e-06,
+      "loss": 0.0142,
+      "step": 14510
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 9.937137330754351e-06,
+      "loss": 0.0064,
+      "step": 14520
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 9.921018697614441e-06,
+      "loss": 0.0201,
+      "step": 14530
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 9.904900064474534e-06,
+      "loss": 0.0129,
+      "step": 14540
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 9.888781431334624e-06,
+      "loss": 0.0189,
+      "step": 14550
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 9.872662798194714e-06,
+      "loss": 0.0184,
+      "step": 14560
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 9.856544165054803e-06,
+      "loss": 0.0111,
+      "step": 14570
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 9.840425531914893e-06,
+      "loss": 0.0195,
+      "step": 14580
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 9.824306898774985e-06,
+      "loss": 0.0224,
+      "step": 14590
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 9.808188265635074e-06,
+      "loss": 0.0175,
+      "step": 14600
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 9.792069632495164e-06,
+      "loss": 0.0216,
+      "step": 14610
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 9.775950999355254e-06,
+      "loss": 0.0223,
+      "step": 14620
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 9.759832366215344e-06,
+      "loss": 0.0111,
+      "step": 14630
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 9.743713733075437e-06,
+      "loss": 0.018,
+      "step": 14640
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 9.727595099935527e-06,
+      "loss": 0.0185,
+      "step": 14650
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 9.711476466795616e-06,
+      "loss": 0.0134,
+      "step": 14660
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 9.695357833655706e-06,
+      "loss": 0.016,
+      "step": 14670
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 9.679239200515796e-06,
+      "loss": 0.0051,
+      "step": 14680
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 9.663120567375886e-06,
+      "loss": 0.0152,
+      "step": 14690
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 9.647001934235977e-06,
+      "loss": 0.0189,
+      "step": 14700
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 9.630883301096067e-06,
+      "loss": 0.0138,
+      "step": 14710
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 9.614764667956157e-06,
+      "loss": 0.0145,
+      "step": 14720
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 9.598646034816248e-06,
+      "loss": 0.0292,
+      "step": 14730
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 9.582527401676338e-06,
+      "loss": 0.0176,
+      "step": 14740
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 9.56640876853643e-06,
+      "loss": 0.0086,
+      "step": 14750
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 9.55029013539652e-06,
+      "loss": 0.0169,
+      "step": 14760
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 9.534171502256609e-06,
+      "loss": 0.0143,
+      "step": 14770
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 9.518052869116699e-06,
+      "loss": 0.0178,
+      "step": 14780
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 9.501934235976788e-06,
+      "loss": 0.0163,
+      "step": 14790
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 9.48581560283688e-06,
+      "loss": 0.0148,
+      "step": 14800
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 9.46969696969697e-06,
+      "loss": 0.0056,
+      "step": 14810
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 9.45357833655706e-06,
+      "loss": 0.016,
+      "step": 14820
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 9.437459703417151e-06,
+      "loss": 0.021,
+      "step": 14830
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 9.42134107027724e-06,
+      "loss": 0.0165,
+      "step": 14840
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 9.405222437137332e-06,
+      "loss": 0.0132,
+      "step": 14850
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 9.389103803997422e-06,
+      "loss": 0.008,
+      "step": 14860
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 9.372985170857512e-06,
+      "loss": 0.023,
+      "step": 14870
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 9.356866537717601e-06,
+      "loss": 0.0275,
+      "step": 14880
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 9.340747904577691e-06,
+      "loss": 0.0397,
+      "step": 14890
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 9.324629271437781e-06,
+      "loss": 0.0138,
+      "step": 14900
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 9.308510638297872e-06,
+      "loss": 0.0093,
+      "step": 14910
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 9.292392005157964e-06,
+      "loss": 0.0292,
+      "step": 14920
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 9.276273372018054e-06,
+      "loss": 0.0222,
+      "step": 14930
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 9.260154738878143e-06,
+      "loss": 0.0099,
+      "step": 14940
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 9.244036105738233e-06,
+      "loss": 0.0272,
+      "step": 14950
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 9.227917472598325e-06,
+      "loss": 0.019,
+      "step": 14960
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 9.211798839458414e-06,
+      "loss": 0.0156,
+      "step": 14970
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 9.195680206318504e-06,
+      "loss": 0.0263,
+      "step": 14980
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 9.179561573178594e-06,
+      "loss": 0.0153,
+      "step": 14990
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 9.163442940038684e-06,
+      "loss": 0.0128,
+      "step": 15000
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 9.147324306898775e-06,
+      "loss": 0.0193,
+      "step": 15010
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 9.131205673758867e-06,
+      "loss": 0.0211,
+      "step": 15020
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 9.115087040618956e-06,
+      "loss": 0.0214,
+      "step": 15030
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 9.098968407479046e-06,
+      "loss": 0.0031,
+      "step": 15040
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 9.082849774339136e-06,
+      "loss": 0.0139,
+      "step": 15050
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 9.066731141199226e-06,
+      "loss": 0.0293,
+      "step": 15060
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 9.050612508059317e-06,
+      "loss": 0.0159,
+      "step": 15070
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 9.034493874919407e-06,
+      "loss": 0.0105,
+      "step": 15080
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 9.018375241779497e-06,
+      "loss": 0.0151,
+      "step": 15090
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 9.002256608639586e-06,
+      "loss": 0.0084,
+      "step": 15100
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 8.986137975499678e-06,
+      "loss": 0.0224,
+      "step": 15110
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 8.97001934235977e-06,
+      "loss": 0.0072,
+      "step": 15120
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 8.953900709219859e-06,
+      "loss": 0.0066,
+      "step": 15130
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 8.937782076079949e-06,
+      "loss": 0.0152,
+      "step": 15140
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 8.921663442940039e-06,
+      "loss": 0.0117,
+      "step": 15150
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 8.905544809800128e-06,
+      "loss": 0.0184,
+      "step": 15160
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 8.88942617666022e-06,
+      "loss": 0.0128,
+      "step": 15170
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 8.87330754352031e-06,
+      "loss": 0.0192,
+      "step": 15180
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 8.8571889103804e-06,
+      "loss": 0.0109,
+      "step": 15190
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 8.84107027724049e-06,
+      "loss": 0.0143,
+      "step": 15200
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 8.82495164410058e-06,
+      "loss": 0.0239,
+      "step": 15210
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 8.808833010960672e-06,
+      "loss": 0.02,
+      "step": 15220
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 8.792714377820762e-06,
+      "loss": 0.0102,
+      "step": 15230
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 8.776595744680852e-06,
+      "loss": 0.0174,
+      "step": 15240
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 8.760477111540941e-06,
+      "loss": 0.0145,
+      "step": 15250
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 8.744358478401031e-06,
+      "loss": 0.0185,
+      "step": 15260
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 8.728239845261121e-06,
+      "loss": 0.0327,
+      "step": 15270
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 8.713733075435204e-06,
+      "loss": 0.0128,
+      "step": 15280
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 8.697614442295294e-06,
+      "loss": 0.0158,
+      "step": 15290
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 8.681495809155383e-06,
+      "loss": 0.0216,
+      "step": 15300
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 8.665377176015473e-06,
+      "loss": 0.0097,
+      "step": 15310
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 8.649258542875565e-06,
+      "loss": 0.0258,
+      "step": 15320
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 8.633139909735654e-06,
+      "loss": 0.0182,
+      "step": 15330
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 8.617021276595746e-06,
+      "loss": 0.0143,
+      "step": 15340
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 8.600902643455835e-06,
+      "loss": 0.0205,
+      "step": 15350
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 8.584784010315925e-06,
+      "loss": 0.0135,
+      "step": 15360
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 8.568665377176017e-06,
+      "loss": 0.0206,
+      "step": 15370
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 8.552546744036106e-06,
+      "loss": 0.0138,
+      "step": 15380
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 8.536428110896196e-06,
+      "loss": 0.0114,
+      "step": 15390
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 8.520309477756286e-06,
+      "loss": 0.0149,
+      "step": 15400
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 8.504190844616376e-06,
+      "loss": 0.0261,
+      "step": 15410
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 8.488072211476466e-06,
+      "loss": 0.0147,
+      "step": 15420
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 8.471953578336557e-06,
+      "loss": 0.0098,
+      "step": 15430
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 8.455834945196648e-06,
+      "loss": 0.0148,
+      "step": 15440
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 8.439716312056738e-06,
+      "loss": 0.0157,
+      "step": 15450
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 8.423597678916828e-06,
+      "loss": 0.0093,
+      "step": 15460
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 8.407479045776918e-06,
+      "loss": 0.0106,
+      "step": 15470
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 8.39136041263701e-06,
+      "loss": 0.009,
+      "step": 15480
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 8.375241779497099e-06,
+      "loss": 0.0138,
+      "step": 15490
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 8.359123146357189e-06,
+      "loss": 0.0161,
+      "step": 15500
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 8.343004513217279e-06,
+      "loss": 0.0075,
+      "step": 15510
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 8.326885880077368e-06,
+      "loss": 0.0076,
+      "step": 15520
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 8.312379110251451e-06,
+      "loss": 0.0247,
+      "step": 15530
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 8.296260477111541e-06,
+      "loss": 0.0059,
+      "step": 15540
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 8.28014184397163e-06,
+      "loss": 0.0222,
+      "step": 15550
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 8.26402321083172e-06,
+      "loss": 0.0097,
+      "step": 15560
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 8.247904577691814e-06,
+      "loss": 0.0144,
+      "step": 15570
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 8.231785944551903e-06,
+      "loss": 0.013,
+      "step": 15580
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 8.215667311411993e-06,
+      "loss": 0.0164,
+      "step": 15590
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 8.199548678272083e-06,
+      "loss": 0.0205,
+      "step": 15600
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 8.183430045132173e-06,
+      "loss": 0.0232,
+      "step": 15610
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 8.167311411992262e-06,
+      "loss": 0.0133,
+      "step": 15620
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 8.151192778852354e-06,
+      "loss": 0.0157,
+      "step": 15630
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 8.135074145712444e-06,
+      "loss": 0.0207,
+      "step": 15640
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 8.118955512572533e-06,
+      "loss": 0.0239,
+      "step": 15650
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 8.102836879432623e-06,
+      "loss": 0.0129,
+      "step": 15660
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 8.086718246292715e-06,
+      "loss": 0.0114,
+      "step": 15670
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 8.070599613152806e-06,
+      "loss": 0.0123,
+      "step": 15680
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 8.054480980012896e-06,
+      "loss": 0.0212,
+      "step": 15690
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 8.038362346872986e-06,
+      "loss": 0.0123,
+      "step": 15700
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 8.022243713733075e-06,
+      "loss": 0.0197,
+      "step": 15710
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 8.006125080593165e-06,
+      "loss": 0.0109,
+      "step": 15720
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 7.990006447453257e-06,
+      "loss": 0.0088,
+      "step": 15730
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 7.973887814313346e-06,
+      "loss": 0.0138,
+      "step": 15740
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 7.957769181173436e-06,
+      "loss": 0.0101,
+      "step": 15750
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 7.941650548033528e-06,
+      "loss": 0.0143,
+      "step": 15760
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 7.925531914893617e-06,
+      "loss": 0.0146,
+      "step": 15770
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 7.909413281753707e-06,
+      "loss": 0.0181,
+      "step": 15780
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 7.893294648613799e-06,
+      "loss": 0.0136,
+      "step": 15790
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 7.877176015473888e-06,
+      "loss": 0.0057,
+      "step": 15800
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 7.861057382333978e-06,
+      "loss": 0.0072,
+      "step": 15810
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 7.844938749194068e-06,
+      "loss": 0.0207,
+      "step": 15820
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 7.828820116054158e-06,
+      "loss": 0.025,
+      "step": 15830
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 7.812701482914249e-06,
+      "loss": 0.0301,
+      "step": 15840
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 7.796582849774339e-06,
+      "loss": 0.0121,
+      "step": 15850
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 7.78046421663443e-06,
+      "loss": 0.0284,
+      "step": 15860
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 7.76434558349452e-06,
+      "loss": 0.013,
+      "step": 15870
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 7.74822695035461e-06,
+      "loss": 0.0128,
+      "step": 15880
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 7.732108317214701e-06,
+      "loss": 0.0186,
+      "step": 15890
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 7.715989684074791e-06,
+      "loss": 0.0097,
+      "step": 15900
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 7.69987105093488e-06,
+      "loss": 0.0142,
+      "step": 15910
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 7.68375241779497e-06,
+      "loss": 0.0167,
+      "step": 15920
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 7.66763378465506e-06,
+      "loss": 0.0149,
+      "step": 15930
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 7.651515151515152e-06,
+      "loss": 0.0214,
+      "step": 15940
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 7.635396518375243e-06,
+      "loss": 0.0179,
+      "step": 15950
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 7.619277885235333e-06,
+      "loss": 0.0334,
+      "step": 15960
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 7.603159252095423e-06,
+      "loss": 0.0045,
+      "step": 15970
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 7.587040618955513e-06,
+      "loss": 0.0191,
+      "step": 15980
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 7.570921985815603e-06,
+      "loss": 0.0221,
+      "step": 15990
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 7.554803352675693e-06,
+      "loss": 0.0331,
+      "step": 16000
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 7.5386847195357836e-06,
+      "loss": 0.0073,
+      "step": 16010
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 7.522566086395873e-06,
+      "loss": 0.022,
+      "step": 16020
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 7.506447453255963e-06,
+      "loss": 0.0127,
+      "step": 16030
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 7.4903288201160546e-06,
+      "loss": 0.0022,
+      "step": 16040
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 7.474210186976144e-06,
+      "loss": 0.0117,
+      "step": 16050
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 7.458091553836235e-06,
+      "loss": 0.0144,
+      "step": 16060
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 7.4419729206963255e-06,
+      "loss": 0.0305,
+      "step": 16070
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 7.425854287556415e-06,
+      "loss": 0.0213,
+      "step": 16080
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 7.409735654416506e-06,
+      "loss": 0.024,
+      "step": 16090
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 7.393617021276596e-06,
+      "loss": 0.0046,
+      "step": 16100
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 7.3774983881366855e-06,
+      "loss": 0.0153,
+      "step": 16110
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 7.361379754996777e-06,
+      "loss": 0.017,
+      "step": 16120
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 7.345261121856867e-06,
+      "loss": 0.0275,
+      "step": 16130
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 7.329142488716957e-06,
+      "loss": 0.0251,
+      "step": 16140
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 7.313023855577047e-06,
+      "loss": 0.0284,
+      "step": 16150
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 7.296905222437137e-06,
+      "loss": 0.0081,
+      "step": 16160
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 7.280786589297228e-06,
+      "loss": 0.0103,
+      "step": 16170
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 7.264667956157318e-06,
+      "loss": 0.0214,
+      "step": 16180
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 7.248549323017409e-06,
+      "loss": 0.0185,
+      "step": 16190
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 7.2324306898774984e-06,
+      "loss": 0.0113,
+      "step": 16200
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 7.216312056737588e-06,
+      "loss": 0.0094,
+      "step": 16210
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 7.20019342359768e-06,
+      "loss": 0.0057,
+      "step": 16220
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 7.184074790457769e-06,
+      "loss": 0.0144,
+      "step": 16230
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 7.167956157317859e-06,
+      "loss": 0.0072,
+      "step": 16240
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 7.15183752417795e-06,
+      "loss": 0.0069,
+      "step": 16250
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 7.13571889103804e-06,
+      "loss": 0.0152,
+      "step": 16260
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 7.119600257898131e-06,
+      "loss": 0.0156,
+      "step": 16270
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 7.103481624758221e-06,
+      "loss": 0.0142,
+      "step": 16280
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 7.0873629916183106e-06,
+      "loss": 0.0082,
+      "step": 16290
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 7.071244358478401e-06,
+      "loss": 0.0192,
+      "step": 16300
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 7.055125725338492e-06,
+      "loss": 0.0085,
+      "step": 16310
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 7.0390070921985815e-06,
+      "loss": 0.0186,
+      "step": 16320
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 7.022888459058672e-06,
+      "loss": 0.0107,
+      "step": 16330
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 7.006769825918762e-06,
+      "loss": 0.013,
+      "step": 16340
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 6.9906511927788525e-06,
+      "loss": 0.0078,
+      "step": 16350
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 6.974532559638943e-06,
+      "loss": 0.0161,
+      "step": 16360
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 6.958413926499033e-06,
+      "loss": 0.0183,
+      "step": 16370
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 6.9422952933591235e-06,
+      "loss": 0.0122,
+      "step": 16380
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 6.926176660219213e-06,
+      "loss": 0.0087,
+      "step": 16390
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 6.910058027079303e-06,
+      "loss": 0.0213,
+      "step": 16400
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 6.8939393939393945e-06,
+      "loss": 0.0064,
+      "step": 16410
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 6.877820760799484e-06,
+      "loss": 0.0235,
+      "step": 16420
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 6.861702127659575e-06,
+      "loss": 0.0046,
+      "step": 16430
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 6.845583494519665e-06,
+      "loss": 0.0116,
+      "step": 16440
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 6.829464861379755e-06,
+      "loss": 0.0317,
+      "step": 16450
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 6.813346228239846e-06,
+      "loss": 0.0189,
+      "step": 16460
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 6.797227595099936e-06,
+      "loss": 0.0172,
+      "step": 16470
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 6.781108961960025e-06,
+      "loss": 0.0219,
+      "step": 16480
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 6.764990328820116e-06,
+      "loss": 0.0088,
+      "step": 16490
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 6.748871695680207e-06,
+      "loss": 0.0128,
+      "step": 16500
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 6.732753062540297e-06,
+      "loss": 0.0143,
+      "step": 16510
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 6.716634429400387e-06,
+      "loss": 0.024,
+      "step": 16520
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 6.700515796260477e-06,
+      "loss": 0.0169,
+      "step": 16530
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 6.684397163120567e-06,
+      "loss": 0.023,
+      "step": 16540
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.9982323146479927,
+      "eval_loss": 0.007050972431898117,
+      "eval_runtime": 68.6203,
+      "eval_samples_per_second": 857.384,
+      "eval_steps_per_second": 26.8,
+      "step": 16547
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 6.668278529980658e-06,
+      "loss": 0.0155,
+      "step": 16550
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 6.652159896840749e-06,
+      "loss": 0.007,
+      "step": 16560
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 6.636041263700838e-06,
+      "loss": 0.0138,
+      "step": 16570
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 6.619922630560928e-06,
+      "loss": 0.0104,
+      "step": 16580
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 6.603803997421019e-06,
+      "loss": 0.0072,
+      "step": 16590
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 6.587685364281109e-06,
+      "loss": 0.008,
+      "step": 16600
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 6.571566731141199e-06,
+      "loss": 0.0058,
+      "step": 16610
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 6.55544809800129e-06,
+      "loss": 0.0083,
+      "step": 16620
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 6.5393294648613795e-06,
+      "loss": 0.0078,
+      "step": 16630
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 6.523210831721471e-06,
+      "loss": 0.0224,
+      "step": 16640
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 6.507092198581561e-06,
+      "loss": 0.0091,
+      "step": 16650
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 6.4909735654416505e-06,
+      "loss": 0.0137,
+      "step": 16660
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 6.474854932301741e-06,
+      "loss": 0.0185,
+      "step": 16670
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 6.458736299161831e-06,
+      "loss": 0.0144,
+      "step": 16680
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 6.4426176660219215e-06,
+      "loss": 0.0421,
+      "step": 16690
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 6.426499032882012e-06,
+      "loss": 0.0141,
+      "step": 16700
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 6.410380399742102e-06,
+      "loss": 0.0226,
+      "step": 16710
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 6.3942617666021925e-06,
+      "loss": 0.0137,
+      "step": 16720
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 6.378143133462282e-06,
+      "loss": 0.0287,
+      "step": 16730
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 6.362024500322373e-06,
+      "loss": 0.0118,
+      "step": 16740
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 6.3459058671824635e-06,
+      "loss": 0.0058,
+      "step": 16750
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 6.329787234042553e-06,
+      "loss": 0.0221,
+      "step": 16760
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 6.313668600902643e-06,
+      "loss": 0.0083,
+      "step": 16770
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 6.297549967762734e-06,
+      "loss": 0.0069,
+      "step": 16780
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 6.281431334622824e-06,
+      "loss": 0.0173,
+      "step": 16790
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 6.265312701482915e-06,
+      "loss": 0.005,
+      "step": 16800
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 6.249194068343005e-06,
+      "loss": 0.0071,
+      "step": 16810
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 6.233075435203094e-06,
+      "loss": 0.011,
+      "step": 16820
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 6.216956802063185e-06,
+      "loss": 0.0199,
+      "step": 16830
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 6.200838168923276e-06,
+      "loss": 0.0121,
+      "step": 16840
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 6.184719535783365e-06,
+      "loss": 0.0085,
+      "step": 16850
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 6.168600902643456e-06,
+      "loss": 0.0196,
+      "step": 16860
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 6.152482269503546e-06,
+      "loss": 0.0147,
+      "step": 16870
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 6.136363636363637e-06,
+      "loss": 0.0218,
+      "step": 16880
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 6.120245003223727e-06,
+      "loss": 0.0127,
+      "step": 16890
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 6.104126370083817e-06,
+      "loss": 0.0215,
+      "step": 16900
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 6.088007736943907e-06,
+      "loss": 0.0056,
+      "step": 16910
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 6.071889103803997e-06,
+      "loss": 0.0092,
+      "step": 16920
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 6.055770470664089e-06,
+      "loss": 0.0081,
+      "step": 16930
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 6.039651837524178e-06,
+      "loss": 0.0079,
+      "step": 16940
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 6.023533204384268e-06,
+      "loss": 0.0245,
+      "step": 16950
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 6.007414571244359e-06,
+      "loss": 0.0071,
+      "step": 16960
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 5.9912959381044485e-06,
+      "loss": 0.0142,
+      "step": 16970
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 5.975177304964539e-06,
+      "loss": 0.0047,
+      "step": 16980
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 5.95905867182463e-06,
+      "loss": 0.0087,
+      "step": 16990
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 5.9429400386847195e-06,
+      "loss": 0.0121,
+      "step": 17000
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 5.92682140554481e-06,
+      "loss": 0.0327,
+      "step": 17010
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 5.9107027724049e-06,
+      "loss": 0.0215,
+      "step": 17020
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 5.8945841392649905e-06,
+      "loss": 0.0168,
+      "step": 17030
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 5.878465506125081e-06,
+      "loss": 0.011,
+      "step": 17040
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 5.862346872985171e-06,
+      "loss": 0.012,
+      "step": 17050
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 5.846228239845261e-06,
+      "loss": 0.0202,
+      "step": 17060
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 5.830109606705352e-06,
+      "loss": 0.0205,
+      "step": 17070
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 5.813990973565442e-06,
+      "loss": 0.0144,
+      "step": 17080
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 5.7978723404255325e-06,
+      "loss": 0.0169,
+      "step": 17090
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 5.781753707285622e-06,
+      "loss": 0.0143,
+      "step": 17100
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 5.765635074145712e-06,
+      "loss": 0.0139,
+      "step": 17110
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 5.7495164410058034e-06,
+      "loss": 0.0208,
+      "step": 17120
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 5.733397807865893e-06,
+      "loss": 0.0141,
+      "step": 17130
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 5.717279174725983e-06,
+      "loss": 0.0136,
+      "step": 17140
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 5.701160541586074e-06,
+      "loss": 0.008,
+      "step": 17150
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 5.685041908446163e-06,
+      "loss": 0.0119,
+      "step": 17160
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 5.668923275306255e-06,
+      "loss": 0.0094,
+      "step": 17170
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 5.652804642166345e-06,
+      "loss": 0.0111,
+      "step": 17180
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 5.636686009026434e-06,
+      "loss": 0.0156,
+      "step": 17190
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 5.620567375886525e-06,
+      "loss": 0.0335,
+      "step": 17200
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 5.604448742746615e-06,
+      "loss": 0.0153,
+      "step": 17210
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 5.588330109606706e-06,
+      "loss": 0.0152,
+      "step": 17220
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 5.572211476466796e-06,
+      "loss": 0.0103,
+      "step": 17230
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 5.556092843326886e-06,
+      "loss": 0.012,
+      "step": 17240
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 5.539974210186976e-06,
+      "loss": 0.0147,
+      "step": 17250
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 5.523855577047067e-06,
+      "loss": 0.025,
+      "step": 17260
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 5.507736943907157e-06,
+      "loss": 0.019,
+      "step": 17270
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 5.491618310767247e-06,
+      "loss": 0.0093,
+      "step": 17280
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 5.475499677627337e-06,
+      "loss": 0.0231,
+      "step": 17290
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 5.459381044487428e-06,
+      "loss": 0.0131,
+      "step": 17300
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 5.443262411347518e-06,
+      "loss": 0.0124,
+      "step": 17310
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 5.427143778207608e-06,
+      "loss": 0.0145,
+      "step": 17320
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 5.411025145067699e-06,
+      "loss": 0.0299,
+      "step": 17330
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 5.3949065119277884e-06,
+      "loss": 0.0095,
+      "step": 17340
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 5.378787878787878e-06,
+      "loss": 0.0067,
+      "step": 17350
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 5.36266924564797e-06,
+      "loss": 0.0148,
+      "step": 17360
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 5.3465506125080594e-06,
+      "loss": 0.0163,
+      "step": 17370
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 5.33043197936815e-06,
+      "loss": 0.0149,
+      "step": 17380
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 5.31431334622824e-06,
+      "loss": 0.0158,
+      "step": 17390
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 5.29819471308833e-06,
+      "loss": 0.0167,
+      "step": 17400
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 5.282076079948421e-06,
+      "loss": 0.0092,
+      "step": 17410
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 5.265957446808511e-06,
+      "loss": 0.0141,
+      "step": 17420
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 5.2498388136686006e-06,
+      "loss": 0.0164,
+      "step": 17430
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 5.233720180528691e-06,
+      "loss": 0.0198,
+      "step": 17440
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 5.217601547388782e-06,
+      "loss": 0.0043,
+      "step": 17450
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 5.201482914248872e-06,
+      "loss": 0.0158,
+      "step": 17460
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 5.185364281108962e-06,
+      "loss": 0.0104,
+      "step": 17470
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 5.169245647969052e-06,
+      "loss": 0.0165,
+      "step": 17480
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 5.1531270148291426e-06,
+      "loss": 0.0168,
+      "step": 17490
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 5.137008381689233e-06,
+      "loss": 0.0091,
+      "step": 17500
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 5.120889748549323e-06,
+      "loss": 0.0168,
+      "step": 17510
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 5.1047711154094135e-06,
+      "loss": 0.0167,
+      "step": 17520
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 5.088652482269503e-06,
+      "loss": 0.0146,
+      "step": 17530
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 5.072533849129594e-06,
+      "loss": 0.0218,
+      "step": 17540
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 5.0564152159896845e-06,
+      "loss": 0.0114,
+      "step": 17550
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 5.040296582849774e-06,
+      "loss": 0.0103,
+      "step": 17560
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 5.024177949709865e-06,
+      "loss": 0.008,
+      "step": 17570
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 5.008059316569955e-06,
+      "loss": 0.0096,
+      "step": 17580
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 4.991940683430045e-06,
+      "loss": 0.0038,
+      "step": 17590
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 4.975822050290136e-06,
+      "loss": 0.0171,
+      "step": 17600
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 4.959703417150226e-06,
+      "loss": 0.02,
+      "step": 17610
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 4.943584784010316e-06,
+      "loss": 0.0208,
+      "step": 17620
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 4.927466150870406e-06,
+      "loss": 0.0205,
+      "step": 17630
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 4.911347517730496e-06,
+      "loss": 0.0249,
+      "step": 17640
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 4.895228884590587e-06,
+      "loss": 0.0116,
+      "step": 17650
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 4.879110251450677e-06,
+      "loss": 0.0085,
+      "step": 17660
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 4.862991618310768e-06,
+      "loss": 0.0124,
+      "step": 17670
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 4.846872985170857e-06,
+      "loss": 0.0139,
+      "step": 17680
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 4.830754352030948e-06,
+      "loss": 0.0125,
+      "step": 17690
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 4.814635718891039e-06,
+      "loss": 0.022,
+      "step": 17700
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 4.798517085751128e-06,
+      "loss": 0.0162,
+      "step": 17710
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 4.782398452611218e-06,
+      "loss": 0.0088,
+      "step": 17720
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 4.766279819471309e-06,
+      "loss": 0.0214,
+      "step": 17730
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 4.750161186331399e-06,
+      "loss": 0.009,
+      "step": 17740
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 4.73404255319149e-06,
+      "loss": 0.0059,
+      "step": 17750
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 4.71792392005158e-06,
+      "loss": 0.0139,
+      "step": 17760
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 4.7018052869116695e-06,
+      "loss": 0.0168,
+      "step": 17770
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 4.68568665377176e-06,
+      "loss": 0.0184,
+      "step": 17780
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 4.669568020631851e-06,
+      "loss": 0.0076,
+      "step": 17790
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 4.6534493874919405e-06,
+      "loss": 0.0141,
+      "step": 17800
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 4.637330754352031e-06,
+      "loss": 0.0118,
+      "step": 17810
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 4.621212121212121e-06,
+      "loss": 0.0084,
+      "step": 17820
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 4.6050934880722115e-06,
+      "loss": 0.0171,
+      "step": 17830
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 4.588974854932302e-06,
+      "loss": 0.0169,
+      "step": 17840
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 4.572856221792392e-06,
+      "loss": 0.0062,
+      "step": 17850
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 4.5567375886524825e-06,
+      "loss": 0.0157,
+      "step": 17860
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 4.540618955512572e-06,
+      "loss": 0.0071,
+      "step": 17870
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 4.524500322372663e-06,
+      "loss": 0.0059,
+      "step": 17880
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 4.5083816892327535e-06,
+      "loss": 0.006,
+      "step": 17890
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 4.492263056092843e-06,
+      "loss": 0.0135,
+      "step": 17900
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 4.476144422952934e-06,
+      "loss": 0.0123,
+      "step": 17910
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 4.460025789813024e-06,
+      "loss": 0.012,
+      "step": 17920
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 4.443907156673114e-06,
+      "loss": 0.0175,
+      "step": 17930
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 4.427788523533205e-06,
+      "loss": 0.0126,
+      "step": 17940
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 4.411669890393295e-06,
+      "loss": 0.0162,
+      "step": 17950
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 4.395551257253385e-06,
+      "loss": 0.0114,
+      "step": 17960
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 4.379432624113475e-06,
+      "loss": 0.0176,
+      "step": 17970
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 4.363313990973566e-06,
+      "loss": 0.009,
+      "step": 17980
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 4.347195357833656e-06,
+      "loss": 0.0145,
+      "step": 17990
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 4.331076724693746e-06,
+      "loss": 0.008,
+      "step": 18000
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 4.314958091553836e-06,
+      "loss": 0.0087,
+      "step": 18010
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 4.298839458413926e-06,
+      "loss": 0.0124,
+      "step": 18020
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 4.282720825274017e-06,
+      "loss": 0.0089,
+      "step": 18030
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 4.266602192134108e-06,
+      "loss": 0.0048,
+      "step": 18040
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 4.250483558994197e-06,
+      "loss": 0.0158,
+      "step": 18050
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 4.234364925854287e-06,
+      "loss": 0.0039,
+      "step": 18060
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 4.218246292714379e-06,
+      "loss": 0.0037,
+      "step": 18070
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 4.202127659574468e-06,
+      "loss": 0.0195,
+      "step": 18080
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 4.186009026434558e-06,
+      "loss": 0.009,
+      "step": 18090
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 4.169890393294649e-06,
+      "loss": 0.0151,
+      "step": 18100
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 4.1537717601547385e-06,
+      "loss": 0.0134,
+      "step": 18110
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 4.13765312701483e-06,
+      "loss": 0.0123,
+      "step": 18120
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 4.12153449387492e-06,
+      "loss": 0.01,
+      "step": 18130
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 4.1054158607350095e-06,
+      "loss": 0.0154,
+      "step": 18140
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 4.0892972275951e-06,
+      "loss": 0.0022,
+      "step": 18150
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 4.07317859445519e-06,
+      "loss": 0.0114,
+      "step": 18160
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 4.0570599613152805e-06,
+      "loss": 0.02,
+      "step": 18170
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 4.040941328175371e-06,
+      "loss": 0.0076,
+      "step": 18180
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 4.024822695035461e-06,
+      "loss": 0.0107,
+      "step": 18190
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 4.0087040618955515e-06,
+      "loss": 0.0086,
+      "step": 18200
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 3.992585428755641e-06,
+      "loss": 0.0236,
+      "step": 18210
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 3.976466795615732e-06,
+      "loss": 0.0061,
+      "step": 18220
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 3.9603481624758225e-06,
+      "loss": 0.0082,
+      "step": 18230
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 3.944229529335912e-06,
+      "loss": 0.0127,
+      "step": 18240
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 3.928110896196003e-06,
+      "loss": 0.0102,
+      "step": 18250
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 3.911992263056093e-06,
+      "loss": 0.0139,
+      "step": 18260
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 3.895873629916183e-06,
+      "loss": 0.0096,
+      "step": 18270
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 3.879754996776274e-06,
+      "loss": 0.0094,
+      "step": 18280
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 3.863636363636364e-06,
+      "loss": 0.0084,
+      "step": 18290
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 3.847517730496453e-06,
+      "loss": 0.0113,
+      "step": 18300
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 3.831399097356545e-06,
+      "loss": 0.0095,
+      "step": 18310
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 3.815280464216635e-06,
+      "loss": 0.0039,
+      "step": 18320
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 3.7991618310767248e-06,
+      "loss": 0.0137,
+      "step": 18330
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 3.783043197936815e-06,
+      "loss": 0.0044,
+      "step": 18340
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 3.766924564796905e-06,
+      "loss": 0.0078,
+      "step": 18350
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 3.7508059316569958e-06,
+      "loss": 0.0114,
+      "step": 18360
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 3.734687298517086e-06,
+      "loss": 0.0122,
+      "step": 18370
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 3.718568665377176e-06,
+      "loss": 0.0141,
+      "step": 18380
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 3.7024500322372663e-06,
+      "loss": 0.0079,
+      "step": 18390
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 3.6863313990973565e-06,
+      "loss": 0.0126,
+      "step": 18400
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 3.6702127659574467e-06,
+      "loss": 0.0069,
+      "step": 18410
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 3.6540941328175373e-06,
+      "loss": 0.0113,
+      "step": 18420
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 3.6379754996776275e-06,
+      "loss": 0.0168,
+      "step": 18430
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 3.6218568665377177e-06,
+      "loss": 0.0194,
+      "step": 18440
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 3.605738233397808e-06,
+      "loss": 0.0277,
+      "step": 18450
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 3.589619600257898e-06,
+      "loss": 0.006,
+      "step": 18460
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 3.5735009671179887e-06,
+      "loss": 0.0229,
+      "step": 18470
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 3.5573823339780785e-06,
+      "loss": 0.0112,
+      "step": 18480
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 3.541263700838169e-06,
+      "loss": 0.015,
+      "step": 18490
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 3.5251450676982593e-06,
+      "loss": 0.0114,
+      "step": 18500
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 3.5090264345583495e-06,
+      "loss": 0.0135,
+      "step": 18510
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 3.4929078014184396e-06,
+      "loss": 0.0156,
+      "step": 18520
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 3.47678916827853e-06,
+      "loss": 0.0069,
+      "step": 18530
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 3.4606705351386204e-06,
+      "loss": 0.0095,
+      "step": 18540
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 3.4445519019987106e-06,
+      "loss": 0.0114,
+      "step": 18550
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 3.4284332688588013e-06,
+      "loss": 0.0147,
+      "step": 18560
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 3.412314635718891e-06,
+      "loss": 0.02,
+      "step": 18570
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 3.396196002578981e-06,
+      "loss": 0.0169,
+      "step": 18580
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 3.380077369439072e-06,
+      "loss": 0.008,
+      "step": 18590
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 3.363958736299162e-06,
+      "loss": 0.0215,
+      "step": 18600
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 3.347840103159252e-06,
+      "loss": 0.0133,
+      "step": 18610
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 3.3317214700193424e-06,
+      "loss": 0.0106,
+      "step": 18620
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 3.315602836879433e-06,
+      "loss": 0.0141,
+      "step": 18630
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 3.299484203739523e-06,
+      "loss": 0.0114,
+      "step": 18640
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 3.283365570599613e-06,
+      "loss": 0.0119,
+      "step": 18650
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 3.2672469374597036e-06,
+      "loss": 0.0229,
+      "step": 18660
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 3.2511283043197938e-06,
+      "loss": 0.0057,
+      "step": 18670
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 3.2350096711798844e-06,
+      "loss": 0.0216,
+      "step": 18680
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 3.218891038039974e-06,
+      "loss": 0.0119,
+      "step": 18690
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 3.2027724049000643e-06,
+      "loss": 0.0163,
+      "step": 18700
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 3.186653771760155e-06,
+      "loss": 0.0106,
+      "step": 18710
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 3.170535138620245e-06,
+      "loss": 0.0137,
+      "step": 18720
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 3.1544165054803353e-06,
+      "loss": 0.013,
+      "step": 18730
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 3.1382978723404255e-06,
+      "loss": 0.0179,
+      "step": 18740
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 3.122179239200516e-06,
+      "loss": 0.0177,
+      "step": 18750
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 3.1060606060606063e-06,
+      "loss": 0.0085,
+      "step": 18760
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 3.089941972920696e-06,
+      "loss": 0.0207,
+      "step": 18770
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 3.0738233397807867e-06,
+      "loss": 0.0217,
+      "step": 18780
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 3.057704706640877e-06,
+      "loss": 0.0064,
+      "step": 18790
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 3.0415860735009675e-06,
+      "loss": 0.0049,
+      "step": 18800
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 3.0254674403610572e-06,
+      "loss": 0.0194,
+      "step": 18810
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 3.0093488072211474e-06,
+      "loss": 0.0087,
+      "step": 18820
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 2.993230174081238e-06,
+      "loss": 0.0114,
+      "step": 18830
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 2.9771115409413282e-06,
+      "loss": 0.0066,
+      "step": 18840
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 2.9609929078014184e-06,
+      "loss": 0.0155,
+      "step": 18850
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 2.9448742746615086e-06,
+      "loss": 0.0079,
+      "step": 18860
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 2.9287556415215992e-06,
+      "loss": 0.0056,
+      "step": 18870
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 2.9126370083816894e-06,
+      "loss": 0.0236,
+      "step": 18880
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 2.8965183752417796e-06,
+      "loss": 0.0166,
+      "step": 18890
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 2.88039974210187e-06,
+      "loss": 0.0063,
+      "step": 18900
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 2.86428110896196e-06,
+      "loss": 0.0144,
+      "step": 18910
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 2.8481624758220506e-06,
+      "loss": 0.0025,
+      "step": 18920
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 2.8320438426821408e-06,
+      "loss": 0.0105,
+      "step": 18930
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 2.815925209542231e-06,
+      "loss": 0.0038,
+      "step": 18940
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 2.799806576402321e-06,
+      "loss": 0.011,
+      "step": 18950
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 2.7836879432624114e-06,
+      "loss": 0.0069,
+      "step": 18960
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 2.767569310122502e-06,
+      "loss": 0.0035,
+      "step": 18970
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 2.7514506769825917e-06,
+      "loss": 0.0045,
+      "step": 18980
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 2.7353320438426823e-06,
+      "loss": 0.0089,
+      "step": 18990
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 2.7192134107027725e-06,
+      "loss": 0.0099,
+      "step": 19000
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 2.7030947775628627e-06,
+      "loss": 0.0154,
+      "step": 19010
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 2.686976144422953e-06,
+      "loss": 0.0165,
+      "step": 19020
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 2.670857511283043e-06,
+      "loss": 0.008,
+      "step": 19030
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 2.6547388781431337e-06,
+      "loss": 0.009,
+      "step": 19040
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 2.638620245003224e-06,
+      "loss": 0.0144,
+      "step": 19050
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 2.622501611863314e-06,
+      "loss": 0.0206,
+      "step": 19060
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 2.6063829787234043e-06,
+      "loss": 0.0107,
+      "step": 19070
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 2.5902643455834945e-06,
+      "loss": 0.0059,
+      "step": 19080
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 2.574145712443585e-06,
+      "loss": 0.0269,
+      "step": 19090
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 2.558027079303675e-06,
+      "loss": 0.0096,
+      "step": 19100
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 2.5419084461637655e-06,
+      "loss": 0.0069,
+      "step": 19110
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 2.5257898130238556e-06,
+      "loss": 0.011,
+      "step": 19120
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 2.509671179883946e-06,
+      "loss": 0.0066,
+      "step": 19130
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 2.493552546744036e-06,
+      "loss": 0.0087,
+      "step": 19140
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 2.4774339136041262e-06,
+      "loss": 0.0129,
+      "step": 19150
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 2.461315280464217e-06,
+      "loss": 0.0101,
+      "step": 19160
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 2.445196647324307e-06,
+      "loss": 0.0062,
+      "step": 19170
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 2.429078014184397e-06,
+      "loss": 0.0115,
+      "step": 19180
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 2.4129593810444874e-06,
+      "loss": 0.0049,
+      "step": 19190
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 2.3968407479045776e-06,
+      "loss": 0.0093,
+      "step": 19200
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 2.380722114764668e-06,
+      "loss": 0.0056,
+      "step": 19210
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 2.364603481624758e-06,
+      "loss": 0.0077,
+      "step": 19220
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 2.3484848484848486e-06,
+      "loss": 0.0107,
+      "step": 19230
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 2.3323662153449388e-06,
+      "loss": 0.0137,
+      "step": 19240
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 2.3162475822050294e-06,
+      "loss": 0.0078,
+      "step": 19250
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 2.3001289490651196e-06,
+      "loss": 0.0085,
+      "step": 19260
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 2.2840103159252093e-06,
+      "loss": 0.01,
+      "step": 19270
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 2.2678916827853e-06,
+      "loss": 0.0282,
+      "step": 19280
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 2.25177304964539e-06,
+      "loss": 0.0107,
+      "step": 19290
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 2.2356544165054807e-06,
+      "loss": 0.0153,
+      "step": 19300
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 2.2195357833655705e-06,
+      "loss": 0.0063,
+      "step": 19310
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 2.2034171502256607e-06,
+      "loss": 0.0121,
+      "step": 19320
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 2.1872985170857513e-06,
+      "loss": 0.0261,
+      "step": 19330
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 2.1711798839458415e-06,
+      "loss": 0.016,
+      "step": 19340
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 2.1550612508059317e-06,
+      "loss": 0.0081,
+      "step": 19350
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 2.138942617666022e-06,
+      "loss": 0.0112,
+      "step": 19360
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 2.1228239845261125e-06,
+      "loss": 0.0053,
+      "step": 19370
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 2.1067053513862027e-06,
+      "loss": 0.0095,
+      "step": 19380
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 2.0905867182462924e-06,
+      "loss": 0.0137,
+      "step": 19390
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 2.074468085106383e-06,
+      "loss": 0.0137,
+      "step": 19400
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 2.0583494519664732e-06,
+      "loss": 0.0103,
+      "step": 19410
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 2.042230818826564e-06,
+      "loss": 0.0105,
+      "step": 19420
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 2.0261121856866536e-06,
+      "loss": 0.0092,
+      "step": 19430
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 2.009993552546744e-06,
+      "loss": 0.0059,
+      "step": 19440
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 1.9938749194068344e-06,
+      "loss": 0.0082,
+      "step": 19450
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 1.9777562862669246e-06,
+      "loss": 0.0208,
+      "step": 19460
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 1.961637653127015e-06,
+      "loss": 0.0025,
+      "step": 19470
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 1.945519019987105e-06,
+      "loss": 0.0088,
+      "step": 19480
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 1.9294003868471956e-06,
+      "loss": 0.01,
+      "step": 19490
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 1.913281753707286e-06,
+      "loss": 0.0073,
+      "step": 19500
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 1.8971631205673758e-06,
+      "loss": 0.0124,
+      "step": 19510
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 1.8810444874274662e-06,
+      "loss": 0.0047,
+      "step": 19520
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 1.8649258542875566e-06,
+      "loss": 0.0078,
+      "step": 19530
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 1.8488072211476466e-06,
+      "loss": 0.0068,
+      "step": 19540
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 1.832688588007737e-06,
+      "loss": 0.0109,
+      "step": 19550
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 1.8165699548678271e-06,
+      "loss": 0.0091,
+      "step": 19560
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 1.8004513217279175e-06,
+      "loss": 0.0169,
+      "step": 19570
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 1.784332688588008e-06,
+      "loss": 0.0174,
+      "step": 19580
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 1.7682140554480981e-06,
+      "loss": 0.0102,
+      "step": 19590
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 1.7520954223081883e-06,
+      "loss": 0.0061,
+      "step": 19600
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 1.7359767891682785e-06,
+      "loss": 0.01,
+      "step": 19610
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 1.719858156028369e-06,
+      "loss": 0.0124,
+      "step": 19620
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 1.703739522888459e-06,
+      "loss": 0.0157,
+      "step": 19630
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 1.6876208897485495e-06,
+      "loss": 0.0061,
+      "step": 19640
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 1.6715022566086397e-06,
+      "loss": 0.0062,
+      "step": 19650
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 1.6553836234687299e-06,
+      "loss": 0.0089,
+      "step": 19660
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 1.63926499032882e-06,
+      "loss": 0.0319,
+      "step": 19670
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 1.6231463571889105e-06,
+      "loss": 0.0115,
+      "step": 19680
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 1.6070277240490007e-06,
+      "loss": 0.0102,
+      "step": 19690
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 1.590909090909091e-06,
+      "loss": 0.011,
+      "step": 19700
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 1.5747904577691812e-06,
+      "loss": 0.0076,
+      "step": 19710
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 1.5586718246292714e-06,
+      "loss": 0.0079,
+      "step": 19720
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 1.5425531914893616e-06,
+      "loss": 0.0247,
+      "step": 19730
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 1.526434558349452e-06,
+      "loss": 0.0126,
+      "step": 19740
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 1.5103159252095422e-06,
+      "loss": 0.017,
+      "step": 19750
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 1.4941972920696326e-06,
+      "loss": 0.0031,
+      "step": 19760
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 1.4780786589297228e-06,
+      "loss": 0.0087,
+      "step": 19770
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 1.4619600257898132e-06,
+      "loss": 0.0259,
+      "step": 19780
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 1.4458413926499032e-06,
+      "loss": 0.0082,
+      "step": 19790
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 1.4297227595099936e-06,
+      "loss": 0.0196,
+      "step": 19800
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 1.4136041263700838e-06,
+      "loss": 0.0076,
+      "step": 19810
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 1.3974854932301742e-06,
+      "loss": 0.0122,
+      "step": 19820
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 1.3813668600902644e-06,
+      "loss": 0.0103,
+      "step": 19830
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 1.3652482269503548e-06,
+      "loss": 0.011,
+      "step": 19840
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 1.3491295938104447e-06,
+      "loss": 0.0087,
+      "step": 19850
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 1.3330109606705351e-06,
+      "loss": 0.0087,
+      "step": 19860
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 1.3168923275306253e-06,
+      "loss": 0.0117,
+      "step": 19870
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 1.3007736943907157e-06,
+      "loss": 0.004,
+      "step": 19880
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 1.284655061250806e-06,
+      "loss": 0.0069,
+      "step": 19890
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 1.2685364281108963e-06,
+      "loss": 0.0073,
+      "step": 19900
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 1.2524177949709865e-06,
+      "loss": 0.0067,
+      "step": 19910
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 1.2362991618310767e-06,
+      "loss": 0.0049,
+      "step": 19920
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 1.220180528691167e-06,
+      "loss": 0.0184,
+      "step": 19930
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 1.2040618955512573e-06,
+      "loss": 0.0133,
+      "step": 19940
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 1.1879432624113477e-06,
+      "loss": 0.0112,
+      "step": 19950
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 1.1718246292714379e-06,
+      "loss": 0.0121,
+      "step": 19960
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 1.155705996131528e-06,
+      "loss": 0.0081,
+      "step": 19970
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 1.1395873629916183e-06,
+      "loss": 0.0054,
+      "step": 19980
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 1.1234687298517087e-06,
+      "loss": 0.0141,
+      "step": 19990
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 1.1073500967117988e-06,
+      "loss": 0.0127,
+      "step": 20000
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 1.0912314635718892e-06,
+      "loss": 0.014,
+      "step": 20010
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 1.0751128304319794e-06,
+      "loss": 0.0255,
+      "step": 20020
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 1.0589941972920696e-06,
+      "loss": 0.0131,
+      "step": 20030
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 1.0428755641521598e-06,
+      "loss": 0.0128,
+      "step": 20040
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 1.0267569310122502e-06,
+      "loss": 0.0042,
+      "step": 20050
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 1.0106382978723404e-06,
+      "loss": 0.0127,
+      "step": 20060
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 9.945196647324308e-07,
+      "loss": 0.0083,
+      "step": 20070
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 9.78401031592521e-07,
+      "loss": 0.0095,
+      "step": 20080
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 9.622823984526114e-07,
+      "loss": 0.0064,
+      "step": 20090
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 9.461637653127015e-07,
+      "loss": 0.0139,
+      "step": 20100
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 9.300451321727918e-07,
+      "loss": 0.0078,
+      "step": 20110
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 9.139264990328821e-07,
+      "loss": 0.0083,
+      "step": 20120
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 8.978078658929724e-07,
+      "loss": 0.0097,
+      "step": 20130
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 8.816892327530626e-07,
+      "loss": 0.0083,
+      "step": 20140
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 8.655705996131528e-07,
+      "loss": 0.0105,
+      "step": 20150
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 8.494519664732431e-07,
+      "loss": 0.0136,
+      "step": 20160
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 0.0084,
+      "step": 20170
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 8.172147001934236e-07,
+      "loss": 0.0132,
+      "step": 20180
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 8.010960670535139e-07,
+      "loss": 0.0149,
+      "step": 20190
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 7.849774339136041e-07,
+      "loss": 0.0155,
+      "step": 20200
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 7.688588007736944e-07,
+      "loss": 0.0048,
+      "step": 20210
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 7.527401676337847e-07,
+      "loss": 0.019,
+      "step": 20220
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 7.36621534493875e-07,
+      "loss": 0.0105,
+      "step": 20230
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 7.205029013539652e-07,
+      "loss": 0.0153,
+      "step": 20240
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 7.043842682140555e-07,
+      "loss": 0.0127,
+      "step": 20250
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 6.882656350741458e-07,
+      "loss": 0.0199,
+      "step": 20260
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 6.72147001934236e-07,
+      "loss": 0.0097,
+      "step": 20270
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 6.560283687943263e-07,
+      "loss": 0.0024,
+      "step": 20280
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 6.399097356544166e-07,
+      "loss": 0.009,
+      "step": 20290
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 6.237911025145067e-07,
+      "loss": 0.0048,
+      "step": 20300
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 6.07672469374597e-07,
+      "loss": 0.0126,
+      "step": 20310
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 5.915538362346873e-07,
+      "loss": 0.0076,
+      "step": 20320
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 5.754352030947775e-07,
+      "loss": 0.0137,
+      "step": 20330
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 5.593165699548678e-07,
+      "loss": 0.0164,
+      "step": 20340
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 5.431979368149581e-07,
+      "loss": 0.0035,
+      "step": 20350
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 5.270793036750483e-07,
+      "loss": 0.0149,
+      "step": 20360
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 5.109606705351386e-07,
+      "loss": 0.0094,
+      "step": 20370
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 4.948420373952289e-07,
+      "loss": 0.0071,
+      "step": 20380
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 4.787234042553191e-07,
+      "loss": 0.0092,
+      "step": 20390
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 4.626047711154094e-07,
+      "loss": 0.0085,
+      "step": 20400
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 4.4648613797549967e-07,
+      "loss": 0.0176,
+      "step": 20410
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 4.3036750483558996e-07,
+      "loss": 0.0113,
+      "step": 20420
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 4.142488716956802e-07,
+      "loss": 0.0043,
+      "step": 20430
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 3.981302385557705e-07,
+      "loss": 0.0051,
+      "step": 20440
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 3.8201160541586074e-07,
+      "loss": 0.0231,
+      "step": 20450
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 3.65892972275951e-07,
+      "loss": 0.009,
+      "step": 20460
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 3.497743391360413e-07,
+      "loss": 0.0057,
+      "step": 20470
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 3.336557059961315e-07,
+      "loss": 0.0119,
+      "step": 20480
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 3.1753707285622176e-07,
+      "loss": 0.0082,
+      "step": 20490
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 3.0141843971631206e-07,
+      "loss": 0.0093,
+      "step": 20500
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 2.8529980657640235e-07,
+      "loss": 0.0073,
+      "step": 20510
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 2.691811734364926e-07,
+      "loss": 0.011,
+      "step": 20520
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 2.530625402965829e-07,
+      "loss": 0.0089,
+      "step": 20530
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 2.369439071566731e-07,
+      "loss": 0.0099,
+      "step": 20540
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 2.208252740167634e-07,
+      "loss": 0.0058,
+      "step": 20550
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 2.0470664087685367e-07,
+      "loss": 0.0165,
+      "step": 20560
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 1.885880077369439e-07,
+      "loss": 0.0132,
+      "step": 20570
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 1.7246937459703418e-07,
+      "loss": 0.0094,
+      "step": 20580
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 1.5635074145712445e-07,
+      "loss": 0.0094,
+      "step": 20590
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 1.402321083172147e-07,
+      "loss": 0.0104,
+      "step": 20600
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 1.2411347517730496e-07,
+      "loss": 0.0061,
+      "step": 20610
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 1.0799484203739524e-07,
+      "loss": 0.0056,
+      "step": 20620
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 9.18762088974855e-08,
+      "loss": 0.006,
+      "step": 20630
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 7.575757575757576e-08,
+      "loss": 0.0083,
+      "step": 20640
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 5.963894261766603e-08,
+      "loss": 0.0089,
+      "step": 20650
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 4.352030947775629e-08,
+      "loss": 0.0092,
+      "step": 20660
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 2.740167633784655e-08,
+      "loss": 0.0145,
+      "step": 20670
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 1.1283043197936816e-08,
+      "loss": 0.0103,
+      "step": 20680
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.9988272087568413,
+      "eval_loss": 0.005595107097178698,
+      "eval_runtime": 68.5663,
+      "eval_samples_per_second": 858.06,
+      "eval_steps_per_second": 26.821,
+      "step": 20680
+    },
+    {
+      "epoch": 5.0,
+      "step": 20680,
+      "total_flos": 4.806704310439244e+19,
+      "train_loss": 0.17228164457584547,
+      "train_runtime": 8820.4548,
+      "train_samples_per_second": 300.156,
+      "train_steps_per_second": 2.345
+    }
+  ],
+  "max_steps": 20680,
+  "num_train_epochs": 5,
+  "total_flos": 4.806704310439244e+19,
+  "trial_name": null,
+  "trial_params": null
+}