diff --git "a/output/version_2/checkpoint-10732/trainer_state.json" "b/output/version_2/checkpoint-10732/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/output/version_2/checkpoint-10732/trainer_state.json"
@@ -0,0 +1,6499 @@
+{
+  "best_metric": 0.9948979591836735,
+  "best_model_checkpoint": "./output/version_2\\checkpoint-10732",
+  "epoch": 4.0,
+  "eval_steps": 500,
+  "global_step": 10732,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.6892,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.6872,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3e-06,
+      "loss": 0.6825,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.6758,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-06,
+      "loss": 0.6768,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6e-06,
+      "loss": 0.6441,
+      "step": 60
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 0.6105,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.5332,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9e-06,
+      "loss": 0.4755,
+      "step": 90
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1e-05,
+      "loss": 0.3745,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.3879,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.2e-05,
+      "loss": 0.2498,
+      "step": 120
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.2274,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 0.2154,
+      "step": 140
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.5e-05,
+      "loss": 0.2528,
+      "step": 150
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.3022,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 0.1149,
+      "step": 170
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.8e-05,
+      "loss": 0.1523,
+      "step": 180
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9e-05,
+      "loss": 0.2482,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 0.1524,
+      "step": 200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.1e-05,
+      "loss": 0.185,
+      "step": 210
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 0.1886,
+      "step": 220
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 0.0569,
+      "step": 230
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.4e-05,
+      "loss": 0.165,
+      "step": 240
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.5e-05,
+      "loss": 0.1309,
+      "step": 250
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 0.1491,
+      "step": 260
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 0.1351,
+      "step": 270
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 0.1292,
+      "step": 280
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9e-05,
+      "loss": 0.1237,
+      "step": 290
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 3e-05,
+      "loss": 0.0656,
+      "step": 300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 3.1e-05,
+      "loss": 0.137,
+      "step": 310
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.1777,
+      "step": 320
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 3.3e-05,
+      "loss": 0.1426,
+      "step": 330
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 0.1596,
+      "step": 340
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 3.5e-05,
+      "loss": 0.1664,
+      "step": 350
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 3.6e-05,
+      "loss": 0.0927,
+      "step": 360
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 3.7e-05,
+      "loss": 0.0272,
+      "step": 370
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 3.8e-05,
+      "loss": 0.0649,
+      "step": 380
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 0.121,
+      "step": 390
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4e-05,
+      "loss": 0.2762,
+      "step": 400
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.1e-05,
+      "loss": 0.1732,
+      "step": 410
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2e-05,
+      "loss": 0.0866,
+      "step": 420
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.3e-05,
+      "loss": 0.0099,
+      "step": 430
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.066,
+      "step": 440
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.5e-05,
+      "loss": 0.0862,
+      "step": 450
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.0859,
+      "step": 460
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7e-05,
+      "loss": 0.0117,
+      "step": 470
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.8e-05,
+      "loss": 0.1539,
+      "step": 480
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9e-05,
+      "loss": 0.1276,
+      "step": 490
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.0697,
+      "step": 500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.996128532713899e-05,
+      "loss": 0.1297,
+      "step": 510
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9922570654277976e-05,
+      "loss": 0.0338,
+      "step": 520
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.988385598141696e-05,
+      "loss": 0.1039,
+      "step": 530
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.984514130855594e-05,
+      "loss": 0.2102,
+      "step": 540
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.980642663569493e-05,
+      "loss": 0.0714,
+      "step": 550
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.9767711962833916e-05,
+      "loss": 0.1337,
+      "step": 560
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.97289972899729e-05,
+      "loss": 0.0891,
+      "step": 570
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.969028261711189e-05,
+      "loss": 0.1179,
+      "step": 580
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.965156794425087e-05,
+      "loss": 0.0721,
+      "step": 590
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.9612853271389856e-05,
+      "loss": 0.1309,
+      "step": 600
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.957413859852884e-05,
+      "loss": 0.1412,
+      "step": 610
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.953542392566783e-05,
+      "loss": 0.1124,
+      "step": 620
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.949670925280682e-05,
+      "loss": 0.1115,
+      "step": 630
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.9457994579945803e-05,
+      "loss": 0.1481,
+      "step": 640
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.9419279907084783e-05,
+      "loss": 0.1063,
+      "step": 650
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.938056523422378e-05,
+      "loss": 0.0524,
+      "step": 660
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.934185056136276e-05,
+      "loss": 0.0351,
+      "step": 670
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.9303135888501744e-05,
+      "loss": 0.1137,
+      "step": 680
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.926442121564073e-05,
+      "loss": 0.1061,
+      "step": 690
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.922570654277972e-05,
+      "loss": 0.0612,
+      "step": 700
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.9186991869918704e-05,
+      "loss": 0.0897,
+      "step": 710
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.9148277197057684e-05,
+      "loss": 0.1368,
+      "step": 720
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.910956252419667e-05,
+      "loss": 0.0617,
+      "step": 730
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.907084785133566e-05,
+      "loss": 0.1055,
+      "step": 740
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.9032133178474644e-05,
+      "loss": 0.1092,
+      "step": 750
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.899341850561363e-05,
+      "loss": 0.0669,
+      "step": 760
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.895470383275261e-05,
+      "loss": 0.0394,
+      "step": 770
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.89159891598916e-05,
+      "loss": 0.0063,
+      "step": 780
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.887727448703059e-05,
+      "loss": 0.117,
+      "step": 790
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.883855981416957e-05,
+      "loss": 0.0789,
+      "step": 800
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.879984514130856e-05,
+      "loss": 0.044,
+      "step": 810
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.876113046844754e-05,
+      "loss": 0.1397,
+      "step": 820
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.872241579558653e-05,
+      "loss": 0.27,
+      "step": 830
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.868370112272552e-05,
+      "loss": 0.1251,
+      "step": 840
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.86449864498645e-05,
+      "loss": 0.0165,
+      "step": 850
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.8606271777003485e-05,
+      "loss": 0.0631,
+      "step": 860
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.856755710414247e-05,
+      "loss": 0.0242,
+      "step": 870
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.852884243128146e-05,
+      "loss": 0.091,
+      "step": 880
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.8490127758420445e-05,
+      "loss": 0.0989,
+      "step": 890
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.8451413085559425e-05,
+      "loss": 0.0949,
+      "step": 900
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.841269841269841e-05,
+      "loss": 0.0405,
+      "step": 910
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.8373983739837406e-05,
+      "loss": 0.0543,
+      "step": 920
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.8335269066976386e-05,
+      "loss": 0.0431,
+      "step": 930
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.829655439411537e-05,
+      "loss": 0.0011,
+      "step": 940
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.825783972125435e-05,
+      "loss": 0.1565,
+      "step": 950
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.8219125048393346e-05,
+      "loss": 0.0831,
+      "step": 960
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.818041037553233e-05,
+      "loss": 0.119,
+      "step": 970
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.814169570267131e-05,
+      "loss": 0.0402,
+      "step": 980
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.81029810298103e-05,
+      "loss": 0.0609,
+      "step": 990
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.8064266356949286e-05,
+      "loss": 0.0652,
+      "step": 1000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.802555168408827e-05,
+      "loss": 0.0841,
+      "step": 1010
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.798683701122726e-05,
+      "loss": 0.0596,
+      "step": 1020
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.794812233836624e-05,
+      "loss": 0.0839,
+      "step": 1030
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.7909407665505226e-05,
+      "loss": 0.057,
+      "step": 1040
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.787069299264422e-05,
+      "loss": 0.0192,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.78319783197832e-05,
+      "loss": 0.053,
+      "step": 1060
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.779326364692219e-05,
+      "loss": 0.0052,
+      "step": 1070
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.775454897406117e-05,
+      "loss": 0.0156,
+      "step": 1080
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.771583430120016e-05,
+      "loss": 0.0571,
+      "step": 1090
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.767711962833915e-05,
+      "loss": 0.0672,
+      "step": 1100
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.763840495547813e-05,
+      "loss": 0.1001,
+      "step": 1110
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.7599690282617114e-05,
+      "loss": 0.0415,
+      "step": 1120
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.75609756097561e-05,
+      "loss": 0.0885,
+      "step": 1130
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.752226093689509e-05,
+      "loss": 0.0054,
+      "step": 1140
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.7483546264034074e-05,
+      "loss": 0.0467,
+      "step": 1150
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.7444831591173054e-05,
+      "loss": 0.0324,
+      "step": 1160
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.740611691831204e-05,
+      "loss": 0.0433,
+      "step": 1170
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.736740224545103e-05,
+      "loss": 0.0874,
+      "step": 1180
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.7328687572590014e-05,
+      "loss": 0.0746,
+      "step": 1190
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.7289972899729e-05,
+      "loss": 0.057,
+      "step": 1200
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.725125822686798e-05,
+      "loss": 0.1251,
+      "step": 1210
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.7212543554006975e-05,
+      "loss": 0.0396,
+      "step": 1220
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.7173828881145955e-05,
+      "loss": 0.1524,
+      "step": 1230
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.713511420828494e-05,
+      "loss": 0.0269,
+      "step": 1240
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.709639953542393e-05,
+      "loss": 0.0882,
+      "step": 1250
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.7057684862562915e-05,
+      "loss": 0.051,
+      "step": 1260
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.70189701897019e-05,
+      "loss": 0.1079,
+      "step": 1270
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.698025551684089e-05,
+      "loss": 0.0952,
+      "step": 1280
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.694154084397987e-05,
+      "loss": 0.0759,
+      "step": 1290
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.6902826171118855e-05,
+      "loss": 0.099,
+      "step": 1300
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.686411149825784e-05,
+      "loss": 0.062,
+      "step": 1310
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.682539682539683e-05,
+      "loss": 0.0976,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.6786682152535815e-05,
+      "loss": 0.1164,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.6747967479674795e-05,
+      "loss": 0.0889,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.670925280681379e-05,
+      "loss": 0.0487,
+      "step": 1350
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.667053813395277e-05,
+      "loss": 0.0434,
+      "step": 1360
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.6631823461091756e-05,
+      "loss": 0.0033,
+      "step": 1370
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.659310878823074e-05,
+      "loss": 0.0833,
+      "step": 1380
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.655439411536973e-05,
+      "loss": 0.0385,
+      "step": 1390
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.6515679442508716e-05,
+      "loss": 0.0027,
+      "step": 1400
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.6476964769647696e-05,
+      "loss": 0.1787,
+      "step": 1410
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.643825009678668e-05,
+      "loss": 0.175,
+      "step": 1420
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.639953542392567e-05,
+      "loss": 0.0056,
+      "step": 1430
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.6360820751064656e-05,
+      "loss": 0.0317,
+      "step": 1440
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.632210607820364e-05,
+      "loss": 0.1109,
+      "step": 1450
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.628339140534262e-05,
+      "loss": 0.11,
+      "step": 1460
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.624467673248161e-05,
+      "loss": 0.0042,
+      "step": 1470
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.62059620596206e-05,
+      "loss": 0.0208,
+      "step": 1480
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.616724738675958e-05,
+      "loss": 0.1056,
+      "step": 1490
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.612853271389857e-05,
+      "loss": 0.0726,
+      "step": 1500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.608981804103756e-05,
+      "loss": 0.0262,
+      "step": 1510
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.6051103368176543e-05,
+      "loss": 0.0712,
+      "step": 1520
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.601238869531553e-05,
+      "loss": 0.0285,
+      "step": 1530
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.597367402245451e-05,
+      "loss": 0.0181,
+      "step": 1540
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.59349593495935e-05,
+      "loss": 0.0444,
+      "step": 1550
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.5896244676732484e-05,
+      "loss": 0.006,
+      "step": 1560
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.585753000387147e-05,
+      "loss": 0.0315,
+      "step": 1570
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.581881533101046e-05,
+      "loss": 0.0827,
+      "step": 1580
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.578010065814944e-05,
+      "loss": 0.0655,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.5741385985288424e-05,
+      "loss": 0.0392,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.570267131242742e-05,
+      "loss": 0.027,
+      "step": 1610
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.56639566395664e-05,
+      "loss": 0.0674,
+      "step": 1620
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.5625241966705384e-05,
+      "loss": 0.0039,
+      "step": 1630
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.5586527293844364e-05,
+      "loss": 0.002,
+      "step": 1640
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.554781262098336e-05,
+      "loss": 0.1919,
+      "step": 1650
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.5509097948122345e-05,
+      "loss": 0.0571,
+      "step": 1660
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.5470383275261325e-05,
+      "loss": 0.0593,
+      "step": 1670
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.543166860240031e-05,
+      "loss": 0.0156,
+      "step": 1680
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.53929539295393e-05,
+      "loss": 0.0675,
+      "step": 1690
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.5354239256678285e-05,
+      "loss": 0.0148,
+      "step": 1700
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.531552458381727e-05,
+      "loss": 0.1182,
+      "step": 1710
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.527680991095625e-05,
+      "loss": 0.1101,
+      "step": 1720
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.523809523809524e-05,
+      "loss": 0.033,
+      "step": 1730
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.5199380565234225e-05,
+      "loss": 0.0549,
+      "step": 1740
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.516066589237321e-05,
+      "loss": 0.0526,
+      "step": 1750
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.51219512195122e-05,
+      "loss": 0.0972,
+      "step": 1760
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.508323654665118e-05,
+      "loss": 0.0463,
+      "step": 1770
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.5044521873790165e-05,
+      "loss": 0.0691,
+      "step": 1780
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.500580720092916e-05,
+      "loss": 0.045,
+      "step": 1790
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.496709252806814e-05,
+      "loss": 0.0805,
+      "step": 1800
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.4928377855207126e-05,
+      "loss": 0.0254,
+      "step": 1810
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.4889663182346106e-05,
+      "loss": 0.167,
+      "step": 1820
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.48509485094851e-05,
+      "loss": 0.0393,
+      "step": 1830
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.4812233836624086e-05,
+      "loss": 0.055,
+      "step": 1840
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.4773519163763066e-05,
+      "loss": 0.0102,
+      "step": 1850
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.473480449090205e-05,
+      "loss": 0.0888,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.469608981804104e-05,
+      "loss": 0.0065,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.4657375145180026e-05,
+      "loss": 0.1025,
+      "step": 1880
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.461866047231901e-05,
+      "loss": 0.0398,
+      "step": 1890
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.457994579945799e-05,
+      "loss": 0.0399,
+      "step": 1900
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.454123112659698e-05,
+      "loss": 0.1081,
+      "step": 1910
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.450251645373597e-05,
+      "loss": 0.0724,
+      "step": 1920
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.446380178087495e-05,
+      "loss": 0.11,
+      "step": 1930
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.442508710801394e-05,
+      "loss": 0.0095,
+      "step": 1940
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.438637243515292e-05,
+      "loss": 0.0271,
+      "step": 1950
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.4347657762291913e-05,
+      "loss": 0.0947,
+      "step": 1960
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.43089430894309e-05,
+      "loss": 0.0164,
+      "step": 1970
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.427022841656988e-05,
+      "loss": 0.0321,
+      "step": 1980
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.423151374370887e-05,
+      "loss": 0.0843,
+      "step": 1990
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.4192799070847854e-05,
+      "loss": 0.0719,
+      "step": 2000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.415408439798684e-05,
+      "loss": 0.0489,
+      "step": 2010
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.411536972512583e-05,
+      "loss": 0.0346,
+      "step": 2020
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.407665505226481e-05,
+      "loss": 0.0419,
+      "step": 2030
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.4037940379403794e-05,
+      "loss": 0.0439,
+      "step": 2040
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.399922570654278e-05,
+      "loss": 0.1516,
+      "step": 2050
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.396051103368177e-05,
+      "loss": 0.0296,
+      "step": 2060
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.3921796360820754e-05,
+      "loss": 0.0594,
+      "step": 2070
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.3883081687959734e-05,
+      "loss": 0.0581,
+      "step": 2080
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.384436701509873e-05,
+      "loss": 0.0079,
+      "step": 2090
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.380565234223771e-05,
+      "loss": 0.0026,
+      "step": 2100
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.3766937669376695e-05,
+      "loss": 0.011,
+      "step": 2110
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.372822299651568e-05,
+      "loss": 0.0042,
+      "step": 2120
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.368950832365467e-05,
+      "loss": 0.0964,
+      "step": 2130
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.3650793650793655e-05,
+      "loss": 0.0765,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.361207897793264e-05,
+      "loss": 0.0457,
+      "step": 2150
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.357336430507162e-05,
+      "loss": 0.038,
+      "step": 2160
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.353464963221061e-05,
+      "loss": 0.0879,
+      "step": 2170
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.3495934959349595e-05,
+      "loss": 0.047,
+      "step": 2180
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.345722028648858e-05,
+      "loss": 0.1213,
+      "step": 2190
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.341850561362757e-05,
+      "loss": 0.0239,
+      "step": 2200
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.337979094076655e-05,
+      "loss": 0.0415,
+      "step": 2210
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.334107626790554e-05,
+      "loss": 0.0876,
+      "step": 2220
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.330236159504452e-05,
+      "loss": 0.1069,
+      "step": 2230
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.326364692218351e-05,
+      "loss": 0.0087,
+      "step": 2240
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.3224932249322496e-05,
+      "loss": 0.0306,
+      "step": 2250
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.318621757646148e-05,
+      "loss": 0.027,
+      "step": 2260
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.314750290360047e-05,
+      "loss": 0.0433,
+      "step": 2270
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.310878823073945e-05,
+      "loss": 0.0713,
+      "step": 2280
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.3070073557878436e-05,
+      "loss": 0.0216,
+      "step": 2290
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.303135888501742e-05,
+      "loss": 0.0022,
+      "step": 2300
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.299264421215641e-05,
+      "loss": 0.0023,
+      "step": 2310
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.2953929539295396e-05,
+      "loss": 0.1148,
+      "step": 2320
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.291521486643438e-05,
+      "loss": 0.0826,
+      "step": 2330
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.287650019357336e-05,
+      "loss": 0.0375,
+      "step": 2340
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.2837785520712356e-05,
+      "loss": 0.0497,
+      "step": 2350
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.2799070847851336e-05,
+      "loss": 0.087,
+      "step": 2360
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.276035617499032e-05,
+      "loss": 0.0169,
+      "step": 2370
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.272164150212931e-05,
+      "loss": 0.0034,
+      "step": 2380
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.26829268292683e-05,
+      "loss": 0.0409,
+      "step": 2390
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.2644212156407283e-05,
+      "loss": 0.0587,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.2605497483546263e-05,
+      "loss": 0.0644,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.256678281068525e-05,
+      "loss": 0.0473,
+      "step": 2420
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.252806813782424e-05,
+      "loss": 0.016,
+      "step": 2430
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.2489353464963224e-05,
+      "loss": 0.0617,
+      "step": 2440
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.245063879210221e-05,
+      "loss": 0.0014,
+      "step": 2450
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.241192411924119e-05,
+      "loss": 0.0622,
+      "step": 2460
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.237320944638018e-05,
+      "loss": 0.0619,
+      "step": 2470
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.233449477351917e-05,
+      "loss": 0.0319,
+      "step": 2480
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.229578010065815e-05,
+      "loss": 0.0219,
+      "step": 2490
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.225706542779714e-05,
+      "loss": 0.0008,
+      "step": 2500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.221835075493612e-05,
+      "loss": 0.0488,
+      "step": 2510
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.217963608207511e-05,
+      "loss": 0.0586,
+      "step": 2520
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.21409214092141e-05,
+      "loss": 0.0867,
+      "step": 2530
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 4.210220673635308e-05,
+      "loss": 0.0748,
+      "step": 2540
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 4.2063492063492065e-05,
+      "loss": 0.0852,
+      "step": 2550
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 4.202477739063105e-05,
+      "loss": 0.094,
+      "step": 2560
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.198606271777004e-05,
+      "loss": 0.0212,
+      "step": 2570
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.1947348044909025e-05,
+      "loss": 0.0148,
+      "step": 2580
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.1908633372048005e-05,
+      "loss": 0.0943,
+      "step": 2590
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.186991869918699e-05,
+      "loss": 0.0711,
+      "step": 2600
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.1831204026325985e-05,
+      "loss": 0.006,
+      "step": 2610
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.1792489353464965e-05,
+      "loss": 0.0006,
+      "step": 2620
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.175377468060395e-05,
+      "loss": 0.0561,
+      "step": 2630
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.171506000774293e-05,
+      "loss": 0.0721,
+      "step": 2640
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.1676345334881925e-05,
+      "loss": 0.0756,
+      "step": 2650
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.163763066202091e-05,
+      "loss": 0.0792,
+      "step": 2660
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.159891598915989e-05,
+      "loss": 0.0516,
+      "step": 2670
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.156020131629888e-05,
+      "loss": 0.064,
+      "step": 2680
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.9914733016494269,
+      "eval_f1": 0.9911773213769163,
+      "eval_loss": 0.03051486425101757,
+      "eval_runtime": 142.497,
+      "eval_samples_per_second": 100.409,
+      "eval_steps_per_second": 1.572,
+      "step": 2683
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1521486643437866e-05,
+      "loss": 0.0885,
+      "step": 2690
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.148277197057685e-05,
+      "loss": 0.0126,
+      "step": 2700
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.144405729771584e-05,
+      "loss": 0.0009,
+      "step": 2710
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.140534262485482e-05,
+      "loss": 0.0004,
+      "step": 2720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1366627951993806e-05,
+      "loss": 0.0118,
+      "step": 2730
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.132791327913279e-05,
+      "loss": 0.028,
+      "step": 2740
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.128919860627178e-05,
+      "loss": 0.0003,
+      "step": 2750
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.1250483933410766e-05,
+      "loss": 0.0033,
+      "step": 2760
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.1211769260549746e-05,
+      "loss": 0.0019,
+      "step": 2770
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.117305458768874e-05,
+      "loss": 0.015,
+      "step": 2780
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.1134339914827726e-05,
+      "loss": 0.0415,
+      "step": 2790
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.1095625241966706e-05,
+      "loss": 0.0002,
+      "step": 2800
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.105691056910569e-05,
+      "loss": 0.0647,
+      "step": 2810
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.101819589624468e-05,
+      "loss": 0.0213,
+      "step": 2820
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.097948122338367e-05,
+      "loss": 0.0004,
+      "step": 2830
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.0940766550522653e-05,
+      "loss": 0.0006,
+      "step": 2840
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.0902051877661633e-05,
+      "loss": 0.0002,
+      "step": 2850
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.086333720480062e-05,
+      "loss": 0.0237,
+      "step": 2860
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.082462253193961e-05,
+      "loss": 0.0638,
+      "step": 2870
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.0785907859078594e-05,
+      "loss": 0.1456,
+      "step": 2880
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.074719318621758e-05,
+      "loss": 0.1325,
+      "step": 2890
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.070847851335656e-05,
+      "loss": 0.0526,
+      "step": 2900
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.066976384049555e-05,
+      "loss": 0.001,
+      "step": 2910
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.0631049167634534e-05,
+      "loss": 0.0006,
+      "step": 2920
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.059233449477352e-05,
+      "loss": 0.0283,
+      "step": 2930
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.055361982191251e-05,
+      "loss": 0.0388,
+      "step": 2940
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.051490514905149e-05,
+      "loss": 0.0323,
+      "step": 2950
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.047619047619048e-05,
+      "loss": 0.0394,
+      "step": 2960
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.043747580332947e-05,
+      "loss": 0.0055,
+      "step": 2970
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.039876113046845e-05,
+      "loss": 0.0387,
+      "step": 2980
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.0360046457607435e-05,
+      "loss": 0.0236,
+      "step": 2990
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.032133178474642e-05,
+      "loss": 0.0033,
+      "step": 3000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.028261711188541e-05,
+      "loss": 0.0566,
+      "step": 3010
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.0243902439024395e-05,
+      "loss": 0.0017,
+      "step": 3020
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.0205187766163375e-05,
+      "loss": 0.0262,
+      "step": 3030
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.016647309330236e-05,
+      "loss": 0.0552,
+      "step": 3040
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.012775842044135e-05,
+      "loss": 0.0006,
+      "step": 3050
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.0089043747580335e-05,
+      "loss": 0.0317,
+      "step": 3060
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.005032907471932e-05,
+      "loss": 0.0006,
+      "step": 3070
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.00116144018583e-05,
+      "loss": 0.0901,
+      "step": 3080
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.9972899728997295e-05,
+      "loss": 0.0105,
+      "step": 3090
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.9934185056136275e-05,
+      "loss": 0.0023,
+      "step": 3100
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.989547038327526e-05,
+      "loss": 0.0008,
+      "step": 3110
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.985675571041425e-05,
+      "loss": 0.0948,
+      "step": 3120
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.9818041037553236e-05,
+      "loss": 0.0007,
+      "step": 3130
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.977932636469222e-05,
+      "loss": 0.0372,
+      "step": 3140
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.97406116918312e-05,
+      "loss": 0.0064,
+      "step": 3150
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.970189701897019e-05,
+      "loss": 0.0338,
+      "step": 3160
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.9663182346109176e-05,
+      "loss": 0.0003,
+      "step": 3170
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.962446767324816e-05,
+      "loss": 0.0002,
+      "step": 3180
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.958575300038715e-05,
+      "loss": 0.0059,
+      "step": 3190
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.9547038327526136e-05,
+      "loss": 0.0399,
+      "step": 3200
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.9508323654665116e-05,
+      "loss": 0.0461,
+      "step": 3210
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.946960898180411e-05,
+      "loss": 0.0055,
+      "step": 3220
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.943089430894309e-05,
+      "loss": 0.0547,
+      "step": 3230
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.9392179636082076e-05,
+      "loss": 0.0007,
+      "step": 3240
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.935346496322106e-05,
+      "loss": 0.0003,
+      "step": 3250
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.931475029036005e-05,
+      "loss": 0.0763,
+      "step": 3260
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.927603561749904e-05,
+      "loss": 0.0651,
+      "step": 3270
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.923732094463802e-05,
+      "loss": 0.0007,
+      "step": 3280
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.9198606271777003e-05,
+      "loss": 0.0521,
+      "step": 3290
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.915989159891599e-05,
+      "loss": 0.0003,
+      "step": 3300
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.912117692605498e-05,
+      "loss": 0.0002,
+      "step": 3310
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.9082462253193964e-05,
+      "loss": 0.0726,
+      "step": 3320
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.9043747580332944e-05,
+      "loss": 0.0433,
+      "step": 3330
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.900503290747193e-05,
+      "loss": 0.0017,
+      "step": 3340
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.8966318234610924e-05,
+      "loss": 0.0016,
+      "step": 3350
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.8927603561749904e-05,
+      "loss": 0.1122,
+      "step": 3360
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 0.0038,
+      "step": 3370
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.885017421602787e-05,
+      "loss": 0.0255,
+      "step": 3380
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.8811459543166864e-05,
+      "loss": 0.0337,
+      "step": 3390
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.877274487030585e-05,
+      "loss": 0.001,
+      "step": 3400
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.873403019744483e-05,
+      "loss": 0.0104,
+      "step": 3410
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.869531552458382e-05,
+      "loss": 0.0152,
+      "step": 3420
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.8656600851722805e-05,
+      "loss": 0.0002,
+      "step": 3430
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.861788617886179e-05,
+      "loss": 0.0659,
+      "step": 3440
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.857917150600078e-05,
+      "loss": 0.0277,
+      "step": 3450
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.854045683313976e-05,
+      "loss": 0.0232,
+      "step": 3460
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.8501742160278745e-05,
+      "loss": 0.0169,
+      "step": 3470
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.846302748741774e-05,
+      "loss": 0.0016,
+      "step": 3480
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.842431281455672e-05,
+      "loss": 0.043,
+      "step": 3490
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.8385598141695705e-05,
+      "loss": 0.0002,
+      "step": 3500
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 3.8346883468834685e-05,
+      "loss": 0.0001,
+      "step": 3510
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 3.830816879597368e-05,
+      "loss": 0.0646,
+      "step": 3520
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.8269454123112665e-05,
+      "loss": 0.0252,
+      "step": 3530
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.8230739450251645e-05,
+      "loss": 0.0896,
+      "step": 3540
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.819202477739063e-05,
+      "loss": 0.0383,
+      "step": 3550
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 3.815331010452962e-05,
+      "loss": 0.0007,
+      "step": 3560
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 3.8114595431668606e-05,
+      "loss": 0.0008,
+      "step": 3570
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 3.807588075880759e-05,
+      "loss": 0.076,
+      "step": 3580
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 3.803716608594657e-05,
+      "loss": 0.029,
+      "step": 3590
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 3.799845141308556e-05,
+      "loss": 0.0072,
+      "step": 3600
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 3.795973674022455e-05,
+      "loss": 0.0595,
+      "step": 3610
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 3.792102206736353e-05,
+      "loss": 0.0117,
+      "step": 3620
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 3.788230739450252e-05,
+      "loss": 0.0628,
+      "step": 3630
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 3.78435927216415e-05,
+      "loss": 0.031,
+      "step": 3640
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 3.780487804878049e-05,
+      "loss": 0.0343,
+      "step": 3650
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 3.776616337591948e-05,
+      "loss": 0.0262,
+      "step": 3660
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 3.772744870305846e-05,
+      "loss": 0.0031,
+      "step": 3670
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 3.7688734030197446e-05,
+      "loss": 0.0154,
+      "step": 3680
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 3.765001935733643e-05,
+      "loss": 0.1325,
+      "step": 3690
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 3.761130468447542e-05,
+      "loss": 0.0016,
+      "step": 3700
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 3.757259001161441e-05,
+      "loss": 0.0008,
+      "step": 3710
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 3.753387533875339e-05,
+      "loss": 0.0004,
+      "step": 3720
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 3.7495160665892373e-05,
+      "loss": 0.0485,
+      "step": 3730
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 3.745644599303136e-05,
+      "loss": 0.0259,
+      "step": 3740
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 3.741773132017035e-05,
+      "loss": 0.0381,
+      "step": 3750
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 3.7379016647309334e-05,
+      "loss": 0.0811,
+      "step": 3760
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 3.7340301974448314e-05,
+      "loss": 0.0006,
+      "step": 3770
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 3.730158730158731e-05,
+      "loss": 0.0536,
+      "step": 3780
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 3.726287262872629e-05,
+      "loss": 0.0583,
+      "step": 3790
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 3.7224157955865274e-05,
+      "loss": 0.0075,
+      "step": 3800
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 3.718544328300426e-05,
+      "loss": 0.0447,
+      "step": 3810
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 3.714672861014325e-05,
+      "loss": 0.0099,
+      "step": 3820
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 3.7108013937282234e-05,
+      "loss": 0.0091,
+      "step": 3830
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 3.706929926442122e-05,
+      "loss": 0.0088,
+      "step": 3840
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 3.70305845915602e-05,
+      "loss": 0.057,
+      "step": 3850
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 3.699186991869919e-05,
+      "loss": 0.0202,
+      "step": 3860
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 3.6953155245838175e-05,
+      "loss": 0.0004,
+      "step": 3870
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 3.691444057297716e-05,
+      "loss": 0.0015,
+      "step": 3880
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 3.687572590011615e-05,
+      "loss": 0.0004,
+      "step": 3890
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 3.683701122725513e-05,
+      "loss": 0.1102,
+      "step": 3900
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 3.679829655439412e-05,
+      "loss": 0.0367,
+      "step": 3910
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 3.67595818815331e-05,
+      "loss": 0.0023,
+      "step": 3920
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 3.672086720867209e-05,
+      "loss": 0.0011,
+      "step": 3930
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 3.6682152535811075e-05,
+      "loss": 0.0181,
+      "step": 3940
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 3.664343786295006e-05,
+      "loss": 0.0062,
+      "step": 3950
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.660472319008905e-05,
+      "loss": 0.0306,
+      "step": 3960
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.656600851722803e-05,
+      "loss": 0.0853,
+      "step": 3970
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.6527293844367015e-05,
+      "loss": 0.0492,
+      "step": 3980
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.6488579171506e-05,
+      "loss": 0.0619,
+      "step": 3990
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.644986449864499e-05,
+      "loss": 0.0013,
+      "step": 4000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.6411149825783976e-05,
+      "loss": 0.001,
+      "step": 4010
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 3.6372435152922956e-05,
+      "loss": 0.0258,
+      "step": 4020
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 3.633372048006194e-05,
+      "loss": 0.0008,
+      "step": 4030
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 3.629500580720093e-05,
+      "loss": 0.0228,
+      "step": 4040
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 3.6256291134339916e-05,
+      "loss": 0.0001,
+      "step": 4050
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 3.62175764614789e-05,
+      "loss": 0.0001,
+      "step": 4060
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 3.617886178861789e-05,
+      "loss": 0.0291,
+      "step": 4070
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 3.614014711575687e-05,
+      "loss": 0.0395,
+      "step": 4080
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 3.610143244289586e-05,
+      "loss": 0.0021,
+      "step": 4090
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 3.606271777003484e-05,
+      "loss": 0.0825,
+      "step": 4100
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 3.602400309717383e-05,
+      "loss": 0.003,
+      "step": 4110
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 3.5985288424312816e-05,
+      "loss": 0.0275,
+      "step": 4120
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 3.59465737514518e-05,
+      "loss": 0.0198,
+      "step": 4130
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 3.590785907859079e-05,
+      "loss": 0.008,
+      "step": 4140
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 3.586914440572977e-05,
+      "loss": 0.001,
+      "step": 4150
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 3.583042973286876e-05,
+      "loss": 0.0005,
+      "step": 4160
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 3.5791715060007743e-05,
+      "loss": 0.0001,
+      "step": 4170
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 3.575300038714673e-05,
+      "loss": 0.0002,
+      "step": 4180
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 3.571428571428572e-05,
+      "loss": 0.0146,
+      "step": 4190
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 3.56755710414247e-05,
+      "loss": 0.0487,
+      "step": 4200
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 3.5636856368563684e-05,
+      "loss": 0.0001,
+      "step": 4210
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 3.559814169570268e-05,
+      "loss": 0.0002,
+      "step": 4220
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.555942702284166e-05,
+      "loss": 0.0299,
+      "step": 4230
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.5520712349980644e-05,
+      "loss": 0.0517,
+      "step": 4240
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.548199767711963e-05,
+      "loss": 0.0196,
+      "step": 4250
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.544328300425862e-05,
+      "loss": 0.0007,
+      "step": 4260
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.5404568331397604e-05,
+      "loss": 0.0005,
+      "step": 4270
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.5365853658536584e-05,
+      "loss": 0.0065,
+      "step": 4280
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.532713898567557e-05,
+      "loss": 0.0214,
+      "step": 4290
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.528842431281456e-05,
+      "loss": 0.0076,
+      "step": 4300
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.5249709639953545e-05,
+      "loss": 0.0153,
+      "step": 4310
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.521099496709253e-05,
+      "loss": 0.0369,
+      "step": 4320
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.517228029423151e-05,
+      "loss": 0.0007,
+      "step": 4330
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.51335656213705e-05,
+      "loss": 0.0981,
+      "step": 4340
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.509485094850949e-05,
+      "loss": 0.0005,
+      "step": 4350
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.505613627564847e-05,
+      "loss": 0.0006,
+      "step": 4360
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.501742160278746e-05,
+      "loss": 0.0158,
+      "step": 4370
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.497870692992644e-05,
+      "loss": 0.0166,
+      "step": 4380
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.493999225706543e-05,
+      "loss": 0.0528,
+      "step": 4390
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.490127758420442e-05,
+      "loss": 0.0013,
+      "step": 4400
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.48625629113434e-05,
+      "loss": 0.0011,
+      "step": 4410
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.4823848238482385e-05,
+      "loss": 0.0304,
+      "step": 4420
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.478513356562137e-05,
+      "loss": 0.0004,
+      "step": 4430
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.474641889276036e-05,
+      "loss": 0.0747,
+      "step": 4440
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.4707704219899346e-05,
+      "loss": 0.118,
+      "step": 4450
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.4668989547038326e-05,
+      "loss": 0.0006,
+      "step": 4460
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.463027487417731e-05,
+      "loss": 0.0498,
+      "step": 4470
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.4591560201316306e-05,
+      "loss": 0.0766,
+      "step": 4480
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.4552845528455286e-05,
+      "loss": 0.0014,
+      "step": 4490
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.451413085559427e-05,
+      "loss": 0.0051,
+      "step": 4500
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.447541618273325e-05,
+      "loss": 0.0045,
+      "step": 4510
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.4436701509872246e-05,
+      "loss": 0.0434,
+      "step": 4520
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.439798683701123e-05,
+      "loss": 0.0001,
+      "step": 4530
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.435927216415021e-05,
+      "loss": 0.0666,
+      "step": 4540
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.43205574912892e-05,
+      "loss": 0.0232,
+      "step": 4550
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.4281842818428186e-05,
+      "loss": 0.0559,
+      "step": 4560
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.424312814556717e-05,
+      "loss": 0.0189,
+      "step": 4570
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.420441347270616e-05,
+      "loss": 0.0336,
+      "step": 4580
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.416569879984514e-05,
+      "loss": 0.0474,
+      "step": 4590
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.412698412698413e-05,
+      "loss": 0.0202,
+      "step": 4600
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.4088269454123113e-05,
+      "loss": 0.0138,
+      "step": 4610
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.40495547812621e-05,
+      "loss": 0.0546,
+      "step": 4620
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.401084010840109e-05,
+      "loss": 0.001,
+      "step": 4630
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.397212543554007e-05,
+      "loss": 0.0191,
+      "step": 4640
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.393341076267906e-05,
+      "loss": 0.0003,
+      "step": 4650
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.389469608981804e-05,
+      "loss": 0.0022,
+      "step": 4660
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.385598141695703e-05,
+      "loss": 0.0023,
+      "step": 4670
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.3817266744096014e-05,
+      "loss": 0.0011,
+      "step": 4680
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.3778552071235e-05,
+      "loss": 0.0002,
+      "step": 4690
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.373983739837399e-05,
+      "loss": 0.0314,
+      "step": 4700
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.3701122725512974e-05,
+      "loss": 0.0205,
+      "step": 4710
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.3662408052651954e-05,
+      "loss": 0.0002,
+      "step": 4720
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.362369337979094e-05,
+      "loss": 0.0122,
+      "step": 4730
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.358497870692993e-05,
+      "loss": 0.0726,
+      "step": 4740
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.3546264034068915e-05,
+      "loss": 0.0205,
+      "step": 4750
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.35075493612079e-05,
+      "loss": 0.0001,
+      "step": 4760
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.346883468834688e-05,
+      "loss": 0.0001,
+      "step": 4770
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.3430120015485875e-05,
+      "loss": 0.0001,
+      "step": 4780
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.3391405342624855e-05,
+      "loss": 0.0616,
+      "step": 4790
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.335269066976384e-05,
+      "loss": 0.0853,
+      "step": 4800
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.331397599690283e-05,
+      "loss": 0.0651,
+      "step": 4810
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.3275261324041815e-05,
+      "loss": 0.0808,
+      "step": 4820
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.32365466511808e-05,
+      "loss": 0.0012,
+      "step": 4830
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.319783197831978e-05,
+      "loss": 0.0013,
+      "step": 4840
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.315911730545877e-05,
+      "loss": 0.0313,
+      "step": 4850
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.3120402632597755e-05,
+      "loss": 0.0283,
+      "step": 4860
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.308168795973674e-05,
+      "loss": 0.0449,
+      "step": 4870
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.304297328687573e-05,
+      "loss": 0.0207,
+      "step": 4880
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.3004258614014716e-05,
+      "loss": 0.017,
+      "step": 4890
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.2965543941153696e-05,
+      "loss": 0.0002,
+      "step": 4900
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.292682926829269e-05,
+      "loss": 0.0003,
+      "step": 4910
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.288811459543167e-05,
+      "loss": 0.0002,
+      "step": 4920
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.2849399922570656e-05,
+      "loss": 0.0028,
+      "step": 4930
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.281068524970964e-05,
+      "loss": 0.0006,
+      "step": 4940
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.277197057684863e-05,
+      "loss": 0.0012,
+      "step": 4950
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.2733255903987616e-05,
+      "loss": 0.0127,
+      "step": 4960
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.2694541231126596e-05,
+      "loss": 0.0758,
+      "step": 4970
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.265582655826558e-05,
+      "loss": 0.0629,
+      "step": 4980
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.261711188540457e-05,
+      "loss": 0.009,
+      "step": 4990
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.2578397212543556e-05,
+      "loss": 0.0343,
+      "step": 5000
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.253968253968254e-05,
+      "loss": 0.019,
+      "step": 5010
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.250096786682152e-05,
+      "loss": 0.0198,
+      "step": 5020
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.246225319396051e-05,
+      "loss": 0.0114,
+      "step": 5030
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.2423538521099503e-05,
+      "loss": 0.083,
+      "step": 5040
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.2384823848238483e-05,
+      "loss": 0.0283,
+      "step": 5050
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.234610917537747e-05,
+      "loss": 0.0007,
+      "step": 5060
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.230739450251645e-05,
+      "loss": 0.0021,
+      "step": 5070
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.2268679829655444e-05,
+      "loss": 0.0014,
+      "step": 5080
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.222996515679443e-05,
+      "loss": 0.0017,
+      "step": 5090
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.219125048393341e-05,
+      "loss": 0.0001,
+      "step": 5100
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.21525358110724e-05,
+      "loss": 0.0378,
+      "step": 5110
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.2113821138211384e-05,
+      "loss": 0.0452,
+      "step": 5120
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.207510646535037e-05,
+      "loss": 0.003,
+      "step": 5130
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.203639179248936e-05,
+      "loss": 0.0005,
+      "step": 5140
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.199767711962834e-05,
+      "loss": 0.0279,
+      "step": 5150
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.1958962446767324e-05,
+      "loss": 0.0113,
+      "step": 5160
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.192024777390631e-05,
+      "loss": 0.0003,
+      "step": 5170
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.18815331010453e-05,
+      "loss": 0.0002,
+      "step": 5180
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.1842818428184285e-05,
+      "loss": 0.0005,
+      "step": 5190
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.1804103755323265e-05,
+      "loss": 0.0023,
+      "step": 5200
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.176538908246225e-05,
+      "loss": 0.0043,
+      "step": 5210
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.1726674409601245e-05,
+      "loss": 0.0001,
+      "step": 5220
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.1687959736740225e-05,
+      "loss": 0.0647,
+      "step": 5230
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.164924506387921e-05,
+      "loss": 0.0546,
+      "step": 5240
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.161053039101819e-05,
+      "loss": 0.0356,
+      "step": 5250
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.1571815718157185e-05,
+      "loss": 0.0154,
+      "step": 5260
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.153310104529617e-05,
+      "loss": 0.0251,
+      "step": 5270
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.149438637243515e-05,
+      "loss": 0.0616,
+      "step": 5280
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.145567169957414e-05,
+      "loss": 0.0577,
+      "step": 5290
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.1416957026713125e-05,
+      "loss": 0.0098,
+      "step": 5300
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.137824235385211e-05,
+      "loss": 0.022,
+      "step": 5310
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.13395276809911e-05,
+      "loss": 0.0041,
+      "step": 5320
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.130081300813008e-05,
+      "loss": 0.0013,
+      "step": 5330
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.1262098335269066e-05,
+      "loss": 0.0564,
+      "step": 5340
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.122338366240806e-05,
+      "loss": 0.0022,
+      "step": 5350
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.118466898954704e-05,
+      "loss": 0.0159,
+      "step": 5360
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9909840648588203,
+      "eval_f1": 0.9906839026503936,
+      "eval_loss": 0.040714751929044724,
+      "eval_runtime": 160.4523,
+      "eval_samples_per_second": 89.173,
+      "eval_steps_per_second": 1.396,
+      "step": 5366
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.1145954316686026e-05,
+      "loss": 0.001,
+      "step": 5370
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1107239643825006e-05,
+      "loss": 0.0004,
+      "step": 5380
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1068524970964e-05,
+      "loss": 0.0001,
+      "step": 5390
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.1029810298102986e-05,
+      "loss": 0.1217,
+      "step": 5400
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0991095625241966e-05,
+      "loss": 0.0008,
+      "step": 5410
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.095238095238095e-05,
+      "loss": 0.0358,
+      "step": 5420
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.091366627951994e-05,
+      "loss": 0.0004,
+      "step": 5430
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 3.0874951606658926e-05,
+      "loss": 0.0082,
+      "step": 5440
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 3.083623693379791e-05,
+      "loss": 0.0537,
+      "step": 5450
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 3.079752226093689e-05,
+      "loss": 0.0003,
+      "step": 5460
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 3.075880758807588e-05,
+      "loss": 0.0178,
+      "step": 5470
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 3.072009291521487e-05,
+      "loss": 0.0311,
+      "step": 5480
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 3.0681378242353853e-05,
+      "loss": 0.0002,
+      "step": 5490
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 3.064266356949284e-05,
+      "loss": 0.0007,
+      "step": 5500
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 3.060394889663182e-05,
+      "loss": 0.0198,
+      "step": 5510
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 3.0565234223770814e-05,
+      "loss": 0.0002,
+      "step": 5520
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 3.05265195509098e-05,
+      "loss": 0.0278,
+      "step": 5530
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 3.048780487804878e-05,
+      "loss": 0.0017,
+      "step": 5540
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 3.044909020518777e-05,
+      "loss": 0.0002,
+      "step": 5550
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 3.041037553232675e-05,
+      "loss": 0.0003,
+      "step": 5560
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 3.037166085946574e-05,
+      "loss": 0.0466,
+      "step": 5570
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 3.0332946186604728e-05,
+      "loss": 0.0001,
+      "step": 5580
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 3.029423151374371e-05,
+      "loss": 0.0001,
+      "step": 5590
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 3.0255516840882698e-05,
+      "loss": 0.0003,
+      "step": 5600
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 3.021680216802168e-05,
+      "loss": 0.0002,
+      "step": 5610
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 3.0178087495160668e-05,
+      "loss": 0.0182,
+      "step": 5620
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 3.0139372822299655e-05,
+      "loss": 0.0375,
+      "step": 5630
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 3.0100658149438638e-05,
+      "loss": 0.0394,
+      "step": 5640
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 3.0061943476577625e-05,
+      "loss": 0.0007,
+      "step": 5650
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 3.0023228803716608e-05,
+      "loss": 0.0016,
+      "step": 5660
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.9984514130855595e-05,
+      "loss": 0.0002,
+      "step": 5670
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.9945799457994585e-05,
+      "loss": 0.0031,
+      "step": 5680
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.9907084785133565e-05,
+      "loss": 0.0001,
+      "step": 5690
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.9868370112272555e-05,
+      "loss": 0.0271,
+      "step": 5700
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.9829655439411535e-05,
+      "loss": 0.0095,
+      "step": 5710
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.9790940766550525e-05,
+      "loss": 0.0104,
+      "step": 5720
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.9752226093689512e-05,
+      "loss": 0.001,
+      "step": 5730
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.9713511420828495e-05,
+      "loss": 0.0537,
+      "step": 5740
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.9674796747967482e-05,
+      "loss": 0.0005,
+      "step": 5750
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.963608207510647e-05,
+      "loss": 0.0001,
+      "step": 5760
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.9597367402245452e-05,
+      "loss": 0.0003,
+      "step": 5770
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.955865272938444e-05,
+      "loss": 0.0001,
+      "step": 5780
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.9519938056523422e-05,
+      "loss": 0.0218,
+      "step": 5790
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.948122338366241e-05,
+      "loss": 0.0,
+      "step": 5800
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.9442508710801396e-05,
+      "loss": 0.0001,
+      "step": 5810
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.940379403794038e-05,
+      "loss": 0.0255,
+      "step": 5820
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.9365079365079366e-05,
+      "loss": 0.0002,
+      "step": 5830
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.932636469221835e-05,
+      "loss": 0.0,
+      "step": 5840
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.9287650019357336e-05,
+      "loss": 0.0,
+      "step": 5850
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.9248935346496326e-05,
+      "loss": 0.0004,
+      "step": 5860
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.9210220673635306e-05,
+      "loss": 0.0196,
+      "step": 5870
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.9171506000774296e-05,
+      "loss": 0.0041,
+      "step": 5880
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.9132791327913276e-05,
+      "loss": 0.0001,
+      "step": 5890
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.9094076655052267e-05,
+      "loss": 0.0001,
+      "step": 5900
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.9055361982191253e-05,
+      "loss": 0.0,
+      "step": 5910
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.9016647309330237e-05,
+      "loss": 0.0001,
+      "step": 5920
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.8977932636469223e-05,
+      "loss": 0.021,
+      "step": 5930
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.8939217963608207e-05,
+      "loss": 0.0002,
+      "step": 5940
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.8900503290747194e-05,
+      "loss": 0.0001,
+      "step": 5950
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.886178861788618e-05,
+      "loss": 0.0001,
+      "step": 5960
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.8823073945025164e-05,
+      "loss": 0.0003,
+      "step": 5970
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.878435927216415e-05,
+      "loss": 0.0005,
+      "step": 5980
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.874564459930314e-05,
+      "loss": 0.0,
+      "step": 5990
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.870692992644212e-05,
+      "loss": 0.0561,
+      "step": 6000
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.866821525358111e-05,
+      "loss": 0.0001,
+      "step": 6010
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.862950058072009e-05,
+      "loss": 0.0088,
+      "step": 6020
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.859078590785908e-05,
+      "loss": 0.0,
+      "step": 6030
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.8552071234998068e-05,
+      "loss": 0.0401,
+      "step": 6040
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.851335656213705e-05,
+      "loss": 0.0001,
+      "step": 6050
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.8474641889276038e-05,
+      "loss": 0.001,
+      "step": 6060
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.843592721641502e-05,
+      "loss": 0.0,
+      "step": 6070
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.8397212543554008e-05,
+      "loss": 0.035,
+      "step": 6080
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.8358497870692995e-05,
+      "loss": 0.0,
+      "step": 6090
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.8319783197831978e-05,
+      "loss": 0.0,
+      "step": 6100
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.8281068524970965e-05,
+      "loss": 0.0016,
+      "step": 6110
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.8242353852109948e-05,
+      "loss": 0.0001,
+      "step": 6120
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.8203639179248935e-05,
+      "loss": 0.0001,
+      "step": 6130
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.8164924506387925e-05,
+      "loss": 0.0,
+      "step": 6140
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.8126209833526905e-05,
+      "loss": 0.0,
+      "step": 6150
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.8087495160665895e-05,
+      "loss": 0.0001,
+      "step": 6160
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.8048780487804882e-05,
+      "loss": 0.0001,
+      "step": 6170
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.8010065814943865e-05,
+      "loss": 0.0,
+      "step": 6180
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.7971351142082852e-05,
+      "loss": 0.0421,
+      "step": 6190
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.7932636469221835e-05,
+      "loss": 0.0,
+      "step": 6200
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.7893921796360822e-05,
+      "loss": 0.0,
+      "step": 6210
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.785520712349981e-05,
+      "loss": 0.0001,
+      "step": 6220
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.7816492450638792e-05,
+      "loss": 0.0,
+      "step": 6230
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.0,
+      "step": 6240
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.7739063104916762e-05,
+      "loss": 0.0035,
+      "step": 6250
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.770034843205575e-05,
+      "loss": 0.0,
+      "step": 6260
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.766163375919474e-05,
+      "loss": 0.0,
+      "step": 6270
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.762291908633372e-05,
+      "loss": 0.0,
+      "step": 6280
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.758420441347271e-05,
+      "loss": 0.0698,
+      "step": 6290
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.754548974061169e-05,
+      "loss": 0.0001,
+      "step": 6300
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.750677506775068e-05,
+      "loss": 0.0005,
+      "step": 6310
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.7468060394889666e-05,
+      "loss": 0.0805,
+      "step": 6320
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.742934572202865e-05,
+      "loss": 0.0005,
+      "step": 6330
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.7390631049167637e-05,
+      "loss": 0.0155,
+      "step": 6340
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.735191637630662e-05,
+      "loss": 0.0448,
+      "step": 6350
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.7313201703445607e-05,
+      "loss": 0.0002,
+      "step": 6360
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.7274487030584593e-05,
+      "loss": 0.0291,
+      "step": 6370
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.7235772357723577e-05,
+      "loss": 0.0006,
+      "step": 6380
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.7197057684862564e-05,
+      "loss": 0.0005,
+      "step": 6390
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.7158343012001554e-05,
+      "loss": 0.0002,
+      "step": 6400
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.7119628339140534e-05,
+      "loss": 0.0298,
+      "step": 6410
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.7080913666279524e-05,
+      "loss": 0.0952,
+      "step": 6420
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.7042198993418504e-05,
+      "loss": 0.0005,
+      "step": 6430
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.7003484320557494e-05,
+      "loss": 0.0099,
+      "step": 6440
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.696476964769648e-05,
+      "loss": 0.0004,
+      "step": 6450
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.6926054974835464e-05,
+      "loss": 0.0002,
+      "step": 6460
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.688734030197445e-05,
+      "loss": 0.0166,
+      "step": 6470
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.6848625629113434e-05,
+      "loss": 0.0019,
+      "step": 6480
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.680991095625242e-05,
+      "loss": 0.0067,
+      "step": 6490
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.6771196283391408e-05,
+      "loss": 0.0,
+      "step": 6500
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.673248161053039e-05,
+      "loss": 0.0002,
+      "step": 6510
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.6693766937669378e-05,
+      "loss": 0.0003,
+      "step": 6520
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.665505226480836e-05,
+      "loss": 0.0001,
+      "step": 6530
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2.6616337591947348e-05,
+      "loss": 0.0,
+      "step": 6540
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2.6577622919086338e-05,
+      "loss": 0.0352,
+      "step": 6550
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 2.6538908246225318e-05,
+      "loss": 0.0002,
+      "step": 6560
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 2.6500193573364308e-05,
+      "loss": 0.0006,
+      "step": 6570
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 2.6461478900503288e-05,
+      "loss": 0.0007,
+      "step": 6580
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 2.642276422764228e-05,
+      "loss": 0.0025,
+      "step": 6590
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 2.6384049554781265e-05,
+      "loss": 0.041,
+      "step": 6600
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 2.634533488192025e-05,
+      "loss": 0.0001,
+      "step": 6610
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 2.6306620209059235e-05,
+      "loss": 0.0155,
+      "step": 6620
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 2.6267905536198222e-05,
+      "loss": 0.0177,
+      "step": 6630
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 2.6229190863337205e-05,
+      "loss": 0.0001,
+      "step": 6640
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 2.6190476190476192e-05,
+      "loss": 0.0003,
+      "step": 6650
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 2.6151761517615176e-05,
+      "loss": 0.0,
+      "step": 6660
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 2.6113046844754162e-05,
+      "loss": 0.0621,
+      "step": 6670
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 2.6074332171893152e-05,
+      "loss": 0.0009,
+      "step": 6680
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 2.6035617499032132e-05,
+      "loss": 0.0002,
+      "step": 6690
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 2.5996902826171123e-05,
+      "loss": 0.0002,
+      "step": 6700
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 2.5958188153310103e-05,
+      "loss": 0.0223,
+      "step": 6710
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 2.5919473480449093e-05,
+      "loss": 0.0002,
+      "step": 6720
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 2.588075880758808e-05,
+      "loss": 0.0002,
+      "step": 6730
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 2.5842044134727063e-05,
+      "loss": 0.0245,
+      "step": 6740
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 2.580332946186605e-05,
+      "loss": 0.0004,
+      "step": 6750
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 2.5764614789005033e-05,
+      "loss": 0.0001,
+      "step": 6760
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 2.572590011614402e-05,
+      "loss": 0.0339,
+      "step": 6770
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 2.5687185443283007e-05,
+      "loss": 0.0001,
+      "step": 6780
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 2.564847077042199e-05,
+      "loss": 0.0002,
+      "step": 6790
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 2.5609756097560977e-05,
+      "loss": 0.0,
+      "step": 6800
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 2.5571041424699967e-05,
+      "loss": 0.0,
+      "step": 6810
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 2.5532326751838947e-05,
+      "loss": 0.0,
+      "step": 6820
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.5493612078977937e-05,
+      "loss": 0.0001,
+      "step": 6830
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.5454897406116917e-05,
+      "loss": 0.0001,
+      "step": 6840
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.5416182733255907e-05,
+      "loss": 0.0001,
+      "step": 6850
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.5377468060394894e-05,
+      "loss": 0.0,
+      "step": 6860
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.5338753387533877e-05,
+      "loss": 0.0065,
+      "step": 6870
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.5300038714672864e-05,
+      "loss": 0.0288,
+      "step": 6880
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.5261324041811847e-05,
+      "loss": 0.0,
+      "step": 6890
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.5222609368950834e-05,
+      "loss": 0.0001,
+      "step": 6900
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.518389469608982e-05,
+      "loss": 0.0006,
+      "step": 6910
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.5145180023228804e-05,
+      "loss": 0.0,
+      "step": 6920
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.510646535036779e-05,
+      "loss": 0.0001,
+      "step": 6930
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.5067750677506774e-05,
+      "loss": 0.0527,
+      "step": 6940
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.502903600464576e-05,
+      "loss": 0.0001,
+      "step": 6950
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.4990321331784748e-05,
+      "loss": 0.0,
+      "step": 6960
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.4951606658923735e-05,
+      "loss": 0.0001,
+      "step": 6970
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.4912891986062718e-05,
+      "loss": 0.0,
+      "step": 6980
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.4874177313201705e-05,
+      "loss": 0.0,
+      "step": 6990
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.4835462640340688e-05,
+      "loss": 0.0007,
+      "step": 7000
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.4796747967479675e-05,
+      "loss": 0.0651,
+      "step": 7010
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.475803329461866e-05,
+      "loss": 0.0166,
+      "step": 7020
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.471931862175765e-05,
+      "loss": 0.0003,
+      "step": 7030
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.4680603948896632e-05,
+      "loss": 0.0002,
+      "step": 7040
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.464188927603562e-05,
+      "loss": 0.0002,
+      "step": 7050
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.4603174603174602e-05,
+      "loss": 0.0451,
+      "step": 7060
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.4564459930313592e-05,
+      "loss": 0.0001,
+      "step": 7070
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.4525745257452575e-05,
+      "loss": 0.0002,
+      "step": 7080
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.4487030584591562e-05,
+      "loss": 0.0334,
+      "step": 7090
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.4448315911730546e-05,
+      "loss": 0.0001,
+      "step": 7100
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.4409601238869532e-05,
+      "loss": 0.0338,
+      "step": 7110
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.437088656600852e-05,
+      "loss": 0.0001,
+      "step": 7120
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.4332171893147502e-05,
+      "loss": 0.0292,
+      "step": 7130
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.429345722028649e-05,
+      "loss": 0.0294,
+      "step": 7140
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.4254742547425473e-05,
+      "loss": 0.0001,
+      "step": 7150
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.4216027874564463e-05,
+      "loss": 0.0002,
+      "step": 7160
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.4177313201703446e-05,
+      "loss": 0.0008,
+      "step": 7170
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.4138598528842433e-05,
+      "loss": 0.0264,
+      "step": 7180
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.4099883855981416e-05,
+      "loss": 0.0001,
+      "step": 7190
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.4061169183120406e-05,
+      "loss": 0.0001,
+      "step": 7200
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.402245451025939e-05,
+      "loss": 0.0496,
+      "step": 7210
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.3983739837398377e-05,
+      "loss": 0.0001,
+      "step": 7220
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.394502516453736e-05,
+      "loss": 0.0001,
+      "step": 7230
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.3906310491676347e-05,
+      "loss": 0.0904,
+      "step": 7240
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.3867595818815333e-05,
+      "loss": 0.0001,
+      "step": 7250
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.3828881145954317e-05,
+      "loss": 0.0001,
+      "step": 7260
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.3790166473093304e-05,
+      "loss": 0.0001,
+      "step": 7270
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.3751451800232287e-05,
+      "loss": 0.0001,
+      "step": 7280
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.3712737127371277e-05,
+      "loss": 0.0252,
+      "step": 7290
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.367402245451026e-05,
+      "loss": 0.0001,
+      "step": 7300
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.3635307781649247e-05,
+      "loss": 0.0369,
+      "step": 7310
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2.359659310878823e-05,
+      "loss": 0.0001,
+      "step": 7320
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2.3557878435927217e-05,
+      "loss": 0.0003,
+      "step": 7330
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 2.3519163763066204e-05,
+      "loss": 0.0044,
+      "step": 7340
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 2.348044909020519e-05,
+      "loss": 0.0148,
+      "step": 7350
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 2.3441734417344174e-05,
+      "loss": 0.0002,
+      "step": 7360
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 2.340301974448316e-05,
+      "loss": 0.0159,
+      "step": 7370
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 2.3364305071622144e-05,
+      "loss": 0.0282,
+      "step": 7380
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 2.332559039876113e-05,
+      "loss": 0.0008,
+      "step": 7390
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 2.3286875725900118e-05,
+      "loss": 0.0001,
+      "step": 7400
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 2.32481610530391e-05,
+      "loss": 0.0001,
+      "step": 7410
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.3209446380178088e-05,
+      "loss": 0.0001,
+      "step": 7420
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.3170731707317075e-05,
+      "loss": 0.0071,
+      "step": 7430
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.313201703445606e-05,
+      "loss": 0.0285,
+      "step": 7440
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.3093302361595045e-05,
+      "loss": 0.0002,
+      "step": 7450
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.305458768873403e-05,
+      "loss": 0.0005,
+      "step": 7460
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.3015873015873015e-05,
+      "loss": 0.028,
+      "step": 7470
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.2977158343012005e-05,
+      "loss": 0.0317,
+      "step": 7480
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.293844367015099e-05,
+      "loss": 0.0012,
+      "step": 7490
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.2899728997289975e-05,
+      "loss": 0.0003,
+      "step": 7500
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.286101432442896e-05,
+      "loss": 0.0002,
+      "step": 7510
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.2822299651567945e-05,
+      "loss": 0.0,
+      "step": 7520
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.2783584978706932e-05,
+      "loss": 0.0333,
+      "step": 7530
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.2744870305845916e-05,
+      "loss": 0.023,
+      "step": 7540
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.2706155632984902e-05,
+      "loss": 0.0,
+      "step": 7550
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.2667440960123886e-05,
+      "loss": 0.0,
+      "step": 7560
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.2628726287262876e-05,
+      "loss": 0.0001,
+      "step": 7570
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.259001161440186e-05,
+      "loss": 0.0003,
+      "step": 7580
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.2551296941540846e-05,
+      "loss": 0.0001,
+      "step": 7590
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.251258226867983e-05,
+      "loss": 0.0037,
+      "step": 7600
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.2473867595818816e-05,
+      "loss": 0.0776,
+      "step": 7610
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.2435152922957803e-05,
+      "loss": 0.0314,
+      "step": 7620
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.2396438250096786e-05,
+      "loss": 0.0282,
+      "step": 7630
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.2357723577235773e-05,
+      "loss": 0.0,
+      "step": 7640
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.2319008904374756e-05,
+      "loss": 0.0457,
+      "step": 7650
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.2280294231513747e-05,
+      "loss": 0.0,
+      "step": 7660
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.224157955865273e-05,
+      "loss": 0.0202,
+      "step": 7670
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.2202864885791717e-05,
+      "loss": 0.0001,
+      "step": 7680
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.21641502129307e-05,
+      "loss": 0.0038,
+      "step": 7690
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.2125435540069687e-05,
+      "loss": 0.0002,
+      "step": 7700
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.2086720867208674e-05,
+      "loss": 0.0001,
+      "step": 7710
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.204800619434766e-05,
+      "loss": 0.0187,
+      "step": 7720
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.2009291521486644e-05,
+      "loss": 0.0001,
+      "step": 7730
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.197057684862563e-05,
+      "loss": 0.0002,
+      "step": 7740
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 2.1931862175764617e-05,
+      "loss": 0.0,
+      "step": 7750
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 2.18931475029036e-05,
+      "loss": 0.0035,
+      "step": 7760
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 2.1854432830042587e-05,
+      "loss": 0.0028,
+      "step": 7770
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 2.181571815718157e-05,
+      "loss": 0.0,
+      "step": 7780
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 2.1777003484320557e-05,
+      "loss": 0.1069,
+      "step": 7790
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 2.1738288811459544e-05,
+      "loss": 0.0001,
+      "step": 7800
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 2.169957413859853e-05,
+      "loss": 0.0004,
+      "step": 7810
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 2.1660859465737514e-05,
+      "loss": 0.0006,
+      "step": 7820
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 2.16221447928765e-05,
+      "loss": 0.0001,
+      "step": 7830
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 2.1583430120015488e-05,
+      "loss": 0.0727,
+      "step": 7840
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.1544715447154475e-05,
+      "loss": 0.0002,
+      "step": 7850
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.1506000774293458e-05,
+      "loss": 0.0025,
+      "step": 7860
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.1467286101432445e-05,
+      "loss": 0.0008,
+      "step": 7870
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 0.0012,
+      "step": 7880
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.1389856755710415e-05,
+      "loss": 0.0003,
+      "step": 7890
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.13511420828494e-05,
+      "loss": 0.0002,
+      "step": 7900
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 2.1312427409988385e-05,
+      "loss": 0.0001,
+      "step": 7910
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 2.1273712737127372e-05,
+      "loss": 0.0123,
+      "step": 7920
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.123499806426636e-05,
+      "loss": 0.0002,
+      "step": 7930
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.1196283391405345e-05,
+      "loss": 0.0005,
+      "step": 7940
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.115756871854433e-05,
+      "loss": 0.0001,
+      "step": 7950
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.1118854045683315e-05,
+      "loss": 0.0001,
+      "step": 7960
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.10801393728223e-05,
+      "loss": 0.0018,
+      "step": 7970
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.104142469996129e-05,
+      "loss": 0.0,
+      "step": 7980
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.1002710027100272e-05,
+      "loss": 0.0,
+      "step": 7990
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.096399535423926e-05,
+      "loss": 0.0,
+      "step": 8000
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 2.0925280681378242e-05,
+      "loss": 0.0228,
+      "step": 8010
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 2.088656600851723e-05,
+      "loss": 0.0001,
+      "step": 8020
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 2.0847851335656216e-05,
+      "loss": 0.0003,
+      "step": 8030
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 2.08091366627952e-05,
+      "loss": 0.0687,
+      "step": 8040
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.9946882862734134,
+      "eval_f1": 0.9944991314418066,
+      "eval_loss": 0.033534225076436996,
+      "eval_runtime": 139.2743,
+      "eval_samples_per_second": 102.732,
+      "eval_steps_per_second": 1.608,
+      "step": 8049
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 2.0770421989934186e-05,
+      "loss": 0.0,
+      "step": 8050
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 2.073170731707317e-05,
+      "loss": 0.0,
+      "step": 8060
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 2.069299264421216e-05,
+      "loss": 0.0,
+      "step": 8070
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 2.0654277971351143e-05,
+      "loss": 0.0,
+      "step": 8080
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 2.061556329849013e-05,
+      "loss": 0.0,
+      "step": 8090
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 2.0576848625629113e-05,
+      "loss": 0.0,
+      "step": 8100
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 2.05381339527681e-05,
+      "loss": 0.0,
+      "step": 8110
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 2.0499419279907087e-05,
+      "loss": 0.0003,
+      "step": 8120
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 2.046070460704607e-05,
+      "loss": 0.0001,
+      "step": 8130
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 2.0421989934185057e-05,
+      "loss": 0.0,
+      "step": 8140
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 2.038327526132404e-05,
+      "loss": 0.0,
+      "step": 8150
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 2.034456058846303e-05,
+      "loss": 0.0,
+      "step": 8160
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 2.0305845915602014e-05,
+      "loss": 0.0094,
+      "step": 8170
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 2.0267131242741e-05,
+      "loss": 0.0,
+      "step": 8180
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 2.0228416569879984e-05,
+      "loss": 0.0827,
+      "step": 8190
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 2.018970189701897e-05,
+      "loss": 0.0,
+      "step": 8200
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 2.0150987224157957e-05,
+      "loss": 0.0805,
+      "step": 8210
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 2.0112272551296944e-05,
+      "loss": 0.0225,
+      "step": 8220
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 2.0073557878435927e-05,
+      "loss": 0.0086,
+      "step": 8230
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 2.0034843205574914e-05,
+      "loss": 0.0001,
+      "step": 8240
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.99961285327139e-05,
+      "loss": 0.0002,
+      "step": 8250
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.9957413859852884e-05,
+      "loss": 0.0005,
+      "step": 8260
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.991869918699187e-05,
+      "loss": 0.0009,
+      "step": 8270
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.9879984514130854e-05,
+      "loss": 0.0,
+      "step": 8280
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.984126984126984e-05,
+      "loss": 0.0001,
+      "step": 8290
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.9802555168408828e-05,
+      "loss": 0.0292,
+      "step": 8300
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.9763840495547815e-05,
+      "loss": 0.0001,
+      "step": 8310
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.9725125822686798e-05,
+      "loss": 0.0,
+      "step": 8320
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.9686411149825785e-05,
+      "loss": 0.0,
+      "step": 8330
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.9647696476964768e-05,
+      "loss": 0.0005,
+      "step": 8340
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.960898180410376e-05,
+      "loss": 0.0352,
+      "step": 8350
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.9570267131242742e-05,
+      "loss": 0.0,
+      "step": 8360
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.953155245838173e-05,
+      "loss": 0.1081,
+      "step": 8370
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.9492837785520712e-05,
+      "loss": 0.0001,
+      "step": 8380
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.94541231126597e-05,
+      "loss": 0.0003,
+      "step": 8390
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.9415408439798685e-05,
+      "loss": 0.0077,
+      "step": 8400
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.937669376693767e-05,
+      "loss": 0.0003,
+      "step": 8410
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.9337979094076656e-05,
+      "loss": 0.0006,
+      "step": 8420
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.929926442121564e-05,
+      "loss": 0.0003,
+      "step": 8430
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.926054974835463e-05,
+      "loss": 0.0058,
+      "step": 8440
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.9221835075493612e-05,
+      "loss": 0.0001,
+      "step": 8450
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.91831204026326e-05,
+      "loss": 0.0055,
+      "step": 8460
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.9144405729771583e-05,
+      "loss": 0.0001,
+      "step": 8470
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.9105691056910573e-05,
+      "loss": 0.0316,
+      "step": 8480
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.9066976384049556e-05,
+      "loss": 0.0003,
+      "step": 8490
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.9028261711188543e-05,
+      "loss": 0.0001,
+      "step": 8500
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.8989547038327526e-05,
+      "loss": 0.0001,
+      "step": 8510
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.8950832365466513e-05,
+      "loss": 0.0039,
+      "step": 8520
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.89121176926055e-05,
+      "loss": 0.0,
+      "step": 8530
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.8873403019744483e-05,
+      "loss": 0.0,
+      "step": 8540
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.883468834688347e-05,
+      "loss": 0.0001,
+      "step": 8550
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.8795973674022453e-05,
+      "loss": 0.0,
+      "step": 8560
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.8757259001161443e-05,
+      "loss": 0.0,
+      "step": 8570
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.8718544328300427e-05,
+      "loss": 0.0,
+      "step": 8580
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.8679829655439414e-05,
+      "loss": 0.0001,
+      "step": 8590
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.8641114982578397e-05,
+      "loss": 0.0,
+      "step": 8600
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.8602400309717384e-05,
+      "loss": 0.0,
+      "step": 8610
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.856368563685637e-05,
+      "loss": 0.0001,
+      "step": 8620
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.8524970963995357e-05,
+      "loss": 0.0,
+      "step": 8630
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.848625629113434e-05,
+      "loss": 0.0,
+      "step": 8640
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.8447541618273327e-05,
+      "loss": 0.0002,
+      "step": 8650
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1.840882694541231e-05,
+      "loss": 0.0,
+      "step": 8660
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1.8370112272551297e-05,
+      "loss": 0.0,
+      "step": 8670
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.8331397599690284e-05,
+      "loss": 0.0,
+      "step": 8680
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.8292682926829268e-05,
+      "loss": 0.0,
+      "step": 8690
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.8253968253968254e-05,
+      "loss": 0.0,
+      "step": 8700
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1.821525358110724e-05,
+      "loss": 0.0,
+      "step": 8710
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1.8176538908246228e-05,
+      "loss": 0.0022,
+      "step": 8720
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1.813782423538521e-05,
+      "loss": 0.0,
+      "step": 8730
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.8099109562524198e-05,
+      "loss": 0.0,
+      "step": 8740
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.806039488966318e-05,
+      "loss": 0.0001,
+      "step": 8750
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.8021680216802168e-05,
+      "loss": 0.0,
+      "step": 8760
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.7982965543941155e-05,
+      "loss": 0.0,
+      "step": 8770
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.7944250871080138e-05,
+      "loss": 0.0004,
+      "step": 8780
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.7905536198219125e-05,
+      "loss": 0.0,
+      "step": 8790
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.7866821525358112e-05,
+      "loss": 0.0,
+      "step": 8800
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.78281068524971e-05,
+      "loss": 0.0,
+      "step": 8810
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.7789392179636082e-05,
+      "loss": 0.0001,
+      "step": 8820
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.775067750677507e-05,
+      "loss": 0.0,
+      "step": 8830
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.7711962833914052e-05,
+      "loss": 0.0,
+      "step": 8840
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.7673248161053042e-05,
+      "loss": 0.0,
+      "step": 8850
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.7634533488192026e-05,
+      "loss": 0.0,
+      "step": 8860
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1.7595818815331012e-05,
+      "loss": 0.0,
+      "step": 8870
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1.7557104142469996e-05,
+      "loss": 0.0,
+      "step": 8880
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1.7518389469608982e-05,
+      "loss": 0.0,
+      "step": 8890
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.747967479674797e-05,
+      "loss": 0.0,
+      "step": 8900
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.7440960123886953e-05,
+      "loss": 0.0,
+      "step": 8910
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.740224545102594e-05,
+      "loss": 0.0,
+      "step": 8920
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1.7363530778164923e-05,
+      "loss": 0.0,
+      "step": 8930
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1.7324816105303913e-05,
+      "loss": 0.0,
+      "step": 8940
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.7286101432442896e-05,
+      "loss": 0.0,
+      "step": 8950
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.7247386759581883e-05,
+      "loss": 0.0,
+      "step": 8960
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.7208672086720866e-05,
+      "loss": 0.0,
+      "step": 8970
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.7169957413859853e-05,
+      "loss": 0.029,
+      "step": 8980
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.713124274099884e-05,
+      "loss": 0.0,
+      "step": 8990
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.7092528068137827e-05,
+      "loss": 0.0,
+      "step": 9000
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.705381339527681e-05,
+      "loss": 0.0,
+      "step": 9010
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.7015098722415797e-05,
+      "loss": 0.0,
+      "step": 9020
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.6976384049554784e-05,
+      "loss": 0.0,
+      "step": 9030
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.6937669376693767e-05,
+      "loss": 0.0398,
+      "step": 9040
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.6898954703832754e-05,
+      "loss": 0.0001,
+      "step": 9050
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.6860240030971737e-05,
+      "loss": 0.0,
+      "step": 9060
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.6821525358110724e-05,
+      "loss": 0.0,
+      "step": 9070
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.678281068524971e-05,
+      "loss": 0.0,
+      "step": 9080
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1.6744096012388697e-05,
+      "loss": 0.0,
+      "step": 9090
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1.670538133952768e-05,
+      "loss": 0.0,
+      "step": 9100
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.0002,
+      "step": 9110
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 1.6627951993805654e-05,
+      "loss": 0.0056,
+      "step": 9120
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 1.658923732094464e-05,
+      "loss": 0.0001,
+      "step": 9130
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 1.6550522648083624e-05,
+      "loss": 0.0,
+      "step": 9140
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 1.651180797522261e-05,
+      "loss": 0.0,
+      "step": 9150
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 1.6473093302361594e-05,
+      "loss": 0.0,
+      "step": 9160
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 1.643437862950058e-05,
+      "loss": 0.0,
+      "step": 9170
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 1.6395663956639568e-05,
+      "loss": 0.0,
+      "step": 9180
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 1.635694928377855e-05,
+      "loss": 0.0115,
+      "step": 9190
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 1.6318234610917538e-05,
+      "loss": 0.0,
+      "step": 9200
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 1.6279519938056525e-05,
+      "loss": 0.0,
+      "step": 9210
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1.624080526519551e-05,
+      "loss": 0.0,
+      "step": 9220
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1.6202090592334495e-05,
+      "loss": 0.0,
+      "step": 9230
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1.6163375919473482e-05,
+      "loss": 0.0001,
+      "step": 9240
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 1.6124661246612465e-05,
+      "loss": 0.0255,
+      "step": 9250
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 1.6085946573751455e-05,
+      "loss": 0.0,
+      "step": 9260
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 1.604723190089044e-05,
+      "loss": 0.0,
+      "step": 9270
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 1.6008517228029425e-05,
+      "loss": 0.0,
+      "step": 9280
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 1.596980255516841e-05,
+      "loss": 0.0002,
+      "step": 9290
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 1.5931087882307396e-05,
+      "loss": 0.0,
+      "step": 9300
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 1.5892373209446382e-05,
+      "loss": 0.0,
+      "step": 9310
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 1.5853658536585366e-05,
+      "loss": 0.0069,
+      "step": 9320
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 1.5814943863724352e-05,
+      "loss": 0.0,
+      "step": 9330
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 1.5776229190863336e-05,
+      "loss": 0.0376,
+      "step": 9340
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 1.5737514518002326e-05,
+      "loss": 0.0,
+      "step": 9350
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 1.569879984514131e-05,
+      "loss": 0.0046,
+      "step": 9360
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 1.5660085172280296e-05,
+      "loss": 0.0,
+      "step": 9370
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 1.562137049941928e-05,
+      "loss": 0.0,
+      "step": 9380
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 1.5582655826558266e-05,
+      "loss": 0.0002,
+      "step": 9390
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 1.5543941153697253e-05,
+      "loss": 0.0007,
+      "step": 9400
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 1.5505226480836236e-05,
+      "loss": 0.0,
+      "step": 9410
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 1.5466511807975223e-05,
+      "loss": 0.0,
+      "step": 9420
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 1.5427797135114207e-05,
+      "loss": 0.0326,
+      "step": 9430
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 1.5389082462253197e-05,
+      "loss": 0.0,
+      "step": 9440
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 1.535036778939218e-05,
+      "loss": 0.0,
+      "step": 9450
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 1.5311653116531167e-05,
+      "loss": 0.0001,
+      "step": 9460
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 1.527293844367015e-05,
+      "loss": 0.0,
+      "step": 9470
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 1.5234223770809137e-05,
+      "loss": 0.0001,
+      "step": 9480
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 1.5195509097948124e-05,
+      "loss": 0.0,
+      "step": 9490
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 1.5156794425087109e-05,
+      "loss": 0.0,
+      "step": 9500
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 1.5118079752226094e-05,
+      "loss": 0.0001,
+      "step": 9510
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 1.5079365079365079e-05,
+      "loss": 0.0,
+      "step": 9520
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 1.5040650406504067e-05,
+      "loss": 0.0,
+      "step": 9530
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 1.5001935733643052e-05,
+      "loss": 0.0002,
+      "step": 9540
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 1.4963221060782037e-05,
+      "loss": 0.0,
+      "step": 9550
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 1.4924506387921023e-05,
+      "loss": 0.0,
+      "step": 9560
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 1.4885791715060008e-05,
+      "loss": 0.0,
+      "step": 9570
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 1.4847077042198994e-05,
+      "loss": 0.0002,
+      "step": 9580
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 1.4808362369337981e-05,
+      "loss": 0.0,
+      "step": 9590
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 1.4769647696476966e-05,
+      "loss": 0.0728,
+      "step": 9600
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 1.4730933023615951e-05,
+      "loss": 0.0001,
+      "step": 9610
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 1.4692218350754936e-05,
+      "loss": 0.0,
+      "step": 9620
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 1.4653503677893923e-05,
+      "loss": 0.0,
+      "step": 9630
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 1.4614789005032908e-05,
+      "loss": 0.0,
+      "step": 9640
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 1.4576074332171893e-05,
+      "loss": 0.0214,
+      "step": 9650
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 1.4537359659310878e-05,
+      "loss": 0.0,
+      "step": 9660
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 1.4498644986449867e-05,
+      "loss": 0.0,
+      "step": 9670
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 1.4459930313588852e-05,
+      "loss": 0.0492,
+      "step": 9680
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 1.4421215640727837e-05,
+      "loss": 0.0001,
+      "step": 9690
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 1.4382500967866822e-05,
+      "loss": 0.0002,
+      "step": 9700
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 1.4343786295005807e-05,
+      "loss": 0.0003,
+      "step": 9710
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 1.4305071622144794e-05,
+      "loss": 0.0376,
+      "step": 9720
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 1.4266356949283779e-05,
+      "loss": 0.0,
+      "step": 9730
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 1.4227642276422764e-05,
+      "loss": 0.0001,
+      "step": 9740
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 1.4188927603561749e-05,
+      "loss": 0.0001,
+      "step": 9750
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 1.4150212930700737e-05,
+      "loss": 0.0,
+      "step": 9760
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 1.4111498257839722e-05,
+      "loss": 0.0,
+      "step": 9770
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 1.4072783584978708e-05,
+      "loss": 0.0,
+      "step": 9780
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 1.4034068912117693e-05,
+      "loss": 0.0,
+      "step": 9790
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 1.3995354239256678e-05,
+      "loss": 0.0001,
+      "step": 9800
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 1.3956639566395666e-05,
+      "loss": 0.0,
+      "step": 9810
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 1.3917924893534651e-05,
+      "loss": 0.0,
+      "step": 9820
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 1.3879210220673636e-05,
+      "loss": 0.0033,
+      "step": 9830
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 1.3840495547812621e-05,
+      "loss": 0.0001,
+      "step": 9840
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 1.3801780874951608e-05,
+      "loss": 0.0372,
+      "step": 9850
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 1.3763066202090593e-05,
+      "loss": 0.0247,
+      "step": 9860
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 1.3724351529229578e-05,
+      "loss": 0.0003,
+      "step": 9870
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 1.3685636856368563e-05,
+      "loss": 0.0002,
+      "step": 9880
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 1.3646922183507548e-05,
+      "loss": 0.0,
+      "step": 9890
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 1.3608207510646537e-05,
+      "loss": 0.0001,
+      "step": 9900
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 1.3569492837785522e-05,
+      "loss": 0.0,
+      "step": 9910
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 1.3530778164924507e-05,
+      "loss": 0.0,
+      "step": 9920
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 1.3492063492063492e-05,
+      "loss": 0.0,
+      "step": 9930
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 1.3453348819202477e-05,
+      "loss": 0.0,
+      "step": 9940
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 1.3414634146341466e-05,
+      "loss": 0.0,
+      "step": 9950
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 1.337591947348045e-05,
+      "loss": 0.0,
+      "step": 9960
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 1.3337204800619436e-05,
+      "loss": 0.0001,
+      "step": 9970
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 1.329849012775842e-05,
+      "loss": 0.0,
+      "step": 9980
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 1.3259775454897407e-05,
+      "loss": 0.0,
+      "step": 9990
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 1.3221060782036393e-05,
+      "loss": 0.0,
+      "step": 10000
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 1.3182346109175378e-05,
+      "loss": 0.0,
+      "step": 10010
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 1.3143631436314363e-05,
+      "loss": 0.0,
+      "step": 10020
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 1.3104916763453348e-05,
+      "loss": 0.0,
+      "step": 10030
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 1.3066202090592336e-05,
+      "loss": 0.0,
+      "step": 10040
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 1.3027487417731321e-05,
+      "loss": 0.0,
+      "step": 10050
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 1.2988772744870306e-05,
+      "loss": 0.0,
+      "step": 10060
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 1.2950058072009291e-05,
+      "loss": 0.0005,
+      "step": 10070
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 1.291134339914828e-05,
+      "loss": 0.0001,
+      "step": 10080
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 1.2872628726287265e-05,
+      "loss": 0.0,
+      "step": 10090
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 1.283391405342625e-05,
+      "loss": 0.0,
+      "step": 10100
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 1.2795199380565235e-05,
+      "loss": 0.0001,
+      "step": 10110
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 1.275648470770422e-05,
+      "loss": 0.0,
+      "step": 10120
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 1.2717770034843207e-05,
+      "loss": 0.0,
+      "step": 10130
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 1.2679055361982192e-05,
+      "loss": 0.0,
+      "step": 10140
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 1.2640340689121177e-05,
+      "loss": 0.0001,
+      "step": 10150
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 1.2601626016260162e-05,
+      "loss": 0.0001,
+      "step": 10160
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 1.256291134339915e-05,
+      "loss": 0.0001,
+      "step": 10170
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 1.2524196670538136e-05,
+      "loss": 0.0,
+      "step": 10180
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 1.248548199767712e-05,
+      "loss": 0.0,
+      "step": 10190
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 1.2446767324816106e-05,
+      "loss": 0.0,
+      "step": 10200
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 1.2408052651955092e-05,
+      "loss": 0.0,
+      "step": 10210
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 1.2369337979094078e-05,
+      "loss": 0.0,
+      "step": 10220
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 1.2330623306233063e-05,
+      "loss": 0.0,
+      "step": 10230
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 1.2291908633372048e-05,
+      "loss": 0.0,
+      "step": 10240
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 1.2253193960511034e-05,
+      "loss": 0.0,
+      "step": 10250
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 1.221447928765002e-05,
+      "loss": 0.002,
+      "step": 10260
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 1.2175764614789005e-05,
+      "loss": 0.0,
+      "step": 10270
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 1.2137049941927991e-05,
+      "loss": 0.0001,
+      "step": 10280
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 1.2098335269066976e-05,
+      "loss": 0.0,
+      "step": 10290
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 1.2059620596205963e-05,
+      "loss": 0.0,
+      "step": 10300
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 1.2020905923344948e-05,
+      "loss": 0.0001,
+      "step": 10310
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 1.1982191250483935e-05,
+      "loss": 0.0,
+      "step": 10320
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 1.194347657762292e-05,
+      "loss": 0.0,
+      "step": 10330
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 1.1904761904761905e-05,
+      "loss": 0.0,
+      "step": 10340
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 1.1866047231900892e-05,
+      "loss": 0.0,
+      "step": 10350
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 1.1827332559039877e-05,
+      "loss": 0.0,
+      "step": 10360
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 1.1788617886178862e-05,
+      "loss": 0.0288,
+      "step": 10370
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 1.1749903213317847e-05,
+      "loss": 0.0004,
+      "step": 10380
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 1.1711188540456834e-05,
+      "loss": 0.0365,
+      "step": 10390
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 1.1672473867595819e-05,
+      "loss": 0.0,
+      "step": 10400
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 1.1633759194734806e-05,
+      "loss": 0.0,
+      "step": 10410
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 1.159504452187379e-05,
+      "loss": 0.0,
+      "step": 10420
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 1.1556329849012776e-05,
+      "loss": 0.0,
+      "step": 10430
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 1.1517615176151763e-05,
+      "loss": 0.0001,
+      "step": 10440
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 1.1478900503290748e-05,
+      "loss": 0.0,
+      "step": 10450
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 1.1440185830429734e-05,
+      "loss": 0.0,
+      "step": 10460
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 1.140147115756872e-05,
+      "loss": 0.0,
+      "step": 10470
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 1.1362756484707704e-05,
+      "loss": 0.0,
+      "step": 10480
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 1.132404181184669e-05,
+      "loss": 0.0,
+      "step": 10490
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 1.1285327138985675e-05,
+      "loss": 0.0004,
+      "step": 10500
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 1.1246612466124661e-05,
+      "loss": 0.0,
+      "step": 10510
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 1.1207897793263646e-05,
+      "loss": 0.0,
+      "step": 10520
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 1.1169183120402633e-05,
+      "loss": 0.0602,
+      "step": 10530
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 1.1130468447541618e-05,
+      "loss": 0.0,
+      "step": 10540
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 1.1091753774680605e-05,
+      "loss": 0.0,
+      "step": 10550
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 1.105303910181959e-05,
+      "loss": 0.0,
+      "step": 10560
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 1.1014324428958577e-05,
+      "loss": 0.0012,
+      "step": 10570
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 1.0975609756097562e-05,
+      "loss": 0.0001,
+      "step": 10580
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 1.0936895083236547e-05,
+      "loss": 0.0,
+      "step": 10590
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 1.0898180410375534e-05,
+      "loss": 0.0,
+      "step": 10600
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 1.0859465737514519e-05,
+      "loss": 0.0,
+      "step": 10610
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 1.0820751064653504e-05,
+      "loss": 0.0,
+      "step": 10620
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 1.0782036391792489e-05,
+      "loss": 0.0,
+      "step": 10630
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 1.0743321718931476e-05,
+      "loss": 0.0,
+      "step": 10640
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 1.070460704607046e-05,
+      "loss": 0.0,
+      "step": 10650
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 1.0665892373209446e-05,
+      "loss": 0.0,
+      "step": 10660
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 1.0627177700348433e-05,
+      "loss": 0.0,
+      "step": 10670
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 1.0588463027487418e-05,
+      "loss": 0.0,
+      "step": 10680
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 1.0549748354626404e-05,
+      "loss": 0.0,
+      "step": 10690
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 1.051103368176539e-05,
+      "loss": 0.0,
+      "step": 10700
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 1.0472319008904376e-05,
+      "loss": 0.0,
+      "step": 10710
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 1.0433604336043361e-05,
+      "loss": 0.0,
+      "step": 10720
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 1.0394889663182348e-05,
+      "loss": 0.0,
+      "step": 10730
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.9948979591836735,
+      "eval_f1": 0.9947143581203389,
+      "eval_loss": 0.03681769222021103,
+      "eval_runtime": 143.4337,
+      "eval_samples_per_second": 99.753,
+      "eval_steps_per_second": 1.562,
+      "step": 10732
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 13415,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "total_flos": 2.2742532870856704e+16,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}