diff --git "a/results/trainer_state.json" "b/results/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/results/trainer_state.json"
@@ -0,0 +1,4723 @@
+{
+  "best_metric": 0.97535,
+  "best_model_checkpoint": "models/camembert-allocine/checkpoint-5500",
+  "epoch": 3.0,
+  "global_step": 7500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 0.6935,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.6941,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.6928,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3e-06,
+      "loss": 0.6896,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.6895,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-06,
+      "loss": 0.6849,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6e-06,
+      "loss": 0.6782,
+      "step": 60
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 0.6604,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.6061,
+      "step": 80
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9e-06,
+      "loss": 0.499,
+      "step": 90
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1e-05,
+      "loss": 0.401,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.3412,
+      "step": 110
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.2e-05,
+      "loss": 0.315,
+      "step": 120
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.2725,
+      "step": 130
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 0.224,
+      "step": 140
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.5e-05,
+      "loss": 0.2441,
+      "step": 150
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.2135,
+      "step": 160
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 0.1807,
+      "step": 170
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.8e-05,
+      "loss": 0.1957,
+      "step": 180
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9e-05,
+      "loss": 0.1938,
+      "step": 190
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2e-05,
+      "loss": 0.1525,
+      "step": 200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.1e-05,
+      "loss": 0.2019,
+      "step": 210
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 0.1541,
+      "step": 220
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 0.1988,
+      "step": 230
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.4e-05,
+      "loss": 0.1404,
+      "step": 240
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.5e-05,
+      "loss": 0.1522,
+      "step": 250
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 0.2005,
+      "step": 260
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 0.1869,
+      "step": 270
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 0.1432,
+      "step": 280
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.9e-05,
+      "loss": 0.1347,
+      "step": 290
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 3e-05,
+      "loss": 0.1426,
+      "step": 300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 3.1e-05,
+      "loss": 0.1813,
+      "step": 310
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.1467,
+      "step": 320
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 3.3e-05,
+      "loss": 0.1573,
+      "step": 330
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 0.1327,
+      "step": 340
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 3.5e-05,
+      "loss": 0.1489,
+      "step": 350
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 3.6e-05,
+      "loss": 0.1574,
+      "step": 360
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 3.7e-05,
+      "loss": 0.1114,
+      "step": 370
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 3.8e-05,
+      "loss": 0.185,
+      "step": 380
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 0.1243,
+      "step": 390
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4e-05,
+      "loss": 0.1322,
+      "step": 400
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1e-05,
+      "loss": 0.0885,
+      "step": 410
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.2e-05,
+      "loss": 0.1102,
+      "step": 420
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.3e-05,
+      "loss": 0.1712,
+      "step": 430
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.1227,
+      "step": 440
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.5e-05,
+      "loss": 0.1395,
+      "step": 450
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.1377,
+      "step": 460
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.7e-05,
+      "loss": 0.1266,
+      "step": 470
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.8e-05,
+      "loss": 0.1347,
+      "step": 480
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.9e-05,
+      "loss": 0.1257,
+      "step": 490
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.1276,
+      "step": 500
+    },
+    {
+      "epoch": 0.2,
+      "eval_accuracy": 0.9623,
+      "eval_f1": 0.9621637896427138,
+      "eval_loss": 0.1187000721693039,
+      "eval_precision": 0.9462100276352151,
+      "eval_recall": 0.9786647611269906,
+      "eval_runtime": 143.7296,
+      "eval_samples_per_second": 139.15,
+      "eval_steps_per_second": 8.697,
+      "step": 500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.992857142857143e-05,
+      "loss": 0.1576,
+      "step": 510
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.985714285714286e-05,
+      "loss": 0.1211,
+      "step": 520
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.978571428571429e-05,
+      "loss": 0.1075,
+      "step": 530
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.971428571428572e-05,
+      "loss": 0.0756,
+      "step": 540
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.964285714285715e-05,
+      "loss": 0.11,
+      "step": 550
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.957142857142857e-05,
+      "loss": 0.0851,
+      "step": 560
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.9500000000000004e-05,
+      "loss": 0.1322,
+      "step": 570
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.942857142857143e-05,
+      "loss": 0.1639,
+      "step": 580
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.935714285714286e-05,
+      "loss": 0.1211,
+      "step": 590
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.928571428571429e-05,
+      "loss": 0.1551,
+      "step": 600
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.921428571428572e-05,
+      "loss": 0.1461,
+      "step": 610
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.9142857142857144e-05,
+      "loss": 0.1033,
+      "step": 620
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.9071428571428574e-05,
+      "loss": 0.1334,
+      "step": 630
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.9e-05,
+      "loss": 0.1236,
+      "step": 640
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.892857142857143e-05,
+      "loss": 0.122,
+      "step": 650
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.885714285714286e-05,
+      "loss": 0.1161,
+      "step": 660
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.878571428571429e-05,
+      "loss": 0.111,
+      "step": 670
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.8714285714285714e-05,
+      "loss": 0.0982,
+      "step": 680
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.8642857142857145e-05,
+      "loss": 0.1537,
+      "step": 690
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.8571428571428576e-05,
+      "loss": 0.0897,
+      "step": 700
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.85e-05,
+      "loss": 0.0882,
+      "step": 710
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.842857142857143e-05,
+      "loss": 0.1019,
+      "step": 720
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.835714285714286e-05,
+      "loss": 0.121,
+      "step": 730
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.828571428571429e-05,
+      "loss": 0.141,
+      "step": 740
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.8214285714285716e-05,
+      "loss": 0.1218,
+      "step": 750
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.8142857142857147e-05,
+      "loss": 0.1323,
+      "step": 760
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.807142857142857e-05,
+      "loss": 0.1311,
+      "step": 770
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.8e-05,
+      "loss": 0.1214,
+      "step": 780
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.7928571428571425e-05,
+      "loss": 0.1176,
+      "step": 790
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.785714285714286e-05,
+      "loss": 0.1178,
+      "step": 800
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.7785714285714287e-05,
+      "loss": 0.1299,
+      "step": 810
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.771428571428572e-05,
+      "loss": 0.1121,
+      "step": 820
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.764285714285715e-05,
+      "loss": 0.0961,
+      "step": 830
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.757142857142857e-05,
+      "loss": 0.0881,
+      "step": 840
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.75e-05,
+      "loss": 0.1506,
+      "step": 850
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.742857142857143e-05,
+      "loss": 0.0964,
+      "step": 860
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.7357142857142864e-05,
+      "loss": 0.1096,
+      "step": 870
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.728571428571429e-05,
+      "loss": 0.1127,
+      "step": 880
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.721428571428572e-05,
+      "loss": 0.1004,
+      "step": 890
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.714285714285714e-05,
+      "loss": 0.1168,
+      "step": 900
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.707142857142857e-05,
+      "loss": 0.0914,
+      "step": 910
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.7e-05,
+      "loss": 0.0927,
+      "step": 920
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.6928571428571435e-05,
+      "loss": 0.1838,
+      "step": 930
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.685714285714286e-05,
+      "loss": 0.1216,
+      "step": 940
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.678571428571429e-05,
+      "loss": 0.0978,
+      "step": 950
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.671428571428571e-05,
+      "loss": 0.1531,
+      "step": 960
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.6642857142857144e-05,
+      "loss": 0.1033,
+      "step": 970
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.6571428571428575e-05,
+      "loss": 0.1049,
+      "step": 980
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.6500000000000005e-05,
+      "loss": 0.1052,
+      "step": 990
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.642857142857143e-05,
+      "loss": 0.1013,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4,
+      "eval_accuracy": 0.9683,
+      "eval_f1": 0.9674738354196594,
+      "eval_loss": 0.09167405217885971,
+      "eval_precision": 0.9724628712871287,
+      "eval_recall": 0.9625357288689261,
+      "eval_runtime": 143.8255,
+      "eval_samples_per_second": 139.057,
+      "eval_steps_per_second": 8.691,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.635714285714286e-05,
+      "loss": 0.12,
+      "step": 1010
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.628571428571429e-05,
+      "loss": 0.1522,
+      "step": 1020
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.6214285714285715e-05,
+      "loss": 0.1069,
+      "step": 1030
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.6142857142857145e-05,
+      "loss": 0.1176,
+      "step": 1040
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.607142857142857e-05,
+      "loss": 0.1285,
+      "step": 1050
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.1146,
+      "step": 1060
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.592857142857143e-05,
+      "loss": 0.0579,
+      "step": 1070
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.585714285714286e-05,
+      "loss": 0.1004,
+      "step": 1080
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.5785714285714285e-05,
+      "loss": 0.1088,
+      "step": 1090
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.5714285714285716e-05,
+      "loss": 0.0889,
+      "step": 1100
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.564285714285714e-05,
+      "loss": 0.0992,
+      "step": 1110
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.557142857142858e-05,
+      "loss": 0.1041,
+      "step": 1120
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.55e-05,
+      "loss": 0.1211,
+      "step": 1130
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.542857142857143e-05,
+      "loss": 0.1122,
+      "step": 1140
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.5357142857142856e-05,
+      "loss": 0.0959,
+      "step": 1150
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.528571428571429e-05,
+      "loss": 0.0772,
+      "step": 1160
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.521428571428572e-05,
+      "loss": 0.1671,
+      "step": 1170
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.514285714285714e-05,
+      "loss": 0.1093,
+      "step": 1180
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.507142857142858e-05,
+      "loss": 0.1037,
+      "step": 1190
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.5e-05,
+      "loss": 0.1003,
+      "step": 1200
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.4928571428571434e-05,
+      "loss": 0.1099,
+      "step": 1210
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.485714285714286e-05,
+      "loss": 0.081,
+      "step": 1220
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.478571428571429e-05,
+      "loss": 0.1598,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.471428571428571e-05,
+      "loss": 0.0956,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.464285714285715e-05,
+      "loss": 0.094,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.4571428571428574e-05,
+      "loss": 0.1015,
+      "step": 1260
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.4500000000000004e-05,
+      "loss": 0.1213,
+      "step": 1270
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.442857142857143e-05,
+      "loss": 0.0803,
+      "step": 1280
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.435714285714286e-05,
+      "loss": 0.1084,
+      "step": 1290
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.428571428571428e-05,
+      "loss": 0.0804,
+      "step": 1300
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.4214285714285714e-05,
+      "loss": 0.0706,
+      "step": 1310
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.4142857142857144e-05,
+      "loss": 0.1034,
+      "step": 1320
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.4071428571428575e-05,
+      "loss": 0.0871,
+      "step": 1330
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.0836,
+      "step": 1340
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.392857142857143e-05,
+      "loss": 0.0856,
+      "step": 1350
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.385714285714286e-05,
+      "loss": 0.0981,
+      "step": 1360
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.3785714285714284e-05,
+      "loss": 0.1054,
+      "step": 1370
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.371428571428572e-05,
+      "loss": 0.0895,
+      "step": 1380
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.3642857142857146e-05,
+      "loss": 0.0943,
+      "step": 1390
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.3571428571428576e-05,
+      "loss": 0.0527,
+      "step": 1400
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.35e-05,
+      "loss": 0.0798,
+      "step": 1410
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.342857142857143e-05,
+      "loss": 0.0954,
+      "step": 1420
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.3357142857142855e-05,
+      "loss": 0.1228,
+      "step": 1430
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.328571428571429e-05,
+      "loss": 0.0912,
+      "step": 1440
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.3214285714285716e-05,
+      "loss": 0.081,
+      "step": 1450
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.314285714285715e-05,
+      "loss": 0.1376,
+      "step": 1460
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.307142857142857e-05,
+      "loss": 0.1039,
+      "step": 1470
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.3e-05,
+      "loss": 0.121,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.292857142857143e-05,
+      "loss": 0.0621,
+      "step": 1490
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.1254,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6,
+      "eval_accuracy": 0.9701,
+      "eval_f1": 0.9697979797979798,
+      "eval_loss": 0.08888383954763412,
+      "eval_precision": 0.9597161135545782,
+      "eval_recall": 0.9800939158840343,
+      "eval_runtime": 143.8997,
+      "eval_samples_per_second": 138.986,
+      "eval_steps_per_second": 8.687,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.278571428571429e-05,
+      "loss": 0.1127,
+      "step": 1510
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.271428571428572e-05,
+      "loss": 0.0865,
+      "step": 1520
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.264285714285715e-05,
+      "loss": 0.1063,
+      "step": 1530
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.257142857142857e-05,
+      "loss": 0.0945,
+      "step": 1540
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.25e-05,
+      "loss": 0.1045,
+      "step": 1550
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.242857142857143e-05,
+      "loss": 0.0997,
+      "step": 1560
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.2357142857142864e-05,
+      "loss": 0.1338,
+      "step": 1570
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.228571428571429e-05,
+      "loss": 0.125,
+      "step": 1580
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.221428571428572e-05,
+      "loss": 0.11,
+      "step": 1590
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.214285714285714e-05,
+      "loss": 0.0997,
+      "step": 1600
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.2071428571428574e-05,
+      "loss": 0.0786,
+      "step": 1610
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.2e-05,
+      "loss": 0.1165,
+      "step": 1620
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.192857142857143e-05,
+      "loss": 0.0917,
+      "step": 1630
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.185714285714286e-05,
+      "loss": 0.1007,
+      "step": 1640
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.178571428571429e-05,
+      "loss": 0.0941,
+      "step": 1650
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.1714285714285714e-05,
+      "loss": 0.1163,
+      "step": 1660
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.1642857142857144e-05,
+      "loss": 0.1007,
+      "step": 1670
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.1571428571428575e-05,
+      "loss": 0.1057,
+      "step": 1680
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.15e-05,
+      "loss": 0.0845,
+      "step": 1690
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.1428571428571437e-05,
+      "loss": 0.1117,
+      "step": 1700
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.135714285714286e-05,
+      "loss": 0.1128,
+      "step": 1710
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.128571428571429e-05,
+      "loss": 0.1154,
+      "step": 1720
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.1214285714285715e-05,
+      "loss": 0.1067,
+      "step": 1730
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.1142857142857146e-05,
+      "loss": 0.0921,
+      "step": 1740
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.107142857142857e-05,
+      "loss": 0.0756,
+      "step": 1750
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.1e-05,
+      "loss": 0.0892,
+      "step": 1760
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.092857142857143e-05,
+      "loss": 0.1378,
+      "step": 1770
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.085714285714286e-05,
+      "loss": 0.085,
+      "step": 1780
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.0785714285714286e-05,
+      "loss": 0.094,
+      "step": 1790
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.0714285714285717e-05,
+      "loss": 0.0891,
+      "step": 1800
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.064285714285714e-05,
+      "loss": 0.1294,
+      "step": 1810
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.057142857142857e-05,
+      "loss": 0.1448,
+      "step": 1820
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.05e-05,
+      "loss": 0.1177,
+      "step": 1830
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.042857142857143e-05,
+      "loss": 0.0911,
+      "step": 1840
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.035714285714286e-05,
+      "loss": 0.1022,
+      "step": 1850
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.028571428571429e-05,
+      "loss": 0.1287,
+      "step": 1860
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.021428571428572e-05,
+      "loss": 0.107,
+      "step": 1870
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.014285714285714e-05,
+      "loss": 0.0787,
+      "step": 1880
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.007142857142857e-05,
+      "loss": 0.1115,
+      "step": 1890
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4e-05,
+      "loss": 0.0864,
+      "step": 1900
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.9928571428571434e-05,
+      "loss": 0.1156,
+      "step": 1910
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.985714285714286e-05,
+      "loss": 0.0898,
+      "step": 1920
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.978571428571429e-05,
+      "loss": 0.0928,
+      "step": 1930
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.971428571428571e-05,
+      "loss": 0.1011,
+      "step": 1940
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.964285714285714e-05,
+      "loss": 0.0862,
+      "step": 1950
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.9571428571428574e-05,
+      "loss": 0.0858,
+      "step": 1960
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.9500000000000005e-05,
+      "loss": 0.0593,
+      "step": 1970
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.942857142857143e-05,
+      "loss": 0.156,
+      "step": 1980
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.935714285714286e-05,
+      "loss": 0.0998,
+      "step": 1990
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.928571428571429e-05,
+      "loss": 0.1004,
+      "step": 2000
+    },
+    {
+      "epoch": 0.8,
+      "eval_accuracy": 0.97155,
+      "eval_f1": 0.9709025824597288,
+      "eval_loss": 0.07919494062662125,
+      "eval_precision": 0.9727431089250947,
+      "eval_recall": 0.9690690077582687,
+      "eval_runtime": 144.0168,
+      "eval_samples_per_second": 138.873,
+      "eval_steps_per_second": 8.68,
+      "step": 2000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.9214285714285714e-05,
+      "loss": 0.11,
+      "step": 2010
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.9142857142857145e-05,
+      "loss": 0.1417,
+      "step": 2020
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.9071428571428575e-05,
+      "loss": 0.09,
+      "step": 2030
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 0.0834,
+      "step": 2040
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.892857142857143e-05,
+      "loss": 0.0806,
+      "step": 2050
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.885714285714286e-05,
+      "loss": 0.1456,
+      "step": 2060
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.8785714285714285e-05,
+      "loss": 0.0933,
+      "step": 2070
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.8714285714285715e-05,
+      "loss": 0.074,
+      "step": 2080
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.8642857142857146e-05,
+      "loss": 0.0742,
+      "step": 2090
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.857142857142858e-05,
+      "loss": 0.1136,
+      "step": 2100
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.85e-05,
+      "loss": 0.0839,
+      "step": 2110
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.842857142857143e-05,
+      "loss": 0.0984,
+      "step": 2120
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.8357142857142855e-05,
+      "loss": 0.074,
+      "step": 2130
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.8285714285714286e-05,
+      "loss": 0.0982,
+      "step": 2140
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.821428571428572e-05,
+      "loss": 0.0817,
+      "step": 2150
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.814285714285715e-05,
+      "loss": 0.0818,
+      "step": 2160
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.807142857142857e-05,
+      "loss": 0.0855,
+      "step": 2170
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.8e-05,
+      "loss": 0.0753,
+      "step": 2180
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.792857142857143e-05,
+      "loss": 0.0952,
+      "step": 2190
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.785714285714286e-05,
+      "loss": 0.1329,
+      "step": 2200
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.778571428571429e-05,
+      "loss": 0.0793,
+      "step": 2210
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.771428571428572e-05,
+      "loss": 0.0841,
+      "step": 2220
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.764285714285715e-05,
+      "loss": 0.1033,
+      "step": 2230
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.757142857142857e-05,
+      "loss": 0.1023,
+      "step": 2240
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.1064,
+      "step": 2250
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.742857142857143e-05,
+      "loss": 0.0712,
+      "step": 2260
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.735714285714286e-05,
+      "loss": 0.0812,
+      "step": 2270
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.728571428571428e-05,
+      "loss": 0.0941,
+      "step": 2280
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.721428571428572e-05,
+      "loss": 0.0872,
+      "step": 2290
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.7142857142857143e-05,
+      "loss": 0.0823,
+      "step": 2300
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.7071428571428574e-05,
+      "loss": 0.0981,
+      "step": 2310
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.7e-05,
+      "loss": 0.0975,
+      "step": 2320
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.692857142857143e-05,
+      "loss": 0.0938,
+      "step": 2330
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.685714285714286e-05,
+      "loss": 0.1076,
+      "step": 2340
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.678571428571429e-05,
+      "loss": 0.1094,
+      "step": 2350
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.671428571428572e-05,
+      "loss": 0.0873,
+      "step": 2360
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.6642857142857145e-05,
+      "loss": 0.096,
+      "step": 2370
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.6571428571428576e-05,
+      "loss": 0.0948,
+      "step": 2380
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.65e-05,
+      "loss": 0.0732,
+      "step": 2390
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.642857142857143e-05,
+      "loss": 0.0766,
+      "step": 2400
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.6357142857142854e-05,
+      "loss": 0.1156,
+      "step": 2410
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.628571428571429e-05,
+      "loss": 0.0823,
+      "step": 2420
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.6214285714285716e-05,
+      "loss": 0.0863,
+      "step": 2430
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.6142857142857146e-05,
+      "loss": 0.0995,
+      "step": 2440
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.607142857142857e-05,
+      "loss": 0.0843,
+      "step": 2450
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.6e-05,
+      "loss": 0.1004,
+      "step": 2460
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.5928571428571425e-05,
+      "loss": 0.0807,
+      "step": 2470
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.585714285714286e-05,
+      "loss": 0.1013,
+      "step": 2480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.5785714285714286e-05,
+      "loss": 0.0728,
+      "step": 2490
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.571428571428572e-05,
+      "loss": 0.1149,
+      "step": 2500
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.9727,
+      "eval_f1": 0.9722758200467146,
+      "eval_loss": 0.07623879611492157,
+      "eval_precision": 0.9672661143665386,
+      "eval_recall": 0.9773376888525929,
+      "eval_runtime": 144.3048,
+      "eval_samples_per_second": 138.595,
+      "eval_steps_per_second": 8.662,
+      "step": 2500
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.564285714285715e-05,
+      "loss": 0.0565,
+      "step": 2510
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.557142857142857e-05,
+      "loss": 0.0491,
+      "step": 2520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.55e-05,
+      "loss": 0.0437,
+      "step": 2530
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5428571428571426e-05,
+      "loss": 0.0413,
+      "step": 2540
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5357142857142864e-05,
+      "loss": 0.0885,
+      "step": 2550
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.528571428571429e-05,
+      "loss": 0.0708,
+      "step": 2560
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.521428571428572e-05,
+      "loss": 0.0559,
+      "step": 2570
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.514285714285714e-05,
+      "loss": 0.0744,
+      "step": 2580
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.507142857142857e-05,
+      "loss": 0.1189,
+      "step": 2590
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.5e-05,
+      "loss": 0.0642,
+      "step": 2600
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.4928571428571434e-05,
+      "loss": 0.0489,
+      "step": 2610
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.485714285714286e-05,
+      "loss": 0.0724,
+      "step": 2620
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.478571428571429e-05,
+      "loss": 0.0659,
+      "step": 2630
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.471428571428571e-05,
+      "loss": 0.0704,
+      "step": 2640
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.4642857142857144e-05,
+      "loss": 0.067,
+      "step": 2650
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.4571428571428574e-05,
+      "loss": 0.0769,
+      "step": 2660
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.45e-05,
+      "loss": 0.0436,
+      "step": 2670
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.442857142857143e-05,
+      "loss": 0.037,
+      "step": 2680
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.435714285714286e-05,
+      "loss": 0.0582,
+      "step": 2690
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.428571428571429e-05,
+      "loss": 0.0721,
+      "step": 2700
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.4214285714285714e-05,
+      "loss": 0.0517,
+      "step": 2710
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.4142857142857145e-05,
+      "loss": 0.0628,
+      "step": 2720
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.407142857142857e-05,
+      "loss": 0.0336,
+      "step": 2730
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 0.0706,
+      "step": 2740
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.392857142857143e-05,
+      "loss": 0.0592,
+      "step": 2750
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.385714285714286e-05,
+      "loss": 0.0371,
+      "step": 2760
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.3785714285714285e-05,
+      "loss": 0.0399,
+      "step": 2770
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.3714285714285716e-05,
+      "loss": 0.0688,
+      "step": 2780
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.364285714285714e-05,
+      "loss": 0.0441,
+      "step": 2790
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.357142857142857e-05,
+      "loss": 0.0748,
+      "step": 2800
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.35e-05,
+      "loss": 0.0674,
+      "step": 2810
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.342857142857143e-05,
+      "loss": 0.048,
+      "step": 2820
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.3357142857142856e-05,
+      "loss": 0.0591,
+      "step": 2830
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.3285714285714286e-05,
+      "loss": 0.0531,
+      "step": 2840
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.321428571428572e-05,
+      "loss": 0.0686,
+      "step": 2850
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.314285714285714e-05,
+      "loss": 0.0838,
+      "step": 2860
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.307142857142858e-05,
+      "loss": 0.0522,
+      "step": 2870
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.3e-05,
+      "loss": 0.0731,
+      "step": 2880
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.292857142857143e-05,
+      "loss": 0.083,
+      "step": 2890
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.285714285714286e-05,
+      "loss": 0.0674,
+      "step": 2900
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.278571428571429e-05,
+      "loss": 0.0663,
+      "step": 2910
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.271428571428571e-05,
+      "loss": 0.0508,
+      "step": 2920
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.264285714285714e-05,
+      "loss": 0.0453,
+      "step": 2930
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.257142857142857e-05,
+      "loss": 0.0885,
+      "step": 2940
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.2500000000000004e-05,
+      "loss": 0.0624,
+      "step": 2950
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.242857142857143e-05,
+      "loss": 0.0683,
+      "step": 2960
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.235714285714286e-05,
+      "loss": 0.0414,
+      "step": 2970
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.228571428571428e-05,
+      "loss": 0.0502,
+      "step": 2980
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.221428571428571e-05,
+      "loss": 0.0723,
+      "step": 2990
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.2142857142857144e-05,
+      "loss": 0.0574,
+      "step": 3000
+    },
+    {
+      "epoch": 1.2,
+      "eval_accuracy": 0.9733,
+      "eval_f1": 0.9728851426830506,
+      "eval_loss": 0.08491361886262894,
+      "eval_precision": 0.967872297433825,
+      "eval_recall": 0.9779501837484688,
+      "eval_runtime": 143.6357,
+      "eval_samples_per_second": 139.241,
+      "eval_steps_per_second": 8.703,
+      "step": 3000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.2071428571428575e-05,
+      "loss": 0.0557,
+      "step": 3010
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.0725,
+      "step": 3020
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.192857142857143e-05,
+      "loss": 0.0385,
+      "step": 3030
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.185714285714286e-05,
+      "loss": 0.0617,
+      "step": 3040
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.1785714285714284e-05,
+      "loss": 0.0701,
+      "step": 3050
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.1714285714285715e-05,
+      "loss": 0.0396,
+      "step": 3060
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.1642857142857145e-05,
+      "loss": 0.0699,
+      "step": 3070
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.1571428571428576e-05,
+      "loss": 0.0897,
+      "step": 3080
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.15e-05,
+      "loss": 0.0548,
+      "step": 3090
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.142857142857143e-05,
+      "loss": 0.0585,
+      "step": 3100
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.1357142857142855e-05,
+      "loss": 0.0447,
+      "step": 3110
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.1285714285714285e-05,
+      "loss": 0.0511,
+      "step": 3120
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.1214285714285716e-05,
+      "loss": 0.0925,
+      "step": 3130
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.114285714285715e-05,
+      "loss": 0.0486,
+      "step": 3140
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.107142857142857e-05,
+      "loss": 0.0983,
+      "step": 3150
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.1e-05,
+      "loss": 0.0499,
+      "step": 3160
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.092857142857143e-05,
+      "loss": 0.0446,
+      "step": 3170
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.0857142857142856e-05,
+      "loss": 0.0739,
+      "step": 3180
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.078571428571429e-05,
+      "loss": 0.0494,
+      "step": 3190
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.071428571428572e-05,
+      "loss": 0.0454,
+      "step": 3200
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.064285714285715e-05,
+      "loss": 0.0686,
+      "step": 3210
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.057142857142857e-05,
+      "loss": 0.053,
+      "step": 3220
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.05e-05,
+      "loss": 0.0569,
+      "step": 3230
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.042857142857143e-05,
+      "loss": 0.07,
+      "step": 3240
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.0357142857142857e-05,
+      "loss": 0.0395,
+      "step": 3250
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.0285714285714288e-05,
+      "loss": 0.0537,
+      "step": 3260
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 3.021428571428572e-05,
+      "loss": 0.0792,
+      "step": 3270
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 3.0142857142857146e-05,
+      "loss": 0.0661,
+      "step": 3280
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.0071428571428573e-05,
+      "loss": 0.0864,
+      "step": 3290
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3e-05,
+      "loss": 0.0585,
+      "step": 3300
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.9928571428571428e-05,
+      "loss": 0.0613,
+      "step": 3310
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.9857142857142862e-05,
+      "loss": 0.0802,
+      "step": 3320
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.978571428571429e-05,
+      "loss": 0.0416,
+      "step": 3330
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.9714285714285717e-05,
+      "loss": 0.0431,
+      "step": 3340
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.9642857142857144e-05,
+      "loss": 0.0397,
+      "step": 3350
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.957142857142857e-05,
+      "loss": 0.0487,
+      "step": 3360
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.95e-05,
+      "loss": 0.0181,
+      "step": 3370
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.9428571428571426e-05,
+      "loss": 0.062,
+      "step": 3380
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.935714285714286e-05,
+      "loss": 0.0544,
+      "step": 3390
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.9285714285714288e-05,
+      "loss": 0.0445,
+      "step": 3400
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.9214285714285715e-05,
+      "loss": 0.0718,
+      "step": 3410
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.9142857142857146e-05,
+      "loss": 0.0539,
+      "step": 3420
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.9071428571428573e-05,
+      "loss": 0.0693,
+      "step": 3430
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.9e-05,
+      "loss": 0.0527,
+      "step": 3440
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.8928571428571434e-05,
+      "loss": 0.0507,
+      "step": 3450
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.885714285714286e-05,
+      "loss": 0.051,
+      "step": 3460
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.878571428571429e-05,
+      "loss": 0.0748,
+      "step": 3470
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.8714285714285716e-05,
+      "loss": 0.0761,
+      "step": 3480
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.8642857142857144e-05,
+      "loss": 0.0482,
+      "step": 3490
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 0.0394,
+      "step": 3500
+    },
+    {
+      "epoch": 1.4,
+      "eval_accuracy": 0.97175,
+      "eval_f1": 0.9715236127211331,
+      "eval_loss": 0.10262948274612427,
+      "eval_precision": 0.9594823295171727,
+      "eval_recall": 0.9838709677419355,
+      "eval_runtime": 143.9773,
+      "eval_samples_per_second": 138.911,
+      "eval_steps_per_second": 8.682,
+      "step": 3500
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.8499999999999998e-05,
+      "loss": 0.0865,
+      "step": 3510
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.8428571428571432e-05,
+      "loss": 0.0864,
+      "step": 3520
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.835714285714286e-05,
+      "loss": 0.0397,
+      "step": 3530
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.8285714285714287e-05,
+      "loss": 0.0672,
+      "step": 3540
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.8214285714285714e-05,
+      "loss": 0.0818,
+      "step": 3550
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.814285714285714e-05,
+      "loss": 0.0534,
+      "step": 3560
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.8071428571428572e-05,
+      "loss": 0.0692,
+      "step": 3570
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 0.0672,
+      "step": 3580
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.792857142857143e-05,
+      "loss": 0.0511,
+      "step": 3590
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.785714285714286e-05,
+      "loss": 0.0561,
+      "step": 3600
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.778571428571429e-05,
+      "loss": 0.0598,
+      "step": 3610
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.7714285714285716e-05,
+      "loss": 0.0845,
+      "step": 3620
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.7642857142857143e-05,
+      "loss": 0.0652,
+      "step": 3630
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.757142857142857e-05,
+      "loss": 0.0541,
+      "step": 3640
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.7500000000000004e-05,
+      "loss": 0.0589,
+      "step": 3650
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.742857142857143e-05,
+      "loss": 0.0451,
+      "step": 3660
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.735714285714286e-05,
+      "loss": 0.0766,
+      "step": 3670
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.7285714285714286e-05,
+      "loss": 0.0479,
+      "step": 3680
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.7214285714285714e-05,
+      "loss": 0.0691,
+      "step": 3690
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.714285714285714e-05,
+      "loss": 0.0606,
+      "step": 3700
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.7071428571428575e-05,
+      "loss": 0.0692,
+      "step": 3710
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 0.0451,
+      "step": 3720
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.692857142857143e-05,
+      "loss": 0.0628,
+      "step": 3730
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.6857142857142857e-05,
+      "loss": 0.0745,
+      "step": 3740
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.6785714285714288e-05,
+      "loss": 0.0616,
+      "step": 3750
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.6714285714285715e-05,
+      "loss": 0.0728,
+      "step": 3760
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.6642857142857142e-05,
+      "loss": 0.0512,
+      "step": 3770
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.6571428571428576e-05,
+      "loss": 0.0511,
+      "step": 3780
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.6500000000000004e-05,
+      "loss": 0.036,
+      "step": 3790
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.642857142857143e-05,
+      "loss": 0.0438,
+      "step": 3800
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.635714285714286e-05,
+      "loss": 0.0602,
+      "step": 3810
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.6285714285714286e-05,
+      "loss": 0.0585,
+      "step": 3820
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.6214285714285713e-05,
+      "loss": 0.0728,
+      "step": 3830
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.6142857142857147e-05,
+      "loss": 0.0711,
+      "step": 3840
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.6071428571428574e-05,
+      "loss": 0.0699,
+      "step": 3850
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 0.0606,
+      "step": 3860
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.592857142857143e-05,
+      "loss": 0.0543,
+      "step": 3870
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.5857142857142856e-05,
+      "loss": 0.0592,
+      "step": 3880
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.5785714285714284e-05,
+      "loss": 0.0604,
+      "step": 3890
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.5714285714285714e-05,
+      "loss": 0.0377,
+      "step": 3900
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.5642857142857145e-05,
+      "loss": 0.0789,
+      "step": 3910
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.5571428571428572e-05,
+      "loss": 0.0526,
+      "step": 3920
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.5500000000000003e-05,
+      "loss": 0.0564,
+      "step": 3930
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.542857142857143e-05,
+      "loss": 0.0671,
+      "step": 3940
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.5357142857142858e-05,
+      "loss": 0.0451,
+      "step": 3950
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.5285714285714285e-05,
+      "loss": 0.0567,
+      "step": 3960
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.521428571428572e-05,
+      "loss": 0.0727,
+      "step": 3970
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.5142857142857147e-05,
+      "loss": 0.0538,
+      "step": 3980
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.5071428571428574e-05,
+      "loss": 0.0394,
+      "step": 3990
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.5e-05,
+      "loss": 0.0401,
+      "step": 4000
+    },
+    {
+      "epoch": 1.6,
+      "eval_accuracy": 0.9698,
+      "eval_f1": 0.9697182392459642,
+      "eval_loss": 0.10654732584953308,
+      "eval_precision": 0.952807881773399,
+      "eval_recall": 0.9872396896692528,
+      "eval_runtime": 143.8494,
+      "eval_samples_per_second": 139.034,
+      "eval_steps_per_second": 8.69,
+      "step": 4000
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.492857142857143e-05,
+      "loss": 0.0522,
+      "step": 4010
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.485714285714286e-05,
+      "loss": 0.0955,
+      "step": 4020
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.4785714285714287e-05,
+      "loss": 0.0654,
+      "step": 4030
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.4714285714285714e-05,
+      "loss": 0.0525,
+      "step": 4040
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.4642857142857145e-05,
+      "loss": 0.0636,
+      "step": 4050
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.4571428571428572e-05,
+      "loss": 0.0665,
+      "step": 4060
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.45e-05,
+      "loss": 0.0909,
+      "step": 4070
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.442857142857143e-05,
+      "loss": 0.0334,
+      "step": 4080
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.4357142857142857e-05,
+      "loss": 0.0505,
+      "step": 4090
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.4285714285714288e-05,
+      "loss": 0.0477,
+      "step": 4100
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.4214285714285715e-05,
+      "loss": 0.0515,
+      "step": 4110
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.4142857142857146e-05,
+      "loss": 0.0435,
+      "step": 4120
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.4071428571428573e-05,
+      "loss": 0.0518,
+      "step": 4130
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.4e-05,
+      "loss": 0.0401,
+      "step": 4140
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.392857142857143e-05,
+      "loss": 0.0504,
+      "step": 4150
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.385714285714286e-05,
+      "loss": 0.0436,
+      "step": 4160
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.3785714285714286e-05,
+      "loss": 0.0539,
+      "step": 4170
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.3714285714285717e-05,
+      "loss": 0.0539,
+      "step": 4180
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.3642857142857144e-05,
+      "loss": 0.0535,
+      "step": 4190
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.357142857142857e-05,
+      "loss": 0.036,
+      "step": 4200
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.35e-05,
+      "loss": 0.0427,
+      "step": 4210
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.342857142857143e-05,
+      "loss": 0.0452,
+      "step": 4220
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.3357142857142857e-05,
+      "loss": 0.0801,
+      "step": 4230
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.3285714285714287e-05,
+      "loss": 0.0835,
+      "step": 4240
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.3214285714285715e-05,
+      "loss": 0.0533,
+      "step": 4250
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.3142857142857145e-05,
+      "loss": 0.0924,
+      "step": 4260
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.3071428571428573e-05,
+      "loss": 0.0522,
+      "step": 4270
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 0.0789,
+      "step": 4280
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.292857142857143e-05,
+      "loss": 0.0463,
+      "step": 4290
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.2857142857142858e-05,
+      "loss": 0.0468,
+      "step": 4300
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.278571428571429e-05,
+      "loss": 0.0781,
+      "step": 4310
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.2714285714285716e-05,
+      "loss": 0.0635,
+      "step": 4320
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.2642857142857143e-05,
+      "loss": 0.0933,
+      "step": 4330
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.257142857142857e-05,
+      "loss": 0.0461,
+      "step": 4340
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.25e-05,
+      "loss": 0.0739,
+      "step": 4350
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.242857142857143e-05,
+      "loss": 0.0615,
+      "step": 4360
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.2357142857142856e-05,
+      "loss": 0.05,
+      "step": 4370
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.2285714285714287e-05,
+      "loss": 0.0409,
+      "step": 4380
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.2214285714285714e-05,
+      "loss": 0.0484,
+      "step": 4390
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.214285714285714e-05,
+      "loss": 0.0763,
+      "step": 4400
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.2071428571428572e-05,
+      "loss": 0.0358,
+      "step": 4410
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 0.0648,
+      "step": 4420
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.192857142857143e-05,
+      "loss": 0.0657,
+      "step": 4430
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.185714285714286e-05,
+      "loss": 0.0663,
+      "step": 4440
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.1785714285714288e-05,
+      "loss": 0.0485,
+      "step": 4450
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.1714285714285715e-05,
+      "loss": 0.0797,
+      "step": 4460
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.1642857142857146e-05,
+      "loss": 0.0536,
+      "step": 4470
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.1571428571428574e-05,
+      "loss": 0.0765,
+      "step": 4480
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.15e-05,
+      "loss": 0.0694,
+      "step": 4490
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 0.0458,
+      "step": 4500
+    },
+    {
+      "epoch": 1.8,
+      "eval_accuracy": 0.9744,
+      "eval_f1": 0.9739334080032583,
+      "eval_loss": 0.08336396515369415,
+      "eval_precision": 0.9714604915701808,
+      "eval_recall": 0.9764189465087791,
+      "eval_runtime": 143.9739,
+      "eval_samples_per_second": 138.914,
+      "eval_steps_per_second": 8.682,
+      "step": 4500
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.135714285714286e-05,
+      "loss": 0.0456,
+      "step": 4510
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.1285714285714286e-05,
+      "loss": 0.0534,
+      "step": 4520
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.1214285714285713e-05,
+      "loss": 0.0756,
+      "step": 4530
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.1142857142857144e-05,
+      "loss": 0.0359,
+      "step": 4540
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.107142857142857e-05,
+      "loss": 0.0574,
+      "step": 4550
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.1e-05,
+      "loss": 0.0754,
+      "step": 4560
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.092857142857143e-05,
+      "loss": 0.0748,
+      "step": 4570
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.0857142857142857e-05,
+      "loss": 0.0521,
+      "step": 4580
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.0785714285714288e-05,
+      "loss": 0.057,
+      "step": 4590
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.0714285714285718e-05,
+      "loss": 0.0367,
+      "step": 4600
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.0642857142857146e-05,
+      "loss": 0.0442,
+      "step": 4610
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.0571428571428573e-05,
+      "loss": 0.0726,
+      "step": 4620
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.05e-05,
+      "loss": 0.0356,
+      "step": 4630
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.042857142857143e-05,
+      "loss": 0.0633,
+      "step": 4640
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.0357142857142858e-05,
+      "loss": 0.0509,
+      "step": 4650
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.0285714285714286e-05,
+      "loss": 0.0533,
+      "step": 4660
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2.0214285714285716e-05,
+      "loss": 0.0256,
+      "step": 4670
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2.0142857142857144e-05,
+      "loss": 0.0411,
+      "step": 4680
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2.007142857142857e-05,
+      "loss": 0.0755,
+      "step": 4690
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2e-05,
+      "loss": 0.0909,
+      "step": 4700
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.992857142857143e-05,
+      "loss": 0.0696,
+      "step": 4710
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.9857142857142856e-05,
+      "loss": 0.0548,
+      "step": 4720
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.9785714285714287e-05,
+      "loss": 0.0481,
+      "step": 4730
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.9714285714285714e-05,
+      "loss": 0.0629,
+      "step": 4740
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.9642857142857145e-05,
+      "loss": 0.0899,
+      "step": 4750
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.9571428571428572e-05,
+      "loss": 0.0659,
+      "step": 4760
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.9500000000000003e-05,
+      "loss": 0.058,
+      "step": 4770
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.942857142857143e-05,
+      "loss": 0.0493,
+      "step": 4780
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.9357142857142858e-05,
+      "loss": 0.1113,
+      "step": 4790
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.928571428571429e-05,
+      "loss": 0.0505,
+      "step": 4800
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.9214285714285716e-05,
+      "loss": 0.0239,
+      "step": 4810
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.9142857142857143e-05,
+      "loss": 0.0626,
+      "step": 4820
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.9071428571428574e-05,
+      "loss": 0.0559,
+      "step": 4830
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.9e-05,
+      "loss": 0.0814,
+      "step": 4840
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.892857142857143e-05,
+      "loss": 0.0688,
+      "step": 4850
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.885714285714286e-05,
+      "loss": 0.1068,
+      "step": 4860
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.8785714285714286e-05,
+      "loss": 0.0828,
+      "step": 4870
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.8714285714285714e-05,
+      "loss": 0.0782,
+      "step": 4880
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.864285714285714e-05,
+      "loss": 0.0679,
+      "step": 4890
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.8571428571428572e-05,
+      "loss": 0.0679,
+      "step": 4900
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.85e-05,
+      "loss": 0.0425,
+      "step": 4910
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.842857142857143e-05,
+      "loss": 0.0721,
+      "step": 4920
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.835714285714286e-05,
+      "loss": 0.0578,
+      "step": 4930
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.8285714285714288e-05,
+      "loss": 0.0463,
+      "step": 4940
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.8214285714285715e-05,
+      "loss": 0.075,
+      "step": 4950
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.8142857142857146e-05,
+      "loss": 0.0584,
+      "step": 4960
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.8071428571428573e-05,
+      "loss": 0.0427,
+      "step": 4970
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.8e-05,
+      "loss": 0.064,
+      "step": 4980
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.792857142857143e-05,
+      "loss": 0.0736,
+      "step": 4990
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.785714285714286e-05,
+      "loss": 0.0554,
+      "step": 5000
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.97195,
+      "eval_f1": 0.9717337632891621,
+      "eval_loss": 0.0872604101896286,
+      "eval_precision": 0.9594070241766989,
+      "eval_recall": 0.9843813801551654,
+      "eval_runtime": 143.7319,
+      "eval_samples_per_second": 139.148,
+      "eval_steps_per_second": 8.697,
+      "step": 5000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.7785714285714286e-05,
+      "loss": 0.0211,
+      "step": 5010
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.7714285714285713e-05,
+      "loss": 0.0437,
+      "step": 5020
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.7642857142857144e-05,
+      "loss": 0.0519,
+      "step": 5030
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.757142857142857e-05,
+      "loss": 0.0269,
+      "step": 5040
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.75e-05,
+      "loss": 0.0335,
+      "step": 5050
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.742857142857143e-05,
+      "loss": 0.0119,
+      "step": 5060
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.7357142857142856e-05,
+      "loss": 0.0249,
+      "step": 5070
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.7285714285714287e-05,
+      "loss": 0.0057,
+      "step": 5080
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.7214285714285715e-05,
+      "loss": 0.0567,
+      "step": 5090
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.7142857142857145e-05,
+      "loss": 0.031,
+      "step": 5100
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.7071428571428573e-05,
+      "loss": 0.0433,
+      "step": 5110
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 0.0288,
+      "step": 5120
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.692857142857143e-05,
+      "loss": 0.0212,
+      "step": 5130
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.6857142857142858e-05,
+      "loss": 0.039,
+      "step": 5140
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.6785714285714285e-05,
+      "loss": 0.0187,
+      "step": 5150
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.6714285714285716e-05,
+      "loss": 0.0079,
+      "step": 5160
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.6642857142857143e-05,
+      "loss": 0.0235,
+      "step": 5170
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.657142857142857e-05,
+      "loss": 0.0397,
+      "step": 5180
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.65e-05,
+      "loss": 0.0444,
+      "step": 5190
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.642857142857143e-05,
+      "loss": 0.0309,
+      "step": 5200
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.6357142857142856e-05,
+      "loss": 0.0346,
+      "step": 5210
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.6285714285714287e-05,
+      "loss": 0.0268,
+      "step": 5220
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.6214285714285714e-05,
+      "loss": 0.0296,
+      "step": 5230
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.614285714285714e-05,
+      "loss": 0.0434,
+      "step": 5240
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.6071428571428572e-05,
+      "loss": 0.0433,
+      "step": 5250
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.0326,
+      "step": 5260
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.592857142857143e-05,
+      "loss": 0.036,
+      "step": 5270
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.5857142857142857e-05,
+      "loss": 0.0131,
+      "step": 5280
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.5785714285714288e-05,
+      "loss": 0.0188,
+      "step": 5290
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.5714285714285715e-05,
+      "loss": 0.0509,
+      "step": 5300
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.5642857142857143e-05,
+      "loss": 0.0366,
+      "step": 5310
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.5571428571428573e-05,
+      "loss": 0.0369,
+      "step": 5320
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.55e-05,
+      "loss": 0.0513,
+      "step": 5330
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.5428571428571428e-05,
+      "loss": 0.0795,
+      "step": 5340
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.535714285714286e-05,
+      "loss": 0.0248,
+      "step": 5350
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.5285714285714286e-05,
+      "loss": 0.0213,
+      "step": 5360
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.5214285714285715e-05,
+      "loss": 0.0321,
+      "step": 5370
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.5142857142857144e-05,
+      "loss": 0.0154,
+      "step": 5380
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.5071428571428573e-05,
+      "loss": 0.0287,
+      "step": 5390
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.5e-05,
+      "loss": 0.059,
+      "step": 5400
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.4928571428571431e-05,
+      "loss": 0.0867,
+      "step": 5410
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.4857142857142858e-05,
+      "loss": 0.0419,
+      "step": 5420
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.4785714285714286e-05,
+      "loss": 0.0209,
+      "step": 5430
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.4714285714285713e-05,
+      "loss": 0.0386,
+      "step": 5440
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.4642857142857144e-05,
+      "loss": 0.0127,
+      "step": 5450
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.4571428571428573e-05,
+      "loss": 0.0339,
+      "step": 5460
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.45e-05,
+      "loss": 0.0267,
+      "step": 5470
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.442857142857143e-05,
+      "loss": 0.0399,
+      "step": 5480
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.4357142857142858e-05,
+      "loss": 0.0292,
+      "step": 5490
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.0516,
+      "step": 5500
+    },
+    {
+      "epoch": 2.2,
+      "eval_accuracy": 0.97535,
+      "eval_f1": 0.9749045558666326,
+      "eval_loss": 0.09278112649917603,
+      "eval_precision": 0.9722814498933902,
+      "eval_recall": 0.9775418538178848,
+      "eval_runtime": 143.6458,
+      "eval_samples_per_second": 139.231,
+      "eval_steps_per_second": 8.702,
+      "step": 5500
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.4214285714285716e-05,
+      "loss": 0.0535,
+      "step": 5510
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.4142857142857143e-05,
+      "loss": 0.0389,
+      "step": 5520
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.407142857142857e-05,
+      "loss": 0.0132,
+      "step": 5530
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 0.0167,
+      "step": 5540
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.392857142857143e-05,
+      "loss": 0.0493,
+      "step": 5550
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.3857142857142858e-05,
+      "loss": 0.0288,
+      "step": 5560
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.3785714285714285e-05,
+      "loss": 0.0268,
+      "step": 5570
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.3714285714285716e-05,
+      "loss": 0.0357,
+      "step": 5580
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.3642857142857143e-05,
+      "loss": 0.0294,
+      "step": 5590
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.357142857142857e-05,
+      "loss": 0.0409,
+      "step": 5600
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.3500000000000001e-05,
+      "loss": 0.0237,
+      "step": 5610
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.3428571428571429e-05,
+      "loss": 0.0338,
+      "step": 5620
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.3357142857142858e-05,
+      "loss": 0.0343,
+      "step": 5630
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.3285714285714288e-05,
+      "loss": 0.0194,
+      "step": 5640
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.3214285714285716e-05,
+      "loss": 0.0397,
+      "step": 5650
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.3142857142857143e-05,
+      "loss": 0.0194,
+      "step": 5660
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.3071428571428574e-05,
+      "loss": 0.048,
+      "step": 5670
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.0276,
+      "step": 5680
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.2928571428571428e-05,
+      "loss": 0.0299,
+      "step": 5690
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.2857142857142857e-05,
+      "loss": 0.0256,
+      "step": 5700
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.2785714285714286e-05,
+      "loss": 0.0334,
+      "step": 5710
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.2714285714285715e-05,
+      "loss": 0.0523,
+      "step": 5720
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.2642857142857143e-05,
+      "loss": 0.04,
+      "step": 5730
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.2571428571428573e-05,
+      "loss": 0.0467,
+      "step": 5740
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.25e-05,
+      "loss": 0.0511,
+      "step": 5750
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.242857142857143e-05,
+      "loss": 0.0487,
+      "step": 5760
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.2357142857142857e-05,
+      "loss": 0.015,
+      "step": 5770
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.2285714285714286e-05,
+      "loss": 0.0387,
+      "step": 5780
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.2214285714285715e-05,
+      "loss": 0.0204,
+      "step": 5790
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.2142857142857144e-05,
+      "loss": 0.0283,
+      "step": 5800
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.2071428571428573e-05,
+      "loss": 0.0226,
+      "step": 5810
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.2e-05,
+      "loss": 0.0207,
+      "step": 5820
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.192857142857143e-05,
+      "loss": 0.0334,
+      "step": 5830
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.1857142857142858e-05,
+      "loss": 0.0186,
+      "step": 5840
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.1785714285714286e-05,
+      "loss": 0.0445,
+      "step": 5850
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.1714285714285715e-05,
+      "loss": 0.0208,
+      "step": 5860
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.1642857142857144e-05,
+      "loss": 0.047,
+      "step": 5870
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.1571428571428573e-05,
+      "loss": 0.0281,
+      "step": 5880
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.1500000000000002e-05,
+      "loss": 0.0429,
+      "step": 5890
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 0.0265,
+      "step": 5900
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.1357142857142858e-05,
+      "loss": 0.0303,
+      "step": 5910
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.1285714285714285e-05,
+      "loss": 0.0363,
+      "step": 5920
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.1214285714285714e-05,
+      "loss": 0.0451,
+      "step": 5930
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.1142857142857143e-05,
+      "loss": 0.0206,
+      "step": 5940
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.107142857142857e-05,
+      "loss": 0.0397,
+      "step": 5950
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.0476,
+      "step": 5960
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.092857142857143e-05,
+      "loss": 0.0217,
+      "step": 5970
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.0857142857142858e-05,
+      "loss": 0.0162,
+      "step": 5980
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.0785714285714287e-05,
+      "loss": 0.0196,
+      "step": 5990
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.0714285714285714e-05,
+      "loss": 0.0355,
+      "step": 6000
+    },
+    {
+      "epoch": 2.4,
+      "eval_accuracy": 0.97435,
+      "eval_f1": 0.9740843647385704,
+      "eval_loss": 0.10171904414892197,
+      "eval_precision": 0.9641964196419642,
+      "eval_recall": 0.9841772151898734,
+      "eval_runtime": 143.9823,
+      "eval_samples_per_second": 138.906,
+      "eval_steps_per_second": 8.682,
+      "step": 6000
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.0642857142857143e-05,
+      "loss": 0.0197,
+      "step": 6010
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.0571428571428572e-05,
+      "loss": 0.0271,
+      "step": 6020
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.05e-05,
+      "loss": 0.0288,
+      "step": 6030
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.0428571428571428e-05,
+      "loss": 0.0257,
+      "step": 6040
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.0357142857142859e-05,
+      "loss": 0.0075,
+      "step": 6050
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.0285714285714286e-05,
+      "loss": 0.0075,
+      "step": 6060
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.0214285714285715e-05,
+      "loss": 0.0532,
+      "step": 6070
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.0142857142857143e-05,
+      "loss": 0.0349,
+      "step": 6080
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.0071428571428572e-05,
+      "loss": 0.043,
+      "step": 6090
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1e-05,
+      "loss": 0.0648,
+      "step": 6100
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 9.928571428571428e-06,
+      "loss": 0.0374,
+      "step": 6110
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 9.857142857142857e-06,
+      "loss": 0.0238,
+      "step": 6120
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 9.785714285714286e-06,
+      "loss": 0.0489,
+      "step": 6130
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 9.714285714285715e-06,
+      "loss": 0.0255,
+      "step": 6140
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 9.642857142857144e-06,
+      "loss": 0.0156,
+      "step": 6150
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 9.571428571428572e-06,
+      "loss": 0.02,
+      "step": 6160
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 9.5e-06,
+      "loss": 0.0311,
+      "step": 6170
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 9.42857142857143e-06,
+      "loss": 0.0208,
+      "step": 6180
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 9.357142857142857e-06,
+      "loss": 0.0185,
+      "step": 6190
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 9.285714285714286e-06,
+      "loss": 0.0241,
+      "step": 6200
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 9.214285714285715e-06,
+      "loss": 0.0528,
+      "step": 6210
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 9.142857142857144e-06,
+      "loss": 0.0251,
+      "step": 6220
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 9.071428571428573e-06,
+      "loss": 0.0282,
+      "step": 6230
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 9e-06,
+      "loss": 0.0298,
+      "step": 6240
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 8.92857142857143e-06,
+      "loss": 0.0134,
+      "step": 6250
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 8.857142857142857e-06,
+      "loss": 0.0437,
+      "step": 6260
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 8.785714285714286e-06,
+      "loss": 0.0379,
+      "step": 6270
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 8.714285714285715e-06,
+      "loss": 0.0317,
+      "step": 6280
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 8.642857142857144e-06,
+      "loss": 0.0121,
+      "step": 6290
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 8.571428571428573e-06,
+      "loss": 0.026,
+      "step": 6300
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 8.500000000000002e-06,
+      "loss": 0.0338,
+      "step": 6310
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 8.428571428571429e-06,
+      "loss": 0.0351,
+      "step": 6320
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 8.357142857142858e-06,
+      "loss": 0.0243,
+      "step": 6330
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 8.285714285714285e-06,
+      "loss": 0.024,
+      "step": 6340
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 8.214285714285714e-06,
+      "loss": 0.031,
+      "step": 6350
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 8.142857142857143e-06,
+      "loss": 0.0263,
+      "step": 6360
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 8.07142857142857e-06,
+      "loss": 0.0506,
+      "step": 6370
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.0175,
+      "step": 6380
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 7.928571428571429e-06,
+      "loss": 0.0467,
+      "step": 6390
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 7.857142857142858e-06,
+      "loss": 0.0422,
+      "step": 6400
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 7.785714285714287e-06,
+      "loss": 0.0575,
+      "step": 6410
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 7.714285714285714e-06,
+      "loss": 0.0292,
+      "step": 6420
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 7.642857142857143e-06,
+      "loss": 0.0486,
+      "step": 6430
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 7.571428571428572e-06,
+      "loss": 0.0328,
+      "step": 6440
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 7.5e-06,
+      "loss": 0.0381,
+      "step": 6450
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 7.428571428571429e-06,
+      "loss": 0.0159,
+      "step": 6460
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 7.3571428571428565e-06,
+      "loss": 0.0434,
+      "step": 6470
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 7.285714285714286e-06,
+      "loss": 0.0195,
+      "step": 6480
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 7.214285714285715e-06,
+      "loss": 0.021,
+      "step": 6490
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 7.142857142857143e-06,
+      "loss": 0.0227,
+      "step": 6500
+    },
+    {
+      "epoch": 2.6,
+      "eval_accuracy": 0.9748,
+      "eval_f1": 0.9743119266055046,
+      "eval_loss": 0.09828508645296097,
+      "eval_precision": 0.9729234527687296,
+      "eval_recall": 0.9757043691302573,
+      "eval_runtime": 143.7114,
+      "eval_samples_per_second": 139.168,
+      "eval_steps_per_second": 8.698,
+      "step": 6500
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 7.071428571428572e-06,
+      "loss": 0.0114,
+      "step": 6510
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 0.0125,
+      "step": 6520
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 6.928571428571429e-06,
+      "loss": 0.0335,
+      "step": 6530
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 6.857142857142858e-06,
+      "loss": 0.0545,
+      "step": 6540
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 6.785714285714285e-06,
+      "loss": 0.0198,
+      "step": 6550
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 6.714285714285714e-06,
+      "loss": 0.0249,
+      "step": 6560
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 6.642857142857144e-06,
+      "loss": 0.034,
+      "step": 6570
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 6.5714285714285714e-06,
+      "loss": 0.0379,
+      "step": 6580
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 0.0301,
+      "step": 6590
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 6.428571428571429e-06,
+      "loss": 0.037,
+      "step": 6600
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 6.357142857142858e-06,
+      "loss": 0.0284,
+      "step": 6610
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.285714285714287e-06,
+      "loss": 0.0276,
+      "step": 6620
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.214285714285715e-06,
+      "loss": 0.0375,
+      "step": 6630
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.142857142857143e-06,
+      "loss": 0.0453,
+      "step": 6640
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.071428571428572e-06,
+      "loss": 0.017,
+      "step": 6650
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6e-06,
+      "loss": 0.0163,
+      "step": 6660
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.928571428571429e-06,
+      "loss": 0.047,
+      "step": 6670
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.857142857142857e-06,
+      "loss": 0.0387,
+      "step": 6680
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.785714285714286e-06,
+      "loss": 0.0244,
+      "step": 6690
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.0282,
+      "step": 6700
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.642857142857143e-06,
+      "loss": 0.0527,
+      "step": 6710
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.571428571428572e-06,
+      "loss": 0.0312,
+      "step": 6720
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 0.027,
+      "step": 6730
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.428571428571429e-06,
+      "loss": 0.0435,
+      "step": 6740
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.357142857142857e-06,
+      "loss": 0.0183,
+      "step": 6750
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.285714285714286e-06,
+      "loss": 0.0194,
+      "step": 6760
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5.214285714285714e-06,
+      "loss": 0.0339,
+      "step": 6770
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5.142857142857143e-06,
+      "loss": 0.013,
+      "step": 6780
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5.071428571428571e-06,
+      "loss": 0.044,
+      "step": 6790
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5e-06,
+      "loss": 0.0303,
+      "step": 6800
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.9285714285714286e-06,
+      "loss": 0.0213,
+      "step": 6810
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.857142857142858e-06,
+      "loss": 0.0118,
+      "step": 6820
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.785714285714286e-06,
+      "loss": 0.0198,
+      "step": 6830
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.714285714285715e-06,
+      "loss": 0.0353,
+      "step": 6840
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.642857142857143e-06,
+      "loss": 0.0162,
+      "step": 6850
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.571428571428572e-06,
+      "loss": 0.0217,
+      "step": 6860
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.5e-06,
+      "loss": 0.0574,
+      "step": 6870
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.428571428571428e-06,
+      "loss": 0.0323,
+      "step": 6880
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.357142857142857e-06,
+      "loss": 0.0278,
+      "step": 6890
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.285714285714286e-06,
+      "loss": 0.0359,
+      "step": 6900
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.2142857142857145e-06,
+      "loss": 0.0287,
+      "step": 6910
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 4.142857142857143e-06,
+      "loss": 0.0323,
+      "step": 6920
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 4.071428571428572e-06,
+      "loss": 0.0305,
+      "step": 6930
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.0114,
+      "step": 6940
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.928571428571429e-06,
+      "loss": 0.031,
+      "step": 6950
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.857142857142857e-06,
+      "loss": 0.0227,
+      "step": 6960
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.785714285714286e-06,
+      "loss": 0.0301,
+      "step": 6970
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.7142857142857146e-06,
+      "loss": 0.0187,
+      "step": 6980
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.642857142857143e-06,
+      "loss": 0.0405,
+      "step": 6990
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.5714285714285714e-06,
+      "loss": 0.0359,
+      "step": 7000
+    },
+    {
+      "epoch": 2.8,
+      "eval_accuracy": 0.97465,
+      "eval_f1": 0.9743330127069305,
+      "eval_loss": 0.0989721491932869,
+      "eval_precision": 0.966455759766998,
+      "eval_recall": 0.9823397305022458,
+      "eval_runtime": 144.719,
+      "eval_samples_per_second": 138.199,
+      "eval_steps_per_second": 8.637,
+      "step": 7000
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.5000000000000004e-06,
+      "loss": 0.0325,
+      "step": 7010
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 3.428571428571429e-06,
+      "loss": 0.0177,
+      "step": 7020
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 3.357142857142857e-06,
+      "loss": 0.0463,
+      "step": 7030
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 3.2857142857142857e-06,
+      "loss": 0.0338,
+      "step": 7040
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 3.2142857142857143e-06,
+      "loss": 0.036,
+      "step": 7050
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 3.1428571428571433e-06,
+      "loss": 0.0384,
+      "step": 7060
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 3.0714285714285715e-06,
+      "loss": 0.0441,
+      "step": 7070
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 3e-06,
+      "loss": 0.0358,
+      "step": 7080
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.9285714285714287e-06,
+      "loss": 0.0317,
+      "step": 7090
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.0246,
+      "step": 7100
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.785714285714286e-06,
+      "loss": 0.0244,
+      "step": 7110
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.7142857142857144e-06,
+      "loss": 0.0172,
+      "step": 7120
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.642857142857143e-06,
+      "loss": 0.0177,
+      "step": 7130
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.5714285714285716e-06,
+      "loss": 0.0256,
+      "step": 7140
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.5e-06,
+      "loss": 0.0286,
+      "step": 7150
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.428571428571429e-06,
+      "loss": 0.0361,
+      "step": 7160
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.3571428571428574e-06,
+      "loss": 0.0075,
+      "step": 7170
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.285714285714286e-06,
+      "loss": 0.0225,
+      "step": 7180
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.214285714285714e-06,
+      "loss": 0.0121,
+      "step": 7190
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.142857142857143e-06,
+      "loss": 0.0264,
+      "step": 7200
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.0714285714285713e-06,
+      "loss": 0.0243,
+      "step": 7210
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.0163,
+      "step": 7220
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.9285714285714285e-06,
+      "loss": 0.0232,
+      "step": 7230
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.8571428571428573e-06,
+      "loss": 0.031,
+      "step": 7240
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.7857142857142857e-06,
+      "loss": 0.041,
+      "step": 7250
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.7142857142857145e-06,
+      "loss": 0.0237,
+      "step": 7260
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.6428571428571429e-06,
+      "loss": 0.0494,
+      "step": 7270
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.5714285714285717e-06,
+      "loss": 0.0588,
+      "step": 7280
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.5e-06,
+      "loss": 0.0147,
+      "step": 7290
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 0.0343,
+      "step": 7300
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.3571428571428572e-06,
+      "loss": 0.028,
+      "step": 7310
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.2857142857142858e-06,
+      "loss": 0.0238,
+      "step": 7320
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.2142857142857144e-06,
+      "loss": 0.0233,
+      "step": 7330
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.142857142857143e-06,
+      "loss": 0.0241,
+      "step": 7340
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.0714285714285716e-06,
+      "loss": 0.0322,
+      "step": 7350
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.0297,
+      "step": 7360
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 9.285714285714287e-07,
+      "loss": 0.007,
+      "step": 7370
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 8.571428571428572e-07,
+      "loss": 0.0426,
+      "step": 7380
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 7.857142857142858e-07,
+      "loss": 0.0241,
+      "step": 7390
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 7.142857142857143e-07,
+      "loss": 0.0234,
+      "step": 7400
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 6.428571428571429e-07,
+      "loss": 0.04,
+      "step": 7410
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.714285714285715e-07,
+      "loss": 0.0163,
+      "step": 7420
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 0.0136,
+      "step": 7430
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 4.285714285714286e-07,
+      "loss": 0.018,
+      "step": 7440
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 3.5714285714285716e-07,
+      "loss": 0.0194,
+      "step": 7450
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.8571428571428575e-07,
+      "loss": 0.0215,
+      "step": 7460
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 2.142857142857143e-07,
+      "loss": 0.028,
+      "step": 7470
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.4285714285714287e-07,
+      "loss": 0.0237,
+      "step": 7480
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 7.142857142857144e-08,
+      "loss": 0.0144,
+      "step": 7490
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.0,
+      "loss": 0.0384,
+      "step": 7500
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.97455,
+      "eval_f1": 0.9742369792984765,
+      "eval_loss": 0.10010936111211777,
+      "eval_precision": 0.9661680554161228,
+      "eval_recall": 0.9824418129848917,
+      "eval_runtime": 143.9014,
+      "eval_samples_per_second": 138.984,
+      "eval_steps_per_second": 8.687,
+      "step": 7500
+    },
+    {
+      "epoch": 3.0,
+      "step": 7500,
+      "total_flos": 9.043732950372576e+16,
+      "train_loss": 0.07549678431898356,
+      "train_runtime": 12137.3214,
+      "train_samples_per_second": 39.547,
+      "train_steps_per_second": 0.618
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.97535,
+      "eval_f1": 0.9749045558666326,
+      "eval_loss": 0.09278112649917603,
+      "eval_precision": 0.9722814498933902,
+      "eval_recall": 0.9775418538178848,
+      "eval_runtime": 144.2966,
+      "eval_samples_per_second": 138.603,
+      "eval_steps_per_second": 8.663,
+      "step": 7500
+    }
+  ],
+  "max_steps": 7500,
+  "num_train_epochs": 3,
+  "total_flos": 9.043732950372576e+16,
+  "trial_name": null,
+  "trial_params": null
+}