diff --git "a/checkpoint-8000/trainer_state.json" "b/checkpoint-8000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-8000/trainer_state.json"
@@ -0,0 +1,4960 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 18.867924528301888,
+  "global_step": 8000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-05,
+      "loss": 0.0325,
+      "step": 10
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 20
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3e-05,
+      "loss": 0.0313,
+      "step": 30
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4e-05,
+      "loss": 0.0401,
+      "step": 40
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 0.0392,
+      "step": 50
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 6e-05,
+      "loss": 0.0417,
+      "step": 60
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 7e-05,
+      "loss": 0.0513,
+      "step": 70
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 8e-05,
+      "loss": 0.0591,
+      "step": 80
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9e-05,
+      "loss": 0.0679,
+      "step": 90
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001,
+      "loss": 0.0808,
+      "step": 100
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 9.988066825775657e-05,
+      "loss": 0.0966,
+      "step": 110
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 9.976133651551312e-05,
+      "loss": 0.1199,
+      "step": 120
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 9.964200477326969e-05,
+      "loss": 0.1203,
+      "step": 130
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 9.952267303102626e-05,
+      "loss": 0.1137,
+      "step": 140
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 9.940334128878282e-05,
+      "loss": 0.1278,
+      "step": 150
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 9.928400954653937e-05,
+      "loss": 0.1318,
+      "step": 160
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 9.916467780429594e-05,
+      "loss": 0.1185,
+      "step": 170
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 9.90453460620525e-05,
+      "loss": 0.1277,
+      "step": 180
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 9.892601431980909e-05,
+      "loss": 0.1277,
+      "step": 190
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 9.880668257756564e-05,
+      "loss": 0.1326,
+      "step": 200
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.86873508353222e-05,
+      "loss": 0.1424,
+      "step": 210
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.856801909307877e-05,
+      "loss": 0.1577,
+      "step": 220
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.844868735083532e-05,
+      "loss": 0.1477,
+      "step": 230
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 9.832935560859189e-05,
+      "loss": 0.1388,
+      "step": 240
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 9.821002386634846e-05,
+      "loss": 0.1458,
+      "step": 250
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.809069212410502e-05,
+      "loss": 0.1589,
+      "step": 260
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.797136038186157e-05,
+      "loss": 0.1611,
+      "step": 270
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 9.785202863961814e-05,
+      "loss": 0.155,
+      "step": 280
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.77326968973747e-05,
+      "loss": 0.1472,
+      "step": 290
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 9.761336515513126e-05,
+      "loss": 0.1394,
+      "step": 300
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.749403341288783e-05,
+      "loss": 0.1577,
+      "step": 310
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.737470167064439e-05,
+      "loss": 0.1582,
+      "step": 320
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 9.725536992840096e-05,
+      "loss": 0.1476,
+      "step": 330
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.713603818615752e-05,
+      "loss": 0.1562,
+      "step": 340
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.701670644391409e-05,
+      "loss": 0.1633,
+      "step": 350
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 9.689737470167066e-05,
+      "loss": 0.1647,
+      "step": 360
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 9.677804295942722e-05,
+      "loss": 0.1633,
+      "step": 370
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 9.665871121718377e-05,
+      "loss": 0.1617,
+      "step": 380
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 9.653937947494034e-05,
+      "loss": 0.1535,
+      "step": 390
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 9.64200477326969e-05,
+      "loss": 0.1524,
+      "step": 400
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 9.630071599045346e-05,
+      "loss": 0.1773,
+      "step": 410
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 9.618138424821003e-05,
+      "loss": 0.1661,
+      "step": 420
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.7361899614334106,
+      "eval_runtime": 26.9369,
+      "eval_samples_per_second": 27.954,
+      "eval_steps_per_second": 1.782,
+      "step": 424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.606205250596659e-05,
+      "loss": 0.1356,
+      "step": 430
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.594272076372316e-05,
+      "loss": 0.119,
+      "step": 440
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.582338902147971e-05,
+      "loss": 0.109,
+      "step": 450
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.570405727923628e-05,
+      "loss": 0.1161,
+      "step": 460
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.558472553699284e-05,
+      "loss": 0.1272,
+      "step": 470
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 9.546539379474941e-05,
+      "loss": 0.1147,
+      "step": 480
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 9.534606205250597e-05,
+      "loss": 0.1268,
+      "step": 490
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 9.522673031026254e-05,
+      "loss": 0.1321,
+      "step": 500
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 9.51073985680191e-05,
+      "loss": 0.1142,
+      "step": 510
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 9.498806682577566e-05,
+      "loss": 0.1211,
+      "step": 520
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 9.486873508353223e-05,
+      "loss": 0.1174,
+      "step": 530
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 9.474940334128879e-05,
+      "loss": 0.1309,
+      "step": 540
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 9.463007159904536e-05,
+      "loss": 0.1224,
+      "step": 550
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 9.451073985680191e-05,
+      "loss": 0.1271,
+      "step": 560
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 9.439140811455848e-05,
+      "loss": 0.1339,
+      "step": 570
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 9.427207637231504e-05,
+      "loss": 0.1116,
+      "step": 580
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 9.41527446300716e-05,
+      "loss": 0.1438,
+      "step": 590
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 9.403341288782816e-05,
+      "loss": 0.1483,
+      "step": 600
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 9.391408114558473e-05,
+      "loss": 0.1312,
+      "step": 610
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 9.379474940334129e-05,
+      "loss": 0.1397,
+      "step": 620
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.367541766109786e-05,
+      "loss": 0.1488,
+      "step": 630
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 9.355608591885443e-05,
+      "loss": 0.1379,
+      "step": 640
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 9.343675417661099e-05,
+      "loss": 0.1344,
+      "step": 650
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.331742243436754e-05,
+      "loss": 0.1241,
+      "step": 660
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.319809069212411e-05,
+      "loss": 0.1208,
+      "step": 670
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.307875894988068e-05,
+      "loss": 0.119,
+      "step": 680
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.295942720763724e-05,
+      "loss": 0.1301,
+      "step": 690
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.28400954653938e-05,
+      "loss": 0.1448,
+      "step": 700
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 9.272076372315036e-05,
+      "loss": 0.1258,
+      "step": 710
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 9.260143198090693e-05,
+      "loss": 0.1306,
+      "step": 720
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 9.248210023866349e-05,
+      "loss": 0.1322,
+      "step": 730
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 9.236276849642005e-05,
+      "loss": 0.1276,
+      "step": 740
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 9.224343675417661e-05,
+      "loss": 0.1462,
+      "step": 750
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 9.212410501193318e-05,
+      "loss": 0.1202,
+      "step": 760
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 9.200477326968973e-05,
+      "loss": 0.145,
+      "step": 770
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.18854415274463e-05,
+      "loss": 0.1472,
+      "step": 780
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.176610978520288e-05,
+      "loss": 0.1297,
+      "step": 790
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.164677804295944e-05,
+      "loss": 0.1433,
+      "step": 800
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 9.1527446300716e-05,
+      "loss": 0.136,
+      "step": 810
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 9.140811455847256e-05,
+      "loss": 0.1667,
+      "step": 820
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 9.128878281622913e-05,
+      "loss": 0.1461,
+      "step": 830
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 9.116945107398569e-05,
+      "loss": 0.1341,
+      "step": 840
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.739912748336792,
+      "eval_runtime": 26.737,
+      "eval_samples_per_second": 28.163,
+      "eval_steps_per_second": 1.795,
+      "step": 848
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 9.105011933174225e-05,
+      "loss": 0.1389,
+      "step": 850
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 9.093078758949881e-05,
+      "loss": 0.0738,
+      "step": 860
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 9.081145584725538e-05,
+      "loss": 0.0694,
+      "step": 870
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 9.069212410501193e-05,
+      "loss": 0.0837,
+      "step": 880
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 9.05727923627685e-05,
+      "loss": 0.0774,
+      "step": 890
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 9.045346062052506e-05,
+      "loss": 0.0781,
+      "step": 900
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 9.033412887828163e-05,
+      "loss": 0.0764,
+      "step": 910
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 9.021479713603818e-05,
+      "loss": 0.0804,
+      "step": 920
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 9.009546539379475e-05,
+      "loss": 0.0832,
+      "step": 930
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 8.997613365155131e-05,
+      "loss": 0.0864,
+      "step": 940
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 8.985680190930788e-05,
+      "loss": 0.0789,
+      "step": 950
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 8.973747016706445e-05,
+      "loss": 0.0873,
+      "step": 960
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 8.961813842482101e-05,
+      "loss": 0.0797,
+      "step": 970
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 8.949880668257758e-05,
+      "loss": 0.0793,
+      "step": 980
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 8.937947494033413e-05,
+      "loss": 0.0894,
+      "step": 990
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 8.92601431980907e-05,
+      "loss": 0.086,
+      "step": 1000
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 8.914081145584726e-05,
+      "loss": 0.087,
+      "step": 1010
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 8.902147971360383e-05,
+      "loss": 0.0904,
+      "step": 1020
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 8.890214797136038e-05,
+      "loss": 0.0836,
+      "step": 1030
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 8.878281622911695e-05,
+      "loss": 0.0851,
+      "step": 1040
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 8.866348448687351e-05,
+      "loss": 0.0801,
+      "step": 1050
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 8.854415274463007e-05,
+      "loss": 0.081,
+      "step": 1060
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 8.842482100238663e-05,
+      "loss": 0.0874,
+      "step": 1070
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 8.83054892601432e-05,
+      "loss": 0.0906,
+      "step": 1080
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 8.818615751789976e-05,
+      "loss": 0.0981,
+      "step": 1090
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 8.806682577565633e-05,
+      "loss": 0.0825,
+      "step": 1100
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.79474940334129e-05,
+      "loss": 0.0919,
+      "step": 1110
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 8.782816229116946e-05,
+      "loss": 0.0908,
+      "step": 1120
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 8.770883054892602e-05,
+      "loss": 0.0858,
+      "step": 1130
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 8.758949880668258e-05,
+      "loss": 0.0968,
+      "step": 1140
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 8.747016706443915e-05,
+      "loss": 0.0883,
+      "step": 1150
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 8.735083532219571e-05,
+      "loss": 0.0824,
+      "step": 1160
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 8.723150357995227e-05,
+      "loss": 0.0875,
+      "step": 1170
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 8.711217183770883e-05,
+      "loss": 0.0834,
+      "step": 1180
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 8.69928400954654e-05,
+      "loss": 0.087,
+      "step": 1190
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 8.687350835322196e-05,
+      "loss": 0.0998,
+      "step": 1200
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 8.675417661097852e-05,
+      "loss": 0.0846,
+      "step": 1210
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.663484486873508e-05,
+      "loss": 0.0931,
+      "step": 1220
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 8.651551312649165e-05,
+      "loss": 0.0859,
+      "step": 1230
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 8.639618138424822e-05,
+      "loss": 0.0907,
+      "step": 1240
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 8.627684964200478e-05,
+      "loss": 0.0966,
+      "step": 1250
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 8.615751789976135e-05,
+      "loss": 0.0924,
+      "step": 1260
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 8.603818615751791e-05,
+      "loss": 0.0921,
+      "step": 1270
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 1.7798256874084473,
+      "eval_runtime": 33.5926,
+      "eval_samples_per_second": 22.416,
+      "eval_steps_per_second": 1.429,
+      "step": 1272
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 8.591885441527447e-05,
+      "loss": 0.0546,
+      "step": 1280
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 8.579952267303103e-05,
+      "loss": 0.0488,
+      "step": 1290
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 8.56801909307876e-05,
+      "loss": 0.0443,
+      "step": 1300
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 8.556085918854416e-05,
+      "loss": 0.0435,
+      "step": 1310
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 8.544152744630072e-05,
+      "loss": 0.0465,
+      "step": 1320
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 8.532219570405728e-05,
+      "loss": 0.0502,
+      "step": 1330
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 8.520286396181385e-05,
+      "loss": 0.0513,
+      "step": 1340
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 8.50835322195704e-05,
+      "loss": 0.0593,
+      "step": 1350
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 8.496420047732697e-05,
+      "loss": 0.0554,
+      "step": 1360
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 8.484486873508353e-05,
+      "loss": 0.056,
+      "step": 1370
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 8.47255369928401e-05,
+      "loss": 0.0538,
+      "step": 1380
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 8.460620525059667e-05,
+      "loss": 0.047,
+      "step": 1390
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 8.448687350835323e-05,
+      "loss": 0.0458,
+      "step": 1400
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 8.43675417661098e-05,
+      "loss": 0.0619,
+      "step": 1410
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 8.424821002386635e-05,
+      "loss": 0.0536,
+      "step": 1420
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 8.412887828162292e-05,
+      "loss": 0.0458,
+      "step": 1430
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 8.400954653937948e-05,
+      "loss": 0.0491,
+      "step": 1440
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 8.389021479713605e-05,
+      "loss": 0.0572,
+      "step": 1450
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 8.37708830548926e-05,
+      "loss": 0.0511,
+      "step": 1460
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 8.365155131264917e-05,
+      "loss": 0.045,
+      "step": 1470
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 8.353221957040573e-05,
+      "loss": 0.0587,
+      "step": 1480
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 8.34128878281623e-05,
+      "loss": 0.0498,
+      "step": 1490
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 8.329355608591885e-05,
+      "loss": 0.0569,
+      "step": 1500
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 8.317422434367542e-05,
+      "loss": 0.0555,
+      "step": 1510
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 8.305489260143198e-05,
+      "loss": 0.0569,
+      "step": 1520
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 8.293556085918854e-05,
+      "loss": 0.0516,
+      "step": 1530
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 8.28162291169451e-05,
+      "loss": 0.0587,
+      "step": 1540
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 8.269689737470167e-05,
+      "loss": 0.0653,
+      "step": 1550
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 8.257756563245825e-05,
+      "loss": 0.0541,
+      "step": 1560
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 8.24582338902148e-05,
+      "loss": 0.0525,
+      "step": 1570
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 8.233890214797137e-05,
+      "loss": 0.0696,
+      "step": 1580
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 8.221957040572793e-05,
+      "loss": 0.0589,
+      "step": 1590
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 8.210023866348449e-05,
+      "loss": 0.0635,
+      "step": 1600
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 8.198090692124105e-05,
+      "loss": 0.0565,
+      "step": 1610
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 8.186157517899762e-05,
+      "loss": 0.0616,
+      "step": 1620
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 8.174224343675418e-05,
+      "loss": 0.066,
+      "step": 1630
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 8.162291169451074e-05,
+      "loss": 0.0669,
+      "step": 1640
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 8.15035799522673e-05,
+      "loss": 0.0652,
+      "step": 1650
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 8.138424821002387e-05,
+      "loss": 0.0653,
+      "step": 1660
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 8.126491646778044e-05,
+      "loss": 0.0489,
+      "step": 1670
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 8.114558472553699e-05,
+      "loss": 0.063,
+      "step": 1680
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 8.102625298329355e-05,
+      "loss": 0.0593,
+      "step": 1690
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 1.8369163274765015,
+      "eval_runtime": 26.4761,
+      "eval_samples_per_second": 28.441,
+      "eval_steps_per_second": 1.813,
+      "step": 1696
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 8.090692124105012e-05,
+      "loss": 0.0566,
+      "step": 1700
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 8.078758949880669e-05,
+      "loss": 0.0352,
+      "step": 1710
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 8.066825775656325e-05,
+      "loss": 0.0317,
+      "step": 1720
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 8.054892601431982e-05,
+      "loss": 0.0341,
+      "step": 1730
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 8.042959427207638e-05,
+      "loss": 0.0361,
+      "step": 1740
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 8.031026252983294e-05,
+      "loss": 0.034,
+      "step": 1750
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 8.01909307875895e-05,
+      "loss": 0.0347,
+      "step": 1760
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 8.007159904534607e-05,
+      "loss": 0.04,
+      "step": 1770
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 7.995226730310262e-05,
+      "loss": 0.0308,
+      "step": 1780
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 7.983293556085919e-05,
+      "loss": 0.0352,
+      "step": 1790
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 7.971360381861575e-05,
+      "loss": 0.0353,
+      "step": 1800
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 7.959427207637232e-05,
+      "loss": 0.0383,
+      "step": 1810
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 7.947494033412887e-05,
+      "loss": 0.037,
+      "step": 1820
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 7.935560859188544e-05,
+      "loss": 0.0389,
+      "step": 1830
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 7.9236276849642e-05,
+      "loss": 0.0371,
+      "step": 1840
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 7.911694510739857e-05,
+      "loss": 0.0373,
+      "step": 1850
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 7.899761336515514e-05,
+      "loss": 0.0418,
+      "step": 1860
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 7.88782816229117e-05,
+      "loss": 0.0464,
+      "step": 1870
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 7.875894988066827e-05,
+      "loss": 0.0402,
+      "step": 1880
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 7.863961813842482e-05,
+      "loss": 0.0326,
+      "step": 1890
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 7.852028639618139e-05,
+      "loss": 0.0361,
+      "step": 1900
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 7.840095465393795e-05,
+      "loss": 0.0364,
+      "step": 1910
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 7.828162291169452e-05,
+      "loss": 0.0348,
+      "step": 1920
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 7.816229116945107e-05,
+      "loss": 0.048,
+      "step": 1930
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 7.804295942720764e-05,
+      "loss": 0.0429,
+      "step": 1940
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 7.79236276849642e-05,
+      "loss": 0.0446,
+      "step": 1950
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 7.780429594272077e-05,
+      "loss": 0.0386,
+      "step": 1960
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 7.768496420047732e-05,
+      "loss": 0.0429,
+      "step": 1970
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 7.756563245823389e-05,
+      "loss": 0.0369,
+      "step": 1980
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 7.744630071599046e-05,
+      "loss": 0.0407,
+      "step": 1990
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 7.732696897374702e-05,
+      "loss": 0.0348,
+      "step": 2000
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 7.720763723150359e-05,
+      "loss": 0.0377,
+      "step": 2010
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 7.708830548926015e-05,
+      "loss": 0.0424,
+      "step": 2020
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 7.696897374701672e-05,
+      "loss": 0.0401,
+      "step": 2030
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 7.684964200477327e-05,
+      "loss": 0.0519,
+      "step": 2040
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 7.673031026252984e-05,
+      "loss": 0.0383,
+      "step": 2050
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 7.66109785202864e-05,
+      "loss": 0.0413,
+      "step": 2060
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 7.649164677804296e-05,
+      "loss": 0.0389,
+      "step": 2070
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 7.637231503579952e-05,
+      "loss": 0.0397,
+      "step": 2080
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 7.625298329355609e-05,
+      "loss": 0.0364,
+      "step": 2090
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 7.613365155131266e-05,
+      "loss": 0.0402,
+      "step": 2100
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 7.601431980906921e-05,
+      "loss": 0.0399,
+      "step": 2110
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 7.589498806682577e-05,
+      "loss": 0.0534,
+      "step": 2120
+    },
+    {
+      "epoch": 5.0,
+      "eval_loss": 1.8871911764144897,
+      "eval_runtime": 32.4548,
+      "eval_samples_per_second": 23.202,
+      "eval_steps_per_second": 1.479,
+      "step": 2120
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 7.577565632458234e-05,
+      "loss": 0.0232,
+      "step": 2130
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 7.565632458233891e-05,
+      "loss": 0.0197,
+      "step": 2140
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 7.553699284009546e-05,
+      "loss": 0.022,
+      "step": 2150
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 7.541766109785204e-05,
+      "loss": 0.027,
+      "step": 2160
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 7.52983293556086e-05,
+      "loss": 0.0236,
+      "step": 2170
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 7.517899761336516e-05,
+      "loss": 0.0222,
+      "step": 2180
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 7.505966587112172e-05,
+      "loss": 0.0255,
+      "step": 2190
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 7.494033412887829e-05,
+      "loss": 0.0214,
+      "step": 2200
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 7.482100238663486e-05,
+      "loss": 0.028,
+      "step": 2210
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 7.470167064439141e-05,
+      "loss": 0.0291,
+      "step": 2220
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 7.458233890214797e-05,
+      "loss": 0.0343,
+      "step": 2230
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 7.446300715990454e-05,
+      "loss": 0.0292,
+      "step": 2240
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 7.43436754176611e-05,
+      "loss": 0.0227,
+      "step": 2250
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 7.422434367541766e-05,
+      "loss": 0.0242,
+      "step": 2260
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 7.410501193317423e-05,
+      "loss": 0.0233,
+      "step": 2270
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 7.398568019093079e-05,
+      "loss": 0.0324,
+      "step": 2280
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 7.386634844868734e-05,
+      "loss": 0.0276,
+      "step": 2290
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 7.374701670644391e-05,
+      "loss": 0.0235,
+      "step": 2300
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 7.362768496420048e-05,
+      "loss": 0.0249,
+      "step": 2310
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 7.350835322195704e-05,
+      "loss": 0.0234,
+      "step": 2320
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 7.338902147971361e-05,
+      "loss": 0.0264,
+      "step": 2330
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 7.326968973747017e-05,
+      "loss": 0.0302,
+      "step": 2340
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 7.315035799522674e-05,
+      "loss": 0.0291,
+      "step": 2350
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 7.30310262529833e-05,
+      "loss": 0.0279,
+      "step": 2360
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 7.291169451073986e-05,
+      "loss": 0.027,
+      "step": 2370
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 7.279236276849643e-05,
+      "loss": 0.026,
+      "step": 2380
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 7.267303102625299e-05,
+      "loss": 0.0294,
+      "step": 2390
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 7.255369928400954e-05,
+      "loss": 0.0247,
+      "step": 2400
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 7.243436754176611e-05,
+      "loss": 0.0291,
+      "step": 2410
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 7.231503579952268e-05,
+      "loss": 0.0263,
+      "step": 2420
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 7.219570405727923e-05,
+      "loss": 0.0315,
+      "step": 2430
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 7.20763723150358e-05,
+      "loss": 0.0313,
+      "step": 2440
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 7.195704057279236e-05,
+      "loss": 0.0294,
+      "step": 2450
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 7.183770883054893e-05,
+      "loss": 0.033,
+      "step": 2460
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 7.17183770883055e-05,
+      "loss": 0.0251,
+      "step": 2470
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 7.159904534606206e-05,
+      "loss": 0.0318,
+      "step": 2480
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 7.147971360381863e-05,
+      "loss": 0.0289,
+      "step": 2490
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 7.136038186157519e-05,
+      "loss": 0.0265,
+      "step": 2500
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 7.124105011933174e-05,
+      "loss": 0.0313,
+      "step": 2510
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 7.112171837708831e-05,
+      "loss": 0.0295,
+      "step": 2520
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 7.100238663484488e-05,
+      "loss": 0.0236,
+      "step": 2530
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 7.088305489260143e-05,
+      "loss": 0.0255,
+      "step": 2540
+    },
+    {
+      "epoch": 6.0,
+      "eval_loss": 1.9039226770401,
+      "eval_runtime": 26.9225,
+      "eval_samples_per_second": 27.969,
+      "eval_steps_per_second": 1.783,
+      "step": 2544
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 7.0763723150358e-05,
+      "loss": 0.0236,
+      "step": 2550
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 7.064439140811456e-05,
+      "loss": 0.0171,
+      "step": 2560
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 7.052505966587113e-05,
+      "loss": 0.0169,
+      "step": 2570
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 7.040572792362768e-05,
+      "loss": 0.0164,
+      "step": 2580
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 7.028639618138425e-05,
+      "loss": 0.0141,
+      "step": 2590
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 7.016706443914081e-05,
+      "loss": 0.0124,
+      "step": 2600
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 7.004773269689738e-05,
+      "loss": 0.0165,
+      "step": 2610
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 6.992840095465394e-05,
+      "loss": 0.0143,
+      "step": 2620
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 6.980906921241051e-05,
+      "loss": 0.02,
+      "step": 2630
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 6.968973747016708e-05,
+      "loss": 0.0167,
+      "step": 2640
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 6.957040572792363e-05,
+      "loss": 0.0158,
+      "step": 2650
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 6.94510739856802e-05,
+      "loss": 0.016,
+      "step": 2660
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 6.933174224343676e-05,
+      "loss": 0.0194,
+      "step": 2670
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 6.921241050119333e-05,
+      "loss": 0.0178,
+      "step": 2680
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 6.909307875894988e-05,
+      "loss": 0.019,
+      "step": 2690
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 6.897374701670645e-05,
+      "loss": 0.0139,
+      "step": 2700
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 6.885441527446301e-05,
+      "loss": 0.0169,
+      "step": 2710
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 6.873508353221956e-05,
+      "loss": 0.0194,
+      "step": 2720
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 6.861575178997613e-05,
+      "loss": 0.0137,
+      "step": 2730
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 6.84964200477327e-05,
+      "loss": 0.0149,
+      "step": 2740
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 6.837708830548926e-05,
+      "loss": 0.0172,
+      "step": 2750
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 6.825775656324583e-05,
+      "loss": 0.0138,
+      "step": 2760
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 6.81384248210024e-05,
+      "loss": 0.0233,
+      "step": 2770
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 6.801909307875896e-05,
+      "loss": 0.0181,
+      "step": 2780
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 6.789976133651553e-05,
+      "loss": 0.0169,
+      "step": 2790
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 6.778042959427208e-05,
+      "loss": 0.0203,
+      "step": 2800
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 6.766109785202865e-05,
+      "loss": 0.0163,
+      "step": 2810
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 6.754176610978521e-05,
+      "loss": 0.0168,
+      "step": 2820
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 6.742243436754176e-05,
+      "loss": 0.0161,
+      "step": 2830
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 6.730310262529833e-05,
+      "loss": 0.0155,
+      "step": 2840
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 6.71837708830549e-05,
+      "loss": 0.016,
+      "step": 2850
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 6.706443914081146e-05,
+      "loss": 0.0207,
+      "step": 2860
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 6.694510739856802e-05,
+      "loss": 0.024,
+      "step": 2870
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 6.682577565632458e-05,
+      "loss": 0.0222,
+      "step": 2880
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 6.670644391408115e-05,
+      "loss": 0.0245,
+      "step": 2890
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 6.65871121718377e-05,
+      "loss": 0.0194,
+      "step": 2900
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 6.646778042959427e-05,
+      "loss": 0.0179,
+      "step": 2910
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 6.634844868735083e-05,
+      "loss": 0.0188,
+      "step": 2920
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 6.622911694510741e-05,
+      "loss": 0.0193,
+      "step": 2930
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 6.610978520286396e-05,
+      "loss": 0.0155,
+      "step": 2940
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 6.599045346062053e-05,
+      "loss": 0.015,
+      "step": 2950
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 6.58711217183771e-05,
+      "loss": 0.0211,
+      "step": 2960
+    },
+    {
+      "epoch": 7.0,
+      "eval_loss": 1.957861304283142,
+      "eval_runtime": 29.2453,
+      "eval_samples_per_second": 25.748,
+      "eval_steps_per_second": 1.641,
+      "step": 2968
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 6.575178997613366e-05,
+      "loss": 0.0141,
+      "step": 2970
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 6.563245823389022e-05,
+      "loss": 0.0104,
+      "step": 2980
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 6.551312649164678e-05,
+      "loss": 0.0086,
+      "step": 2990
+    },
+    {
+      "epoch": 7.08,
+      "learning_rate": 6.539379474940335e-05,
+      "loss": 0.0125,
+      "step": 3000
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 6.52744630071599e-05,
+      "loss": 0.0132,
+      "step": 3010
+    },
+    {
+      "epoch": 7.12,
+      "learning_rate": 6.515513126491647e-05,
+      "loss": 0.0125,
+      "step": 3020
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 6.503579952267303e-05,
+      "loss": 0.0128,
+      "step": 3030
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 6.49164677804296e-05,
+      "loss": 0.014,
+      "step": 3040
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 6.479713603818615e-05,
+      "loss": 0.0112,
+      "step": 3050
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 6.467780429594272e-05,
+      "loss": 0.0092,
+      "step": 3060
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 6.455847255369928e-05,
+      "loss": 0.0097,
+      "step": 3070
+    },
+    {
+      "epoch": 7.26,
+      "learning_rate": 6.443914081145585e-05,
+      "loss": 0.0151,
+      "step": 3080
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 6.431980906921242e-05,
+      "loss": 0.01,
+      "step": 3090
+    },
+    {
+      "epoch": 7.31,
+      "learning_rate": 6.420047732696898e-05,
+      "loss": 0.01,
+      "step": 3100
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 6.408114558472555e-05,
+      "loss": 0.0119,
+      "step": 3110
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 6.39618138424821e-05,
+      "loss": 0.0143,
+      "step": 3120
+    },
+    {
+      "epoch": 7.38,
+      "learning_rate": 6.384248210023867e-05,
+      "loss": 0.0099,
+      "step": 3130
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 6.372315035799523e-05,
+      "loss": 0.0175,
+      "step": 3140
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 6.36038186157518e-05,
+      "loss": 0.0112,
+      "step": 3150
+    },
+    {
+      "epoch": 7.45,
+      "learning_rate": 6.348448687350835e-05,
+      "loss": 0.0113,
+      "step": 3160
+    },
+    {
+      "epoch": 7.48,
+      "learning_rate": 6.336515513126492e-05,
+      "loss": 0.0107,
+      "step": 3170
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 6.324582338902148e-05,
+      "loss": 0.0096,
+      "step": 3180
+    },
+    {
+      "epoch": 7.52,
+      "learning_rate": 6.312649164677804e-05,
+      "loss": 0.0132,
+      "step": 3190
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 6.30071599045346e-05,
+      "loss": 0.0154,
+      "step": 3200
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 6.288782816229117e-05,
+      "loss": 0.0126,
+      "step": 3210
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 6.276849642004773e-05,
+      "loss": 0.0099,
+      "step": 3220
+    },
+    {
+      "epoch": 7.62,
+      "learning_rate": 6.26491646778043e-05,
+      "loss": 0.0181,
+      "step": 3230
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 6.252983293556087e-05,
+      "loss": 0.0094,
+      "step": 3240
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 6.241050119331743e-05,
+      "loss": 0.0113,
+      "step": 3250
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 6.2291169451074e-05,
+      "loss": 0.0104,
+      "step": 3260
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 6.217183770883055e-05,
+      "loss": 0.0127,
+      "step": 3270
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 6.205250596658712e-05,
+      "loss": 0.0121,
+      "step": 3280
+    },
+    {
+      "epoch": 7.76,
+      "learning_rate": 6.193317422434368e-05,
+      "loss": 0.0103,
+      "step": 3290
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 6.181384248210024e-05,
+      "loss": 0.0138,
+      "step": 3300
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 6.16945107398568e-05,
+      "loss": 0.0161,
+      "step": 3310
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 6.157517899761337e-05,
+      "loss": 0.0104,
+      "step": 3320
+    },
+    {
+      "epoch": 7.85,
+      "learning_rate": 6.145584725536993e-05,
+      "loss": 0.0096,
+      "step": 3330
+    },
+    {
+      "epoch": 7.88,
+      "learning_rate": 6.133651551312649e-05,
+      "loss": 0.0095,
+      "step": 3340
+    },
+    {
+      "epoch": 7.9,
+      "learning_rate": 6.121718377088305e-05,
+      "loss": 0.0139,
+      "step": 3350
+    },
+    {
+      "epoch": 7.92,
+      "learning_rate": 6.109785202863962e-05,
+      "loss": 0.0123,
+      "step": 3360
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 6.097852028639618e-05,
+      "loss": 0.0129,
+      "step": 3370
+    },
+    {
+      "epoch": 7.97,
+      "learning_rate": 6.0859188544152745e-05,
+      "loss": 0.009,
+      "step": 3380
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 6.073985680190931e-05,
+      "loss": 0.0122,
+      "step": 3390
+    },
+    {
+      "epoch": 8.0,
+      "eval_loss": 1.9987685680389404,
+      "eval_runtime": 31.4837,
+      "eval_samples_per_second": 23.917,
+      "eval_steps_per_second": 1.525,
+      "step": 3392
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 6.062052505966588e-05,
+      "loss": 0.0075,
+      "step": 3400
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 6.0501193317422436e-05,
+      "loss": 0.0082,
+      "step": 3410
+    },
+    {
+      "epoch": 8.07,
+      "learning_rate": 6.0381861575179e-05,
+      "loss": 0.0054,
+      "step": 3420
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 6.026252983293557e-05,
+      "loss": 0.0049,
+      "step": 3430
+    },
+    {
+      "epoch": 8.11,
+      "learning_rate": 6.0143198090692134e-05,
+      "loss": 0.0084,
+      "step": 3440
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 6.002386634844869e-05,
+      "loss": 0.0107,
+      "step": 3450
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 5.990453460620525e-05,
+      "loss": 0.0074,
+      "step": 3460
+    },
+    {
+      "epoch": 8.18,
+      "learning_rate": 5.978520286396182e-05,
+      "loss": 0.0076,
+      "step": 3470
+    },
+    {
+      "epoch": 8.21,
+      "learning_rate": 5.966587112171838e-05,
+      "loss": 0.0056,
+      "step": 3480
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 5.9546539379474945e-05,
+      "loss": 0.0113,
+      "step": 3490
+    },
+    {
+      "epoch": 8.25,
+      "learning_rate": 5.942720763723151e-05,
+      "loss": 0.0052,
+      "step": 3500
+    },
+    {
+      "epoch": 8.28,
+      "learning_rate": 5.930787589498808e-05,
+      "loss": 0.0065,
+      "step": 3510
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 5.918854415274463e-05,
+      "loss": 0.0048,
+      "step": 3520
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 5.9069212410501195e-05,
+      "loss": 0.0093,
+      "step": 3530
+    },
+    {
+      "epoch": 8.35,
+      "learning_rate": 5.894988066825776e-05,
+      "loss": 0.0062,
+      "step": 3540
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 5.8830548926014314e-05,
+      "loss": 0.0075,
+      "step": 3550
+    },
+    {
+      "epoch": 8.4,
+      "learning_rate": 5.871121718377088e-05,
+      "loss": 0.0087,
+      "step": 3560
+    },
+    {
+      "epoch": 8.42,
+      "learning_rate": 5.859188544152745e-05,
+      "loss": 0.0086,
+      "step": 3570
+    },
+    {
+      "epoch": 8.44,
+      "learning_rate": 5.847255369928402e-05,
+      "loss": 0.0074,
+      "step": 3580
+    },
+    {
+      "epoch": 8.47,
+      "learning_rate": 5.835322195704057e-05,
+      "loss": 0.0063,
+      "step": 3590
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 5.823389021479714e-05,
+      "loss": 0.0093,
+      "step": 3600
+    },
+    {
+      "epoch": 8.51,
+      "learning_rate": 5.8114558472553704e-05,
+      "loss": 0.0075,
+      "step": 3610
+    },
+    {
+      "epoch": 8.54,
+      "learning_rate": 5.799522673031027e-05,
+      "loss": 0.0076,
+      "step": 3620
+    },
+    {
+      "epoch": 8.56,
+      "learning_rate": 5.787589498806682e-05,
+      "loss": 0.014,
+      "step": 3630
+    },
+    {
+      "epoch": 8.58,
+      "learning_rate": 5.775656324582339e-05,
+      "loss": 0.0099,
+      "step": 3640
+    },
+    {
+      "epoch": 8.61,
+      "learning_rate": 5.7637231503579955e-05,
+      "loss": 0.0074,
+      "step": 3650
+    },
+    {
+      "epoch": 8.63,
+      "learning_rate": 5.7517899761336514e-05,
+      "loss": 0.0071,
+      "step": 3660
+    },
+    {
+      "epoch": 8.66,
+      "learning_rate": 5.739856801909308e-05,
+      "loss": 0.0116,
+      "step": 3670
+    },
+    {
+      "epoch": 8.68,
+      "learning_rate": 5.7279236276849646e-05,
+      "loss": 0.0078,
+      "step": 3680
+    },
+    {
+      "epoch": 8.7,
+      "learning_rate": 5.715990453460621e-05,
+      "loss": 0.0069,
+      "step": 3690
+    },
+    {
+      "epoch": 8.73,
+      "learning_rate": 5.7040572792362765e-05,
+      "loss": 0.0081,
+      "step": 3700
+    },
+    {
+      "epoch": 8.75,
+      "learning_rate": 5.692124105011933e-05,
+      "loss": 0.0084,
+      "step": 3710
+    },
+    {
+      "epoch": 8.77,
+      "learning_rate": 5.68019093078759e-05,
+      "loss": 0.0098,
+      "step": 3720
+    },
+    {
+      "epoch": 8.8,
+      "learning_rate": 5.668257756563246e-05,
+      "loss": 0.007,
+      "step": 3730
+    },
+    {
+      "epoch": 8.82,
+      "learning_rate": 5.656324582338902e-05,
+      "loss": 0.007,
+      "step": 3740
+    },
+    {
+      "epoch": 8.84,
+      "learning_rate": 5.644391408114559e-05,
+      "loss": 0.0133,
+      "step": 3750
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 5.6324582338902155e-05,
+      "loss": 0.008,
+      "step": 3760
+    },
+    {
+      "epoch": 8.89,
+      "learning_rate": 5.620525059665871e-05,
+      "loss": 0.0079,
+      "step": 3770
+    },
+    {
+      "epoch": 8.92,
+      "learning_rate": 5.6085918854415273e-05,
+      "loss": 0.0061,
+      "step": 3780
+    },
+    {
+      "epoch": 8.94,
+      "learning_rate": 5.596658711217184e-05,
+      "loss": 0.0068,
+      "step": 3790
+    },
+    {
+      "epoch": 8.96,
+      "learning_rate": 5.5847255369928406e-05,
+      "loss": 0.0105,
+      "step": 3800
+    },
+    {
+      "epoch": 8.99,
+      "learning_rate": 5.5727923627684965e-05,
+      "loss": 0.0052,
+      "step": 3810
+    },
+    {
+      "epoch": 9.0,
+      "eval_loss": 2.0036983489990234,
+      "eval_runtime": 26.569,
+      "eval_samples_per_second": 28.341,
+      "eval_steps_per_second": 1.807,
+      "step": 3816
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 5.560859188544153e-05,
+      "loss": 0.0083,
+      "step": 3820
+    },
+    {
+      "epoch": 9.03,
+      "learning_rate": 5.54892601431981e-05,
+      "loss": 0.0041,
+      "step": 3830
+    },
+    {
+      "epoch": 9.06,
+      "learning_rate": 5.536992840095465e-05,
+      "loss": 0.0046,
+      "step": 3840
+    },
+    {
+      "epoch": 9.08,
+      "learning_rate": 5.5250596658711216e-05,
+      "loss": 0.0062,
+      "step": 3850
+    },
+    {
+      "epoch": 9.1,
+      "learning_rate": 5.513126491646778e-05,
+      "loss": 0.0039,
+      "step": 3860
+    },
+    {
+      "epoch": 9.13,
+      "learning_rate": 5.501193317422435e-05,
+      "loss": 0.0041,
+      "step": 3870
+    },
+    {
+      "epoch": 9.15,
+      "learning_rate": 5.489260143198091e-05,
+      "loss": 0.0072,
+      "step": 3880
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 5.4773269689737473e-05,
+      "loss": 0.0088,
+      "step": 3890
+    },
+    {
+      "epoch": 9.2,
+      "learning_rate": 5.465393794749404e-05,
+      "loss": 0.0054,
+      "step": 3900
+    },
+    {
+      "epoch": 9.22,
+      "learning_rate": 5.4534606205250606e-05,
+      "loss": 0.0056,
+      "step": 3910
+    },
+    {
+      "epoch": 9.25,
+      "learning_rate": 5.441527446300716e-05,
+      "loss": 0.0035,
+      "step": 3920
+    },
+    {
+      "epoch": 9.27,
+      "learning_rate": 5.4295942720763724e-05,
+      "loss": 0.005,
+      "step": 3930
+    },
+    {
+      "epoch": 9.29,
+      "learning_rate": 5.417661097852029e-05,
+      "loss": 0.0038,
+      "step": 3940
+    },
+    {
+      "epoch": 9.32,
+      "learning_rate": 5.405727923627685e-05,
+      "loss": 0.0038,
+      "step": 3950
+    },
+    {
+      "epoch": 9.34,
+      "learning_rate": 5.3937947494033416e-05,
+      "loss": 0.006,
+      "step": 3960
+    },
+    {
+      "epoch": 9.36,
+      "learning_rate": 5.381861575178998e-05,
+      "loss": 0.0035,
+      "step": 3970
+    },
+    {
+      "epoch": 9.39,
+      "learning_rate": 5.369928400954655e-05,
+      "loss": 0.0046,
+      "step": 3980
+    },
+    {
+      "epoch": 9.41,
+      "learning_rate": 5.35799522673031e-05,
+      "loss": 0.0031,
+      "step": 3990
+    },
+    {
+      "epoch": 9.43,
+      "learning_rate": 5.346062052505967e-05,
+      "loss": 0.0044,
+      "step": 4000
+    },
+    {
+      "epoch": 9.46,
+      "learning_rate": 5.334128878281623e-05,
+      "loss": 0.003,
+      "step": 4010
+    },
+    {
+      "epoch": 9.48,
+      "learning_rate": 5.322195704057279e-05,
+      "loss": 0.0043,
+      "step": 4020
+    },
+    {
+      "epoch": 9.5,
+      "learning_rate": 5.310262529832936e-05,
+      "loss": 0.008,
+      "step": 4030
+    },
+    {
+      "epoch": 9.53,
+      "learning_rate": 5.2983293556085924e-05,
+      "loss": 0.0045,
+      "step": 4040
+    },
+    {
+      "epoch": 9.55,
+      "learning_rate": 5.286396181384249e-05,
+      "loss": 0.006,
+      "step": 4050
+    },
+    {
+      "epoch": 9.58,
+      "learning_rate": 5.274463007159904e-05,
+      "loss": 0.0053,
+      "step": 4060
+    },
+    {
+      "epoch": 9.6,
+      "learning_rate": 5.262529832935561e-05,
+      "loss": 0.006,
+      "step": 4070
+    },
+    {
+      "epoch": 9.62,
+      "learning_rate": 5.2505966587112175e-05,
+      "loss": 0.004,
+      "step": 4080
+    },
+    {
+      "epoch": 9.65,
+      "learning_rate": 5.238663484486874e-05,
+      "loss": 0.0047,
+      "step": 4090
+    },
+    {
+      "epoch": 9.67,
+      "learning_rate": 5.22673031026253e-05,
+      "loss": 0.0037,
+      "step": 4100
+    },
+    {
+      "epoch": 9.69,
+      "learning_rate": 5.214797136038187e-05,
+      "loss": 0.0036,
+      "step": 4110
+    },
+    {
+      "epoch": 9.72,
+      "learning_rate": 5.202863961813843e-05,
+      "loss": 0.0039,
+      "step": 4120
+    },
+    {
+      "epoch": 9.74,
+      "learning_rate": 5.1909307875894985e-05,
+      "loss": 0.0063,
+      "step": 4130
+    },
+    {
+      "epoch": 9.76,
+      "learning_rate": 5.178997613365155e-05,
+      "loss": 0.0047,
+      "step": 4140
+    },
+    {
+      "epoch": 9.79,
+      "learning_rate": 5.167064439140812e-05,
+      "loss": 0.0048,
+      "step": 4150
+    },
+    {
+      "epoch": 9.81,
+      "learning_rate": 5.1551312649164684e-05,
+      "loss": 0.0045,
+      "step": 4160
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 5.143198090692124e-05,
+      "loss": 0.0093,
+      "step": 4170
+    },
+    {
+      "epoch": 9.86,
+      "learning_rate": 5.131264916467781e-05,
+      "loss": 0.0056,
+      "step": 4180
+    },
+    {
+      "epoch": 9.88,
+      "learning_rate": 5.1193317422434375e-05,
+      "loss": 0.0056,
+      "step": 4190
+    },
+    {
+      "epoch": 9.91,
+      "learning_rate": 5.107398568019093e-05,
+      "loss": 0.0057,
+      "step": 4200
+    },
+    {
+      "epoch": 9.93,
+      "learning_rate": 5.0954653937947494e-05,
+      "loss": 0.0051,
+      "step": 4210
+    },
+    {
+      "epoch": 9.95,
+      "learning_rate": 5.083532219570406e-05,
+      "loss": 0.0044,
+      "step": 4220
+    },
+    {
+      "epoch": 9.98,
+      "learning_rate": 5.0715990453460626e-05,
+      "loss": 0.0067,
+      "step": 4230
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 5.059665871121718e-05,
+      "loss": 0.0055,
+      "step": 4240
+    },
+    {
+      "epoch": 10.0,
+      "eval_loss": 2.0297234058380127,
+      "eval_runtime": 31.7229,
+      "eval_samples_per_second": 23.737,
+      "eval_steps_per_second": 1.513,
+      "step": 4240
+    },
+    {
+      "epoch": 10.02,
+      "learning_rate": 5.047732696897375e-05,
+      "loss": 0.0072,
+      "step": 4250
+    },
+    {
+      "epoch": 10.05,
+      "learning_rate": 5.035799522673032e-05,
+      "loss": 0.0043,
+      "step": 4260
+    },
+    {
+      "epoch": 10.07,
+      "learning_rate": 5.0238663484486884e-05,
+      "loss": 0.0054,
+      "step": 4270
+    },
+    {
+      "epoch": 10.09,
+      "learning_rate": 5.0119331742243436e-05,
+      "loss": 0.0025,
+      "step": 4280
+    },
+    {
+      "epoch": 10.12,
+      "learning_rate": 5e-05,
+      "loss": 0.0037,
+      "step": 4290
+    },
+    {
+      "epoch": 10.14,
+      "learning_rate": 4.988066825775656e-05,
+      "loss": 0.0034,
+      "step": 4300
+    },
+    {
+      "epoch": 10.17,
+      "learning_rate": 4.976133651551313e-05,
+      "loss": 0.0075,
+      "step": 4310
+    },
+    {
+      "epoch": 10.19,
+      "learning_rate": 4.964200477326969e-05,
+      "loss": 0.0036,
+      "step": 4320
+    },
+    {
+      "epoch": 10.21,
+      "learning_rate": 4.952267303102625e-05,
+      "loss": 0.0035,
+      "step": 4330
+    },
+    {
+      "epoch": 10.24,
+      "learning_rate": 4.940334128878282e-05,
+      "loss": 0.0054,
+      "step": 4340
+    },
+    {
+      "epoch": 10.26,
+      "learning_rate": 4.9284009546539385e-05,
+      "loss": 0.0025,
+      "step": 4350
+    },
+    {
+      "epoch": 10.28,
+      "learning_rate": 4.9164677804295945e-05,
+      "loss": 0.0042,
+      "step": 4360
+    },
+    {
+      "epoch": 10.31,
+      "learning_rate": 4.904534606205251e-05,
+      "loss": 0.0026,
+      "step": 4370
+    },
+    {
+      "epoch": 10.33,
+      "learning_rate": 4.892601431980907e-05,
+      "loss": 0.0049,
+      "step": 4380
+    },
+    {
+      "epoch": 10.35,
+      "learning_rate": 4.880668257756563e-05,
+      "loss": 0.0032,
+      "step": 4390
+    },
+    {
+      "epoch": 10.38,
+      "learning_rate": 4.8687350835322196e-05,
+      "loss": 0.0031,
+      "step": 4400
+    },
+    {
+      "epoch": 10.4,
+      "learning_rate": 4.856801909307876e-05,
+      "loss": 0.0023,
+      "step": 4410
+    },
+    {
+      "epoch": 10.42,
+      "learning_rate": 4.844868735083533e-05,
+      "loss": 0.0034,
+      "step": 4420
+    },
+    {
+      "epoch": 10.45,
+      "learning_rate": 4.832935560859189e-05,
+      "loss": 0.0029,
+      "step": 4430
+    },
+    {
+      "epoch": 10.47,
+      "learning_rate": 4.821002386634845e-05,
+      "loss": 0.0032,
+      "step": 4440
+    },
+    {
+      "epoch": 10.5,
+      "learning_rate": 4.809069212410501e-05,
+      "loss": 0.0051,
+      "step": 4450
+    },
+    {
+      "epoch": 10.52,
+      "learning_rate": 4.797136038186158e-05,
+      "loss": 0.0058,
+      "step": 4460
+    },
+    {
+      "epoch": 10.54,
+      "learning_rate": 4.785202863961814e-05,
+      "loss": 0.0029,
+      "step": 4470
+    },
+    {
+      "epoch": 10.57,
+      "learning_rate": 4.7732696897374704e-05,
+      "loss": 0.0034,
+      "step": 4480
+    },
+    {
+      "epoch": 10.59,
+      "learning_rate": 4.761336515513127e-05,
+      "loss": 0.0027,
+      "step": 4490
+    },
+    {
+      "epoch": 10.61,
+      "learning_rate": 4.749403341288783e-05,
+      "loss": 0.0022,
+      "step": 4500
+    },
+    {
+      "epoch": 10.64,
+      "learning_rate": 4.7374701670644396e-05,
+      "loss": 0.003,
+      "step": 4510
+    },
+    {
+      "epoch": 10.66,
+      "learning_rate": 4.7255369928400955e-05,
+      "loss": 0.0042,
+      "step": 4520
+    },
+    {
+      "epoch": 10.68,
+      "learning_rate": 4.713603818615752e-05,
+      "loss": 0.003,
+      "step": 4530
+    },
+    {
+      "epoch": 10.71,
+      "learning_rate": 4.701670644391408e-05,
+      "loss": 0.0023,
+      "step": 4540
+    },
+    {
+      "epoch": 10.73,
+      "learning_rate": 4.6897374701670647e-05,
+      "loss": 0.0055,
+      "step": 4550
+    },
+    {
+      "epoch": 10.75,
+      "learning_rate": 4.677804295942721e-05,
+      "loss": 0.0023,
+      "step": 4560
+    },
+    {
+      "epoch": 10.78,
+      "learning_rate": 4.665871121718377e-05,
+      "loss": 0.0038,
+      "step": 4570
+    },
+    {
+      "epoch": 10.8,
+      "learning_rate": 4.653937947494034e-05,
+      "loss": 0.0041,
+      "step": 4580
+    },
+    {
+      "epoch": 10.83,
+      "learning_rate": 4.64200477326969e-05,
+      "loss": 0.0043,
+      "step": 4590
+    },
+    {
+      "epoch": 10.85,
+      "learning_rate": 4.6300715990453463e-05,
+      "loss": 0.0045,
+      "step": 4600
+    },
+    {
+      "epoch": 10.87,
+      "learning_rate": 4.618138424821002e-05,
+      "loss": 0.0031,
+      "step": 4610
+    },
+    {
+      "epoch": 10.9,
+      "learning_rate": 4.606205250596659e-05,
+      "loss": 0.0039,
+      "step": 4620
+    },
+    {
+      "epoch": 10.92,
+      "learning_rate": 4.594272076372315e-05,
+      "loss": 0.0065,
+      "step": 4630
+    },
+    {
+      "epoch": 10.94,
+      "learning_rate": 4.582338902147972e-05,
+      "loss": 0.0032,
+      "step": 4640
+    },
+    {
+      "epoch": 10.97,
+      "learning_rate": 4.570405727923628e-05,
+      "loss": 0.003,
+      "step": 4650
+    },
+    {
+      "epoch": 10.99,
+      "learning_rate": 4.5584725536992847e-05,
+      "loss": 0.0042,
+      "step": 4660
+    },
+    {
+      "epoch": 11.0,
+      "eval_loss": 2.059401273727417,
+      "eval_runtime": 26.8366,
+      "eval_samples_per_second": 28.059,
+      "eval_steps_per_second": 1.789,
+      "step": 4664
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 4.5465393794749406e-05,
+      "loss": 0.0019,
+      "step": 4670
+    },
+    {
+      "epoch": 11.04,
+      "learning_rate": 4.5346062052505965e-05,
+      "loss": 0.0021,
+      "step": 4680
+    },
+    {
+      "epoch": 11.06,
+      "learning_rate": 4.522673031026253e-05,
+      "loss": 0.0019,
+      "step": 4690
+    },
+    {
+      "epoch": 11.08,
+      "learning_rate": 4.510739856801909e-05,
+      "loss": 0.0023,
+      "step": 4700
+    },
+    {
+      "epoch": 11.11,
+      "learning_rate": 4.498806682577566e-05,
+      "loss": 0.0017,
+      "step": 4710
+    },
+    {
+      "epoch": 11.13,
+      "learning_rate": 4.486873508353222e-05,
+      "loss": 0.0016,
+      "step": 4720
+    },
+    {
+      "epoch": 11.16,
+      "learning_rate": 4.474940334128879e-05,
+      "loss": 0.0029,
+      "step": 4730
+    },
+    {
+      "epoch": 11.18,
+      "learning_rate": 4.463007159904535e-05,
+      "loss": 0.0017,
+      "step": 4740
+    },
+    {
+      "epoch": 11.2,
+      "learning_rate": 4.4510739856801914e-05,
+      "loss": 0.0018,
+      "step": 4750
+    },
+    {
+      "epoch": 11.23,
+      "learning_rate": 4.4391408114558474e-05,
+      "loss": 0.0015,
+      "step": 4760
+    },
+    {
+      "epoch": 11.25,
+      "learning_rate": 4.427207637231503e-05,
+      "loss": 0.0019,
+      "step": 4770
+    },
+    {
+      "epoch": 11.27,
+      "learning_rate": 4.41527446300716e-05,
+      "loss": 0.0032,
+      "step": 4780
+    },
+    {
+      "epoch": 11.3,
+      "learning_rate": 4.4033412887828165e-05,
+      "loss": 0.002,
+      "step": 4790
+    },
+    {
+      "epoch": 11.32,
+      "learning_rate": 4.391408114558473e-05,
+      "loss": 0.004,
+      "step": 4800
+    },
+    {
+      "epoch": 11.34,
+      "learning_rate": 4.379474940334129e-05,
+      "loss": 0.0016,
+      "step": 4810
+    },
+    {
+      "epoch": 11.37,
+      "learning_rate": 4.367541766109786e-05,
+      "loss": 0.0029,
+      "step": 4820
+    },
+    {
+      "epoch": 11.39,
+      "learning_rate": 4.3556085918854416e-05,
+      "loss": 0.0015,
+      "step": 4830
+    },
+    {
+      "epoch": 11.42,
+      "learning_rate": 4.343675417661098e-05,
+      "loss": 0.0016,
+      "step": 4840
+    },
+    {
+      "epoch": 11.44,
+      "learning_rate": 4.331742243436754e-05,
+      "loss": 0.0027,
+      "step": 4850
+    },
+    {
+      "epoch": 11.46,
+      "learning_rate": 4.319809069212411e-05,
+      "loss": 0.0017,
+      "step": 4860
+    },
+    {
+      "epoch": 11.49,
+      "learning_rate": 4.3078758949880674e-05,
+      "loss": 0.0033,
+      "step": 4870
+    },
+    {
+      "epoch": 11.51,
+      "learning_rate": 4.295942720763723e-05,
+      "loss": 0.0027,
+      "step": 4880
+    },
+    {
+      "epoch": 11.53,
+      "learning_rate": 4.28400954653938e-05,
+      "loss": 0.002,
+      "step": 4890
+    },
+    {
+      "epoch": 11.56,
+      "learning_rate": 4.272076372315036e-05,
+      "loss": 0.0018,
+      "step": 4900
+    },
+    {
+      "epoch": 11.58,
+      "learning_rate": 4.2601431980906925e-05,
+      "loss": 0.002,
+      "step": 4910
+    },
+    {
+      "epoch": 11.6,
+      "learning_rate": 4.2482100238663484e-05,
+      "loss": 0.0032,
+      "step": 4920
+    },
+    {
+      "epoch": 11.63,
+      "learning_rate": 4.236276849642005e-05,
+      "loss": 0.0019,
+      "step": 4930
+    },
+    {
+      "epoch": 11.65,
+      "learning_rate": 4.2243436754176616e-05,
+      "loss": 0.0039,
+      "step": 4940
+    },
+    {
+      "epoch": 11.67,
+      "learning_rate": 4.2124105011933175e-05,
+      "loss": 0.003,
+      "step": 4950
+    },
+    {
+      "epoch": 11.7,
+      "learning_rate": 4.200477326968974e-05,
+      "loss": 0.0071,
+      "step": 4960
+    },
+    {
+      "epoch": 11.72,
+      "learning_rate": 4.18854415274463e-05,
+      "loss": 0.0025,
+      "step": 4970
+    },
+    {
+      "epoch": 11.75,
+      "learning_rate": 4.176610978520287e-05,
+      "loss": 0.0021,
+      "step": 4980
+    },
+    {
+      "epoch": 11.77,
+      "learning_rate": 4.1646778042959426e-05,
+      "loss": 0.0035,
+      "step": 4990
+    },
+    {
+      "epoch": 11.79,
+      "learning_rate": 4.152744630071599e-05,
+      "loss": 0.0026,
+      "step": 5000
+    },
+    {
+      "epoch": 11.82,
+      "learning_rate": 4.140811455847255e-05,
+      "loss": 0.0059,
+      "step": 5010
+    },
+    {
+      "epoch": 11.84,
+      "learning_rate": 4.1288782816229125e-05,
+      "loss": 0.0034,
+      "step": 5020
+    },
+    {
+      "epoch": 11.86,
+      "learning_rate": 4.1169451073985684e-05,
+      "loss": 0.0028,
+      "step": 5030
+    },
+    {
+      "epoch": 11.89,
+      "learning_rate": 4.105011933174224e-05,
+      "loss": 0.0032,
+      "step": 5040
+    },
+    {
+      "epoch": 11.91,
+      "learning_rate": 4.093078758949881e-05,
+      "loss": 0.0022,
+      "step": 5050
+    },
+    {
+      "epoch": 11.93,
+      "learning_rate": 4.081145584725537e-05,
+      "loss": 0.0048,
+      "step": 5060
+    },
+    {
+      "epoch": 11.96,
+      "learning_rate": 4.0692124105011935e-05,
+      "loss": 0.0041,
+      "step": 5070
+    },
+    {
+      "epoch": 11.98,
+      "learning_rate": 4.0572792362768494e-05,
+      "loss": 0.0025,
+      "step": 5080
+    },
+    {
+      "epoch": 12.0,
+      "eval_loss": 2.078885316848755,
+      "eval_runtime": 28.5234,
+      "eval_samples_per_second": 26.399,
+      "eval_steps_per_second": 1.683,
+      "step": 5088
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 4.045346062052506e-05,
+      "loss": 0.0029,
+      "step": 5090
+    },
+    {
+      "epoch": 12.03,
+      "learning_rate": 4.0334128878281626e-05,
+      "loss": 0.0023,
+      "step": 5100
+    },
+    {
+      "epoch": 12.05,
+      "learning_rate": 4.021479713603819e-05,
+      "loss": 0.0019,
+      "step": 5110
+    },
+    {
+      "epoch": 12.08,
+      "learning_rate": 4.009546539379475e-05,
+      "loss": 0.0016,
+      "step": 5120
+    },
+    {
+      "epoch": 12.1,
+      "learning_rate": 3.997613365155131e-05,
+      "loss": 0.0031,
+      "step": 5130
+    },
+    {
+      "epoch": 12.12,
+      "learning_rate": 3.985680190930788e-05,
+      "loss": 0.0057,
+      "step": 5140
+    },
+    {
+      "epoch": 12.15,
+      "learning_rate": 3.9737470167064437e-05,
+      "loss": 0.0015,
+      "step": 5150
+    },
+    {
+      "epoch": 12.17,
+      "learning_rate": 3.9618138424821e-05,
+      "loss": 0.0017,
+      "step": 5160
+    },
+    {
+      "epoch": 12.19,
+      "learning_rate": 3.949880668257757e-05,
+      "loss": 0.0024,
+      "step": 5170
+    },
+    {
+      "epoch": 12.22,
+      "learning_rate": 3.9379474940334135e-05,
+      "loss": 0.0023,
+      "step": 5180
+    },
+    {
+      "epoch": 12.24,
+      "learning_rate": 3.9260143198090694e-05,
+      "loss": 0.0035,
+      "step": 5190
+    },
+    {
+      "epoch": 12.26,
+      "learning_rate": 3.914081145584726e-05,
+      "loss": 0.0017,
+      "step": 5200
+    },
+    {
+      "epoch": 12.29,
+      "learning_rate": 3.902147971360382e-05,
+      "loss": 0.0016,
+      "step": 5210
+    },
+    {
+      "epoch": 12.31,
+      "learning_rate": 3.8902147971360386e-05,
+      "loss": 0.0016,
+      "step": 5220
+    },
+    {
+      "epoch": 12.33,
+      "learning_rate": 3.8782816229116945e-05,
+      "loss": 0.0027,
+      "step": 5230
+    },
+    {
+      "epoch": 12.36,
+      "learning_rate": 3.866348448687351e-05,
+      "loss": 0.0027,
+      "step": 5240
+    },
+    {
+      "epoch": 12.38,
+      "learning_rate": 3.854415274463008e-05,
+      "loss": 0.0016,
+      "step": 5250
+    },
+    {
+      "epoch": 12.41,
+      "learning_rate": 3.8424821002386637e-05,
+      "loss": 0.0017,
+      "step": 5260
+    },
+    {
+      "epoch": 12.43,
+      "learning_rate": 3.83054892601432e-05,
+      "loss": 0.0014,
+      "step": 5270
+    },
+    {
+      "epoch": 12.45,
+      "learning_rate": 3.818615751789976e-05,
+      "loss": 0.0014,
+      "step": 5280
+    },
+    {
+      "epoch": 12.48,
+      "learning_rate": 3.806682577565633e-05,
+      "loss": 0.0022,
+      "step": 5290
+    },
+    {
+      "epoch": 12.5,
+      "learning_rate": 3.794749403341289e-05,
+      "loss": 0.0014,
+      "step": 5300
+    },
+    {
+      "epoch": 12.52,
+      "learning_rate": 3.7828162291169453e-05,
+      "loss": 0.0019,
+      "step": 5310
+    },
+    {
+      "epoch": 12.55,
+      "learning_rate": 3.770883054892602e-05,
+      "loss": 0.0013,
+      "step": 5320
+    },
+    {
+      "epoch": 12.57,
+      "learning_rate": 3.758949880668258e-05,
+      "loss": 0.0016,
+      "step": 5330
+    },
+    {
+      "epoch": 12.59,
+      "learning_rate": 3.7470167064439145e-05,
+      "loss": 0.0024,
+      "step": 5340
+    },
+    {
+      "epoch": 12.62,
+      "learning_rate": 3.7350835322195704e-05,
+      "loss": 0.0016,
+      "step": 5350
+    },
+    {
+      "epoch": 12.64,
+      "learning_rate": 3.723150357995227e-05,
+      "loss": 0.0019,
+      "step": 5360
+    },
+    {
+      "epoch": 12.67,
+      "learning_rate": 3.711217183770883e-05,
+      "loss": 0.0014,
+      "step": 5370
+    },
+    {
+      "epoch": 12.69,
+      "learning_rate": 3.6992840095465396e-05,
+      "loss": 0.0012,
+      "step": 5380
+    },
+    {
+      "epoch": 12.71,
+      "learning_rate": 3.6873508353221955e-05,
+      "loss": 0.0023,
+      "step": 5390
+    },
+    {
+      "epoch": 12.74,
+      "learning_rate": 3.675417661097852e-05,
+      "loss": 0.0012,
+      "step": 5400
+    },
+    {
+      "epoch": 12.76,
+      "learning_rate": 3.663484486873509e-05,
+      "loss": 0.0021,
+      "step": 5410
+    },
+    {
+      "epoch": 12.78,
+      "learning_rate": 3.651551312649165e-05,
+      "loss": 0.0035,
+      "step": 5420
+    },
+    {
+      "epoch": 12.81,
+      "learning_rate": 3.639618138424821e-05,
+      "loss": 0.0028,
+      "step": 5430
+    },
+    {
+      "epoch": 12.83,
+      "learning_rate": 3.627684964200477e-05,
+      "loss": 0.0012,
+      "step": 5440
+    },
+    {
+      "epoch": 12.85,
+      "learning_rate": 3.615751789976134e-05,
+      "loss": 0.002,
+      "step": 5450
+    },
+    {
+      "epoch": 12.88,
+      "learning_rate": 3.60381861575179e-05,
+      "loss": 0.0018,
+      "step": 5460
+    },
+    {
+      "epoch": 12.9,
+      "learning_rate": 3.5918854415274464e-05,
+      "loss": 0.0018,
+      "step": 5470
+    },
+    {
+      "epoch": 12.92,
+      "learning_rate": 3.579952267303103e-05,
+      "loss": 0.0044,
+      "step": 5480
+    },
+    {
+      "epoch": 12.95,
+      "learning_rate": 3.5680190930787596e-05,
+      "loss": 0.0018,
+      "step": 5490
+    },
+    {
+      "epoch": 12.97,
+      "learning_rate": 3.5560859188544155e-05,
+      "loss": 0.0014,
+      "step": 5500
+    },
+    {
+      "epoch": 13.0,
+      "learning_rate": 3.5441527446300715e-05,
+      "loss": 0.0013,
+      "step": 5510
+    },
+    {
+      "epoch": 13.0,
+      "eval_loss": 2.102971315383911,
+      "eval_runtime": 31.002,
+      "eval_samples_per_second": 24.289,
+      "eval_steps_per_second": 1.548,
+      "step": 5512
+    },
+    {
+      "epoch": 13.02,
+      "learning_rate": 3.532219570405728e-05,
+      "loss": 0.0025,
+      "step": 5520
+    },
+    {
+      "epoch": 13.04,
+      "learning_rate": 3.520286396181384e-05,
+      "loss": 0.0011,
+      "step": 5530
+    },
+    {
+      "epoch": 13.07,
+      "learning_rate": 3.5083532219570406e-05,
+      "loss": 0.0013,
+      "step": 5540
+    },
+    {
+      "epoch": 13.09,
+      "learning_rate": 3.496420047732697e-05,
+      "loss": 0.0015,
+      "step": 5550
+    },
+    {
+      "epoch": 13.11,
+      "learning_rate": 3.484486873508354e-05,
+      "loss": 0.001,
+      "step": 5560
+    },
+    {
+      "epoch": 13.14,
+      "learning_rate": 3.47255369928401e-05,
+      "loss": 0.0015,
+      "step": 5570
+    },
+    {
+      "epoch": 13.16,
+      "learning_rate": 3.4606205250596664e-05,
+      "loss": 0.0045,
+      "step": 5580
+    },
+    {
+      "epoch": 13.18,
+      "learning_rate": 3.448687350835322e-05,
+      "loss": 0.001,
+      "step": 5590
+    },
+    {
+      "epoch": 13.21,
+      "learning_rate": 3.436754176610978e-05,
+      "loss": 0.0011,
+      "step": 5600
+    },
+    {
+      "epoch": 13.23,
+      "learning_rate": 3.424821002386635e-05,
+      "loss": 0.0022,
+      "step": 5610
+    },
+    {
+      "epoch": 13.25,
+      "learning_rate": 3.4128878281622915e-05,
+      "loss": 0.0021,
+      "step": 5620
+    },
+    {
+      "epoch": 13.28,
+      "learning_rate": 3.400954653937948e-05,
+      "loss": 0.0015,
+      "step": 5630
+    },
+    {
+      "epoch": 13.3,
+      "learning_rate": 3.389021479713604e-05,
+      "loss": 0.0025,
+      "step": 5640
+    },
+    {
+      "epoch": 13.33,
+      "learning_rate": 3.3770883054892606e-05,
+      "loss": 0.0012,
+      "step": 5650
+    },
+    {
+      "epoch": 13.35,
+      "learning_rate": 3.3651551312649165e-05,
+      "loss": 0.0012,
+      "step": 5660
+    },
+    {
+      "epoch": 13.37,
+      "learning_rate": 3.353221957040573e-05,
+      "loss": 0.0017,
+      "step": 5670
+    },
+    {
+      "epoch": 13.4,
+      "learning_rate": 3.341288782816229e-05,
+      "loss": 0.0012,
+      "step": 5680
+    },
+    {
+      "epoch": 13.42,
+      "learning_rate": 3.329355608591885e-05,
+      "loss": 0.0014,
+      "step": 5690
+    },
+    {
+      "epoch": 13.44,
+      "learning_rate": 3.3174224343675416e-05,
+      "loss": 0.0015,
+      "step": 5700
+    },
+    {
+      "epoch": 13.47,
+      "learning_rate": 3.305489260143198e-05,
+      "loss": 0.0015,
+      "step": 5710
+    },
+    {
+      "epoch": 13.49,
+      "learning_rate": 3.293556085918855e-05,
+      "loss": 0.0011,
+      "step": 5720
+    },
+    {
+      "epoch": 13.51,
+      "learning_rate": 3.281622911694511e-05,
+      "loss": 0.0016,
+      "step": 5730
+    },
+    {
+      "epoch": 13.54,
+      "learning_rate": 3.2696897374701674e-05,
+      "loss": 0.0033,
+      "step": 5740
+    },
+    {
+      "epoch": 13.56,
+      "learning_rate": 3.257756563245823e-05,
+      "loss": 0.0012,
+      "step": 5750
+    },
+    {
+      "epoch": 13.58,
+      "learning_rate": 3.24582338902148e-05,
+      "loss": 0.0012,
+      "step": 5760
+    },
+    {
+      "epoch": 13.61,
+      "learning_rate": 3.233890214797136e-05,
+      "loss": 0.0021,
+      "step": 5770
+    },
+    {
+      "epoch": 13.63,
+      "learning_rate": 3.2219570405727925e-05,
+      "loss": 0.0011,
+      "step": 5780
+    },
+    {
+      "epoch": 13.66,
+      "learning_rate": 3.210023866348449e-05,
+      "loss": 0.0012,
+      "step": 5790
+    },
+    {
+      "epoch": 13.68,
+      "learning_rate": 3.198090692124105e-05,
+      "loss": 0.0012,
+      "step": 5800
+    },
+    {
+      "epoch": 13.7,
+      "learning_rate": 3.1861575178997616e-05,
+      "loss": 0.0011,
+      "step": 5810
+    },
+    {
+      "epoch": 13.73,
+      "learning_rate": 3.1742243436754176e-05,
+      "loss": 0.0011,
+      "step": 5820
+    },
+    {
+      "epoch": 13.75,
+      "learning_rate": 3.162291169451074e-05,
+      "loss": 0.001,
+      "step": 5830
+    },
+    {
+      "epoch": 13.77,
+      "learning_rate": 3.15035799522673e-05,
+      "loss": 0.0021,
+      "step": 5840
+    },
+    {
+      "epoch": 13.8,
+      "learning_rate": 3.138424821002387e-05,
+      "loss": 0.001,
+      "step": 5850
+    },
+    {
+      "epoch": 13.82,
+      "learning_rate": 3.126491646778043e-05,
+      "loss": 0.0016,
+      "step": 5860
+    },
+    {
+      "epoch": 13.84,
+      "learning_rate": 3.1145584725537e-05,
+      "loss": 0.001,
+      "step": 5870
+    },
+    {
+      "epoch": 13.87,
+      "learning_rate": 3.102625298329356e-05,
+      "loss": 0.0013,
+      "step": 5880
+    },
+    {
+      "epoch": 13.89,
+      "learning_rate": 3.090692124105012e-05,
+      "loss": 0.0016,
+      "step": 5890
+    },
+    {
+      "epoch": 13.92,
+      "learning_rate": 3.0787589498806684e-05,
+      "loss": 0.0014,
+      "step": 5900
+    },
+    {
+      "epoch": 13.94,
+      "learning_rate": 3.0668257756563243e-05,
+      "loss": 0.0011,
+      "step": 5910
+    },
+    {
+      "epoch": 13.96,
+      "learning_rate": 3.054892601431981e-05,
+      "loss": 0.0012,
+      "step": 5920
+    },
+    {
+      "epoch": 13.99,
+      "learning_rate": 3.0429594272076372e-05,
+      "loss": 0.0011,
+      "step": 5930
+    },
+    {
+      "epoch": 14.0,
+      "eval_loss": 2.127690553665161,
+      "eval_runtime": 27.0511,
+      "eval_samples_per_second": 27.836,
+      "eval_steps_per_second": 1.774,
+      "step": 5936
+    },
+    {
+      "epoch": 14.01,
+      "learning_rate": 3.031026252983294e-05,
+      "loss": 0.0016,
+      "step": 5940
+    },
+    {
+      "epoch": 14.03,
+      "learning_rate": 3.01909307875895e-05,
+      "loss": 0.0009,
+      "step": 5950
+    },
+    {
+      "epoch": 14.06,
+      "learning_rate": 3.0071599045346067e-05,
+      "loss": 0.0009,
+      "step": 5960
+    },
+    {
+      "epoch": 14.08,
+      "learning_rate": 2.9952267303102627e-05,
+      "loss": 0.0009,
+      "step": 5970
+    },
+    {
+      "epoch": 14.1,
+      "learning_rate": 2.983293556085919e-05,
+      "loss": 0.0017,
+      "step": 5980
+    },
+    {
+      "epoch": 14.13,
+      "learning_rate": 2.9713603818615755e-05,
+      "loss": 0.0008,
+      "step": 5990
+    },
+    {
+      "epoch": 14.15,
+      "learning_rate": 2.9594272076372315e-05,
+      "loss": 0.0041,
+      "step": 6000
+    },
+    {
+      "epoch": 14.17,
+      "learning_rate": 2.947494033412888e-05,
+      "loss": 0.0008,
+      "step": 6010
+    },
+    {
+      "epoch": 14.2,
+      "learning_rate": 2.935560859188544e-05,
+      "loss": 0.0009,
+      "step": 6020
+    },
+    {
+      "epoch": 14.22,
+      "learning_rate": 2.923627684964201e-05,
+      "loss": 0.0013,
+      "step": 6030
+    },
+    {
+      "epoch": 14.25,
+      "learning_rate": 2.911694510739857e-05,
+      "loss": 0.0015,
+      "step": 6040
+    },
+    {
+      "epoch": 14.27,
+      "learning_rate": 2.8997613365155135e-05,
+      "loss": 0.0009,
+      "step": 6050
+    },
+    {
+      "epoch": 14.29,
+      "learning_rate": 2.8878281622911694e-05,
+      "loss": 0.0009,
+      "step": 6060
+    },
+    {
+      "epoch": 14.32,
+      "learning_rate": 2.8758949880668257e-05,
+      "loss": 0.0009,
+      "step": 6070
+    },
+    {
+      "epoch": 14.34,
+      "learning_rate": 2.8639618138424823e-05,
+      "loss": 0.0011,
+      "step": 6080
+    },
+    {
+      "epoch": 14.36,
+      "learning_rate": 2.8520286396181382e-05,
+      "loss": 0.0009,
+      "step": 6090
+    },
+    {
+      "epoch": 14.39,
+      "learning_rate": 2.840095465393795e-05,
+      "loss": 0.0009,
+      "step": 6100
+    },
+    {
+      "epoch": 14.41,
+      "learning_rate": 2.828162291169451e-05,
+      "loss": 0.0008,
+      "step": 6110
+    },
+    {
+      "epoch": 14.43,
+      "learning_rate": 2.8162291169451077e-05,
+      "loss": 0.0014,
+      "step": 6120
+    },
+    {
+      "epoch": 14.46,
+      "learning_rate": 2.8042959427207637e-05,
+      "loss": 0.0009,
+      "step": 6130
+    },
+    {
+      "epoch": 14.48,
+      "learning_rate": 2.7923627684964203e-05,
+      "loss": 0.001,
+      "step": 6140
+    },
+    {
+      "epoch": 14.5,
+      "learning_rate": 2.7804295942720766e-05,
+      "loss": 0.0008,
+      "step": 6150
+    },
+    {
+      "epoch": 14.53,
+      "learning_rate": 2.7684964200477325e-05,
+      "loss": 0.0008,
+      "step": 6160
+    },
+    {
+      "epoch": 14.55,
+      "learning_rate": 2.756563245823389e-05,
+      "loss": 0.0008,
+      "step": 6170
+    },
+    {
+      "epoch": 14.58,
+      "learning_rate": 2.7446300715990454e-05,
+      "loss": 0.0009,
+      "step": 6180
+    },
+    {
+      "epoch": 14.6,
+      "learning_rate": 2.732696897374702e-05,
+      "loss": 0.0009,
+      "step": 6190
+    },
+    {
+      "epoch": 14.62,
+      "learning_rate": 2.720763723150358e-05,
+      "loss": 0.0009,
+      "step": 6200
+    },
+    {
+      "epoch": 14.65,
+      "learning_rate": 2.7088305489260145e-05,
+      "loss": 0.0013,
+      "step": 6210
+    },
+    {
+      "epoch": 14.67,
+      "learning_rate": 2.6968973747016708e-05,
+      "loss": 0.0008,
+      "step": 6220
+    },
+    {
+      "epoch": 14.69,
+      "learning_rate": 2.6849642004773274e-05,
+      "loss": 0.0008,
+      "step": 6230
+    },
+    {
+      "epoch": 14.72,
+      "learning_rate": 2.6730310262529833e-05,
+      "loss": 0.0012,
+      "step": 6240
+    },
+    {
+      "epoch": 14.74,
+      "learning_rate": 2.6610978520286396e-05,
+      "loss": 0.0008,
+      "step": 6250
+    },
+    {
+      "epoch": 14.76,
+      "learning_rate": 2.6491646778042962e-05,
+      "loss": 0.0015,
+      "step": 6260
+    },
+    {
+      "epoch": 14.79,
+      "learning_rate": 2.637231503579952e-05,
+      "loss": 0.0008,
+      "step": 6270
+    },
+    {
+      "epoch": 14.81,
+      "learning_rate": 2.6252983293556088e-05,
+      "loss": 0.0012,
+      "step": 6280
+    },
+    {
+      "epoch": 14.83,
+      "learning_rate": 2.613365155131265e-05,
+      "loss": 0.0009,
+      "step": 6290
+    },
+    {
+      "epoch": 14.86,
+      "learning_rate": 2.6014319809069216e-05,
+      "loss": 0.0009,
+      "step": 6300
+    },
+    {
+      "epoch": 14.88,
+      "learning_rate": 2.5894988066825776e-05,
+      "loss": 0.0013,
+      "step": 6310
+    },
+    {
+      "epoch": 14.91,
+      "learning_rate": 2.5775656324582342e-05,
+      "loss": 0.0017,
+      "step": 6320
+    },
+    {
+      "epoch": 14.93,
+      "learning_rate": 2.5656324582338905e-05,
+      "loss": 0.0008,
+      "step": 6330
+    },
+    {
+      "epoch": 14.95,
+      "learning_rate": 2.5536992840095464e-05,
+      "loss": 0.0007,
+      "step": 6340
+    },
+    {
+      "epoch": 14.98,
+      "learning_rate": 2.541766109785203e-05,
+      "loss": 0.0008,
+      "step": 6350
+    },
+    {
+      "epoch": 15.0,
+      "learning_rate": 2.529832935560859e-05,
+      "loss": 0.0008,
+      "step": 6360
+    },
+    {
+      "epoch": 15.0,
+      "eval_loss": 2.145920991897583,
+      "eval_runtime": 29.6422,
+      "eval_samples_per_second": 25.403,
+      "eval_steps_per_second": 1.619,
+      "step": 6360
+    },
+    {
+      "epoch": 15.02,
+      "learning_rate": 2.517899761336516e-05,
+      "loss": 0.0011,
+      "step": 6370
+    },
+    {
+      "epoch": 15.05,
+      "learning_rate": 2.5059665871121718e-05,
+      "loss": 0.0008,
+      "step": 6380
+    },
+    {
+      "epoch": 15.07,
+      "learning_rate": 2.494033412887828e-05,
+      "loss": 0.0009,
+      "step": 6390
+    },
+    {
+      "epoch": 15.09,
+      "learning_rate": 2.4821002386634844e-05,
+      "loss": 0.0007,
+      "step": 6400
+    },
+    {
+      "epoch": 15.12,
+      "learning_rate": 2.470167064439141e-05,
+      "loss": 0.0007,
+      "step": 6410
+    },
+    {
+      "epoch": 15.14,
+      "learning_rate": 2.4582338902147972e-05,
+      "loss": 0.0007,
+      "step": 6420
+    },
+    {
+      "epoch": 15.17,
+      "learning_rate": 2.4463007159904535e-05,
+      "loss": 0.0007,
+      "step": 6430
+    },
+    {
+      "epoch": 15.19,
+      "learning_rate": 2.4343675417661098e-05,
+      "loss": 0.0012,
+      "step": 6440
+    },
+    {
+      "epoch": 15.21,
+      "learning_rate": 2.4224343675417664e-05,
+      "loss": 0.0007,
+      "step": 6450
+    },
+    {
+      "epoch": 15.24,
+      "learning_rate": 2.4105011933174227e-05,
+      "loss": 0.0007,
+      "step": 6460
+    },
+    {
+      "epoch": 15.26,
+      "learning_rate": 2.398568019093079e-05,
+      "loss": 0.0007,
+      "step": 6470
+    },
+    {
+      "epoch": 15.28,
+      "learning_rate": 2.3866348448687352e-05,
+      "loss": 0.0006,
+      "step": 6480
+    },
+    {
+      "epoch": 15.31,
+      "learning_rate": 2.3747016706443915e-05,
+      "loss": 0.0007,
+      "step": 6490
+    },
+    {
+      "epoch": 15.33,
+      "learning_rate": 2.3627684964200477e-05,
+      "loss": 0.0007,
+      "step": 6500
+    },
+    {
+      "epoch": 15.35,
+      "learning_rate": 2.350835322195704e-05,
+      "loss": 0.0007,
+      "step": 6510
+    },
+    {
+      "epoch": 15.38,
+      "learning_rate": 2.3389021479713606e-05,
+      "loss": 0.0007,
+      "step": 6520
+    },
+    {
+      "epoch": 15.4,
+      "learning_rate": 2.326968973747017e-05,
+      "loss": 0.0007,
+      "step": 6530
+    },
+    {
+      "epoch": 15.42,
+      "learning_rate": 2.3150357995226732e-05,
+      "loss": 0.0006,
+      "step": 6540
+    },
+    {
+      "epoch": 15.45,
+      "learning_rate": 2.3031026252983294e-05,
+      "loss": 0.0007,
+      "step": 6550
+    },
+    {
+      "epoch": 15.47,
+      "learning_rate": 2.291169451073986e-05,
+      "loss": 0.0007,
+      "step": 6560
+    },
+    {
+      "epoch": 15.5,
+      "learning_rate": 2.2792362768496423e-05,
+      "loss": 0.0014,
+      "step": 6570
+    },
+    {
+      "epoch": 15.52,
+      "learning_rate": 2.2673031026252983e-05,
+      "loss": 0.0007,
+      "step": 6580
+    },
+    {
+      "epoch": 15.54,
+      "learning_rate": 2.2553699284009545e-05,
+      "loss": 0.001,
+      "step": 6590
+    },
+    {
+      "epoch": 15.57,
+      "learning_rate": 2.243436754176611e-05,
+      "loss": 0.0007,
+      "step": 6600
+    },
+    {
+      "epoch": 15.59,
+      "learning_rate": 2.2315035799522674e-05,
+      "loss": 0.0006,
+      "step": 6610
+    },
+    {
+      "epoch": 15.61,
+      "learning_rate": 2.2195704057279237e-05,
+      "loss": 0.0006,
+      "step": 6620
+    },
+    {
+      "epoch": 15.64,
+      "learning_rate": 2.20763723150358e-05,
+      "loss": 0.0011,
+      "step": 6630
+    },
+    {
+      "epoch": 15.66,
+      "learning_rate": 2.1957040572792366e-05,
+      "loss": 0.0006,
+      "step": 6640
+    },
+    {
+      "epoch": 15.68,
+      "learning_rate": 2.183770883054893e-05,
+      "loss": 0.0006,
+      "step": 6650
+    },
+    {
+      "epoch": 15.71,
+      "learning_rate": 2.171837708830549e-05,
+      "loss": 0.0006,
+      "step": 6660
+    },
+    {
+      "epoch": 15.73,
+      "learning_rate": 2.1599045346062054e-05,
+      "loss": 0.0012,
+      "step": 6670
+    },
+    {
+      "epoch": 15.75,
+      "learning_rate": 2.1479713603818617e-05,
+      "loss": 0.0007,
+      "step": 6680
+    },
+    {
+      "epoch": 15.78,
+      "learning_rate": 2.136038186157518e-05,
+      "loss": 0.001,
+      "step": 6690
+    },
+    {
+      "epoch": 15.8,
+      "learning_rate": 2.1241050119331742e-05,
+      "loss": 0.0007,
+      "step": 6700
+    },
+    {
+      "epoch": 15.83,
+      "learning_rate": 2.1121718377088308e-05,
+      "loss": 0.0007,
+      "step": 6710
+    },
+    {
+      "epoch": 15.85,
+      "learning_rate": 2.100238663484487e-05,
+      "loss": 0.0015,
+      "step": 6720
+    },
+    {
+      "epoch": 15.87,
+      "learning_rate": 2.0883054892601433e-05,
+      "loss": 0.0007,
+      "step": 6730
+    },
+    {
+      "epoch": 15.9,
+      "learning_rate": 2.0763723150357996e-05,
+      "loss": 0.0007,
+      "step": 6740
+    },
+    {
+      "epoch": 15.92,
+      "learning_rate": 2.0644391408114562e-05,
+      "loss": 0.0006,
+      "step": 6750
+    },
+    {
+      "epoch": 15.94,
+      "learning_rate": 2.052505966587112e-05,
+      "loss": 0.0007,
+      "step": 6760
+    },
+    {
+      "epoch": 15.97,
+      "learning_rate": 2.0405727923627684e-05,
+      "loss": 0.0008,
+      "step": 6770
+    },
+    {
+      "epoch": 15.99,
+      "learning_rate": 2.0286396181384247e-05,
+      "loss": 0.0006,
+      "step": 6780
+    },
+    {
+      "epoch": 16.0,
+      "eval_loss": 2.1572394371032715,
+      "eval_runtime": 31.4998,
+      "eval_samples_per_second": 23.905,
+      "eval_steps_per_second": 1.524,
+      "step": 6784
+    },
+    {
+      "epoch": 16.01,
+      "learning_rate": 2.0167064439140813e-05,
+      "loss": 0.0008,
+      "step": 6790
+    },
+    {
+      "epoch": 16.04,
+      "learning_rate": 2.0047732696897376e-05,
+      "loss": 0.0006,
+      "step": 6800
+    },
+    {
+      "epoch": 16.06,
+      "learning_rate": 1.992840095465394e-05,
+      "loss": 0.0006,
+      "step": 6810
+    },
+    {
+      "epoch": 16.08,
+      "learning_rate": 1.98090692124105e-05,
+      "loss": 0.0006,
+      "step": 6820
+    },
+    {
+      "epoch": 16.11,
+      "learning_rate": 1.9689737470167067e-05,
+      "loss": 0.0014,
+      "step": 6830
+    },
+    {
+      "epoch": 16.13,
+      "learning_rate": 1.957040572792363e-05,
+      "loss": 0.0006,
+      "step": 6840
+    },
+    {
+      "epoch": 16.16,
+      "learning_rate": 1.9451073985680193e-05,
+      "loss": 0.0006,
+      "step": 6850
+    },
+    {
+      "epoch": 16.18,
+      "learning_rate": 1.9331742243436756e-05,
+      "loss": 0.0006,
+      "step": 6860
+    },
+    {
+      "epoch": 16.2,
+      "learning_rate": 1.9212410501193318e-05,
+      "loss": 0.0006,
+      "step": 6870
+    },
+    {
+      "epoch": 16.23,
+      "learning_rate": 1.909307875894988e-05,
+      "loss": 0.0006,
+      "step": 6880
+    },
+    {
+      "epoch": 16.25,
+      "learning_rate": 1.8973747016706444e-05,
+      "loss": 0.0013,
+      "step": 6890
+    },
+    {
+      "epoch": 16.27,
+      "learning_rate": 1.885441527446301e-05,
+      "loss": 0.0005,
+      "step": 6900
+    },
+    {
+      "epoch": 16.3,
+      "learning_rate": 1.8735083532219573e-05,
+      "loss": 0.0005,
+      "step": 6910
+    },
+    {
+      "epoch": 16.32,
+      "learning_rate": 1.8615751789976135e-05,
+      "loss": 0.0012,
+      "step": 6920
+    },
+    {
+      "epoch": 16.34,
+      "learning_rate": 1.8496420047732698e-05,
+      "loss": 0.0006,
+      "step": 6930
+    },
+    {
+      "epoch": 16.37,
+      "learning_rate": 1.837708830548926e-05,
+      "loss": 0.0006,
+      "step": 6940
+    },
+    {
+      "epoch": 16.39,
+      "learning_rate": 1.8257756563245823e-05,
+      "loss": 0.0006,
+      "step": 6950
+    },
+    {
+      "epoch": 16.42,
+      "learning_rate": 1.8138424821002386e-05,
+      "loss": 0.0006,
+      "step": 6960
+    },
+    {
+      "epoch": 16.44,
+      "learning_rate": 1.801909307875895e-05,
+      "loss": 0.0006,
+      "step": 6970
+    },
+    {
+      "epoch": 16.46,
+      "learning_rate": 1.7899761336515515e-05,
+      "loss": 0.0006,
+      "step": 6980
+    },
+    {
+      "epoch": 16.49,
+      "learning_rate": 1.7780429594272078e-05,
+      "loss": 0.0006,
+      "step": 6990
+    },
+    {
+      "epoch": 16.51,
+      "learning_rate": 1.766109785202864e-05,
+      "loss": 0.0006,
+      "step": 7000
+    },
+    {
+      "epoch": 16.53,
+      "learning_rate": 1.7541766109785203e-05,
+      "loss": 0.002,
+      "step": 7010
+    },
+    {
+      "epoch": 16.56,
+      "learning_rate": 1.742243436754177e-05,
+      "loss": 0.001,
+      "step": 7020
+    },
+    {
+      "epoch": 16.58,
+      "learning_rate": 1.7303102625298332e-05,
+      "loss": 0.0006,
+      "step": 7030
+    },
+    {
+      "epoch": 16.6,
+      "learning_rate": 1.718377088305489e-05,
+      "loss": 0.0006,
+      "step": 7040
+    },
+    {
+      "epoch": 16.63,
+      "learning_rate": 1.7064439140811457e-05,
+      "loss": 0.0006,
+      "step": 7050
+    },
+    {
+      "epoch": 16.65,
+      "learning_rate": 1.694510739856802e-05,
+      "loss": 0.0006,
+      "step": 7060
+    },
+    {
+      "epoch": 16.67,
+      "learning_rate": 1.6825775656324583e-05,
+      "loss": 0.0006,
+      "step": 7070
+    },
+    {
+      "epoch": 16.7,
+      "learning_rate": 1.6706443914081145e-05,
+      "loss": 0.001,
+      "step": 7080
+    },
+    {
+      "epoch": 16.72,
+      "learning_rate": 1.6587112171837708e-05,
+      "loss": 0.0006,
+      "step": 7090
+    },
+    {
+      "epoch": 16.75,
+      "learning_rate": 1.6467780429594274e-05,
+      "loss": 0.0011,
+      "step": 7100
+    },
+    {
+      "epoch": 16.77,
+      "learning_rate": 1.6348448687350837e-05,
+      "loss": 0.0006,
+      "step": 7110
+    },
+    {
+      "epoch": 16.79,
+      "learning_rate": 1.62291169451074e-05,
+      "loss": 0.0006,
+      "step": 7120
+    },
+    {
+      "epoch": 16.82,
+      "learning_rate": 1.6109785202863962e-05,
+      "loss": 0.0006,
+      "step": 7130
+    },
+    {
+      "epoch": 16.84,
+      "learning_rate": 1.5990453460620525e-05,
+      "loss": 0.0006,
+      "step": 7140
+    },
+    {
+      "epoch": 16.86,
+      "learning_rate": 1.5871121718377088e-05,
+      "loss": 0.0006,
+      "step": 7150
+    },
+    {
+      "epoch": 16.89,
+      "learning_rate": 1.575178997613365e-05,
+      "loss": 0.0006,
+      "step": 7160
+    },
+    {
+      "epoch": 16.91,
+      "learning_rate": 1.5632458233890217e-05,
+      "loss": 0.0006,
+      "step": 7170
+    },
+    {
+      "epoch": 16.93,
+      "learning_rate": 1.551312649164678e-05,
+      "loss": 0.0005,
+      "step": 7180
+    },
+    {
+      "epoch": 16.96,
+      "learning_rate": 1.5393794749403342e-05,
+      "loss": 0.0005,
+      "step": 7190
+    },
+    {
+      "epoch": 16.98,
+      "learning_rate": 1.5274463007159905e-05,
+      "loss": 0.0006,
+      "step": 7200
+    },
+    {
+      "epoch": 17.0,
+      "eval_loss": 2.1699986457824707,
+      "eval_runtime": 28.0353,
+      "eval_samples_per_second": 26.859,
+      "eval_steps_per_second": 1.712,
+      "step": 7208
+    },
+    {
+      "epoch": 17.0,
+      "learning_rate": 1.515513126491647e-05,
+      "loss": 0.0005,
+      "step": 7210
+    },
+    {
+      "epoch": 17.03,
+      "learning_rate": 1.5035799522673034e-05,
+      "loss": 0.0005,
+      "step": 7220
+    },
+    {
+      "epoch": 17.05,
+      "learning_rate": 1.4916467780429595e-05,
+      "loss": 0.0005,
+      "step": 7230
+    },
+    {
+      "epoch": 17.08,
+      "learning_rate": 1.4797136038186157e-05,
+      "loss": 0.0005,
+      "step": 7240
+    },
+    {
+      "epoch": 17.1,
+      "learning_rate": 1.467780429594272e-05,
+      "loss": 0.0011,
+      "step": 7250
+    },
+    {
+      "epoch": 17.12,
+      "learning_rate": 1.4558472553699284e-05,
+      "loss": 0.0005,
+      "step": 7260
+    },
+    {
+      "epoch": 17.15,
+      "learning_rate": 1.4439140811455847e-05,
+      "loss": 0.0006,
+      "step": 7270
+    },
+    {
+      "epoch": 17.17,
+      "learning_rate": 1.4319809069212412e-05,
+      "loss": 0.0007,
+      "step": 7280
+    },
+    {
+      "epoch": 17.19,
+      "learning_rate": 1.4200477326968974e-05,
+      "loss": 0.0006,
+      "step": 7290
+    },
+    {
+      "epoch": 17.22,
+      "learning_rate": 1.4081145584725539e-05,
+      "loss": 0.0008,
+      "step": 7300
+    },
+    {
+      "epoch": 17.24,
+      "learning_rate": 1.3961813842482101e-05,
+      "loss": 0.0005,
+      "step": 7310
+    },
+    {
+      "epoch": 17.26,
+      "learning_rate": 1.3842482100238662e-05,
+      "loss": 0.0006,
+      "step": 7320
+    },
+    {
+      "epoch": 17.29,
+      "learning_rate": 1.3723150357995227e-05,
+      "loss": 0.0006,
+      "step": 7330
+    },
+    {
+      "epoch": 17.31,
+      "learning_rate": 1.360381861575179e-05,
+      "loss": 0.0005,
+      "step": 7340
+    },
+    {
+      "epoch": 17.33,
+      "learning_rate": 1.3484486873508354e-05,
+      "loss": 0.0005,
+      "step": 7350
+    },
+    {
+      "epoch": 17.36,
+      "learning_rate": 1.3365155131264917e-05,
+      "loss": 0.0005,
+      "step": 7360
+    },
+    {
+      "epoch": 17.38,
+      "learning_rate": 1.3245823389021481e-05,
+      "loss": 0.0005,
+      "step": 7370
+    },
+    {
+      "epoch": 17.41,
+      "learning_rate": 1.3126491646778044e-05,
+      "loss": 0.0006,
+      "step": 7380
+    },
+    {
+      "epoch": 17.43,
+      "learning_rate": 1.3007159904534608e-05,
+      "loss": 0.0005,
+      "step": 7390
+    },
+    {
+      "epoch": 17.45,
+      "learning_rate": 1.2887828162291171e-05,
+      "loss": 0.002,
+      "step": 7400
+    },
+    {
+      "epoch": 17.48,
+      "learning_rate": 1.2768496420047732e-05,
+      "loss": 0.0005,
+      "step": 7410
+    },
+    {
+      "epoch": 17.5,
+      "learning_rate": 1.2649164677804295e-05,
+      "loss": 0.0005,
+      "step": 7420
+    },
+    {
+      "epoch": 17.52,
+      "learning_rate": 1.2529832935560859e-05,
+      "loss": 0.0006,
+      "step": 7430
+    },
+    {
+      "epoch": 17.55,
+      "learning_rate": 1.2410501193317422e-05,
+      "loss": 0.0005,
+      "step": 7440
+    },
+    {
+      "epoch": 17.57,
+      "learning_rate": 1.2291169451073986e-05,
+      "loss": 0.0006,
+      "step": 7450
+    },
+    {
+      "epoch": 17.59,
+      "learning_rate": 1.2171837708830549e-05,
+      "loss": 0.0006,
+      "step": 7460
+    },
+    {
+      "epoch": 17.62,
+      "learning_rate": 1.2052505966587113e-05,
+      "loss": 0.0005,
+      "step": 7470
+    },
+    {
+      "epoch": 17.64,
+      "learning_rate": 1.1933174224343676e-05,
+      "loss": 0.0005,
+      "step": 7480
+    },
+    {
+      "epoch": 17.67,
+      "learning_rate": 1.1813842482100239e-05,
+      "loss": 0.0005,
+      "step": 7490
+    },
+    {
+      "epoch": 17.69,
+      "learning_rate": 1.1694510739856803e-05,
+      "loss": 0.0006,
+      "step": 7500
+    },
+    {
+      "epoch": 17.71,
+      "learning_rate": 1.1575178997613366e-05,
+      "loss": 0.0009,
+      "step": 7510
+    },
+    {
+      "epoch": 17.74,
+      "learning_rate": 1.145584725536993e-05,
+      "loss": 0.0005,
+      "step": 7520
+    },
+    {
+      "epoch": 17.76,
+      "learning_rate": 1.1336515513126491e-05,
+      "loss": 0.0005,
+      "step": 7530
+    },
+    {
+      "epoch": 17.78,
+      "learning_rate": 1.1217183770883056e-05,
+      "loss": 0.0005,
+      "step": 7540
+    },
+    {
+      "epoch": 17.81,
+      "learning_rate": 1.1097852028639618e-05,
+      "loss": 0.0005,
+      "step": 7550
+    },
+    {
+      "epoch": 17.83,
+      "learning_rate": 1.0978520286396183e-05,
+      "loss": 0.0007,
+      "step": 7560
+    },
+    {
+      "epoch": 17.85,
+      "learning_rate": 1.0859188544152746e-05,
+      "loss": 0.0005,
+      "step": 7570
+    },
+    {
+      "epoch": 17.88,
+      "learning_rate": 1.0739856801909308e-05,
+      "loss": 0.0012,
+      "step": 7580
+    },
+    {
+      "epoch": 17.9,
+      "learning_rate": 1.0620525059665871e-05,
+      "loss": 0.0005,
+      "step": 7590
+    },
+    {
+      "epoch": 17.92,
+      "learning_rate": 1.0501193317422435e-05,
+      "loss": 0.0009,
+      "step": 7600
+    },
+    {
+      "epoch": 17.95,
+      "learning_rate": 1.0381861575178998e-05,
+      "loss": 0.0007,
+      "step": 7610
+    },
+    {
+      "epoch": 17.97,
+      "learning_rate": 1.026252983293556e-05,
+      "loss": 0.0005,
+      "step": 7620
+    },
+    {
+      "epoch": 18.0,
+      "learning_rate": 1.0143198090692124e-05,
+      "loss": 0.0005,
+      "step": 7630
+    },
+    {
+      "epoch": 18.0,
+      "eval_loss": 2.181823253631592,
+      "eval_runtime": 26.8733,
+      "eval_samples_per_second": 28.02,
+      "eval_steps_per_second": 1.786,
+      "step": 7632
+    },
+    {
+      "epoch": 18.02,
+      "learning_rate": 1.0023866348448688e-05,
+      "loss": 0.0005,
+      "step": 7640
+    },
+    {
+      "epoch": 18.04,
+      "learning_rate": 9.90453460620525e-06,
+      "loss": 0.0005,
+      "step": 7650
+    },
+    {
+      "epoch": 18.07,
+      "learning_rate": 9.785202863961815e-06,
+      "loss": 0.0004,
+      "step": 7660
+    },
+    {
+      "epoch": 18.09,
+      "learning_rate": 9.665871121718378e-06,
+      "loss": 0.0005,
+      "step": 7670
+    },
+    {
+      "epoch": 18.11,
+      "learning_rate": 9.54653937947494e-06,
+      "loss": 0.0005,
+      "step": 7680
+    },
+    {
+      "epoch": 18.14,
+      "learning_rate": 9.427207637231505e-06,
+      "loss": 0.0007,
+      "step": 7690
+    },
+    {
+      "epoch": 18.16,
+      "learning_rate": 9.307875894988068e-06,
+      "loss": 0.0007,
+      "step": 7700
+    },
+    {
+      "epoch": 18.18,
+      "learning_rate": 9.18854415274463e-06,
+      "loss": 0.0005,
+      "step": 7710
+    },
+    {
+      "epoch": 18.21,
+      "learning_rate": 9.069212410501193e-06,
+      "loss": 0.0005,
+      "step": 7720
+    },
+    {
+      "epoch": 18.23,
+      "learning_rate": 8.949880668257757e-06,
+      "loss": 0.0005,
+      "step": 7730
+    },
+    {
+      "epoch": 18.25,
+      "learning_rate": 8.83054892601432e-06,
+      "loss": 0.0005,
+      "step": 7740
+    },
+    {
+      "epoch": 18.28,
+      "learning_rate": 8.711217183770885e-06,
+      "loss": 0.0005,
+      "step": 7750
+    },
+    {
+      "epoch": 18.3,
+      "learning_rate": 8.591885441527446e-06,
+      "loss": 0.0005,
+      "step": 7760
+    },
+    {
+      "epoch": 18.33,
+      "learning_rate": 8.47255369928401e-06,
+      "loss": 0.0005,
+      "step": 7770
+    },
+    {
+      "epoch": 18.35,
+      "learning_rate": 8.353221957040573e-06,
+      "loss": 0.0005,
+      "step": 7780
+    },
+    {
+      "epoch": 18.37,
+      "learning_rate": 8.233890214797137e-06,
+      "loss": 0.0005,
+      "step": 7790
+    },
+    {
+      "epoch": 18.4,
+      "learning_rate": 8.1145584725537e-06,
+      "loss": 0.0005,
+      "step": 7800
+    },
+    {
+      "epoch": 18.42,
+      "learning_rate": 7.995226730310263e-06,
+      "loss": 0.0005,
+      "step": 7810
+    },
+    {
+      "epoch": 18.44,
+      "learning_rate": 7.875894988066825e-06,
+      "loss": 0.0005,
+      "step": 7820
+    },
+    {
+      "epoch": 18.47,
+      "learning_rate": 7.75656324582339e-06,
+      "loss": 0.0005,
+      "step": 7830
+    },
+    {
+      "epoch": 18.49,
+      "learning_rate": 7.637231503579952e-06,
+      "loss": 0.0005,
+      "step": 7840
+    },
+    {
+      "epoch": 18.51,
+      "learning_rate": 7.517899761336517e-06,
+      "loss": 0.0004,
+      "step": 7850
+    },
+    {
+      "epoch": 18.54,
+      "learning_rate": 7.398568019093079e-06,
+      "loss": 0.0004,
+      "step": 7860
+    },
+    {
+      "epoch": 18.56,
+      "learning_rate": 7.279236276849642e-06,
+      "loss": 0.0005,
+      "step": 7870
+    },
+    {
+      "epoch": 18.58,
+      "learning_rate": 7.159904534606206e-06,
+      "loss": 0.0008,
+      "step": 7880
+    },
+    {
+      "epoch": 18.61,
+      "learning_rate": 7.040572792362769e-06,
+      "loss": 0.0007,
+      "step": 7890
+    },
+    {
+      "epoch": 18.63,
+      "learning_rate": 6.921241050119331e-06,
+      "loss": 0.0005,
+      "step": 7900
+    },
+    {
+      "epoch": 18.66,
+      "learning_rate": 6.801909307875895e-06,
+      "loss": 0.001,
+      "step": 7910
+    },
+    {
+      "epoch": 18.68,
+      "learning_rate": 6.682577565632458e-06,
+      "loss": 0.0005,
+      "step": 7920
+    },
+    {
+      "epoch": 18.7,
+      "learning_rate": 6.563245823389022e-06,
+      "loss": 0.0004,
+      "step": 7930
+    },
+    {
+      "epoch": 18.73,
+      "learning_rate": 6.4439140811455855e-06,
+      "loss": 0.0005,
+      "step": 7940
+    },
+    {
+      "epoch": 18.75,
+      "learning_rate": 6.324582338902147e-06,
+      "loss": 0.0005,
+      "step": 7950
+    },
+    {
+      "epoch": 18.77,
+      "learning_rate": 6.205250596658711e-06,
+      "loss": 0.0004,
+      "step": 7960
+    },
+    {
+      "epoch": 18.8,
+      "learning_rate": 6.0859188544152745e-06,
+      "loss": 0.0005,
+      "step": 7970
+    },
+    {
+      "epoch": 18.82,
+      "learning_rate": 5.966587112171838e-06,
+      "loss": 0.0005,
+      "step": 7980
+    },
+    {
+      "epoch": 18.84,
+      "learning_rate": 5.847255369928402e-06,
+      "loss": 0.0005,
+      "step": 7990
+    },
+    {
+      "epoch": 18.87,
+      "learning_rate": 5.727923627684965e-06,
+      "loss": 0.0005,
+      "step": 8000
+    }
+  ],
+  "max_steps": 8480,
+  "num_train_epochs": 20,
+  "total_flos": 0.0,
+  "trial_name": null,
+  "trial_params": null
+}