diff --git "a/checkpoint-17000/trainer_state.json" "b/checkpoint-17000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-17000/trainer_state.json"
@@ -0,0 +1,10491 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.6746031746031746,
+  "eval_steps": 500,
+  "global_step": 17000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 8.4848,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9e-07,
+      "loss": 8.4485,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 8.1681,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9e-06,
+      "loss": 7.9101,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 7.5507,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9e-06,
+      "loss": 7.303,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 7.1892,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9e-06,
+      "loss": 7.133,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4e-06,
+      "loss": 7.1948,
+      "step": 90
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 7.162,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.4e-06,
+      "loss": 7.1766,
+      "step": 110
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.9e-06,
+      "loss": 7.2304,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 7.2542,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.900000000000001e-06,
+      "loss": 7.2441,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.4e-06,
+      "loss": 7.2798,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.9e-06,
+      "loss": 7.3108,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 7.3133,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.9e-06,
+      "loss": 7.2967,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.4e-06,
+      "loss": 7.241,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.900000000000002e-06,
+      "loss": 7.2576,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.04e-05,
+      "loss": 7.1878,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.09e-05,
+      "loss": 7.1305,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1400000000000001e-05,
+      "loss": 7.1155,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.19e-05,
+      "loss": 7.146,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.24e-05,
+      "loss": 7.0695,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.29e-05,
+      "loss": 6.9668,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3400000000000002e-05,
+      "loss": 6.9292,
+      "step": 270
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3900000000000002e-05,
+      "loss": 6.8995,
+      "step": 280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.44e-05,
+      "loss": 6.7996,
+      "step": 290
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.49e-05,
+      "loss": 6.7778,
+      "step": 300
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.54e-05,
+      "loss": 6.7461,
+      "step": 310
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.59e-05,
+      "loss": 6.6594,
+      "step": 320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6400000000000002e-05,
+      "loss": 6.463,
+      "step": 330
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.69e-05,
+      "loss": 6.4716,
+      "step": 340
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.74e-05,
+      "loss": 6.4418,
+      "step": 350
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.79e-05,
+      "loss": 6.3907,
+      "step": 360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.84e-05,
+      "loss": 6.3509,
+      "step": 370
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8900000000000002e-05,
+      "loss": 6.3025,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.94e-05,
+      "loss": 6.2211,
+      "step": 390
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9900000000000003e-05,
+      "loss": 6.1932,
+      "step": 400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.04e-05,
+      "loss": 6.0846,
+      "step": 410
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.09e-05,
+      "loss": 6.0304,
+      "step": 420
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1400000000000002e-05,
+      "loss": 6.0075,
+      "step": 430
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.19e-05,
+      "loss": 6.0828,
+      "step": 440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2400000000000002e-05,
+      "loss": 5.9428,
+      "step": 450
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.29e-05,
+      "loss": 5.9078,
+      "step": 460
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3400000000000003e-05,
+      "loss": 5.9284,
+      "step": 470
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.39e-05,
+      "loss": 5.8741,
+      "step": 480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.44e-05,
+      "loss": 5.7947,
+      "step": 490
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4900000000000002e-05,
+      "loss": 5.6575,
+      "step": 500
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 5.361832141876221,
+      "eval_runtime": 418.3831,
+      "eval_samples_per_second": 50.721,
+      "eval_steps_per_second": 0.397,
+      "step": 500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.54e-05,
+      "loss": 5.7587,
+      "step": 510
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5900000000000003e-05,
+      "loss": 5.6477,
+      "step": 520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.64e-05,
+      "loss": 5.5106,
+      "step": 530
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6900000000000003e-05,
+      "loss": 5.6785,
+      "step": 540
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7400000000000002e-05,
+      "loss": 5.5936,
+      "step": 550
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7900000000000004e-05,
+      "loss": 5.5427,
+      "step": 560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.84e-05,
+      "loss": 5.4491,
+      "step": 570
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8899999999999998e-05,
+      "loss": 5.5095,
+      "step": 580
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.94e-05,
+      "loss": 5.5513,
+      "step": 590
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9900000000000002e-05,
+      "loss": 5.3501,
+      "step": 600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.04e-05,
+      "loss": 5.2687,
+      "step": 610
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.09e-05,
+      "loss": 5.2777,
+      "step": 620
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.1400000000000004e-05,
+      "loss": 5.3233,
+      "step": 630
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.19e-05,
+      "loss": 5.279,
+      "step": 640
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.24e-05,
+      "loss": 5.2254,
+      "step": 650
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.29e-05,
+      "loss": 5.1651,
+      "step": 660
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.3400000000000005e-05,
+      "loss": 5.0647,
+      "step": 670
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.3900000000000004e-05,
+      "loss": 5.0709,
+      "step": 680
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.4399999999999996e-05,
+      "loss": 4.9399,
+      "step": 690
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.49e-05,
+      "loss": 5.006,
+      "step": 700
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.54e-05,
+      "loss": 4.8655,
+      "step": 710
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.59e-05,
+      "loss": 4.8941,
+      "step": 720
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.6400000000000004e-05,
+      "loss": 4.761,
+      "step": 730
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.69e-05,
+      "loss": 4.8677,
+      "step": 740
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.74e-05,
+      "loss": 4.7086,
+      "step": 750
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.79e-05,
+      "loss": 4.5536,
+      "step": 760
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 4.569,
+      "step": 770
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.8900000000000004e-05,
+      "loss": 4.5551,
+      "step": 780
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.94e-05,
+      "loss": 4.4652,
+      "step": 790
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.99e-05,
+      "loss": 4.5472,
+      "step": 800
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.0400000000000006e-05,
+      "loss": 4.4666,
+      "step": 810
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.09e-05,
+      "loss": 4.3739,
+      "step": 820
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.14e-05,
+      "loss": 4.3631,
+      "step": 830
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.19e-05,
+      "loss": 4.2101,
+      "step": 840
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.24e-05,
+      "loss": 4.1703,
+      "step": 850
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.29e-05,
+      "loss": 4.049,
+      "step": 860
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.3400000000000005e-05,
+      "loss": 4.1063,
+      "step": 870
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.39e-05,
+      "loss": 4.1917,
+      "step": 880
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.44e-05,
+      "loss": 4.1338,
+      "step": 890
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.4850000000000006e-05,
+      "loss": 3.9918,
+      "step": 900
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.5350000000000005e-05,
+      "loss": 3.9524,
+      "step": 910
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.585e-05,
+      "loss": 3.9808,
+      "step": 920
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.635e-05,
+      "loss": 4.0589,
+      "step": 930
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.685000000000001e-05,
+      "loss": 3.9574,
+      "step": 940
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.735e-05,
+      "loss": 3.8789,
+      "step": 950
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.785e-05,
+      "loss": 3.8236,
+      "step": 960
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.835e-05,
+      "loss": 3.7172,
+      "step": 970
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.885e-05,
+      "loss": 3.7982,
+      "step": 980
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.935e-05,
+      "loss": 3.7415,
+      "step": 990
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9850000000000006e-05,
+      "loss": 3.7155,
+      "step": 1000
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 3.0474321842193604,
+      "eval_runtime": 418.2758,
+      "eval_samples_per_second": 50.734,
+      "eval_steps_per_second": 0.397,
+      "step": 1000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.999999752285213e-05,
+      "loss": 3.6935,
+      "step": 1010
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.999998538988411e-05,
+      "loss": 3.4524,
+      "step": 1020
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9999963146114506e-05,
+      "loss": 3.5505,
+      "step": 1030
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.999993079155232e-05,
+      "loss": 3.5052,
+      "step": 1040
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.999988832621063e-05,
+      "loss": 3.631,
+      "step": 1050
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.999983575010662e-05,
+      "loss": 3.4188,
+      "step": 1060
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.999977306326154e-05,
+      "loss": 3.4033,
+      "step": 1070
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9999700265700746e-05,
+      "loss": 3.3126,
+      "step": 1080
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.999961735745368e-05,
+      "loss": 3.2439,
+      "step": 1090
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.999952433855388e-05,
+      "loss": 3.2104,
+      "step": 1100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9999421209038955e-05,
+      "loss": 3.3126,
+      "step": 1110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.999930796895063e-05,
+      "loss": 3.325,
+      "step": 1120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9999184618334685e-05,
+      "loss": 3.2966,
+      "step": 1130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9999051157241014e-05,
+      "loss": 3.1917,
+      "step": 1140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.99989075857236e-05,
+      "loss": 3.3806,
+      "step": 1150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9998753903840504e-05,
+      "loss": 3.1974,
+      "step": 1160
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.999859011165386e-05,
+      "loss": 3.192,
+      "step": 1170
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.999841620922995e-05,
+      "loss": 3.0923,
+      "step": 1180
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.999823219663907e-05,
+      "loss": 3.1304,
+      "step": 1190
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9998038073955655e-05,
+      "loss": 3.1287,
+      "step": 1200
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9997833841258223e-05,
+      "loss": 3.1111,
+      "step": 1210
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.999761949862936e-05,
+      "loss": 3.0336,
+      "step": 1220
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.999739504615576e-05,
+      "loss": 2.943,
+      "step": 1230
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9997160483928185e-05,
+      "loss": 3.1012,
+      "step": 1240
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.999691581204152e-05,
+      "loss": 3.0045,
+      "step": 1250
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.99966610305947e-05,
+      "loss": 2.9925,
+      "step": 1260
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.999639613969078e-05,
+      "loss": 3.0126,
+      "step": 1270
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.999612113943689e-05,
+      "loss": 2.9458,
+      "step": 1280
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.999583602994424e-05,
+      "loss": 2.78,
+      "step": 1290
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.999554081132813e-05,
+      "loss": 2.9488,
+      "step": 1300
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.999523548370799e-05,
+      "loss": 2.9124,
+      "step": 1310
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.999492004720727e-05,
+      "loss": 2.7713,
+      "step": 1320
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9994594501953555e-05,
+      "loss": 2.6387,
+      "step": 1330
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.999425884807851e-05,
+      "loss": 2.7555,
+      "step": 1340
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.999391308571788e-05,
+      "loss": 2.6673,
+      "step": 1350
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.99935572150115e-05,
+      "loss": 2.7384,
+      "step": 1360
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9993191236103295e-05,
+      "loss": 2.6289,
+      "step": 1370
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9992815149141294e-05,
+      "loss": 2.6144,
+      "step": 1380
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9992428954277573e-05,
+      "loss": 2.7018,
+      "step": 1390
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.999203265166834e-05,
+      "loss": 2.6399,
+      "step": 1400
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9991626241473876e-05,
+      "loss": 2.6183,
+      "step": 1410
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.999120972385853e-05,
+      "loss": 2.7058,
+      "step": 1420
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9990783098990765e-05,
+      "loss": 2.6614,
+      "step": 1430
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9990346367043114e-05,
+      "loss": 2.5442,
+      "step": 1440
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.998989952819223e-05,
+      "loss": 2.543,
+      "step": 1450
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.998944258261879e-05,
+      "loss": 2.4622,
+      "step": 1460
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.998897553050763e-05,
+      "loss": 2.5788,
+      "step": 1470
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9988498372047633e-05,
+      "loss": 2.7313,
+      "step": 1480
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.998801110743177e-05,
+      "loss": 2.664,
+      "step": 1490
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9987513736857106e-05,
+      "loss": 2.4293,
+      "step": 1500
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 1.9413985013961792,
+      "eval_runtime": 418.5254,
+      "eval_samples_per_second": 50.704,
+      "eval_steps_per_second": 0.397,
+      "step": 1500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.998700626052481e-05,
+      "loss": 2.5566,
+      "step": 1510
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.99864886786401e-05,
+      "loss": 2.3453,
+      "step": 1520
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.998596099141231e-05,
+      "loss": 2.4617,
+      "step": 1530
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.998542319905486e-05,
+      "loss": 2.4163,
+      "step": 1540
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.998487530178525e-05,
+      "loss": 2.3288,
+      "step": 1550
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.998431729982506e-05,
+      "loss": 2.3888,
+      "step": 1560
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.998374919339998e-05,
+      "loss": 2.3015,
+      "step": 1570
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9983170982739745e-05,
+      "loss": 2.2867,
+      "step": 1580
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.998258266807822e-05,
+      "loss": 2.4161,
+      "step": 1590
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.998198424965333e-05,
+      "loss": 2.343,
+      "step": 1600
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9981375727707114e-05,
+      "loss": 2.4815,
+      "step": 1610
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.998075710248565e-05,
+      "loss": 2.2715,
+      "step": 1620
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9980128374239156e-05,
+      "loss": 2.113,
+      "step": 1630
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.997948954322189e-05,
+      "loss": 2.2174,
+      "step": 1640
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.997884060969222e-05,
+      "loss": 2.2034,
+      "step": 1650
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.99781815739126e-05,
+      "loss": 2.3214,
+      "step": 1660
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9977512436149563e-05,
+      "loss": 2.1852,
+      "step": 1670
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9976833196673735e-05,
+      "loss": 2.3061,
+      "step": 1680
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9976143855759814e-05,
+      "loss": 2.1955,
+      "step": 1690
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.99754444136866e-05,
+      "loss": 2.2442,
+      "step": 1700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9974734870736975e-05,
+      "loss": 2.1702,
+      "step": 1710
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.997401522719788e-05,
+      "loss": 2.1317,
+      "step": 1720
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.997328548336038e-05,
+      "loss": 1.975,
+      "step": 1730
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.99725456395196e-05,
+      "loss": 2.0007,
+      "step": 1740
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.997179569597475e-05,
+      "loss": 2.0239,
+      "step": 1750
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9971035653029153e-05,
+      "loss": 1.9269,
+      "step": 1760
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.997026551099017e-05,
+      "loss": 2.0069,
+      "step": 1770
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.996948527016929e-05,
+      "loss": 2.2301,
+      "step": 1780
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.996869493088205e-05,
+      "loss": 1.9459,
+      "step": 1790
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.996789449344812e-05,
+      "loss": 2.0352,
+      "step": 1800
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9967083958191187e-05,
+      "loss": 2.0872,
+      "step": 1810
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.996626332543907e-05,
+      "loss": 2.0663,
+      "step": 1820
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.996543259552367e-05,
+      "loss": 1.9918,
+      "step": 1830
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.996459176878095e-05,
+      "loss": 1.9385,
+      "step": 1840
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9963740845550965e-05,
+      "loss": 2.094,
+      "step": 1850
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.996287982617787e-05,
+      "loss": 1.7899,
+      "step": 1860
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.996200871100988e-05,
+      "loss": 1.9819,
+      "step": 1870
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9961127500399295e-05,
+      "loss": 2.0245,
+      "step": 1880
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.996023619470252e-05,
+      "loss": 1.8344,
+      "step": 1890
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.9959334794280014e-05,
+      "loss": 1.99,
+      "step": 1900
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.9958423299496344e-05,
+      "loss": 1.7641,
+      "step": 1910
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.995750171072014e-05,
+      "loss": 1.8968,
+      "step": 1920
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.9956570028324124e-05,
+      "loss": 1.8585,
+      "step": 1930
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.99556282526851e-05,
+      "loss": 1.7962,
+      "step": 1940
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.995467638418395e-05,
+      "loss": 1.8851,
+      "step": 1950
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.995371442320565e-05,
+      "loss": 1.8522,
+      "step": 1960
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.9952742370139224e-05,
+      "loss": 1.734,
+      "step": 1970
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.995176022537784e-05,
+      "loss": 1.7785,
+      "step": 1980
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.995076798931867e-05,
+      "loss": 1.7153,
+      "step": 1990
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.994976566236303e-05,
+      "loss": 1.7285,
+      "step": 2000
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 1.385947585105896,
+      "eval_runtime": 418.4462,
+      "eval_samples_per_second": 50.714,
+      "eval_steps_per_second": 0.397,
+      "step": 2000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.994875324491629e-05,
+      "loss": 1.779,
+      "step": 2010
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.9947730737387885e-05,
+      "loss": 1.7023,
+      "step": 2020
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.994669814019138e-05,
+      "loss": 1.7549,
+      "step": 2030
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.994565545374438e-05,
+      "loss": 1.6152,
+      "step": 2040
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.9944602678468575e-05,
+      "loss": 1.7418,
+      "step": 2050
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.994353981478974e-05,
+      "loss": 1.8069,
+      "step": 2060
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.994246686313774e-05,
+      "loss": 1.7131,
+      "step": 2070
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.99413838239465e-05,
+      "loss": 1.6513,
+      "step": 2080
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.9940290697654045e-05,
+      "loss": 1.7468,
+      "step": 2090
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.993918748470246e-05,
+      "loss": 1.6187,
+      "step": 2100
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.993807418553795e-05,
+      "loss": 1.6701,
+      "step": 2110
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.993695080061074e-05,
+      "loss": 1.6918,
+      "step": 2120
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.993581733037516e-05,
+      "loss": 1.5082,
+      "step": 2130
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.9934673775289644e-05,
+      "loss": 1.8572,
+      "step": 2140
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9933520135816665e-05,
+      "loss": 1.6094,
+      "step": 2150
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.99323564124228e-05,
+      "loss": 1.6666,
+      "step": 2160
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9931182605578696e-05,
+      "loss": 1.7766,
+      "step": 2170
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9929998715759076e-05,
+      "loss": 1.6558,
+      "step": 2180
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.992880474344274e-05,
+      "loss": 1.6491,
+      "step": 2190
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.992760068911258e-05,
+      "loss": 1.5249,
+      "step": 2200
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9926386553255547e-05,
+      "loss": 1.3977,
+      "step": 2210
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9925162336362676e-05,
+      "loss": 1.5947,
+      "step": 2220
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.992392803892908e-05,
+      "loss": 1.6167,
+      "step": 2230
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9922683661453953e-05,
+      "loss": 1.4321,
+      "step": 2240
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9921429204440555e-05,
+      "loss": 1.5326,
+      "step": 2250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.992016466839624e-05,
+      "loss": 1.542,
+      "step": 2260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.991889005383241e-05,
+      "loss": 1.6529,
+      "step": 2270
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.991760536126457e-05,
+      "loss": 1.478,
+      "step": 2280
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.99163105912123e-05,
+      "loss": 1.4306,
+      "step": 2290
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.991500574419924e-05,
+      "loss": 1.5033,
+      "step": 2300
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.99136908207531e-05,
+      "loss": 1.6476,
+      "step": 2310
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.991236582140569e-05,
+      "loss": 1.5609,
+      "step": 2320
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.991103074669289e-05,
+      "loss": 1.5151,
+      "step": 2330
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.990968559715462e-05,
+      "loss": 1.5199,
+      "step": 2340
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9908330373334935e-05,
+      "loss": 1.4894,
+      "step": 2350
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.990696507578191e-05,
+      "loss": 1.4853,
+      "step": 2360
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.990558970504772e-05,
+      "loss": 1.4917,
+      "step": 2370
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.990434325927671e-05,
+      "loss": 1.4449,
+      "step": 2380
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.990294875103413e-05,
+      "loss": 1.3534,
+      "step": 2390
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.990154417123472e-05,
+      "loss": 1.5047,
+      "step": 2400
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.990012952044654e-05,
+      "loss": 1.485,
+      "step": 2410
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.98987047992417e-05,
+      "loss": 1.4522,
+      "step": 2420
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.989727000819644e-05,
+      "loss": 1.4396,
+      "step": 2430
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9895825147891e-05,
+      "loss": 1.4213,
+      "step": 2440
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.989437021890975e-05,
+      "loss": 1.4682,
+      "step": 2450
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.98929052218411e-05,
+      "loss": 1.2974,
+      "step": 2460
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.989143015727754e-05,
+      "loss": 1.3692,
+      "step": 2470
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.988994502581565e-05,
+      "loss": 1.5167,
+      "step": 2480
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.988844982805605e-05,
+      "loss": 1.3565,
+      "step": 2490
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9886944564603444e-05,
+      "loss": 1.4537,
+      "step": 2500
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 1.0673185586929321,
+      "eval_runtime": 418.0872,
+      "eval_samples_per_second": 50.757,
+      "eval_steps_per_second": 0.397,
+      "step": 2500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.988542923606663e-05,
+      "loss": 1.3416,
+      "step": 2510
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.988390384305843e-05,
+      "loss": 1.1869,
+      "step": 2520
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9882368386195776e-05,
+      "loss": 1.2827,
+      "step": 2530
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.988082286609965e-05,
+      "loss": 1.3322,
+      "step": 2540
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9879267283395125e-05,
+      "loss": 1.3467,
+      "step": 2550
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9877701638711314e-05,
+      "loss": 1.2807,
+      "step": 2560
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.987612593268142e-05,
+      "loss": 1.2851,
+      "step": 2570
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.98745401659427e-05,
+      "loss": 1.3327,
+      "step": 2580
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9872944339136503e-05,
+      "loss": 1.4023,
+      "step": 2590
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.987133845290822e-05,
+      "loss": 1.2509,
+      "step": 2600
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9869722507907343e-05,
+      "loss": 1.2828,
+      "step": 2610
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9868096504787395e-05,
+      "loss": 1.3894,
+      "step": 2620
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9866460444205983e-05,
+      "loss": 1.269,
+      "step": 2630
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.986481432682479e-05,
+      "loss": 1.3342,
+      "step": 2640
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.986315815330956e-05,
+      "loss": 1.215,
+      "step": 2650
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.98614919243301e-05,
+      "loss": 1.1793,
+      "step": 2660
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.98598156405603e-05,
+      "loss": 1.3144,
+      "step": 2670
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.985812930267808e-05,
+      "loss": 1.2003,
+      "step": 2680
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.985643291136546e-05,
+      "loss": 1.2853,
+      "step": 2690
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.985472646730852e-05,
+      "loss": 1.1849,
+      "step": 2700
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.985300997119739e-05,
+      "loss": 1.4755,
+      "step": 2710
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.985128342372629e-05,
+      "loss": 1.2081,
+      "step": 2720
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.9849546825593476e-05,
+      "loss": 1.1917,
+      "step": 2730
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.9847800177501305e-05,
+      "loss": 1.1097,
+      "step": 2740
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.984604348015616e-05,
+      "loss": 1.1659,
+      "step": 2750
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.984427673426851e-05,
+      "loss": 1.16,
+      "step": 2760
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.9842499940552886e-05,
+      "loss": 1.2824,
+      "step": 2770
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.9840713099727885e-05,
+      "loss": 1.0822,
+      "step": 2780
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.983891621251616e-05,
+      "loss": 1.1324,
+      "step": 2790
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.983710927964442e-05,
+      "loss": 1.3446,
+      "step": 2800
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.9835292301843476e-05,
+      "loss": 1.0565,
+      "step": 2810
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.983346527984813e-05,
+      "loss": 1.1089,
+      "step": 2820
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.983162821439733e-05,
+      "loss": 1.1193,
+      "step": 2830
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.982978110623402e-05,
+      "loss": 1.1459,
+      "step": 2840
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.982792395610524e-05,
+      "loss": 1.1643,
+      "step": 2850
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.9826056764762074e-05,
+      "loss": 1.1106,
+      "step": 2860
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.982417953295968e-05,
+      "loss": 1.1448,
+      "step": 2870
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.982229226145727e-05,
+      "loss": 1.1846,
+      "step": 2880
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.982039495101812e-05,
+      "loss": 1.2616,
+      "step": 2890
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9818487602409556e-05,
+      "loss": 1.0037,
+      "step": 2900
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.981657021640298e-05,
+      "loss": 1.0965,
+      "step": 2910
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.981464279377384e-05,
+      "loss": 1.213,
+      "step": 2920
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.981270533530164e-05,
+      "loss": 1.0509,
+      "step": 2930
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9810757841769976e-05,
+      "loss": 1.1545,
+      "step": 2940
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.980880031396645e-05,
+      "loss": 1.0516,
+      "step": 2950
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.980683275268277e-05,
+      "loss": 1.1009,
+      "step": 2960
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.980485515871467e-05,
+      "loss": 1.0394,
+      "step": 2970
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.980286753286195e-05,
+      "loss": 1.1427,
+      "step": 2980
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9800869875928475e-05,
+      "loss": 1.1794,
+      "step": 2990
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9798862188722164e-05,
+      "loss": 1.0714,
+      "step": 3000
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 0.8538320660591125,
+      "eval_runtime": 418.1568,
+      "eval_samples_per_second": 50.749,
+      "eval_steps_per_second": 0.397,
+      "step": 3000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.979684447205499e-05,
+      "loss": 1.0871,
+      "step": 3010
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9794816726742976e-05,
+      "loss": 1.0599,
+      "step": 3020
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.979277895360622e-05,
+      "loss": 1.1874,
+      "step": 3030
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.979073115346885e-05,
+      "loss": 1.1884,
+      "step": 3040
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9788673327159076e-05,
+      "loss": 1.2292,
+      "step": 3050
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.978660547550914e-05,
+      "loss": 0.9754,
+      "step": 3060
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.978452759935535e-05,
+      "loss": 1.0247,
+      "step": 3070
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.978243969953806e-05,
+      "loss": 1.0904,
+      "step": 3080
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.978055202016805e-05,
+      "loss": 1.2127,
+      "step": 3090
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9778445077719834e-05,
+      "loss": 1.0942,
+      "step": 3100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9776328114068094e-05,
+      "loss": 1.0569,
+      "step": 3110
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9774201130069e-05,
+      "loss": 1.1967,
+      "step": 3120
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.977206412658277e-05,
+      "loss": 0.884,
+      "step": 3130
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9769917104473687e-05,
+      "loss": 0.9447,
+      "step": 3140
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.976776006461006e-05,
+      "loss": 1.1597,
+      "step": 3150
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9765593007864265e-05,
+      "loss": 1.0985,
+      "step": 3160
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9763415935112745e-05,
+      "loss": 1.0807,
+      "step": 3170
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.976122884723596e-05,
+      "loss": 0.9753,
+      "step": 3180
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.975903174511846e-05,
+      "loss": 0.9865,
+      "step": 3190
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9756824629648816e-05,
+      "loss": 0.8852,
+      "step": 3200
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.975460750171965e-05,
+      "loss": 1.0028,
+      "step": 3210
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9752380362227636e-05,
+      "loss": 0.9919,
+      "step": 3220
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.975014321207352e-05,
+      "loss": 0.9709,
+      "step": 3230
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9747896052162065e-05,
+      "loss": 1.0648,
+      "step": 3240
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.974563888340209e-05,
+      "loss": 0.9859,
+      "step": 3250
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9743371706706486e-05,
+      "loss": 1.0467,
+      "step": 3260
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.974109452299216e-05,
+      "loss": 0.9843,
+      "step": 3270
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.973880733318007e-05,
+      "loss": 1.0496,
+      "step": 3280
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.973651013819525e-05,
+      "loss": 0.9739,
+      "step": 3290
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.973420293896674e-05,
+      "loss": 0.9777,
+      "step": 3300
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.973188573642765e-05,
+      "loss": 0.9028,
+      "step": 3310
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.972955853151515e-05,
+      "loss": 1.0436,
+      "step": 3320
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.972722132517042e-05,
+      "loss": 0.9751,
+      "step": 3330
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.97248741183387e-05,
+      "loss": 0.9398,
+      "step": 3340
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.972251691196928e-05,
+      "loss": 1.026,
+      "step": 3350
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.972014970701549e-05,
+      "loss": 0.9012,
+      "step": 3360
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9717772504434714e-05,
+      "loss": 0.9352,
+      "step": 3370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.971538530518836e-05,
+      "loss": 0.8128,
+      "step": 3380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9712988110241884e-05,
+      "loss": 0.8543,
+      "step": 3390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9710580920564796e-05,
+      "loss": 0.8571,
+      "step": 3400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9708163737130644e-05,
+      "loss": 0.9688,
+      "step": 3410
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.970573656091701e-05,
+      "loss": 0.887,
+      "step": 3420
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9703299392905516e-05,
+      "loss": 0.8429,
+      "step": 3430
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9700852234081843e-05,
+      "loss": 1.0118,
+      "step": 3440
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.96983950854357e-05,
+      "loss": 0.9119,
+      "step": 3450
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.969592794796083e-05,
+      "loss": 0.8758,
+      "step": 3460
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.969345082265503e-05,
+      "loss": 1.0163,
+      "step": 3470
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9690963710520125e-05,
+      "loss": 0.8004,
+      "step": 3480
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.968846661256199e-05,
+      "loss": 1.0429,
+      "step": 3490
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9685959529790526e-05,
+      "loss": 1.0045,
+      "step": 3500
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 0.726789116859436,
+      "eval_runtime": 418.0987,
+      "eval_samples_per_second": 50.756,
+      "eval_steps_per_second": 0.397,
+      "step": 3500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9683442463219674e-05,
+      "loss": 0.9721,
+      "step": 3510
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.968091541386742e-05,
+      "loss": 0.8624,
+      "step": 3520
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.96783783827558e-05,
+      "loss": 0.9204,
+      "step": 3530
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.967583137091085e-05,
+      "loss": 1.1513,
+      "step": 3540
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.967327437936268e-05,
+      "loss": 0.7905,
+      "step": 3550
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.96707074091454e-05,
+      "loss": 1.0437,
+      "step": 3560
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9668130461297194e-05,
+      "loss": 0.9948,
+      "step": 3570
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9665543536860255e-05,
+      "loss": 0.9434,
+      "step": 3580
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.966294663688083e-05,
+      "loss": 0.9568,
+      "step": 3590
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.966033976240916e-05,
+      "loss": 0.8783,
+      "step": 3600
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.965772291449958e-05,
+      "loss": 0.9411,
+      "step": 3610
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.965509609421042e-05,
+      "loss": 1.0,
+      "step": 3620
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9652459302604046e-05,
+      "loss": 0.8513,
+      "step": 3630
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.964981254074686e-05,
+      "loss": 0.8841,
+      "step": 3640
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9647155809709306e-05,
+      "loss": 0.8358,
+      "step": 3650
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9644489110565845e-05,
+      "loss": 1.0356,
+      "step": 3660
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.964181244439498e-05,
+      "loss": 0.8238,
+      "step": 3670
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.963912581227924e-05,
+      "loss": 0.9625,
+      "step": 3680
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.963642921530518e-05,
+      "loss": 0.8891,
+      "step": 3690
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9633722654563405e-05,
+      "loss": 0.8561,
+      "step": 3700
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.963100613114852e-05,
+      "loss": 0.8179,
+      "step": 3710
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.962827964615919e-05,
+      "loss": 0.8611,
+      "step": 3720
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.962554320069808e-05,
+      "loss": 0.7999,
+      "step": 3730
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9622796795871904e-05,
+      "loss": 0.8694,
+      "step": 3740
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9620040432791395e-05,
+      "loss": 0.7623,
+      "step": 3750
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.961727411257131e-05,
+      "loss": 0.7975,
+      "step": 3760
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.961449783633045e-05,
+      "loss": 0.7171,
+      "step": 3770
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9611711605191625e-05,
+      "loss": 0.9011,
+      "step": 3780
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9608915420281675e-05,
+      "loss": 0.801,
+      "step": 3790
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9606109282731463e-05,
+      "loss": 0.9375,
+      "step": 3800
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.96032931936759e-05,
+      "loss": 0.8559,
+      "step": 3810
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9600467154253885e-05,
+      "loss": 0.8106,
+      "step": 3820
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.959763116560836e-05,
+      "loss": 0.8521,
+      "step": 3830
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.95947852288863e-05,
+      "loss": 0.7385,
+      "step": 3840
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.95919293452387e-05,
+      "loss": 0.8008,
+      "step": 3850
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.958906351582054e-05,
+      "loss": 0.7899,
+      "step": 3860
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.958618774179089e-05,
+      "loss": 0.9726,
+      "step": 3870
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9583302024312796e-05,
+      "loss": 0.9544,
+      "step": 3880
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9580406364553325e-05,
+      "loss": 0.8564,
+      "step": 3890
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.957750076368357e-05,
+      "loss": 0.7692,
+      "step": 3900
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.957458522287868e-05,
+      "loss": 0.7267,
+      "step": 3910
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9571659743317754e-05,
+      "loss": 0.7888,
+      "step": 3920
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.956872432618399e-05,
+      "loss": 0.9033,
+      "step": 3930
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9565778972664534e-05,
+      "loss": 0.6797,
+      "step": 3940
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.95628236839506e-05,
+      "loss": 0.753,
+      "step": 3950
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.95598584612374e-05,
+      "loss": 0.769,
+      "step": 3960
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9556883305724155e-05,
+      "loss": 0.7502,
+      "step": 3970
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.955389821861412e-05,
+      "loss": 0.714,
+      "step": 3980
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.955090320111457e-05,
+      "loss": 0.8226,
+      "step": 3990
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.954789825443678e-05,
+      "loss": 0.8212,
+      "step": 4000
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 0.6174027323722839,
+      "eval_runtime": 418.2799,
+      "eval_samples_per_second": 50.734,
+      "eval_steps_per_second": 0.397,
+      "step": 4000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.954488337979604e-05,
+      "loss": 0.7718,
+      "step": 4010
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9541858578411674e-05,
+      "loss": 0.8504,
+      "step": 4020
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9538823851507006e-05,
+      "loss": 0.8015,
+      "step": 4030
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.953577920030937e-05,
+      "loss": 0.7355,
+      "step": 4040
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.953272462605013e-05,
+      "loss": 0.6763,
+      "step": 4050
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.952966012996466e-05,
+      "loss": 0.8094,
+      "step": 4060
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.952658571329233e-05,
+      "loss": 0.63,
+      "step": 4070
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.952350137727653e-05,
+      "loss": 0.9106,
+      "step": 4080
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9520407123164676e-05,
+      "loss": 0.8204,
+      "step": 4090
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9517302952208174e-05,
+      "loss": 0.8734,
+      "step": 4100
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.951418886566247e-05,
+      "loss": 0.7548,
+      "step": 4110
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.951106486478698e-05,
+      "loss": 0.8011,
+      "step": 4120
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.950793095084516e-05,
+      "loss": 0.7024,
+      "step": 4130
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.950478712510446e-05,
+      "loss": 0.8575,
+      "step": 4140
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9501633388836354e-05,
+      "loss": 0.8867,
+      "step": 4150
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.949846974331631e-05,
+      "loss": 0.8235,
+      "step": 4160
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.949529618982381e-05,
+      "loss": 0.8188,
+      "step": 4170
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.9492112729642346e-05,
+      "loss": 0.7565,
+      "step": 4180
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.948891936405941e-05,
+      "loss": 0.6754,
+      "step": 4190
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.948571609436649e-05,
+      "loss": 0.7474,
+      "step": 4200
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.94825029218591e-05,
+      "loss": 0.8623,
+      "step": 4210
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.947927984783674e-05,
+      "loss": 0.7378,
+      "step": 4220
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.947604687360295e-05,
+      "loss": 0.819,
+      "step": 4230
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.947280400046525e-05,
+      "loss": 0.7274,
+      "step": 4240
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.9469551229735126e-05,
+      "loss": 0.7652,
+      "step": 4250
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.946628856272814e-05,
+      "loss": 0.695,
+      "step": 4260
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.9463016000763794e-05,
+      "loss": 0.8428,
+      "step": 4270
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.945973354516564e-05,
+      "loss": 0.7881,
+      "step": 4280
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.9456441197261195e-05,
+      "loss": 0.8882,
+      "step": 4290
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.9453138958381994e-05,
+      "loss": 0.7668,
+      "step": 4300
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.944982682986357e-05,
+      "loss": 0.7228,
+      "step": 4310
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.944650481304545e-05,
+      "loss": 0.6518,
+      "step": 4320
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.944317290927117e-05,
+      "loss": 0.7608,
+      "step": 4330
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.943983111988827e-05,
+      "loss": 0.6956,
+      "step": 4340
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.943647944624826e-05,
+      "loss": 0.6995,
+      "step": 4350
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.943311788970667e-05,
+      "loss": 0.6156,
+      "step": 4360
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.9429746451623024e-05,
+      "loss": 0.6318,
+      "step": 4370
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.9426365133360834e-05,
+      "loss": 0.7453,
+      "step": 4380
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.942297393628763e-05,
+      "loss": 0.7357,
+      "step": 4390
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.941957286177491e-05,
+      "loss": 0.6936,
+      "step": 4400
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.9416161911198176e-05,
+      "loss": 0.7319,
+      "step": 4410
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.941274108593693e-05,
+      "loss": 0.6867,
+      "step": 4420
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.940931038737468e-05,
+      "loss": 0.6808,
+      "step": 4430
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.940586981689888e-05,
+      "loss": 0.6318,
+      "step": 4440
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.940241937590102e-05,
+      "loss": 0.7534,
+      "step": 4450
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.939895906577658e-05,
+      "loss": 0.693,
+      "step": 4460
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.939548888792502e-05,
+      "loss": 0.7632,
+      "step": 4470
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.939200884374979e-05,
+      "loss": 0.8196,
+      "step": 4480
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.938851893465832e-05,
+      "loss": 0.6438,
+      "step": 4490
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.938501916206206e-05,
+      "loss": 0.699,
+      "step": 4500
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 0.5351088643074036,
+      "eval_runtime": 417.6608,
+      "eval_samples_per_second": 50.809,
+      "eval_steps_per_second": 0.397,
+      "step": 4500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.938150952737642e-05,
+      "loss": 0.7139,
+      "step": 4510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.937799003202081e-05,
+      "loss": 0.7223,
+      "step": 4520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9374460677418635e-05,
+      "loss": 0.7339,
+      "step": 4530
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.937092146499727e-05,
+      "loss": 0.7045,
+      "step": 4540
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9367372396188095e-05,
+      "loss": 0.6417,
+      "step": 4550
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9363813472426465e-05,
+      "loss": 0.7393,
+      "step": 4560
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9360244695151716e-05,
+      "loss": 0.7757,
+      "step": 4570
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.935666606580719e-05,
+      "loss": 0.8525,
+      "step": 4580
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.935307758584019e-05,
+      "loss": 0.6333,
+      "step": 4590
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9349479256702023e-05,
+      "loss": 0.7163,
+      "step": 4600
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.934587107984796e-05,
+      "loss": 0.8321,
+      "step": 4610
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.934225305673728e-05,
+      "loss": 0.7177,
+      "step": 4620
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.93386251888332e-05,
+      "loss": 0.6257,
+      "step": 4630
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9334987477602975e-05,
+      "loss": 0.7032,
+      "step": 4640
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.933133992451779e-05,
+      "loss": 0.6951,
+      "step": 4650
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.932768253105286e-05,
+      "loss": 0.6543,
+      "step": 4660
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.932401529868733e-05,
+      "loss": 0.5692,
+      "step": 4670
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.932033822890436e-05,
+      "loss": 0.6059,
+      "step": 4680
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9316651323191067e-05,
+      "loss": 0.5913,
+      "step": 4690
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.931295458303856e-05,
+      "loss": 0.6108,
+      "step": 4700
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9309248009941914e-05,
+      "loss": 0.6693,
+      "step": 4710
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.93055316054002e-05,
+      "loss": 0.6593,
+      "step": 4720
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.930180537091644e-05,
+      "loss": 0.5848,
+      "step": 4730
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.929806930799765e-05,
+      "loss": 0.5926,
+      "step": 4740
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9294323418154805e-05,
+      "loss": 0.637,
+      "step": 4750
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.929056770290288e-05,
+      "loss": 0.5765,
+      "step": 4760
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.928680216376079e-05,
+      "loss": 0.6537,
+      "step": 4770
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9283026802251454e-05,
+      "loss": 0.7517,
+      "step": 4780
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.927924161990175e-05,
+      "loss": 0.6206,
+      "step": 4790
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9275446618242516e-05,
+      "loss": 0.6658,
+      "step": 4800
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.927164179880858e-05,
+      "loss": 0.689,
+      "step": 4810
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.926782716313874e-05,
+      "loss": 0.7423,
+      "step": 4820
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9264002712775745e-05,
+      "loss": 0.6781,
+      "step": 4830
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9260168449266335e-05,
+      "loss": 0.6123,
+      "step": 4840
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.925632437416121e-05,
+      "loss": 0.6607,
+      "step": 4850
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.925247048901504e-05,
+      "loss": 0.579,
+      "step": 4860
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.924860679538645e-05,
+      "loss": 0.619,
+      "step": 4870
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.924473329483806e-05,
+      "loss": 0.6327,
+      "step": 4880
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.924084998893642e-05,
+      "loss": 0.5618,
+      "step": 4890
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.923695687925207e-05,
+      "loss": 0.6511,
+      "step": 4900
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.923305396735952e-05,
+      "loss": 0.7701,
+      "step": 4910
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.9229141254837226e-05,
+      "loss": 0.7293,
+      "step": 4920
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.922521874326761e-05,
+      "loss": 0.6308,
+      "step": 4930
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.922128643423707e-05,
+      "loss": 0.6569,
+      "step": 4940
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.921734432933596e-05,
+      "loss": 0.7299,
+      "step": 4950
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.921339243015858e-05,
+      "loss": 0.687,
+      "step": 4960
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.920943073830322e-05,
+      "loss": 0.5625,
+      "step": 4970
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.920545925537212e-05,
+      "loss": 0.5562,
+      "step": 4980
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.9201477982971466e-05,
+      "loss": 0.584,
+      "step": 4990
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.919748692271141e-05,
+      "loss": 0.5762,
+      "step": 5000
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 0.48749640583992004,
+      "eval_runtime": 418.3454,
+      "eval_samples_per_second": 50.726,
+      "eval_steps_per_second": 0.397,
+      "step": 5000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.9193486076206075e-05,
+      "loss": 0.5461,
+      "step": 5010
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.918947544507353e-05,
+      "loss": 0.6191,
+      "step": 5020
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.91854550309358e-05,
+      "loss": 0.6065,
+      "step": 5030
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.9181424835418874e-05,
+      "loss": 0.6403,
+      "step": 5040
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.91773848601527e-05,
+      "loss": 0.6746,
+      "step": 5050
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.9173335106771154e-05,
+      "loss": 0.6852,
+      "step": 5060
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.916927557691211e-05,
+      "loss": 0.813,
+      "step": 5070
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.916520627221736e-05,
+      "loss": 0.6969,
+      "step": 5080
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.9161127194332655e-05,
+      "loss": 0.567,
+      "step": 5090
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.915703834490773e-05,
+      "loss": 0.5092,
+      "step": 5100
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.915293972559623e-05,
+      "loss": 0.637,
+      "step": 5110
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.9148831338055766e-05,
+      "loss": 0.5444,
+      "step": 5120
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.9144713183947915e-05,
+      "loss": 0.5532,
+      "step": 5130
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.914058526493819e-05,
+      "loss": 0.5989,
+      "step": 5140
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.913644758269604e-05,
+      "loss": 0.617,
+      "step": 5150
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.9132300138894895e-05,
+      "loss": 0.6167,
+      "step": 5160
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.912814293521211e-05,
+      "loss": 0.735,
+      "step": 5170
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.9123975973328986e-05,
+      "loss": 0.7821,
+      "step": 5180
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.911979925493079e-05,
+      "loss": 0.5826,
+      "step": 5190
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.91156127817067e-05,
+      "loss": 0.6216,
+      "step": 5200
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.911141655534989e-05,
+      "loss": 0.5913,
+      "step": 5210
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.9107210577557415e-05,
+      "loss": 0.5662,
+      "step": 5220
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.9102994850030344e-05,
+      "loss": 0.6377,
+      "step": 5230
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.909876937447363e-05,
+      "loss": 0.5323,
+      "step": 5240
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.90945341525962e-05,
+      "loss": 0.5167,
+      "step": 5250
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.909028918611091e-05,
+      "loss": 0.681,
+      "step": 5260
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.908603447673457e-05,
+      "loss": 0.6329,
+      "step": 5270
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.9081770026187914e-05,
+      "loss": 0.6075,
+      "step": 5280
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.907749583619562e-05,
+      "loss": 0.6072,
+      "step": 5290
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.907321190848632e-05,
+      "loss": 0.6488,
+      "step": 5300
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.9068918244792565e-05,
+      "loss": 0.6291,
+      "step": 5310
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.906461484685085e-05,
+      "loss": 0.6862,
+      "step": 5320
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.906030171640163e-05,
+      "loss": 0.4359,
+      "step": 5330
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.905597885518923e-05,
+      "loss": 0.5581,
+      "step": 5340
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.9051646264962005e-05,
+      "loss": 0.5444,
+      "step": 5350
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.904730394747216e-05,
+      "loss": 0.5766,
+      "step": 5360
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.904295190447587e-05,
+      "loss": 0.5677,
+      "step": 5370
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.903859013773325e-05,
+      "loss": 0.5348,
+      "step": 5380
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.903421864900835e-05,
+      "loss": 0.6141,
+      "step": 5390
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.902983744006913e-05,
+      "loss": 0.6015,
+      "step": 5400
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.902544651268749e-05,
+      "loss": 0.6371,
+      "step": 5410
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.902104586863927e-05,
+      "loss": 0.5884,
+      "step": 5420
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.901663550970422e-05,
+      "loss": 0.4732,
+      "step": 5430
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.901221543766604e-05,
+      "loss": 0.6512,
+      "step": 5440
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.900778565431236e-05,
+      "loss": 0.5253,
+      "step": 5450
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.900334616143471e-05,
+      "loss": 0.6781,
+      "step": 5460
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.8998896960828576e-05,
+      "loss": 0.5466,
+      "step": 5470
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.899443805429336e-05,
+      "loss": 0.4637,
+      "step": 5480
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.8989969443632366e-05,
+      "loss": 0.4792,
+      "step": 5490
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.898549113065287e-05,
+      "loss": 0.6649,
+      "step": 5500
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 0.44898268580436707,
+      "eval_runtime": 418.3764,
+      "eval_samples_per_second": 50.722,
+      "eval_steps_per_second": 0.397,
+      "step": 5500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.898100311716604e-05,
+      "loss": 0.5524,
+      "step": 5510
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.897650540498697e-05,
+      "loss": 0.5248,
+      "step": 5520
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.8971997995934675e-05,
+      "loss": 0.6922,
+      "step": 5530
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.896748089183211e-05,
+      "loss": 0.5127,
+      "step": 5540
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.896295409450612e-05,
+      "loss": 0.6378,
+      "step": 5550
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.895841760578751e-05,
+      "loss": 0.6006,
+      "step": 5560
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.8953871427510965e-05,
+      "loss": 0.5499,
+      "step": 5570
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.894931556151512e-05,
+      "loss": 0.546,
+      "step": 5580
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.894475000964249e-05,
+      "loss": 0.5088,
+      "step": 5590
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.8940174773739564e-05,
+      "loss": 0.4459,
+      "step": 5600
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.8935589855656694e-05,
+      "loss": 0.6628,
+      "step": 5610
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.893099525724818e-05,
+      "loss": 0.526,
+      "step": 5620
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.892639098037221e-05,
+      "loss": 0.5136,
+      "step": 5630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.892177702689091e-05,
+      "loss": 0.5404,
+      "step": 5640
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.891715339867033e-05,
+      "loss": 0.5318,
+      "step": 5650
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.8912520097580395e-05,
+      "loss": 0.5548,
+      "step": 5660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.890787712549496e-05,
+      "loss": 0.5183,
+      "step": 5670
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.89032244842918e-05,
+      "loss": 0.68,
+      "step": 5680
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.8898562175852594e-05,
+      "loss": 0.4871,
+      "step": 5690
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.8893890202062924e-05,
+      "loss": 0.4945,
+      "step": 5700
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.88892085648123e-05,
+      "loss": 0.4752,
+      "step": 5710
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.888451726599411e-05,
+      "loss": 0.4715,
+      "step": 5720
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.887981630750568e-05,
+      "loss": 0.4668,
+      "step": 5730
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.887510569124822e-05,
+      "loss": 0.5806,
+      "step": 5740
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.887038541912686e-05,
+      "loss": 0.5533,
+      "step": 5750
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.886565549305064e-05,
+      "loss": 0.6291,
+      "step": 5760
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.886091591493248e-05,
+      "loss": 0.6165,
+      "step": 5770
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.885616668668922e-05,
+      "loss": 0.5649,
+      "step": 5780
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.8851407810241614e-05,
+      "loss": 0.5761,
+      "step": 5790
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.884663928751429e-05,
+      "loss": 0.5108,
+      "step": 5800
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.88418611204358e-05,
+      "loss": 0.4481,
+      "step": 5810
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.883707331093859e-05,
+      "loss": 0.6043,
+      "step": 5820
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.8832275860959004e-05,
+      "loss": 0.5655,
+      "step": 5830
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.882746877243728e-05,
+      "loss": 0.5544,
+      "step": 5840
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.8822652047317576e-05,
+      "loss": 0.523,
+      "step": 5850
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.881782568754791e-05,
+      "loss": 0.5768,
+      "step": 5860
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.881298969508023e-05,
+      "loss": 0.5601,
+      "step": 5870
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.8808144071870364e-05,
+      "loss": 0.49,
+      "step": 5880
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.880328881987805e-05,
+      "loss": 0.4818,
+      "step": 5890
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.8798423941066886e-05,
+      "loss": 0.3413,
+      "step": 5900
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.87935494374044e-05,
+      "loss": 0.5083,
+      "step": 5910
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.8788665310862e-05,
+      "loss": 0.5453,
+      "step": 5920
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.878377156341498e-05,
+      "loss": 0.5733,
+      "step": 5930
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.877886819704254e-05,
+      "loss": 0.4787,
+      "step": 5940
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.8773955213727745e-05,
+      "loss": 0.5693,
+      "step": 5950
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.876903261545757e-05,
+      "loss": 0.4449,
+      "step": 5960
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.876410040422288e-05,
+      "loss": 0.5425,
+      "step": 5970
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.8759158582018406e-05,
+      "loss": 0.4394,
+      "step": 5980
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.875420715084279e-05,
+      "loss": 0.6024,
+      "step": 5990
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.874924611269854e-05,
+      "loss": 0.5543,
+      "step": 6000
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 0.40883690118789673,
+      "eval_runtime": 418.0326,
+      "eval_samples_per_second": 50.764,
+      "eval_steps_per_second": 0.397,
+      "step": 6000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.874427546959208e-05,
+      "loss": 0.4781,
+      "step": 6010
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.8739295223533684e-05,
+      "loss": 0.4425,
+      "step": 6020
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.873430537653754e-05,
+      "loss": 0.4825,
+      "step": 6030
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.872930593062168e-05,
+      "loss": 0.4632,
+      "step": 6040
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.872429688780805e-05,
+      "loss": 0.5248,
+      "step": 6050
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.871927825012248e-05,
+      "loss": 0.3932,
+      "step": 6060
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.8714250019594656e-05,
+      "loss": 0.624,
+      "step": 6070
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.8709212198258155e-05,
+      "loss": 0.4842,
+      "step": 6080
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.870416478815045e-05,
+      "loss": 0.6339,
+      "step": 6090
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.8699107791312865e-05,
+      "loss": 0.4489,
+      "step": 6100
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.869404120979062e-05,
+      "loss": 0.6883,
+      "step": 6110
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.868896504563279e-05,
+      "loss": 0.6182,
+      "step": 6120
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.868387930089235e-05,
+      "loss": 0.4031,
+      "step": 6130
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.8678783977626155e-05,
+      "loss": 0.3859,
+      "step": 6140
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.867367907789488e-05,
+      "loss": 0.6153,
+      "step": 6150
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.866856460376315e-05,
+      "loss": 0.4983,
+      "step": 6160
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.86634405572994e-05,
+      "loss": 0.5714,
+      "step": 6170
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.865830694057597e-05,
+      "loss": 0.5521,
+      "step": 6180
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.8653163755669055e-05,
+      "loss": 0.5279,
+      "step": 6190
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.864801100465873e-05,
+      "loss": 0.5233,
+      "step": 6200
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.8642848689628936e-05,
+      "loss": 0.5213,
+      "step": 6210
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.863767681266746e-05,
+      "loss": 0.4372,
+      "step": 6220
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.8632495375866004e-05,
+      "loss": 0.4636,
+      "step": 6230
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.8627304381320096e-05,
+      "loss": 0.4069,
+      "step": 6240
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.862210383112914e-05,
+      "loss": 0.5319,
+      "step": 6250
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.861689372739641e-05,
+      "loss": 0.3865,
+      "step": 6260
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.8611674072229044e-05,
+      "loss": 0.5073,
+      "step": 6270
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.860644486773802e-05,
+      "loss": 0.4799,
+      "step": 6280
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.8601206116038216e-05,
+      "loss": 0.447,
+      "step": 6290
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.859595781924835e-05,
+      "loss": 0.428,
+      "step": 6300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.8590699979491e-05,
+      "loss": 0.4732,
+      "step": 6310
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.85854325988926e-05,
+      "loss": 0.4316,
+      "step": 6320
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.858015567958346e-05,
+      "loss": 0.5887,
+      "step": 6330
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.857486922369773e-05,
+      "loss": 0.3951,
+      "step": 6340
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.856957323337343e-05,
+      "loss": 0.4887,
+      "step": 6350
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.856426771075242e-05,
+      "loss": 0.475,
+      "step": 6360
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.855895265798043e-05,
+      "loss": 0.3888,
+      "step": 6370
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.8553628077207045e-05,
+      "loss": 0.4112,
+      "step": 6380
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.854829397058568e-05,
+      "loss": 0.4162,
+      "step": 6390
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.854295034027364e-05,
+      "loss": 0.4891,
+      "step": 6400
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.8537597188432047e-05,
+      "loss": 0.4153,
+      "step": 6410
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.853223451722589e-05,
+      "loss": 0.4251,
+      "step": 6420
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.852686232882402e-05,
+      "loss": 0.4573,
+      "step": 6430
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.852148062539911e-05,
+      "loss": 0.4512,
+      "step": 6440
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.85160894091277e-05,
+      "loss": 0.4009,
+      "step": 6450
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.8510688682190175e-05,
+      "loss": 0.3795,
+      "step": 6460
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.850527844677075e-05,
+      "loss": 0.5127,
+      "step": 6470
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.849985870505752e-05,
+      "loss": 0.4448,
+      "step": 6480
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.849442945924238e-05,
+      "loss": 0.4285,
+      "step": 6490
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.848899071152111e-05,
+      "loss": 0.459,
+      "step": 6500
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 0.37777015566825867,
+      "eval_runtime": 418.0357,
+      "eval_samples_per_second": 50.764,
+      "eval_steps_per_second": 0.397,
+      "step": 6500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.848354246409331e-05,
+      "loss": 0.5527,
+      "step": 6510
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.847808471916242e-05,
+      "loss": 0.5115,
+      "step": 6520
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.8472617478935746e-05,
+      "loss": 0.3688,
+      "step": 6530
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.846714074562439e-05,
+      "loss": 0.3861,
+      "step": 6540
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.8461654521443347e-05,
+      "loss": 0.3681,
+      "step": 6550
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.845615880861141e-05,
+      "loss": 0.6873,
+      "step": 6560
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.8450653609351225e-05,
+      "loss": 0.4661,
+      "step": 6570
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.844513892588927e-05,
+      "loss": 0.5365,
+      "step": 6580
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.843961476045587e-05,
+      "loss": 0.4212,
+      "step": 6590
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.8434081115285154e-05,
+      "loss": 0.3769,
+      "step": 6600
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.842853799261513e-05,
+      "loss": 0.5944,
+      "step": 6610
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.8422985394687605e-05,
+      "loss": 0.5788,
+      "step": 6620
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.8417423323748235e-05,
+      "loss": 0.3855,
+      "step": 6630
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.8411851782046495e-05,
+      "loss": 0.4508,
+      "step": 6640
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.84062707718357e-05,
+      "loss": 0.5117,
+      "step": 6650
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.840068029537299e-05,
+      "loss": 0.4715,
+      "step": 6660
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.839508035491934e-05,
+      "loss": 0.5006,
+      "step": 6670
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.8389470952739535e-05,
+      "loss": 0.5785,
+      "step": 6680
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.83838520911022e-05,
+      "loss": 0.4302,
+      "step": 6690
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.8378223772279796e-05,
+      "loss": 0.4764,
+      "step": 6700
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.837258599854858e-05,
+      "loss": 0.4053,
+      "step": 6710
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.836693877218867e-05,
+      "loss": 0.5234,
+      "step": 6720
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.8361282095483976e-05,
+      "loss": 0.4821,
+      "step": 6730
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.835561597072224e-05,
+      "loss": 0.4271,
+      "step": 6740
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.834994040019501e-05,
+      "loss": 0.4853,
+      "step": 6750
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.83442553861977e-05,
+      "loss": 0.5864,
+      "step": 6760
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.83385609310295e-05,
+      "loss": 0.5347,
+      "step": 6770
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.833285703699343e-05,
+      "loss": 0.455,
+      "step": 6780
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.832714370639633e-05,
+      "loss": 0.3714,
+      "step": 6790
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.832142094154886e-05,
+      "loss": 0.3763,
+      "step": 6800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.8315688744765485e-05,
+      "loss": 0.391,
+      "step": 6810
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.8309947118364484e-05,
+      "loss": 0.4394,
+      "step": 6820
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.830419606466797e-05,
+      "loss": 0.5285,
+      "step": 6830
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.8298435586001856e-05,
+      "loss": 0.313,
+      "step": 6840
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.8292665684695854e-05,
+      "loss": 0.4531,
+      "step": 6850
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.828688636308351e-05,
+      "loss": 0.3305,
+      "step": 6860
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.8281097623502144e-05,
+      "loss": 0.4083,
+      "step": 6870
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.8275299468292945e-05,
+      "loss": 0.4293,
+      "step": 6880
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.826949189980085e-05,
+      "loss": 0.603,
+      "step": 6890
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.826367492037463e-05,
+      "loss": 0.3804,
+      "step": 6900
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.825784853236687e-05,
+      "loss": 0.4902,
+      "step": 6910
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.825201273813393e-05,
+      "loss": 0.4689,
+      "step": 6920
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.824616754003602e-05,
+      "loss": 0.4018,
+      "step": 6930
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.8240312940437106e-05,
+      "loss": 0.4347,
+      "step": 6940
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.8234448941704984e-05,
+      "loss": 0.4068,
+      "step": 6950
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.822857554621125e-05,
+      "loss": 0.4685,
+      "step": 6960
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.8222692756331283e-05,
+      "loss": 0.536,
+      "step": 6970
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.821680057444428e-05,
+      "loss": 0.3602,
+      "step": 6980
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.8210899002933235e-05,
+      "loss": 0.4639,
+      "step": 6990
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.820498804418493e-05,
+      "loss": 0.5873,
+      "step": 7000
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 0.35154759883880615,
+      "eval_runtime": 418.4869,
+      "eval_samples_per_second": 50.709,
+      "eval_steps_per_second": 0.397,
+      "step": 7000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.8199067700589936e-05,
+      "loss": 0.4068,
+      "step": 7010
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.819313797454265e-05,
+      "loss": 0.4871,
+      "step": 7020
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.818719886844124e-05,
+      "loss": 0.3975,
+      "step": 7030
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.818125038468766e-05,
+      "loss": 0.4742,
+      "step": 7040
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.817529252568768e-05,
+      "loss": 0.5177,
+      "step": 7050
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.816932529385084e-05,
+      "loss": 0.3987,
+      "step": 7060
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.816334869159051e-05,
+      "loss": 0.4887,
+      "step": 7070
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.815736272132378e-05,
+      "loss": 0.5346,
+      "step": 7080
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.815136738547159e-05,
+      "loss": 0.5989,
+      "step": 7090
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.814536268645866e-05,
+      "loss": 0.4946,
+      "step": 7100
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.813934862671347e-05,
+      "loss": 0.2868,
+      "step": 7110
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.81333252086683e-05,
+      "loss": 0.3188,
+      "step": 7120
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.812729243475921e-05,
+      "loss": 0.3084,
+      "step": 7130
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.812125030742607e-05,
+      "loss": 0.4687,
+      "step": 7140
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.8115198829112486e-05,
+      "loss": 0.3589,
+      "step": 7150
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.810913800226588e-05,
+      "loss": 0.5004,
+      "step": 7160
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.810306782933745e-05,
+      "loss": 0.4089,
+      "step": 7170
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.8096988312782174e-05,
+      "loss": 0.457,
+      "step": 7180
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.809089945505879e-05,
+      "loss": 0.5278,
+      "step": 7190
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.808480125862984e-05,
+      "loss": 0.3891,
+      "step": 7200
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.807869372596164e-05,
+      "loss": 0.3798,
+      "step": 7210
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.8072576859524256e-05,
+      "loss": 0.4176,
+      "step": 7220
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.8066450661791546e-05,
+      "loss": 0.3696,
+      "step": 7230
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.8060315135241155e-05,
+      "loss": 0.4555,
+      "step": 7240
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.8054170282354485e-05,
+      "loss": 0.5163,
+      "step": 7250
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.804801610561671e-05,
+      "loss": 0.4524,
+      "step": 7260
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.804185260751678e-05,
+      "loss": 0.3723,
+      "step": 7270
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.8035679790547404e-05,
+      "loss": 0.5742,
+      "step": 7280
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.802949765720508e-05,
+      "loss": 0.3586,
+      "step": 7290
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.802330620999005e-05,
+      "loss": 0.3838,
+      "step": 7300
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.801710545140635e-05,
+      "loss": 0.3558,
+      "step": 7310
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.801089538396176e-05,
+      "loss": 0.4985,
+      "step": 7320
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.800467601016784e-05,
+      "loss": 0.4349,
+      "step": 7330
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.799844733253988e-05,
+      "loss": 0.4238,
+      "step": 7340
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.7992209353596996e-05,
+      "loss": 0.3519,
+      "step": 7350
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.7985962075862e-05,
+      "loss": 0.4092,
+      "step": 7360
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.79797055018615e-05,
+      "loss": 0.3657,
+      "step": 7370
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.797343963412586e-05,
+      "loss": 0.4374,
+      "step": 7380
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.7967164475189195e-05,
+      "loss": 0.3754,
+      "step": 7390
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.796088002758938e-05,
+      "loss": 0.3715,
+      "step": 7400
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.7954586293868064e-05,
+      "loss": 0.4302,
+      "step": 7410
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.794828327657062e-05,
+      "loss": 0.4806,
+      "step": 7420
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.79419709782462e-05,
+      "loss": 0.3517,
+      "step": 7430
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.793564940144769e-05,
+      "loss": 0.5377,
+      "step": 7440
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.792931854873175e-05,
+      "loss": 0.4147,
+      "step": 7450
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.792297842265878e-05,
+      "loss": 0.4127,
+      "step": 7460
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.791662902579293e-05,
+      "loss": 0.3598,
+      "step": 7470
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.791027036070211e-05,
+      "loss": 0.3575,
+      "step": 7480
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.7903902429957955e-05,
+      "loss": 0.3582,
+      "step": 7490
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.7897525236135874e-05,
+      "loss": 0.3679,
+      "step": 7500
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 0.32194384932518005,
+      "eval_runtime": 417.9212,
+      "eval_samples_per_second": 50.778,
+      "eval_steps_per_second": 0.397,
+      "step": 7500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.789113878181499e-05,
+      "loss": 0.4129,
+      "step": 7510
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.788474306957822e-05,
+      "loss": 0.3853,
+      "step": 7520
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.7878338102012185e-05,
+      "loss": 0.5193,
+      "step": 7530
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.7871923881707245e-05,
+      "loss": 0.402,
+      "step": 7540
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.7865500411257546e-05,
+      "loss": 0.3756,
+      "step": 7550
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.7859067693260926e-05,
+      "loss": 0.4442,
+      "step": 7560
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.785262573031899e-05,
+      "loss": 0.4374,
+      "step": 7570
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.784617452503708e-05,
+      "loss": 0.4228,
+      "step": 7580
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.783971408002426e-05,
+      "loss": 0.4584,
+      "step": 7590
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.783324439789335e-05,
+      "loss": 0.3035,
+      "step": 7600
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.7826765481260904e-05,
+      "loss": 0.3636,
+      "step": 7610
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.78202773327472e-05,
+      "loss": 0.3109,
+      "step": 7620
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.781377995497626e-05,
+      "loss": 0.382,
+      "step": 7630
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.780727335057582e-05,
+      "loss": 0.4424,
+      "step": 7640
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.7800757522177374e-05,
+      "loss": 0.4035,
+      "step": 7650
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.779423247241613e-05,
+      "loss": 0.3559,
+      "step": 7660
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.778769820393102e-05,
+      "loss": 0.361,
+      "step": 7670
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.778115471936473e-05,
+      "loss": 0.3457,
+      "step": 7680
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.777460202136365e-05,
+      "loss": 0.312,
+      "step": 7690
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.7768040112577885e-05,
+      "loss": 0.4919,
+      "step": 7700
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.776146899566131e-05,
+      "loss": 0.3933,
+      "step": 7710
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.775488867327148e-05,
+      "loss": 0.4155,
+      "step": 7720
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.7748299148069696e-05,
+      "loss": 0.4706,
+      "step": 7730
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.774170042272098e-05,
+      "loss": 0.4326,
+      "step": 7740
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.773509249989406e-05,
+      "loss": 0.4091,
+      "step": 7750
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.772847538226139e-05,
+      "loss": 0.3552,
+      "step": 7760
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.772184907249916e-05,
+      "loss": 0.4213,
+      "step": 7770
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.7715213573287256e-05,
+      "loss": 0.3572,
+      "step": 7780
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.770856888730929e-05,
+      "loss": 0.421,
+      "step": 7790
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.770191501725258e-05,
+      "loss": 0.4397,
+      "step": 7800
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.769525196580817e-05,
+      "loss": 0.4168,
+      "step": 7810
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.768857973567083e-05,
+      "loss": 0.3568,
+      "step": 7820
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.7681898329539e-05,
+      "loss": 0.2783,
+      "step": 7830
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.767520775011487e-05,
+      "loss": 0.4276,
+      "step": 7840
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.766850800010433e-05,
+      "loss": 0.3645,
+      "step": 7850
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.7661799082216964e-05,
+      "loss": 0.3618,
+      "step": 7860
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.765508099916608e-05,
+      "loss": 0.3337,
+      "step": 7870
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.764835375366869e-05,
+      "loss": 0.4016,
+      "step": 7880
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.7641617348445497e-05,
+      "loss": 0.4969,
+      "step": 7890
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.763487178622095e-05,
+      "loss": 0.3982,
+      "step": 7900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.762811706972313e-05,
+      "loss": 0.2844,
+      "step": 7910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.762135320168391e-05,
+      "loss": 0.4956,
+      "step": 7920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.761458018483877e-05,
+      "loss": 0.528,
+      "step": 7930
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.760779802192698e-05,
+      "loss": 0.3502,
+      "step": 7940
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.760100671569144e-05,
+      "loss": 0.4082,
+      "step": 7950
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.759420626887877e-05,
+      "loss": 0.3614,
+      "step": 7960
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.7587396684239314e-05,
+      "loss": 0.353,
+      "step": 7970
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.758057796452706e-05,
+      "loss": 0.4929,
+      "step": 7980
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.7573750112499736e-05,
+      "loss": 0.342,
+      "step": 7990
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.756691313091874e-05,
+      "loss": 0.316,
+      "step": 8000
+    },
+    {
+      "epoch": 0.32,
+      "eval_loss": 0.31540176272392273,
+      "eval_runtime": 418.1435,
+      "eval_samples_per_second": 50.751,
+      "eval_steps_per_second": 0.397,
+      "step": 8000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.756006702254917e-05,
+      "loss": 0.347,
+      "step": 8010
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.7553211790159815e-05,
+      "loss": 0.3275,
+      "step": 8020
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.754634743652315e-05,
+      "loss": 0.2993,
+      "step": 8030
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.753947396441535e-05,
+      "loss": 0.3933,
+      "step": 8040
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.7532591376616254e-05,
+      "loss": 0.4461,
+      "step": 8050
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.752569967590941e-05,
+      "loss": 0.4575,
+      "step": 8060
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.751879886508204e-05,
+      "loss": 0.4256,
+      "step": 8070
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.751188894692506e-05,
+      "loss": 0.3652,
+      "step": 8080
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.750496992423307e-05,
+      "loss": 0.3424,
+      "step": 8090
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.7498041799804325e-05,
+      "loss": 0.4356,
+      "step": 8100
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.749110457644081e-05,
+      "loss": 0.2382,
+      "step": 8110
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.748415825694815e-05,
+      "loss": 0.4469,
+      "step": 8120
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.7477202844135646e-05,
+      "loss": 0.4062,
+      "step": 8130
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.747023834081632e-05,
+      "loss": 0.4464,
+      "step": 8140
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.7463264749806824e-05,
+      "loss": 0.3861,
+      "step": 8150
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.74562820739275e-05,
+      "loss": 0.3953,
+      "step": 8160
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.7449290316002384e-05,
+      "loss": 0.3564,
+      "step": 8170
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.744228947885915e-05,
+      "loss": 0.3464,
+      "step": 8180
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.743527956532918e-05,
+      "loss": 0.4324,
+      "step": 8190
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.74282605782475e-05,
+      "loss": 0.3053,
+      "step": 8200
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.742123252045281e-05,
+      "loss": 0.2968,
+      "step": 8210
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.74141953947875e-05,
+      "loss": 0.4083,
+      "step": 8220
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.740714920409759e-05,
+      "loss": 0.3365,
+      "step": 8230
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.740009395123281e-05,
+      "loss": 0.3791,
+      "step": 8240
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.739302963904651e-05,
+      "loss": 0.3428,
+      "step": 8250
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.7385956270395735e-05,
+      "loss": 0.2716,
+      "step": 8260
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.737887384814119e-05,
+      "loss": 0.4324,
+      "step": 8270
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.737178237514724e-05,
+      "loss": 0.2427,
+      "step": 8280
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.736468185428188e-05,
+      "loss": 0.3261,
+      "step": 8290
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.735757228841682e-05,
+      "loss": 0.3679,
+      "step": 8300
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.735045368042738e-05,
+      "loss": 0.4742,
+      "step": 8310
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.734332603319255e-05,
+      "loss": 0.3608,
+      "step": 8320
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.733618934959499e-05,
+      "loss": 0.4825,
+      "step": 8330
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.732904363252102e-05,
+      "loss": 0.2265,
+      "step": 8340
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.732188888486057e-05,
+      "loss": 0.3751,
+      "step": 8350
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.731472510950728e-05,
+      "loss": 0.2804,
+      "step": 8360
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.7307552309358384e-05,
+      "loss": 0.3737,
+      "step": 8370
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.73003704873148e-05,
+      "loss": 0.3221,
+      "step": 8380
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.7293179646281117e-05,
+      "loss": 0.3493,
+      "step": 8390
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.7285979789165517e-05,
+      "loss": 0.4198,
+      "step": 8400
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.727877091887985e-05,
+      "loss": 0.3985,
+      "step": 8410
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.727155303833963e-05,
+      "loss": 0.4404,
+      "step": 8420
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.7264326150464e-05,
+      "loss": 0.34,
+      "step": 8430
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.725709025817574e-05,
+      "loss": 0.3819,
+      "step": 8440
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.724984536440129e-05,
+      "loss": 0.3809,
+      "step": 8450
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.724259147207071e-05,
+      "loss": 0.3328,
+      "step": 8460
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.723532858411771e-05,
+      "loss": 0.3414,
+      "step": 8470
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.722805670347963e-05,
+      "loss": 0.3858,
+      "step": 8480
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.7220775833097465e-05,
+      "loss": 0.4932,
+      "step": 8490
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.721348597591583e-05,
+      "loss": 0.4315,
+      "step": 8500
+    },
+    {
+      "epoch": 0.34,
+      "eval_loss": 0.2860631048679352,
+      "eval_runtime": 417.7621,
+      "eval_samples_per_second": 50.797,
+      "eval_steps_per_second": 0.397,
+      "step": 8500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.7206187134882977e-05,
+      "loss": 0.3599,
+      "step": 8510
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.719887931295079e-05,
+      "loss": 0.3814,
+      "step": 8520
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.71915625130748e-05,
+      "loss": 0.3903,
+      "step": 8530
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.718423673821415e-05,
+      "loss": 0.3916,
+      "step": 8540
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.717690199133161e-05,
+      "loss": 0.4121,
+      "step": 8550
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.716955827539361e-05,
+      "loss": 0.2989,
+      "step": 8560
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.7162205593370155e-05,
+      "loss": 0.3213,
+      "step": 8570
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.715484394823494e-05,
+      "loss": 0.3213,
+      "step": 8580
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.7147473342965234e-05,
+      "loss": 0.2916,
+      "step": 8590
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.714009378054195e-05,
+      "loss": 0.2974,
+      "step": 8600
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.7132705263949626e-05,
+      "loss": 0.3117,
+      "step": 8610
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.712530779617642e-05,
+      "loss": 0.3203,
+      "step": 8620
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.71179013802141e-05,
+      "loss": 0.3879,
+      "step": 8630
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.7110486019058056e-05,
+      "loss": 0.2667,
+      "step": 8640
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.71030617157073e-05,
+      "loss": 0.4758,
+      "step": 8650
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.709562847316446e-05,
+      "loss": 0.3329,
+      "step": 8660
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.70881862944358e-05,
+      "loss": 0.4659,
+      "step": 8670
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.7080735182531155e-05,
+      "loss": 0.4398,
+      "step": 8680
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.7073275140464e-05,
+      "loss": 0.535,
+      "step": 8690
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.7065806171251416e-05,
+      "loss": 0.4053,
+      "step": 8700
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.70583282779141e-05,
+      "loss": 0.3622,
+      "step": 8710
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.705084146347635e-05,
+      "loss": 0.454,
+      "step": 8720
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.704334573096608e-05,
+      "loss": 0.3315,
+      "step": 8730
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.703584108341481e-05,
+      "loss": 0.4026,
+      "step": 8740
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.702832752385764e-05,
+      "loss": 0.3361,
+      "step": 8750
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.702080505533332e-05,
+      "loss": 0.2869,
+      "step": 8760
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.7013273680884174e-05,
+      "loss": 0.3038,
+      "step": 8770
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.7005733403556115e-05,
+      "loss": 0.3016,
+      "step": 8780
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.69981842263987e-05,
+      "loss": 0.3292,
+      "step": 8790
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.699062615246506e-05,
+      "loss": 0.2753,
+      "step": 8800
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.69830591848119e-05,
+      "loss": 0.3193,
+      "step": 8810
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.697548332649956e-05,
+      "loss": 0.4299,
+      "step": 8820
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.696789858059196e-05,
+      "loss": 0.2904,
+      "step": 8830
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.696030495015663e-05,
+      "loss": 0.3724,
+      "step": 8840
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.695270243826466e-05,
+      "loss": 0.3208,
+      "step": 8850
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.694509104799076e-05,
+      "loss": 0.36,
+      "step": 8860
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.693747078241322e-05,
+      "loss": 0.2768,
+      "step": 8870
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.692984164461392e-05,
+      "loss": 0.4259,
+      "step": 8880
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.692220363767833e-05,
+      "loss": 0.4621,
+      "step": 8890
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.6914556764695517e-05,
+      "loss": 0.2619,
+      "step": 8900
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.690690102875811e-05,
+      "loss": 0.337,
+      "step": 8910
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.6899236432962346e-05,
+      "loss": 0.3201,
+      "step": 8920
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.689156298040802e-05,
+      "loss": 0.3273,
+      "step": 8930
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.688388067419854e-05,
+      "loss": 0.2428,
+      "step": 8940
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.6876189517440874e-05,
+      "loss": 0.325,
+      "step": 8950
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.686848951324557e-05,
+      "loss": 0.3229,
+      "step": 8960
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.686078066472677e-05,
+      "loss": 0.3566,
+      "step": 8970
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.6853062975002166e-05,
+      "loss": 0.448,
+      "step": 8980
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.684533644719305e-05,
+      "loss": 0.2828,
+      "step": 8990
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.683760108442428e-05,
+      "loss": 0.3733,
+      "step": 9000
+    },
+    {
+      "epoch": 0.36,
+      "eval_loss": 0.27112504839897156,
+      "eval_runtime": 418.2349,
+      "eval_samples_per_second": 50.739,
+      "eval_steps_per_second": 0.397,
+      "step": 9000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.6829856889824276e-05,
+      "loss": 0.2642,
+      "step": 9010
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.6822103866525054e-05,
+      "loss": 0.3161,
+      "step": 9020
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.681434201766218e-05,
+      "loss": 0.4087,
+      "step": 9030
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.680657134637479e-05,
+      "loss": 0.3215,
+      "step": 9040
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.679879185580562e-05,
+      "loss": 0.2724,
+      "step": 9050
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.679100354910092e-05,
+      "loss": 0.2951,
+      "step": 9060
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.678320642941054e-05,
+      "loss": 0.2948,
+      "step": 9070
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.677540049988789e-05,
+      "loss": 0.356,
+      "step": 9080
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.676758576368994e-05,
+      "loss": 0.3685,
+      "step": 9090
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.6759762223977224e-05,
+      "loss": 0.3844,
+      "step": 9100
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.6751929883913834e-05,
+      "loss": 0.3448,
+      "step": 9110
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.6744088746667416e-05,
+      "loss": 0.3643,
+      "step": 9120
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.6736238815409174e-05,
+      "loss": 0.4536,
+      "step": 9130
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.6729166361020526e-05,
+      "loss": 0.3612,
+      "step": 9140
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.672129972988926e-05,
+      "loss": 0.4602,
+      "step": 9150
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.671342431396279e-05,
+      "loss": 0.3286,
+      "step": 9160
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.670554011642619e-05,
+      "loss": 0.2872,
+      "step": 9170
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.669764714046808e-05,
+      "loss": 0.443,
+      "step": 9180
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.6689745389280646e-05,
+      "loss": 0.3235,
+      "step": 9190
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.66818348660596e-05,
+      "loss": 0.4508,
+      "step": 9200
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.6673915574004225e-05,
+      "loss": 0.3091,
+      "step": 9210
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.666598751631733e-05,
+      "loss": 0.3964,
+      "step": 9220
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.6658050696205276e-05,
+      "loss": 0.3217,
+      "step": 9230
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.665010511687797e-05,
+      "loss": 0.253,
+      "step": 9240
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.6642150781548866e-05,
+      "loss": 0.2719,
+      "step": 9250
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.663418769343495e-05,
+      "loss": 0.254,
+      "step": 9260
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.662621585575676e-05,
+      "loss": 0.3581,
+      "step": 9270
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.661823527173835e-05,
+      "loss": 0.2885,
+      "step": 9280
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.6610245944607334e-05,
+      "loss": 0.3114,
+      "step": 9290
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.660224787759486e-05,
+      "loss": 0.2502,
+      "step": 9300
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.6594241073935596e-05,
+      "loss": 0.2833,
+      "step": 9310
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.658622553686776e-05,
+      "loss": 0.3415,
+      "step": 9320
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.6578201269633074e-05,
+      "loss": 0.2826,
+      "step": 9330
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.6570168275476835e-05,
+      "loss": 0.2739,
+      "step": 9340
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.656212655764783e-05,
+      "loss": 0.3696,
+      "step": 9350
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.6554076119398404e-05,
+      "loss": 0.292,
+      "step": 9360
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.654601696398441e-05,
+      "loss": 0.2558,
+      "step": 9370
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.653794909466521e-05,
+      "loss": 0.303,
+      "step": 9380
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.6529872514703744e-05,
+      "loss": 0.3714,
+      "step": 9390
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.652259614783889e-05,
+      "loss": 0.2222,
+      "step": 9400
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.6514503026659005e-05,
+      "loss": 0.2817,
+      "step": 9410
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.650640120431918e-05,
+      "loss": 0.2228,
+      "step": 9420
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.6498290684096064e-05,
+      "loss": 0.3488,
+      "step": 9430
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.6490171469269803e-05,
+      "loss": 0.3142,
+      "step": 9440
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.6482043563124075e-05,
+      "loss": 0.3621,
+      "step": 9450
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.6473906968946066e-05,
+      "loss": 0.2929,
+      "step": 9460
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.6465761690026485e-05,
+      "loss": 0.286,
+      "step": 9470
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.645760772965954e-05,
+      "loss": 0.2405,
+      "step": 9480
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.644944509114294e-05,
+      "loss": 0.4306,
+      "step": 9490
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.644127377777795e-05,
+      "loss": 0.3735,
+      "step": 9500
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 0.26075831055641174,
+      "eval_runtime": 417.9926,
+      "eval_samples_per_second": 50.769,
+      "eval_steps_per_second": 0.397,
+      "step": 9500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.6433093792869285e-05,
+      "loss": 0.3406,
+      "step": 9510
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.642490513972522e-05,
+      "loss": 0.3247,
+      "step": 9520
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.641670782165749e-05,
+      "loss": 0.2977,
+      "step": 9530
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.640850184198137e-05,
+      "loss": 0.38,
+      "step": 9540
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.640028720401562e-05,
+      "loss": 0.3872,
+      "step": 9550
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.6392063911082494e-05,
+      "loss": 0.3314,
+      "step": 9560
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.638383196650777e-05,
+      "loss": 0.3301,
+      "step": 9570
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.637559137362071e-05,
+      "loss": 0.3195,
+      "step": 9580
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.6367342135754066e-05,
+      "loss": 0.3514,
+      "step": 9590
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.635908425624411e-05,
+      "loss": 0.2632,
+      "step": 9600
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.6350817738430594e-05,
+      "loss": 0.2799,
+      "step": 9610
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.6342542585656754e-05,
+      "loss": 0.299,
+      "step": 9620
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.6334258801269336e-05,
+      "loss": 0.29,
+      "step": 9630
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.632596638861857e-05,
+      "loss": 0.2172,
+      "step": 9640
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.631766535105817e-05,
+      "loss": 0.3141,
+      "step": 9650
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.630935569194535e-05,
+      "loss": 0.2964,
+      "step": 9660
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.63010374146408e-05,
+      "loss": 0.409,
+      "step": 9670
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.629271052250871e-05,
+      "loss": 0.3562,
+      "step": 9680
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.628437501891674e-05,
+      "loss": 0.3286,
+      "step": 9690
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.627603090723602e-05,
+      "loss": 0.3932,
+      "step": 9700
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.62676781908412e-05,
+      "loss": 0.3177,
+      "step": 9710
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.6259316873110384e-05,
+      "loss": 0.3259,
+      "step": 9720
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.625094695742516e-05,
+      "loss": 0.2867,
+      "step": 9730
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.6242568447170576e-05,
+      "loss": 0.3716,
+      "step": 9740
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.6234181345735186e-05,
+      "loss": 0.392,
+      "step": 9750
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.6225785656511e-05,
+      "loss": 0.3015,
+      "step": 9760
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.6217381382893516e-05,
+      "loss": 0.311,
+      "step": 9770
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.620896852828168e-05,
+      "loss": 0.3418,
+      "step": 9780
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.6200547096077925e-05,
+      "loss": 0.3661,
+      "step": 9790
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.619211708968817e-05,
+      "loss": 0.2459,
+      "step": 9800
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.618367851252175e-05,
+      "loss": 0.3364,
+      "step": 9810
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.6175231367991515e-05,
+      "loss": 0.3878,
+      "step": 9820
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.616677565951377e-05,
+      "loss": 0.4193,
+      "step": 9830
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.615831139050826e-05,
+      "loss": 0.2775,
+      "step": 9840
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.6149838564398226e-05,
+      "loss": 0.4107,
+      "step": 9850
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.614135718461034e-05,
+      "loss": 0.3571,
+      "step": 9860
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.613286725457474e-05,
+      "loss": 0.2986,
+      "step": 9870
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.6124368777725046e-05,
+      "loss": 0.2805,
+      "step": 9880
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.61158617574983e-05,
+      "loss": 0.4153,
+      "step": 9890
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.610734619733503e-05,
+      "loss": 0.3451,
+      "step": 9900
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.6098822100679203e-05,
+      "loss": 0.3163,
+      "step": 9910
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.609028947097823e-05,
+      "loss": 0.3261,
+      "step": 9920
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.608174831168298e-05,
+      "loss": 0.2511,
+      "step": 9930
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.6073198626247785e-05,
+      "loss": 0.2431,
+      "step": 9940
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.60646404181304e-05,
+      "loss": 0.3688,
+      "step": 9950
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.6056073690792056e-05,
+      "loss": 0.2687,
+      "step": 9960
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.6047498447697405e-05,
+      "loss": 0.3475,
+      "step": 9970
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.603891469231455e-05,
+      "loss": 0.3253,
+      "step": 9980
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.603032242811505e-05,
+      "loss": 0.3304,
+      "step": 9990
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.602172165857389e-05,
+      "loss": 0.2522,
+      "step": 10000
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 0.242937833070755,
+      "eval_runtime": 417.9609,
+      "eval_samples_per_second": 50.773,
+      "eval_steps_per_second": 0.397,
+      "step": 10000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.601311238716949e-05,
+      "loss": 0.3334,
+      "step": 10010
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.600449461738372e-05,
+      "loss": 0.259,
+      "step": 10020
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.59958683527019e-05,
+      "loss": 0.3929,
+      "step": 10030
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.598723359661274e-05,
+      "loss": 0.2661,
+      "step": 10040
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.597859035260845e-05,
+      "loss": 0.4051,
+      "step": 10050
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.5969938624184616e-05,
+      "loss": 0.2525,
+      "step": 10060
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.596127841484028e-05,
+      "loss": 0.2723,
+      "step": 10070
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.5952609728077906e-05,
+      "loss": 0.2254,
+      "step": 10080
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.59439325674034e-05,
+      "loss": 0.3768,
+      "step": 10090
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.5935246936326076e-05,
+      "loss": 0.2933,
+      "step": 10100
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.59265528383587e-05,
+      "loss": 0.3771,
+      "step": 10110
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.5917850277017415e-05,
+      "loss": 0.2795,
+      "step": 10120
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.590913925582186e-05,
+      "loss": 0.3644,
+      "step": 10130
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.5900419778295015e-05,
+      "loss": 0.2976,
+      "step": 10140
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.589169184796335e-05,
+      "loss": 0.2111,
+      "step": 10150
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.5882955468356695e-05,
+      "loss": 0.2877,
+      "step": 10160
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.5874210643008334e-05,
+      "loss": 0.25,
+      "step": 10170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.586545737545497e-05,
+      "loss": 0.3163,
+      "step": 10180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.585669566923669e-05,
+      "loss": 0.2228,
+      "step": 10190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.5847925527897015e-05,
+      "loss": 0.3223,
+      "step": 10200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.583914695498288e-05,
+      "loss": 0.4331,
+      "step": 10210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.583035995404461e-05,
+      "loss": 0.3516,
+      "step": 10220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.582156452863597e-05,
+      "loss": 0.2665,
+      "step": 10230
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.5812760682314096e-05,
+      "loss": 0.3126,
+      "step": 10240
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.5803948418639565e-05,
+      "loss": 0.3282,
+      "step": 10250
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.5795127741176325e-05,
+      "loss": 0.2087,
+      "step": 10260
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.5786298653491753e-05,
+      "loss": 0.3262,
+      "step": 10270
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.5777461159156623e-05,
+      "loss": 0.3702,
+      "step": 10280
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.576861526174508e-05,
+      "loss": 0.3947,
+      "step": 10290
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.575976096483472e-05,
+      "loss": 0.3133,
+      "step": 10300
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.575089827200649e-05,
+      "loss": 0.3047,
+      "step": 10310
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.5742027186844745e-05,
+      "loss": 0.3666,
+      "step": 10320
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.573314771293724e-05,
+      "loss": 0.2344,
+      "step": 10330
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.572425985387514e-05,
+      "loss": 0.3484,
+      "step": 10340
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.5715363613252957e-05,
+      "loss": 0.2752,
+      "step": 10350
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.570645899466864e-05,
+      "loss": 0.3478,
+      "step": 10360
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.569754600172348e-05,
+      "loss": 0.2665,
+      "step": 10370
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.56886246380222e-05,
+      "loss": 0.333,
+      "step": 10380
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.567969490717289e-05,
+      "loss": 0.2743,
+      "step": 10390
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.5670756812786995e-05,
+      "loss": 0.2733,
+      "step": 10400
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.56618103584794e-05,
+      "loss": 0.4219,
+      "step": 10410
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.565285554786831e-05,
+      "loss": 0.4045,
+      "step": 10420
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.564389238457537e-05,
+      "loss": 0.3422,
+      "step": 10430
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.563492087222556e-05,
+      "loss": 0.274,
+      "step": 10440
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.5625941014447236e-05,
+      "loss": 0.2901,
+      "step": 10450
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.561695281487216e-05,
+      "loss": 0.314,
+      "step": 10460
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.5607956277135446e-05,
+      "loss": 0.313,
+      "step": 10470
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.5598951404875577e-05,
+      "loss": 0.2365,
+      "step": 10480
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.558993820173443e-05,
+      "loss": 0.2133,
+      "step": 10490
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.558091667135722e-05,
+      "loss": 0.3149,
+      "step": 10500
+    },
+    {
+      "epoch": 0.42,
+      "eval_loss": 0.24121485650539398,
+      "eval_runtime": 418.2538,
+      "eval_samples_per_second": 50.737,
+      "eval_steps_per_second": 0.397,
+      "step": 10500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.557188681739256e-05,
+      "loss": 0.3959,
+      "step": 10510
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.5562848643492395e-05,
+      "loss": 0.329,
+      "step": 10520
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.5553802153312065e-05,
+      "loss": 0.2977,
+      "step": 10530
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.554474735051027e-05,
+      "loss": 0.3745,
+      "step": 10540
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.553568423874905e-05,
+      "loss": 0.5012,
+      "step": 10550
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.552661282169384e-05,
+      "loss": 0.2026,
+      "step": 10560
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.5517533103013385e-05,
+      "loss": 0.2982,
+      "step": 10570
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.550844508637985e-05,
+      "loss": 0.3731,
+      "step": 10580
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.54993487754687e-05,
+      "loss": 0.2941,
+      "step": 10590
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.5490244173958776e-05,
+      "loss": 0.313,
+      "step": 10600
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.5481131285532285e-05,
+      "loss": 0.249,
+      "step": 10610
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.5472010113874766e-05,
+      "loss": 0.2765,
+      "step": 10620
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.5462880662675115e-05,
+      "loss": 0.2644,
+      "step": 10630
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.545374293562559e-05,
+      "loss": 0.2764,
+      "step": 10640
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.544459693642177e-05,
+      "loss": 0.3042,
+      "step": 10650
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.543544266876258e-05,
+      "loss": 0.2831,
+      "step": 10660
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.542628013635034e-05,
+      "loss": 0.273,
+      "step": 10670
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.541710934289063e-05,
+      "loss": 0.4041,
+      "step": 10680
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.5407930292092446e-05,
+      "loss": 0.3134,
+      "step": 10690
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.539874298766808e-05,
+      "loss": 0.239,
+      "step": 10700
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.538954743333318e-05,
+      "loss": 0.3031,
+      "step": 10710
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.538034363280672e-05,
+      "loss": 0.2875,
+      "step": 10720
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.537113158981101e-05,
+      "loss": 0.3123,
+      "step": 10730
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.536191130807171e-05,
+      "loss": 0.2419,
+      "step": 10740
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.53526827913178e-05,
+      "loss": 0.3613,
+      "step": 10750
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.534344604328157e-05,
+      "loss": 0.2814,
+      "step": 10760
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.533420106769868e-05,
+      "loss": 0.3387,
+      "step": 10770
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.5324947868308096e-05,
+      "loss": 0.2335,
+      "step": 10780
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.53156864488521e-05,
+      "loss": 0.285,
+      "step": 10790
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.530641681307631e-05,
+      "loss": 0.2223,
+      "step": 10800
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.529713896472968e-05,
+      "loss": 0.2491,
+      "step": 10810
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.5287852907564455e-05,
+      "loss": 0.274,
+      "step": 10820
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.5278558645336224e-05,
+      "loss": 0.2613,
+      "step": 10830
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.526925618180388e-05,
+      "loss": 0.2319,
+      "step": 10840
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.525994552072965e-05,
+      "loss": 0.3251,
+      "step": 10850
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.525062666587906e-05,
+      "loss": 0.2676,
+      "step": 10860
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.524129962102097e-05,
+      "loss": 0.2367,
+      "step": 10870
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.523196438992751e-05,
+      "loss": 0.2414,
+      "step": 10880
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.522262097637418e-05,
+      "loss": 0.2246,
+      "step": 10890
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.521326938413972e-05,
+      "loss": 0.4593,
+      "step": 10900
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.520390961700626e-05,
+      "loss": 0.2566,
+      "step": 10910
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.519454167875916e-05,
+      "loss": 0.2522,
+      "step": 10920
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.5185165573187134e-05,
+      "loss": 0.3219,
+      "step": 10930
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.517578130408217e-05,
+      "loss": 0.2272,
+      "step": 10940
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.516638887523958e-05,
+      "loss": 0.3827,
+      "step": 10950
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.515698829045796e-05,
+      "loss": 0.3493,
+      "step": 10960
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.514757955353921e-05,
+      "loss": 0.3103,
+      "step": 10970
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.513816266828853e-05,
+      "loss": 0.233,
+      "step": 10980
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.512873763851441e-05,
+      "loss": 0.3589,
+      "step": 10990
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.5119304468028635e-05,
+      "loss": 0.2791,
+      "step": 11000
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 0.23625946044921875,
+      "eval_runtime": 419.0561,
+      "eval_samples_per_second": 50.64,
+      "eval_steps_per_second": 0.396,
+      "step": 11000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.510986316064629e-05,
+      "loss": 0.2868,
+      "step": 11010
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.510041372018573e-05,
+      "loss": 0.3129,
+      "step": 11020
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.5090956150468646e-05,
+      "loss": 0.2409,
+      "step": 11030
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.508149045531994e-05,
+      "loss": 0.2923,
+      "step": 11040
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.507201663856788e-05,
+      "loss": 0.2246,
+      "step": 11050
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.506253470404397e-05,
+      "loss": 0.2345,
+      "step": 11060
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.505304465558302e-05,
+      "loss": 0.2791,
+      "step": 11070
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.504354649702309e-05,
+      "loss": 0.271,
+      "step": 11080
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.5034040232205564e-05,
+      "loss": 0.3207,
+      "step": 11090
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.5024525864975074e-05,
+      "loss": 0.2934,
+      "step": 11100
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.501500339917955e-05,
+      "loss": 0.2384,
+      "step": 11110
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.5005472838670155e-05,
+      "loss": 0.2569,
+      "step": 11120
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.499593418730138e-05,
+      "loss": 0.2454,
+      "step": 11130
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.4986387448930956e-05,
+      "loss": 0.27,
+      "step": 11140
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.4976832627419896e-05,
+      "loss": 0.2404,
+      "step": 11150
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.496726972663248e-05,
+      "loss": 0.2864,
+      "step": 11160
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.4957698750436245e-05,
+      "loss": 0.2629,
+      "step": 11170
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.4948119702702005e-05,
+      "loss": 0.1688,
+      "step": 11180
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.493853258730385e-05,
+      "loss": 0.3921,
+      "step": 11190
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.49289374081191e-05,
+      "loss": 0.4271,
+      "step": 11200
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.491933416902837e-05,
+      "loss": 0.2344,
+      "step": 11210
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.4909722873915504e-05,
+      "loss": 0.3114,
+      "step": 11220
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.490010352666763e-05,
+      "loss": 0.3139,
+      "step": 11230
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.4890476131175135e-05,
+      "loss": 0.2721,
+      "step": 11240
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.488084069133163e-05,
+      "loss": 0.3234,
+      "step": 11250
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.4871197211034014e-05,
+      "loss": 0.3028,
+      "step": 11260
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.4861545694182396e-05,
+      "loss": 0.2338,
+      "step": 11270
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.485188614468019e-05,
+      "loss": 0.2631,
+      "step": 11280
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.484221856643401e-05,
+      "loss": 0.2845,
+      "step": 11290
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.483254296335375e-05,
+      "loss": 0.2829,
+      "step": 11300
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.4822859339352524e-05,
+      "loss": 0.3305,
+      "step": 11310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.481316769834671e-05,
+      "loss": 0.2288,
+      "step": 11320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.4803468044255924e-05,
+      "loss": 0.3006,
+      "step": 11330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.479376038100301e-05,
+      "loss": 0.2839,
+      "step": 11340
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.478404471251407e-05,
+      "loss": 0.2562,
+      "step": 11350
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.477432104271844e-05,
+      "loss": 0.2592,
+      "step": 11360
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.476458937554866e-05,
+      "loss": 0.2437,
+      "step": 11370
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.475484971494055e-05,
+      "loss": 0.2622,
+      "step": 11380
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.474510206483315e-05,
+      "loss": 0.364,
+      "step": 11390
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.473534642916871e-05,
+      "loss": 0.3147,
+      "step": 11400
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.4725582811892716e-05,
+      "loss": 0.2872,
+      "step": 11410
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.4715811216953916e-05,
+      "loss": 0.3939,
+      "step": 11420
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.4706031648304246e-05,
+      "loss": 0.2465,
+      "step": 11430
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.469624410989888e-05,
+      "loss": 0.2673,
+      "step": 11440
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.468644860569621e-05,
+      "loss": 0.3001,
+      "step": 11450
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.4676645139657845e-05,
+      "loss": 0.259,
+      "step": 11460
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.466683371574865e-05,
+      "loss": 0.4362,
+      "step": 11470
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.4657014337936664e-05,
+      "loss": 0.2505,
+      "step": 11480
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.464718701019317e-05,
+      "loss": 0.4108,
+      "step": 11490
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.463735173649265e-05,
+      "loss": 0.2409,
+      "step": 11500
+    },
+    {
+      "epoch": 0.46,
+      "eval_loss": 0.2285226732492447,
+      "eval_runtime": 417.8639,
+      "eval_samples_per_second": 50.784,
+      "eval_steps_per_second": 0.397,
+      "step": 11500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.46275085208128e-05,
+      "loss": 0.249,
+      "step": 11510
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.4617657367134546e-05,
+      "loss": 0.398,
+      "step": 11520
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.4607798279442015e-05,
+      "loss": 0.2894,
+      "step": 11530
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.459793126172253e-05,
+      "loss": 0.3604,
+      "step": 11540
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.458805631796663e-05,
+      "loss": 0.2509,
+      "step": 11550
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.4578173452168074e-05,
+      "loss": 0.2936,
+      "step": 11560
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.4568282668323804e-05,
+      "loss": 0.2997,
+      "step": 11570
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.455838397043397e-05,
+      "loss": 0.3713,
+      "step": 11580
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.454847736250194e-05,
+      "loss": 0.2575,
+      "step": 11590
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.4538562848534246e-05,
+      "loss": 0.3296,
+      "step": 11600
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.452864043254065e-05,
+      "loss": 0.3477,
+      "step": 11610
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.451871011853409e-05,
+      "loss": 0.2552,
+      "step": 11620
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.450877191053071e-05,
+      "loss": 0.2681,
+      "step": 11630
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.449882581254984e-05,
+      "loss": 0.3606,
+      "step": 11640
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.448887182861401e-05,
+      "loss": 0.2225,
+      "step": 11650
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.447890996274893e-05,
+      "loss": 0.2933,
+      "step": 11660
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.4468940218983495e-05,
+      "loss": 0.3829,
+      "step": 11670
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.44589626013498e-05,
+      "loss": 0.2162,
+      "step": 11680
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.44489771138831e-05,
+      "loss": 0.4018,
+      "step": 11690
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.4438983760621864e-05,
+      "loss": 0.252,
+      "step": 11700
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.442898254560773e-05,
+      "loss": 0.2617,
+      "step": 11710
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.441897347288549e-05,
+      "loss": 0.2662,
+      "step": 11720
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.4408956546503165e-05,
+      "loss": 0.264,
+      "step": 11730
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.439893177051191e-05,
+      "loss": 0.2628,
+      "step": 11740
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.438889914896606e-05,
+      "loss": 0.3476,
+      "step": 11750
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.437885868592314e-05,
+      "loss": 0.2045,
+      "step": 11760
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.4368810385443835e-05,
+      "loss": 0.2709,
+      "step": 11770
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.435875425159201e-05,
+      "loss": 0.1685,
+      "step": 11780
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.434869028843468e-05,
+      "loss": 0.2683,
+      "step": 11790
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.433861850004204e-05,
+      "loss": 0.2069,
+      "step": 11800
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.432853889048745e-05,
+      "loss": 0.3237,
+      "step": 11810
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.431845146384743e-05,
+      "loss": 0.2029,
+      "step": 11820
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.430835622420165e-05,
+      "loss": 0.2304,
+      "step": 11830
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.4298253175632967e-05,
+      "loss": 0.2708,
+      "step": 11840
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.428814232222737e-05,
+      "loss": 0.1854,
+      "step": 11850
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.4278023668074016e-05,
+      "loss": 0.2048,
+      "step": 11860
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.426789721726522e-05,
+      "loss": 0.4315,
+      "step": 11870
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.425776297389643e-05,
+      "loss": 0.2235,
+      "step": 11880
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.4247620942066284e-05,
+      "loss": 0.3601,
+      "step": 11890
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.423747112587654e-05,
+      "loss": 0.3081,
+      "step": 11900
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.42273135294321e-05,
+      "loss": 0.2509,
+      "step": 11910
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.421714815684104e-05,
+      "loss": 0.2872,
+      "step": 11920
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.420697501221456e-05,
+      "loss": 0.2778,
+      "step": 11930
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.419679409966699e-05,
+      "loss": 0.2501,
+      "step": 11940
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.418660542331585e-05,
+      "loss": 0.1815,
+      "step": 11950
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.417640898728176e-05,
+      "loss": 0.1513,
+      "step": 11960
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.416620479568847e-05,
+      "loss": 0.274,
+      "step": 11970
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.415599285266291e-05,
+      "loss": 0.2325,
+      "step": 11980
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.41457731623351e-05,
+      "loss": 0.2299,
+      "step": 11990
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.413554572883822e-05,
+      "loss": 0.2355,
+      "step": 12000
+    },
+    {
+      "epoch": 0.48,
+      "eval_loss": 0.22013269364833832,
+      "eval_runtime": 418.2119,
+      "eval_samples_per_second": 50.742,
+      "eval_steps_per_second": 0.397,
+      "step": 12000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.4125310556308576e-05,
+      "loss": 0.2985,
+      "step": 12010
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.4115067648885605e-05,
+      "loss": 0.3294,
+      "step": 12020
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.410481701071186e-05,
+      "loss": 0.3906,
+      "step": 12030
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.409455864593305e-05,
+      "loss": 0.1957,
+      "step": 12040
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.4084292558697964e-05,
+      "loss": 0.2096,
+      "step": 12050
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.407401875315855e-05,
+      "loss": 0.2658,
+      "step": 12060
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.406373723346987e-05,
+      "loss": 0.2649,
+      "step": 12070
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.4053448003790105e-05,
+      "loss": 0.3154,
+      "step": 12080
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.4043151068280544e-05,
+      "loss": 0.257,
+      "step": 12090
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.40328464311056e-05,
+      "loss": 0.2027,
+      "step": 12100
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.402253409643281e-05,
+      "loss": 0.2089,
+      "step": 12110
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.4012214068432805e-05,
+      "loss": 0.3156,
+      "step": 12120
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.400188635127934e-05,
+      "loss": 0.4385,
+      "step": 12130
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.399155094914929e-05,
+      "loss": 0.2317,
+      "step": 12140
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.3981207866222605e-05,
+      "loss": 0.335,
+      "step": 12150
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.3970857106682375e-05,
+      "loss": 0.206,
+      "step": 12160
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.3960498674714775e-05,
+      "loss": 0.3536,
+      "step": 12170
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.395013257450911e-05,
+      "loss": 0.1917,
+      "step": 12180
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.393975881025773e-05,
+      "loss": 0.2163,
+      "step": 12190
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.3929377386156144e-05,
+      "loss": 0.2391,
+      "step": 12200
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.391898830640293e-05,
+      "loss": 0.2252,
+      "step": 12210
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.3908591575199777e-05,
+      "loss": 0.214,
+      "step": 12220
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.389818719675144e-05,
+      "loss": 0.3494,
+      "step": 12230
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.388777517526579e-05,
+      "loss": 0.2315,
+      "step": 12240
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.38773555149538e-05,
+      "loss": 0.2348,
+      "step": 12250
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.38669282200295e-05,
+      "loss": 0.2268,
+      "step": 12260
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.385649329471004e-05,
+      "loss": 0.2337,
+      "step": 12270
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.3846050743215626e-05,
+      "loss": 0.3385,
+      "step": 12280
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.383560056976958e-05,
+      "loss": 0.2745,
+      "step": 12290
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.3825142778598265e-05,
+      "loss": 0.2586,
+      "step": 12300
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.3814677373931175e-05,
+      "loss": 0.1934,
+      "step": 12310
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.380420436000084e-05,
+      "loss": 0.3108,
+      "step": 12320
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.37937237410429e-05,
+      "loss": 0.1911,
+      "step": 12330
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.378323552129604e-05,
+      "loss": 0.1881,
+      "step": 12340
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.377273970500205e-05,
+      "loss": 0.277,
+      "step": 12350
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.376223629640577e-05,
+      "loss": 0.3446,
+      "step": 12360
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.3751725299755106e-05,
+      "loss": 0.2468,
+      "step": 12370
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.374120671930107e-05,
+      "loss": 0.2176,
+      "step": 12380
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.373068055929769e-05,
+      "loss": 0.4059,
+      "step": 12390
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.372014682400211e-05,
+      "loss": 0.2584,
+      "step": 12400
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.3709605517674485e-05,
+      "loss": 0.3409,
+      "step": 12410
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.369905664457808e-05,
+      "loss": 0.2292,
+      "step": 12420
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.3688500208979196e-05,
+      "loss": 0.1868,
+      "step": 12430
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.36779362151472e-05,
+      "loss": 0.2209,
+      "step": 12440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.3667364667354494e-05,
+      "loss": 0.1841,
+      "step": 12450
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.365678556987658e-05,
+      "loss": 0.2414,
+      "step": 12460
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.3646198926991967e-05,
+      "loss": 0.2001,
+      "step": 12470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.363560474298224e-05,
+      "loss": 0.3616,
+      "step": 12480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.362500302213204e-05,
+      "loss": 0.3376,
+      "step": 12490
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.361439376872902e-05,
+      "loss": 0.2589,
+      "step": 12500
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 0.21298615634441376,
+      "eval_runtime": 418.2735,
+      "eval_samples_per_second": 50.735,
+      "eval_steps_per_second": 0.397,
+      "step": 12500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.360377698706394e-05,
+      "loss": 0.3342,
+      "step": 12510
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.3593152681430535e-05,
+      "loss": 0.2365,
+      "step": 12520
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.358252085612564e-05,
+      "loss": 0.2947,
+      "step": 12530
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.3571881515449096e-05,
+      "loss": 0.2789,
+      "step": 12540
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.356123466370381e-05,
+      "loss": 0.245,
+      "step": 12550
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.3550580305195696e-05,
+      "loss": 0.2026,
+      "step": 12560
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.3539918444233736e-05,
+      "loss": 0.4313,
+      "step": 12570
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.352924908512992e-05,
+      "loss": 0.1862,
+      "step": 12580
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.351857223219929e-05,
+      "loss": 0.2102,
+      "step": 12590
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.350788788975989e-05,
+      "loss": 0.3079,
+      "step": 12600
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.3497196062132845e-05,
+      "loss": 0.2444,
+      "step": 12610
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.348649675364225e-05,
+      "loss": 0.2257,
+      "step": 12620
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.347578996861527e-05,
+      "loss": 0.3998,
+      "step": 12630
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.3465075711382054e-05,
+      "loss": 0.1705,
+      "step": 12640
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.3454353986275806e-05,
+      "loss": 0.2764,
+      "step": 12650
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.344362479763274e-05,
+      "loss": 0.3394,
+      "step": 12660
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.343288814979208e-05,
+      "loss": 0.2938,
+      "step": 12670
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.342214404709608e-05,
+      "loss": 0.3014,
+      "step": 12680
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.341139249388999e-05,
+      "loss": 0.2947,
+      "step": 12690
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.34006334945221e-05,
+      "loss": 0.1608,
+      "step": 12700
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.338986705334369e-05,
+      "loss": 0.1966,
+      "step": 12710
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.337909317470905e-05,
+      "loss": 0.1925,
+      "step": 12720
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.3368311862975496e-05,
+      "loss": 0.2176,
+      "step": 12730
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.335752312250334e-05,
+      "loss": 0.2482,
+      "step": 12740
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.334672695765588e-05,
+      "loss": 0.2974,
+      "step": 12750
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.333592337279945e-05,
+      "loss": 0.2011,
+      "step": 12760
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.3325112372303367e-05,
+      "loss": 0.2358,
+      "step": 12770
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.3314293960539943e-05,
+      "loss": 0.358,
+      "step": 12780
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.33034681418845e-05,
+      "loss": 0.2744,
+      "step": 12790
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.329263492071533e-05,
+      "loss": 0.1627,
+      "step": 12800
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.328179430141377e-05,
+      "loss": 0.3395,
+      "step": 12810
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.327094628836409e-05,
+      "loss": 0.3314,
+      "step": 12820
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.326009088595358e-05,
+      "loss": 0.273,
+      "step": 12830
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.324922809857253e-05,
+      "loss": 0.3055,
+      "step": 12840
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.323835793061418e-05,
+      "loss": 0.1936,
+      "step": 12850
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.32274803864748e-05,
+      "loss": 0.2266,
+      "step": 12860
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.321659547055361e-05,
+      "loss": 0.2255,
+      "step": 12870
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.320570318725282e-05,
+      "loss": 0.2426,
+      "step": 12880
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.3194803540977626e-05,
+      "loss": 0.2298,
+      "step": 12890
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.318389653613618e-05,
+      "loss": 0.3301,
+      "step": 12900
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.3172982177139654e-05,
+      "loss": 0.2165,
+      "step": 12910
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.316206046840215e-05,
+      "loss": 0.2066,
+      "step": 12920
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.315113141434076e-05,
+      "loss": 0.1956,
+      "step": 12930
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.3140195019375564e-05,
+      "loss": 0.2454,
+      "step": 12940
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.3129251287929565e-05,
+      "loss": 0.3691,
+      "step": 12950
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.311830022442879e-05,
+      "loss": 0.1916,
+      "step": 12960
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.3107341833302176e-05,
+      "loss": 0.2429,
+      "step": 12970
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.3096376118981665e-05,
+      "loss": 0.2137,
+      "step": 12980
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.3085403085902145e-05,
+      "loss": 0.3688,
+      "step": 12990
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.3074422738501466e-05,
+      "loss": 0.2839,
+      "step": 13000
+    },
+    {
+      "epoch": 0.52,
+      "eval_loss": 0.2154642939567566,
+      "eval_runtime": 418.9167,
+      "eval_samples_per_second": 50.657,
+      "eval_steps_per_second": 0.396,
+      "step": 13000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.306343508122044e-05,
+      "loss": 0.2105,
+      "step": 13010
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.3052440118502814e-05,
+      "loss": 0.2411,
+      "step": 13020
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.304143785479532e-05,
+      "loss": 0.2659,
+      "step": 13030
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.303042829454763e-05,
+      "loss": 0.2667,
+      "step": 13040
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.3019411442212354e-05,
+      "loss": 0.4302,
+      "step": 13050
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.300838730224507e-05,
+      "loss": 0.2536,
+      "step": 13060
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.29973558791043e-05,
+      "loss": 0.2187,
+      "step": 13070
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.29863171772515e-05,
+      "loss": 0.3478,
+      "step": 13080
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.297527120115108e-05,
+      "loss": 0.3,
+      "step": 13090
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.296421795527039e-05,
+      "loss": 0.2498,
+      "step": 13100
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.295315744407972e-05,
+      "loss": 0.3484,
+      "step": 13110
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.2942089672052305e-05,
+      "loss": 0.313,
+      "step": 13120
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.2931014643664305e-05,
+      "loss": 0.2503,
+      "step": 13130
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.29199323633948e-05,
+      "loss": 0.395,
+      "step": 13140
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.290884283572586e-05,
+      "loss": 0.2637,
+      "step": 13150
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.289774606514241e-05,
+      "loss": 0.2008,
+      "step": 13160
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.288664205613237e-05,
+      "loss": 0.2469,
+      "step": 13170
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.287553081318655e-05,
+      "loss": 0.2709,
+      "step": 13180
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.286441234079869e-05,
+      "loss": 0.2395,
+      "step": 13190
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.2853286643465465e-05,
+      "loss": 0.1838,
+      "step": 13200
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.284215372568647e-05,
+      "loss": 0.2242,
+      "step": 13210
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.283101359196421e-05,
+      "loss": 0.1872,
+      "step": 13220
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.281986624680412e-05,
+      "loss": 0.274,
+      "step": 13230
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.280871169471454e-05,
+      "loss": 0.2898,
+      "step": 13240
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.279754994020674e-05,
+      "loss": 0.197,
+      "step": 13250
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.278638098779488e-05,
+      "loss": 0.2235,
+      "step": 13260
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.2775204841996063e-05,
+      "loss": 0.3346,
+      "step": 13270
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.276402150733027e-05,
+      "loss": 0.4034,
+      "step": 13280
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.275283098832041e-05,
+      "loss": 0.2065,
+      "step": 13290
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.274163328949229e-05,
+      "loss": 0.2313,
+      "step": 13300
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.273042841537461e-05,
+      "loss": 0.2259,
+      "step": 13310
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.2719216370499e-05,
+      "loss": 0.2038,
+      "step": 13320
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.270799715939996e-05,
+      "loss": 0.2416,
+      "step": 13330
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.2696770786614906e-05,
+      "loss": 0.1405,
+      "step": 13340
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.268553725668415e-05,
+      "loss": 0.1845,
+      "step": 13350
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.267429657415088e-05,
+      "loss": 0.2307,
+      "step": 13360
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.266304874356122e-05,
+      "loss": 0.2398,
+      "step": 13370
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.265179376946412e-05,
+      "loss": 0.2761,
+      "step": 13380
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.264053165641147e-05,
+      "loss": 0.2496,
+      "step": 13390
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.262926240895805e-05,
+      "loss": 0.2076,
+      "step": 13400
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.261798603166148e-05,
+      "loss": 0.2246,
+      "step": 13410
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.260670252908232e-05,
+      "loss": 0.1725,
+      "step": 13420
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.259541190578396e-05,
+      "loss": 0.2438,
+      "step": 13430
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.258411416633269e-05,
+      "loss": 0.2151,
+      "step": 13440
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.257280931529769e-05,
+      "loss": 0.2686,
+      "step": 13450
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.2561497357251013e-05,
+      "loss": 0.2475,
+      "step": 13460
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.255017829676757e-05,
+      "loss": 0.2763,
+      "step": 13470
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.2538852138425156e-05,
+      "loss": 0.198,
+      "step": 13480
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.252751888680445e-05,
+      "loss": 0.3391,
+      "step": 13490
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.251617854648896e-05,
+      "loss": 0.2519,
+      "step": 13500
+    },
+    {
+      "epoch": 0.54,
+      "eval_loss": 0.2056160569190979,
+      "eval_runtime": 417.9749,
+      "eval_samples_per_second": 50.771,
+      "eval_steps_per_second": 0.397,
+      "step": 13500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.25048311220651e-05,
+      "loss": 0.3165,
+      "step": 13510
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.249347661812214e-05,
+      "loss": 0.2754,
+      "step": 13520
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.2482115039252203e-05,
+      "loss": 0.3362,
+      "step": 13530
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.2470746390050275e-05,
+      "loss": 0.2698,
+      "step": 13540
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.245937067511421e-05,
+      "loss": 0.1827,
+      "step": 13550
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.24479878990447e-05,
+      "loss": 0.2291,
+      "step": 13560
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.243659806644534e-05,
+      "loss": 0.2624,
+      "step": 13570
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.2425201181922515e-05,
+      "loss": 0.1847,
+      "step": 13580
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.24137972500855e-05,
+      "loss": 0.2794,
+      "step": 13590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.240238627554641e-05,
+      "loss": 0.2526,
+      "step": 13600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.239096826292023e-05,
+      "loss": 0.2407,
+      "step": 13610
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.237954321682476e-05,
+      "loss": 0.2054,
+      "step": 13620
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.236811114188066e-05,
+      "loss": 0.1901,
+      "step": 13630
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.235667204271143e-05,
+      "loss": 0.2364,
+      "step": 13640
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.234522592394341e-05,
+      "loss": 0.3236,
+      "step": 13650
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.233377279020578e-05,
+      "loss": 0.1645,
+      "step": 13660
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.2322312646130555e-05,
+      "loss": 0.3133,
+      "step": 13670
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.23108454963526e-05,
+      "loss": 0.2286,
+      "step": 13680
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.229937134550959e-05,
+      "loss": 0.2134,
+      "step": 13690
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.2287890198242035e-05,
+      "loss": 0.2149,
+      "step": 13700
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.22764020591933e-05,
+      "loss": 0.1988,
+      "step": 13710
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.226490693300954e-05,
+      "loss": 0.3344,
+      "step": 13720
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.225340482433977e-05,
+      "loss": 0.2382,
+      "step": 13730
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.224189573783581e-05,
+      "loss": 0.3492,
+      "step": 13740
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.223037967815231e-05,
+      "loss": 0.2188,
+      "step": 13750
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.2218856649946734e-05,
+      "loss": 0.1817,
+      "step": 13760
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.220732665787936e-05,
+      "loss": 0.2007,
+      "step": 13770
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.219578970661329e-05,
+      "loss": 0.2416,
+      "step": 13780
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.218424580081445e-05,
+      "loss": 0.2488,
+      "step": 13790
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.217269494515156e-05,
+      "loss": 0.3818,
+      "step": 13800
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.216113714429616e-05,
+      "loss": 0.1958,
+      "step": 13810
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.2149572402922606e-05,
+      "loss": 0.2613,
+      "step": 13820
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.213800072570804e-05,
+      "loss": 0.3022,
+      "step": 13830
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.212642211733243e-05,
+      "loss": 0.2999,
+      "step": 13840
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.211483658247855e-05,
+      "loss": 0.2695,
+      "step": 13850
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.2103244125831936e-05,
+      "loss": 0.2643,
+      "step": 13860
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.209164475208097e-05,
+      "loss": 0.171,
+      "step": 13870
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.2080038465916816e-05,
+      "loss": 0.1458,
+      "step": 13880
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.206842527203343e-05,
+      "loss": 0.275,
+      "step": 13890
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.205680517512756e-05,
+      "loss": 0.1795,
+      "step": 13900
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.2045178179898734e-05,
+      "loss": 0.3648,
+      "step": 13910
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.203354429104931e-05,
+      "loss": 0.1719,
+      "step": 13920
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.20219035132844e-05,
+      "loss": 0.2129,
+      "step": 13930
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.20102558513119e-05,
+      "loss": 0.2361,
+      "step": 13940
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.199860130984251e-05,
+      "loss": 0.2654,
+      "step": 13950
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.1986939893589696e-05,
+      "loss": 0.1486,
+      "step": 13960
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.1975271607269726e-05,
+      "loss": 0.3391,
+      "step": 13970
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.1963596455601616e-05,
+      "loss": 0.1369,
+      "step": 13980
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.195191444330718e-05,
+      "loss": 0.1624,
+      "step": 13990
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.1940225575111e-05,
+      "loss": 0.2097,
+      "step": 14000
+    },
+    {
+      "epoch": 0.56,
+      "eval_loss": 0.1943461149930954,
+      "eval_runtime": 418.6112,
+      "eval_samples_per_second": 50.694,
+      "eval_steps_per_second": 0.397,
+      "step": 14000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.192852985574044e-05,
+      "loss": 0.2262,
+      "step": 14010
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.1916827289925614e-05,
+      "loss": 0.2447,
+      "step": 14020
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.1905117882399425e-05,
+      "loss": 0.3826,
+      "step": 14030
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.189340163789754e-05,
+      "loss": 0.2021,
+      "step": 14040
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.1881678561158374e-05,
+      "loss": 0.2034,
+      "step": 14050
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.186994865692313e-05,
+      "loss": 0.2368,
+      "step": 14060
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.185821192993575e-05,
+      "loss": 0.2541,
+      "step": 14070
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.184646838494296e-05,
+      "loss": 0.2463,
+      "step": 14080
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.183471802669421e-05,
+      "loss": 0.2991,
+      "step": 14090
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.182296085994174e-05,
+      "loss": 0.3223,
+      "step": 14100
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.181119688944052e-05,
+      "loss": 0.1822,
+      "step": 14110
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.1799426119948284e-05,
+      "loss": 0.2515,
+      "step": 14120
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.178764855622551e-05,
+      "loss": 0.178,
+      "step": 14130
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.177586420303542e-05,
+      "loss": 0.207,
+      "step": 14140
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.176407306514399e-05,
+      "loss": 0.3104,
+      "step": 14150
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.175227514731994e-05,
+      "loss": 0.3336,
+      "step": 14160
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.1740470454334726e-05,
+      "loss": 0.3108,
+      "step": 14170
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.172865899096255e-05,
+      "loss": 0.2151,
+      "step": 14180
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.171684076198035e-05,
+      "loss": 0.2348,
+      "step": 14190
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.17050157721678e-05,
+      "loss": 0.1405,
+      "step": 14200
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.16931840263073e-05,
+      "loss": 0.2833,
+      "step": 14210
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.168134552918399e-05,
+      "loss": 0.2294,
+      "step": 14220
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.1669500285585755e-05,
+      "loss": 0.2429,
+      "step": 14230
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.165764830030318e-05,
+      "loss": 0.2143,
+      "step": 14240
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.1645789578129596e-05,
+      "loss": 0.2426,
+      "step": 14250
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.163392412386106e-05,
+      "loss": 0.2706,
+      "step": 14260
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.1622051942296326e-05,
+      "loss": 0.2889,
+      "step": 14270
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.161017303823691e-05,
+      "loss": 0.1868,
+      "step": 14280
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.1598287416487004e-05,
+      "loss": 0.2544,
+      "step": 14290
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.1586395081853556e-05,
+      "loss": 0.2527,
+      "step": 14300
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.15744960391462e-05,
+      "loss": 0.2033,
+      "step": 14310
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.1562590293177294e-05,
+      "loss": 0.2584,
+      "step": 14320
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.15506778487619e-05,
+      "loss": 0.2086,
+      "step": 14330
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.1539950925598155e-05,
+      "loss": 0.2377,
+      "step": 14340
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.152802576740966e-05,
+      "loss": 0.2588,
+      "step": 14350
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.151609392475371e-05,
+      "loss": 0.2317,
+      "step": 14360
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.15041554024559e-05,
+      "loss": 0.2742,
+      "step": 14370
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.1492210205344574e-05,
+      "loss": 0.3086,
+      "step": 14380
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.1480258338250744e-05,
+      "loss": 0.2377,
+      "step": 14390
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.1468299806008134e-05,
+      "loss": 0.202,
+      "step": 14400
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.145633461345316e-05,
+      "loss": 0.2564,
+      "step": 14410
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.1444362765424936e-05,
+      "loss": 0.2096,
+      "step": 14420
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.1432384266765264e-05,
+      "loss": 0.2748,
+      "step": 14430
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.142039912231864e-05,
+      "loss": 0.2425,
+      "step": 14440
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.140840733693223e-05,
+      "loss": 0.2281,
+      "step": 14450
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.139640891545591e-05,
+      "loss": 0.1757,
+      "step": 14460
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.138440386274222e-05,
+      "loss": 0.1937,
+      "step": 14470
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.13723921836464e-05,
+      "loss": 0.2556,
+      "step": 14480
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.136037388302636e-05,
+      "loss": 0.2376,
+      "step": 14490
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.134834896574269e-05,
+      "loss": 0.2532,
+      "step": 14500
+    },
+    {
+      "epoch": 0.58,
+      "eval_loss": 0.1903214007616043,
+      "eval_runtime": 418.5676,
+      "eval_samples_per_second": 50.699,
+      "eval_steps_per_second": 0.397,
+      "step": 14500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.1336317436658636e-05,
+      "loss": 0.1726,
+      "step": 14510
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.1324279300640165e-05,
+      "loss": 0.2322,
+      "step": 14520
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.1312234562555875e-05,
+      "loss": 0.2021,
+      "step": 14530
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.130018322727705e-05,
+      "loss": 0.2606,
+      "step": 14540
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.128812529967763e-05,
+      "loss": 0.3163,
+      "step": 14550
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.127606078463424e-05,
+      "loss": 0.1976,
+      "step": 14560
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.126398968702616e-05,
+      "loss": 0.3172,
+      "step": 14570
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.1251912011735326e-05,
+      "loss": 0.2341,
+      "step": 14580
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.123982776364635e-05,
+      "loss": 0.2797,
+      "step": 14590
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.122773694764648e-05,
+      "loss": 0.1968,
+      "step": 14600
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.1215639568625644e-05,
+      "loss": 0.2469,
+      "step": 14610
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.12035356314764e-05,
+      "loss": 0.2495,
+      "step": 14620
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.119142514109397e-05,
+      "loss": 0.1653,
+      "step": 14630
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.117930810237625e-05,
+      "loss": 0.2299,
+      "step": 14640
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.1167184520223734e-05,
+      "loss": 0.3069,
+      "step": 14650
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.115505439953961e-05,
+      "loss": 0.2387,
+      "step": 14660
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.1142917745229686e-05,
+      "loss": 0.169,
+      "step": 14670
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.113077456220241e-05,
+      "loss": 0.1681,
+      "step": 14680
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.111862485536888e-05,
+      "loss": 0.1703,
+      "step": 14690
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.110646862964284e-05,
+      "loss": 0.2813,
+      "step": 14700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.1094305889940646e-05,
+      "loss": 0.149,
+      "step": 14710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.108213664118131e-05,
+      "loss": 0.2005,
+      "step": 14720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.1069960888286476e-05,
+      "loss": 0.3199,
+      "step": 14730
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.10577786361804e-05,
+      "loss": 0.2457,
+      "step": 14740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.104558988978997e-05,
+      "loss": 0.1448,
+      "step": 14750
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.103339465404474e-05,
+      "loss": 0.2459,
+      "step": 14760
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.102119293387683e-05,
+      "loss": 0.1989,
+      "step": 14770
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.100898473422101e-05,
+      "loss": 0.3401,
+      "step": 14780
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.099677006001469e-05,
+      "loss": 0.2778,
+      "step": 14790
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.098454891619787e-05,
+      "loss": 0.232,
+      "step": 14800
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.097232130771317e-05,
+      "loss": 0.185,
+      "step": 14810
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.096008723950582e-05,
+      "loss": 0.1612,
+      "step": 14820
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.09478467165237e-05,
+      "loss": 0.2458,
+      "step": 14830
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.093559974371725e-05,
+      "loss": 0.2182,
+      "step": 14840
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.0923346326039543e-05,
+      "loss": 0.2154,
+      "step": 14850
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.091108646844627e-05,
+      "loss": 0.2493,
+      "step": 14860
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.0898820175895706e-05,
+      "loss": 0.1843,
+      "step": 14870
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.0886547453348734e-05,
+      "loss": 0.217,
+      "step": 14880
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.0874268305768847e-05,
+      "loss": 0.1674,
+      "step": 14890
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.0861982738122115e-05,
+      "loss": 0.226,
+      "step": 14900
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.0849690755377235e-05,
+      "loss": 0.2691,
+      "step": 14910
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.0837392362505464e-05,
+      "loss": 0.2215,
+      "step": 14920
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.0825087564480694e-05,
+      "loss": 0.2606,
+      "step": 14930
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.081277636627936e-05,
+      "loss": 0.2866,
+      "step": 14940
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.0800458772880514e-05,
+      "loss": 0.1769,
+      "step": 14950
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.0788134789265784e-05,
+      "loss": 0.2749,
+      "step": 14960
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.0775804420419404e-05,
+      "loss": 0.1725,
+      "step": 14970
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.076346767132816e-05,
+      "loss": 0.1566,
+      "step": 14980
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.0751124546981435e-05,
+      "loss": 0.2697,
+      "step": 14990
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.073877505237118e-05,
+      "loss": 0.1996,
+      "step": 15000
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 0.1962420493364334,
+      "eval_runtime": 418.0883,
+      "eval_samples_per_second": 50.757,
+      "eval_steps_per_second": 0.397,
+      "step": 15000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.0726419192491946e-05,
+      "loss": 0.2014,
+      "step": 15010
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.071405697234082e-05,
+      "loss": 0.2558,
+      "step": 15020
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.070168839691751e-05,
+      "loss": 0.1687,
+      "step": 15030
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.068931347122424e-05,
+      "loss": 0.1823,
+      "step": 15040
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.0676932200265854e-05,
+      "loss": 0.2021,
+      "step": 15050
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.066454458904973e-05,
+      "loss": 0.2005,
+      "step": 15060
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.0652150642585804e-05,
+      "loss": 0.2901,
+      "step": 15070
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.0639750365886595e-05,
+      "loss": 0.216,
+      "step": 15080
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.0627343763967185e-05,
+      "loss": 0.1637,
+      "step": 15090
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.06149308418452e-05,
+      "loss": 0.1915,
+      "step": 15100
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.0602511604540824e-05,
+      "loss": 0.2158,
+      "step": 15110
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.05900860570768e-05,
+      "loss": 0.3233,
+      "step": 15120
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.057765420447842e-05,
+      "loss": 0.1529,
+      "step": 15130
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.0565216051773526e-05,
+      "loss": 0.2633,
+      "step": 15140
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.05527716039925e-05,
+      "loss": 0.2637,
+      "step": 15150
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.05403208661683e-05,
+      "loss": 0.2649,
+      "step": 15160
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.052786384333639e-05,
+      "loss": 0.2078,
+      "step": 15170
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.0515400540534795e-05,
+      "loss": 0.2082,
+      "step": 15180
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.050293096280408e-05,
+      "loss": 0.186,
+      "step": 15190
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.0490455115187345e-05,
+      "loss": 0.1616,
+      "step": 15200
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.047797300273023e-05,
+      "loss": 0.3062,
+      "step": 15210
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.046548463048089e-05,
+      "loss": 0.1998,
+      "step": 15220
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.0452990003490047e-05,
+      "loss": 0.2465,
+      "step": 15230
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.0440489126810905e-05,
+      "loss": 0.2273,
+      "step": 15240
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.042798200549924e-05,
+      "loss": 0.2074,
+      "step": 15250
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.041546864461334e-05,
+      "loss": 0.1532,
+      "step": 15260
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.0402949049214e-05,
+      "loss": 0.2018,
+      "step": 15270
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.039042322436456e-05,
+      "loss": 0.1822,
+      "step": 15280
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.0377891175130866e-05,
+      "loss": 0.239,
+      "step": 15290
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.036535290658127e-05,
+      "loss": 0.2357,
+      "step": 15300
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.0352808423786656e-05,
+      "loss": 0.282,
+      "step": 15310
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.034025773182044e-05,
+      "loss": 0.1637,
+      "step": 15320
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.032770083575851e-05,
+      "loss": 0.1815,
+      "step": 15330
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.031513774067928e-05,
+      "loss": 0.2052,
+      "step": 15340
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.030256845166367e-05,
+      "loss": 0.2037,
+      "step": 15350
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.028999297379511e-05,
+      "loss": 0.3643,
+      "step": 15360
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.027741131215953e-05,
+      "loss": 0.1886,
+      "step": 15370
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.026482347184536e-05,
+      "loss": 0.198,
+      "step": 15380
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.025222945794354e-05,
+      "loss": 0.2331,
+      "step": 15390
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.023962927554747e-05,
+      "loss": 0.2216,
+      "step": 15400
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.022702292975309e-05,
+      "loss": 0.2021,
+      "step": 15410
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.021441042565881e-05,
+      "loss": 0.1685,
+      "step": 15420
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.0201791768365535e-05,
+      "loss": 0.1956,
+      "step": 15430
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.018916696297665e-05,
+      "loss": 0.2003,
+      "step": 15440
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.0176536014598046e-05,
+      "loss": 0.113,
+      "step": 15450
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.0163898928338086e-05,
+      "loss": 0.22,
+      "step": 15460
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.01512557093076e-05,
+      "loss": 0.2027,
+      "step": 15470
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.0138606362619926e-05,
+      "loss": 0.1874,
+      "step": 15480
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.0125950893390876e-05,
+      "loss": 0.1865,
+      "step": 15490
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.011328930673871e-05,
+      "loss": 0.1916,
+      "step": 15500
+    },
+    {
+      "epoch": 0.62,
+      "eval_loss": 0.17877347767353058,
+      "eval_runtime": 418.0328,
+      "eval_samples_per_second": 50.764,
+      "eval_steps_per_second": 0.397,
+      "step": 15500
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.010062160778419e-05,
+      "loss": 0.1748,
+      "step": 15510
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.008794780165056e-05,
+      "loss": 0.0862,
+      "step": 15520
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.0075267893463494e-05,
+      "loss": 0.1667,
+      "step": 15530
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.006258188835117e-05,
+      "loss": 0.2448,
+      "step": 15540
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.0049889791444215e-05,
+      "loss": 0.2721,
+      "step": 15550
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.003719160787572e-05,
+      "loss": 0.1923,
+      "step": 15560
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.0024487342781244e-05,
+      "loss": 0.1552,
+      "step": 15570
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.00117770012988e-05,
+      "loss": 0.2733,
+      "step": 15580
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.999906058856886e-05,
+      "loss": 0.2428,
+      "step": 15590
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.998633810973435e-05,
+      "loss": 0.174,
+      "step": 15600
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.9973609569940665e-05,
+      "loss": 0.1926,
+      "step": 15610
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.996087497433562e-05,
+      "loss": 0.1799,
+      "step": 15620
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.99481343280695e-05,
+      "loss": 0.3514,
+      "step": 15630
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.993538763629504e-05,
+      "loss": 0.3367,
+      "step": 15640
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.992263490416741e-05,
+      "loss": 0.173,
+      "step": 15650
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.9909876136844234e-05,
+      "loss": 0.1306,
+      "step": 15660
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.989711133948556e-05,
+      "loss": 0.1925,
+      "step": 15670
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.988434051725389e-05,
+      "loss": 0.3303,
+      "step": 15680
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.987156367531415e-05,
+      "loss": 0.2463,
+      "step": 15690
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.985878081883373e-05,
+      "loss": 0.4195,
+      "step": 15700
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.9845991952982396e-05,
+      "loss": 0.363,
+      "step": 15710
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.98331970829324e-05,
+      "loss": 0.2345,
+      "step": 15720
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.9820396213858396e-05,
+      "loss": 0.2152,
+      "step": 15730
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.980758935093747e-05,
+      "loss": 0.2138,
+      "step": 15740
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.9794776499349126e-05,
+      "loss": 0.2165,
+      "step": 15750
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.9781957664275305e-05,
+      "loss": 0.2083,
+      "step": 15760
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.976913285090035e-05,
+      "loss": 0.1945,
+      "step": 15770
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.975630206441103e-05,
+      "loss": 0.3239,
+      "step": 15780
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.974346530999653e-05,
+      "loss": 0.2629,
+      "step": 15790
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.973062259284845e-05,
+      "loss": 0.1978,
+      "step": 15800
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.971777391816081e-05,
+      "loss": 0.2133,
+      "step": 15810
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.970491929113001e-05,
+      "loss": 0.221,
+      "step": 15820
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.969205871695488e-05,
+      "loss": 0.1931,
+      "step": 15830
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.967919220083667e-05,
+      "loss": 0.2523,
+      "step": 15840
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.9666319747978985e-05,
+      "loss": 0.1967,
+      "step": 15850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.9653441363587884e-05,
+      "loss": 0.2043,
+      "step": 15860
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.9640557052871793e-05,
+      "loss": 0.2141,
+      "step": 15870
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.9627666821041545e-05,
+      "loss": 0.2076,
+      "step": 15880
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.961477067331036e-05,
+      "loss": 0.2236,
+      "step": 15890
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.9601868614893875e-05,
+      "loss": 0.2102,
+      "step": 15900
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.958896065101007e-05,
+      "loss": 0.1511,
+      "step": 15910
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.957604678687937e-05,
+      "loss": 0.1777,
+      "step": 15920
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.956312702772454e-05,
+      "loss": 0.2397,
+      "step": 15930
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.955020137877077e-05,
+      "loss": 0.1782,
+      "step": 15940
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.9537269845245586e-05,
+      "loss": 0.2032,
+      "step": 15950
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.952433243237893e-05,
+      "loss": 0.1708,
+      "step": 15960
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.95113891454031e-05,
+      "loss": 0.2845,
+      "step": 15970
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.9498439989552794e-05,
+      "loss": 0.279,
+      "step": 15980
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.9485484970065055e-05,
+      "loss": 0.3012,
+      "step": 15990
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.947252409217933e-05,
+      "loss": 0.143,
+      "step": 16000
+    },
+    {
+      "epoch": 0.63,
+      "eval_loss": 0.1801401823759079,
+      "eval_runtime": 417.8974,
+      "eval_samples_per_second": 50.78,
+      "eval_steps_per_second": 0.397,
+      "step": 16000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.945955736113739e-05,
+      "loss": 0.3489,
+      "step": 16010
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.944658478218342e-05,
+      "loss": 0.1844,
+      "step": 16020
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.9433606360563946e-05,
+      "loss": 0.1442,
+      "step": 16030
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.942062210152786e-05,
+      "loss": 0.1472,
+      "step": 16040
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.940763201032641e-05,
+      "loss": 0.2269,
+      "step": 16050
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.939463609221321e-05,
+      "loss": 0.1673,
+      "step": 16060
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.938163435244424e-05,
+      "loss": 0.1835,
+      "step": 16070
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.936862679627781e-05,
+      "loss": 0.1993,
+      "step": 16080
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.93556134289746e-05,
+      "loss": 0.1533,
+      "step": 16090
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.934259425579764e-05,
+      "loss": 0.2945,
+      "step": 16100
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.932956928201231e-05,
+      "loss": 0.1682,
+      "step": 16110
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.9316538512886316e-05,
+      "loss": 0.2424,
+      "step": 16120
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.930350195368973e-05,
+      "loss": 0.2045,
+      "step": 16130
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.929045960969494e-05,
+      "loss": 0.1497,
+      "step": 16140
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.9277411486176726e-05,
+      "loss": 0.1769,
+      "step": 16150
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.9264357588412144e-05,
+      "loss": 0.1715,
+      "step": 16160
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.925129792168062e-05,
+      "loss": 0.2104,
+      "step": 16170
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.9238232491263904e-05,
+      "loss": 0.1531,
+      "step": 16180
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.9225161302446075e-05,
+      "loss": 0.1504,
+      "step": 16190
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.9212084360513546e-05,
+      "loss": 0.2598,
+      "step": 16200
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.919900167075506e-05,
+      "loss": 0.1634,
+      "step": 16210
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.918591323846167e-05,
+      "loss": 0.2585,
+      "step": 16220
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.9172819068926775e-05,
+      "loss": 0.2227,
+      "step": 16230
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.915971916744606e-05,
+      "loss": 0.2505,
+      "step": 16240
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.914661353931757e-05,
+      "loss": 0.1472,
+      "step": 16250
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.913350218984163e-05,
+      "loss": 0.1481,
+      "step": 16260
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.9120385124320894e-05,
+      "loss": 0.2447,
+      "step": 16270
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.910726234806034e-05,
+      "loss": 0.1757,
+      "step": 16280
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.909413386636722e-05,
+      "loss": 0.2973,
+      "step": 16290
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.9080999684551135e-05,
+      "loss": 0.2309,
+      "step": 16300
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.906785980792396e-05,
+      "loss": 0.1751,
+      "step": 16310
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.9054714241799896e-05,
+      "loss": 0.1684,
+      "step": 16320
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.904156299149543e-05,
+      "loss": 0.3728,
+      "step": 16330
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.902840606232935e-05,
+      "loss": 0.2806,
+      "step": 16340
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.901524345962275e-05,
+      "loss": 0.2635,
+      "step": 16350
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.900207518869901e-05,
+      "loss": 0.3082,
+      "step": 16360
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.898890125488378e-05,
+      "loss": 0.1963,
+      "step": 16370
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.897572166350506e-05,
+      "loss": 0.2389,
+      "step": 16380
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.8962536419893094e-05,
+      "loss": 0.1989,
+      "step": 16390
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.894934552938041e-05,
+      "loss": 0.2131,
+      "step": 16400
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.893614899730183e-05,
+      "loss": 0.1599,
+      "step": 16410
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.8922946828994465e-05,
+      "loss": 0.1773,
+      "step": 16420
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.8909739029797695e-05,
+      "loss": 0.2714,
+      "step": 16430
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.889652560505318e-05,
+      "loss": 0.2115,
+      "step": 16440
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.888330656010486e-05,
+      "loss": 0.1637,
+      "step": 16450
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.887008190029895e-05,
+      "loss": 0.238,
+      "step": 16460
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.88568516309839e-05,
+      "loss": 0.2658,
+      "step": 16470
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.884361575751049e-05,
+      "loss": 0.1886,
+      "step": 16480
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.883037428523173e-05,
+      "loss": 0.3092,
+      "step": 16490
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.881712721950288e-05,
+      "loss": 0.1789,
+      "step": 16500
+    },
+    {
+      "epoch": 0.65,
+      "eval_loss": 0.1752716451883316,
+      "eval_runtime": 418.3011,
+      "eval_samples_per_second": 50.731,
+      "eval_steps_per_second": 0.397,
+      "step": 16500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.88038745656815e-05,
+      "loss": 0.1959,
+      "step": 16510
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.8790616329127375e-05,
+      "loss": 0.1533,
+      "step": 16520
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.877735251520258e-05,
+      "loss": 0.2211,
+      "step": 16530
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.8764083129271435e-05,
+      "loss": 0.3124,
+      "step": 16540
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.875080817670048e-05,
+      "loss": 0.1561,
+      "step": 16550
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.873752766285856e-05,
+      "loss": 0.1935,
+      "step": 16560
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.872424159311674e-05,
+      "loss": 0.3091,
+      "step": 16570
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.8710949972848324e-05,
+      "loss": 0.227,
+      "step": 16580
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.869765280742889e-05,
+      "loss": 0.2095,
+      "step": 16590
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.8684350102236214e-05,
+      "loss": 0.2354,
+      "step": 16600
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.867104186265037e-05,
+      "loss": 0.1885,
+      "step": 16610
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.865772809405362e-05,
+      "loss": 0.1956,
+      "step": 16620
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.864440880183049e-05,
+      "loss": 0.2298,
+      "step": 16630
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.863108399136773e-05,
+      "loss": 0.2026,
+      "step": 16640
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.861775366805432e-05,
+      "loss": 0.2263,
+      "step": 16650
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.860441783728149e-05,
+      "loss": 0.2215,
+      "step": 16660
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.859107650444266e-05,
+      "loss": 0.1512,
+      "step": 16670
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.8577729674933506e-05,
+      "loss": 0.192,
+      "step": 16680
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.856437735415191e-05,
+      "loss": 0.354,
+      "step": 16690
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.855101954749799e-05,
+      "loss": 0.1936,
+      "step": 16700
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.853765626037408e-05,
+      "loss": 0.189,
+      "step": 16710
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.85242874981847e-05,
+      "loss": 0.2457,
+      "step": 16720
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.8510913266336635e-05,
+      "loss": 0.2103,
+      "step": 16730
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.849753357023885e-05,
+      "loss": 0.1582,
+      "step": 16740
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.8484148415302515e-05,
+      "loss": 0.192,
+      "step": 16750
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.8470757806941035e-05,
+      "loss": 0.2872,
+      "step": 16760
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.845736175056999e-05,
+      "loss": 0.1835,
+      "step": 16770
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.844396025160719e-05,
+      "loss": 0.1355,
+      "step": 16780
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.8430553315472626e-05,
+      "loss": 0.2624,
+      "step": 16790
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.84171409475885e-05,
+      "loss": 0.283,
+      "step": 16800
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.840372315337921e-05,
+      "loss": 0.2807,
+      "step": 16810
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.839029993827133e-05,
+      "loss": 0.1936,
+      "step": 16820
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.8376871307693665e-05,
+      "loss": 0.1698,
+      "step": 16830
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.836343726707716e-05,
+      "loss": 0.2596,
+      "step": 16840
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.8349997821855e-05,
+      "loss": 0.275,
+      "step": 16850
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.833655297746251e-05,
+      "loss": 0.2106,
+      "step": 16860
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.8323102739337246e-05,
+      "loss": 0.2141,
+      "step": 16870
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.830964711291889e-05,
+      "loss": 0.2442,
+      "step": 16880
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.8296186103649344e-05,
+      "loss": 0.3838,
+      "step": 16890
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.828271971697268e-05,
+      "loss": 0.1915,
+      "step": 16900
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.826924795833513e-05,
+      "loss": 0.2168,
+      "step": 16910
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.825577083318512e-05,
+      "loss": 0.1755,
+      "step": 16920
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.824228834697323e-05,
+      "loss": 0.1907,
+      "step": 16930
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.8228800505152204e-05,
+      "loss": 0.1758,
+      "step": 16940
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.8215307313176974e-05,
+      "loss": 0.2474,
+      "step": 16950
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.820180877650461e-05,
+      "loss": 0.2514,
+      "step": 16960
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.818830490059436e-05,
+      "loss": 0.1943,
+      "step": 16970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.817479569090763e-05,
+      "loss": 0.3307,
+      "step": 16980
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.8161281152907985e-05,
+      "loss": 0.2499,
+      "step": 16990
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.8147761292061135e-05,
+      "loss": 0.2105,
+      "step": 17000
+    },
+    {
+      "epoch": 0.67,
+      "eval_loss": 0.17629510164260864,
+      "eval_runtime": 418.2527,
+      "eval_samples_per_second": 50.737,
+      "eval_steps_per_second": 0.397,
+      "step": 17000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 50400,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "total_flos": 0.0,
+  "trial_name": null,
+  "trial_params": null
+}