diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12239 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.999942624361696,
+  "eval_steps": 500,
+  "global_step": 8714,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.2935779816513764e-07,
+      "loss": 1.0605,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.1467889908256882e-06,
+      "loss": 1.151,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.2935779816513764e-06,
+      "loss": 1.1878,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.369140625,
+      "learning_rate": 3.4403669724770644e-06,
+      "loss": 1.1512,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.318359375,
+      "learning_rate": 4.587155963302753e-06,
+      "loss": 1.1618,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.30078125,
+      "learning_rate": 5.733944954128441e-06,
+      "loss": 1.1177,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.267578125,
+      "learning_rate": 6.880733944954129e-06,
+      "loss": 1.0947,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.255859375,
+      "learning_rate": 8.027522935779817e-06,
+      "loss": 1.1256,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.25,
+      "learning_rate": 9.174311926605506e-06,
+      "loss": 1.1013,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.0321100917431194e-05,
+      "loss": 1.1046,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.1467889908256882e-05,
+      "loss": 1.1076,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.2614678899082569e-05,
+      "loss": 1.0508,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.3761467889908258e-05,
+      "loss": 1.0812,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.4908256880733946e-05,
+      "loss": 1.0587,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.240234375,
+      "learning_rate": 1.6055045871559634e-05,
+      "loss": 1.0215,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.7201834862385323e-05,
+      "loss": 1.0493,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.834862385321101e-05,
+      "loss": 1.0221,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.94954128440367e-05,
+      "loss": 1.069,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.0642201834862388e-05,
+      "loss": 1.0516,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.1788990825688073e-05,
+      "loss": 1.088,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.2935779816513765e-05,
+      "loss": 1.0212,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.408256880733945e-05,
+      "loss": 1.0614,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.5229357798165138e-05,
+      "loss": 0.9889,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.6376146788990823e-05,
+      "loss": 1.0308,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.7522935779816515e-05,
+      "loss": 1.0306,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.8669724770642203e-05,
+      "loss": 1.0098,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.9816513761467892e-05,
+      "loss": 1.0313,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.197265625,
+      "learning_rate": 3.096330275229358e-05,
+      "loss": 0.9705,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.21484375,
+      "learning_rate": 3.211009174311927e-05,
+      "loss": 1.0609,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.224609375,
+      "learning_rate": 3.325688073394496e-05,
+      "loss": 1.0414,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 3.4403669724770645e-05,
+      "loss": 1.0439,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 3.555045871559633e-05,
+      "loss": 0.9848,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.224609375,
+      "learning_rate": 3.669724770642202e-05,
+      "loss": 0.9856,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 3.784403669724771e-05,
+      "loss": 0.96,
+      "step": 165
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.212890625,
+      "learning_rate": 3.89908256880734e-05,
+      "loss": 1.0011,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 4.013761467889908e-05,
+      "loss": 0.9926,
+      "step": 175
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.23046875,
+      "learning_rate": 4.1284403669724776e-05,
+      "loss": 1.0263,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 4.2431192660550464e-05,
+      "loss": 1.0145,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.21875,
+      "learning_rate": 4.3577981651376146e-05,
+      "loss": 1.0163,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.27734375,
+      "learning_rate": 4.4724770642201834e-05,
+      "loss": 1.0453,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 4.587155963302753e-05,
+      "loss": 1.0237,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 4.701834862385321e-05,
+      "loss": 0.9497,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.20703125,
+      "learning_rate": 4.81651376146789e-05,
+      "loss": 0.9389,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.21875,
+      "learning_rate": 4.931192660550459e-05,
+      "loss": 1.0078,
+      "step": 215
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.203125,
+      "learning_rate": 5.0458715596330276e-05,
+      "loss": 0.9668,
+      "step": 220
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 5.1605504587155965e-05,
+      "loss": 1.0083,
+      "step": 225
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 5.2752293577981646e-05,
+      "loss": 0.9589,
+      "step": 230
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 5.389908256880735e-05,
+      "loss": 0.9918,
+      "step": 235
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2109375,
+      "learning_rate": 5.504587155963303e-05,
+      "loss": 0.9467,
+      "step": 240
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 5.619266055045872e-05,
+      "loss": 1.0294,
+      "step": 245
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 5.733944954128441e-05,
+      "loss": 0.9983,
+      "step": 250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2265625,
+      "learning_rate": 5.8486238532110095e-05,
+      "loss": 0.9869,
+      "step": 255
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.220703125,
+      "learning_rate": 5.9633027522935784e-05,
+      "loss": 0.9782,
+      "step": 260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 6.0779816513761465e-05,
+      "loss": 1.0076,
+      "step": 265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.20703125,
+      "learning_rate": 6.192660550458716e-05,
+      "loss": 0.9639,
+      "step": 270
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 6.307339449541286e-05,
+      "loss": 1.0112,
+      "step": 275
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 6.422018348623854e-05,
+      "loss": 0.9492,
+      "step": 280
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.22265625,
+      "learning_rate": 6.536697247706422e-05,
+      "loss": 0.9467,
+      "step": 285
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.212890625,
+      "learning_rate": 6.651376146788991e-05,
+      "loss": 1.0238,
+      "step": 290
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 6.76605504587156e-05,
+      "loss": 1.0118,
+      "step": 295
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 6.880733944954129e-05,
+      "loss": 0.9937,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 6.995412844036697e-05,
+      "loss": 1.0102,
+      "step": 305
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 7.110091743119265e-05,
+      "loss": 0.951,
+      "step": 310
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 7.224770642201836e-05,
+      "loss": 0.9713,
+      "step": 315
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2109375,
+      "learning_rate": 7.339449541284404e-05,
+      "loss": 0.9837,
+      "step": 320
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 7.454128440366973e-05,
+      "loss": 0.9261,
+      "step": 325
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.19921875,
+      "learning_rate": 7.568807339449542e-05,
+      "loss": 0.9606,
+      "step": 330
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.220703125,
+      "learning_rate": 7.68348623853211e-05,
+      "loss": 0.9607,
+      "step": 335
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.20703125,
+      "learning_rate": 7.79816513761468e-05,
+      "loss": 0.9969,
+      "step": 340
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.205078125,
+      "learning_rate": 7.912844036697248e-05,
+      "loss": 1.0295,
+      "step": 345
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.220703125,
+      "learning_rate": 8.027522935779816e-05,
+      "loss": 0.985,
+      "step": 350
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 8.142201834862386e-05,
+      "loss": 1.0003,
+      "step": 355
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 8.256880733944955e-05,
+      "loss": 0.9673,
+      "step": 360
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.19921875,
+      "learning_rate": 8.371559633027523e-05,
+      "loss": 0.9255,
+      "step": 365
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.19921875,
+      "learning_rate": 8.486238532110093e-05,
+      "loss": 0.9656,
+      "step": 370
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.20703125,
+      "learning_rate": 8.600917431192661e-05,
+      "loss": 0.9941,
+      "step": 375
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 8.715596330275229e-05,
+      "loss": 0.9551,
+      "step": 380
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.234375,
+      "learning_rate": 8.830275229357799e-05,
+      "loss": 1.004,
+      "step": 385
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.208984375,
+      "learning_rate": 8.944954128440367e-05,
+      "loss": 1.032,
+      "step": 390
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 9.059633027522936e-05,
+      "loss": 0.9599,
+      "step": 395
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 9.174311926605506e-05,
+      "loss": 0.9825,
+      "step": 400
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 9.288990825688074e-05,
+      "loss": 0.9986,
+      "step": 405
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.203125,
+      "learning_rate": 9.403669724770642e-05,
+      "loss": 0.9708,
+      "step": 410
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.518348623853212e-05,
+      "loss": 1.0041,
+      "step": 415
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.63302752293578e-05,
+      "loss": 0.9834,
+      "step": 420
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 9.74770642201835e-05,
+      "loss": 1.0396,
+      "step": 425
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.197265625,
+      "learning_rate": 9.862385321100918e-05,
+      "loss": 1.0168,
+      "step": 430
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.977064220183486e-05,
+      "loss": 0.9267,
+      "step": 435
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00010091743119266055,
+      "loss": 0.9569,
+      "step": 440
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00010206422018348625,
+      "loss": 0.9834,
+      "step": 445
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00010321100917431193,
+      "loss": 0.956,
+      "step": 450
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00010435779816513762,
+      "loss": 0.9971,
+      "step": 455
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00010550458715596329,
+      "loss": 0.9447,
+      "step": 460
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.000106651376146789,
+      "loss": 0.9298,
+      "step": 465
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001077981651376147,
+      "loss": 1.0319,
+      "step": 470
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00010894495412844036,
+      "loss": 0.9869,
+      "step": 475
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00011009174311926606,
+      "loss": 0.9707,
+      "step": 480
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00011123853211009174,
+      "loss": 1.0084,
+      "step": 485
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00011238532110091744,
+      "loss": 1.0212,
+      "step": 490
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00011353211009174312,
+      "loss": 0.9996,
+      "step": 495
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00011467889908256881,
+      "loss": 1.0164,
+      "step": 500
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00011582568807339451,
+      "loss": 0.9616,
+      "step": 505
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00011697247706422019,
+      "loss": 1.0129,
+      "step": 510
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00011811926605504589,
+      "loss": 0.9769,
+      "step": 515
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00011926605504587157,
+      "loss": 0.9869,
+      "step": 520
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00012041284403669726,
+      "loss": 0.9503,
+      "step": 525
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00012155963302752293,
+      "loss": 0.9757,
+      "step": 530
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00012270642201834864,
+      "loss": 0.9539,
+      "step": 535
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00012385321100917432,
+      "loss": 0.9872,
+      "step": 540
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.000125,
+      "loss": 0.9614,
+      "step": 545
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001261467889908257,
+      "loss": 1.0016,
+      "step": 550
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001272935779816514,
+      "loss": 0.9464,
+      "step": 555
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00012844036697247707,
+      "loss": 0.9367,
+      "step": 560
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00012958715596330276,
+      "loss": 1.0023,
+      "step": 565
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00013073394495412844,
+      "loss": 1.0137,
+      "step": 570
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00013188073394495412,
+      "loss": 0.9737,
+      "step": 575
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00013302752293577983,
+      "loss": 0.9705,
+      "step": 580
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001341743119266055,
+      "loss": 0.9556,
+      "step": 585
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001353211009174312,
+      "loss": 0.952,
+      "step": 590
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001364678899082569,
+      "loss": 0.9486,
+      "step": 595
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00013761467889908258,
+      "loss": 0.971,
+      "step": 600
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00013876146788990826,
+      "loss": 1.0046,
+      "step": 605
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00013990825688073395,
+      "loss": 0.9297,
+      "step": 610
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00014105504587155965,
+      "loss": 1.0238,
+      "step": 615
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001422018348623853,
+      "loss": 0.9935,
+      "step": 620
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00014334862385321102,
+      "loss": 0.9899,
+      "step": 625
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00014449541284403673,
+      "loss": 0.9657,
+      "step": 630
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00014564220183486238,
+      "loss": 1.0127,
+      "step": 635
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001467889908256881,
+      "loss": 0.9565,
+      "step": 640
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00014793577981651377,
+      "loss": 1.0164,
+      "step": 645
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00014908256880733945,
+      "loss": 0.9938,
+      "step": 650
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00015022935779816513,
+      "loss": 0.9961,
+      "step": 655
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00015137614678899084,
+      "loss": 0.9774,
+      "step": 660
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00015252293577981652,
+      "loss": 1.0042,
+      "step": 665
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001536697247706422,
+      "loss": 0.9304,
+      "step": 670
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00015481651376146791,
+      "loss": 0.9921,
+      "step": 675
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001559633027522936,
+      "loss": 0.9679,
+      "step": 680
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00015711009174311928,
+      "loss": 0.9292,
+      "step": 685
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00015825688073394496,
+      "loss": 0.9849,
+      "step": 690
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00015940366972477064,
+      "loss": 1.012,
+      "step": 695
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00016055045871559632,
+      "loss": 0.9902,
+      "step": 700
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00016169724770642203,
+      "loss": 0.9778,
+      "step": 705
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001628440366972477,
+      "loss": 0.9252,
+      "step": 710
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001639908256880734,
+      "loss": 0.9662,
+      "step": 715
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001651376146788991,
+      "loss": 0.941,
+      "step": 720
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00016628440366972479,
+      "loss": 1.0185,
+      "step": 725
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00016743119266055047,
+      "loss": 0.9532,
+      "step": 730
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00016857798165137615,
+      "loss": 0.975,
+      "step": 735
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00016972477064220186,
+      "loss": 0.9521,
+      "step": 740
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001708715596330275,
+      "loss": 0.9932,
+      "step": 745
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00017201834862385322,
+      "loss": 0.9432,
+      "step": 750
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00017316513761467893,
+      "loss": 0.9458,
+      "step": 755
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00017431192660550458,
+      "loss": 0.9715,
+      "step": 760
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001754587155963303,
+      "loss": 1.012,
+      "step": 765
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017660550458715597,
+      "loss": 0.9465,
+      "step": 770
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00017775229357798166,
+      "loss": 0.9804,
+      "step": 775
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00017889908256880734,
+      "loss": 0.9905,
+      "step": 780
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00018004587155963305,
+      "loss": 0.9361,
+      "step": 785
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00018119266055045873,
+      "loss": 0.9817,
+      "step": 790
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001823394495412844,
+      "loss": 0.9195,
+      "step": 795
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00018348623853211012,
+      "loss": 0.9655,
+      "step": 800
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001846330275229358,
+      "loss": 0.9728,
+      "step": 805
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00018577981651376148,
+      "loss": 0.9195,
+      "step": 810
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00018692660550458716,
+      "loss": 0.9886,
+      "step": 815
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00018807339449541284,
+      "loss": 0.9954,
+      "step": 820
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00018922018348623853,
+      "loss": 1.0054,
+      "step": 825
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019036697247706424,
+      "loss": 0.9706,
+      "step": 830
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019151376146788992,
+      "loss": 0.9439,
+      "step": 835
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001926605504587156,
+      "loss": 1.0154,
+      "step": 840
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001938073394495413,
+      "loss": 1.0052,
+      "step": 845
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.000194954128440367,
+      "loss": 0.9093,
+      "step": 850
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00019610091743119267,
+      "loss": 1.0025,
+      "step": 855
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00019724770642201835,
+      "loss": 0.9491,
+      "step": 860
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019839449541284406,
+      "loss": 0.9754,
+      "step": 865
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019954128440366972,
+      "loss": 0.9863,
+      "step": 870
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019999992777982591,
+      "loss": 0.9368,
+      "step": 875
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00019999948643469536,
+      "loss": 0.9568,
+      "step": 880
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019999864386852187,
+      "loss": 0.969,
+      "step": 885
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019999740008468594,
+      "loss": 0.9279,
+      "step": 890
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019999575508817795,
+      "loss": 0.9784,
+      "step": 895
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019999370888559804,
+      "loss": 0.981,
+      "step": 900
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019999126148515601,
+      "loss": 0.9495,
+      "step": 905
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001999884128966714,
+      "loss": 0.968,
+      "step": 910
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001999851631315734,
+      "loss": 0.9799,
+      "step": 915
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019998151220290082,
+      "loss": 0.9995,
+      "step": 920
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019997746012530206,
+      "loss": 0.9432,
+      "step": 925
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019997300691503497,
+      "loss": 1.0105,
+      "step": 930
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00019996815258996686,
+      "loss": 0.9422,
+      "step": 935
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001999628971695744,
+      "loss": 0.9056,
+      "step": 940
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019995724067494353,
+      "loss": 0.9646,
+      "step": 945
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019995118312876944,
+      "loss": 0.9543,
+      "step": 950
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001999447245553564,
+      "loss": 1.011,
+      "step": 955
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001999378649806177,
+      "loss": 0.95,
+      "step": 960
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019993060443207554,
+      "loss": 0.9927,
+      "step": 965
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00019992294293886095,
+      "loss": 1.0095,
+      "step": 970
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00019991488053171353,
+      "loss": 0.9915,
+      "step": 975
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019990641724298156,
+      "loss": 0.9816,
+      "step": 980
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019989755310662174,
+      "loss": 0.9515,
+      "step": 985
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.000199888288158199,
+      "loss": 0.957,
+      "step": 990
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019987862243488647,
+      "loss": 0.9783,
+      "step": 995
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019986855597546526,
+      "loss": 0.9576,
+      "step": 1000
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001998580888203243,
+      "loss": 0.9497,
+      "step": 1005
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019984722101146029,
+      "loss": 0.9157,
+      "step": 1010
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019983595259247735,
+      "loss": 0.954,
+      "step": 1015
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.000199824283608587,
+      "loss": 0.9945,
+      "step": 1020
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00019981221410660788,
+      "loss": 0.9631,
+      "step": 1025
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019979974413496566,
+      "loss": 1.0138,
+      "step": 1030
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00019978687374369271,
+      "loss": 0.9555,
+      "step": 1035
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00019977360298442803,
+      "loss": 0.9725,
+      "step": 1040
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019975993191041695,
+      "loss": 1.0321,
+      "step": 1045
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019974586057651102,
+      "loss": 0.9974,
+      "step": 1050
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001997313890391676,
+      "loss": 1.0188,
+      "step": 1055
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00019971651735644995,
+      "loss": 0.9457,
+      "step": 1060
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00019970124558802663,
+      "loss": 0.9522,
+      "step": 1065
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019968557379517152,
+      "loss": 0.926,
+      "step": 1070
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019966950204076346,
+      "loss": 0.9678,
+      "step": 1075
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019965303038928608,
+      "loss": 1.0012,
+      "step": 1080
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0001996361589068274,
+      "loss": 0.9966,
+      "step": 1085
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019961888766107972,
+      "loss": 0.956,
+      "step": 1090
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001996012167213393,
+      "loss": 0.903,
+      "step": 1095
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019958314615850598,
+      "loss": 0.9292,
+      "step": 1100
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00019956467604508312,
+      "loss": 0.9788,
+      "step": 1105
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019954580645517697,
+      "loss": 0.9705,
+      "step": 1110
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019952653746449676,
+      "loss": 0.9652,
+      "step": 1115
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0001995068691503541,
+      "loss": 1.0108,
+      "step": 1120
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00019948680159166283,
+      "loss": 0.9594,
+      "step": 1125
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00019946633486893865,
+      "loss": 1.0025,
+      "step": 1130
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019944546906429874,
+      "loss": 0.9474,
+      "step": 1135
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019942420426146153,
+      "loss": 0.9728,
+      "step": 1140
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019940254054574637,
+      "loss": 0.958,
+      "step": 1145
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019938047800407302,
+      "loss": 0.9844,
+      "step": 1150
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019935801672496157,
+      "loss": 1.0021,
+      "step": 1155
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019933515679853182,
+      "loss": 0.9412,
+      "step": 1160
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00019931189831650303,
+      "loss": 0.9642,
+      "step": 1165
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0001992882413721937,
+      "loss": 0.9182,
+      "step": 1170
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019926418606052088,
+      "loss": 0.9886,
+      "step": 1175
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001992397324780001,
+      "loss": 0.9945,
+      "step": 1180
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00019921488072274475,
+      "loss": 0.9431,
+      "step": 1185
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019918963089446577,
+      "loss": 0.986,
+      "step": 1190
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001991639830944714,
+      "loss": 0.9287,
+      "step": 1195
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00019913793742566647,
+      "loss": 0.949,
+      "step": 1200
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019911149399255224,
+      "loss": 0.9563,
+      "step": 1205
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019908465290122585,
+      "loss": 0.9945,
+      "step": 1210
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019905741425938002,
+      "loss": 0.9838,
+      "step": 1215
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019902977817630243,
+      "loss": 0.9648,
+      "step": 1220
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001990017447628754,
+      "loss": 0.9129,
+      "step": 1225
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019897331413157548,
+      "loss": 1.0074,
+      "step": 1230
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00019894448639647304,
+      "loss": 0.9718,
+      "step": 1235
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019891526167323145,
+      "loss": 0.9342,
+      "step": 1240
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00019888564007910714,
+      "loss": 0.9772,
+      "step": 1245
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001988556217329488,
+      "loss": 0.953,
+      "step": 1250
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001988252067551969,
+      "loss": 0.9969,
+      "step": 1255
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019879439526788341,
+      "loss": 0.9575,
+      "step": 1260
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019876318739463114,
+      "loss": 0.9778,
+      "step": 1265
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019873158326065327,
+      "loss": 0.9934,
+      "step": 1270
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00019869958299275289,
+      "loss": 0.9699,
+      "step": 1275
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00019866718671932249,
+      "loss": 1.0175,
+      "step": 1280
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00019863439457034333,
+      "loss": 0.9513,
+      "step": 1285
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019860120667738516,
+      "loss": 1.0107,
+      "step": 1290
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019856762317360546,
+      "loss": 0.965,
+      "step": 1295
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00019853364419374902,
+      "loss": 0.9439,
+      "step": 1300
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019849926987414732,
+      "loss": 0.9695,
+      "step": 1305
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019846450035271808,
+      "loss": 0.9242,
+      "step": 1310
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0001984293357689647,
+      "loss": 0.9807,
+      "step": 1315
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019839377626397554,
+      "loss": 0.9666,
+      "step": 1320
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019835782198042364,
+      "loss": 0.9653,
+      "step": 1325
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019832147306256576,
+      "loss": 0.9482,
+      "step": 1330
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001982847296562423,
+      "loss": 0.9439,
+      "step": 1335
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019824759190887622,
+      "loss": 0.9696,
+      "step": 1340
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019821005996947273,
+      "loss": 0.947,
+      "step": 1345
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019817213398861866,
+      "loss": 0.9564,
+      "step": 1350
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019813381411848185,
+      "loss": 1.0109,
+      "step": 1355
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0001980951005128104,
+      "loss": 0.9487,
+      "step": 1360
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0001980559933269323,
+      "loss": 0.9799,
+      "step": 1365
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019801649271775459,
+      "loss": 0.967,
+      "step": 1370
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019797659884376282,
+      "loss": 0.9259,
+      "step": 1375
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019793631186502047,
+      "loss": 0.9562,
+      "step": 1380
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019789563194316817,
+      "loss": 0.9456,
+      "step": 1385
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019785455924142318,
+      "loss": 0.9771,
+      "step": 1390
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019781309392457867,
+      "loss": 0.9406,
+      "step": 1395
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001977712361590031,
+      "loss": 0.981,
+      "step": 1400
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019772898611263942,
+      "loss": 0.9472,
+      "step": 1405
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019768634395500465,
+      "loss": 0.9453,
+      "step": 1410
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019764330985718896,
+      "loss": 0.9404,
+      "step": 1415
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019759988399185505,
+      "loss": 0.9435,
+      "step": 1420
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001975560665332376,
+      "loss": 0.984,
+      "step": 1425
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019751185765714234,
+      "loss": 0.9327,
+      "step": 1430
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019746725754094544,
+      "loss": 1.0195,
+      "step": 1435
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019742226636359296,
+      "loss": 0.9728,
+      "step": 1440
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019737688430559986,
+      "loss": 1.0601,
+      "step": 1445
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019733111154904943,
+      "loss": 0.9826,
+      "step": 1450
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019728494827759256,
+      "loss": 0.9481,
+      "step": 1455
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019723839467644699,
+      "loss": 1.0373,
+      "step": 1460
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019719145093239645,
+      "loss": 0.9958,
+      "step": 1465
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00019714411723379015,
+      "loss": 1.0146,
+      "step": 1470
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019709639377054177,
+      "loss": 0.98,
+      "step": 1475
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001970482807341289,
+      "loss": 0.9914,
+      "step": 1480
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019699977831759218,
+      "loss": 0.9927,
+      "step": 1485
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001969508867155345,
+      "loss": 0.9551,
+      "step": 1490
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019690160612412022,
+      "loss": 0.9567,
+      "step": 1495
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019685193674107452,
+      "loss": 1.0195,
+      "step": 1500
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001968018787656824,
+      "loss": 0.9472,
+      "step": 1505
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019675143239878805,
+      "loss": 0.9845,
+      "step": 1510
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019670059784279394,
+      "loss": 1.0282,
+      "step": 1515
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019664937530166002,
+      "loss": 0.9489,
+      "step": 1520
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.000196597764980903,
+      "loss": 0.9755,
+      "step": 1525
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019654576708759538,
+      "loss": 0.9942,
+      "step": 1530
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019649338183036475,
+      "loss": 0.9597,
+      "step": 1535
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00019644060941939286,
+      "loss": 0.9735,
+      "step": 1540
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0001963874500664148,
+      "loss": 0.982,
+      "step": 1545
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019633390398471817,
+      "loss": 0.948,
+      "step": 1550
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019627997138914224,
+      "loss": 0.9767,
+      "step": 1555
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019622565249607704,
+      "loss": 0.9506,
+      "step": 1560
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019617094752346254,
+      "loss": 0.9403,
+      "step": 1565
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001961158566907877,
+      "loss": 0.9344,
+      "step": 1570
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019606038021908963,
+      "loss": 0.9416,
+      "step": 1575
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019600451833095287,
+      "loss": 0.9499,
+      "step": 1580
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019594827125050807,
+      "loss": 0.9864,
+      "step": 1585
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019589163920343163,
+      "loss": 0.9493,
+      "step": 1590
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019583462241694436,
+      "loss": 0.9822,
+      "step": 1595
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00019577722111981078,
+      "loss": 0.9847,
+      "step": 1600
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019571943554233819,
+      "loss": 1.0152,
+      "step": 1605
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019566126591637558,
+      "loss": 0.9449,
+      "step": 1610
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019560271247531308,
+      "loss": 0.9329,
+      "step": 1615
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001955437754540805,
+      "loss": 1.0026,
+      "step": 1620
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019548445508914681,
+      "loss": 1.0132,
+      "step": 1625
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019542475161851906,
+      "loss": 0.9635,
+      "step": 1630
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019536466528174133,
+      "loss": 0.9352,
+      "step": 1635
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019530419631989392,
+      "loss": 0.9849,
+      "step": 1640
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019524334497559226,
+      "loss": 0.9475,
+      "step": 1645
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019518211149298595,
+      "loss": 0.9999,
+      "step": 1650
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019512049611775795,
+      "loss": 0.9432,
+      "step": 1655
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019505849909712332,
+      "loss": 0.9787,
+      "step": 1660
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019499612067982848,
+      "loss": 0.989,
+      "step": 1665
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019493336111615003,
+      "loss": 0.9421,
+      "step": 1670
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00019487022065789385,
+      "loss": 0.9757,
+      "step": 1675
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019480669955839402,
+      "loss": 0.9923,
+      "step": 1680
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00019474279807251186,
+      "loss": 0.9455,
+      "step": 1685
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019467851645663494,
+      "loss": 0.9216,
+      "step": 1690
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019461385496867594,
+      "loss": 0.9623,
+      "step": 1695
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019454881386807163,
+      "loss": 0.969,
+      "step": 1700
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.000194483393415782,
+      "loss": 0.9648,
+      "step": 1705
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019441759387428903,
+      "loss": 0.957,
+      "step": 1710
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0001943514155075956,
+      "loss": 0.9866,
+      "step": 1715
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019428485858122472,
+      "loss": 0.9538,
+      "step": 1720
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019421792336221807,
+      "loss": 0.9369,
+      "step": 1725
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019415061011913523,
+      "loss": 1.0101,
+      "step": 1730
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019408291912205246,
+      "loss": 0.9173,
+      "step": 1735
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019401485064256176,
+      "loss": 1.0116,
+      "step": 1740
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019394640495376954,
+      "loss": 0.9867,
+      "step": 1745
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001938775823302957,
+      "loss": 0.9117,
+      "step": 1750
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00019380838304827256,
+      "loss": 0.9466,
+      "step": 1755
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019373880738534358,
+      "loss": 0.9731,
+      "step": 1760
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001936688556206624,
+      "loss": 0.9643,
+      "step": 1765
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019359852803489168,
+      "loss": 0.9608,
+      "step": 1770
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019352782491020195,
+      "loss": 0.9447,
+      "step": 1775
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001934567465302704,
+      "loss": 0.9874,
+      "step": 1780
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019338529318028,
+      "loss": 0.9387,
+      "step": 1785
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019331346514691813,
+      "loss": 0.9288,
+      "step": 1790
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019324126271837538,
+      "loss": 0.9475,
+      "step": 1795
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019316868618434455,
+      "loss": 0.9658,
+      "step": 1800
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019309573583601954,
+      "loss": 0.942,
+      "step": 1805
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019302241196609397,
+      "loss": 0.9382,
+      "step": 1810
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019294871486876013,
+      "loss": 0.9777,
+      "step": 1815
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001928746448397078,
+      "loss": 0.9431,
+      "step": 1820
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001928002021761231,
+      "loss": 0.9449,
+      "step": 1825
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019272538717668715,
+      "loss": 0.9613,
+      "step": 1830
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.000192650200141575,
+      "loss": 0.9659,
+      "step": 1835
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019257464137245446,
+      "loss": 0.9767,
+      "step": 1840
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001924987111724847,
+      "loss": 0.9568,
+      "step": 1845
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0001924224098463153,
+      "loss": 0.969,
+      "step": 1850
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019234573770008474,
+      "loss": 0.9678,
+      "step": 1855
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019226869504141943,
+      "loss": 0.9344,
+      "step": 1860
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019219128217943233,
+      "loss": 0.9624,
+      "step": 1865
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019211349942472165,
+      "loss": 0.966,
+      "step": 1870
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019203534708936985,
+      "loss": 1.0469,
+      "step": 1875
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019195682548694208,
+      "loss": 0.9285,
+      "step": 1880
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00019187793493248524,
+      "loss": 0.9235,
+      "step": 1885
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019179867574252638,
+      "loss": 0.9598,
+      "step": 1890
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019171904823507178,
+      "loss": 0.9784,
+      "step": 1895
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019163905272960528,
+      "loss": 0.9909,
+      "step": 1900
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0001915586895470874,
+      "loss": 0.9529,
+      "step": 1905
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0001914779590099538,
+      "loss": 0.9368,
+      "step": 1910
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019139686144211405,
+      "loss": 1.0026,
+      "step": 1915
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019131539716895024,
+      "loss": 0.9895,
+      "step": 1920
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019123356651731593,
+      "loss": 1.0119,
+      "step": 1925
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019115136981553464,
+      "loss": 1.0007,
+      "step": 1930
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00019106880739339844,
+      "loss": 0.9512,
+      "step": 1935
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019098587958216688,
+      "loss": 0.9515,
+      "step": 1940
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0001909025867145655,
+      "loss": 0.9643,
+      "step": 1945
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019081892912478456,
+      "loss": 0.9237,
+      "step": 1950
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019073490714847759,
+      "loss": 0.9493,
+      "step": 1955
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019065052112276018,
+      "loss": 0.9399,
+      "step": 1960
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00019056577138620856,
+      "loss": 0.9452,
+      "step": 1965
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019048065827885827,
+      "loss": 0.9864,
+      "step": 1970
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019039518214220276,
+      "loss": 0.9163,
+      "step": 1975
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.000190309343319192,
+      "loss": 0.9867,
+      "step": 1980
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019022314215423123,
+      "loss": 0.966,
+      "step": 1985
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019013657899317942,
+      "loss": 0.928,
+      "step": 1990
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019004965418334798,
+      "loss": 0.9204,
+      "step": 1995
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001899623680734993,
+      "loss": 1.0279,
+      "step": 2000
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00018987472101384545,
+      "loss": 0.9373,
+      "step": 2005
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00018978671335604665,
+      "loss": 0.9348,
+      "step": 2010
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00018969834545321,
+      "loss": 0.9843,
+      "step": 2015
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00018960961765988792,
+      "loss": 0.9544,
+      "step": 2020
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00018952053033207683,
+      "loss": 0.9533,
+      "step": 2025
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00018943108382721562,
+      "loss": 0.931,
+      "step": 2030
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001893412785041844,
+      "loss": 0.9711,
+      "step": 2035
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00018925111472330283,
+      "loss": 0.9499,
+      "step": 2040
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00018916059284632884,
+      "loss": 0.977,
+      "step": 2045
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00018906971323645713,
+      "loss": 0.9472,
+      "step": 2050
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001889784762583177,
+      "loss": 0.949,
+      "step": 2055
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00018888688227797432,
+      "loss": 0.9756,
+      "step": 2060
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001887949316629233,
+      "loss": 0.984,
+      "step": 2065
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00018870262478209163,
+      "loss": 0.9647,
+      "step": 2070
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00018860996200583595,
+      "loss": 0.9458,
+      "step": 2075
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00018851694370594069,
+      "loss": 0.9515,
+      "step": 2080
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00018842357025561674,
+      "loss": 0.9702,
+      "step": 2085
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00018832984202949996,
+      "loss": 0.9536,
+      "step": 2090
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001882357594036497,
+      "loss": 1.0003,
+      "step": 2095
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00018814132275554713,
+      "loss": 0.9873,
+      "step": 2100
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0001880465324640939,
+      "loss": 0.9468,
+      "step": 2105
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001879513889096106,
+      "loss": 0.91,
+      "step": 2110
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00018785589247383513,
+      "loss": 0.963,
+      "step": 2115
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00018776004353992124,
+      "loss": 0.9428,
+      "step": 2120
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00018766384249243703,
+      "loss": 0.9523,
+      "step": 2125
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00018756728971736327,
+      "loss": 0.9452,
+      "step": 2130
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00018747038560209208,
+      "loss": 0.9214,
+      "step": 2135
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00018737313053542512,
+      "loss": 0.9716,
+      "step": 2140
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00018727552490757217,
+      "loss": 0.9555,
+      "step": 2145
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001871775691101496,
+      "loss": 0.9227,
+      "step": 2150
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00018707926353617865,
+      "loss": 0.9556,
+      "step": 2155
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00018698060858008403,
+      "loss": 0.9862,
+      "step": 2160
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00018688160463769222,
+      "loss": 0.9951,
+      "step": 2165
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00018678225210622986,
+      "loss": 0.8858,
+      "step": 2170
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00018668255138432237,
+      "loss": 0.9466,
+      "step": 2175
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00018658250287199196,
+      "loss": 0.99,
+      "step": 2180
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001864821069706564,
+      "loss": 0.966,
+      "step": 2185
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00018638136408312728,
+      "loss": 0.9413,
+      "step": 2190
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00018628027461360834,
+      "loss": 0.9428,
+      "step": 2195
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0001861788389676939,
+      "loss": 0.9792,
+      "step": 2200
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00018607705755236716,
+      "loss": 0.9799,
+      "step": 2205
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00018597493077599867,
+      "loss": 0.9549,
+      "step": 2210
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001858724590483447,
+      "loss": 0.9422,
+      "step": 2215
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00018576964278054544,
+      "loss": 0.9438,
+      "step": 2220
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00018566648238512357,
+      "loss": 0.9763,
+      "step": 2225
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00018556297827598242,
+      "loss": 0.9515,
+      "step": 2230
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00018545913086840434,
+      "loss": 0.9293,
+      "step": 2235
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00018535494057904915,
+      "loss": 0.9969,
+      "step": 2240
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00018525040782595235,
+      "loss": 0.9516,
+      "step": 2245
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00018514553302852356,
+      "loss": 0.9467,
+      "step": 2250
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001850403166075446,
+      "loss": 0.9982,
+      "step": 2255
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00018493475898516813,
+      "loss": 0.9582,
+      "step": 2260
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001848288605849157,
+      "loss": 0.9908,
+      "step": 2265
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00018472262183167614,
+      "loss": 0.9683,
+      "step": 2270
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00018461604315170389,
+      "loss": 0.9615,
+      "step": 2275
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00018450912497261723,
+      "loss": 1.0142,
+      "step": 2280
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00018440186772339663,
+      "loss": 1.0032,
+      "step": 2285
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00018429427183438288,
+      "loss": 0.986,
+      "step": 2290
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00018418633773727562,
+      "loss": 0.9617,
+      "step": 2295
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00018407806586513134,
+      "loss": 1.0442,
+      "step": 2300
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00018396945665236184,
+      "loss": 0.9725,
+      "step": 2305
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00018386051053473232,
+      "loss": 0.9812,
+      "step": 2310
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00018375122794935984,
+      "loss": 0.9325,
+      "step": 2315
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00018364160933471134,
+      "loss": 0.973,
+      "step": 2320
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00018353165513060202,
+      "loss": 0.9885,
+      "step": 2325
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001834213657781936,
+      "loss": 0.959,
+      "step": 2330
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00018331074171999237,
+      "loss": 0.955,
+      "step": 2335
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00018319978339984767,
+      "loss": 0.9766,
+      "step": 2340
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001830884912629499,
+      "loss": 0.9408,
+      "step": 2345
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001829768657558288,
+      "loss": 0.9502,
+      "step": 2350
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00018286490732635172,
+      "loss": 0.9516,
+      "step": 2355
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00018275261642372175,
+      "loss": 0.9545,
+      "step": 2360
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001826399934984759,
+      "loss": 0.9735,
+      "step": 2365
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001825270390024834,
+      "loss": 0.9447,
+      "step": 2370
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001824137533889438,
+      "loss": 0.9672,
+      "step": 2375
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00018230013711238513,
+      "loss": 0.9548,
+      "step": 2380
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00018218619062866211,
+      "loss": 1.011,
+      "step": 2385
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00018207191439495438,
+      "loss": 0.9822,
+      "step": 2390
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00018195730886976462,
+      "loss": 0.9096,
+      "step": 2395
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00018184237451291665,
+      "loss": 0.9577,
+      "step": 2400
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00018172711178555368,
+      "loss": 0.945,
+      "step": 2405
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00018161152115013637,
+      "loss": 0.9408,
+      "step": 2410
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001814956030704411,
+      "loss": 0.9714,
+      "step": 2415
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00018137935801155794,
+      "loss": 0.9548,
+      "step": 2420
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00018126278643988895,
+      "loss": 0.9534,
+      "step": 2425
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001811458888231462,
+      "loss": 0.9586,
+      "step": 2430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00018102866563034995,
+      "loss": 0.9891,
+      "step": 2435
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0001809111173318267,
+      "loss": 0.9938,
+      "step": 2440
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00018079324439920742,
+      "loss": 0.9306,
+      "step": 2445
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00018067504730542551,
+      "loss": 0.9616,
+      "step": 2450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00018055652652471506,
+      "loss": 0.9537,
+      "step": 2455
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0001804376825326088,
+      "loss": 0.9976,
+      "step": 2460
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00018031851580593627,
+      "loss": 0.9823,
+      "step": 2465
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00018019902682282193,
+      "loss": 0.9452,
+      "step": 2470
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001800792160626832,
+      "loss": 0.9457,
+      "step": 2475
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0001799590840062285,
+      "loss": 0.9093,
+      "step": 2480
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001798386311354554,
+      "loss": 0.9901,
+      "step": 2485
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00017971785793364866,
+      "loss": 0.9244,
+      "step": 2490
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00017959676488537825,
+      "loss": 1.0049,
+      "step": 2495
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001794753524764975,
+      "loss": 0.9679,
+      "step": 2500
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00017935362119414098,
+      "loss": 0.9085,
+      "step": 2505
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00017923157152672278,
+      "loss": 0.9318,
+      "step": 2510
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0001791092039639343,
+      "loss": 1.0041,
+      "step": 2515
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00017898651899674254,
+      "loss": 0.9653,
+      "step": 2520
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001788635171173879,
+      "loss": 0.9806,
+      "step": 2525
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00017874019881938233,
+      "loss": 0.9317,
+      "step": 2530
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00017861656459750735,
+      "loss": 0.9239,
+      "step": 2535
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.000178492614947812,
+      "loss": 0.9849,
+      "step": 2540
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00017836835036761094,
+      "loss": 0.9522,
+      "step": 2545
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00017824377135548236,
+      "loss": 0.957,
+      "step": 2550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.000178118878411266,
+      "loss": 0.9839,
+      "step": 2555
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00017799367203606128,
+      "loss": 0.9661,
+      "step": 2560
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001778681527322251,
+      "loss": 0.9682,
+      "step": 2565
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00017774232100336982,
+      "loss": 0.9406,
+      "step": 2570
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00017761617735436152,
+      "loss": 0.9704,
+      "step": 2575
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00017748972229131757,
+      "loss": 0.9781,
+      "step": 2580
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00017736295632160494,
+      "loss": 0.9033,
+      "step": 2585
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.000177235879953838,
+      "loss": 0.9342,
+      "step": 2590
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00017710849369787646,
+      "loss": 0.996,
+      "step": 2595
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00017698079806482343,
+      "loss": 0.9728,
+      "step": 2600
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0001768527935670233,
+      "loss": 0.9596,
+      "step": 2605
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001767244807180597,
+      "loss": 0.9935,
+      "step": 2610
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0001765958600327534,
+      "loss": 0.9791,
+      "step": 2615
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00017646693202716033,
+      "loss": 0.9885,
+      "step": 2620
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00017633769721856938,
+      "loss": 0.9904,
+      "step": 2625
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001762081561255005,
+      "loss": 0.9769,
+      "step": 2630
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0001760783092677025,
+      "loss": 0.9886,
+      "step": 2635
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00017594815716615093,
+      "loss": 0.9557,
+      "step": 2640
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00017581770034304613,
+      "loss": 0.9793,
+      "step": 2645
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.000175686939321811,
+      "loss": 0.969,
+      "step": 2650
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00017555587462708894,
+      "loss": 0.9779,
+      "step": 2655
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00017542450678474184,
+      "loss": 0.9713,
+      "step": 2660
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00017529283632184784,
+      "loss": 1.0001,
+      "step": 2665
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00017516086376669917,
+      "loss": 0.9495,
+      "step": 2670
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00017502858964880028,
+      "loss": 0.9374,
+      "step": 2675
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00017489601449886547,
+      "loss": 0.9121,
+      "step": 2680
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0001747631388488169,
+      "loss": 1.0079,
+      "step": 2685
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00017462996323178235,
+      "loss": 0.9612,
+      "step": 2690
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00017449648818209314,
+      "loss": 0.9592,
+      "step": 2695
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00017436271423528206,
+      "loss": 0.9785,
+      "step": 2700
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00017422864192808107,
+      "loss": 0.9565,
+      "step": 2705
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001740942717984192,
+      "loss": 0.9596,
+      "step": 2710
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00017395960438542057,
+      "loss": 0.953,
+      "step": 2715
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00017382464022940182,
+      "loss": 0.9306,
+      "step": 2720
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0001736893798718704,
+      "loss": 0.9798,
+      "step": 2725
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00017355382385552206,
+      "loss": 0.9611,
+      "step": 2730
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00017341797272423895,
+      "loss": 0.9406,
+      "step": 2735
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0001732818270230871,
+      "loss": 0.9695,
+      "step": 2740
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00017314538729831455,
+      "loss": 0.939,
+      "step": 2745
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.000173008654097349,
+      "loss": 0.9651,
+      "step": 2750
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00017287162796879568,
+      "loss": 0.9339,
+      "step": 2755
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.000172734309462435,
+      "loss": 0.986,
+      "step": 2760
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00017259669912922062,
+      "loss": 0.9875,
+      "step": 2765
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00017245879752127692,
+      "loss": 0.9468,
+      "step": 2770
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00017232060519189703,
+      "loss": 0.9657,
+      "step": 2775
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001721821226955405,
+      "loss": 0.9782,
+      "step": 2780
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00017204335058783112,
+      "loss": 0.9734,
+      "step": 2785
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00017190428942555463,
+      "loss": 0.9407,
+      "step": 2790
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00017176493976665656,
+      "loss": 0.9671,
+      "step": 2795
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001716253021702399,
+      "loss": 0.9668,
+      "step": 2800
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.000171485377196563,
+      "loss": 0.9813,
+      "step": 2805
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0001713451654070371,
+      "loss": 0.9691,
+      "step": 2810
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00017120466736422437,
+      "loss": 0.9934,
+      "step": 2815
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0001710638836318354,
+      "loss": 0.9472,
+      "step": 2820
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00017092281477472704,
+      "loss": 0.9592,
+      "step": 2825
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00017078146135890014,
+      "loss": 0.9529,
+      "step": 2830
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001706398239514973,
+      "loss": 0.9202,
+      "step": 2835
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0001704979031208004,
+      "loss": 0.9266,
+      "step": 2840
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00017035569943622876,
+      "loss": 0.9419,
+      "step": 2845
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001702132134683363,
+      "loss": 0.9281,
+      "step": 2850
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00017007044578880968,
+      "loss": 0.9769,
+      "step": 2855
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00016992739697046586,
+      "loss": 0.9802,
+      "step": 2860
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00016978406758724968,
+      "loss": 0.9337,
+      "step": 2865
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00016964045821423178,
+      "loss": 0.9348,
+      "step": 2870
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001694965694276061,
+      "loss": 0.9377,
+      "step": 2875
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00016935240180468775,
+      "loss": 0.9952,
+      "step": 2880
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0001692079559239105,
+      "loss": 0.9321,
+      "step": 2885
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00016906323236482465,
+      "loss": 0.9443,
+      "step": 2890
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00016891823170809447,
+      "loss": 0.95,
+      "step": 2895
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00016877295453549614,
+      "loss": 0.902,
+      "step": 2900
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00016862740142991527,
+      "loss": 0.9792,
+      "step": 2905
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00016848157297534453,
+      "loss": 1.0632,
+      "step": 2910
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00016833546975688135,
+      "loss": 0.9456,
+      "step": 2915
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001681890923607256,
+      "loss": 0.9323,
+      "step": 2920
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00016804244137417725,
+      "loss": 0.9667,
+      "step": 2925
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00016789551738563384,
+      "loss": 1.0013,
+      "step": 2930
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00016774832098458847,
+      "loss": 0.9697,
+      "step": 2935
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00016760085276162708,
+      "loss": 0.9564,
+      "step": 2940
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00016745311330842622,
+      "loss": 0.9877,
+      "step": 2945
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00016730510321775075,
+      "loss": 0.9751,
+      "step": 2950
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00016715682308345131,
+      "loss": 0.9653,
+      "step": 2955
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00016700827350046206,
+      "loss": 0.9419,
+      "step": 2960
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00016685945506479829,
+      "loss": 0.967,
+      "step": 2965
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00016671036837355386,
+      "loss": 0.9686,
+      "step": 2970
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00016656101402489913,
+      "loss": 0.9312,
+      "step": 2975
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00016641139261807818,
+      "loss": 0.9774,
+      "step": 2980
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00016626150475340667,
+      "loss": 0.9456,
+      "step": 2985
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00016611135103226937,
+      "loss": 0.9756,
+      "step": 2990
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00016596093205711765,
+      "loss": 0.9411,
+      "step": 2995
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00016581024843146725,
+      "loss": 1.0284,
+      "step": 3000
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001656593007598956,
+      "loss": 0.9242,
+      "step": 3005
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00016550808964803978,
+      "loss": 0.9531,
+      "step": 3010
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00016535661570259359,
+      "loss": 0.9997,
+      "step": 3015
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00016520487953130552,
+      "loss": 0.9521,
+      "step": 3020
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00016505288174297616,
+      "loss": 0.9336,
+      "step": 3025
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00016490062294745571,
+      "loss": 0.9519,
+      "step": 3030
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001647481037556417,
+      "loss": 0.9587,
+      "step": 3035
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00016459532477947634,
+      "loss": 0.9376,
+      "step": 3040
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00016444228663194417,
+      "loss": 0.9735,
+      "step": 3045
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00016428898992706955,
+      "loss": 0.9576,
+      "step": 3050
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001641354352799144,
+      "loss": 0.9539,
+      "step": 3055
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00016398162330657533,
+      "loss": 0.9193,
+      "step": 3060
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00016382755462418153,
+      "loss": 0.9475,
+      "step": 3065
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001636732298508922,
+      "loss": 0.9172,
+      "step": 3070
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00016351864960589393,
+      "loss": 0.9774,
+      "step": 3075
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0001633638145093984,
+      "loss": 0.9509,
+      "step": 3080
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00016320872518263982,
+      "loss": 0.9406,
+      "step": 3085
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00016305338224787235,
+      "loss": 0.9813,
+      "step": 3090
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00016289778632836774,
+      "loss": 0.9551,
+      "step": 3095
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001627419380484128,
+      "loss": 0.9364,
+      "step": 3100
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00016258583803330681,
+      "loss": 0.9944,
+      "step": 3105
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00016242948690935912,
+      "loss": 0.9569,
+      "step": 3110
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001622728853038865,
+      "loss": 0.9911,
+      "step": 3115
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00016211603384521083,
+      "loss": 1.0201,
+      "step": 3120
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001619589331626564,
+      "loss": 1.0051,
+      "step": 3125
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00016180158388654742,
+      "loss": 0.9384,
+      "step": 3130
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00016164398664820557,
+      "loss": 0.9649,
+      "step": 3135
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00016148614207994735,
+      "loss": 0.9263,
+      "step": 3140
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00016132805081508163,
+      "loss": 0.9732,
+      "step": 3145
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00016116971348790712,
+      "loss": 0.9238,
+      "step": 3150
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001610111307337097,
+      "loss": 0.9184,
+      "step": 3155
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001608523031887601,
+      "loss": 0.9524,
+      "step": 3160
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00016069323149031105,
+      "loss": 0.9733,
+      "step": 3165
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00016053391627659505,
+      "loss": 0.9432,
+      "step": 3170
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00016037435818682142,
+      "loss": 0.9818,
+      "step": 3175
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001602145578611742,
+      "loss": 0.9223,
+      "step": 3180
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00016005451594080911,
+      "loss": 0.9526,
+      "step": 3185
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00015989423306785142,
+      "loss": 0.9617,
+      "step": 3190
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00015973370988539294,
+      "loss": 0.9638,
+      "step": 3195
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00015957294703748982,
+      "loss": 0.9456,
+      "step": 3200
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00015941194516915977,
+      "loss": 0.9243,
+      "step": 3205
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00015925070492637944,
+      "loss": 0.9099,
+      "step": 3210
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00015908922695608192,
+      "loss": 0.9662,
+      "step": 3215
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001589275119061542,
+      "loss": 0.9263,
+      "step": 3220
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00015876556042543433,
+      "loss": 0.985,
+      "step": 3225
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00015860337316370916,
+      "loss": 0.9458,
+      "step": 3230
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00015844095077171142,
+      "loss": 0.9781,
+      "step": 3235
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001582782939011173,
+      "loss": 0.971,
+      "step": 3240
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0001581154032045437,
+      "loss": 0.9509,
+      "step": 3245
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00015795227933554568,
+      "loss": 0.9443,
+      "step": 3250
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.000157788922948614,
+      "loss": 0.9901,
+      "step": 3255
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00015762533469917216,
+      "loss": 0.9864,
+      "step": 3260
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00015746151524357403,
+      "loss": 0.9442,
+      "step": 3265
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00015729746523910113,
+      "loss": 0.9724,
+      "step": 3270
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00015713318534395996,
+      "loss": 0.935,
+      "step": 3275
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00015696867621727942,
+      "loss": 0.9613,
+      "step": 3280
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00015680393851910811,
+      "loss": 0.9971,
+      "step": 3285
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00015663897291041175,
+      "loss": 0.9521,
+      "step": 3290
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0001564737800530705,
+      "loss": 0.9817,
+      "step": 3295
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00015630836060987624,
+      "loss": 0.9148,
+      "step": 3300
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00015614271524453,
+      "loss": 1.0246,
+      "step": 3305
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00015597684462163923,
+      "loss": 0.9814,
+      "step": 3310
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00015581074940671527,
+      "loss": 0.964,
+      "step": 3315
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0001556444302661704,
+      "loss": 0.9563,
+      "step": 3320
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00015547788786731553,
+      "loss": 0.918,
+      "step": 3325
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00015531112287835717,
+      "loss": 0.9815,
+      "step": 3330
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00015514413596839498,
+      "loss": 0.9684,
+      "step": 3335
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00015497692780741908,
+      "loss": 0.9207,
+      "step": 3340
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00015480949906630724,
+      "loss": 0.9295,
+      "step": 3345
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001546418504168222,
+      "loss": 0.9619,
+      "step": 3350
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001544739825316091,
+      "loss": 0.9399,
+      "step": 3355
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00015430589608419264,
+      "loss": 0.9613,
+      "step": 3360
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00015413759174897455,
+      "loss": 0.978,
+      "step": 3365
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00015396907020123068,
+      "loss": 0.9336,
+      "step": 3370
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00015380033211710842,
+      "loss": 0.9627,
+      "step": 3375
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00015363137817362392,
+      "loss": 0.9808,
+      "step": 3380
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00015346220904865945,
+      "loss": 0.9829,
+      "step": 3385
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00015329282542096064,
+      "loss": 0.9156,
+      "step": 3390
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001531232279701337,
+      "loss": 0.9248,
+      "step": 3395
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00015295341737664285,
+      "loss": 0.9772,
+      "step": 3400
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00015278339432180736,
+      "loss": 0.9806,
+      "step": 3405
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.000152613159487799,
+      "loss": 0.9854,
+      "step": 3410
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00015244271355763933,
+      "loss": 0.9877,
+      "step": 3415
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00015227205721519675,
+      "loss": 0.939,
+      "step": 3420
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001521011911451839,
+      "loss": 0.9387,
+      "step": 3425
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00015193011603315503,
+      "loss": 0.9922,
+      "step": 3430
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00015175883256550291,
+      "loss": 0.9371,
+      "step": 3435
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00015158734142945644,
+      "loss": 0.9549,
+      "step": 3440
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00015141564331307768,
+      "loss": 0.9199,
+      "step": 3445
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001512437389052591,
+      "loss": 0.9866,
+      "step": 3450
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00015107162889572092,
+      "loss": 0.9582,
+      "step": 3455
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001508993139750083,
+      "loss": 0.9177,
+      "step": 3460
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00015072679483448852,
+      "loss": 0.9264,
+      "step": 3465
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0001505540721663481,
+      "loss": 0.9603,
+      "step": 3470
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00015038114666359042,
+      "loss": 0.8968,
+      "step": 3475
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001502080190200325,
+      "loss": 0.9603,
+      "step": 3480
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00015003468993030248,
+      "loss": 0.9649,
+      "step": 3485
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00014986116008983664,
+      "loss": 0.9562,
+      "step": 3490
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00014968743019487686,
+      "loss": 0.9489,
+      "step": 3495
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00014951350094246762,
+      "loss": 0.9261,
+      "step": 3500
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00014933937303045325,
+      "loss": 0.9313,
+      "step": 3505
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0001491650471574751,
+      "loss": 0.9792,
+      "step": 3510
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00014899052402296895,
+      "loss": 0.9988,
+      "step": 3515
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00014881580432716182,
+      "loss": 0.9285,
+      "step": 3520
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001486408887710696,
+      "loss": 1.0013,
+      "step": 3525
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00014846577805649388,
+      "loss": 0.9516,
+      "step": 3530
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001482904728860193,
+      "loss": 0.9365,
+      "step": 3535
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00014811497396301072,
+      "loss": 1.0238,
+      "step": 3540
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001479392819916103,
+      "loss": 0.9546,
+      "step": 3545
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00014776339767673491,
+      "loss": 0.9308,
+      "step": 3550
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00014758732172407302,
+      "loss": 0.8988,
+      "step": 3555
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.000147411054840082,
+      "loss": 0.9015,
+      "step": 3560
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00014723459773198535,
+      "loss": 0.9554,
+      "step": 3565
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00014705795110776974,
+      "loss": 0.9538,
+      "step": 3570
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00014688111567618225,
+      "loss": 0.9782,
+      "step": 3575
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0001467040921467275,
+      "loss": 0.9925,
+      "step": 3580
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00014652688122966475,
+      "loss": 0.9505,
+      "step": 3585
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00014634948363600518,
+      "loss": 0.9953,
+      "step": 3590
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0001461719000775089,
+      "loss": 0.9487,
+      "step": 3595
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00014599413126668213,
+      "loss": 0.9231,
+      "step": 3600
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00014581617791677452,
+      "loss": 0.9661,
+      "step": 3605
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00014563804074177588,
+      "loss": 0.9161,
+      "step": 3610
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0001454597204564139,
+      "loss": 0.9306,
+      "step": 3615
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00014528121777615058,
+      "loss": 0.9724,
+      "step": 3620
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00014510253341718,
+      "loss": 0.905,
+      "step": 3625
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001449236680964251,
+      "loss": 0.9708,
+      "step": 3630
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00014474462253153482,
+      "loss": 0.9457,
+      "step": 3635
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00014456539744088138,
+      "loss": 0.983,
+      "step": 3640
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00014438599354355723,
+      "loss": 0.9451,
+      "step": 3645
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00014420641155937224,
+      "loss": 0.9913,
+      "step": 3650
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00014402665220885087,
+      "loss": 0.9705,
+      "step": 3655
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00014384671621322915,
+      "loss": 0.9837,
+      "step": 3660
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00014366660429445183,
+      "loss": 0.9603,
+      "step": 3665
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00014348631717516953,
+      "loss": 0.9084,
+      "step": 3670
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00014330585557873588,
+      "loss": 0.9671,
+      "step": 3675
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00014312522022920444,
+      "loss": 1.0088,
+      "step": 3680
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00014294441185132607,
+      "loss": 0.9046,
+      "step": 3685
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00014276343117054563,
+      "loss": 0.9526,
+      "step": 3690
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00014258227891299948,
+      "loss": 0.9559,
+      "step": 3695
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00014240095580551234,
+      "loss": 0.9475,
+      "step": 3700
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00014221946257559438,
+      "loss": 0.9121,
+      "step": 3705
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0001420377999514384,
+      "loss": 0.9302,
+      "step": 3710
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001418559686619168,
+      "loss": 0.9858,
+      "step": 3715
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00014167396943657874,
+      "loss": 1.0036,
+      "step": 3720
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00014149180300564713,
+      "loss": 0.9651,
+      "step": 3725
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001413094701000158,
+      "loss": 1.0074,
+      "step": 3730
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001411269714512465,
+      "loss": 0.9187,
+      "step": 3735
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00014094430779156597,
+      "loss": 0.9897,
+      "step": 3740
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00014076147985386296,
+      "loss": 0.9479,
+      "step": 3745
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00014057848837168547,
+      "loss": 0.9694,
+      "step": 3750
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00014039533407923758,
+      "loss": 0.9581,
+      "step": 3755
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00014021201771137663,
+      "loss": 0.9638,
+      "step": 3760
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001400285400036102,
+      "loss": 0.9051,
+      "step": 3765
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00013984490169209333,
+      "loss": 0.9816,
+      "step": 3770
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00013966110351362523,
+      "loss": 0.9417,
+      "step": 3775
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00013947714620564678,
+      "loss": 1.0037,
+      "step": 3780
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001392930305062371,
+      "loss": 0.9123,
+      "step": 3785
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00013910875715411098,
+      "loss": 0.9404,
+      "step": 3790
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00013892432688861566,
+      "loss": 1.0267,
+      "step": 3795
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.000138739740449728,
+      "loss": 0.9702,
+      "step": 3800
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00013855499857805137,
+      "loss": 0.9276,
+      "step": 3805
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0001383701020148129,
+      "loss": 0.9404,
+      "step": 3810
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001381850515018603,
+      "loss": 0.9507,
+      "step": 3815
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.000137999847781659,
+      "loss": 0.9401,
+      "step": 3820
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00013781449159728907,
+      "loss": 0.9393,
+      "step": 3825
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00013762898369244238,
+      "loss": 0.9554,
+      "step": 3830
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00013744332481141952,
+      "loss": 0.9233,
+      "step": 3835
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00013725751569912682,
+      "loss": 0.9486,
+      "step": 3840
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00013707155710107326,
+      "loss": 0.9328,
+      "step": 3845
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00013688544976336783,
+      "loss": 1.0144,
+      "step": 3850
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0001366991944327161,
+      "loss": 0.9676,
+      "step": 3855
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00013651279185641752,
+      "loss": 0.9442,
+      "step": 3860
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00013632624278236223,
+      "loss": 0.9467,
+      "step": 3865
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001361395479590283,
+      "loss": 0.9365,
+      "step": 3870
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00013595270813547845,
+      "loss": 0.9367,
+      "step": 3875
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00013576572406135722,
+      "loss": 0.9691,
+      "step": 3880
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0001355785964868879,
+      "loss": 0.9681,
+      "step": 3885
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00013539132616286956,
+      "loss": 0.9321,
+      "step": 3890
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.000135203913840674,
+      "loss": 0.9851,
+      "step": 3895
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00013501636027224282,
+      "loss": 0.9517,
+      "step": 3900
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00013482866621008417,
+      "loss": 0.9395,
+      "step": 3905
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001346408324072701,
+      "loss": 0.9875,
+      "step": 3910
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001344528596174331,
+      "loss": 0.9648,
+      "step": 3915
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0001342647485947635,
+      "loss": 0.9321,
+      "step": 3920
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001340765000940062,
+      "loss": 0.9561,
+      "step": 3925
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00013388811487045766,
+      "loss": 0.8957,
+      "step": 3930
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00013369959367996286,
+      "loss": 0.952,
+      "step": 3935
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0001335109372789125,
+      "loss": 0.9556,
+      "step": 3940
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00013332214642423952,
+      "loss": 0.9346,
+      "step": 3945
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00013313322187341652,
+      "loss": 0.9259,
+      "step": 3950
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00013294416438445244,
+      "loss": 0.918,
+      "step": 3955
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00013275497471588953,
+      "loss": 0.9664,
+      "step": 3960
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00013256565362680057,
+      "loss": 0.9847,
+      "step": 3965
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001323762018767854,
+      "loss": 0.9926,
+      "step": 3970
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00013218662022596828,
+      "loss": 0.9912,
+      "step": 3975
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00013199690943499457,
+      "loss": 0.9541,
+      "step": 3980
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001318070702650278,
+      "loss": 0.909,
+      "step": 3985
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001316171034777466,
+      "loss": 0.9476,
+      "step": 3990
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00013142700983534157,
+      "loss": 0.9356,
+      "step": 3995
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00013123679010051232,
+      "loss": 0.9414,
+      "step": 4000
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001310464450364644,
+      "loss": 0.8908,
+      "step": 4005
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00013085597540690618,
+      "loss": 0.9375,
+      "step": 4010
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0001306653819760458,
+      "loss": 0.8908,
+      "step": 4015
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00013047466550858812,
+      "loss": 0.9261,
+      "step": 4020
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00013028382676973178,
+      "loss": 1.0016,
+      "step": 4025
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00013009286652516575,
+      "loss": 0.915,
+      "step": 4030
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00012990178554106674,
+      "loss": 0.9558,
+      "step": 4035
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00012971058458409576,
+      "loss": 0.9889,
+      "step": 4040
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001295192644213952,
+      "loss": 0.98,
+      "step": 4045
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00012932782582058584,
+      "loss": 0.9573,
+      "step": 4050
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00012913626954976355,
+      "loss": 0.9243,
+      "step": 4055
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00012894459637749627,
+      "loss": 1.0168,
+      "step": 4060
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00012875280707282116,
+      "loss": 0.967,
+      "step": 4065
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001285609024052411,
+      "loss": 0.986,
+      "step": 4070
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00012836888314472208,
+      "loss": 0.903,
+      "step": 4075
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00012817675006168963,
+      "loss": 0.9617,
+      "step": 4080
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00012798450392702615,
+      "loss": 1.0141,
+      "step": 4085
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00012779214551206746,
+      "loss": 0.9681,
+      "step": 4090
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00012759967558860006,
+      "loss": 0.9053,
+      "step": 4095
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001274070949288577,
+      "loss": 0.951,
+      "step": 4100
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00012721440430551849,
+      "loss": 0.9198,
+      "step": 4105
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00012702160449170165,
+      "loss": 0.9335,
+      "step": 4110
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001268286962609647,
+      "loss": 0.9297,
+      "step": 4115
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0001266356803873,
+      "loss": 0.99,
+      "step": 4120
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00012644255764513178,
+      "loss": 0.9614,
+      "step": 4125
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001262493288093131,
+      "loss": 0.96,
+      "step": 4130
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00012605599465512278,
+      "loss": 0.9421,
+      "step": 4135
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.000125862555958262,
+      "loss": 0.9729,
+      "step": 4140
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00012566901349485158,
+      "loss": 0.9906,
+      "step": 4145
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00012547536804142858,
+      "loss": 0.9593,
+      "step": 4150
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00012528162037494332,
+      "loss": 0.9116,
+      "step": 4155
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001250877712727561,
+      "loss": 0.9287,
+      "step": 4160
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00012489382151263442,
+      "loss": 0.955,
+      "step": 4165
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00012469977187274945,
+      "loss": 0.9838,
+      "step": 4170
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00012450562313167326,
+      "loss": 0.952,
+      "step": 4175
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00012431137606837538,
+      "loss": 0.9538,
+      "step": 4180
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00012411703146221993,
+      "loss": 0.9594,
+      "step": 4185
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001239225900929624,
+      "loss": 0.9902,
+      "step": 4190
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00012372805274074643,
+      "loss": 0.9899,
+      "step": 4195
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00012353342018610084,
+      "loss": 0.9749,
+      "step": 4200
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0001233386932099364,
+      "loss": 0.9629,
+      "step": 4205
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00012314387259354282,
+      "loss": 0.9576,
+      "step": 4210
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00012294895911858528,
+      "loss": 0.9349,
+      "step": 4215
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00012275395356710177,
+      "loss": 0.9125,
+      "step": 4220
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00012255885672149953,
+      "loss": 0.9121,
+      "step": 4225
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0001223636693645523,
+      "loss": 0.9513,
+      "step": 4230
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00012216839227939676,
+      "loss": 0.9783,
+      "step": 4235
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00012197302624952971,
+      "loss": 0.9486,
+      "step": 4240
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001217775720588048,
+      "loss": 0.9371,
+      "step": 4245
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00012158203049142947,
+      "loss": 0.9579,
+      "step": 4250
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00012138640233196152,
+      "loss": 1.0184,
+      "step": 4255
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00012119068836530644,
+      "loss": 0.9227,
+      "step": 4260
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00012099488937671384,
+      "loss": 0.9478,
+      "step": 4265
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00012079900615177449,
+      "loss": 0.9149,
+      "step": 4270
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00012060303947641715,
+      "loss": 0.9394,
+      "step": 4275
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00012040699013690543,
+      "loss": 0.9162,
+      "step": 4280
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00012021085891983456,
+      "loss": 0.9467,
+      "step": 4285
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00012001464661212827,
+      "loss": 0.9434,
+      "step": 4290
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00011981835400103568,
+      "loss": 0.9312,
+      "step": 4295
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001196219818741281,
+      "loss": 1.0016,
+      "step": 4300
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00011942553101929585,
+      "loss": 0.9505,
+      "step": 4305
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00011922900222474523,
+      "loss": 0.9585,
+      "step": 4310
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00011903239627899503,
+      "loss": 0.9276,
+      "step": 4315
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00011883571397087387,
+      "loss": 0.9423,
+      "step": 4320
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00011863895608951654,
+      "loss": 0.9189,
+      "step": 4325
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00011844212342436112,
+      "loss": 0.9851,
+      "step": 4330
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001182452167651457,
+      "loss": 0.9874,
+      "step": 4335
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00011804823690190538,
+      "loss": 0.986,
+      "step": 4340
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0001178511846249688,
+      "loss": 0.9263,
+      "step": 4345
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00011765406072495528,
+      "loss": 0.9252,
+      "step": 4350
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001174568659927714,
+      "loss": 0.9696,
+      "step": 4355
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00011725960121960806,
+      "loss": 0.9126,
+      "step": 4360
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00011706226719693698,
+      "loss": 0.9451,
+      "step": 4365
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00011686486471650798,
+      "loss": 0.999,
+      "step": 4370
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00011666739457034532,
+      "loss": 0.9888,
+      "step": 4375
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0001164698575507449,
+      "loss": 0.9613,
+      "step": 4380
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00011627225445027086,
+      "loss": 0.9938,
+      "step": 4385
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00011607458606175253,
+      "loss": 0.9332,
+      "step": 4390
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00011587685317828112,
+      "loss": 0.9306,
+      "step": 4395
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00011567905659320663,
+      "loss": 0.9353,
+      "step": 4400
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001154811971001347,
+      "loss": 0.9365,
+      "step": 4405
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00011528327549292326,
+      "loss": 0.968,
+      "step": 4410
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00011508529256567961,
+      "loss": 0.9369,
+      "step": 4415
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00011488724911275694,
+      "loss": 0.978,
+      "step": 4420
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00011468914592875135,
+      "loss": 0.9493,
+      "step": 4425
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00011449098380849858,
+      "loss": 0.9679,
+      "step": 4430
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00011429276354707086,
+      "loss": 0.9584,
+      "step": 4435
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00011409448593977363,
+      "loss": 0.9717,
+      "step": 4440
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00011389615178214253,
+      "loss": 0.9412,
+      "step": 4445
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00011369776186993994,
+      "loss": 0.9613,
+      "step": 4450
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00011349931699915209,
+      "loss": 1.009,
+      "step": 4455
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001133008179659856,
+      "loss": 0.9168,
+      "step": 4460
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001131022655668645,
+      "loss": 0.9805,
+      "step": 4465
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00011290366059842683,
+      "loss": 0.9751,
+      "step": 4470
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00011270500385752167,
+      "loss": 0.9511,
+      "step": 4475
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00011250629614120571,
+      "loss": 0.9661,
+      "step": 4480
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00011230753824674025,
+      "loss": 0.94,
+      "step": 4485
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00011210873097158786,
+      "loss": 0.9012,
+      "step": 4490
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00011190987511340928,
+      "loss": 0.9374,
+      "step": 4495
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00011171097147006013,
+      "loss": 0.9617,
+      "step": 4500
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0001115120208395878,
+      "loss": 1.0004,
+      "step": 4505
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00011131302402022821,
+      "loss": 0.9257,
+      "step": 4510
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00011111398181040251,
+      "loss": 0.9265,
+      "step": 4515
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00011091489500871408,
+      "loss": 0.9412,
+      "step": 4520
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00011071576441394516,
+      "loss": 0.9631,
+      "step": 4525
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00011051659082505366,
+      "loss": 0.9245,
+      "step": 4530
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0001103173750411701,
+      "loss": 0.9441,
+      "step": 4535
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00011011811786159416,
+      "loss": 0.992,
+      "step": 4540
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00010991882008579177,
+      "loss": 0.9648,
+      "step": 4545
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00010971948251339157,
+      "loss": 0.9309,
+      "step": 4550
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00010952010594418201,
+      "loss": 0.9653,
+      "step": 4555
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00010932069117810787,
+      "loss": 0.9478,
+      "step": 4560
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00010912123901526736,
+      "loss": 0.9484,
+      "step": 4565
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00010892175025590856,
+      "loss": 0.9292,
+      "step": 4570
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00010872222570042652,
+      "loss": 0.9283,
+      "step": 4575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00010852266614935982,
+      "loss": 0.9676,
+      "step": 4580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00010832307240338752,
+      "loss": 0.9498,
+      "step": 4585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00010812344526332578,
+      "loss": 0.9799,
+      "step": 4590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00010792378553012488,
+      "loss": 0.9604,
+      "step": 4595
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00010772409400486571,
+      "loss": 0.9349,
+      "step": 4600
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001075243714887569,
+      "loss": 0.929,
+      "step": 4605
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00010732461878313125,
+      "loss": 0.9429,
+      "step": 4610
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0001071248366894428,
+      "loss": 0.8954,
+      "step": 4615
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00010692502600926348,
+      "loss": 0.9028,
+      "step": 4620
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00010672518754427988,
+      "loss": 0.9128,
+      "step": 4625
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00010652532209629011,
+      "loss": 0.9641,
+      "step": 4630
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001063254304672005,
+      "loss": 0.9866,
+      "step": 4635
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00010612551345902244,
+      "loss": 1.0028,
+      "step": 4640
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001059255718738692,
+      "loss": 0.9608,
+      "step": 4645
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00010572560651395258,
+      "loss": 0.9699,
+      "step": 4650
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001055256181815798,
+      "loss": 0.948,
+      "step": 4655
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001053256076791503,
+      "loss": 0.9667,
+      "step": 4660
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00010512557580915235,
+      "loss": 0.917,
+      "step": 4665
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00010492552337416007,
+      "loss": 0.9426,
+      "step": 4670
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00010472545117683004,
+      "loss": 0.9653,
+      "step": 4675
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00010452536001989815,
+      "loss": 0.9518,
+      "step": 4680
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00010432525070617633,
+      "loss": 0.9645,
+      "step": 4685
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00010412512403854942,
+      "loss": 0.9644,
+      "step": 4690
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00010392498081997182,
+      "loss": 0.9523,
+      "step": 4695
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00010372482185346435,
+      "loss": 0.9408,
+      "step": 4700
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00010352464794211104,
+      "loss": 0.9521,
+      "step": 4705
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00010332445988905586,
+      "loss": 0.9245,
+      "step": 4710
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00010312425849749951,
+      "loss": 0.9627,
+      "step": 4715
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00010292404457069631,
+      "loss": 0.9672,
+      "step": 4720
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00010272381891195071,
+      "loss": 0.9933,
+      "step": 4725
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00010252358232461436,
+      "loss": 0.9642,
+      "step": 4730
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00010232333561208264,
+      "loss": 0.9775,
+      "step": 4735
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00010212307957779173,
+      "loss": 0.9309,
+      "step": 4740
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00010192281502521499,
+      "loss": 0.9637,
+      "step": 4745
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00010172254275786017,
+      "loss": 0.9439,
+      "step": 4750
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00010152226357926582,
+      "loss": 0.9763,
+      "step": 4755
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001013219782929983,
+      "loss": 0.9333,
+      "step": 4760
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00010112168770264843,
+      "loss": 0.9137,
+      "step": 4765
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001009213926118284,
+      "loss": 0.9108,
+      "step": 4770
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00010072109382416828,
+      "loss": 0.9475,
+      "step": 4775
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00010052079214331318,
+      "loss": 0.9641,
+      "step": 4780
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00010032048837291963,
+      "loss": 0.9474,
+      "step": 4785
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00010012018331665272,
+      "loss": 0.9768,
+      "step": 4790
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 9.991987777818258e-05,
+      "loss": 0.9267,
+      "step": 4795
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 9.971957256118129e-05,
+      "loss": 0.9629,
+      "step": 4800
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1953125,
+      "learning_rate": 9.951926846931969e-05,
+      "loss": 0.9381,
+      "step": 4805
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1875,
+      "learning_rate": 9.931896630626402e-05,
+      "loss": 0.9326,
+      "step": 4810
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 9.911866687567291e-05,
+      "loss": 0.905,
+      "step": 4815
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.197265625,
+      "learning_rate": 9.891837098119389e-05,
+      "loss": 1.0023,
+      "step": 4820
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 9.87180794264604e-05,
+      "loss": 0.9583,
+      "step": 4825
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1796875,
+      "learning_rate": 9.851779301508842e-05,
+      "loss": 0.9945,
+      "step": 4830
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 9.831751255067329e-05,
+      "loss": 0.9292,
+      "step": 4835
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.811723883678654e-05,
+      "loss": 1.0147,
+      "step": 4840
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 9.791697267697255e-05,
+      "loss": 0.9691,
+      "step": 4845
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.185546875,
+      "learning_rate": 9.771671487474546e-05,
+      "loss": 0.9563,
+      "step": 4850
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 9.751646623358576e-05,
+      "loss": 0.9499,
+      "step": 4855
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 9.731622755693737e-05,
+      "loss": 0.9847,
+      "step": 4860
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 9.711599964820405e-05,
+      "loss": 0.9937,
+      "step": 4865
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.691578331074643e-05,
+      "loss": 0.9582,
+      "step": 4870
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 9.671557934787874e-05,
+      "loss": 0.9172,
+      "step": 4875
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 9.651538856286551e-05,
+      "loss": 0.9437,
+      "step": 4880
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 9.631521175891844e-05,
+      "loss": 0.9188,
+      "step": 4885
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.611504973919311e-05,
+      "loss": 0.9397,
+      "step": 4890
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.19140625,
+      "learning_rate": 9.591490330678579e-05,
+      "loss": 1.0114,
+      "step": 4895
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.571477326473021e-05,
+      "loss": 0.9385,
+      "step": 4900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1875,
+      "learning_rate": 9.551466041599432e-05,
+      "loss": 0.9427,
+      "step": 4905
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.531456556347712e-05,
+      "loss": 0.9701,
+      "step": 4910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.511448951000535e-05,
+      "loss": 0.9269,
+      "step": 4915
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 9.491443305833043e-05,
+      "loss": 1.0192,
+      "step": 4920
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1875,
+      "learning_rate": 9.4714397011125e-05,
+      "loss": 0.9431,
+      "step": 4925
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.451438217097994e-05,
+      "loss": 0.9808,
+      "step": 4930
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1796875,
+      "learning_rate": 9.4314389340401e-05,
+      "loss": 0.9503,
+      "step": 4935
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.411441932180554e-05,
+      "loss": 0.9156,
+      "step": 4940
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.391447291751961e-05,
+      "loss": 0.9239,
+      "step": 4945
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.371455092977423e-05,
+      "loss": 0.8946,
+      "step": 4950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1875,
+      "learning_rate": 9.351465416070274e-05,
+      "loss": 0.9369,
+      "step": 4955
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.331478341233706e-05,
+      "loss": 0.9686,
+      "step": 4960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 9.311493948660488e-05,
+      "loss": 0.955,
+      "step": 4965
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1796875,
+      "learning_rate": 9.291512318532614e-05,
+      "loss": 0.9624,
+      "step": 4970
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.271533531021005e-05,
+      "loss": 0.9807,
+      "step": 4975
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.251557666285174e-05,
+      "loss": 0.9558,
+      "step": 4980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.231584804472898e-05,
+      "loss": 0.879,
+      "step": 4985
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.211615025719919e-05,
+      "loss": 0.9336,
+      "step": 4990
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.185546875,
+      "learning_rate": 9.1916484101496e-05,
+      "loss": 0.9256,
+      "step": 4995
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.17168503787262e-05,
+      "loss": 0.9757,
+      "step": 5000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 9.151724988986635e-05,
+      "loss": 0.9466,
+      "step": 5005
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.131768343575979e-05,
+      "loss": 0.9148,
+      "step": 5010
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 9.11181518171132e-05,
+      "loss": 1.008,
+      "step": 5015
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.091865583449351e-05,
+      "loss": 0.8901,
+      "step": 5020
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.071919628832476e-05,
+      "loss": 0.9584,
+      "step": 5025
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.051977397888464e-05,
+      "loss": 0.9067,
+      "step": 5030
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.032038970630163e-05,
+      "loss": 1.0218,
+      "step": 5035
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.01210442705514e-05,
+      "loss": 0.9347,
+      "step": 5040
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.193359375,
+      "learning_rate": 8.992173847145401e-05,
+      "loss": 0.9076,
+      "step": 5045
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.972247310867027e-05,
+      "loss": 0.9324,
+      "step": 5050
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.952324898169888e-05,
+      "loss": 0.9548,
+      "step": 5055
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 8.932406688987309e-05,
+      "loss": 0.9436,
+      "step": 5060
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 8.912492763235744e-05,
+      "loss": 0.981,
+      "step": 5065
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.189453125,
+      "learning_rate": 8.892583200814466e-05,
+      "loss": 0.9681,
+      "step": 5070
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 8.872678081605236e-05,
+      "loss": 0.9521,
+      "step": 5075
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 8.852777485471997e-05,
+      "loss": 0.9332,
+      "step": 5080
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.832881492260535e-05,
+      "loss": 0.9508,
+      "step": 5085
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.208984375,
+      "learning_rate": 8.81299018179817e-05,
+      "loss": 0.9635,
+      "step": 5090
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18359375,
+      "learning_rate": 8.793103633893437e-05,
+      "loss": 0.9189,
+      "step": 5095
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.773221928335759e-05,
+      "loss": 0.9554,
+      "step": 5100
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 8.753345144895136e-05,
+      "loss": 0.9469,
+      "step": 5105
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.73347336332181e-05,
+      "loss": 0.962,
+      "step": 5110
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.713606663345968e-05,
+      "loss": 1.0239,
+      "step": 5115
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 8.693745124677386e-05,
+      "loss": 0.9611,
+      "step": 5120
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.197265625,
+      "learning_rate": 8.673888827005164e-05,
+      "loss": 0.919,
+      "step": 5125
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1953125,
+      "learning_rate": 8.654037849997342e-05,
+      "loss": 0.9146,
+      "step": 5130
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.634192273300629e-05,
+      "loss": 0.9725,
+      "step": 5135
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.614352176540067e-05,
+      "loss": 0.923,
+      "step": 5140
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18359375,
+      "learning_rate": 8.594517639318705e-05,
+      "loss": 0.9446,
+      "step": 5145
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.57468874121729e-05,
+      "loss": 0.9708,
+      "step": 5150
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 8.554865561793942e-05,
+      "loss": 0.994,
+      "step": 5155
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 8.535048180583838e-05,
+      "loss": 0.9119,
+      "step": 5160
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 8.515236677098889e-05,
+      "loss": 0.9196,
+      "step": 5165
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.193359375,
+      "learning_rate": 8.495431130827422e-05,
+      "loss": 0.9464,
+      "step": 5170
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 8.475631621233869e-05,
+      "loss": 0.9656,
+      "step": 5175
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 8.455838227758432e-05,
+      "loss": 0.9434,
+      "step": 5180
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.436051029816783e-05,
+      "loss": 1.0145,
+      "step": 5185
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 8.416270106799726e-05,
+      "loss": 0.9732,
+      "step": 5190
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.171875,
+      "learning_rate": 8.396495538072902e-05,
+      "loss": 0.96,
+      "step": 5195
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.197265625,
+      "learning_rate": 8.376727402976447e-05,
+      "loss": 0.9475,
+      "step": 5200
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.189453125,
+      "learning_rate": 8.356965780824677e-05,
+      "loss": 0.9615,
+      "step": 5205
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1875,
+      "learning_rate": 8.3372107509058e-05,
+      "loss": 0.9265,
+      "step": 5210
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.317462392481546e-05,
+      "loss": 0.9579,
+      "step": 5215
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.297720784786906e-05,
+      "loss": 0.9899,
+      "step": 5220
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 8.277986007029756e-05,
+      "loss": 0.9242,
+      "step": 5225
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 8.2582581383906e-05,
+      "loss": 0.9504,
+      "step": 5230
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.189453125,
+      "learning_rate": 8.238537258022194e-05,
+      "loss": 0.9578,
+      "step": 5235
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 8.218823445049265e-05,
+      "loss": 0.9664,
+      "step": 5240
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18359375,
+      "learning_rate": 8.199116778568192e-05,
+      "loss": 0.9685,
+      "step": 5245
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 8.179417337646669e-05,
+      "loss": 0.9711,
+      "step": 5250
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.19140625,
+      "learning_rate": 8.159725201323408e-05,
+      "loss": 0.9146,
+      "step": 5255
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 8.140040448607804e-05,
+      "loss": 0.9694,
+      "step": 5260
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 8.12036315847964e-05,
+      "loss": 0.9633,
+      "step": 5265
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18359375,
+      "learning_rate": 8.100693409888748e-05,
+      "loss": 0.9796,
+      "step": 5270
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 8.081031281754695e-05,
+      "loss": 0.9705,
+      "step": 5275
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.061376852966495e-05,
+      "loss": 0.9391,
+      "step": 5280
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1796875,
+      "learning_rate": 8.041730202382245e-05,
+      "loss": 0.8999,
+      "step": 5285
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.02209140882886e-05,
+      "loss": 0.8907,
+      "step": 5290
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1796875,
+      "learning_rate": 8.002460551101702e-05,
+      "loss": 0.9771,
+      "step": 5295
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 7.982837707964321e-05,
+      "loss": 0.9144,
+      "step": 5300
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.963222958148085e-05,
+      "loss": 0.9878,
+      "step": 5305
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.943616380351913e-05,
+      "loss": 1.0033,
+      "step": 5310
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.92401805324192e-05,
+      "loss": 0.9517,
+      "step": 5315
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.185546875,
+      "learning_rate": 7.904428055451118e-05,
+      "loss": 1.0219,
+      "step": 5320
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.884846465579113e-05,
+      "loss": 0.9133,
+      "step": 5325
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 7.865273362191759e-05,
+      "loss": 0.9543,
+      "step": 5330
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.845708823820876e-05,
+      "loss": 0.9329,
+      "step": 5335
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.826152928963904e-05,
+      "loss": 0.9753,
+      "step": 5340
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.806605756083622e-05,
+      "loss": 0.9324,
+      "step": 5345
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.787067383607796e-05,
+      "loss": 1.0059,
+      "step": 5350
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.767537889928889e-05,
+      "loss": 0.939,
+      "step": 5355
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 7.748017353403748e-05,
+      "loss": 0.9298,
+      "step": 5360
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.72850585235327e-05,
+      "loss": 1.0243,
+      "step": 5365
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.70900346506211e-05,
+      "loss": 0.9523,
+      "step": 5370
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.201171875,
+      "learning_rate": 7.689510269778347e-05,
+      "loss": 0.9428,
+      "step": 5375
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.670026344713189e-05,
+      "loss": 0.9624,
+      "step": 5380
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.650551768040641e-05,
+      "loss": 0.9296,
+      "step": 5385
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.631086617897203e-05,
+      "loss": 0.9786,
+      "step": 5390
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.185546875,
+      "learning_rate": 7.611630972381557e-05,
+      "loss": 0.9551,
+      "step": 5395
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.193359375,
+      "learning_rate": 7.592184909554245e-05,
+      "loss": 0.9591,
+      "step": 5400
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.572748507437368e-05,
+      "loss": 0.9729,
+      "step": 5405
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.553321844014258e-05,
+      "loss": 0.9482,
+      "step": 5410
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 7.533904997229183e-05,
+      "loss": 0.9894,
+      "step": 5415
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.514498044987009e-05,
+      "loss": 0.9682,
+      "step": 5420
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.495101065152917e-05,
+      "loss": 0.9402,
+      "step": 5425
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.475714135552074e-05,
+      "loss": 0.9921,
+      "step": 5430
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 7.456337333969316e-05,
+      "loss": 0.9541,
+      "step": 5435
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.224609375,
+      "learning_rate": 7.43697073814885e-05,
+      "loss": 0.9567,
+      "step": 5440
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 7.417614425793932e-05,
+      "loss": 0.9909,
+      "step": 5445
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.201171875,
+      "learning_rate": 7.39826847456656e-05,
+      "loss": 1.007,
+      "step": 5450
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 7.37893296208716e-05,
+      "loss": 0.9779,
+      "step": 5455
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.359607965934274e-05,
+      "loss": 0.9178,
+      "step": 5460
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.20703125,
+      "learning_rate": 7.340293563644256e-05,
+      "loss": 0.9477,
+      "step": 5465
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.22265625,
+      "learning_rate": 7.32098983271094e-05,
+      "loss": 0.9965,
+      "step": 5470
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 7.301696850585366e-05,
+      "loss": 0.9738,
+      "step": 5475
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.282414694675426e-05,
+      "loss": 0.9809,
+      "step": 5480
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.263143442345592e-05,
+      "loss": 0.9294,
+      "step": 5485
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 7.243883170916574e-05,
+      "loss": 0.9068,
+      "step": 5490
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 7.22463395766504e-05,
+      "loss": 0.9642,
+      "step": 5495
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.185546875,
+      "learning_rate": 7.205395879823271e-05,
+      "loss": 0.9434,
+      "step": 5500
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.186169014578883e-05,
+      "loss": 0.9572,
+      "step": 5505
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.166953439074504e-05,
+      "loss": 0.9084,
+      "step": 5510
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 7.14774923040746e-05,
+      "loss": 0.9545,
+      "step": 5515
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.128556465629475e-05,
+      "loss": 0.9395,
+      "step": 5520
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 7.109375221746352e-05,
+      "loss": 0.9312,
+      "step": 5525
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.090205575717678e-05,
+      "loss": 0.943,
+      "step": 5530
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.185546875,
+      "learning_rate": 7.071047604456499e-05,
+      "loss": 0.9178,
+      "step": 5535
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.19921875,
+      "learning_rate": 7.051901384829021e-05,
+      "loss": 0.9456,
+      "step": 5540
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.032766993654303e-05,
+      "loss": 0.98,
+      "step": 5545
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.013644507703937e-05,
+      "loss": 0.9122,
+      "step": 5550
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.193359375,
+      "learning_rate": 6.994534003701765e-05,
+      "loss": 0.9898,
+      "step": 5555
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.975435558323532e-05,
+      "loss": 0.9486,
+      "step": 5560
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 6.956349248196627e-05,
+      "loss": 0.9626,
+      "step": 5565
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 6.937275149899725e-05,
+      "loss": 0.9448,
+      "step": 5570
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.918213339962518e-05,
+      "loss": 1.0067,
+      "step": 5575
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18359375,
+      "learning_rate": 6.899163894865395e-05,
+      "loss": 0.9252,
+      "step": 5580
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.880126891039124e-05,
+      "loss": 0.9502,
+      "step": 5585
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.86110240486457e-05,
+      "loss": 0.9001,
+      "step": 5590
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.842090512672358e-05,
+      "loss": 0.967,
+      "step": 5595
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.823091290742602e-05,
+      "loss": 0.9918,
+      "step": 5600
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.185546875,
+      "learning_rate": 6.804104815304566e-05,
+      "loss": 0.9647,
+      "step": 5605
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.785131162536374e-05,
+      "loss": 0.9447,
+      "step": 5610
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.766170408564705e-05,
+      "loss": 0.9132,
+      "step": 5615
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.747222629464484e-05,
+      "loss": 0.926,
+      "step": 5620
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18359375,
+      "learning_rate": 6.728287901258581e-05,
+      "loss": 0.9387,
+      "step": 5625
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.709366299917497e-05,
+      "loss": 0.9901,
+      "step": 5630
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.690457901359073e-05,
+      "loss": 0.9857,
+      "step": 5635
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.671562781448166e-05,
+      "loss": 0.9281,
+      "step": 5640
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 6.652681015996369e-05,
+      "loss": 0.9267,
+      "step": 5645
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18359375,
+      "learning_rate": 6.633812680761684e-05,
+      "loss": 0.9685,
+      "step": 5650
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.614957851448227e-05,
+      "loss": 0.895,
+      "step": 5655
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.59611660370594e-05,
+      "loss": 0.9486,
+      "step": 5660
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.19140625,
+      "learning_rate": 6.577289013130252e-05,
+      "loss": 0.9667,
+      "step": 5665
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.558475155261811e-05,
+      "loss": 0.9238,
+      "step": 5670
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.539675105586158e-05,
+      "loss": 0.9313,
+      "step": 5675
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.52088893953344e-05,
+      "loss": 0.9187,
+      "step": 5680
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.502116732478096e-05,
+      "loss": 0.9357,
+      "step": 5685
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.48335855973855e-05,
+      "loss": 0.9794,
+      "step": 5690
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.464614496576935e-05,
+      "loss": 0.9375,
+      "step": 5695
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 6.445884618198754e-05,
+      "loss": 0.9485,
+      "step": 5700
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 6.427168999752614e-05,
+      "loss": 1.0271,
+      "step": 5705
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.193359375,
+      "learning_rate": 6.408467716329894e-05,
+      "loss": 0.9713,
+      "step": 5710
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.19140625,
+      "learning_rate": 6.389780842964468e-05,
+      "loss": 0.93,
+      "step": 5715
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.197265625,
+      "learning_rate": 6.371108454632391e-05,
+      "loss": 0.9489,
+      "step": 5720
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.352450626251587e-05,
+      "loss": 0.9221,
+      "step": 5725
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.19140625,
+      "learning_rate": 6.33380743268159e-05,
+      "loss": 0.9411,
+      "step": 5730
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 6.315178948723186e-05,
+      "loss": 0.9548,
+      "step": 5735
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.29656524911817e-05,
+      "loss": 0.9121,
+      "step": 5740
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 6.277966408548992e-05,
+      "loss": 0.957,
+      "step": 5745
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.259382501638509e-05,
+      "loss": 0.9261,
+      "step": 5750
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.240813602949641e-05,
+      "loss": 0.9032,
+      "step": 5755
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.222259786985101e-05,
+      "loss": 0.9069,
+      "step": 5760
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.20372112818709e-05,
+      "loss": 0.9158,
+      "step": 5765
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.185197700936982e-05,
+      "loss": 0.926,
+      "step": 5770
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.166689579555053e-05,
+      "loss": 0.9858,
+      "step": 5775
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.185546875,
+      "learning_rate": 6.14819683830016e-05,
+      "loss": 0.9931,
+      "step": 5780
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 6.129719551369456e-05,
+      "loss": 0.9457,
+      "step": 5785
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.111257792898082e-05,
+      "loss": 0.9593,
+      "step": 5790
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 6.092811636958877e-05,
+      "loss": 0.991,
+      "step": 5795
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.0743811575620846e-05,
+      "loss": 1.0029,
+      "step": 5800
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 6.055966428655042e-05,
+      "loss": 0.9506,
+      "step": 5805
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.0375675241219e-05,
+      "loss": 0.9009,
+      "step": 5810
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.019184517783311e-05,
+      "loss": 0.8785,
+      "step": 5815
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 6.000817483396148e-05,
+      "loss": 0.9345,
+      "step": 5820
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 5.982466494653187e-05,
+      "loss": 0.909,
+      "step": 5825
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 5.96413162518285e-05,
+      "loss": 0.9035,
+      "step": 5830
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.185546875,
+      "learning_rate": 5.945812948548855e-05,
+      "loss": 0.9752,
+      "step": 5835
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 5.9275105382499694e-05,
+      "loss": 0.9315,
+      "step": 5840
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.19140625,
+      "learning_rate": 5.909224467719694e-05,
+      "loss": 0.9753,
+      "step": 5845
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.181640625,
+      "learning_rate": 5.890954810325966e-05,
+      "loss": 0.9314,
+      "step": 5850
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.8727016393708746e-05,
+      "loss": 0.9488,
+      "step": 5855
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1796875,
+      "learning_rate": 5.854465028090355e-05,
+      "loss": 0.9126,
+      "step": 5860
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 5.836245049653908e-05,
+      "loss": 0.933,
+      "step": 5865
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 5.8180417771643006e-05,
+      "loss": 0.9855,
+      "step": 5870
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18359375,
+      "learning_rate": 5.799855283657254e-05,
+      "loss": 0.9595,
+      "step": 5875
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.781685642101196e-05,
+      "loss": 0.9593,
+      "step": 5880
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 5.7635329253969195e-05,
+      "loss": 0.9815,
+      "step": 5885
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1875,
+      "learning_rate": 5.7453972063773184e-05,
+      "loss": 0.9813,
+      "step": 5890
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1796875,
+      "learning_rate": 5.727278557807085e-05,
+      "loss": 0.8825,
+      "step": 5895
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1796875,
+      "learning_rate": 5.7091770523824317e-05,
+      "loss": 0.9529,
+      "step": 5900
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 5.691092762730774e-05,
+      "loss": 0.9978,
+      "step": 5905
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.189453125,
+      "learning_rate": 5.673025761410462e-05,
+      "loss": 0.932,
+      "step": 5910
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.21484375,
+      "learning_rate": 5.654976120910478e-05,
+      "loss": 0.9421,
+      "step": 5915
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1953125,
+      "learning_rate": 5.636943913650147e-05,
+      "loss": 0.9461,
+      "step": 5920
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 5.618929211978857e-05,
+      "loss": 0.9497,
+      "step": 5925
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 5.60093208817575e-05,
+      "loss": 0.9658,
+      "step": 5930
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 5.582952614449445e-05,
+      "loss": 0.919,
+      "step": 5935
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1875,
+      "learning_rate": 5.564990862937744e-05,
+      "loss": 0.9407,
+      "step": 5940
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1875,
+      "learning_rate": 5.5470469057073404e-05,
+      "loss": 0.9422,
+      "step": 5945
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 5.5291208147535466e-05,
+      "loss": 0.9276,
+      "step": 5950
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18359375,
+      "learning_rate": 5.511212661999967e-05,
+      "loss": 0.9815,
+      "step": 5955
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.4933225192982586e-05,
+      "loss": 0.9177,
+      "step": 5960
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18359375,
+      "learning_rate": 5.4754504584278e-05,
+      "loss": 0.9294,
+      "step": 5965
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 5.457596551095441e-05,
+      "loss": 0.9828,
+      "step": 5970
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.19921875,
+      "learning_rate": 5.4397608689351656e-05,
+      "loss": 0.9501,
+      "step": 5975
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 5.421943483507863e-05,
+      "loss": 1.0297,
+      "step": 5980
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 5.404144466300998e-05,
+      "loss": 0.9216,
+      "step": 5985
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 5.3863638887283364e-05,
+      "loss": 0.9293,
+      "step": 5990
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 5.3686018221296665e-05,
+      "loss": 0.95,
+      "step": 5995
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.350858337770498e-05,
+      "loss": 0.8833,
+      "step": 6000
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.333133506841797e-05,
+      "loss": 0.9642,
+      "step": 6005
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 5.315427400459678e-05,
+      "loss": 0.9833,
+      "step": 6010
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.189453125,
+      "learning_rate": 5.297740089665129e-05,
+      "loss": 0.9212,
+      "step": 6015
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 5.280071645423726e-05,
+      "loss": 0.9469,
+      "step": 6020
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.262422138625349e-05,
+      "loss": 0.8959,
+      "step": 6025
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.244791640083906e-05,
+      "loss": 0.9473,
+      "step": 6030
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.193359375,
+      "learning_rate": 5.227180220537016e-05,
+      "loss": 0.9685,
+      "step": 6035
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.181640625,
+      "learning_rate": 5.2095879506457736e-05,
+      "loss": 0.9495,
+      "step": 6040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 5.192014900994423e-05,
+      "loss": 0.9516,
+      "step": 6045
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.181640625,
+      "learning_rate": 5.174461142090111e-05,
+      "loss": 0.9893,
+      "step": 6050
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.156926744362562e-05,
+      "loss": 0.9476,
+      "step": 6055
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.139411778163832e-05,
+      "loss": 0.9632,
+      "step": 6060
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 5.121916313768018e-05,
+      "loss": 0.953,
+      "step": 6065
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.19140625,
+      "learning_rate": 5.104440421370962e-05,
+      "loss": 0.969,
+      "step": 6070
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1875,
+      "learning_rate": 5.0869841710899815e-05,
+      "loss": 0.9233,
+      "step": 6075
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.0695476329635825e-05,
+      "loss": 0.9095,
+      "step": 6080
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1875,
+      "learning_rate": 5.052130876951192e-05,
+      "loss": 0.9356,
+      "step": 6085
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 5.034733972932855e-05,
+      "loss": 0.9314,
+      "step": 6090
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.017356990708969e-05,
+      "loss": 0.9438,
+      "step": 6095
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.000000000000002e-05,
+      "loss": 0.8893,
+      "step": 6100
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.185546875,
+      "learning_rate": 4.982663070446207e-05,
+      "loss": 0.9236,
+      "step": 6105
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.9653462716073594e-05,
+      "loss": 0.9417,
+      "step": 6110
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.185546875,
+      "learning_rate": 4.9480496729624515e-05,
+      "loss": 0.9293,
+      "step": 6115
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.177734375,
+      "learning_rate": 4.930773343909434e-05,
+      "loss": 0.9214,
+      "step": 6120
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.91351735376493e-05,
+      "loss": 0.918,
+      "step": 6125
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.181640625,
+      "learning_rate": 4.8962817717639555e-05,
+      "loss": 0.9821,
+      "step": 6130
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 4.879066667059659e-05,
+      "loss": 0.9732,
+      "step": 6135
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.185546875,
+      "learning_rate": 4.8618721087230014e-05,
+      "loss": 0.9026,
+      "step": 6140
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1953125,
+      "learning_rate": 4.844698165742536e-05,
+      "loss": 0.9547,
+      "step": 6145
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.177734375,
+      "learning_rate": 4.8275449070240854e-05,
+      "loss": 0.9191,
+      "step": 6150
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.810412401390487e-05,
+      "loss": 0.8428,
+      "step": 6155
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 4.793300717581308e-05,
+      "loss": 0.8993,
+      "step": 6160
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 4.7762099242525847e-05,
+      "loss": 0.9833,
+      "step": 6165
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 4.7591400899765234e-05,
+      "loss": 0.9107,
+      "step": 6170
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 4.7420912832412445e-05,
+      "loss": 0.9322,
+      "step": 6175
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 4.7250635724505e-05,
+      "loss": 0.9575,
+      "step": 6180
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 4.7080570259233935e-05,
+      "loss": 0.968,
+      "step": 6185
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19921875,
+      "learning_rate": 4.6910717118941286e-05,
+      "loss": 0.9626,
+      "step": 6190
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 4.674107698511704e-05,
+      "loss": 0.9467,
+      "step": 6195
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 4.6571650538396615e-05,
+      "loss": 0.8788,
+      "step": 6200
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1796875,
+      "learning_rate": 4.640243845855806e-05,
+      "loss": 0.9311,
+      "step": 6205
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 4.6233441424519295e-05,
+      "loss": 0.9713,
+      "step": 6210
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 4.606466011433557e-05,
+      "loss": 0.97,
+      "step": 6215
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 4.5896095205196356e-05,
+      "loss": 0.9904,
+      "step": 6220
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 4.5727747373423116e-05,
+      "loss": 0.9918,
+      "step": 6225
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 4.5559617294466176e-05,
+      "loss": 0.9482,
+      "step": 6230
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 4.539170564290237e-05,
+      "loss": 0.9262,
+      "step": 6235
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.189453125,
+      "learning_rate": 4.522401309243193e-05,
+      "loss": 0.9817,
+      "step": 6240
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 4.50565403158761e-05,
+      "loss": 0.8938,
+      "step": 6245
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 4.488928798517442e-05,
+      "loss": 0.9392,
+      "step": 6250
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 4.472225677138186e-05,
+      "loss": 0.9194,
+      "step": 6255
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.185546875,
+      "learning_rate": 4.455544734466624e-05,
+      "loss": 0.9411,
+      "step": 6260
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.185546875,
+      "learning_rate": 4.4388860374305474e-05,
+      "loss": 0.9871,
+      "step": 6265
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.275390625,
+      "learning_rate": 4.422249652868506e-05,
+      "loss": 0.9315,
+      "step": 6270
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 4.405635647529514e-05,
+      "loss": 0.9464,
+      "step": 6275
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 4.389044088072798e-05,
+      "loss": 0.9525,
+      "step": 6280
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.3724750410675287e-05,
+      "loss": 0.8944,
+      "step": 6285
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 4.355928572992547e-05,
+      "loss": 0.9523,
+      "step": 6290
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.181640625,
+      "learning_rate": 4.339404750236117e-05,
+      "loss": 0.934,
+      "step": 6295
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 4.322903639095619e-05,
+      "loss": 0.9886,
+      "step": 6300
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 4.306425305777333e-05,
+      "loss": 0.9596,
+      "step": 6305
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 4.289969816396132e-05,
+      "loss": 0.9387,
+      "step": 6310
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.2735372369752535e-05,
+      "loss": 0.9387,
+      "step": 6315
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.197265625,
+      "learning_rate": 4.2571276334459895e-05,
+      "loss": 0.9439,
+      "step": 6320
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.240741071647464e-05,
+      "loss": 0.9445,
+      "step": 6325
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 4.224377617326353e-05,
+      "loss": 0.9121,
+      "step": 6330
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.181640625,
+      "learning_rate": 4.208037336136612e-05,
+      "loss": 0.9763,
+      "step": 6335
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.185546875,
+      "learning_rate": 4.1917202936392265e-05,
+      "loss": 0.9658,
+      "step": 6340
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 4.175426555301937e-05,
+      "loss": 0.9722,
+      "step": 6345
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.15915618649899e-05,
+      "loss": 0.9671,
+      "step": 6350
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.189453125,
+      "learning_rate": 4.142909252510866e-05,
+      "loss": 0.9453,
+      "step": 6355
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.19140625,
+      "learning_rate": 4.126685818524013e-05,
+      "loss": 0.949,
+      "step": 6360
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.177734375,
+      "learning_rate": 4.1104859496305994e-05,
+      "loss": 0.9617,
+      "step": 6365
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1875,
+      "learning_rate": 4.094309710828236e-05,
+      "loss": 0.9842,
+      "step": 6370
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.208984375,
+      "learning_rate": 4.078157167019738e-05,
+      "loss": 0.9351,
+      "step": 6375
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.181640625,
+      "learning_rate": 4.0620283830128414e-05,
+      "loss": 0.9748,
+      "step": 6380
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1796875,
+      "learning_rate": 4.0459234235199515e-05,
+      "loss": 1.0083,
+      "step": 6385
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.181640625,
+      "learning_rate": 4.029842353157888e-05,
+      "loss": 0.967,
+      "step": 6390
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.177734375,
+      "learning_rate": 4.013785236447616e-05,
+      "loss": 0.924,
+      "step": 6395
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.9977521378140084e-05,
+      "loss": 0.9537,
+      "step": 6400
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.981743121585547e-05,
+      "loss": 0.9567,
+      "step": 6405
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 3.965758251994115e-05,
+      "loss": 0.9325,
+      "step": 6410
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.949797593174698e-05,
+      "loss": 0.9687,
+      "step": 6415
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.933861209165146e-05,
+      "loss": 0.9731,
+      "step": 6420
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.917949163905914e-05,
+      "loss": 0.95,
+      "step": 6425
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.181640625,
+      "learning_rate": 3.9020615212398016e-05,
+      "loss": 0.9823,
+      "step": 6430
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.886198344911707e-05,
+      "loss": 0.9272,
+      "step": 6435
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 3.8703596985683556e-05,
+      "loss": 0.8984,
+      "step": 6440
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.197265625,
+      "learning_rate": 3.8545456457580566e-05,
+      "loss": 0.9515,
+      "step": 6445
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 3.838756249930439e-05,
+      "loss": 0.9613,
+      "step": 6450
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.185546875,
+      "learning_rate": 3.822991574436213e-05,
+      "loss": 0.9774,
+      "step": 6455
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.807251682526902e-05,
+      "loss": 0.9645,
+      "step": 6460
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.791536637354576e-05,
+      "loss": 0.9584,
+      "step": 6465
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.775846501971636e-05,
+      "loss": 0.9622,
+      "step": 6470
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 3.760181339330526e-05,
+      "loss": 0.8987,
+      "step": 6475
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 3.7445412122835077e-05,
+      "loss": 0.9277,
+      "step": 6480
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 3.7289261835823695e-05,
+      "loss": 0.9665,
+      "step": 6485
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.713336315878224e-05,
+      "loss": 0.8984,
+      "step": 6490
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 3.6977716717212165e-05,
+      "loss": 0.9439,
+      "step": 6495
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18359375,
+      "learning_rate": 3.6822323135603054e-05,
+      "loss": 0.9428,
+      "step": 6500
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 3.6667183037429756e-05,
+      "loss": 0.9747,
+      "step": 6505
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1953125,
+      "learning_rate": 3.651229704515018e-05,
+      "loss": 0.9646,
+      "step": 6510
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.193359375,
+      "learning_rate": 3.635766578020279e-05,
+      "loss": 0.9733,
+      "step": 6515
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18359375,
+      "learning_rate": 3.6203289863003905e-05,
+      "loss": 0.916,
+      "step": 6520
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 3.604916991294537e-05,
+      "loss": 0.9703,
+      "step": 6525
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.5895306548392005e-05,
+      "loss": 0.9546,
+      "step": 6530
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.574170038667923e-05,
+      "loss": 0.9571,
+      "step": 6535
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 3.558835204411044e-05,
+      "loss": 0.9374,
+      "step": 6540
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18359375,
+      "learning_rate": 3.543526213595461e-05,
+      "loss": 0.9203,
+      "step": 6545
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 3.52824312764438e-05,
+      "loss": 0.9333,
+      "step": 6550
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.185546875,
+      "learning_rate": 3.512986007877072e-05,
+      "loss": 0.9077,
+      "step": 6555
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.185546875,
+      "learning_rate": 3.497754915508632e-05,
+      "loss": 0.9262,
+      "step": 6560
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.4825499116497176e-05,
+      "loss": 0.9597,
+      "step": 6565
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.181640625,
+      "learning_rate": 3.467371057306318e-05,
+      "loss": 0.9578,
+      "step": 6570
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.452218413379504e-05,
+      "loss": 0.941,
+      "step": 6575
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.189453125,
+      "learning_rate": 3.437092040665183e-05,
+      "loss": 0.9687,
+      "step": 6580
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 3.421991999853857e-05,
+      "loss": 0.9264,
+      "step": 6585
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.406918351530376e-05,
+      "loss": 0.9096,
+      "step": 6590
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.3918711561737046e-05,
+      "loss": 0.9541,
+      "step": 6595
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.3768504741566664e-05,
+      "loss": 0.9201,
+      "step": 6600
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 3.361856365745705e-05,
+      "loss": 0.9519,
+      "step": 6605
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.189453125,
+      "learning_rate": 3.346888891100649e-05,
+      "loss": 0.9442,
+      "step": 6610
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 3.331948110274462e-05,
+      "loss": 0.9115,
+      "step": 6615
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.185546875,
+      "learning_rate": 3.3170340832130134e-05,
+      "loss": 0.9761,
+      "step": 6620
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.185546875,
+      "learning_rate": 3.302146869754823e-05,
+      "loss": 0.9547,
+      "step": 6625
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.189453125,
+      "learning_rate": 3.287286529630832e-05,
+      "loss": 0.9867,
+      "step": 6630
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.272453122464152e-05,
+      "loss": 0.9566,
+      "step": 6635
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.2576467077698493e-05,
+      "loss": 0.938,
+      "step": 6640
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 3.242867344954674e-05,
+      "loss": 0.9476,
+      "step": 6645
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.228115093316848e-05,
+      "loss": 0.9288,
+      "step": 6650
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 3.213390012045811e-05,
+      "loss": 0.9603,
+      "step": 6655
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1875,
+      "learning_rate": 3.198692160221987e-05,
+      "loss": 0.911,
+      "step": 6660
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18359375,
+      "learning_rate": 3.184021596816563e-05,
+      "loss": 1.0001,
+      "step": 6665
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.169378380691218e-05,
+      "loss": 0.9743,
+      "step": 6670
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.1547625705979265e-05,
+      "loss": 0.9113,
+      "step": 6675
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 3.140174225178692e-05,
+      "loss": 0.9722,
+      "step": 6680
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1875,
+      "learning_rate": 3.1256134029653275e-05,
+      "loss": 0.9743,
+      "step": 6685
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.111080162379215e-05,
+      "loss": 0.8996,
+      "step": 6690
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.096574561731072e-05,
+      "loss": 0.925,
+      "step": 6695
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.082096659220722e-05,
+      "loss": 0.9859,
+      "step": 6700
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18359375,
+      "learning_rate": 3.0676465129368556e-05,
+      "loss": 0.984,
+      "step": 6705
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.177734375,
+      "learning_rate": 3.0532241808567966e-05,
+      "loss": 0.9618,
+      "step": 6710
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 3.0388297208462703e-05,
+      "loss": 0.9563,
+      "step": 6715
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.0244631906591825e-05,
+      "loss": 0.9311,
+      "step": 6720
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 3.01012464793737e-05,
+      "loss": 0.9555,
+      "step": 6725
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.9958141502103722e-05,
+      "loss": 0.9164,
+      "step": 6730
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 2.9815317548952192e-05,
+      "loss": 0.9166,
+      "step": 6735
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.9672775192961756e-05,
+      "loss": 0.9433,
+      "step": 6740
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.9530515006045368e-05,
+      "loss": 0.9627,
+      "step": 6745
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19140625,
+      "learning_rate": 2.938853755898364e-05,
+      "loss": 0.9391,
+      "step": 6750
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.9246843421422998e-05,
+      "loss": 0.9186,
+      "step": 6755
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.910543316187301e-05,
+      "loss": 0.9401,
+      "step": 6760
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.896430734770431e-05,
+      "loss": 0.9041,
+      "step": 6765
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.882346654514627e-05,
+      "loss": 0.9367,
+      "step": 6770
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.8682911319284712e-05,
+      "loss": 0.8709,
+      "step": 6775
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.8542642234059725e-05,
+      "loss": 0.9867,
+      "step": 6780
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.8402659852263257e-05,
+      "loss": 0.9308,
+      "step": 6785
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.826296473553697e-05,
+      "loss": 0.9116,
+      "step": 6790
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.812355744436993e-05,
+      "loss": 0.9537,
+      "step": 6795
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.7984438538096392e-05,
+      "loss": 0.9502,
+      "step": 6800
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 2.784560857489358e-05,
+      "loss": 0.9706,
+      "step": 6805
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.7707068111779377e-05,
+      "loss": 0.9433,
+      "step": 6810
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.7568817704610116e-05,
+      "loss": 0.9487,
+      "step": 6815
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 2.7430857908078345e-05,
+      "loss": 0.9408,
+      "step": 6820
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.7293189275710706e-05,
+      "loss": 0.9269,
+      "step": 6825
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.7155812359865517e-05,
+      "loss": 0.9716,
+      "step": 6830
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.7018727711730706e-05,
+      "loss": 0.906,
+      "step": 6835
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.6881935881321563e-05,
+      "loss": 0.8983,
+      "step": 6840
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.6745437417478502e-05,
+      "loss": 0.9094,
+      "step": 6845
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.6609232867864896e-05,
+      "loss": 1.0059,
+      "step": 6850
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.6473322778964847e-05,
+      "loss": 1.0066,
+      "step": 6855
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 2.6337707696081094e-05,
+      "loss": 0.9076,
+      "step": 6860
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.6202388163332637e-05,
+      "loss": 0.9695,
+      "step": 6865
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.606736472365272e-05,
+      "loss": 0.9594,
+      "step": 6870
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 2.5932637918786563e-05,
+      "loss": 0.944,
+      "step": 6875
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.5798208289289204e-05,
+      "loss": 0.9092,
+      "step": 6880
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.566407637452345e-05,
+      "loss": 0.9355,
+      "step": 6885
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.5530242712657492e-05,
+      "loss": 0.9607,
+      "step": 6890
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.5396707840662903e-05,
+      "loss": 0.9421,
+      "step": 6895
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.526347229431242e-05,
+      "loss": 0.9299,
+      "step": 6900
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.513053660817788e-05,
+      "loss": 1.0149,
+      "step": 6905
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.499790131562797e-05,
+      "loss": 0.9583,
+      "step": 6910
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 2.4865566948826048e-05,
+      "loss": 0.95,
+      "step": 6915
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.4733534038728257e-05,
+      "loss": 0.9341,
+      "step": 6920
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 2.4601803115081068e-05,
+      "loss": 0.9515,
+      "step": 6925
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19140625,
+      "learning_rate": 2.4470374706419485e-05,
+      "loss": 0.9775,
+      "step": 6930
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.4339249340064507e-05,
+      "loss": 0.9294,
+      "step": 6935
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.4208427542121504e-05,
+      "loss": 0.9606,
+      "step": 6940
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.407790983747773e-05,
+      "loss": 0.9173,
+      "step": 6945
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.394769674980035e-05,
+      "loss": 0.9214,
+      "step": 6950
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19140625,
+      "learning_rate": 2.3817788801534367e-05,
+      "loss": 0.9737,
+      "step": 6955
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 2.3688186513900455e-05,
+      "loss": 0.9523,
+      "step": 6960
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19140625,
+      "learning_rate": 2.3558890406892986e-05,
+      "loss": 0.9941,
+      "step": 6965
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 2.34299009992778e-05,
+      "loss": 0.9817,
+      "step": 6970
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.3301218808590176e-05,
+      "loss": 0.9577,
+      "step": 6975
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.317284435113278e-05,
+      "loss": 0.9816,
+      "step": 6980
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.3044778141973655e-05,
+      "loss": 0.9451,
+      "step": 6985
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.2917020694944023e-05,
+      "loss": 0.9186,
+      "step": 6990
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.278957252263617e-05,
+      "loss": 0.9286,
+      "step": 6995
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.2662434136401722e-05,
+      "loss": 0.9249,
+      "step": 7000
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.2535606046349177e-05,
+      "loss": 0.8731,
+      "step": 7005
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.2409088761342235e-05,
+      "loss": 0.9873,
+      "step": 7010
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.228288278899735e-05,
+      "loss": 0.9734,
+      "step": 7015
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.215698863568213e-05,
+      "loss": 0.9737,
+      "step": 7020
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.203140680651298e-05,
+      "loss": 0.9152,
+      "step": 7025
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.1906137805353212e-05,
+      "loss": 0.9442,
+      "step": 7030
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 2.1781182134810997e-05,
+      "loss": 0.9409,
+      "step": 7035
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.1656540296237316e-05,
+      "loss": 0.9027,
+      "step": 7040
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.1532212789724094e-05,
+      "loss": 0.9258,
+      "step": 7045
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.1408200114101985e-05,
+      "loss": 0.9651,
+      "step": 7050
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.1284502766938475e-05,
+      "loss": 0.9362,
+      "step": 7055
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.116112124453592e-05,
+      "loss": 0.9467,
+      "step": 7060
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.1038056041929456e-05,
+      "loss": 0.9919,
+      "step": 7065
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 2.0915307652885164e-05,
+      "loss": 0.926,
+      "step": 7070
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1953125,
+      "learning_rate": 2.0792876569897912e-05,
+      "loss": 0.948,
+      "step": 7075
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.067076328418949e-05,
+      "loss": 0.984,
+      "step": 7080
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.0548968285706593e-05,
+      "loss": 0.9223,
+      "step": 7085
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.0427492063118935e-05,
+      "loss": 0.9921,
+      "step": 7090
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.0306335103817208e-05,
+      "loss": 0.9457,
+      "step": 7095
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.018549789391102e-05,
+      "loss": 0.9918,
+      "step": 7100
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.006498091822726e-05,
+      "loss": 0.9386,
+      "step": 7105
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.994478466030787e-05,
+      "loss": 0.9116,
+      "step": 7110
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.982490960240798e-05,
+      "loss": 0.9345,
+      "step": 7115
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 1.970535622549401e-05,
+      "loss": 0.9185,
+      "step": 7120
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.185546875,
+      "learning_rate": 1.9586125009241774e-05,
+      "loss": 0.9691,
+      "step": 7125
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.201171875,
+      "learning_rate": 1.946721643203443e-05,
+      "loss": 0.9646,
+      "step": 7130
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18359375,
+      "learning_rate": 1.934863097096067e-05,
+      "loss": 0.9284,
+      "step": 7135
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.193359375,
+      "learning_rate": 1.923036910181275e-05,
+      "loss": 0.9732,
+      "step": 7140
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.9112431299084598e-05,
+      "loss": 0.9394,
+      "step": 7145
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 1.8994818035969975e-05,
+      "loss": 0.9554,
+      "step": 7150
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.8877529784360437e-05,
+      "loss": 0.9861,
+      "step": 7155
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18359375,
+      "learning_rate": 1.8760567014843545e-05,
+      "loss": 0.8772,
+      "step": 7160
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.864393019670092e-05,
+      "loss": 0.9258,
+      "step": 7165
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 1.8527619797906494e-05,
+      "loss": 0.8904,
+      "step": 7170
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.8411636285124457e-05,
+      "loss": 0.9438,
+      "step": 7175
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.8295980123707357e-05,
+      "loss": 0.9734,
+      "step": 7180
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.8180651777694535e-05,
+      "loss": 0.9641,
+      "step": 7185
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.8065651709809905e-05,
+      "loss": 0.953,
+      "step": 7190
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.205078125,
+      "learning_rate": 1.795098038146038e-05,
+      "loss": 0.9441,
+      "step": 7195
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 1.783663825273372e-05,
+      "loss": 0.9674,
+      "step": 7200
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.772262578239704e-05,
+      "loss": 0.9597,
+      "step": 7205
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.7608943427894686e-05,
+      "loss": 0.9111,
+      "step": 7210
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18359375,
+      "learning_rate": 1.7495591645346533e-05,
+      "loss": 0.9676,
+      "step": 7215
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18359375,
+      "learning_rate": 1.7382570889546124e-05,
+      "loss": 0.946,
+      "step": 7220
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.7269881613958805e-05,
+      "loss": 0.9481,
+      "step": 7225
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18359375,
+      "learning_rate": 1.7157524270720036e-05,
+      "loss": 0.9301,
+      "step": 7230
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 1.7045499310633428e-05,
+      "loss": 0.9599,
+      "step": 7235
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.6933807183168994e-05,
+      "loss": 0.9333,
+      "step": 7240
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.682244833646135e-05,
+      "loss": 0.9387,
+      "step": 7245
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.193359375,
+      "learning_rate": 1.6711423217307885e-05,
+      "loss": 0.9474,
+      "step": 7250
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.6600732271167098e-05,
+      "loss": 0.895,
+      "step": 7255
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.64903759421566e-05,
+      "loss": 0.944,
+      "step": 7260
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.638035467305148e-05,
+      "loss": 0.9113,
+      "step": 7265
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.627066890528247e-05,
+      "loss": 0.9302,
+      "step": 7270
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.6161319078934278e-05,
+      "loss": 0.9291,
+      "step": 7275
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.6052305632743592e-05,
+      "loss": 0.9693,
+      "step": 7280
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.594362900409756e-05,
+      "loss": 0.9234,
+      "step": 7285
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.583528962903197e-05,
+      "loss": 0.9372,
+      "step": 7290
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.5727287942229387e-05,
+      "loss": 0.9087,
+      "step": 7295
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.189453125,
+      "learning_rate": 1.5619624377017537e-05,
+      "loss": 0.957,
+      "step": 7300
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 1.55122993653675e-05,
+      "loss": 0.9158,
+      "step": 7305
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.540531333789207e-05,
+      "loss": 0.8897,
+      "step": 7310
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.5298666723843867e-05,
+      "loss": 0.8997,
+      "step": 7315
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.5192359951113755e-05,
+      "loss": 0.9355,
+      "step": 7320
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.5086393446229063e-05,
+      "loss": 0.9448,
+      "step": 7325
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.4980767634351877e-05,
+      "loss": 0.9978,
+      "step": 7330
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.4875482939277396e-05,
+      "loss": 1.0184,
+      "step": 7335
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.4770539783432113e-05,
+      "loss": 0.9291,
+      "step": 7340
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.189453125,
+      "learning_rate": 1.466593858787223e-05,
+      "loss": 0.9722,
+      "step": 7345
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.4561679772281877e-05,
+      "loss": 0.9643,
+      "step": 7350
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.4457763754971553e-05,
+      "loss": 0.9647,
+      "step": 7355
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.4354190952876334e-05,
+      "loss": 0.9138,
+      "step": 7360
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.425096178155415e-05,
+      "loss": 0.9588,
+      "step": 7365
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.4148076655184373e-05,
+      "loss": 0.9397,
+      "step": 7370
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.404553598656585e-05,
+      "loss": 0.9406,
+      "step": 7375
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18359375,
+      "learning_rate": 1.3943340187115494e-05,
+      "loss": 0.9484,
+      "step": 7380
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.3841489666866369e-05,
+      "loss": 1.01,
+      "step": 7385
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.185546875,
+      "learning_rate": 1.373998483446638e-05,
+      "loss": 0.933,
+      "step": 7390
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 1.3638826097176328e-05,
+      "loss": 0.9798,
+      "step": 7395
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.3538013860868436e-05,
+      "loss": 0.941,
+      "step": 7400
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 1.3437548530024691e-05,
+      "loss": 0.9279,
+      "step": 7405
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.185546875,
+      "learning_rate": 1.3337430507735205e-05,
+      "loss": 0.9817,
+      "step": 7410
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.3237660195696633e-05,
+      "loss": 0.9868,
+      "step": 7415
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.313823799421051e-05,
+      "loss": 0.9241,
+      "step": 7420
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.3039164302181683e-05,
+      "loss": 0.9881,
+      "step": 7425
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.2940439517116676e-05,
+      "loss": 0.9063,
+      "step": 7430
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.2842064035122125e-05,
+      "loss": 0.9329,
+      "step": 7435
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.2744038250903267e-05,
+      "loss": 0.9975,
+      "step": 7440
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.264636255776208e-05,
+      "loss": 0.9458,
+      "step": 7445
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.2549037347596115e-05,
+      "loss": 0.8736,
+      "step": 7450
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 1.245206301089652e-05,
+      "loss": 0.9527,
+      "step": 7455
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.2355439936746827e-05,
+      "loss": 0.9283,
+      "step": 7460
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.2259168512821062e-05,
+      "loss": 0.9606,
+      "step": 7465
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.2163249125382426e-05,
+      "loss": 0.982,
+      "step": 7470
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.206768215928169e-05,
+      "loss": 0.931,
+      "step": 7475
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.1972467997955595e-05,
+      "loss": 0.9234,
+      "step": 7480
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.187760702342534e-05,
+      "loss": 0.9514,
+      "step": 7485
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.1783099616295056e-05,
+      "loss": 0.9941,
+      "step": 7490
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.1688946155750347e-05,
+      "loss": 0.9438,
+      "step": 7495
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.1595147019556607e-05,
+      "loss": 0.9474,
+      "step": 7500
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18359375,
+      "learning_rate": 1.1501702584057661e-05,
+      "loss": 0.9334,
+      "step": 7505
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.140861322417417e-05,
+      "loss": 1.0021,
+      "step": 7510
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.1315879313402123e-05,
+      "loss": 0.9094,
+      "step": 7515
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.1223501223811451e-05,
+      "loss": 0.934,
+      "step": 7520
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.1131479326044348e-05,
+      "loss": 0.9176,
+      "step": 7525
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.1039813989313951e-05,
+      "loss": 0.9808,
+      "step": 7530
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.0948505581402735e-05,
+      "loss": 0.9731,
+      "step": 7535
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.085755446866119e-05,
+      "loss": 0.9826,
+      "step": 7540
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.076696101600615e-05,
+      "loss": 0.9289,
+      "step": 7545
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.0676725586919457e-05,
+      "loss": 0.9442,
+      "step": 7550
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.0586848543446537e-05,
+      "loss": 0.9553,
+      "step": 7555
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.0497330246194848e-05,
+      "loss": 0.9037,
+      "step": 7560
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.185546875,
+      "learning_rate": 1.0408171054332483e-05,
+      "loss": 0.9526,
+      "step": 7565
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.031937132558668e-05,
+      "loss": 0.967,
+      "step": 7570
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.185546875,
+      "learning_rate": 1.0230931416242518e-05,
+      "loss": 0.9567,
+      "step": 7575
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.014285168114133e-05,
+      "loss": 0.9413,
+      "step": 7580
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 1.0055132473679363e-05,
+      "loss": 0.9979,
+      "step": 7585
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.96777414580633e-06,
+      "loss": 0.913,
+      "step": 7590
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.880777048024014e-06,
+      "loss": 0.9232,
+      "step": 7595
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.794141529384915e-06,
+      "loss": 0.9419,
+      "step": 7600
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.707867937490722e-06,
+      "loss": 0.9402,
+      "step": 7605
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 9.621956618491024e-06,
+      "loss": 0.958,
+      "step": 7610
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 9.536407917081869e-06,
+      "loss": 0.9503,
+      "step": 7615
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1953125,
+      "learning_rate": 9.451222176504414e-06,
+      "loss": 0.9652,
+      "step": 7620
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.366399738543574e-06,
+      "loss": 0.9568,
+      "step": 7625
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 9.281940943526491e-06,
+      "loss": 0.9612,
+      "step": 7630
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.185546875,
+      "learning_rate": 9.197846130321419e-06,
+      "loss": 0.9385,
+      "step": 7635
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 9.114115636336152e-06,
+      "loss": 0.9141,
+      "step": 7640
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 9.030749797516825e-06,
+      "loss": 0.9535,
+      "step": 7645
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.197265625,
+      "learning_rate": 8.947748948346357e-06,
+      "loss": 0.8915,
+      "step": 7650
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.865113421843407e-06,
+      "loss": 0.9905,
+      "step": 7655
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.782843549560771e-06,
+      "loss": 0.9076,
+      "step": 7660
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 8.700939661584184e-06,
+      "loss": 0.996,
+      "step": 7665
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.61940208653097e-06,
+      "loss": 0.9741,
+      "step": 7670
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.538231151548693e-06,
+      "loss": 0.9005,
+      "step": 7675
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 8.457427182313937e-06,
+      "loss": 0.9117,
+      "step": 7680
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.37699050303089e-06,
+      "loss": 0.9099,
+      "step": 7685
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18359375,
+      "learning_rate": 8.296921436430071e-06,
+      "loss": 0.9472,
+      "step": 7690
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.171875,
+      "learning_rate": 8.217220303767092e-06,
+      "loss": 0.9763,
+      "step": 7695
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18359375,
+      "learning_rate": 8.137887424821277e-06,
+      "loss": 0.9201,
+      "step": 7700
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.17578125,
+      "learning_rate": 8.058923117894534e-06,
+      "loss": 0.9278,
+      "step": 7705
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.980327699809832e-06,
+      "loss": 0.9282,
+      "step": 7710
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 7.902101485910185e-06,
+      "loss": 0.9873,
+      "step": 7715
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 7.824244790057223e-06,
+      "loss": 0.9301,
+      "step": 7720
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19921875,
+      "learning_rate": 7.746757924630033e-06,
+      "loss": 0.9648,
+      "step": 7725
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.66964120052377e-06,
+      "loss": 0.9401,
+      "step": 7730
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.592894927148553e-06,
+      "loss": 0.9304,
+      "step": 7735
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.516519412428203e-06,
+      "loss": 0.9685,
+      "step": 7740
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 7.440514962798905e-06,
+      "loss": 0.958,
+      "step": 7745
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.3648818832080745e-06,
+      "loss": 0.9361,
+      "step": 7750
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.289620477113068e-06,
+      "loss": 0.9719,
+      "step": 7755
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.214731046480094e-06,
+      "loss": 0.9499,
+      "step": 7760
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.140213891782821e-06,
+      "loss": 0.9577,
+      "step": 7765
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.066069312001289e-06,
+      "loss": 0.9519,
+      "step": 7770
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 6.992297604620679e-06,
+      "loss": 0.9976,
+      "step": 7775
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19140625,
+      "learning_rate": 6.918899065630113e-06,
+      "loss": 0.9751,
+      "step": 7780
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.845873989521523e-06,
+      "loss": 0.9353,
+      "step": 7785
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 6.773222669288359e-06,
+      "loss": 0.9473,
+      "step": 7790
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 6.700945396424518e-06,
+      "loss": 0.9311,
+      "step": 7795
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.629042460923096e-06,
+      "loss": 0.9307,
+      "step": 7800
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.205078125,
+      "learning_rate": 6.5575141512753015e-06,
+      "loss": 0.9714,
+      "step": 7805
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 6.486360754469234e-06,
+      "loss": 0.9419,
+      "step": 7810
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 6.415582555988742e-06,
+      "loss": 0.9297,
+      "step": 7815
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.345179839812343e-06,
+      "loss": 0.9209,
+      "step": 7820
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.275152888411984e-06,
+      "loss": 0.9888,
+      "step": 7825
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18359375,
+      "learning_rate": 6.205501982751971e-06,
+      "loss": 0.9304,
+      "step": 7830
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.136227402287809e-06,
+      "loss": 0.9514,
+      "step": 7835
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.067329424965162e-06,
+      "loss": 0.9081,
+      "step": 7840
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.998808327218619e-06,
+      "loss": 0.9615,
+      "step": 7845
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.189453125,
+      "learning_rate": 5.930664383970641e-06,
+      "loss": 0.9369,
+      "step": 7850
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 5.862897868630468e-06,
+      "loss": 0.9581,
+      "step": 7855
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 5.795509053093029e-06,
+      "loss": 0.9426,
+      "step": 7860
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.72849820773782e-06,
+      "loss": 0.8658,
+      "step": 7865
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 5.6618656014278406e-06,
+      "loss": 0.9303,
+      "step": 7870
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1875,
+      "learning_rate": 5.595611501508491e-06,
+      "loss": 0.9989,
+      "step": 7875
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.5297361738065325e-06,
+      "loss": 0.9053,
+      "step": 7880
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.181640625,
+      "learning_rate": 5.464239882628985e-06,
+      "loss": 0.9741,
+      "step": 7885
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 5.399122890762143e-06,
+      "loss": 0.9414,
+      "step": 7890
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1796875,
+      "learning_rate": 5.334385459470359e-06,
+      "loss": 0.8792,
+      "step": 7895
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.193359375,
+      "learning_rate": 5.270027848495207e-06,
+      "loss": 0.9348,
+      "step": 7900
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.2060503160542785e-06,
+      "loss": 0.9175,
+      "step": 7905
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.1424531188402405e-06,
+      "loss": 0.9214,
+      "step": 7910
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.079236512019703e-06,
+      "loss": 1.0005,
+      "step": 7915
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1875,
+      "learning_rate": 5.016400749232297e-06,
+      "loss": 0.9413,
+      "step": 7920
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 4.953946082589655e-06,
+      "loss": 0.9344,
+      "step": 7925
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 4.89187276267431e-06,
+      "loss": 0.9874,
+      "step": 7930
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.185546875,
+      "learning_rate": 4.83018103853875e-06,
+      "loss": 1.0028,
+      "step": 7935
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.7688711577044354e-06,
+      "loss": 0.9403,
+      "step": 7940
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1875,
+      "learning_rate": 4.707943366160794e-06,
+      "loss": 0.9021,
+      "step": 7945
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 4.647397908364182e-06,
+      "loss": 0.9949,
+      "step": 7950
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 4.587235027236958e-06,
+      "loss": 0.9576,
+      "step": 7955
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 4.5274549641665105e-06,
+      "loss": 0.937,
+      "step": 7960
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 4.468057959004246e-06,
+      "loss": 0.8949,
+      "step": 7965
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 4.40904425006472e-06,
+      "loss": 0.931,
+      "step": 7970
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 4.3504140741245095e-06,
+      "loss": 0.9385,
+      "step": 7975
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.2921676664214535e-06,
+      "loss": 0.9761,
+      "step": 7980
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1796875,
+      "learning_rate": 4.234305260653604e-06,
+      "loss": 0.9486,
+      "step": 7985
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.181640625,
+      "learning_rate": 4.176827088978297e-06,
+      "loss": 0.9197,
+      "step": 7990
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 4.119733382011215e-06,
+      "loss": 0.928,
+      "step": 7995
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1875,
+      "learning_rate": 4.0630243688255185e-06,
+      "loss": 0.9408,
+      "step": 8000
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 4.00670027695087e-06,
+      "loss": 0.9565,
+      "step": 8005
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18359375,
+      "learning_rate": 3.950761332372543e-06,
+      "loss": 0.9043,
+      "step": 8010
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.185546875,
+      "learning_rate": 3.8952077595305055e-06,
+      "loss": 0.9867,
+      "step": 8015
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18359375,
+      "learning_rate": 3.8400397813185054e-06,
+      "loss": 0.967,
+      "step": 8020
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.177734375,
+      "learning_rate": 3.785257619083249e-06,
+      "loss": 0.9493,
+      "step": 8025
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.7308614926234165e-06,
+      "loss": 0.9425,
+      "step": 8030
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.181640625,
+      "learning_rate": 3.676851620188826e-06,
+      "loss": 0.9513,
+      "step": 8035
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.6232282184795794e-06,
+      "loss": 0.9305,
+      "step": 8040
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 3.56999150264512e-06,
+      "loss": 0.8898,
+      "step": 8045
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.185546875,
+      "learning_rate": 3.517141686283498e-06,
+      "loss": 0.9281,
+      "step": 8050
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.19921875,
+      "learning_rate": 3.4646789814403503e-06,
+      "loss": 0.9702,
+      "step": 8055
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.412603598608188e-06,
+      "loss": 0.9329,
+      "step": 8060
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 3.360915746725479e-06,
+      "loss": 0.8976,
+      "step": 8065
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 3.3096156331758e-06,
+      "loss": 0.9532,
+      "step": 8070
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1875,
+      "learning_rate": 3.258703463787105e-06,
+      "loss": 0.9792,
+      "step": 8075
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.2081794428307278e-06,
+      "loss": 0.9745,
+      "step": 8080
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 3.1580437730207578e-06,
+      "loss": 0.944,
+      "step": 8085
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.1082966555130654e-06,
+      "loss": 0.9542,
+      "step": 8090
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.058938289904578e-06,
+      "loss": 0.9291,
+      "step": 8095
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.181640625,
+      "learning_rate": 3.0099688742324715e-06,
+      "loss": 0.9879,
+      "step": 8100
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19140625,
+      "learning_rate": 2.9613886049733365e-06,
+      "loss": 0.984,
+      "step": 8105
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.913197677042456e-06,
+      "loss": 0.9277,
+      "step": 8110
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.865396283792965e-06,
+      "loss": 0.9794,
+      "step": 8115
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.8179846170150903e-06,
+      "loss": 0.9043,
+      "step": 8120
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.7709628669353895e-06,
+      "loss": 0.9177,
+      "step": 8125
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.7243312222159924e-06,
+      "loss": 0.9242,
+      "step": 8130
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.6780898699538483e-06,
+      "loss": 1.0109,
+      "step": 8135
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 2.6322389956799143e-06,
+      "loss": 0.9075,
+      "step": 8140
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.5867787833585124e-06,
+      "loss": 0.9205,
+      "step": 8145
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 2.541709415386495e-06,
+      "loss": 0.9378,
+      "step": 8150
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.4970310725926148e-06,
+      "loss": 0.9403,
+      "step": 8155
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.4527439342366785e-06,
+      "loss": 0.9588,
+      "step": 8160
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 2.4088481780089267e-06,
+      "loss": 0.952,
+      "step": 8165
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.3653439800292556e-06,
+      "loss": 0.9714,
+      "step": 8170
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.3222315148465956e-06,
+      "loss": 0.9316,
+      "step": 8175
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.2795109554381024e-06,
+      "loss": 0.9406,
+      "step": 8180
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.237182473208499e-06,
+      "loss": 0.9059,
+      "step": 8185
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.195246237989479e-06,
+      "loss": 1.0483,
+      "step": 8190
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.1537024180388834e-06,
+      "loss": 0.9307,
+      "step": 8195
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.1125511800401234e-06,
+      "loss": 0.9557,
+      "step": 8200
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.0717926891014706e-06,
+      "loss": 0.8885,
+      "step": 8205
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.0314271087554126e-06,
+      "loss": 0.9263,
+      "step": 8210
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.991454600957976e-06,
+      "loss": 0.9167,
+      "step": 8215
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 1.951875326088104e-06,
+      "loss": 0.9438,
+      "step": 8220
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.9126894429469912e-06,
+      "loss": 0.8987,
+      "step": 8225
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.8738971087574275e-06,
+      "loss": 0.9953,
+      "step": 8230
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.193359375,
+      "learning_rate": 1.8354984791632778e-06,
+      "loss": 0.9986,
+      "step": 8235
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.797493708228659e-06,
+      "loss": 0.9941,
+      "step": 8240
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.759882948437519e-06,
+      "loss": 0.9877,
+      "step": 8245
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.7226663506929142e-06,
+      "loss": 0.941,
+      "step": 8250
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.685844064316433e-06,
+      "loss": 0.9541,
+      "step": 8255
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 1.6494162370475852e-06,
+      "loss": 0.9602,
+      "step": 8260
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.6133830150432349e-06,
+      "loss": 0.9152,
+      "step": 8265
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.185546875,
+      "learning_rate": 1.5777445428770022e-06,
+      "loss": 0.9283,
+      "step": 8270
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.5425009635386622e-06,
+      "loss": 0.9179,
+      "step": 8275
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.5076524184336027e-06,
+      "loss": 0.9523,
+      "step": 8280
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.4731990473822454e-06,
+      "loss": 0.9898,
+      "step": 8285
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.189453125,
+      "learning_rate": 1.4391409886194474e-06,
+      "loss": 0.9237,
+      "step": 8290
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18359375,
+      "learning_rate": 1.405478378794034e-06,
+      "loss": 0.9489,
+      "step": 8295
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.3722113529681668e-06,
+      "loss": 0.9042,
+      "step": 8300
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.3393400446168435e-06,
+      "loss": 0.9168,
+      "step": 8305
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.306864585627332e-06,
+      "loss": 0.9866,
+      "step": 8310
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 1.2747851062986926e-06,
+      "loss": 0.9297,
+      "step": 8315
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.189453125,
+      "learning_rate": 1.2431017353412233e-06,
+      "loss": 0.9472,
+      "step": 8320
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.2118145998758824e-06,
+      "loss": 0.9502,
+      "step": 8325
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.1809238254339105e-06,
+      "loss": 0.9025,
+      "step": 8330
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.1504295359562434e-06,
+      "loss": 0.9621,
+      "step": 8335
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.1203318537929996e-06,
+      "loss": 0.9844,
+      "step": 8340
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 1.090630899703038e-06,
+      "loss": 0.961,
+      "step": 8345
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.189453125,
+      "learning_rate": 1.0613267928534453e-06,
+      "loss": 1.0294,
+      "step": 8350
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.0324196508190832e-06,
+      "loss": 0.9599,
+      "step": 8355
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.0039095895820639e-06,
+      "loss": 0.9601,
+      "step": 8360
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1875,
+      "learning_rate": 9.757967235313526e-07,
+      "loss": 0.9332,
+      "step": 8365
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.185546875,
+      "learning_rate": 9.480811654622557e-07,
+      "loss": 0.9132,
+      "step": 8370
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 9.207630265760103e-07,
+      "loss": 0.9203,
+      "step": 8375
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.938424164792736e-07,
+      "loss": 0.9423,
+      "step": 8380
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1796875,
+      "learning_rate": 8.673194431837784e-07,
+      "loss": 0.9558,
+      "step": 8385
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 8.411942131058115e-07,
+      "loss": 0.9531,
+      "step": 8390
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 8.154668310658253e-07,
+      "loss": 0.9631,
+      "step": 8395
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.90137400288049e-07,
+      "loss": 0.9279,
+      "step": 8400
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.652060223999669e-07,
+      "loss": 0.8924,
+      "step": 8405
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.406727974320627e-07,
+      "loss": 0.9217,
+      "step": 8410
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 7.16537823817276e-07,
+      "loss": 0.9582,
+      "step": 8415
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.928011983907245e-07,
+      "loss": 0.9871,
+      "step": 8420
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.694630163892046e-07,
+      "loss": 0.9488,
+      "step": 8425
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 6.465233714509245e-07,
+      "loss": 0.9788,
+      "step": 8430
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18359375,
+      "learning_rate": 6.239823556150159e-07,
+      "loss": 0.9567,
+      "step": 8435
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.01840059321257e-07,
+      "loss": 0.9152,
+      "step": 8440
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 5.800965714096496e-07,
+      "loss": 0.9505,
+      "step": 8445
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 5.587519791200869e-07,
+      "loss": 0.9529,
+      "step": 8450
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 5.378063680920087e-07,
+      "loss": 0.9628,
+      "step": 8455
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.172598223640468e-07,
+      "loss": 0.9805,
+      "step": 8460
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.193359375,
+      "learning_rate": 4.971124243736913e-07,
+      "loss": 0.9674,
+      "step": 8465
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1875,
+      "learning_rate": 4.773642549569579e-07,
+      "loss": 0.9235,
+      "step": 8470
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.181640625,
+      "learning_rate": 4.5801539334805466e-07,
+      "loss": 0.9514,
+      "step": 8475
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1875,
+      "learning_rate": 4.390659171790934e-07,
+      "loss": 0.9045,
+      "step": 8480
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 4.205159024797456e-07,
+      "loss": 0.9634,
+      "step": 8485
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18359375,
+      "learning_rate": 4.023654236769647e-07,
+      "loss": 0.9239,
+      "step": 8490
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.8461455359466435e-07,
+      "loss": 0.9787,
+      "step": 8495
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.672633634534295e-07,
+      "loss": 0.9538,
+      "step": 8500
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.5031192287023895e-07,
+      "loss": 0.9254,
+      "step": 8505
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.3376029985819903e-07,
+      "loss": 0.9176,
+      "step": 8510
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.189453125,
+      "learning_rate": 3.176085608262436e-07,
+      "loss": 0.9247,
+      "step": 8515
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 3.0185677057887885e-07,
+      "loss": 0.9522,
+      "step": 8520
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 2.8650499231591685e-07,
+      "loss": 0.9726,
+      "step": 8525
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.715532876322646e-07,
+      "loss": 0.9405,
+      "step": 8530
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.570017165175909e-07,
+      "loss": 0.9377,
+      "step": 8535
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.428503373561708e-07,
+      "loss": 0.9515,
+      "step": 8540
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.2909920692660847e-07,
+      "loss": 0.9543,
+      "step": 8545
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.1574838040161473e-07,
+      "loss": 0.9612,
+      "step": 8550
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.0279791134778515e-07,
+      "loss": 0.9845,
+      "step": 8555
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.9024785172541136e-07,
+      "loss": 0.9681,
+      "step": 8560
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.780982518882035e-07,
+      "loss": 0.9659,
+      "step": 8565
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.6634916058319018e-07,
+      "loss": 0.9751,
+      "step": 8570
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.5500062495041878e-07,
+      "loss": 0.9723,
+      "step": 8575
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.193359375,
+      "learning_rate": 1.4405269052284455e-07,
+      "loss": 0.9445,
+      "step": 8580
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 1.3350540122611942e-07,
+      "loss": 0.9415,
+      "step": 8585
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.2335879937839246e-07,
+      "loss": 0.9441,
+      "step": 8590
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.1361292569017635e-07,
+      "loss": 0.9352,
+      "step": 8595
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 1.0426781926416996e-07,
+      "loss": 0.9541,
+      "step": 8600
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.532351759510283e-08,
+      "loss": 0.9471,
+      "step": 8605
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 8.678005656957977e-08,
+      "loss": 0.9756,
+      "step": 8610
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.86374704659254e-08,
+      "loss": 0.9784,
+      "step": 8615
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.089579195409534e-08,
+      "loss": 0.9212,
+      "step": 8620
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.355505209548751e-08,
+      "loss": 0.955,
+      "step": 8625
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18359375,
+      "learning_rate": 5.6615280342842135e-08,
+      "loss": 0.9341,
+      "step": 8630
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 5.00765045401197e-08,
+      "loss": 0.939,
+      "step": 8635
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18359375,
+      "learning_rate": 4.3938750922412064e-08,
+      "loss": 0.9634,
+      "step": 8640
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.173828125,
+      "learning_rate": 3.82020441158093e-08,
+      "loss": 0.9063,
+      "step": 8645
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 3.286640713727751e-08,
+      "loss": 0.9496,
+      "step": 8650
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.7931861394658865e-08,
+      "loss": 0.883,
+      "step": 8655
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.3398426686471743e-08,
+      "loss": 0.9478,
+      "step": 8660
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.193359375,
+      "learning_rate": 1.9266121201899634e-08,
+      "loss": 0.9193,
+      "step": 8665
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.5534961520724533e-08,
+      "loss": 0.9255,
+      "step": 8670
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 1.2204962613204807e-08,
+      "loss": 0.8715,
+      "step": 8675
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.185546875,
+      "learning_rate": 9.276137840075194e-09,
+      "loss": 0.9714,
+      "step": 8680
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.748498952446891e-09,
+      "loss": 0.9396,
+      "step": 8685
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.6220560918075474e-09,
+      "loss": 0.9383,
+      "step": 8690
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 2.8968177899213465e-09,
+      "loss": 0.921,
+      "step": 8695
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.189453125,
+      "learning_rate": 1.572790968851212e-09,
+      "loss": 0.9909,
+      "step": 8700
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.19921875,
+      "learning_rate": 6.49980940892192e-10,
+      "loss": 0.9556,
+      "step": 8705
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.2839140858256215e-10,
+      "loss": 0.939,
+      "step": 8710
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.9498655200004578,
+      "eval_runtime": 959.6662,
+      "eval_samples_per_second": 16.08,
+      "eval_steps_per_second": 1.006,
+      "step": 8714
+    },
+    {
+      "epoch": 1.0,
+      "step": 8714,
+      "total_flos": 1.2254141370364592e+19,
+      "train_loss": 0.9222531057372673,
+      "train_runtime": 31670.6049,
+      "train_samples_per_second": 4.402,
+      "train_steps_per_second": 0.275
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 8714,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "total_flos": 1.2254141370364592e+19,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}