diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,24440 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999713113578335,
+  "eval_steps": 500,
+  "global_step": 17428,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.1474469305794606e-07,
+      "loss": 1.3368,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.376953125,
+      "learning_rate": 5.737234652897304e-07,
+      "loss": 1.0791,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.1474469305794607e-06,
+      "loss": 1.1069,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.721170395869191e-06,
+      "loss": 1.1566,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.384765625,
+      "learning_rate": 2.2948938611589215e-06,
+      "loss": 1.1799,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.868617326448652e-06,
+      "loss": 1.111,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.306640625,
+      "learning_rate": 3.442340791738382e-06,
+      "loss": 1.0485,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.016064257028113e-06,
+      "loss": 1.1027,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.3125,
+      "learning_rate": 4.589787722317843e-06,
+      "loss": 1.1293,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.163511187607573e-06,
+      "loss": 1.0885,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.3203125,
+      "learning_rate": 5.737234652897304e-06,
+      "loss": 1.1255,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.310958118187034e-06,
+      "loss": 1.1155,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.298828125,
+      "learning_rate": 6.884681583476764e-06,
+      "loss": 1.0925,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.314453125,
+      "learning_rate": 7.4584050487664955e-06,
+      "loss": 1.1294,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.283203125,
+      "learning_rate": 8.032128514056226e-06,
+      "loss": 1.1594,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2734375,
+      "learning_rate": 8.605851979345956e-06,
+      "loss": 1.0828,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.179575444635686e-06,
+      "loss": 1.1212,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 9.753298909925416e-06,
+      "loss": 1.0751,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.0327022375215146e-05,
+      "loss": 1.0247,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.0900745840504876e-05,
+      "loss": 1.0963,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.1474469305794608e-05,
+      "loss": 1.0563,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2048192771084338e-05,
+      "loss": 1.1441,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.2621916236374069e-05,
+      "loss": 0.9927,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.3195639701663797e-05,
+      "loss": 0.9616,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 1.3769363166953527e-05,
+      "loss": 0.9632,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2421875,
+      "learning_rate": 1.434308663224326e-05,
+      "loss": 1.0317,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.4916810097532991e-05,
+      "loss": 1.0095,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.549053356282272e-05,
+      "loss": 0.9795,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2421875,
+      "learning_rate": 1.606425702811245e-05,
+      "loss": 0.9755,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.23828125,
+      "learning_rate": 1.663798049340218e-05,
+      "loss": 0.978,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.721170395869191e-05,
+      "loss": 0.9761,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 1.7785427423981642e-05,
+      "loss": 0.9764,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.8359150889271372e-05,
+      "loss": 0.9912,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.8932874354561102e-05,
+      "loss": 1.0147,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9506597819850832e-05,
+      "loss": 1.0047,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.0080321285140562e-05,
+      "loss": 0.9577,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 2.0654044750430293e-05,
+      "loss": 1.0258,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.451171875,
+      "learning_rate": 2.1227768215720023e-05,
+      "loss": 1.0307,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.27734375,
+      "learning_rate": 2.1801491681009753e-05,
+      "loss": 1.0406,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.2375215146299486e-05,
+      "loss": 0.9139,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.2948938611589217e-05,
+      "loss": 1.0682,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.322265625,
+      "learning_rate": 2.3522662076878947e-05,
+      "loss": 1.0834,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.4096385542168677e-05,
+      "loss": 1.0576,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.4670109007458407e-05,
+      "loss": 1.0109,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.279296875,
+      "learning_rate": 2.5243832472748137e-05,
+      "loss": 0.9636,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.5817555938037867e-05,
+      "loss": 1.016,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.283203125,
+      "learning_rate": 2.6391279403327594e-05,
+      "loss": 1.0972,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.6965002868617328e-05,
+      "loss": 0.9953,
+      "step": 235
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.283203125,
+      "learning_rate": 2.7538726333907055e-05,
+      "loss": 1.0365,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.275390625,
+      "learning_rate": 2.8112449799196788e-05,
+      "loss": 1.0454,
+      "step": 245
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.310546875,
+      "learning_rate": 2.868617326448652e-05,
+      "loss": 1.0073,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.925989672977625e-05,
+      "loss": 0.9828,
+      "step": 255
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.9833620195065982e-05,
+      "loss": 0.9571,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.31640625,
+      "learning_rate": 3.040734366035571e-05,
+      "loss": 1.0254,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.283203125,
+      "learning_rate": 3.098106712564544e-05,
+      "loss": 0.9365,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2734375,
+      "learning_rate": 3.155479059093517e-05,
+      "loss": 0.9507,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.296875,
+      "learning_rate": 3.21285140562249e-05,
+      "loss": 0.9689,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.27734375,
+      "learning_rate": 3.2702237521514636e-05,
+      "loss": 1.0112,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.29296875,
+      "learning_rate": 3.327596098680436e-05,
+      "loss": 1.0417,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.27734375,
+      "learning_rate": 3.3849684452094096e-05,
+      "loss": 1.0556,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.28515625,
+      "learning_rate": 3.442340791738382e-05,
+      "loss": 0.9538,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.27734375,
+      "learning_rate": 3.499713138267356e-05,
+      "loss": 0.9891,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.267578125,
+      "learning_rate": 3.5570854847963284e-05,
+      "loss": 0.9103,
+      "step": 310
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.267578125,
+      "learning_rate": 3.614457831325301e-05,
+      "loss": 1.0865,
+      "step": 315
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.294921875,
+      "learning_rate": 3.6718301778542744e-05,
+      "loss": 0.9835,
+      "step": 320
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.28515625,
+      "learning_rate": 3.729202524383247e-05,
+      "loss": 1.0064,
+      "step": 325
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.30078125,
+      "learning_rate": 3.7865748709122204e-05,
+      "loss": 1.0383,
+      "step": 330
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.27734375,
+      "learning_rate": 3.843947217441193e-05,
+      "loss": 1.0254,
+      "step": 335
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2890625,
+      "learning_rate": 3.9013195639701665e-05,
+      "loss": 1.0076,
+      "step": 340
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.263671875,
+      "learning_rate": 3.958691910499139e-05,
+      "loss": 1.0027,
+      "step": 345
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2890625,
+      "learning_rate": 4.0160642570281125e-05,
+      "loss": 1.0093,
+      "step": 350
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.283203125,
+      "learning_rate": 4.073436603557086e-05,
+      "loss": 1.0416,
+      "step": 355
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2890625,
+      "learning_rate": 4.1308089500860585e-05,
+      "loss": 1.0127,
+      "step": 360
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.271484375,
+      "learning_rate": 4.188181296615032e-05,
+      "loss": 1.0044,
+      "step": 365
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.275390625,
+      "learning_rate": 4.2455536431440046e-05,
+      "loss": 0.9113,
+      "step": 370
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.26171875,
+      "learning_rate": 4.302925989672978e-05,
+      "loss": 0.9618,
+      "step": 375
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.26953125,
+      "learning_rate": 4.3602983362019506e-05,
+      "loss": 0.9842,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.318359375,
+      "learning_rate": 4.417670682730924e-05,
+      "loss": 0.9779,
+      "step": 385
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.29296875,
+      "learning_rate": 4.475043029259897e-05,
+      "loss": 0.9694,
+      "step": 390
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.53241537578887e-05,
+      "loss": 1.0035,
+      "step": 395
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.259765625,
+      "learning_rate": 4.589787722317843e-05,
+      "loss": 0.9444,
+      "step": 400
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.26953125,
+      "learning_rate": 4.647160068846816e-05,
+      "loss": 0.9722,
+      "step": 405
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.283203125,
+      "learning_rate": 4.7045324153757894e-05,
+      "loss": 0.9614,
+      "step": 410
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.26953125,
+      "learning_rate": 4.761904761904762e-05,
+      "loss": 1.0325,
+      "step": 415
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.279296875,
+      "learning_rate": 4.8192771084337354e-05,
+      "loss": 0.9545,
+      "step": 420
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.263671875,
+      "learning_rate": 4.876649454962709e-05,
+      "loss": 0.9912,
+      "step": 425
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.29296875,
+      "learning_rate": 4.9340218014916814e-05,
+      "loss": 1.0194,
+      "step": 430
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.263671875,
+      "learning_rate": 4.991394148020654e-05,
+      "loss": 1.048,
+      "step": 435
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.240234375,
+      "learning_rate": 5.0487664945496275e-05,
+      "loss": 0.936,
+      "step": 440
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2421875,
+      "learning_rate": 5.1061388410786e-05,
+      "loss": 1.0078,
+      "step": 445
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.263671875,
+      "learning_rate": 5.1635111876075735e-05,
+      "loss": 0.9624,
+      "step": 450
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.220883534136547e-05,
+      "loss": 0.977,
+      "step": 455
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.271484375,
+      "learning_rate": 5.278255880665519e-05,
+      "loss": 1.0434,
+      "step": 460
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.335628227194492e-05,
+      "loss": 0.9327,
+      "step": 465
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.3930005737234656e-05,
+      "loss": 1.0557,
+      "step": 470
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.27734375,
+      "learning_rate": 5.450372920252439e-05,
+      "loss": 0.9911,
+      "step": 475
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.507745266781411e-05,
+      "loss": 0.9561,
+      "step": 480
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.279296875,
+      "learning_rate": 5.565117613310384e-05,
+      "loss": 1.0194,
+      "step": 485
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.279296875,
+      "learning_rate": 5.6224899598393576e-05,
+      "loss": 0.9214,
+      "step": 490
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.283203125,
+      "learning_rate": 5.679862306368331e-05,
+      "loss": 0.9457,
+      "step": 495
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.287109375,
+      "learning_rate": 5.737234652897304e-05,
+      "loss": 0.9919,
+      "step": 500
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.298828125,
+      "learning_rate": 5.794606999426276e-05,
+      "loss": 0.9558,
+      "step": 505
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2578125,
+      "learning_rate": 5.85197934595525e-05,
+      "loss": 0.9813,
+      "step": 510
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.27734375,
+      "learning_rate": 5.909351692484223e-05,
+      "loss": 1.0898,
+      "step": 515
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.302734375,
+      "learning_rate": 5.9667240390131964e-05,
+      "loss": 1.0559,
+      "step": 520
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.25,
+      "learning_rate": 6.02409638554217e-05,
+      "loss": 1.0005,
+      "step": 525
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.267578125,
+      "learning_rate": 6.081468732071142e-05,
+      "loss": 0.9529,
+      "step": 530
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.244140625,
+      "learning_rate": 6.138841078600115e-05,
+      "loss": 0.9576,
+      "step": 535
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 6.196213425129088e-05,
+      "loss": 0.9927,
+      "step": 540
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.251953125,
+      "learning_rate": 6.253585771658062e-05,
+      "loss": 0.9369,
+      "step": 545
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.265625,
+      "learning_rate": 6.310958118187034e-05,
+      "loss": 1.0384,
+      "step": 550
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 6.368330464716007e-05,
+      "loss": 0.9414,
+      "step": 555
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.265625,
+      "learning_rate": 6.42570281124498e-05,
+      "loss": 0.9954,
+      "step": 560
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.267578125,
+      "learning_rate": 6.483075157773954e-05,
+      "loss": 0.9833,
+      "step": 565
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 6.540447504302927e-05,
+      "loss": 1.0212,
+      "step": 570
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.25,
+      "learning_rate": 6.597819850831899e-05,
+      "loss": 1.0068,
+      "step": 575
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.655192197360873e-05,
+      "loss": 0.9296,
+      "step": 580
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2421875,
+      "learning_rate": 6.712564543889846e-05,
+      "loss": 1.0144,
+      "step": 585
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 6.769936890418819e-05,
+      "loss": 1.0089,
+      "step": 590
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.248046875,
+      "learning_rate": 6.827309236947793e-05,
+      "loss": 1.0388,
+      "step": 595
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.25390625,
+      "learning_rate": 6.884681583476765e-05,
+      "loss": 1.055,
+      "step": 600
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 6.942053930005738e-05,
+      "loss": 0.9395,
+      "step": 605
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 6.999426276534711e-05,
+      "loss": 0.9684,
+      "step": 610
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 7.056798623063683e-05,
+      "loss": 0.9351,
+      "step": 615
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.26171875,
+      "learning_rate": 7.114170969592657e-05,
+      "loss": 1.0539,
+      "step": 620
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.234375,
+      "learning_rate": 7.17154331612163e-05,
+      "loss": 1.0311,
+      "step": 625
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 7.228915662650602e-05,
+      "loss": 0.9639,
+      "step": 630
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 7.286288009179575e-05,
+      "loss": 0.9518,
+      "step": 635
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.343660355708549e-05,
+      "loss": 0.9754,
+      "step": 640
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.25,
+      "learning_rate": 7.401032702237521e-05,
+      "loss": 0.9694,
+      "step": 645
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 7.458405048766494e-05,
+      "loss": 0.9101,
+      "step": 650
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 7.515777395295467e-05,
+      "loss": 1.0275,
+      "step": 655
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 7.573149741824441e-05,
+      "loss": 0.9585,
+      "step": 660
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.25,
+      "learning_rate": 7.630522088353414e-05,
+      "loss": 0.9789,
+      "step": 665
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.23828125,
+      "learning_rate": 7.687894434882386e-05,
+      "loss": 0.8663,
+      "step": 670
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 7.74526678141136e-05,
+      "loss": 1.0568,
+      "step": 675
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 7.802639127940333e-05,
+      "loss": 1.0539,
+      "step": 680
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.255859375,
+      "learning_rate": 7.860011474469306e-05,
+      "loss": 0.9589,
+      "step": 685
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.263671875,
+      "learning_rate": 7.917383820998278e-05,
+      "loss": 0.9545,
+      "step": 690
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.24609375,
+      "learning_rate": 7.974756167527252e-05,
+      "loss": 0.9511,
+      "step": 695
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.25390625,
+      "learning_rate": 8.032128514056225e-05,
+      "loss": 1.0379,
+      "step": 700
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.232421875,
+      "learning_rate": 8.089500860585198e-05,
+      "loss": 0.9617,
+      "step": 705
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 8.146873207114172e-05,
+      "loss": 0.9202,
+      "step": 710
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 8.204245553643144e-05,
+      "loss": 1.0169,
+      "step": 715
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.240234375,
+      "learning_rate": 8.261617900172117e-05,
+      "loss": 0.955,
+      "step": 720
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.26171875,
+      "learning_rate": 8.31899024670109e-05,
+      "loss": 0.9753,
+      "step": 725
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2265625,
+      "learning_rate": 8.376362593230064e-05,
+      "loss": 1.0299,
+      "step": 730
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.271484375,
+      "learning_rate": 8.433734939759037e-05,
+      "loss": 1.0099,
+      "step": 735
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.232421875,
+      "learning_rate": 8.491107286288009e-05,
+      "loss": 0.9463,
+      "step": 740
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.228515625,
+      "learning_rate": 8.548479632816982e-05,
+      "loss": 1.0125,
+      "step": 745
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 8.605851979345956e-05,
+      "loss": 0.9646,
+      "step": 750
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.23046875,
+      "learning_rate": 8.663224325874929e-05,
+      "loss": 0.9458,
+      "step": 755
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.232421875,
+      "learning_rate": 8.720596672403901e-05,
+      "loss": 0.9823,
+      "step": 760
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 8.777969018932875e-05,
+      "loss": 0.9312,
+      "step": 765
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.232421875,
+      "learning_rate": 8.835341365461848e-05,
+      "loss": 1.0313,
+      "step": 770
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 8.892713711990821e-05,
+      "loss": 0.9286,
+      "step": 775
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.950086058519795e-05,
+      "loss": 0.9921,
+      "step": 780
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.224609375,
+      "learning_rate": 9.007458405048767e-05,
+      "loss": 0.9907,
+      "step": 785
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.228515625,
+      "learning_rate": 9.06483075157774e-05,
+      "loss": 0.9607,
+      "step": 790
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 9.122203098106713e-05,
+      "loss": 1.0123,
+      "step": 795
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.179575444635687e-05,
+      "loss": 0.9162,
+      "step": 800
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 9.23694779116466e-05,
+      "loss": 0.94,
+      "step": 805
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.228515625,
+      "learning_rate": 9.294320137693632e-05,
+      "loss": 0.9643,
+      "step": 810
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.232421875,
+      "learning_rate": 9.351692484222605e-05,
+      "loss": 0.9354,
+      "step": 815
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2265625,
+      "learning_rate": 9.409064830751579e-05,
+      "loss": 0.9763,
+      "step": 820
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.466437177280552e-05,
+      "loss": 1.0257,
+      "step": 825
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.236328125,
+      "learning_rate": 9.523809523809524e-05,
+      "loss": 1.0173,
+      "step": 830
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 9.581181870338497e-05,
+      "loss": 0.9775,
+      "step": 835
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 9.638554216867471e-05,
+      "loss": 0.9463,
+      "step": 840
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 9.695926563396444e-05,
+      "loss": 0.9372,
+      "step": 845
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.753298909925417e-05,
+      "loss": 1.0044,
+      "step": 850
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.81067125645439e-05,
+      "loss": 0.9879,
+      "step": 855
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.868043602983363e-05,
+      "loss": 1.0205,
+      "step": 860
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 9.925415949512336e-05,
+      "loss": 0.9758,
+      "step": 865
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.982788296041308e-05,
+      "loss": 0.9762,
+      "step": 870
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00010040160642570282,
+      "loss": 0.9892,
+      "step": 875
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00010097532989099255,
+      "loss": 1.0356,
+      "step": 880
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00010154905335628228,
+      "loss": 1.0237,
+      "step": 885
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.25,
+      "learning_rate": 0.000102122776821572,
+      "loss": 1.0023,
+      "step": 890
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00010269650028686174,
+      "loss": 0.9528,
+      "step": 895
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00010327022375215147,
+      "loss": 1.0398,
+      "step": 900
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00010384394721744119,
+      "loss": 0.9367,
+      "step": 905
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00010441767068273094,
+      "loss": 1.0829,
+      "step": 910
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00010499139414802066,
+      "loss": 0.9457,
+      "step": 915
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00010556511761331038,
+      "loss": 1.0245,
+      "step": 920
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00010613884107860012,
+      "loss": 0.9817,
+      "step": 925
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00010671256454388984,
+      "loss": 0.9489,
+      "step": 930
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00010728628800917956,
+      "loss": 1.0089,
+      "step": 935
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00010786001147446931,
+      "loss": 0.9619,
+      "step": 940
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00010843373493975903,
+      "loss": 0.959,
+      "step": 945
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00010900745840504878,
+      "loss": 0.9085,
+      "step": 950
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0001095811818703385,
+      "loss": 0.9786,
+      "step": 955
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00011015490533562822,
+      "loss": 0.9537,
+      "step": 960
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00011072862880091797,
+      "loss": 0.9376,
+      "step": 965
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00011130235226620769,
+      "loss": 0.958,
+      "step": 970
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00011187607573149743,
+      "loss": 0.9215,
+      "step": 975
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00011244979919678715,
+      "loss": 0.9653,
+      "step": 980
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00011302352266207687,
+      "loss": 0.9603,
+      "step": 985
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00011359724612736662,
+      "loss": 0.9093,
+      "step": 990
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00011417096959265634,
+      "loss": 0.9243,
+      "step": 995
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00011474469305794609,
+      "loss": 1.0176,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001153184165232358,
+      "loss": 0.9657,
+      "step": 1005
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00011589213998852553,
+      "loss": 0.9962,
+      "step": 1010
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00011646586345381527,
+      "loss": 1.022,
+      "step": 1015
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.000117039586919105,
+      "loss": 0.9741,
+      "step": 1020
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00011761331038439474,
+      "loss": 0.9569,
+      "step": 1025
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00011818703384968446,
+      "loss": 0.9687,
+      "step": 1030
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00011876075731497418,
+      "loss": 0.9592,
+      "step": 1035
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00011933448078026393,
+      "loss": 0.9529,
+      "step": 1040
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00011990820424555365,
+      "loss": 1.077,
+      "step": 1045
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001204819277108434,
+      "loss": 0.9984,
+      "step": 1050
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00012105565117613311,
+      "loss": 0.9589,
+      "step": 1055
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00012162937464142283,
+      "loss": 0.9386,
+      "step": 1060
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00012220309810671257,
+      "loss": 1.0137,
+      "step": 1065
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0001227768215720023,
+      "loss": 0.8944,
+      "step": 1070
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.000123350545037292,
+      "loss": 0.9208,
+      "step": 1075
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00012392426850258177,
+      "loss": 0.9763,
+      "step": 1080
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00012449799196787148,
+      "loss": 0.9631,
+      "step": 1085
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00012507171543316124,
+      "loss": 0.9173,
+      "step": 1090
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012564543889845094,
+      "loss": 0.9671,
+      "step": 1095
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00012621916236374068,
+      "loss": 0.9751,
+      "step": 1100
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0001267928858290304,
+      "loss": 0.9351,
+      "step": 1105
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00012736660929432014,
+      "loss": 0.9701,
+      "step": 1110
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00012794033275960988,
+      "loss": 0.9568,
+      "step": 1115
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001285140562248996,
+      "loss": 0.9354,
+      "step": 1120
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00012908777969018932,
+      "loss": 0.9803,
+      "step": 1125
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00012966150315547908,
+      "loss": 0.9745,
+      "step": 1130
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00013023522662076878,
+      "loss": 1.0447,
+      "step": 1135
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00013080895008605854,
+      "loss": 0.9392,
+      "step": 1140
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00013138267355134825,
+      "loss": 1.0016,
+      "step": 1145
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00013195639701663798,
+      "loss": 1.0061,
+      "step": 1150
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00013253012048192772,
+      "loss": 0.9837,
+      "step": 1155
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00013310384394721745,
+      "loss": 1.0022,
+      "step": 1160
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00013367756741250719,
+      "loss": 0.978,
+      "step": 1165
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00013425129087779692,
+      "loss": 0.9534,
+      "step": 1170
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00013482501434308663,
+      "loss": 0.9628,
+      "step": 1175
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00013539873780837639,
+      "loss": 0.9247,
+      "step": 1180
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0001359724612736661,
+      "loss": 0.9551,
+      "step": 1185
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00013654618473895585,
+      "loss": 0.9261,
+      "step": 1190
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00013711990820424556,
+      "loss": 1.0057,
+      "step": 1195
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001376936316695353,
+      "loss": 0.9528,
+      "step": 1200
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00013826735513482503,
+      "loss": 0.9443,
+      "step": 1205
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00013884107860011476,
+      "loss": 0.9167,
+      "step": 1210
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00013941480206540447,
+      "loss": 0.95,
+      "step": 1215
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00013998852553069423,
+      "loss": 0.9668,
+      "step": 1220
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00014056224899598393,
+      "loss": 0.9145,
+      "step": 1225
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00014113597246127367,
+      "loss": 0.9576,
+      "step": 1230
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001417096959265634,
+      "loss": 0.9274,
+      "step": 1235
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00014228341939185313,
+      "loss": 0.9507,
+      "step": 1240
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00014285714285714287,
+      "loss": 0.9324,
+      "step": 1245
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001434308663224326,
+      "loss": 1.0164,
+      "step": 1250
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00014400458978772233,
+      "loss": 0.9863,
+      "step": 1255
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00014457831325301204,
+      "loss": 0.9704,
+      "step": 1260
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00014515203671830177,
+      "loss": 1.0213,
+      "step": 1265
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001457257601835915,
+      "loss": 0.9403,
+      "step": 1270
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00014629948364888124,
+      "loss": 0.9601,
+      "step": 1275
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00014687320711417098,
+      "loss": 1.0241,
+      "step": 1280
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0001474469305794607,
+      "loss": 1.0017,
+      "step": 1285
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00014802065404475042,
+      "loss": 0.9326,
+      "step": 1290
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00014859437751004018,
+      "loss": 1.0238,
+      "step": 1295
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00014916810097532988,
+      "loss": 0.9449,
+      "step": 1300
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00014974182444061964,
+      "loss": 0.9592,
+      "step": 1305
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00015031554790590935,
+      "loss": 0.9213,
+      "step": 1310
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00015088927137119908,
+      "loss": 0.9286,
+      "step": 1315
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00015146299483648882,
+      "loss": 0.9357,
+      "step": 1320
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00015203671830177855,
+      "loss": 0.9788,
+      "step": 1325
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00015261044176706828,
+      "loss": 0.9794,
+      "step": 1330
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00015318416523235802,
+      "loss": 0.9329,
+      "step": 1335
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00015375788869764772,
+      "loss": 0.9709,
+      "step": 1340
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00015433161216293748,
+      "loss": 1.0187,
+      "step": 1345
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0001549053356282272,
+      "loss": 0.9459,
+      "step": 1350
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00015547905909351695,
+      "loss": 0.9776,
+      "step": 1355
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00015605278255880666,
+      "loss": 0.9328,
+      "step": 1360
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001566265060240964,
+      "loss": 1.001,
+      "step": 1365
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00015720022948938613,
+      "loss": 0.9298,
+      "step": 1370
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00015777395295467586,
+      "loss": 0.9574,
+      "step": 1375
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00015834767641996557,
+      "loss": 0.9998,
+      "step": 1380
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00015892139988525533,
+      "loss": 0.9435,
+      "step": 1385
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00015949512335054503,
+      "loss": 0.9459,
+      "step": 1390
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0001600688468158348,
+      "loss": 0.9581,
+      "step": 1395
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001606425702811245,
+      "loss": 0.9255,
+      "step": 1400
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00016121629374641423,
+      "loss": 1.0634,
+      "step": 1405
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00016179001721170397,
+      "loss": 1.0368,
+      "step": 1410
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0001623637406769937,
+      "loss": 0.9813,
+      "step": 1415
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00016293746414228343,
+      "loss": 0.9923,
+      "step": 1420
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00016351118760757317,
+      "loss": 0.9478,
+      "step": 1425
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00016408491107286287,
+      "loss": 1.034,
+      "step": 1430
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00016465863453815263,
+      "loss": 0.9275,
+      "step": 1435
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00016523235800344234,
+      "loss": 1.0333,
+      "step": 1440
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0001658060814687321,
+      "loss": 0.9415,
+      "step": 1445
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001663798049340218,
+      "loss": 0.9903,
+      "step": 1450
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00016695352839931154,
+      "loss": 0.9824,
+      "step": 1455
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00016752725186460127,
+      "loss": 0.8981,
+      "step": 1460
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.000168100975329891,
+      "loss": 0.9891,
+      "step": 1465
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00016867469879518074,
+      "loss": 1.0248,
+      "step": 1470
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00016924842226047048,
+      "loss": 1.0793,
+      "step": 1475
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00016982214572576018,
+      "loss": 0.9942,
+      "step": 1480
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00017039586919104992,
+      "loss": 0.9166,
+      "step": 1485
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00017096959265633965,
+      "loss": 0.9555,
+      "step": 1490
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00017154331612162938,
+      "loss": 0.897,
+      "step": 1495
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00017211703958691912,
+      "loss": 1.0034,
+      "step": 1500
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00017269076305220885,
+      "loss": 1.0096,
+      "step": 1505
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017326448651749858,
+      "loss": 1.0415,
+      "step": 1510
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001738382099827883,
+      "loss": 0.9825,
+      "step": 1515
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00017441193344807802,
+      "loss": 0.994,
+      "step": 1520
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00017498565691336776,
+      "loss": 1.0082,
+      "step": 1525
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001755593803786575,
+      "loss": 0.9467,
+      "step": 1530
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00017613310384394722,
+      "loss": 0.895,
+      "step": 1535
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00017670682730923696,
+      "loss": 0.9505,
+      "step": 1540
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00017728055077452666,
+      "loss": 1.0304,
+      "step": 1545
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00017785427423981642,
+      "loss": 0.9931,
+      "step": 1550
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00017842799770510613,
+      "loss": 0.9413,
+      "step": 1555
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001790017211703959,
+      "loss": 0.9221,
+      "step": 1560
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001795754446356856,
+      "loss": 0.9809,
+      "step": 1565
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00018014916810097533,
+      "loss": 0.9561,
+      "step": 1570
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00018072289156626507,
+      "loss": 0.9034,
+      "step": 1575
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0001812966150315548,
+      "loss": 0.9123,
+      "step": 1580
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00018187033849684453,
+      "loss": 0.9413,
+      "step": 1585
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00018244406196213427,
+      "loss": 0.9556,
+      "step": 1590
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018301778542742397,
+      "loss": 0.9516,
+      "step": 1595
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018359150889271373,
+      "loss": 1.0159,
+      "step": 1600
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018416523235800344,
+      "loss": 0.9851,
+      "step": 1605
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001847389558232932,
+      "loss": 0.9938,
+      "step": 1610
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001853126792885829,
+      "loss": 1.0143,
+      "step": 1615
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00018588640275387264,
+      "loss": 0.9578,
+      "step": 1620
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00018646012621916237,
+      "loss": 0.9961,
+      "step": 1625
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001870338496844521,
+      "loss": 0.9975,
+      "step": 1630
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00018760757314974184,
+      "loss": 0.9234,
+      "step": 1635
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00018818129661503157,
+      "loss": 0.9473,
+      "step": 1640
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00018875502008032128,
+      "loss": 1.0179,
+      "step": 1645
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018932874354561104,
+      "loss": 0.8971,
+      "step": 1650
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00018990246701090075,
+      "loss": 0.9953,
+      "step": 1655
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019047619047619048,
+      "loss": 0.9251,
+      "step": 1660
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00019104991394148021,
+      "loss": 0.9521,
+      "step": 1665
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00019162363740676995,
+      "loss": 0.9647,
+      "step": 1670
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00019219736087205968,
+      "loss": 0.9525,
+      "step": 1675
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00019277108433734942,
+      "loss": 0.9947,
+      "step": 1680
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00019334480780263912,
+      "loss": 0.9591,
+      "step": 1685
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00019391853126792888,
+      "loss": 0.9734,
+      "step": 1690
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001944922547332186,
+      "loss": 1.0063,
+      "step": 1695
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00019506597819850835,
+      "loss": 0.9684,
+      "step": 1700
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019563970166379806,
+      "loss": 0.9499,
+      "step": 1705
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0001962134251290878,
+      "loss": 0.9455,
+      "step": 1710
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019678714859437752,
+      "loss": 0.9922,
+      "step": 1715
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00019736087205966726,
+      "loss": 1.0273,
+      "step": 1720
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.000197934595524957,
+      "loss": 1.0015,
+      "step": 1725
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00019850831899024672,
+      "loss": 0.9356,
+      "step": 1730
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00019908204245553643,
+      "loss": 0.9609,
+      "step": 1735
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00019965576592082616,
+      "loss": 0.9577,
+      "step": 1740
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00019999999197655853,
+      "loss": 0.9813,
+      "step": 1745
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001999999017128567,
+      "loss": 0.9655,
+      "step": 1750
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019999971115624204,
+      "loss": 0.9365,
+      "step": 1755
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00019999942030690567,
+      "loss": 1.0163,
+      "step": 1760
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001999990291651393,
+      "loss": 0.9904,
+      "step": 1765
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019999853773133513,
+      "loss": 0.9746,
+      "step": 1770
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019999794600598616,
+      "loss": 1.0293,
+      "step": 1775
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00019999725398968577,
+      "loss": 0.928,
+      "step": 1780
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.000199996461683128,
+      "loss": 0.967,
+      "step": 1785
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019999556908710753,
+      "loss": 0.9306,
+      "step": 1790
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00019999457620251953,
+      "loss": 0.8565,
+      "step": 1795
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0001999934830303598,
+      "loss": 0.9218,
+      "step": 1800
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00019999228957172477,
+      "loss": 0.8999,
+      "step": 1805
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001999909958278113,
+      "loss": 0.958,
+      "step": 1810
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.000199989601799917,
+      "loss": 0.9169,
+      "step": 1815
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00019998810748943994,
+      "loss": 0.9944,
+      "step": 1820
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019998651289787885,
+      "loss": 0.9755,
+      "step": 1825
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019998481802683293,
+      "loss": 0.9872,
+      "step": 1830
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019998302287800208,
+      "loss": 0.9734,
+      "step": 1835
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00019998112745318669,
+      "loss": 0.9889,
+      "step": 1840
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019997913175428771,
+      "loss": 0.9554,
+      "step": 1845
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00019997703578330674,
+      "loss": 1.0364,
+      "step": 1850
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019997483954234583,
+      "loss": 0.9727,
+      "step": 1855
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00019997254303360772,
+      "loss": 0.897,
+      "step": 1860
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001999701462593956,
+      "loss": 0.9793,
+      "step": 1865
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00019996764922211328,
+      "loss": 0.9048,
+      "step": 1870
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001999650519242651,
+      "loss": 0.977,
+      "step": 1875
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00019996235436845603,
+      "loss": 0.9698,
+      "step": 1880
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00019995955655739147,
+      "loss": 0.9077,
+      "step": 1885
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00019995665849387742,
+      "loss": 0.9419,
+      "step": 1890
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001999536601808205,
+      "loss": 0.9445,
+      "step": 1895
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019995056162122775,
+      "loss": 1.0352,
+      "step": 1900
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0001999473628182068,
+      "loss": 1.0013,
+      "step": 1905
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00019994406377496588,
+      "loss": 0.9071,
+      "step": 1910
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019994066449481368,
+      "loss": 0.9322,
+      "step": 1915
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001999371649811594,
+      "loss": 0.8917,
+      "step": 1920
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019993356523751287,
+      "loss": 0.9989,
+      "step": 1925
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00019992986526748432,
+      "loss": 0.8982,
+      "step": 1930
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00019992606507478463,
+      "loss": 0.9933,
+      "step": 1935
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001999221646632251,
+      "loss": 0.9206,
+      "step": 1940
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001999181640367175,
+      "loss": 1.0054,
+      "step": 1945
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019991406319927428,
+      "loss": 0.9391,
+      "step": 1950
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019990986215500826,
+      "loss": 0.9102,
+      "step": 1955
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00019990556090813277,
+      "loss": 0.9721,
+      "step": 1960
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019990115946296168,
+      "loss": 0.9735,
+      "step": 1965
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00019989665782390933,
+      "loss": 0.9573,
+      "step": 1970
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019989205599549057,
+      "loss": 0.972,
+      "step": 1975
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00019988735398232066,
+      "loss": 0.9919,
+      "step": 1980
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00019988255178911543,
+      "loss": 0.9882,
+      "step": 1985
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00019987764942069113,
+      "loss": 0.9546,
+      "step": 1990
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00019987264688196453,
+      "loss": 0.994,
+      "step": 1995
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00019986754417795278,
+      "loss": 0.9044,
+      "step": 2000
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019986234131377353,
+      "loss": 0.9218,
+      "step": 2005
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019985703829464492,
+      "loss": 1.0188,
+      "step": 2010
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001998516351258855,
+      "loss": 0.9972,
+      "step": 2015
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00019984613181291426,
+      "loss": 0.9235,
+      "step": 2020
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00019984052836125065,
+      "loss": 0.9297,
+      "step": 2025
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00019983482477651455,
+      "loss": 0.8868,
+      "step": 2030
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019982902106442622,
+      "loss": 0.972,
+      "step": 2035
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019982311723080643,
+      "loss": 0.96,
+      "step": 2040
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00019981711328157626,
+      "loss": 0.8752,
+      "step": 2045
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00019981100922275728,
+      "loss": 0.9826,
+      "step": 2050
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00019980480506047143,
+      "loss": 1.0445,
+      "step": 2055
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019979850080094108,
+      "loss": 0.9625,
+      "step": 2060
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00019979209645048889,
+      "loss": 0.9338,
+      "step": 2065
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019978559201553803,
+      "loss": 0.9259,
+      "step": 2070
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.000199778987502612,
+      "loss": 0.9573,
+      "step": 2075
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00019977228291833462,
+      "loss": 1.0202,
+      "step": 2080
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019976547826943019,
+      "loss": 0.9039,
+      "step": 2085
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019975857356272322,
+      "loss": 0.9118,
+      "step": 2090
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019975156880513867,
+      "loss": 0.9077,
+      "step": 2095
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019974446400370185,
+      "loss": 0.9197,
+      "step": 2100
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019973725916553838,
+      "loss": 0.9025,
+      "step": 2105
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00019972995429787415,
+      "loss": 0.9744,
+      "step": 2110
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00019972254940803553,
+      "loss": 0.922,
+      "step": 2115
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00019971504450344902,
+      "loss": 0.994,
+      "step": 2120
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00019970743959164155,
+      "loss": 0.9548,
+      "step": 2125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019969973468024032,
+      "loss": 0.9639,
+      "step": 2130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001996919297769728,
+      "loss": 0.9711,
+      "step": 2135
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019968402488966677,
+      "loss": 0.8984,
+      "step": 2140
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019967602002625027,
+      "loss": 0.9091,
+      "step": 2145
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019966791519475169,
+      "loss": 0.9957,
+      "step": 2150
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00019965971040329948,
+      "loss": 0.9527,
+      "step": 2155
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00019965140566012253,
+      "loss": 0.8977,
+      "step": 2160
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019964300097355,
+      "loss": 0.9534,
+      "step": 2165
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001996344963520111,
+      "loss": 0.9924,
+      "step": 2170
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00019962589180403542,
+      "loss": 0.9923,
+      "step": 2175
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019961718733825268,
+      "loss": 0.957,
+      "step": 2180
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00019960838296339287,
+      "loss": 0.9071,
+      "step": 2185
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019959947868828618,
+      "loss": 0.9991,
+      "step": 2190
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019959047452186294,
+      "loss": 1.0085,
+      "step": 2195
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00019958137047315375,
+      "loss": 0.9302,
+      "step": 2200
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001995721665512893,
+      "loss": 0.9795,
+      "step": 2205
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00019956286276550048,
+      "loss": 0.943,
+      "step": 2210
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019955345912511837,
+      "loss": 0.9711,
+      "step": 2215
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0001995439556395741,
+      "loss": 0.891,
+      "step": 2220
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00019953435231839906,
+      "loss": 0.9991,
+      "step": 2225
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0001995246491712247,
+      "loss": 0.8978,
+      "step": 2230
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00019951484620778258,
+      "loss": 0.989,
+      "step": 2235
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00019950494343790438,
+      "loss": 0.9474,
+      "step": 2240
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019949494087152193,
+      "loss": 0.8953,
+      "step": 2245
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00019948483851866703,
+      "loss": 1.0249,
+      "step": 2250
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0001994746363894717,
+      "loss": 0.9422,
+      "step": 2255
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.000199464334494168,
+      "loss": 0.8769,
+      "step": 2260
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019945393284308792,
+      "loss": 1.0703,
+      "step": 2265
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001994434314466636,
+      "loss": 0.921,
+      "step": 2270
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00019943283031542726,
+      "loss": 0.9315,
+      "step": 2275
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019942212946001103,
+      "loss": 1.0263,
+      "step": 2280
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00019941132889114718,
+      "loss": 0.956,
+      "step": 2285
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001994004286196679,
+      "loss": 0.9692,
+      "step": 2290
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0001993894286565054,
+      "loss": 0.882,
+      "step": 2295
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019937832901269187,
+      "loss": 0.9997,
+      "step": 2300
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019936712969935948,
+      "loss": 0.98,
+      "step": 2305
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019935583072774039,
+      "loss": 1.0199,
+      "step": 2310
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001993444321091666,
+      "loss": 1.0151,
+      "step": 2315
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0001993329338550702,
+      "loss": 0.9283,
+      "step": 2320
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019932133597698313,
+      "loss": 0.9331,
+      "step": 2325
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0001993096384865372,
+      "loss": 0.9331,
+      "step": 2330
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019929784139546426,
+      "loss": 0.9636,
+      "step": 2335
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019928594471559587,
+      "loss": 1.0361,
+      "step": 2340
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019927394845886361,
+      "loss": 1.0051,
+      "step": 2345
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019926185263729896,
+      "loss": 0.9548,
+      "step": 2350
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00019924965726303304,
+      "loss": 0.8994,
+      "step": 2355
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001992373623482971,
+      "loss": 0.9032,
+      "step": 2360
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00019922496790542195,
+      "loss": 0.9512,
+      "step": 2365
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019921247394683844,
+      "loss": 0.9407,
+      "step": 2370
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00019919988048507713,
+      "loss": 1.025,
+      "step": 2375
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019918718753276834,
+      "loss": 0.9967,
+      "step": 2380
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00019917439510264227,
+      "loss": 0.9103,
+      "step": 2385
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001991615032075288,
+      "loss": 0.982,
+      "step": 2390
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019914851186035755,
+      "loss": 0.9849,
+      "step": 2395
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019913542107415801,
+      "loss": 0.9227,
+      "step": 2400
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00019912223086205932,
+      "loss": 0.9488,
+      "step": 2405
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00019910894123729032,
+      "loss": 0.9723,
+      "step": 2410
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019909555221317955,
+      "loss": 1.0172,
+      "step": 2415
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00019908206380315528,
+      "loss": 0.9484,
+      "step": 2420
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00019906847602074547,
+      "loss": 0.8921,
+      "step": 2425
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00019905478887957773,
+      "loss": 0.9606,
+      "step": 2430
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019904100239337923,
+      "loss": 0.9709,
+      "step": 2435
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019902711657597693,
+      "loss": 1.0452,
+      "step": 2440
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019901313144129727,
+      "loss": 0.9562,
+      "step": 2445
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019899904700336642,
+      "loss": 1.0301,
+      "step": 2450
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00019898486327631006,
+      "loss": 1.0033,
+      "step": 2455
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019897058027435347,
+      "loss": 1.0009,
+      "step": 2460
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00019895619801182148,
+      "loss": 1.0021,
+      "step": 2465
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019894171650313856,
+      "loss": 0.9669,
+      "step": 2470
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019892713576282863,
+      "loss": 1.0139,
+      "step": 2475
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00019891245580551509,
+      "loss": 0.9629,
+      "step": 2480
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.000198897676645921,
+      "loss": 0.9778,
+      "step": 2485
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019888279829886877,
+      "loss": 1.0,
+      "step": 2490
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019886782077928038,
+      "loss": 0.9069,
+      "step": 2495
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001988527441021772,
+      "loss": 0.9978,
+      "step": 2500
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001988375682826801,
+      "loss": 0.9191,
+      "step": 2505
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019882229333600932,
+      "loss": 0.9824,
+      "step": 2510
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019880691927748467,
+      "loss": 1.0422,
+      "step": 2515
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019879144612252515,
+      "loss": 0.938,
+      "step": 2520
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019877587388664934,
+      "loss": 0.9222,
+      "step": 2525
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.000198760202585475,
+      "loss": 0.9571,
+      "step": 2530
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019874443223471945,
+      "loss": 0.968,
+      "step": 2535
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0001987285628501992,
+      "loss": 0.9952,
+      "step": 2540
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019871259444783016,
+      "loss": 0.98,
+      "step": 2545
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001986965270436275,
+      "loss": 0.9747,
+      "step": 2550
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00019868036065370567,
+      "loss": 0.9699,
+      "step": 2555
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00019866409529427852,
+      "loss": 0.9908,
+      "step": 2560
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019864773098165898,
+      "loss": 1.029,
+      "step": 2565
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00019863126773225935,
+      "loss": 0.9535,
+      "step": 2570
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019861470556259113,
+      "loss": 0.9764,
+      "step": 2575
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019859804448926503,
+      "loss": 1.0262,
+      "step": 2580
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001985812845289909,
+      "loss": 0.9952,
+      "step": 2585
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001985644256985778,
+      "loss": 0.9349,
+      "step": 2590
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.000198547468014934,
+      "loss": 0.9153,
+      "step": 2595
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019853041149506687,
+      "loss": 0.9486,
+      "step": 2600
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001985132561560829,
+      "loss": 1.0409,
+      "step": 2605
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019849600201518768,
+      "loss": 1.0015,
+      "step": 2610
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019847864908968592,
+      "loss": 1.043,
+      "step": 2615
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001984611973969814,
+      "loss": 0.9974,
+      "step": 2620
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001984436469545769,
+      "loss": 0.9339,
+      "step": 2625
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00019842599778007437,
+      "loss": 0.9445,
+      "step": 2630
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00019840824989117464,
+      "loss": 0.9528,
+      "step": 2635
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001983904033056776,
+      "loss": 1.0025,
+      "step": 2640
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019837245804148218,
+      "loss": 0.9132,
+      "step": 2645
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00019835441411658616,
+      "loss": 1.0168,
+      "step": 2650
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00019833627154908635,
+      "loss": 1.026,
+      "step": 2655
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00019831803035717853,
+      "loss": 0.986,
+      "step": 2660
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00019829969055915728,
+      "loss": 0.9764,
+      "step": 2665
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019828125217341618,
+      "loss": 0.9559,
+      "step": 2670
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001982627152184476,
+      "loss": 0.8965,
+      "step": 2675
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00019824407971284286,
+      "loss": 0.9161,
+      "step": 2680
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.000198225345675292,
+      "loss": 0.9899,
+      "step": 2685
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019820651312458403,
+      "loss": 0.9215,
+      "step": 2690
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019818758207960663,
+      "loss": 0.9574,
+      "step": 2695
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019816855255934635,
+      "loss": 1.0076,
+      "step": 2700
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001981494245828884,
+      "loss": 1.008,
+      "step": 2705
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019813019816941689,
+      "loss": 1.0307,
+      "step": 2710
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019811087333821453,
+      "loss": 0.9577,
+      "step": 2715
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019809145010866276,
+      "loss": 0.9487,
+      "step": 2720
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019807192850024175,
+      "loss": 0.9503,
+      "step": 2725
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019805230853253033,
+      "loss": 0.9102,
+      "step": 2730
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019803259022520587,
+      "loss": 0.9899,
+      "step": 2735
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0001980127735980445,
+      "loss": 0.9321,
+      "step": 2740
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019799285867092096,
+      "loss": 0.9606,
+      "step": 2745
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00019797284546380848,
+      "loss": 0.9441,
+      "step": 2750
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019795273399677893,
+      "loss": 1.0307,
+      "step": 2755
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00019793252429000266,
+      "loss": 0.9548,
+      "step": 2760
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019791221636374865,
+      "loss": 0.9304,
+      "step": 2765
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001978918102383843,
+      "loss": 0.952,
+      "step": 2770
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019787130593437553,
+      "loss": 1.0108,
+      "step": 2775
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019785070347228673,
+      "loss": 0.951,
+      "step": 2780
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019783000287278078,
+      "loss": 1.0625,
+      "step": 2785
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019780920415661882,
+      "loss": 0.9996,
+      "step": 2790
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001977883073446606,
+      "loss": 0.9658,
+      "step": 2795
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00019776731245786414,
+      "loss": 1.027,
+      "step": 2800
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00019774621951728583,
+      "loss": 0.9845,
+      "step": 2805
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00019772502854408042,
+      "loss": 0.9474,
+      "step": 2810
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.000197703739559501,
+      "loss": 0.9708,
+      "step": 2815
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019768235258489888,
+      "loss": 0.9832,
+      "step": 2820
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019766086764172377,
+      "loss": 0.9671,
+      "step": 2825
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00019763928475152352,
+      "loss": 1.004,
+      "step": 2830
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00019761760393594425,
+      "loss": 0.9782,
+      "step": 2835
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019759582521673035,
+      "loss": 0.951,
+      "step": 2840
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019757394861572432,
+      "loss": 0.9236,
+      "step": 2845
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019755197415486685,
+      "loss": 1.0101,
+      "step": 2850
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00019752990185619682,
+      "loss": 0.9481,
+      "step": 2855
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00019750773174185123,
+      "loss": 0.9601,
+      "step": 2860
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019748546383406508,
+      "loss": 0.9757,
+      "step": 2865
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019746309815517153,
+      "loss": 0.9169,
+      "step": 2870
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0001974406347276019,
+      "loss": 0.9888,
+      "step": 2875
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001974180735738853,
+      "loss": 0.9517,
+      "step": 2880
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019739541471664907,
+      "loss": 0.9753,
+      "step": 2885
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00019737265817861845,
+      "loss": 0.9833,
+      "step": 2890
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019734980398261666,
+      "loss": 1.0542,
+      "step": 2895
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019732685215156483,
+      "loss": 0.9272,
+      "step": 2900
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00019730380270848209,
+      "loss": 0.8694,
+      "step": 2905
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00019728065567648536,
+      "loss": 0.9686,
+      "step": 2910
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00019725741107878958,
+      "loss": 0.9636,
+      "step": 2915
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019723406893870738,
+      "loss": 0.9482,
+      "step": 2920
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001972106292796493,
+      "loss": 0.949,
+      "step": 2925
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019718709212512373,
+      "loss": 0.9562,
+      "step": 2930
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019716345749873674,
+      "loss": 0.9978,
+      "step": 2935
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019713972542419227,
+      "loss": 0.9335,
+      "step": 2940
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019711589592529187,
+      "loss": 1.0099,
+      "step": 2945
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0001970919690259349,
+      "loss": 0.9736,
+      "step": 2950
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019706794475011835,
+      "loss": 1.0408,
+      "step": 2955
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019704382312193687,
+      "loss": 0.983,
+      "step": 2960
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019701960416558282,
+      "loss": 0.9751,
+      "step": 2965
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001969952879053461,
+      "loss": 0.9368,
+      "step": 2970
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019697087436561418,
+      "loss": 0.9285,
+      "step": 2975
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001969463635708722,
+      "loss": 0.9645,
+      "step": 2980
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001969217555457027,
+      "loss": 0.9815,
+      "step": 2985
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019689705031478586,
+      "loss": 1.0308,
+      "step": 2990
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019687224790289933,
+      "loss": 0.9747,
+      "step": 2995
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019684734833491811,
+      "loss": 0.9034,
+      "step": 3000
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001968223516358148,
+      "loss": 0.963,
+      "step": 3005
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001967972578306593,
+      "loss": 1.0035,
+      "step": 3010
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00019677206694461896,
+      "loss": 0.9707,
+      "step": 3015
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001967467790029585,
+      "loss": 1.0332,
+      "step": 3020
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001967213940310399,
+      "loss": 1.0415,
+      "step": 3025
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019669591205432254,
+      "loss": 0.9611,
+      "step": 3030
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001966703330983631,
+      "loss": 0.9171,
+      "step": 3035
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019664465718881543,
+      "loss": 0.9974,
+      "step": 3040
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00019661888435143073,
+      "loss": 0.9486,
+      "step": 3045
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00019659301461205728,
+      "loss": 0.9646,
+      "step": 3050
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001965670479966407,
+      "loss": 0.964,
+      "step": 3055
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001965409845312236,
+      "loss": 0.9701,
+      "step": 3060
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0001965148242419459,
+      "loss": 0.9235,
+      "step": 3065
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001964885671550445,
+      "loss": 0.9457,
+      "step": 3070
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019646221329685344,
+      "loss": 0.9532,
+      "step": 3075
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001964357626938038,
+      "loss": 0.9198,
+      "step": 3080
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00019640921537242365,
+      "loss": 0.984,
+      "step": 3085
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019638257135933814,
+      "loss": 0.932,
+      "step": 3090
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019635583068126935,
+      "loss": 1.0043,
+      "step": 3095
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001963289933650363,
+      "loss": 0.9815,
+      "step": 3100
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001963020594375549,
+      "loss": 0.9163,
+      "step": 3105
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019627502892583806,
+      "loss": 0.9536,
+      "step": 3110
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019624790185699548,
+      "loss": 0.954,
+      "step": 3115
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001962206782582337,
+      "loss": 0.9576,
+      "step": 3120
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001961933581568561,
+      "loss": 0.9739,
+      "step": 3125
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019616594158026283,
+      "loss": 0.8902,
+      "step": 3130
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001961384285559508,
+      "loss": 0.9157,
+      "step": 3135
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001961108191115136,
+      "loss": 0.9462,
+      "step": 3140
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019608311327464167,
+      "loss": 0.9893,
+      "step": 3145
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00019605531107312195,
+      "loss": 0.9836,
+      "step": 3150
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019602741253483817,
+      "loss": 0.8958,
+      "step": 3155
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019599941768777055,
+      "loss": 0.9296,
+      "step": 3160
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.000195971326559996,
+      "loss": 0.9593,
+      "step": 3165
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019594313917968795,
+      "loss": 0.8882,
+      "step": 3170
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019591485557511636,
+      "loss": 1.0049,
+      "step": 3175
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00019588647577464775,
+      "loss": 0.9233,
+      "step": 3180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.000195857999806745,
+      "loss": 0.9671,
+      "step": 3185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001958294276999676,
+      "loss": 0.9873,
+      "step": 3190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019580075948297135,
+      "loss": 0.9511,
+      "step": 3195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019577199518450847,
+      "loss": 1.0246,
+      "step": 3200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019574313483342748,
+      "loss": 0.9028,
+      "step": 3205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019571417845867337,
+      "loss": 0.9715,
+      "step": 3210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00019568512608928736,
+      "loss": 1.0416,
+      "step": 3215
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019565597775440688,
+      "loss": 0.9841,
+      "step": 3220
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019562673348326573,
+      "loss": 0.9476,
+      "step": 3225
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019559739330519388,
+      "loss": 0.9602,
+      "step": 3230
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019556795724961742,
+      "loss": 0.9761,
+      "step": 3235
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019553842534605868,
+      "loss": 0.9506,
+      "step": 3240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00019550879762413615,
+      "loss": 0.9009,
+      "step": 3245
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019547907411356427,
+      "loss": 0.9218,
+      "step": 3250
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019544925484415372,
+      "loss": 0.9365,
+      "step": 3255
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001954193398458111,
+      "loss": 1.0058,
+      "step": 3260
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001953893291485391,
+      "loss": 0.9287,
+      "step": 3265
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019535922278243634,
+      "loss": 0.989,
+      "step": 3270
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019532902077769735,
+      "loss": 0.9226,
+      "step": 3275
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019529872316461272,
+      "loss": 0.9946,
+      "step": 3280
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019526832997356875,
+      "loss": 1.0205,
+      "step": 3285
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00019523784123504775,
+      "loss": 0.9722,
+      "step": 3290
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019520725697962777,
+      "loss": 0.9066,
+      "step": 3295
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019517657723798268,
+      "loss": 0.9514,
+      "step": 3300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019514580204088212,
+      "loss": 0.9312,
+      "step": 3305
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00019511493141919145,
+      "loss": 0.9376,
+      "step": 3310
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00019508396540387178,
+      "loss": 0.9941,
+      "step": 3315
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001950529040259798,
+      "loss": 0.964,
+      "step": 3320
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019502174731666797,
+      "loss": 0.9354,
+      "step": 3325
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019499049530718424,
+      "loss": 0.8546,
+      "step": 3330
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019495914802887226,
+      "loss": 0.9735,
+      "step": 3335
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019492770551317106,
+      "loss": 0.9428,
+      "step": 3340
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00019489616779161533,
+      "loss": 1.007,
+      "step": 3345
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00019486453489583525,
+      "loss": 0.9552,
+      "step": 3350
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001948328068575563,
+      "loss": 0.9544,
+      "step": 3355
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001948009837085996,
+      "loss": 0.9965,
+      "step": 3360
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019476906548088148,
+      "loss": 0.9454,
+      "step": 3365
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019473705220641367,
+      "loss": 1.0041,
+      "step": 3370
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001947049439173033,
+      "loss": 0.9858,
+      "step": 3375
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019467274064575275,
+      "loss": 0.9584,
+      "step": 3380
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001946404424240596,
+      "loss": 0.9498,
+      "step": 3385
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001946080492846167,
+      "loss": 0.9957,
+      "step": 3390
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019457556125991216,
+      "loss": 0.9955,
+      "step": 3395
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00019454297838252918,
+      "loss": 0.9109,
+      "step": 3400
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019451030068514609,
+      "loss": 0.9387,
+      "step": 3405
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019447752820053634,
+      "loss": 0.9458,
+      "step": 3410
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019444466096156846,
+      "loss": 0.9691,
+      "step": 3415
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019441169900120598,
+      "loss": 0.9299,
+      "step": 3420
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019437864235250744,
+      "loss": 0.9265,
+      "step": 3425
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00019434549104862639,
+      "loss": 1.0198,
+      "step": 3430
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001943122451228112,
+      "loss": 0.9645,
+      "step": 3435
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00019427890460840526,
+      "loss": 1.0366,
+      "step": 3440
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019424546953884675,
+      "loss": 0.9585,
+      "step": 3445
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019421193994766873,
+      "loss": 0.9712,
+      "step": 3450
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.000194178315868499,
+      "loss": 1.0027,
+      "step": 3455
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019414459733506023,
+      "loss": 0.9572,
+      "step": 3460
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019411078438116969,
+      "loss": 0.9517,
+      "step": 3465
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019407687704073943,
+      "loss": 0.9082,
+      "step": 3470
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019404287534777615,
+      "loss": 0.9686,
+      "step": 3475
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019400877933638114,
+      "loss": 0.9654,
+      "step": 3480
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019397458904075036,
+      "loss": 1.0102,
+      "step": 3485
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019394030449517428,
+      "loss": 0.8971,
+      "step": 3490
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019390592573403787,
+      "loss": 1.029,
+      "step": 3495
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001938714527918207,
+      "loss": 0.9432,
+      "step": 3500
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001938368857030966,
+      "loss": 1.0492,
+      "step": 3505
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019380222450253405,
+      "loss": 1.0287,
+      "step": 3510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019376746922489577,
+      "loss": 0.9945,
+      "step": 3515
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019373261990503888,
+      "loss": 0.9362,
+      "step": 3520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019369767657791479,
+      "loss": 0.9723,
+      "step": 3525
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019366263927856928,
+      "loss": 0.986,
+      "step": 3530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00019362750804214222,
+      "loss": 0.9137,
+      "step": 3535
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001935922829038679,
+      "loss": 0.8947,
+      "step": 3540
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019355696389907455,
+      "loss": 0.9909,
+      "step": 3545
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019352155106318471,
+      "loss": 0.9157,
+      "step": 3550
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019348604443171502,
+      "loss": 0.9311,
+      "step": 3555
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019345044404027613,
+      "loss": 1.043,
+      "step": 3560
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001934147499245727,
+      "loss": 0.9324,
+      "step": 3565
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001933789621204035,
+      "loss": 0.9573,
+      "step": 3570
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019334308066366114,
+      "loss": 0.9804,
+      "step": 3575
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019330710559033225,
+      "loss": 0.948,
+      "step": 3580
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001932710369364973,
+      "loss": 0.9988,
+      "step": 3585
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019323487473833062,
+      "loss": 0.9151,
+      "step": 3590
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001931986190321004,
+      "loss": 0.9649,
+      "step": 3595
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019316226985416853,
+      "loss": 0.9589,
+      "step": 3600
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019312582724099076,
+      "loss": 0.9673,
+      "step": 3605
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00019308929122911642,
+      "loss": 0.8674,
+      "step": 3610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001930526618551886,
+      "loss": 1.0047,
+      "step": 3615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019301593915594403,
+      "loss": 0.9402,
+      "step": 3620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019297912316821298,
+      "loss": 0.9739,
+      "step": 3625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019294221392891932,
+      "loss": 0.9754,
+      "step": 3630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019290521147508042,
+      "loss": 0.9692,
+      "step": 3635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001928681158438072,
+      "loss": 1.0017,
+      "step": 3640
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019283092707230392,
+      "loss": 1.0303,
+      "step": 3645
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001927936451978684,
+      "loss": 1.0064,
+      "step": 3650
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001927562702578917,
+      "loss": 0.9485,
+      "step": 3655
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019271880228985828,
+      "loss": 0.9458,
+      "step": 3660
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019268124133134588,
+      "loss": 0.9705,
+      "step": 3665
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019264358742002556,
+      "loss": 0.9049,
+      "step": 3670
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019260584059366153,
+      "loss": 1.0574,
+      "step": 3675
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019256800089011123,
+      "loss": 1.0533,
+      "step": 3680
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0001925300683473252,
+      "loss": 0.9216,
+      "step": 3685
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001924920430033472,
+      "loss": 0.9334,
+      "step": 3690
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019245392489631392,
+      "loss": 0.9523,
+      "step": 3695
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019241571406445525,
+      "loss": 0.9449,
+      "step": 3700
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019237741054609387,
+      "loss": 0.9897,
+      "step": 3705
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019233901437964562,
+      "loss": 0.9448,
+      "step": 3710
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00019230052560361913,
+      "loss": 0.9926,
+      "step": 3715
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019226194425661598,
+      "loss": 0.9142,
+      "step": 3720
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019222327037733052,
+      "loss": 1.0046,
+      "step": 3725
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00019218450400454998,
+      "loss": 0.9166,
+      "step": 3730
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019214564517715433,
+      "loss": 0.9597,
+      "step": 3735
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019210669393411624,
+      "loss": 0.9981,
+      "step": 3740
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019206765031450112,
+      "loss": 0.9568,
+      "step": 3745
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00019202851435746695,
+      "loss": 0.9513,
+      "step": 3750
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019198928610226435,
+      "loss": 0.943,
+      "step": 3755
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019194996558823655,
+      "loss": 1.0194,
+      "step": 3760
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019191055285481927,
+      "loss": 1.0096,
+      "step": 3765
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019187104794154074,
+      "loss": 0.9833,
+      "step": 3770
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019183145088802158,
+      "loss": 0.9513,
+      "step": 3775
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019179176173397494,
+      "loss": 0.9946,
+      "step": 3780
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0001917519805192062,
+      "loss": 0.9526,
+      "step": 3785
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019171210728361317,
+      "loss": 1.0067,
+      "step": 3790
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019167214206718594,
+      "loss": 0.9788,
+      "step": 3795
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001916320849100068,
+      "loss": 0.9777,
+      "step": 3800
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019159193585225026,
+      "loss": 0.9702,
+      "step": 3805
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019155169493418304,
+      "loss": 0.952,
+      "step": 3810
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.000191511362196164,
+      "loss": 0.9244,
+      "step": 3815
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00019147093767864402,
+      "loss": 0.9819,
+      "step": 3820
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019143042142216607,
+      "loss": 0.9759,
+      "step": 3825
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019138981346736514,
+      "loss": 0.8902,
+      "step": 3830
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019134911385496815,
+      "loss": 0.9328,
+      "step": 3835
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019130832262579398,
+      "loss": 0.8924,
+      "step": 3840
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019126743982075337,
+      "loss": 0.9471,
+      "step": 3845
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019122646548084892,
+      "loss": 0.9705,
+      "step": 3850
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019118539964717505,
+      "loss": 1.0149,
+      "step": 3855
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001911442423609179,
+      "loss": 0.9364,
+      "step": 3860
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00019110299366335536,
+      "loss": 0.9572,
+      "step": 3865
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00019106165359585698,
+      "loss": 0.9038,
+      "step": 3870
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00019102022219988398,
+      "loss": 0.9761,
+      "step": 3875
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00019097869951698913,
+      "loss": 1.0293,
+      "step": 3880
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001909370855888168,
+      "loss": 0.8963,
+      "step": 3885
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019089538045710284,
+      "loss": 0.9691,
+      "step": 3890
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019085358416367457,
+      "loss": 0.9661,
+      "step": 3895
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001908116967504508,
+      "loss": 0.9243,
+      "step": 3900
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019076971825944164,
+      "loss": 0.9582,
+      "step": 3905
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019072764873274856,
+      "loss": 0.973,
+      "step": 3910
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001906854882125644,
+      "loss": 0.9518,
+      "step": 3915
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019064323674117318,
+      "loss": 0.998,
+      "step": 3920
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001906008943609502,
+      "loss": 0.8808,
+      "step": 3925
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001905584611143619,
+      "loss": 0.9755,
+      "step": 3930
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019051593704396587,
+      "loss": 0.9383,
+      "step": 3935
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019047332219241078,
+      "loss": 0.9663,
+      "step": 3940
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019043061660243632,
+      "loss": 0.9642,
+      "step": 3945
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019038782031687325,
+      "loss": 1.0157,
+      "step": 3950
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001903449333786432,
+      "loss": 0.9732,
+      "step": 3955
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019030195583075881,
+      "loss": 0.9632,
+      "step": 3960
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019025888771632355,
+      "loss": 0.9818,
+      "step": 3965
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019021572907853177,
+      "loss": 0.9512,
+      "step": 3970
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019017247996066852,
+      "loss": 0.8931,
+      "step": 3975
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019012914040610963,
+      "loss": 1.0358,
+      "step": 3980
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019008571045832167,
+      "loss": 0.9163,
+      "step": 3985
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019004219016086188,
+      "loss": 0.9641,
+      "step": 3990
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00018999857955737798,
+      "loss": 0.9079,
+      "step": 3995
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00018995487869160845,
+      "loss": 0.9652,
+      "step": 4000
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00018991108760738214,
+      "loss": 0.8911,
+      "step": 4005
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00018986720634861848,
+      "loss": 0.9227,
+      "step": 4010
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018982323495932732,
+      "loss": 0.9767,
+      "step": 4015
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00018977917348360888,
+      "loss": 0.9541,
+      "step": 4020
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001897350219656537,
+      "loss": 0.9738,
+      "step": 4025
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001896907804497427,
+      "loss": 0.9014,
+      "step": 4030
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00018964644898024707,
+      "loss": 1.0062,
+      "step": 4035
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001896020276016281,
+      "loss": 0.9882,
+      "step": 4040
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00018955751635843737,
+      "loss": 0.9511,
+      "step": 4045
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001895129152953165,
+      "loss": 0.8994,
+      "step": 4050
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00018946822445699735,
+      "loss": 0.9227,
+      "step": 4055
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00018942344388830158,
+      "loss": 0.9347,
+      "step": 4060
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00018937857363414106,
+      "loss": 0.9333,
+      "step": 4065
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00018933361373951746,
+      "loss": 0.9831,
+      "step": 4070
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018928856424952245,
+      "loss": 0.8975,
+      "step": 4075
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001892434252093375,
+      "loss": 1.0322,
+      "step": 4080
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00018919819666423396,
+      "loss": 0.9409,
+      "step": 4085
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018915287865957277,
+      "loss": 1.0335,
+      "step": 4090
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001891074712408049,
+      "loss": 0.9877,
+      "step": 4095
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018906197445347068,
+      "loss": 0.9743,
+      "step": 4100
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001890163883432003,
+      "loss": 0.935,
+      "step": 4105
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00018897071295571335,
+      "loss": 0.9519,
+      "step": 4110
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00018892494833681913,
+      "loss": 0.9171,
+      "step": 4115
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00018887909453241632,
+      "loss": 0.9584,
+      "step": 4120
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001888331515884931,
+      "loss": 0.9841,
+      "step": 4125
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.000188787119551127,
+      "loss": 1.0196,
+      "step": 4130
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018874099846648496,
+      "loss": 0.9448,
+      "step": 4135
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001886947883808232,
+      "loss": 0.9824,
+      "step": 4140
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001886484893404872,
+      "loss": 0.9046,
+      "step": 4145
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001886021013919117,
+      "loss": 0.9443,
+      "step": 4150
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00018855562458162055,
+      "loss": 0.9724,
+      "step": 4155
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0001885090589562267,
+      "loss": 0.9864,
+      "step": 4160
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018846240456243225,
+      "loss": 0.9661,
+      "step": 4165
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00018841566144702833,
+      "loss": 0.9449,
+      "step": 4170
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00018836882965689493,
+      "loss": 0.9158,
+      "step": 4175
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018832190923900112,
+      "loss": 0.9386,
+      "step": 4180
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018827490024040484,
+      "loss": 0.9941,
+      "step": 4185
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018822780270825277,
+      "loss": 0.948,
+      "step": 4190
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00018818061668978046,
+      "loss": 1.0238,
+      "step": 4195
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001881333422323122,
+      "loss": 0.9964,
+      "step": 4200
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018808597938326093,
+      "loss": 0.9123,
+      "step": 4205
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00018803852819012832,
+      "loss": 0.9153,
+      "step": 4210
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001879909887005046,
+      "loss": 0.956,
+      "step": 4215
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00018794336096206852,
+      "loss": 0.9875,
+      "step": 4220
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00018789564502258741,
+      "loss": 0.9332,
+      "step": 4225
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.000187847840929917,
+      "loss": 1.0428,
+      "step": 4230
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00018779994873200146,
+      "loss": 0.8838,
+      "step": 4235
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00018775196847687332,
+      "loss": 0.9732,
+      "step": 4240
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0001877039002126534,
+      "loss": 0.9697,
+      "step": 4245
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00018765574398755085,
+      "loss": 1.0307,
+      "step": 4250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00018760749984986298,
+      "loss": 0.9221,
+      "step": 4255
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001875591678479753,
+      "loss": 0.911,
+      "step": 4260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00018751074803036142,
+      "loss": 0.9499,
+      "step": 4265
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00018746224044558302,
+      "loss": 0.931,
+      "step": 4270
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001874136451422898,
+      "loss": 0.9934,
+      "step": 4275
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001873649621692195,
+      "loss": 1.0263,
+      "step": 4280
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00018731619157519774,
+      "loss": 0.9163,
+      "step": 4285
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00018726733340913797,
+      "loss": 0.9006,
+      "step": 4290
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018721838772004157,
+      "loss": 0.9588,
+      "step": 4295
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001871693545569976,
+      "loss": 0.972,
+      "step": 4300
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00018712023396918293,
+      "loss": 0.8951,
+      "step": 4305
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001870710260058621,
+      "loss": 0.9148,
+      "step": 4310
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00018702173071638716,
+      "loss": 0.8892,
+      "step": 4315
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018697234815019792,
+      "loss": 1.0288,
+      "step": 4320
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018692287835682164,
+      "loss": 1.0217,
+      "step": 4325
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018687332138587302,
+      "loss": 0.9623,
+      "step": 4330
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001868236772870543,
+      "loss": 0.9821,
+      "step": 4335
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00018677394611015498,
+      "loss": 0.9868,
+      "step": 4340
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.000186724127905052,
+      "loss": 0.9652,
+      "step": 4345
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018667422272170955,
+      "loss": 0.9466,
+      "step": 4350
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018662423061017896,
+      "loss": 1.0052,
+      "step": 4355
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00018657415162059892,
+      "loss": 0.9959,
+      "step": 4360
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001865239858031951,
+      "loss": 1.0139,
+      "step": 4365
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018647373320828035,
+      "loss": 0.9677,
+      "step": 4370
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00018642339388625444,
+      "loss": 0.9707,
+      "step": 4375
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001863729678876043,
+      "loss": 0.9591,
+      "step": 4380
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018632245526290352,
+      "loss": 1.0214,
+      "step": 4385
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001862718560628129,
+      "loss": 0.9707,
+      "step": 4390
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018622117033807985,
+      "loss": 0.9585,
+      "step": 4395
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00018617039813953855,
+      "loss": 0.8711,
+      "step": 4400
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00018611953951811004,
+      "loss": 1.0133,
+      "step": 4405
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00018606859452480193,
+      "loss": 0.9349,
+      "step": 4410
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00018601756321070845,
+      "loss": 0.979,
+      "step": 4415
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001859664456270105,
+      "loss": 0.9793,
+      "step": 4420
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00018591524182497547,
+      "loss": 0.8638,
+      "step": 4425
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001858639518559571,
+      "loss": 0.9971,
+      "step": 4430
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018581257577139572,
+      "loss": 0.9375,
+      "step": 4435
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018576111362281794,
+      "loss": 1.0139,
+      "step": 4440
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00018570956546183666,
+      "loss": 1.0013,
+      "step": 4445
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00018565793134015115,
+      "loss": 0.9572,
+      "step": 4450
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018560621130954674,
+      "loss": 1.0208,
+      "step": 4455
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00018555440542189508,
+      "loss": 0.9392,
+      "step": 4460
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00018550251372915382,
+      "loss": 1.012,
+      "step": 4465
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00018545053628336668,
+      "loss": 0.924,
+      "step": 4470
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00018539847313666345,
+      "loss": 0.9599,
+      "step": 4475
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00018534632434125982,
+      "loss": 0.9028,
+      "step": 4480
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00018529408994945738,
+      "loss": 1.0321,
+      "step": 4485
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001852417700136436,
+      "loss": 0.9446,
+      "step": 4490
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00018518936458629165,
+      "loss": 0.9477,
+      "step": 4495
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00018513687371996058,
+      "loss": 0.9671,
+      "step": 4500
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.000185084297467295,
+      "loss": 0.8988,
+      "step": 4505
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001850316358810253,
+      "loss": 0.9879,
+      "step": 4510
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0001849788890139673,
+      "loss": 0.9617,
+      "step": 4515
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00018492605691902242,
+      "loss": 0.9869,
+      "step": 4520
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00018487313964917761,
+      "loss": 0.9515,
+      "step": 4525
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00018482013725750512,
+      "loss": 0.9748,
+      "step": 4530
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00018476704979716275,
+      "loss": 1.0375,
+      "step": 4535
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00018471387732139344,
+      "loss": 0.9816,
+      "step": 4540
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00018466061988352546,
+      "loss": 0.999,
+      "step": 4545
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018460727753697234,
+      "loss": 0.9573,
+      "step": 4550
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00018455385033523268,
+      "loss": 0.9211,
+      "step": 4555
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018450033833189027,
+      "loss": 0.9198,
+      "step": 4560
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001844467415806139,
+      "loss": 0.9629,
+      "step": 4565
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00018439306013515733,
+      "loss": 0.9674,
+      "step": 4570
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018433929404935935,
+      "loss": 0.9872,
+      "step": 4575
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00018428544337714358,
+      "loss": 1.0281,
+      "step": 4580
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018423150817251845,
+      "loss": 0.9846,
+      "step": 4585
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001841774884895772,
+      "loss": 0.9727,
+      "step": 4590
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00018412338438249782,
+      "loss": 0.9988,
+      "step": 4595
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018406919590554296,
+      "loss": 0.9363,
+      "step": 4600
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018401492311305985,
+      "loss": 0.9301,
+      "step": 4605
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018396056605948032,
+      "loss": 0.9556,
+      "step": 4610
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00018390612479932066,
+      "loss": 1.0037,
+      "step": 4615
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018385159938718172,
+      "loss": 1.0031,
+      "step": 4620
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00018379698987774858,
+      "loss": 0.9881,
+      "step": 4625
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018374229632579087,
+      "loss": 0.9284,
+      "step": 4630
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018368751878616234,
+      "loss": 0.9625,
+      "step": 4635
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00018363265731380102,
+      "loss": 0.9116,
+      "step": 4640
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018357771196372916,
+      "loss": 0.9498,
+      "step": 4645
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018352268279105314,
+      "loss": 0.9198,
+      "step": 4650
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001834675698509633,
+      "loss": 0.9423,
+      "step": 4655
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001834123731987341,
+      "loss": 0.9621,
+      "step": 4660
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018335709288972395,
+      "loss": 0.9525,
+      "step": 4665
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00018330172897937513,
+      "loss": 0.9146,
+      "step": 4670
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018324628152321373,
+      "loss": 0.9209,
+      "step": 4675
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018319075057684968,
+      "loss": 1.0023,
+      "step": 4680
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001831351361959767,
+      "loss": 0.9152,
+      "step": 4685
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001830794384363721,
+      "loss": 0.9415,
+      "step": 4690
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018302365735389678,
+      "loss": 1.0131,
+      "step": 4695
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018296779300449535,
+      "loss": 0.9834,
+      "step": 4700
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00018291184544419578,
+      "loss": 0.8716,
+      "step": 4705
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018285581472910964,
+      "loss": 1.0339,
+      "step": 4710
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00018279970091543174,
+      "loss": 0.9977,
+      "step": 4715
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001827435040594404,
+      "loss": 0.9412,
+      "step": 4720
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018268722421749703,
+      "loss": 1.0184,
+      "step": 4725
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001826308614460465,
+      "loss": 0.9689,
+      "step": 4730
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00018257441580161663,
+      "loss": 0.9036,
+      "step": 4735
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00018251788734081849,
+      "loss": 0.9537,
+      "step": 4740
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001824612761203462,
+      "loss": 0.9667,
+      "step": 4745
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00018240458219697685,
+      "loss": 1.0325,
+      "step": 4750
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00018234780562757045,
+      "loss": 0.96,
+      "step": 4755
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018229094646906997,
+      "loss": 0.9985,
+      "step": 4760
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00018223400477850117,
+      "loss": 0.9671,
+      "step": 4765
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018217698061297254,
+      "loss": 0.9163,
+      "step": 4770
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018211987402967536,
+      "loss": 0.9751,
+      "step": 4775
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001820626850858836,
+      "loss": 0.9423,
+      "step": 4780
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00018200541383895367,
+      "loss": 0.9343,
+      "step": 4785
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001819480603463247,
+      "loss": 0.9399,
+      "step": 4790
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018189062466551824,
+      "loss": 1.035,
+      "step": 4795
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001818331068541382,
+      "loss": 0.9444,
+      "step": 4800
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018177550696987096,
+      "loss": 0.9632,
+      "step": 4805
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001817178250704852,
+      "loss": 0.913,
+      "step": 4810
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00018166006121383185,
+      "loss": 0.918,
+      "step": 4815
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018160221545784392,
+      "loss": 0.958,
+      "step": 4820
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018154428786053677,
+      "loss": 1.1622,
+      "step": 4825
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00018148627848000768,
+      "loss": 0.9385,
+      "step": 4830
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00018142818737443603,
+      "loss": 0.9235,
+      "step": 4835
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00018137001460208309,
+      "loss": 1.0112,
+      "step": 4840
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018131176022129214,
+      "loss": 0.9345,
+      "step": 4845
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00018125342429048825,
+      "loss": 0.9295,
+      "step": 4850
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018119500686817824,
+      "loss": 0.9728,
+      "step": 4855
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00018113650801295073,
+      "loss": 0.9295,
+      "step": 4860
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.000181077927783476,
+      "loss": 0.9564,
+      "step": 4865
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00018101926623850586,
+      "loss": 0.9301,
+      "step": 4870
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00018096052343687382,
+      "loss": 0.9672,
+      "step": 4875
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018090169943749476,
+      "loss": 0.9522,
+      "step": 4880
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018084279429936504,
+      "loss": 1.0778,
+      "step": 4885
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00018078380808156245,
+      "loss": 0.9789,
+      "step": 4890
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00018072474084324593,
+      "loss": 0.9687,
+      "step": 4895
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00018066559264365593,
+      "loss": 1.0009,
+      "step": 4900
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018060636354211385,
+      "loss": 0.9497,
+      "step": 4905
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001805470535980224,
+      "loss": 0.9586,
+      "step": 4910
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001804876628708653,
+      "loss": 0.9354,
+      "step": 4915
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018042819142020727,
+      "loss": 0.9981,
+      "step": 4920
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018036863930569408,
+      "loss": 0.9645,
+      "step": 4925
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018030900658705227,
+      "loss": 0.9754,
+      "step": 4930
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00018024929332408933,
+      "loss": 0.9265,
+      "step": 4935
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00018018949957669347,
+      "loss": 0.9608,
+      "step": 4940
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00018012962540483364,
+      "loss": 0.9572,
+      "step": 4945
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018006967086855948,
+      "loss": 0.9657,
+      "step": 4950
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00018000963602800117,
+      "loss": 1.0154,
+      "step": 4955
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017994952094336946,
+      "loss": 0.9399,
+      "step": 4960
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001798893256749556,
+      "loss": 0.9801,
+      "step": 4965
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001798290502831312,
+      "loss": 0.9232,
+      "step": 4970
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001797686948283483,
+      "loss": 0.9539,
+      "step": 4975
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017970825937113923,
+      "loss": 0.9298,
+      "step": 4980
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017964774397211643,
+      "loss": 0.9667,
+      "step": 4985
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00017958714869197273,
+      "loss": 0.9768,
+      "step": 4990
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017952647359148087,
+      "loss": 0.9718,
+      "step": 4995
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00017946571873149377,
+      "loss": 0.9546,
+      "step": 5000
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017940488417294437,
+      "loss": 1.0762,
+      "step": 5005
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00017934396997684537,
+      "loss": 0.9056,
+      "step": 5010
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00017928297620428953,
+      "loss": 0.9799,
+      "step": 5015
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00017922190291644934,
+      "loss": 0.9785,
+      "step": 5020
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017916075017457698,
+      "loss": 0.9463,
+      "step": 5025
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00017909951804000445,
+      "loss": 0.9665,
+      "step": 5030
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001790382065741432,
+      "loss": 0.9876,
+      "step": 5035
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00017897681583848449,
+      "loss": 0.9929,
+      "step": 5040
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00017891534589459883,
+      "loss": 0.991,
+      "step": 5045
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00017885379680413627,
+      "loss": 0.9451,
+      "step": 5050
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001787921686288263,
+      "loss": 0.9316,
+      "step": 5055
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00017873046143047767,
+      "loss": 0.9187,
+      "step": 5060
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017866867527097837,
+      "loss": 0.9836,
+      "step": 5065
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001786068102122956,
+      "loss": 0.9655,
+      "step": 5070
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00017854486631647569,
+      "loss": 0.9407,
+      "step": 5075
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017848284364564406,
+      "loss": 0.9331,
+      "step": 5080
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00017842074226200505,
+      "loss": 1.0075,
+      "step": 5085
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001783585622278421,
+      "loss": 0.8936,
+      "step": 5090
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00017829630360551737,
+      "loss": 0.963,
+      "step": 5095
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001782339664574719,
+      "loss": 0.9445,
+      "step": 5100
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00017817155084622562,
+      "loss": 0.9558,
+      "step": 5105
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00017810905683437683,
+      "loss": 0.9419,
+      "step": 5110
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0001780464844846028,
+      "loss": 1.0337,
+      "step": 5115
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00017798383385965918,
+      "loss": 0.9724,
+      "step": 5120
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00017792110502238016,
+      "loss": 0.971,
+      "step": 5125
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001778582980356784,
+      "loss": 0.9743,
+      "step": 5130
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00017779541296254487,
+      "loss": 1.0086,
+      "step": 5135
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00017773244986604895,
+      "loss": 0.9543,
+      "step": 5140
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00017766940880933825,
+      "loss": 0.9424,
+      "step": 5145
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017760628985563845,
+      "loss": 0.9988,
+      "step": 5150
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00017754309306825357,
+      "loss": 0.8561,
+      "step": 5155
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017747981851056548,
+      "loss": 0.9413,
+      "step": 5160
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00017741646624603417,
+      "loss": 0.9704,
+      "step": 5165
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00017735303633819753,
+      "loss": 1.0081,
+      "step": 5170
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00017728952885067133,
+      "loss": 0.9601,
+      "step": 5175
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017722594384714916,
+      "loss": 0.9446,
+      "step": 5180
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00017716228139140228,
+      "loss": 0.9643,
+      "step": 5185
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00017709854154727975,
+      "loss": 0.9975,
+      "step": 5190
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00017703472437870813,
+      "loss": 1.0159,
+      "step": 5195
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00017697082994969158,
+      "loss": 1.0124,
+      "step": 5200
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001769068583243118,
+      "loss": 0.9084,
+      "step": 5205
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001768428095667278,
+      "loss": 0.8991,
+      "step": 5210
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00017677868374117606,
+      "loss": 0.9662,
+      "step": 5215
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00017671448091197026,
+      "loss": 0.9955,
+      "step": 5220
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00017665020114350136,
+      "loss": 0.9421,
+      "step": 5225
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00017658584450023747,
+      "loss": 0.9946,
+      "step": 5230
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001765214110467238,
+      "loss": 0.9487,
+      "step": 5235
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00017645690084758267,
+      "loss": 0.9412,
+      "step": 5240
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00017639231396751322,
+      "loss": 0.9802,
+      "step": 5245
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017632765047129157,
+      "loss": 0.9389,
+      "step": 5250
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00017626291042377077,
+      "loss": 0.9462,
+      "step": 5255
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00017619809388988049,
+      "loss": 0.9653,
+      "step": 5260
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00017613320093462723,
+      "loss": 0.9345,
+      "step": 5265
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00017606823162309406,
+      "loss": 0.9675,
+      "step": 5270
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00017600318602044066,
+      "loss": 1.0095,
+      "step": 5275
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00017593806419190325,
+      "loss": 0.9232,
+      "step": 5280
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00017587286620279443,
+      "loss": 0.9346,
+      "step": 5285
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00017580759211850323,
+      "loss": 0.9148,
+      "step": 5290
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00017574224200449506,
+      "loss": 0.9724,
+      "step": 5295
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00017567681592631145,
+      "loss": 0.958,
+      "step": 5300
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00017561131394957022,
+      "loss": 0.9913,
+      "step": 5305
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00017554573613996524,
+      "loss": 0.9496,
+      "step": 5310
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017548008256326655,
+      "loss": 0.9953,
+      "step": 5315
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00017541435328531996,
+      "loss": 0.9762,
+      "step": 5320
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00017534854837204745,
+      "loss": 0.9767,
+      "step": 5325
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017528266788944676,
+      "loss": 0.9449,
+      "step": 5330
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00017521671190359132,
+      "loss": 0.9,
+      "step": 5335
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00017515068048063048,
+      "loss": 0.9287,
+      "step": 5340
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00017508457368678904,
+      "loss": 0.9636,
+      "step": 5345
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00017501839158836756,
+      "loss": 0.9587,
+      "step": 5350
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00017495213425174205,
+      "loss": 0.9104,
+      "step": 5355
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.000174885801743364,
+      "loss": 0.9747,
+      "step": 5360
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00017481939412976024,
+      "loss": 0.9149,
+      "step": 5365
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00017475291147753299,
+      "loss": 0.9608,
+      "step": 5370
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001746863538533597,
+      "loss": 0.9185,
+      "step": 5375
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.000174619721323993,
+      "loss": 0.9215,
+      "step": 5380
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001745530139562607,
+      "loss": 0.929,
+      "step": 5385
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001744862318170656,
+      "loss": 0.9378,
+      "step": 5390
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00017441937497338552,
+      "loss": 0.9807,
+      "step": 5395
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001743524434922732,
+      "loss": 0.9641,
+      "step": 5400
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017428543744085623,
+      "loss": 1.0143,
+      "step": 5405
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00017421835688633704,
+      "loss": 0.9707,
+      "step": 5410
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001741512018959927,
+      "loss": 1.0347,
+      "step": 5415
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00017408397253717496,
+      "loss": 1.0184,
+      "step": 5420
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001740166688773102,
+      "loss": 0.9867,
+      "step": 5425
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00017394929098389929,
+      "loss": 0.986,
+      "step": 5430
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00017388183892451755,
+      "loss": 0.9799,
+      "step": 5435
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00017381431276681464,
+      "loss": 0.9343,
+      "step": 5440
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001737467125785146,
+      "loss": 0.9886,
+      "step": 5445
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001736790384274157,
+      "loss": 1.0098,
+      "step": 5450
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00017361129038139038,
+      "loss": 0.9989,
+      "step": 5455
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001735434685083852,
+      "loss": 0.9337,
+      "step": 5460
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00017347557287642076,
+      "loss": 1.0127,
+      "step": 5465
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00017340760355359161,
+      "loss": 0.9243,
+      "step": 5470
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001733395606080663,
+      "loss": 0.9355,
+      "step": 5475
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00017327144410808707,
+      "loss": 0.9836,
+      "step": 5480
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017320325412197,
+      "loss": 0.8971,
+      "step": 5485
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00017313499071810497,
+      "loss": 0.937,
+      "step": 5490
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00017306665396495534,
+      "loss": 1.0036,
+      "step": 5495
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001729982439310581,
+      "loss": 0.9265,
+      "step": 5500
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017292976068502376,
+      "loss": 0.9741,
+      "step": 5505
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0001728612042955362,
+      "loss": 1.0222,
+      "step": 5510
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00017279257483135272,
+      "loss": 0.9135,
+      "step": 5515
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017272387236130383,
+      "loss": 0.983,
+      "step": 5520
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00017265509695429335,
+      "loss": 0.9202,
+      "step": 5525
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00017258624867929817,
+      "loss": 0.9667,
+      "step": 5530
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00017251732760536833,
+      "loss": 0.9566,
+      "step": 5535
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00017244833380162687,
+      "loss": 0.9045,
+      "step": 5540
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001723792673372697,
+      "loss": 0.9259,
+      "step": 5545
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00017231012828156566,
+      "loss": 0.9829,
+      "step": 5550
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00017224091670385642,
+      "loss": 0.8636,
+      "step": 5555
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017217163267355638,
+      "loss": 0.9075,
+      "step": 5560
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00017210227626015252,
+      "loss": 0.9639,
+      "step": 5565
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00017203284753320447,
+      "loss": 0.9425,
+      "step": 5570
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00017196334656234446,
+      "loss": 0.9542,
+      "step": 5575
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00017189377341727708,
+      "loss": 0.9048,
+      "step": 5580
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00017182412816777931,
+      "loss": 0.924,
+      "step": 5585
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00017175441088370045,
+      "loss": 0.9078,
+      "step": 5590
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00017168462163496214,
+      "loss": 0.9432,
+      "step": 5595
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017161476049155807,
+      "loss": 0.9207,
+      "step": 5600
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00017154482752355406,
+      "loss": 1.008,
+      "step": 5605
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00017147482280108802,
+      "loss": 0.9298,
+      "step": 5610
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017140474639436981,
+      "loss": 0.9983,
+      "step": 5615
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001713345983736811,
+      "loss": 0.9208,
+      "step": 5620
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00017126437880937557,
+      "loss": 0.9903,
+      "step": 5625
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00017119408777187842,
+      "loss": 0.959,
+      "step": 5630
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017112372533168672,
+      "loss": 0.9756,
+      "step": 5635
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00017105329155936905,
+      "loss": 0.9773,
+      "step": 5640
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001709827865255656,
+      "loss": 0.9547,
+      "step": 5645
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001709122103009879,
+      "loss": 1.0082,
+      "step": 5650
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00017084156295641906,
+      "loss": 1.0108,
+      "step": 5655
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001707708445627134,
+      "loss": 0.9202,
+      "step": 5660
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00017070005519079652,
+      "loss": 0.9213,
+      "step": 5665
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00017062919491166523,
+      "loss": 0.9814,
+      "step": 5670
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00017055826379638742,
+      "loss": 0.9815,
+      "step": 5675
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017048726191610202,
+      "loss": 0.9683,
+      "step": 5680
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00017041618934201904,
+      "loss": 0.9397,
+      "step": 5685
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001703450461454192,
+      "loss": 1.0024,
+      "step": 5690
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017027383239765422,
+      "loss": 0.9648,
+      "step": 5695
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001702025481701465,
+      "loss": 0.924,
+      "step": 5700
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00017013119353438913,
+      "loss": 0.9824,
+      "step": 5705
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00017005976856194582,
+      "loss": 1.0145,
+      "step": 5710
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00016998827332445084,
+      "loss": 0.9222,
+      "step": 5715
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001699167078936089,
+      "loss": 0.8669,
+      "step": 5720
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001698450723411951,
+      "loss": 0.8831,
+      "step": 5725
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00016977336673905497,
+      "loss": 0.9277,
+      "step": 5730
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00016970159115910417,
+      "loss": 0.9618,
+      "step": 5735
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00016962974567332858,
+      "loss": 0.9225,
+      "step": 5740
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00016955783035378424,
+      "loss": 0.9228,
+      "step": 5745
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00016948584527259715,
+      "loss": 0.9805,
+      "step": 5750
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001694137905019633,
+      "loss": 0.9336,
+      "step": 5755
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00016934166611414867,
+      "loss": 0.9763,
+      "step": 5760
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001692694721814889,
+      "loss": 0.9667,
+      "step": 5765
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001691972087763895,
+      "loss": 0.9347,
+      "step": 5770
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001691248759713256,
+      "loss": 1.0006,
+      "step": 5775
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00016905247383884196,
+      "loss": 0.9602,
+      "step": 5780
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016898000245155282,
+      "loss": 0.9398,
+      "step": 5785
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.000168907461882142,
+      "loss": 0.9302,
+      "step": 5790
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00016883485220336257,
+      "loss": 0.916,
+      "step": 5795
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00016876217348803693,
+      "loss": 0.889,
+      "step": 5800
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00016868942580905677,
+      "loss": 0.9243,
+      "step": 5805
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.000168616609239383,
+      "loss": 0.9543,
+      "step": 5810
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001685437238520455,
+      "loss": 1.0033,
+      "step": 5815
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00016847076972014316,
+      "loss": 1.0096,
+      "step": 5820
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00016839774691684395,
+      "loss": 0.9758,
+      "step": 5825
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00016832465551538465,
+      "loss": 0.9468,
+      "step": 5830
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00016825149558907074,
+      "loss": 0.9649,
+      "step": 5835
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001681782672112766,
+      "loss": 0.9563,
+      "step": 5840
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00016810497045544515,
+      "loss": 0.9937,
+      "step": 5845
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001680316053950879,
+      "loss": 0.9506,
+      "step": 5850
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00016795817210378487,
+      "loss": 0.9616,
+      "step": 5855
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001678846706551846,
+      "loss": 0.9319,
+      "step": 5860
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00016781110112300377,
+      "loss": 0.9158,
+      "step": 5865
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001677374635810276,
+      "loss": 1.0142,
+      "step": 5870
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00016766375810310934,
+      "loss": 0.9922,
+      "step": 5875
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00016758998476317042,
+      "loss": 0.8845,
+      "step": 5880
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001675161436352004,
+      "loss": 0.9652,
+      "step": 5885
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001674422347932567,
+      "loss": 0.9629,
+      "step": 5890
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00016736825831146482,
+      "loss": 0.9261,
+      "step": 5895
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00016729421426401787,
+      "loss": 0.8938,
+      "step": 5900
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.000167220102725177,
+      "loss": 0.9534,
+      "step": 5905
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001671459237692708,
+      "loss": 0.8848,
+      "step": 5910
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00016707167747069562,
+      "loss": 1.0258,
+      "step": 5915
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001669973639039153,
+      "loss": 0.9314,
+      "step": 5920
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001669229831434611,
+      "loss": 1.0198,
+      "step": 5925
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00016684853526393185,
+      "loss": 1.0186,
+      "step": 5930
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00016677402033999346,
+      "loss": 0.947,
+      "step": 5935
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00016669943844637924,
+      "loss": 0.9626,
+      "step": 5940
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00016662478965788962,
+      "loss": 1.0421,
+      "step": 5945
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00016655007404939212,
+      "loss": 0.8739,
+      "step": 5950
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00016647529169582122,
+      "loss": 0.9454,
+      "step": 5955
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00016640044267217846,
+      "loss": 0.918,
+      "step": 5960
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00016632552705353213,
+      "loss": 1.0082,
+      "step": 5965
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00016625054491501738,
+      "loss": 0.9736,
+      "step": 5970
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00016617549633183608,
+      "loss": 1.0407,
+      "step": 5975
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00016610038137925668,
+      "loss": 0.9043,
+      "step": 5980
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00016602520013261424,
+      "loss": 0.9397,
+      "step": 5985
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001659499526673103,
+      "loss": 0.9678,
+      "step": 5990
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001658746390588128,
+      "loss": 1.0472,
+      "step": 5995
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00016579925938265606,
+      "loss": 0.9778,
+      "step": 6000
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00016572381371444058,
+      "loss": 0.9727,
+      "step": 6005
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001656483021298331,
+      "loss": 1.0312,
+      "step": 6010
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00016557272470456646,
+      "loss": 0.9426,
+      "step": 6015
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00016549708151443956,
+      "loss": 0.8717,
+      "step": 6020
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00016542137263531723,
+      "loss": 0.9256,
+      "step": 6025
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00016534559814313017,
+      "loss": 0.9841,
+      "step": 6030
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00016526975811387493,
+      "loss": 1.0084,
+      "step": 6035
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00016519385262361372,
+      "loss": 1.003,
+      "step": 6040
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00016511788174847444,
+      "loss": 0.9585,
+      "step": 6045
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001650418455646506,
+      "loss": 0.9436,
+      "step": 6050
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00016496574414840117,
+      "loss": 0.9489,
+      "step": 6055
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00016488957757605056,
+      "loss": 0.9298,
+      "step": 6060
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001648133459239885,
+      "loss": 1.0384,
+      "step": 6065
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00016473704926867,
+      "loss": 0.9922,
+      "step": 6070
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001646606876866153,
+      "loss": 0.8736,
+      "step": 6075
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00016458426125440974,
+      "loss": 0.9582,
+      "step": 6080
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001645077700487036,
+      "loss": 1.0029,
+      "step": 6085
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00016443121414621236,
+      "loss": 0.9921,
+      "step": 6090
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00016435459362371612,
+      "loss": 0.9999,
+      "step": 6095
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00016427790855805995,
+      "loss": 0.9272,
+      "step": 6100
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00016420115902615365,
+      "loss": 0.9221,
+      "step": 6105
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00016412434510497157,
+      "loss": 1.0148,
+      "step": 6110
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00016404746687155277,
+      "loss": 1.0202,
+      "step": 6115
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00016397052440300067,
+      "loss": 0.9146,
+      "step": 6120
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00016389351777648325,
+      "loss": 0.9159,
+      "step": 6125
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00016381644706923277,
+      "loss": 0.9181,
+      "step": 6130
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00016373931235854573,
+      "loss": 0.9705,
+      "step": 6135
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00016366211372178285,
+      "loss": 0.9669,
+      "step": 6140
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00016358485123636903,
+      "loss": 0.9533,
+      "step": 6145
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00016350752497979308,
+      "loss": 0.9237,
+      "step": 6150
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00016343013502960786,
+      "loss": 0.9883,
+      "step": 6155
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016335268146343008,
+      "loss": 0.9447,
+      "step": 6160
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00016327516435894025,
+      "loss": 0.9629,
+      "step": 6165
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001631975837938826,
+      "loss": 1.0215,
+      "step": 6170
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00016311993984606505,
+      "loss": 0.9506,
+      "step": 6175
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00016304223259335898,
+      "loss": 1.0332,
+      "step": 6180
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00016296446211369942,
+      "loss": 0.9143,
+      "step": 6185
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00016288662848508467,
+      "loss": 0.9682,
+      "step": 6190
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00016280873178557643,
+      "loss": 0.9127,
+      "step": 6195
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00016273077209329968,
+      "loss": 0.8634,
+      "step": 6200
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001626527494864425,
+      "loss": 0.9811,
+      "step": 6205
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001625746640432561,
+      "loss": 0.9667,
+      "step": 6210
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001624965158420548,
+      "loss": 0.9561,
+      "step": 6215
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001624183049612157,
+      "loss": 0.9321,
+      "step": 6220
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00016234003147917888,
+      "loss": 0.9751,
+      "step": 6225
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001622616954744472,
+      "loss": 0.998,
+      "step": 6230
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00016218329702558616,
+      "loss": 0.9544,
+      "step": 6235
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00016210483621122395,
+      "loss": 0.9151,
+      "step": 6240
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00016202631311005124,
+      "loss": 0.9362,
+      "step": 6245
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00016194772780082125,
+      "loss": 0.976,
+      "step": 6250
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001618690803623496,
+      "loss": 0.988,
+      "step": 6255
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00016179037087351406,
+      "loss": 0.9854,
+      "step": 6260
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00016171159941325483,
+      "loss": 0.966,
+      "step": 6265
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00016163276606057415,
+      "loss": 0.942,
+      "step": 6270
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016155387089453638,
+      "loss": 0.9386,
+      "step": 6275
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00016147491399426785,
+      "loss": 0.9389,
+      "step": 6280
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001613958954389568,
+      "loss": 0.9415,
+      "step": 6285
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00016131681530785335,
+      "loss": 0.9897,
+      "step": 6290
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00016123767368026929,
+      "loss": 0.8812,
+      "step": 6295
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001611584706355782,
+      "loss": 0.9468,
+      "step": 6300
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001610792062532152,
+      "loss": 0.9353,
+      "step": 6305
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00016099988061267688,
+      "loss": 0.9126,
+      "step": 6310
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00016092049379352132,
+      "loss": 0.9287,
+      "step": 6315
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.000160841045875368,
+      "loss": 0.9711,
+      "step": 6320
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001607615369378976,
+      "loss": 0.9874,
+      "step": 6325
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00016068196706085197,
+      "loss": 0.9632,
+      "step": 6330
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00016060233632403422,
+      "loss": 0.8828,
+      "step": 6335
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001605226448073084,
+      "loss": 0.9771,
+      "step": 6340
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001604428925905995,
+      "loss": 0.9693,
+      "step": 6345
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00016036307975389344,
+      "loss": 0.8772,
+      "step": 6350
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00016028320637723694,
+      "loss": 0.8691,
+      "step": 6355
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00016020327254073736,
+      "loss": 0.9865,
+      "step": 6360
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001601232783245628,
+      "loss": 0.9231,
+      "step": 6365
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00016004322380894182,
+      "loss": 0.9726,
+      "step": 6370
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00015996310907416355,
+      "loss": 0.9754,
+      "step": 6375
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00015988293420057744,
+      "loss": 1.03,
+      "step": 6380
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001598026992685933,
+      "loss": 0.9943,
+      "step": 6385
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00015972240435868117,
+      "loss": 0.9568,
+      "step": 6390
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00015964204955137124,
+      "loss": 0.9483,
+      "step": 6395
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00015956163492725372,
+      "loss": 0.9173,
+      "step": 6400
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00015948116056697888,
+      "loss": 0.9607,
+      "step": 6405
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001594006265512569,
+      "loss": 0.9838,
+      "step": 6410
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00015932003296085774,
+      "loss": 0.9236,
+      "step": 6415
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00015923937987661116,
+      "loss": 0.9918,
+      "step": 6420
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00015915866737940655,
+      "loss": 0.9185,
+      "step": 6425
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00015907789555019296,
+      "loss": 0.9472,
+      "step": 6430
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001589970644699788,
+      "loss": 0.9831,
+      "step": 6435
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00015891617421983205,
+      "loss": 0.8943,
+      "step": 6440
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00015883522488087994,
+      "loss": 0.9508,
+      "step": 6445
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00015875421653430903,
+      "loss": 0.9309,
+      "step": 6450
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00015867314926136509,
+      "loss": 0.9801,
+      "step": 6455
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001585920231433528,
+      "loss": 0.9322,
+      "step": 6460
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00015851083826163607,
+      "loss": 0.9407,
+      "step": 6465
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00015842959469763765,
+      "loss": 0.9791,
+      "step": 6470
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00015834829253283915,
+      "loss": 0.9557,
+      "step": 6475
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00015826693184878095,
+      "loss": 0.9298,
+      "step": 6480
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00015818551272706217,
+      "loss": 0.9324,
+      "step": 6485
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00015810403524934042,
+      "loss": 0.9836,
+      "step": 6490
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00015802249949733202,
+      "loss": 1.0264,
+      "step": 6495
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00015794090555281155,
+      "loss": 0.9183,
+      "step": 6500
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00015785925349761208,
+      "loss": 0.9037,
+      "step": 6505
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00015777754341362487,
+      "loss": 0.9699,
+      "step": 6510
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00015769577538279949,
+      "loss": 0.9243,
+      "step": 6515
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001576139494871435,
+      "loss": 0.9242,
+      "step": 6520
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00015753206580872256,
+      "loss": 1.0183,
+      "step": 6525
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001574501244296603,
+      "loss": 0.9684,
+      "step": 6530
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001573681254321382,
+      "loss": 0.9501,
+      "step": 6535
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00015728606889839553,
+      "loss": 0.901,
+      "step": 6540
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00015720395491072918,
+      "loss": 0.9925,
+      "step": 6545
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00015712178355149385,
+      "loss": 0.9256,
+      "step": 6550
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00015703955490310162,
+      "loss": 0.9016,
+      "step": 6555
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00015695726904802208,
+      "loss": 0.9426,
+      "step": 6560
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001568749260687822,
+      "loss": 1.0187,
+      "step": 6565
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00015679252604796623,
+      "loss": 0.9141,
+      "step": 6570
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001567100690682156,
+      "loss": 0.9538,
+      "step": 6575
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00015662755521222895,
+      "loss": 0.9471,
+      "step": 6580
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00015654498456276188,
+      "loss": 0.9182,
+      "step": 6585
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.000156462357202627,
+      "loss": 1.0089,
+      "step": 6590
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001563796732146938,
+      "loss": 0.9403,
+      "step": 6595
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00015629693268188842,
+      "loss": 1.0023,
+      "step": 6600
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.000156214135687194,
+      "loss": 1.0552,
+      "step": 6605
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00015613128231365002,
+      "loss": 0.9476,
+      "step": 6610
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00015604837264435268,
+      "loss": 0.9862,
+      "step": 6615
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00015596540676245454,
+      "loss": 1.0152,
+      "step": 6620
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00015588238475116464,
+      "loss": 0.9548,
+      "step": 6625
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001557993066937482,
+      "loss": 0.9643,
+      "step": 6630
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001557161726735268,
+      "loss": 0.9176,
+      "step": 6635
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001556329827738779,
+      "loss": 0.9208,
+      "step": 6640
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00015554973707823525,
+      "loss": 0.9301,
+      "step": 6645
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00015546643567008848,
+      "loss": 0.9482,
+      "step": 6650
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00015538307863298303,
+      "loss": 0.9623,
+      "step": 6655
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00015529966605052023,
+      "loss": 0.9909,
+      "step": 6660
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00015521619800635704,
+      "loss": 1.0055,
+      "step": 6665
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00015513267458420606,
+      "loss": 0.9519,
+      "step": 6670
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001550490958678355,
+      "loss": 0.994,
+      "step": 6675
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00015496546194106888,
+      "loss": 0.9019,
+      "step": 6680
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00015488177288778532,
+      "loss": 0.9355,
+      "step": 6685
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00015479802879191898,
+      "loss": 0.9583,
+      "step": 6690
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00015471422973745936,
+      "loss": 0.9922,
+      "step": 6695
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00015463037580845107,
+      "loss": 0.9591,
+      "step": 6700
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00015454646708899374,
+      "loss": 0.9839,
+      "step": 6705
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00015446250366324196,
+      "loss": 0.9443,
+      "step": 6710
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00015437848561540517,
+      "loss": 0.961,
+      "step": 6715
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00015429441302974755,
+      "loss": 0.9097,
+      "step": 6720
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00015421028599058812,
+      "loss": 0.8755,
+      "step": 6725
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00015412610458230037,
+      "loss": 0.9329,
+      "step": 6730
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00015404186888931233,
+      "loss": 0.9379,
+      "step": 6735
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00015395757899610662,
+      "loss": 1.0126,
+      "step": 6740
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00015387323498722,
+      "loss": 0.9752,
+      "step": 6745
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00015378883694724369,
+      "loss": 0.9499,
+      "step": 6750
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00015370438496082302,
+      "loss": 0.9756,
+      "step": 6755
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001536198791126574,
+      "loss": 0.932,
+      "step": 6760
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00015353531948750026,
+      "loss": 0.8913,
+      "step": 6765
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001534507061701591,
+      "loss": 0.8923,
+      "step": 6770
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00015336603924549503,
+      "loss": 1.0337,
+      "step": 6775
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001532813187984232,
+      "loss": 0.9402,
+      "step": 6780
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001531965449139122,
+      "loss": 0.9171,
+      "step": 6785
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00015311171767698435,
+      "loss": 0.9178,
+      "step": 6790
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001530268371727154,
+      "loss": 0.9138,
+      "step": 6795
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001529419034862346,
+      "loss": 0.992,
+      "step": 6800
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00015285691670272451,
+      "loss": 0.9323,
+      "step": 6805
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001527718769074209,
+      "loss": 0.9513,
+      "step": 6810
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00015268678418561276,
+      "loss": 0.9996,
+      "step": 6815
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00015260163862264217,
+      "loss": 0.9446,
+      "step": 6820
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00015251644030390415,
+      "loss": 0.9033,
+      "step": 6825
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00015243118931484667,
+      "loss": 1.0265,
+      "step": 6830
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001523458857409705,
+      "loss": 0.9354,
+      "step": 6835
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00015226052966782914,
+      "loss": 0.9134,
+      "step": 6840
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0001521751211810288,
+      "loss": 0.9379,
+      "step": 6845
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00015208966036622825,
+      "loss": 0.9665,
+      "step": 6850
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00015200414730913865,
+      "loss": 1.0182,
+      "step": 6855
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00015191858209552368,
+      "loss": 1.046,
+      "step": 6860
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00015183296481119924,
+      "loss": 1.0084,
+      "step": 6865
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00015174729554203348,
+      "loss": 0.9653,
+      "step": 6870
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00015166157437394672,
+      "loss": 0.9834,
+      "step": 6875
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00015157580139291124,
+      "loss": 1.0267,
+      "step": 6880
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00015148997668495143,
+      "loss": 0.931,
+      "step": 6885
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001514041003361434,
+      "loss": 0.966,
+      "step": 6890
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00015131817243261512,
+      "loss": 0.8916,
+      "step": 6895
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00015123219306054634,
+      "loss": 0.909,
+      "step": 6900
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00015114616230616835,
+      "loss": 1.0168,
+      "step": 6905
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00015106008025576393,
+      "loss": 0.8856,
+      "step": 6910
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00015097394699566737,
+      "loss": 0.9172,
+      "step": 6915
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001508877626122644,
+      "loss": 0.9176,
+      "step": 6920
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00015080152719199183,
+      "loss": 0.9964,
+      "step": 6925
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00015071524082133778,
+      "loss": 0.8834,
+      "step": 6930
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00015062890358684148,
+      "loss": 1.0168,
+      "step": 6935
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001505425155750931,
+      "loss": 1.0014,
+      "step": 6940
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00015045607687273383,
+      "loss": 0.9284,
+      "step": 6945
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00015036958756645564,
+      "loss": 1.0146,
+      "step": 6950
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00015028304774300123,
+      "loss": 0.8889,
+      "step": 6955
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00015019645748916402,
+      "loss": 0.9944,
+      "step": 6960
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00015010981689178796,
+      "loss": 0.9689,
+      "step": 6965
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00015002312603776754,
+      "loss": 0.8948,
+      "step": 6970
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00014993638501404762,
+      "loss": 0.9505,
+      "step": 6975
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00014984959390762335,
+      "loss": 1.0177,
+      "step": 6980
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00014976275280554016,
+      "loss": 0.9758,
+      "step": 6985
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00014967586179489366,
+      "loss": 1.0012,
+      "step": 6990
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001495889209628294,
+      "loss": 0.944,
+      "step": 6995
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00014950193039654297,
+      "loss": 0.9734,
+      "step": 7000
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00014941489018327988,
+      "loss": 0.9297,
+      "step": 7005
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001493278004103353,
+      "loss": 0.9897,
+      "step": 7010
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00014924066116505427,
+      "loss": 0.9069,
+      "step": 7015
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00014915347253483126,
+      "loss": 1.0033,
+      "step": 7020
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00014906623460711046,
+      "loss": 0.9473,
+      "step": 7025
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00014897894746938536,
+      "loss": 0.8782,
+      "step": 7030
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00014889161120919893,
+      "loss": 0.9315,
+      "step": 7035
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00014880422591414323,
+      "loss": 0.9325,
+      "step": 7040
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00014871679167185973,
+      "loss": 0.9548,
+      "step": 7045
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00014862930857003877,
+      "loss": 0.9158,
+      "step": 7050
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00014854177669641983,
+      "loss": 0.9412,
+      "step": 7055
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001484541961387912,
+      "loss": 0.998,
+      "step": 7060
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00014836656698499016,
+      "loss": 0.9125,
+      "step": 7065
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00014827888932290257,
+      "loss": 0.9262,
+      "step": 7070
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.000148191163240463,
+      "loss": 0.9809,
+      "step": 7075
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00014810338882565455,
+      "loss": 0.9246,
+      "step": 7080
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00014801556616650887,
+      "loss": 0.9429,
+      "step": 7085
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00014792769535110597,
+      "loss": 0.895,
+      "step": 7090
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00014783977646757403,
+      "loss": 0.9291,
+      "step": 7095
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00014775180960408966,
+      "loss": 0.9427,
+      "step": 7100
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00014766379484887744,
+      "loss": 1.0229,
+      "step": 7105
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00014757573229021002,
+      "loss": 1.005,
+      "step": 7110
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00014748762201640796,
+      "loss": 0.8976,
+      "step": 7115
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00014739946411583977,
+      "loss": 0.9826,
+      "step": 7120
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00014731125867692158,
+      "loss": 0.9575,
+      "step": 7125
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00014722300578811734,
+      "loss": 0.9073,
+      "step": 7130
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00014713470553793853,
+      "loss": 0.9769,
+      "step": 7135
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001470463580149441,
+      "loss": 0.9957,
+      "step": 7140
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00014695796330774048,
+      "loss": 0.9545,
+      "step": 7145
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00014686952150498134,
+      "loss": 0.9559,
+      "step": 7150
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00014678103269536762,
+      "loss": 0.9678,
+      "step": 7155
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00014669249696764748,
+      "loss": 0.9252,
+      "step": 7160
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00014660391441061603,
+      "loss": 0.9327,
+      "step": 7165
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00014651528511311538,
+      "loss": 0.9919,
+      "step": 7170
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001464266091640345,
+      "loss": 0.9298,
+      "step": 7175
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001463378866523092,
+      "loss": 0.9694,
+      "step": 7180
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00014624911766692196,
+      "loss": 0.9764,
+      "step": 7185
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00014616030229690186,
+      "loss": 0.9309,
+      "step": 7190
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001460714406313245,
+      "loss": 0.917,
+      "step": 7195
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00014598253275931197,
+      "loss": 0.9378,
+      "step": 7200
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00014589357877003257,
+      "loss": 0.9691,
+      "step": 7205
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00014580457875270098,
+      "loss": 0.9381,
+      "step": 7210
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00014571553279657803,
+      "loss": 0.9212,
+      "step": 7215
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00014562644099097048,
+      "loss": 1.0157,
+      "step": 7220
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00014553730342523134,
+      "loss": 0.9484,
+      "step": 7225
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001454481201887592,
+      "loss": 0.9287,
+      "step": 7230
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00014535889137099877,
+      "loss": 0.9912,
+      "step": 7235
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00014526961706144023,
+      "loss": 0.9336,
+      "step": 7240
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00014518029734961947,
+      "loss": 0.9718,
+      "step": 7245
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00014509093232511791,
+      "loss": 0.9316,
+      "step": 7250
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00014500152207756246,
+      "loss": 0.9038,
+      "step": 7255
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00014491206669662533,
+      "loss": 1.002,
+      "step": 7260
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00014482256627202405,
+      "loss": 0.9443,
+      "step": 7265
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00014473302089352123,
+      "loss": 0.8925,
+      "step": 7270
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00014464343065092466,
+      "loss": 0.9866,
+      "step": 7275
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00014455379563408713,
+      "loss": 0.9338,
+      "step": 7280
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00014446411593290625,
+      "loss": 0.879,
+      "step": 7285
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001443743916373245,
+      "loss": 0.9915,
+      "step": 7290
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00014428462283732908,
+      "loss": 0.9512,
+      "step": 7295
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001441948096229518,
+      "loss": 0.9097,
+      "step": 7300
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00014410495208426908,
+      "loss": 0.9026,
+      "step": 7305
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00014401505031140171,
+      "loss": 0.9441,
+      "step": 7310
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00014392510439451494,
+      "loss": 0.9621,
+      "step": 7315
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00014383511442381822,
+      "loss": 0.977,
+      "step": 7320
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00014374508048956515,
+      "loss": 0.9762,
+      "step": 7325
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00014365500268205352,
+      "loss": 0.971,
+      "step": 7330
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001435648810916251,
+      "loss": 0.9624,
+      "step": 7335
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001434747158086655,
+      "loss": 0.9771,
+      "step": 7340
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00014338450692360418,
+      "loss": 0.96,
+      "step": 7345
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001432942545269144,
+      "loss": 0.9897,
+      "step": 7350
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.000143203958709113,
+      "loss": 0.9156,
+      "step": 7355
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00014311361956076036,
+      "loss": 0.9731,
+      "step": 7360
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00014302323717246032,
+      "loss": 0.9721,
+      "step": 7365
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001429328116348601,
+      "loss": 0.9479,
+      "step": 7370
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00014284234303865026,
+      "loss": 0.9593,
+      "step": 7375
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001427518314745644,
+      "loss": 0.9217,
+      "step": 7380
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001426612770333793,
+      "loss": 0.9589,
+      "step": 7385
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00014257067980591475,
+      "loss": 0.982,
+      "step": 7390
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00014248003988303346,
+      "loss": 0.9983,
+      "step": 7395
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00014238935735564094,
+      "loss": 1.0515,
+      "step": 7400
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00014229863231468538,
+      "loss": 0.9091,
+      "step": 7405
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00014220786485115772,
+      "loss": 0.9742,
+      "step": 7410
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001421170550560913,
+      "loss": 0.915,
+      "step": 7415
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0001420262030205621,
+      "loss": 1.0453,
+      "step": 7420
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001419353088356883,
+      "loss": 1.0303,
+      "step": 7425
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00014184437259263038,
+      "loss": 0.9718,
+      "step": 7430
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00014175339438259112,
+      "loss": 0.9389,
+      "step": 7435
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00014166237429681525,
+      "loss": 0.9127,
+      "step": 7440
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00014157131242658957,
+      "loss": 0.9791,
+      "step": 7445
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001414802088632428,
+      "loss": 0.8747,
+      "step": 7450
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00014138906369814538,
+      "loss": 0.8839,
+      "step": 7455
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001412978770227096,
+      "loss": 0.9465,
+      "step": 7460
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00014120664892838933,
+      "loss": 0.9703,
+      "step": 7465
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001411153795066799,
+      "loss": 0.9488,
+      "step": 7470
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00014102406884911826,
+      "loss": 0.9921,
+      "step": 7475
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00014093271704728252,
+      "loss": 0.9758,
+      "step": 7480
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00014084132419279224,
+      "loss": 0.9741,
+      "step": 7485
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00014074989037730798,
+      "loss": 0.9938,
+      "step": 7490
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00014065841569253155,
+      "loss": 0.9672,
+      "step": 7495
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00014056690023020566,
+      "loss": 0.9742,
+      "step": 7500
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00014047534408211383,
+      "loss": 0.9657,
+      "step": 7505
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00014038374734008058,
+      "loss": 0.9535,
+      "step": 7510
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00014029211009597097,
+      "loss": 0.9175,
+      "step": 7515
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00014020043244169082,
+      "loss": 0.9896,
+      "step": 7520
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00014010871446918635,
+      "loss": 0.8928,
+      "step": 7525
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00014001695627044428,
+      "loss": 0.8972,
+      "step": 7530
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00013992515793749172,
+      "loss": 0.952,
+      "step": 7535
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00013983331956239596,
+      "loss": 0.9458,
+      "step": 7540
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00013974144123726442,
+      "loss": 0.9653,
+      "step": 7545
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00013964952305424474,
+      "loss": 0.9689,
+      "step": 7550
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00013955756510552437,
+      "loss": 1.0045,
+      "step": 7555
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001394655674833307,
+      "loss": 1.0394,
+      "step": 7560
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00013937353027993092,
+      "loss": 0.8979,
+      "step": 7565
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00013928145358763194,
+      "loss": 0.9154,
+      "step": 7570
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00013918933749878024,
+      "loss": 0.9568,
+      "step": 7575
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00013909718210576179,
+      "loss": 0.9796,
+      "step": 7580
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.000139004987501002,
+      "loss": 0.941,
+      "step": 7585
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0001389127537769657,
+      "loss": 0.923,
+      "step": 7590
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00013882048102615676,
+      "loss": 0.9264,
+      "step": 7595
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00013872816934111838,
+      "loss": 0.9101,
+      "step": 7600
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00013863581881443275,
+      "loss": 0.9212,
+      "step": 7605
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001385434295387209,
+      "loss": 0.9427,
+      "step": 7610
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00013845100160664287,
+      "loss": 0.9683,
+      "step": 7615
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001383585351108974,
+      "loss": 0.945,
+      "step": 7620
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00013826603014422192,
+      "loss": 1.0078,
+      "step": 7625
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001381734867993925,
+      "loss": 1.0363,
+      "step": 7630
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001380809051692236,
+      "loss": 0.9902,
+      "step": 7635
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00013798828534656812,
+      "loss": 1.0163,
+      "step": 7640
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00013789562742431727,
+      "loss": 0.9083,
+      "step": 7645
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00013780293149540053,
+      "loss": 0.9632,
+      "step": 7650
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00013771019765278537,
+      "loss": 0.943,
+      "step": 7655
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00013761742598947734,
+      "loss": 0.9423,
+      "step": 7660
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00013752461659852,
+      "loss": 0.9826,
+      "step": 7665
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00013743176957299464,
+      "loss": 0.9697,
+      "step": 7670
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00013733888500602038,
+      "loss": 0.9645,
+      "step": 7675
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00013724596299075388,
+      "loss": 0.9397,
+      "step": 7680
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001371530036203895,
+      "loss": 0.9476,
+      "step": 7685
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00013706000698815893,
+      "loss": 0.929,
+      "step": 7690
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00013696697318733134,
+      "loss": 0.9219,
+      "step": 7695
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00013687390231121314,
+      "loss": 0.9162,
+      "step": 7700
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00013678079445314783,
+      "loss": 0.8962,
+      "step": 7705
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00013668764970651615,
+      "loss": 0.9636,
+      "step": 7710
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001365944681647358,
+      "loss": 0.9463,
+      "step": 7715
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00013650124992126128,
+      "loss": 0.9982,
+      "step": 7720
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00013640799506958403,
+      "loss": 0.9802,
+      "step": 7725
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00013631470370323214,
+      "loss": 0.9338,
+      "step": 7730
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001362213759157703,
+      "loss": 0.9789,
+      "step": 7735
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001361280118007998,
+      "loss": 1.0204,
+      "step": 7740
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001360346114519583,
+      "loss": 0.949,
+      "step": 7745
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001359411749629198,
+      "loss": 0.9641,
+      "step": 7750
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001358477024273946,
+      "loss": 0.9711,
+      "step": 7755
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001357541939391291,
+      "loss": 0.9231,
+      "step": 7760
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00013566064959190583,
+      "loss": 1.0005,
+      "step": 7765
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001355670694795432,
+      "loss": 0.9869,
+      "step": 7770
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001354734536958955,
+      "loss": 0.9578,
+      "step": 7775
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001353798023348528,
+      "loss": 0.9296,
+      "step": 7780
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00013528611549034096,
+      "loss": 0.9171,
+      "step": 7785
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00013519239325632124,
+      "loss": 0.9803,
+      "step": 7790
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00013509863572679057,
+      "loss": 0.9256,
+      "step": 7795
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00013500484299578116,
+      "loss": 0.968,
+      "step": 7800
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00013491101515736057,
+      "loss": 0.9171,
+      "step": 7805
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00013481715230563153,
+      "loss": 1.0461,
+      "step": 7810
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00013472325453473197,
+      "loss": 0.9204,
+      "step": 7815
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00013462932193883482,
+      "loss": 0.9476,
+      "step": 7820
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001345353546121478,
+      "loss": 0.8913,
+      "step": 7825
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00013444135264891371,
+      "loss": 0.9357,
+      "step": 7830
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00013434731614340984,
+      "loss": 0.8973,
+      "step": 7835
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00013425324518994826,
+      "loss": 0.9029,
+      "step": 7840
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001341591398828756,
+      "loss": 1.0038,
+      "step": 7845
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00013406500031657283,
+      "loss": 1.0186,
+      "step": 7850
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00013397082658545543,
+      "loss": 0.9331,
+      "step": 7855
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00013387661878397307,
+      "loss": 0.9009,
+      "step": 7860
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00013378237700660957,
+      "loss": 0.9224,
+      "step": 7865
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00013368810134788278,
+      "loss": 0.972,
+      "step": 7870
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00013359379190234472,
+      "loss": 0.8709,
+      "step": 7875
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00013349944876458108,
+      "loss": 0.9213,
+      "step": 7880
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00013340507202921152,
+      "loss": 0.9116,
+      "step": 7885
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0001333106617908892,
+      "loss": 0.9288,
+      "step": 7890
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00013321621814430106,
+      "loss": 0.996,
+      "step": 7895
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001331217411841675,
+      "loss": 0.96,
+      "step": 7900
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013302723100524222,
+      "loss": 0.9032,
+      "step": 7905
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001329326877023124,
+      "loss": 0.9752,
+      "step": 7910
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00013283811137019836,
+      "loss": 0.9543,
+      "step": 7915
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00013274350210375357,
+      "loss": 0.9367,
+      "step": 7920
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00013264885999786442,
+      "loss": 0.9281,
+      "step": 7925
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0001325541851474504,
+      "loss": 0.9697,
+      "step": 7930
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001324594776474638,
+      "loss": 0.9704,
+      "step": 7935
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001323647375928895,
+      "loss": 0.9694,
+      "step": 7940
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00013226996507874526,
+      "loss": 1.0088,
+      "step": 7945
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00013217516020008128,
+      "loss": 0.9742,
+      "step": 7950
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00013208032305198018,
+      "loss": 1.0108,
+      "step": 7955
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00013198545372955706,
+      "loss": 0.9823,
+      "step": 7960
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00013189055232795915,
+      "loss": 0.974,
+      "step": 7965
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00013179561894236598,
+      "loss": 1.0433,
+      "step": 7970
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00013170065366798907,
+      "loss": 0.9458,
+      "step": 7975
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00013160565660007195,
+      "loss": 0.9387,
+      "step": 7980
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00013151062783389007,
+      "loss": 0.8602,
+      "step": 7985
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00013141556746475058,
+      "loss": 0.9077,
+      "step": 7990
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00013132047558799242,
+      "loss": 0.9475,
+      "step": 7995
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00013122535229898613,
+      "loss": 0.9877,
+      "step": 8000
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0001311301976931337,
+      "loss": 0.8818,
+      "step": 8005
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00013103501186586855,
+      "loss": 0.9365,
+      "step": 8010
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00013093979491265542,
+      "loss": 0.9283,
+      "step": 8015
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001308445469289902,
+      "loss": 0.9965,
+      "step": 8020
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00013074926801040005,
+      "loss": 0.9523,
+      "step": 8025
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.000130653958252443,
+      "loss": 0.9938,
+      "step": 8030
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001305586177507081,
+      "loss": 0.954,
+      "step": 8035
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00013046324660081525,
+      "loss": 0.9538,
+      "step": 8040
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00013036784489841495,
+      "loss": 0.9105,
+      "step": 8045
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00013027241273918855,
+      "loss": 0.9565,
+      "step": 8050
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00013017695021884777,
+      "loss": 0.877,
+      "step": 8055
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00013008145743313487,
+      "loss": 0.937,
+      "step": 8060
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00012998593447782246,
+      "loss": 0.9396,
+      "step": 8065
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001298903814487133,
+      "loss": 0.925,
+      "step": 8070
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001297947984416405,
+      "loss": 0.9002,
+      "step": 8075
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001296991855524671,
+      "loss": 0.9851,
+      "step": 8080
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001296035428770861,
+      "loss": 0.9482,
+      "step": 8085
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00012950787051142052,
+      "loss": 0.9381,
+      "step": 8090
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00012941216855142298,
+      "loss": 0.9905,
+      "step": 8095
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00012931643709307588,
+      "loss": 0.9794,
+      "step": 8100
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00012922067623239117,
+      "loss": 0.9296,
+      "step": 8105
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00012912488606541035,
+      "loss": 0.9246,
+      "step": 8110
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00012902906668820418,
+      "loss": 0.9497,
+      "step": 8115
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00012893321819687286,
+      "loss": 0.9739,
+      "step": 8120
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0001288373406875457,
+      "loss": 0.884,
+      "step": 8125
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00012874143425638116,
+      "loss": 0.9579,
+      "step": 8130
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001286454989995667,
+      "loss": 0.9465,
+      "step": 8135
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00012854953501331863,
+      "loss": 1.0287,
+      "step": 8140
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001284535423938822,
+      "loss": 0.9739,
+      "step": 8145
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001283575212375312,
+      "loss": 0.9755,
+      "step": 8150
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00012826147164056822,
+      "loss": 0.9555,
+      "step": 8155
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00012816539369932422,
+      "loss": 0.9819,
+      "step": 8160
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00012806928751015874,
+      "loss": 0.9665,
+      "step": 8165
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001279731531694595,
+      "loss": 0.9912,
+      "step": 8170
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00012787699077364262,
+      "loss": 0.8946,
+      "step": 8175
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00012778080041915215,
+      "loss": 1.0273,
+      "step": 8180
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00012768458220246035,
+      "loss": 1.0033,
+      "step": 8185
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00012758833622006737,
+      "loss": 0.9711,
+      "step": 8190
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001274920625685012,
+      "loss": 0.9333,
+      "step": 8195
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001273957613443176,
+      "loss": 0.9855,
+      "step": 8200
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00012729943264409992,
+      "loss": 0.9678,
+      "step": 8205
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00012720307656445914,
+      "loss": 0.9855,
+      "step": 8210
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001271066932020337,
+      "loss": 0.9305,
+      "step": 8215
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00012701028265348934,
+      "loss": 0.9011,
+      "step": 8220
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001269138450155191,
+      "loss": 0.938,
+      "step": 8225
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00012681738038484324,
+      "loss": 0.9451,
+      "step": 8230
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00012672088885820897,
+      "loss": 0.957,
+      "step": 8235
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00012662437053239062,
+      "loss": 0.9565,
+      "step": 8240
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001265278255041893,
+      "loss": 0.9258,
+      "step": 8245
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001264312538704329,
+      "loss": 0.9812,
+      "step": 8250
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00012633465572797604,
+      "loss": 0.9499,
+      "step": 8255
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001262380311736999,
+      "loss": 0.9338,
+      "step": 8260
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001261413803045122,
+      "loss": 0.9084,
+      "step": 8265
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00012604470321734694,
+      "loss": 1.024,
+      "step": 8270
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00012594800000916448,
+      "loss": 0.996,
+      "step": 8275
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00012585127077695144,
+      "loss": 0.9739,
+      "step": 8280
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00012575451561772048,
+      "loss": 0.9378,
+      "step": 8285
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00012565773462851017,
+      "loss": 0.9997,
+      "step": 8290
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00012556092790638518,
+      "loss": 0.9145,
+      "step": 8295
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00012546409554843585,
+      "loss": 0.9176,
+      "step": 8300
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00012536723765177826,
+      "loss": 0.927,
+      "step": 8305
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0001252703543135541,
+      "loss": 1.0053,
+      "step": 8310
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001251734456309306,
+      "loss": 0.8807,
+      "step": 8315
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00012507651170110042,
+      "loss": 1.0664,
+      "step": 8320
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001249795526212815,
+      "loss": 0.9316,
+      "step": 8325
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.000124882568488717,
+      "loss": 0.9411,
+      "step": 8330
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00012478555940067528,
+      "loss": 0.9834,
+      "step": 8335
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00012468852545444961,
+      "loss": 0.9721,
+      "step": 8340
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001245914667473583,
+      "loss": 0.9796,
+      "step": 8345
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00012449438337674447,
+      "loss": 0.9297,
+      "step": 8350
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00012439727543997586,
+      "loss": 0.9277,
+      "step": 8355
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00012430014303444503,
+      "loss": 0.8868,
+      "step": 8360
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00012420298625756898,
+      "loss": 0.9015,
+      "step": 8365
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001241058052067892,
+      "loss": 1.0216,
+      "step": 8370
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001240085999795714,
+      "loss": 0.904,
+      "step": 8375
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00012391137067340572,
+      "loss": 0.9708,
+      "step": 8380
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001238141173858063,
+      "loss": 0.8768,
+      "step": 8385
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00012371684021431144,
+      "loss": 0.9909,
+      "step": 8390
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00012361953925648327,
+      "loss": 0.9845,
+      "step": 8395
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001235222146099079,
+      "loss": 0.9625,
+      "step": 8400
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00012342486637219517,
+      "loss": 0.9705,
+      "step": 8405
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00012332749464097855,
+      "loss": 0.9871,
+      "step": 8410
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00012323009951391504,
+      "loss": 0.934,
+      "step": 8415
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00012313268108868518,
+      "loss": 0.9593,
+      "step": 8420
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00012303523946299285,
+      "loss": 0.9758,
+      "step": 8425
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00012293777473456518,
+      "loss": 0.9393,
+      "step": 8430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00012284028700115245,
+      "loss": 0.977,
+      "step": 8435
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001227427763605281,
+      "loss": 0.9755,
+      "step": 8440
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001226452429104884,
+      "loss": 0.982,
+      "step": 8445
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001225476867488527,
+      "loss": 1.0214,
+      "step": 8450
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00012245010797346296,
+      "loss": 0.9816,
+      "step": 8455
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00012235250668218386,
+      "loss": 1.0024,
+      "step": 8460
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00012225488297290266,
+      "loss": 0.9436,
+      "step": 8465
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00012215723694352916,
+      "loss": 0.9496,
+      "step": 8470
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00012205956869199549,
+      "loss": 0.9747,
+      "step": 8475
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00012196187831625605,
+      "loss": 0.9595,
+      "step": 8480
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00012186416591428751,
+      "loss": 0.9554,
+      "step": 8485
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00012176643158408853,
+      "loss": 0.8985,
+      "step": 8490
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00012166867542367985,
+      "loss": 0.9213,
+      "step": 8495
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00012157089753110406,
+      "loss": 0.9098,
+      "step": 8500
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00012147309800442555,
+      "loss": 1.0076,
+      "step": 8505
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00012137527694173038,
+      "loss": 0.9378,
+      "step": 8510
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00012127743444112629,
+      "loss": 0.9233,
+      "step": 8515
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00012117957060074245,
+      "loss": 0.9537,
+      "step": 8520
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00012108168551872944,
+      "loss": 0.9529,
+      "step": 8525
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00012098377929325917,
+      "loss": 0.9558,
+      "step": 8530
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00012088585202252474,
+      "loss": 1.0213,
+      "step": 8535
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00012078790380474037,
+      "loss": 0.9165,
+      "step": 8540
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00012068993473814126,
+      "loss": 0.8604,
+      "step": 8545
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00012059194492098351,
+      "loss": 0.879,
+      "step": 8550
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00012049393445154411,
+      "loss": 0.9013,
+      "step": 8555
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001203959034281207,
+      "loss": 0.9588,
+      "step": 8560
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00012029785194903153,
+      "loss": 0.9096,
+      "step": 8565
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00012019978011261541,
+      "loss": 0.9206,
+      "step": 8570
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00012010168801723149,
+      "loss": 0.9477,
+      "step": 8575
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00012000357576125932,
+      "loss": 0.8503,
+      "step": 8580
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00011990544344309865,
+      "loss": 0.9125,
+      "step": 8585
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00011980729116116927,
+      "loss": 0.9419,
+      "step": 8590
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00011970911901391113,
+      "loss": 0.9902,
+      "step": 8595
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00011961092709978402,
+      "loss": 0.9452,
+      "step": 8600
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00011951271551726755,
+      "loss": 0.9609,
+      "step": 8605
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00011941448436486106,
+      "loss": 1.0057,
+      "step": 8610
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00011931623374108358,
+      "loss": 0.9763,
+      "step": 8615
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00011921796374447356,
+      "loss": 0.9368,
+      "step": 8620
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00011911967447358901,
+      "loss": 0.8984,
+      "step": 8625
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00011902136602700711,
+      "loss": 0.9687,
+      "step": 8630
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00011892303850332443,
+      "loss": 0.9462,
+      "step": 8635
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00011882469200115656,
+      "loss": 0.9771,
+      "step": 8640
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00011872632661913823,
+      "loss": 1.0017,
+      "step": 8645
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00011862794245592301,
+      "loss": 0.8716,
+      "step": 8650
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00011852953961018332,
+      "loss": 0.9954,
+      "step": 8655
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00011843111818061036,
+      "loss": 0.9749,
+      "step": 8660
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00011833267826591394,
+      "loss": 0.9266,
+      "step": 8665
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001182342199648224,
+      "loss": 0.9462,
+      "step": 8670
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00011813574337608258,
+      "loss": 0.9736,
+      "step": 8675
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00011803724859845952,
+      "loss": 1.0284,
+      "step": 8680
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00011793873573073673,
+      "loss": 0.9386,
+      "step": 8685
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00011784020487171566,
+      "loss": 0.9723,
+      "step": 8690
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00011774165612021585,
+      "loss": 0.9502,
+      "step": 8695
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00011764308957507488,
+      "loss": 0.9251,
+      "step": 8700
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00011754450533514807,
+      "loss": 0.9441,
+      "step": 8705
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00011744590349930849,
+      "loss": 0.995,
+      "step": 8710
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00011734728416644694,
+      "loss": 0.8906,
+      "step": 8715
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00011724864743547168,
+      "loss": 0.8851,
+      "step": 8720
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001171499934053085,
+      "loss": 0.9489,
+      "step": 8725
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00011705132217490047,
+      "loss": 0.8942,
+      "step": 8730
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00011695263384320794,
+      "loss": 0.9441,
+      "step": 8735
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00011685392850920842,
+      "loss": 0.9637,
+      "step": 8740
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00011675520627189648,
+      "loss": 0.9091,
+      "step": 8745
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00011665646723028359,
+      "loss": 0.9633,
+      "step": 8750
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00011655771148339812,
+      "loss": 0.9663,
+      "step": 8755
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00011645893913028514,
+      "loss": 0.9058,
+      "step": 8760
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00011636015027000651,
+      "loss": 0.8779,
+      "step": 8765
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00011626134500164047,
+      "loss": 0.9222,
+      "step": 8770
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00011616252342428184,
+      "loss": 0.9287,
+      "step": 8775
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00011606368563704177,
+      "loss": 0.9116,
+      "step": 8780
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00011596483173904762,
+      "loss": 0.9408,
+      "step": 8785
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00011586596182944293,
+      "loss": 0.9107,
+      "step": 8790
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00011576707600738739,
+      "loss": 0.9608,
+      "step": 8795
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00011566817437205643,
+      "loss": 0.9313,
+      "step": 8800
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001155692570226416,
+      "loss": 0.9032,
+      "step": 8805
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00011547032405835005,
+      "loss": 0.9539,
+      "step": 8810
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00011537137557840463,
+      "loss": 0.9214,
+      "step": 8815
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00011527241168204375,
+      "loss": 0.9505,
+      "step": 8820
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00011517343246852126,
+      "loss": 1.0026,
+      "step": 8825
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00011507443803710643,
+      "loss": 1.0072,
+      "step": 8830
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00011497542848708374,
+      "loss": 0.8672,
+      "step": 8835
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00011487640391775283,
+      "loss": 0.9045,
+      "step": 8840
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00011477736442842846,
+      "loss": 0.9884,
+      "step": 8845
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00011467831011844027,
+      "loss": 0.955,
+      "step": 8850
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00011457924108713287,
+      "loss": 0.9621,
+      "step": 8855
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00011448015743386553,
+      "loss": 0.9314,
+      "step": 8860
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00011438105925801224,
+      "loss": 0.9508,
+      "step": 8865
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00011428194665896155,
+      "loss": 0.8861,
+      "step": 8870
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00011418281973611647,
+      "loss": 0.9121,
+      "step": 8875
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00011408367858889437,
+      "loss": 0.9316,
+      "step": 8880
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00011398452331672689,
+      "loss": 1.0033,
+      "step": 8885
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00011388535401905985,
+      "loss": 0.8959,
+      "step": 8890
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00011378617079535312,
+      "loss": 0.9182,
+      "step": 8895
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00011368697374508052,
+      "loss": 0.9436,
+      "step": 8900
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00011358776296772982,
+      "loss": 0.9697,
+      "step": 8905
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00011348853856280244,
+      "loss": 0.9119,
+      "step": 8910
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00011338930062981352,
+      "loss": 0.9646,
+      "step": 8915
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001132900492682918,
+      "loss": 0.9617,
+      "step": 8920
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00011319078457777947,
+      "loss": 0.9387,
+      "step": 8925
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00011309150665783204,
+      "loss": 0.9628,
+      "step": 8930
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00011299221560801836,
+      "loss": 0.9331,
+      "step": 8935
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00011289291152792038,
+      "loss": 0.9422,
+      "step": 8940
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00011279359451713318,
+      "loss": 0.9675,
+      "step": 8945
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00011269426467526477,
+      "loss": 0.9467,
+      "step": 8950
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00011259492210193603,
+      "loss": 0.9142,
+      "step": 8955
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00011249556689678063,
+      "loss": 0.9927,
+      "step": 8960
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00011239619915944488,
+      "loss": 0.9563,
+      "step": 8965
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00011229681898958775,
+      "loss": 0.9659,
+      "step": 8970
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001121974264868805,
+      "loss": 0.8972,
+      "step": 8975
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00011209802175100692,
+      "loss": 0.882,
+      "step": 8980
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00011199860488166302,
+      "loss": 1.0123,
+      "step": 8985
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00011189917597855694,
+      "loss": 0.984,
+      "step": 8990
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00011179973514140896,
+      "loss": 0.9422,
+      "step": 8995
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00011170028246995123,
+      "loss": 0.9223,
+      "step": 9000
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00011160081806392788,
+      "loss": 1.0244,
+      "step": 9005
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00011150134202309474,
+      "loss": 0.8835,
+      "step": 9010
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00011140185444721937,
+      "loss": 0.9615,
+      "step": 9015
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00011130235543608081,
+      "loss": 0.9861,
+      "step": 9020
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00011120284508946959,
+      "loss": 0.9579,
+      "step": 9025
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00011110332350718768,
+      "loss": 0.8844,
+      "step": 9030
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00011100379078904828,
+      "loss": 0.9529,
+      "step": 9035
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00011090424703487569,
+      "loss": 0.9963,
+      "step": 9040
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001108046923445054,
+      "loss": 0.9544,
+      "step": 9045
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00011070512681778375,
+      "loss": 0.9834,
+      "step": 9050
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00011060555055456807,
+      "loss": 0.9431,
+      "step": 9055
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00011050596365472637,
+      "loss": 0.9676,
+      "step": 9060
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00011040636621813736,
+      "loss": 0.948,
+      "step": 9065
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00011030675834469026,
+      "loss": 0.9499,
+      "step": 9070
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00011020714013428484,
+      "loss": 0.9566,
+      "step": 9075
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001101075116868312,
+      "loss": 0.9713,
+      "step": 9080
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0001100078731022497,
+      "loss": 0.9098,
+      "step": 9085
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00010990822448047089,
+      "loss": 0.9014,
+      "step": 9090
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00010980856592143538,
+      "loss": 0.964,
+      "step": 9095
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00010970889752509374,
+      "loss": 0.9408,
+      "step": 9100
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00010960921939140638,
+      "loss": 0.9547,
+      "step": 9105
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00010950953162034357,
+      "loss": 0.9402,
+      "step": 9110
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010940983431188508,
+      "loss": 0.9,
+      "step": 9115
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00010931012756602039,
+      "loss": 0.9944,
+      "step": 9120
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00010921041148274838,
+      "loss": 0.9479,
+      "step": 9125
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00010911068616207736,
+      "loss": 0.9428,
+      "step": 9130
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00010901095170402479,
+      "loss": 0.9801,
+      "step": 9135
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00010891120820861745,
+      "loss": 0.9957,
+      "step": 9140
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00010881145577589103,
+      "loss": 0.9769,
+      "step": 9145
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00010871169450589025,
+      "loss": 0.9695,
+      "step": 9150
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00010861192449866871,
+      "loss": 0.9266,
+      "step": 9155
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00010851214585428878,
+      "loss": 0.9607,
+      "step": 9160
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00010841235867282137,
+      "loss": 0.957,
+      "step": 9165
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00010831256305434616,
+      "loss": 0.8895,
+      "step": 9170
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00010821275909895115,
+      "loss": 0.9494,
+      "step": 9175
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00010811294690673271,
+      "loss": 0.9248,
+      "step": 9180
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00010801312657779547,
+      "loss": 0.9844,
+      "step": 9185
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00010791329821225232,
+      "loss": 0.9136,
+      "step": 9190
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00010781346191022405,
+      "loss": 0.9677,
+      "step": 9195
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00010771361777183957,
+      "loss": 1.0098,
+      "step": 9200
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00010761376589723553,
+      "loss": 1.0026,
+      "step": 9205
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00010751390638655638,
+      "loss": 0.9912,
+      "step": 9210
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00010741403933995424,
+      "loss": 0.9464,
+      "step": 9215
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00010731416485758879,
+      "loss": 0.8838,
+      "step": 9220
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00010721428303962713,
+      "loss": 1.0435,
+      "step": 9225
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00010711439398624377,
+      "loss": 0.8534,
+      "step": 9230
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00010701449779762046,
+      "loss": 0.9371,
+      "step": 9235
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00010691459457394604,
+      "loss": 0.9594,
+      "step": 9240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00010681468441541648,
+      "loss": 0.9701,
+      "step": 9245
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00010671476742223474,
+      "loss": 0.9005,
+      "step": 9250
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00010661484369461052,
+      "loss": 0.9064,
+      "step": 9255
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00010651491333276036,
+      "loss": 0.9756,
+      "step": 9260
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00010641497643690743,
+      "loss": 0.9448,
+      "step": 9265
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00010631503310728146,
+      "loss": 0.9097,
+      "step": 9270
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00010621508344411861,
+      "loss": 0.9309,
+      "step": 9275
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001061151275476614,
+      "loss": 0.89,
+      "step": 9280
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00010601516551815865,
+      "loss": 0.8777,
+      "step": 9285
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00010591519745586522,
+      "loss": 0.8984,
+      "step": 9290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00010581522346104215,
+      "loss": 0.9201,
+      "step": 9295
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00010571524363395635,
+      "loss": 0.9349,
+      "step": 9300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00010561525807488062,
+      "loss": 0.9182,
+      "step": 9305
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00010551526688409346,
+      "loss": 0.9809,
+      "step": 9310
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00010541527016187903,
+      "loss": 0.9387,
+      "step": 9315
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00010531526800852709,
+      "loss": 0.9576,
+      "step": 9320
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00010521526052433282,
+      "loss": 0.933,
+      "step": 9325
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00010511524780959667,
+      "loss": 0.9386,
+      "step": 9330
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001050152299646245,
+      "loss": 0.95,
+      "step": 9335
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00010491520708972716,
+      "loss": 0.9805,
+      "step": 9340
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001048151792852206,
+      "loss": 0.8701,
+      "step": 9345
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00010471514665142572,
+      "loss": 1.011,
+      "step": 9350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00010461510928866828,
+      "loss": 0.9755,
+      "step": 9355
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00010451506729727875,
+      "loss": 0.9517,
+      "step": 9360
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00010441502077759229,
+      "loss": 0.9555,
+      "step": 9365
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00010431496982994848,
+      "loss": 0.9624,
+      "step": 9370
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00010421491455469153,
+      "loss": 0.9754,
+      "step": 9375
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00010411485505216984,
+      "loss": 1.0126,
+      "step": 9380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00010401479142273611,
+      "loss": 0.9281,
+      "step": 9385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00010391472376674716,
+      "loss": 0.9074,
+      "step": 9390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00010381465218456383,
+      "loss": 0.9708,
+      "step": 9395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00010371457677655096,
+      "loss": 0.9924,
+      "step": 9400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001036144976430772,
+      "loss": 0.9766,
+      "step": 9405
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00010351441488451486,
+      "loss": 0.8865,
+      "step": 9410
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00010341432860124003,
+      "loss": 0.939,
+      "step": 9415
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00010331423889363223,
+      "loss": 0.9585,
+      "step": 9420
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00010321414586207443,
+      "loss": 0.9802,
+      "step": 9425
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00010311404960695299,
+      "loss": 0.9173,
+      "step": 9430
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00010301395022865738,
+      "loss": 0.96,
+      "step": 9435
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00010291384782758034,
+      "loss": 0.9746,
+      "step": 9440
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00010281374250411755,
+      "loss": 0.9577,
+      "step": 9445
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00010271363435866765,
+      "loss": 0.9737,
+      "step": 9450
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00010261352349163218,
+      "loss": 0.9668,
+      "step": 9455
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00010251341000341528,
+      "loss": 0.9797,
+      "step": 9460
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00010241329399442379,
+      "loss": 0.9331,
+      "step": 9465
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00010231317556506708,
+      "loss": 1.0343,
+      "step": 9470
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00010221305481575696,
+      "loss": 0.985,
+      "step": 9475
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00010211293184690751,
+      "loss": 0.9484,
+      "step": 9480
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00010201280675893507,
+      "loss": 0.9083,
+      "step": 9485
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00010191267965225811,
+      "loss": 0.9323,
+      "step": 9490
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00010181255062729713,
+      "loss": 0.8674,
+      "step": 9495
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00010171241978447455,
+      "loss": 0.9209,
+      "step": 9500
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001016122872242146,
+      "loss": 0.9076,
+      "step": 9505
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00010151215304694324,
+      "loss": 0.92,
+      "step": 9510
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00010141201735308805,
+      "loss": 0.9009,
+      "step": 9515
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00010131188024307817,
+      "loss": 0.8967,
+      "step": 9520
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00010121174181734405,
+      "loss": 0.959,
+      "step": 9525
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00010111160217631756,
+      "loss": 0.9591,
+      "step": 9530
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00010101146142043178,
+      "loss": 0.9943,
+      "step": 9535
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001009113196501209,
+      "loss": 0.893,
+      "step": 9540
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001008111769658201,
+      "loss": 0.8963,
+      "step": 9545
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00010071103346796549,
+      "loss": 0.902,
+      "step": 9550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.000100610889256994,
+      "loss": 0.9108,
+      "step": 9555
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00010051074443334327,
+      "loss": 0.9926,
+      "step": 9560
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00010041059909745156,
+      "loss": 0.9072,
+      "step": 9565
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00010031045334975768,
+      "loss": 0.9536,
+      "step": 9570
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00010021030729070076,
+      "loss": 0.9899,
+      "step": 9575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00010011016102072033,
+      "loss": 0.904,
+      "step": 9580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001000100146402561,
+      "loss": 0.9465,
+      "step": 9585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.25,
+      "learning_rate": 9.990986824974788e-05,
+      "loss": 0.862,
+      "step": 9590
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.980972194963552e-05,
+      "loss": 0.9563,
+      "step": 9595
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.970957584035873e-05,
+      "loss": 0.9042,
+      "step": 9600
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.96094300223571e-05,
+      "loss": 1.0266,
+      "step": 9605
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.950928459606984e-05,
+      "loss": 0.9222,
+      "step": 9610
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.940913966193586e-05,
+      "loss": 0.8897,
+      "step": 9615
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.930899532039347e-05,
+      "loss": 1.022,
+      "step": 9620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.920885167188054e-05,
+      "loss": 0.9223,
+      "step": 9625
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 9.910870881683402e-05,
+      "loss": 0.9112,
+      "step": 9630
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.900856685569027e-05,
+      "loss": 0.9659,
+      "step": 9635
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.25,
+      "learning_rate": 9.890842588888474e-05,
+      "loss": 0.8959,
+      "step": 9640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.88082860168517e-05,
+      "loss": 0.9726,
+      "step": 9645
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.248046875,
+      "learning_rate": 9.870814734002456e-05,
+      "loss": 1.028,
+      "step": 9650
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.860800995883533e-05,
+      "loss": 0.9658,
+      "step": 9655
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.850787397371482e-05,
+      "loss": 0.9798,
+      "step": 9660
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.840773948509243e-05,
+      "loss": 0.9827,
+      "step": 9665
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.83076065933961e-05,
+      "loss": 0.914,
+      "step": 9670
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.820747539905202e-05,
+      "loss": 1.0198,
+      "step": 9675
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.810734600248486e-05,
+      "loss": 0.882,
+      "step": 9680
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.800721850411743e-05,
+      "loss": 0.921,
+      "step": 9685
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.790709300437052e-05,
+      "loss": 0.9465,
+      "step": 9690
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.780696960366311e-05,
+      "loss": 0.9131,
+      "step": 9695
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.770684840241191e-05,
+      "loss": 0.9612,
+      "step": 9700
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.76067295010315e-05,
+      "loss": 0.8921,
+      "step": 9705
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 9.750661299993415e-05,
+      "loss": 0.8779,
+      "step": 9710
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.740649899952967e-05,
+      "loss": 0.9104,
+      "step": 9715
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.248046875,
+      "learning_rate": 9.73063876002255e-05,
+      "loss": 0.9074,
+      "step": 9720
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.720627890242628e-05,
+      "loss": 0.8993,
+      "step": 9725
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.710617300653412e-05,
+      "loss": 1.0271,
+      "step": 9730
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.700607001294814e-05,
+      "loss": 0.8672,
+      "step": 9735
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.690597002206477e-05,
+      "loss": 0.9625,
+      "step": 9740
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.68058731342772e-05,
+      "loss": 0.9321,
+      "step": 9745
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.670577944997566e-05,
+      "loss": 0.9249,
+      "step": 9750
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.660568906954711e-05,
+      "loss": 0.9381,
+      "step": 9755
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.65056020933752e-05,
+      "loss": 0.948,
+      "step": 9760
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.640551862184021e-05,
+      "loss": 0.97,
+      "step": 9765
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.630543875531879e-05,
+      "loss": 0.9537,
+      "step": 9770
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.620536259418416e-05,
+      "loss": 0.8846,
+      "step": 9775
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.610529023880561e-05,
+      "loss": 1.0599,
+      "step": 9780
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.600522178954879e-05,
+      "loss": 0.9412,
+      "step": 9785
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.590515734677531e-05,
+      "loss": 0.9605,
+      "step": 9790
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.580509701084286e-05,
+      "loss": 0.9098,
+      "step": 9795
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.570504088210496e-05,
+      "loss": 0.9473,
+      "step": 9800
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.25,
+      "learning_rate": 9.560498906091085e-05,
+      "loss": 0.9205,
+      "step": 9805
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.550494164760562e-05,
+      "loss": 0.9717,
+      "step": 9810
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.540489874252972e-05,
+      "loss": 0.9545,
+      "step": 9815
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.53048604460193e-05,
+      "loss": 0.988,
+      "step": 9820
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.52048268584057e-05,
+      "loss": 0.8669,
+      "step": 9825
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.510479808001566e-05,
+      "loss": 0.9643,
+      "step": 9830
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.500477421117102e-05,
+      "loss": 0.9156,
+      "step": 9835
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.490475535218875e-05,
+      "loss": 0.9286,
+      "step": 9840
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.480474160338082e-05,
+      "loss": 0.9137,
+      "step": 9845
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.470473306505392e-05,
+      "loss": 0.8888,
+      "step": 9850
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.460472983750977e-05,
+      "loss": 0.9228,
+      "step": 9855
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.450473202104448e-05,
+      "loss": 0.9923,
+      "step": 9860
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.440473971594895e-05,
+      "loss": 0.9869,
+      "step": 9865
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.430475302250844e-05,
+      "loss": 0.9735,
+      "step": 9870
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.420477204100264e-05,
+      "loss": 0.9472,
+      "step": 9875
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.41047968717055e-05,
+      "loss": 0.9805,
+      "step": 9880
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 9.400482761488507e-05,
+      "loss": 0.9235,
+      "step": 9885
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.390486437080361e-05,
+      "loss": 0.8878,
+      "step": 9890
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.380490723971717e-05,
+      "loss": 0.9191,
+      "step": 9895
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.370495632187587e-05,
+      "loss": 0.9409,
+      "step": 9900
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.360501171752339e-05,
+      "loss": 0.9563,
+      "step": 9905
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.35050735268973e-05,
+      "loss": 0.9377,
+      "step": 9910
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.340514185022851e-05,
+      "loss": 0.9467,
+      "step": 9915
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.330521678774157e-05,
+      "loss": 0.9456,
+      "step": 9920
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 9.320529843965432e-05,
+      "loss": 0.904,
+      "step": 9925
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.310538690617788e-05,
+      "loss": 0.9357,
+      "step": 9930
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.300548228751657e-05,
+      "loss": 0.9386,
+      "step": 9935
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.290558468386765e-05,
+      "loss": 0.9631,
+      "step": 9940
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.280569419542154e-05,
+      "loss": 1.0054,
+      "step": 9945
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.270581092236134e-05,
+      "loss": 0.9002,
+      "step": 9950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.260593496486302e-05,
+      "loss": 0.9869,
+      "step": 9955
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.250606642309523e-05,
+      "loss": 0.9615,
+      "step": 9960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.240620539721904e-05,
+      "loss": 0.9069,
+      "step": 9965
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.23063519873882e-05,
+      "loss": 0.9112,
+      "step": 9970
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.220650629374862e-05,
+      "loss": 0.9132,
+      "step": 9975
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.210666841643857e-05,
+      "loss": 0.9454,
+      "step": 9980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.200683845558845e-05,
+      "loss": 0.9629,
+      "step": 9985
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.190701651132079e-05,
+      "loss": 0.9644,
+      "step": 9990
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.180720268374992e-05,
+      "loss": 0.9684,
+      "step": 9995
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.24609375,
+      "learning_rate": 9.170739707298221e-05,
+      "loss": 0.9239,
+      "step": 10000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.160759977911576e-05,
+      "loss": 0.8766,
+      "step": 10005
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.150781090224015e-05,
+      "loss": 0.973,
+      "step": 10010
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.140803054243677e-05,
+      "loss": 0.9588,
+      "step": 10015
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.130825879977828e-05,
+      "loss": 0.9172,
+      "step": 10020
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.12084957743288e-05,
+      "loss": 0.9477,
+      "step": 10025
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.110874156614362e-05,
+      "loss": 0.9445,
+      "step": 10030
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.100899627526933e-05,
+      "loss": 0.9084,
+      "step": 10035
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.25,
+      "learning_rate": 9.090926000174338e-05,
+      "loss": 0.8938,
+      "step": 10040
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.080953284559433e-05,
+      "loss": 0.9041,
+      "step": 10045
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.070981490684159e-05,
+      "loss": 0.9502,
+      "step": 10050
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.061010628549522e-05,
+      "loss": 0.9194,
+      "step": 10055
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.240234375,
+      "learning_rate": 9.051040708155606e-05,
+      "loss": 0.9056,
+      "step": 10060
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.041071739501538e-05,
+      "loss": 0.9462,
+      "step": 10065
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.0311037325855e-05,
+      "loss": 0.8849,
+      "step": 10070
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.021136697404706e-05,
+      "loss": 0.9928,
+      "step": 10075
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.011170643955394e-05,
+      "loss": 0.91,
+      "step": 10080
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.001205582232825e-05,
+      "loss": 0.913,
+      "step": 10085
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.991241522231252e-05,
+      "loss": 0.9421,
+      "step": 10090
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2421875,
+      "learning_rate": 8.981278473943936e-05,
+      "loss": 0.9712,
+      "step": 10095
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.271484375,
+      "learning_rate": 8.971316447363115e-05,
+      "loss": 0.9144,
+      "step": 10100
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.96135545248001e-05,
+      "loss": 0.9127,
+      "step": 10105
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.267578125,
+      "learning_rate": 8.951395499284797e-05,
+      "loss": 0.9771,
+      "step": 10110
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.263671875,
+      "learning_rate": 8.941436597766616e-05,
+      "loss": 0.9841,
+      "step": 10115
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 8.93147875791355e-05,
+      "loss": 0.9138,
+      "step": 10120
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.23828125,
+      "learning_rate": 8.921521989712611e-05,
+      "loss": 0.9643,
+      "step": 10125
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 8.91156630314975e-05,
+      "loss": 0.9428,
+      "step": 10130
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.267578125,
+      "learning_rate": 8.901611708209818e-05,
+      "loss": 0.9558,
+      "step": 10135
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 8.891658214876585e-05,
+      "loss": 0.9425,
+      "step": 10140
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.23828125,
+      "learning_rate": 8.8817058331327e-05,
+      "loss": 0.9372,
+      "step": 10145
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.271484375,
+      "learning_rate": 8.871754572959716e-05,
+      "loss": 0.9263,
+      "step": 10150
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 8.861804444338045e-05,
+      "loss": 0.9587,
+      "step": 10155
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 8.851855457246972e-05,
+      "loss": 0.9165,
+      "step": 10160
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.841907621664638e-05,
+      "loss": 0.9305,
+      "step": 10165
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.83196094756802e-05,
+      "loss": 0.8434,
+      "step": 10170
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.27734375,
+      "learning_rate": 8.82201544493295e-05,
+      "loss": 0.9661,
+      "step": 10175
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.244140625,
+      "learning_rate": 8.812071123734058e-05,
+      "loss": 0.9175,
+      "step": 10180
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.263671875,
+      "learning_rate": 8.802127993944814e-05,
+      "loss": 0.9114,
+      "step": 10185
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 8.792186065537473e-05,
+      "loss": 0.9737,
+      "step": 10190
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.255859375,
+      "learning_rate": 8.782245348483104e-05,
+      "loss": 0.992,
+      "step": 10195
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 8.772305852751542e-05,
+      "loss": 0.9458,
+      "step": 10200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.294921875,
+      "learning_rate": 8.762367588311414e-05,
+      "loss": 0.8808,
+      "step": 10205
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.263671875,
+      "learning_rate": 8.752430565130103e-05,
+      "loss": 0.9249,
+      "step": 10210
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.265625,
+      "learning_rate": 8.742494793173743e-05,
+      "loss": 0.9526,
+      "step": 10215
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.25390625,
+      "learning_rate": 8.73256028240723e-05,
+      "loss": 1.0145,
+      "step": 10220
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.265625,
+      "learning_rate": 8.722627042794171e-05,
+      "loss": 0.9798,
+      "step": 10225
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.255859375,
+      "learning_rate": 8.712695084296924e-05,
+      "loss": 0.8368,
+      "step": 10230
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.702764416876537e-05,
+      "loss": 0.9953,
+      "step": 10235
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.255859375,
+      "learning_rate": 8.692835050492785e-05,
+      "loss": 0.9289,
+      "step": 10240
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 8.682906995104125e-05,
+      "loss": 0.8819,
+      "step": 10245
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2890625,
+      "learning_rate": 8.672980260667702e-05,
+      "loss": 0.9383,
+      "step": 10250
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.25,
+      "learning_rate": 8.663054857139339e-05,
+      "loss": 0.9766,
+      "step": 10255
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.271484375,
+      "learning_rate": 8.653130794473517e-05,
+      "loss": 0.9114,
+      "step": 10260
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 8.643208082623386e-05,
+      "loss": 0.8984,
+      "step": 10265
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2734375,
+      "learning_rate": 8.633286731540722e-05,
+      "loss": 0.9411,
+      "step": 10270
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.259765625,
+      "learning_rate": 8.623366751175958e-05,
+      "loss": 0.9607,
+      "step": 10275
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 8.613448151478131e-05,
+      "loss": 0.9819,
+      "step": 10280
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.603530942394908e-05,
+      "loss": 0.9658,
+      "step": 10285
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.593615133872558e-05,
+      "loss": 0.9323,
+      "step": 10290
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.26171875,
+      "learning_rate": 8.583700735855941e-05,
+      "loss": 0.9315,
+      "step": 10295
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.294921875,
+      "learning_rate": 8.57378775828851e-05,
+      "loss": 0.9985,
+      "step": 10300
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.255859375,
+      "learning_rate": 8.563876211112282e-05,
+      "loss": 0.9708,
+      "step": 10305
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.553966104267852e-05,
+      "loss": 0.9249,
+      "step": 10310
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.259765625,
+      "learning_rate": 8.544057447694358e-05,
+      "loss": 0.9349,
+      "step": 10315
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.259765625,
+      "learning_rate": 8.534150251329494e-05,
+      "loss": 0.9589,
+      "step": 10320
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.259765625,
+      "learning_rate": 8.52424452510949e-05,
+      "loss": 0.926,
+      "step": 10325
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.248046875,
+      "learning_rate": 8.514340278969089e-05,
+      "loss": 0.87,
+      "step": 10330
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.265625,
+      "learning_rate": 8.504437522841566e-05,
+      "loss": 0.966,
+      "step": 10335
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 8.494536266658687e-05,
+      "loss": 0.9709,
+      "step": 10340
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.255859375,
+      "learning_rate": 8.484636520350724e-05,
+      "loss": 1.087,
+      "step": 10345
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.474738293846424e-05,
+      "loss": 0.9734,
+      "step": 10350
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.464841597073024e-05,
+      "loss": 0.9621,
+      "step": 10355
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.267578125,
+      "learning_rate": 8.454946439956213e-05,
+      "loss": 0.9551,
+      "step": 10360
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2734375,
+      "learning_rate": 8.445052832420146e-05,
+      "loss": 0.9701,
+      "step": 10365
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.435160784387423e-05,
+      "loss": 0.9783,
+      "step": 10370
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28125,
+      "learning_rate": 8.425270305779069e-05,
+      "loss": 0.9546,
+      "step": 10375
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2734375,
+      "learning_rate": 8.415381406514551e-05,
+      "loss": 0.9028,
+      "step": 10380
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.24609375,
+      "learning_rate": 8.405494096511737e-05,
+      "loss": 0.9537,
+      "step": 10385
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.395608385686911e-05,
+      "loss": 0.8804,
+      "step": 10390
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.25,
+      "learning_rate": 8.38572428395475e-05,
+      "loss": 0.9221,
+      "step": 10395
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.265625,
+      "learning_rate": 8.37584180122832e-05,
+      "loss": 0.9675,
+      "step": 10400
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.365960947419054e-05,
+      "loss": 0.9677,
+      "step": 10405
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.356081732436759e-05,
+      "loss": 0.9545,
+      "step": 10410
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 8.346204166189607e-05,
+      "loss": 0.8864,
+      "step": 10415
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 8.336328258584093e-05,
+      "loss": 0.9734,
+      "step": 10420
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.25,
+      "learning_rate": 8.326454019525072e-05,
+      "loss": 0.9281,
+      "step": 10425
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.26171875,
+      "learning_rate": 8.316581458915711e-05,
+      "loss": 0.9169,
+      "step": 10430
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.267578125,
+      "learning_rate": 8.3067105866575e-05,
+      "loss": 0.9249,
+      "step": 10435
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.271484375,
+      "learning_rate": 8.296841412650233e-05,
+      "loss": 1.0116,
+      "step": 10440
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.265625,
+      "learning_rate": 8.286973946792e-05,
+      "loss": 0.9494,
+      "step": 10445
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.277108198979188e-05,
+      "loss": 0.9606,
+      "step": 10450
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.265625,
+      "learning_rate": 8.267244179106441e-05,
+      "loss": 0.9395,
+      "step": 10455
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.257381897066691e-05,
+      "loss": 0.9427,
+      "step": 10460
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.265625,
+      "learning_rate": 8.24752136275111e-05,
+      "loss": 0.9139,
+      "step": 10465
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.259765625,
+      "learning_rate": 8.237662586049133e-05,
+      "loss": 1.0039,
+      "step": 10470
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.306640625,
+      "learning_rate": 8.227805576848418e-05,
+      "loss": 0.9437,
+      "step": 10475
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28125,
+      "learning_rate": 8.217950345034858e-05,
+      "loss": 0.9363,
+      "step": 10480
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.263671875,
+      "learning_rate": 8.208096900492562e-05,
+      "loss": 0.9247,
+      "step": 10485
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.198245253103843e-05,
+      "loss": 0.9966,
+      "step": 10490
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.188395412749223e-05,
+      "loss": 1.0299,
+      "step": 10495
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.255859375,
+      "learning_rate": 8.178547389307393e-05,
+      "loss": 0.9266,
+      "step": 10500
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.168701192655243e-05,
+      "loss": 1.0203,
+      "step": 10505
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.283203125,
+      "learning_rate": 8.158856832667811e-05,
+      "loss": 0.9132,
+      "step": 10510
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.14901431921831e-05,
+      "loss": 0.9272,
+      "step": 10515
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.139173662178086e-05,
+      "loss": 0.9588,
+      "step": 10520
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.271484375,
+      "learning_rate": 8.129334871416632e-05,
+      "loss": 0.9455,
+      "step": 10525
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.25,
+      "learning_rate": 8.119497956801571e-05,
+      "loss": 0.9381,
+      "step": 10530
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.109662928198638e-05,
+      "loss": 1.0069,
+      "step": 10535
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.287109375,
+      "learning_rate": 8.099829795471683e-05,
+      "loss": 0.9718,
+      "step": 10540
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.26171875,
+      "learning_rate": 8.089998568482643e-05,
+      "loss": 0.9662,
+      "step": 10545
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.080169257091562e-05,
+      "loss": 0.9271,
+      "step": 10550
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2734375,
+      "learning_rate": 8.070341871156541e-05,
+      "loss": 1.0115,
+      "step": 10555
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.255859375,
+      "learning_rate": 8.060516420533774e-05,
+      "loss": 0.938,
+      "step": 10560
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.050692915077489e-05,
+      "loss": 0.8801,
+      "step": 10565
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.26171875,
+      "learning_rate": 8.040871364639983e-05,
+      "loss": 0.9114,
+      "step": 10570
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.265625,
+      "learning_rate": 8.031051779071587e-05,
+      "loss": 0.9507,
+      "step": 10575
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.021234168220649e-05,
+      "loss": 0.9484,
+      "step": 10580
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.011418541933558e-05,
+      "loss": 0.9998,
+      "step": 10585
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.259765625,
+      "learning_rate": 8.00160491005469e-05,
+      "loss": 0.9445,
+      "step": 10590
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.251953125,
+      "learning_rate": 7.991793282426442e-05,
+      "loss": 0.9108,
+      "step": 10595
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.271484375,
+      "learning_rate": 7.981983668889182e-05,
+      "loss": 0.9833,
+      "step": 10600
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2734375,
+      "learning_rate": 7.972176079281275e-05,
+      "loss": 0.9454,
+      "step": 10605
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.271484375,
+      "learning_rate": 7.962370523439044e-05,
+      "loss": 0.9443,
+      "step": 10610
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.28515625,
+      "learning_rate": 7.952567011196774e-05,
+      "loss": 0.9786,
+      "step": 10615
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2734375,
+      "learning_rate": 7.942765552386709e-05,
+      "loss": 0.9386,
+      "step": 10620
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.255859375,
+      "learning_rate": 7.932966156839018e-05,
+      "loss": 1.0092,
+      "step": 10625
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 7.923168834381822e-05,
+      "loss": 0.8734,
+      "step": 10630
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.913373594841139e-05,
+      "loss": 0.9287,
+      "step": 10635
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 7.903580448040917e-05,
+      "loss": 0.9007,
+      "step": 10640
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.25,
+      "learning_rate": 7.893789403802992e-05,
+      "loss": 0.9592,
+      "step": 10645
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.25390625,
+      "learning_rate": 7.884000471947104e-05,
+      "loss": 0.9578,
+      "step": 10650
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.265625,
+      "learning_rate": 7.874213662290862e-05,
+      "loss": 0.9031,
+      "step": 10655
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.26171875,
+      "learning_rate": 7.864428984649757e-05,
+      "loss": 0.9133,
+      "step": 10660
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.251953125,
+      "learning_rate": 7.854646448837134e-05,
+      "loss": 0.9158,
+      "step": 10665
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.25,
+      "learning_rate": 7.844866064664189e-05,
+      "loss": 0.9334,
+      "step": 10670
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.255859375,
+      "learning_rate": 7.835087841939973e-05,
+      "loss": 0.9653,
+      "step": 10675
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.26171875,
+      "learning_rate": 7.82531179047135e-05,
+      "loss": 0.9604,
+      "step": 10680
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 7.815537920063019e-05,
+      "loss": 1.0108,
+      "step": 10685
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 7.805766240517498e-05,
+      "loss": 0.9327,
+      "step": 10690
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.795996761635087e-05,
+      "loss": 0.9369,
+      "step": 10695
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.786229493213901e-05,
+      "loss": 0.9592,
+      "step": 10700
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.776464445049817e-05,
+      "loss": 0.8983,
+      "step": 10705
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.263671875,
+      "learning_rate": 7.766701626936505e-05,
+      "loss": 0.9935,
+      "step": 10710
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.275390625,
+      "learning_rate": 7.75694104866538e-05,
+      "loss": 0.9206,
+      "step": 10715
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.74718272002563e-05,
+      "loss": 0.9622,
+      "step": 10720
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.737426650804168e-05,
+      "loss": 0.9379,
+      "step": 10725
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.263671875,
+      "learning_rate": 7.727672850785651e-05,
+      "loss": 0.9329,
+      "step": 10730
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.244140625,
+      "learning_rate": 7.717921329752466e-05,
+      "loss": 1.0257,
+      "step": 10735
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.259765625,
+      "learning_rate": 7.708172097484699e-05,
+      "loss": 0.8839,
+      "step": 10740
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.26171875,
+      "learning_rate": 7.698425163760156e-05,
+      "loss": 0.9445,
+      "step": 10745
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.26171875,
+      "learning_rate": 7.688680538354323e-05,
+      "loss": 0.9536,
+      "step": 10750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.255859375,
+      "learning_rate": 7.678938231040383e-05,
+      "loss": 0.8843,
+      "step": 10755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.24609375,
+      "learning_rate": 7.669198251589188e-05,
+      "loss": 0.9955,
+      "step": 10760
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.240234375,
+      "learning_rate": 7.659460609769252e-05,
+      "loss": 1.0015,
+      "step": 10765
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.255859375,
+      "learning_rate": 7.649725315346761e-05,
+      "loss": 0.9057,
+      "step": 10770
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.639992378085521e-05,
+      "loss": 0.9174,
+      "step": 10775
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.25390625,
+      "learning_rate": 7.630261807747e-05,
+      "loss": 0.929,
+      "step": 10780
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 7.620533614090269e-05,
+      "loss": 0.9768,
+      "step": 10785
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.263671875,
+      "learning_rate": 7.610807806872038e-05,
+      "loss": 1.0064,
+      "step": 10790
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.275390625,
+      "learning_rate": 7.601084395846603e-05,
+      "loss": 0.9434,
+      "step": 10795
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.248046875,
+      "learning_rate": 7.591363390765868e-05,
+      "loss": 0.9428,
+      "step": 10800
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 7.581644801379324e-05,
+      "loss": 0.9157,
+      "step": 10805
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.265625,
+      "learning_rate": 7.571928637434031e-05,
+      "loss": 0.948,
+      "step": 10810
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.26171875,
+      "learning_rate": 7.562214908674633e-05,
+      "loss": 1.0172,
+      "step": 10815
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.26171875,
+      "learning_rate": 7.55250362484331e-05,
+      "loss": 0.9701,
+      "step": 10820
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.251953125,
+      "learning_rate": 7.542794795679811e-05,
+      "loss": 0.9276,
+      "step": 10825
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.26171875,
+      "learning_rate": 7.533088430921402e-05,
+      "loss": 0.9317,
+      "step": 10830
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.279296875,
+      "learning_rate": 7.5233845403029e-05,
+      "loss": 0.9199,
+      "step": 10835
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 7.51368313355662e-05,
+      "loss": 0.8962,
+      "step": 10840
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.5039842204124e-05,
+      "loss": 0.8965,
+      "step": 10845
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.248046875,
+      "learning_rate": 7.49428781059757e-05,
+      "loss": 0.9054,
+      "step": 10850
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.26171875,
+      "learning_rate": 7.484593913836951e-05,
+      "loss": 0.9142,
+      "step": 10855
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.474902539852848e-05,
+      "loss": 0.9123,
+      "step": 10860
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.26171875,
+      "learning_rate": 7.465213698365026e-05,
+      "loss": 0.9701,
+      "step": 10865
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.255859375,
+      "learning_rate": 7.455527399090721e-05,
+      "loss": 0.97,
+      "step": 10870
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.251953125,
+      "learning_rate": 7.445843651744609e-05,
+      "loss": 0.9103,
+      "step": 10875
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.26171875,
+      "learning_rate": 7.436162466038818e-05,
+      "loss": 0.9494,
+      "step": 10880
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.251953125,
+      "learning_rate": 7.426483851682898e-05,
+      "loss": 0.9303,
+      "step": 10885
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.251953125,
+      "learning_rate": 7.416807818383817e-05,
+      "loss": 0.9832,
+      "step": 10890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 7.407134375845972e-05,
+      "loss": 0.8896,
+      "step": 10895
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.25,
+      "learning_rate": 7.397463533771139e-05,
+      "loss": 0.9505,
+      "step": 10900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.255859375,
+      "learning_rate": 7.387795301858504e-05,
+      "loss": 0.9017,
+      "step": 10905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.26171875,
+      "learning_rate": 7.378129689804623e-05,
+      "loss": 0.9594,
+      "step": 10910
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.259765625,
+      "learning_rate": 7.368466707303434e-05,
+      "loss": 0.9913,
+      "step": 10915
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.265625,
+      "learning_rate": 7.358806364046226e-05,
+      "loss": 1.0053,
+      "step": 10920
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.279296875,
+      "learning_rate": 7.349148669721658e-05,
+      "loss": 0.9282,
+      "step": 10925
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.25,
+      "learning_rate": 7.339493634015711e-05,
+      "loss": 0.9456,
+      "step": 10930
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.244140625,
+      "learning_rate": 7.329841266611721e-05,
+      "loss": 0.9272,
+      "step": 10935
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.283203125,
+      "learning_rate": 7.320191577190336e-05,
+      "loss": 0.9377,
+      "step": 10940
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.25390625,
+      "learning_rate": 7.310544575429514e-05,
+      "loss": 0.8861,
+      "step": 10945
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.259765625,
+      "learning_rate": 7.300900271004534e-05,
+      "loss": 0.9117,
+      "step": 10950
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.296875,
+      "learning_rate": 7.291258673587947e-05,
+      "loss": 0.9802,
+      "step": 10955
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.281619792849612e-05,
+      "loss": 0.9195,
+      "step": 10960
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.271983638456644e-05,
+      "loss": 1.0314,
+      "step": 10965
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.251953125,
+      "learning_rate": 7.26235022007344e-05,
+      "loss": 0.9812,
+      "step": 10970
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.252719547361641e-05,
+      "loss": 0.9425,
+      "step": 10975
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.243091629980141e-05,
+      "loss": 0.9824,
+      "step": 10980
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.271484375,
+      "learning_rate": 7.233466477585068e-05,
+      "loss": 0.8939,
+      "step": 10985
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.26171875,
+      "learning_rate": 7.223844099829773e-05,
+      "loss": 0.9015,
+      "step": 10990
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.234375,
+      "learning_rate": 7.214224506364834e-05,
+      "loss": 0.8971,
+      "step": 10995
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.283203125,
+      "learning_rate": 7.204607706838026e-05,
+      "loss": 0.9336,
+      "step": 11000
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.275390625,
+      "learning_rate": 7.194993710894335e-05,
+      "loss": 1.0268,
+      "step": 11005
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.259765625,
+      "learning_rate": 7.185382528175917e-05,
+      "loss": 0.8822,
+      "step": 11010
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.26171875,
+      "learning_rate": 7.175774168322123e-05,
+      "loss": 1.0541,
+      "step": 11015
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.25,
+      "learning_rate": 7.166168640969464e-05,
+      "loss": 0.9136,
+      "step": 11020
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.255859375,
+      "learning_rate": 7.156565955751616e-05,
+      "loss": 0.9599,
+      "step": 11025
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.25390625,
+      "learning_rate": 7.146966122299396e-05,
+      "loss": 0.9775,
+      "step": 11030
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.265625,
+      "learning_rate": 7.137369150240769e-05,
+      "loss": 0.977,
+      "step": 11035
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.28515625,
+      "learning_rate": 7.127775049200828e-05,
+      "loss": 0.9294,
+      "step": 11040
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.118183828801781e-05,
+      "loss": 0.9713,
+      "step": 11045
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.108595498662956e-05,
+      "loss": 0.924,
+      "step": 11050
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.255859375,
+      "learning_rate": 7.099010068400781e-05,
+      "loss": 0.9111,
+      "step": 11055
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 7.089427547628766e-05,
+      "loss": 0.9591,
+      "step": 11060
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.24609375,
+      "learning_rate": 7.079847945957516e-05,
+      "loss": 0.8856,
+      "step": 11065
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27734375,
+      "learning_rate": 7.070271272994698e-05,
+      "loss": 0.9428,
+      "step": 11070
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.248046875,
+      "learning_rate": 7.060697538345048e-05,
+      "loss": 0.9358,
+      "step": 11075
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.251953125,
+      "learning_rate": 7.051126751610346e-05,
+      "loss": 0.9376,
+      "step": 11080
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.26171875,
+      "learning_rate": 7.041558922389434e-05,
+      "loss": 0.9481,
+      "step": 11085
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.263671875,
+      "learning_rate": 7.031994060278162e-05,
+      "loss": 0.9089,
+      "step": 11090
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.26953125,
+      "learning_rate": 7.02243217486943e-05,
+      "loss": 0.9229,
+      "step": 11095
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.248046875,
+      "learning_rate": 7.012873275753137e-05,
+      "loss": 0.8838,
+      "step": 11100
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.271484375,
+      "learning_rate": 7.003317372516189e-05,
+      "loss": 0.9375,
+      "step": 11105
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.25,
+      "learning_rate": 6.993764474742493e-05,
+      "loss": 0.8994,
+      "step": 11110
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 6.984214592012935e-05,
+      "loss": 0.9055,
+      "step": 11115
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.28515625,
+      "learning_rate": 6.974667733905377e-05,
+      "loss": 0.9884,
+      "step": 11120
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.251953125,
+      "learning_rate": 6.965123909994658e-05,
+      "loss": 0.9248,
+      "step": 11125
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 6.955583129852559e-05,
+      "loss": 0.9173,
+      "step": 11130
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.244140625,
+      "learning_rate": 6.946045403047821e-05,
+      "loss": 0.9609,
+      "step": 11135
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.279296875,
+      "learning_rate": 6.936510739146113e-05,
+      "loss": 0.9058,
+      "step": 11140
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.28515625,
+      "learning_rate": 6.926979147710044e-05,
+      "loss": 0.9753,
+      "step": 11145
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.25390625,
+      "learning_rate": 6.917450638299123e-05,
+      "loss": 0.9487,
+      "step": 11150
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.255859375,
+      "learning_rate": 6.90792522046979e-05,
+      "loss": 0.9352,
+      "step": 11155
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.259765625,
+      "learning_rate": 6.898402903775369e-05,
+      "loss": 0.9578,
+      "step": 11160
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.279296875,
+      "learning_rate": 6.888883697766076e-05,
+      "loss": 0.9074,
+      "step": 11165
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.275390625,
+      "learning_rate": 6.87936761198901e-05,
+      "loss": 1.0342,
+      "step": 11170
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.259765625,
+      "learning_rate": 6.869854655988139e-05,
+      "loss": 0.9576,
+      "step": 11175
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.283203125,
+      "learning_rate": 6.860344839304299e-05,
+      "loss": 0.9478,
+      "step": 11180
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.267578125,
+      "learning_rate": 6.850838171475165e-05,
+      "loss": 0.9573,
+      "step": 11185
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.251953125,
+      "learning_rate": 6.841334662035266e-05,
+      "loss": 0.963,
+      "step": 11190
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.259765625,
+      "learning_rate": 6.83183432051595e-05,
+      "loss": 0.9702,
+      "step": 11195
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.263671875,
+      "learning_rate": 6.822337156445406e-05,
+      "loss": 0.939,
+      "step": 11200
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.263671875,
+      "learning_rate": 6.812843179348618e-05,
+      "loss": 0.9908,
+      "step": 11205
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.271484375,
+      "learning_rate": 6.803352398747384e-05,
+      "loss": 0.9683,
+      "step": 11210
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 6.793864824160295e-05,
+      "loss": 0.9921,
+      "step": 11215
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.279296875,
+      "learning_rate": 6.78438046510272e-05,
+      "loss": 0.9828,
+      "step": 11220
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.265625,
+      "learning_rate": 6.774899331086814e-05,
+      "loss": 1.0154,
+      "step": 11225
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.267578125,
+      "learning_rate": 6.765421431621491e-05,
+      "loss": 0.9471,
+      "step": 11230
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2578125,
+      "learning_rate": 6.755946776212421e-05,
+      "loss": 0.9154,
+      "step": 11235
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.244140625,
+      "learning_rate": 6.746475374362018e-05,
+      "loss": 0.8874,
+      "step": 11240
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.265625,
+      "learning_rate": 6.737007235569442e-05,
+      "loss": 0.9187,
+      "step": 11245
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.283203125,
+      "learning_rate": 6.727542369330571e-05,
+      "loss": 0.9596,
+      "step": 11250
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 6.718080785138002e-05,
+      "loss": 0.9327,
+      "step": 11255
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.259765625,
+      "learning_rate": 6.708622492481051e-05,
+      "loss": 0.9895,
+      "step": 11260
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.251953125,
+      "learning_rate": 6.699167500845714e-05,
+      "loss": 0.9862,
+      "step": 11265
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.248046875,
+      "learning_rate": 6.689715819714697e-05,
+      "loss": 0.9216,
+      "step": 11270
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.283203125,
+      "learning_rate": 6.680267458567366e-05,
+      "loss": 0.9501,
+      "step": 11275
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.248046875,
+      "learning_rate": 6.670822426879776e-05,
+      "loss": 0.916,
+      "step": 11280
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.275390625,
+      "learning_rate": 6.661380734124625e-05,
+      "loss": 0.9404,
+      "step": 11285
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 6.65194238977128e-05,
+      "loss": 0.9192,
+      "step": 11290
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 6.642507403285732e-05,
+      "loss": 0.9325,
+      "step": 11295
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.26953125,
+      "learning_rate": 6.633075784130619e-05,
+      "loss": 0.9313,
+      "step": 11300
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.32421875,
+      "learning_rate": 6.623647541765195e-05,
+      "loss": 1.0401,
+      "step": 11305
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.614222685645324e-05,
+      "loss": 1.0222,
+      "step": 11310
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.255859375,
+      "learning_rate": 6.604801225223486e-05,
+      "loss": 0.892,
+      "step": 11315
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.24609375,
+      "learning_rate": 6.595383169948738e-05,
+      "loss": 0.8829,
+      "step": 11320
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 6.58596852926674e-05,
+      "loss": 0.9097,
+      "step": 11325
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.267578125,
+      "learning_rate": 6.576557312619711e-05,
+      "loss": 0.9305,
+      "step": 11330
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.267578125,
+      "learning_rate": 6.567149529446447e-05,
+      "loss": 0.9914,
+      "step": 11335
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.26953125,
+      "learning_rate": 6.5577451891823e-05,
+      "loss": 0.9554,
+      "step": 11340
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2578125,
+      "learning_rate": 6.548344301259161e-05,
+      "loss": 0.9622,
+      "step": 11345
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.53894687510547e-05,
+      "loss": 0.9283,
+      "step": 11350
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.259765625,
+      "learning_rate": 6.52955292014618e-05,
+      "loss": 0.926,
+      "step": 11355
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2734375,
+      "learning_rate": 6.52016244580278e-05,
+      "loss": 0.9567,
+      "step": 11360
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.259765625,
+      "learning_rate": 6.51077546149325e-05,
+      "loss": 0.9689,
+      "step": 11365
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23828125,
+      "learning_rate": 6.50139197663209e-05,
+      "loss": 0.9252,
+      "step": 11370
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2578125,
+      "learning_rate": 6.492012000630269e-05,
+      "loss": 0.9499,
+      "step": 11375
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.482635542895255e-05,
+      "loss": 0.9449,
+      "step": 11380
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.267578125,
+      "learning_rate": 6.473262612830977e-05,
+      "loss": 1.0348,
+      "step": 11385
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.255859375,
+      "learning_rate": 6.46389321983783e-05,
+      "loss": 0.9567,
+      "step": 11390
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 6.45452737331266e-05,
+      "loss": 0.9624,
+      "step": 11395
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 6.445165082648755e-05,
+      "loss": 0.9057,
+      "step": 11400
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23828125,
+      "learning_rate": 6.43580635723584e-05,
+      "loss": 0.9551,
+      "step": 11405
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2890625,
+      "learning_rate": 6.426451206460061e-05,
+      "loss": 0.9804,
+      "step": 11410
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.244140625,
+      "learning_rate": 6.417099639703979e-05,
+      "loss": 0.9483,
+      "step": 11415
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.407751666346569e-05,
+      "loss": 0.953,
+      "step": 11420
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.251953125,
+      "learning_rate": 6.398407295763187e-05,
+      "loss": 0.9259,
+      "step": 11425
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26953125,
+      "learning_rate": 6.38906653732559e-05,
+      "loss": 0.9765,
+      "step": 11430
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.251953125,
+      "learning_rate": 6.3797294004019e-05,
+      "loss": 0.9349,
+      "step": 11435
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.28125,
+      "learning_rate": 6.37039589435662e-05,
+      "loss": 0.9406,
+      "step": 11440
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2578125,
+      "learning_rate": 6.361066028550593e-05,
+      "loss": 0.8959,
+      "step": 11445
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.351739812341036e-05,
+      "loss": 0.947,
+      "step": 11450
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.267578125,
+      "learning_rate": 6.342417255081479e-05,
+      "loss": 0.9025,
+      "step": 11455
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.271484375,
+      "learning_rate": 6.333098366121804e-05,
+      "loss": 0.9347,
+      "step": 11460
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 6.323783154808205e-05,
+      "loss": 0.9275,
+      "step": 11465
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.255859375,
+      "learning_rate": 6.314471630483183e-05,
+      "loss": 0.9963,
+      "step": 11470
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.275390625,
+      "learning_rate": 6.305163802485554e-05,
+      "loss": 0.9659,
+      "step": 11475
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.244140625,
+      "learning_rate": 6.29585968015041e-05,
+      "loss": 0.9556,
+      "step": 11480
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.271484375,
+      "learning_rate": 6.286559272809142e-05,
+      "loss": 1.0343,
+      "step": 11485
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.248046875,
+      "learning_rate": 6.277262589789406e-05,
+      "loss": 0.9564,
+      "step": 11490
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.251953125,
+      "learning_rate": 6.267969640415124e-05,
+      "loss": 0.9114,
+      "step": 11495
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26953125,
+      "learning_rate": 6.258680434006478e-05,
+      "loss": 1.0162,
+      "step": 11500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.255859375,
+      "learning_rate": 6.24939497987989e-05,
+      "loss": 1.0089,
+      "step": 11505
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26953125,
+      "learning_rate": 6.240113287348026e-05,
+      "loss": 0.9537,
+      "step": 11510
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.251953125,
+      "learning_rate": 6.230835365719767e-05,
+      "loss": 0.9135,
+      "step": 11515
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.263671875,
+      "learning_rate": 6.22156122430023e-05,
+      "loss": 0.9519,
+      "step": 11520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.259765625,
+      "learning_rate": 6.212290872390722e-05,
+      "loss": 0.9276,
+      "step": 11525
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 6.203024319288762e-05,
+      "loss": 0.8321,
+      "step": 11530
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.255859375,
+      "learning_rate": 6.193761574288057e-05,
+      "loss": 0.9405,
+      "step": 11535
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.265625,
+      "learning_rate": 6.184502646678486e-05,
+      "loss": 0.9904,
+      "step": 11540
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.175247545746116e-05,
+      "loss": 1.006,
+      "step": 11545
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2578125,
+      "learning_rate": 6.165996280773157e-05,
+      "loss": 0.8975,
+      "step": 11550
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.28515625,
+      "learning_rate": 6.156748861037991e-05,
+      "loss": 0.9428,
+      "step": 11555
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2578125,
+      "learning_rate": 6.147505295815124e-05,
+      "loss": 0.923,
+      "step": 11560
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.138265594375212e-05,
+      "loss": 1.0096,
+      "step": 11565
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.259765625,
+      "learning_rate": 6.129029765985028e-05,
+      "loss": 0.9237,
+      "step": 11570
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.119797819907463e-05,
+      "loss": 0.925,
+      "step": 11575
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.265625,
+      "learning_rate": 6.110569765401513e-05,
+      "loss": 0.9749,
+      "step": 11580
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.1013456117222686e-05,
+      "loss": 0.9196,
+      "step": 11585
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2734375,
+      "learning_rate": 6.092125368120921e-05,
+      "loss": 0.8837,
+      "step": 11590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.28125,
+      "learning_rate": 6.082909043844719e-05,
+      "loss": 0.9113,
+      "step": 11595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.24609375,
+      "learning_rate": 6.073696648137001e-05,
+      "loss": 0.9584,
+      "step": 11600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.27734375,
+      "learning_rate": 6.0644881902371474e-05,
+      "loss": 0.9302,
+      "step": 11605
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.263671875,
+      "learning_rate": 6.055283679380605e-05,
+      "loss": 0.9446,
+      "step": 11610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.263671875,
+      "learning_rate": 6.046083124798851e-05,
+      "loss": 1.0009,
+      "step": 11615
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.271484375,
+      "learning_rate": 6.036886535719399e-05,
+      "loss": 0.9066,
+      "step": 11620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.259765625,
+      "learning_rate": 6.027693921365789e-05,
+      "loss": 0.9608,
+      "step": 11625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.25390625,
+      "learning_rate": 6.018505290957565e-05,
+      "loss": 0.9562,
+      "step": 11630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.25390625,
+      "learning_rate": 6.0093206537102866e-05,
+      "loss": 0.9836,
+      "step": 11635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.244140625,
+      "learning_rate": 6.000140018835497e-05,
+      "loss": 0.9908,
+      "step": 11640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.26953125,
+      "learning_rate": 5.990963395540739e-05,
+      "loss": 0.9516,
+      "step": 11645
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.251953125,
+      "learning_rate": 5.9817907930295155e-05,
+      "loss": 0.934,
+      "step": 11650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.26953125,
+      "learning_rate": 5.972622220501315e-05,
+      "loss": 0.9158,
+      "step": 11655
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.265625,
+      "learning_rate": 5.9634576871515656e-05,
+      "loss": 0.8874,
+      "step": 11660
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.263671875,
+      "learning_rate": 5.9542972021716616e-05,
+      "loss": 0.9071,
+      "step": 11665
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.265625,
+      "learning_rate": 5.945140774748929e-05,
+      "loss": 0.9595,
+      "step": 11670
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.26953125,
+      "learning_rate": 5.935988414066617e-05,
+      "loss": 0.9756,
+      "step": 11675
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.259765625,
+      "learning_rate": 5.9268401293039125e-05,
+      "loss": 1.0034,
+      "step": 11680
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.917695929635898e-05,
+      "loss": 0.8618,
+      "step": 11685
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.26953125,
+      "learning_rate": 5.908555824233575e-05,
+      "loss": 0.9354,
+      "step": 11690
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.899419822263822e-05,
+      "loss": 0.9256,
+      "step": 11695
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.26953125,
+      "learning_rate": 5.8902879328894156e-05,
+      "loss": 0.8939,
+      "step": 11700
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.263671875,
+      "learning_rate": 5.881160165269004e-05,
+      "loss": 0.9736,
+      "step": 11705
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.28515625,
+      "learning_rate": 5.872036528557096e-05,
+      "loss": 0.9653,
+      "step": 11710
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.28125,
+      "learning_rate": 5.862917031904066e-05,
+      "loss": 0.9689,
+      "step": 11715
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2578125,
+      "learning_rate": 5.853801684456126e-05,
+      "loss": 0.9334,
+      "step": 11720
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.287109375,
+      "learning_rate": 5.844690495355338e-05,
+      "loss": 0.9702,
+      "step": 11725
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.259765625,
+      "learning_rate": 5.8355834737395856e-05,
+      "loss": 0.8879,
+      "step": 11730
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 5.8264806287425724e-05,
+      "loss": 0.9326,
+      "step": 11735
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.817381969493823e-05,
+      "loss": 0.9541,
+      "step": 11740
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.251953125,
+      "learning_rate": 5.808287505118647e-05,
+      "loss": 0.9941,
+      "step": 11745
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.26953125,
+      "learning_rate": 5.799197244738166e-05,
+      "loss": 0.9653,
+      "step": 11750
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.259765625,
+      "learning_rate": 5.790111197469269e-05,
+      "loss": 0.9109,
+      "step": 11755
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.27734375,
+      "learning_rate": 5.781029372424633e-05,
+      "loss": 1.0813,
+      "step": 11760
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2734375,
+      "learning_rate": 5.7719517787126856e-05,
+      "loss": 0.9185,
+      "step": 11765
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.25390625,
+      "learning_rate": 5.762878425437627e-05,
+      "loss": 0.842,
+      "step": 11770
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.244140625,
+      "learning_rate": 5.753809321699388e-05,
+      "loss": 0.9481,
+      "step": 11775
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.744744476593652e-05,
+      "loss": 0.9382,
+      "step": 11780
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.255859375,
+      "learning_rate": 5.7356838992118277e-05,
+      "loss": 1.0021,
+      "step": 11785
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.267578125,
+      "learning_rate": 5.7266275986410324e-05,
+      "loss": 1.0197,
+      "step": 11790
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.251953125,
+      "learning_rate": 5.717575583964111e-05,
+      "loss": 0.9619,
+      "step": 11795
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.708527864259594e-05,
+      "loss": 0.9145,
+      "step": 11800
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.263671875,
+      "learning_rate": 5.6994844486017204e-05,
+      "loss": 0.865,
+      "step": 11805
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.255859375,
+      "learning_rate": 5.6904453460603955e-05,
+      "loss": 1.0117,
+      "step": 11810
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.259765625,
+      "learning_rate": 5.681410565701215e-05,
+      "loss": 0.9478,
+      "step": 11815
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.271484375,
+      "learning_rate": 5.672380116585425e-05,
+      "loss": 0.9194,
+      "step": 11820
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.263671875,
+      "learning_rate": 5.663354007769943e-05,
+      "loss": 0.9231,
+      "step": 11825
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.251953125,
+      "learning_rate": 5.654332248307319e-05,
+      "loss": 1.0146,
+      "step": 11830
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.6453148472457476e-05,
+      "loss": 0.923,
+      "step": 11835
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.259765625,
+      "learning_rate": 5.636301813629057e-05,
+      "loss": 0.9278,
+      "step": 11840
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.255859375,
+      "learning_rate": 5.62729315649668e-05,
+      "loss": 0.9252,
+      "step": 11845
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.259765625,
+      "learning_rate": 5.618288884883684e-05,
+      "loss": 0.8796,
+      "step": 11850
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.271484375,
+      "learning_rate": 5.6092890078207107e-05,
+      "loss": 0.9037,
+      "step": 11855
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.600293534334014e-05,
+      "loss": 0.8821,
+      "step": 11860
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.591302473445429e-05,
+      "loss": 0.9258,
+      "step": 11865
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.259765625,
+      "learning_rate": 5.582315834172353e-05,
+      "loss": 0.9576,
+      "step": 11870
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.263671875,
+      "learning_rate": 5.573333625527767e-05,
+      "loss": 0.9547,
+      "step": 11875
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2578125,
+      "learning_rate": 5.564355856520189e-05,
+      "loss": 0.9878,
+      "step": 11880
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.25390625,
+      "learning_rate": 5.555382536153702e-05,
+      "loss": 0.8999,
+      "step": 11885
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.248046875,
+      "learning_rate": 5.5464136734279094e-05,
+      "loss": 0.9551,
+      "step": 11890
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2578125,
+      "learning_rate": 5.537449277337965e-05,
+      "loss": 0.9651,
+      "step": 11895
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.240234375,
+      "learning_rate": 5.528489356874522e-05,
+      "loss": 0.9024,
+      "step": 11900
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.25,
+      "learning_rate": 5.5195339210237626e-05,
+      "loss": 0.9697,
+      "step": 11905
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.251953125,
+      "learning_rate": 5.510582978767356e-05,
+      "loss": 0.99,
+      "step": 11910
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.291015625,
+      "learning_rate": 5.501636539082478e-05,
+      "loss": 0.9663,
+      "step": 11915
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 5.4926946109417775e-05,
+      "loss": 0.9425,
+      "step": 11920
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.251953125,
+      "learning_rate": 5.483757203313383e-05,
+      "loss": 1.0466,
+      "step": 11925
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26953125,
+      "learning_rate": 5.4748243251608965e-05,
+      "loss": 0.9748,
+      "step": 11930
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.251953125,
+      "learning_rate": 5.465895985443361e-05,
+      "loss": 0.9863,
+      "step": 11935
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.248046875,
+      "learning_rate": 5.4569721931152864e-05,
+      "loss": 0.8764,
+      "step": 11940
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.28515625,
+      "learning_rate": 5.448052957126606e-05,
+      "loss": 1.0002,
+      "step": 11945
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.28125,
+      "learning_rate": 5.4391382864226916e-05,
+      "loss": 0.9917,
+      "step": 11950
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.248046875,
+      "learning_rate": 5.4302281899443394e-05,
+      "loss": 0.9495,
+      "step": 11955
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.291015625,
+      "learning_rate": 5.421322676627747e-05,
+      "loss": 0.9429,
+      "step": 11960
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.263671875,
+      "learning_rate": 5.412421755404529e-05,
+      "loss": 1.0265,
+      "step": 11965
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.25,
+      "learning_rate": 5.40352543520168e-05,
+      "loss": 0.9534,
+      "step": 11970
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.23828125,
+      "learning_rate": 5.3946337249415936e-05,
+      "loss": 0.9935,
+      "step": 11975
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.283203125,
+      "learning_rate": 5.385746633542027e-05,
+      "loss": 0.9642,
+      "step": 11980
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.279296875,
+      "learning_rate": 5.376864169916116e-05,
+      "loss": 0.9807,
+      "step": 11985
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.267578125,
+      "learning_rate": 5.367986342972355e-05,
+      "loss": 0.9962,
+      "step": 11990
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30859375,
+      "learning_rate": 5.359113161614576e-05,
+      "loss": 0.924,
+      "step": 11995
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.26953125,
+      "learning_rate": 5.3502446347419674e-05,
+      "loss": 0.9266,
+      "step": 12000
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.341380771249037e-05,
+      "loss": 0.9088,
+      "step": 12005
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.259765625,
+      "learning_rate": 5.332521580025622e-05,
+      "loss": 1.0162,
+      "step": 12010
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.26953125,
+      "learning_rate": 5.323667069956868e-05,
+      "loss": 0.954,
+      "step": 12015
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.25390625,
+      "learning_rate": 5.314817249923236e-05,
+      "loss": 0.9517,
+      "step": 12020
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.25,
+      "learning_rate": 5.3059721288004714e-05,
+      "loss": 0.9392,
+      "step": 12025
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.267578125,
+      "learning_rate": 5.297131715459614e-05,
+      "loss": 0.9615,
+      "step": 12030
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 5.288296018766987e-05,
+      "loss": 1.0129,
+      "step": 12035
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.287109375,
+      "learning_rate": 5.2794650475841664e-05,
+      "loss": 0.9961,
+      "step": 12040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.240234375,
+      "learning_rate": 5.2706388107680095e-05,
+      "loss": 0.8726,
+      "step": 12045
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.279296875,
+      "learning_rate": 5.2618173171706064e-05,
+      "loss": 0.9517,
+      "step": 12050
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.267578125,
+      "learning_rate": 5.253000575639305e-05,
+      "loss": 1.0232,
+      "step": 12055
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.2441885950166746e-05,
+      "loss": 0.9498,
+      "step": 12060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.244140625,
+      "learning_rate": 5.235381384140519e-05,
+      "loss": 0.9127,
+      "step": 12065
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.271484375,
+      "learning_rate": 5.226578951843859e-05,
+      "loss": 0.9645,
+      "step": 12070
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.25,
+      "learning_rate": 5.217781306954912e-05,
+      "loss": 0.906,
+      "step": 12075
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.265625,
+      "learning_rate": 5.208988458297109e-05,
+      "loss": 0.9287,
+      "step": 12080
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.25390625,
+      "learning_rate": 5.2002004146890535e-05,
+      "loss": 0.9332,
+      "step": 12085
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.263671875,
+      "learning_rate": 5.191417184944549e-05,
+      "loss": 0.9164,
+      "step": 12090
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.182638777872555e-05,
+      "loss": 0.8766,
+      "step": 12095
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.279296875,
+      "learning_rate": 5.1738652022771974e-05,
+      "loss": 0.9041,
+      "step": 12100
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.275390625,
+      "learning_rate": 5.165096466957769e-05,
+      "loss": 0.9226,
+      "step": 12105
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.267578125,
+      "learning_rate": 5.1563325807086856e-05,
+      "loss": 0.8697,
+      "step": 12110
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.147573552319526e-05,
+      "loss": 0.8859,
+      "step": 12115
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2890625,
+      "learning_rate": 5.138819390574972e-05,
+      "loss": 0.8888,
+      "step": 12120
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.251953125,
+      "learning_rate": 5.130070104254847e-05,
+      "loss": 0.9413,
+      "step": 12125
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2578125,
+      "learning_rate": 5.121325702134063e-05,
+      "loss": 0.944,
+      "step": 12130
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.28125,
+      "learning_rate": 5.112586192982653e-05,
+      "loss": 0.9692,
+      "step": 12135
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.25390625,
+      "learning_rate": 5.1038515855657264e-05,
+      "loss": 0.9133,
+      "step": 12140
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2734375,
+      "learning_rate": 5.0951218886434884e-05,
+      "loss": 0.969,
+      "step": 12145
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.086397110971218e-05,
+      "loss": 0.9552,
+      "step": 12150
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.265625,
+      "learning_rate": 5.077677261299251e-05,
+      "loss": 0.8627,
+      "step": 12155
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.068962348372992e-05,
+      "loss": 0.9822,
+      "step": 12160
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.060252380932886e-05,
+      "loss": 0.936,
+      "step": 12165
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.248046875,
+      "learning_rate": 5.0515473677144254e-05,
+      "loss": 0.9528,
+      "step": 12170
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.25,
+      "learning_rate": 5.042847317448125e-05,
+      "loss": 0.9435,
+      "step": 12175
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.034152238859533e-05,
+      "loss": 0.9473,
+      "step": 12180
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.25,
+      "learning_rate": 5.025462140669204e-05,
+      "loss": 0.937,
+      "step": 12185
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.265625,
+      "learning_rate": 5.016777031592694e-05,
+      "loss": 0.9567,
+      "step": 12190
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.251953125,
+      "learning_rate": 5.008096920340568e-05,
+      "loss": 0.9716,
+      "step": 12195
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.999421815618364e-05,
+      "loss": 0.9504,
+      "step": 12200
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 4.990751726126612e-05,
+      "loss": 0.9038,
+      "step": 12205
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.25390625,
+      "learning_rate": 4.9820866605607994e-05,
+      "loss": 0.8834,
+      "step": 12210
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.27734375,
+      "learning_rate": 4.973426627611389e-05,
+      "loss": 0.9035,
+      "step": 12215
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.25390625,
+      "learning_rate": 4.964771635963781e-05,
+      "loss": 0.9357,
+      "step": 12220
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.259765625,
+      "learning_rate": 4.95612169429833e-05,
+      "loss": 0.9755,
+      "step": 12225
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.240234375,
+      "learning_rate": 4.9474768112903293e-05,
+      "loss": 0.9856,
+      "step": 12230
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 4.9388369956099815e-05,
+      "loss": 0.898,
+      "step": 12235
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2578125,
+      "learning_rate": 4.930202255922427e-05,
+      "loss": 0.9254,
+      "step": 12240
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.26953125,
+      "learning_rate": 4.9215726008876995e-05,
+      "loss": 0.945,
+      "step": 12245
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.26171875,
+      "learning_rate": 4.9129480391607465e-05,
+      "loss": 0.9477,
+      "step": 12250
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.904328579391393e-05,
+      "loss": 0.9186,
+      "step": 12255
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2421875,
+      "learning_rate": 4.895714230224363e-05,
+      "loss": 0.9673,
+      "step": 12260
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.887105000299239e-05,
+      "loss": 1.0013,
+      "step": 12265
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.25390625,
+      "learning_rate": 4.8785008982504845e-05,
+      "loss": 0.9292,
+      "step": 12270
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.267578125,
+      "learning_rate": 4.8699019327074035e-05,
+      "loss": 0.9545,
+      "step": 12275
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2421875,
+      "learning_rate": 4.861308112294168e-05,
+      "loss": 1.0284,
+      "step": 12280
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.28125,
+      "learning_rate": 4.852719445629773e-05,
+      "loss": 1.0357,
+      "step": 12285
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.244140625,
+      "learning_rate": 4.844135941328048e-05,
+      "loss": 0.9705,
+      "step": 12290
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.279296875,
+      "learning_rate": 4.835557607997656e-05,
+      "loss": 1.0149,
+      "step": 12295
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.259765625,
+      "learning_rate": 4.826984454242057e-05,
+      "loss": 0.9071,
+      "step": 12300
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.271484375,
+      "learning_rate": 4.818416488659534e-05,
+      "loss": 1.006,
+      "step": 12305
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.255859375,
+      "learning_rate": 4.80985371984315e-05,
+      "loss": 1.007,
+      "step": 12310
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.26171875,
+      "learning_rate": 4.801296156380767e-05,
+      "loss": 0.9192,
+      "step": 12315
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.275390625,
+      "learning_rate": 4.7927438068550256e-05,
+      "loss": 0.991,
+      "step": 12320
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.78419667984333e-05,
+      "loss": 0.9496,
+      "step": 12325
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.251953125,
+      "learning_rate": 4.7756547839178564e-05,
+      "loss": 0.9835,
+      "step": 12330
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.26171875,
+      "learning_rate": 4.767118127645524e-05,
+      "loss": 0.9501,
+      "step": 12335
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.25390625,
+      "learning_rate": 4.758586719588007e-05,
+      "loss": 0.9286,
+      "step": 12340
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.251953125,
+      "learning_rate": 4.7500605683017076e-05,
+      "loss": 0.9161,
+      "step": 12345
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.255859375,
+      "learning_rate": 4.74153968233776e-05,
+      "loss": 0.9634,
+      "step": 12350
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 4.733024070242024e-05,
+      "loss": 0.8658,
+      "step": 12355
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.263671875,
+      "learning_rate": 4.724513740555053e-05,
+      "loss": 0.9709,
+      "step": 12360
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.263671875,
+      "learning_rate": 4.716008701812123e-05,
+      "loss": 0.8628,
+      "step": 12365
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.25390625,
+      "learning_rate": 4.707508962543188e-05,
+      "loss": 0.9155,
+      "step": 12370
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.28125,
+      "learning_rate": 4.699014531272894e-05,
+      "loss": 0.9683,
+      "step": 12375
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2734375,
+      "learning_rate": 4.690525416520557e-05,
+      "loss": 1.0008,
+      "step": 12380
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.275390625,
+      "learning_rate": 4.6820416268001747e-05,
+      "loss": 0.9406,
+      "step": 12385
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2890625,
+      "learning_rate": 4.673563170620385e-05,
+      "loss": 0.8862,
+      "step": 12390
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.25390625,
+      "learning_rate": 4.6650900564844935e-05,
+      "loss": 0.9025,
+      "step": 12395
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 4.6566222928904436e-05,
+      "loss": 0.9169,
+      "step": 12400
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.267578125,
+      "learning_rate": 4.648159888330804e-05,
+      "loss": 0.9738,
+      "step": 12405
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 4.639702851292782e-05,
+      "loss": 0.9669,
+      "step": 12410
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 4.631251190258187e-05,
+      "loss": 0.9728,
+      "step": 12415
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 4.622804913703452e-05,
+      "loss": 0.9234,
+      "step": 12420
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.248046875,
+      "learning_rate": 4.614364030099596e-05,
+      "loss": 0.94,
+      "step": 12425
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.605928547912237e-05,
+      "loss": 0.9276,
+      "step": 12430
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.267578125,
+      "learning_rate": 4.597498475601579e-05,
+      "loss": 0.9564,
+      "step": 12435
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.255859375,
+      "learning_rate": 4.5890738216223884e-05,
+      "loss": 1.0038,
+      "step": 12440
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.279296875,
+      "learning_rate": 4.58065459442401e-05,
+      "loss": 1.0051,
+      "step": 12445
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.572240802450335e-05,
+      "loss": 0.901,
+      "step": 12450
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.25390625,
+      "learning_rate": 4.5638324541398136e-05,
+      "loss": 0.9423,
+      "step": 12455
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.26171875,
+      "learning_rate": 4.55542955792543e-05,
+      "loss": 0.9089,
+      "step": 12460
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.271484375,
+      "learning_rate": 4.547032122234698e-05,
+      "loss": 0.9291,
+      "step": 12465
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.255859375,
+      "learning_rate": 4.538640155489666e-05,
+      "loss": 0.9689,
+      "step": 12470
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.5302536661068816e-05,
+      "loss": 0.9536,
+      "step": 12475
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.228515625,
+      "learning_rate": 4.521872662497416e-05,
+      "loss": 0.8805,
+      "step": 12480
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.23828125,
+      "learning_rate": 4.513497153066822e-05,
+      "loss": 0.8875,
+      "step": 12485
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.263671875,
+      "learning_rate": 4.505127146215159e-05,
+      "loss": 0.9778,
+      "step": 12490
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 4.49676265033695e-05,
+      "loss": 0.9782,
+      "step": 12495
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.25,
+      "learning_rate": 4.4884036738212074e-05,
+      "loss": 0.9139,
+      "step": 12500
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.275390625,
+      "learning_rate": 4.480050225051394e-05,
+      "loss": 0.9136,
+      "step": 12505
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2578125,
+      "learning_rate": 4.4717023124054394e-05,
+      "loss": 0.9731,
+      "step": 12510
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.25,
+      "learning_rate": 4.463359944255718e-05,
+      "loss": 0.9265,
+      "step": 12515
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.271484375,
+      "learning_rate": 4.455023128969036e-05,
+      "loss": 0.9152,
+      "step": 12520
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.271484375,
+      "learning_rate": 4.446691874906645e-05,
+      "loss": 0.9031,
+      "step": 12525
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.248046875,
+      "learning_rate": 4.4383661904242e-05,
+      "loss": 0.9367,
+      "step": 12530
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.28125,
+      "learning_rate": 4.430046083871791e-05,
+      "loss": 0.9354,
+      "step": 12535
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.26953125,
+      "learning_rate": 4.421731563593895e-05,
+      "loss": 0.959,
+      "step": 12540
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2578125,
+      "learning_rate": 4.413422637929402e-05,
+      "loss": 0.9779,
+      "step": 12545
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.25390625,
+      "learning_rate": 4.40511931521158e-05,
+      "loss": 0.9275,
+      "step": 12550
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.259765625,
+      "learning_rate": 4.396821603768079e-05,
+      "loss": 0.9925,
+      "step": 12555
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2578125,
+      "learning_rate": 4.3885295119209294e-05,
+      "loss": 0.9756,
+      "step": 12560
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.25,
+      "learning_rate": 4.380243047986513e-05,
+      "loss": 1.0194,
+      "step": 12565
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.3719622202755816e-05,
+      "loss": 0.9061,
+      "step": 12570
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.275390625,
+      "learning_rate": 4.3636870370932194e-05,
+      "loss": 0.9297,
+      "step": 12575
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.259765625,
+      "learning_rate": 4.3554175067388636e-05,
+      "loss": 0.9902,
+      "step": 12580
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 4.3471536375062696e-05,
+      "loss": 0.9526,
+      "step": 12585
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.27734375,
+      "learning_rate": 4.338895437683521e-05,
+      "loss": 0.9041,
+      "step": 12590
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.330642915553023e-05,
+      "loss": 0.9641,
+      "step": 12595
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.259765625,
+      "learning_rate": 4.322396079391467e-05,
+      "loss": 0.9234,
+      "step": 12600
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.3141549374698645e-05,
+      "loss": 0.8561,
+      "step": 12605
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.305919498053495e-05,
+      "loss": 0.9899,
+      "step": 12610
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.2976897694019356e-05,
+      "loss": 0.9444,
+      "step": 12615
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.271484375,
+      "learning_rate": 4.289465759769025e-05,
+      "loss": 0.9067,
+      "step": 12620
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.259765625,
+      "learning_rate": 4.2812474774028735e-05,
+      "loss": 0.9728,
+      "step": 12625
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.25,
+      "learning_rate": 4.27303493054584e-05,
+      "loss": 0.955,
+      "step": 12630
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.26171875,
+      "learning_rate": 4.264828127434539e-05,
+      "loss": 0.9825,
+      "step": 12635
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 4.256627076299816e-05,
+      "loss": 0.9644,
+      "step": 12640
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.251953125,
+      "learning_rate": 4.248431785366759e-05,
+      "loss": 0.9211,
+      "step": 12645
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.267578125,
+      "learning_rate": 4.2402422628546666e-05,
+      "loss": 0.9046,
+      "step": 12650
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.279296875,
+      "learning_rate": 4.2320585169770565e-05,
+      "loss": 0.9737,
+      "step": 12655
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 4.2238805559416594e-05,
+      "loss": 0.9187,
+      "step": 12660
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.263671875,
+      "learning_rate": 4.215708387950391e-05,
+      "loss": 0.9978,
+      "step": 12665
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.267578125,
+      "learning_rate": 4.20754202119937e-05,
+      "loss": 0.9516,
+      "step": 12670
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.251953125,
+      "learning_rate": 4.1993814638788944e-05,
+      "loss": 0.9569,
+      "step": 12675
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 4.191226724173426e-05,
+      "loss": 0.9413,
+      "step": 12680
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.25390625,
+      "learning_rate": 4.1830778102616055e-05,
+      "loss": 0.9646,
+      "step": 12685
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 4.174934730316216e-05,
+      "loss": 0.9325,
+      "step": 12690
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.255859375,
+      "learning_rate": 4.166797492504206e-05,
+      "loss": 0.9078,
+      "step": 12695
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 4.1586661049866496e-05,
+      "loss": 0.8876,
+      "step": 12700
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.26953125,
+      "learning_rate": 4.1505405759187666e-05,
+      "loss": 0.9473,
+      "step": 12705
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.271484375,
+      "learning_rate": 4.142420913449887e-05,
+      "loss": 0.9425,
+      "step": 12710
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.259765625,
+      "learning_rate": 4.13430712572347e-05,
+      "loss": 0.924,
+      "step": 12715
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.259765625,
+      "learning_rate": 4.12619922087708e-05,
+      "loss": 0.8847,
+      "step": 12720
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.26171875,
+      "learning_rate": 4.118097207042373e-05,
+      "loss": 0.8964,
+      "step": 12725
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 4.1100010923451084e-05,
+      "loss": 0.928,
+      "step": 12730
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.25390625,
+      "learning_rate": 4.10191088490512e-05,
+      "loss": 0.958,
+      "step": 12735
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.271484375,
+      "learning_rate": 4.093826592836322e-05,
+      "loss": 0.9281,
+      "step": 12740
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2578125,
+      "learning_rate": 4.0857482242466885e-05,
+      "loss": 0.9126,
+      "step": 12745
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.077675787238267e-05,
+      "loss": 0.9593,
+      "step": 12750
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2578125,
+      "learning_rate": 4.0696092899071416e-05,
+      "loss": 0.8777,
+      "step": 12755
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.25390625,
+      "learning_rate": 4.061548740343446e-05,
+      "loss": 0.9838,
+      "step": 12760
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.26953125,
+      "learning_rate": 4.053494146631355e-05,
+      "loss": 0.9303,
+      "step": 12765
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 4.045445516849055e-05,
+      "loss": 0.8684,
+      "step": 12770
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.244140625,
+      "learning_rate": 4.037402859068764e-05,
+      "loss": 0.9017,
+      "step": 12775
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.255859375,
+      "learning_rate": 4.029366181356702e-05,
+      "loss": 0.9531,
+      "step": 12780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.26171875,
+      "learning_rate": 4.0213354917731004e-05,
+      "loss": 0.9097,
+      "step": 12785
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.263671875,
+      "learning_rate": 4.0133107983721726e-05,
+      "loss": 0.9436,
+      "step": 12790
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.283203125,
+      "learning_rate": 4.005292109202129e-05,
+      "loss": 0.9268,
+      "step": 12795
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2578125,
+      "learning_rate": 3.997279432305158e-05,
+      "loss": 0.874,
+      "step": 12800
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.248046875,
+      "learning_rate": 3.9892727757174074e-05,
+      "loss": 0.965,
+      "step": 12805
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.981272147469002e-05,
+      "loss": 0.9038,
+      "step": 12810
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 3.973277555584004e-05,
+      "loss": 0.9387,
+      "step": 12815
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.965289008080438e-05,
+      "loss": 0.899,
+      "step": 12820
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 3.957306512970258e-05,
+      "loss": 0.9678,
+      "step": 12825
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.275390625,
+      "learning_rate": 3.9493300782593415e-05,
+      "loss": 1.016,
+      "step": 12830
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2578125,
+      "learning_rate": 3.9413597119475044e-05,
+      "loss": 0.9981,
+      "step": 12835
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 3.9333954220284586e-05,
+      "loss": 0.9549,
+      "step": 12840
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.291015625,
+      "learning_rate": 3.925437216489838e-05,
+      "loss": 0.9696,
+      "step": 12845
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.271484375,
+      "learning_rate": 3.91748510331316e-05,
+      "loss": 1.0139,
+      "step": 12850
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.909539090473845e-05,
+      "loss": 0.9212,
+      "step": 12855
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.26953125,
+      "learning_rate": 3.9015991859411815e-05,
+      "loss": 1.0049,
+      "step": 12860
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.263671875,
+      "learning_rate": 3.8936653976783454e-05,
+      "loss": 0.9443,
+      "step": 12865
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2734375,
+      "learning_rate": 3.885737733642366e-05,
+      "loss": 0.9899,
+      "step": 12870
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.263671875,
+      "learning_rate": 3.877816201784139e-05,
+      "loss": 0.9079,
+      "step": 12875
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.86990081004841e-05,
+      "loss": 0.9355,
+      "step": 12880
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2890625,
+      "learning_rate": 3.861991566373759e-05,
+      "loss": 0.9222,
+      "step": 12885
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 3.85408847869261e-05,
+      "loss": 0.8832,
+      "step": 12890
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.26171875,
+      "learning_rate": 3.846191554931201e-05,
+      "loss": 0.9305,
+      "step": 12895
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2734375,
+      "learning_rate": 3.838300803009601e-05,
+      "loss": 0.939,
+      "step": 12900
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.26953125,
+      "learning_rate": 3.8304162308416766e-05,
+      "loss": 0.9762,
+      "step": 12905
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2890625,
+      "learning_rate": 3.822537846335109e-05,
+      "loss": 0.975,
+      "step": 12910
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.267578125,
+      "learning_rate": 3.814665657391365e-05,
+      "loss": 0.9453,
+      "step": 12915
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.25,
+      "learning_rate": 3.806799671905695e-05,
+      "loss": 0.9061,
+      "step": 12920
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.263671875,
+      "learning_rate": 3.798939897767141e-05,
+      "loss": 0.903,
+      "step": 12925
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.25,
+      "learning_rate": 3.7910863428584985e-05,
+      "loss": 0.8856,
+      "step": 12930
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.25390625,
+      "learning_rate": 3.783239015056343e-05,
+      "loss": 0.9018,
+      "step": 12935
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.7753979222309876e-05,
+      "loss": 0.9687,
+      "step": 12940
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.767563072246508e-05,
+      "loss": 0.9176,
+      "step": 12945
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.279296875,
+      "learning_rate": 3.7597344729607056e-05,
+      "loss": 0.874,
+      "step": 12950
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.263671875,
+      "learning_rate": 3.751912132225118e-05,
+      "loss": 0.9264,
+      "step": 12955
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.265625,
+      "learning_rate": 3.744096057885014e-05,
+      "loss": 0.9561,
+      "step": 12960
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.263671875,
+      "learning_rate": 3.73628625777936e-05,
+      "loss": 0.9813,
+      "step": 12965
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.27734375,
+      "learning_rate": 3.7284827397408485e-05,
+      "loss": 0.9638,
+      "step": 12970
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.271484375,
+      "learning_rate": 3.720685511595855e-05,
+      "loss": 0.9361,
+      "step": 12975
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.314453125,
+      "learning_rate": 3.712894581164461e-05,
+      "loss": 0.9472,
+      "step": 12980
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.26171875,
+      "learning_rate": 3.705109956260419e-05,
+      "loss": 0.9043,
+      "step": 12985
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 3.69733164469117e-05,
+      "loss": 0.98,
+      "step": 12990
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.265625,
+      "learning_rate": 3.68955965425781e-05,
+      "loss": 0.9291,
+      "step": 12995
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2578125,
+      "learning_rate": 3.6817939927551105e-05,
+      "loss": 0.9215,
+      "step": 13000
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.67403466797148e-05,
+      "loss": 0.9316,
+      "step": 13005
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.25390625,
+      "learning_rate": 3.6662816876889837e-05,
+      "loss": 0.8962,
+      "step": 13010
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.658535059683318e-05,
+      "loss": 0.9216,
+      "step": 13015
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.650794791723805e-05,
+      "loss": 0.8916,
+      "step": 13020
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.263671875,
+      "learning_rate": 3.6430608915734e-05,
+      "loss": 0.9796,
+      "step": 13025
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.248046875,
+      "learning_rate": 3.635333366988657e-05,
+      "loss": 0.9619,
+      "step": 13030
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.6276122257197465e-05,
+      "loss": 0.9352,
+      "step": 13035
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.6198974755104366e-05,
+      "loss": 0.8827,
+      "step": 13040
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2578125,
+      "learning_rate": 3.6121891240980764e-05,
+      "loss": 0.9641,
+      "step": 13045
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.279296875,
+      "learning_rate": 3.604487179213612e-05,
+      "loss": 0.9531,
+      "step": 13050
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2734375,
+      "learning_rate": 3.596791648581546e-05,
+      "loss": 0.9712,
+      "step": 13055
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.271484375,
+      "learning_rate": 3.589102539919965e-05,
+      "loss": 0.9869,
+      "step": 13060
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.24609375,
+      "learning_rate": 3.5814198609405024e-05,
+      "loss": 0.9768,
+      "step": 13065
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2578125,
+      "learning_rate": 3.5737436193483555e-05,
+      "loss": 0.9148,
+      "step": 13070
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.56607382284225e-05,
+      "loss": 0.9513,
+      "step": 13075
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.5584104791144603e-05,
+      "loss": 0.929,
+      "step": 13080
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.24609375,
+      "learning_rate": 3.5507535958507864e-05,
+      "loss": 0.8616,
+      "step": 13085
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.543103180730541e-05,
+      "loss": 0.9664,
+      "step": 13090
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.248046875,
+      "learning_rate": 3.535459241426563e-05,
+      "loss": 0.8815,
+      "step": 13095
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.26953125,
+      "learning_rate": 3.5278217856051866e-05,
+      "loss": 0.8886,
+      "step": 13100
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.5201908209262445e-05,
+      "loss": 0.9652,
+      "step": 13105
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.25390625,
+      "learning_rate": 3.5125663550430585e-05,
+      "loss": 0.9308,
+      "step": 13110
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 3.504948395602442e-05,
+      "loss": 0.8708,
+      "step": 13115
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.271484375,
+      "learning_rate": 3.4973369502446685e-05,
+      "loss": 0.9243,
+      "step": 13120
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.4897320266034905e-05,
+      "loss": 0.95,
+      "step": 13125
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.240234375,
+      "learning_rate": 3.482133632306117e-05,
+      "loss": 0.8964,
+      "step": 13130
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.25,
+      "learning_rate": 3.4745417749732003e-05,
+      "loss": 0.9514,
+      "step": 13135
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 3.466956462218849e-05,
+      "loss": 0.9212,
+      "step": 13140
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.248046875,
+      "learning_rate": 3.4593777016505946e-05,
+      "loss": 0.9294,
+      "step": 13145
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.451805500869413e-05,
+      "loss": 0.9664,
+      "step": 13150
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 3.444239867469683e-05,
+      "loss": 0.9726,
+      "step": 13155
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28515625,
+      "learning_rate": 3.4366808090392123e-05,
+      "loss": 0.9226,
+      "step": 13160
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 3.429128333159208e-05,
+      "loss": 0.8796,
+      "step": 13165
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.267578125,
+      "learning_rate": 3.421582447404273e-05,
+      "loss": 0.9727,
+      "step": 13170
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2578125,
+      "learning_rate": 3.414043159342408e-05,
+      "loss": 0.9514,
+      "step": 13175
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2578125,
+      "learning_rate": 3.406510476534985e-05,
+      "loss": 0.9228,
+      "step": 13180
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.25390625,
+      "learning_rate": 3.398984406536765e-05,
+      "loss": 0.8952,
+      "step": 13185
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.391464956895869e-05,
+      "loss": 0.8966,
+      "step": 13190
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.275390625,
+      "learning_rate": 3.3839521351537726e-05,
+      "loss": 0.9749,
+      "step": 13195
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.263671875,
+      "learning_rate": 3.376445948845322e-05,
+      "loss": 0.9342,
+      "step": 13200
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.267578125,
+      "learning_rate": 3.368946405498686e-05,
+      "loss": 0.9269,
+      "step": 13205
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.248046875,
+      "learning_rate": 3.361453512635393e-05,
+      "loss": 0.9147,
+      "step": 13210
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.25390625,
+      "learning_rate": 3.353967277770282e-05,
+      "loss": 0.9704,
+      "step": 13215
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.265625,
+      "learning_rate": 3.346487708411532e-05,
+      "loss": 0.9207,
+      "step": 13220
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2578125,
+      "learning_rate": 3.3390148120606204e-05,
+      "loss": 0.9769,
+      "step": 13225
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 3.331548596212347e-05,
+      "loss": 0.9078,
+      "step": 13230
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.24609375,
+      "learning_rate": 3.324089068354797e-05,
+      "loss": 0.9256,
+      "step": 13235
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2578125,
+      "learning_rate": 3.3166362359693596e-05,
+      "loss": 0.9599,
+      "step": 13240
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.263671875,
+      "learning_rate": 3.3091901065307084e-05,
+      "loss": 0.903,
+      "step": 13245
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.271484375,
+      "learning_rate": 3.301750687506784e-05,
+      "loss": 0.9432,
+      "step": 13250
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.26171875,
+      "learning_rate": 3.29431798635881e-05,
+      "loss": 0.9166,
+      "step": 13255
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.27734375,
+      "learning_rate": 3.2868920105412594e-05,
+      "loss": 0.9601,
+      "step": 13260
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.279472767501876e-05,
+      "loss": 0.8922,
+      "step": 13265
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.271484375,
+      "learning_rate": 3.272060264681631e-05,
+      "loss": 0.9163,
+      "step": 13270
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.264654509514757e-05,
+      "loss": 0.8981,
+      "step": 13275
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.25390625,
+      "learning_rate": 3.257255509428705e-05,
+      "loss": 0.9022,
+      "step": 13280
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.25,
+      "learning_rate": 3.24986327184415e-05,
+      "loss": 0.9099,
+      "step": 13285
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2890625,
+      "learning_rate": 3.2424778041749984e-05,
+      "loss": 0.9285,
+      "step": 13290
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.235099113828351e-05,
+      "loss": 0.9232,
+      "step": 13295
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.227727208204523e-05,
+      "loss": 0.8646,
+      "step": 13300
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 3.2203620946970156e-05,
+      "loss": 0.957,
+      "step": 13305
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.267578125,
+      "learning_rate": 3.213003780692531e-05,
+      "loss": 0.9429,
+      "step": 13310
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.2056522735709346e-05,
+      "loss": 0.9812,
+      "step": 13315
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2734375,
+      "learning_rate": 3.198307580705281e-05,
+      "loss": 0.9726,
+      "step": 13320
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.265625,
+      "learning_rate": 3.190969709461783e-05,
+      "loss": 0.9495,
+      "step": 13325
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 3.183638667199809e-05,
+      "loss": 0.9097,
+      "step": 13330
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.240234375,
+      "learning_rate": 3.176314461271887e-05,
+      "loss": 1.0239,
+      "step": 13335
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.1689970990236784e-05,
+      "loss": 0.9087,
+      "step": 13340
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.1616865877939915e-05,
+      "loss": 0.9057,
+      "step": 13345
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.1543829349147523e-05,
+      "loss": 0.8985,
+      "step": 13350
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2421875,
+      "learning_rate": 3.147086147711022e-05,
+      "loss": 0.9505,
+      "step": 13355
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.267578125,
+      "learning_rate": 3.139796233500958e-05,
+      "loss": 0.9484,
+      "step": 13360
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.294921875,
+      "learning_rate": 3.132513199595846e-05,
+      "loss": 0.9627,
+      "step": 13365
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 3.1252370533000494e-05,
+      "loss": 0.9296,
+      "step": 13370
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.1179678019110434e-05,
+      "loss": 0.9281,
+      "step": 13375
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.110705452719376e-05,
+      "loss": 0.9775,
+      "step": 13380
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.1034500130086706e-05,
+      "loss": 0.9013,
+      "step": 13385
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.267578125,
+      "learning_rate": 3.096201490055635e-05,
+      "loss": 0.8999,
+      "step": 13390
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.088959891130022e-05,
+      "loss": 0.902,
+      "step": 13395
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.081725223494656e-05,
+      "loss": 0.9296,
+      "step": 13400
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 3.074497494405404e-05,
+      "loss": 0.944,
+      "step": 13405
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 3.0672767111111666e-05,
+      "loss": 0.9,
+      "step": 13410
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.0600628808538915e-05,
+      "loss": 0.9412,
+      "step": 13415
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2421875,
+      "learning_rate": 3.05285601086854e-05,
+      "loss": 0.9297,
+      "step": 13420
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.25,
+      "learning_rate": 3.045656108383106e-05,
+      "loss": 0.9706,
+      "step": 13425
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.240234375,
+      "learning_rate": 3.0384631806185815e-05,
+      "loss": 0.9361,
+      "step": 13430
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.263671875,
+      "learning_rate": 3.0312772347889773e-05,
+      "loss": 0.9369,
+      "step": 13435
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.263671875,
+      "learning_rate": 3.0240982781012873e-05,
+      "loss": 0.9819,
+      "step": 13440
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.265625,
+      "learning_rate": 3.0169263177555085e-05,
+      "loss": 0.9965,
+      "step": 13445
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.267578125,
+      "learning_rate": 3.0097613609446172e-05,
+      "loss": 0.9073,
+      "step": 13450
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.26953125,
+      "learning_rate": 3.002603414854559e-05,
+      "loss": 1.0237,
+      "step": 13455
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.9954524866642585e-05,
+      "loss": 0.9512,
+      "step": 13460
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.988308583545596e-05,
+      "loss": 0.9565,
+      "step": 13465
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.9811717126634066e-05,
+      "loss": 0.9656,
+      "step": 13470
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.271484375,
+      "learning_rate": 2.974041881175468e-05,
+      "loss": 0.9785,
+      "step": 13475
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2734375,
+      "learning_rate": 2.9669190962325112e-05,
+      "loss": 0.9613,
+      "step": 13480
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.959803364978184e-05,
+      "loss": 0.9328,
+      "step": 13485
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.265625,
+      "learning_rate": 2.952694694549073e-05,
+      "loss": 0.9668,
+      "step": 13490
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.9455930920746778e-05,
+      "loss": 0.8782,
+      "step": 13495
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.265625,
+      "learning_rate": 2.9384985646774053e-05,
+      "loss": 1.0128,
+      "step": 13500
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.27734375,
+      "learning_rate": 2.9314111194725757e-05,
+      "loss": 0.9231,
+      "step": 13505
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.24609375,
+      "learning_rate": 2.9243307635683957e-05,
+      "loss": 0.9255,
+      "step": 13510
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.9172575040659744e-05,
+      "loss": 0.9449,
+      "step": 13515
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.910191348059289e-05,
+      "loss": 0.8508,
+      "step": 13520
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.248046875,
+      "learning_rate": 2.9031323026352053e-05,
+      "loss": 0.9159,
+      "step": 13525
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.248046875,
+      "learning_rate": 2.8960803748734534e-05,
+      "loss": 0.9562,
+      "step": 13530
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.271484375,
+      "learning_rate": 2.8890355718466177e-05,
+      "loss": 0.8926,
+      "step": 13535
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.8819979006201526e-05,
+      "loss": 0.9727,
+      "step": 13540
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.8749673682523404e-05,
+      "loss": 0.954,
+      "step": 13545
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.8679439817943232e-05,
+      "loss": 0.8983,
+      "step": 13550
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.248046875,
+      "learning_rate": 2.860927748290061e-05,
+      "loss": 0.9354,
+      "step": 13555
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.853918674776345e-05,
+      "loss": 0.9717,
+      "step": 13560
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.84691676828279e-05,
+      "loss": 0.9196,
+      "step": 13565
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.8399220358318148e-05,
+      "loss": 0.956,
+      "step": 13570
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.832934484438652e-05,
+      "loss": 1.0523,
+      "step": 13575
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.8259541211113216e-05,
+      "loss": 1.0049,
+      "step": 13580
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.8189809528506462e-05,
+      "loss": 0.9359,
+      "step": 13585
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.81201498665022e-05,
+      "loss": 0.8533,
+      "step": 13590
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.8050562294964267e-05,
+      "loss": 0.8861,
+      "step": 13595
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.798104688368407e-05,
+      "loss": 0.9037,
+      "step": 13600
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.25,
+      "learning_rate": 2.791160370238075e-05,
+      "loss": 0.9423,
+      "step": 13605
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.23828125,
+      "learning_rate": 2.7842232820700977e-05,
+      "loss": 0.9029,
+      "step": 13610
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.294921875,
+      "learning_rate": 2.7772934308218846e-05,
+      "loss": 0.9684,
+      "step": 13615
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.265625,
+      "learning_rate": 2.7703708234435988e-05,
+      "loss": 0.9452,
+      "step": 13620
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.27734375,
+      "learning_rate": 2.7634554668781242e-05,
+      "loss": 0.9624,
+      "step": 13625
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2890625,
+      "learning_rate": 2.7565473680610887e-05,
+      "loss": 0.9741,
+      "step": 13630
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.7496465339208233e-05,
+      "loss": 0.9307,
+      "step": 13635
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.7427529713783905e-05,
+      "loss": 0.9435,
+      "step": 13640
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.7358666873475493e-05,
+      "loss": 0.8875,
+      "step": 13645
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.7289876887347554e-05,
+      "loss": 0.9351,
+      "step": 13650
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.722115982439173e-05,
+      "loss": 0.928,
+      "step": 13655
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.248046875,
+      "learning_rate": 2.7152515753526364e-05,
+      "loss": 0.8899,
+      "step": 13660
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.70839447435967e-05,
+      "loss": 1.0374,
+      "step": 13665
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.28515625,
+      "learning_rate": 2.7015446863374637e-05,
+      "loss": 0.984,
+      "step": 13670
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.6947022181558813e-05,
+      "loss": 0.8674,
+      "step": 13675
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.6878670766774328e-05,
+      "loss": 0.9809,
+      "step": 13680
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.6810392687572928e-05,
+      "loss": 0.9736,
+      "step": 13685
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.24609375,
+      "learning_rate": 2.6742188012432767e-05,
+      "loss": 0.9327,
+      "step": 13690
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.667405680975831e-05,
+      "loss": 0.923,
+      "step": 13695
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.265625,
+      "learning_rate": 2.6605999147880456e-05,
+      "loss": 1.0264,
+      "step": 13700
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.6538015095056223e-05,
+      "loss": 0.9748,
+      "step": 13705
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.271484375,
+      "learning_rate": 2.6470104719468925e-05,
+      "loss": 0.8896,
+      "step": 13710
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.6402268089227866e-05,
+      "loss": 0.9706,
+      "step": 13715
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.24609375,
+      "learning_rate": 2.6334505272368493e-05,
+      "loss": 0.968,
+      "step": 13720
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 2.626681633685213e-05,
+      "loss": 0.9703,
+      "step": 13725
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.6199201350566104e-05,
+      "loss": 0.9082,
+      "step": 13730
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.613166038132345e-05,
+      "loss": 0.9308,
+      "step": 13735
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 2.606419349686312e-05,
+      "loss": 0.9935,
+      "step": 13740
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 2.5996800764849638e-05,
+      "loss": 0.8884,
+      "step": 13745
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.5929482252873183e-05,
+      "loss": 0.9865,
+      "step": 13750
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.23828125,
+      "learning_rate": 2.5862238028449582e-05,
+      "loss": 0.9271,
+      "step": 13755
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.25,
+      "learning_rate": 2.579506815902002e-05,
+      "loss": 0.9695,
+      "step": 13760
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.5727972711951208e-05,
+      "loss": 0.9152,
+      "step": 13765
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.5660951754535245e-05,
+      "loss": 0.914,
+      "step": 13770
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.559400535398938e-05,
+      "loss": 0.9339,
+      "step": 13775
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.5527133577456254e-05,
+      "loss": 1.0445,
+      "step": 13780
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.25,
+      "learning_rate": 2.5460336492003522e-05,
+      "loss": 0.921,
+      "step": 13785
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.5393614164624047e-05,
+      "loss": 0.9167,
+      "step": 13790
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.25,
+      "learning_rate": 2.5326966662235597e-05,
+      "loss": 0.9875,
+      "step": 13795
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.5260394051681024e-05,
+      "loss": 1.0246,
+      "step": 13800
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.5193896399727945e-05,
+      "loss": 0.9762,
+      "step": 13805
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.5127473773068888e-05,
+      "loss": 0.9957,
+      "step": 13810
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.506112623832113e-05,
+      "loss": 0.9724,
+      "step": 13815
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.499485386202659e-05,
+      "loss": 0.9355,
+      "step": 13820
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.265625,
+      "learning_rate": 2.49286567106518e-05,
+      "loss": 0.9687,
+      "step": 13825
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.4862534850587925e-05,
+      "loss": 0.9089,
+      "step": 13830
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 2.4796488348150548e-05,
+      "loss": 0.8917,
+      "step": 13835
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.4730517269579667e-05,
+      "loss": 0.9085,
+      "step": 13840
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2734375,
+      "learning_rate": 2.4664621681039723e-05,
+      "loss": 0.9635,
+      "step": 13845
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.459880164861932e-05,
+      "loss": 1.0015,
+      "step": 13850
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.271484375,
+      "learning_rate": 2.453305723833139e-05,
+      "loss": 0.9528,
+      "step": 13855
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.4467388516113e-05,
+      "loss": 0.9126,
+      "step": 13860
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.248046875,
+      "learning_rate": 2.4401795547825234e-05,
+      "loss": 0.91,
+      "step": 13865
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.433627839925332e-05,
+      "loss": 0.9091,
+      "step": 13870
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.427083713610632e-05,
+      "loss": 0.9037,
+      "step": 13875
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.42054718240173e-05,
+      "loss": 0.9868,
+      "step": 13880
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.4140182528543044e-05,
+      "loss": 0.9237,
+      "step": 13885
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.265625,
+      "learning_rate": 2.4074969315164176e-05,
+      "loss": 0.9442,
+      "step": 13890
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.4009832249285035e-05,
+      "loss": 0.9621,
+      "step": 13895
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.28515625,
+      "learning_rate": 2.3944771396233467e-05,
+      "loss": 1.0088,
+      "step": 13900
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 2.387978682126104e-05,
+      "loss": 0.955,
+      "step": 13905
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.3814878589542678e-05,
+      "loss": 0.9292,
+      "step": 13910
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 2.3750046766176846e-05,
+      "loss": 0.9506,
+      "step": 13915
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.240234375,
+      "learning_rate": 2.368529141618533e-05,
+      "loss": 0.9709,
+      "step": 13920
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.362061260451316e-05,
+      "loss": 0.9349,
+      "step": 13925
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2734375,
+      "learning_rate": 2.3556010396028737e-05,
+      "loss": 0.9998,
+      "step": 13930
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.3491484855523504e-05,
+      "loss": 0.9845,
+      "step": 13935
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.3427036047712125e-05,
+      "loss": 0.9669,
+      "step": 13940
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.3362664037232184e-05,
+      "loss": 0.9731,
+      "step": 13945
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.275390625,
+      "learning_rate": 2.3298368888644386e-05,
+      "loss": 0.9392,
+      "step": 13950
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.248046875,
+      "learning_rate": 2.323415066643221e-05,
+      "loss": 0.9111,
+      "step": 13955
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.3170009435002083e-05,
+      "loss": 0.9801,
+      "step": 13960
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.3105945258683125e-05,
+      "loss": 0.9287,
+      "step": 13965
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.248046875,
+      "learning_rate": 2.3041958201727275e-05,
+      "loss": 0.9081,
+      "step": 13970
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.24609375,
+      "learning_rate": 2.29780483283091e-05,
+      "loss": 0.856,
+      "step": 13975
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.298828125,
+      "learning_rate": 2.2914215702525677e-05,
+      "loss": 0.9447,
+      "step": 13980
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.287109375,
+      "learning_rate": 2.285046038839672e-05,
+      "loss": 1.0183,
+      "step": 13985
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.2786782449864298e-05,
+      "loss": 0.9925,
+      "step": 13990
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.2723181950793004e-05,
+      "loss": 0.9696,
+      "step": 13995
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.27734375,
+      "learning_rate": 2.265965895496962e-05,
+      "loss": 0.9292,
+      "step": 14000
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.25,
+      "learning_rate": 2.259621352610333e-05,
+      "loss": 0.9397,
+      "step": 14005
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.2532845727825445e-05,
+      "loss": 0.89,
+      "step": 14010
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 2.246955562368941e-05,
+      "loss": 0.954,
+      "step": 14015
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.2406343277170827e-05,
+      "loss": 0.9743,
+      "step": 14020
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2421875,
+      "learning_rate": 2.234320875166721e-05,
+      "loss": 0.9589,
+      "step": 14025
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 2.228015211049813e-05,
+      "loss": 0.9826,
+      "step": 14030
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.2217173416904935e-05,
+      "loss": 0.9069,
+      "step": 14035
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.2154272734050908e-05,
+      "loss": 0.9242,
+      "step": 14040
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.29296875,
+      "learning_rate": 2.2091450125020986e-05,
+      "loss": 0.9874,
+      "step": 14045
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.271484375,
+      "learning_rate": 2.202870565282188e-05,
+      "loss": 0.9641,
+      "step": 14050
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.1966039380381944e-05,
+      "loss": 0.9674,
+      "step": 14055
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.30078125,
+      "learning_rate": 2.1903451370551e-05,
+      "loss": 0.9647,
+      "step": 14060
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 2.1840941686100524e-05,
+      "loss": 0.9176,
+      "step": 14065
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.1778510389723283e-05,
+      "loss": 0.923,
+      "step": 14070
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.1716157544033578e-05,
+      "loss": 0.8969,
+      "step": 14075
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 2.1653883211566895e-05,
+      "loss": 0.9502,
+      "step": 14080
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.1591687454780087e-05,
+      "loss": 0.985,
+      "step": 14085
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.152957033605111e-05,
+      "loss": 0.9258,
+      "step": 14090
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.30078125,
+      "learning_rate": 2.146753191767913e-05,
+      "loss": 0.8852,
+      "step": 14095
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.265625,
+      "learning_rate": 2.14055722618843e-05,
+      "loss": 0.8649,
+      "step": 14100
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2734375,
+      "learning_rate": 2.1343691430807878e-05,
+      "loss": 0.9535,
+      "step": 14105
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.275390625,
+      "learning_rate": 2.1281889486511984e-05,
+      "loss": 1.0796,
+      "step": 14110
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.122016649097962e-05,
+      "loss": 0.937,
+      "step": 14115
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.1158522506114696e-05,
+      "loss": 0.866,
+      "step": 14120
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.1096957593741772e-05,
+      "loss": 1.0118,
+      "step": 14125
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.265625,
+      "learning_rate": 2.1035471815606178e-05,
+      "loss": 0.916,
+      "step": 14130
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.0974065233373864e-05,
+      "loss": 0.9609,
+      "step": 14135
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.275390625,
+      "learning_rate": 2.0912737908631296e-05,
+      "loss": 0.9456,
+      "step": 14140
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.30078125,
+      "learning_rate": 2.0851489902885556e-05,
+      "loss": 0.9138,
+      "step": 14145
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.079032127756405e-05,
+      "loss": 0.9634,
+      "step": 14150
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.244140625,
+      "learning_rate": 2.0729232094014693e-05,
+      "loss": 0.9542,
+      "step": 14155
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.066822241350561e-05,
+      "loss": 0.9578,
+      "step": 14160
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.06072922972253e-05,
+      "loss": 0.9623,
+      "step": 14165
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 2.0546441806282367e-05,
+      "loss": 0.9994,
+      "step": 14170
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.048567100170562e-05,
+      "loss": 0.9908,
+      "step": 14175
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.0424979944443946e-05,
+      "loss": 0.9765,
+      "step": 14180
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.271484375,
+      "learning_rate": 2.0364368695366233e-05,
+      "loss": 1.0086,
+      "step": 14185
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.0303837315261277e-05,
+      "loss": 0.9459,
+      "step": 14190
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.29296875,
+      "learning_rate": 2.0243385864837884e-05,
+      "loss": 0.9684,
+      "step": 14195
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.25,
+      "learning_rate": 2.018301440472461e-05,
+      "loss": 1.0211,
+      "step": 14200
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.012272299546978e-05,
+      "loss": 0.9033,
+      "step": 14205
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.0062511697541485e-05,
+      "loss": 0.8911,
+      "step": 14210
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.0002380571327482e-05,
+      "loss": 0.9106,
+      "step": 14215
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.994232967713505e-05,
+      "loss": 0.8815,
+      "step": 14220
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.988235907519107e-05,
+      "loss": 0.9802,
+      "step": 14225
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9822468825641816e-05,
+      "loss": 0.9102,
+      "step": 14230
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.25,
+      "learning_rate": 1.9762658988553086e-05,
+      "loss": 0.9642,
+      "step": 14235
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9702929623909903e-05,
+      "loss": 0.9312,
+      "step": 14240
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9643280791616702e-05,
+      "loss": 0.9633,
+      "step": 14245
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9583712551497036e-05,
+      "loss": 0.9299,
+      "step": 14250
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9524224963293726e-05,
+      "loss": 0.9632,
+      "step": 14255
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 1.946481808666868e-05,
+      "loss": 0.9104,
+      "step": 14260
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 1.9405491981202794e-05,
+      "loss": 0.9729,
+      "step": 14265
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9346246706396066e-05,
+      "loss": 1.0376,
+      "step": 14270
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9287082321667295e-05,
+      "loss": 0.892,
+      "step": 14275
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.9227998886354305e-05,
+      "loss": 0.9092,
+      "step": 14280
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9168996459713594e-05,
+      "loss": 0.9831,
+      "step": 14285
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9110075100920466e-05,
+      "loss": 0.9946,
+      "step": 14290
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.9051234869068978e-05,
+      "loss": 0.9851,
+      "step": 14295
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.8992475823171717e-05,
+      "loss": 0.9481,
+      "step": 14300
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.8933798022159943e-05,
+      "loss": 0.9748,
+      "step": 14305
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.8875201524883347e-05,
+      "loss": 0.9562,
+      "step": 14310
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.8816686390110172e-05,
+      "loss": 0.9724,
+      "step": 14315
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 1.8758252676526945e-05,
+      "loss": 0.8649,
+      "step": 14320
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.869990044273867e-05,
+      "loss": 0.9095,
+      "step": 14325
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 1.8641629747268486e-05,
+      "loss": 0.8604,
+      "step": 14330
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.858344064855787e-05,
+      "loss": 0.9063,
+      "step": 14335
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.25,
+      "learning_rate": 1.8525333204966443e-05,
+      "loss": 0.9234,
+      "step": 14340
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.8467307474771855e-05,
+      "loss": 0.9857,
+      "step": 14345
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.8409363516169907e-05,
+      "loss": 0.9724,
+      "step": 14350
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.8351501387274307e-05,
+      "loss": 0.9585,
+      "step": 14355
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.829372114611675e-05,
+      "loss": 0.8978,
+      "step": 14360
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.823602285064675e-05,
+      "loss": 0.9764,
+      "step": 14365
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.817840655873171e-05,
+      "loss": 0.8856,
+      "step": 14370
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 1.8120872328156713e-05,
+      "loss": 0.9027,
+      "step": 14375
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.8063420216624548e-05,
+      "loss": 0.845,
+      "step": 14380
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.8006050281755725e-05,
+      "loss": 0.8853,
+      "step": 14385
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.794876258108823e-05,
+      "loss": 0.8895,
+      "step": 14390
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.7891557172077666e-05,
+      "loss": 1.0076,
+      "step": 14395
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.25,
+      "learning_rate": 1.783443411209701e-05,
+      "loss": 0.9455,
+      "step": 14400
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.7777393458436753e-05,
+      "loss": 0.915,
+      "step": 14405
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.7720435268304625e-05,
+      "loss": 1.0042,
+      "step": 14410
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.766355959882575e-05,
+      "loss": 0.9392,
+      "step": 14415
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.7606766507042473e-05,
+      "loss": 0.9593,
+      "step": 14420
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.755005604991423e-05,
+      "loss": 0.9684,
+      "step": 14425
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.7493428284317716e-05,
+      "loss": 0.885,
+      "step": 14430
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.7436883267046567e-05,
+      "loss": 0.9329,
+      "step": 14435
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.7380421054811524e-05,
+      "loss": 1.0126,
+      "step": 14440
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 1.732404170424019e-05,
+      "loss": 0.8792,
+      "step": 14445
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.7267745271877177e-05,
+      "loss": 0.9785,
+      "step": 14450
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.7211531814183825e-05,
+      "loss": 0.8623,
+      "step": 14455
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.7155401387538327e-05,
+      "loss": 0.9199,
+      "step": 14460
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.25,
+      "learning_rate": 1.709935404823555e-05,
+      "loss": 0.8532,
+      "step": 14465
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.70433898524871e-05,
+      "loss": 0.9295,
+      "step": 14470
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.698750885642114e-05,
+      "loss": 0.9423,
+      "step": 14475
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.693171111608237e-05,
+      "loss": 0.9283,
+      "step": 14480
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.6875996687432073e-05,
+      "loss": 0.9966,
+      "step": 14485
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.6820365626347878e-05,
+      "loss": 0.9173,
+      "step": 14490
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.676481798862387e-05,
+      "loss": 0.8821,
+      "step": 14495
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.6709353829970496e-05,
+      "loss": 0.9018,
+      "step": 14500
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.665397320601436e-05,
+      "loss": 0.8973,
+      "step": 14505
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.25,
+      "learning_rate": 1.659867617229841e-05,
+      "loss": 0.8835,
+      "step": 14510
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.654346278428166e-05,
+      "loss": 0.9423,
+      "step": 14515
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.25,
+      "learning_rate": 1.6488333097339335e-05,
+      "loss": 0.9255,
+      "step": 14520
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.64332871667626e-05,
+      "loss": 0.8827,
+      "step": 14525
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.6378325047758723e-05,
+      "loss": 0.9508,
+      "step": 14530
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 1.6323446795450826e-05,
+      "loss": 0.9225,
+      "step": 14535
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.6268652464877988e-05,
+      "loss": 1.091,
+      "step": 14540
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.6213942110995105e-05,
+      "loss": 1.0095,
+      "step": 14545
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.6159315788672825e-05,
+      "loss": 1.0021,
+      "step": 14550
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 1.6104773552697517e-05,
+      "loss": 0.9367,
+      "step": 14555
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.6050315457771257e-05,
+      "loss": 0.954,
+      "step": 14560
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.24609375,
+      "learning_rate": 1.5995941558511695e-05,
+      "loss": 0.9226,
+      "step": 14565
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.5941651909452028e-05,
+      "loss": 0.9882,
+      "step": 14570
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.5887446565041007e-05,
+      "loss": 0.9663,
+      "step": 14575
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.583332557964282e-05,
+      "loss": 0.9891,
+      "step": 14580
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.5779289007537e-05,
+      "loss": 0.9168,
+      "step": 14585
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.25,
+      "learning_rate": 1.5725336902918486e-05,
+      "loss": 0.952,
+      "step": 14590
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.25,
+      "learning_rate": 1.5671469319897425e-05,
+      "loss": 0.9711,
+      "step": 14595
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.561768631249929e-05,
+      "loss": 0.9587,
+      "step": 14600
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.5563987934664624e-05,
+      "loss": 0.9205,
+      "step": 14605
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.5510374240249205e-05,
+      "loss": 0.9584,
+      "step": 14610
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.5456845283023758e-05,
+      "loss": 0.9561,
+      "step": 14615
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.540340111667413e-05,
+      "loss": 0.8794,
+      "step": 14620
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.5350041794801097e-05,
+      "loss": 0.9399,
+      "step": 14625
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.5296767370920273e-05,
+      "loss": 0.9739,
+      "step": 14630
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.5243577898462246e-05,
+      "loss": 0.9206,
+      "step": 14635
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.5190473430772289e-05,
+      "loss": 1.0323,
+      "step": 14640
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.5137454021110508e-05,
+      "loss": 0.9263,
+      "step": 14645
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.5084519722651658e-05,
+      "loss": 0.8936,
+      "step": 14650
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.5031670588485103e-05,
+      "loss": 0.9072,
+      "step": 14655
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.497890667161489e-05,
+      "loss": 0.9722,
+      "step": 14660
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 1.4926228024959487e-05,
+      "loss": 0.8629,
+      "step": 14665
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 1.4873634701351946e-05,
+      "loss": 0.8868,
+      "step": 14670
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.4821126753539638e-05,
+      "loss": 0.98,
+      "step": 14675
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.4768704234184428e-05,
+      "loss": 0.9752,
+      "step": 14680
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.471636719586238e-05,
+      "loss": 0.9944,
+      "step": 14685
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.466411569106393e-05,
+      "loss": 0.9097,
+      "step": 14690
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.4611949772193657e-05,
+      "loss": 0.8881,
+      "step": 14695
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.4559869491570332e-05,
+      "loss": 0.8736,
+      "step": 14700
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.4507874901426877e-05,
+      "loss": 0.9889,
+      "step": 14705
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.4455966053910187e-05,
+      "loss": 0.9122,
+      "step": 14710
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.4404143001081238e-05,
+      "loss": 0.9335,
+      "step": 14715
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.4352405794914892e-05,
+      "loss": 0.934,
+      "step": 14720
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.4300754487300006e-05,
+      "loss": 0.9494,
+      "step": 14725
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.4249189130039175e-05,
+      "loss": 0.9182,
+      "step": 14730
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.24609375,
+      "learning_rate": 1.419770977484891e-05,
+      "loss": 0.9858,
+      "step": 14735
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.4146316473359366e-05,
+      "loss": 0.9127,
+      "step": 14740
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.4095009277114412e-05,
+      "loss": 0.9651,
+      "step": 14745
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.4043788237571632e-05,
+      "loss": 0.9079,
+      "step": 14750
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.3992653406102097e-05,
+      "loss": 0.9679,
+      "step": 14755
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.3941604833990528e-05,
+      "loss": 0.9932,
+      "step": 14760
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.389064257243502e-05,
+      "loss": 0.8945,
+      "step": 14765
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.3839766672547206e-05,
+      "loss": 0.9169,
+      "step": 14770
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.3788977185352003e-05,
+      "loss": 0.9185,
+      "step": 14775
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.3738274161787768e-05,
+      "loss": 0.9105,
+      "step": 14780
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.3687657652706076e-05,
+      "loss": 0.9342,
+      "step": 14785
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 1.3637127708871734e-05,
+      "loss": 0.9612,
+      "step": 14790
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.3586684380962778e-05,
+      "loss": 0.9476,
+      "step": 14795
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.244140625,
+      "learning_rate": 1.3536327719570286e-05,
+      "loss": 0.895,
+      "step": 14800
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.25,
+      "learning_rate": 1.3486057775198535e-05,
+      "loss": 0.9524,
+      "step": 14805
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.3435874598264709e-05,
+      "loss": 1.0229,
+      "step": 14810
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.3385778239099067e-05,
+      "loss": 0.9131,
+      "step": 14815
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.3335768747944722e-05,
+      "loss": 0.8996,
+      "step": 14820
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.3285846174957728e-05,
+      "loss": 0.9554,
+      "step": 14825
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.3236010570206914e-05,
+      "loss": 0.9014,
+      "step": 14830
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2421875,
+      "learning_rate": 1.3186261983673942e-05,
+      "loss": 0.9185,
+      "step": 14835
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.240234375,
+      "learning_rate": 1.3136600465253147e-05,
+      "loss": 0.9184,
+      "step": 14840
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.308702606475154e-05,
+      "loss": 0.9709,
+      "step": 14845
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.3037538831888819e-05,
+      "loss": 0.9465,
+      "step": 14850
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.298813881629718e-05,
+      "loss": 0.8841,
+      "step": 14855
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.2938826067521404e-05,
+      "loss": 0.8585,
+      "step": 14860
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.2889600635018762e-05,
+      "loss": 1.0142,
+      "step": 14865
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.2840462568158874e-05,
+      "loss": 0.8854,
+      "step": 14870
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.244140625,
+      "learning_rate": 1.2791411916223827e-05,
+      "loss": 0.9188,
+      "step": 14875
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.2742448728407963e-05,
+      "loss": 0.9166,
+      "step": 14880
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.2693573053817976e-05,
+      "loss": 0.8888,
+      "step": 14885
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.2644784941472699e-05,
+      "loss": 0.9281,
+      "step": 14890
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 1.2596084440303258e-05,
+      "loss": 0.9147,
+      "step": 14895
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.2547471599152804e-05,
+      "loss": 0.9626,
+      "step": 14900
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.2498946466776639e-05,
+      "loss": 0.9859,
+      "step": 14905
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.24505090918421e-05,
+      "loss": 0.9697,
+      "step": 14910
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.240215952292847e-05,
+      "loss": 0.9919,
+      "step": 14915
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.235389780852696e-05,
+      "loss": 0.9465,
+      "step": 14920
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.2305723997040752e-05,
+      "loss": 0.9491,
+      "step": 14925
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.2257638136784777e-05,
+      "loss": 0.931,
+      "step": 14930
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.2209640275985779e-05,
+      "loss": 0.9354,
+      "step": 14935
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.2161730462782283e-05,
+      "loss": 0.9203,
+      "step": 14940
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.211390874522449e-05,
+      "loss": 1.0151,
+      "step": 14945
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.2066175171274219e-05,
+      "loss": 0.947,
+      "step": 14950
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 1.2018529788804932e-05,
+      "loss": 0.8716,
+      "step": 14955
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.1970972645601587e-05,
+      "loss": 0.9291,
+      "step": 14960
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.1923503789360712e-05,
+      "loss": 0.952,
+      "step": 14965
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.1876123267690209e-05,
+      "loss": 0.9042,
+      "step": 14970
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.1828831128109475e-05,
+      "loss": 0.9913,
+      "step": 14975
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.1781627418049179e-05,
+      "loss": 0.9471,
+      "step": 14980
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.1734512184851377e-05,
+      "loss": 0.9306,
+      "step": 14985
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.228515625,
+      "learning_rate": 1.1687485475769343e-05,
+      "loss": 0.9082,
+      "step": 14990
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.1640547337967577e-05,
+      "loss": 0.9856,
+      "step": 14995
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.1593697818521765e-05,
+      "loss": 0.8916,
+      "step": 15000
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.1546936964418664e-05,
+      "loss": 0.9843,
+      "step": 15005
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 1.1500264822556194e-05,
+      "loss": 0.8997,
+      "step": 15010
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 1.1453681439743224e-05,
+      "loss": 0.9233,
+      "step": 15015
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 1.1407186862699614e-05,
+      "loss": 0.8995,
+      "step": 15020
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.1360781138056209e-05,
+      "loss": 0.9615,
+      "step": 15025
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 1.1314464312354678e-05,
+      "loss": 1.0071,
+      "step": 15030
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.1268236432047596e-05,
+      "loss": 0.9332,
+      "step": 15035
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.1222097543498244e-05,
+      "loss": 0.9835,
+      "step": 15040
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.1176047692980773e-05,
+      "loss": 0.929,
+      "step": 15045
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.1130086926679894e-05,
+      "loss": 0.9613,
+      "step": 15050
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.1084215290691092e-05,
+      "loss": 0.9264,
+      "step": 15055
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.1038432831020384e-05,
+      "loss": 1.0485,
+      "step": 15060
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.099273959358439e-05,
+      "loss": 0.9286,
+      "step": 15065
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 1.0947135624210247e-05,
+      "loss": 0.85,
+      "step": 15070
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.0901620968635517e-05,
+      "loss": 0.9062,
+      "step": 15075
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.0856195672508262e-05,
+      "loss": 0.9725,
+      "step": 15080
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.0810859781386828e-05,
+      "loss": 0.947,
+      "step": 15085
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.0765613340739989e-05,
+      "loss": 0.9843,
+      "step": 15090
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.0720456395946732e-05,
+      "loss": 1.0915,
+      "step": 15095
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 1.0675388992296353e-05,
+      "loss": 0.8977,
+      "step": 15100
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.0630411174988275e-05,
+      "loss": 0.9742,
+      "step": 15105
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.0585522989132102e-05,
+      "loss": 0.9579,
+      "step": 15110
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.0540724479747587e-05,
+      "loss": 0.9284,
+      "step": 15115
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.0496015691764461e-05,
+      "loss": 0.9307,
+      "step": 15120
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.045139667002254e-05,
+      "loss": 0.9269,
+      "step": 15125
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.0406867459271564e-05,
+      "loss": 0.8834,
+      "step": 15130
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.036242810417124e-05,
+      "loss": 0.8787,
+      "step": 15135
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.0318078649291119e-05,
+      "loss": 0.9375,
+      "step": 15140
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.0273819139110608e-05,
+      "loss": 0.9264,
+      "step": 15145
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 1.0229649618018933e-05,
+      "loss": 0.9261,
+      "step": 15150
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.0185570130314991e-05,
+      "loss": 0.9527,
+      "step": 15155
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.0141580720207466e-05,
+      "loss": 0.9477,
+      "step": 15160
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.0097681431814621e-05,
+      "loss": 1.0199,
+      "step": 15165
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.0053872309164414e-05,
+      "loss": 1.0186,
+      "step": 15170
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.00101533961943e-05,
+      "loss": 0.9518,
+      "step": 15175
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.966524736751337e-06,
+      "loss": 0.9136,
+      "step": 15180
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.922986374591969e-06,
+      "loss": 0.923,
+      "step": 15185
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.25,
+      "learning_rate": 9.879538353382178e-06,
+      "loss": 0.9558,
+      "step": 15190
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.836180716697251e-06,
+      "loss": 0.9489,
+      "step": 15195
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.792913508021906e-06,
+      "loss": 0.9487,
+      "step": 15200
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.749736770750106e-06,
+      "loss": 0.953,
+      "step": 15205
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.706650548185091e-06,
+      "loss": 0.938,
+      "step": 15210
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.663654883539364e-06,
+      "loss": 0.9389,
+      "step": 15215
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.620749819934539e-06,
+      "loss": 0.9893,
+      "step": 15220
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.577935400401406e-06,
+      "loss": 0.9918,
+      "step": 15225
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.535211667879861e-06,
+      "loss": 0.9645,
+      "step": 15230
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.25,
+      "learning_rate": 9.492578665218788e-06,
+      "loss": 0.9379,
+      "step": 15235
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.450036435176136e-06,
+      "loss": 0.9619,
+      "step": 15240
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.407585020418763e-06,
+      "loss": 0.9583,
+      "step": 15245
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.365224463522492e-06,
+      "loss": 0.8578,
+      "step": 15250
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.322954806971985e-06,
+      "loss": 0.9276,
+      "step": 15255
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.25,
+      "learning_rate": 9.280776093160782e-06,
+      "loss": 0.9916,
+      "step": 15260
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.238688364391135e-06,
+      "loss": 0.9697,
+      "step": 15265
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.196691662874135e-06,
+      "loss": 0.9859,
+      "step": 15270
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.248046875,
+      "learning_rate": 9.154786030729545e-06,
+      "loss": 0.9831,
+      "step": 15275
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.112971509985757e-06,
+      "loss": 0.9049,
+      "step": 15280
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.071248142579825e-06,
+      "loss": 0.9565,
+      "step": 15285
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.029615970357375e-06,
+      "loss": 0.9601,
+      "step": 15290
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.255859375,
+      "learning_rate": 8.988075035072562e-06,
+      "loss": 0.8574,
+      "step": 15295
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.267578125,
+      "learning_rate": 8.946625378388029e-06,
+      "loss": 0.9204,
+      "step": 15300
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2734375,
+      "learning_rate": 8.905267041874887e-06,
+      "loss": 0.941,
+      "step": 15305
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.26171875,
+      "learning_rate": 8.864000067012702e-06,
+      "loss": 0.9745,
+      "step": 15310
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.25390625,
+      "learning_rate": 8.822824495189297e-06,
+      "loss": 0.9844,
+      "step": 15315
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.255859375,
+      "learning_rate": 8.781740367700941e-06,
+      "loss": 0.8501,
+      "step": 15320
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.244140625,
+      "learning_rate": 8.740747725752118e-06,
+      "loss": 0.9523,
+      "step": 15325
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.69984661045562e-06,
+      "loss": 0.952,
+      "step": 15330
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.659037062832342e-06,
+      "loss": 0.9723,
+      "step": 15335
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.25,
+      "learning_rate": 8.61831912381148e-06,
+      "loss": 0.9169,
+      "step": 15340
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.25,
+      "learning_rate": 8.577692834230223e-06,
+      "loss": 0.9249,
+      "step": 15345
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.29296875,
+      "learning_rate": 8.53715823483393e-06,
+      "loss": 0.9725,
+      "step": 15350
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.496715366275976e-06,
+      "loss": 0.9398,
+      "step": 15355
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.456364269117711e-06,
+      "loss": 0.8724,
+      "step": 15360
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.265625,
+      "learning_rate": 8.416104983828499e-06,
+      "loss": 0.9182,
+      "step": 15365
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.265625,
+      "learning_rate": 8.375937550785539e-06,
+      "loss": 0.9904,
+      "step": 15370
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.263671875,
+      "learning_rate": 8.335862010274009e-06,
+      "loss": 1.0124,
+      "step": 15375
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.295878402486856e-06,
+      "loss": 0.9373,
+      "step": 15380
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 8.255986767524826e-06,
+      "loss": 0.892,
+      "step": 15385
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.26171875,
+      "learning_rate": 8.21618714539647e-06,
+      "loss": 0.9043,
+      "step": 15390
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.176479576018003e-06,
+      "loss": 0.933,
+      "step": 15395
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.267578125,
+      "learning_rate": 8.136864099213359e-06,
+      "loss": 0.9752,
+      "step": 15400
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.27734375,
+      "learning_rate": 8.097340754714078e-06,
+      "loss": 0.9664,
+      "step": 15405
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 8.05790958215934e-06,
+      "loss": 0.8931,
+      "step": 15410
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.25390625,
+      "learning_rate": 8.018570621095822e-06,
+      "loss": 0.9852,
+      "step": 15415
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.979323910977787e-06,
+      "loss": 0.95,
+      "step": 15420
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.265625,
+      "learning_rate": 7.940169491166904e-06,
+      "loss": 0.9871,
+      "step": 15425
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.26171875,
+      "learning_rate": 7.90110740093234e-06,
+      "loss": 0.8466,
+      "step": 15430
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.251953125,
+      "learning_rate": 7.862137679450653e-06,
+      "loss": 0.8754,
+      "step": 15435
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.27734375,
+      "learning_rate": 7.823260365805717e-06,
+      "loss": 0.9045,
+      "step": 15440
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.251953125,
+      "learning_rate": 7.784475498988808e-06,
+      "loss": 0.9642,
+      "step": 15445
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.255859375,
+      "learning_rate": 7.745783117898397e-06,
+      "loss": 0.9163,
+      "step": 15450
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.25390625,
+      "learning_rate": 7.707183261340255e-06,
+      "loss": 1.0132,
+      "step": 15455
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.25390625,
+      "learning_rate": 7.668675968027328e-06,
+      "loss": 0.9522,
+      "step": 15460
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.25390625,
+      "learning_rate": 7.630261276579765e-06,
+      "loss": 0.9135,
+      "step": 15465
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.244140625,
+      "learning_rate": 7.5919392255248025e-06,
+      "loss": 0.983,
+      "step": 15470
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.271484375,
+      "learning_rate": 7.553709853296764e-06,
+      "loss": 0.9665,
+      "step": 15475
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.25390625,
+      "learning_rate": 7.515573198237069e-06,
+      "loss": 0.9158,
+      "step": 15480
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 7.477529298594077e-06,
+      "loss": 0.9633,
+      "step": 15485
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.26171875,
+      "learning_rate": 7.439578192523211e-06,
+      "loss": 0.9665,
+      "step": 15490
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.251953125,
+      "learning_rate": 7.4017199180867246e-06,
+      "loss": 1.0238,
+      "step": 15495
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2421875,
+      "learning_rate": 7.363954513253879e-06,
+      "loss": 0.955,
+      "step": 15500
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.248046875,
+      "learning_rate": 7.3262820159006765e-06,
+      "loss": 0.857,
+      "step": 15505
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.244140625,
+      "learning_rate": 7.288702463810026e-06,
+      "loss": 0.958,
+      "step": 15510
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.28125,
+      "learning_rate": 7.2512158946716145e-06,
+      "loss": 0.9374,
+      "step": 15515
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 7.213822346081822e-06,
+      "loss": 0.9358,
+      "step": 15520
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.265625,
+      "learning_rate": 7.176521855543772e-06,
+      "loss": 0.962,
+      "step": 15525
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2734375,
+      "learning_rate": 7.13931446046725e-06,
+      "loss": 0.9796,
+      "step": 15530
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.1022001981686845e-06,
+      "loss": 0.9309,
+      "step": 15535
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.28125,
+      "learning_rate": 7.065179105871056e-06,
+      "loss": 0.9524,
+      "step": 15540
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.255859375,
+      "learning_rate": 7.028251220703974e-06,
+      "loss": 0.9123,
+      "step": 15545
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.26953125,
+      "learning_rate": 6.99141657970348e-06,
+      "loss": 0.9028,
+      "step": 15550
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 6.954675219812201e-06,
+      "loss": 0.9712,
+      "step": 15555
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.918027177879094e-06,
+      "loss": 0.9409,
+      "step": 15560
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.275390625,
+      "learning_rate": 6.881472490659635e-06,
+      "loss": 0.9266,
+      "step": 15565
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.255859375,
+      "learning_rate": 6.845011194815598e-06,
+      "loss": 0.9021,
+      "step": 15570
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.263671875,
+      "learning_rate": 6.808643326915087e-06,
+      "loss": 0.9928,
+      "step": 15575
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.267578125,
+      "learning_rate": 6.772368923432593e-06,
+      "loss": 0.9648,
+      "step": 15580
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.271484375,
+      "learning_rate": 6.736188020748746e-06,
+      "loss": 0.9538,
+      "step": 15585
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.263671875,
+      "learning_rate": 6.700100655150487e-06,
+      "loss": 0.9885,
+      "step": 15590
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.255859375,
+      "learning_rate": 6.6641068628309545e-06,
+      "loss": 0.9485,
+      "step": 15595
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.287109375,
+      "learning_rate": 6.628206679889349e-06,
+      "loss": 0.9249,
+      "step": 15600
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.28125,
+      "learning_rate": 6.5924001423311014e-06,
+      "loss": 0.9412,
+      "step": 15605
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 6.55668728606762e-06,
+      "loss": 0.8892,
+      "step": 15610
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 6.521068146916432e-06,
+      "loss": 0.9258,
+      "step": 15615
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.24609375,
+      "learning_rate": 6.485542760601027e-06,
+      "loss": 0.9151,
+      "step": 15620
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.27734375,
+      "learning_rate": 6.450111162750905e-06,
+      "loss": 0.9967,
+      "step": 15625
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 6.414773388901452e-06,
+      "loss": 0.872,
+      "step": 15630
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.379529474494006e-06,
+      "loss": 0.8878,
+      "step": 15635
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.251953125,
+      "learning_rate": 6.344379454875771e-06,
+      "loss": 0.9422,
+      "step": 15640
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 6.309323365299724e-06,
+      "loss": 0.9509,
+      "step": 15645
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2734375,
+      "learning_rate": 6.274361240924676e-06,
+      "loss": 0.9587,
+      "step": 15650
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 6.239493116815231e-06,
+      "loss": 0.9111,
+      "step": 15655
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.204719027941641e-06,
+      "loss": 0.8894,
+      "step": 15660
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.248046875,
+      "learning_rate": 6.170039009179895e-06,
+      "loss": 0.8962,
+      "step": 15665
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2890625,
+      "learning_rate": 6.135453095311627e-06,
+      "loss": 0.923,
+      "step": 15670
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 6.100961321024112e-06,
+      "loss": 0.9691,
+      "step": 15675
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.25,
+      "learning_rate": 6.066563720910168e-06,
+      "loss": 1.0003,
+      "step": 15680
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2578125,
+      "learning_rate": 6.032260329468198e-06,
+      "loss": 0.9405,
+      "step": 15685
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.998051181102082e-06,
+      "loss": 0.8734,
+      "step": 15690
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.255859375,
+      "learning_rate": 5.963936310121243e-06,
+      "loss": 0.9619,
+      "step": 15695
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.248046875,
+      "learning_rate": 5.929915750740478e-06,
+      "loss": 0.9116,
+      "step": 15700
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 5.89598953708006e-06,
+      "loss": 0.9071,
+      "step": 15705
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.862157703165583e-06,
+      "loss": 0.9501,
+      "step": 15710
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.251953125,
+      "learning_rate": 5.828420282928016e-06,
+      "loss": 0.9906,
+      "step": 15715
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.263671875,
+      "learning_rate": 5.794777310203658e-06,
+      "loss": 0.9415,
+      "step": 15720
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.25,
+      "learning_rate": 5.761228818734032e-06,
+      "loss": 0.9114,
+      "step": 15725
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 5.727774842165956e-06,
+      "loss": 0.9124,
+      "step": 15730
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.267578125,
+      "learning_rate": 5.694415414051402e-06,
+      "loss": 0.8599,
+      "step": 15735
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.263671875,
+      "learning_rate": 5.6611505678475726e-06,
+      "loss": 0.9767,
+      "step": 15740
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.259765625,
+      "learning_rate": 5.627980336916772e-06,
+      "loss": 0.9614,
+      "step": 15745
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.263671875,
+      "learning_rate": 5.594904754526398e-06,
+      "loss": 0.9834,
+      "step": 15750
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2578125,
+      "learning_rate": 5.561923853848994e-06,
+      "loss": 0.962,
+      "step": 15755
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.25,
+      "learning_rate": 5.529037667962067e-06,
+      "loss": 0.9628,
+      "step": 15760
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.322265625,
+      "learning_rate": 5.496246229848179e-06,
+      "loss": 0.9037,
+      "step": 15765
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.255859375,
+      "learning_rate": 5.463549572394833e-06,
+      "loss": 0.8848,
+      "step": 15770
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.244140625,
+      "learning_rate": 5.4309477283945194e-06,
+      "loss": 0.89,
+      "step": 15775
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2578125,
+      "learning_rate": 5.3984407305445736e-06,
+      "loss": 0.9684,
+      "step": 15780
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.263671875,
+      "learning_rate": 5.366028611447282e-06,
+      "loss": 1.0014,
+      "step": 15785
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.244140625,
+      "learning_rate": 5.333711403609698e-06,
+      "loss": 0.9274,
+      "step": 15790
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2421875,
+      "learning_rate": 5.301489139443738e-06,
+      "loss": 0.9214,
+      "step": 15795
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.271484375,
+      "learning_rate": 5.269361851266097e-06,
+      "loss": 1.0203,
+      "step": 15800
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.27734375,
+      "learning_rate": 5.237329571298166e-06,
+      "loss": 0.9583,
+      "step": 15805
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.28515625,
+      "learning_rate": 5.205392331666126e-06,
+      "loss": 0.9585,
+      "step": 15810
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.173550164400753e-06,
+      "loss": 0.9791,
+      "step": 15815
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.287109375,
+      "learning_rate": 5.141803101437559e-06,
+      "loss": 0.9917,
+      "step": 15820
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.234375,
+      "learning_rate": 5.110151174616584e-06,
+      "loss": 0.851,
+      "step": 15825
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.271484375,
+      "learning_rate": 5.078594415682536e-06,
+      "loss": 0.9701,
+      "step": 15830
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.287109375,
+      "learning_rate": 5.04713285628462e-06,
+      "loss": 0.9642,
+      "step": 15835
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2578125,
+      "learning_rate": 5.015766527976573e-06,
+      "loss": 0.9995,
+      "step": 15840
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.263671875,
+      "learning_rate": 4.984495462216654e-06,
+      "loss": 0.9138,
+      "step": 15845
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.259765625,
+      "learning_rate": 4.953319690367531e-06,
+      "loss": 1.0434,
+      "step": 15850
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.283203125,
+      "learning_rate": 4.922239243696358e-06,
+      "loss": 0.9751,
+      "step": 15855
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 4.891254153374614e-06,
+      "loss": 0.8599,
+      "step": 15860
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2734375,
+      "learning_rate": 4.860364450478206e-06,
+      "loss": 1.0107,
+      "step": 15865
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.251953125,
+      "learning_rate": 4.829570165987318e-06,
+      "loss": 0.9392,
+      "step": 15870
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2421875,
+      "learning_rate": 4.798871330786492e-06,
+      "loss": 0.9362,
+      "step": 15875
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.248046875,
+      "learning_rate": 4.7682679756645195e-06,
+      "loss": 0.9688,
+      "step": 15880
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 4.737760131314406e-06,
+      "loss": 0.8777,
+      "step": 15885
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.27734375,
+      "learning_rate": 4.707347828333408e-06,
+      "loss": 0.9369,
+      "step": 15890
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2734375,
+      "learning_rate": 4.677031097222906e-06,
+      "loss": 0.8431,
+      "step": 15895
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.26171875,
+      "learning_rate": 4.646809968388499e-06,
+      "loss": 0.946,
+      "step": 15900
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.26171875,
+      "learning_rate": 4.616684472139842e-06,
+      "loss": 0.9039,
+      "step": 15905
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2578125,
+      "learning_rate": 4.586654638690724e-06,
+      "loss": 0.9346,
+      "step": 15910
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.271484375,
+      "learning_rate": 4.55672049815895e-06,
+      "loss": 0.9267,
+      "step": 15915
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.267578125,
+      "learning_rate": 4.5268820805663855e-06,
+      "loss": 0.885,
+      "step": 15920
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.26171875,
+      "learning_rate": 4.497139415838858e-06,
+      "loss": 0.9064,
+      "step": 15925
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.25,
+      "learning_rate": 4.467492533806217e-06,
+      "loss": 0.9358,
+      "step": 15930
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.255859375,
+      "learning_rate": 4.437941464202177e-06,
+      "loss": 0.933,
+      "step": 15935
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.26171875,
+      "learning_rate": 4.408486236664411e-06,
+      "loss": 0.9991,
+      "step": 15940
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.251953125,
+      "learning_rate": 4.379126880734463e-06,
+      "loss": 0.8015,
+      "step": 15945
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 4.349863425857681e-06,
+      "loss": 0.8728,
+      "step": 15950
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.248046875,
+      "learning_rate": 4.320695901383298e-06,
+      "loss": 0.9851,
+      "step": 15955
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 4.291624336564304e-06,
+      "loss": 0.919,
+      "step": 15960
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.255859375,
+      "learning_rate": 4.26264876055743e-06,
+      "loss": 0.9507,
+      "step": 15965
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.248046875,
+      "learning_rate": 4.233769202423166e-06,
+      "loss": 0.9137,
+      "step": 15970
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.25390625,
+      "learning_rate": 4.204985691125674e-06,
+      "loss": 0.9431,
+      "step": 15975
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.24609375,
+      "learning_rate": 4.176298255532829e-06,
+      "loss": 0.9374,
+      "step": 15980
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.263671875,
+      "learning_rate": 4.147706924416095e-06,
+      "loss": 0.9224,
+      "step": 15985
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.26953125,
+      "learning_rate": 4.119211726450611e-06,
+      "loss": 0.9349,
+      "step": 15990
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.0908126902150404e-06,
+      "loss": 0.995,
+      "step": 15995
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 4.062509844191631e-06,
+      "loss": 0.9509,
+      "step": 16000
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.26953125,
+      "learning_rate": 4.0343032167662e-06,
+      "loss": 0.9903,
+      "step": 16005
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.29296875,
+      "learning_rate": 4.0061928362279846e-06,
+      "loss": 0.9883,
+      "step": 16010
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 3.978178730769733e-06,
+      "loss": 0.9506,
+      "step": 16015
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.25390625,
+      "learning_rate": 3.95026092848767e-06,
+      "loss": 0.8995,
+      "step": 16020
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.25,
+      "learning_rate": 3.922439457381355e-06,
+      "loss": 0.9902,
+      "step": 16025
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.894714345353801e-06,
+      "loss": 0.8852,
+      "step": 16030
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.265625,
+      "learning_rate": 3.867085620211352e-06,
+      "loss": 0.991,
+      "step": 16035
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.8395533096637015e-06,
+      "loss": 0.9524,
+      "step": 16040
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.8121174413238257e-06,
+      "loss": 0.8911,
+      "step": 16045
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.7847780427079814e-06,
+      "loss": 0.953,
+      "step": 16050
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.7575351412356576e-06,
+      "loss": 0.9258,
+      "step": 16055
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.263671875,
+      "learning_rate": 3.73038876422962e-06,
+      "loss": 0.935,
+      "step": 16060
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.267578125,
+      "learning_rate": 3.7033389389157567e-06,
+      "loss": 0.9546,
+      "step": 16065
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.26171875,
+      "learning_rate": 3.6763856924231678e-06,
+      "loss": 0.9483,
+      "step": 16070
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.649529051784051e-06,
+      "loss": 0.9712,
+      "step": 16075
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.25390625,
+      "learning_rate": 3.62276904393376e-06,
+      "loss": 0.9029,
+      "step": 16080
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.25390625,
+      "learning_rate": 3.5961056957107273e-06,
+      "loss": 0.9028,
+      "step": 16085
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.569539033856406e-06,
+      "loss": 0.9829,
+      "step": 16090
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.5430690850153046e-06,
+      "loss": 1.0096,
+      "step": 16095
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.516695875734932e-06,
+      "loss": 0.8646,
+      "step": 16100
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.271484375,
+      "learning_rate": 3.4904194324657748e-06,
+      "loss": 0.9332,
+      "step": 16105
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.4642397815612847e-06,
+      "loss": 0.9654,
+      "step": 16110
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.27734375,
+      "learning_rate": 3.438156949277782e-06,
+      "loss": 0.9824,
+      "step": 16115
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.26171875,
+      "learning_rate": 3.4121709617745745e-06,
+      "loss": 0.8956,
+      "step": 16120
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.275390625,
+      "learning_rate": 3.386281845113748e-06,
+      "loss": 0.9375,
+      "step": 16125
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 3.3604896252603104e-06,
+      "loss": 0.9401,
+      "step": 16130
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.287109375,
+      "learning_rate": 3.334794328082025e-06,
+      "loss": 0.9871,
+      "step": 16135
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.265625,
+      "learning_rate": 3.3091959793495107e-06,
+      "loss": 0.9283,
+      "step": 16140
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.263671875,
+      "learning_rate": 3.2836946047360982e-06,
+      "loss": 0.9401,
+      "step": 16145
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.24609375,
+      "learning_rate": 3.2582902298178953e-06,
+      "loss": 0.935,
+      "step": 16150
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.263671875,
+      "learning_rate": 3.2329828800737096e-06,
+      "loss": 0.9749,
+      "step": 16155
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.26953125,
+      "learning_rate": 3.207772580885049e-06,
+      "loss": 0.9388,
+      "step": 16160
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.25390625,
+      "learning_rate": 3.1826593575360887e-06,
+      "loss": 0.9138,
+      "step": 16165
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.25390625,
+      "learning_rate": 3.1576432352136144e-06,
+      "loss": 0.9225,
+      "step": 16170
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.27734375,
+      "learning_rate": 3.1327242390070677e-06,
+      "loss": 0.9162,
+      "step": 16175
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.265625,
+      "learning_rate": 3.1079023939084573e-06,
+      "loss": 0.899,
+      "step": 16180
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.0831777248123584e-06,
+      "loss": 0.9189,
+      "step": 16185
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.25390625,
+      "learning_rate": 3.0585502565158687e-06,
+      "loss": 0.8877,
+      "step": 16190
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.0340200137186526e-06,
+      "loss": 0.8927,
+      "step": 16195
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.0095870210228083e-06,
+      "loss": 0.9733,
+      "step": 16200
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.248046875,
+      "learning_rate": 2.985251302932912e-06,
+      "loss": 0.9776,
+      "step": 16205
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.9610128838560068e-06,
+      "loss": 0.8292,
+      "step": 16210
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.936871788101514e-06,
+      "loss": 0.9053,
+      "step": 16215
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.912828039881288e-06,
+      "loss": 0.8882,
+      "step": 16220
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.8888816633095063e-06,
+      "loss": 0.9554,
+      "step": 16225
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.265625,
+      "learning_rate": 2.865032682402724e-06,
+      "loss": 0.9225,
+      "step": 16230
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.8412811210797975e-06,
+      "loss": 0.9057,
+      "step": 16235
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2734375,
+      "learning_rate": 2.817627003161882e-06,
+      "loss": 0.9084,
+      "step": 16240
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2890625,
+      "learning_rate": 2.7940703523724134e-06,
+      "loss": 0.8942,
+      "step": 16245
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.770611192337047e-06,
+      "loss": 0.9854,
+      "step": 16250
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.265625,
+      "learning_rate": 2.747249546583708e-06,
+      "loss": 0.9788,
+      "step": 16255
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.7239854385424645e-06,
+      "loss": 0.848,
+      "step": 16260
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.275390625,
+      "learning_rate": 2.70081889154562e-06,
+      "loss": 0.9951,
+      "step": 16265
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.6777499288275775e-06,
+      "loss": 1.004,
+      "step": 16270
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.6547785735249187e-06,
+      "loss": 0.9352,
+      "step": 16275
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.265625,
+      "learning_rate": 2.631904848676281e-06,
+      "loss": 0.9755,
+      "step": 16280
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.271484375,
+      "learning_rate": 2.609128777222447e-06,
+      "loss": 1.0406,
+      "step": 16285
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.586450382006178e-06,
+      "loss": 0.89,
+      "step": 16290
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.279296875,
+      "learning_rate": 2.563869685772358e-06,
+      "loss": 0.9417,
+      "step": 16295
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.24609375,
+      "learning_rate": 2.5413867111678367e-06,
+      "loss": 0.978,
+      "step": 16300
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.5190014807414454e-06,
+      "loss": 0.9096,
+      "step": 16305
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.24609375,
+      "learning_rate": 2.4967140169440464e-06,
+      "loss": 0.9473,
+      "step": 16310
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.4745243421283706e-06,
+      "loss": 0.9828,
+      "step": 16315
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.265625,
+      "learning_rate": 2.452432478549116e-06,
+      "loss": 0.9684,
+      "step": 16320
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.430438448362915e-06,
+      "loss": 0.9518,
+      "step": 16325
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.298828125,
+      "learning_rate": 2.408542273628189e-06,
+      "loss": 0.9734,
+      "step": 16330
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.3867439763053166e-06,
+      "loss": 0.8628,
+      "step": 16335
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.271484375,
+      "learning_rate": 2.3650435782564427e-06,
+      "loss": 1.007,
+      "step": 16340
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.25,
+      "learning_rate": 2.343441101245558e-06,
+      "loss": 0.9013,
+      "step": 16345
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.265625,
+      "learning_rate": 2.3219365669384206e-06,
+      "loss": 0.9517,
+      "step": 16350
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.3005299969026117e-06,
+      "loss": 1.0107,
+      "step": 16355
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 2.2792214126073797e-06,
+      "loss": 0.9161,
+      "step": 16360
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 2.258010835423774e-06,
+      "loss": 0.9082,
+      "step": 16365
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.2368982866245225e-06,
+      "loss": 0.8977,
+      "step": 16370
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.215883787384032e-06,
+      "loss": 0.8715,
+      "step": 16375
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.194967358778366e-06,
+      "loss": 0.9551,
+      "step": 16380
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.1741490217852545e-06,
+      "loss": 0.9018,
+      "step": 16385
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.1534287972840294e-06,
+      "loss": 0.9079,
+      "step": 16390
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.1328067060556235e-06,
+      "loss": 0.8878,
+      "step": 16395
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.1122827687825473e-06,
+      "loss": 0.8754,
+      "step": 16400
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.271484375,
+      "learning_rate": 2.091857006048903e-06,
+      "loss": 0.9388,
+      "step": 16405
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.0715294383402695e-06,
+      "loss": 0.955,
+      "step": 16410
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28125,
+      "learning_rate": 2.051300086043806e-06,
+      "loss": 0.9801,
+      "step": 16415
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.031168969448116e-06,
+      "loss": 0.9802,
+      "step": 16420
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.0111361087433165e-06,
+      "loss": 0.938,
+      "step": 16425
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.9912015240209583e-06,
+      "loss": 0.9356,
+      "step": 16430
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.9713652352740364e-06,
+      "loss": 0.9112,
+      "step": 16435
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.951627262396971e-06,
+      "loss": 0.9399,
+      "step": 16440
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.9319876251855606e-06,
+      "loss": 0.9381,
+      "step": 16445
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9124463433370045e-06,
+      "loss": 0.9783,
+      "step": 16450
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.893003436449814e-06,
+      "loss": 0.9608,
+      "step": 16455
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.8736589240239022e-06,
+      "loss": 0.955,
+      "step": 16460
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.8544128254604277e-06,
+      "loss": 0.9459,
+      "step": 16465
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.8352651600619165e-06,
+      "loss": 0.9227,
+      "step": 16470
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.8162159470321072e-06,
+      "loss": 0.9127,
+      "step": 16475
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.7972652054760508e-06,
+      "loss": 0.9188,
+      "step": 16480
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.7784129544000106e-06,
+      "loss": 0.8966,
+      "step": 16485
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.7596592127114619e-06,
+      "loss": 0.9293,
+      "step": 16490
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.741003999219115e-06,
+      "loss": 0.989,
+      "step": 16495
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.7224473326328484e-06,
+      "loss": 0.9269,
+      "step": 16500
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.703989231563685e-06,
+      "loss": 0.9396,
+      "step": 16505
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.6856297145238177e-06,
+      "loss": 0.9334,
+      "step": 16510
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.667368799926572e-06,
+      "loss": 0.8955,
+      "step": 16515
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.6492065060863536e-06,
+      "loss": 0.9297,
+      "step": 16520
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.6311428512186699e-06,
+      "loss": 0.9916,
+      "step": 16525
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.6131778534401176e-06,
+      "loss": 0.9735,
+      "step": 16530
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.5953115307683287e-06,
+      "loss": 0.9515,
+      "step": 16535
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.5775439011219806e-06,
+      "loss": 0.9161,
+      "step": 16540
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.5598749823207525e-06,
+      "loss": 0.901,
+      "step": 16545
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.5423047920853473e-06,
+      "loss": 0.9621,
+      "step": 16550
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 1.5248333480374355e-06,
+      "loss": 0.9058,
+      "step": 16555
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.5074606676996561e-06,
+      "loss": 0.9201,
+      "step": 16560
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.4901867684955829e-06,
+      "loss": 0.9491,
+      "step": 16565
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.4730116677497351e-06,
+      "loss": 0.9164,
+      "step": 16570
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 1.4559353826875344e-06,
+      "loss": 0.9667,
+      "step": 16575
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.24609375,
+      "learning_rate": 1.438957930435314e-06,
+      "loss": 0.8973,
+      "step": 16580
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.422079328020265e-06,
+      "loss": 0.9339,
+      "step": 16585
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.405299592370435e-06,
+      "loss": 0.9648,
+      "step": 16590
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.388618740314751e-06,
+      "loss": 0.9488,
+      "step": 16595
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 1.37203678858292e-06,
+      "loss": 0.9638,
+      "step": 16600
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.355553753805483e-06,
+      "loss": 1.001,
+      "step": 16605
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.3391696525137831e-06,
+      "loss": 0.9786,
+      "step": 16610
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.3228845011399204e-06,
+      "loss": 0.9358,
+      "step": 16615
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.3066983160167746e-06,
+      "loss": 1.0021,
+      "step": 16620
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.2906111133779376e-06,
+      "loss": 0.9945,
+      "step": 16625
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.27462290935777e-06,
+      "loss": 0.928,
+      "step": 16630
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.2587337199913118e-06,
+      "loss": 0.843,
+      "step": 16635
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.2429435612143158e-06,
+      "loss": 0.9762,
+      "step": 16640
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.227252448863192e-06,
+      "loss": 0.9254,
+      "step": 16645
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.211660398675052e-06,
+      "loss": 1.0486,
+      "step": 16650
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.1961674262876199e-06,
+      "loss": 0.9478,
+      "step": 16655
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.1807735472392778e-06,
+      "loss": 0.916,
+      "step": 16660
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.1654787769689868e-06,
+      "loss": 0.9788,
+      "step": 16665
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.150283130816343e-06,
+      "loss": 0.9357,
+      "step": 16670
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.1351866240215336e-06,
+      "loss": 0.9498,
+      "step": 16675
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.1201892717252692e-06,
+      "loss": 0.9669,
+      "step": 16680
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.1052910889688629e-06,
+      "loss": 0.9039,
+      "step": 16685
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.0904920906941618e-06,
+      "loss": 0.964,
+      "step": 16690
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.075792291743516e-06,
+      "loss": 0.8932,
+      "step": 16695
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.0611917068597877e-06,
+      "loss": 0.9485,
+      "step": 16700
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.0466903506863523e-06,
+      "loss": 0.9478,
+      "step": 16705
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.032288237767065e-06,
+      "loss": 0.9896,
+      "step": 16710
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.0179853825462271e-06,
+      "loss": 0.9185,
+      "step": 16715
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.0037817993686084e-06,
+      "loss": 0.9289,
+      "step": 16720
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.896775024793914e-07,
+      "loss": 0.9296,
+      "step": 16725
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 9.756725060242277e-07,
+      "loss": 0.9143,
+      "step": 16730
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.617668240491372e-07,
+      "loss": 1.0418,
+      "step": 16735
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.479604705005529e-07,
+      "loss": 0.9322,
+      "step": 16740
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.342534592252761e-07,
+      "loss": 0.9627,
+      "step": 16745
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.206458039704768e-07,
+      "loss": 0.9865,
+      "step": 16750
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.240234375,
+      "learning_rate": 9.07137518383705e-07,
+      "loss": 0.9431,
+      "step": 16755
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.9372861601279e-07,
+      "loss": 0.9944,
+      "step": 16760
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.804191103059523e-07,
+      "loss": 0.9314,
+      "step": 16765
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.263671875,
+      "learning_rate": 8.672090146116917e-07,
+      "loss": 0.9187,
+      "step": 16770
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.540983421787996e-07,
+      "loss": 0.976,
+      "step": 16775
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 8.410871061563797e-07,
+      "loss": 0.9552,
+      "step": 16780
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.240234375,
+      "learning_rate": 8.281753195937714e-07,
+      "loss": 0.9202,
+      "step": 16785
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.259765625,
+      "learning_rate": 8.153629954406161e-07,
+      "loss": 0.8923,
+      "step": 16790
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.244140625,
+      "learning_rate": 8.026501465467684e-07,
+      "loss": 1.0348,
+      "step": 16795
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.900367856623403e-07,
+      "loss": 0.8979,
+      "step": 16800
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.248046875,
+      "learning_rate": 7.775229254376348e-07,
+      "loss": 0.8703,
+      "step": 16805
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.25,
+      "learning_rate": 7.651085784231793e-07,
+      "loss": 0.9397,
+      "step": 16810
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.52793757069703e-07,
+      "loss": 0.9422,
+      "step": 16815
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.27734375,
+      "learning_rate": 7.405784737281151e-07,
+      "loss": 0.991,
+      "step": 16820
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.25,
+      "learning_rate": 7.284627406494826e-07,
+      "loss": 0.9515,
+      "step": 16825
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.25390625,
+      "learning_rate": 7.164465699850409e-07,
+      "loss": 0.9394,
+      "step": 16830
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.26953125,
+      "learning_rate": 7.045299737861832e-07,
+      "loss": 0.9742,
+      "step": 16835
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.263671875,
+      "learning_rate": 6.92712964004405e-07,
+      "loss": 0.949,
+      "step": 16840
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2578125,
+      "learning_rate": 6.809955524913369e-07,
+      "loss": 0.9727,
+      "step": 16845
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2578125,
+      "learning_rate": 6.693777509987453e-07,
+      "loss": 0.9082,
+      "step": 16850
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.248046875,
+      "learning_rate": 6.578595711784541e-07,
+      "loss": 0.9473,
+      "step": 16855
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2421875,
+      "learning_rate": 6.464410245824004e-07,
+      "loss": 1.002,
+      "step": 16860
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.251953125,
+      "learning_rate": 6.351221226625903e-07,
+      "loss": 0.99,
+      "step": 16865
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.251953125,
+      "learning_rate": 6.239028767710986e-07,
+      "loss": 0.992,
+      "step": 16870
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.26953125,
+      "learning_rate": 6.127832981600246e-07,
+      "loss": 0.8812,
+      "step": 16875
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.25390625,
+      "learning_rate": 6.017633979815363e-07,
+      "loss": 0.9415,
+      "step": 16880
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.265625,
+      "learning_rate": 5.908431872878372e-07,
+      "loss": 0.911,
+      "step": 16885
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2578125,
+      "learning_rate": 5.800226770311113e-07,
+      "loss": 0.9454,
+      "step": 16890
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.25390625,
+      "learning_rate": 5.693018780635995e-07,
+      "loss": 1.0156,
+      "step": 16895
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.259765625,
+      "learning_rate": 5.58680801137501e-07,
+      "loss": 0.9276,
+      "step": 16900
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.25,
+      "learning_rate": 5.481594569050174e-07,
+      "loss": 0.8796,
+      "step": 16905
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2890625,
+      "learning_rate": 5.377378559183077e-07,
+      "loss": 0.9183,
+      "step": 16910
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.259765625,
+      "learning_rate": 5.274160086295332e-07,
+      "loss": 0.9421,
+      "step": 16915
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 5.171939253907687e-07,
+      "loss": 0.9335,
+      "step": 16920
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2578125,
+      "learning_rate": 5.070716164540579e-07,
+      "loss": 0.9423,
+      "step": 16925
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.267578125,
+      "learning_rate": 4.970490919713577e-07,
+      "loss": 0.9731,
+      "step": 16930
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.26171875,
+      "learning_rate": 4.871263619945721e-07,
+      "loss": 0.9762,
+      "step": 16935
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 4.773034364754958e-07,
+      "loss": 0.9654,
+      "step": 16940
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.26171875,
+      "learning_rate": 4.675803252658484e-07,
+      "loss": 0.9329,
+      "step": 16945
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.259765625,
+      "learning_rate": 4.5795703811721825e-07,
+      "loss": 0.9249,
+      "step": 16950
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 4.484335846810961e-07,
+      "loss": 0.8427,
+      "step": 16955
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.251953125,
+      "learning_rate": 4.3900997450885274e-07,
+      "loss": 0.8893,
+      "step": 16960
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 4.2968621705168354e-07,
+      "loss": 0.8505,
+      "step": 16965
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.27734375,
+      "learning_rate": 4.204623216606751e-07,
+      "loss": 0.9157,
+      "step": 16970
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.248046875,
+      "learning_rate": 4.113382975867608e-07,
+      "loss": 0.9722,
+      "step": 16975
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2578125,
+      "learning_rate": 4.023141539806985e-07,
+      "loss": 0.9496,
+      "step": 16980
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.279296875,
+      "learning_rate": 3.9338989989307073e-07,
+      "loss": 0.9032,
+      "step": 16985
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.845655442742624e-07,
+      "loss": 0.9208,
+      "step": 16990
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.263671875,
+      "learning_rate": 3.7584109597451623e-07,
+      "loss": 0.9329,
+      "step": 16995
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.283203125,
+      "learning_rate": 3.672165637438218e-07,
+      "loss": 0.9515,
+      "step": 17000
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.28125,
+      "learning_rate": 3.586919562319935e-07,
+      "loss": 0.9471,
+      "step": 17005
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.26953125,
+      "learning_rate": 3.5026728198860324e-07,
+      "loss": 0.9335,
+      "step": 17010
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 3.4194254946302573e-07,
+      "loss": 0.9268,
+      "step": 17015
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.23828125,
+      "learning_rate": 3.337177670043823e-07,
+      "loss": 0.9364,
+      "step": 17020
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.255929428615523e-07,
+      "loss": 0.8839,
+      "step": 17025
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2734375,
+      "learning_rate": 3.175680851831619e-07,
+      "loss": 0.9537,
+      "step": 17030
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.24609375,
+      "learning_rate": 3.0964320201759545e-07,
+      "loss": 0.9085,
+      "step": 17035
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2578125,
+      "learning_rate": 3.0181830131295053e-07,
+      "loss": 0.9741,
+      "step": 17040
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 2.9409339091703844e-07,
+      "loss": 0.9369,
+      "step": 17045
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.8646847857742854e-07,
+      "loss": 0.9125,
+      "step": 17050
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.789435719413813e-07,
+      "loss": 0.9309,
+      "step": 17055
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.283203125,
+      "learning_rate": 2.7151867855581546e-07,
+      "loss": 0.9311,
+      "step": 17060
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.283203125,
+      "learning_rate": 2.641938058674187e-07,
+      "loss": 0.9312,
+      "step": 17065
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.569689612225035e-07,
+      "loss": 0.9185,
+      "step": 17070
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.279296875,
+      "learning_rate": 2.4984415186709576e-07,
+      "loss": 0.9021,
+      "step": 17075
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.4281938494686853e-07,
+      "loss": 0.9878,
+      "step": 17080
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.3589466750718604e-07,
+      "loss": 0.935,
+      "step": 17085
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.2907000649304845e-07,
+      "loss": 0.9045,
+      "step": 17090
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.27734375,
+      "learning_rate": 2.2234540874911392e-07,
+      "loss": 0.9137,
+      "step": 17095
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.1572088101968758e-07,
+      "loss": 1.008,
+      "step": 17100
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.0919642994869925e-07,
+      "loss": 0.8877,
+      "step": 17105
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.0277206207972576e-07,
+      "loss": 0.9412,
+      "step": 17110
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9644778385596864e-07,
+      "loss": 0.9598,
+      "step": 17115
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.902236016202208e-07,
+      "loss": 0.9411,
+      "step": 17120
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.8409952161489997e-07,
+      "loss": 0.9106,
+      "step": 17125
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.7807554998203747e-07,
+      "loss": 0.8809,
+      "step": 17130
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 1.7215169276325605e-07,
+      "loss": 0.9383,
+      "step": 17135
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.663279558997699e-07,
+      "loss": 0.9645,
+      "step": 17140
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.6060434523238466e-07,
+      "loss": 0.8795,
+      "step": 17145
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 1.5498086650147513e-07,
+      "loss": 0.9248,
+      "step": 17150
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.4945752534699653e-07,
+      "loss": 0.9705,
+      "step": 17155
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.4403432730847323e-07,
+      "loss": 0.9542,
+      "step": 17160
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.3871127782500993e-07,
+      "loss": 0.9194,
+      "step": 17165
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.3348838223523618e-07,
+      "loss": 0.8496,
+      "step": 17170
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.2836564577735078e-07,
+      "loss": 0.8582,
+      "step": 17175
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.234375,
+      "learning_rate": 1.2334307358911056e-07,
+      "loss": 0.9176,
+      "step": 17180
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.1842067070779728e-07,
+      "loss": 0.9633,
+      "step": 17185
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.24609375,
+      "learning_rate": 1.1359844207023962e-07,
+      "loss": 0.8306,
+      "step": 17190
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.0887639251280224e-07,
+      "loss": 0.947,
+      "step": 17195
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.0425452677135238e-07,
+      "loss": 0.9984,
+      "step": 17200
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.973284948132656e-08,
+      "loss": 0.9801,
+      "step": 17205
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 9.531136517761941e-08,
+      "loss": 0.9025,
+      "step": 17210
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.099007829469486e-08,
+      "loss": 0.8792,
+      "step": 17215
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.267578125,
+      "learning_rate": 8.676899316648613e-08,
+      "loss": 0.9554,
+      "step": 17220
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.263671875,
+      "learning_rate": 8.264811402646233e-08,
+      "loss": 1.0131,
+      "step": 17225
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.263671875,
+      "learning_rate": 7.862744500756192e-08,
+      "loss": 0.9268,
+      "step": 17230
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.259765625,
+      "learning_rate": 7.470699014223703e-08,
+      "loss": 0.9519,
+      "step": 17235
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.265625,
+      "learning_rate": 7.088675336244244e-08,
+      "loss": 0.9131,
+      "step": 17240
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.265625,
+      "learning_rate": 6.71667384995911e-08,
+      "loss": 0.9502,
+      "step": 17245
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.25390625,
+      "learning_rate": 6.35469492846208e-08,
+      "loss": 0.9628,
+      "step": 17250
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.255859375,
+      "learning_rate": 6.00273893479053e-08,
+      "loss": 0.9062,
+      "step": 17255
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.28515625,
+      "learning_rate": 5.660806221932102e-08,
+      "loss": 0.9433,
+      "step": 17260
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.3288971328224747e-08,
+      "loss": 0.9557,
+      "step": 17265
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.259765625,
+      "learning_rate": 5.0070120003420375e-08,
+      "loss": 0.9473,
+      "step": 17270
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2578125,
+      "learning_rate": 4.6951511473203316e-08,
+      "loss": 0.9675,
+      "step": 17275
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.259765625,
+      "learning_rate": 4.3933148865316075e-08,
+      "loss": 0.9835,
+      "step": 17280
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.25,
+      "learning_rate": 4.101503520695937e-08,
+      "loss": 0.9784,
+      "step": 17285
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.26171875,
+      "learning_rate": 3.819717342480322e-08,
+      "loss": 0.9332,
+      "step": 17290
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.24609375,
+      "learning_rate": 3.547956634495364e-08,
+      "loss": 0.976,
+      "step": 17295
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.244140625,
+      "learning_rate": 3.286221669299705e-08,
+      "loss": 0.9327,
+      "step": 17300
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2578125,
+      "learning_rate": 3.0345127093955875e-08,
+      "loss": 0.8962,
+      "step": 17305
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 2.7928300072277424e-08,
+      "loss": 0.8845,
+      "step": 17310
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.28125,
+      "learning_rate": 2.561173805186723e-08,
+      "loss": 0.9619,
+      "step": 17315
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.339544335610011e-08,
+      "loss": 0.9725,
+      "step": 17320
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.1279418207742486e-08,
+      "loss": 0.9017,
+      "step": 17325
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.9263664729030073e-08,
+      "loss": 0.8888,
+      "step": 17330
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.7348184941623492e-08,
+      "loss": 0.9509,
+      "step": 17335
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.5532980766608252e-08,
+      "loss": 1.0119,
+      "step": 17340
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.3818054024516969e-08,
+      "loss": 0.9384,
+      "step": 17345
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.2203406435284948e-08,
+      "loss": 0.9307,
+      "step": 17350
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.244140625,
+      "learning_rate": 1.0689039618305696e-08,
+      "loss": 0.9041,
+      "step": 17355
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.274955092386516e-09,
+      "loss": 0.9935,
+      "step": 17360
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.255859375,
+      "learning_rate": 7.961154275737403e-09,
+      "loss": 0.9374,
+      "step": 17365
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 6.747638486026553e-09,
+      "loss": 0.91,
+      "step": 17370
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.287109375,
+      "learning_rate": 5.634408940313751e-09,
+      "loss": 0.997,
+      "step": 17375
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.28515625,
+      "learning_rate": 4.621466755094784e-09,
+      "loss": 0.9748,
+      "step": 17380
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.7088129462792277e-09,
+      "loss": 0.9361,
+      "step": 17385
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.896448429201559e-09,
+      "loss": 0.9747,
+      "step": 17390
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2890625,
+      "learning_rate": 2.1843740185878423e-09,
+      "loss": 0.9583,
+      "step": 17395
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.5725904286223446e-09,
+      "loss": 0.9044,
+      "step": 17400
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.0610982728698203e-09,
+      "loss": 0.9078,
+      "step": 17405
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.26953125,
+      "learning_rate": 6.498980643199204e-10,
+      "loss": 0.9411,
+      "step": 17410
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.263671875,
+      "learning_rate": 3.389902153760893e-10,
+      "loss": 0.9685,
+      "step": 17415
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.2837503786666815e-10,
+      "loss": 0.937,
+      "step": 17420
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.8052743022689556e-11,
+      "loss": 0.9411,
+      "step": 17425
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.9471394419670105,
+      "eval_runtime": 7952.5701,
+      "eval_samples_per_second": 1.94,
+      "eval_steps_per_second": 0.243,
+      "step": 17428
+    },
+    {
+      "epoch": 1.0,
+      "step": 17428,
+      "total_flos": 1.2254141370633028e+19,
+      "train_loss": 0.9532137394142939,
+      "train_runtime": 251062.1903,
+      "train_samples_per_second": 0.555,
+      "train_steps_per_second": 0.069
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 17428,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "total_flos": 1.2254141370633028e+19,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}