diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,70045 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 50000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 2e-05,
+      "grad_norm": 10.4375,
+      "learning_rate": 4e-08,
+      "loss": 7.7211,
+      "step": 1
+    },
+    {
+      "epoch": 0.0001,
+      "grad_norm": 11.125,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 7.5988,
+      "step": 5
+    },
+    {
+      "epoch": 0.0002,
+      "grad_norm": 11.25,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 7.6992,
+      "step": 10
+    },
+    {
+      "epoch": 0.0003,
+      "grad_norm": 10.9375,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 7.6052,
+      "step": 15
+    },
+    {
+      "epoch": 0.0004,
+      "grad_norm": 11.0625,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 7.6137,
+      "step": 20
+    },
+    {
+      "epoch": 0.0005,
+      "grad_norm": 10.875,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 7.5752,
+      "step": 25
+    },
+    {
+      "epoch": 0.0006,
+      "grad_norm": 11.0,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 7.6467,
+      "step": 30
+    },
+    {
+      "epoch": 0.0007,
+      "grad_norm": 11.125,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 7.5769,
+      "step": 35
+    },
+    {
+      "epoch": 0.0008,
+      "grad_norm": 10.9375,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 7.6557,
+      "step": 40
+    },
+    {
+      "epoch": 0.0009,
+      "grad_norm": 10.75,
+      "learning_rate": 1.8e-06,
+      "loss": 7.6391,
+      "step": 45
+    },
+    {
+      "epoch": 0.001,
+      "grad_norm": 11.25,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 7.5252,
+      "step": 50
+    },
+    {
+      "epoch": 0.0011,
+      "grad_norm": 10.8125,
+      "learning_rate": 2.2e-06,
+      "loss": 7.5162,
+      "step": 55
+    },
+    {
+      "epoch": 0.0012,
+      "grad_norm": 10.9375,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 7.457,
+      "step": 60
+    },
+    {
+      "epoch": 0.0013,
+      "grad_norm": 10.3125,
+      "learning_rate": 2.6e-06,
+      "loss": 7.3021,
+      "step": 65
+    },
+    {
+      "epoch": 0.0014,
+      "grad_norm": 10.875,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 7.2525,
+      "step": 70
+    },
+    {
+      "epoch": 0.0015,
+      "grad_norm": 11.8125,
+      "learning_rate": 3e-06,
+      "loss": 7.1694,
+      "step": 75
+    },
+    {
+      "epoch": 0.0016,
+      "grad_norm": 11.375,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 7.0152,
+      "step": 80
+    },
+    {
+      "epoch": 0.0017,
+      "grad_norm": 11.75,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 6.9306,
+      "step": 85
+    },
+    {
+      "epoch": 0.0018,
+      "grad_norm": 12.1875,
+      "learning_rate": 3.6e-06,
+      "loss": 6.8075,
+      "step": 90
+    },
+    {
+      "epoch": 0.0019,
+      "grad_norm": 11.5625,
+      "learning_rate": 3.8e-06,
+      "loss": 6.5017,
+      "step": 95
+    },
+    {
+      "epoch": 0.002,
+      "grad_norm": 12.25,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 6.3431,
+      "step": 100
+    },
+    {
+      "epoch": 0.0021,
+      "grad_norm": 11.8125,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 6.1335,
+      "step": 105
+    },
+    {
+      "epoch": 0.0022,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.4e-06,
+      "loss": 5.9512,
+      "step": 110
+    },
+    {
+      "epoch": 0.0023,
+      "grad_norm": 6.9375,
+      "learning_rate": 4.6e-06,
+      "loss": 5.7415,
+      "step": 115
+    },
+    {
+      "epoch": 0.0024,
+      "grad_norm": 6.59375,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 5.5713,
+      "step": 120
+    },
+    {
+      "epoch": 0.0025,
+      "grad_norm": 6.28125,
+      "learning_rate": 5e-06,
+      "loss": 5.4064,
+      "step": 125
+    },
+    {
+      "epoch": 0.0026,
+      "grad_norm": 6.65625,
+      "learning_rate": 5.2e-06,
+      "loss": 5.2156,
+      "step": 130
+    },
+    {
+      "epoch": 0.0027,
+      "grad_norm": 7.125,
+      "learning_rate": 5.4e-06,
+      "loss": 5.0183,
+      "step": 135
+    },
+    {
+      "epoch": 0.0028,
+      "grad_norm": 7.03125,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 4.7652,
+      "step": 140
+    },
+    {
+      "epoch": 0.0029,
+      "grad_norm": 7.6875,
+      "learning_rate": 5.8e-06,
+      "loss": 4.4785,
+      "step": 145
+    },
+    {
+      "epoch": 0.003,
+      "grad_norm": 7.9375,
+      "learning_rate": 6e-06,
+      "loss": 4.1766,
+      "step": 150
+    },
+    {
+      "epoch": 0.0031,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.2e-06,
+      "loss": 3.893,
+      "step": 155
+    },
+    {
+      "epoch": 0.0032,
+      "grad_norm": 8.25,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 3.5528,
+      "step": 160
+    },
+    {
+      "epoch": 0.0033,
+      "grad_norm": 7.71875,
+      "learning_rate": 6.6e-06,
+      "loss": 3.2487,
+      "step": 165
+    },
+    {
+      "epoch": 0.0034,
+      "grad_norm": 5.375,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 2.9867,
+      "step": 170
+    },
+    {
+      "epoch": 0.0035,
+      "grad_norm": 5.125,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 2.9225,
+      "step": 175
+    },
+    {
+      "epoch": 0.0036,
+      "grad_norm": 5.625,
+      "learning_rate": 7.2e-06,
+      "loss": 2.8023,
+      "step": 180
+    },
+    {
+      "epoch": 0.0037,
+      "grad_norm": 5.75,
+      "learning_rate": 7.4e-06,
+      "loss": 2.7618,
+      "step": 185
+    },
+    {
+      "epoch": 0.0038,
+      "grad_norm": 6.25,
+      "learning_rate": 7.6e-06,
+      "loss": 2.7105,
+      "step": 190
+    },
+    {
+      "epoch": 0.0039,
+      "grad_norm": 6.46875,
+      "learning_rate": 7.8e-06,
+      "loss": 2.6358,
+      "step": 195
+    },
+    {
+      "epoch": 0.004,
+      "grad_norm": 7.21875,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 2.6077,
+      "step": 200
+    },
+    {
+      "epoch": 0.0041,
+      "grad_norm": 7.90625,
+      "learning_rate": 8.200000000000001e-06,
+      "loss": 2.6066,
+      "step": 205
+    },
+    {
+      "epoch": 0.0042,
+      "grad_norm": 7.28125,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 2.5544,
+      "step": 210
+    },
+    {
+      "epoch": 0.0043,
+      "grad_norm": 7.90625,
+      "learning_rate": 8.599999999999999e-06,
+      "loss": 2.5405,
+      "step": 215
+    },
+    {
+      "epoch": 0.0044,
+      "grad_norm": 7.75,
+      "learning_rate": 8.8e-06,
+      "loss": 2.5352,
+      "step": 220
+    },
+    {
+      "epoch": 0.0045,
+      "grad_norm": 8.0625,
+      "learning_rate": 9e-06,
+      "loss": 2.5186,
+      "step": 225
+    },
+    {
+      "epoch": 0.0046,
+      "grad_norm": 9.0625,
+      "learning_rate": 9.2e-06,
+      "loss": 2.4445,
+      "step": 230
+    },
+    {
+      "epoch": 0.0047,
+      "grad_norm": 8.75,
+      "learning_rate": 9.4e-06,
+      "loss": 2.4411,
+      "step": 235
+    },
+    {
+      "epoch": 0.0048,
+      "grad_norm": 8.9375,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 2.4281,
+      "step": 240
+    },
+    {
+      "epoch": 0.0049,
+      "grad_norm": 9.25,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 2.3953,
+      "step": 245
+    },
+    {
+      "epoch": 0.005,
+      "grad_norm": 9.875,
+      "learning_rate": 1e-05,
+      "loss": 2.3903,
+      "step": 250
+    },
+    {
+      "epoch": 0.0051,
+      "grad_norm": 10.0,
+      "learning_rate": 1.02e-05,
+      "loss": 2.3481,
+      "step": 255
+    },
+    {
+      "epoch": 0.0052,
+      "grad_norm": 10.5625,
+      "learning_rate": 1.04e-05,
+      "loss": 2.3561,
+      "step": 260
+    },
+    {
+      "epoch": 0.0053,
+      "grad_norm": 10.625,
+      "learning_rate": 1.06e-05,
+      "loss": 2.3282,
+      "step": 265
+    },
+    {
+      "epoch": 0.0054,
+      "grad_norm": 10.4375,
+      "learning_rate": 1.08e-05,
+      "loss": 2.3154,
+      "step": 270
+    },
+    {
+      "epoch": 0.0055,
+      "grad_norm": 10.125,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 2.2352,
+      "step": 275
+    },
+    {
+      "epoch": 0.0056,
+      "grad_norm": 10.25,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 2.2167,
+      "step": 280
+    },
+    {
+      "epoch": 0.0057,
+      "grad_norm": 9.375,
+      "learning_rate": 1.1400000000000001e-05,
+      "loss": 2.2214,
+      "step": 285
+    },
+    {
+      "epoch": 0.0058,
+      "grad_norm": 10.8125,
+      "learning_rate": 1.16e-05,
+      "loss": 2.2491,
+      "step": 290
+    },
+    {
+      "epoch": 0.0059,
+      "grad_norm": 9.9375,
+      "learning_rate": 1.18e-05,
+      "loss": 2.1925,
+      "step": 295
+    },
+    {
+      "epoch": 0.006,
+      "grad_norm": 9.25,
+      "learning_rate": 1.2e-05,
+      "loss": 2.1493,
+      "step": 300
+    },
+    {
+      "epoch": 0.0061,
+      "grad_norm": 8.625,
+      "learning_rate": 1.22e-05,
+      "loss": 2.1356,
+      "step": 305
+    },
+    {
+      "epoch": 0.0062,
+      "grad_norm": 9.3125,
+      "learning_rate": 1.24e-05,
+      "loss": 2.1395,
+      "step": 310
+    },
+    {
+      "epoch": 0.0063,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.2600000000000001e-05,
+      "loss": 2.1276,
+      "step": 315
+    },
+    {
+      "epoch": 0.0064,
+      "grad_norm": 8.75,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 2.1206,
+      "step": 320
+    },
+    {
+      "epoch": 0.0065,
+      "grad_norm": 6.65625,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 2.0828,
+      "step": 325
+    },
+    {
+      "epoch": 0.0066,
+      "grad_norm": 8.125,
+      "learning_rate": 1.32e-05,
+      "loss": 2.0734,
+      "step": 330
+    },
+    {
+      "epoch": 0.0067,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.3400000000000002e-05,
+      "loss": 2.0281,
+      "step": 335
+    },
+    {
+      "epoch": 0.0068,
+      "grad_norm": 6.65625,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 2.029,
+      "step": 340
+    },
+    {
+      "epoch": 0.0069,
+      "grad_norm": 6.3125,
+      "learning_rate": 1.3800000000000002e-05,
+      "loss": 1.9719,
+      "step": 345
+    },
+    {
+      "epoch": 0.007,
+      "grad_norm": 5.96875,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 1.9584,
+      "step": 350
+    },
+    {
+      "epoch": 0.0071,
+      "grad_norm": 5.25,
+      "learning_rate": 1.42e-05,
+      "loss": 1.9177,
+      "step": 355
+    },
+    {
+      "epoch": 0.0072,
+      "grad_norm": 5.03125,
+      "learning_rate": 1.44e-05,
+      "loss": 1.939,
+      "step": 360
+    },
+    {
+      "epoch": 0.0073,
+      "grad_norm": 4.40625,
+      "learning_rate": 1.4599999999999999e-05,
+      "loss": 1.8974,
+      "step": 365
+    },
+    {
+      "epoch": 0.0074,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.48e-05,
+      "loss": 1.8671,
+      "step": 370
+    },
+    {
+      "epoch": 0.0075,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.5e-05,
+      "loss": 1.8443,
+      "step": 375
+    },
+    {
+      "epoch": 0.0076,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.52e-05,
+      "loss": 1.8036,
+      "step": 380
+    },
+    {
+      "epoch": 0.0077,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.54e-05,
+      "loss": 1.7554,
+      "step": 385
+    },
+    {
+      "epoch": 0.0078,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.56e-05,
+      "loss": 1.7402,
+      "step": 390
+    },
+    {
+      "epoch": 0.0079,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.58e-05,
+      "loss": 1.7023,
+      "step": 395
+    },
+    {
+      "epoch": 0.008,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.611,
+      "step": 400
+    },
+    {
+      "epoch": 0.0081,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.62e-05,
+      "loss": 1.6018,
+      "step": 405
+    },
+    {
+      "epoch": 0.0082,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.6400000000000002e-05,
+      "loss": 1.5639,
+      "step": 410
+    },
+    {
+      "epoch": 0.0083,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.66e-05,
+      "loss": 1.5868,
+      "step": 415
+    },
+    {
+      "epoch": 0.0084,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 1.5643,
+      "step": 420
+    },
+    {
+      "epoch": 0.0085,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 1.5462,
+      "step": 425
+    },
+    {
+      "epoch": 0.0086,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.7199999999999998e-05,
+      "loss": 1.5432,
+      "step": 430
+    },
+    {
+      "epoch": 0.0087,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.74e-05,
+      "loss": 1.534,
+      "step": 435
+    },
+    {
+      "epoch": 0.0088,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.76e-05,
+      "loss": 1.5213,
+      "step": 440
+    },
+    {
+      "epoch": 0.0089,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.78e-05,
+      "loss": 1.5282,
+      "step": 445
+    },
+    {
+      "epoch": 0.009,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.8e-05,
+      "loss": 1.5067,
+      "step": 450
+    },
+    {
+      "epoch": 0.0091,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.8200000000000002e-05,
+      "loss": 1.5178,
+      "step": 455
+    },
+    {
+      "epoch": 0.0092,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.84e-05,
+      "loss": 1.4998,
+      "step": 460
+    },
+    {
+      "epoch": 0.0093,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.86e-05,
+      "loss": 1.476,
+      "step": 465
+    },
+    {
+      "epoch": 0.0094,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.88e-05,
+      "loss": 1.4943,
+      "step": 470
+    },
+    {
+      "epoch": 0.0095,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9e-05,
+      "loss": 1.4803,
+      "step": 475
+    },
+    {
+      "epoch": 0.0096,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 1.4766,
+      "step": 480
+    },
+    {
+      "epoch": 0.0097,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.94e-05,
+      "loss": 1.4773,
+      "step": 485
+    },
+    {
+      "epoch": 0.0098,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 1.4634,
+      "step": 490
+    },
+    {
+      "epoch": 0.0099,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.9800000000000004e-05,
+      "loss": 1.4541,
+      "step": 495
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4140625,
+      "learning_rate": 2e-05,
+      "loss": 1.4519,
+      "step": 500
+    },
+    {
+      "epoch": 0.0101,
+      "grad_norm": 1.40625,
+      "learning_rate": 2.0200000000000003e-05,
+      "loss": 1.4589,
+      "step": 505
+    },
+    {
+      "epoch": 0.0102,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.04e-05,
+      "loss": 1.4407,
+      "step": 510
+    },
+    {
+      "epoch": 0.0103,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.06e-05,
+      "loss": 1.4394,
+      "step": 515
+    },
+    {
+      "epoch": 0.0104,
+      "grad_norm": 1.171875,
+      "learning_rate": 2.08e-05,
+      "loss": 1.4188,
+      "step": 520
+    },
+    {
+      "epoch": 0.0105,
+      "grad_norm": 1.5546875,
+      "learning_rate": 2.1e-05,
+      "loss": 1.4213,
+      "step": 525
+    },
+    {
+      "epoch": 0.0106,
+      "grad_norm": 1.1796875,
+      "learning_rate": 2.12e-05,
+      "loss": 1.4338,
+      "step": 530
+    },
+    {
+      "epoch": 0.0107,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.1400000000000002e-05,
+      "loss": 1.407,
+      "step": 535
+    },
+    {
+      "epoch": 0.0108,
+      "grad_norm": 1.25,
+      "learning_rate": 2.16e-05,
+      "loss": 1.422,
+      "step": 540
+    },
+    {
+      "epoch": 0.0109,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.18e-05,
+      "loss": 1.3851,
+      "step": 545
+    },
+    {
+      "epoch": 0.011,
+      "grad_norm": 1.3671875,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 1.3691,
+      "step": 550
+    },
+    {
+      "epoch": 0.0111,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.22e-05,
+      "loss": 1.3359,
+      "step": 555
+    },
+    {
+      "epoch": 0.0112,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.2400000000000002e-05,
+      "loss": 1.3413,
+      "step": 560
+    },
+    {
+      "epoch": 0.0113,
+      "grad_norm": 1.125,
+      "learning_rate": 2.26e-05,
+      "loss": 1.3346,
+      "step": 565
+    },
+    {
+      "epoch": 0.0114,
+      "grad_norm": 1.2734375,
+      "learning_rate": 2.2800000000000002e-05,
+      "loss": 1.3402,
+      "step": 570
+    },
+    {
+      "epoch": 0.0115,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 1.3191,
+      "step": 575
+    },
+    {
+      "epoch": 0.0116,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.32e-05,
+      "loss": 1.3331,
+      "step": 580
+    },
+    {
+      "epoch": 0.0117,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.3400000000000003e-05,
+      "loss": 1.3402,
+      "step": 585
+    },
+    {
+      "epoch": 0.0118,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.36e-05,
+      "loss": 1.3445,
+      "step": 590
+    },
+    {
+      "epoch": 0.0119,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.38e-05,
+      "loss": 1.3232,
+      "step": 595
+    },
+    {
+      "epoch": 0.012,
+      "grad_norm": 1.1484375,
+      "learning_rate": 2.4e-05,
+      "loss": 1.3162,
+      "step": 600
+    },
+    {
+      "epoch": 0.0121,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.4200000000000002e-05,
+      "loss": 1.3005,
+      "step": 605
+    },
+    {
+      "epoch": 0.0122,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.44e-05,
+      "loss": 1.2987,
+      "step": 610
+    },
+    {
+      "epoch": 0.0123,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.46e-05,
+      "loss": 1.2934,
+      "step": 615
+    },
+    {
+      "epoch": 0.0124,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.48e-05,
+      "loss": 1.2979,
+      "step": 620
+    },
+    {
+      "epoch": 0.0125,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.5e-05,
+      "loss": 1.3077,
+      "step": 625
+    },
+    {
+      "epoch": 0.0126,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.5200000000000003e-05,
+      "loss": 1.2867,
+      "step": 630
+    },
+    {
+      "epoch": 0.0127,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.54e-05,
+      "loss": 1.2905,
+      "step": 635
+    },
+    {
+      "epoch": 0.0128,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.5600000000000002e-05,
+      "loss": 1.2952,
+      "step": 640
+    },
+    {
+      "epoch": 0.0129,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.58e-05,
+      "loss": 1.2943,
+      "step": 645
+    },
+    {
+      "epoch": 0.013,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 1.3066,
+      "step": 650
+    },
+    {
+      "epoch": 0.0131,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.6200000000000003e-05,
+      "loss": 1.2996,
+      "step": 655
+    },
+    {
+      "epoch": 0.0132,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.64e-05,
+      "loss": 1.2991,
+      "step": 660
+    },
+    {
+      "epoch": 0.0133,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.6600000000000003e-05,
+      "loss": 1.3189,
+      "step": 665
+    },
+    {
+      "epoch": 0.0134,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.6800000000000004e-05,
+      "loss": 1.2904,
+      "step": 670
+    },
+    {
+      "epoch": 0.0135,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 1.2887,
+      "step": 675
+    },
+    {
+      "epoch": 0.0136,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.7200000000000004e-05,
+      "loss": 1.2996,
+      "step": 680
+    },
+    {
+      "epoch": 0.0137,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.7400000000000002e-05,
+      "loss": 1.2775,
+      "step": 685
+    },
+    {
+      "epoch": 0.0138,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.7600000000000003e-05,
+      "loss": 1.3152,
+      "step": 690
+    },
+    {
+      "epoch": 0.0139,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.7800000000000005e-05,
+      "loss": 1.2881,
+      "step": 695
+    },
+    {
+      "epoch": 0.014,
+      "grad_norm": 1.5390625,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 1.3061,
+      "step": 700
+    },
+    {
+      "epoch": 0.0141,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.8199999999999998e-05,
+      "loss": 1.2836,
+      "step": 705
+    },
+    {
+      "epoch": 0.0142,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.84e-05,
+      "loss": 1.284,
+      "step": 710
+    },
+    {
+      "epoch": 0.0143,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.86e-05,
+      "loss": 1.2783,
+      "step": 715
+    },
+    {
+      "epoch": 0.0144,
+      "grad_norm": 1.375,
+      "learning_rate": 2.88e-05,
+      "loss": 1.2979,
+      "step": 720
+    },
+    {
+      "epoch": 0.0145,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.9e-05,
+      "loss": 1.2927,
+      "step": 725
+    },
+    {
+      "epoch": 0.0146,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.9199999999999998e-05,
+      "loss": 1.2912,
+      "step": 730
+    },
+    {
+      "epoch": 0.0147,
+      "grad_norm": 0.44921875,
+      "learning_rate": 2.94e-05,
+      "loss": 1.2914,
+      "step": 735
+    },
+    {
+      "epoch": 0.0148,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.96e-05,
+      "loss": 1.2884,
+      "step": 740
+    },
+    {
+      "epoch": 0.0149,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.98e-05,
+      "loss": 1.2859,
+      "step": 745
+    },
+    {
+      "epoch": 0.015,
+      "grad_norm": 0.453125,
+      "learning_rate": 3e-05,
+      "loss": 1.2754,
+      "step": 750
+    },
+    {
+      "epoch": 0.0151,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.02e-05,
+      "loss": 1.2759,
+      "step": 755
+    },
+    {
+      "epoch": 0.0152,
+      "grad_norm": 0.4765625,
+      "learning_rate": 3.04e-05,
+      "loss": 1.2778,
+      "step": 760
+    },
+    {
+      "epoch": 0.0153,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.06e-05,
+      "loss": 1.2911,
+      "step": 765
+    },
+    {
+      "epoch": 0.0154,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.08e-05,
+      "loss": 1.2925,
+      "step": 770
+    },
+    {
+      "epoch": 0.0155,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.1e-05,
+      "loss": 1.2886,
+      "step": 775
+    },
+    {
+      "epoch": 0.0156,
+      "grad_norm": 0.375,
+      "learning_rate": 3.12e-05,
+      "loss": 1.2766,
+      "step": 780
+    },
+    {
+      "epoch": 0.0157,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.1400000000000004e-05,
+      "loss": 1.2861,
+      "step": 785
+    },
+    {
+      "epoch": 0.0158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.16e-05,
+      "loss": 1.2759,
+      "step": 790
+    },
+    {
+      "epoch": 0.0159,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.18e-05,
+      "loss": 1.2797,
+      "step": 795
+    },
+    {
+      "epoch": 0.016,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 1.2718,
+      "step": 800
+    },
+    {
+      "epoch": 0.0161,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.2200000000000003e-05,
+      "loss": 1.2832,
+      "step": 805
+    },
+    {
+      "epoch": 0.0162,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.24e-05,
+      "loss": 1.2777,
+      "step": 810
+    },
+    {
+      "epoch": 0.0163,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.26e-05,
+      "loss": 1.2797,
+      "step": 815
+    },
+    {
+      "epoch": 0.0164,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.2800000000000004e-05,
+      "loss": 1.2861,
+      "step": 820
+    },
+    {
+      "epoch": 0.0165,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.3e-05,
+      "loss": 1.2852,
+      "step": 825
+    },
+    {
+      "epoch": 0.0166,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.32e-05,
+      "loss": 1.2813,
+      "step": 830
+    },
+    {
+      "epoch": 0.0167,
+      "grad_norm": 0.318359375,
+      "learning_rate": 3.3400000000000005e-05,
+      "loss": 1.2765,
+      "step": 835
+    },
+    {
+      "epoch": 0.0168,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.3600000000000004e-05,
+      "loss": 1.2945,
+      "step": 840
+    },
+    {
+      "epoch": 0.0169,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.38e-05,
+      "loss": 1.2825,
+      "step": 845
+    },
+    {
+      "epoch": 0.017,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 1.2905,
+      "step": 850
+    },
+    {
+      "epoch": 0.0171,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.4200000000000005e-05,
+      "loss": 1.3032,
+      "step": 855
+    },
+    {
+      "epoch": 0.0172,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.4399999999999996e-05,
+      "loss": 1.2758,
+      "step": 860
+    },
+    {
+      "epoch": 0.0173,
+      "grad_norm": 0.49609375,
+      "learning_rate": 3.46e-05,
+      "loss": 1.2831,
+      "step": 865
+    },
+    {
+      "epoch": 0.0174,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.48e-05,
+      "loss": 1.2828,
+      "step": 870
+    },
+    {
+      "epoch": 0.0175,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.5e-05,
+      "loss": 1.2846,
+      "step": 875
+    },
+    {
+      "epoch": 0.0176,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.52e-05,
+      "loss": 1.2829,
+      "step": 880
+    },
+    {
+      "epoch": 0.0177,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.54e-05,
+      "loss": 1.2801,
+      "step": 885
+    },
+    {
+      "epoch": 0.0178,
+      "grad_norm": 0.44921875,
+      "learning_rate": 3.56e-05,
+      "loss": 1.2946,
+      "step": 890
+    },
+    {
+      "epoch": 0.0179,
+      "grad_norm": 0.330078125,
+      "learning_rate": 3.58e-05,
+      "loss": 1.2757,
+      "step": 895
+    },
+    {
+      "epoch": 0.018,
+      "grad_norm": 0.287109375,
+      "learning_rate": 3.6e-05,
+      "loss": 1.2646,
+      "step": 900
+    },
+    {
+      "epoch": 0.0181,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.62e-05,
+      "loss": 1.2974,
+      "step": 905
+    },
+    {
+      "epoch": 0.0182,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.6400000000000004e-05,
+      "loss": 1.2955,
+      "step": 910
+    },
+    {
+      "epoch": 0.0183,
+      "grad_norm": 0.37890625,
+      "learning_rate": 3.66e-05,
+      "loss": 1.2691,
+      "step": 915
+    },
+    {
+      "epoch": 0.0184,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.68e-05,
+      "loss": 1.2937,
+      "step": 920
+    },
+    {
+      "epoch": 0.0185,
+      "grad_norm": 0.26953125,
+      "learning_rate": 3.7e-05,
+      "loss": 1.2876,
+      "step": 925
+    },
+    {
+      "epoch": 0.0186,
+      "grad_norm": 0.27734375,
+      "learning_rate": 3.72e-05,
+      "loss": 1.2768,
+      "step": 930
+    },
+    {
+      "epoch": 0.0187,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.74e-05,
+      "loss": 1.2729,
+      "step": 935
+    },
+    {
+      "epoch": 0.0188,
+      "grad_norm": 0.275390625,
+      "learning_rate": 3.76e-05,
+      "loss": 1.2823,
+      "step": 940
+    },
+    {
+      "epoch": 0.0189,
+      "grad_norm": 0.283203125,
+      "learning_rate": 3.7800000000000004e-05,
+      "loss": 1.2785,
+      "step": 945
+    },
+    {
+      "epoch": 0.019,
+      "grad_norm": 0.375,
+      "learning_rate": 3.8e-05,
+      "loss": 1.2864,
+      "step": 950
+    },
+    {
+      "epoch": 0.0191,
+      "grad_norm": 0.29296875,
+      "learning_rate": 3.82e-05,
+      "loss": 1.292,
+      "step": 955
+    },
+    {
+      "epoch": 0.0192,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 1.2802,
+      "step": 960
+    },
+    {
+      "epoch": 0.0193,
+      "grad_norm": 0.30078125,
+      "learning_rate": 3.86e-05,
+      "loss": 1.2641,
+      "step": 965
+    },
+    {
+      "epoch": 0.0194,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.88e-05,
+      "loss": 1.2784,
+      "step": 970
+    },
+    {
+      "epoch": 0.0195,
+      "grad_norm": 0.458984375,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 1.3087,
+      "step": 975
+    },
+    {
+      "epoch": 0.0196,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.9200000000000004e-05,
+      "loss": 1.3011,
+      "step": 980
+    },
+    {
+      "epoch": 0.0197,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.94e-05,
+      "loss": 1.296,
+      "step": 985
+    },
+    {
+      "epoch": 0.0198,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.960000000000001e-05,
+      "loss": 1.2965,
+      "step": 990
+    },
+    {
+      "epoch": 0.0199,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.9800000000000005e-05,
+      "loss": 1.28,
+      "step": 995
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4e-05,
+      "loss": 1.2904,
+      "step": 1000
+    },
+    {
+      "epoch": 0.0201,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.02e-05,
+      "loss": 1.2698,
+      "step": 1005
+    },
+    {
+      "epoch": 0.0202,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.0400000000000006e-05,
+      "loss": 1.274,
+      "step": 1010
+    },
+    {
+      "epoch": 0.0203,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.0600000000000004e-05,
+      "loss": 1.2872,
+      "step": 1015
+    },
+    {
+      "epoch": 0.0204,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.08e-05,
+      "loss": 1.2748,
+      "step": 1020
+    },
+    {
+      "epoch": 0.0205,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.1e-05,
+      "loss": 1.2713,
+      "step": 1025
+    },
+    {
+      "epoch": 0.0206,
+      "grad_norm": 0.3046875,
+      "learning_rate": 4.12e-05,
+      "loss": 1.2999,
+      "step": 1030
+    },
+    {
+      "epoch": 0.0207,
+      "grad_norm": 0.365234375,
+      "learning_rate": 4.14e-05,
+      "loss": 1.2866,
+      "step": 1035
+    },
+    {
+      "epoch": 0.0208,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.16e-05,
+      "loss": 1.2662,
+      "step": 1040
+    },
+    {
+      "epoch": 0.0209,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.18e-05,
+      "loss": 1.2901,
+      "step": 1045
+    },
+    {
+      "epoch": 0.021,
+      "grad_norm": 0.3125,
+      "learning_rate": 4.2e-05,
+      "loss": 1.2859,
+      "step": 1050
+    },
+    {
+      "epoch": 0.0211,
+      "grad_norm": 0.359375,
+      "learning_rate": 4.22e-05,
+      "loss": 1.2891,
+      "step": 1055
+    },
+    {
+      "epoch": 0.0212,
+      "grad_norm": 0.3203125,
+      "learning_rate": 4.24e-05,
+      "loss": 1.2752,
+      "step": 1060
+    },
+    {
+      "epoch": 0.0213,
+      "grad_norm": 0.29296875,
+      "learning_rate": 4.26e-05,
+      "loss": 1.2744,
+      "step": 1065
+    },
+    {
+      "epoch": 0.0214,
+      "grad_norm": 0.255859375,
+      "learning_rate": 4.2800000000000004e-05,
+      "loss": 1.2744,
+      "step": 1070
+    },
+    {
+      "epoch": 0.0215,
+      "grad_norm": 0.357421875,
+      "learning_rate": 4.3e-05,
+      "loss": 1.2824,
+      "step": 1075
+    },
+    {
+      "epoch": 0.0216,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.32e-05,
+      "loss": 1.2733,
+      "step": 1080
+    },
+    {
+      "epoch": 0.0217,
+      "grad_norm": 0.375,
+      "learning_rate": 4.3400000000000005e-05,
+      "loss": 1.2975,
+      "step": 1085
+    },
+    {
+      "epoch": 0.0218,
+      "grad_norm": 0.384765625,
+      "learning_rate": 4.36e-05,
+      "loss": 1.2753,
+      "step": 1090
+    },
+    {
+      "epoch": 0.0219,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.38e-05,
+      "loss": 1.2807,
+      "step": 1095
+    },
+    {
+      "epoch": 0.022,
+      "grad_norm": 0.279296875,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 1.2617,
+      "step": 1100
+    },
+    {
+      "epoch": 0.0221,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.4200000000000004e-05,
+      "loss": 1.3028,
+      "step": 1105
+    },
+    {
+      "epoch": 0.0222,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.44e-05,
+      "loss": 1.2973,
+      "step": 1110
+    },
+    {
+      "epoch": 0.0223,
+      "grad_norm": 0.376953125,
+      "learning_rate": 4.46e-05,
+      "loss": 1.285,
+      "step": 1115
+    },
+    {
+      "epoch": 0.0224,
+      "grad_norm": 0.296875,
+      "learning_rate": 4.4800000000000005e-05,
+      "loss": 1.2769,
+      "step": 1120
+    },
+    {
+      "epoch": 0.0225,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.5e-05,
+      "loss": 1.279,
+      "step": 1125
+    },
+    {
+      "epoch": 0.0226,
+      "grad_norm": 0.28515625,
+      "learning_rate": 4.52e-05,
+      "loss": 1.2887,
+      "step": 1130
+    },
+    {
+      "epoch": 0.0227,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.5400000000000006e-05,
+      "loss": 1.2769,
+      "step": 1135
+    },
+    {
+      "epoch": 0.0228,
+      "grad_norm": 0.380859375,
+      "learning_rate": 4.5600000000000004e-05,
+      "loss": 1.2911,
+      "step": 1140
+    },
+    {
+      "epoch": 0.0229,
+      "grad_norm": 0.30859375,
+      "learning_rate": 4.58e-05,
+      "loss": 1.2891,
+      "step": 1145
+    },
+    {
+      "epoch": 0.023,
+      "grad_norm": 0.376953125,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 1.2981,
+      "step": 1150
+    },
+    {
+      "epoch": 0.0231,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.6200000000000005e-05,
+      "loss": 1.2796,
+      "step": 1155
+    },
+    {
+      "epoch": 0.0232,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.64e-05,
+      "loss": 1.2823,
+      "step": 1160
+    },
+    {
+      "epoch": 0.0233,
+      "grad_norm": 21.5,
+      "learning_rate": 4.660000000000001e-05,
+      "loss": 1.2755,
+      "step": 1165
+    },
+    {
+      "epoch": 0.0234,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.6800000000000006e-05,
+      "loss": 1.2689,
+      "step": 1170
+    },
+    {
+      "epoch": 0.0235,
+      "grad_norm": 0.208984375,
+      "learning_rate": 4.7e-05,
+      "loss": 1.2674,
+      "step": 1175
+    },
+    {
+      "epoch": 0.0236,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.72e-05,
+      "loss": 1.2875,
+      "step": 1180
+    },
+    {
+      "epoch": 0.0237,
+      "grad_norm": 0.263671875,
+      "learning_rate": 4.74e-05,
+      "loss": 1.2653,
+      "step": 1185
+    },
+    {
+      "epoch": 0.0238,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.76e-05,
+      "loss": 1.2629,
+      "step": 1190
+    },
+    {
+      "epoch": 0.0239,
+      "grad_norm": 0.3203125,
+      "learning_rate": 4.78e-05,
+      "loss": 1.2766,
+      "step": 1195
+    },
+    {
+      "epoch": 0.024,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.8e-05,
+      "loss": 1.296,
+      "step": 1200
+    },
+    {
+      "epoch": 0.0241,
+      "grad_norm": 0.298828125,
+      "learning_rate": 4.82e-05,
+      "loss": 1.2924,
+      "step": 1205
+    },
+    {
+      "epoch": 0.0242,
+      "grad_norm": 0.28125,
+      "learning_rate": 4.8400000000000004e-05,
+      "loss": 1.2838,
+      "step": 1210
+    },
+    {
+      "epoch": 0.0243,
+      "grad_norm": 0.2890625,
+      "learning_rate": 4.86e-05,
+      "loss": 1.2859,
+      "step": 1215
+    },
+    {
+      "epoch": 0.0244,
+      "grad_norm": 0.310546875,
+      "learning_rate": 4.88e-05,
+      "loss": 1.2927,
+      "step": 1220
+    },
+    {
+      "epoch": 0.0245,
+      "grad_norm": 0.3203125,
+      "learning_rate": 4.9e-05,
+      "loss": 1.2823,
+      "step": 1225
+    },
+    {
+      "epoch": 0.0246,
+      "grad_norm": 0.2890625,
+      "learning_rate": 4.92e-05,
+      "loss": 1.2751,
+      "step": 1230
+    },
+    {
+      "epoch": 0.0247,
+      "grad_norm": 0.361328125,
+      "learning_rate": 4.94e-05,
+      "loss": 1.2822,
+      "step": 1235
+    },
+    {
+      "epoch": 0.0248,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 4.96e-05,
+      "loss": 1.2759,
+      "step": 1240
+    },
+    {
+      "epoch": 0.0249,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.9800000000000004e-05,
+      "loss": 1.2772,
+      "step": 1245
+    },
+    {
+      "epoch": 0.025,
+      "grad_norm": 0.306640625,
+      "learning_rate": 5e-05,
+      "loss": 1.2781,
+      "step": 1250
+    },
+    {
+      "epoch": 0.0251,
+      "grad_norm": 0.314453125,
+      "learning_rate": 5.02e-05,
+      "loss": 1.2697,
+      "step": 1255
+    },
+    {
+      "epoch": 0.0252,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.0400000000000005e-05,
+      "loss": 1.2927,
+      "step": 1260
+    },
+    {
+      "epoch": 0.0253,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.0600000000000003e-05,
+      "loss": 1.2767,
+      "step": 1265
+    },
+    {
+      "epoch": 0.0254,
+      "grad_norm": 0.2890625,
+      "learning_rate": 5.08e-05,
+      "loss": 1.2734,
+      "step": 1270
+    },
+    {
+      "epoch": 0.0255,
+      "grad_norm": 0.283203125,
+      "learning_rate": 5.1000000000000006e-05,
+      "loss": 1.2851,
+      "step": 1275
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.1200000000000004e-05,
+      "loss": 1.2764,
+      "step": 1280
+    },
+    {
+      "epoch": 0.0257,
+      "grad_norm": 0.35546875,
+      "learning_rate": 5.14e-05,
+      "loss": 1.2728,
+      "step": 1285
+    },
+    {
+      "epoch": 0.0258,
+      "grad_norm": 0.2265625,
+      "learning_rate": 5.16e-05,
+      "loss": 1.2888,
+      "step": 1290
+    },
+    {
+      "epoch": 0.0259,
+      "grad_norm": 0.283203125,
+      "learning_rate": 5.1800000000000005e-05,
+      "loss": 1.2764,
+      "step": 1295
+    },
+    {
+      "epoch": 0.026,
+      "grad_norm": 0.28515625,
+      "learning_rate": 5.2000000000000004e-05,
+      "loss": 1.277,
+      "step": 1300
+    },
+    {
+      "epoch": 0.0261,
+      "grad_norm": 0.25390625,
+      "learning_rate": 5.22e-05,
+      "loss": 1.2801,
+      "step": 1305
+    },
+    {
+      "epoch": 0.0262,
+      "grad_norm": 0.306640625,
+      "learning_rate": 5.2400000000000007e-05,
+      "loss": 1.2958,
+      "step": 1310
+    },
+    {
+      "epoch": 0.0263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 5.2600000000000005e-05,
+      "loss": 1.2963,
+      "step": 1315
+    },
+    {
+      "epoch": 0.0264,
+      "grad_norm": 0.416015625,
+      "learning_rate": 5.28e-05,
+      "loss": 1.2809,
+      "step": 1320
+    },
+    {
+      "epoch": 0.0265,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.300000000000001e-05,
+      "loss": 1.2758,
+      "step": 1325
+    },
+    {
+      "epoch": 0.0266,
+      "grad_norm": 0.25390625,
+      "learning_rate": 5.3200000000000006e-05,
+      "loss": 1.2758,
+      "step": 1330
+    },
+    {
+      "epoch": 0.0267,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.3400000000000004e-05,
+      "loss": 1.2788,
+      "step": 1335
+    },
+    {
+      "epoch": 0.0268,
+      "grad_norm": 0.267578125,
+      "learning_rate": 5.360000000000001e-05,
+      "loss": 1.2839,
+      "step": 1340
+    },
+    {
+      "epoch": 0.0269,
+      "grad_norm": 0.384765625,
+      "learning_rate": 5.380000000000001e-05,
+      "loss": 1.2982,
+      "step": 1345
+    },
+    {
+      "epoch": 0.027,
+      "grad_norm": 0.390625,
+      "learning_rate": 5.4000000000000005e-05,
+      "loss": 1.2729,
+      "step": 1350
+    },
+    {
+      "epoch": 0.0271,
+      "grad_norm": 0.2578125,
+      "learning_rate": 5.420000000000001e-05,
+      "loss": 1.284,
+      "step": 1355
+    },
+    {
+      "epoch": 0.0272,
+      "grad_norm": 0.25390625,
+      "learning_rate": 5.440000000000001e-05,
+      "loss": 1.2775,
+      "step": 1360
+    },
+    {
+      "epoch": 0.0273,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.4600000000000006e-05,
+      "loss": 1.2813,
+      "step": 1365
+    },
+    {
+      "epoch": 0.0274,
+      "grad_norm": 0.283203125,
+      "learning_rate": 5.4800000000000004e-05,
+      "loss": 1.2761,
+      "step": 1370
+    },
+    {
+      "epoch": 0.0275,
+      "grad_norm": 0.29296875,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 1.2694,
+      "step": 1375
+    },
+    {
+      "epoch": 0.0276,
+      "grad_norm": 0.380859375,
+      "learning_rate": 5.520000000000001e-05,
+      "loss": 1.2791,
+      "step": 1380
+    },
+    {
+      "epoch": 0.0277,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.5400000000000005e-05,
+      "loss": 1.2768,
+      "step": 1385
+    },
+    {
+      "epoch": 0.0278,
+      "grad_norm": 0.224609375,
+      "learning_rate": 5.560000000000001e-05,
+      "loss": 1.2986,
+      "step": 1390
+    },
+    {
+      "epoch": 0.0279,
+      "grad_norm": 0.30859375,
+      "learning_rate": 5.580000000000001e-05,
+      "loss": 1.2638,
+      "step": 1395
+    },
+    {
+      "epoch": 0.028,
+      "grad_norm": 0.3671875,
+      "learning_rate": 5.6000000000000006e-05,
+      "loss": 1.2769,
+      "step": 1400
+    },
+    {
+      "epoch": 0.0281,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 5.620000000000001e-05,
+      "loss": 1.2844,
+      "step": 1405
+    },
+    {
+      "epoch": 0.0282,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.6399999999999995e-05,
+      "loss": 1.2715,
+      "step": 1410
+    },
+    {
+      "epoch": 0.0283,
+      "grad_norm": 0.271484375,
+      "learning_rate": 5.66e-05,
+      "loss": 1.2876,
+      "step": 1415
+    },
+    {
+      "epoch": 0.0284,
+      "grad_norm": 0.287109375,
+      "learning_rate": 5.68e-05,
+      "loss": 1.2686,
+      "step": 1420
+    },
+    {
+      "epoch": 0.0285,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 5.6999999999999996e-05,
+      "loss": 1.2717,
+      "step": 1425
+    },
+    {
+      "epoch": 0.0286,
+      "grad_norm": 0.44140625,
+      "learning_rate": 5.72e-05,
+      "loss": 1.2998,
+      "step": 1430
+    },
+    {
+      "epoch": 0.0287,
+      "grad_norm": 0.294921875,
+      "learning_rate": 5.74e-05,
+      "loss": 1.2718,
+      "step": 1435
+    },
+    {
+      "epoch": 0.0288,
+      "grad_norm": 0.29296875,
+      "learning_rate": 5.76e-05,
+      "loss": 1.2874,
+      "step": 1440
+    },
+    {
+      "epoch": 0.0289,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.7799999999999995e-05,
+      "loss": 1.2685,
+      "step": 1445
+    },
+    {
+      "epoch": 0.029,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 5.8e-05,
+      "loss": 1.276,
+      "step": 1450
+    },
+    {
+      "epoch": 0.0291,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 5.82e-05,
+      "loss": 1.275,
+      "step": 1455
+    },
+    {
+      "epoch": 0.0292,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 5.8399999999999997e-05,
+      "loss": 1.278,
+      "step": 1460
+    },
+    {
+      "epoch": 0.0293,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 5.86e-05,
+      "loss": 1.2612,
+      "step": 1465
+    },
+    {
+      "epoch": 0.0294,
+      "grad_norm": 0.310546875,
+      "learning_rate": 5.88e-05,
+      "loss": 1.2845,
+      "step": 1470
+    },
+    {
+      "epoch": 0.0295,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.9e-05,
+      "loss": 1.2944,
+      "step": 1475
+    },
+    {
+      "epoch": 0.0296,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.92e-05,
+      "loss": 1.2727,
+      "step": 1480
+    },
+    {
+      "epoch": 0.0297,
+      "grad_norm": 0.275390625,
+      "learning_rate": 5.94e-05,
+      "loss": 1.2852,
+      "step": 1485
+    },
+    {
+      "epoch": 0.0298,
+      "grad_norm": 0.2890625,
+      "learning_rate": 5.96e-05,
+      "loss": 1.2815,
+      "step": 1490
+    },
+    {
+      "epoch": 0.0299,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.9800000000000003e-05,
+      "loss": 1.2824,
+      "step": 1495
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 6e-05,
+      "loss": 1.2868,
+      "step": 1500
+    },
+    {
+      "epoch": 0.0301,
+      "grad_norm": 0.244140625,
+      "learning_rate": 6.02e-05,
+      "loss": 1.2764,
+      "step": 1505
+    },
+    {
+      "epoch": 0.0302,
+      "grad_norm": 0.380859375,
+      "learning_rate": 6.04e-05,
+      "loss": 1.2736,
+      "step": 1510
+    },
+    {
+      "epoch": 0.0303,
+      "grad_norm": 0.359375,
+      "learning_rate": 6.06e-05,
+      "loss": 1.2959,
+      "step": 1515
+    },
+    {
+      "epoch": 0.0304,
+      "grad_norm": 0.380859375,
+      "learning_rate": 6.08e-05,
+      "loss": 1.2902,
+      "step": 1520
+    },
+    {
+      "epoch": 0.0305,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 6.1e-05,
+      "loss": 1.2661,
+      "step": 1525
+    },
+    {
+      "epoch": 0.0306,
+      "grad_norm": 0.298828125,
+      "learning_rate": 6.12e-05,
+      "loss": 1.2746,
+      "step": 1530
+    },
+    {
+      "epoch": 0.0307,
+      "grad_norm": 0.357421875,
+      "learning_rate": 6.14e-05,
+      "loss": 1.2796,
+      "step": 1535
+    },
+    {
+      "epoch": 0.0308,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 6.16e-05,
+      "loss": 1.2843,
+      "step": 1540
+    },
+    {
+      "epoch": 0.0309,
+      "grad_norm": 0.255859375,
+      "learning_rate": 6.18e-05,
+      "loss": 1.2977,
+      "step": 1545
+    },
+    {
+      "epoch": 0.031,
+      "grad_norm": 0.27734375,
+      "learning_rate": 6.2e-05,
+      "loss": 1.2822,
+      "step": 1550
+    },
+    {
+      "epoch": 0.0311,
+      "grad_norm": 0.255859375,
+      "learning_rate": 6.220000000000001e-05,
+      "loss": 1.2882,
+      "step": 1555
+    },
+    {
+      "epoch": 0.0312,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.24e-05,
+      "loss": 1.29,
+      "step": 1560
+    },
+    {
+      "epoch": 0.0313,
+      "grad_norm": 0.216796875,
+      "learning_rate": 6.26e-05,
+      "loss": 1.2745,
+      "step": 1565
+    },
+    {
+      "epoch": 0.0314,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.280000000000001e-05,
+      "loss": 1.2704,
+      "step": 1570
+    },
+    {
+      "epoch": 0.0315,
+      "grad_norm": 0.208984375,
+      "learning_rate": 6.3e-05,
+      "loss": 1.2702,
+      "step": 1575
+    },
+    {
+      "epoch": 0.0316,
+      "grad_norm": 0.279296875,
+      "learning_rate": 6.32e-05,
+      "loss": 1.2718,
+      "step": 1580
+    },
+    {
+      "epoch": 0.0317,
+      "grad_norm": 0.3046875,
+      "learning_rate": 6.340000000000001e-05,
+      "loss": 1.2832,
+      "step": 1585
+    },
+    {
+      "epoch": 0.0318,
+      "grad_norm": 0.279296875,
+      "learning_rate": 6.36e-05,
+      "loss": 1.278,
+      "step": 1590
+    },
+    {
+      "epoch": 0.0319,
+      "grad_norm": 0.30859375,
+      "learning_rate": 6.38e-05,
+      "loss": 1.2852,
+      "step": 1595
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 0.37109375,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 1.284,
+      "step": 1600
+    },
+    {
+      "epoch": 0.0321,
+      "grad_norm": 0.21875,
+      "learning_rate": 6.42e-05,
+      "loss": 1.2889,
+      "step": 1605
+    },
+    {
+      "epoch": 0.0322,
+      "grad_norm": 0.82421875,
+      "learning_rate": 6.440000000000001e-05,
+      "loss": 1.2825,
+      "step": 1610
+    },
+    {
+      "epoch": 0.0323,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 6.460000000000001e-05,
+      "loss": 1.2801,
+      "step": 1615
+    },
+    {
+      "epoch": 0.0324,
+      "grad_norm": 0.28125,
+      "learning_rate": 6.48e-05,
+      "loss": 1.2759,
+      "step": 1620
+    },
+    {
+      "epoch": 0.0325,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.500000000000001e-05,
+      "loss": 1.2855,
+      "step": 1625
+    },
+    {
+      "epoch": 0.0326,
+      "grad_norm": 0.203125,
+      "learning_rate": 6.52e-05,
+      "loss": 1.2651,
+      "step": 1630
+    },
+    {
+      "epoch": 0.0327,
+      "grad_norm": 1.3125,
+      "learning_rate": 6.54e-05,
+      "loss": 1.2849,
+      "step": 1635
+    },
+    {
+      "epoch": 0.0328,
+      "grad_norm": 0.302734375,
+      "learning_rate": 6.560000000000001e-05,
+      "loss": 1.2896,
+      "step": 1640
+    },
+    {
+      "epoch": 0.0329,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.58e-05,
+      "loss": 1.2933,
+      "step": 1645
+    },
+    {
+      "epoch": 0.033,
+      "grad_norm": 0.259765625,
+      "learning_rate": 6.6e-05,
+      "loss": 1.2889,
+      "step": 1650
+    },
+    {
+      "epoch": 0.0331,
+      "grad_norm": 0.275390625,
+      "learning_rate": 6.620000000000001e-05,
+      "loss": 1.29,
+      "step": 1655
+    },
+    {
+      "epoch": 0.0332,
+      "grad_norm": 0.384765625,
+      "learning_rate": 6.64e-05,
+      "loss": 1.2891,
+      "step": 1660
+    },
+    {
+      "epoch": 0.0333,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.66e-05,
+      "loss": 1.2671,
+      "step": 1665
+    },
+    {
+      "epoch": 0.0334,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 6.680000000000001e-05,
+      "loss": 1.2791,
+      "step": 1670
+    },
+    {
+      "epoch": 0.0335,
+      "grad_norm": 0.400390625,
+      "learning_rate": 6.7e-05,
+      "loss": 1.2796,
+      "step": 1675
+    },
+    {
+      "epoch": 0.0336,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 6.720000000000001e-05,
+      "loss": 1.274,
+      "step": 1680
+    },
+    {
+      "epoch": 0.0337,
+      "grad_norm": 2.515625,
+      "learning_rate": 6.740000000000001e-05,
+      "loss": 1.2853,
+      "step": 1685
+    },
+    {
+      "epoch": 0.0338,
+      "grad_norm": 0.22265625,
+      "learning_rate": 6.76e-05,
+      "loss": 1.286,
+      "step": 1690
+    },
+    {
+      "epoch": 0.0339,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 6.780000000000001e-05,
+      "loss": 1.2778,
+      "step": 1695
+    },
+    {
+      "epoch": 0.034,
+      "grad_norm": 0.216796875,
+      "learning_rate": 6.800000000000001e-05,
+      "loss": 1.2863,
+      "step": 1700
+    },
+    {
+      "epoch": 0.0341,
+      "grad_norm": 0.36328125,
+      "learning_rate": 6.82e-05,
+      "loss": 1.2913,
+      "step": 1705
+    },
+    {
+      "epoch": 0.0342,
+      "grad_norm": 0.291015625,
+      "learning_rate": 6.840000000000001e-05,
+      "loss": 1.2698,
+      "step": 1710
+    },
+    {
+      "epoch": 0.0343,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 6.860000000000001e-05,
+      "loss": 1.2805,
+      "step": 1715
+    },
+    {
+      "epoch": 0.0344,
+      "grad_norm": 0.314453125,
+      "learning_rate": 6.879999999999999e-05,
+      "loss": 1.2579,
+      "step": 1720
+    },
+    {
+      "epoch": 0.0345,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 6.9e-05,
+      "loss": 1.2875,
+      "step": 1725
+    },
+    {
+      "epoch": 0.0346,
+      "grad_norm": 0.265625,
+      "learning_rate": 6.92e-05,
+      "loss": 1.2847,
+      "step": 1730
+    },
+    {
+      "epoch": 0.0347,
+      "grad_norm": 0.287109375,
+      "learning_rate": 6.939999999999999e-05,
+      "loss": 1.2917,
+      "step": 1735
+    },
+    {
+      "epoch": 0.0348,
+      "grad_norm": 0.275390625,
+      "learning_rate": 6.96e-05,
+      "loss": 1.2965,
+      "step": 1740
+    },
+    {
+      "epoch": 0.0349,
+      "grad_norm": 0.240234375,
+      "learning_rate": 6.98e-05,
+      "loss": 1.2752,
+      "step": 1745
+    },
+    {
+      "epoch": 0.035,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 7e-05,
+      "loss": 1.2728,
+      "step": 1750
+    },
+    {
+      "epoch": 0.0351,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.02e-05,
+      "loss": 1.2754,
+      "step": 1755
+    },
+    {
+      "epoch": 0.0352,
+      "grad_norm": 0.234375,
+      "learning_rate": 7.04e-05,
+      "loss": 1.2747,
+      "step": 1760
+    },
+    {
+      "epoch": 0.0353,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.06e-05,
+      "loss": 1.2707,
+      "step": 1765
+    },
+    {
+      "epoch": 0.0354,
+      "grad_norm": 0.24609375,
+      "learning_rate": 7.08e-05,
+      "loss": 1.2717,
+      "step": 1770
+    },
+    {
+      "epoch": 0.0355,
+      "grad_norm": 0.220703125,
+      "learning_rate": 7.1e-05,
+      "loss": 1.285,
+      "step": 1775
+    },
+    {
+      "epoch": 0.0356,
+      "grad_norm": 0.359375,
+      "learning_rate": 7.12e-05,
+      "loss": 1.2825,
+      "step": 1780
+    },
+    {
+      "epoch": 0.0357,
+      "grad_norm": 0.24609375,
+      "learning_rate": 7.14e-05,
+      "loss": 1.2825,
+      "step": 1785
+    },
+    {
+      "epoch": 0.0358,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.16e-05,
+      "loss": 1.2624,
+      "step": 1790
+    },
+    {
+      "epoch": 0.0359,
+      "grad_norm": 0.20703125,
+      "learning_rate": 7.18e-05,
+      "loss": 1.2677,
+      "step": 1795
+    },
+    {
+      "epoch": 0.036,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.2e-05,
+      "loss": 1.2809,
+      "step": 1800
+    },
+    {
+      "epoch": 0.0361,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 7.22e-05,
+      "loss": 1.2812,
+      "step": 1805
+    },
+    {
+      "epoch": 0.0362,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 7.24e-05,
+      "loss": 1.2617,
+      "step": 1810
+    },
+    {
+      "epoch": 0.0363,
+      "grad_norm": 0.279296875,
+      "learning_rate": 7.26e-05,
+      "loss": 1.2677,
+      "step": 1815
+    },
+    {
+      "epoch": 0.0364,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.280000000000001e-05,
+      "loss": 1.2792,
+      "step": 1820
+    },
+    {
+      "epoch": 0.0365,
+      "grad_norm": 0.212890625,
+      "learning_rate": 7.3e-05,
+      "loss": 1.2823,
+      "step": 1825
+    },
+    {
+      "epoch": 0.0366,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.32e-05,
+      "loss": 1.2965,
+      "step": 1830
+    },
+    {
+      "epoch": 0.0367,
+      "grad_norm": 0.330078125,
+      "learning_rate": 7.340000000000001e-05,
+      "loss": 1.2895,
+      "step": 1835
+    },
+    {
+      "epoch": 0.0368,
+      "grad_norm": 0.3203125,
+      "learning_rate": 7.36e-05,
+      "loss": 1.276,
+      "step": 1840
+    },
+    {
+      "epoch": 0.0369,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.38e-05,
+      "loss": 1.2825,
+      "step": 1845
+    },
+    {
+      "epoch": 0.037,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.4e-05,
+      "loss": 1.2845,
+      "step": 1850
+    },
+    {
+      "epoch": 0.0371,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 7.42e-05,
+      "loss": 1.2943,
+      "step": 1855
+    },
+    {
+      "epoch": 0.0372,
+      "grad_norm": 0.244140625,
+      "learning_rate": 7.44e-05,
+      "loss": 1.2733,
+      "step": 1860
+    },
+    {
+      "epoch": 0.0373,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 7.46e-05,
+      "loss": 1.2825,
+      "step": 1865
+    },
+    {
+      "epoch": 0.0374,
+      "grad_norm": 0.486328125,
+      "learning_rate": 7.48e-05,
+      "loss": 1.2815,
+      "step": 1870
+    },
+    {
+      "epoch": 0.0375,
+      "grad_norm": 0.302734375,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 1.2881,
+      "step": 1875
+    },
+    {
+      "epoch": 0.0376,
+      "grad_norm": 0.20703125,
+      "learning_rate": 7.52e-05,
+      "loss": 1.2671,
+      "step": 1880
+    },
+    {
+      "epoch": 0.0377,
+      "grad_norm": 0.25,
+      "learning_rate": 7.54e-05,
+      "loss": 1.2822,
+      "step": 1885
+    },
+    {
+      "epoch": 0.0378,
+      "grad_norm": 0.265625,
+      "learning_rate": 7.560000000000001e-05,
+      "loss": 1.2915,
+      "step": 1890
+    },
+    {
+      "epoch": 0.0379,
+      "grad_norm": 0.265625,
+      "learning_rate": 7.58e-05,
+      "loss": 1.2753,
+      "step": 1895
+    },
+    {
+      "epoch": 0.038,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.6e-05,
+      "loss": 1.2691,
+      "step": 1900
+    },
+    {
+      "epoch": 0.0381,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 7.620000000000001e-05,
+      "loss": 1.3047,
+      "step": 1905
+    },
+    {
+      "epoch": 0.0382,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 7.64e-05,
+      "loss": 1.2838,
+      "step": 1910
+    },
+    {
+      "epoch": 0.0383,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.66e-05,
+      "loss": 1.2925,
+      "step": 1915
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 0.2734375,
+      "learning_rate": 7.680000000000001e-05,
+      "loss": 1.2866,
+      "step": 1920
+    },
+    {
+      "epoch": 0.0385,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 7.7e-05,
+      "loss": 1.2818,
+      "step": 1925
+    },
+    {
+      "epoch": 0.0386,
+      "grad_norm": 0.2109375,
+      "learning_rate": 7.72e-05,
+      "loss": 1.272,
+      "step": 1930
+    },
+    {
+      "epoch": 0.0387,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 7.740000000000001e-05,
+      "loss": 1.269,
+      "step": 1935
+    },
+    {
+      "epoch": 0.0388,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 7.76e-05,
+      "loss": 1.2562,
+      "step": 1940
+    },
+    {
+      "epoch": 0.0389,
+      "grad_norm": 0.306640625,
+      "learning_rate": 7.780000000000001e-05,
+      "loss": 1.3013,
+      "step": 1945
+    },
+    {
+      "epoch": 0.039,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 7.800000000000001e-05,
+      "loss": 1.2788,
+      "step": 1950
+    },
+    {
+      "epoch": 0.0391,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 7.82e-05,
+      "loss": 1.2918,
+      "step": 1955
+    },
+    {
+      "epoch": 0.0392,
+      "grad_norm": 0.25390625,
+      "learning_rate": 7.840000000000001e-05,
+      "loss": 1.2791,
+      "step": 1960
+    },
+    {
+      "epoch": 0.0393,
+      "grad_norm": 0.310546875,
+      "learning_rate": 7.860000000000001e-05,
+      "loss": 1.2749,
+      "step": 1965
+    },
+    {
+      "epoch": 0.0394,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.88e-05,
+      "loss": 1.2873,
+      "step": 1970
+    },
+    {
+      "epoch": 0.0395,
+      "grad_norm": 0.45703125,
+      "learning_rate": 7.900000000000001e-05,
+      "loss": 1.3044,
+      "step": 1975
+    },
+    {
+      "epoch": 0.0396,
+      "grad_norm": 0.25390625,
+      "learning_rate": 7.920000000000001e-05,
+      "loss": 1.2672,
+      "step": 1980
+    },
+    {
+      "epoch": 0.0397,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 7.94e-05,
+      "loss": 1.2666,
+      "step": 1985
+    },
+    {
+      "epoch": 0.0398,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 7.960000000000001e-05,
+      "loss": 1.2739,
+      "step": 1990
+    },
+    {
+      "epoch": 0.0399,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 7.98e-05,
+      "loss": 1.299,
+      "step": 1995
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 8e-05,
+      "loss": 1.2749,
+      "step": 2000
+    },
+    {
+      "epoch": 0.0401,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 8.020000000000001e-05,
+      "loss": 1.2739,
+      "step": 2005
+    },
+    {
+      "epoch": 0.0402,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.04e-05,
+      "loss": 1.2655,
+      "step": 2010
+    },
+    {
+      "epoch": 0.0403,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 8.060000000000001e-05,
+      "loss": 1.2842,
+      "step": 2015
+    },
+    {
+      "epoch": 0.0404,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.080000000000001e-05,
+      "loss": 1.2841,
+      "step": 2020
+    },
+    {
+      "epoch": 0.0405,
+      "grad_norm": 0.28125,
+      "learning_rate": 8.1e-05,
+      "loss": 1.3023,
+      "step": 2025
+    },
+    {
+      "epoch": 0.0406,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 8.120000000000001e-05,
+      "loss": 1.2888,
+      "step": 2030
+    },
+    {
+      "epoch": 0.0407,
+      "grad_norm": 0.25,
+      "learning_rate": 8.14e-05,
+      "loss": 1.2784,
+      "step": 2035
+    },
+    {
+      "epoch": 0.0408,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 8.16e-05,
+      "loss": 1.2769,
+      "step": 2040
+    },
+    {
+      "epoch": 0.0409,
+      "grad_norm": 0.20703125,
+      "learning_rate": 8.18e-05,
+      "loss": 1.2758,
+      "step": 2045
+    },
+    {
+      "epoch": 0.041,
+      "grad_norm": 0.234375,
+      "learning_rate": 8.2e-05,
+      "loss": 1.2859,
+      "step": 2050
+    },
+    {
+      "epoch": 0.0411,
+      "grad_norm": 0.296875,
+      "learning_rate": 8.22e-05,
+      "loss": 1.2871,
+      "step": 2055
+    },
+    {
+      "epoch": 0.0412,
+      "grad_norm": 0.296875,
+      "learning_rate": 8.24e-05,
+      "loss": 1.2624,
+      "step": 2060
+    },
+    {
+      "epoch": 0.0413,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 8.26e-05,
+      "loss": 1.2708,
+      "step": 2065
+    },
+    {
+      "epoch": 0.0414,
+      "grad_norm": 0.2421875,
+      "learning_rate": 8.28e-05,
+      "loss": 1.2871,
+      "step": 2070
+    },
+    {
+      "epoch": 0.0415,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 8.3e-05,
+      "loss": 1.3016,
+      "step": 2075
+    },
+    {
+      "epoch": 0.0416,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 8.32e-05,
+      "loss": 1.2695,
+      "step": 2080
+    },
+    {
+      "epoch": 0.0417,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.34e-05,
+      "loss": 1.2734,
+      "step": 2085
+    },
+    {
+      "epoch": 0.0418,
+      "grad_norm": 0.24609375,
+      "learning_rate": 8.36e-05,
+      "loss": 1.2874,
+      "step": 2090
+    },
+    {
+      "epoch": 0.0419,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 8.38e-05,
+      "loss": 1.2736,
+      "step": 2095
+    },
+    {
+      "epoch": 0.042,
+      "grad_norm": 0.265625,
+      "learning_rate": 8.4e-05,
+      "loss": 1.2754,
+      "step": 2100
+    },
+    {
+      "epoch": 0.0421,
+      "grad_norm": 0.287109375,
+      "learning_rate": 8.42e-05,
+      "loss": 1.2778,
+      "step": 2105
+    },
+    {
+      "epoch": 0.0422,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.44e-05,
+      "loss": 1.2911,
+      "step": 2110
+    },
+    {
+      "epoch": 0.0423,
+      "grad_norm": 0.314453125,
+      "learning_rate": 8.46e-05,
+      "loss": 1.2763,
+      "step": 2115
+    },
+    {
+      "epoch": 0.0424,
+      "grad_norm": 0.240234375,
+      "learning_rate": 8.48e-05,
+      "loss": 1.2841,
+      "step": 2120
+    },
+    {
+      "epoch": 0.0425,
+      "grad_norm": 0.291015625,
+      "learning_rate": 8.5e-05,
+      "loss": 1.2772,
+      "step": 2125
+    },
+    {
+      "epoch": 0.0426,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.52e-05,
+      "loss": 1.2762,
+      "step": 2130
+    },
+    {
+      "epoch": 0.0427,
+      "grad_norm": 0.236328125,
+      "learning_rate": 8.54e-05,
+      "loss": 1.2789,
+      "step": 2135
+    },
+    {
+      "epoch": 0.0428,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 8.560000000000001e-05,
+      "loss": 1.2648,
+      "step": 2140
+    },
+    {
+      "epoch": 0.0429,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 8.58e-05,
+      "loss": 1.276,
+      "step": 2145
+    },
+    {
+      "epoch": 0.043,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 8.6e-05,
+      "loss": 1.2944,
+      "step": 2150
+    },
+    {
+      "epoch": 0.0431,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 8.620000000000001e-05,
+      "loss": 1.2883,
+      "step": 2155
+    },
+    {
+      "epoch": 0.0432,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.64e-05,
+      "loss": 1.2796,
+      "step": 2160
+    },
+    {
+      "epoch": 0.0433,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 8.66e-05,
+      "loss": 1.2715,
+      "step": 2165
+    },
+    {
+      "epoch": 0.0434,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.680000000000001e-05,
+      "loss": 1.2815,
+      "step": 2170
+    },
+    {
+      "epoch": 0.0435,
+      "grad_norm": 0.35546875,
+      "learning_rate": 8.7e-05,
+      "loss": 1.286,
+      "step": 2175
+    },
+    {
+      "epoch": 0.0436,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.72e-05,
+      "loss": 1.2751,
+      "step": 2180
+    },
+    {
+      "epoch": 0.0437,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.740000000000001e-05,
+      "loss": 1.266,
+      "step": 2185
+    },
+    {
+      "epoch": 0.0438,
+      "grad_norm": 0.392578125,
+      "learning_rate": 8.76e-05,
+      "loss": 1.2652,
+      "step": 2190
+    },
+    {
+      "epoch": 0.0439,
+      "grad_norm": 0.265625,
+      "learning_rate": 8.78e-05,
+      "loss": 1.2854,
+      "step": 2195
+    },
+    {
+      "epoch": 0.044,
+      "grad_norm": 0.357421875,
+      "learning_rate": 8.800000000000001e-05,
+      "loss": 1.2795,
+      "step": 2200
+    },
+    {
+      "epoch": 0.0441,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.82e-05,
+      "loss": 1.2862,
+      "step": 2205
+    },
+    {
+      "epoch": 0.0442,
+      "grad_norm": 0.318359375,
+      "learning_rate": 8.840000000000001e-05,
+      "loss": 1.2711,
+      "step": 2210
+    },
+    {
+      "epoch": 0.0443,
+      "grad_norm": 0.240234375,
+      "learning_rate": 8.86e-05,
+      "loss": 1.2942,
+      "step": 2215
+    },
+    {
+      "epoch": 0.0444,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 8.88e-05,
+      "loss": 1.2756,
+      "step": 2220
+    },
+    {
+      "epoch": 0.0445,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.900000000000001e-05,
+      "loss": 1.2685,
+      "step": 2225
+    },
+    {
+      "epoch": 0.0446,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.92e-05,
+      "loss": 1.2933,
+      "step": 2230
+    },
+    {
+      "epoch": 0.0447,
+      "grad_norm": 0.263671875,
+      "learning_rate": 8.94e-05,
+      "loss": 1.2905,
+      "step": 2235
+    },
+    {
+      "epoch": 0.0448,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 8.960000000000001e-05,
+      "loss": 1.2822,
+      "step": 2240
+    },
+    {
+      "epoch": 0.0449,
+      "grad_norm": 0.1484375,
+      "learning_rate": 8.98e-05,
+      "loss": 1.267,
+      "step": 2245
+    },
+    {
+      "epoch": 0.045,
+      "grad_norm": 0.1796875,
+      "learning_rate": 9e-05,
+      "loss": 1.2686,
+      "step": 2250
+    },
+    {
+      "epoch": 0.0451,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 9.020000000000001e-05,
+      "loss": 1.2836,
+      "step": 2255
+    },
+    {
+      "epoch": 0.0452,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.04e-05,
+      "loss": 1.2879,
+      "step": 2260
+    },
+    {
+      "epoch": 0.0453,
+      "grad_norm": 0.41015625,
+      "learning_rate": 9.06e-05,
+      "loss": 1.2788,
+      "step": 2265
+    },
+    {
+      "epoch": 0.0454,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.080000000000001e-05,
+      "loss": 1.2888,
+      "step": 2270
+    },
+    {
+      "epoch": 0.0455,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.1e-05,
+      "loss": 1.2712,
+      "step": 2275
+    },
+    {
+      "epoch": 0.0456,
+      "grad_norm": 0.205078125,
+      "learning_rate": 9.120000000000001e-05,
+      "loss": 1.2817,
+      "step": 2280
+    },
+    {
+      "epoch": 0.0457,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.140000000000001e-05,
+      "loss": 1.2776,
+      "step": 2285
+    },
+    {
+      "epoch": 0.0458,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.16e-05,
+      "loss": 1.2704,
+      "step": 2290
+    },
+    {
+      "epoch": 0.0459,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 9.180000000000001e-05,
+      "loss": 1.2827,
+      "step": 2295
+    },
+    {
+      "epoch": 0.046,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.200000000000001e-05,
+      "loss": 1.2775,
+      "step": 2300
+    },
+    {
+      "epoch": 0.0461,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.22e-05,
+      "loss": 1.2812,
+      "step": 2305
+    },
+    {
+      "epoch": 0.0462,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.240000000000001e-05,
+      "loss": 1.2747,
+      "step": 2310
+    },
+    {
+      "epoch": 0.0463,
+      "grad_norm": 0.1875,
+      "learning_rate": 9.260000000000001e-05,
+      "loss": 1.2777,
+      "step": 2315
+    },
+    {
+      "epoch": 0.0464,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 9.28e-05,
+      "loss": 1.2791,
+      "step": 2320
+    },
+    {
+      "epoch": 0.0465,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.300000000000001e-05,
+      "loss": 1.2757,
+      "step": 2325
+    },
+    {
+      "epoch": 0.0466,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 9.320000000000002e-05,
+      "loss": 1.283,
+      "step": 2330
+    },
+    {
+      "epoch": 0.0467,
+      "grad_norm": 0.234375,
+      "learning_rate": 9.340000000000001e-05,
+      "loss": 1.2982,
+      "step": 2335
+    },
+    {
+      "epoch": 0.0468,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 9.360000000000001e-05,
+      "loss": 1.2738,
+      "step": 2340
+    },
+    {
+      "epoch": 0.0469,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.38e-05,
+      "loss": 1.2952,
+      "step": 2345
+    },
+    {
+      "epoch": 0.047,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 9.4e-05,
+      "loss": 1.2766,
+      "step": 2350
+    },
+    {
+      "epoch": 0.0471,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.42e-05,
+      "loss": 1.2716,
+      "step": 2355
+    },
+    {
+      "epoch": 0.0472,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.44e-05,
+      "loss": 1.283,
+      "step": 2360
+    },
+    {
+      "epoch": 0.0473,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.46e-05,
+      "loss": 1.2843,
+      "step": 2365
+    },
+    {
+      "epoch": 0.0474,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.48e-05,
+      "loss": 1.2879,
+      "step": 2370
+    },
+    {
+      "epoch": 0.0475,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.5e-05,
+      "loss": 1.272,
+      "step": 2375
+    },
+    {
+      "epoch": 0.0476,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.52e-05,
+      "loss": 1.2869,
+      "step": 2380
+    },
+    {
+      "epoch": 0.0477,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 9.54e-05,
+      "loss": 1.2973,
+      "step": 2385
+    },
+    {
+      "epoch": 0.0478,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 9.56e-05,
+      "loss": 1.275,
+      "step": 2390
+    },
+    {
+      "epoch": 0.0479,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.58e-05,
+      "loss": 1.2778,
+      "step": 2395
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.6e-05,
+      "loss": 1.2664,
+      "step": 2400
+    },
+    {
+      "epoch": 0.0481,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.620000000000001e-05,
+      "loss": 1.2818,
+      "step": 2405
+    },
+    {
+      "epoch": 0.0482,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.64e-05,
+      "loss": 1.2936,
+      "step": 2410
+    },
+    {
+      "epoch": 0.0483,
+      "grad_norm": 0.185546875,
+      "learning_rate": 9.66e-05,
+      "loss": 1.2726,
+      "step": 2415
+    },
+    {
+      "epoch": 0.0484,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.680000000000001e-05,
+      "loss": 1.2693,
+      "step": 2420
+    },
+    {
+      "epoch": 0.0485,
+      "grad_norm": 0.373046875,
+      "learning_rate": 9.7e-05,
+      "loss": 1.2985,
+      "step": 2425
+    },
+    {
+      "epoch": 0.0486,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.72e-05,
+      "loss": 1.2828,
+      "step": 2430
+    },
+    {
+      "epoch": 0.0487,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.74e-05,
+      "loss": 1.2789,
+      "step": 2435
+    },
+    {
+      "epoch": 0.0488,
+      "grad_norm": 0.388671875,
+      "learning_rate": 9.76e-05,
+      "loss": 1.2814,
+      "step": 2440
+    },
+    {
+      "epoch": 0.0489,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.78e-05,
+      "loss": 1.2816,
+      "step": 2445
+    },
+    {
+      "epoch": 0.049,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 9.8e-05,
+      "loss": 1.2694,
+      "step": 2450
+    },
+    {
+      "epoch": 0.0491,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.82e-05,
+      "loss": 1.3064,
+      "step": 2455
+    },
+    {
+      "epoch": 0.0492,
+      "grad_norm": 0.38671875,
+      "learning_rate": 9.84e-05,
+      "loss": 1.2972,
+      "step": 2460
+    },
+    {
+      "epoch": 0.0493,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.86e-05,
+      "loss": 1.2906,
+      "step": 2465
+    },
+    {
+      "epoch": 0.0494,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.88e-05,
+      "loss": 1.2739,
+      "step": 2470
+    },
+    {
+      "epoch": 0.0495,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.900000000000001e-05,
+      "loss": 1.271,
+      "step": 2475
+    },
+    {
+      "epoch": 0.0496,
+      "grad_norm": 0.228515625,
+      "learning_rate": 9.92e-05,
+      "loss": 1.2956,
+      "step": 2480
+    },
+    {
+      "epoch": 0.0497,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.94e-05,
+      "loss": 1.2795,
+      "step": 2485
+    },
+    {
+      "epoch": 0.0498,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 9.960000000000001e-05,
+      "loss": 1.2661,
+      "step": 2490
+    },
+    {
+      "epoch": 0.0499,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.98e-05,
+      "loss": 1.2855,
+      "step": 2495
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.0001,
+      "loss": 1.2797,
+      "step": 2500
+    },
+    {
+      "epoch": 0.0501,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00010020000000000001,
+      "loss": 1.304,
+      "step": 2505
+    },
+    {
+      "epoch": 0.0502,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001004,
+      "loss": 1.2935,
+      "step": 2510
+    },
+    {
+      "epoch": 0.0503,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0001006,
+      "loss": 1.2863,
+      "step": 2515
+    },
+    {
+      "epoch": 0.0504,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00010080000000000001,
+      "loss": 1.2674,
+      "step": 2520
+    },
+    {
+      "epoch": 0.0505,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.000101,
+      "loss": 1.2854,
+      "step": 2525
+    },
+    {
+      "epoch": 0.0506,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00010120000000000001,
+      "loss": 1.2641,
+      "step": 2530
+    },
+    {
+      "epoch": 0.0507,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00010140000000000001,
+      "loss": 1.2872,
+      "step": 2535
+    },
+    {
+      "epoch": 0.0508,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0001016,
+      "loss": 1.2833,
+      "step": 2540
+    },
+    {
+      "epoch": 0.0509,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00010180000000000001,
+      "loss": 1.289,
+      "step": 2545
+    },
+    {
+      "epoch": 0.051,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00010200000000000001,
+      "loss": 1.2814,
+      "step": 2550
+    },
+    {
+      "epoch": 0.0511,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0001022,
+      "loss": 1.2742,
+      "step": 2555
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00010240000000000001,
+      "loss": 1.2766,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0513,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00010260000000000001,
+      "loss": 1.2922,
+      "step": 2565
+    },
+    {
+      "epoch": 0.0514,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001028,
+      "loss": 1.2798,
+      "step": 2570
+    },
+    {
+      "epoch": 0.0515,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010300000000000001,
+      "loss": 1.2708,
+      "step": 2575
+    },
+    {
+      "epoch": 0.0516,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001032,
+      "loss": 1.2895,
+      "step": 2580
+    },
+    {
+      "epoch": 0.0517,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001034,
+      "loss": 1.2935,
+      "step": 2585
+    },
+    {
+      "epoch": 0.0518,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00010360000000000001,
+      "loss": 1.2739,
+      "step": 2590
+    },
+    {
+      "epoch": 0.0519,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001038,
+      "loss": 1.2771,
+      "step": 2595
+    },
+    {
+      "epoch": 0.052,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00010400000000000001,
+      "loss": 1.2755,
+      "step": 2600
+    },
+    {
+      "epoch": 0.0521,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00010420000000000001,
+      "loss": 1.2746,
+      "step": 2605
+    },
+    {
+      "epoch": 0.0522,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001044,
+      "loss": 1.281,
+      "step": 2610
+    },
+    {
+      "epoch": 0.0523,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00010460000000000001,
+      "loss": 1.2849,
+      "step": 2615
+    },
+    {
+      "epoch": 0.0524,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00010480000000000001,
+      "loss": 1.2987,
+      "step": 2620
+    },
+    {
+      "epoch": 0.0525,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.000105,
+      "loss": 1.3004,
+      "step": 2625
+    },
+    {
+      "epoch": 0.0526,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00010520000000000001,
+      "loss": 1.2926,
+      "step": 2630
+    },
+    {
+      "epoch": 0.0527,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00010540000000000001,
+      "loss": 1.2872,
+      "step": 2635
+    },
+    {
+      "epoch": 0.0528,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001056,
+      "loss": 1.2765,
+      "step": 2640
+    },
+    {
+      "epoch": 0.0529,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00010580000000000001,
+      "loss": 1.2818,
+      "step": 2645
+    },
+    {
+      "epoch": 0.053,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00010600000000000002,
+      "loss": 1.275,
+      "step": 2650
+    },
+    {
+      "epoch": 0.0531,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001062,
+      "loss": 1.28,
+      "step": 2655
+    },
+    {
+      "epoch": 0.0532,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00010640000000000001,
+      "loss": 1.2956,
+      "step": 2660
+    },
+    {
+      "epoch": 0.0533,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00010660000000000002,
+      "loss": 1.2845,
+      "step": 2665
+    },
+    {
+      "epoch": 0.0534,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00010680000000000001,
+      "loss": 1.2791,
+      "step": 2670
+    },
+    {
+      "epoch": 0.0535,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00010700000000000001,
+      "loss": 1.271,
+      "step": 2675
+    },
+    {
+      "epoch": 0.0536,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00010720000000000002,
+      "loss": 1.2772,
+      "step": 2680
+    },
+    {
+      "epoch": 0.0537,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00010740000000000001,
+      "loss": 1.2735,
+      "step": 2685
+    },
+    {
+      "epoch": 0.0538,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00010760000000000001,
+      "loss": 1.2682,
+      "step": 2690
+    },
+    {
+      "epoch": 0.0539,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00010780000000000002,
+      "loss": 1.2632,
+      "step": 2695
+    },
+    {
+      "epoch": 0.054,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00010800000000000001,
+      "loss": 1.2783,
+      "step": 2700
+    },
+    {
+      "epoch": 0.0541,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00010820000000000001,
+      "loss": 1.2639,
+      "step": 2705
+    },
+    {
+      "epoch": 0.0542,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00010840000000000002,
+      "loss": 1.2868,
+      "step": 2710
+    },
+    {
+      "epoch": 0.0543,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00010860000000000001,
+      "loss": 1.2721,
+      "step": 2715
+    },
+    {
+      "epoch": 0.0544,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00010880000000000002,
+      "loss": 1.289,
+      "step": 2720
+    },
+    {
+      "epoch": 0.0545,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.000109,
+      "loss": 1.2817,
+      "step": 2725
+    },
+    {
+      "epoch": 0.0546,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00010920000000000001,
+      "loss": 1.262,
+      "step": 2730
+    },
+    {
+      "epoch": 0.0547,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00010940000000000002,
+      "loss": 1.2844,
+      "step": 2735
+    },
+    {
+      "epoch": 0.0548,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00010960000000000001,
+      "loss": 1.2825,
+      "step": 2740
+    },
+    {
+      "epoch": 0.0549,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00010980000000000001,
+      "loss": 1.2804,
+      "step": 2745
+    },
+    {
+      "epoch": 0.055,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00011000000000000002,
+      "loss": 1.2852,
+      "step": 2750
+    },
+    {
+      "epoch": 0.0551,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00011020000000000001,
+      "loss": 1.285,
+      "step": 2755
+    },
+    {
+      "epoch": 0.0552,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00011040000000000001,
+      "loss": 1.2669,
+      "step": 2760
+    },
+    {
+      "epoch": 0.0553,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00011060000000000002,
+      "loss": 1.2927,
+      "step": 2765
+    },
+    {
+      "epoch": 0.0554,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00011080000000000001,
+      "loss": 1.2847,
+      "step": 2770
+    },
+    {
+      "epoch": 0.0555,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00011100000000000001,
+      "loss": 1.2758,
+      "step": 2775
+    },
+    {
+      "epoch": 0.0556,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00011120000000000002,
+      "loss": 1.2941,
+      "step": 2780
+    },
+    {
+      "epoch": 0.0557,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00011140000000000001,
+      "loss": 1.2724,
+      "step": 2785
+    },
+    {
+      "epoch": 0.0558,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00011160000000000002,
+      "loss": 1.2733,
+      "step": 2790
+    },
+    {
+      "epoch": 0.0559,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00011180000000000002,
+      "loss": 1.2709,
+      "step": 2795
+    },
+    {
+      "epoch": 0.056,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00011200000000000001,
+      "loss": 1.281,
+      "step": 2800
+    },
+    {
+      "epoch": 0.0561,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00011220000000000002,
+      "loss": 1.299,
+      "step": 2805
+    },
+    {
+      "epoch": 0.0562,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011240000000000002,
+      "loss": 1.2688,
+      "step": 2810
+    },
+    {
+      "epoch": 0.0563,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0001126,
+      "loss": 1.2802,
+      "step": 2815
+    },
+    {
+      "epoch": 0.0564,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00011279999999999999,
+      "loss": 1.2699,
+      "step": 2820
+    },
+    {
+      "epoch": 0.0565,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.000113,
+      "loss": 1.2626,
+      "step": 2825
+    },
+    {
+      "epoch": 0.0566,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001132,
+      "loss": 1.2685,
+      "step": 2830
+    },
+    {
+      "epoch": 0.0567,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00011339999999999999,
+      "loss": 1.2698,
+      "step": 2835
+    },
+    {
+      "epoch": 0.0568,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0001136,
+      "loss": 1.2864,
+      "step": 2840
+    },
+    {
+      "epoch": 0.0569,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001138,
+      "loss": 1.2949,
+      "step": 2845
+    },
+    {
+      "epoch": 0.057,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00011399999999999999,
+      "loss": 1.2669,
+      "step": 2850
+    },
+    {
+      "epoch": 0.0571,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001142,
+      "loss": 1.2674,
+      "step": 2855
+    },
+    {
+      "epoch": 0.0572,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0001144,
+      "loss": 1.2775,
+      "step": 2860
+    },
+    {
+      "epoch": 0.0573,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0001146,
+      "loss": 1.2799,
+      "step": 2865
+    },
+    {
+      "epoch": 0.0574,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001148,
+      "loss": 1.275,
+      "step": 2870
+    },
+    {
+      "epoch": 0.0575,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00011499999999999999,
+      "loss": 1.2862,
+      "step": 2875
+    },
+    {
+      "epoch": 0.0576,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001152,
+      "loss": 1.2734,
+      "step": 2880
+    },
+    {
+      "epoch": 0.0577,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0001154,
+      "loss": 1.275,
+      "step": 2885
+    },
+    {
+      "epoch": 0.0578,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00011559999999999999,
+      "loss": 1.2841,
+      "step": 2890
+    },
+    {
+      "epoch": 0.0579,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001158,
+      "loss": 1.2757,
+      "step": 2895
+    },
+    {
+      "epoch": 0.058,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.000116,
+      "loss": 1.27,
+      "step": 2900
+    },
+    {
+      "epoch": 0.0581,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00011619999999999999,
+      "loss": 1.283,
+      "step": 2905
+    },
+    {
+      "epoch": 0.0582,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001164,
+      "loss": 1.2681,
+      "step": 2910
+    },
+    {
+      "epoch": 0.0583,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001166,
+      "loss": 1.2866,
+      "step": 2915
+    },
+    {
+      "epoch": 0.0584,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00011679999999999999,
+      "loss": 1.2932,
+      "step": 2920
+    },
+    {
+      "epoch": 0.0585,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.000117,
+      "loss": 1.2683,
+      "step": 2925
+    },
+    {
+      "epoch": 0.0586,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001172,
+      "loss": 1.2831,
+      "step": 2930
+    },
+    {
+      "epoch": 0.0587,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001174,
+      "loss": 1.2847,
+      "step": 2935
+    },
+    {
+      "epoch": 0.0588,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001176,
+      "loss": 1.2771,
+      "step": 2940
+    },
+    {
+      "epoch": 0.0589,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001178,
+      "loss": 1.2951,
+      "step": 2945
+    },
+    {
+      "epoch": 0.059,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.000118,
+      "loss": 1.2725,
+      "step": 2950
+    },
+    {
+      "epoch": 0.0591,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001182,
+      "loss": 1.276,
+      "step": 2955
+    },
+    {
+      "epoch": 0.0592,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001184,
+      "loss": 1.3125,
+      "step": 2960
+    },
+    {
+      "epoch": 0.0593,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001186,
+      "loss": 1.2838,
+      "step": 2965
+    },
+    {
+      "epoch": 0.0594,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0001188,
+      "loss": 1.2778,
+      "step": 2970
+    },
+    {
+      "epoch": 0.0595,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.000119,
+      "loss": 1.2741,
+      "step": 2975
+    },
+    {
+      "epoch": 0.0596,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001192,
+      "loss": 1.265,
+      "step": 2980
+    },
+    {
+      "epoch": 0.0597,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001194,
+      "loss": 1.2799,
+      "step": 2985
+    },
+    {
+      "epoch": 0.0598,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00011960000000000001,
+      "loss": 1.2964,
+      "step": 2990
+    },
+    {
+      "epoch": 0.0599,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001198,
+      "loss": 1.2637,
+      "step": 2995
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012,
+      "loss": 1.2762,
+      "step": 3000
+    },
+    {
+      "epoch": 0.0601,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012020000000000001,
+      "loss": 1.2836,
+      "step": 3005
+    },
+    {
+      "epoch": 0.0602,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001204,
+      "loss": 1.2892,
+      "step": 3010
+    },
+    {
+      "epoch": 0.0603,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001206,
+      "loss": 1.2862,
+      "step": 3015
+    },
+    {
+      "epoch": 0.0604,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001208,
+      "loss": 1.2955,
+      "step": 3020
+    },
+    {
+      "epoch": 0.0605,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.000121,
+      "loss": 1.2829,
+      "step": 3025
+    },
+    {
+      "epoch": 0.0606,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0001212,
+      "loss": 1.2846,
+      "step": 3030
+    },
+    {
+      "epoch": 0.0607,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0001214,
+      "loss": 1.2637,
+      "step": 3035
+    },
+    {
+      "epoch": 0.0608,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001216,
+      "loss": 1.2641,
+      "step": 3040
+    },
+    {
+      "epoch": 0.0609,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001218,
+      "loss": 1.2698,
+      "step": 3045
+    },
+    {
+      "epoch": 0.061,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.000122,
+      "loss": 1.292,
+      "step": 3050
+    },
+    {
+      "epoch": 0.0611,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00012220000000000002,
+      "loss": 1.2716,
+      "step": 3055
+    },
+    {
+      "epoch": 0.0612,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001224,
+      "loss": 1.2782,
+      "step": 3060
+    },
+    {
+      "epoch": 0.0613,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001226,
+      "loss": 1.2876,
+      "step": 3065
+    },
+    {
+      "epoch": 0.0614,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0001228,
+      "loss": 1.2615,
+      "step": 3070
+    },
+    {
+      "epoch": 0.0615,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.000123,
+      "loss": 1.2675,
+      "step": 3075
+    },
+    {
+      "epoch": 0.0616,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001232,
+      "loss": 1.2561,
+      "step": 3080
+    },
+    {
+      "epoch": 0.0617,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00012340000000000002,
+      "loss": 1.2833,
+      "step": 3085
+    },
+    {
+      "epoch": 0.0618,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0001236,
+      "loss": 1.2941,
+      "step": 3090
+    },
+    {
+      "epoch": 0.0619,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001238,
+      "loss": 1.2875,
+      "step": 3095
+    },
+    {
+      "epoch": 0.062,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.000124,
+      "loss": 1.2806,
+      "step": 3100
+    },
+    {
+      "epoch": 0.0621,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001242,
+      "loss": 1.2979,
+      "step": 3105
+    },
+    {
+      "epoch": 0.0622,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00012440000000000002,
+      "loss": 1.2711,
+      "step": 3110
+    },
+    {
+      "epoch": 0.0623,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001246,
+      "loss": 1.2803,
+      "step": 3115
+    },
+    {
+      "epoch": 0.0624,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001248,
+      "loss": 1.2727,
+      "step": 3120
+    },
+    {
+      "epoch": 0.0625,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.000125,
+      "loss": 1.2779,
+      "step": 3125
+    },
+    {
+      "epoch": 0.0626,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001252,
+      "loss": 1.277,
+      "step": 3130
+    },
+    {
+      "epoch": 0.0627,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001254,
+      "loss": 1.2787,
+      "step": 3135
+    },
+    {
+      "epoch": 0.0628,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00012560000000000002,
+      "loss": 1.2717,
+      "step": 3140
+    },
+    {
+      "epoch": 0.0629,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001258,
+      "loss": 1.282,
+      "step": 3145
+    },
+    {
+      "epoch": 0.063,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.000126,
+      "loss": 1.2833,
+      "step": 3150
+    },
+    {
+      "epoch": 0.0631,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001262,
+      "loss": 1.2869,
+      "step": 3155
+    },
+    {
+      "epoch": 0.0632,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0001264,
+      "loss": 1.2877,
+      "step": 3160
+    },
+    {
+      "epoch": 0.0633,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00012660000000000001,
+      "loss": 1.2695,
+      "step": 3165
+    },
+    {
+      "epoch": 0.0634,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00012680000000000002,
+      "loss": 1.2643,
+      "step": 3170
+    },
+    {
+      "epoch": 0.0635,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.000127,
+      "loss": 1.2716,
+      "step": 3175
+    },
+    {
+      "epoch": 0.0636,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001272,
+      "loss": 1.2685,
+      "step": 3180
+    },
+    {
+      "epoch": 0.0637,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001274,
+      "loss": 1.2769,
+      "step": 3185
+    },
+    {
+      "epoch": 0.0638,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001276,
+      "loss": 1.2828,
+      "step": 3190
+    },
+    {
+      "epoch": 0.0639,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012780000000000002,
+      "loss": 1.2563,
+      "step": 3195
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00012800000000000002,
+      "loss": 1.2637,
+      "step": 3200
+    },
+    {
+      "epoch": 0.0641,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001282,
+      "loss": 1.2726,
+      "step": 3205
+    },
+    {
+      "epoch": 0.0642,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001284,
+      "loss": 1.2784,
+      "step": 3210
+    },
+    {
+      "epoch": 0.0643,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0001286,
+      "loss": 1.3122,
+      "step": 3215
+    },
+    {
+      "epoch": 0.0644,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00012880000000000001,
+      "loss": 1.2951,
+      "step": 3220
+    },
+    {
+      "epoch": 0.0645,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00012900000000000002,
+      "loss": 1.2702,
+      "step": 3225
+    },
+    {
+      "epoch": 0.0646,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00012920000000000002,
+      "loss": 1.2813,
+      "step": 3230
+    },
+    {
+      "epoch": 0.0647,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001294,
+      "loss": 1.2887,
+      "step": 3235
+    },
+    {
+      "epoch": 0.0648,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001296,
+      "loss": 1.2634,
+      "step": 3240
+    },
+    {
+      "epoch": 0.0649,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001298,
+      "loss": 1.2839,
+      "step": 3245
+    },
+    {
+      "epoch": 0.065,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00013000000000000002,
+      "loss": 1.2974,
+      "step": 3250
+    },
+    {
+      "epoch": 0.0651,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00013020000000000002,
+      "loss": 1.2758,
+      "step": 3255
+    },
+    {
+      "epoch": 0.0652,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001304,
+      "loss": 1.2724,
+      "step": 3260
+    },
+    {
+      "epoch": 0.0653,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001306,
+      "loss": 1.2622,
+      "step": 3265
+    },
+    {
+      "epoch": 0.0654,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001308,
+      "loss": 1.286,
+      "step": 3270
+    },
+    {
+      "epoch": 0.0655,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.000131,
+      "loss": 1.2736,
+      "step": 3275
+    },
+    {
+      "epoch": 0.0656,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00013120000000000002,
+      "loss": 1.2632,
+      "step": 3280
+    },
+    {
+      "epoch": 0.0657,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00013140000000000002,
+      "loss": 1.2841,
+      "step": 3285
+    },
+    {
+      "epoch": 0.0658,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001316,
+      "loss": 1.2578,
+      "step": 3290
+    },
+    {
+      "epoch": 0.0659,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0001318,
+      "loss": 1.2898,
+      "step": 3295
+    },
+    {
+      "epoch": 0.066,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.000132,
+      "loss": 1.2895,
+      "step": 3300
+    },
+    {
+      "epoch": 0.0661,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00013220000000000001,
+      "loss": 1.2728,
+      "step": 3305
+    },
+    {
+      "epoch": 0.0662,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00013240000000000002,
+      "loss": 1.2805,
+      "step": 3310
+    },
+    {
+      "epoch": 0.0663,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00013260000000000002,
+      "loss": 1.2796,
+      "step": 3315
+    },
+    {
+      "epoch": 0.0664,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001328,
+      "loss": 1.2708,
+      "step": 3320
+    },
+    {
+      "epoch": 0.0665,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.000133,
+      "loss": 1.2699,
+      "step": 3325
+    },
+    {
+      "epoch": 0.0666,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0001332,
+      "loss": 1.2839,
+      "step": 3330
+    },
+    {
+      "epoch": 0.0667,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00013340000000000002,
+      "loss": 1.2596,
+      "step": 3335
+    },
+    {
+      "epoch": 0.0668,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00013360000000000002,
+      "loss": 1.2748,
+      "step": 3340
+    },
+    {
+      "epoch": 0.0669,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00013380000000000003,
+      "loss": 1.3046,
+      "step": 3345
+    },
+    {
+      "epoch": 0.067,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.000134,
+      "loss": 1.2784,
+      "step": 3350
+    },
+    {
+      "epoch": 0.0671,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001342,
+      "loss": 1.2756,
+      "step": 3355
+    },
+    {
+      "epoch": 0.0672,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00013440000000000001,
+      "loss": 1.2623,
+      "step": 3360
+    },
+    {
+      "epoch": 0.0673,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00013460000000000002,
+      "loss": 1.2779,
+      "step": 3365
+    },
+    {
+      "epoch": 0.0674,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00013480000000000002,
+      "loss": 1.2767,
+      "step": 3370
+    },
+    {
+      "epoch": 0.0675,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00013500000000000003,
+      "loss": 1.2729,
+      "step": 3375
+    },
+    {
+      "epoch": 0.0676,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001352,
+      "loss": 1.2751,
+      "step": 3380
+    },
+    {
+      "epoch": 0.0677,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001354,
+      "loss": 1.268,
+      "step": 3385
+    },
+    {
+      "epoch": 0.0678,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00013560000000000002,
+      "loss": 1.2826,
+      "step": 3390
+    },
+    {
+      "epoch": 0.0679,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00013580000000000002,
+      "loss": 1.2885,
+      "step": 3395
+    },
+    {
+      "epoch": 0.068,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00013600000000000003,
+      "loss": 1.2881,
+      "step": 3400
+    },
+    {
+      "epoch": 0.0681,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001362,
+      "loss": 1.2674,
+      "step": 3405
+    },
+    {
+      "epoch": 0.0682,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0001364,
+      "loss": 1.277,
+      "step": 3410
+    },
+    {
+      "epoch": 0.0683,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0001366,
+      "loss": 1.2776,
+      "step": 3415
+    },
+    {
+      "epoch": 0.0684,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00013680000000000002,
+      "loss": 1.2903,
+      "step": 3420
+    },
+    {
+      "epoch": 0.0685,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00013700000000000002,
+      "loss": 1.2841,
+      "step": 3425
+    },
+    {
+      "epoch": 0.0686,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00013720000000000003,
+      "loss": 1.2874,
+      "step": 3430
+    },
+    {
+      "epoch": 0.0687,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001374,
+      "loss": 1.2798,
+      "step": 3435
+    },
+    {
+      "epoch": 0.0688,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00013759999999999998,
+      "loss": 1.2714,
+      "step": 3440
+    },
+    {
+      "epoch": 0.0689,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001378,
+      "loss": 1.2897,
+      "step": 3445
+    },
+    {
+      "epoch": 0.069,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.000138,
+      "loss": 1.2854,
+      "step": 3450
+    },
+    {
+      "epoch": 0.0691,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0001382,
+      "loss": 1.2775,
+      "step": 3455
+    },
+    {
+      "epoch": 0.0692,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001384,
+      "loss": 1.2804,
+      "step": 3460
+    },
+    {
+      "epoch": 0.0693,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.0001386,
+      "loss": 1.2688,
+      "step": 3465
+    },
+    {
+      "epoch": 0.0694,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00013879999999999999,
+      "loss": 1.3098,
+      "step": 3470
+    },
+    {
+      "epoch": 0.0695,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.000139,
+      "loss": 1.2832,
+      "step": 3475
+    },
+    {
+      "epoch": 0.0696,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001392,
+      "loss": 1.2914,
+      "step": 3480
+    },
+    {
+      "epoch": 0.0697,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001394,
+      "loss": 1.2872,
+      "step": 3485
+    },
+    {
+      "epoch": 0.0698,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0001396,
+      "loss": 1.2813,
+      "step": 3490
+    },
+    {
+      "epoch": 0.0699,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001398,
+      "loss": 1.2778,
+      "step": 3495
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00014,
+      "loss": 1.2745,
+      "step": 3500
+    },
+    {
+      "epoch": 0.0701,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001402,
+      "loss": 1.2679,
+      "step": 3505
+    },
+    {
+      "epoch": 0.0702,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001404,
+      "loss": 1.2903,
+      "step": 3510
+    },
+    {
+      "epoch": 0.0703,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0001406,
+      "loss": 1.2763,
+      "step": 3515
+    },
+    {
+      "epoch": 0.0704,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001408,
+      "loss": 1.2933,
+      "step": 3520
+    },
+    {
+      "epoch": 0.0705,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.000141,
+      "loss": 1.2776,
+      "step": 3525
+    },
+    {
+      "epoch": 0.0706,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0001412,
+      "loss": 1.2795,
+      "step": 3530
+    },
+    {
+      "epoch": 0.0707,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001414,
+      "loss": 1.292,
+      "step": 3535
+    },
+    {
+      "epoch": 0.0708,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001416,
+      "loss": 1.2867,
+      "step": 3540
+    },
+    {
+      "epoch": 0.0709,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001418,
+      "loss": 1.2698,
+      "step": 3545
+    },
+    {
+      "epoch": 0.071,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.000142,
+      "loss": 1.2749,
+      "step": 3550
+    },
+    {
+      "epoch": 0.0711,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001422,
+      "loss": 1.2671,
+      "step": 3555
+    },
+    {
+      "epoch": 0.0712,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001424,
+      "loss": 1.277,
+      "step": 3560
+    },
+    {
+      "epoch": 0.0713,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001426,
+      "loss": 1.2791,
+      "step": 3565
+    },
+    {
+      "epoch": 0.0714,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001428,
+      "loss": 1.2649,
+      "step": 3570
+    },
+    {
+      "epoch": 0.0715,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.000143,
+      "loss": 1.2789,
+      "step": 3575
+    },
+    {
+      "epoch": 0.0716,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001432,
+      "loss": 1.2795,
+      "step": 3580
+    },
+    {
+      "epoch": 0.0717,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001434,
+      "loss": 1.258,
+      "step": 3585
+    },
+    {
+      "epoch": 0.0718,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001436,
+      "loss": 1.2658,
+      "step": 3590
+    },
+    {
+      "epoch": 0.0719,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001438,
+      "loss": 1.2891,
+      "step": 3595
+    },
+    {
+      "epoch": 0.072,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.000144,
+      "loss": 1.2695,
+      "step": 3600
+    },
+    {
+      "epoch": 0.0721,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001442,
+      "loss": 1.2815,
+      "step": 3605
+    },
+    {
+      "epoch": 0.0722,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001444,
+      "loss": 1.2628,
+      "step": 3610
+    },
+    {
+      "epoch": 0.0723,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0001446,
+      "loss": 1.2826,
+      "step": 3615
+    },
+    {
+      "epoch": 0.0724,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001448,
+      "loss": 1.2797,
+      "step": 3620
+    },
+    {
+      "epoch": 0.0725,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.000145,
+      "loss": 1.2826,
+      "step": 3625
+    },
+    {
+      "epoch": 0.0726,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001452,
+      "loss": 1.2728,
+      "step": 3630
+    },
+    {
+      "epoch": 0.0727,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001454,
+      "loss": 1.269,
+      "step": 3635
+    },
+    {
+      "epoch": 0.0728,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00014560000000000002,
+      "loss": 1.27,
+      "step": 3640
+    },
+    {
+      "epoch": 0.0729,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001458,
+      "loss": 1.2683,
+      "step": 3645
+    },
+    {
+      "epoch": 0.073,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.000146,
+      "loss": 1.3017,
+      "step": 3650
+    },
+    {
+      "epoch": 0.0731,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001462,
+      "loss": 1.2929,
+      "step": 3655
+    },
+    {
+      "epoch": 0.0732,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0001464,
+      "loss": 1.2691,
+      "step": 3660
+    },
+    {
+      "epoch": 0.0733,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001466,
+      "loss": 1.2711,
+      "step": 3665
+    },
+    {
+      "epoch": 0.0734,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00014680000000000002,
+      "loss": 1.2955,
+      "step": 3670
+    },
+    {
+      "epoch": 0.0735,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.000147,
+      "loss": 1.2901,
+      "step": 3675
+    },
+    {
+      "epoch": 0.0736,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001472,
+      "loss": 1.276,
+      "step": 3680
+    },
+    {
+      "epoch": 0.0737,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001474,
+      "loss": 1.2773,
+      "step": 3685
+    },
+    {
+      "epoch": 0.0738,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001476,
+      "loss": 1.2751,
+      "step": 3690
+    },
+    {
+      "epoch": 0.0739,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00014780000000000001,
+      "loss": 1.2721,
+      "step": 3695
+    },
+    {
+      "epoch": 0.074,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.000148,
+      "loss": 1.2748,
+      "step": 3700
+    },
+    {
+      "epoch": 0.0741,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001482,
+      "loss": 1.2626,
+      "step": 3705
+    },
+    {
+      "epoch": 0.0742,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001484,
+      "loss": 1.2844,
+      "step": 3710
+    },
+    {
+      "epoch": 0.0743,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001486,
+      "loss": 1.2786,
+      "step": 3715
+    },
+    {
+      "epoch": 0.0744,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001488,
+      "loss": 1.281,
+      "step": 3720
+    },
+    {
+      "epoch": 0.0745,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00014900000000000002,
+      "loss": 1.2755,
+      "step": 3725
+    },
+    {
+      "epoch": 0.0746,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001492,
+      "loss": 1.2749,
+      "step": 3730
+    },
+    {
+      "epoch": 0.0747,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001494,
+      "loss": 1.2824,
+      "step": 3735
+    },
+    {
+      "epoch": 0.0748,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001496,
+      "loss": 1.2868,
+      "step": 3740
+    },
+    {
+      "epoch": 0.0749,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0001498,
+      "loss": 1.2897,
+      "step": 3745
+    },
+    {
+      "epoch": 0.075,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 1.2742,
+      "step": 3750
+    },
+    {
+      "epoch": 0.0751,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00015020000000000002,
+      "loss": 1.2728,
+      "step": 3755
+    },
+    {
+      "epoch": 0.0752,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001504,
+      "loss": 1.2739,
+      "step": 3760
+    },
+    {
+      "epoch": 0.0753,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0001506,
+      "loss": 1.2755,
+      "step": 3765
+    },
+    {
+      "epoch": 0.0754,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0001508,
+      "loss": 1.2716,
+      "step": 3770
+    },
+    {
+      "epoch": 0.0755,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.000151,
+      "loss": 1.2817,
+      "step": 3775
+    },
+    {
+      "epoch": 0.0756,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00015120000000000002,
+      "loss": 1.2851,
+      "step": 3780
+    },
+    {
+      "epoch": 0.0757,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00015140000000000002,
+      "loss": 1.2878,
+      "step": 3785
+    },
+    {
+      "epoch": 0.0758,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001516,
+      "loss": 1.2881,
+      "step": 3790
+    },
+    {
+      "epoch": 0.0759,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001518,
+      "loss": 1.2769,
+      "step": 3795
+    },
+    {
+      "epoch": 0.076,
+      "grad_norm": 0.75,
+      "learning_rate": 0.000152,
+      "loss": 1.274,
+      "step": 3800
+    },
+    {
+      "epoch": 0.0761,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0001522,
+      "loss": 1.3003,
+      "step": 3805
+    },
+    {
+      "epoch": 0.0762,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00015240000000000002,
+      "loss": 1.2774,
+      "step": 3810
+    },
+    {
+      "epoch": 0.0763,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00015260000000000002,
+      "loss": 1.2687,
+      "step": 3815
+    },
+    {
+      "epoch": 0.0764,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001528,
+      "loss": 1.2712,
+      "step": 3820
+    },
+    {
+      "epoch": 0.0765,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.000153,
+      "loss": 1.2612,
+      "step": 3825
+    },
+    {
+      "epoch": 0.0766,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001532,
+      "loss": 1.298,
+      "step": 3830
+    },
+    {
+      "epoch": 0.0767,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00015340000000000002,
+      "loss": 1.298,
+      "step": 3835
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00015360000000000002,
+      "loss": 1.2609,
+      "step": 3840
+    },
+    {
+      "epoch": 0.0769,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001538,
+      "loss": 1.2892,
+      "step": 3845
+    },
+    {
+      "epoch": 0.077,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.000154,
+      "loss": 1.2868,
+      "step": 3850
+    },
+    {
+      "epoch": 0.0771,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0001542,
+      "loss": 1.2854,
+      "step": 3855
+    },
+    {
+      "epoch": 0.0772,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001544,
+      "loss": 1.2711,
+      "step": 3860
+    },
+    {
+      "epoch": 0.0773,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00015460000000000002,
+      "loss": 1.2844,
+      "step": 3865
+    },
+    {
+      "epoch": 0.0774,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00015480000000000002,
+      "loss": 1.2699,
+      "step": 3870
+    },
+    {
+      "epoch": 0.0775,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.000155,
+      "loss": 1.2834,
+      "step": 3875
+    },
+    {
+      "epoch": 0.0776,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001552,
+      "loss": 1.272,
+      "step": 3880
+    },
+    {
+      "epoch": 0.0777,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001554,
+      "loss": 1.2759,
+      "step": 3885
+    },
+    {
+      "epoch": 0.0778,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00015560000000000001,
+      "loss": 1.273,
+      "step": 3890
+    },
+    {
+      "epoch": 0.0779,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00015580000000000002,
+      "loss": 1.2747,
+      "step": 3895
+    },
+    {
+      "epoch": 0.078,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00015600000000000002,
+      "loss": 1.2862,
+      "step": 3900
+    },
+    {
+      "epoch": 0.0781,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001562,
+      "loss": 1.2652,
+      "step": 3905
+    },
+    {
+      "epoch": 0.0782,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001564,
+      "loss": 1.2648,
+      "step": 3910
+    },
+    {
+      "epoch": 0.0783,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001566,
+      "loss": 1.2784,
+      "step": 3915
+    },
+    {
+      "epoch": 0.0784,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00015680000000000002,
+      "loss": 1.2742,
+      "step": 3920
+    },
+    {
+      "epoch": 0.0785,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00015700000000000002,
+      "loss": 1.2749,
+      "step": 3925
+    },
+    {
+      "epoch": 0.0786,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00015720000000000003,
+      "loss": 1.2678,
+      "step": 3930
+    },
+    {
+      "epoch": 0.0787,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001574,
+      "loss": 1.2754,
+      "step": 3935
+    },
+    {
+      "epoch": 0.0788,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001576,
+      "loss": 1.2681,
+      "step": 3940
+    },
+    {
+      "epoch": 0.0789,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00015780000000000001,
+      "loss": 1.2814,
+      "step": 3945
+    },
+    {
+      "epoch": 0.079,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00015800000000000002,
+      "loss": 1.2796,
+      "step": 3950
+    },
+    {
+      "epoch": 0.0791,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00015820000000000002,
+      "loss": 1.2727,
+      "step": 3955
+    },
+    {
+      "epoch": 0.0792,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00015840000000000003,
+      "loss": 1.272,
+      "step": 3960
+    },
+    {
+      "epoch": 0.0793,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0001586,
+      "loss": 1.2734,
+      "step": 3965
+    },
+    {
+      "epoch": 0.0794,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001588,
+      "loss": 1.2766,
+      "step": 3970
+    },
+    {
+      "epoch": 0.0795,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00015900000000000002,
+      "loss": 1.262,
+      "step": 3975
+    },
+    {
+      "epoch": 0.0796,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00015920000000000002,
+      "loss": 1.2616,
+      "step": 3980
+    },
+    {
+      "epoch": 0.0797,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00015940000000000003,
+      "loss": 1.28,
+      "step": 3985
+    },
+    {
+      "epoch": 0.0798,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0001596,
+      "loss": 1.2748,
+      "step": 3990
+    },
+    {
+      "epoch": 0.0799,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0001598,
+      "loss": 1.2886,
+      "step": 3995
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00016,
+      "loss": 1.2739,
+      "step": 4000
+    },
+    {
+      "epoch": 0.0801,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016020000000000002,
+      "loss": 1.2772,
+      "step": 4005
+    },
+    {
+      "epoch": 0.0802,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00016040000000000002,
+      "loss": 1.2926,
+      "step": 4010
+    },
+    {
+      "epoch": 0.0803,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00016060000000000003,
+      "loss": 1.2837,
+      "step": 4015
+    },
+    {
+      "epoch": 0.0804,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001608,
+      "loss": 1.2672,
+      "step": 4020
+    },
+    {
+      "epoch": 0.0805,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.000161,
+      "loss": 1.2826,
+      "step": 4025
+    },
+    {
+      "epoch": 0.0806,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00016120000000000002,
+      "loss": 1.2919,
+      "step": 4030
+    },
+    {
+      "epoch": 0.0807,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00016140000000000002,
+      "loss": 1.277,
+      "step": 4035
+    },
+    {
+      "epoch": 0.0808,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00016160000000000002,
+      "loss": 1.2847,
+      "step": 4040
+    },
+    {
+      "epoch": 0.0809,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00016180000000000003,
+      "loss": 1.2692,
+      "step": 4045
+    },
+    {
+      "epoch": 0.081,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.000162,
+      "loss": 1.2972,
+      "step": 4050
+    },
+    {
+      "epoch": 0.0811,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0001622,
+      "loss": 1.2671,
+      "step": 4055
+    },
+    {
+      "epoch": 0.0812,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00016240000000000002,
+      "loss": 1.2643,
+      "step": 4060
+    },
+    {
+      "epoch": 0.0813,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0001626,
+      "loss": 1.2721,
+      "step": 4065
+    },
+    {
+      "epoch": 0.0814,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001628,
+      "loss": 1.2954,
+      "step": 4070
+    },
+    {
+      "epoch": 0.0815,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.000163,
+      "loss": 1.2687,
+      "step": 4075
+    },
+    {
+      "epoch": 0.0816,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001632,
+      "loss": 1.2796,
+      "step": 4080
+    },
+    {
+      "epoch": 0.0817,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001634,
+      "loss": 1.2804,
+      "step": 4085
+    },
+    {
+      "epoch": 0.0818,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001636,
+      "loss": 1.2839,
+      "step": 4090
+    },
+    {
+      "epoch": 0.0819,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0001638,
+      "loss": 1.2719,
+      "step": 4095
+    },
+    {
+      "epoch": 0.082,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.000164,
+      "loss": 1.2762,
+      "step": 4100
+    },
+    {
+      "epoch": 0.0821,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0001642,
+      "loss": 1.2804,
+      "step": 4105
+    },
+    {
+      "epoch": 0.0822,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001644,
+      "loss": 1.2652,
+      "step": 4110
+    },
+    {
+      "epoch": 0.0823,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0001646,
+      "loss": 1.268,
+      "step": 4115
+    },
+    {
+      "epoch": 0.0824,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0001648,
+      "loss": 1.2737,
+      "step": 4120
+    },
+    {
+      "epoch": 0.0825,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.000165,
+      "loss": 1.2723,
+      "step": 4125
+    },
+    {
+      "epoch": 0.0826,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001652,
+      "loss": 1.2775,
+      "step": 4130
+    },
+    {
+      "epoch": 0.0827,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001654,
+      "loss": 1.2815,
+      "step": 4135
+    },
+    {
+      "epoch": 0.0828,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0001656,
+      "loss": 1.2725,
+      "step": 4140
+    },
+    {
+      "epoch": 0.0829,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001658,
+      "loss": 1.2759,
+      "step": 4145
+    },
+    {
+      "epoch": 0.083,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.000166,
+      "loss": 1.2649,
+      "step": 4150
+    },
+    {
+      "epoch": 0.0831,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001662,
+      "loss": 1.2833,
+      "step": 4155
+    },
+    {
+      "epoch": 0.0832,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001664,
+      "loss": 1.271,
+      "step": 4160
+    },
+    {
+      "epoch": 0.0833,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001666,
+      "loss": 1.3122,
+      "step": 4165
+    },
+    {
+      "epoch": 0.0834,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0001668,
+      "loss": 1.2976,
+      "step": 4170
+    },
+    {
+      "epoch": 0.0835,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.000167,
+      "loss": 1.2778,
+      "step": 4175
+    },
+    {
+      "epoch": 0.0836,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001672,
+      "loss": 1.2957,
+      "step": 4180
+    },
+    {
+      "epoch": 0.0837,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0001674,
+      "loss": 1.2716,
+      "step": 4185
+    },
+    {
+      "epoch": 0.0838,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001676,
+      "loss": 1.2678,
+      "step": 4190
+    },
+    {
+      "epoch": 0.0839,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0001678,
+      "loss": 1.3049,
+      "step": 4195
+    },
+    {
+      "epoch": 0.084,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.000168,
+      "loss": 1.2603,
+      "step": 4200
+    },
+    {
+      "epoch": 0.0841,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001682,
+      "loss": 1.283,
+      "step": 4205
+    },
+    {
+      "epoch": 0.0842,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001684,
+      "loss": 1.269,
+      "step": 4210
+    },
+    {
+      "epoch": 0.0843,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001686,
+      "loss": 1.2793,
+      "step": 4215
+    },
+    {
+      "epoch": 0.0844,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001688,
+      "loss": 1.2814,
+      "step": 4220
+    },
+    {
+      "epoch": 0.0845,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00016900000000000002,
+      "loss": 1.2605,
+      "step": 4225
+    },
+    {
+      "epoch": 0.0846,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001692,
+      "loss": 1.2807,
+      "step": 4230
+    },
+    {
+      "epoch": 0.0847,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0001694,
+      "loss": 1.2791,
+      "step": 4235
+    },
+    {
+      "epoch": 0.0848,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0001696,
+      "loss": 1.2885,
+      "step": 4240
+    },
+    {
+      "epoch": 0.0849,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0001698,
+      "loss": 1.2784,
+      "step": 4245
+    },
+    {
+      "epoch": 0.085,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00017,
+      "loss": 1.2785,
+      "step": 4250
+    },
+    {
+      "epoch": 0.0851,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00017020000000000002,
+      "loss": 1.2759,
+      "step": 4255
+    },
+    {
+      "epoch": 0.0852,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0001704,
+      "loss": 1.2933,
+      "step": 4260
+    },
+    {
+      "epoch": 0.0853,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001706,
+      "loss": 1.2817,
+      "step": 4265
+    },
+    {
+      "epoch": 0.0854,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001708,
+      "loss": 1.2863,
+      "step": 4270
+    },
+    {
+      "epoch": 0.0855,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.000171,
+      "loss": 1.2869,
+      "step": 4275
+    },
+    {
+      "epoch": 0.0856,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00017120000000000001,
+      "loss": 1.261,
+      "step": 4280
+    },
+    {
+      "epoch": 0.0857,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0001714,
+      "loss": 1.2806,
+      "step": 4285
+    },
+    {
+      "epoch": 0.0858,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001716,
+      "loss": 1.2787,
+      "step": 4290
+    },
+    {
+      "epoch": 0.0859,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001718,
+      "loss": 1.2674,
+      "step": 4295
+    },
+    {
+      "epoch": 0.086,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.000172,
+      "loss": 1.284,
+      "step": 4300
+    },
+    {
+      "epoch": 0.0861,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001722,
+      "loss": 1.285,
+      "step": 4305
+    },
+    {
+      "epoch": 0.0862,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00017240000000000002,
+      "loss": 1.2805,
+      "step": 4310
+    },
+    {
+      "epoch": 0.0863,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0001726,
+      "loss": 1.2892,
+      "step": 4315
+    },
+    {
+      "epoch": 0.0864,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001728,
+      "loss": 1.2744,
+      "step": 4320
+    },
+    {
+      "epoch": 0.0865,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000173,
+      "loss": 1.2618,
+      "step": 4325
+    },
+    {
+      "epoch": 0.0866,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001732,
+      "loss": 1.2819,
+      "step": 4330
+    },
+    {
+      "epoch": 0.0867,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001734,
+      "loss": 1.2994,
+      "step": 4335
+    },
+    {
+      "epoch": 0.0868,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017360000000000002,
+      "loss": 1.2742,
+      "step": 4340
+    },
+    {
+      "epoch": 0.0869,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.0001738,
+      "loss": 1.2767,
+      "step": 4345
+    },
+    {
+      "epoch": 0.087,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.000174,
+      "loss": 1.2636,
+      "step": 4350
+    },
+    {
+      "epoch": 0.0871,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001742,
+      "loss": 1.271,
+      "step": 4355
+    },
+    {
+      "epoch": 0.0872,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001744,
+      "loss": 1.2746,
+      "step": 4360
+    },
+    {
+      "epoch": 0.0873,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00017460000000000002,
+      "loss": 1.2739,
+      "step": 4365
+    },
+    {
+      "epoch": 0.0874,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00017480000000000002,
+      "loss": 1.2609,
+      "step": 4370
+    },
+    {
+      "epoch": 0.0875,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.000175,
+      "loss": 1.2906,
+      "step": 4375
+    },
+    {
+      "epoch": 0.0876,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0001752,
+      "loss": 1.2818,
+      "step": 4380
+    },
+    {
+      "epoch": 0.0877,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001754,
+      "loss": 1.2688,
+      "step": 4385
+    },
+    {
+      "epoch": 0.0878,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001756,
+      "loss": 1.2797,
+      "step": 4390
+    },
+    {
+      "epoch": 0.0879,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017580000000000002,
+      "loss": 1.2997,
+      "step": 4395
+    },
+    {
+      "epoch": 0.088,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017600000000000002,
+      "loss": 1.288,
+      "step": 4400
+    },
+    {
+      "epoch": 0.0881,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001762,
+      "loss": 1.2926,
+      "step": 4405
+    },
+    {
+      "epoch": 0.0882,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001764,
+      "loss": 1.2776,
+      "step": 4410
+    },
+    {
+      "epoch": 0.0883,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001766,
+      "loss": 1.2661,
+      "step": 4415
+    },
+    {
+      "epoch": 0.0884,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00017680000000000001,
+      "loss": 1.2722,
+      "step": 4420
+    },
+    {
+      "epoch": 0.0885,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017700000000000002,
+      "loss": 1.2767,
+      "step": 4425
+    },
+    {
+      "epoch": 0.0886,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001772,
+      "loss": 1.2872,
+      "step": 4430
+    },
+    {
+      "epoch": 0.0887,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001774,
+      "loss": 1.2786,
+      "step": 4435
+    },
+    {
+      "epoch": 0.0888,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001776,
+      "loss": 1.2788,
+      "step": 4440
+    },
+    {
+      "epoch": 0.0889,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001778,
+      "loss": 1.289,
+      "step": 4445
+    },
+    {
+      "epoch": 0.089,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00017800000000000002,
+      "loss": 1.2882,
+      "step": 4450
+    },
+    {
+      "epoch": 0.0891,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00017820000000000002,
+      "loss": 1.2796,
+      "step": 4455
+    },
+    {
+      "epoch": 0.0892,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0001784,
+      "loss": 1.2727,
+      "step": 4460
+    },
+    {
+      "epoch": 0.0893,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001786,
+      "loss": 1.2923,
+      "step": 4465
+    },
+    {
+      "epoch": 0.0894,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0001788,
+      "loss": 1.2663,
+      "step": 4470
+    },
+    {
+      "epoch": 0.0895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00017900000000000001,
+      "loss": 1.2648,
+      "step": 4475
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00017920000000000002,
+      "loss": 1.2669,
+      "step": 4480
+    },
+    {
+      "epoch": 0.0897,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00017940000000000002,
+      "loss": 1.2722,
+      "step": 4485
+    },
+    {
+      "epoch": 0.0898,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001796,
+      "loss": 1.2681,
+      "step": 4490
+    },
+    {
+      "epoch": 0.0899,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001798,
+      "loss": 1.2722,
+      "step": 4495
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018,
+      "loss": 1.2926,
+      "step": 4500
+    },
+    {
+      "epoch": 0.0901,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018020000000000002,
+      "loss": 1.2717,
+      "step": 4505
+    },
+    {
+      "epoch": 0.0902,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00018040000000000002,
+      "loss": 1.2727,
+      "step": 4510
+    },
+    {
+      "epoch": 0.0903,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00018060000000000003,
+      "loss": 1.271,
+      "step": 4515
+    },
+    {
+      "epoch": 0.0904,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001808,
+      "loss": 1.2828,
+      "step": 4520
+    },
+    {
+      "epoch": 0.0905,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.000181,
+      "loss": 1.2774,
+      "step": 4525
+    },
+    {
+      "epoch": 0.0906,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001812,
+      "loss": 1.2674,
+      "step": 4530
+    },
+    {
+      "epoch": 0.0907,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018140000000000002,
+      "loss": 1.289,
+      "step": 4535
+    },
+    {
+      "epoch": 0.0908,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00018160000000000002,
+      "loss": 1.2795,
+      "step": 4540
+    },
+    {
+      "epoch": 0.0909,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00018180000000000003,
+      "loss": 1.2713,
+      "step": 4545
+    },
+    {
+      "epoch": 0.091,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.000182,
+      "loss": 1.2776,
+      "step": 4550
+    },
+    {
+      "epoch": 0.0911,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001822,
+      "loss": 1.2927,
+      "step": 4555
+    },
+    {
+      "epoch": 0.0912,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00018240000000000002,
+      "loss": 1.2679,
+      "step": 4560
+    },
+    {
+      "epoch": 0.0913,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00018260000000000002,
+      "loss": 1.2725,
+      "step": 4565
+    },
+    {
+      "epoch": 0.0914,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00018280000000000003,
+      "loss": 1.2827,
+      "step": 4570
+    },
+    {
+      "epoch": 0.0915,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000183,
+      "loss": 1.2676,
+      "step": 4575
+    },
+    {
+      "epoch": 0.0916,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0001832,
+      "loss": 1.2786,
+      "step": 4580
+    },
+    {
+      "epoch": 0.0917,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001834,
+      "loss": 1.2678,
+      "step": 4585
+    },
+    {
+      "epoch": 0.0918,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00018360000000000002,
+      "loss": 1.2673,
+      "step": 4590
+    },
+    {
+      "epoch": 0.0919,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00018380000000000002,
+      "loss": 1.2656,
+      "step": 4595
+    },
+    {
+      "epoch": 0.092,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00018400000000000003,
+      "loss": 1.2794,
+      "step": 4600
+    },
+    {
+      "epoch": 0.0921,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001842,
+      "loss": 1.2991,
+      "step": 4605
+    },
+    {
+      "epoch": 0.0922,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0001844,
+      "loss": 1.2715,
+      "step": 4610
+    },
+    {
+      "epoch": 0.0923,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00018460000000000001,
+      "loss": 1.3009,
+      "step": 4615
+    },
+    {
+      "epoch": 0.0924,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018480000000000002,
+      "loss": 1.2914,
+      "step": 4620
+    },
+    {
+      "epoch": 0.0925,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00018500000000000002,
+      "loss": 1.2837,
+      "step": 4625
+    },
+    {
+      "epoch": 0.0926,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018520000000000003,
+      "loss": 1.2724,
+      "step": 4630
+    },
+    {
+      "epoch": 0.0927,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001854,
+      "loss": 1.2765,
+      "step": 4635
+    },
+    {
+      "epoch": 0.0928,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001856,
+      "loss": 1.2784,
+      "step": 4640
+    },
+    {
+      "epoch": 0.0929,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00018580000000000002,
+      "loss": 1.2711,
+      "step": 4645
+    },
+    {
+      "epoch": 0.093,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00018600000000000002,
+      "loss": 1.2632,
+      "step": 4650
+    },
+    {
+      "epoch": 0.0931,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00018620000000000003,
+      "loss": 1.2872,
+      "step": 4655
+    },
+    {
+      "epoch": 0.0932,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018640000000000003,
+      "loss": 1.288,
+      "step": 4660
+    },
+    {
+      "epoch": 0.0933,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001866,
+      "loss": 1.2662,
+      "step": 4665
+    },
+    {
+      "epoch": 0.0934,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00018680000000000001,
+      "loss": 1.2851,
+      "step": 4670
+    },
+    {
+      "epoch": 0.0935,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00018700000000000002,
+      "loss": 1.2799,
+      "step": 4675
+    },
+    {
+      "epoch": 0.0936,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018720000000000002,
+      "loss": 1.2773,
+      "step": 4680
+    },
+    {
+      "epoch": 0.0937,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018740000000000003,
+      "loss": 1.2941,
+      "step": 4685
+    },
+    {
+      "epoch": 0.0938,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001876,
+      "loss": 1.2802,
+      "step": 4690
+    },
+    {
+      "epoch": 0.0939,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001878,
+      "loss": 1.2764,
+      "step": 4695
+    },
+    {
+      "epoch": 0.094,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.000188,
+      "loss": 1.2867,
+      "step": 4700
+    },
+    {
+      "epoch": 0.0941,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001882,
+      "loss": 1.286,
+      "step": 4705
+    },
+    {
+      "epoch": 0.0942,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001884,
+      "loss": 1.2802,
+      "step": 4710
+    },
+    {
+      "epoch": 0.0943,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0001886,
+      "loss": 1.2787,
+      "step": 4715
+    },
+    {
+      "epoch": 0.0944,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0001888,
+      "loss": 1.2589,
+      "step": 4720
+    },
+    {
+      "epoch": 0.0945,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00018899999999999999,
+      "loss": 1.2721,
+      "step": 4725
+    },
+    {
+      "epoch": 0.0946,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001892,
+      "loss": 1.2649,
+      "step": 4730
+    },
+    {
+      "epoch": 0.0947,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001894,
+      "loss": 1.2611,
+      "step": 4735
+    },
+    {
+      "epoch": 0.0948,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0001896,
+      "loss": 1.2738,
+      "step": 4740
+    },
+    {
+      "epoch": 0.0949,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001898,
+      "loss": 1.2856,
+      "step": 4745
+    },
+    {
+      "epoch": 0.095,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019,
+      "loss": 1.2751,
+      "step": 4750
+    },
+    {
+      "epoch": 0.0951,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001902,
+      "loss": 1.2734,
+      "step": 4755
+    },
+    {
+      "epoch": 0.0952,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0001904,
+      "loss": 1.2705,
+      "step": 4760
+    },
+    {
+      "epoch": 0.0953,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001906,
+      "loss": 1.2691,
+      "step": 4765
+    },
+    {
+      "epoch": 0.0954,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001908,
+      "loss": 1.2713,
+      "step": 4770
+    },
+    {
+      "epoch": 0.0955,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.000191,
+      "loss": 1.2829,
+      "step": 4775
+    },
+    {
+      "epoch": 0.0956,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001912,
+      "loss": 1.2835,
+      "step": 4780
+    },
+    {
+      "epoch": 0.0957,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001914,
+      "loss": 1.2763,
+      "step": 4785
+    },
+    {
+      "epoch": 0.0958,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001916,
+      "loss": 1.2705,
+      "step": 4790
+    },
+    {
+      "epoch": 0.0959,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001918,
+      "loss": 1.2824,
+      "step": 4795
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.000192,
+      "loss": 1.2905,
+      "step": 4800
+    },
+    {
+      "epoch": 0.0961,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001922,
+      "loss": 1.2807,
+      "step": 4805
+    },
+    {
+      "epoch": 0.0962,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019240000000000001,
+      "loss": 1.2839,
+      "step": 4810
+    },
+    {
+      "epoch": 0.0963,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0001926,
+      "loss": 1.2728,
+      "step": 4815
+    },
+    {
+      "epoch": 0.0964,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001928,
+      "loss": 1.2725,
+      "step": 4820
+    },
+    {
+      "epoch": 0.0965,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.000193,
+      "loss": 1.2591,
+      "step": 4825
+    },
+    {
+      "epoch": 0.0966,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0001932,
+      "loss": 1.2819,
+      "step": 4830
+    },
+    {
+      "epoch": 0.0967,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0001934,
+      "loss": 1.274,
+      "step": 4835
+    },
+    {
+      "epoch": 0.0968,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019360000000000002,
+      "loss": 1.2751,
+      "step": 4840
+    },
+    {
+      "epoch": 0.0969,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001938,
+      "loss": 1.2784,
+      "step": 4845
+    },
+    {
+      "epoch": 0.097,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.000194,
+      "loss": 1.2764,
+      "step": 4850
+    },
+    {
+      "epoch": 0.0971,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001942,
+      "loss": 1.273,
+      "step": 4855
+    },
+    {
+      "epoch": 0.0972,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001944,
+      "loss": 1.2812,
+      "step": 4860
+    },
+    {
+      "epoch": 0.0973,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019460000000000001,
+      "loss": 1.267,
+      "step": 4865
+    },
+    {
+      "epoch": 0.0974,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001948,
+      "loss": 1.2874,
+      "step": 4870
+    },
+    {
+      "epoch": 0.0975,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.000195,
+      "loss": 1.279,
+      "step": 4875
+    },
+    {
+      "epoch": 0.0976,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001952,
+      "loss": 1.2724,
+      "step": 4880
+    },
+    {
+      "epoch": 0.0977,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001954,
+      "loss": 1.2779,
+      "step": 4885
+    },
+    {
+      "epoch": 0.0978,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001956,
+      "loss": 1.2691,
+      "step": 4890
+    },
+    {
+      "epoch": 0.0979,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00019580000000000002,
+      "loss": 1.2759,
+      "step": 4895
+    },
+    {
+      "epoch": 0.098,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.000196,
+      "loss": 1.2942,
+      "step": 4900
+    },
+    {
+      "epoch": 0.0981,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0001962,
+      "loss": 1.2701,
+      "step": 4905
+    },
+    {
+      "epoch": 0.0982,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001964,
+      "loss": 1.2815,
+      "step": 4910
+    },
+    {
+      "epoch": 0.0983,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001966,
+      "loss": 1.2724,
+      "step": 4915
+    },
+    {
+      "epoch": 0.0984,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0001968,
+      "loss": 1.2684,
+      "step": 4920
+    },
+    {
+      "epoch": 0.0985,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019700000000000002,
+      "loss": 1.2713,
+      "step": 4925
+    },
+    {
+      "epoch": 0.0986,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001972,
+      "loss": 1.2784,
+      "step": 4930
+    },
+    {
+      "epoch": 0.0987,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001974,
+      "loss": 1.2701,
+      "step": 4935
+    },
+    {
+      "epoch": 0.0988,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0001976,
+      "loss": 1.2692,
+      "step": 4940
+    },
+    {
+      "epoch": 0.0989,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001978,
+      "loss": 1.2928,
+      "step": 4945
+    },
+    {
+      "epoch": 0.099,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00019800000000000002,
+      "loss": 1.2809,
+      "step": 4950
+    },
+    {
+      "epoch": 0.0991,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019820000000000002,
+      "loss": 1.2699,
+      "step": 4955
+    },
+    {
+      "epoch": 0.0992,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001984,
+      "loss": 1.2973,
+      "step": 4960
+    },
+    {
+      "epoch": 0.0993,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001986,
+      "loss": 1.2821,
+      "step": 4965
+    },
+    {
+      "epoch": 0.0994,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001988,
+      "loss": 1.2699,
+      "step": 4970
+    },
+    {
+      "epoch": 0.0995,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.000199,
+      "loss": 1.2749,
+      "step": 4975
+    },
+    {
+      "epoch": 0.0996,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00019920000000000002,
+      "loss": 1.2709,
+      "step": 4980
+    },
+    {
+      "epoch": 0.0997,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00019940000000000002,
+      "loss": 1.2664,
+      "step": 4985
+    },
+    {
+      "epoch": 0.0998,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001996,
+      "loss": 1.2793,
+      "step": 4990
+    },
+    {
+      "epoch": 0.0999,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001998,
+      "loss": 1.2776,
+      "step": 4995
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002,
+      "loss": 1.2754,
+      "step": 5000
+    },
+    {
+      "epoch": 0.1001,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019999999390765168,
+      "loss": 1.2774,
+      "step": 5005
+    },
+    {
+      "epoch": 0.1002,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0001999999756306074,
+      "loss": 1.2908,
+      "step": 5010
+    },
+    {
+      "epoch": 0.1003,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00019999994516886946,
+      "loss": 1.2649,
+      "step": 5015
+    },
+    {
+      "epoch": 0.1004,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00019999990252244151,
+      "loss": 1.2956,
+      "step": 5020
+    },
+    {
+      "epoch": 0.1005,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019999984769132877,
+      "loss": 1.2584,
+      "step": 5025
+    },
+    {
+      "epoch": 0.1006,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00019999978067553796,
+      "loss": 1.288,
+      "step": 5030
+    },
+    {
+      "epoch": 0.1007,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019999970147507713,
+      "loss": 1.2733,
+      "step": 5035
+    },
+    {
+      "epoch": 0.1008,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00019999961008995606,
+      "loss": 1.2819,
+      "step": 5040
+    },
+    {
+      "epoch": 0.1009,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019999950652018584,
+      "loss": 1.2775,
+      "step": 5045
+    },
+    {
+      "epoch": 0.101,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019999939076577905,
+      "loss": 1.2851,
+      "step": 5050
+    },
+    {
+      "epoch": 0.1011,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019999926282674983,
+      "loss": 1.277,
+      "step": 5055
+    },
+    {
+      "epoch": 0.1012,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019999912270311375,
+      "loss": 1.2835,
+      "step": 5060
+    },
+    {
+      "epoch": 0.1013,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019999897039488796,
+      "loss": 1.2654,
+      "step": 5065
+    },
+    {
+      "epoch": 0.1014,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001999988059020909,
+      "loss": 1.2839,
+      "step": 5070
+    },
+    {
+      "epoch": 0.1015,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019999862922474268,
+      "loss": 1.2739,
+      "step": 5075
+    },
+    {
+      "epoch": 0.1016,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00019999844036286483,
+      "loss": 1.2684,
+      "step": 5080
+    },
+    {
+      "epoch": 0.1017,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00019999823931648035,
+      "loss": 1.3092,
+      "step": 5085
+    },
+    {
+      "epoch": 0.1018,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0001999980260856137,
+      "loss": 1.278,
+      "step": 5090
+    },
+    {
+      "epoch": 0.1019,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019999780067029094,
+      "loss": 1.2708,
+      "step": 5095
+    },
+    {
+      "epoch": 0.102,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00019999756307053948,
+      "loss": 1.3041,
+      "step": 5100
+    },
+    {
+      "epoch": 0.1021,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019999731328638828,
+      "loss": 1.2686,
+      "step": 5105
+    },
+    {
+      "epoch": 0.1022,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0001999970513178678,
+      "loss": 1.2832,
+      "step": 5110
+    },
+    {
+      "epoch": 0.1023,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00019999677716500992,
+      "loss": 1.2849,
+      "step": 5115
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001999964908278481,
+      "loss": 1.2805,
+      "step": 5120
+    },
+    {
+      "epoch": 0.1025,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019999619230641713,
+      "loss": 1.2801,
+      "step": 5125
+    },
+    {
+      "epoch": 0.1026,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00019999588160075348,
+      "loss": 1.2976,
+      "step": 5130
+    },
+    {
+      "epoch": 0.1027,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.000199995558710895,
+      "loss": 1.2664,
+      "step": 5135
+    },
+    {
+      "epoch": 0.1028,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.000199995223636881,
+      "loss": 1.2663,
+      "step": 5140
+    },
+    {
+      "epoch": 0.1029,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001999948763787523,
+      "loss": 1.2906,
+      "step": 5145
+    },
+    {
+      "epoch": 0.103,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019999451693655123,
+      "loss": 1.2751,
+      "step": 5150
+    },
+    {
+      "epoch": 0.1031,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019999414531032158,
+      "loss": 1.2703,
+      "step": 5155
+    },
+    {
+      "epoch": 0.1032,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00019999376150010866,
+      "loss": 1.2879,
+      "step": 5160
+    },
+    {
+      "epoch": 0.1033,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.0001999933655059592,
+      "loss": 1.2822,
+      "step": 5165
+    },
+    {
+      "epoch": 0.1034,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019999295732792146,
+      "loss": 1.2934,
+      "step": 5170
+    },
+    {
+      "epoch": 0.1035,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019999253696604522,
+      "loss": 1.2698,
+      "step": 5175
+    },
+    {
+      "epoch": 0.1036,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00019999210442038162,
+      "loss": 1.2906,
+      "step": 5180
+    },
+    {
+      "epoch": 0.1037,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019999165969098343,
+      "loss": 1.268,
+      "step": 5185
+    },
+    {
+      "epoch": 0.1038,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00019999120277790477,
+      "loss": 1.2685,
+      "step": 5190
+    },
+    {
+      "epoch": 0.1039,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0001999907336812014,
+      "loss": 1.2856,
+      "step": 5195
+    },
+    {
+      "epoch": 0.104,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00019999025240093044,
+      "loss": 1.2886,
+      "step": 5200
+    },
+    {
+      "epoch": 0.1041,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001999897589371505,
+      "loss": 1.2798,
+      "step": 5205
+    },
+    {
+      "epoch": 0.1042,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00019998925328992175,
+      "loss": 1.2835,
+      "step": 5210
+    },
+    {
+      "epoch": 0.1043,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001999887354593058,
+      "loss": 1.2746,
+      "step": 5215
+    },
+    {
+      "epoch": 0.1044,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001999882054453657,
+      "loss": 1.2737,
+      "step": 5220
+    },
+    {
+      "epoch": 0.1045,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019998766324816607,
+      "loss": 1.267,
+      "step": 5225
+    },
+    {
+      "epoch": 0.1046,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.000199987108867773,
+      "loss": 1.2728,
+      "step": 5230
+    },
+    {
+      "epoch": 0.1047,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019998654230425395,
+      "loss": 1.2663,
+      "step": 5235
+    },
+    {
+      "epoch": 0.1048,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00019998596355767805,
+      "loss": 1.2893,
+      "step": 5240
+    },
+    {
+      "epoch": 0.1049,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00019998537262811577,
+      "loss": 1.2655,
+      "step": 5245
+    },
+    {
+      "epoch": 0.105,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00019998476951563915,
+      "loss": 1.2939,
+      "step": 5250
+    },
+    {
+      "epoch": 0.1051,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019998415422032163,
+      "loss": 1.2817,
+      "step": 5255
+    },
+    {
+      "epoch": 0.1052,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019998352674223816,
+      "loss": 1.2766,
+      "step": 5260
+    },
+    {
+      "epoch": 0.1053,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00019998288708146527,
+      "loss": 1.2793,
+      "step": 5265
+    },
+    {
+      "epoch": 0.1054,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0001999822352380809,
+      "loss": 1.2866,
+      "step": 5270
+    },
+    {
+      "epoch": 0.1055,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019998157121216444,
+      "loss": 1.2911,
+      "step": 5275
+    },
+    {
+      "epoch": 0.1056,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00019998089500379676,
+      "loss": 1.274,
+      "step": 5280
+    },
+    {
+      "epoch": 0.1057,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00019998020661306037,
+      "loss": 1.2727,
+      "step": 5285
+    },
+    {
+      "epoch": 0.1058,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.000199979506040039,
+      "loss": 1.2733,
+      "step": 5290
+    },
+    {
+      "epoch": 0.1059,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00019997879328481814,
+      "loss": 1.2692,
+      "step": 5295
+    },
+    {
+      "epoch": 0.106,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00019997806834748456,
+      "loss": 1.2805,
+      "step": 5300
+    },
+    {
+      "epoch": 0.1061,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0001999773312281266,
+      "loss": 1.2811,
+      "step": 5305
+    },
+    {
+      "epoch": 0.1062,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00019997658192683412,
+      "loss": 1.2667,
+      "step": 5310
+    },
+    {
+      "epoch": 0.1063,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019997582044369843,
+      "loss": 1.2858,
+      "step": 5315
+    },
+    {
+      "epoch": 0.1064,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019997504677881224,
+      "loss": 1.286,
+      "step": 5320
+    },
+    {
+      "epoch": 0.1065,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019997426093226986,
+      "loss": 1.2801,
+      "step": 5325
+    },
+    {
+      "epoch": 0.1066,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.000199973462904167,
+      "loss": 1.2713,
+      "step": 5330
+    },
+    {
+      "epoch": 0.1067,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.000199972652694601,
+      "loss": 1.2707,
+      "step": 5335
+    },
+    {
+      "epoch": 0.1068,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00019997183030367048,
+      "loss": 1.2796,
+      "step": 5340
+    },
+    {
+      "epoch": 0.1069,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001999709957314757,
+      "loss": 1.2802,
+      "step": 5345
+    },
+    {
+      "epoch": 0.107,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019997014897811833,
+      "loss": 1.2675,
+      "step": 5350
+    },
+    {
+      "epoch": 0.1071,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00019996929004370152,
+      "loss": 1.2705,
+      "step": 5355
+    },
+    {
+      "epoch": 0.1072,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00019996841892833,
+      "loss": 1.2902,
+      "step": 5360
+    },
+    {
+      "epoch": 0.1073,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00019996753563210985,
+      "loss": 1.2943,
+      "step": 5365
+    },
+    {
+      "epoch": 0.1074,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00019996664015514871,
+      "loss": 1.2742,
+      "step": 5370
+    },
+    {
+      "epoch": 0.1075,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00019996573249755572,
+      "loss": 1.301,
+      "step": 5375
+    },
+    {
+      "epoch": 0.1076,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019996481265944146,
+      "loss": 1.2813,
+      "step": 5380
+    },
+    {
+      "epoch": 0.1077,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.000199963880640918,
+      "loss": 1.2757,
+      "step": 5385
+    },
+    {
+      "epoch": 0.1078,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019996293644209887,
+      "loss": 1.2669,
+      "step": 5390
+    },
+    {
+      "epoch": 0.1079,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001999619800630992,
+      "loss": 1.2644,
+      "step": 5395
+    },
+    {
+      "epoch": 0.108,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019996101150403543,
+      "loss": 1.2813,
+      "step": 5400
+    },
+    {
+      "epoch": 0.1081,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00019996003076502565,
+      "loss": 1.2853,
+      "step": 5405
+    },
+    {
+      "epoch": 0.1082,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00019995903784618936,
+      "loss": 1.2777,
+      "step": 5410
+    },
+    {
+      "epoch": 0.1083,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00019995803274764747,
+      "loss": 1.2929,
+      "step": 5415
+    },
+    {
+      "epoch": 0.1084,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001999570154695225,
+      "loss": 1.2787,
+      "step": 5420
+    },
+    {
+      "epoch": 0.1085,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00019995598601193842,
+      "loss": 1.2765,
+      "step": 5425
+    },
+    {
+      "epoch": 0.1086,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00019995494437502064,
+      "loss": 1.2752,
+      "step": 5430
+    },
+    {
+      "epoch": 0.1087,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019995389055889605,
+      "loss": 1.2637,
+      "step": 5435
+    },
+    {
+      "epoch": 0.1088,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001999528245636931,
+      "loss": 1.2865,
+      "step": 5440
+    },
+    {
+      "epoch": 0.1089,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001999517463895417,
+      "loss": 1.2882,
+      "step": 5445
+    },
+    {
+      "epoch": 0.109,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00019995065603657316,
+      "loss": 1.277,
+      "step": 5450
+    },
+    {
+      "epoch": 0.1091,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00019994955350492038,
+      "loss": 1.276,
+      "step": 5455
+    },
+    {
+      "epoch": 0.1092,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0001999484387947177,
+      "loss": 1.2715,
+      "step": 5460
+    },
+    {
+      "epoch": 0.1093,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00019994731190610087,
+      "loss": 1.2743,
+      "step": 5465
+    },
+    {
+      "epoch": 0.1094,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001999461728392073,
+      "loss": 1.267,
+      "step": 5470
+    },
+    {
+      "epoch": 0.1095,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00019994502159417573,
+      "loss": 1.2801,
+      "step": 5475
+    },
+    {
+      "epoch": 0.1096,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00019994385817114646,
+      "loss": 1.2834,
+      "step": 5480
+    },
+    {
+      "epoch": 0.1097,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00019994268257026118,
+      "loss": 1.2867,
+      "step": 5485
+    },
+    {
+      "epoch": 0.1098,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019994149479166324,
+      "loss": 1.2917,
+      "step": 5490
+    },
+    {
+      "epoch": 0.1099,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0001999402948354973,
+      "loss": 1.2716,
+      "step": 5495
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0001999390827019096,
+      "loss": 1.2686,
+      "step": 5500
+    },
+    {
+      "epoch": 0.1101,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001999378583910478,
+      "loss": 1.2878,
+      "step": 5505
+    },
+    {
+      "epoch": 0.1102,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001999366219030611,
+      "loss": 1.2779,
+      "step": 5510
+    },
+    {
+      "epoch": 0.1103,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019993537323810014,
+      "loss": 1.2869,
+      "step": 5515
+    },
+    {
+      "epoch": 0.1104,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001999341123963171,
+      "loss": 1.2714,
+      "step": 5520
+    },
+    {
+      "epoch": 0.1105,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019993283937786563,
+      "loss": 1.2605,
+      "step": 5525
+    },
+    {
+      "epoch": 0.1106,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001999315541829008,
+      "loss": 1.2887,
+      "step": 5530
+    },
+    {
+      "epoch": 0.1107,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0001999302568115792,
+      "loss": 1.2603,
+      "step": 5535
+    },
+    {
+      "epoch": 0.1108,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019992894726405893,
+      "loss": 1.2923,
+      "step": 5540
+    },
+    {
+      "epoch": 0.1109,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00019992762554049955,
+      "loss": 1.2802,
+      "step": 5545
+    },
+    {
+      "epoch": 0.111,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001999262916410621,
+      "loss": 1.2739,
+      "step": 5550
+    },
+    {
+      "epoch": 0.1111,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019992494556590916,
+      "loss": 1.2788,
+      "step": 5555
+    },
+    {
+      "epoch": 0.1112,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00019992358731520468,
+      "loss": 1.2814,
+      "step": 5560
+    },
+    {
+      "epoch": 0.1113,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001999222168891142,
+      "loss": 1.2621,
+      "step": 5565
+    },
+    {
+      "epoch": 0.1114,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001999208342878047,
+      "loss": 1.2716,
+      "step": 5570
+    },
+    {
+      "epoch": 0.1115,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.0001999194395114446,
+      "loss": 1.2746,
+      "step": 5575
+    },
+    {
+      "epoch": 0.1116,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00019991803256020393,
+      "loss": 1.2859,
+      "step": 5580
+    },
+    {
+      "epoch": 0.1117,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.000199916613434254,
+      "loss": 1.267,
+      "step": 5585
+    },
+    {
+      "epoch": 0.1118,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019991518213376787,
+      "loss": 1.2692,
+      "step": 5590
+    },
+    {
+      "epoch": 0.1119,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019991373865891986,
+      "loss": 1.2788,
+      "step": 5595
+    },
+    {
+      "epoch": 0.112,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019991228300988585,
+      "loss": 1.2673,
+      "step": 5600
+    },
+    {
+      "epoch": 0.1121,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00019991081518684321,
+      "loss": 1.2676,
+      "step": 5605
+    },
+    {
+      "epoch": 0.1122,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00019990933518997084,
+      "loss": 1.2798,
+      "step": 5610
+    },
+    {
+      "epoch": 0.1123,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00019990784301944902,
+      "loss": 1.2755,
+      "step": 5615
+    },
+    {
+      "epoch": 0.1124,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019990633867545955,
+      "loss": 1.2944,
+      "step": 5620
+    },
+    {
+      "epoch": 0.1125,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0001999048221581858,
+      "loss": 1.273,
+      "step": 5625
+    },
+    {
+      "epoch": 0.1126,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001999032934678125,
+      "loss": 1.2769,
+      "step": 5630
+    },
+    {
+      "epoch": 0.1127,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001999017526045259,
+      "loss": 1.2933,
+      "step": 5635
+    },
+    {
+      "epoch": 0.1128,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019990019956851382,
+      "loss": 1.2841,
+      "step": 5640
+    },
+    {
+      "epoch": 0.1129,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019989863435996544,
+      "loss": 1.2746,
+      "step": 5645
+    },
+    {
+      "epoch": 0.113,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019989705697907149,
+      "loss": 1.2842,
+      "step": 5650
+    },
+    {
+      "epoch": 0.1131,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019989546742602414,
+      "loss": 1.2839,
+      "step": 5655
+    },
+    {
+      "epoch": 0.1132,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019989386570101714,
+      "loss": 1.2773,
+      "step": 5660
+    },
+    {
+      "epoch": 0.1133,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001998922518042456,
+      "loss": 1.2757,
+      "step": 5665
+    },
+    {
+      "epoch": 0.1134,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00019989062573590616,
+      "loss": 1.2697,
+      "step": 5670
+    },
+    {
+      "epoch": 0.1135,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019988898749619702,
+      "loss": 1.2739,
+      "step": 5675
+    },
+    {
+      "epoch": 0.1136,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001998873370853177,
+      "loss": 1.2649,
+      "step": 5680
+    },
+    {
+      "epoch": 0.1137,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019988567450346937,
+      "loss": 1.2789,
+      "step": 5685
+    },
+    {
+      "epoch": 0.1138,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001998839997508546,
+      "loss": 1.2652,
+      "step": 5690
+    },
+    {
+      "epoch": 0.1139,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00019988231282767744,
+      "loss": 1.2893,
+      "step": 5695
+    },
+    {
+      "epoch": 0.114,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0001998806137341434,
+      "loss": 1.2777,
+      "step": 5700
+    },
+    {
+      "epoch": 0.1141,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00019987890247045957,
+      "loss": 1.2775,
+      "step": 5705
+    },
+    {
+      "epoch": 0.1142,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019987717903683448,
+      "loss": 1.284,
+      "step": 5710
+    },
+    {
+      "epoch": 0.1143,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019987544343347803,
+      "loss": 1.2738,
+      "step": 5715
+    },
+    {
+      "epoch": 0.1144,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00019987369566060176,
+      "loss": 1.2838,
+      "step": 5720
+    },
+    {
+      "epoch": 0.1145,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001998719357184186,
+      "loss": 1.2753,
+      "step": 5725
+    },
+    {
+      "epoch": 0.1146,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019987016360714307,
+      "loss": 1.2748,
+      "step": 5730
+    },
+    {
+      "epoch": 0.1147,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00019986837932699103,
+      "loss": 1.2825,
+      "step": 5735
+    },
+    {
+      "epoch": 0.1148,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00019986658287817987,
+      "loss": 1.2657,
+      "step": 5740
+    },
+    {
+      "epoch": 0.1149,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00019986477426092855,
+      "loss": 1.2667,
+      "step": 5745
+    },
+    {
+      "epoch": 0.115,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0001998629534754574,
+      "loss": 1.2755,
+      "step": 5750
+    },
+    {
+      "epoch": 0.1151,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001998611205219883,
+      "loss": 1.2707,
+      "step": 5755
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00019985927540074454,
+      "loss": 1.2767,
+      "step": 5760
+    },
+    {
+      "epoch": 0.1153,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00019985741811195097,
+      "loss": 1.2725,
+      "step": 5765
+    },
+    {
+      "epoch": 0.1154,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00019985554865583393,
+      "loss": 1.2636,
+      "step": 5770
+    },
+    {
+      "epoch": 0.1155,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001998536670326212,
+      "loss": 1.2739,
+      "step": 5775
+    },
+    {
+      "epoch": 0.1156,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.000199851773242542,
+      "loss": 1.2845,
+      "step": 5780
+    },
+    {
+      "epoch": 0.1157,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001998498672858271,
+      "loss": 1.2776,
+      "step": 5785
+    },
+    {
+      "epoch": 0.1158,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019984794916270876,
+      "loss": 1.2674,
+      "step": 5790
+    },
+    {
+      "epoch": 0.1159,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019984601887342073,
+      "loss": 1.266,
+      "step": 5795
+    },
+    {
+      "epoch": 0.116,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00019984407641819812,
+      "loss": 1.3164,
+      "step": 5800
+    },
+    {
+      "epoch": 0.1161,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019984212179727766,
+      "loss": 1.2683,
+      "step": 5805
+    },
+    {
+      "epoch": 0.1162,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019984015501089752,
+      "loss": 1.269,
+      "step": 5810
+    },
+    {
+      "epoch": 0.1163,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019983817605929733,
+      "loss": 1.293,
+      "step": 5815
+    },
+    {
+      "epoch": 0.1164,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019983618494271826,
+      "loss": 1.2658,
+      "step": 5820
+    },
+    {
+      "epoch": 0.1165,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00019983418166140285,
+      "loss": 1.2856,
+      "step": 5825
+    },
+    {
+      "epoch": 0.1166,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00019983216621559524,
+      "loss": 1.2774,
+      "step": 5830
+    },
+    {
+      "epoch": 0.1167,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019983013860554101,
+      "loss": 1.2688,
+      "step": 5835
+    },
+    {
+      "epoch": 0.1168,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019982809883148722,
+      "loss": 1.2848,
+      "step": 5840
+    },
+    {
+      "epoch": 0.1169,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00019982604689368239,
+      "loss": 1.2742,
+      "step": 5845
+    },
+    {
+      "epoch": 0.117,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019982398279237655,
+      "loss": 1.2914,
+      "step": 5850
+    },
+    {
+      "epoch": 0.1171,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001998219065278212,
+      "loss": 1.2717,
+      "step": 5855
+    },
+    {
+      "epoch": 0.1172,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00019981981810026934,
+      "loss": 1.2791,
+      "step": 5860
+    },
+    {
+      "epoch": 0.1173,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001998177175099754,
+      "loss": 1.2794,
+      "step": 5865
+    },
+    {
+      "epoch": 0.1174,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00019981560475719538,
+      "loss": 1.2734,
+      "step": 5870
+    },
+    {
+      "epoch": 0.1175,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.0001998134798421867,
+      "loss": 1.2802,
+      "step": 5875
+    },
+    {
+      "epoch": 0.1176,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00019981134276520827,
+      "loss": 1.2651,
+      "step": 5880
+    },
+    {
+      "epoch": 0.1177,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00019980919352652048,
+      "loss": 1.2764,
+      "step": 5885
+    },
+    {
+      "epoch": 0.1178,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019980703212638522,
+      "loss": 1.2844,
+      "step": 5890
+    },
+    {
+      "epoch": 0.1179,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00019980485856506582,
+      "loss": 1.2998,
+      "step": 5895
+    },
+    {
+      "epoch": 0.118,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00019980267284282717,
+      "loss": 1.2762,
+      "step": 5900
+    },
+    {
+      "epoch": 0.1181,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019980047495993554,
+      "loss": 1.2869,
+      "step": 5905
+    },
+    {
+      "epoch": 0.1182,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00019979826491665881,
+      "loss": 1.2879,
+      "step": 5910
+    },
+    {
+      "epoch": 0.1183,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00019979604271326616,
+      "loss": 1.2685,
+      "step": 5915
+    },
+    {
+      "epoch": 0.1184,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019979380835002846,
+      "loss": 1.2693,
+      "step": 5920
+    },
+    {
+      "epoch": 0.1185,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001997915618272179,
+      "loss": 1.2881,
+      "step": 5925
+    },
+    {
+      "epoch": 0.1186,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00019978930314510824,
+      "loss": 1.2841,
+      "step": 5930
+    },
+    {
+      "epoch": 0.1187,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001997870323039747,
+      "loss": 1.277,
+      "step": 5935
+    },
+    {
+      "epoch": 0.1188,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00019978474930409395,
+      "loss": 1.276,
+      "step": 5940
+    },
+    {
+      "epoch": 0.1189,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019978245414574417,
+      "loss": 1.2588,
+      "step": 5945
+    },
+    {
+      "epoch": 0.119,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.000199780146829205,
+      "loss": 1.2776,
+      "step": 5950
+    },
+    {
+      "epoch": 0.1191,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019977782735475764,
+      "loss": 1.279,
+      "step": 5955
+    },
+    {
+      "epoch": 0.1192,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00019977549572268468,
+      "loss": 1.2871,
+      "step": 5960
+    },
+    {
+      "epoch": 0.1193,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00019977315193327018,
+      "loss": 1.272,
+      "step": 5965
+    },
+    {
+      "epoch": 0.1194,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00019977079598679977,
+      "loss": 1.2711,
+      "step": 5970
+    },
+    {
+      "epoch": 0.1195,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00019976842788356055,
+      "loss": 1.2722,
+      "step": 5975
+    },
+    {
+      "epoch": 0.1196,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.000199766047623841,
+      "loss": 1.2719,
+      "step": 5980
+    },
+    {
+      "epoch": 0.1197,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019976365520793114,
+      "loss": 1.2691,
+      "step": 5985
+    },
+    {
+      "epoch": 0.1198,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00019976125063612252,
+      "loss": 1.264,
+      "step": 5990
+    },
+    {
+      "epoch": 0.1199,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00019975883390870817,
+      "loss": 1.2847,
+      "step": 5995
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00019975640502598244,
+      "loss": 1.2713,
+      "step": 6000
+    },
+    {
+      "epoch": 0.1201,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001997539639882414,
+      "loss": 1.2828,
+      "step": 6005
+    },
+    {
+      "epoch": 0.1202,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019975151079578237,
+      "loss": 1.2863,
+      "step": 6010
+    },
+    {
+      "epoch": 0.1203,
+      "grad_norm": 1.625,
+      "learning_rate": 0.0001997490454489044,
+      "loss": 1.2754,
+      "step": 6015
+    },
+    {
+      "epoch": 0.1204,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019974656794790775,
+      "loss": 1.2844,
+      "step": 6020
+    },
+    {
+      "epoch": 0.1205,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0001997440782930944,
+      "loss": 1.2768,
+      "step": 6025
+    },
+    {
+      "epoch": 0.1206,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00019974157648476766,
+      "loss": 1.2685,
+      "step": 6030
+    },
+    {
+      "epoch": 0.1207,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00019973906252323238,
+      "loss": 1.2796,
+      "step": 6035
+    },
+    {
+      "epoch": 0.1208,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00019973653640879485,
+      "loss": 1.2856,
+      "step": 6040
+    },
+    {
+      "epoch": 0.1209,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019973399814176293,
+      "loss": 1.266,
+      "step": 6045
+    },
+    {
+      "epoch": 0.121,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019973144772244582,
+      "loss": 1.2738,
+      "step": 6050
+    },
+    {
+      "epoch": 0.1211,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00019972888515115434,
+      "loss": 1.2819,
+      "step": 6055
+    },
+    {
+      "epoch": 0.1212,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001997263104282007,
+      "loss": 1.2663,
+      "step": 6060
+    },
+    {
+      "epoch": 0.1213,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00019972372355389867,
+      "loss": 1.2789,
+      "step": 6065
+    },
+    {
+      "epoch": 0.1214,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00019972112452856339,
+      "loss": 1.2835,
+      "step": 6070
+    },
+    {
+      "epoch": 0.1215,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019971851335251158,
+      "loss": 1.289,
+      "step": 6075
+    },
+    {
+      "epoch": 0.1216,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001997158900260614,
+      "loss": 1.2674,
+      "step": 6080
+    },
+    {
+      "epoch": 0.1217,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019971325454953248,
+      "loss": 1.281,
+      "step": 6085
+    },
+    {
+      "epoch": 0.1218,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00019971060692324598,
+      "loss": 1.2735,
+      "step": 6090
+    },
+    {
+      "epoch": 0.1219,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00019970794714752445,
+      "loss": 1.2822,
+      "step": 6095
+    },
+    {
+      "epoch": 0.122,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019970527522269205,
+      "loss": 1.2687,
+      "step": 6100
+    },
+    {
+      "epoch": 0.1221,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00019970259114907425,
+      "loss": 1.2857,
+      "step": 6105
+    },
+    {
+      "epoch": 0.1222,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001996998949269982,
+      "loss": 1.2624,
+      "step": 6110
+    },
+    {
+      "epoch": 0.1223,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00019969718655679232,
+      "loss": 1.2871,
+      "step": 6115
+    },
+    {
+      "epoch": 0.1224,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00019969446603878673,
+      "loss": 1.2774,
+      "step": 6120
+    },
+    {
+      "epoch": 0.1225,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001996917333733128,
+      "loss": 1.2725,
+      "step": 6125
+    },
+    {
+      "epoch": 0.1226,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001996889885607036,
+      "loss": 1.2749,
+      "step": 6130
+    },
+    {
+      "epoch": 0.1227,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001996862316012935,
+      "loss": 1.2779,
+      "step": 6135
+    },
+    {
+      "epoch": 0.1228,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00019968346249541846,
+      "loss": 1.2767,
+      "step": 6140
+    },
+    {
+      "epoch": 0.1229,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.0001996806812434159,
+      "loss": 1.2698,
+      "step": 6145
+    },
+    {
+      "epoch": 0.123,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019967788784562473,
+      "loss": 1.2687,
+      "step": 6150
+    },
+    {
+      "epoch": 0.1231,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019967508230238522,
+      "loss": 1.2724,
+      "step": 6155
+    },
+    {
+      "epoch": 0.1232,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00019967226461403933,
+      "loss": 1.2574,
+      "step": 6160
+    },
+    {
+      "epoch": 0.1233,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001996694347809303,
+      "loss": 1.2786,
+      "step": 6165
+    },
+    {
+      "epoch": 0.1234,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019966659280340297,
+      "loss": 1.285,
+      "step": 6170
+    },
+    {
+      "epoch": 0.1235,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00019966373868180365,
+      "loss": 1.2926,
+      "step": 6175
+    },
+    {
+      "epoch": 0.1236,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001996608724164801,
+      "loss": 1.2959,
+      "step": 6180
+    },
+    {
+      "epoch": 0.1237,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00019965799400778152,
+      "loss": 1.2703,
+      "step": 6185
+    },
+    {
+      "epoch": 0.1238,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00019965510345605866,
+      "loss": 1.295,
+      "step": 6190
+    },
+    {
+      "epoch": 0.1239,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00019965220076166376,
+      "loss": 1.281,
+      "step": 6195
+    },
+    {
+      "epoch": 0.124,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019964928592495045,
+      "loss": 1.2801,
+      "step": 6200
+    },
+    {
+      "epoch": 0.1241,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001996463589462739,
+      "loss": 1.2886,
+      "step": 6205
+    },
+    {
+      "epoch": 0.1242,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019964341982599078,
+      "loss": 1.2744,
+      "step": 6210
+    },
+    {
+      "epoch": 0.1243,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00019964046856445924,
+      "loss": 1.2862,
+      "step": 6215
+    },
+    {
+      "epoch": 0.1244,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019963750516203884,
+      "loss": 1.2711,
+      "step": 6220
+    },
+    {
+      "epoch": 0.1245,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00019963452961909063,
+      "loss": 1.2904,
+      "step": 6225
+    },
+    {
+      "epoch": 0.1246,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019963154193597727,
+      "loss": 1.2646,
+      "step": 6230
+    },
+    {
+      "epoch": 0.1247,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00019962854211306267,
+      "loss": 1.2782,
+      "step": 6235
+    },
+    {
+      "epoch": 0.1248,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001996255301507125,
+      "loss": 1.2882,
+      "step": 6240
+    },
+    {
+      "epoch": 0.1249,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001996225060492936,
+      "loss": 1.2678,
+      "step": 6245
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00019961946980917456,
+      "loss": 1.295,
+      "step": 6250
+    },
+    {
+      "epoch": 0.1251,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00019961642143072529,
+      "loss": 1.2697,
+      "step": 6255
+    },
+    {
+      "epoch": 0.1252,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00019961336091431727,
+      "loss": 1.2718,
+      "step": 6260
+    },
+    {
+      "epoch": 0.1253,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00019961028826032332,
+      "loss": 1.2631,
+      "step": 6265
+    },
+    {
+      "epoch": 0.1254,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00019960720346911797,
+      "loss": 1.2784,
+      "step": 6270
+    },
+    {
+      "epoch": 0.1255,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00019960410654107697,
+      "loss": 1.277,
+      "step": 6275
+    },
+    {
+      "epoch": 0.1256,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00019960099747657774,
+      "loss": 1.2752,
+      "step": 6280
+    },
+    {
+      "epoch": 0.1257,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00019959787627599906,
+      "loss": 1.2692,
+      "step": 6285
+    },
+    {
+      "epoch": 0.1258,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00019959474293972129,
+      "loss": 1.279,
+      "step": 6290
+    },
+    {
+      "epoch": 0.1259,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001995915974681262,
+      "loss": 1.2716,
+      "step": 6295
+    },
+    {
+      "epoch": 0.126,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019958843986159704,
+      "loss": 1.2663,
+      "step": 6300
+    },
+    {
+      "epoch": 0.1261,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019958527012051857,
+      "loss": 1.2761,
+      "step": 6305
+    },
+    {
+      "epoch": 0.1262,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.000199582088245277,
+      "loss": 1.2674,
+      "step": 6310
+    },
+    {
+      "epoch": 0.1263,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019957889423626005,
+      "loss": 1.2837,
+      "step": 6315
+    },
+    {
+      "epoch": 0.1264,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019957568809385694,
+      "loss": 1.2713,
+      "step": 6320
+    },
+    {
+      "epoch": 0.1265,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019957246981845822,
+      "loss": 1.2671,
+      "step": 6325
+    },
+    {
+      "epoch": 0.1266,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001995692394104561,
+      "loss": 1.2865,
+      "step": 6330
+    },
+    {
+      "epoch": 0.1267,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001995659968702442,
+      "loss": 1.2662,
+      "step": 6335
+    },
+    {
+      "epoch": 0.1268,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019956274219821757,
+      "loss": 1.2805,
+      "step": 6340
+    },
+    {
+      "epoch": 0.1269,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00019955947539477284,
+      "loss": 1.2808,
+      "step": 6345
+    },
+    {
+      "epoch": 0.127,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019955619646030802,
+      "loss": 1.2862,
+      "step": 6350
+    },
+    {
+      "epoch": 0.1271,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001995529053952226,
+      "loss": 1.297,
+      "step": 6355
+    },
+    {
+      "epoch": 0.1272,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001995496021999177,
+      "loss": 1.2751,
+      "step": 6360
+    },
+    {
+      "epoch": 0.1273,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00019954628687479572,
+      "loss": 1.268,
+      "step": 6365
+    },
+    {
+      "epoch": 0.1274,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019954295942026064,
+      "loss": 1.2707,
+      "step": 6370
+    },
+    {
+      "epoch": 0.1275,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00019953961983671788,
+      "loss": 1.2673,
+      "step": 6375
+    },
+    {
+      "epoch": 0.1276,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001995362681245744,
+      "loss": 1.2688,
+      "step": 6380
+    },
+    {
+      "epoch": 0.1277,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00019953290428423857,
+      "loss": 1.275,
+      "step": 6385
+    },
+    {
+      "epoch": 0.1278,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00019952952831612026,
+      "loss": 1.297,
+      "step": 6390
+    },
+    {
+      "epoch": 0.1279,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00019952614022063084,
+      "loss": 1.2874,
+      "step": 6395
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001995227399981831,
+      "loss": 1.2739,
+      "step": 6400
+    },
+    {
+      "epoch": 0.1281,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019951932764919144,
+      "loss": 1.255,
+      "step": 6405
+    },
+    {
+      "epoch": 0.1282,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.00019951590317407152,
+      "loss": 1.3022,
+      "step": 6410
+    },
+    {
+      "epoch": 0.1283,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001995124665732407,
+      "loss": 1.2898,
+      "step": 6415
+    },
+    {
+      "epoch": 0.1284,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00019950901784711764,
+      "loss": 1.2874,
+      "step": 6420
+    },
+    {
+      "epoch": 0.1285,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00019950555699612267,
+      "loss": 1.2764,
+      "step": 6425
+    },
+    {
+      "epoch": 0.1286,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019950208402067733,
+      "loss": 1.2817,
+      "step": 6430
+    },
+    {
+      "epoch": 0.1287,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00019949859892120491,
+      "loss": 1.2733,
+      "step": 6435
+    },
+    {
+      "epoch": 0.1288,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00019949510169813003,
+      "loss": 1.2853,
+      "step": 6440
+    },
+    {
+      "epoch": 0.1289,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001994915923518788,
+      "loss": 1.2874,
+      "step": 6445
+    },
+    {
+      "epoch": 0.129,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00019948807088287883,
+      "loss": 1.2663,
+      "step": 6450
+    },
+    {
+      "epoch": 0.1291,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019948453729155922,
+      "loss": 1.2738,
+      "step": 6455
+    },
+    {
+      "epoch": 0.1292,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00019948099157835047,
+      "loss": 1.2669,
+      "step": 6460
+    },
+    {
+      "epoch": 0.1293,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00019947743374368467,
+      "loss": 1.2709,
+      "step": 6465
+    },
+    {
+      "epoch": 0.1294,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00019947386378799532,
+      "loss": 1.2722,
+      "step": 6470
+    },
+    {
+      "epoch": 0.1295,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00019947028171171742,
+      "loss": 1.2701,
+      "step": 6475
+    },
+    {
+      "epoch": 0.1296,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00019946668751528744,
+      "loss": 1.2767,
+      "step": 6480
+    },
+    {
+      "epoch": 0.1297,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00019946308119914323,
+      "loss": 1.2815,
+      "step": 6485
+    },
+    {
+      "epoch": 0.1298,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00019945946276372434,
+      "loss": 1.2768,
+      "step": 6490
+    },
+    {
+      "epoch": 0.1299,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019945583220947158,
+      "loss": 1.2604,
+      "step": 6495
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00019945218953682734,
+      "loss": 1.2808,
+      "step": 6500
+    },
+    {
+      "epoch": 0.1301,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00019944853474623548,
+      "loss": 1.3083,
+      "step": 6505
+    },
+    {
+      "epoch": 0.1302,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00019944486783814134,
+      "loss": 1.2751,
+      "step": 6510
+    },
+    {
+      "epoch": 0.1303,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00019944118881299168,
+      "loss": 1.299,
+      "step": 6515
+    },
+    {
+      "epoch": 0.1304,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001994374976712348,
+      "loss": 1.2789,
+      "step": 6520
+    },
+    {
+      "epoch": 0.1305,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019943379441332047,
+      "loss": 1.2837,
+      "step": 6525
+    },
+    {
+      "epoch": 0.1306,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001994300790396999,
+      "loss": 1.2806,
+      "step": 6530
+    },
+    {
+      "epoch": 0.1307,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001994263515508258,
+      "loss": 1.2786,
+      "step": 6535
+    },
+    {
+      "epoch": 0.1308,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019942261194715236,
+      "loss": 1.2835,
+      "step": 6540
+    },
+    {
+      "epoch": 0.1309,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00019941886022913522,
+      "loss": 1.2828,
+      "step": 6545
+    },
+    {
+      "epoch": 0.131,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019941509639723155,
+      "loss": 1.2782,
+      "step": 6550
+    },
+    {
+      "epoch": 0.1311,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.0001994113204518999,
+      "loss": 1.2725,
+      "step": 6555
+    },
+    {
+      "epoch": 0.1312,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019940753239360047,
+      "loss": 1.263,
+      "step": 6560
+    },
+    {
+      "epoch": 0.1313,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00019940373222279473,
+      "loss": 1.2955,
+      "step": 6565
+    },
+    {
+      "epoch": 0.1314,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001993999199399457,
+      "loss": 1.2657,
+      "step": 6570
+    },
+    {
+      "epoch": 0.1315,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.000199396095545518,
+      "loss": 1.285,
+      "step": 6575
+    },
+    {
+      "epoch": 0.1316,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001993922590399775,
+      "loss": 1.2733,
+      "step": 6580
+    },
+    {
+      "epoch": 0.1317,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019938841042379174,
+      "loss": 1.2771,
+      "step": 6585
+    },
+    {
+      "epoch": 0.1318,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00019938454969742968,
+      "loss": 1.2789,
+      "step": 6590
+    },
+    {
+      "epoch": 0.1319,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019938067686136167,
+      "loss": 1.2737,
+      "step": 6595
+    },
+    {
+      "epoch": 0.132,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019937679191605963,
+      "loss": 1.2798,
+      "step": 6600
+    },
+    {
+      "epoch": 0.1321,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019937289486199696,
+      "loss": 1.2941,
+      "step": 6605
+    },
+    {
+      "epoch": 0.1322,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00019936898569964848,
+      "loss": 1.2696,
+      "step": 6610
+    },
+    {
+      "epoch": 0.1323,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.0001993650644294905,
+      "loss": 1.2899,
+      "step": 6615
+    },
+    {
+      "epoch": 0.1324,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00019936113105200085,
+      "loss": 1.2615,
+      "step": 6620
+    },
+    {
+      "epoch": 0.1325,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019935718556765876,
+      "loss": 1.2766,
+      "step": 6625
+    },
+    {
+      "epoch": 0.1326,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.000199353227976945,
+      "loss": 1.2847,
+      "step": 6630
+    },
+    {
+      "epoch": 0.1327,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00019934925828034175,
+      "loss": 1.2675,
+      "step": 6635
+    },
+    {
+      "epoch": 0.1328,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00019934527647833276,
+      "loss": 1.2825,
+      "step": 6640
+    },
+    {
+      "epoch": 0.1329,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001993412825714032,
+      "loss": 1.2788,
+      "step": 6645
+    },
+    {
+      "epoch": 0.133,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019933727656003963,
+      "loss": 1.2785,
+      "step": 6650
+    },
+    {
+      "epoch": 0.1331,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001993332584447303,
+      "loss": 1.2705,
+      "step": 6655
+    },
+    {
+      "epoch": 0.1332,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019932922822596473,
+      "loss": 1.2811,
+      "step": 6660
+    },
+    {
+      "epoch": 0.1333,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00019932518590423394,
+      "loss": 1.2811,
+      "step": 6665
+    },
+    {
+      "epoch": 0.1334,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019932113148003058,
+      "loss": 1.2809,
+      "step": 6670
+    },
+    {
+      "epoch": 0.1335,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00019931706495384863,
+      "loss": 1.2626,
+      "step": 6675
+    },
+    {
+      "epoch": 0.1336,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00019931298632618356,
+      "loss": 1.2561,
+      "step": 6680
+    },
+    {
+      "epoch": 0.1337,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00019930889559753234,
+      "loss": 1.2721,
+      "step": 6685
+    },
+    {
+      "epoch": 0.1338,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00019930479276839344,
+      "loss": 1.2692,
+      "step": 6690
+    },
+    {
+      "epoch": 0.1339,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00019930067783926675,
+      "loss": 1.2778,
+      "step": 6695
+    },
+    {
+      "epoch": 0.134,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0001992965508106537,
+      "loss": 1.2741,
+      "step": 6700
+    },
+    {
+      "epoch": 0.1341,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00019929241168305714,
+      "loss": 1.268,
+      "step": 6705
+    },
+    {
+      "epoch": 0.1342,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00019928826045698136,
+      "loss": 1.2796,
+      "step": 6710
+    },
+    {
+      "epoch": 0.1343,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00019928409713293227,
+      "loss": 1.2618,
+      "step": 6715
+    },
+    {
+      "epoch": 0.1344,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00019927992171141708,
+      "loss": 1.2777,
+      "step": 6720
+    },
+    {
+      "epoch": 0.1345,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00019927573419294456,
+      "loss": 1.2694,
+      "step": 6725
+    },
+    {
+      "epoch": 0.1346,
+      "grad_norm": 0.625,
+      "learning_rate": 0.000199271534578025,
+      "loss": 1.2753,
+      "step": 6730
+    },
+    {
+      "epoch": 0.1347,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00019926732286717003,
+      "loss": 1.2883,
+      "step": 6735
+    },
+    {
+      "epoch": 0.1348,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001992630990608929,
+      "loss": 1.2783,
+      "step": 6740
+    },
+    {
+      "epoch": 0.1349,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019925886315970824,
+      "loss": 1.2833,
+      "step": 6745
+    },
+    {
+      "epoch": 0.135,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00019925461516413223,
+      "loss": 1.2866,
+      "step": 6750
+    },
+    {
+      "epoch": 0.1351,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001992503550746824,
+      "loss": 1.2834,
+      "step": 6755
+    },
+    {
+      "epoch": 0.1352,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00019924608289187786,
+      "loss": 1.2979,
+      "step": 6760
+    },
+    {
+      "epoch": 0.1353,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00019924179861623915,
+      "loss": 1.2692,
+      "step": 6765
+    },
+    {
+      "epoch": 0.1354,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019923750224828832,
+      "loss": 1.2813,
+      "step": 6770
+    },
+    {
+      "epoch": 0.1355,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00019923319378854887,
+      "loss": 1.2697,
+      "step": 6775
+    },
+    {
+      "epoch": 0.1356,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00019922887323754577,
+      "loss": 1.2857,
+      "step": 6780
+    },
+    {
+      "epoch": 0.1357,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00019922454059580544,
+      "loss": 1.2899,
+      "step": 6785
+    },
+    {
+      "epoch": 0.1358,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019922019586385585,
+      "loss": 1.281,
+      "step": 6790
+    },
+    {
+      "epoch": 0.1359,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00019921583904222633,
+      "loss": 1.284,
+      "step": 6795
+    },
+    {
+      "epoch": 0.136,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001992114701314478,
+      "loss": 1.2644,
+      "step": 6800
+    },
+    {
+      "epoch": 0.1361,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019920708913205256,
+      "loss": 1.2802,
+      "step": 6805
+    },
+    {
+      "epoch": 0.1362,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00019920269604457446,
+      "loss": 1.2868,
+      "step": 6810
+    },
+    {
+      "epoch": 0.1363,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0001991982908695487,
+      "loss": 1.2693,
+      "step": 6815
+    },
+    {
+      "epoch": 0.1364,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00019919387360751217,
+      "loss": 1.2795,
+      "step": 6820
+    },
+    {
+      "epoch": 0.1365,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.000199189444259003,
+      "loss": 1.2731,
+      "step": 6825
+    },
+    {
+      "epoch": 0.1366,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001991850028245609,
+      "loss": 1.2826,
+      "step": 6830
+    },
+    {
+      "epoch": 0.1367,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00019918054930472706,
+      "loss": 1.2741,
+      "step": 6835
+    },
+    {
+      "epoch": 0.1368,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019917608370004417,
+      "loss": 1.2751,
+      "step": 6840
+    },
+    {
+      "epoch": 0.1369,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001991716060110563,
+      "loss": 1.2722,
+      "step": 6845
+    },
+    {
+      "epoch": 0.137,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019916711623830903,
+      "loss": 1.285,
+      "step": 6850
+    },
+    {
+      "epoch": 0.1371,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001991626143823495,
+      "loss": 1.2664,
+      "step": 6855
+    },
+    {
+      "epoch": 0.1372,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019915810044372618,
+      "loss": 1.276,
+      "step": 6860
+    },
+    {
+      "epoch": 0.1373,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001991535744229891,
+      "loss": 1.2868,
+      "step": 6865
+    },
+    {
+      "epoch": 0.1374,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019914903632068973,
+      "loss": 1.274,
+      "step": 6870
+    },
+    {
+      "epoch": 0.1375,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00019914448613738106,
+      "loss": 1.2954,
+      "step": 6875
+    },
+    {
+      "epoch": 0.1376,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00019913992387361745,
+      "loss": 1.2692,
+      "step": 6880
+    },
+    {
+      "epoch": 0.1377,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00019913534952995486,
+      "loss": 1.2752,
+      "step": 6885
+    },
+    {
+      "epoch": 0.1378,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00019913076310695068,
+      "loss": 1.2775,
+      "step": 6890
+    },
+    {
+      "epoch": 0.1379,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00019912616460516364,
+      "loss": 1.2662,
+      "step": 6895
+    },
+    {
+      "epoch": 0.138,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00019912155402515417,
+      "loss": 1.2895,
+      "step": 6900
+    },
+    {
+      "epoch": 0.1381,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00019911693136748403,
+      "loss": 1.2923,
+      "step": 6905
+    },
+    {
+      "epoch": 0.1382,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001991122966327164,
+      "loss": 1.2827,
+      "step": 6910
+    },
+    {
+      "epoch": 0.1383,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.0001991076498214161,
+      "loss": 1.2825,
+      "step": 6915
+    },
+    {
+      "epoch": 0.1384,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001991029909341493,
+      "loss": 1.272,
+      "step": 6920
+    },
+    {
+      "epoch": 0.1385,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00019909831997148362,
+      "loss": 1.2853,
+      "step": 6925
+    },
+    {
+      "epoch": 0.1386,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00019909363693398828,
+      "loss": 1.2671,
+      "step": 6930
+    },
+    {
+      "epoch": 0.1387,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019908894182223388,
+      "loss": 1.2566,
+      "step": 6935
+    },
+    {
+      "epoch": 0.1388,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00019908423463679248,
+      "loss": 1.2724,
+      "step": 6940
+    },
+    {
+      "epoch": 0.1389,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001990795153782376,
+      "loss": 1.2657,
+      "step": 6945
+    },
+    {
+      "epoch": 0.139,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019907478404714436,
+      "loss": 1.2619,
+      "step": 6950
+    },
+    {
+      "epoch": 0.1391,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001990700406440892,
+      "loss": 1.2818,
+      "step": 6955
+    },
+    {
+      "epoch": 0.1392,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00019906528516965008,
+      "loss": 1.2735,
+      "step": 6960
+    },
+    {
+      "epoch": 0.1393,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001990605176244065,
+      "loss": 1.2864,
+      "step": 6965
+    },
+    {
+      "epoch": 0.1394,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.0001990557380089393,
+      "loss": 1.2702,
+      "step": 6970
+    },
+    {
+      "epoch": 0.1395,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001990509463238309,
+      "loss": 1.2774,
+      "step": 6975
+    },
+    {
+      "epoch": 0.1396,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00019904614256966512,
+      "loss": 1.2736,
+      "step": 6980
+    },
+    {
+      "epoch": 0.1397,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00019904132674702734,
+      "loss": 1.2696,
+      "step": 6985
+    },
+    {
+      "epoch": 0.1398,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001990364988565043,
+      "loss": 1.2742,
+      "step": 6990
+    },
+    {
+      "epoch": 0.1399,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001990316588986843,
+      "loss": 1.2689,
+      "step": 6995
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019902680687415705,
+      "loss": 1.2777,
+      "step": 7000
+    },
+    {
+      "epoch": 0.1401,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019902194278351374,
+      "loss": 1.2715,
+      "step": 7005
+    },
+    {
+      "epoch": 0.1402,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019901706662734712,
+      "loss": 1.2777,
+      "step": 7010
+    },
+    {
+      "epoch": 0.1403,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001990121784062512,
+      "loss": 1.2792,
+      "step": 7015
+    },
+    {
+      "epoch": 0.1404,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00019900727812082177,
+      "loss": 1.2653,
+      "step": 7020
+    },
+    {
+      "epoch": 0.1405,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00019900236577165576,
+      "loss": 1.2901,
+      "step": 7025
+    },
+    {
+      "epoch": 0.1406,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001989974413593518,
+      "loss": 1.2692,
+      "step": 7030
+    },
+    {
+      "epoch": 0.1407,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001989925048845099,
+      "loss": 1.2786,
+      "step": 7035
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019898755634773158,
+      "loss": 1.2723,
+      "step": 7040
+    },
+    {
+      "epoch": 0.1409,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00019898259574961978,
+      "loss": 1.2632,
+      "step": 7045
+    },
+    {
+      "epoch": 0.141,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001989776230907789,
+      "loss": 1.2843,
+      "step": 7050
+    },
+    {
+      "epoch": 0.1411,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019897263837181491,
+      "loss": 1.2657,
+      "step": 7055
+    },
+    {
+      "epoch": 0.1412,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001989676415933351,
+      "loss": 1.2672,
+      "step": 7060
+    },
+    {
+      "epoch": 0.1413,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00019896263275594842,
+      "loss": 1.284,
+      "step": 7065
+    },
+    {
+      "epoch": 0.1414,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0001989576118602651,
+      "loss": 1.2754,
+      "step": 7070
+    },
+    {
+      "epoch": 0.1415,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00019895257890689696,
+      "loss": 1.2655,
+      "step": 7075
+    },
+    {
+      "epoch": 0.1416,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00019894753389645723,
+      "loss": 1.276,
+      "step": 7080
+    },
+    {
+      "epoch": 0.1417,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001989424768295606,
+      "loss": 1.2696,
+      "step": 7085
+    },
+    {
+      "epoch": 0.1418,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00019893740770682335,
+      "loss": 1.2887,
+      "step": 7090
+    },
+    {
+      "epoch": 0.1419,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019893232652886306,
+      "loss": 1.2578,
+      "step": 7095
+    },
+    {
+      "epoch": 0.142,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019892723329629887,
+      "loss": 1.2602,
+      "step": 7100
+    },
+    {
+      "epoch": 0.1421,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00019892212800975135,
+      "loss": 1.2761,
+      "step": 7105
+    },
+    {
+      "epoch": 0.1422,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019891701066984262,
+      "loss": 1.2788,
+      "step": 7110
+    },
+    {
+      "epoch": 0.1423,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00019891188127719618,
+      "loss": 1.2584,
+      "step": 7115
+    },
+    {
+      "epoch": 0.1424,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019890673983243706,
+      "loss": 1.2713,
+      "step": 7120
+    },
+    {
+      "epoch": 0.1425,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001989015863361917,
+      "loss": 1.2837,
+      "step": 7125
+    },
+    {
+      "epoch": 0.1426,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00019889642078908804,
+      "loss": 1.2787,
+      "step": 7130
+    },
+    {
+      "epoch": 0.1427,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019889124319175547,
+      "loss": 1.2776,
+      "step": 7135
+    },
+    {
+      "epoch": 0.1428,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001988860535448249,
+      "loss": 1.2767,
+      "step": 7140
+    },
+    {
+      "epoch": 0.1429,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019888085184892868,
+      "loss": 1.2831,
+      "step": 7145
+    },
+    {
+      "epoch": 0.143,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001988756381047006,
+      "loss": 1.284,
+      "step": 7150
+    },
+    {
+      "epoch": 0.1431,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00019887041231277593,
+      "loss": 1.2633,
+      "step": 7155
+    },
+    {
+      "epoch": 0.1432,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0001988651744737914,
+      "loss": 1.266,
+      "step": 7160
+    },
+    {
+      "epoch": 0.1433,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019885992458838528,
+      "loss": 1.2667,
+      "step": 7165
+    },
+    {
+      "epoch": 0.1434,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001988546626571972,
+      "loss": 1.2821,
+      "step": 7170
+    },
+    {
+      "epoch": 0.1435,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019884938868086835,
+      "loss": 1.2702,
+      "step": 7175
+    },
+    {
+      "epoch": 0.1436,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019884410266004135,
+      "loss": 1.2685,
+      "step": 7180
+    },
+    {
+      "epoch": 0.1437,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019883880459536024,
+      "loss": 1.2713,
+      "step": 7185
+    },
+    {
+      "epoch": 0.1438,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00019883349448747062,
+      "loss": 1.2683,
+      "step": 7190
+    },
+    {
+      "epoch": 0.1439,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00019882817233701948,
+      "loss": 1.2751,
+      "step": 7195
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001988228381446553,
+      "loss": 1.2719,
+      "step": 7200
+    },
+    {
+      "epoch": 0.1441,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019881749191102808,
+      "loss": 1.2806,
+      "step": 7205
+    },
+    {
+      "epoch": 0.1442,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001988121336367892,
+      "loss": 1.2749,
+      "step": 7210
+    },
+    {
+      "epoch": 0.1443,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00019880676332259154,
+      "loss": 1.2742,
+      "step": 7215
+    },
+    {
+      "epoch": 0.1444,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00019880138096908952,
+      "loss": 1.2716,
+      "step": 7220
+    },
+    {
+      "epoch": 0.1445,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00019879598657693891,
+      "loss": 1.2861,
+      "step": 7225
+    },
+    {
+      "epoch": 0.1446,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00019879058014679704,
+      "loss": 1.2842,
+      "step": 7230
+    },
+    {
+      "epoch": 0.1447,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019878516167932261,
+      "loss": 1.2907,
+      "step": 7235
+    },
+    {
+      "epoch": 0.1448,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001987797311751759,
+      "loss": 1.2844,
+      "step": 7240
+    },
+    {
+      "epoch": 0.1449,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019877428863501856,
+      "loss": 1.2769,
+      "step": 7245
+    },
+    {
+      "epoch": 0.145,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019876883405951377,
+      "loss": 1.2741,
+      "step": 7250
+    },
+    {
+      "epoch": 0.1451,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00019876336744932614,
+      "loss": 1.3016,
+      "step": 7255
+    },
+    {
+      "epoch": 0.1452,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.0001987578888051218,
+      "loss": 1.2728,
+      "step": 7260
+    },
+    {
+      "epoch": 0.1453,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019875239812756825,
+      "loss": 1.2737,
+      "step": 7265
+    },
+    {
+      "epoch": 0.1454,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00019874689541733457,
+      "loss": 1.2758,
+      "step": 7270
+    },
+    {
+      "epoch": 0.1455,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019874138067509117,
+      "loss": 1.2678,
+      "step": 7275
+    },
+    {
+      "epoch": 0.1456,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019873585390151003,
+      "loss": 1.2761,
+      "step": 7280
+    },
+    {
+      "epoch": 0.1457,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019873031509726462,
+      "loss": 1.2609,
+      "step": 7285
+    },
+    {
+      "epoch": 0.1458,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00019872476426302982,
+      "loss": 1.2916,
+      "step": 7290
+    },
+    {
+      "epoch": 0.1459,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019871920139948192,
+      "loss": 1.2793,
+      "step": 7295
+    },
+    {
+      "epoch": 0.146,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001987136265072988,
+      "loss": 1.2919,
+      "step": 7300
+    },
+    {
+      "epoch": 0.1461,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019870803958715972,
+      "loss": 1.2722,
+      "step": 7305
+    },
+    {
+      "epoch": 0.1462,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001987024406397454,
+      "loss": 1.2782,
+      "step": 7310
+    },
+    {
+      "epoch": 0.1463,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019869682966573813,
+      "loss": 1.2742,
+      "step": 7315
+    },
+    {
+      "epoch": 0.1464,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019869120666582153,
+      "loss": 1.2674,
+      "step": 7320
+    },
+    {
+      "epoch": 0.1465,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00019868557164068074,
+      "loss": 1.2903,
+      "step": 7325
+    },
+    {
+      "epoch": 0.1466,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001986799245910024,
+      "loss": 1.2872,
+      "step": 7330
+    },
+    {
+      "epoch": 0.1467,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00019867426551747457,
+      "loss": 1.2782,
+      "step": 7335
+    },
+    {
+      "epoch": 0.1468,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001986685944207868,
+      "loss": 1.3026,
+      "step": 7340
+    },
+    {
+      "epoch": 0.1469,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001986629113016301,
+      "loss": 1.2609,
+      "step": 7345
+    },
+    {
+      "epoch": 0.147,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00019865721616069696,
+      "loss": 1.2917,
+      "step": 7350
+    },
+    {
+      "epoch": 0.1471,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00019865150899868125,
+      "loss": 1.269,
+      "step": 7355
+    },
+    {
+      "epoch": 0.1472,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019864578981627844,
+      "loss": 1.2762,
+      "step": 7360
+    },
+    {
+      "epoch": 0.1473,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019864005861418535,
+      "loss": 1.2757,
+      "step": 7365
+    },
+    {
+      "epoch": 0.1474,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001986343153931003,
+      "loss": 1.2776,
+      "step": 7370
+    },
+    {
+      "epoch": 0.1475,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00019862856015372317,
+      "loss": 1.2751,
+      "step": 7375
+    },
+    {
+      "epoch": 0.1476,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00019862279289675509,
+      "loss": 1.2778,
+      "step": 7380
+    },
+    {
+      "epoch": 0.1477,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001986170136228989,
+      "loss": 1.2864,
+      "step": 7385
+    },
+    {
+      "epoch": 0.1478,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001986112223328587,
+      "loss": 1.2736,
+      "step": 7390
+    },
+    {
+      "epoch": 0.1479,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019860541902734022,
+      "loss": 1.2676,
+      "step": 7395
+    },
+    {
+      "epoch": 0.148,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0001985996037070505,
+      "loss": 1.2848,
+      "step": 7400
+    },
+    {
+      "epoch": 0.1481,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00019859377637269815,
+      "loss": 1.2613,
+      "step": 7405
+    },
+    {
+      "epoch": 0.1482,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019858793702499323,
+      "loss": 1.2707,
+      "step": 7410
+    },
+    {
+      "epoch": 0.1483,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019858208566464724,
+      "loss": 1.277,
+      "step": 7415
+    },
+    {
+      "epoch": 0.1484,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019857622229237313,
+      "loss": 1.2874,
+      "step": 7420
+    },
+    {
+      "epoch": 0.1485,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00019857034690888537,
+      "loss": 1.2708,
+      "step": 7425
+    },
+    {
+      "epoch": 0.1486,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019856445951489982,
+      "loss": 1.2689,
+      "step": 7430
+    },
+    {
+      "epoch": 0.1487,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019855856011113384,
+      "loss": 1.2771,
+      "step": 7435
+    },
+    {
+      "epoch": 0.1488,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019855264869830629,
+      "loss": 1.2873,
+      "step": 7440
+    },
+    {
+      "epoch": 0.1489,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019854672527713744,
+      "loss": 1.2584,
+      "step": 7445
+    },
+    {
+      "epoch": 0.149,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019854078984834903,
+      "loss": 1.2604,
+      "step": 7450
+    },
+    {
+      "epoch": 0.1491,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00019853484241266428,
+      "loss": 1.2931,
+      "step": 7455
+    },
+    {
+      "epoch": 0.1492,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019852888297080786,
+      "loss": 1.2804,
+      "step": 7460
+    },
+    {
+      "epoch": 0.1493,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019852291152350592,
+      "loss": 1.2759,
+      "step": 7465
+    },
+    {
+      "epoch": 0.1494,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001985169280714861,
+      "loss": 1.2763,
+      "step": 7470
+    },
+    {
+      "epoch": 0.1495,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001985109326154774,
+      "loss": 1.2735,
+      "step": 7475
+    },
+    {
+      "epoch": 0.1496,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00019850492515621038,
+      "loss": 1.2723,
+      "step": 7480
+    },
+    {
+      "epoch": 0.1497,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00019849890569441703,
+      "loss": 1.2841,
+      "step": 7485
+    },
+    {
+      "epoch": 0.1498,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019849287423083078,
+      "loss": 1.2814,
+      "step": 7490
+    },
+    {
+      "epoch": 0.1499,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00019848683076618658,
+      "loss": 1.2645,
+      "step": 7495
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019848077530122083,
+      "loss": 1.2775,
+      "step": 7500
+    },
+    {
+      "epoch": 0.1501,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00019847470783667127,
+      "loss": 1.285,
+      "step": 7505
+    },
+    {
+      "epoch": 0.1502,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001984686283732773,
+      "loss": 1.2755,
+      "step": 7510
+    },
+    {
+      "epoch": 0.1503,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00019846253691177966,
+      "loss": 1.2793,
+      "step": 7515
+    },
+    {
+      "epoch": 0.1504,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019845643345292054,
+      "loss": 1.2707,
+      "step": 7520
+    },
+    {
+      "epoch": 0.1505,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019845031799744367,
+      "loss": 1.2776,
+      "step": 7525
+    },
+    {
+      "epoch": 0.1506,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001984441905460942,
+      "loss": 1.2858,
+      "step": 7530
+    },
+    {
+      "epoch": 0.1507,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00019843805109961868,
+      "loss": 1.2789,
+      "step": 7535
+    },
+    {
+      "epoch": 0.1508,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019843189965876526,
+      "loss": 1.29,
+      "step": 7540
+    },
+    {
+      "epoch": 0.1509,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00019842573622428345,
+      "loss": 1.278,
+      "step": 7545
+    },
+    {
+      "epoch": 0.151,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0001984195607969242,
+      "loss": 1.2715,
+      "step": 7550
+    },
+    {
+      "epoch": 0.1511,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00019841337337744004,
+      "loss": 1.2804,
+      "step": 7555
+    },
+    {
+      "epoch": 0.1512,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019840717396658484,
+      "loss": 1.2683,
+      "step": 7560
+    },
+    {
+      "epoch": 0.1513,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019840096256511398,
+      "loss": 1.2714,
+      "step": 7565
+    },
+    {
+      "epoch": 0.1514,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019839473917378434,
+      "loss": 1.2719,
+      "step": 7570
+    },
+    {
+      "epoch": 0.1515,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00019838850379335417,
+      "loss": 1.2706,
+      "step": 7575
+    },
+    {
+      "epoch": 0.1516,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00019838225642458327,
+      "loss": 1.2822,
+      "step": 7580
+    },
+    {
+      "epoch": 0.1517,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019837599706823284,
+      "loss": 1.2907,
+      "step": 7585
+    },
+    {
+      "epoch": 0.1518,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00019836972572506557,
+      "loss": 1.2777,
+      "step": 7590
+    },
+    {
+      "epoch": 0.1519,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019836344239584564,
+      "loss": 1.2679,
+      "step": 7595
+    },
+    {
+      "epoch": 0.152,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019835714708133862,
+      "loss": 1.2955,
+      "step": 7600
+    },
+    {
+      "epoch": 0.1521,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00019835083978231156,
+      "loss": 1.2864,
+      "step": 7605
+    },
+    {
+      "epoch": 0.1522,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00019834452049953297,
+      "loss": 1.2792,
+      "step": 7610
+    },
+    {
+      "epoch": 0.1523,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0001983381892337729,
+      "loss": 1.2934,
+      "step": 7615
+    },
+    {
+      "epoch": 0.1524,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019833184598580276,
+      "loss": 1.2734,
+      "step": 7620
+    },
+    {
+      "epoch": 0.1525,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0001983254907563955,
+      "loss": 1.2608,
+      "step": 7625
+    },
+    {
+      "epoch": 0.1526,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00019831912354632535,
+      "loss": 1.2693,
+      "step": 7630
+    },
+    {
+      "epoch": 0.1527,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001983127443563683,
+      "loss": 1.2563,
+      "step": 7635
+    },
+    {
+      "epoch": 0.1528,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019830635318730154,
+      "loss": 1.2838,
+      "step": 7640
+    },
+    {
+      "epoch": 0.1529,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019829995003990388,
+      "loss": 1.2792,
+      "step": 7645
+    },
+    {
+      "epoch": 0.153,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019829353491495545,
+      "loss": 1.273,
+      "step": 7650
+    },
+    {
+      "epoch": 0.1531,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019828710781323792,
+      "loss": 1.2776,
+      "step": 7655
+    },
+    {
+      "epoch": 0.1532,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019828066873553448,
+      "loss": 1.2829,
+      "step": 7660
+    },
+    {
+      "epoch": 0.1533,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019827421768262967,
+      "loss": 1.2824,
+      "step": 7665
+    },
+    {
+      "epoch": 0.1534,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0001982677546553095,
+      "loss": 1.2714,
+      "step": 7670
+    },
+    {
+      "epoch": 0.1535,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00019826127965436152,
+      "loss": 1.2692,
+      "step": 7675
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019825479268057467,
+      "loss": 1.2581,
+      "step": 7680
+    },
+    {
+      "epoch": 0.1537,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001982482937347394,
+      "loss": 1.2773,
+      "step": 7685
+    },
+    {
+      "epoch": 0.1538,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00019824178281764753,
+      "loss": 1.2617,
+      "step": 7690
+    },
+    {
+      "epoch": 0.1539,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019823525993009243,
+      "loss": 1.2676,
+      "step": 7695
+    },
+    {
+      "epoch": 0.154,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001982287250728689,
+      "loss": 1.2746,
+      "step": 7700
+    },
+    {
+      "epoch": 0.1541,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019822217824677315,
+      "loss": 1.301,
+      "step": 7705
+    },
+    {
+      "epoch": 0.1542,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001982156194526029,
+      "loss": 1.2803,
+      "step": 7710
+    },
+    {
+      "epoch": 0.1543,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001982090486911574,
+      "loss": 1.303,
+      "step": 7715
+    },
+    {
+      "epoch": 0.1544,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001982024659632372,
+      "loss": 1.2989,
+      "step": 7720
+    },
+    {
+      "epoch": 0.1545,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019819587126964437,
+      "loss": 1.2836,
+      "step": 7725
+    },
+    {
+      "epoch": 0.1546,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019818926461118253,
+      "loss": 1.27,
+      "step": 7730
+    },
+    {
+      "epoch": 0.1547,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0001981826459886566,
+      "loss": 1.2927,
+      "step": 7735
+    },
+    {
+      "epoch": 0.1548,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00019817601540287306,
+      "loss": 1.272,
+      "step": 7740
+    },
+    {
+      "epoch": 0.1549,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001981693728546399,
+      "loss": 1.2876,
+      "step": 7745
+    },
+    {
+      "epoch": 0.155,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019816271834476642,
+      "loss": 1.2708,
+      "step": 7750
+    },
+    {
+      "epoch": 0.1551,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019815605187406345,
+      "loss": 1.266,
+      "step": 7755
+    },
+    {
+      "epoch": 0.1552,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001981493734433433,
+      "loss": 1.2833,
+      "step": 7760
+    },
+    {
+      "epoch": 0.1553,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001981426830534197,
+      "loss": 1.2736,
+      "step": 7765
+    },
+    {
+      "epoch": 0.1554,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00019813598070510792,
+      "loss": 1.2893,
+      "step": 7770
+    },
+    {
+      "epoch": 0.1555,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001981292663992245,
+      "loss": 1.2617,
+      "step": 7775
+    },
+    {
+      "epoch": 0.1556,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00019812254013658768,
+      "loss": 1.2591,
+      "step": 7780
+    },
+    {
+      "epoch": 0.1557,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019811580191801697,
+      "loss": 1.2719,
+      "step": 7785
+    },
+    {
+      "epoch": 0.1558,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0001981090517443334,
+      "loss": 1.2924,
+      "step": 7790
+    },
+    {
+      "epoch": 0.1559,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001981022896163595,
+      "loss": 1.2779,
+      "step": 7795
+    },
+    {
+      "epoch": 0.156,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00019809551553491916,
+      "loss": 1.2872,
+      "step": 7800
+    },
+    {
+      "epoch": 0.1561,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00019808872950083782,
+      "loss": 1.2495,
+      "step": 7805
+    },
+    {
+      "epoch": 0.1562,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019808193151494232,
+      "loss": 1.26,
+      "step": 7810
+    },
+    {
+      "epoch": 0.1563,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.000198075121578061,
+      "loss": 1.2997,
+      "step": 7815
+    },
+    {
+      "epoch": 0.1564,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00019806829969102357,
+      "loss": 1.2813,
+      "step": 7820
+    },
+    {
+      "epoch": 0.1565,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001980614658546613,
+      "loss": 1.281,
+      "step": 7825
+    },
+    {
+      "epoch": 0.1566,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019805462006980689,
+      "loss": 1.285,
+      "step": 7830
+    },
+    {
+      "epoch": 0.1567,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019804776233729444,
+      "loss": 1.269,
+      "step": 7835
+    },
+    {
+      "epoch": 0.1568,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0001980408926579596,
+      "loss": 1.2773,
+      "step": 7840
+    },
+    {
+      "epoch": 0.1569,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019803401103263933,
+      "loss": 1.2831,
+      "step": 7845
+    },
+    {
+      "epoch": 0.157,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00019802711746217218,
+      "loss": 1.2593,
+      "step": 7850
+    },
+    {
+      "epoch": 0.1571,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00019802021194739814,
+      "loss": 1.2795,
+      "step": 7855
+    },
+    {
+      "epoch": 0.1572,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00019801329448915862,
+      "loss": 1.275,
+      "step": 7860
+    },
+    {
+      "epoch": 0.1573,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019800636508829643,
+      "loss": 1.2819,
+      "step": 7865
+    },
+    {
+      "epoch": 0.1574,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00019799942374565597,
+      "loss": 1.2757,
+      "step": 7870
+    },
+    {
+      "epoch": 0.1575,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00019799247046208297,
+      "loss": 1.2702,
+      "step": 7875
+    },
+    {
+      "epoch": 0.1576,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001979855052384247,
+      "loss": 1.2855,
+      "step": 7880
+    },
+    {
+      "epoch": 0.1577,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019797852807552983,
+      "loss": 1.2784,
+      "step": 7885
+    },
+    {
+      "epoch": 0.1578,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00019797153897424852,
+      "loss": 1.264,
+      "step": 7890
+    },
+    {
+      "epoch": 0.1579,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00019796453793543238,
+      "loss": 1.2863,
+      "step": 7895
+    },
+    {
+      "epoch": 0.158,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.0001979575249599344,
+      "loss": 1.2745,
+      "step": 7900
+    },
+    {
+      "epoch": 0.1581,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019795050004860917,
+      "loss": 1.2763,
+      "step": 7905
+    },
+    {
+      "epoch": 0.1582,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00019794346320231265,
+      "loss": 1.2931,
+      "step": 7910
+    },
+    {
+      "epoch": 0.1583,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019793641442190221,
+      "loss": 1.2949,
+      "step": 7915
+    },
+    {
+      "epoch": 0.1584,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019792935370823675,
+      "loss": 1.2758,
+      "step": 7920
+    },
+    {
+      "epoch": 0.1585,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019792228106217658,
+      "loss": 1.2933,
+      "step": 7925
+    },
+    {
+      "epoch": 0.1586,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00019791519648458352,
+      "loss": 1.2848,
+      "step": 7930
+    },
+    {
+      "epoch": 0.1587,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019790809997632076,
+      "loss": 1.2898,
+      "step": 7935
+    },
+    {
+      "epoch": 0.1588,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.000197900991538253,
+      "loss": 1.2642,
+      "step": 7940
+    },
+    {
+      "epoch": 0.1589,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00019789387117124637,
+      "loss": 1.2602,
+      "step": 7945
+    },
+    {
+      "epoch": 0.159,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001978867388761685,
+      "loss": 1.2823,
+      "step": 7950
+    },
+    {
+      "epoch": 0.1591,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00019787959465388842,
+      "loss": 1.262,
+      "step": 7955
+    },
+    {
+      "epoch": 0.1592,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00019787243850527664,
+      "loss": 1.2793,
+      "step": 7960
+    },
+    {
+      "epoch": 0.1593,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0001978652704312051,
+      "loss": 1.2789,
+      "step": 7965
+    },
+    {
+      "epoch": 0.1594,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019785809043254722,
+      "loss": 1.2733,
+      "step": 7970
+    },
+    {
+      "epoch": 0.1595,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00019785089851017787,
+      "loss": 1.2956,
+      "step": 7975
+    },
+    {
+      "epoch": 0.1596,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0001978436946649733,
+      "loss": 1.2872,
+      "step": 7980
+    },
+    {
+      "epoch": 0.1597,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00019783647889781136,
+      "loss": 1.2763,
+      "step": 7985
+    },
+    {
+      "epoch": 0.1598,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00019782925120957124,
+      "loss": 1.2779,
+      "step": 7990
+    },
+    {
+      "epoch": 0.1599,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0001978220116011336,
+      "loss": 1.2777,
+      "step": 7995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00019781476007338058,
+      "loss": 1.2688,
+      "step": 8000
+    },
+    {
+      "epoch": 0.1601,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00019780749662719573,
+      "loss": 1.2899,
+      "step": 8005
+    },
+    {
+      "epoch": 0.1602,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0001978002212634641,
+      "loss": 1.2675,
+      "step": 8010
+    },
+    {
+      "epoch": 0.1603,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0001977929339830722,
+      "loss": 1.2669,
+      "step": 8015
+    },
+    {
+      "epoch": 0.1604,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0001977856347869079,
+      "loss": 1.2745,
+      "step": 8020
+    },
+    {
+      "epoch": 0.1605,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00019777832367586063,
+      "loss": 1.2753,
+      "step": 8025
+    },
+    {
+      "epoch": 0.1606,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00019777100065082118,
+      "loss": 1.2677,
+      "step": 8030
+    },
+    {
+      "epoch": 0.1607,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00019776366571268192,
+      "loss": 1.2703,
+      "step": 8035
+    },
+    {
+      "epoch": 0.1608,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00019775631886233654,
+      "loss": 1.2799,
+      "step": 8040
+    },
+    {
+      "epoch": 0.1609,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0001977489601006802,
+      "loss": 1.2672,
+      "step": 8045
+    },
+    {
+      "epoch": 0.161,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001977415894286096,
+      "loss": 1.2823,
+      "step": 8050
+    },
+    {
+      "epoch": 0.1611,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001977342068470228,
+      "loss": 1.281,
+      "step": 8055
+    },
+    {
+      "epoch": 0.1612,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019772681235681936,
+      "loss": 1.2685,
+      "step": 8060
+    },
+    {
+      "epoch": 0.1613,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00019771940595890027,
+      "loss": 1.2692,
+      "step": 8065
+    },
+    {
+      "epoch": 0.1614,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.000197711987654168,
+      "loss": 1.2746,
+      "step": 8070
+    },
+    {
+      "epoch": 0.1615,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001977045574435264,
+      "loss": 1.2924,
+      "step": 8075
+    },
+    {
+      "epoch": 0.1616,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019769711532788083,
+      "loss": 1.2779,
+      "step": 8080
+    },
+    {
+      "epoch": 0.1617,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0001976896613081381,
+      "loss": 1.2569,
+      "step": 8085
+    },
+    {
+      "epoch": 0.1618,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001976821953852065,
+      "loss": 1.2714,
+      "step": 8090
+    },
+    {
+      "epoch": 0.1619,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001976747175599957,
+      "loss": 1.2942,
+      "step": 8095
+    },
+    {
+      "epoch": 0.162,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0001976672278334168,
+      "loss": 1.2866,
+      "step": 8100
+    },
+    {
+      "epoch": 0.1621,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019765972620638248,
+      "loss": 1.274,
+      "step": 8105
+    },
+    {
+      "epoch": 0.1622,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019765221267980675,
+      "loss": 1.2711,
+      "step": 8110
+    },
+    {
+      "epoch": 0.1623,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019764468725460508,
+      "loss": 1.2655,
+      "step": 8115
+    },
+    {
+      "epoch": 0.1624,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019763714993169452,
+      "loss": 1.2583,
+      "step": 8120
+    },
+    {
+      "epoch": 0.1625,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00019762960071199333,
+      "loss": 1.2796,
+      "step": 8125
+    },
+    {
+      "epoch": 0.1626,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001976220395964215,
+      "loss": 1.2738,
+      "step": 8130
+    },
+    {
+      "epoch": 0.1627,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019761446658590024,
+      "loss": 1.2649,
+      "step": 8135
+    },
+    {
+      "epoch": 0.1628,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00019760688168135232,
+      "loss": 1.265,
+      "step": 8140
+    },
+    {
+      "epoch": 0.1629,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00019759928488370193,
+      "loss": 1.2762,
+      "step": 8145
+    },
+    {
+      "epoch": 0.163,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00019759167619387476,
+      "loss": 1.2712,
+      "step": 8150
+    },
+    {
+      "epoch": 0.1631,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019758405561279784,
+      "loss": 1.2811,
+      "step": 8155
+    },
+    {
+      "epoch": 0.1632,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00019757642314139977,
+      "loss": 1.2614,
+      "step": 8160
+    },
+    {
+      "epoch": 0.1633,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00019756877878061052,
+      "loss": 1.2688,
+      "step": 8165
+    },
+    {
+      "epoch": 0.1634,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001975611225313615,
+      "loss": 1.2963,
+      "step": 8170
+    },
+    {
+      "epoch": 0.1635,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00019755345439458565,
+      "loss": 1.2844,
+      "step": 8175
+    },
+    {
+      "epoch": 0.1636,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00019754577437121733,
+      "loss": 1.2774,
+      "step": 8180
+    },
+    {
+      "epoch": 0.1637,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019753808246219224,
+      "loss": 1.2811,
+      "step": 8185
+    },
+    {
+      "epoch": 0.1638,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019753037866844771,
+      "loss": 1.2653,
+      "step": 8190
+    },
+    {
+      "epoch": 0.1639,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00019752266299092236,
+      "loss": 1.2688,
+      "step": 8195
+    },
+    {
+      "epoch": 0.164,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019751493543055632,
+      "loss": 1.2754,
+      "step": 8200
+    },
+    {
+      "epoch": 0.1641,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019750719598829122,
+      "loss": 1.2814,
+      "step": 8205
+    },
+    {
+      "epoch": 0.1642,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00019749944466507008,
+      "loss": 1.2794,
+      "step": 8210
+    },
+    {
+      "epoch": 0.1643,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019749168146183731,
+      "loss": 1.2734,
+      "step": 8215
+    },
+    {
+      "epoch": 0.1644,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001974839063795389,
+      "loss": 1.2838,
+      "step": 8220
+    },
+    {
+      "epoch": 0.1645,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001974761194191222,
+      "loss": 1.276,
+      "step": 8225
+    },
+    {
+      "epoch": 0.1646,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019746832058153602,
+      "loss": 1.2771,
+      "step": 8230
+    },
+    {
+      "epoch": 0.1647,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001974605098677306,
+      "loss": 1.2693,
+      "step": 8235
+    },
+    {
+      "epoch": 0.1648,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00019745268727865774,
+      "loss": 1.2638,
+      "step": 8240
+    },
+    {
+      "epoch": 0.1649,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019744485281527049,
+      "loss": 1.2823,
+      "step": 8245
+    },
+    {
+      "epoch": 0.165,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019743700647852354,
+      "loss": 1.2766,
+      "step": 8250
+    },
+    {
+      "epoch": 0.1651,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019742914826937288,
+      "loss": 1.2935,
+      "step": 8255
+    },
+    {
+      "epoch": 0.1652,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00019742127818877606,
+      "loss": 1.2784,
+      "step": 8260
+    },
+    {
+      "epoch": 0.1653,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.000197413396237692,
+      "loss": 1.287,
+      "step": 8265
+    },
+    {
+      "epoch": 0.1654,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019740550241708108,
+      "loss": 1.2687,
+      "step": 8270
+    },
+    {
+      "epoch": 0.1655,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001973975967279052,
+      "loss": 1.2782,
+      "step": 8275
+    },
+    {
+      "epoch": 0.1656,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0001973896791711275,
+      "loss": 1.2872,
+      "step": 8280
+    },
+    {
+      "epoch": 0.1657,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001973817497477129,
+      "loss": 1.3247,
+      "step": 8285
+    },
+    {
+      "epoch": 0.1658,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019737380845862745,
+      "loss": 1.2914,
+      "step": 8290
+    },
+    {
+      "epoch": 0.1659,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001973658553048388,
+      "loss": 1.2731,
+      "step": 8295
+    },
+    {
+      "epoch": 0.166,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019735789028731604,
+      "loss": 1.277,
+      "step": 8300
+    },
+    {
+      "epoch": 0.1661,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019734991340702966,
+      "loss": 1.2664,
+      "step": 8305
+    },
+    {
+      "epoch": 0.1662,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00019734192466495162,
+      "loss": 1.2574,
+      "step": 8310
+    },
+    {
+      "epoch": 0.1663,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001973339240620553,
+      "loss": 1.2735,
+      "step": 8315
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001973259115993156,
+      "loss": 1.2682,
+      "step": 8320
+    },
+    {
+      "epoch": 0.1665,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00019731788727770885,
+      "loss": 1.2775,
+      "step": 8325
+    },
+    {
+      "epoch": 0.1666,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019730985109821266,
+      "loss": 1.2775,
+      "step": 8330
+    },
+    {
+      "epoch": 0.1667,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001973018030618063,
+      "loss": 1.2882,
+      "step": 8335
+    },
+    {
+      "epoch": 0.1668,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0001972937431694704,
+      "loss": 1.2938,
+      "step": 8340
+    },
+    {
+      "epoch": 0.1669,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019728567142218703,
+      "loss": 1.269,
+      "step": 8345
+    },
+    {
+      "epoch": 0.167,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00019727758782093967,
+      "loss": 1.3138,
+      "step": 8350
+    },
+    {
+      "epoch": 0.1671,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00019726949236671332,
+      "loss": 1.3034,
+      "step": 8355
+    },
+    {
+      "epoch": 0.1672,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00019726138506049438,
+      "loss": 1.2799,
+      "step": 8360
+    },
+    {
+      "epoch": 0.1673,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00019725326590327066,
+      "loss": 1.2843,
+      "step": 8365
+    },
+    {
+      "epoch": 0.1674,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019724513489603155,
+      "loss": 1.2772,
+      "step": 8370
+    },
+    {
+      "epoch": 0.1675,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019723699203976766,
+      "loss": 1.2744,
+      "step": 8375
+    },
+    {
+      "epoch": 0.1676,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019722883733547128,
+      "loss": 1.2873,
+      "step": 8380
+    },
+    {
+      "epoch": 0.1677,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00019722067078413599,
+      "loss": 1.2744,
+      "step": 8385
+    },
+    {
+      "epoch": 0.1678,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019721249238675688,
+      "loss": 1.2683,
+      "step": 8390
+    },
+    {
+      "epoch": 0.1679,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019720430214433042,
+      "loss": 1.2619,
+      "step": 8395
+    },
+    {
+      "epoch": 0.168,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00019719610005785465,
+      "loss": 1.2663,
+      "step": 8400
+    },
+    {
+      "epoch": 0.1681,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00019718788612832887,
+      "loss": 1.2697,
+      "step": 8405
+    },
+    {
+      "epoch": 0.1682,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00019717966035675397,
+      "loss": 1.2786,
+      "step": 8410
+    },
+    {
+      "epoch": 0.1683,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00019717142274413223,
+      "loss": 1.2782,
+      "step": 8415
+    },
+    {
+      "epoch": 0.1684,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001971631732914674,
+      "loss": 1.2941,
+      "step": 8420
+    },
+    {
+      "epoch": 0.1685,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001971549119997646,
+      "loss": 1.2813,
+      "step": 8425
+    },
+    {
+      "epoch": 0.1686,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00019714663887003054,
+      "loss": 1.2894,
+      "step": 8430
+    },
+    {
+      "epoch": 0.1687,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019713835390327316,
+      "loss": 1.2786,
+      "step": 8435
+    },
+    {
+      "epoch": 0.1688,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.000197130057100502,
+      "loss": 1.262,
+      "step": 8440
+    },
+    {
+      "epoch": 0.1689,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019712174846272805,
+      "loss": 1.2934,
+      "step": 8445
+    },
+    {
+      "epoch": 0.169,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00019711342799096361,
+      "loss": 1.2778,
+      "step": 8450
+    },
+    {
+      "epoch": 0.1691,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019710509568622258,
+      "loss": 1.2767,
+      "step": 8455
+    },
+    {
+      "epoch": 0.1692,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019709675154952017,
+      "loss": 1.2771,
+      "step": 8460
+    },
+    {
+      "epoch": 0.1693,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.0001970883955818731,
+      "loss": 1.2739,
+      "step": 8465
+    },
+    {
+      "epoch": 0.1694,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00019708002778429955,
+      "loss": 1.3082,
+      "step": 8470
+    },
+    {
+      "epoch": 0.1695,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00019707164815781908,
+      "loss": 1.2704,
+      "step": 8475
+    },
+    {
+      "epoch": 0.1696,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019706325670345275,
+      "loss": 1.2711,
+      "step": 8480
+    },
+    {
+      "epoch": 0.1697,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.000197054853422223,
+      "loss": 1.2717,
+      "step": 8485
+    },
+    {
+      "epoch": 0.1698,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019704643831515374,
+      "loss": 1.2763,
+      "step": 8490
+    },
+    {
+      "epoch": 0.1699,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019703801138327038,
+      "loss": 1.2815,
+      "step": 8495
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019702957262759965,
+      "loss": 1.2579,
+      "step": 8500
+    },
+    {
+      "epoch": 0.1701,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00019702112204916984,
+      "loss": 1.2664,
+      "step": 8505
+    },
+    {
+      "epoch": 0.1702,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001970126596490106,
+      "loss": 1.2675,
+      "step": 8510
+    },
+    {
+      "epoch": 0.1703,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019700418542815306,
+      "loss": 1.2646,
+      "step": 8515
+    },
+    {
+      "epoch": 0.1704,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00019699569938762973,
+      "loss": 1.2767,
+      "step": 8520
+    },
+    {
+      "epoch": 0.1705,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019698720152847468,
+      "loss": 1.2785,
+      "step": 8525
+    },
+    {
+      "epoch": 0.1706,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00019697869185172331,
+      "loss": 1.2787,
+      "step": 8530
+    },
+    {
+      "epoch": 0.1707,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00019697017035841252,
+      "loss": 1.2698,
+      "step": 8535
+    },
+    {
+      "epoch": 0.1708,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0001969616370495806,
+      "loss": 1.2644,
+      "step": 8540
+    },
+    {
+      "epoch": 0.1709,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019695309192626734,
+      "loss": 1.2751,
+      "step": 8545
+    },
+    {
+      "epoch": 0.171,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0001969445349895139,
+      "loss": 1.267,
+      "step": 8550
+    },
+    {
+      "epoch": 0.1711,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00019693596624036292,
+      "loss": 1.2785,
+      "step": 8555
+    },
+    {
+      "epoch": 0.1712,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00019692738567985853,
+      "loss": 1.2654,
+      "step": 8560
+    },
+    {
+      "epoch": 0.1713,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0001969187933090462,
+      "loss": 1.2687,
+      "step": 8565
+    },
+    {
+      "epoch": 0.1714,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00019691018912897286,
+      "loss": 1.2832,
+      "step": 8570
+    },
+    {
+      "epoch": 0.1715,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00019690157314068696,
+      "loss": 1.2771,
+      "step": 8575
+    },
+    {
+      "epoch": 0.1716,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001968929453452383,
+      "loss": 1.2638,
+      "step": 8580
+    },
+    {
+      "epoch": 0.1717,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00019688430574367819,
+      "loss": 1.2714,
+      "step": 8585
+    },
+    {
+      "epoch": 0.1718,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019687565433705926,
+      "loss": 1.259,
+      "step": 8590
+    },
+    {
+      "epoch": 0.1719,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019686699112643572,
+      "loss": 1.2709,
+      "step": 8595
+    },
+    {
+      "epoch": 0.172,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001968583161128631,
+      "loss": 1.2596,
+      "step": 8600
+    },
+    {
+      "epoch": 0.1721,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019684962929739853,
+      "loss": 1.2843,
+      "step": 8605
+    },
+    {
+      "epoch": 0.1722,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019684093068110038,
+      "loss": 1.2817,
+      "step": 8610
+    },
+    {
+      "epoch": 0.1723,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019683222026502858,
+      "loss": 1.2611,
+      "step": 8615
+    },
+    {
+      "epoch": 0.1724,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019682349805024446,
+      "loss": 1.2709,
+      "step": 8620
+    },
+    {
+      "epoch": 0.1725,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001968147640378108,
+      "loss": 1.2774,
+      "step": 8625
+    },
+    {
+      "epoch": 0.1726,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00019680601822879182,
+      "loss": 1.273,
+      "step": 8630
+    },
+    {
+      "epoch": 0.1727,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00019679726062425316,
+      "loss": 1.2762,
+      "step": 8635
+    },
+    {
+      "epoch": 0.1728,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019678849122526187,
+      "loss": 1.2656,
+      "step": 8640
+    },
+    {
+      "epoch": 0.1729,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019677971003288655,
+      "loss": 1.2944,
+      "step": 8645
+    },
+    {
+      "epoch": 0.173,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00019677091704819715,
+      "loss": 1.283,
+      "step": 8650
+    },
+    {
+      "epoch": 0.1731,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.000196762112272265,
+      "loss": 1.2715,
+      "step": 8655
+    },
+    {
+      "epoch": 0.1732,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00019675329570616298,
+      "loss": 1.2668,
+      "step": 8660
+    },
+    {
+      "epoch": 0.1733,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001967444673509654,
+      "loss": 1.2685,
+      "step": 8665
+    },
+    {
+      "epoch": 0.1734,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00019673562720774792,
+      "loss": 1.2682,
+      "step": 8670
+    },
+    {
+      "epoch": 0.1735,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001967267752775877,
+      "loss": 1.2804,
+      "step": 8675
+    },
+    {
+      "epoch": 0.1736,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001967179115615633,
+      "loss": 1.2795,
+      "step": 8680
+    },
+    {
+      "epoch": 0.1737,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019670903606075474,
+      "loss": 1.2879,
+      "step": 8685
+    },
+    {
+      "epoch": 0.1738,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019670014877624353,
+      "loss": 1.2885,
+      "step": 8690
+    },
+    {
+      "epoch": 0.1739,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019669124970911247,
+      "loss": 1.2709,
+      "step": 8695
+    },
+    {
+      "epoch": 0.174,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00019668233886044597,
+      "loss": 1.27,
+      "step": 8700
+    },
+    {
+      "epoch": 0.1741,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001966734162313297,
+      "loss": 1.2667,
+      "step": 8705
+    },
+    {
+      "epoch": 0.1742,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00019666448182285094,
+      "loss": 1.2627,
+      "step": 8710
+    },
+    {
+      "epoch": 0.1743,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019665553563609825,
+      "loss": 1.2781,
+      "step": 8715
+    },
+    {
+      "epoch": 0.1744,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019664657767216176,
+      "loss": 1.2872,
+      "step": 8720
+    },
+    {
+      "epoch": 0.1745,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019663760793213296,
+      "loss": 1.2726,
+      "step": 8725
+    },
+    {
+      "epoch": 0.1746,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001966286264171047,
+      "loss": 1.2645,
+      "step": 8730
+    },
+    {
+      "epoch": 0.1747,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00019661963312817148,
+      "loss": 1.2775,
+      "step": 8735
+    },
+    {
+      "epoch": 0.1748,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019661062806642903,
+      "loss": 1.2669,
+      "step": 8740
+    },
+    {
+      "epoch": 0.1749,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019660161123297458,
+      "loss": 1.2767,
+      "step": 8745
+    },
+    {
+      "epoch": 0.175,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019659258262890683,
+      "loss": 1.2549,
+      "step": 8750
+    },
+    {
+      "epoch": 0.1751,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019658354225532589,
+      "loss": 1.2711,
+      "step": 8755
+    },
+    {
+      "epoch": 0.1752,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00019657449011333328,
+      "loss": 1.268,
+      "step": 8760
+    },
+    {
+      "epoch": 0.1753,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00019656542620403203,
+      "loss": 1.293,
+      "step": 8765
+    },
+    {
+      "epoch": 0.1754,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019655635052852647,
+      "loss": 1.2634,
+      "step": 8770
+    },
+    {
+      "epoch": 0.1755,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.0001965472630879225,
+      "loss": 1.2787,
+      "step": 8775
+    },
+    {
+      "epoch": 0.1756,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001965381638833274,
+      "loss": 1.2719,
+      "step": 8780
+    },
+    {
+      "epoch": 0.1757,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00019652905291584984,
+      "loss": 1.272,
+      "step": 8785
+    },
+    {
+      "epoch": 0.1758,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.0001965199301866,
+      "loss": 1.2757,
+      "step": 8790
+    },
+    {
+      "epoch": 0.1759,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00019651079569668945,
+      "loss": 1.2605,
+      "step": 8795
+    },
+    {
+      "epoch": 0.176,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019650164944723115,
+      "loss": 1.2806,
+      "step": 8800
+    },
+    {
+      "epoch": 0.1761,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00019649249143933962,
+      "loss": 1.292,
+      "step": 8805
+    },
+    {
+      "epoch": 0.1762,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00019648332167413067,
+      "loss": 1.2545,
+      "step": 8810
+    },
+    {
+      "epoch": 0.1763,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001964741401527217,
+      "loss": 1.2629,
+      "step": 8815
+    },
+    {
+      "epoch": 0.1764,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019646494687623135,
+      "loss": 1.2846,
+      "step": 8820
+    },
+    {
+      "epoch": 0.1765,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019645574184577982,
+      "loss": 1.2665,
+      "step": 8825
+    },
+    {
+      "epoch": 0.1766,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019644652506248874,
+      "loss": 1.2865,
+      "step": 8830
+    },
+    {
+      "epoch": 0.1767,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00019643729652748113,
+      "loss": 1.2883,
+      "step": 8835
+    },
+    {
+      "epoch": 0.1768,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00019642805624188147,
+      "loss": 1.2807,
+      "step": 8840
+    },
+    {
+      "epoch": 0.1769,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019641880420681566,
+      "loss": 1.2834,
+      "step": 8845
+    },
+    {
+      "epoch": 0.177,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019640954042341103,
+      "loss": 1.2881,
+      "step": 8850
+    },
+    {
+      "epoch": 0.1771,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0001964002648927963,
+      "loss": 1.279,
+      "step": 8855
+    },
+    {
+      "epoch": 0.1772,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019639097761610174,
+      "loss": 1.2706,
+      "step": 8860
+    },
+    {
+      "epoch": 0.1773,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019638167859445895,
+      "loss": 1.2711,
+      "step": 8865
+    },
+    {
+      "epoch": 0.1774,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.000196372367829001,
+      "loss": 1.2822,
+      "step": 8870
+    },
+    {
+      "epoch": 0.1775,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001963630453208623,
+      "loss": 1.2952,
+      "step": 8875
+    },
+    {
+      "epoch": 0.1776,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019635371107117888,
+      "loss": 1.2846,
+      "step": 8880
+    },
+    {
+      "epoch": 0.1777,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.000196344365081088,
+      "loss": 1.2707,
+      "step": 8885
+    },
+    {
+      "epoch": 0.1778,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001963350073517285,
+      "loss": 1.2641,
+      "step": 8890
+    },
+    {
+      "epoch": 0.1779,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019632563788424053,
+      "loss": 1.2679,
+      "step": 8895
+    },
+    {
+      "epoch": 0.178,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00019631625667976583,
+      "loss": 1.2771,
+      "step": 8900
+    },
+    {
+      "epoch": 0.1781,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00019630686373944738,
+      "loss": 1.2729,
+      "step": 8905
+    },
+    {
+      "epoch": 0.1782,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001962974590644297,
+      "loss": 1.2782,
+      "step": 8910
+    },
+    {
+      "epoch": 0.1783,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00019628804265585877,
+      "loss": 1.2721,
+      "step": 8915
+    },
+    {
+      "epoch": 0.1784,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00019627861451488189,
+      "loss": 1.2646,
+      "step": 8920
+    },
+    {
+      "epoch": 0.1785,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001962691746426479,
+      "loss": 1.2857,
+      "step": 8925
+    },
+    {
+      "epoch": 0.1786,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019625972304030697,
+      "loss": 1.2803,
+      "step": 8930
+    },
+    {
+      "epoch": 0.1787,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019625025970901078,
+      "loss": 1.2864,
+      "step": 8935
+    },
+    {
+      "epoch": 0.1788,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001962407846499124,
+      "loss": 1.2604,
+      "step": 8940
+    },
+    {
+      "epoch": 0.1789,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00019623129786416635,
+      "loss": 1.2799,
+      "step": 8945
+    },
+    {
+      "epoch": 0.179,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00019622179935292855,
+      "loss": 1.2696,
+      "step": 8950
+    },
+    {
+      "epoch": 0.1791,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00019621228911735636,
+      "loss": 1.2532,
+      "step": 8955
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0001962027671586086,
+      "loss": 1.2987,
+      "step": 8960
+    },
+    {
+      "epoch": 0.1793,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00019619323347784548,
+      "loss": 1.2921,
+      "step": 8965
+    },
+    {
+      "epoch": 0.1794,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00019618368807622862,
+      "loss": 1.2858,
+      "step": 8970
+    },
+    {
+      "epoch": 0.1795,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019617413095492114,
+      "loss": 1.2762,
+      "step": 8975
+    },
+    {
+      "epoch": 0.1796,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00019616456211508752,
+      "loss": 1.2641,
+      "step": 8980
+    },
+    {
+      "epoch": 0.1797,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001961549815578937,
+      "loss": 1.267,
+      "step": 8985
+    },
+    {
+      "epoch": 0.1798,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0001961453892845071,
+      "loss": 1.2831,
+      "step": 8990
+    },
+    {
+      "epoch": 0.1799,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0001961357852960964,
+      "loss": 1.2677,
+      "step": 8995
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001961261695938319,
+      "loss": 1.2918,
+      "step": 9000
+    },
+    {
+      "epoch": 0.1801,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001961165421788852,
+      "loss": 1.2717,
+      "step": 9005
+    },
+    {
+      "epoch": 0.1802,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0001961069030524294,
+      "loss": 1.2649,
+      "step": 9010
+    },
+    {
+      "epoch": 0.1803,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019609725221563897,
+      "loss": 1.2866,
+      "step": 9015
+    },
+    {
+      "epoch": 0.1804,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00019608758966968988,
+      "loss": 1.266,
+      "step": 9020
+    },
+    {
+      "epoch": 0.1805,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019607791541575943,
+      "loss": 1.2772,
+      "step": 9025
+    },
+    {
+      "epoch": 0.1806,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0001960682294550264,
+      "loss": 1.2694,
+      "step": 9030
+    },
+    {
+      "epoch": 0.1807,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019605853178867105,
+      "loss": 1.2645,
+      "step": 9035
+    },
+    {
+      "epoch": 0.1808,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019604882241787498,
+      "loss": 1.2783,
+      "step": 9040
+    },
+    {
+      "epoch": 0.1809,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019603910134382123,
+      "loss": 1.2759,
+      "step": 9045
+    },
+    {
+      "epoch": 0.181,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0001960293685676943,
+      "loss": 1.2799,
+      "step": 9050
+    },
+    {
+      "epoch": 0.1811,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019601962409068012,
+      "loss": 1.2591,
+      "step": 9055
+    },
+    {
+      "epoch": 0.1812,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.000196009867913966,
+      "loss": 1.2535,
+      "step": 9060
+    },
+    {
+      "epoch": 0.1813,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019600010003874069,
+      "loss": 1.268,
+      "step": 9065
+    },
+    {
+      "epoch": 0.1814,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019599032046619438,
+      "loss": 1.2716,
+      "step": 9070
+    },
+    {
+      "epoch": 0.1815,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001959805291975187,
+      "loss": 1.2732,
+      "step": 9075
+    },
+    {
+      "epoch": 0.1816,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019597072623390668,
+      "loss": 1.2601,
+      "step": 9080
+    },
+    {
+      "epoch": 0.1817,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00019596091157655278,
+      "loss": 1.2731,
+      "step": 9085
+    },
+    {
+      "epoch": 0.1818,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001959510852266529,
+      "loss": 1.2626,
+      "step": 9090
+    },
+    {
+      "epoch": 0.1819,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001959412471854043,
+      "loss": 1.2693,
+      "step": 9095
+    },
+    {
+      "epoch": 0.182,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00019593139745400576,
+      "loss": 1.2754,
+      "step": 9100
+    },
+    {
+      "epoch": 0.1821,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019592153603365743,
+      "loss": 1.2606,
+      "step": 9105
+    },
+    {
+      "epoch": 0.1822,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001959116629255609,
+      "loss": 1.2576,
+      "step": 9110
+    },
+    {
+      "epoch": 0.1823,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001959017781309192,
+      "loss": 1.2712,
+      "step": 9115
+    },
+    {
+      "epoch": 0.1824,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.0001958918816509367,
+      "loss": 1.2769,
+      "step": 9120
+    },
+    {
+      "epoch": 0.1825,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0001958819734868193,
+      "loss": 1.2862,
+      "step": 9125
+    },
+    {
+      "epoch": 0.1826,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00019587205363977427,
+      "loss": 1.2763,
+      "step": 9130
+    },
+    {
+      "epoch": 0.1827,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019586212211101037,
+      "loss": 1.2796,
+      "step": 9135
+    },
+    {
+      "epoch": 0.1828,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0001958521789017376,
+      "loss": 1.2971,
+      "step": 9140
+    },
+    {
+      "epoch": 0.1829,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0001958422240131676,
+      "loss": 1.2768,
+      "step": 9145
+    },
+    {
+      "epoch": 0.183,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019583225744651333,
+      "loss": 1.2867,
+      "step": 9150
+    },
+    {
+      "epoch": 0.1831,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019582227920298916,
+      "loss": 1.2874,
+      "step": 9155
+    },
+    {
+      "epoch": 0.1832,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019581228928381098,
+      "loss": 1.2784,
+      "step": 9160
+    },
+    {
+      "epoch": 0.1833,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019580228769019593,
+      "loss": 1.2639,
+      "step": 9165
+    },
+    {
+      "epoch": 0.1834,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019579227442336278,
+      "loss": 1.2699,
+      "step": 9170
+    },
+    {
+      "epoch": 0.1835,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001957822494845315,
+      "loss": 1.2679,
+      "step": 9175
+    },
+    {
+      "epoch": 0.1836,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00019577221287492367,
+      "loss": 1.3061,
+      "step": 9180
+    },
+    {
+      "epoch": 0.1837,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019576216459576222,
+      "loss": 1.2811,
+      "step": 9185
+    },
+    {
+      "epoch": 0.1838,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00019575210464827149,
+      "loss": 1.2792,
+      "step": 9190
+    },
+    {
+      "epoch": 0.1839,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00019574203303367727,
+      "loss": 1.2616,
+      "step": 9195
+    },
+    {
+      "epoch": 0.184,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019573194975320673,
+      "loss": 1.2762,
+      "step": 9200
+    },
+    {
+      "epoch": 0.1841,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019572185480808848,
+      "loss": 1.2691,
+      "step": 9205
+    },
+    {
+      "epoch": 0.1842,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00019571174819955263,
+      "loss": 1.2603,
+      "step": 9210
+    },
+    {
+      "epoch": 0.1843,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019570162992883054,
+      "loss": 1.2749,
+      "step": 9215
+    },
+    {
+      "epoch": 0.1844,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019569149999715515,
+      "loss": 1.2718,
+      "step": 9220
+    },
+    {
+      "epoch": 0.1845,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00019568135840576076,
+      "loss": 1.2775,
+      "step": 9225
+    },
+    {
+      "epoch": 0.1846,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00019567120515588308,
+      "loss": 1.268,
+      "step": 9230
+    },
+    {
+      "epoch": 0.1847,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001956610402487592,
+      "loss": 1.268,
+      "step": 9235
+    },
+    {
+      "epoch": 0.1848,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001956508636856278,
+      "loss": 1.2825,
+      "step": 9240
+    },
+    {
+      "epoch": 0.1849,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00019564067546772878,
+      "loss": 1.2777,
+      "step": 9245
+    },
+    {
+      "epoch": 0.185,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019563047559630357,
+      "loss": 1.283,
+      "step": 9250
+    },
+    {
+      "epoch": 0.1851,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00019562026407259495,
+      "loss": 1.2584,
+      "step": 9255
+    },
+    {
+      "epoch": 0.1852,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00019561004089784723,
+      "loss": 1.2589,
+      "step": 9260
+    },
+    {
+      "epoch": 0.1853,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019559980607330605,
+      "loss": 1.2584,
+      "step": 9265
+    },
+    {
+      "epoch": 0.1854,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00019558955960021849,
+      "loss": 1.2894,
+      "step": 9270
+    },
+    {
+      "epoch": 0.1855,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00019557930147983302,
+      "loss": 1.2681,
+      "step": 9275
+    },
+    {
+      "epoch": 0.1856,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00019556903171339963,
+      "loss": 1.2716,
+      "step": 9280
+    },
+    {
+      "epoch": 0.1857,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00019555875030216954,
+      "loss": 1.293,
+      "step": 9285
+    },
+    {
+      "epoch": 0.1858,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00019554845724739566,
+      "loss": 1.2827,
+      "step": 9290
+    },
+    {
+      "epoch": 0.1859,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019553815255033205,
+      "loss": 1.2744,
+      "step": 9295
+    },
+    {
+      "epoch": 0.186,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019552783621223436,
+      "loss": 1.2763,
+      "step": 9300
+    },
+    {
+      "epoch": 0.1861,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001955175082343596,
+      "loss": 1.2676,
+      "step": 9305
+    },
+    {
+      "epoch": 0.1862,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00019550716861796623,
+      "loss": 1.2687,
+      "step": 9310
+    },
+    {
+      "epoch": 0.1863,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00019549681736431404,
+      "loss": 1.2696,
+      "step": 9315
+    },
+    {
+      "epoch": 0.1864,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019548645447466431,
+      "loss": 1.2609,
+      "step": 9320
+    },
+    {
+      "epoch": 0.1865,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00019547607995027978,
+      "loss": 1.2884,
+      "step": 9325
+    },
+    {
+      "epoch": 0.1866,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00019546569379242444,
+      "loss": 1.2763,
+      "step": 9330
+    },
+    {
+      "epoch": 0.1867,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00019545529600236398,
+      "loss": 1.2765,
+      "step": 9335
+    },
+    {
+      "epoch": 0.1868,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019544488658136523,
+      "loss": 1.2691,
+      "step": 9340
+    },
+    {
+      "epoch": 0.1869,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001954344655306965,
+      "loss": 1.2689,
+      "step": 9345
+    },
+    {
+      "epoch": 0.187,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001954240328516277,
+      "loss": 1.2701,
+      "step": 9350
+    },
+    {
+      "epoch": 0.1871,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019541358854542991,
+      "loss": 1.2708,
+      "step": 9355
+    },
+    {
+      "epoch": 0.1872,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001954031326133758,
+      "loss": 1.2694,
+      "step": 9360
+    },
+    {
+      "epoch": 0.1873,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019539266505673938,
+      "loss": 1.2736,
+      "step": 9365
+    },
+    {
+      "epoch": 0.1874,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019538218587679605,
+      "loss": 1.2805,
+      "step": 9370
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001953716950748227,
+      "loss": 1.2719,
+      "step": 9375
+    },
+    {
+      "epoch": 0.1876,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001953611926520976,
+      "loss": 1.2628,
+      "step": 9380
+    },
+    {
+      "epoch": 0.1877,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00019535067860990046,
+      "loss": 1.2777,
+      "step": 9385
+    },
+    {
+      "epoch": 0.1878,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00019534015294951233,
+      "loss": 1.2935,
+      "step": 9390
+    },
+    {
+      "epoch": 0.1879,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019532961567221576,
+      "loss": 1.2916,
+      "step": 9395
+    },
+    {
+      "epoch": 0.188,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001953190667792947,
+      "loss": 1.2875,
+      "step": 9400
+    },
+    {
+      "epoch": 0.1881,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001953085062720345,
+      "loss": 1.2741,
+      "step": 9405
+    },
+    {
+      "epoch": 0.1882,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00019529793415172192,
+      "loss": 1.2778,
+      "step": 9410
+    },
+    {
+      "epoch": 0.1883,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019528735041964509,
+      "loss": 1.2695,
+      "step": 9415
+    },
+    {
+      "epoch": 0.1884,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00019527675507709366,
+      "loss": 1.2814,
+      "step": 9420
+    },
+    {
+      "epoch": 0.1885,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00019526614812535864,
+      "loss": 1.2711,
+      "step": 9425
+    },
+    {
+      "epoch": 0.1886,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019525552956573244,
+      "loss": 1.2991,
+      "step": 9430
+    },
+    {
+      "epoch": 0.1887,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0001952448993995089,
+      "loss": 1.2802,
+      "step": 9435
+    },
+    {
+      "epoch": 0.1888,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019523425762798329,
+      "loss": 1.2766,
+      "step": 9440
+    },
+    {
+      "epoch": 0.1889,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00019522360425245226,
+      "loss": 1.2831,
+      "step": 9445
+    },
+    {
+      "epoch": 0.189,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00019521293927421388,
+      "loss": 1.2617,
+      "step": 9450
+    },
+    {
+      "epoch": 0.1891,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019520226269456768,
+      "loss": 1.2675,
+      "step": 9455
+    },
+    {
+      "epoch": 0.1892,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019519157451481454,
+      "loss": 1.3054,
+      "step": 9460
+    },
+    {
+      "epoch": 0.1893,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001951808747362568,
+      "loss": 1.2714,
+      "step": 9465
+    },
+    {
+      "epoch": 0.1894,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001951701633601982,
+      "loss": 1.2725,
+      "step": 9470
+    },
+    {
+      "epoch": 0.1895,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019515944038794384,
+      "loss": 1.2733,
+      "step": 9475
+    },
+    {
+      "epoch": 0.1896,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019514870582080032,
+      "loss": 1.2653,
+      "step": 9480
+    },
+    {
+      "epoch": 0.1897,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00019513795966007562,
+      "loss": 1.2782,
+      "step": 9485
+    },
+    {
+      "epoch": 0.1898,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019512720190707913,
+      "loss": 1.2767,
+      "step": 9490
+    },
+    {
+      "epoch": 0.1899,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019511643256312164,
+      "loss": 1.2573,
+      "step": 9495
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00019510565162951537,
+      "loss": 1.287,
+      "step": 9500
+    },
+    {
+      "epoch": 0.1901,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001950948591075739,
+      "loss": 1.2604,
+      "step": 9505
+    },
+    {
+      "epoch": 0.1902,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019508405499861232,
+      "loss": 1.2779,
+      "step": 9510
+    },
+    {
+      "epoch": 0.1903,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019507323930394708,
+      "loss": 1.2703,
+      "step": 9515
+    },
+    {
+      "epoch": 0.1904,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00019506241202489602,
+      "loss": 1.2791,
+      "step": 9520
+    },
+    {
+      "epoch": 0.1905,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00019505157316277837,
+      "loss": 1.2736,
+      "step": 9525
+    },
+    {
+      "epoch": 0.1906,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019504072271891488,
+      "loss": 1.2768,
+      "step": 9530
+    },
+    {
+      "epoch": 0.1907,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00019502986069462762,
+      "loss": 1.2717,
+      "step": 9535
+    },
+    {
+      "epoch": 0.1908,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019501898709124008,
+      "loss": 1.2992,
+      "step": 9540
+    },
+    {
+      "epoch": 0.1909,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00019500810191007718,
+      "loss": 1.2618,
+      "step": 9545
+    },
+    {
+      "epoch": 0.191,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00019499720515246525,
+      "loss": 1.2825,
+      "step": 9550
+    },
+    {
+      "epoch": 0.1911,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019498629681973206,
+      "loss": 1.2791,
+      "step": 9555
+    },
+    {
+      "epoch": 0.1912,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00019497537691320668,
+      "loss": 1.2764,
+      "step": 9560
+    },
+    {
+      "epoch": 0.1913,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00019496444543421975,
+      "loss": 1.2851,
+      "step": 9565
+    },
+    {
+      "epoch": 0.1914,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001949535023841032,
+      "loss": 1.2727,
+      "step": 9570
+    },
+    {
+      "epoch": 0.1915,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0001949425477641904,
+      "loss": 1.2818,
+      "step": 9575
+    },
+    {
+      "epoch": 0.1916,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019493158157581615,
+      "loss": 1.2823,
+      "step": 9580
+    },
+    {
+      "epoch": 0.1917,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00019492060382031662,
+      "loss": 1.2709,
+      "step": 9585
+    },
+    {
+      "epoch": 0.1918,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00019490961449902946,
+      "loss": 1.2669,
+      "step": 9590
+    },
+    {
+      "epoch": 0.1919,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00019489861361329366,
+      "loss": 1.2647,
+      "step": 9595
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00019488760116444966,
+      "loss": 1.2787,
+      "step": 9600
+    },
+    {
+      "epoch": 0.1921,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019487657715383926,
+      "loss": 1.2656,
+      "step": 9605
+    },
+    {
+      "epoch": 0.1922,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019486554158280574,
+      "loss": 1.2753,
+      "step": 9610
+    },
+    {
+      "epoch": 0.1923,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019485449445269377,
+      "loss": 1.2624,
+      "step": 9615
+    },
+    {
+      "epoch": 0.1924,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00019484343576484933,
+      "loss": 1.2676,
+      "step": 9620
+    },
+    {
+      "epoch": 0.1925,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00019483236552061994,
+      "loss": 1.2601,
+      "step": 9625
+    },
+    {
+      "epoch": 0.1926,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019482128372135446,
+      "loss": 1.2731,
+      "step": 9630
+    },
+    {
+      "epoch": 0.1927,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001948101903684032,
+      "loss": 1.2962,
+      "step": 9635
+    },
+    {
+      "epoch": 0.1928,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00019479908546311781,
+      "loss": 1.2615,
+      "step": 9640
+    },
+    {
+      "epoch": 0.1929,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019478796900685146,
+      "loss": 1.2803,
+      "step": 9645
+    },
+    {
+      "epoch": 0.193,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001947768410009586,
+      "loss": 1.2902,
+      "step": 9650
+    },
+    {
+      "epoch": 0.1931,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0001947657014467951,
+      "loss": 1.2751,
+      "step": 9655
+    },
+    {
+      "epoch": 0.1932,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00019475455034571838,
+      "loss": 1.2665,
+      "step": 9660
+    },
+    {
+      "epoch": 0.1933,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0001947433876990871,
+      "loss": 1.3043,
+      "step": 9665
+    },
+    {
+      "epoch": 0.1934,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00019473221350826142,
+      "loss": 1.2749,
+      "step": 9670
+    },
+    {
+      "epoch": 0.1935,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001947210277746029,
+      "loss": 1.2747,
+      "step": 9675
+    },
+    {
+      "epoch": 0.1936,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019470983049947444,
+      "loss": 1.2845,
+      "step": 9680
+    },
+    {
+      "epoch": 0.1937,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00019469862168424042,
+      "loss": 1.2711,
+      "step": 9685
+    },
+    {
+      "epoch": 0.1938,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001946874013302666,
+      "loss": 1.2547,
+      "step": 9690
+    },
+    {
+      "epoch": 0.1939,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001946761694389202,
+      "loss": 1.2814,
+      "step": 9695
+    },
+    {
+      "epoch": 0.194,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019466492601156966,
+      "loss": 1.2698,
+      "step": 9700
+    },
+    {
+      "epoch": 0.1941,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00019465367104958505,
+      "loss": 1.2709,
+      "step": 9705
+    },
+    {
+      "epoch": 0.1942,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00019464240455433775,
+      "loss": 1.262,
+      "step": 9710
+    },
+    {
+      "epoch": 0.1943,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019463112652720054,
+      "loss": 1.2754,
+      "step": 9715
+    },
+    {
+      "epoch": 0.1944,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00019461983696954758,
+      "loss": 1.2804,
+      "step": 9720
+    },
+    {
+      "epoch": 0.1945,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019460853588275454,
+      "loss": 1.2999,
+      "step": 9725
+    },
+    {
+      "epoch": 0.1946,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00019459722326819838,
+      "loss": 1.2747,
+      "step": 9730
+    },
+    {
+      "epoch": 0.1947,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019458589912725748,
+      "loss": 1.2864,
+      "step": 9735
+    },
+    {
+      "epoch": 0.1948,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001945745634613117,
+      "loss": 1.2783,
+      "step": 9740
+    },
+    {
+      "epoch": 0.1949,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00019456321627174221,
+      "loss": 1.2751,
+      "step": 9745
+    },
+    {
+      "epoch": 0.195,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001945518575599317,
+      "loss": 1.2696,
+      "step": 9750
+    },
+    {
+      "epoch": 0.1951,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019454048732726412,
+      "loss": 1.2678,
+      "step": 9755
+    },
+    {
+      "epoch": 0.1952,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00019452910557512496,
+      "loss": 1.2724,
+      "step": 9760
+    },
+    {
+      "epoch": 0.1953,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.000194517712304901,
+      "loss": 1.2662,
+      "step": 9765
+    },
+    {
+      "epoch": 0.1954,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00019450630751798048,
+      "loss": 1.287,
+      "step": 9770
+    },
+    {
+      "epoch": 0.1955,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0001944948912157531,
+      "loss": 1.2712,
+      "step": 9775
+    },
+    {
+      "epoch": 0.1956,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00019448346339960982,
+      "loss": 1.2644,
+      "step": 9780
+    },
+    {
+      "epoch": 0.1957,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00019447202407094316,
+      "loss": 1.2848,
+      "step": 9785
+    },
+    {
+      "epoch": 0.1958,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001944605732311469,
+      "loss": 1.281,
+      "step": 9790
+    },
+    {
+      "epoch": 0.1959,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019444911088161636,
+      "loss": 1.2606,
+      "step": 9795
+    },
+    {
+      "epoch": 0.196,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00019443763702374812,
+      "loss": 1.2567,
+      "step": 9800
+    },
+    {
+      "epoch": 0.1961,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019442615165894027,
+      "loss": 1.2824,
+      "step": 9805
+    },
+    {
+      "epoch": 0.1962,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00019441465478859228,
+      "loss": 1.2762,
+      "step": 9810
+    },
+    {
+      "epoch": 0.1963,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.000194403146414105,
+      "loss": 1.2763,
+      "step": 9815
+    },
+    {
+      "epoch": 0.1964,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019439162653688065,
+      "loss": 1.2789,
+      "step": 9820
+    },
+    {
+      "epoch": 0.1965,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019438009515832297,
+      "loss": 1.2886,
+      "step": 9825
+    },
+    {
+      "epoch": 0.1966,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019436855227983695,
+      "loss": 1.297,
+      "step": 9830
+    },
+    {
+      "epoch": 0.1967,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00019435699790282908,
+      "loss": 1.2572,
+      "step": 9835
+    },
+    {
+      "epoch": 0.1968,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00019434543202870725,
+      "loss": 1.274,
+      "step": 9840
+    },
+    {
+      "epoch": 0.1969,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0001943338546588807,
+      "loss": 1.2746,
+      "step": 9845
+    },
+    {
+      "epoch": 0.197,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001943222657947601,
+      "loss": 1.2876,
+      "step": 9850
+    },
+    {
+      "epoch": 0.1971,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019431066543775752,
+      "loss": 1.2762,
+      "step": 9855
+    },
+    {
+      "epoch": 0.1972,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019429905358928646,
+      "loss": 1.2748,
+      "step": 9860
+    },
+    {
+      "epoch": 0.1973,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019428743025076174,
+      "loss": 1.2842,
+      "step": 9865
+    },
+    {
+      "epoch": 0.1974,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00019427579542359965,
+      "loss": 1.2809,
+      "step": 9870
+    },
+    {
+      "epoch": 0.1975,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00019426414910921787,
+      "loss": 1.2746,
+      "step": 9875
+    },
+    {
+      "epoch": 0.1976,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019425249130903543,
+      "loss": 1.275,
+      "step": 9880
+    },
+    {
+      "epoch": 0.1977,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019424082202447283,
+      "loss": 1.2923,
+      "step": 9885
+    },
+    {
+      "epoch": 0.1978,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001942291412569519,
+      "loss": 1.2727,
+      "step": 9890
+    },
+    {
+      "epoch": 0.1979,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019421744900789597,
+      "loss": 1.2614,
+      "step": 9895
+    },
+    {
+      "epoch": 0.198,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019420574527872968,
+      "loss": 1.2856,
+      "step": 9900
+    },
+    {
+      "epoch": 0.1981,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019419403007087907,
+      "loss": 1.2642,
+      "step": 9905
+    },
+    {
+      "epoch": 0.1982,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001941823033857716,
+      "loss": 1.2721,
+      "step": 9910
+    },
+    {
+      "epoch": 0.1983,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001941705652248362,
+      "loss": 1.2886,
+      "step": 9915
+    },
+    {
+      "epoch": 0.1984,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00019415881558950302,
+      "loss": 1.2745,
+      "step": 9920
+    },
+    {
+      "epoch": 0.1985,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001941470544812038,
+      "loss": 1.2862,
+      "step": 9925
+    },
+    {
+      "epoch": 0.1986,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001941352819013716,
+      "loss": 1.26,
+      "step": 9930
+    },
+    {
+      "epoch": 0.1987,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019412349785144078,
+      "loss": 1.2852,
+      "step": 9935
+    },
+    {
+      "epoch": 0.1988,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00019411170233284727,
+      "loss": 1.2615,
+      "step": 9940
+    },
+    {
+      "epoch": 0.1989,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00019409989534702833,
+      "loss": 1.2895,
+      "step": 9945
+    },
+    {
+      "epoch": 0.199,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019408807689542257,
+      "loss": 1.2772,
+      "step": 9950
+    },
+    {
+      "epoch": 0.1991,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019407624697947003,
+      "loss": 1.2728,
+      "step": 9955
+    },
+    {
+      "epoch": 0.1992,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019406440560061216,
+      "loss": 1.2744,
+      "step": 9960
+    },
+    {
+      "epoch": 0.1993,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001940525527602918,
+      "loss": 1.2702,
+      "step": 9965
+    },
+    {
+      "epoch": 0.1994,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00019404068845995317,
+      "loss": 1.2642,
+      "step": 9970
+    },
+    {
+      "epoch": 0.1995,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0001940288127010419,
+      "loss": 1.2726,
+      "step": 9975
+    },
+    {
+      "epoch": 0.1996,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00019401692548500502,
+      "loss": 1.2883,
+      "step": 9980
+    },
+    {
+      "epoch": 0.1997,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00019400502681329098,
+      "loss": 1.2878,
+      "step": 9985
+    },
+    {
+      "epoch": 0.1998,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019399311668734956,
+      "loss": 1.2789,
+      "step": 9990
+    },
+    {
+      "epoch": 0.1999,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019398119510863197,
+      "loss": 1.2824,
+      "step": 9995
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00019396926207859084,
+      "loss": 1.2769,
+      "step": 10000
+    },
+    {
+      "epoch": 0.2001,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019395731759868018,
+      "loss": 1.2733,
+      "step": 10005
+    },
+    {
+      "epoch": 0.2002,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019394536167035534,
+      "loss": 1.2711,
+      "step": 10010
+    },
+    {
+      "epoch": 0.2003,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00019393339429507318,
+      "loss": 1.2619,
+      "step": 10015
+    },
+    {
+      "epoch": 0.2004,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00019392141547429183,
+      "loss": 1.2863,
+      "step": 10020
+    },
+    {
+      "epoch": 0.2005,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001939094252094709,
+      "loss": 1.2568,
+      "step": 10025
+    },
+    {
+      "epoch": 0.2006,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019389742350207141,
+      "loss": 1.2874,
+      "step": 10030
+    },
+    {
+      "epoch": 0.2007,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019388541035355564,
+      "loss": 1.2687,
+      "step": 10035
+    },
+    {
+      "epoch": 0.2008,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019387338576538744,
+      "loss": 1.2926,
+      "step": 10040
+    },
+    {
+      "epoch": 0.2009,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001938613497390319,
+      "loss": 1.2756,
+      "step": 10045
+    },
+    {
+      "epoch": 0.201,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001938493022759556,
+      "loss": 1.2677,
+      "step": 10050
+    },
+    {
+      "epoch": 0.2011,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001938372433776265,
+      "loss": 1.2673,
+      "step": 10055
+    },
+    {
+      "epoch": 0.2012,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019382517304551396,
+      "loss": 1.2897,
+      "step": 10060
+    },
+    {
+      "epoch": 0.2013,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019381309128108865,
+      "loss": 1.2777,
+      "step": 10065
+    },
+    {
+      "epoch": 0.2014,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00019380099808582278,
+      "loss": 1.2932,
+      "step": 10070
+    },
+    {
+      "epoch": 0.2015,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001937888934611898,
+      "loss": 1.2608,
+      "step": 10075
+    },
+    {
+      "epoch": 0.2016,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001937767774086646,
+      "loss": 1.2614,
+      "step": 10080
+    },
+    {
+      "epoch": 0.2017,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019376464992972356,
+      "loss": 1.288,
+      "step": 10085
+    },
+    {
+      "epoch": 0.2018,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001937525110258444,
+      "loss": 1.26,
+      "step": 10090
+    },
+    {
+      "epoch": 0.2019,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00019374036069850608,
+      "loss": 1.2696,
+      "step": 10095
+    },
+    {
+      "epoch": 0.202,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00019372819894918915,
+      "loss": 1.2853,
+      "step": 10100
+    },
+    {
+      "epoch": 0.2021,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00019371602577937554,
+      "loss": 1.2824,
+      "step": 10105
+    },
+    {
+      "epoch": 0.2022,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001937038411905484,
+      "loss": 1.2631,
+      "step": 10110
+    },
+    {
+      "epoch": 0.2023,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001936916451841925,
+      "loss": 1.2753,
+      "step": 10115
+    },
+    {
+      "epoch": 0.2024,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0001936794377617938,
+      "loss": 1.2721,
+      "step": 10120
+    },
+    {
+      "epoch": 0.2025,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00019366721892483978,
+      "loss": 1.2766,
+      "step": 10125
+    },
+    {
+      "epoch": 0.2026,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019365498867481923,
+      "loss": 1.2782,
+      "step": 10130
+    },
+    {
+      "epoch": 0.2027,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019364274701322244,
+      "loss": 1.2872,
+      "step": 10135
+    },
+    {
+      "epoch": 0.2028,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019363049394154094,
+      "loss": 1.2707,
+      "step": 10140
+    },
+    {
+      "epoch": 0.2029,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001936182294612678,
+      "loss": 1.2782,
+      "step": 10145
+    },
+    {
+      "epoch": 0.203,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00019360595357389735,
+      "loss": 1.2857,
+      "step": 10150
+    },
+    {
+      "epoch": 0.2031,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00019359366628092539,
+      "loss": 1.262,
+      "step": 10155
+    },
+    {
+      "epoch": 0.2032,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00019358136758384912,
+      "loss": 1.258,
+      "step": 10160
+    },
+    {
+      "epoch": 0.2033,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019356905748416702,
+      "loss": 1.2719,
+      "step": 10165
+    },
+    {
+      "epoch": 0.2034,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00019355673598337914,
+      "loss": 1.2774,
+      "step": 10170
+    },
+    {
+      "epoch": 0.2035,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019354440308298675,
+      "loss": 1.2854,
+      "step": 10175
+    },
+    {
+      "epoch": 0.2036,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019353205878449258,
+      "loss": 1.2769,
+      "step": 10180
+    },
+    {
+      "epoch": 0.2037,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001935197030894008,
+      "loss": 1.2695,
+      "step": 10185
+    },
+    {
+      "epoch": 0.2038,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019350733599921683,
+      "loss": 1.259,
+      "step": 10190
+    },
+    {
+      "epoch": 0.2039,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00019349495751544763,
+      "loss": 1.2954,
+      "step": 10195
+    },
+    {
+      "epoch": 0.204,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00019348256763960145,
+      "loss": 1.2792,
+      "step": 10200
+    },
+    {
+      "epoch": 0.2041,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.000193470166373188,
+      "loss": 1.2664,
+      "step": 10205
+    },
+    {
+      "epoch": 0.2042,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019345775371771824,
+      "loss": 1.2659,
+      "step": 10210
+    },
+    {
+      "epoch": 0.2043,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001934453296747047,
+      "loss": 1.2682,
+      "step": 10215
+    },
+    {
+      "epoch": 0.2044,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019343289424566122,
+      "loss": 1.2631,
+      "step": 10220
+    },
+    {
+      "epoch": 0.2045,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019342044743210295,
+      "loss": 1.2869,
+      "step": 10225
+    },
+    {
+      "epoch": 0.2046,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00019340798923554657,
+      "loss": 1.2807,
+      "step": 10230
+    },
+    {
+      "epoch": 0.2047,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00019339551965751002,
+      "loss": 1.2618,
+      "step": 10235
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00019338303869951269,
+      "loss": 1.2881,
+      "step": 10240
+    },
+    {
+      "epoch": 0.2049,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019337054636307536,
+      "loss": 1.2922,
+      "step": 10245
+    },
+    {
+      "epoch": 0.205,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019335804264972018,
+      "loss": 1.2808,
+      "step": 10250
+    },
+    {
+      "epoch": 0.2051,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001933455275609707,
+      "loss": 1.292,
+      "step": 10255
+    },
+    {
+      "epoch": 0.2052,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.0001933330010983518,
+      "loss": 1.2881,
+      "step": 10260
+    },
+    {
+      "epoch": 0.2053,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019332046326338986,
+      "loss": 1.2885,
+      "step": 10265
+    },
+    {
+      "epoch": 0.2054,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019330791405761252,
+      "loss": 1.2944,
+      "step": 10270
+    },
+    {
+      "epoch": 0.2055,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00019329535348254893,
+      "loss": 1.2837,
+      "step": 10275
+    },
+    {
+      "epoch": 0.2056,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019328278153972947,
+      "loss": 1.28,
+      "step": 10280
+    },
+    {
+      "epoch": 0.2057,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019327019823068604,
+      "loss": 1.2884,
+      "step": 10285
+    },
+    {
+      "epoch": 0.2058,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00019325760355695188,
+      "loss": 1.2578,
+      "step": 10290
+    },
+    {
+      "epoch": 0.2059,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001932449975200616,
+      "loss": 1.2849,
+      "step": 10295
+    },
+    {
+      "epoch": 0.206,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019323238012155123,
+      "loss": 1.2689,
+      "step": 10300
+    },
+    {
+      "epoch": 0.2061,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00019321975136295813,
+      "loss": 1.275,
+      "step": 10305
+    },
+    {
+      "epoch": 0.2062,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001932071112458211,
+      "loss": 1.2702,
+      "step": 10310
+    },
+    {
+      "epoch": 0.2063,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001931944597716803,
+      "loss": 1.2841,
+      "step": 10315
+    },
+    {
+      "epoch": 0.2064,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019318179694207725,
+      "loss": 1.2741,
+      "step": 10320
+    },
+    {
+      "epoch": 0.2065,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001931691227585549,
+      "loss": 1.2791,
+      "step": 10325
+    },
+    {
+      "epoch": 0.2066,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00019315643722265757,
+      "loss": 1.2628,
+      "step": 10330
+    },
+    {
+      "epoch": 0.2067,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001931437403359309,
+      "loss": 1.2675,
+      "step": 10335
+    },
+    {
+      "epoch": 0.2068,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00019313103209992204,
+      "loss": 1.274,
+      "step": 10340
+    },
+    {
+      "epoch": 0.2069,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001931183125161794,
+      "loss": 1.2872,
+      "step": 10345
+    },
+    {
+      "epoch": 0.207,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00019310558158625285,
+      "loss": 1.271,
+      "step": 10350
+    },
+    {
+      "epoch": 0.2071,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019309283931169356,
+      "loss": 1.2621,
+      "step": 10355
+    },
+    {
+      "epoch": 0.2072,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00019308008569405422,
+      "loss": 1.2763,
+      "step": 10360
+    },
+    {
+      "epoch": 0.2073,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0001930673207348888,
+      "loss": 1.2744,
+      "step": 10365
+    },
+    {
+      "epoch": 0.2074,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00019305454443575262,
+      "loss": 1.2798,
+      "step": 10370
+    },
+    {
+      "epoch": 0.2075,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019304175679820247,
+      "loss": 1.2768,
+      "step": 10375
+    },
+    {
+      "epoch": 0.2076,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001930289578237965,
+      "loss": 1.2843,
+      "step": 10380
+    },
+    {
+      "epoch": 0.2077,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00019301614751409416,
+      "loss": 1.2948,
+      "step": 10385
+    },
+    {
+      "epoch": 0.2078,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001930033258706564,
+      "loss": 1.2728,
+      "step": 10390
+    },
+    {
+      "epoch": 0.2079,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019299049289504553,
+      "loss": 1.2993,
+      "step": 10395
+    },
+    {
+      "epoch": 0.208,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019297764858882514,
+      "loss": 1.2747,
+      "step": 10400
+    },
+    {
+      "epoch": 0.2081,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00019296479295356035,
+      "loss": 1.2727,
+      "step": 10405
+    },
+    {
+      "epoch": 0.2082,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00019295192599081746,
+      "loss": 1.2589,
+      "step": 10410
+    },
+    {
+      "epoch": 0.2083,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00019293904770216437,
+      "loss": 1.2772,
+      "step": 10415
+    },
+    {
+      "epoch": 0.2084,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00019292615808917026,
+      "loss": 1.2792,
+      "step": 10420
+    },
+    {
+      "epoch": 0.2085,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019291325715340563,
+      "loss": 1.2805,
+      "step": 10425
+    },
+    {
+      "epoch": 0.2086,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019290034489644246,
+      "loss": 1.2649,
+      "step": 10430
+    },
+    {
+      "epoch": 0.2087,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00019288742131985407,
+      "loss": 1.2842,
+      "step": 10435
+    },
+    {
+      "epoch": 0.2088,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019287448642521513,
+      "loss": 1.2807,
+      "step": 10440
+    },
+    {
+      "epoch": 0.2089,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019286154021410173,
+      "loss": 1.2712,
+      "step": 10445
+    },
+    {
+      "epoch": 0.209,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019284858268809137,
+      "loss": 1.2739,
+      "step": 10450
+    },
+    {
+      "epoch": 0.2091,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019283561384876284,
+      "loss": 1.2617,
+      "step": 10455
+    },
+    {
+      "epoch": 0.2092,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019282263369769633,
+      "loss": 1.2862,
+      "step": 10460
+    },
+    {
+      "epoch": 0.2093,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019280964223647348,
+      "loss": 1.2774,
+      "step": 10465
+    },
+    {
+      "epoch": 0.2094,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019279663946667727,
+      "loss": 1.2674,
+      "step": 10470
+    },
+    {
+      "epoch": 0.2095,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.000192783625389892,
+      "loss": 1.2808,
+      "step": 10475
+    },
+    {
+      "epoch": 0.2096,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019277060000770342,
+      "loss": 1.2685,
+      "step": 10480
+    },
+    {
+      "epoch": 0.2097,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019275756332169867,
+      "loss": 1.2621,
+      "step": 10485
+    },
+    {
+      "epoch": 0.2098,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00019274451533346615,
+      "loss": 1.2764,
+      "step": 10490
+    },
+    {
+      "epoch": 0.2099,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00019273145604459577,
+      "loss": 1.2793,
+      "step": 10495
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00019271838545667876,
+      "loss": 1.2681,
+      "step": 10500
+    },
+    {
+      "epoch": 0.2101,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0001927053035713077,
+      "loss": 1.2671,
+      "step": 10505
+    },
+    {
+      "epoch": 0.2102,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00019269221039007665,
+      "loss": 1.2775,
+      "step": 10510
+    },
+    {
+      "epoch": 0.2103,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001926791059145809,
+      "loss": 1.2802,
+      "step": 10515
+    },
+    {
+      "epoch": 0.2104,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001926659901464172,
+      "loss": 1.2807,
+      "step": 10520
+    },
+    {
+      "epoch": 0.2105,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019265286308718372,
+      "loss": 1.2637,
+      "step": 10525
+    },
+    {
+      "epoch": 0.2106,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00019263972473847993,
+      "loss": 1.2828,
+      "step": 10530
+    },
+    {
+      "epoch": 0.2107,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019262657510190666,
+      "loss": 1.2527,
+      "step": 10535
+    },
+    {
+      "epoch": 0.2108,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00019261341417906621,
+      "loss": 1.2813,
+      "step": 10540
+    },
+    {
+      "epoch": 0.2109,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00019260024197156214,
+      "loss": 1.2727,
+      "step": 10545
+    },
+    {
+      "epoch": 0.211,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001925870584809995,
+      "loss": 1.2738,
+      "step": 10550
+    },
+    {
+      "epoch": 0.2111,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019257386370898457,
+      "loss": 1.2717,
+      "step": 10555
+    },
+    {
+      "epoch": 0.2112,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00019256065765712522,
+      "loss": 1.2794,
+      "step": 10560
+    },
+    {
+      "epoch": 0.2113,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00019254744032703048,
+      "loss": 1.2769,
+      "step": 10565
+    },
+    {
+      "epoch": 0.2114,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00019253421172031086,
+      "loss": 1.2665,
+      "step": 10570
+    },
+    {
+      "epoch": 0.2115,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019252097183857823,
+      "loss": 1.266,
+      "step": 10575
+    },
+    {
+      "epoch": 0.2116,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0001925077206834458,
+      "loss": 1.2677,
+      "step": 10580
+    },
+    {
+      "epoch": 0.2117,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019249445825652824,
+      "loss": 1.2772,
+      "step": 10585
+    },
+    {
+      "epoch": 0.2118,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001924811845594415,
+      "loss": 1.2825,
+      "step": 10590
+    },
+    {
+      "epoch": 0.2119,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00019246789959380295,
+      "loss": 1.2654,
+      "step": 10595
+    },
+    {
+      "epoch": 0.212,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019245460336123134,
+      "loss": 1.2576,
+      "step": 10600
+    },
+    {
+      "epoch": 0.2121,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00019244129586334672,
+      "loss": 1.2667,
+      "step": 10605
+    },
+    {
+      "epoch": 0.2122,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001924279771017706,
+      "loss": 1.2709,
+      "step": 10610
+    },
+    {
+      "epoch": 0.2123,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019241464707812585,
+      "loss": 1.273,
+      "step": 10615
+    },
+    {
+      "epoch": 0.2124,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001924013057940367,
+      "loss": 1.2754,
+      "step": 10620
+    },
+    {
+      "epoch": 0.2125,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001923879532511287,
+      "loss": 1.2721,
+      "step": 10625
+    },
+    {
+      "epoch": 0.2126,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00019237458945102882,
+      "loss": 1.3034,
+      "step": 10630
+    },
+    {
+      "epoch": 0.2127,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00019236121439536542,
+      "loss": 1.2749,
+      "step": 10635
+    },
+    {
+      "epoch": 0.2128,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019234782808576824,
+      "loss": 1.2794,
+      "step": 10640
+    },
+    {
+      "epoch": 0.2129,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001923344305238683,
+      "loss": 1.2762,
+      "step": 10645
+    },
+    {
+      "epoch": 0.213,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00019232102171129811,
+      "loss": 1.2927,
+      "step": 10650
+    },
+    {
+      "epoch": 0.2131,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00019230760164969143,
+      "loss": 1.2781,
+      "step": 10655
+    },
+    {
+      "epoch": 0.2132,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001922941703406835,
+      "loss": 1.2729,
+      "step": 10660
+    },
+    {
+      "epoch": 0.2133,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00019228072778591088,
+      "loss": 1.2865,
+      "step": 10665
+    },
+    {
+      "epoch": 0.2134,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001922672739870115,
+      "loss": 1.2795,
+      "step": 10670
+    },
+    {
+      "epoch": 0.2135,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00019225380894562463,
+      "loss": 1.2793,
+      "step": 10675
+    },
+    {
+      "epoch": 0.2136,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00019224033266339102,
+      "loss": 1.2716,
+      "step": 10680
+    },
+    {
+      "epoch": 0.2137,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019222684514195264,
+      "loss": 1.2835,
+      "step": 10685
+    },
+    {
+      "epoch": 0.2138,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00019221334638295294,
+      "loss": 1.2735,
+      "step": 10690
+    },
+    {
+      "epoch": 0.2139,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001921998363880367,
+      "loss": 1.2762,
+      "step": 10695
+    },
+    {
+      "epoch": 0.214,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019218631515885006,
+      "loss": 1.2706,
+      "step": 10700
+    },
+    {
+      "epoch": 0.2141,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00019217278269704056,
+      "loss": 1.2814,
+      "step": 10705
+    },
+    {
+      "epoch": 0.2142,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00019215923900425707,
+      "loss": 1.2852,
+      "step": 10710
+    },
+    {
+      "epoch": 0.2143,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019214568408214985,
+      "loss": 1.2629,
+      "step": 10715
+    },
+    {
+      "epoch": 0.2144,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019213211793237057,
+      "loss": 1.2669,
+      "step": 10720
+    },
+    {
+      "epoch": 0.2145,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00019211854055657215,
+      "loss": 1.29,
+      "step": 10725
+    },
+    {
+      "epoch": 0.2146,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019210495195640895,
+      "loss": 1.2699,
+      "step": 10730
+    },
+    {
+      "epoch": 0.2147,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0001920913521335368,
+      "loss": 1.2685,
+      "step": 10735
+    },
+    {
+      "epoch": 0.2148,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019207774108961272,
+      "loss": 1.292,
+      "step": 10740
+    },
+    {
+      "epoch": 0.2149,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019206411882629517,
+      "loss": 1.2727,
+      "step": 10745
+    },
+    {
+      "epoch": 0.215,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00019205048534524406,
+      "loss": 1.2874,
+      "step": 10750
+    },
+    {
+      "epoch": 0.2151,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019203684064812045,
+      "loss": 1.2604,
+      "step": 10755
+    },
+    {
+      "epoch": 0.2152,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019202318473658705,
+      "loss": 1.2611,
+      "step": 10760
+    },
+    {
+      "epoch": 0.2153,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0001920095176123077,
+      "loss": 1.2719,
+      "step": 10765
+    },
+    {
+      "epoch": 0.2154,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00019199583927694772,
+      "loss": 1.2937,
+      "step": 10770
+    },
+    {
+      "epoch": 0.2155,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00019198214973217378,
+      "loss": 1.2747,
+      "step": 10775
+    },
+    {
+      "epoch": 0.2156,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00019196844897965393,
+      "loss": 1.293,
+      "step": 10780
+    },
+    {
+      "epoch": 0.2157,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00019195473702105748,
+      "loss": 1.2817,
+      "step": 10785
+    },
+    {
+      "epoch": 0.2158,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001919410138580553,
+      "loss": 1.2926,
+      "step": 10790
+    },
+    {
+      "epoch": 0.2159,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00019192727949231945,
+      "loss": 1.2743,
+      "step": 10795
+    },
+    {
+      "epoch": 0.216,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019191353392552344,
+      "loss": 1.2652,
+      "step": 10800
+    },
+    {
+      "epoch": 0.2161,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00019189977715934213,
+      "loss": 1.2712,
+      "step": 10805
+    },
+    {
+      "epoch": 0.2162,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00019188600919545174,
+      "loss": 1.2713,
+      "step": 10810
+    },
+    {
+      "epoch": 0.2163,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019187223003552985,
+      "loss": 1.2827,
+      "step": 10815
+    },
+    {
+      "epoch": 0.2164,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0001918584396812554,
+      "loss": 1.2867,
+      "step": 10820
+    },
+    {
+      "epoch": 0.2165,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00019184463813430873,
+      "loss": 1.2597,
+      "step": 10825
+    },
+    {
+      "epoch": 0.2166,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019183082539637146,
+      "loss": 1.2815,
+      "step": 10830
+    },
+    {
+      "epoch": 0.2167,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001918170014691267,
+      "loss": 1.2751,
+      "step": 10835
+    },
+    {
+      "epoch": 0.2168,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001918031663542588,
+      "loss": 1.2824,
+      "step": 10840
+    },
+    {
+      "epoch": 0.2169,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001917893200534536,
+      "loss": 1.2825,
+      "step": 10845
+    },
+    {
+      "epoch": 0.217,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019177546256839812,
+      "loss": 1.276,
+      "step": 10850
+    },
+    {
+      "epoch": 0.2171,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019176159390078094,
+      "loss": 1.2834,
+      "step": 10855
+    },
+    {
+      "epoch": 0.2172,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019174771405229186,
+      "loss": 1.2771,
+      "step": 10860
+    },
+    {
+      "epoch": 0.2173,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00019173382302462214,
+      "loss": 1.2725,
+      "step": 10865
+    },
+    {
+      "epoch": 0.2174,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00019171992081946435,
+      "loss": 1.2756,
+      "step": 10870
+    },
+    {
+      "epoch": 0.2175,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001917060074385124,
+      "loss": 1.2619,
+      "step": 10875
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019169208288346166,
+      "loss": 1.2662,
+      "step": 10880
+    },
+    {
+      "epoch": 0.2177,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001916781471560087,
+      "loss": 1.2562,
+      "step": 10885
+    },
+    {
+      "epoch": 0.2178,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019166420025785164,
+      "loss": 1.2912,
+      "step": 10890
+    },
+    {
+      "epoch": 0.2179,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.0001916502421906898,
+      "loss": 1.265,
+      "step": 10895
+    },
+    {
+      "epoch": 0.218,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00019163627295622397,
+      "loss": 1.2902,
+      "step": 10900
+    },
+    {
+      "epoch": 0.2181,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.00019162229255615624,
+      "loss": 1.2792,
+      "step": 10905
+    },
+    {
+      "epoch": 0.2182,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019160830099219006,
+      "loss": 1.27,
+      "step": 10910
+    },
+    {
+      "epoch": 0.2183,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019159429826603032,
+      "loss": 1.2746,
+      "step": 10915
+    },
+    {
+      "epoch": 0.2184,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019158028437938317,
+      "loss": 1.2873,
+      "step": 10920
+    },
+    {
+      "epoch": 0.2185,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019156625933395614,
+      "loss": 1.2934,
+      "step": 10925
+    },
+    {
+      "epoch": 0.2186,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019155222313145816,
+      "loss": 1.2979,
+      "step": 10930
+    },
+    {
+      "epoch": 0.2187,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001915381757735995,
+      "loss": 1.2877,
+      "step": 10935
+    },
+    {
+      "epoch": 0.2188,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019152411726209176,
+      "loss": 1.2627,
+      "step": 10940
+    },
+    {
+      "epoch": 0.2189,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.000191510047598648,
+      "loss": 1.2816,
+      "step": 10945
+    },
+    {
+      "epoch": 0.219,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.0001914959667849825,
+      "loss": 1.2676,
+      "step": 10950
+    },
+    {
+      "epoch": 0.2191,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019148187482281097,
+      "loss": 1.2759,
+      "step": 10955
+    },
+    {
+      "epoch": 0.2192,
+      "grad_norm": 1.625,
+      "learning_rate": 0.0001914677717138505,
+      "loss": 1.2799,
+      "step": 10960
+    },
+    {
+      "epoch": 0.2193,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00019145365745981948,
+      "loss": 1.2868,
+      "step": 10965
+    },
+    {
+      "epoch": 0.2194,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00019143953206243776,
+      "loss": 1.2862,
+      "step": 10970
+    },
+    {
+      "epoch": 0.2195,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00019142539552342638,
+      "loss": 1.2696,
+      "step": 10975
+    },
+    {
+      "epoch": 0.2196,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001914112478445079,
+      "loss": 1.2765,
+      "step": 10980
+    },
+    {
+      "epoch": 0.2197,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00019139708902740613,
+      "loss": 1.2714,
+      "step": 10985
+    },
+    {
+      "epoch": 0.2198,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001913829190738463,
+      "loss": 1.2846,
+      "step": 10990
+    },
+    {
+      "epoch": 0.2199,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.000191368737985555,
+      "loss": 1.2598,
+      "step": 10995
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001913545457642601,
+      "loss": 1.2858,
+      "step": 11000
+    },
+    {
+      "epoch": 0.2201,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0001913403424116909,
+      "loss": 1.2813,
+      "step": 11005
+    },
+    {
+      "epoch": 0.2202,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00019132612792957808,
+      "loss": 1.2662,
+      "step": 11010
+    },
+    {
+      "epoch": 0.2203,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00019131190231965356,
+      "loss": 1.2901,
+      "step": 11015
+    },
+    {
+      "epoch": 0.2204,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019129766558365074,
+      "loss": 1.2643,
+      "step": 11020
+    },
+    {
+      "epoch": 0.2205,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0001912834177233043,
+      "loss": 1.2632,
+      "step": 11025
+    },
+    {
+      "epoch": 0.2206,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001912691587403503,
+      "loss": 1.2718,
+      "step": 11030
+    },
+    {
+      "epoch": 0.2207,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00019125488863652615,
+      "loss": 1.2641,
+      "step": 11035
+    },
+    {
+      "epoch": 0.2208,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019124060741357063,
+      "loss": 1.2832,
+      "step": 11040
+    },
+    {
+      "epoch": 0.2209,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00019122631507322387,
+      "loss": 1.2741,
+      "step": 11045
+    },
+    {
+      "epoch": 0.221,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001912120116172273,
+      "loss": 1.2735,
+      "step": 11050
+    },
+    {
+      "epoch": 0.2211,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019119769704732382,
+      "loss": 1.2763,
+      "step": 11055
+    },
+    {
+      "epoch": 0.2212,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001911833713652576,
+      "loss": 1.2651,
+      "step": 11060
+    },
+    {
+      "epoch": 0.2213,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00019116903457277413,
+      "loss": 1.2713,
+      "step": 11065
+    },
+    {
+      "epoch": 0.2214,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019115468667162038,
+      "loss": 1.2707,
+      "step": 11070
+    },
+    {
+      "epoch": 0.2215,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019114032766354453,
+      "loss": 1.2792,
+      "step": 11075
+    },
+    {
+      "epoch": 0.2216,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019112595755029624,
+      "loss": 1.2801,
+      "step": 11080
+    },
+    {
+      "epoch": 0.2217,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001911115763336264,
+      "loss": 1.2824,
+      "step": 11085
+    },
+    {
+      "epoch": 0.2218,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0001910971840152874,
+      "loss": 1.2784,
+      "step": 11090
+    },
+    {
+      "epoch": 0.2219,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001910827805970328,
+      "loss": 1.2967,
+      "step": 11095
+    },
+    {
+      "epoch": 0.222,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00019106836608061772,
+      "loss": 1.2717,
+      "step": 11100
+    },
+    {
+      "epoch": 0.2221,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019105394046779845,
+      "loss": 1.2777,
+      "step": 11105
+    },
+    {
+      "epoch": 0.2222,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019103950376033276,
+      "loss": 1.2762,
+      "step": 11110
+    },
+    {
+      "epoch": 0.2223,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019102505595997965,
+      "loss": 1.269,
+      "step": 11115
+    },
+    {
+      "epoch": 0.2224,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019101059706849957,
+      "loss": 1.2835,
+      "step": 11120
+    },
+    {
+      "epoch": 0.2225,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00019099612708765434,
+      "loss": 1.2763,
+      "step": 11125
+    },
+    {
+      "epoch": 0.2226,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.000190981646019207,
+      "loss": 1.282,
+      "step": 11130
+    },
+    {
+      "epoch": 0.2227,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001909671538649221,
+      "loss": 1.2803,
+      "step": 11135
+    },
+    {
+      "epoch": 0.2228,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019095265062656544,
+      "loss": 1.288,
+      "step": 11140
+    },
+    {
+      "epoch": 0.2229,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00019093813630590418,
+      "loss": 1.2719,
+      "step": 11145
+    },
+    {
+      "epoch": 0.223,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019092361090470688,
+      "loss": 1.2801,
+      "step": 11150
+    },
+    {
+      "epoch": 0.2231,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00019090907442474334,
+      "loss": 1.2796,
+      "step": 11155
+    },
+    {
+      "epoch": 0.2232,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00019089452686778488,
+      "loss": 1.2712,
+      "step": 11160
+    },
+    {
+      "epoch": 0.2233,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00019087996823560402,
+      "loss": 1.2716,
+      "step": 11165
+    },
+    {
+      "epoch": 0.2234,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001908653985299747,
+      "loss": 1.2727,
+      "step": 11170
+    },
+    {
+      "epoch": 0.2235,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001908508177526722,
+      "loss": 1.2937,
+      "step": 11175
+    },
+    {
+      "epoch": 0.2236,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00019083622590547312,
+      "loss": 1.2946,
+      "step": 11180
+    },
+    {
+      "epoch": 0.2237,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00019082162299015546,
+      "loss": 1.2654,
+      "step": 11185
+    },
+    {
+      "epoch": 0.2238,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00019080700900849851,
+      "loss": 1.2723,
+      "step": 11190
+    },
+    {
+      "epoch": 0.2239,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.000190792383962283,
+      "loss": 1.2633,
+      "step": 11195
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00019077774785329087,
+      "loss": 1.2635,
+      "step": 11200
+    },
+    {
+      "epoch": 0.2241,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00019076310068330554,
+      "loss": 1.2734,
+      "step": 11205
+    },
+    {
+      "epoch": 0.2242,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001907484424541117,
+      "loss": 1.2709,
+      "step": 11210
+    },
+    {
+      "epoch": 0.2243,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019073377316749542,
+      "loss": 1.2917,
+      "step": 11215
+    },
+    {
+      "epoch": 0.2244,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00019071909282524413,
+      "loss": 1.2881,
+      "step": 11220
+    },
+    {
+      "epoch": 0.2245,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001907044014291465,
+      "loss": 1.2729,
+      "step": 11225
+    },
+    {
+      "epoch": 0.2246,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001906896989809927,
+      "loss": 1.269,
+      "step": 11230
+    },
+    {
+      "epoch": 0.2247,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00019067498548257423,
+      "loss": 1.2575,
+      "step": 11235
+    },
+    {
+      "epoch": 0.2248,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00019066026093568378,
+      "loss": 1.3064,
+      "step": 11240
+    },
+    {
+      "epoch": 0.2249,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00019064552534211554,
+      "loss": 1.2662,
+      "step": 11245
+    },
+    {
+      "epoch": 0.225,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.000190630778703665,
+      "loss": 1.2726,
+      "step": 11250
+    },
+    {
+      "epoch": 0.2251,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00019061602102212898,
+      "loss": 1.2845,
+      "step": 11255
+    },
+    {
+      "epoch": 0.2252,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001906012522993057,
+      "loss": 1.2765,
+      "step": 11260
+    },
+    {
+      "epoch": 0.2253,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001905864725369946,
+      "loss": 1.2713,
+      "step": 11265
+    },
+    {
+      "epoch": 0.2254,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019057168173699664,
+      "loss": 1.2837,
+      "step": 11270
+    },
+    {
+      "epoch": 0.2255,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019055687990111398,
+      "loss": 1.2667,
+      "step": 11275
+    },
+    {
+      "epoch": 0.2256,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.0001905420670311502,
+      "loss": 1.2929,
+      "step": 11280
+    },
+    {
+      "epoch": 0.2257,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00019052724312891014,
+      "loss": 1.2869,
+      "step": 11285
+    },
+    {
+      "epoch": 0.2258,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00019051240819620014,
+      "loss": 1.2694,
+      "step": 11290
+    },
+    {
+      "epoch": 0.2259,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0001904975622348278,
+      "loss": 1.2792,
+      "step": 11295
+    },
+    {
+      "epoch": 0.226,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019048270524660196,
+      "loss": 1.272,
+      "step": 11300
+    },
+    {
+      "epoch": 0.2261,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00019046783723333297,
+      "loss": 1.2941,
+      "step": 11305
+    },
+    {
+      "epoch": 0.2262,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019045295819683242,
+      "loss": 1.2748,
+      "step": 11310
+    },
+    {
+      "epoch": 0.2263,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001904380681389133,
+      "loss": 1.2815,
+      "step": 11315
+    },
+    {
+      "epoch": 0.2264,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00019042316706138987,
+      "loss": 1.2735,
+      "step": 11320
+    },
+    {
+      "epoch": 0.2265,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00019040825496607786,
+      "loss": 1.2714,
+      "step": 11325
+    },
+    {
+      "epoch": 0.2266,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019039333185479418,
+      "loss": 1.2654,
+      "step": 11330
+    },
+    {
+      "epoch": 0.2267,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0001903783977293572,
+      "loss": 1.2678,
+      "step": 11335
+    },
+    {
+      "epoch": 0.2268,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019036345259158667,
+      "loss": 1.2708,
+      "step": 11340
+    },
+    {
+      "epoch": 0.2269,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0001903484964433035,
+      "loss": 1.2666,
+      "step": 11345
+    },
+    {
+      "epoch": 0.227,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0001903335292863301,
+      "loss": 1.2821,
+      "step": 11350
+    },
+    {
+      "epoch": 0.2271,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019031855112249015,
+      "loss": 1.2649,
+      "step": 11355
+    },
+    {
+      "epoch": 0.2272,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00019030356195360874,
+      "loss": 1.2597,
+      "step": 11360
+    },
+    {
+      "epoch": 0.2273,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001902885617815122,
+      "loss": 1.2797,
+      "step": 11365
+    },
+    {
+      "epoch": 0.2274,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001902735506080283,
+      "loss": 1.2682,
+      "step": 11370
+    },
+    {
+      "epoch": 0.2275,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00019025852843498607,
+      "loss": 1.2672,
+      "step": 11375
+    },
+    {
+      "epoch": 0.2276,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00019024349526421594,
+      "loss": 1.2925,
+      "step": 11380
+    },
+    {
+      "epoch": 0.2277,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019022845109754966,
+      "loss": 1.2805,
+      "step": 11385
+    },
+    {
+      "epoch": 0.2278,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00019021339593682028,
+      "loss": 1.266,
+      "step": 11390
+    },
+    {
+      "epoch": 0.2279,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019019832978386228,
+      "loss": 1.2861,
+      "step": 11395
+    },
+    {
+      "epoch": 0.228,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0001901832526405114,
+      "loss": 1.2812,
+      "step": 11400
+    },
+    {
+      "epoch": 0.2281,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019016816450860474,
+      "loss": 1.2709,
+      "step": 11405
+    },
+    {
+      "epoch": 0.2282,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001901530653899807,
+      "loss": 1.281,
+      "step": 11410
+    },
+    {
+      "epoch": 0.2283,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00019013795528647912,
+      "loss": 1.2714,
+      "step": 11415
+    },
+    {
+      "epoch": 0.2284,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019012283419994115,
+      "loss": 1.2746,
+      "step": 11420
+    },
+    {
+      "epoch": 0.2285,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00019010770213220916,
+      "loss": 1.2657,
+      "step": 11425
+    },
+    {
+      "epoch": 0.2286,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.000190092559085127,
+      "loss": 1.2674,
+      "step": 11430
+    },
+    {
+      "epoch": 0.2287,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019007740506053983,
+      "loss": 1.2821,
+      "step": 11435
+    },
+    {
+      "epoch": 0.2288,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019006224006029406,
+      "loss": 1.2724,
+      "step": 11440
+    },
+    {
+      "epoch": 0.2289,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001900470640862375,
+      "loss": 1.2689,
+      "step": 11445
+    },
+    {
+      "epoch": 0.229,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00019003187714021938,
+      "loss": 1.2902,
+      "step": 11450
+    },
+    {
+      "epoch": 0.2291,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00019001667922409008,
+      "loss": 1.2839,
+      "step": 11455
+    },
+    {
+      "epoch": 0.2292,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00019000147033970148,
+      "loss": 1.2686,
+      "step": 11460
+    },
+    {
+      "epoch": 0.2293,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018998625048890672,
+      "loss": 1.2747,
+      "step": 11465
+    },
+    {
+      "epoch": 0.2294,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.0001899710196735603,
+      "loss": 1.2738,
+      "step": 11470
+    },
+    {
+      "epoch": 0.2295,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00018995577789551803,
+      "loss": 1.2655,
+      "step": 11475
+    },
+    {
+      "epoch": 0.2296,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001899405251566371,
+      "loss": 1.2857,
+      "step": 11480
+    },
+    {
+      "epoch": 0.2297,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.000189925261458776,
+      "loss": 1.2761,
+      "step": 11485
+    },
+    {
+      "epoch": 0.2298,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00018990998680379456,
+      "loss": 1.2984,
+      "step": 11490
+    },
+    {
+      "epoch": 0.2299,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018989470119355398,
+      "loss": 1.2758,
+      "step": 11495
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.0001898794046299167,
+      "loss": 1.2883,
+      "step": 11500
+    },
+    {
+      "epoch": 0.2301,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00018986409711474665,
+      "loss": 1.2863,
+      "step": 11505
+    },
+    {
+      "epoch": 0.2302,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00018984877864990888,
+      "loss": 1.2762,
+      "step": 11510
+    },
+    {
+      "epoch": 0.2303,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00018983344923727003,
+      "loss": 1.2624,
+      "step": 11515
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00018981810887869785,
+      "loss": 1.28,
+      "step": 11520
+    },
+    {
+      "epoch": 0.2305,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00018980275757606157,
+      "loss": 1.2879,
+      "step": 11525
+    },
+    {
+      "epoch": 0.2306,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001897873953312317,
+      "loss": 1.2712,
+      "step": 11530
+    },
+    {
+      "epoch": 0.2307,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00018977202214608,
+      "loss": 1.269,
+      "step": 11535
+    },
+    {
+      "epoch": 0.2308,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00018975663802247976,
+      "loss": 1.2749,
+      "step": 11540
+    },
+    {
+      "epoch": 0.2309,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001897412429623054,
+      "loss": 1.284,
+      "step": 11545
+    },
+    {
+      "epoch": 0.231,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00018972583696743285,
+      "loss": 1.2711,
+      "step": 11550
+    },
+    {
+      "epoch": 0.2311,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00018971042003973924,
+      "loss": 1.2634,
+      "step": 11555
+    },
+    {
+      "epoch": 0.2312,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.000189694992181103,
+      "loss": 1.2941,
+      "step": 11560
+    },
+    {
+      "epoch": 0.2313,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018967955339340407,
+      "loss": 1.2643,
+      "step": 11565
+    },
+    {
+      "epoch": 0.2314,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018966410367852362,
+      "loss": 1.2801,
+      "step": 11570
+    },
+    {
+      "epoch": 0.2315,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00018964864303834406,
+      "loss": 1.2796,
+      "step": 11575
+    },
+    {
+      "epoch": 0.2316,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0001896331714747493,
+      "loss": 1.2902,
+      "step": 11580
+    },
+    {
+      "epoch": 0.2317,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001896176889896245,
+      "loss": 1.2961,
+      "step": 11585
+    },
+    {
+      "epoch": 0.2318,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001896021955848561,
+      "loss": 1.2764,
+      "step": 11590
+    },
+    {
+      "epoch": 0.2319,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00018958669126233199,
+      "loss": 1.2629,
+      "step": 11595
+    },
+    {
+      "epoch": 0.232,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001895711760239413,
+      "loss": 1.268,
+      "step": 11600
+    },
+    {
+      "epoch": 0.2321,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001895556498715745,
+      "loss": 1.2843,
+      "step": 11605
+    },
+    {
+      "epoch": 0.2322,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001895401128071234,
+      "loss": 1.2742,
+      "step": 11610
+    },
+    {
+      "epoch": 0.2323,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00018952456483248119,
+      "loss": 1.2693,
+      "step": 11615
+    },
+    {
+      "epoch": 0.2324,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018950900594954227,
+      "loss": 1.285,
+      "step": 11620
+    },
+    {
+      "epoch": 0.2325,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00018949343616020252,
+      "loss": 1.2646,
+      "step": 11625
+    },
+    {
+      "epoch": 0.2326,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018947785546635904,
+      "loss": 1.292,
+      "step": 11630
+    },
+    {
+      "epoch": 0.2327,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00018946226386991027,
+      "loss": 1.2673,
+      "step": 11635
+    },
+    {
+      "epoch": 0.2328,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.000189446661372756,
+      "loss": 1.2798,
+      "step": 11640
+    },
+    {
+      "epoch": 0.2329,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001894310479767974,
+      "loss": 1.2679,
+      "step": 11645
+    },
+    {
+      "epoch": 0.233,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001894154236839368,
+      "loss": 1.2776,
+      "step": 11650
+    },
+    {
+      "epoch": 0.2331,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00018939978849607814,
+      "loss": 1.284,
+      "step": 11655
+    },
+    {
+      "epoch": 0.2332,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001893841424151264,
+      "loss": 1.2581,
+      "step": 11660
+    },
+    {
+      "epoch": 0.2333,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.000189368485442988,
+      "loss": 1.2659,
+      "step": 11665
+    },
+    {
+      "epoch": 0.2334,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018935281758157078,
+      "loss": 1.2799,
+      "step": 11670
+    },
+    {
+      "epoch": 0.2335,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00018933713883278376,
+      "loss": 1.2703,
+      "step": 11675
+    },
+    {
+      "epoch": 0.2336,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001893214491985374,
+      "loss": 1.2701,
+      "step": 11680
+    },
+    {
+      "epoch": 0.2337,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00018930574868074334,
+      "loss": 1.2693,
+      "step": 11685
+    },
+    {
+      "epoch": 0.2338,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001892900372813147,
+      "loss": 1.2884,
+      "step": 11690
+    },
+    {
+      "epoch": 0.2339,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00018927431500216586,
+      "loss": 1.2653,
+      "step": 11695
+    },
+    {
+      "epoch": 0.234,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018925858184521256,
+      "loss": 1.2788,
+      "step": 11700
+    },
+    {
+      "epoch": 0.2341,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001892428378123718,
+      "loss": 1.2751,
+      "step": 11705
+    },
+    {
+      "epoch": 0.2342,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00018922708290556198,
+      "loss": 1.2785,
+      "step": 11710
+    },
+    {
+      "epoch": 0.2343,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001892113171267027,
+      "loss": 1.2845,
+      "step": 11715
+    },
+    {
+      "epoch": 0.2344,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001891955404777151,
+      "loss": 1.2677,
+      "step": 11720
+    },
+    {
+      "epoch": 0.2345,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00018917975296052142,
+      "loss": 1.2727,
+      "step": 11725
+    },
+    {
+      "epoch": 0.2346,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00018916395457704534,
+      "loss": 1.2699,
+      "step": 11730
+    },
+    {
+      "epoch": 0.2347,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00018914814532921187,
+      "loss": 1.2854,
+      "step": 11735
+    },
+    {
+      "epoch": 0.2348,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00018913232521894732,
+      "loss": 1.2671,
+      "step": 11740
+    },
+    {
+      "epoch": 0.2349,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018911649424817933,
+      "loss": 1.2712,
+      "step": 11745
+    },
+    {
+      "epoch": 0.235,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001891006524188368,
+      "loss": 1.2645,
+      "step": 11750
+    },
+    {
+      "epoch": 0.2351,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018908479973285005,
+      "loss": 1.2551,
+      "step": 11755
+    },
+    {
+      "epoch": 0.2352,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018906893619215066,
+      "loss": 1.2725,
+      "step": 11760
+    },
+    {
+      "epoch": 0.2353,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001890530617986716,
+      "loss": 1.2886,
+      "step": 11765
+    },
+    {
+      "epoch": 0.2354,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018903717655434707,
+      "loss": 1.2698,
+      "step": 11770
+    },
+    {
+      "epoch": 0.2355,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00018902128046111266,
+      "loss": 1.2697,
+      "step": 11775
+    },
+    {
+      "epoch": 0.2356,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00018900537352090524,
+      "loss": 1.2709,
+      "step": 11780
+    },
+    {
+      "epoch": 0.2357,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00018898945573566308,
+      "loss": 1.2974,
+      "step": 11785
+    },
+    {
+      "epoch": 0.2358,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00018897352710732564,
+      "loss": 1.2821,
+      "step": 11790
+    },
+    {
+      "epoch": 0.2359,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00018895758763783383,
+      "loss": 1.2817,
+      "step": 11795
+    },
+    {
+      "epoch": 0.236,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00018894163732912977,
+      "loss": 1.2872,
+      "step": 11800
+    },
+    {
+      "epoch": 0.2361,
+      "grad_norm": 0.75,
+      "learning_rate": 0.000188925676183157,
+      "loss": 1.2719,
+      "step": 11805
+    },
+    {
+      "epoch": 0.2362,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00018890970420186033,
+      "loss": 1.2973,
+      "step": 11810
+    },
+    {
+      "epoch": 0.2363,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001888937213871859,
+      "loss": 1.2812,
+      "step": 11815
+    },
+    {
+      "epoch": 0.2364,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00018887772774108116,
+      "loss": 1.2838,
+      "step": 11820
+    },
+    {
+      "epoch": 0.2365,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0001888617232654949,
+      "loss": 1.2821,
+      "step": 11825
+    },
+    {
+      "epoch": 0.2366,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00018884570796237718,
+      "loss": 1.2819,
+      "step": 11830
+    },
+    {
+      "epoch": 0.2367,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018882968183367947,
+      "loss": 1.2606,
+      "step": 11835
+    },
+    {
+      "epoch": 0.2368,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018881364488135448,
+      "loss": 1.2709,
+      "step": 11840
+    },
+    {
+      "epoch": 0.2369,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018879759710735622,
+      "loss": 1.2728,
+      "step": 11845
+    },
+    {
+      "epoch": 0.237,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00018878153851364013,
+      "loss": 1.2705,
+      "step": 11850
+    },
+    {
+      "epoch": 0.2371,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018876546910216288,
+      "loss": 1.2654,
+      "step": 11855
+    },
+    {
+      "epoch": 0.2372,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00018874938887488248,
+      "loss": 1.2831,
+      "step": 11860
+    },
+    {
+      "epoch": 0.2373,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00018873329783375824,
+      "loss": 1.2818,
+      "step": 11865
+    },
+    {
+      "epoch": 0.2374,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0001887171959807508,
+      "loss": 1.2781,
+      "step": 11870
+    },
+    {
+      "epoch": 0.2375,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00018870108331782217,
+      "loss": 1.2629,
+      "step": 11875
+    },
+    {
+      "epoch": 0.2376,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0001886849598469356,
+      "loss": 1.2592,
+      "step": 11880
+    },
+    {
+      "epoch": 0.2377,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018866882557005567,
+      "loss": 1.2649,
+      "step": 11885
+    },
+    {
+      "epoch": 0.2378,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018865268048914828,
+      "loss": 1.2722,
+      "step": 11890
+    },
+    {
+      "epoch": 0.2379,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001886365246061807,
+      "loss": 1.2653,
+      "step": 11895
+    },
+    {
+      "epoch": 0.238,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00018862035792312147,
+      "loss": 1.2751,
+      "step": 11900
+    },
+    {
+      "epoch": 0.2381,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00018860418044194045,
+      "loss": 1.2787,
+      "step": 11905
+    },
+    {
+      "epoch": 0.2382,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00018858799216460881,
+      "loss": 1.2673,
+      "step": 11910
+    },
+    {
+      "epoch": 0.2383,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00018857179309309901,
+      "loss": 1.2963,
+      "step": 11915
+    },
+    {
+      "epoch": 0.2384,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00018855558322938493,
+      "loss": 1.274,
+      "step": 11920
+    },
+    {
+      "epoch": 0.2385,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0001885393625754416,
+      "loss": 1.2696,
+      "step": 11925
+    },
+    {
+      "epoch": 0.2386,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00018852313113324552,
+      "loss": 1.2779,
+      "step": 11930
+    },
+    {
+      "epoch": 0.2387,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00018850688890477445,
+      "loss": 1.291,
+      "step": 11935
+    },
+    {
+      "epoch": 0.2388,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00018849063589200743,
+      "loss": 1.2713,
+      "step": 11940
+    },
+    {
+      "epoch": 0.2389,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00018847437209692486,
+      "loss": 1.2747,
+      "step": 11945
+    },
+    {
+      "epoch": 0.239,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001884580975215084,
+      "loss": 1.2701,
+      "step": 11950
+    },
+    {
+      "epoch": 0.2391,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001884418121677411,
+      "loss": 1.2788,
+      "step": 11955
+    },
+    {
+      "epoch": 0.2392,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00018842551603760724,
+      "loss": 1.261,
+      "step": 11960
+    },
+    {
+      "epoch": 0.2393,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001884092091330925,
+      "loss": 1.2894,
+      "step": 11965
+    },
+    {
+      "epoch": 0.2394,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00018839289145618378,
+      "loss": 1.2813,
+      "step": 11970
+    },
+    {
+      "epoch": 0.2395,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00018837656300886937,
+      "loss": 1.2695,
+      "step": 11975
+    },
+    {
+      "epoch": 0.2396,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00018836022379313883,
+      "loss": 1.2741,
+      "step": 11980
+    },
+    {
+      "epoch": 0.2397,
+      "grad_norm": 0.375,
+      "learning_rate": 0.000188343873810983,
+      "loss": 1.2804,
+      "step": 11985
+    },
+    {
+      "epoch": 0.2398,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018832751306439418,
+      "loss": 1.2852,
+      "step": 11990
+    },
+    {
+      "epoch": 0.2399,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001883111415553658,
+      "loss": 1.2708,
+      "step": 11995
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00018829475928589271,
+      "loss": 1.2715,
+      "step": 12000
+    },
+    {
+      "epoch": 0.2401,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018827836625797103,
+      "loss": 1.279,
+      "step": 12005
+    },
+    {
+      "epoch": 0.2402,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00018826196247359817,
+      "loss": 1.2686,
+      "step": 12010
+    },
+    {
+      "epoch": 0.2403,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018824554793477294,
+      "loss": 1.2717,
+      "step": 12015
+    },
+    {
+      "epoch": 0.2404,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00018822912264349534,
+      "loss": 1.2779,
+      "step": 12020
+    },
+    {
+      "epoch": 0.2405,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00018821268660176678,
+      "loss": 1.2813,
+      "step": 12025
+    },
+    {
+      "epoch": 0.2406,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00018819623981158995,
+      "loss": 1.2713,
+      "step": 12030
+    },
+    {
+      "epoch": 0.2407,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00018817978227496883,
+      "loss": 1.2863,
+      "step": 12035
+    },
+    {
+      "epoch": 0.2408,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001881633139939087,
+      "loss": 1.2649,
+      "step": 12040
+    },
+    {
+      "epoch": 0.2409,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.0001881468349704162,
+      "loss": 1.2769,
+      "step": 12045
+    },
+    {
+      "epoch": 0.241,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001881303452064992,
+      "loss": 1.2918,
+      "step": 12050
+    },
+    {
+      "epoch": 0.2411,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00018811384470416705,
+      "loss": 1.283,
+      "step": 12055
+    },
+    {
+      "epoch": 0.2412,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00018809733346543013,
+      "loss": 1.273,
+      "step": 12060
+    },
+    {
+      "epoch": 0.2413,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00018808081149230036,
+      "loss": 1.2716,
+      "step": 12065
+    },
+    {
+      "epoch": 0.2414,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00018806427878679093,
+      "loss": 1.2632,
+      "step": 12070
+    },
+    {
+      "epoch": 0.2415,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001880477353509162,
+      "loss": 1.2862,
+      "step": 12075
+    },
+    {
+      "epoch": 0.2416,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00018803118118669202,
+      "loss": 1.2764,
+      "step": 12080
+    },
+    {
+      "epoch": 0.2417,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018801461629613546,
+      "loss": 1.2914,
+      "step": 12085
+    },
+    {
+      "epoch": 0.2418,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018799804068126485,
+      "loss": 1.2814,
+      "step": 12090
+    },
+    {
+      "epoch": 0.2419,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001879814543440999,
+      "loss": 1.2741,
+      "step": 12095
+    },
+    {
+      "epoch": 0.242,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00018796485728666165,
+      "loss": 1.2886,
+      "step": 12100
+    },
+    {
+      "epoch": 0.2421,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00018794824951097236,
+      "loss": 1.2883,
+      "step": 12105
+    },
+    {
+      "epoch": 0.2422,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018793163101905563,
+      "loss": 1.2766,
+      "step": 12110
+    },
+    {
+      "epoch": 0.2423,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001879150018129364,
+      "loss": 1.2744,
+      "step": 12115
+    },
+    {
+      "epoch": 0.2424,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00018789836189464086,
+      "loss": 1.2837,
+      "step": 12120
+    },
+    {
+      "epoch": 0.2425,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018788171126619653,
+      "loss": 1.2563,
+      "step": 12125
+    },
+    {
+      "epoch": 0.2426,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0001878650499296323,
+      "loss": 1.266,
+      "step": 12130
+    },
+    {
+      "epoch": 0.2427,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018784837788697823,
+      "loss": 1.2743,
+      "step": 12135
+    },
+    {
+      "epoch": 0.2428,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00018783169514026578,
+      "loss": 1.3062,
+      "step": 12140
+    },
+    {
+      "epoch": 0.2429,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00018781500169152773,
+      "loss": 1.2728,
+      "step": 12145
+    },
+    {
+      "epoch": 0.243,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00018779829754279805,
+      "loss": 1.2867,
+      "step": 12150
+    },
+    {
+      "epoch": 0.2431,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018778158269611218,
+      "loss": 1.2659,
+      "step": 12155
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00018776485715350671,
+      "loss": 1.2737,
+      "step": 12160
+    },
+    {
+      "epoch": 0.2433,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00018774812091701962,
+      "loss": 1.2714,
+      "step": 12165
+    },
+    {
+      "epoch": 0.2434,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00018773137398869015,
+      "loss": 1.2749,
+      "step": 12170
+    },
+    {
+      "epoch": 0.2435,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00018771461637055888,
+      "loss": 1.2684,
+      "step": 12175
+    },
+    {
+      "epoch": 0.2436,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001876978480646677,
+      "loss": 1.2893,
+      "step": 12180
+    },
+    {
+      "epoch": 0.2437,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018768106907305973,
+      "loss": 1.287,
+      "step": 12185
+    },
+    {
+      "epoch": 0.2438,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018766427939777945,
+      "loss": 1.2879,
+      "step": 12190
+    },
+    {
+      "epoch": 0.2439,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018764747904087263,
+      "loss": 1.2624,
+      "step": 12195
+    },
+    {
+      "epoch": 0.244,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018763066800438636,
+      "loss": 1.2753,
+      "step": 12200
+    },
+    {
+      "epoch": 0.2441,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00018761384629036902,
+      "loss": 1.2641,
+      "step": 12205
+    },
+    {
+      "epoch": 0.2442,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018759701390087027,
+      "loss": 1.2704,
+      "step": 12210
+    },
+    {
+      "epoch": 0.2443,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001875801708379411,
+      "loss": 1.2701,
+      "step": 12215
+    },
+    {
+      "epoch": 0.2444,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00018756331710363374,
+      "loss": 1.2799,
+      "step": 12220
+    },
+    {
+      "epoch": 0.2445,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001875464527000018,
+      "loss": 1.2794,
+      "step": 12225
+    },
+    {
+      "epoch": 0.2446,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018752957762910018,
+      "loss": 1.2896,
+      "step": 12230
+    },
+    {
+      "epoch": 0.2447,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.000187512691892985,
+      "loss": 1.2782,
+      "step": 12235
+    },
+    {
+      "epoch": 0.2448,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.0001874957954937138,
+      "loss": 1.2632,
+      "step": 12240
+    },
+    {
+      "epoch": 0.2449,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001874788884333453,
+      "loss": 1.268,
+      "step": 12245
+    },
+    {
+      "epoch": 0.245,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018746197071393958,
+      "loss": 1.2822,
+      "step": 12250
+    },
+    {
+      "epoch": 0.2451,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00018744504233755805,
+      "loss": 1.2765,
+      "step": 12255
+    },
+    {
+      "epoch": 0.2452,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018742810330626337,
+      "loss": 1.2736,
+      "step": 12260
+    },
+    {
+      "epoch": 0.2453,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00018741115362211949,
+      "loss": 1.274,
+      "step": 12265
+    },
+    {
+      "epoch": 0.2454,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001873941932871917,
+      "loss": 1.2753,
+      "step": 12270
+    },
+    {
+      "epoch": 0.2455,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018737722230354655,
+      "loss": 1.2646,
+      "step": 12275
+    },
+    {
+      "epoch": 0.2456,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00018736024067325188,
+      "loss": 1.2579,
+      "step": 12280
+    },
+    {
+      "epoch": 0.2457,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.0001873432483983769,
+      "loss": 1.2682,
+      "step": 12285
+    },
+    {
+      "epoch": 0.2458,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00018732624548099204,
+      "loss": 1.2939,
+      "step": 12290
+    },
+    {
+      "epoch": 0.2459,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018730923192316902,
+      "loss": 1.2911,
+      "step": 12295
+    },
+    {
+      "epoch": 0.246,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00018729220772698097,
+      "loss": 1.2863,
+      "step": 12300
+    },
+    {
+      "epoch": 0.2461,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001872751728945022,
+      "loss": 1.2921,
+      "step": 12305
+    },
+    {
+      "epoch": 0.2462,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018725812742780834,
+      "loss": 1.2835,
+      "step": 12310
+    },
+    {
+      "epoch": 0.2463,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001872410713289763,
+      "loss": 1.2811,
+      "step": 12315
+    },
+    {
+      "epoch": 0.2464,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001872240046000844,
+      "loss": 1.2756,
+      "step": 12320
+    },
+    {
+      "epoch": 0.2465,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018720692724321207,
+      "loss": 1.2745,
+      "step": 12325
+    },
+    {
+      "epoch": 0.2466,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0001871898392604402,
+      "loss": 1.2759,
+      "step": 12330
+    },
+    {
+      "epoch": 0.2467,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001871727406538509,
+      "loss": 1.2753,
+      "step": 12335
+    },
+    {
+      "epoch": 0.2468,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00018715563142552758,
+      "loss": 1.285,
+      "step": 12340
+    },
+    {
+      "epoch": 0.2469,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00018713851157755492,
+      "loss": 1.2755,
+      "step": 12345
+    },
+    {
+      "epoch": 0.247,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00018712138111201895,
+      "loss": 1.2789,
+      "step": 12350
+    },
+    {
+      "epoch": 0.2471,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00018710424003100698,
+      "loss": 1.299,
+      "step": 12355
+    },
+    {
+      "epoch": 0.2472,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00018708708833660754,
+      "loss": 1.2641,
+      "step": 12360
+    },
+    {
+      "epoch": 0.2473,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00018706992603091058,
+      "loss": 1.2712,
+      "step": 12365
+    },
+    {
+      "epoch": 0.2474,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00018705275311600722,
+      "loss": 1.2687,
+      "step": 12370
+    },
+    {
+      "epoch": 0.2475,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018703556959398998,
+      "loss": 1.2612,
+      "step": 12375
+    },
+    {
+      "epoch": 0.2476,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0001870183754669526,
+      "loss": 1.2893,
+      "step": 12380
+    },
+    {
+      "epoch": 0.2477,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001870011707369901,
+      "loss": 1.2786,
+      "step": 12385
+    },
+    {
+      "epoch": 0.2478,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.0001869839554061988,
+      "loss": 1.2564,
+      "step": 12390
+    },
+    {
+      "epoch": 0.2479,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00018696672947667646,
+      "loss": 1.281,
+      "step": 12395
+    },
+    {
+      "epoch": 0.248,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0001869494929505219,
+      "loss": 1.2625,
+      "step": 12400
+    },
+    {
+      "epoch": 0.2481,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0001869322458298354,
+      "loss": 1.289,
+      "step": 12405
+    },
+    {
+      "epoch": 0.2482,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001869149881167184,
+      "loss": 1.252,
+      "step": 12410
+    },
+    {
+      "epoch": 0.2483,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00018689771981327376,
+      "loss": 1.2577,
+      "step": 12415
+    },
+    {
+      "epoch": 0.2484,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00018688044092160551,
+      "loss": 1.2897,
+      "step": 12420
+    },
+    {
+      "epoch": 0.2485,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00018686315144381913,
+      "loss": 1.2854,
+      "step": 12425
+    },
+    {
+      "epoch": 0.2486,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018684585138202122,
+      "loss": 1.2699,
+      "step": 12430
+    },
+    {
+      "epoch": 0.2487,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00018682854073831973,
+      "loss": 1.2681,
+      "step": 12435
+    },
+    {
+      "epoch": 0.2488,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00018681121951482393,
+      "loss": 1.277,
+      "step": 12440
+    },
+    {
+      "epoch": 0.2489,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00018679388771364436,
+      "loss": 1.2783,
+      "step": 12445
+    },
+    {
+      "epoch": 0.249,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00018677654533689287,
+      "loss": 1.2533,
+      "step": 12450
+    },
+    {
+      "epoch": 0.2491,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.0001867591923866825,
+      "loss": 1.2697,
+      "step": 12455
+    },
+    {
+      "epoch": 0.2492,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00018674182886512774,
+      "loss": 1.2894,
+      "step": 12460
+    },
+    {
+      "epoch": 0.2493,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018672445477434425,
+      "loss": 1.2617,
+      "step": 12465
+    },
+    {
+      "epoch": 0.2494,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.000186707070116449,
+      "loss": 1.2966,
+      "step": 12470
+    },
+    {
+      "epoch": 0.2495,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00018668967489356028,
+      "loss": 1.273,
+      "step": 12475
+    },
+    {
+      "epoch": 0.2496,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018667226910779765,
+      "loss": 1.2828,
+      "step": 12480
+    },
+    {
+      "epoch": 0.2497,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018665485276128188,
+      "loss": 1.2778,
+      "step": 12485
+    },
+    {
+      "epoch": 0.2498,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018663742585613518,
+      "loss": 1.2737,
+      "step": 12490
+    },
+    {
+      "epoch": 0.2499,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00018661998839448094,
+      "loss": 1.2674,
+      "step": 12495
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00018660254037844388,
+      "loss": 1.2801,
+      "step": 12500
+    },
+    {
+      "epoch": 0.2501,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00018658508181014995,
+      "loss": 1.2787,
+      "step": 12505
+    },
+    {
+      "epoch": 0.2502,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00018656761269172643,
+      "loss": 1.2594,
+      "step": 12510
+    },
+    {
+      "epoch": 0.2503,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.0001865501330253019,
+      "loss": 1.2824,
+      "step": 12515
+    },
+    {
+      "epoch": 0.2504,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00018653264281300622,
+      "loss": 1.281,
+      "step": 12520
+    },
+    {
+      "epoch": 0.2505,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00018651514205697046,
+      "loss": 1.2638,
+      "step": 12525
+    },
+    {
+      "epoch": 0.2506,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00018649763075932708,
+      "loss": 1.258,
+      "step": 12530
+    },
+    {
+      "epoch": 0.2507,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00018648010892220978,
+      "loss": 1.2667,
+      "step": 12535
+    },
+    {
+      "epoch": 0.2508,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.0001864625765477535,
+      "loss": 1.2724,
+      "step": 12540
+    },
+    {
+      "epoch": 0.2509,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00018644503363809457,
+      "loss": 1.2807,
+      "step": 12545
+    },
+    {
+      "epoch": 0.251,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001864274801953705,
+      "loss": 1.2781,
+      "step": 12550
+    },
+    {
+      "epoch": 0.2511,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001864099162217201,
+      "loss": 1.2774,
+      "step": 12555
+    },
+    {
+      "epoch": 0.2512,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00018639234171928353,
+      "loss": 1.2861,
+      "step": 12560
+    },
+    {
+      "epoch": 0.2513,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001863747566902022,
+      "loss": 1.2832,
+      "step": 12565
+    },
+    {
+      "epoch": 0.2514,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00018635716113661873,
+      "loss": 1.2881,
+      "step": 12570
+    },
+    {
+      "epoch": 0.2515,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018633955506067718,
+      "loss": 1.2824,
+      "step": 12575
+    },
+    {
+      "epoch": 0.2516,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0001863219384645227,
+      "loss": 1.2781,
+      "step": 12580
+    },
+    {
+      "epoch": 0.2517,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001863043113503019,
+      "loss": 1.2845,
+      "step": 12585
+    },
+    {
+      "epoch": 0.2518,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001862866737201625,
+      "loss": 1.2684,
+      "step": 12590
+    },
+    {
+      "epoch": 0.2519,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00018626902557625368,
+      "loss": 1.2809,
+      "step": 12595
+    },
+    {
+      "epoch": 0.252,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018625136692072575,
+      "loss": 1.276,
+      "step": 12600
+    },
+    {
+      "epoch": 0.2521,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0001862336977557304,
+      "loss": 1.2675,
+      "step": 12605
+    },
+    {
+      "epoch": 0.2522,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018621601808342056,
+      "loss": 1.2642,
+      "step": 12610
+    },
+    {
+      "epoch": 0.2523,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00018619832790595043,
+      "loss": 1.2785,
+      "step": 12615
+    },
+    {
+      "epoch": 0.2524,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001861806272254755,
+      "loss": 1.2716,
+      "step": 12620
+    },
+    {
+      "epoch": 0.2525,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00018616291604415258,
+      "loss": 1.2612,
+      "step": 12625
+    },
+    {
+      "epoch": 0.2526,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001861451943641397,
+      "loss": 1.2716,
+      "step": 12630
+    },
+    {
+      "epoch": 0.2527,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00018612746218759618,
+      "loss": 1.275,
+      "step": 12635
+    },
+    {
+      "epoch": 0.2528,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018610971951668265,
+      "loss": 1.2858,
+      "step": 12640
+    },
+    {
+      "epoch": 0.2529,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.000186091966353561,
+      "loss": 1.286,
+      "step": 12645
+    },
+    {
+      "epoch": 0.253,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0001860742027003944,
+      "loss": 1.2736,
+      "step": 12650
+    },
+    {
+      "epoch": 0.2531,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00018605642855934725,
+      "loss": 1.2848,
+      "step": 12655
+    },
+    {
+      "epoch": 0.2532,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00018603864393258534,
+      "loss": 1.28,
+      "step": 12660
+    },
+    {
+      "epoch": 0.2533,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00018602084882227566,
+      "loss": 1.2792,
+      "step": 12665
+    },
+    {
+      "epoch": 0.2534,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00018600304323058647,
+      "loss": 1.2929,
+      "step": 12670
+    },
+    {
+      "epoch": 0.2535,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00018598522715968736,
+      "loss": 1.2779,
+      "step": 12675
+    },
+    {
+      "epoch": 0.2536,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001859674006117491,
+      "loss": 1.2794,
+      "step": 12680
+    },
+    {
+      "epoch": 0.2537,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018594956358894388,
+      "loss": 1.2769,
+      "step": 12685
+    },
+    {
+      "epoch": 0.2538,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018593171609344503,
+      "loss": 1.2546,
+      "step": 12690
+    },
+    {
+      "epoch": 0.2539,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00018591385812742725,
+      "loss": 1.2764,
+      "step": 12695
+    },
+    {
+      "epoch": 0.254,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018589598969306645,
+      "loss": 1.2779,
+      "step": 12700
+    },
+    {
+      "epoch": 0.2541,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00018587811079253985,
+      "loss": 1.2612,
+      "step": 12705
+    },
+    {
+      "epoch": 0.2542,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00018586022142802597,
+      "loss": 1.263,
+      "step": 12710
+    },
+    {
+      "epoch": 0.2543,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00018584232160170452,
+      "loss": 1.268,
+      "step": 12715
+    },
+    {
+      "epoch": 0.2544,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001858244113157566,
+      "loss": 1.2649,
+      "step": 12720
+    },
+    {
+      "epoch": 0.2545,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00018580649057236447,
+      "loss": 1.2716,
+      "step": 12725
+    },
+    {
+      "epoch": 0.2546,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00018578855937371173,
+      "loss": 1.2752,
+      "step": 12730
+    },
+    {
+      "epoch": 0.2547,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001857706177219833,
+      "loss": 1.2771,
+      "step": 12735
+    },
+    {
+      "epoch": 0.2548,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018575266561936523,
+      "loss": 1.2932,
+      "step": 12740
+    },
+    {
+      "epoch": 0.2549,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00018573470306804498,
+      "loss": 1.2724,
+      "step": 12745
+    },
+    {
+      "epoch": 0.255,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018571673007021123,
+      "loss": 1.2665,
+      "step": 12750
+    },
+    {
+      "epoch": 0.2551,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00018569874662805393,
+      "loss": 1.2799,
+      "step": 12755
+    },
+    {
+      "epoch": 0.2552,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0001856807527437643,
+      "loss": 1.2684,
+      "step": 12760
+    },
+    {
+      "epoch": 0.2553,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00018566274841953483,
+      "loss": 1.2686,
+      "step": 12765
+    },
+    {
+      "epoch": 0.2554,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00018564473365755935,
+      "loss": 1.287,
+      "step": 12770
+    },
+    {
+      "epoch": 0.2555,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00018562670846003284,
+      "loss": 1.2597,
+      "step": 12775
+    },
+    {
+      "epoch": 0.2556,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.0001856086728291516,
+      "loss": 1.2711,
+      "step": 12780
+    },
+    {
+      "epoch": 0.2557,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00018559062676711332,
+      "loss": 1.2701,
+      "step": 12785
+    },
+    {
+      "epoch": 0.2558,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018557257027611675,
+      "loss": 1.2636,
+      "step": 12790
+    },
+    {
+      "epoch": 0.2559,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00018555450335836206,
+      "loss": 1.2654,
+      "step": 12795
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00018553642601605068,
+      "loss": 1.2674,
+      "step": 12800
+    },
+    {
+      "epoch": 0.2561,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001855183382513852,
+      "loss": 1.282,
+      "step": 12805
+    },
+    {
+      "epoch": 0.2562,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018550024006656966,
+      "loss": 1.2663,
+      "step": 12810
+    },
+    {
+      "epoch": 0.2563,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018548213146380918,
+      "loss": 1.2686,
+      "step": 12815
+    },
+    {
+      "epoch": 0.2564,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001854640124453103,
+      "loss": 1.2783,
+      "step": 12820
+    },
+    {
+      "epoch": 0.2565,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00018544588301328075,
+      "loss": 1.2638,
+      "step": 12825
+    },
+    {
+      "epoch": 0.2566,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0001854277431699295,
+      "loss": 1.2643,
+      "step": 12830
+    },
+    {
+      "epoch": 0.2567,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00018540959291746693,
+      "loss": 1.2695,
+      "step": 12835
+    },
+    {
+      "epoch": 0.2568,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001853914322581045,
+      "loss": 1.2796,
+      "step": 12840
+    },
+    {
+      "epoch": 0.2569,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00018537326119405506,
+      "loss": 1.2923,
+      "step": 12845
+    },
+    {
+      "epoch": 0.257,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00018535507972753274,
+      "loss": 1.2725,
+      "step": 12850
+    },
+    {
+      "epoch": 0.2571,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00018533688786075288,
+      "loss": 1.2767,
+      "step": 12855
+    },
+    {
+      "epoch": 0.2572,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00018531868559593204,
+      "loss": 1.2676,
+      "step": 12860
+    },
+    {
+      "epoch": 0.2573,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00018530047293528819,
+      "loss": 1.2613,
+      "step": 12865
+    },
+    {
+      "epoch": 0.2574,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00018528224988104044,
+      "loss": 1.264,
+      "step": 12870
+    },
+    {
+      "epoch": 0.2575,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018526401643540922,
+      "loss": 1.2949,
+      "step": 12875
+    },
+    {
+      "epoch": 0.2576,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018524577260061627,
+      "loss": 1.2686,
+      "step": 12880
+    },
+    {
+      "epoch": 0.2577,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001852275183788845,
+      "loss": 1.2974,
+      "step": 12885
+    },
+    {
+      "epoch": 0.2578,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.0001852092537724381,
+      "loss": 1.2753,
+      "step": 12890
+    },
+    {
+      "epoch": 0.2579,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00018519097878350263,
+      "loss": 1.284,
+      "step": 12895
+    },
+    {
+      "epoch": 0.258,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00018517269341430476,
+      "loss": 1.2756,
+      "step": 12900
+    },
+    {
+      "epoch": 0.2581,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018515439766707262,
+      "loss": 1.2707,
+      "step": 12905
+    },
+    {
+      "epoch": 0.2582,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00018513609154403534,
+      "loss": 1.2681,
+      "step": 12910
+    },
+    {
+      "epoch": 0.2583,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00018511777504742362,
+      "loss": 1.267,
+      "step": 12915
+    },
+    {
+      "epoch": 0.2584,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00018509944817946922,
+      "loss": 1.2856,
+      "step": 12920
+    },
+    {
+      "epoch": 0.2585,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018508111094240514,
+      "loss": 1.2785,
+      "step": 12925
+    },
+    {
+      "epoch": 0.2586,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018506276333846579,
+      "loss": 1.2767,
+      "step": 12930
+    },
+    {
+      "epoch": 0.2587,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00018504440536988673,
+      "loss": 1.2964,
+      "step": 12935
+    },
+    {
+      "epoch": 0.2588,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00018502603703890488,
+      "loss": 1.2674,
+      "step": 12940
+    },
+    {
+      "epoch": 0.2589,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018500765834775828,
+      "loss": 1.2899,
+      "step": 12945
+    },
+    {
+      "epoch": 0.259,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00018498926929868642,
+      "loss": 1.2753,
+      "step": 12950
+    },
+    {
+      "epoch": 0.2591,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00018497086989392988,
+      "loss": 1.2775,
+      "step": 12955
+    },
+    {
+      "epoch": 0.2592,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00018495246013573054,
+      "loss": 1.2784,
+      "step": 12960
+    },
+    {
+      "epoch": 0.2593,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00018493404002633166,
+      "loss": 1.2678,
+      "step": 12965
+    },
+    {
+      "epoch": 0.2594,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00018491560956797765,
+      "loss": 1.2702,
+      "step": 12970
+    },
+    {
+      "epoch": 0.2595,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00018489716876291415,
+      "loss": 1.281,
+      "step": 12975
+    },
+    {
+      "epoch": 0.2596,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001848787176133882,
+      "loss": 1.272,
+      "step": 12980
+    },
+    {
+      "epoch": 0.2597,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018486025612164794,
+      "loss": 1.2668,
+      "step": 12985
+    },
+    {
+      "epoch": 0.2598,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0001848417842899429,
+      "loss": 1.2885,
+      "step": 12990
+    },
+    {
+      "epoch": 0.2599,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00018482330212052378,
+      "loss": 1.2808,
+      "step": 12995
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001848048096156426,
+      "loss": 1.2967,
+      "step": 13000
+    },
+    {
+      "epoch": 0.2601,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00018478630677755262,
+      "loss": 1.2785,
+      "step": 13005
+    },
+    {
+      "epoch": 0.2602,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00018476779360850832,
+      "loss": 1.29,
+      "step": 13010
+    },
+    {
+      "epoch": 0.2603,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018474927011076552,
+      "loss": 1.2818,
+      "step": 13015
+    },
+    {
+      "epoch": 0.2604,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001847307362865812,
+      "loss": 1.2679,
+      "step": 13020
+    },
+    {
+      "epoch": 0.2605,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018471219213821375,
+      "loss": 1.2741,
+      "step": 13025
+    },
+    {
+      "epoch": 0.2606,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018469363766792255,
+      "loss": 1.2903,
+      "step": 13030
+    },
+    {
+      "epoch": 0.2607,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00018467507287796856,
+      "loss": 1.2743,
+      "step": 13035
+    },
+    {
+      "epoch": 0.2608,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0001846564977706138,
+      "loss": 1.2841,
+      "step": 13040
+    },
+    {
+      "epoch": 0.2609,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018463791234812153,
+      "loss": 1.2797,
+      "step": 13045
+    },
+    {
+      "epoch": 0.261,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018461931661275643,
+      "loss": 1.2794,
+      "step": 13050
+    },
+    {
+      "epoch": 0.2611,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00018460071056678422,
+      "loss": 1.2775,
+      "step": 13055
+    },
+    {
+      "epoch": 0.2612,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00018458209421247208,
+      "loss": 1.2611,
+      "step": 13060
+    },
+    {
+      "epoch": 0.2613,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018456346755208833,
+      "loss": 1.2745,
+      "step": 13065
+    },
+    {
+      "epoch": 0.2614,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00018454483058790255,
+      "loss": 1.2656,
+      "step": 13070
+    },
+    {
+      "epoch": 0.2615,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018452618332218563,
+      "loss": 1.2752,
+      "step": 13075
+    },
+    {
+      "epoch": 0.2616,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00018450752575720967,
+      "loss": 1.274,
+      "step": 13080
+    },
+    {
+      "epoch": 0.2617,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00018448885789524802,
+      "loss": 1.2622,
+      "step": 13085
+    },
+    {
+      "epoch": 0.2618,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018447017973857532,
+      "loss": 1.2745,
+      "step": 13090
+    },
+    {
+      "epoch": 0.2619,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00018445149128946744,
+      "loss": 1.2579,
+      "step": 13095
+    },
+    {
+      "epoch": 0.262,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00018443279255020152,
+      "loss": 1.2727,
+      "step": 13100
+    },
+    {
+      "epoch": 0.2621,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00018441408352305594,
+      "loss": 1.2681,
+      "step": 13105
+    },
+    {
+      "epoch": 0.2622,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00018439536421031033,
+      "loss": 1.2662,
+      "step": 13110
+    },
+    {
+      "epoch": 0.2623,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001843766346142456,
+      "loss": 1.2694,
+      "step": 13115
+    },
+    {
+      "epoch": 0.2624,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001843578947371439,
+      "loss": 1.2643,
+      "step": 13120
+    },
+    {
+      "epoch": 0.2625,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001843391445812886,
+      "loss": 1.2874,
+      "step": 13125
+    },
+    {
+      "epoch": 0.2626,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018432038414896434,
+      "loss": 1.2828,
+      "step": 13130
+    },
+    {
+      "epoch": 0.2627,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00018430161344245707,
+      "loss": 1.2924,
+      "step": 13135
+    },
+    {
+      "epoch": 0.2628,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001842828324640539,
+      "loss": 1.2783,
+      "step": 13140
+    },
+    {
+      "epoch": 0.2629,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00018426404121604323,
+      "loss": 1.266,
+      "step": 13145
+    },
+    {
+      "epoch": 0.263,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00018424523970071477,
+      "loss": 1.2752,
+      "step": 13150
+    },
+    {
+      "epoch": 0.2631,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0001842264279203594,
+      "loss": 1.2672,
+      "step": 13155
+    },
+    {
+      "epoch": 0.2632,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00018420760587726923,
+      "loss": 1.2743,
+      "step": 13160
+    },
+    {
+      "epoch": 0.2633,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00018418877357373776,
+      "loss": 1.2835,
+      "step": 13165
+    },
+    {
+      "epoch": 0.2634,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018416993101205958,
+      "loss": 1.2765,
+      "step": 13170
+    },
+    {
+      "epoch": 0.2635,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018415107819453062,
+      "loss": 1.2819,
+      "step": 13175
+    },
+    {
+      "epoch": 0.2636,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00018413221512344805,
+      "loss": 1.2822,
+      "step": 13180
+    },
+    {
+      "epoch": 0.2637,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00018411334180111027,
+      "loss": 1.296,
+      "step": 13185
+    },
+    {
+      "epoch": 0.2638,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018409445822981693,
+      "loss": 1.2825,
+      "step": 13190
+    },
+    {
+      "epoch": 0.2639,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018407556441186893,
+      "loss": 1.2786,
+      "step": 13195
+    },
+    {
+      "epoch": 0.264,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00018405666034956844,
+      "loss": 1.3017,
+      "step": 13200
+    },
+    {
+      "epoch": 0.2641,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00018403774604521886,
+      "loss": 1.2748,
+      "step": 13205
+    },
+    {
+      "epoch": 0.2642,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00018401882150112484,
+      "loss": 1.2781,
+      "step": 13210
+    },
+    {
+      "epoch": 0.2643,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018399988671959227,
+      "loss": 1.2836,
+      "step": 13215
+    },
+    {
+      "epoch": 0.2644,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001839809417029283,
+      "loss": 1.2759,
+      "step": 13220
+    },
+    {
+      "epoch": 0.2645,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018396198645344135,
+      "loss": 1.2633,
+      "step": 13225
+    },
+    {
+      "epoch": 0.2646,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.000183943020973441,
+      "loss": 1.2725,
+      "step": 13230
+    },
+    {
+      "epoch": 0.2647,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00018392404526523817,
+      "loss": 1.2739,
+      "step": 13235
+    },
+    {
+      "epoch": 0.2648,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.000183905059331145,
+      "loss": 1.2741,
+      "step": 13240
+    },
+    {
+      "epoch": 0.2649,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001838860631734749,
+      "loss": 1.2945,
+      "step": 13245
+    },
+    {
+      "epoch": 0.265,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018386705679454242,
+      "loss": 1.2704,
+      "step": 13250
+    },
+    {
+      "epoch": 0.2651,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00018384804019666345,
+      "loss": 1.2592,
+      "step": 13255
+    },
+    {
+      "epoch": 0.2652,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00018382901338215516,
+      "loss": 1.2715,
+      "step": 13260
+    },
+    {
+      "epoch": 0.2653,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00018380997635333585,
+      "loss": 1.253,
+      "step": 13265
+    },
+    {
+      "epoch": 0.2654,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00018379092911252514,
+      "loss": 1.2628,
+      "step": 13270
+    },
+    {
+      "epoch": 0.2655,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001837718716620439,
+      "loss": 1.278,
+      "step": 13275
+    },
+    {
+      "epoch": 0.2656,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001837528040042142,
+      "loss": 1.2744,
+      "step": 13280
+    },
+    {
+      "epoch": 0.2657,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018373372614135936,
+      "loss": 1.2796,
+      "step": 13285
+    },
+    {
+      "epoch": 0.2658,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.000183714638075804,
+      "loss": 1.2809,
+      "step": 13290
+    },
+    {
+      "epoch": 0.2659,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001836955398098739,
+      "loss": 1.2895,
+      "step": 13295
+    },
+    {
+      "epoch": 0.266,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00018367643134589617,
+      "loss": 1.2515,
+      "step": 13300
+    },
+    {
+      "epoch": 0.2661,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0001836573126861991,
+      "loss": 1.2863,
+      "step": 13305
+    },
+    {
+      "epoch": 0.2662,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00018363818383311225,
+      "loss": 1.274,
+      "step": 13310
+    },
+    {
+      "epoch": 0.2663,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001836190447889664,
+      "loss": 1.3,
+      "step": 13315
+    },
+    {
+      "epoch": 0.2664,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00018359989555609353,
+      "loss": 1.2724,
+      "step": 13320
+    },
+    {
+      "epoch": 0.2665,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018358073613682706,
+      "loss": 1.2828,
+      "step": 13325
+    },
+    {
+      "epoch": 0.2666,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00018356156653350137,
+      "loss": 1.2689,
+      "step": 13330
+    },
+    {
+      "epoch": 0.2667,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00018354238674845225,
+      "loss": 1.2766,
+      "step": 13335
+    },
+    {
+      "epoch": 0.2668,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00018352319678401676,
+      "loss": 1.283,
+      "step": 13340
+    },
+    {
+      "epoch": 0.2669,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00018350399664253305,
+      "loss": 1.264,
+      "step": 13345
+    },
+    {
+      "epoch": 0.267,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018348478632634066,
+      "loss": 1.2978,
+      "step": 13350
+    },
+    {
+      "epoch": 0.2671,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001834655658377803,
+      "loss": 1.2859,
+      "step": 13355
+    },
+    {
+      "epoch": 0.2672,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00018344633517919392,
+      "loss": 1.2876,
+      "step": 13360
+    },
+    {
+      "epoch": 0.2673,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00018342709435292473,
+      "loss": 1.2845,
+      "step": 13365
+    },
+    {
+      "epoch": 0.2674,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00018340784336131713,
+      "loss": 1.2848,
+      "step": 13370
+    },
+    {
+      "epoch": 0.2675,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018338858220671682,
+      "loss": 1.291,
+      "step": 13375
+    },
+    {
+      "epoch": 0.2676,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00018336931089147073,
+      "loss": 1.2883,
+      "step": 13380
+    },
+    {
+      "epoch": 0.2677,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00018335002941792698,
+      "loss": 1.269,
+      "step": 13385
+    },
+    {
+      "epoch": 0.2678,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.000183330737788435,
+      "loss": 1.2764,
+      "step": 13390
+    },
+    {
+      "epoch": 0.2679,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00018331143600534535,
+      "loss": 1.2983,
+      "step": 13395
+    },
+    {
+      "epoch": 0.268,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00018329212407100994,
+      "loss": 1.2701,
+      "step": 13400
+    },
+    {
+      "epoch": 0.2681,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001832728019877819,
+      "loss": 1.298,
+      "step": 13405
+    },
+    {
+      "epoch": 0.2682,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001832534697580155,
+      "loss": 1.2702,
+      "step": 13410
+    },
+    {
+      "epoch": 0.2683,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018323412738406635,
+      "loss": 1.2672,
+      "step": 13415
+    },
+    {
+      "epoch": 0.2684,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018321477486829126,
+      "loss": 1.2811,
+      "step": 13420
+    },
+    {
+      "epoch": 0.2685,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018319541221304827,
+      "loss": 1.2758,
+      "step": 13425
+    },
+    {
+      "epoch": 0.2686,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00018317603942069664,
+      "loss": 1.3042,
+      "step": 13430
+    },
+    {
+      "epoch": 0.2687,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00018315665649359692,
+      "loss": 1.271,
+      "step": 13435
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00018313726343411086,
+      "loss": 1.2752,
+      "step": 13440
+    },
+    {
+      "epoch": 0.2689,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001831178602446014,
+      "loss": 1.2574,
+      "step": 13445
+    },
+    {
+      "epoch": 0.269,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018309844692743283,
+      "loss": 1.2681,
+      "step": 13450
+    },
+    {
+      "epoch": 0.2691,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00018307902348497056,
+      "loss": 1.308,
+      "step": 13455
+    },
+    {
+      "epoch": 0.2692,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00018305958991958127,
+      "loss": 1.2647,
+      "step": 13460
+    },
+    {
+      "epoch": 0.2693,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0001830401462336329,
+      "loss": 1.2867,
+      "step": 13465
+    },
+    {
+      "epoch": 0.2694,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001830206924294946,
+      "loss": 1.2779,
+      "step": 13470
+    },
+    {
+      "epoch": 0.2695,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00018300122850953675,
+      "loss": 1.2949,
+      "step": 13475
+    },
+    {
+      "epoch": 0.2696,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00018298175447613096,
+      "loss": 1.2824,
+      "step": 13480
+    },
+    {
+      "epoch": 0.2697,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018296227033165013,
+      "loss": 1.2675,
+      "step": 13485
+    },
+    {
+      "epoch": 0.2698,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00018294277607846832,
+      "loss": 1.2715,
+      "step": 13490
+    },
+    {
+      "epoch": 0.2699,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0001829232717189608,
+      "loss": 1.2725,
+      "step": 13495
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00018290375725550417,
+      "loss": 1.2773,
+      "step": 13500
+    },
+    {
+      "epoch": 0.2701,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001828842326904762,
+      "loss": 1.2911,
+      "step": 13505
+    },
+    {
+      "epoch": 0.2702,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018286469802625589,
+      "loss": 1.2759,
+      "step": 13510
+    },
+    {
+      "epoch": 0.2703,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018284515326522346,
+      "loss": 1.2665,
+      "step": 13515
+    },
+    {
+      "epoch": 0.2704,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018282559840976042,
+      "loss": 1.2686,
+      "step": 13520
+    },
+    {
+      "epoch": 0.2705,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018280603346224945,
+      "loss": 1.2836,
+      "step": 13525
+    },
+    {
+      "epoch": 0.2706,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00018278645842507448,
+      "loss": 1.2775,
+      "step": 13530
+    },
+    {
+      "epoch": 0.2707,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00018276687330062065,
+      "loss": 1.2708,
+      "step": 13535
+    },
+    {
+      "epoch": 0.2708,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00018274727809127438,
+      "loss": 1.2926,
+      "step": 13540
+    },
+    {
+      "epoch": 0.2709,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00018272767279942328,
+      "loss": 1.2685,
+      "step": 13545
+    },
+    {
+      "epoch": 0.271,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00018270805742745617,
+      "loss": 1.2784,
+      "step": 13550
+    },
+    {
+      "epoch": 0.2711,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00018268843197776318,
+      "loss": 1.2682,
+      "step": 13555
+    },
+    {
+      "epoch": 0.2712,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00018266879645273556,
+      "loss": 1.2762,
+      "step": 13560
+    },
+    {
+      "epoch": 0.2713,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00018264915085476583,
+      "loss": 1.278,
+      "step": 13565
+    },
+    {
+      "epoch": 0.2714,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001826294951862478,
+      "loss": 1.2772,
+      "step": 13570
+    },
+    {
+      "epoch": 0.2715,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00018260982944957638,
+      "loss": 1.2689,
+      "step": 13575
+    },
+    {
+      "epoch": 0.2716,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018259015364714787,
+      "loss": 1.2821,
+      "step": 13580
+    },
+    {
+      "epoch": 0.2717,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00018257046778135964,
+      "loss": 1.2792,
+      "step": 13585
+    },
+    {
+      "epoch": 0.2718,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00018255077185461038,
+      "loss": 1.2746,
+      "step": 13590
+    },
+    {
+      "epoch": 0.2719,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00018253106586929997,
+      "loss": 1.2634,
+      "step": 13595
+    },
+    {
+      "epoch": 0.272,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00018251134982782952,
+      "loss": 1.2933,
+      "step": 13600
+    },
+    {
+      "epoch": 0.2721,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018249162373260141,
+      "loss": 1.2783,
+      "step": 13605
+    },
+    {
+      "epoch": 0.2722,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001824718875860191,
+      "loss": 1.274,
+      "step": 13610
+    },
+    {
+      "epoch": 0.2723,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00018245214139048753,
+      "loss": 1.2879,
+      "step": 13615
+    },
+    {
+      "epoch": 0.2724,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001824323851484126,
+      "loss": 1.284,
+      "step": 13620
+    },
+    {
+      "epoch": 0.2725,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00018241261886220154,
+      "loss": 1.2871,
+      "step": 13625
+    },
+    {
+      "epoch": 0.2726,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018239284253426295,
+      "loss": 1.2646,
+      "step": 13630
+    },
+    {
+      "epoch": 0.2727,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00018237305616700637,
+      "loss": 1.2696,
+      "step": 13635
+    },
+    {
+      "epoch": 0.2728,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018235325976284275,
+      "loss": 1.279,
+      "step": 13640
+    },
+    {
+      "epoch": 0.2729,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018233345332418423,
+      "loss": 1.2715,
+      "step": 13645
+    },
+    {
+      "epoch": 0.273,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001823136368534442,
+      "loss": 1.2723,
+      "step": 13650
+    },
+    {
+      "epoch": 0.2731,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00018229381035303718,
+      "loss": 1.279,
+      "step": 13655
+    },
+    {
+      "epoch": 0.2732,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.000182273973825379,
+      "loss": 1.2725,
+      "step": 13660
+    },
+    {
+      "epoch": 0.2733,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00018225412727288667,
+      "loss": 1.258,
+      "step": 13665
+    },
+    {
+      "epoch": 0.2734,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00018223427069797844,
+      "loss": 1.2733,
+      "step": 13670
+    },
+    {
+      "epoch": 0.2735,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018221440410307374,
+      "loss": 1.2866,
+      "step": 13675
+    },
+    {
+      "epoch": 0.2736,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001821945274905933,
+      "loss": 1.2649,
+      "step": 13680
+    },
+    {
+      "epoch": 0.2737,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00018217464086295904,
+      "loss": 1.2802,
+      "step": 13685
+    },
+    {
+      "epoch": 0.2738,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018215474422259402,
+      "loss": 1.2699,
+      "step": 13690
+    },
+    {
+      "epoch": 0.2739,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018213483757192263,
+      "loss": 1.2794,
+      "step": 13695
+    },
+    {
+      "epoch": 0.274,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00018211492091337042,
+      "loss": 1.2778,
+      "step": 13700
+    },
+    {
+      "epoch": 0.2741,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00018209499424936415,
+      "loss": 1.2748,
+      "step": 13705
+    },
+    {
+      "epoch": 0.2742,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001820750575823319,
+      "loss": 1.3,
+      "step": 13710
+    },
+    {
+      "epoch": 0.2743,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018205511091470283,
+      "loss": 1.268,
+      "step": 13715
+    },
+    {
+      "epoch": 0.2744,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001820351542489074,
+      "loss": 1.2808,
+      "step": 13720
+    },
+    {
+      "epoch": 0.2745,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018201518758737724,
+      "loss": 1.2942,
+      "step": 13725
+    },
+    {
+      "epoch": 0.2746,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00018199521093254523,
+      "loss": 1.2941,
+      "step": 13730
+    },
+    {
+      "epoch": 0.2747,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00018197522428684552,
+      "loss": 1.2586,
+      "step": 13735
+    },
+    {
+      "epoch": 0.2748,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001819552276527134,
+      "loss": 1.2722,
+      "step": 13740
+    },
+    {
+      "epoch": 0.2749,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018193522103258537,
+      "loss": 1.2761,
+      "step": 13745
+    },
+    {
+      "epoch": 0.275,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001819152044288992,
+      "loss": 1.2729,
+      "step": 13750
+    },
+    {
+      "epoch": 0.2751,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00018189517784409381,
+      "loss": 1.2726,
+      "step": 13755
+    },
+    {
+      "epoch": 0.2752,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00018187514128060946,
+      "loss": 1.2928,
+      "step": 13760
+    },
+    {
+      "epoch": 0.2753,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001818550947408875,
+      "loss": 1.2968,
+      "step": 13765
+    },
+    {
+      "epoch": 0.2754,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001818350382273705,
+      "loss": 1.2725,
+      "step": 13770
+    },
+    {
+      "epoch": 0.2755,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018181497174250236,
+      "loss": 1.2815,
+      "step": 13775
+    },
+    {
+      "epoch": 0.2756,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00018179489528872807,
+      "loss": 1.2551,
+      "step": 13780
+    },
+    {
+      "epoch": 0.2757,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00018177480886849388,
+      "loss": 1.2731,
+      "step": 13785
+    },
+    {
+      "epoch": 0.2758,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001817547124842473,
+      "loss": 1.2723,
+      "step": 13790
+    },
+    {
+      "epoch": 0.2759,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00018173460613843701,
+      "loss": 1.2856,
+      "step": 13795
+    },
+    {
+      "epoch": 0.276,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00018171448983351284,
+      "loss": 1.2759,
+      "step": 13800
+    },
+    {
+      "epoch": 0.2761,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018169436357192602,
+      "loss": 1.2684,
+      "step": 13805
+    },
+    {
+      "epoch": 0.2762,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00018167422735612877,
+      "loss": 1.2848,
+      "step": 13810
+    },
+    {
+      "epoch": 0.2763,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00018165408118857464,
+      "loss": 1.2725,
+      "step": 13815
+    },
+    {
+      "epoch": 0.2764,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00018163392507171842,
+      "loss": 1.2796,
+      "step": 13820
+    },
+    {
+      "epoch": 0.2765,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00018161375900801604,
+      "loss": 1.2763,
+      "step": 13825
+    },
+    {
+      "epoch": 0.2766,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00018159358299992467,
+      "loss": 1.265,
+      "step": 13830
+    },
+    {
+      "epoch": 0.2767,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00018157339704990275,
+      "loss": 1.2751,
+      "step": 13835
+    },
+    {
+      "epoch": 0.2768,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00018155320116040982,
+      "loss": 1.2669,
+      "step": 13840
+    },
+    {
+      "epoch": 0.2769,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018153299533390672,
+      "loss": 1.2663,
+      "step": 13845
+    },
+    {
+      "epoch": 0.277,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00018151277957285543,
+      "loss": 1.2702,
+      "step": 13850
+    },
+    {
+      "epoch": 0.2771,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018149255387971922,
+      "loss": 1.2641,
+      "step": 13855
+    },
+    {
+      "epoch": 0.2772,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018147231825696252,
+      "loss": 1.2771,
+      "step": 13860
+    },
+    {
+      "epoch": 0.2773,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00018145207270705096,
+      "loss": 1.2642,
+      "step": 13865
+    },
+    {
+      "epoch": 0.2774,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001814318172324514,
+      "loss": 1.2926,
+      "step": 13870
+    },
+    {
+      "epoch": 0.2775,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00018141155183563193,
+      "loss": 1.2715,
+      "step": 13875
+    },
+    {
+      "epoch": 0.2776,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00018139127651906184,
+      "loss": 1.2719,
+      "step": 13880
+    },
+    {
+      "epoch": 0.2777,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00018137099128521156,
+      "loss": 1.2577,
+      "step": 13885
+    },
+    {
+      "epoch": 0.2778,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001813506961365528,
+      "loss": 1.2711,
+      "step": 13890
+    },
+    {
+      "epoch": 0.2779,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00018133039107555852,
+      "loss": 1.2686,
+      "step": 13895
+    },
+    {
+      "epoch": 0.278,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00018131007610470276,
+      "loss": 1.2703,
+      "step": 13900
+    },
+    {
+      "epoch": 0.2781,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001812897512264609,
+      "loss": 1.2718,
+      "step": 13905
+    },
+    {
+      "epoch": 0.2782,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0001812694164433094,
+      "loss": 1.2709,
+      "step": 13910
+    },
+    {
+      "epoch": 0.2783,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00018124907175772604,
+      "loss": 1.275,
+      "step": 13915
+    },
+    {
+      "epoch": 0.2784,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001812287171721897,
+      "loss": 1.279,
+      "step": 13920
+    },
+    {
+      "epoch": 0.2785,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00018120835268918063,
+      "loss": 1.2835,
+      "step": 13925
+    },
+    {
+      "epoch": 0.2786,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0001811879783111801,
+      "loss": 1.2926,
+      "step": 13930
+    },
+    {
+      "epoch": 0.2787,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00018116759404067064,
+      "loss": 1.2825,
+      "step": 13935
+    },
+    {
+      "epoch": 0.2788,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00018114719988013612,
+      "loss": 1.2772,
+      "step": 13940
+    },
+    {
+      "epoch": 0.2789,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018112679583206137,
+      "loss": 1.2717,
+      "step": 13945
+    },
+    {
+      "epoch": 0.279,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00018110638189893267,
+      "loss": 1.2659,
+      "step": 13950
+    },
+    {
+      "epoch": 0.2791,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00018108595808323736,
+      "loss": 1.2981,
+      "step": 13955
+    },
+    {
+      "epoch": 0.2792,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.000181065524387464,
+      "loss": 1.2865,
+      "step": 13960
+    },
+    {
+      "epoch": 0.2793,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0001810450808141024,
+      "loss": 1.2766,
+      "step": 13965
+    },
+    {
+      "epoch": 0.2794,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018102462736564355,
+      "loss": 1.2702,
+      "step": 13970
+    },
+    {
+      "epoch": 0.2795,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018100416404457961,
+      "loss": 1.2813,
+      "step": 13975
+    },
+    {
+      "epoch": 0.2796,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00018098369085340398,
+      "loss": 1.2684,
+      "step": 13980
+    },
+    {
+      "epoch": 0.2797,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00018096320779461132,
+      "loss": 1.273,
+      "step": 13985
+    },
+    {
+      "epoch": 0.2798,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00018094271487069735,
+      "loss": 1.2998,
+      "step": 13990
+    },
+    {
+      "epoch": 0.2799,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018092221208415907,
+      "loss": 1.2773,
+      "step": 13995
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00018090169943749476,
+      "loss": 1.2811,
+      "step": 14000
+    },
+    {
+      "epoch": 0.2801,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00018088117693320374,
+      "loss": 1.274,
+      "step": 14005
+    },
+    {
+      "epoch": 0.2802,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00018086064457378665,
+      "loss": 1.2709,
+      "step": 14010
+    },
+    {
+      "epoch": 0.2803,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00018084010236174534,
+      "loss": 1.2807,
+      "step": 14015
+    },
+    {
+      "epoch": 0.2804,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00018081955029958274,
+      "loss": 1.2638,
+      "step": 14020
+    },
+    {
+      "epoch": 0.2805,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018079898838980305,
+      "loss": 1.2822,
+      "step": 14025
+    },
+    {
+      "epoch": 0.2806,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00018077841663491175,
+      "loss": 1.2686,
+      "step": 14030
+    },
+    {
+      "epoch": 0.2807,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001807578350374154,
+      "loss": 1.2937,
+      "step": 14035
+    },
+    {
+      "epoch": 0.2808,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018073724359982186,
+      "loss": 1.2769,
+      "step": 14040
+    },
+    {
+      "epoch": 0.2809,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00018071664232464002,
+      "loss": 1.2768,
+      "step": 14045
+    },
+    {
+      "epoch": 0.281,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018069603121438022,
+      "loss": 1.2844,
+      "step": 14050
+    },
+    {
+      "epoch": 0.2811,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00018067541027155375,
+      "loss": 1.2666,
+      "step": 14055
+    },
+    {
+      "epoch": 0.2812,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018065477949867327,
+      "loss": 1.2855,
+      "step": 14060
+    },
+    {
+      "epoch": 0.2813,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00018063413889825254,
+      "loss": 1.2818,
+      "step": 14065
+    },
+    {
+      "epoch": 0.2814,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0001806134884728066,
+      "loss": 1.2736,
+      "step": 14070
+    },
+    {
+      "epoch": 0.2815,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00018059282822485158,
+      "loss": 1.2709,
+      "step": 14075
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018057215815690494,
+      "loss": 1.2829,
+      "step": 14080
+    },
+    {
+      "epoch": 0.2817,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018055147827148523,
+      "loss": 1.2795,
+      "step": 14085
+    },
+    {
+      "epoch": 0.2818,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001805307885711122,
+      "loss": 1.2749,
+      "step": 14090
+    },
+    {
+      "epoch": 0.2819,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001805100890583069,
+      "loss": 1.2754,
+      "step": 14095
+    },
+    {
+      "epoch": 0.282,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001804893797355914,
+      "loss": 1.2819,
+      "step": 14100
+    },
+    {
+      "epoch": 0.2821,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00018046866060548918,
+      "loss": 1.2646,
+      "step": 14105
+    },
+    {
+      "epoch": 0.2822,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018044793167052477,
+      "loss": 1.2765,
+      "step": 14110
+    },
+    {
+      "epoch": 0.2823,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018042719293322388,
+      "loss": 1.2705,
+      "step": 14115
+    },
+    {
+      "epoch": 0.2824,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00018040644439611348,
+      "loss": 1.2782,
+      "step": 14120
+    },
+    {
+      "epoch": 0.2825,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00018038568606172173,
+      "loss": 1.2889,
+      "step": 14125
+    },
+    {
+      "epoch": 0.2826,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00018036491793257798,
+      "loss": 1.2837,
+      "step": 14130
+    },
+    {
+      "epoch": 0.2827,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00018034414001121278,
+      "loss": 1.265,
+      "step": 14135
+    },
+    {
+      "epoch": 0.2828,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001803233523001578,
+      "loss": 1.2863,
+      "step": 14140
+    },
+    {
+      "epoch": 0.2829,
+      "grad_norm": 1.0,
+      "learning_rate": 0.000180302554801946,
+      "loss": 1.2713,
+      "step": 14145
+    },
+    {
+      "epoch": 0.283,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00018028174751911146,
+      "loss": 1.2687,
+      "step": 14150
+    },
+    {
+      "epoch": 0.2831,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00018026093045418954,
+      "loss": 1.2827,
+      "step": 14155
+    },
+    {
+      "epoch": 0.2832,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001802401036097167,
+      "loss": 1.2688,
+      "step": 14160
+    },
+    {
+      "epoch": 0.2833,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00018021926698823059,
+      "loss": 1.2849,
+      "step": 14165
+    },
+    {
+      "epoch": 0.2834,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00018019842059227012,
+      "loss": 1.2676,
+      "step": 14170
+    },
+    {
+      "epoch": 0.2835,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0001801775644243754,
+      "loss": 1.2788,
+      "step": 14175
+    },
+    {
+      "epoch": 0.2836,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018015669848708767,
+      "loss": 1.2856,
+      "step": 14180
+    },
+    {
+      "epoch": 0.2837,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00018013582278294935,
+      "loss": 1.2822,
+      "step": 14185
+    },
+    {
+      "epoch": 0.2838,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00018011493731450413,
+      "loss": 1.2918,
+      "step": 14190
+    },
+    {
+      "epoch": 0.2839,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0001800940420842968,
+      "loss": 1.2682,
+      "step": 14195
+    },
+    {
+      "epoch": 0.284,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00018007313709487334,
+      "loss": 1.2731,
+      "step": 14200
+    },
+    {
+      "epoch": 0.2841,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001800522223487811,
+      "loss": 1.2776,
+      "step": 14205
+    },
+    {
+      "epoch": 0.2842,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001800312978485683,
+      "loss": 1.272,
+      "step": 14210
+    },
+    {
+      "epoch": 0.2843,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018001036359678469,
+      "loss": 1.2845,
+      "step": 14215
+    },
+    {
+      "epoch": 0.2844,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00017998941959598095,
+      "loss": 1.2758,
+      "step": 14220
+    },
+    {
+      "epoch": 0.2845,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00017996846584870908,
+      "loss": 1.2849,
+      "step": 14225
+    },
+    {
+      "epoch": 0.2846,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001799475023575222,
+      "loss": 1.2658,
+      "step": 14230
+    },
+    {
+      "epoch": 0.2847,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00017992652912497464,
+      "loss": 1.2686,
+      "step": 14235
+    },
+    {
+      "epoch": 0.2848,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017990554615362198,
+      "loss": 1.283,
+      "step": 14240
+    },
+    {
+      "epoch": 0.2849,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00017988455344602092,
+      "loss": 1.2836,
+      "step": 14245
+    },
+    {
+      "epoch": 0.285,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00017986355100472928,
+      "loss": 1.2748,
+      "step": 14250
+    },
+    {
+      "epoch": 0.2851,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00017984253883230627,
+      "loss": 1.2853,
+      "step": 14255
+    },
+    {
+      "epoch": 0.2852,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00017982151693131203,
+      "loss": 1.2736,
+      "step": 14260
+    },
+    {
+      "epoch": 0.2853,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001798004853043081,
+      "loss": 1.2884,
+      "step": 14265
+    },
+    {
+      "epoch": 0.2854,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001797794439538571,
+      "loss": 1.2718,
+      "step": 14270
+    },
+    {
+      "epoch": 0.2855,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00017975839288252287,
+      "loss": 1.2878,
+      "step": 14275
+    },
+    {
+      "epoch": 0.2856,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00017973733209287036,
+      "loss": 1.2746,
+      "step": 14280
+    },
+    {
+      "epoch": 0.2857,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00017971626158746584,
+      "loss": 1.2796,
+      "step": 14285
+    },
+    {
+      "epoch": 0.2858,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00017969518136887663,
+      "loss": 1.2706,
+      "step": 14290
+    },
+    {
+      "epoch": 0.2859,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00017967409143967132,
+      "loss": 1.2798,
+      "step": 14295
+    },
+    {
+      "epoch": 0.286,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00017965299180241963,
+      "loss": 1.2803,
+      "step": 14300
+    },
+    {
+      "epoch": 0.2861,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00017963188245969253,
+      "loss": 1.2752,
+      "step": 14305
+    },
+    {
+      "epoch": 0.2862,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00017961076341406208,
+      "loss": 1.2802,
+      "step": 14310
+    },
+    {
+      "epoch": 0.2863,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001795896346681016,
+      "loss": 1.2785,
+      "step": 14315
+    },
+    {
+      "epoch": 0.2864,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017956849622438554,
+      "loss": 1.2848,
+      "step": 14320
+    },
+    {
+      "epoch": 0.2865,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00017954734808548958,
+      "loss": 1.2699,
+      "step": 14325
+    },
+    {
+      "epoch": 0.2866,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017952619025399057,
+      "loss": 1.2622,
+      "step": 14330
+    },
+    {
+      "epoch": 0.2867,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00017950502273246649,
+      "loss": 1.2744,
+      "step": 14335
+    },
+    {
+      "epoch": 0.2868,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00017948384552349657,
+      "loss": 1.2617,
+      "step": 14340
+    },
+    {
+      "epoch": 0.2869,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00017946265862966114,
+      "loss": 1.2846,
+      "step": 14345
+    },
+    {
+      "epoch": 0.287,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00017944146205354182,
+      "loss": 1.2749,
+      "step": 14350
+    },
+    {
+      "epoch": 0.2871,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00017942025579772132,
+      "loss": 1.2651,
+      "step": 14355
+    },
+    {
+      "epoch": 0.2872,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00017939903986478355,
+      "loss": 1.2953,
+      "step": 14360
+    },
+    {
+      "epoch": 0.2873,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001793778142573136,
+      "loss": 1.2785,
+      "step": 14365
+    },
+    {
+      "epoch": 0.2874,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001793565789778978,
+      "loss": 1.2689,
+      "step": 14370
+    },
+    {
+      "epoch": 0.2875,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00017933533402912354,
+      "loss": 1.2763,
+      "step": 14375
+    },
+    {
+      "epoch": 0.2876,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017931407941357947,
+      "loss": 1.2905,
+      "step": 14380
+    },
+    {
+      "epoch": 0.2877,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001792928151338554,
+      "loss": 1.2726,
+      "step": 14385
+    },
+    {
+      "epoch": 0.2878,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00017927154119254236,
+      "loss": 1.2758,
+      "step": 14390
+    },
+    {
+      "epoch": 0.2879,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00017925025759223245,
+      "loss": 1.264,
+      "step": 14395
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00017922896433551907,
+      "loss": 1.2697,
+      "step": 14400
+    },
+    {
+      "epoch": 0.2881,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00017920766142499672,
+      "loss": 1.2848,
+      "step": 14405
+    },
+    {
+      "epoch": 0.2882,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017918634886326108,
+      "loss": 1.2602,
+      "step": 14410
+    },
+    {
+      "epoch": 0.2883,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017916502665290903,
+      "loss": 1.2796,
+      "step": 14415
+    },
+    {
+      "epoch": 0.2884,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0001791436947965386,
+      "loss": 1.2725,
+      "step": 14420
+    },
+    {
+      "epoch": 0.2885,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017912235329674902,
+      "loss": 1.279,
+      "step": 14425
+    },
+    {
+      "epoch": 0.2886,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001791010021561407,
+      "loss": 1.2837,
+      "step": 14430
+    },
+    {
+      "epoch": 0.2887,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0001790796413773152,
+      "loss": 1.273,
+      "step": 14435
+    },
+    {
+      "epoch": 0.2888,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001790582709628753,
+      "loss": 1.2806,
+      "step": 14440
+    },
+    {
+      "epoch": 0.2889,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001790368909154249,
+      "loss": 1.2582,
+      "step": 14445
+    },
+    {
+      "epoch": 0.289,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00017901550123756906,
+      "loss": 1.2734,
+      "step": 14450
+    },
+    {
+      "epoch": 0.2891,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017899410193191406,
+      "loss": 1.275,
+      "step": 14455
+    },
+    {
+      "epoch": 0.2892,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00017897269300106737,
+      "loss": 1.269,
+      "step": 14460
+    },
+    {
+      "epoch": 0.2893,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001789512744476376,
+      "loss": 1.2706,
+      "step": 14465
+    },
+    {
+      "epoch": 0.2894,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001789298462742345,
+      "loss": 1.2803,
+      "step": 14470
+    },
+    {
+      "epoch": 0.2895,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00017890840848346908,
+      "loss": 1.2724,
+      "step": 14475
+    },
+    {
+      "epoch": 0.2896,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00017888696107795342,
+      "loss": 1.269,
+      "step": 14480
+    },
+    {
+      "epoch": 0.2897,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00017886550406030085,
+      "loss": 1.2649,
+      "step": 14485
+    },
+    {
+      "epoch": 0.2898,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00017884403743312582,
+      "loss": 1.2964,
+      "step": 14490
+    },
+    {
+      "epoch": 0.2899,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00017882256119904403,
+      "loss": 1.2873,
+      "step": 14495
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00017880107536067218,
+      "loss": 1.2547,
+      "step": 14500
+    },
+    {
+      "epoch": 0.2901,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001787795799206284,
+      "loss": 1.2598,
+      "step": 14505
+    },
+    {
+      "epoch": 0.2902,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00017875807488153175,
+      "loss": 1.2798,
+      "step": 14510
+    },
+    {
+      "epoch": 0.2903,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00017873656024600254,
+      "loss": 1.2776,
+      "step": 14515
+    },
+    {
+      "epoch": 0.2904,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017871503601666233,
+      "loss": 1.2793,
+      "step": 14520
+    },
+    {
+      "epoch": 0.2905,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017869350219613375,
+      "loss": 1.2768,
+      "step": 14525
+    },
+    {
+      "epoch": 0.2906,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0001786719587870406,
+      "loss": 1.2697,
+      "step": 14530
+    },
+    {
+      "epoch": 0.2907,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00017865040579200794,
+      "loss": 1.2756,
+      "step": 14535
+    },
+    {
+      "epoch": 0.2908,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017862884321366188,
+      "loss": 1.2625,
+      "step": 14540
+    },
+    {
+      "epoch": 0.2909,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001786072710546298,
+      "loss": 1.2717,
+      "step": 14545
+    },
+    {
+      "epoch": 0.291,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0001785856893175402,
+      "loss": 1.2601,
+      "step": 14550
+    },
+    {
+      "epoch": 0.2911,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00017856409800502272,
+      "loss": 1.2878,
+      "step": 14555
+    },
+    {
+      "epoch": 0.2912,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00017854249711970818,
+      "loss": 1.2864,
+      "step": 14560
+    },
+    {
+      "epoch": 0.2913,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00017852088666422863,
+      "loss": 1.2705,
+      "step": 14565
+    },
+    {
+      "epoch": 0.2914,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017849926664121726,
+      "loss": 1.2919,
+      "step": 14570
+    },
+    {
+      "epoch": 0.2915,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001784776370533083,
+      "loss": 1.2675,
+      "step": 14575
+    },
+    {
+      "epoch": 0.2916,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00017845599790313735,
+      "loss": 1.2899,
+      "step": 14580
+    },
+    {
+      "epoch": 0.2917,
+      "grad_norm": 0.375,
+      "learning_rate": 0.000178434349193341,
+      "loss": 1.263,
+      "step": 14585
+    },
+    {
+      "epoch": 0.2918,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00017841269092655715,
+      "loss": 1.2877,
+      "step": 14590
+    },
+    {
+      "epoch": 0.2919,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017839102310542477,
+      "loss": 1.2757,
+      "step": 14595
+    },
+    {
+      "epoch": 0.292,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.000178369345732584,
+      "loss": 1.2557,
+      "step": 14600
+    },
+    {
+      "epoch": 0.2921,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00017834765881067616,
+      "loss": 1.2862,
+      "step": 14605
+    },
+    {
+      "epoch": 0.2922,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00017832596234234376,
+      "loss": 1.2761,
+      "step": 14610
+    },
+    {
+      "epoch": 0.2923,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00017830425633023043,
+      "loss": 1.289,
+      "step": 14615
+    },
+    {
+      "epoch": 0.2924,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.000178282540776981,
+      "loss": 1.2746,
+      "step": 14620
+    },
+    {
+      "epoch": 0.2925,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001782608156852414,
+      "loss": 1.2817,
+      "step": 14625
+    },
+    {
+      "epoch": 0.2926,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001782390810576588,
+      "loss": 1.2749,
+      "step": 14630
+    },
+    {
+      "epoch": 0.2927,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00017821733689688153,
+      "loss": 1.2784,
+      "step": 14635
+    },
+    {
+      "epoch": 0.2928,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.000178195583205559,
+      "loss": 1.2795,
+      "step": 14640
+    },
+    {
+      "epoch": 0.2929,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017817381998634185,
+      "loss": 1.2871,
+      "step": 14645
+    },
+    {
+      "epoch": 0.293,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00017815204724188187,
+      "loss": 1.2784,
+      "step": 14650
+    },
+    {
+      "epoch": 0.2931,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.000178130264974832,
+      "loss": 1.2773,
+      "step": 14655
+    },
+    {
+      "epoch": 0.2932,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001781084731878463,
+      "loss": 1.2665,
+      "step": 14660
+    },
+    {
+      "epoch": 0.2933,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00017808667188358012,
+      "loss": 1.2734,
+      "step": 14665
+    },
+    {
+      "epoch": 0.2934,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00017806486106468981,
+      "loss": 1.2633,
+      "step": 14670
+    },
+    {
+      "epoch": 0.2935,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.000178043040733833,
+      "loss": 1.2884,
+      "step": 14675
+    },
+    {
+      "epoch": 0.2936,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00017802121089366836,
+      "loss": 1.2948,
+      "step": 14680
+    },
+    {
+      "epoch": 0.2937,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00017799937154685586,
+      "loss": 1.2851,
+      "step": 14685
+    },
+    {
+      "epoch": 0.2938,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00017797752269605653,
+      "loss": 1.2751,
+      "step": 14690
+    },
+    {
+      "epoch": 0.2939,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00017795566434393258,
+      "loss": 1.2967,
+      "step": 14695
+    },
+    {
+      "epoch": 0.294,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00017793379649314744,
+      "loss": 1.2845,
+      "step": 14700
+    },
+    {
+      "epoch": 0.2941,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00017791191914636554,
+      "loss": 1.2765,
+      "step": 14705
+    },
+    {
+      "epoch": 0.2942,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00017789003230625266,
+      "loss": 1.2749,
+      "step": 14710
+    },
+    {
+      "epoch": 0.2943,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00017786813597547562,
+      "loss": 1.2778,
+      "step": 14715
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00017784623015670238,
+      "loss": 1.2892,
+      "step": 14720
+    },
+    {
+      "epoch": 0.2945,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00017782431485260212,
+      "loss": 1.273,
+      "step": 14725
+    },
+    {
+      "epoch": 0.2946,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017780239006584515,
+      "loss": 1.2755,
+      "step": 14730
+    },
+    {
+      "epoch": 0.2947,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00017778045579910302,
+      "loss": 1.2814,
+      "step": 14735
+    },
+    {
+      "epoch": 0.2948,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001777585120550482,
+      "loss": 1.2728,
+      "step": 14740
+    },
+    {
+      "epoch": 0.2949,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001777365588363546,
+      "loss": 1.2826,
+      "step": 14745
+    },
+    {
+      "epoch": 0.295,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001777145961456971,
+      "loss": 1.2709,
+      "step": 14750
+    },
+    {
+      "epoch": 0.2951,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001776926239857518,
+      "loss": 1.2757,
+      "step": 14755
+    },
+    {
+      "epoch": 0.2952,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00017767064235919592,
+      "loss": 1.2704,
+      "step": 14760
+    },
+    {
+      "epoch": 0.2953,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00017764865126870786,
+      "loss": 1.2768,
+      "step": 14765
+    },
+    {
+      "epoch": 0.2954,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001776266507169672,
+      "loss": 1.2757,
+      "step": 14770
+    },
+    {
+      "epoch": 0.2955,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001776046407066546,
+      "loss": 1.2748,
+      "step": 14775
+    },
+    {
+      "epoch": 0.2956,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00017758262124045195,
+      "loss": 1.2739,
+      "step": 14780
+    },
+    {
+      "epoch": 0.2957,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001775605923210422,
+      "loss": 1.2669,
+      "step": 14785
+    },
+    {
+      "epoch": 0.2958,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.0001775385539511096,
+      "loss": 1.2754,
+      "step": 14790
+    },
+    {
+      "epoch": 0.2959,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017751650613333935,
+      "loss": 1.2705,
+      "step": 14795
+    },
+    {
+      "epoch": 0.296,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00017749444887041799,
+      "loss": 1.2834,
+      "step": 14800
+    },
+    {
+      "epoch": 0.2961,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00017747238216503307,
+      "loss": 1.2986,
+      "step": 14805
+    },
+    {
+      "epoch": 0.2962,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017745030601987337,
+      "loss": 1.2809,
+      "step": 14810
+    },
+    {
+      "epoch": 0.2963,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00017742822043762888,
+      "loss": 1.2956,
+      "step": 14815
+    },
+    {
+      "epoch": 0.2964,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00017740612542099053,
+      "loss": 1.2924,
+      "step": 14820
+    },
+    {
+      "epoch": 0.2965,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00017738402097265064,
+      "loss": 1.2759,
+      "step": 14825
+    },
+    {
+      "epoch": 0.2966,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0001773619070953025,
+      "loss": 1.2864,
+      "step": 14830
+    },
+    {
+      "epoch": 0.2967,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00017733978379164066,
+      "loss": 1.2729,
+      "step": 14835
+    },
+    {
+      "epoch": 0.2968,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00017731765106436073,
+      "loss": 1.2693,
+      "step": 14840
+    },
+    {
+      "epoch": 0.2969,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00017729550891615957,
+      "loss": 1.2757,
+      "step": 14845
+    },
+    {
+      "epoch": 0.297,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00017727335734973512,
+      "loss": 1.2887,
+      "step": 14850
+    },
+    {
+      "epoch": 0.2971,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00017725119636778644,
+      "loss": 1.2616,
+      "step": 14855
+    },
+    {
+      "epoch": 0.2972,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00017722902597301383,
+      "loss": 1.2704,
+      "step": 14860
+    },
+    {
+      "epoch": 0.2973,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017720684616811866,
+      "loss": 1.2476,
+      "step": 14865
+    },
+    {
+      "epoch": 0.2974,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0001771846569558035,
+      "loss": 1.2728,
+      "step": 14870
+    },
+    {
+      "epoch": 0.2975,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00017716245833877201,
+      "loss": 1.2647,
+      "step": 14875
+    },
+    {
+      "epoch": 0.2976,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00017714025031972903,
+      "loss": 1.2713,
+      "step": 14880
+    },
+    {
+      "epoch": 0.2977,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00017711803290138052,
+      "loss": 1.2732,
+      "step": 14885
+    },
+    {
+      "epoch": 0.2978,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00017709580608643363,
+      "loss": 1.2828,
+      "step": 14890
+    },
+    {
+      "epoch": 0.2979,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001770735698775966,
+      "loss": 1.2711,
+      "step": 14895
+    },
+    {
+      "epoch": 0.298,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00017705132427757895,
+      "loss": 1.2592,
+      "step": 14900
+    },
+    {
+      "epoch": 0.2981,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00017702906928909108,
+      "loss": 1.2756,
+      "step": 14905
+    },
+    {
+      "epoch": 0.2982,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.0001770068049148448,
+      "loss": 1.2883,
+      "step": 14910
+    },
+    {
+      "epoch": 0.2983,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00017698453115755293,
+      "loss": 1.2738,
+      "step": 14915
+    },
+    {
+      "epoch": 0.2984,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00017696224801992945,
+      "loss": 1.2671,
+      "step": 14920
+    },
+    {
+      "epoch": 0.2985,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001769399555046895,
+      "loss": 1.2617,
+      "step": 14925
+    },
+    {
+      "epoch": 0.2986,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00017691765361454938,
+      "loss": 1.2834,
+      "step": 14930
+    },
+    {
+      "epoch": 0.2987,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017689534235222648,
+      "loss": 1.2793,
+      "step": 14935
+    },
+    {
+      "epoch": 0.2988,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00017687302172043933,
+      "loss": 1.2658,
+      "step": 14940
+    },
+    {
+      "epoch": 0.2989,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00017685069172190766,
+      "loss": 1.2884,
+      "step": 14945
+    },
+    {
+      "epoch": 0.299,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00017682835235935236,
+      "loss": 1.2636,
+      "step": 14950
+    },
+    {
+      "epoch": 0.2991,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00017680600363549533,
+      "loss": 1.2718,
+      "step": 14955
+    },
+    {
+      "epoch": 0.2992,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00017678364555305978,
+      "loss": 1.2795,
+      "step": 14960
+    },
+    {
+      "epoch": 0.2993,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00017676127811476987,
+      "loss": 1.2754,
+      "step": 14965
+    },
+    {
+      "epoch": 0.2994,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001767389013233511,
+      "loss": 1.268,
+      "step": 14970
+    },
+    {
+      "epoch": 0.2995,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00017671651518153,
+      "loss": 1.2768,
+      "step": 14975
+    },
+    {
+      "epoch": 0.2996,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00017669411969203417,
+      "loss": 1.2774,
+      "step": 14980
+    },
+    {
+      "epoch": 0.2997,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00017667171485759252,
+      "loss": 1.2867,
+      "step": 14985
+    },
+    {
+      "epoch": 0.2998,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00017664930068093498,
+      "loss": 1.274,
+      "step": 14990
+    },
+    {
+      "epoch": 0.2999,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00017662687716479266,
+      "loss": 1.2993,
+      "step": 14995
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0001766044443118978,
+      "loss": 1.2793,
+      "step": 15000
+    },
+    {
+      "epoch": 0.3001,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017658200212498378,
+      "loss": 1.2666,
+      "step": 15005
+    },
+    {
+      "epoch": 0.3002,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00017655955060678506,
+      "loss": 1.2557,
+      "step": 15010
+    },
+    {
+      "epoch": 0.3003,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00017653708976003737,
+      "loss": 1.2865,
+      "step": 15015
+    },
+    {
+      "epoch": 0.3004,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00017651461958747745,
+      "loss": 1.2781,
+      "step": 15020
+    },
+    {
+      "epoch": 0.3005,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001764921400918432,
+      "loss": 1.2741,
+      "step": 15025
+    },
+    {
+      "epoch": 0.3006,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001764696512758737,
+      "loss": 1.2691,
+      "step": 15030
+    },
+    {
+      "epoch": 0.3007,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00017644715314230918,
+      "loss": 1.2715,
+      "step": 15035
+    },
+    {
+      "epoch": 0.3008,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001764246456938909,
+      "loss": 1.2712,
+      "step": 15040
+    },
+    {
+      "epoch": 0.3009,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017640212893336142,
+      "loss": 1.2573,
+      "step": 15045
+    },
+    {
+      "epoch": 0.301,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00017637960286346425,
+      "loss": 1.2872,
+      "step": 15050
+    },
+    {
+      "epoch": 0.3011,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00017635706748694413,
+      "loss": 1.2622,
+      "step": 15055
+    },
+    {
+      "epoch": 0.3012,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.000176334522806547,
+      "loss": 1.2769,
+      "step": 15060
+    },
+    {
+      "epoch": 0.3013,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00017631196882501973,
+      "loss": 1.2777,
+      "step": 15065
+    },
+    {
+      "epoch": 0.3014,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00017628940554511061,
+      "loss": 1.2869,
+      "step": 15070
+    },
+    {
+      "epoch": 0.3015,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017626683296956882,
+      "loss": 1.2877,
+      "step": 15075
+    },
+    {
+      "epoch": 0.3016,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.0001762442511011448,
+      "loss": 1.267,
+      "step": 15080
+    },
+    {
+      "epoch": 0.3017,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017622165994259,
+      "loss": 1.274,
+      "step": 15085
+    },
+    {
+      "epoch": 0.3018,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001761990594966572,
+      "loss": 1.2706,
+      "step": 15090
+    },
+    {
+      "epoch": 0.3019,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001761764497661001,
+      "loss": 1.2824,
+      "step": 15095
+    },
+    {
+      "epoch": 0.302,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001761538307536737,
+      "loss": 1.277,
+      "step": 15100
+    },
+    {
+      "epoch": 0.3021,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.000176131202462134,
+      "loss": 1.2751,
+      "step": 15105
+    },
+    {
+      "epoch": 0.3022,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001761085648942382,
+      "loss": 1.2859,
+      "step": 15110
+    },
+    {
+      "epoch": 0.3023,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00017608591805274464,
+      "loss": 1.2994,
+      "step": 15115
+    },
+    {
+      "epoch": 0.3024,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00017606326194041273,
+      "loss": 1.2752,
+      "step": 15120
+    },
+    {
+      "epoch": 0.3025,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0001760405965600031,
+      "loss": 1.2811,
+      "step": 15125
+    },
+    {
+      "epoch": 0.3026,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00017601792191427741,
+      "loss": 1.2793,
+      "step": 15130
+    },
+    {
+      "epoch": 0.3027,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001759952380059986,
+      "loss": 1.2791,
+      "step": 15135
+    },
+    {
+      "epoch": 0.3028,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00017597254483793048,
+      "loss": 1.2651,
+      "step": 15140
+    },
+    {
+      "epoch": 0.3029,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00017594984241283825,
+      "loss": 1.2696,
+      "step": 15145
+    },
+    {
+      "epoch": 0.303,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00017592713073348807,
+      "loss": 1.2772,
+      "step": 15150
+    },
+    {
+      "epoch": 0.3031,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00017590440980264738,
+      "loss": 1.2629,
+      "step": 15155
+    },
+    {
+      "epoch": 0.3032,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017588167962308458,
+      "loss": 1.2732,
+      "step": 15160
+    },
+    {
+      "epoch": 0.3033,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00017585894019756925,
+      "loss": 1.2758,
+      "step": 15165
+    },
+    {
+      "epoch": 0.3034,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001758361915288722,
+      "loss": 1.2845,
+      "step": 15170
+    },
+    {
+      "epoch": 0.3035,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00017581343361976524,
+      "loss": 1.2976,
+      "step": 15175
+    },
+    {
+      "epoch": 0.3036,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00017579066647302133,
+      "loss": 1.262,
+      "step": 15180
+    },
+    {
+      "epoch": 0.3037,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00017576789009141465,
+      "loss": 1.283,
+      "step": 15185
+    },
+    {
+      "epoch": 0.3038,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00017574510447772039,
+      "loss": 1.2721,
+      "step": 15190
+    },
+    {
+      "epoch": 0.3039,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00017572230963471488,
+      "loss": 1.2738,
+      "step": 15195
+    },
+    {
+      "epoch": 0.304,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00017569950556517566,
+      "loss": 1.2657,
+      "step": 15200
+    },
+    {
+      "epoch": 0.3041,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00017567669227188128,
+      "loss": 1.2833,
+      "step": 15205
+    },
+    {
+      "epoch": 0.3042,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001756538697576115,
+      "loss": 1.2801,
+      "step": 15210
+    },
+    {
+      "epoch": 0.3043,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0001756310380251472,
+      "loss": 1.2709,
+      "step": 15215
+    },
+    {
+      "epoch": 0.3044,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017560819707727033,
+      "loss": 1.271,
+      "step": 15220
+    },
+    {
+      "epoch": 0.3045,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017558534691676397,
+      "loss": 1.2934,
+      "step": 15225
+    },
+    {
+      "epoch": 0.3046,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00017556248754641235,
+      "loss": 1.2833,
+      "step": 15230
+    },
+    {
+      "epoch": 0.3047,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00017553961896900087,
+      "loss": 1.2853,
+      "step": 15235
+    },
+    {
+      "epoch": 0.3048,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017551674118731591,
+      "loss": 1.2762,
+      "step": 15240
+    },
+    {
+      "epoch": 0.3049,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017549385420414514,
+      "loss": 1.2714,
+      "step": 15245
+    },
+    {
+      "epoch": 0.305,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00017547095802227723,
+      "loss": 1.2767,
+      "step": 15250
+    },
+    {
+      "epoch": 0.3051,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00017544805264450196,
+      "loss": 1.2809,
+      "step": 15255
+    },
+    {
+      "epoch": 0.3052,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00017542513807361037,
+      "loss": 1.2878,
+      "step": 15260
+    },
+    {
+      "epoch": 0.3053,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00017540221431239453,
+      "loss": 1.2757,
+      "step": 15265
+    },
+    {
+      "epoch": 0.3054,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017537928136364755,
+      "loss": 1.2848,
+      "step": 15270
+    },
+    {
+      "epoch": 0.3055,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001753563392301638,
+      "loss": 1.2933,
+      "step": 15275
+    },
+    {
+      "epoch": 0.3056,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001753333879147387,
+      "loss": 1.2775,
+      "step": 15280
+    },
+    {
+      "epoch": 0.3057,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00017531042742016876,
+      "loss": 1.261,
+      "step": 15285
+    },
+    {
+      "epoch": 0.3058,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00017528745774925172,
+      "loss": 1.2739,
+      "step": 15290
+    },
+    {
+      "epoch": 0.3059,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017526447890478633,
+      "loss": 1.2726,
+      "step": 15295
+    },
+    {
+      "epoch": 0.306,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00017524149088957245,
+      "loss": 1.2742,
+      "step": 15300
+    },
+    {
+      "epoch": 0.3061,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00017521849370641114,
+      "loss": 1.2841,
+      "step": 15305
+    },
+    {
+      "epoch": 0.3062,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017519548735810456,
+      "loss": 1.2553,
+      "step": 15310
+    },
+    {
+      "epoch": 0.3063,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00017517247184745593,
+      "loss": 1.2793,
+      "step": 15315
+    },
+    {
+      "epoch": 0.3064,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00017514944717726962,
+      "loss": 1.2574,
+      "step": 15320
+    },
+    {
+      "epoch": 0.3065,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00017512641335035113,
+      "loss": 1.2782,
+      "step": 15325
+    },
+    {
+      "epoch": 0.3066,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00017510337036950703,
+      "loss": 1.2707,
+      "step": 15330
+    },
+    {
+      "epoch": 0.3067,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001750803182375451,
+      "loss": 1.2937,
+      "step": 15335
+    },
+    {
+      "epoch": 0.3068,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017505725695727412,
+      "loss": 1.2696,
+      "step": 15340
+    },
+    {
+      "epoch": 0.3069,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00017503418653150405,
+      "loss": 1.2669,
+      "step": 15345
+    },
+    {
+      "epoch": 0.307,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00017501110696304596,
+      "loss": 1.2686,
+      "step": 15350
+    },
+    {
+      "epoch": 0.3071,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00017498801825471203,
+      "loss": 1.2772,
+      "step": 15355
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00017496492040931552,
+      "loss": 1.2567,
+      "step": 15360
+    },
+    {
+      "epoch": 0.3073,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00017494181342967083,
+      "loss": 1.2684,
+      "step": 15365
+    },
+    {
+      "epoch": 0.3074,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00017491869731859353,
+      "loss": 1.2711,
+      "step": 15370
+    },
+    {
+      "epoch": 0.3075,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00017489557207890023,
+      "loss": 1.2726,
+      "step": 15375
+    },
+    {
+      "epoch": 0.3076,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0001748724377134086,
+      "loss": 1.2968,
+      "step": 15380
+    },
+    {
+      "epoch": 0.3077,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001748492942249376,
+      "loss": 1.2801,
+      "step": 15385
+    },
+    {
+      "epoch": 0.3078,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00017482614161630714,
+      "loss": 1.2642,
+      "step": 15390
+    },
+    {
+      "epoch": 0.3079,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00017480297989033825,
+      "loss": 1.2695,
+      "step": 15395
+    },
+    {
+      "epoch": 0.308,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.0001747798090498532,
+      "loss": 1.3031,
+      "step": 15400
+    },
+    {
+      "epoch": 0.3081,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00017475662909767522,
+      "loss": 1.2677,
+      "step": 15405
+    },
+    {
+      "epoch": 0.3082,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00017473344003662877,
+      "loss": 1.2727,
+      "step": 15410
+    },
+    {
+      "epoch": 0.3083,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00017471024186953936,
+      "loss": 1.2634,
+      "step": 15415
+    },
+    {
+      "epoch": 0.3084,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00017468703459923356,
+      "loss": 1.271,
+      "step": 15420
+    },
+    {
+      "epoch": 0.3085,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00017466381822853915,
+      "loss": 1.2678,
+      "step": 15425
+    },
+    {
+      "epoch": 0.3086,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00017464059276028497,
+      "loss": 1.273,
+      "step": 15430
+    },
+    {
+      "epoch": 0.3087,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00017461735819730096,
+      "loss": 1.2972,
+      "step": 15435
+    },
+    {
+      "epoch": 0.3088,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00017459411454241822,
+      "loss": 1.2564,
+      "step": 15440
+    },
+    {
+      "epoch": 0.3089,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00017457086179846888,
+      "loss": 1.2822,
+      "step": 15445
+    },
+    {
+      "epoch": 0.309,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017454759996828623,
+      "loss": 1.2736,
+      "step": 15450
+    },
+    {
+      "epoch": 0.3091,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00017452432905470464,
+      "loss": 1.263,
+      "step": 15455
+    },
+    {
+      "epoch": 0.3092,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00017450104906055963,
+      "loss": 1.2742,
+      "step": 15460
+    },
+    {
+      "epoch": 0.3093,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00017447775998868776,
+      "loss": 1.2747,
+      "step": 15465
+    },
+    {
+      "epoch": 0.3094,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00017445446184192676,
+      "loss": 1.2778,
+      "step": 15470
+    },
+    {
+      "epoch": 0.3095,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001744311546231154,
+      "loss": 1.2704,
+      "step": 15475
+    },
+    {
+      "epoch": 0.3096,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00017440783833509366,
+      "loss": 1.2724,
+      "step": 15480
+    },
+    {
+      "epoch": 0.3097,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00017438451298070252,
+      "loss": 1.28,
+      "step": 15485
+    },
+    {
+      "epoch": 0.3098,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001743611785627841,
+      "loss": 1.2646,
+      "step": 15490
+    },
+    {
+      "epoch": 0.3099,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00017433783508418162,
+      "loss": 1.2803,
+      "step": 15495
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00017431448254773944,
+      "loss": 1.273,
+      "step": 15500
+    },
+    {
+      "epoch": 0.3101,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00017429112095630295,
+      "loss": 1.277,
+      "step": 15505
+    },
+    {
+      "epoch": 0.3102,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017426775031271875,
+      "loss": 1.2799,
+      "step": 15510
+    },
+    {
+      "epoch": 0.3103,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00017424437061983446,
+      "loss": 1.2801,
+      "step": 15515
+    },
+    {
+      "epoch": 0.3104,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00017422098188049883,
+      "loss": 1.2728,
+      "step": 15520
+    },
+    {
+      "epoch": 0.3105,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00017419758409756164,
+      "loss": 1.2751,
+      "step": 15525
+    },
+    {
+      "epoch": 0.3106,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00017417417727387394,
+      "loss": 1.2661,
+      "step": 15530
+    },
+    {
+      "epoch": 0.3107,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001741507614122877,
+      "loss": 1.2764,
+      "step": 15535
+    },
+    {
+      "epoch": 0.3108,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001741273365156561,
+      "loss": 1.2856,
+      "step": 15540
+    },
+    {
+      "epoch": 0.3109,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017410390258683345,
+      "loss": 1.257,
+      "step": 15545
+    },
+    {
+      "epoch": 0.311,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.000174080459628675,
+      "loss": 1.2728,
+      "step": 15550
+    },
+    {
+      "epoch": 0.3111,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017405700764403726,
+      "loss": 1.2581,
+      "step": 15555
+    },
+    {
+      "epoch": 0.3112,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00017403354663577783,
+      "loss": 1.2588,
+      "step": 15560
+    },
+    {
+      "epoch": 0.3113,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017401007660675525,
+      "loss": 1.2783,
+      "step": 15565
+    },
+    {
+      "epoch": 0.3114,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017398659755982936,
+      "loss": 1.2759,
+      "step": 15570
+    },
+    {
+      "epoch": 0.3115,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.000173963109497861,
+      "loss": 1.2834,
+      "step": 15575
+    },
+    {
+      "epoch": 0.3116,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00017393961242371205,
+      "loss": 1.2823,
+      "step": 15580
+    },
+    {
+      "epoch": 0.3117,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017391610634024564,
+      "loss": 1.269,
+      "step": 15585
+    },
+    {
+      "epoch": 0.3118,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0001738925912503259,
+      "loss": 1.2859,
+      "step": 15590
+    },
+    {
+      "epoch": 0.3119,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.000173869067156818,
+      "loss": 1.2801,
+      "step": 15595
+    },
+    {
+      "epoch": 0.312,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00017384553406258842,
+      "loss": 1.2792,
+      "step": 15600
+    },
+    {
+      "epoch": 0.3121,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001738219919705044,
+      "loss": 1.2775,
+      "step": 15605
+    },
+    {
+      "epoch": 0.3122,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00017379844088343468,
+      "loss": 1.2694,
+      "step": 15610
+    },
+    {
+      "epoch": 0.3123,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017377488080424876,
+      "loss": 1.2962,
+      "step": 15615
+    },
+    {
+      "epoch": 0.3124,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001737513117358174,
+      "loss": 1.2639,
+      "step": 15620
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001737277336810124,
+      "loss": 1.2715,
+      "step": 15625
+    },
+    {
+      "epoch": 0.3126,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017370414664270674,
+      "loss": 1.2714,
+      "step": 15630
+    },
+    {
+      "epoch": 0.3127,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017368055062377434,
+      "loss": 1.2647,
+      "step": 15635
+    },
+    {
+      "epoch": 0.3128,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00017365694562709034,
+      "loss": 1.2867,
+      "step": 15640
+    },
+    {
+      "epoch": 0.3129,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017363333165553092,
+      "loss": 1.2728,
+      "step": 15645
+    },
+    {
+      "epoch": 0.313,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00017360970871197346,
+      "loss": 1.2653,
+      "step": 15650
+    },
+    {
+      "epoch": 0.3131,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0001735860767992962,
+      "loss": 1.2768,
+      "step": 15655
+    },
+    {
+      "epoch": 0.3132,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017356243592037875,
+      "loss": 1.2822,
+      "step": 15660
+    },
+    {
+      "epoch": 0.3133,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001735387860781016,
+      "loss": 1.2628,
+      "step": 15665
+    },
+    {
+      "epoch": 0.3134,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017351512727534644,
+      "loss": 1.267,
+      "step": 15670
+    },
+    {
+      "epoch": 0.3135,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.000173491459514996,
+      "loss": 1.2603,
+      "step": 15675
+    },
+    {
+      "epoch": 0.3136,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017346778279993415,
+      "loss": 1.2806,
+      "step": 15680
+    },
+    {
+      "epoch": 0.3137,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0001734440971330458,
+      "loss": 1.2794,
+      "step": 15685
+    },
+    {
+      "epoch": 0.3138,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017342040251721702,
+      "loss": 1.2543,
+      "step": 15690
+    },
+    {
+      "epoch": 0.3139,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001733966989553349,
+      "loss": 1.2737,
+      "step": 15695
+    },
+    {
+      "epoch": 0.314,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00017337298645028764,
+      "loss": 1.2778,
+      "step": 15700
+    },
+    {
+      "epoch": 0.3141,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00017334926500496456,
+      "loss": 1.2832,
+      "step": 15705
+    },
+    {
+      "epoch": 0.3142,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017332553462225602,
+      "loss": 1.2871,
+      "step": 15710
+    },
+    {
+      "epoch": 0.3143,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00017330179530505348,
+      "loss": 1.2605,
+      "step": 15715
+    },
+    {
+      "epoch": 0.3144,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00017327804705624957,
+      "loss": 1.2808,
+      "step": 15720
+    },
+    {
+      "epoch": 0.3145,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001732542898787379,
+      "loss": 1.2778,
+      "step": 15725
+    },
+    {
+      "epoch": 0.3146,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00017323052377541317,
+      "loss": 1.2697,
+      "step": 15730
+    },
+    {
+      "epoch": 0.3147,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0001732067487491713,
+      "loss": 1.2684,
+      "step": 15735
+    },
+    {
+      "epoch": 0.3148,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0001731829648029091,
+      "loss": 1.2736,
+      "step": 15740
+    },
+    {
+      "epoch": 0.3149,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.0001731591719395247,
+      "loss": 1.2714,
+      "step": 15745
+    },
+    {
+      "epoch": 0.315,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017313537016191706,
+      "loss": 1.2806,
+      "step": 15750
+    },
+    {
+      "epoch": 0.3151,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00017311155947298643,
+      "loss": 1.2589,
+      "step": 15755
+    },
+    {
+      "epoch": 0.3152,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00017308773987563406,
+      "loss": 1.2591,
+      "step": 15760
+    },
+    {
+      "epoch": 0.3153,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017306391137276224,
+      "loss": 1.2695,
+      "step": 15765
+    },
+    {
+      "epoch": 0.3154,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00017304007396727448,
+      "loss": 1.266,
+      "step": 15770
+    },
+    {
+      "epoch": 0.3155,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00017301622766207527,
+      "loss": 1.2819,
+      "step": 15775
+    },
+    {
+      "epoch": 0.3156,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00017299237246007015,
+      "loss": 1.2608,
+      "step": 15780
+    },
+    {
+      "epoch": 0.3157,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00017296850836416588,
+      "loss": 1.2704,
+      "step": 15785
+    },
+    {
+      "epoch": 0.3158,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00017294463537727024,
+      "loss": 1.2834,
+      "step": 15790
+    },
+    {
+      "epoch": 0.3159,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.000172920753502292,
+      "loss": 1.2758,
+      "step": 15795
+    },
+    {
+      "epoch": 0.316,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017289686274214118,
+      "loss": 1.2687,
+      "step": 15800
+    },
+    {
+      "epoch": 0.3161,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0001728729630997287,
+      "loss": 1.2878,
+      "step": 15805
+    },
+    {
+      "epoch": 0.3162,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017284905457796675,
+      "loss": 1.2561,
+      "step": 15810
+    },
+    {
+      "epoch": 0.3163,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00017282513717976848,
+      "loss": 1.2687,
+      "step": 15815
+    },
+    {
+      "epoch": 0.3164,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00017280121090804812,
+      "loss": 1.2854,
+      "step": 15820
+    },
+    {
+      "epoch": 0.3165,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00017277727576572107,
+      "loss": 1.2653,
+      "step": 15825
+    },
+    {
+      "epoch": 0.3166,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00017275333175570368,
+      "loss": 1.2797,
+      "step": 15830
+    },
+    {
+      "epoch": 0.3167,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00017272937888091353,
+      "loss": 1.2851,
+      "step": 15835
+    },
+    {
+      "epoch": 0.3168,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001727054171442692,
+      "loss": 1.2715,
+      "step": 15840
+    },
+    {
+      "epoch": 0.3169,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001726814465486903,
+      "loss": 1.2575,
+      "step": 15845
+    },
+    {
+      "epoch": 0.317,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001726574670970976,
+      "loss": 1.2738,
+      "step": 15850
+    },
+    {
+      "epoch": 0.3171,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00017263347879241291,
+      "loss": 1.2613,
+      "step": 15855
+    },
+    {
+      "epoch": 0.3172,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00017260948163755918,
+      "loss": 1.2835,
+      "step": 15860
+    },
+    {
+      "epoch": 0.3173,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00017258547563546038,
+      "loss": 1.2727,
+      "step": 15865
+    },
+    {
+      "epoch": 0.3174,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00017256146078904153,
+      "loss": 1.2887,
+      "step": 15870
+    },
+    {
+      "epoch": 0.3175,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00017253743710122875,
+      "loss": 1.2594,
+      "step": 15875
+    },
+    {
+      "epoch": 0.3176,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00017251340457494934,
+      "loss": 1.2854,
+      "step": 15880
+    },
+    {
+      "epoch": 0.3177,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001724893632131315,
+      "loss": 1.2695,
+      "step": 15885
+    },
+    {
+      "epoch": 0.3178,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.0001724653130187047,
+      "loss": 1.2825,
+      "step": 15890
+    },
+    {
+      "epoch": 0.3179,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00017244125399459926,
+      "loss": 1.2649,
+      "step": 15895
+    },
+    {
+      "epoch": 0.318,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017241718614374678,
+      "loss": 1.2749,
+      "step": 15900
+    },
+    {
+      "epoch": 0.3181,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0001723931094690798,
+      "loss": 1.2636,
+      "step": 15905
+    },
+    {
+      "epoch": 0.3182,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00017236902397353205,
+      "loss": 1.283,
+      "step": 15910
+    },
+    {
+      "epoch": 0.3183,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00017234492966003824,
+      "loss": 1.2979,
+      "step": 15915
+    },
+    {
+      "epoch": 0.3184,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00017232082653153422,
+      "loss": 1.2681,
+      "step": 15920
+    },
+    {
+      "epoch": 0.3185,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00017229671459095683,
+      "loss": 1.2722,
+      "step": 15925
+    },
+    {
+      "epoch": 0.3186,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0001722725938412441,
+      "loss": 1.2687,
+      "step": 15930
+    },
+    {
+      "epoch": 0.3187,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017224846428533499,
+      "loss": 1.2793,
+      "step": 15935
+    },
+    {
+      "epoch": 0.3188,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0001722243259261697,
+      "loss": 1.2799,
+      "step": 15940
+    },
+    {
+      "epoch": 0.3189,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00017220017876668934,
+      "loss": 1.3004,
+      "step": 15945
+    },
+    {
+      "epoch": 0.319,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00017217602280983623,
+      "loss": 1.2717,
+      "step": 15950
+    },
+    {
+      "epoch": 0.3191,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00017215185805855368,
+      "loss": 1.2777,
+      "step": 15955
+    },
+    {
+      "epoch": 0.3192,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00017212768451578607,
+      "loss": 1.2663,
+      "step": 15960
+    },
+    {
+      "epoch": 0.3193,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00017210350218447887,
+      "loss": 1.2716,
+      "step": 15965
+    },
+    {
+      "epoch": 0.3194,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017207931106757868,
+      "loss": 1.2592,
+      "step": 15970
+    },
+    {
+      "epoch": 0.3195,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00017205511116803306,
+      "loss": 1.2792,
+      "step": 15975
+    },
+    {
+      "epoch": 0.3196,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0001720309024887907,
+      "loss": 1.2633,
+      "step": 15980
+    },
+    {
+      "epoch": 0.3197,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00017200668503280136,
+      "loss": 1.2762,
+      "step": 15985
+    },
+    {
+      "epoch": 0.3198,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001719824588030159,
+      "loss": 1.2883,
+      "step": 15990
+    },
+    {
+      "epoch": 0.3199,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00017195822380238615,
+      "loss": 1.2686,
+      "step": 15995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001719339800338651,
+      "loss": 1.2566,
+      "step": 16000
+    },
+    {
+      "epoch": 0.3201,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00017190972750040682,
+      "loss": 1.287,
+      "step": 16005
+    },
+    {
+      "epoch": 0.3202,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00017188546620496635,
+      "loss": 1.287,
+      "step": 16010
+    },
+    {
+      "epoch": 0.3203,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00017186119615049988,
+      "loss": 1.2819,
+      "step": 16015
+    },
+    {
+      "epoch": 0.3204,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017183691733996462,
+      "loss": 1.2681,
+      "step": 16020
+    },
+    {
+      "epoch": 0.3205,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00017181262977631888,
+      "loss": 1.2843,
+      "step": 16025
+    },
+    {
+      "epoch": 0.3206,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00017178833346252206,
+      "loss": 1.2743,
+      "step": 16030
+    },
+    {
+      "epoch": 0.3207,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017176402840153455,
+      "loss": 1.2923,
+      "step": 16035
+    },
+    {
+      "epoch": 0.3208,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00017173971459631787,
+      "loss": 1.2639,
+      "step": 16040
+    },
+    {
+      "epoch": 0.3209,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001717153920498346,
+      "loss": 1.2709,
+      "step": 16045
+    },
+    {
+      "epoch": 0.321,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0001716910607650483,
+      "loss": 1.2929,
+      "step": 16050
+    },
+    {
+      "epoch": 0.3211,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001716667207449237,
+      "loss": 1.2709,
+      "step": 16055
+    },
+    {
+      "epoch": 0.3212,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0001716423719924266,
+      "loss": 1.2655,
+      "step": 16060
+    },
+    {
+      "epoch": 0.3213,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017161801451052376,
+      "loss": 1.2797,
+      "step": 16065
+    },
+    {
+      "epoch": 0.3214,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00017159364830218312,
+      "loss": 1.2912,
+      "step": 16070
+    },
+    {
+      "epoch": 0.3215,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0001715692733703736,
+      "loss": 1.2695,
+      "step": 16075
+    },
+    {
+      "epoch": 0.3216,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017154488971806518,
+      "loss": 1.2767,
+      "step": 16080
+    },
+    {
+      "epoch": 0.3217,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00017152049734822902,
+      "loss": 1.2826,
+      "step": 16085
+    },
+    {
+      "epoch": 0.3218,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017149609626383717,
+      "loss": 1.2738,
+      "step": 16090
+    },
+    {
+      "epoch": 0.3219,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00017147168646786286,
+      "loss": 1.2728,
+      "step": 16095
+    },
+    {
+      "epoch": 0.322,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00017144726796328034,
+      "loss": 1.2764,
+      "step": 16100
+    },
+    {
+      "epoch": 0.3221,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00017142284075306497,
+      "loss": 1.279,
+      "step": 16105
+    },
+    {
+      "epoch": 0.3222,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0001713984048401931,
+      "loss": 1.2891,
+      "step": 16110
+    },
+    {
+      "epoch": 0.3223,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00017137396022764214,
+      "loss": 1.2714,
+      "step": 16115
+    },
+    {
+      "epoch": 0.3224,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00017134950691839065,
+      "loss": 1.2747,
+      "step": 16120
+    },
+    {
+      "epoch": 0.3225,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017132504491541818,
+      "loss": 1.2872,
+      "step": 16125
+    },
+    {
+      "epoch": 0.3226,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001713005742217053,
+      "loss": 1.2844,
+      "step": 16130
+    },
+    {
+      "epoch": 0.3227,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00017127609484023377,
+      "loss": 1.2764,
+      "step": 16135
+    },
+    {
+      "epoch": 0.3228,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017125160677398626,
+      "loss": 1.2731,
+      "step": 16140
+    },
+    {
+      "epoch": 0.3229,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001712271100259466,
+      "loss": 1.2772,
+      "step": 16145
+    },
+    {
+      "epoch": 0.323,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00017120260459909967,
+      "loss": 1.2892,
+      "step": 16150
+    },
+    {
+      "epoch": 0.3231,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001711780904964313,
+      "loss": 1.2768,
+      "step": 16155
+    },
+    {
+      "epoch": 0.3232,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00017115356772092857,
+      "loss": 1.2764,
+      "step": 16160
+    },
+    {
+      "epoch": 0.3233,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001711290362755794,
+      "loss": 1.2721,
+      "step": 16165
+    },
+    {
+      "epoch": 0.3234,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00017110449616337289,
+      "loss": 1.2685,
+      "step": 16170
+    },
+    {
+      "epoch": 0.3235,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00017107994738729926,
+      "loss": 1.2767,
+      "step": 16175
+    },
+    {
+      "epoch": 0.3236,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00017105538995034963,
+      "loss": 1.2666,
+      "step": 16180
+    },
+    {
+      "epoch": 0.3237,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00017103082385551627,
+      "loss": 1.2669,
+      "step": 16185
+    },
+    {
+      "epoch": 0.3238,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.0001710062491057925,
+      "loss": 1.263,
+      "step": 16190
+    },
+    {
+      "epoch": 0.3239,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00017098166570417262,
+      "loss": 1.2751,
+      "step": 16195
+    },
+    {
+      "epoch": 0.324,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001709570736536521,
+      "loss": 1.2705,
+      "step": 16200
+    },
+    {
+      "epoch": 0.3241,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001709324729572274,
+      "loss": 1.265,
+      "step": 16205
+    },
+    {
+      "epoch": 0.3242,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.000170907863617896,
+      "loss": 1.2841,
+      "step": 16210
+    },
+    {
+      "epoch": 0.3243,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00017088324563865656,
+      "loss": 1.2605,
+      "step": 16215
+    },
+    {
+      "epoch": 0.3244,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00017085861902250865,
+      "loss": 1.2686,
+      "step": 16220
+    },
+    {
+      "epoch": 0.3245,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001708339837724529,
+      "loss": 1.273,
+      "step": 16225
+    },
+    {
+      "epoch": 0.3246,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0001708093398914911,
+      "loss": 1.2857,
+      "step": 16230
+    },
+    {
+      "epoch": 0.3247,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00017078468738262602,
+      "loss": 1.2791,
+      "step": 16235
+    },
+    {
+      "epoch": 0.3248,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00017076002624886154,
+      "loss": 1.2983,
+      "step": 16240
+    },
+    {
+      "epoch": 0.3249,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00017073535649320248,
+      "loss": 1.2837,
+      "step": 16245
+    },
+    {
+      "epoch": 0.325,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00017071067811865476,
+      "loss": 1.2597,
+      "step": 16250
+    },
+    {
+      "epoch": 0.3251,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00017068599112822543,
+      "loss": 1.2754,
+      "step": 16255
+    },
+    {
+      "epoch": 0.3252,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001706612955249225,
+      "loss": 1.2754,
+      "step": 16260
+    },
+    {
+      "epoch": 0.3253,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.000170636591311755,
+      "loss": 1.2675,
+      "step": 16265
+    },
+    {
+      "epoch": 0.3254,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00017061187849173317,
+      "loss": 1.2764,
+      "step": 16270
+    },
+    {
+      "epoch": 0.3255,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001705871570678681,
+      "loss": 1.2666,
+      "step": 16275
+    },
+    {
+      "epoch": 0.3256,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001705624270431721,
+      "loss": 1.2755,
+      "step": 16280
+    },
+    {
+      "epoch": 0.3257,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017053768842065833,
+      "loss": 1.2733,
+      "step": 16285
+    },
+    {
+      "epoch": 0.3258,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00017051294120334125,
+      "loss": 1.2656,
+      "step": 16290
+    },
+    {
+      "epoch": 0.3259,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00017048818539423615,
+      "loss": 1.2868,
+      "step": 16295
+    },
+    {
+      "epoch": 0.326,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00017046342099635948,
+      "loss": 1.2672,
+      "step": 16300
+    },
+    {
+      "epoch": 0.3261,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017043864801272868,
+      "loss": 1.2774,
+      "step": 16305
+    },
+    {
+      "epoch": 0.3262,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001704138664463623,
+      "loss": 1.2901,
+      "step": 16310
+    },
+    {
+      "epoch": 0.3263,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00017038907630027988,
+      "loss": 1.2794,
+      "step": 16315
+    },
+    {
+      "epoch": 0.3264,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00017036427757750205,
+      "loss": 1.2678,
+      "step": 16320
+    },
+    {
+      "epoch": 0.3265,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00017033947028105039,
+      "loss": 1.2872,
+      "step": 16325
+    },
+    {
+      "epoch": 0.3266,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00017031465441394767,
+      "loss": 1.2769,
+      "step": 16330
+    },
+    {
+      "epoch": 0.3267,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017028982997921758,
+      "loss": 1.2607,
+      "step": 16335
+    },
+    {
+      "epoch": 0.3268,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00017026499697988493,
+      "loss": 1.287,
+      "step": 16340
+    },
+    {
+      "epoch": 0.3269,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001702401554189755,
+      "loss": 1.2676,
+      "step": 16345
+    },
+    {
+      "epoch": 0.327,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00017021530529951625,
+      "loss": 1.2706,
+      "step": 16350
+    },
+    {
+      "epoch": 0.3271,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.000170190446624535,
+      "loss": 1.2724,
+      "step": 16355
+    },
+    {
+      "epoch": 0.3272,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00017016557939706075,
+      "loss": 1.2712,
+      "step": 16360
+    },
+    {
+      "epoch": 0.3273,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017014070362012348,
+      "loss": 1.2849,
+      "step": 16365
+    },
+    {
+      "epoch": 0.3274,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00017011581929675425,
+      "loss": 1.2787,
+      "step": 16370
+    },
+    {
+      "epoch": 0.3275,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001700909264299851,
+      "loss": 1.2554,
+      "step": 16375
+    },
+    {
+      "epoch": 0.3276,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00017006602502284918,
+      "loss": 1.2823,
+      "step": 16380
+    },
+    {
+      "epoch": 0.3277,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017004111507838064,
+      "loss": 1.2692,
+      "step": 16385
+    },
+    {
+      "epoch": 0.3278,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017001619659961467,
+      "loss": 1.2933,
+      "step": 16390
+    },
+    {
+      "epoch": 0.3279,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00016999126958958756,
+      "loss": 1.2723,
+      "step": 16395
+    },
+    {
+      "epoch": 0.328,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00016996633405133655,
+      "loss": 1.2635,
+      "step": 16400
+    },
+    {
+      "epoch": 0.3281,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00016994138998789997,
+      "loss": 1.2813,
+      "step": 16405
+    },
+    {
+      "epoch": 0.3282,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00016991643740231716,
+      "loss": 1.2878,
+      "step": 16410
+    },
+    {
+      "epoch": 0.3283,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001698914762976285,
+      "loss": 1.2659,
+      "step": 16415
+    },
+    {
+      "epoch": 0.3284,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00016986650667687552,
+      "loss": 1.2904,
+      "step": 16420
+    },
+    {
+      "epoch": 0.3285,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001698415285431006,
+      "loss": 1.2922,
+      "step": 16425
+    },
+    {
+      "epoch": 0.3286,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00016981654189934727,
+      "loss": 1.2614,
+      "step": 16430
+    },
+    {
+      "epoch": 0.3287,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001697915467486601,
+      "loss": 1.2644,
+      "step": 16435
+    },
+    {
+      "epoch": 0.3288,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00016976654309408464,
+      "loss": 1.2834,
+      "step": 16440
+    },
+    {
+      "epoch": 0.3289,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00016974153093866757,
+      "loss": 1.2779,
+      "step": 16445
+    },
+    {
+      "epoch": 0.329,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00016971651028545648,
+      "loss": 1.2688,
+      "step": 16450
+    },
+    {
+      "epoch": 0.3291,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00016969148113750007,
+      "loss": 1.2786,
+      "step": 16455
+    },
+    {
+      "epoch": 0.3292,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00016966644349784808,
+      "loss": 1.2828,
+      "step": 16460
+    },
+    {
+      "epoch": 0.3293,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001696413973695513,
+      "loss": 1.2738,
+      "step": 16465
+    },
+    {
+      "epoch": 0.3294,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00016961634275566146,
+      "loss": 1.2953,
+      "step": 16470
+    },
+    {
+      "epoch": 0.3295,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00016959127965923142,
+      "loss": 1.2914,
+      "step": 16475
+    },
+    {
+      "epoch": 0.3296,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0001695662080833151,
+      "loss": 1.2552,
+      "step": 16480
+    },
+    {
+      "epoch": 0.3297,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00016954112803096728,
+      "loss": 1.2805,
+      "step": 16485
+    },
+    {
+      "epoch": 0.3298,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.000169516039505244,
+      "loss": 1.2763,
+      "step": 16490
+    },
+    {
+      "epoch": 0.3299,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00016949094250920217,
+      "loss": 1.2859,
+      "step": 16495
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00016946583704589973,
+      "loss": 1.2735,
+      "step": 16500
+    },
+    {
+      "epoch": 0.3301,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00016944072311839581,
+      "loss": 1.2627,
+      "step": 16505
+    },
+    {
+      "epoch": 0.3302,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.0001694156007297504,
+      "loss": 1.2649,
+      "step": 16510
+    },
+    {
+      "epoch": 0.3303,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00016939046988302458,
+      "loss": 1.2754,
+      "step": 16515
+    },
+    {
+      "epoch": 0.3304,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.0001693653305812805,
+      "loss": 1.2809,
+      "step": 16520
+    },
+    {
+      "epoch": 0.3305,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001693401828275813,
+      "loss": 1.2656,
+      "step": 16525
+    },
+    {
+      "epoch": 0.3306,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00016931502662499118,
+      "loss": 1.2586,
+      "step": 16530
+    },
+    {
+      "epoch": 0.3307,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00016928986197657525,
+      "loss": 1.276,
+      "step": 16535
+    },
+    {
+      "epoch": 0.3308,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001692646888853999,
+      "loss": 1.2855,
+      "step": 16540
+    },
+    {
+      "epoch": 0.3309,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016923950735453226,
+      "loss": 1.2728,
+      "step": 16545
+    },
+    {
+      "epoch": 0.331,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001692143173870407,
+      "loss": 1.2705,
+      "step": 16550
+    },
+    {
+      "epoch": 0.3311,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001691891189859945,
+      "loss": 1.2613,
+      "step": 16555
+    },
+    {
+      "epoch": 0.3312,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.000169163912154464,
+      "loss": 1.2671,
+      "step": 16560
+    },
+    {
+      "epoch": 0.3313,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00016913869689552064,
+      "loss": 1.2724,
+      "step": 16565
+    },
+    {
+      "epoch": 0.3314,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001691134732122368,
+      "loss": 1.2637,
+      "step": 16570
+    },
+    {
+      "epoch": 0.3315,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00016908824110768584,
+      "loss": 1.2903,
+      "step": 16575
+    },
+    {
+      "epoch": 0.3316,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00016906300058494228,
+      "loss": 1.2656,
+      "step": 16580
+    },
+    {
+      "epoch": 0.3317,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00016903775164708163,
+      "loss": 1.2662,
+      "step": 16585
+    },
+    {
+      "epoch": 0.3318,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00016901249429718032,
+      "loss": 1.2749,
+      "step": 16590
+    },
+    {
+      "epoch": 0.3319,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00016898722853831593,
+      "loss": 1.2724,
+      "step": 16595
+    },
+    {
+      "epoch": 0.332,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.000168961954373567,
+      "loss": 1.2724,
+      "step": 16600
+    },
+    {
+      "epoch": 0.3321,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00016893667180601312,
+      "loss": 1.2831,
+      "step": 16605
+    },
+    {
+      "epoch": 0.3322,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00016891138083873487,
+      "loss": 1.264,
+      "step": 16610
+    },
+    {
+      "epoch": 0.3323,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00016888608147481388,
+      "loss": 1.2802,
+      "step": 16615
+    },
+    {
+      "epoch": 0.3324,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00016886077371733283,
+      "loss": 1.2805,
+      "step": 16620
+    },
+    {
+      "epoch": 0.3325,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001688354575693754,
+      "loss": 1.2644,
+      "step": 16625
+    },
+    {
+      "epoch": 0.3326,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0001688101330340263,
+      "loss": 1.2851,
+      "step": 16630
+    },
+    {
+      "epoch": 0.3327,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001687848001143711,
+      "loss": 1.2768,
+      "step": 16635
+    },
+    {
+      "epoch": 0.3328,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016875945881349676,
+      "loss": 1.2868,
+      "step": 16640
+    },
+    {
+      "epoch": 0.3329,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00016873410913449091,
+      "loss": 1.2774,
+      "step": 16645
+    },
+    {
+      "epoch": 0.333,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001687087510804423,
+      "loss": 1.2766,
+      "step": 16650
+    },
+    {
+      "epoch": 0.3331,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00016868338465444085,
+      "loss": 1.2802,
+      "step": 16655
+    },
+    {
+      "epoch": 0.3332,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016865800985957726,
+      "loss": 1.2804,
+      "step": 16660
+    },
+    {
+      "epoch": 0.3333,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0001686326266989435,
+      "loss": 1.2596,
+      "step": 16665
+    },
+    {
+      "epoch": 0.3334,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001686072351756323,
+      "loss": 1.265,
+      "step": 16670
+    },
+    {
+      "epoch": 0.3335,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00016858183529273765,
+      "loss": 1.2609,
+      "step": 16675
+    },
+    {
+      "epoch": 0.3336,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00016855642705335437,
+      "loss": 1.2725,
+      "step": 16680
+    },
+    {
+      "epoch": 0.3337,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001685310104605784,
+      "loss": 1.2649,
+      "step": 16685
+    },
+    {
+      "epoch": 0.3338,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001685055855175067,
+      "loss": 1.267,
+      "step": 16690
+    },
+    {
+      "epoch": 0.3339,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001684801522272372,
+      "loss": 1.2709,
+      "step": 16695
+    },
+    {
+      "epoch": 0.334,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00016845471059286887,
+      "loss": 1.2825,
+      "step": 16700
+    },
+    {
+      "epoch": 0.3341,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001684292606175017,
+      "loss": 1.2796,
+      "step": 16705
+    },
+    {
+      "epoch": 0.3342,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00016840380230423668,
+      "loss": 1.2747,
+      "step": 16710
+    },
+    {
+      "epoch": 0.3343,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001683783356561759,
+      "loss": 1.2749,
+      "step": 16715
+    },
+    {
+      "epoch": 0.3344,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00016835286067642228,
+      "loss": 1.2944,
+      "step": 16720
+    },
+    {
+      "epoch": 0.3345,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00016832737736807994,
+      "loss": 1.2746,
+      "step": 16725
+    },
+    {
+      "epoch": 0.3346,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00016830188573425389,
+      "loss": 1.2732,
+      "step": 16730
+    },
+    {
+      "epoch": 0.3347,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00016827638577805026,
+      "loss": 1.273,
+      "step": 16735
+    },
+    {
+      "epoch": 0.3348,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001682508775025762,
+      "loss": 1.2841,
+      "step": 16740
+    },
+    {
+      "epoch": 0.3349,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00016822536091093965,
+      "loss": 1.2937,
+      "step": 16745
+    },
+    {
+      "epoch": 0.335,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00016819983600624986,
+      "loss": 1.2934,
+      "step": 16750
+    },
+    {
+      "epoch": 0.3351,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001681743027916169,
+      "loss": 1.2673,
+      "step": 16755
+    },
+    {
+      "epoch": 0.3352,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.000168148761270152,
+      "loss": 1.2755,
+      "step": 16760
+    },
+    {
+      "epoch": 0.3353,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001681232114449672,
+      "loss": 1.2677,
+      "step": 16765
+    },
+    {
+      "epoch": 0.3354,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00016809765331917575,
+      "loss": 1.2805,
+      "step": 16770
+    },
+    {
+      "epoch": 0.3355,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001680720868958918,
+      "loss": 1.2675,
+      "step": 16775
+    },
+    {
+      "epoch": 0.3356,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00016804651217823053,
+      "loss": 1.2783,
+      "step": 16780
+    },
+    {
+      "epoch": 0.3357,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00016802092916930818,
+      "loss": 1.2763,
+      "step": 16785
+    },
+    {
+      "epoch": 0.3358,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00016799533787224192,
+      "loss": 1.2799,
+      "step": 16790
+    },
+    {
+      "epoch": 0.3359,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00016796973829015,
+      "loss": 1.2868,
+      "step": 16795
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00016794413042615168,
+      "loss": 1.2686,
+      "step": 16800
+    },
+    {
+      "epoch": 0.3361,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00016791851428336711,
+      "loss": 1.2817,
+      "step": 16805
+    },
+    {
+      "epoch": 0.3362,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00016789288986491762,
+      "loss": 1.2932,
+      "step": 16810
+    },
+    {
+      "epoch": 0.3363,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00016786725717392545,
+      "loss": 1.2588,
+      "step": 16815
+    },
+    {
+      "epoch": 0.3364,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00016784161621351382,
+      "loss": 1.2715,
+      "step": 16820
+    },
+    {
+      "epoch": 0.3365,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0001678159669868071,
+      "loss": 1.2674,
+      "step": 16825
+    },
+    {
+      "epoch": 0.3366,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00016779030949693044,
+      "loss": 1.2979,
+      "step": 16830
+    },
+    {
+      "epoch": 0.3367,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00016776464374701025,
+      "loss": 1.272,
+      "step": 16835
+    },
+    {
+      "epoch": 0.3368,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00016773896974017373,
+      "loss": 1.2711,
+      "step": 16840
+    },
+    {
+      "epoch": 0.3369,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00016771328747954925,
+      "loss": 1.2604,
+      "step": 16845
+    },
+    {
+      "epoch": 0.337,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00016768759696826608,
+      "loss": 1.2736,
+      "step": 16850
+    },
+    {
+      "epoch": 0.3371,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00016766189820945456,
+      "loss": 1.3015,
+      "step": 16855
+    },
+    {
+      "epoch": 0.3372,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00016763619120624594,
+      "loss": 1.2751,
+      "step": 16860
+    },
+    {
+      "epoch": 0.3373,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00016761047596177263,
+      "loss": 1.2854,
+      "step": 16865
+    },
+    {
+      "epoch": 0.3374,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00016758475247916787,
+      "loss": 1.2741,
+      "step": 16870
+    },
+    {
+      "epoch": 0.3375,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00016755902076156604,
+      "loss": 1.2687,
+      "step": 16875
+    },
+    {
+      "epoch": 0.3376,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016753328081210245,
+      "loss": 1.2781,
+      "step": 16880
+    },
+    {
+      "epoch": 0.3377,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00016750753263391346,
+      "loss": 1.2769,
+      "step": 16885
+    },
+    {
+      "epoch": 0.3378,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016748177623013638,
+      "loss": 1.2805,
+      "step": 16890
+    },
+    {
+      "epoch": 0.3379,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00016745601160390958,
+      "loss": 1.2726,
+      "step": 16895
+    },
+    {
+      "epoch": 0.338,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00016743023875837233,
+      "loss": 1.2743,
+      "step": 16900
+    },
+    {
+      "epoch": 0.3381,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00016740445769666509,
+      "loss": 1.2686,
+      "step": 16905
+    },
+    {
+      "epoch": 0.3382,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001673786684219291,
+      "loss": 1.2708,
+      "step": 16910
+    },
+    {
+      "epoch": 0.3383,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00016735287093730676,
+      "loss": 1.2711,
+      "step": 16915
+    },
+    {
+      "epoch": 0.3384,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00016732706524594137,
+      "loss": 1.271,
+      "step": 16920
+    },
+    {
+      "epoch": 0.3385,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016730125135097735,
+      "loss": 1.2772,
+      "step": 16925
+    },
+    {
+      "epoch": 0.3386,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00016727542925555996,
+      "loss": 1.2914,
+      "step": 16930
+    },
+    {
+      "epoch": 0.3387,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00016724959896283559,
+      "loss": 1.2739,
+      "step": 16935
+    },
+    {
+      "epoch": 0.3388,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00016722376047595164,
+      "loss": 1.283,
+      "step": 16940
+    },
+    {
+      "epoch": 0.3389,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001671979137980563,
+      "loss": 1.2796,
+      "step": 16945
+    },
+    {
+      "epoch": 0.339,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00016717205893229903,
+      "loss": 1.2653,
+      "step": 16950
+    },
+    {
+      "epoch": 0.3391,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00016714619588183014,
+      "loss": 1.2674,
+      "step": 16955
+    },
+    {
+      "epoch": 0.3392,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00016712032464980095,
+      "loss": 1.2823,
+      "step": 16960
+    },
+    {
+      "epoch": 0.3393,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001670944452393638,
+      "loss": 1.2842,
+      "step": 16965
+    },
+    {
+      "epoch": 0.3394,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.000167068557653672,
+      "loss": 1.2759,
+      "step": 16970
+    },
+    {
+      "epoch": 0.3395,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001670426618958799,
+      "loss": 1.2692,
+      "step": 16975
+    },
+    {
+      "epoch": 0.3396,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00016701675796914286,
+      "loss": 1.2547,
+      "step": 16980
+    },
+    {
+      "epoch": 0.3397,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001669908458766171,
+      "loss": 1.2675,
+      "step": 16985
+    },
+    {
+      "epoch": 0.3398,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00016696492562145996,
+      "loss": 1.2797,
+      "step": 16990
+    },
+    {
+      "epoch": 0.3399,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00016693899720682977,
+      "loss": 1.2799,
+      "step": 16995
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00016691306063588583,
+      "loss": 1.2805,
+      "step": 17000
+    },
+    {
+      "epoch": 0.3401,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00016688711591178842,
+      "loss": 1.2704,
+      "step": 17005
+    },
+    {
+      "epoch": 0.3402,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00016686116303769882,
+      "loss": 1.2715,
+      "step": 17010
+    },
+    {
+      "epoch": 0.3403,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001668352020167793,
+      "loss": 1.2822,
+      "step": 17015
+    },
+    {
+      "epoch": 0.3404,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00016680923285219317,
+      "loss": 1.2689,
+      "step": 17020
+    },
+    {
+      "epoch": 0.3405,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00016678325554710468,
+      "loss": 1.2938,
+      "step": 17025
+    },
+    {
+      "epoch": 0.3406,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016675727010467906,
+      "loss": 1.2714,
+      "step": 17030
+    },
+    {
+      "epoch": 0.3407,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001667312765280826,
+      "loss": 1.2898,
+      "step": 17035
+    },
+    {
+      "epoch": 0.3408,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00016670527482048246,
+      "loss": 1.276,
+      "step": 17040
+    },
+    {
+      "epoch": 0.3409,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00016667926498504696,
+      "loss": 1.2909,
+      "step": 17045
+    },
+    {
+      "epoch": 0.341,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00016665324702494524,
+      "loss": 1.2787,
+      "step": 17050
+    },
+    {
+      "epoch": 0.3411,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001666272209433476,
+      "loss": 1.286,
+      "step": 17055
+    },
+    {
+      "epoch": 0.3412,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00016660118674342517,
+      "loss": 1.2687,
+      "step": 17060
+    },
+    {
+      "epoch": 0.3413,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00016657514442835014,
+      "loss": 1.2934,
+      "step": 17065
+    },
+    {
+      "epoch": 0.3414,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00016654909400129575,
+      "loss": 1.2794,
+      "step": 17070
+    },
+    {
+      "epoch": 0.3415,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00016652303546543608,
+      "loss": 1.2692,
+      "step": 17075
+    },
+    {
+      "epoch": 0.3416,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00016649696882394633,
+      "loss": 1.274,
+      "step": 17080
+    },
+    {
+      "epoch": 0.3417,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00016647089408000266,
+      "loss": 1.2826,
+      "step": 17085
+    },
+    {
+      "epoch": 0.3418,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00016644481123678217,
+      "loss": 1.2883,
+      "step": 17090
+    },
+    {
+      "epoch": 0.3419,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00016641872029746297,
+      "loss": 1.2796,
+      "step": 17095
+    },
+    {
+      "epoch": 0.342,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00016639262126522418,
+      "loss": 1.2833,
+      "step": 17100
+    },
+    {
+      "epoch": 0.3421,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00016636651414324587,
+      "loss": 1.2831,
+      "step": 17105
+    },
+    {
+      "epoch": 0.3422,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00016634039893470912,
+      "loss": 1.2777,
+      "step": 17110
+    },
+    {
+      "epoch": 0.3423,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.000166314275642796,
+      "loss": 1.267,
+      "step": 17115
+    },
+    {
+      "epoch": 0.3424,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00016628814427068953,
+      "loss": 1.2753,
+      "step": 17120
+    },
+    {
+      "epoch": 0.3425,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00016626200482157378,
+      "loss": 1.2961,
+      "step": 17125
+    },
+    {
+      "epoch": 0.3426,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00016623585729863368,
+      "loss": 1.2636,
+      "step": 17130
+    },
+    {
+      "epoch": 0.3427,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00016620970170505534,
+      "loss": 1.2827,
+      "step": 17135
+    },
+    {
+      "epoch": 0.3428,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00016618353804402568,
+      "loss": 1.2761,
+      "step": 17140
+    },
+    {
+      "epoch": 0.3429,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00016615736631873262,
+      "loss": 1.2771,
+      "step": 17145
+    },
+    {
+      "epoch": 0.343,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00016613118653236518,
+      "loss": 1.2679,
+      "step": 17150
+    },
+    {
+      "epoch": 0.3431,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00016610499868811328,
+      "loss": 1.2778,
+      "step": 17155
+    },
+    {
+      "epoch": 0.3432,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00016607880278916777,
+      "loss": 1.281,
+      "step": 17160
+    },
+    {
+      "epoch": 0.3433,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001660525988387206,
+      "loss": 1.2767,
+      "step": 17165
+    },
+    {
+      "epoch": 0.3434,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00016602638683996463,
+      "loss": 1.2879,
+      "step": 17170
+    },
+    {
+      "epoch": 0.3435,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001660001667960937,
+      "loss": 1.2681,
+      "step": 17175
+    },
+    {
+      "epoch": 0.3436,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00016597393871030264,
+      "loss": 1.2894,
+      "step": 17180
+    },
+    {
+      "epoch": 0.3437,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0001659477025857872,
+      "loss": 1.2732,
+      "step": 17185
+    },
+    {
+      "epoch": 0.3438,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00016592145842574433,
+      "loss": 1.2632,
+      "step": 17190
+    },
+    {
+      "epoch": 0.3439,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001658952062333717,
+      "loss": 1.2752,
+      "step": 17195
+    },
+    {
+      "epoch": 0.344,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016586894601186805,
+      "loss": 1.2769,
+      "step": 17200
+    },
+    {
+      "epoch": 0.3441,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00016584267776443317,
+      "loss": 1.2739,
+      "step": 17205
+    },
+    {
+      "epoch": 0.3442,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016581640149426768,
+      "loss": 1.2825,
+      "step": 17210
+    },
+    {
+      "epoch": 0.3443,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00016579011720457333,
+      "loss": 1.3013,
+      "step": 17215
+    },
+    {
+      "epoch": 0.3444,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00016576382489855274,
+      "loss": 1.259,
+      "step": 17220
+    },
+    {
+      "epoch": 0.3445,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001657375245794096,
+      "loss": 1.2735,
+      "step": 17225
+    },
+    {
+      "epoch": 0.3446,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00016571121625034847,
+      "loss": 1.2871,
+      "step": 17230
+    },
+    {
+      "epoch": 0.3447,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016568489991457497,
+      "loss": 1.2805,
+      "step": 17235
+    },
+    {
+      "epoch": 0.3448,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00016565857557529566,
+      "loss": 1.2989,
+      "step": 17240
+    },
+    {
+      "epoch": 0.3449,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016563224323571806,
+      "loss": 1.2662,
+      "step": 17245
+    },
+    {
+      "epoch": 0.345,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00016560590289905073,
+      "loss": 1.2742,
+      "step": 17250
+    },
+    {
+      "epoch": 0.3451,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00016557955456850313,
+      "loss": 1.2879,
+      "step": 17255
+    },
+    {
+      "epoch": 0.3452,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00016555319824728575,
+      "loss": 1.282,
+      "step": 17260
+    },
+    {
+      "epoch": 0.3453,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00016552683393860997,
+      "loss": 1.28,
+      "step": 17265
+    },
+    {
+      "epoch": 0.3454,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00016550046164568827,
+      "loss": 1.2747,
+      "step": 17270
+    },
+    {
+      "epoch": 0.3455,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00016547408137173396,
+      "loss": 1.271,
+      "step": 17275
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00016544769311996148,
+      "loss": 1.2738,
+      "step": 17280
+    },
+    {
+      "epoch": 0.3457,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00016542129689358612,
+      "loss": 1.2827,
+      "step": 17285
+    },
+    {
+      "epoch": 0.3458,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00016539489269582416,
+      "loss": 1.276,
+      "step": 17290
+    },
+    {
+      "epoch": 0.3459,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00016536848052989291,
+      "loss": 1.2659,
+      "step": 17295
+    },
+    {
+      "epoch": 0.346,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00016534206039901057,
+      "loss": 1.2754,
+      "step": 17300
+    },
+    {
+      "epoch": 0.3461,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00016531563230639637,
+      "loss": 1.2846,
+      "step": 17305
+    },
+    {
+      "epoch": 0.3462,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001652891962552705,
+      "loss": 1.2779,
+      "step": 17310
+    },
+    {
+      "epoch": 0.3463,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00016526275224885411,
+      "loss": 1.2727,
+      "step": 17315
+    },
+    {
+      "epoch": 0.3464,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00016523630029036931,
+      "loss": 1.304,
+      "step": 17320
+    },
+    {
+      "epoch": 0.3465,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00016520984038303924,
+      "loss": 1.2671,
+      "step": 17325
+    },
+    {
+      "epoch": 0.3466,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0001651833725300879,
+      "loss": 1.2665,
+      "step": 17330
+    },
+    {
+      "epoch": 0.3467,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016515689673474033,
+      "loss": 1.2886,
+      "step": 17335
+    },
+    {
+      "epoch": 0.3468,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00016513041300022255,
+      "loss": 1.2824,
+      "step": 17340
+    },
+    {
+      "epoch": 0.3469,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0001651039213297615,
+      "loss": 1.2744,
+      "step": 17345
+    },
+    {
+      "epoch": 0.347,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001650774217265851,
+      "loss": 1.2727,
+      "step": 17350
+    },
+    {
+      "epoch": 0.3471,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016505091419392228,
+      "loss": 1.2665,
+      "step": 17355
+    },
+    {
+      "epoch": 0.3472,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00016502439873500289,
+      "loss": 1.2581,
+      "step": 17360
+    },
+    {
+      "epoch": 0.3473,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00016499787535305776,
+      "loss": 1.267,
+      "step": 17365
+    },
+    {
+      "epoch": 0.3474,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016497134405131866,
+      "loss": 1.2742,
+      "step": 17370
+    },
+    {
+      "epoch": 0.3475,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00016494480483301836,
+      "loss": 1.2747,
+      "step": 17375
+    },
+    {
+      "epoch": 0.3476,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00016491825770139062,
+      "loss": 1.2833,
+      "step": 17380
+    },
+    {
+      "epoch": 0.3477,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016489170265967008,
+      "loss": 1.272,
+      "step": 17385
+    },
+    {
+      "epoch": 0.3478,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016486513971109243,
+      "loss": 1.2787,
+      "step": 17390
+    },
+    {
+      "epoch": 0.3479,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.0001648385688588942,
+      "loss": 1.29,
+      "step": 17395
+    },
+    {
+      "epoch": 0.348,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001648119901063131,
+      "loss": 1.274,
+      "step": 17400
+    },
+    {
+      "epoch": 0.3481,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00016478540345658759,
+      "loss": 1.2597,
+      "step": 17405
+    },
+    {
+      "epoch": 0.3482,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00016475880891295716,
+      "loss": 1.271,
+      "step": 17410
+    },
+    {
+      "epoch": 0.3483,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001647322064786623,
+      "loss": 1.2659,
+      "step": 17415
+    },
+    {
+      "epoch": 0.3484,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00016470559615694446,
+      "loss": 1.2842,
+      "step": 17420
+    },
+    {
+      "epoch": 0.3485,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.000164678977951046,
+      "loss": 1.2826,
+      "step": 17425
+    },
+    {
+      "epoch": 0.3486,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0001646523518642102,
+      "loss": 1.2658,
+      "step": 17430
+    },
+    {
+      "epoch": 0.3487,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00016462571789968152,
+      "loss": 1.2674,
+      "step": 17435
+    },
+    {
+      "epoch": 0.3488,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001645990760607051,
+      "loss": 1.2902,
+      "step": 17440
+    },
+    {
+      "epoch": 0.3489,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00016457242635052724,
+      "loss": 1.2783,
+      "step": 17445
+    },
+    {
+      "epoch": 0.349,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00016454576877239507,
+      "loss": 1.2672,
+      "step": 17450
+    },
+    {
+      "epoch": 0.3491,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00016451910332955679,
+      "loss": 1.2708,
+      "step": 17455
+    },
+    {
+      "epoch": 0.3492,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00016449243002526144,
+      "loss": 1.2701,
+      "step": 17460
+    },
+    {
+      "epoch": 0.3493,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00016446574886275913,
+      "loss": 1.2593,
+      "step": 17465
+    },
+    {
+      "epoch": 0.3494,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.0001644390598453009,
+      "loss": 1.2886,
+      "step": 17470
+    },
+    {
+      "epoch": 0.3495,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00016441236297613866,
+      "loss": 1.278,
+      "step": 17475
+    },
+    {
+      "epoch": 0.3496,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001643856582585254,
+      "loss": 1.2793,
+      "step": 17480
+    },
+    {
+      "epoch": 0.3497,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016435894569571496,
+      "loss": 1.2836,
+      "step": 17485
+    },
+    {
+      "epoch": 0.3498,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001643322252909622,
+      "loss": 1.2658,
+      "step": 17490
+    },
+    {
+      "epoch": 0.3499,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016430549704752294,
+      "loss": 1.2795,
+      "step": 17495
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00016427876096865394,
+      "loss": 1.2835,
+      "step": 17500
+    },
+    {
+      "epoch": 0.3501,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00016425201705761288,
+      "loss": 1.2753,
+      "step": 17505
+    },
+    {
+      "epoch": 0.3502,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00016422526531765846,
+      "loss": 1.2945,
+      "step": 17510
+    },
+    {
+      "epoch": 0.3503,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00016419850575205024,
+      "loss": 1.2847,
+      "step": 17515
+    },
+    {
+      "epoch": 0.3504,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00016417173836404887,
+      "loss": 1.2751,
+      "step": 17520
+    },
+    {
+      "epoch": 0.3505,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00016414496315691581,
+      "loss": 1.2698,
+      "step": 17525
+    },
+    {
+      "epoch": 0.3506,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00016411818013391355,
+      "loss": 1.2723,
+      "step": 17530
+    },
+    {
+      "epoch": 0.3507,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00016409138929830553,
+      "loss": 1.2672,
+      "step": 17535
+    },
+    {
+      "epoch": 0.3508,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00016406459065335615,
+      "loss": 1.2673,
+      "step": 17540
+    },
+    {
+      "epoch": 0.3509,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00016403778420233075,
+      "loss": 1.2746,
+      "step": 17545
+    },
+    {
+      "epoch": 0.351,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00016401096994849557,
+      "loss": 1.2729,
+      "step": 17550
+    },
+    {
+      "epoch": 0.3511,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00016398414789511786,
+      "loss": 1.2866,
+      "step": 17555
+    },
+    {
+      "epoch": 0.3512,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0001639573180454658,
+      "loss": 1.2735,
+      "step": 17560
+    },
+    {
+      "epoch": 0.3513,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00016393048040280855,
+      "loss": 1.2864,
+      "step": 17565
+    },
+    {
+      "epoch": 0.3514,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00016390363497041622,
+      "loss": 1.2763,
+      "step": 17570
+    },
+    {
+      "epoch": 0.3515,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00016387678175155978,
+      "loss": 1.2696,
+      "step": 17575
+    },
+    {
+      "epoch": 0.3516,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00016384992074951123,
+      "loss": 1.2638,
+      "step": 17580
+    },
+    {
+      "epoch": 0.3517,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00016382305196754356,
+      "loss": 1.268,
+      "step": 17585
+    },
+    {
+      "epoch": 0.3518,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00016379617540893056,
+      "loss": 1.273,
+      "step": 17590
+    },
+    {
+      "epoch": 0.3519,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001637692910769471,
+      "loss": 1.2586,
+      "step": 17595
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.000163742398974869,
+      "loss": 1.2563,
+      "step": 17600
+    },
+    {
+      "epoch": 0.3521,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00016371549910597287,
+      "loss": 1.2588,
+      "step": 17605
+    },
+    {
+      "epoch": 0.3522,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001636885914735365,
+      "loss": 1.2841,
+      "step": 17610
+    },
+    {
+      "epoch": 0.3523,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016366167608083843,
+      "loss": 1.2859,
+      "step": 17615
+    },
+    {
+      "epoch": 0.3524,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00016363475293115824,
+      "loss": 1.2887,
+      "step": 17620
+    },
+    {
+      "epoch": 0.3525,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001636078220277764,
+      "loss": 1.263,
+      "step": 17625
+    },
+    {
+      "epoch": 0.3526,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00016358088337397442,
+      "loss": 1.2706,
+      "step": 17630
+    },
+    {
+      "epoch": 0.3527,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00016355393697303465,
+      "loss": 1.2572,
+      "step": 17635
+    },
+    {
+      "epoch": 0.3528,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016352698282824044,
+      "loss": 1.2654,
+      "step": 17640
+    },
+    {
+      "epoch": 0.3529,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00016350002094287609,
+      "loss": 1.2826,
+      "step": 17645
+    },
+    {
+      "epoch": 0.353,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00016347305132022677,
+      "loss": 1.294,
+      "step": 17650
+    },
+    {
+      "epoch": 0.3531,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001634460739635787,
+      "loss": 1.2589,
+      "step": 17655
+    },
+    {
+      "epoch": 0.3532,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00016341908887621895,
+      "loss": 1.2697,
+      "step": 17660
+    },
+    {
+      "epoch": 0.3533,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00016339209606143563,
+      "loss": 1.2763,
+      "step": 17665
+    },
+    {
+      "epoch": 0.3534,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016336509552251766,
+      "loss": 1.2638,
+      "step": 17670
+    },
+    {
+      "epoch": 0.3535,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.000163338087262755,
+      "loss": 1.2688,
+      "step": 17675
+    },
+    {
+      "epoch": 0.3536,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00016331107128543857,
+      "loss": 1.2732,
+      "step": 17680
+    },
+    {
+      "epoch": 0.3537,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00016328404759386014,
+      "loss": 1.2805,
+      "step": 17685
+    },
+    {
+      "epoch": 0.3538,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00016325701619131246,
+      "loss": 1.2839,
+      "step": 17690
+    },
+    {
+      "epoch": 0.3539,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00016322997708108922,
+      "loss": 1.2706,
+      "step": 17695
+    },
+    {
+      "epoch": 0.354,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001632029302664851,
+      "loss": 1.2646,
+      "step": 17700
+    },
+    {
+      "epoch": 0.3541,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00016317587575079563,
+      "loss": 1.2798,
+      "step": 17705
+    },
+    {
+      "epoch": 0.3542,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00016314881353731732,
+      "loss": 1.256,
+      "step": 17710
+    },
+    {
+      "epoch": 0.3543,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016312174362934765,
+      "loss": 1.2847,
+      "step": 17715
+    },
+    {
+      "epoch": 0.3544,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016309466603018496,
+      "loss": 1.2766,
+      "step": 17720
+    },
+    {
+      "epoch": 0.3545,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00016306758074312864,
+      "loss": 1.2828,
+      "step": 17725
+    },
+    {
+      "epoch": 0.3546,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0001630404877714789,
+      "loss": 1.2927,
+      "step": 17730
+    },
+    {
+      "epoch": 0.3547,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00016301338711853693,
+      "loss": 1.2847,
+      "step": 17735
+    },
+    {
+      "epoch": 0.3548,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00016298627878760487,
+      "loss": 1.2644,
+      "step": 17740
+    },
+    {
+      "epoch": 0.3549,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00016295916278198584,
+      "loss": 1.272,
+      "step": 17745
+    },
+    {
+      "epoch": 0.355,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00016293203910498376,
+      "loss": 1.2635,
+      "step": 17750
+    },
+    {
+      "epoch": 0.3551,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001629049077599036,
+      "loss": 1.269,
+      "step": 17755
+    },
+    {
+      "epoch": 0.3552,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001628777687500513,
+      "loss": 1.2611,
+      "step": 17760
+    },
+    {
+      "epoch": 0.3553,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00016285062207873355,
+      "loss": 1.3011,
+      "step": 17765
+    },
+    {
+      "epoch": 0.3554,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00016282346774925817,
+      "loss": 1.2647,
+      "step": 17770
+    },
+    {
+      "epoch": 0.3555,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00016279630576493382,
+      "loss": 1.285,
+      "step": 17775
+    },
+    {
+      "epoch": 0.3556,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00016276913612907007,
+      "loss": 1.2873,
+      "step": 17780
+    },
+    {
+      "epoch": 0.3557,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001627419588449775,
+      "loss": 1.2679,
+      "step": 17785
+    },
+    {
+      "epoch": 0.3558,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016271477391596753,
+      "loss": 1.2674,
+      "step": 17790
+    },
+    {
+      "epoch": 0.3559,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0001626875813453526,
+      "loss": 1.279,
+      "step": 17795
+    },
+    {
+      "epoch": 0.356,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00016266038113644607,
+      "loss": 1.2737,
+      "step": 17800
+    },
+    {
+      "epoch": 0.3561,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00016263317329256213,
+      "loss": 1.2746,
+      "step": 17805
+    },
+    {
+      "epoch": 0.3562,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00016260595781701604,
+      "loss": 1.2723,
+      "step": 17810
+    },
+    {
+      "epoch": 0.3563,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001625787347131239,
+      "loss": 1.2839,
+      "step": 17815
+    },
+    {
+      "epoch": 0.3564,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00016255150398420271,
+      "loss": 1.2871,
+      "step": 17820
+    },
+    {
+      "epoch": 0.3565,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00016252426563357055,
+      "loss": 1.2643,
+      "step": 17825
+    },
+    {
+      "epoch": 0.3566,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00016249701966454625,
+      "loss": 1.2741,
+      "step": 17830
+    },
+    {
+      "epoch": 0.3567,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0001624697660804497,
+      "loss": 1.2763,
+      "step": 17835
+    },
+    {
+      "epoch": 0.3568,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016244250488460158,
+      "loss": 1.2808,
+      "step": 17840
+    },
+    {
+      "epoch": 0.3569,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00016241523608032373,
+      "loss": 1.2855,
+      "step": 17845
+    },
+    {
+      "epoch": 0.357,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00016238795967093864,
+      "loss": 1.2828,
+      "step": 17850
+    },
+    {
+      "epoch": 0.3571,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00016236067565976992,
+      "loss": 1.2771,
+      "step": 17855
+    },
+    {
+      "epoch": 0.3572,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016233338405014202,
+      "loss": 1.2766,
+      "step": 17860
+    },
+    {
+      "epoch": 0.3573,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016230608484538034,
+      "loss": 1.2758,
+      "step": 17865
+    },
+    {
+      "epoch": 0.3574,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016227877804881127,
+      "loss": 1.2643,
+      "step": 17870
+    },
+    {
+      "epoch": 0.3575,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00016225146366376198,
+      "loss": 1.2571,
+      "step": 17875
+    },
+    {
+      "epoch": 0.3576,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00016222414169356065,
+      "loss": 1.2686,
+      "step": 17880
+    },
+    {
+      "epoch": 0.3577,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00016219681214153643,
+      "loss": 1.279,
+      "step": 17885
+    },
+    {
+      "epoch": 0.3578,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.0001621694750110193,
+      "loss": 1.2849,
+      "step": 17890
+    },
+    {
+      "epoch": 0.3579,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001621421303053402,
+      "loss": 1.2824,
+      "step": 17895
+    },
+    {
+      "epoch": 0.358,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00016211477802783103,
+      "loss": 1.2467,
+      "step": 17900
+    },
+    {
+      "epoch": 0.3581,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001620874181818246,
+      "loss": 1.2774,
+      "step": 17905
+    },
+    {
+      "epoch": 0.3582,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016206005077065458,
+      "loss": 1.2815,
+      "step": 17910
+    },
+    {
+      "epoch": 0.3583,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00016203267579765563,
+      "loss": 1.2768,
+      "step": 17915
+    },
+    {
+      "epoch": 0.3584,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00016200529326616328,
+      "loss": 1.2614,
+      "step": 17920
+    },
+    {
+      "epoch": 0.3585,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00016197790317951403,
+      "loss": 1.27,
+      "step": 17925
+    },
+    {
+      "epoch": 0.3586,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00016195050554104528,
+      "loss": 1.2821,
+      "step": 17930
+    },
+    {
+      "epoch": 0.3587,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00016192310035409536,
+      "loss": 1.2778,
+      "step": 17935
+    },
+    {
+      "epoch": 0.3588,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00016189568762200348,
+      "loss": 1.2776,
+      "step": 17940
+    },
+    {
+      "epoch": 0.3589,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016186826734810979,
+      "loss": 1.27,
+      "step": 17945
+    },
+    {
+      "epoch": 0.359,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0001618408395357554,
+      "loss": 1.2699,
+      "step": 17950
+    },
+    {
+      "epoch": 0.3591,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00016181340418828233,
+      "loss": 1.2664,
+      "step": 17955
+    },
+    {
+      "epoch": 0.3592,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016178596130903344,
+      "loss": 1.2859,
+      "step": 17960
+    },
+    {
+      "epoch": 0.3593,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0001617585109013526,
+      "loss": 1.2814,
+      "step": 17965
+    },
+    {
+      "epoch": 0.3594,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00016173105296858452,
+      "loss": 1.2761,
+      "step": 17970
+    },
+    {
+      "epoch": 0.3595,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00016170358751407487,
+      "loss": 1.2731,
+      "step": 17975
+    },
+    {
+      "epoch": 0.3596,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00016167611454117025,
+      "loss": 1.28,
+      "step": 17980
+    },
+    {
+      "epoch": 0.3597,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001616486340532182,
+      "loss": 1.2679,
+      "step": 17985
+    },
+    {
+      "epoch": 0.3598,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00016162114605356703,
+      "loss": 1.2686,
+      "step": 17990
+    },
+    {
+      "epoch": 0.3599,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0001615936505455662,
+      "loss": 1.2936,
+      "step": 17995
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001615661475325658,
+      "loss": 1.2739,
+      "step": 18000
+    },
+    {
+      "epoch": 0.3601,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00016153863701791717,
+      "loss": 1.2801,
+      "step": 18005
+    },
+    {
+      "epoch": 0.3602,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00016151111900497225,
+      "loss": 1.2776,
+      "step": 18010
+    },
+    {
+      "epoch": 0.3603,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00016148359349708402,
+      "loss": 1.2781,
+      "step": 18015
+    },
+    {
+      "epoch": 0.3604,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00016145606049760644,
+      "loss": 1.2769,
+      "step": 18020
+    },
+    {
+      "epoch": 0.3605,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016142852000989433,
+      "loss": 1.269,
+      "step": 18025
+    },
+    {
+      "epoch": 0.3606,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00016140097203730337,
+      "loss": 1.288,
+      "step": 18030
+    },
+    {
+      "epoch": 0.3607,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00016137341658319023,
+      "loss": 1.2688,
+      "step": 18035
+    },
+    {
+      "epoch": 0.3608,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00016134585365091243,
+      "loss": 1.2712,
+      "step": 18040
+    },
+    {
+      "epoch": 0.3609,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00016131828324382846,
+      "loss": 1.2785,
+      "step": 18045
+    },
+    {
+      "epoch": 0.361,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00016129070536529766,
+      "loss": 1.2893,
+      "step": 18050
+    },
+    {
+      "epoch": 0.3611,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00016126312001868033,
+      "loss": 1.2802,
+      "step": 18055
+    },
+    {
+      "epoch": 0.3612,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00016123552720733765,
+      "loss": 1.2745,
+      "step": 18060
+    },
+    {
+      "epoch": 0.3613,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016120792693463174,
+      "loss": 1.2817,
+      "step": 18065
+    },
+    {
+      "epoch": 0.3614,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00016118031920392558,
+      "loss": 1.2691,
+      "step": 18070
+    },
+    {
+      "epoch": 0.3615,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00016115270401858314,
+      "loss": 1.267,
+      "step": 18075
+    },
+    {
+      "epoch": 0.3616,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00016112508138196917,
+      "loss": 1.2787,
+      "step": 18080
+    },
+    {
+      "epoch": 0.3617,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016109745129744946,
+      "loss": 1.2832,
+      "step": 18085
+    },
+    {
+      "epoch": 0.3618,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00016106981376839066,
+      "loss": 1.2822,
+      "step": 18090
+    },
+    {
+      "epoch": 0.3619,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00016104216879816026,
+      "loss": 1.2688,
+      "step": 18095
+    },
+    {
+      "epoch": 0.362,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001610145163901268,
+      "loss": 1.2713,
+      "step": 18100
+    },
+    {
+      "epoch": 0.3621,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00016098685654765955,
+      "loss": 1.2825,
+      "step": 18105
+    },
+    {
+      "epoch": 0.3622,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001609591892741288,
+      "loss": 1.2769,
+      "step": 18110
+    },
+    {
+      "epoch": 0.3623,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001609315145729058,
+      "loss": 1.2686,
+      "step": 18115
+    },
+    {
+      "epoch": 0.3624,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00016090383244736256,
+      "loss": 1.2638,
+      "step": 18120
+    },
+    {
+      "epoch": 0.3625,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00016087614290087208,
+      "loss": 1.2612,
+      "step": 18125
+    },
+    {
+      "epoch": 0.3626,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001608484459368082,
+      "loss": 1.2695,
+      "step": 18130
+    },
+    {
+      "epoch": 0.3627,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00016082074155854582,
+      "loss": 1.2653,
+      "step": 18135
+    },
+    {
+      "epoch": 0.3628,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00016079302976946055,
+      "loss": 1.2809,
+      "step": 18140
+    },
+    {
+      "epoch": 0.3629,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.000160765310572929,
+      "loss": 1.2623,
+      "step": 18145
+    },
+    {
+      "epoch": 0.363,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00016073758397232868,
+      "loss": 1.2627,
+      "step": 18150
+    },
+    {
+      "epoch": 0.3631,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.000160709849971038,
+      "loss": 1.2875,
+      "step": 18155
+    },
+    {
+      "epoch": 0.3632,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00016068210857243624,
+      "loss": 1.2739,
+      "step": 18160
+    },
+    {
+      "epoch": 0.3633,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0001606543597799036,
+      "loss": 1.2746,
+      "step": 18165
+    },
+    {
+      "epoch": 0.3634,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00016062660359682124,
+      "loss": 1.2617,
+      "step": 18170
+    },
+    {
+      "epoch": 0.3635,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.0001605988400265711,
+      "loss": 1.2669,
+      "step": 18175
+    },
+    {
+      "epoch": 0.3636,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00016057106907253616,
+      "loss": 1.2625,
+      "step": 18180
+    },
+    {
+      "epoch": 0.3637,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00016054329073810015,
+      "loss": 1.2807,
+      "step": 18185
+    },
+    {
+      "epoch": 0.3638,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001605155050266478,
+      "loss": 1.2784,
+      "step": 18190
+    },
+    {
+      "epoch": 0.3639,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00016048771194156477,
+      "loss": 1.266,
+      "step": 18195
+    },
+    {
+      "epoch": 0.364,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001604599114862375,
+      "loss": 1.2821,
+      "step": 18200
+    },
+    {
+      "epoch": 0.3641,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0001604321036640534,
+      "loss": 1.265,
+      "step": 18205
+    },
+    {
+      "epoch": 0.3642,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00016040428847840079,
+      "loss": 1.275,
+      "step": 18210
+    },
+    {
+      "epoch": 0.3643,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00016037646593266883,
+      "loss": 1.2733,
+      "step": 18215
+    },
+    {
+      "epoch": 0.3644,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00016034863603024767,
+      "loss": 1.2837,
+      "step": 18220
+    },
+    {
+      "epoch": 0.3645,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00016032079877452825,
+      "loss": 1.29,
+      "step": 18225
+    },
+    {
+      "epoch": 0.3646,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00016029295416890248,
+      "loss": 1.2859,
+      "step": 18230
+    },
+    {
+      "epoch": 0.3647,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00016026510221676311,
+      "loss": 1.2716,
+      "step": 18235
+    },
+    {
+      "epoch": 0.3648,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00016023724292150385,
+      "loss": 1.2601,
+      "step": 18240
+    },
+    {
+      "epoch": 0.3649,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00016020937628651927,
+      "loss": 1.2607,
+      "step": 18245
+    },
+    {
+      "epoch": 0.365,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00016018150231520486,
+      "loss": 1.2663,
+      "step": 18250
+    },
+    {
+      "epoch": 0.3651,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001601536210109569,
+      "loss": 1.2797,
+      "step": 18255
+    },
+    {
+      "epoch": 0.3652,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001601257323771727,
+      "loss": 1.2841,
+      "step": 18260
+    },
+    {
+      "epoch": 0.3653,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00016009783641725034,
+      "loss": 1.2699,
+      "step": 18265
+    },
+    {
+      "epoch": 0.3654,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00016006993313458896,
+      "loss": 1.2662,
+      "step": 18270
+    },
+    {
+      "epoch": 0.3655,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00016004202253258842,
+      "loss": 1.2654,
+      "step": 18275
+    },
+    {
+      "epoch": 0.3656,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00016001410461464956,
+      "loss": 1.2724,
+      "step": 18280
+    },
+    {
+      "epoch": 0.3657,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001599861793841741,
+      "loss": 1.2734,
+      "step": 18285
+    },
+    {
+      "epoch": 0.3658,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00015995824684456465,
+      "loss": 1.2846,
+      "step": 18290
+    },
+    {
+      "epoch": 0.3659,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00015993030699922468,
+      "loss": 1.2824,
+      "step": 18295
+    },
+    {
+      "epoch": 0.366,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001599023598515586,
+      "loss": 1.2746,
+      "step": 18300
+    },
+    {
+      "epoch": 0.3661,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00015987440540497167,
+      "loss": 1.2677,
+      "step": 18305
+    },
+    {
+      "epoch": 0.3662,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00015984644366287006,
+      "loss": 1.2753,
+      "step": 18310
+    },
+    {
+      "epoch": 0.3663,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00015981847462866083,
+      "loss": 1.2761,
+      "step": 18315
+    },
+    {
+      "epoch": 0.3664,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001597904983057519,
+      "loss": 1.2859,
+      "step": 18320
+    },
+    {
+      "epoch": 0.3665,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00015976251469755214,
+      "loss": 1.2825,
+      "step": 18325
+    },
+    {
+      "epoch": 0.3666,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015973452380747122,
+      "loss": 1.263,
+      "step": 18330
+    },
+    {
+      "epoch": 0.3667,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00015970652563891978,
+      "loss": 1.2641,
+      "step": 18335
+    },
+    {
+      "epoch": 0.3668,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00015967852019530926,
+      "loss": 1.2742,
+      "step": 18340
+    },
+    {
+      "epoch": 0.3669,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00015965050748005215,
+      "loss": 1.2718,
+      "step": 18345
+    },
+    {
+      "epoch": 0.367,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001596224874965616,
+      "loss": 1.2681,
+      "step": 18350
+    },
+    {
+      "epoch": 0.3671,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001595944602482518,
+      "loss": 1.2541,
+      "step": 18355
+    },
+    {
+      "epoch": 0.3672,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001595664257385378,
+      "loss": 1.2624,
+      "step": 18360
+    },
+    {
+      "epoch": 0.3673,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00015953838397083552,
+      "loss": 1.2788,
+      "step": 18365
+    },
+    {
+      "epoch": 0.3674,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00015951033494856175,
+      "loss": 1.2693,
+      "step": 18370
+    },
+    {
+      "epoch": 0.3675,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00015948227867513415,
+      "loss": 1.2737,
+      "step": 18375
+    },
+    {
+      "epoch": 0.3676,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00015945421515397133,
+      "loss": 1.2917,
+      "step": 18380
+    },
+    {
+      "epoch": 0.3677,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00015942614438849275,
+      "loss": 1.2767,
+      "step": 18385
+    },
+    {
+      "epoch": 0.3678,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001593980663821187,
+      "loss": 1.2795,
+      "step": 18390
+    },
+    {
+      "epoch": 0.3679,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00015936998113827048,
+      "loss": 1.2754,
+      "step": 18395
+    },
+    {
+      "epoch": 0.368,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00015934188866037016,
+      "loss": 1.2673,
+      "step": 18400
+    },
+    {
+      "epoch": 0.3681,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001593137889518407,
+      "loss": 1.2805,
+      "step": 18405
+    },
+    {
+      "epoch": 0.3682,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00015928568201610595,
+      "loss": 1.277,
+      "step": 18410
+    },
+    {
+      "epoch": 0.3683,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00015925756785659069,
+      "loss": 1.2846,
+      "step": 18415
+    },
+    {
+      "epoch": 0.3684,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00015922944647672052,
+      "loss": 1.2697,
+      "step": 18420
+    },
+    {
+      "epoch": 0.3685,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00015920131787992197,
+      "loss": 1.2882,
+      "step": 18425
+    },
+    {
+      "epoch": 0.3686,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0001591731820696224,
+      "loss": 1.2702,
+      "step": 18430
+    },
+    {
+      "epoch": 0.3687,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0001591450390492501,
+      "loss": 1.2701,
+      "step": 18435
+    },
+    {
+      "epoch": 0.3688,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001591168888222342,
+      "loss": 1.2746,
+      "step": 18440
+    },
+    {
+      "epoch": 0.3689,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00015908873139200473,
+      "loss": 1.27,
+      "step": 18445
+    },
+    {
+      "epoch": 0.369,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00015906056676199255,
+      "loss": 1.2761,
+      "step": 18450
+    },
+    {
+      "epoch": 0.3691,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015903239493562948,
+      "loss": 1.2809,
+      "step": 18455
+    },
+    {
+      "epoch": 0.3692,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00015900421591634814,
+      "loss": 1.2606,
+      "step": 18460
+    },
+    {
+      "epoch": 0.3693,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00015897602970758206,
+      "loss": 1.277,
+      "step": 18465
+    },
+    {
+      "epoch": 0.3694,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00015894783631276567,
+      "loss": 1.2848,
+      "step": 18470
+    },
+    {
+      "epoch": 0.3695,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001589196357353342,
+      "loss": 1.262,
+      "step": 18475
+    },
+    {
+      "epoch": 0.3696,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00015889142797872387,
+      "loss": 1.2771,
+      "step": 18480
+    },
+    {
+      "epoch": 0.3697,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001588632130463717,
+      "loss": 1.2658,
+      "step": 18485
+    },
+    {
+      "epoch": 0.3698,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00015883499094171554,
+      "loss": 1.2606,
+      "step": 18490
+    },
+    {
+      "epoch": 0.3699,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0001588067616681942,
+      "loss": 1.2618,
+      "step": 18495
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00015877852522924732,
+      "loss": 1.2699,
+      "step": 18500
+    },
+    {
+      "epoch": 0.3701,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00015875028162831546,
+      "loss": 1.2709,
+      "step": 18505
+    },
+    {
+      "epoch": 0.3702,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015872203086883996,
+      "loss": 1.2559,
+      "step": 18510
+    },
+    {
+      "epoch": 0.3703,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00015869377295426316,
+      "loss": 1.3023,
+      "step": 18515
+    },
+    {
+      "epoch": 0.3704,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00015866550788802813,
+      "loss": 1.2863,
+      "step": 18520
+    },
+    {
+      "epoch": 0.3705,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00015863723567357892,
+      "loss": 1.2842,
+      "step": 18525
+    },
+    {
+      "epoch": 0.3706,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00015860895631436043,
+      "loss": 1.263,
+      "step": 18530
+    },
+    {
+      "epoch": 0.3707,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0001585806698138184,
+      "loss": 1.2694,
+      "step": 18535
+    },
+    {
+      "epoch": 0.3708,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001585523761753994,
+      "loss": 1.292,
+      "step": 18540
+    },
+    {
+      "epoch": 0.3709,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00015852407540255104,
+      "loss": 1.2835,
+      "step": 18545
+    },
+    {
+      "epoch": 0.371,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00015849576749872157,
+      "loss": 1.2743,
+      "step": 18550
+    },
+    {
+      "epoch": 0.3711,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00015846745246736026,
+      "loss": 1.2899,
+      "step": 18555
+    },
+    {
+      "epoch": 0.3712,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00015843913031191723,
+      "loss": 1.2818,
+      "step": 18560
+    },
+    {
+      "epoch": 0.3713,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00015841080103584342,
+      "loss": 1.2764,
+      "step": 18565
+    },
+    {
+      "epoch": 0.3714,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0001583824646425907,
+      "loss": 1.268,
+      "step": 18570
+    },
+    {
+      "epoch": 0.3715,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00015835412113561175,
+      "loss": 1.2724,
+      "step": 18575
+    },
+    {
+      "epoch": 0.3716,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00015832577051836015,
+      "loss": 1.2689,
+      "step": 18580
+    },
+    {
+      "epoch": 0.3717,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00015829741279429035,
+      "loss": 1.279,
+      "step": 18585
+    },
+    {
+      "epoch": 0.3718,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00015826904796685762,
+      "loss": 1.2705,
+      "step": 18590
+    },
+    {
+      "epoch": 0.3719,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00015824067603951812,
+      "loss": 1.2605,
+      "step": 18595
+    },
+    {
+      "epoch": 0.372,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00015821229701572896,
+      "loss": 1.2808,
+      "step": 18600
+    },
+    {
+      "epoch": 0.3721,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00015818391089894796,
+      "loss": 1.2595,
+      "step": 18605
+    },
+    {
+      "epoch": 0.3722,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00015815551769263387,
+      "loss": 1.2825,
+      "step": 18610
+    },
+    {
+      "epoch": 0.3723,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001581271174002464,
+      "loss": 1.2852,
+      "step": 18615
+    },
+    {
+      "epoch": 0.3724,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.000158098710025246,
+      "loss": 1.2666,
+      "step": 18620
+    },
+    {
+      "epoch": 0.3725,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00015807029557109398,
+      "loss": 1.272,
+      "step": 18625
+    },
+    {
+      "epoch": 0.3726,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001580418740412526,
+      "loss": 1.2662,
+      "step": 18630
+    },
+    {
+      "epoch": 0.3727,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00015801344543918495,
+      "loss": 1.271,
+      "step": 18635
+    },
+    {
+      "epoch": 0.3728,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00015798500976835493,
+      "loss": 1.284,
+      "step": 18640
+    },
+    {
+      "epoch": 0.3729,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00015795656703222736,
+      "loss": 1.2844,
+      "step": 18645
+    },
+    {
+      "epoch": 0.373,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001579281172342679,
+      "loss": 1.2883,
+      "step": 18650
+    },
+    {
+      "epoch": 0.3731,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00015789966037794306,
+      "loss": 1.2887,
+      "step": 18655
+    },
+    {
+      "epoch": 0.3732,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015787119646672025,
+      "loss": 1.2611,
+      "step": 18660
+    },
+    {
+      "epoch": 0.3733,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00015784272550406765,
+      "loss": 1.2854,
+      "step": 18665
+    },
+    {
+      "epoch": 0.3734,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00015781424749345446,
+      "loss": 1.2771,
+      "step": 18670
+    },
+    {
+      "epoch": 0.3735,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00015778576243835054,
+      "loss": 1.2708,
+      "step": 18675
+    },
+    {
+      "epoch": 0.3736,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00015775727034222675,
+      "loss": 1.2925,
+      "step": 18680
+    },
+    {
+      "epoch": 0.3737,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001577287712085548,
+      "loss": 1.2764,
+      "step": 18685
+    },
+    {
+      "epoch": 0.3738,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00015770026504080718,
+      "loss": 1.2826,
+      "step": 18690
+    },
+    {
+      "epoch": 0.3739,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00015767175184245726,
+      "loss": 1.2821,
+      "step": 18695
+    },
+    {
+      "epoch": 0.374,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00015764323161697935,
+      "loss": 1.2793,
+      "step": 18700
+    },
+    {
+      "epoch": 0.3741,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00015761470436784846,
+      "loss": 1.267,
+      "step": 18705
+    },
+    {
+      "epoch": 0.3742,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001575861700985407,
+      "loss": 1.2868,
+      "step": 18710
+    },
+    {
+      "epoch": 0.3743,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001575576288125327,
+      "loss": 1.2855,
+      "step": 18715
+    },
+    {
+      "epoch": 0.3744,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00015752908051330227,
+      "loss": 1.2745,
+      "step": 18720
+    },
+    {
+      "epoch": 0.3745,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00015750052520432787,
+      "loss": 1.2621,
+      "step": 18725
+    },
+    {
+      "epoch": 0.3746,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00015747196288908887,
+      "loss": 1.2738,
+      "step": 18730
+    },
+    {
+      "epoch": 0.3747,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00015744339357106558,
+      "loss": 1.267,
+      "step": 18735
+    },
+    {
+      "epoch": 0.3748,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.000157414817253739,
+      "loss": 1.283,
+      "step": 18740
+    },
+    {
+      "epoch": 0.3749,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00015738623394059107,
+      "loss": 1.2686,
+      "step": 18745
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.0001573576436351046,
+      "loss": 1.2761,
+      "step": 18750
+    },
+    {
+      "epoch": 0.3751,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00015732904634076328,
+      "loss": 1.2891,
+      "step": 18755
+    },
+    {
+      "epoch": 0.3752,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00015730044206105156,
+      "loss": 1.2771,
+      "step": 18760
+    },
+    {
+      "epoch": 0.3753,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00015727183079945476,
+      "loss": 1.2861,
+      "step": 18765
+    },
+    {
+      "epoch": 0.3754,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001572432125594591,
+      "loss": 1.2592,
+      "step": 18770
+    },
+    {
+      "epoch": 0.3755,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00015721458734455163,
+      "loss": 1.277,
+      "step": 18775
+    },
+    {
+      "epoch": 0.3756,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00015718595515822027,
+      "loss": 1.2884,
+      "step": 18780
+    },
+    {
+      "epoch": 0.3757,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001571573160039537,
+      "loss": 1.2824,
+      "step": 18785
+    },
+    {
+      "epoch": 0.3758,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00015712866988524155,
+      "loss": 1.2653,
+      "step": 18790
+    },
+    {
+      "epoch": 0.3759,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001571000168055743,
+      "loss": 1.2863,
+      "step": 18795
+    },
+    {
+      "epoch": 0.376,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0001570713567684432,
+      "loss": 1.2722,
+      "step": 18800
+    },
+    {
+      "epoch": 0.3761,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00015704268977734037,
+      "loss": 1.285,
+      "step": 18805
+    },
+    {
+      "epoch": 0.3762,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00015701401583575884,
+      "loss": 1.2628,
+      "step": 18810
+    },
+    {
+      "epoch": 0.3763,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00015698533494719238,
+      "loss": 1.2947,
+      "step": 18815
+    },
+    {
+      "epoch": 0.3764,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00015695664711513576,
+      "loss": 1.2718,
+      "step": 18820
+    },
+    {
+      "epoch": 0.3765,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00015692795234308445,
+      "loss": 1.266,
+      "step": 18825
+    },
+    {
+      "epoch": 0.3766,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001568992506345348,
+      "loss": 1.2918,
+      "step": 18830
+    },
+    {
+      "epoch": 0.3767,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00015687054199298408,
+      "loss": 1.2695,
+      "step": 18835
+    },
+    {
+      "epoch": 0.3768,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001568418264219303,
+      "loss": 1.2741,
+      "step": 18840
+    },
+    {
+      "epoch": 0.3769,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001568131039248724,
+      "loss": 1.2665,
+      "step": 18845
+    },
+    {
+      "epoch": 0.377,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00015678437450531013,
+      "loss": 1.2671,
+      "step": 18850
+    },
+    {
+      "epoch": 0.3771,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00015675563816674407,
+      "loss": 1.2755,
+      "step": 18855
+    },
+    {
+      "epoch": 0.3772,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00015672689491267567,
+      "loss": 1.2768,
+      "step": 18860
+    },
+    {
+      "epoch": 0.3773,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00015669814474660718,
+      "loss": 1.2725,
+      "step": 18865
+    },
+    {
+      "epoch": 0.3774,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0001566693876720417,
+      "loss": 1.2754,
+      "step": 18870
+    },
+    {
+      "epoch": 0.3775,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00015664062369248328,
+      "loss": 1.2784,
+      "step": 18875
+    },
+    {
+      "epoch": 0.3776,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00015661185281143667,
+      "loss": 1.2759,
+      "step": 18880
+    },
+    {
+      "epoch": 0.3777,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001565830750324075,
+      "loss": 1.2771,
+      "step": 18885
+    },
+    {
+      "epoch": 0.3778,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001565542903589023,
+      "loss": 1.2582,
+      "step": 18890
+    },
+    {
+      "epoch": 0.3779,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00015652549879442834,
+      "loss": 1.2688,
+      "step": 18895
+    },
+    {
+      "epoch": 0.378,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0001564967003424938,
+      "loss": 1.2853,
+      "step": 18900
+    },
+    {
+      "epoch": 0.3781,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015646789500660773,
+      "loss": 1.2817,
+      "step": 18905
+    },
+    {
+      "epoch": 0.3782,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00015643908279027992,
+      "loss": 1.2765,
+      "step": 18910
+    },
+    {
+      "epoch": 0.3783,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00015641026369702106,
+      "loss": 1.2925,
+      "step": 18915
+    },
+    {
+      "epoch": 0.3784,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015638143773034267,
+      "loss": 1.2708,
+      "step": 18920
+    },
+    {
+      "epoch": 0.3785,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00015635260489375714,
+      "loss": 1.2739,
+      "step": 18925
+    },
+    {
+      "epoch": 0.3786,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00015632376519077767,
+      "loss": 1.2708,
+      "step": 18930
+    },
+    {
+      "epoch": 0.3787,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0001562949186249182,
+      "loss": 1.2794,
+      "step": 18935
+    },
+    {
+      "epoch": 0.3788,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0001562660651996937,
+      "loss": 1.2682,
+      "step": 18940
+    },
+    {
+      "epoch": 0.3789,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0001562372049186198,
+      "loss": 1.2736,
+      "step": 18945
+    },
+    {
+      "epoch": 0.379,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00015620833778521307,
+      "loss": 1.2572,
+      "step": 18950
+    },
+    {
+      "epoch": 0.3791,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00015617946380299088,
+      "loss": 1.2707,
+      "step": 18955
+    },
+    {
+      "epoch": 0.3792,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00015615058297547145,
+      "loss": 1.2647,
+      "step": 18960
+    },
+    {
+      "epoch": 0.3793,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001561216953061738,
+      "loss": 1.2722,
+      "step": 18965
+    },
+    {
+      "epoch": 0.3794,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001560928007986178,
+      "loss": 1.288,
+      "step": 18970
+    },
+    {
+      "epoch": 0.3795,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001560638994563242,
+      "loss": 1.2815,
+      "step": 18975
+    },
+    {
+      "epoch": 0.3796,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00015603499128281448,
+      "loss": 1.2816,
+      "step": 18980
+    },
+    {
+      "epoch": 0.3797,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00015600607628161103,
+      "loss": 1.2788,
+      "step": 18985
+    },
+    {
+      "epoch": 0.3798,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00015597715445623712,
+      "loss": 1.2678,
+      "step": 18990
+    },
+    {
+      "epoch": 0.3799,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001559482258102167,
+      "loss": 1.2751,
+      "step": 18995
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0001559192903470747,
+      "loss": 1.2684,
+      "step": 19000
+    },
+    {
+      "epoch": 0.3801,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00015589034807033677,
+      "loss": 1.2563,
+      "step": 19005
+    },
+    {
+      "epoch": 0.3802,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00015586139898352946,
+      "loss": 1.2769,
+      "step": 19010
+    },
+    {
+      "epoch": 0.3803,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00015583244309018014,
+      "loss": 1.2626,
+      "step": 19015
+    },
+    {
+      "epoch": 0.3804,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.000155803480393817,
+      "loss": 1.2685,
+      "step": 19020
+    },
+    {
+      "epoch": 0.3805,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00015577451089796905,
+      "loss": 1.28,
+      "step": 19025
+    },
+    {
+      "epoch": 0.3806,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00015574553460616608,
+      "loss": 1.2871,
+      "step": 19030
+    },
+    {
+      "epoch": 0.3807,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00015571655152193885,
+      "loss": 1.2539,
+      "step": 19035
+    },
+    {
+      "epoch": 0.3808,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00015568756164881882,
+      "loss": 1.2694,
+      "step": 19040
+    },
+    {
+      "epoch": 0.3809,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00015565856499033832,
+      "loss": 1.2746,
+      "step": 19045
+    },
+    {
+      "epoch": 0.381,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0001556295615500305,
+      "loss": 1.2979,
+      "step": 19050
+    },
+    {
+      "epoch": 0.3811,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0001556005513314293,
+      "loss": 1.2752,
+      "step": 19055
+    },
+    {
+      "epoch": 0.3812,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00015557153433806966,
+      "loss": 1.273,
+      "step": 19060
+    },
+    {
+      "epoch": 0.3813,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00015554251057348713,
+      "loss": 1.2737,
+      "step": 19065
+    },
+    {
+      "epoch": 0.3814,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.0001555134800412181,
+      "loss": 1.285,
+      "step": 19070
+    },
+    {
+      "epoch": 0.3815,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00015548444274479995,
+      "loss": 1.2885,
+      "step": 19075
+    },
+    {
+      "epoch": 0.3816,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00015545539868777074,
+      "loss": 1.2852,
+      "step": 19080
+    },
+    {
+      "epoch": 0.3817,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00015542634787366942,
+      "loss": 1.2866,
+      "step": 19085
+    },
+    {
+      "epoch": 0.3818,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00015539729030603574,
+      "loss": 1.2608,
+      "step": 19090
+    },
+    {
+      "epoch": 0.3819,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00015536822598841024,
+      "loss": 1.2755,
+      "step": 19095
+    },
+    {
+      "epoch": 0.382,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00015533915492433443,
+      "loss": 1.2788,
+      "step": 19100
+    },
+    {
+      "epoch": 0.3821,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0001553100771173504,
+      "loss": 1.2651,
+      "step": 19105
+    },
+    {
+      "epoch": 0.3822,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00015528099257100127,
+      "loss": 1.2706,
+      "step": 19110
+    },
+    {
+      "epoch": 0.3823,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00015525190128883083,
+      "loss": 1.3004,
+      "step": 19115
+    },
+    {
+      "epoch": 0.3824,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015522280327438388,
+      "loss": 1.2852,
+      "step": 19120
+    },
+    {
+      "epoch": 0.3825,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0001551936985312058,
+      "loss": 1.2781,
+      "step": 19125
+    },
+    {
+      "epoch": 0.3826,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00015516458706284303,
+      "loss": 1.2776,
+      "step": 19130
+    },
+    {
+      "epoch": 0.3827,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00015513546887284264,
+      "loss": 1.2816,
+      "step": 19135
+    },
+    {
+      "epoch": 0.3828,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.0001551063439647526,
+      "loss": 1.2734,
+      "step": 19140
+    },
+    {
+      "epoch": 0.3829,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00015507721234212172,
+      "loss": 1.2806,
+      "step": 19145
+    },
+    {
+      "epoch": 0.383,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00015504807400849958,
+      "loss": 1.295,
+      "step": 19150
+    },
+    {
+      "epoch": 0.3831,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001550189289674366,
+      "loss": 1.2599,
+      "step": 19155
+    },
+    {
+      "epoch": 0.3832,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.000154989777222484,
+      "loss": 1.2555,
+      "step": 19160
+    },
+    {
+      "epoch": 0.3833,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00015496061877719384,
+      "loss": 1.2831,
+      "step": 19165
+    },
+    {
+      "epoch": 0.3834,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.000154931453635119,
+      "loss": 1.2667,
+      "step": 19170
+    },
+    {
+      "epoch": 0.3835,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001549022817998132,
+      "loss": 1.2816,
+      "step": 19175
+    },
+    {
+      "epoch": 0.3836,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00015487310327483086,
+      "loss": 1.2758,
+      "step": 19180
+    },
+    {
+      "epoch": 0.3837,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015484391806372733,
+      "loss": 1.2817,
+      "step": 19185
+    },
+    {
+      "epoch": 0.3838,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00015481472617005876,
+      "loss": 1.272,
+      "step": 19190
+    },
+    {
+      "epoch": 0.3839,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00015478552759738207,
+      "loss": 1.2955,
+      "step": 19195
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00015475632234925504,
+      "loss": 1.2752,
+      "step": 19200
+    },
+    {
+      "epoch": 0.3841,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00015472711042923621,
+      "loss": 1.2722,
+      "step": 19205
+    },
+    {
+      "epoch": 0.3842,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00015469789184088497,
+      "loss": 1.2757,
+      "step": 19210
+    },
+    {
+      "epoch": 0.3843,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00015466866658776155,
+      "loss": 1.3063,
+      "step": 19215
+    },
+    {
+      "epoch": 0.3844,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00015463943467342693,
+      "loss": 1.2687,
+      "step": 19220
+    },
+    {
+      "epoch": 0.3845,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001546101961014429,
+      "loss": 1.2801,
+      "step": 19225
+    },
+    {
+      "epoch": 0.3846,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00015458095087537218,
+      "loss": 1.27,
+      "step": 19230
+    },
+    {
+      "epoch": 0.3847,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00015455169899877813,
+      "loss": 1.2866,
+      "step": 19235
+    },
+    {
+      "epoch": 0.3848,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00015452244047522502,
+      "loss": 1.2745,
+      "step": 19240
+    },
+    {
+      "epoch": 0.3849,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00015449317530827794,
+      "loss": 1.2817,
+      "step": 19245
+    },
+    {
+      "epoch": 0.385,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00015446390350150273,
+      "loss": 1.2667,
+      "step": 19250
+    },
+    {
+      "epoch": 0.3851,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001544346250584661,
+      "loss": 1.2688,
+      "step": 19255
+    },
+    {
+      "epoch": 0.3852,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00015440533998273547,
+      "loss": 1.2681,
+      "step": 19260
+    },
+    {
+      "epoch": 0.3853,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00015437604827787927,
+      "loss": 1.2614,
+      "step": 19265
+    },
+    {
+      "epoch": 0.3854,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001543467499474665,
+      "loss": 1.2938,
+      "step": 19270
+    },
+    {
+      "epoch": 0.3855,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00015431744499506706,
+      "loss": 1.2567,
+      "step": 19275
+    },
+    {
+      "epoch": 0.3856,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00015428813342425177,
+      "loss": 1.2852,
+      "step": 19280
+    },
+    {
+      "epoch": 0.3857,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00015425881523859207,
+      "loss": 1.2679,
+      "step": 19285
+    },
+    {
+      "epoch": 0.3858,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001542294904416603,
+      "loss": 1.2869,
+      "step": 19290
+    },
+    {
+      "epoch": 0.3859,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00015420015903702962,
+      "loss": 1.2798,
+      "step": 19295
+    },
+    {
+      "epoch": 0.386,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.000154170821028274,
+      "loss": 1.2639,
+      "step": 19300
+    },
+    {
+      "epoch": 0.3861,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00015414147641896813,
+      "loss": 1.2678,
+      "step": 19305
+    },
+    {
+      "epoch": 0.3862,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015411212521268758,
+      "loss": 1.2798,
+      "step": 19310
+    },
+    {
+      "epoch": 0.3863,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00015408276741300873,
+      "loss": 1.2712,
+      "step": 19315
+    },
+    {
+      "epoch": 0.3864,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00015405340302350871,
+      "loss": 1.2717,
+      "step": 19320
+    },
+    {
+      "epoch": 0.3865,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0001540240320477655,
+      "loss": 1.2813,
+      "step": 19325
+    },
+    {
+      "epoch": 0.3866,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00015399465448935788,
+      "loss": 1.2896,
+      "step": 19330
+    },
+    {
+      "epoch": 0.3867,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00015396527035186537,
+      "loss": 1.2953,
+      "step": 19335
+    },
+    {
+      "epoch": 0.3868,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00015393587963886835,
+      "loss": 1.2851,
+      "step": 19340
+    },
+    {
+      "epoch": 0.3869,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00015390648235394803,
+      "loss": 1.2557,
+      "step": 19345
+    },
+    {
+      "epoch": 0.387,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0001538770785006863,
+      "loss": 1.2821,
+      "step": 19350
+    },
+    {
+      "epoch": 0.3871,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00015384766808266602,
+      "loss": 1.296,
+      "step": 19355
+    },
+    {
+      "epoch": 0.3872,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00015381825110347074,
+      "loss": 1.2796,
+      "step": 19360
+    },
+    {
+      "epoch": 0.3873,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00015378882756668478,
+      "loss": 1.2739,
+      "step": 19365
+    },
+    {
+      "epoch": 0.3874,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00015375939747589335,
+      "loss": 1.262,
+      "step": 19370
+    },
+    {
+      "epoch": 0.3875,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001537299608346824,
+      "loss": 1.2644,
+      "step": 19375
+    },
+    {
+      "epoch": 0.3876,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001537005176466387,
+      "loss": 1.2794,
+      "step": 19380
+    },
+    {
+      "epoch": 0.3877,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00015367106791534983,
+      "loss": 1.2589,
+      "step": 19385
+    },
+    {
+      "epoch": 0.3878,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0001536416116444041,
+      "loss": 1.2632,
+      "step": 19390
+    },
+    {
+      "epoch": 0.3879,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00015361214883739076,
+      "loss": 1.2919,
+      "step": 19395
+    },
+    {
+      "epoch": 0.388,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00015358267949789966,
+      "loss": 1.2738,
+      "step": 19400
+    },
+    {
+      "epoch": 0.3881,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00015355320362952162,
+      "loss": 1.2726,
+      "step": 19405
+    },
+    {
+      "epoch": 0.3882,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00015352372123584814,
+      "loss": 1.2782,
+      "step": 19410
+    },
+    {
+      "epoch": 0.3883,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00015349423232047162,
+      "loss": 1.2838,
+      "step": 19415
+    },
+    {
+      "epoch": 0.3884,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00015346473688698513,
+      "loss": 1.2876,
+      "step": 19420
+    },
+    {
+      "epoch": 0.3885,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015343523493898265,
+      "loss": 1.2508,
+      "step": 19425
+    },
+    {
+      "epoch": 0.3886,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00015340572648005888,
+      "loss": 1.2814,
+      "step": 19430
+    },
+    {
+      "epoch": 0.3887,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001533762115138093,
+      "loss": 1.2595,
+      "step": 19435
+    },
+    {
+      "epoch": 0.3888,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0001533466900438303,
+      "loss": 1.2737,
+      "step": 19440
+    },
+    {
+      "epoch": 0.3889,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00015331716207371888,
+      "loss": 1.256,
+      "step": 19445
+    },
+    {
+      "epoch": 0.389,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.000153287627607073,
+      "loss": 1.2713,
+      "step": 19450
+    },
+    {
+      "epoch": 0.3891,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00015325808664749135,
+      "loss": 1.2809,
+      "step": 19455
+    },
+    {
+      "epoch": 0.3892,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001532285391985734,
+      "loss": 1.2763,
+      "step": 19460
+    },
+    {
+      "epoch": 0.3893,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0001531989852639194,
+      "loss": 1.2843,
+      "step": 19465
+    },
+    {
+      "epoch": 0.3894,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001531694248471304,
+      "loss": 1.2737,
+      "step": 19470
+    },
+    {
+      "epoch": 0.3895,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00015313985795180828,
+      "loss": 1.2781,
+      "step": 19475
+    },
+    {
+      "epoch": 0.3896,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00015311028458155567,
+      "loss": 1.2751,
+      "step": 19480
+    },
+    {
+      "epoch": 0.3897,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00015308070473997598,
+      "loss": 1.2882,
+      "step": 19485
+    },
+    {
+      "epoch": 0.3898,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001530511184306734,
+      "loss": 1.2761,
+      "step": 19490
+    },
+    {
+      "epoch": 0.3899,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00015302152565725298,
+      "loss": 1.2801,
+      "step": 19495
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001529919264233205,
+      "loss": 1.2707,
+      "step": 19500
+    },
+    {
+      "epoch": 0.3901,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00015296232073248251,
+      "loss": 1.2757,
+      "step": 19505
+    },
+    {
+      "epoch": 0.3902,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00015293270858834644,
+      "loss": 1.2719,
+      "step": 19510
+    },
+    {
+      "epoch": 0.3903,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00015290308999452032,
+      "loss": 1.2636,
+      "step": 19515
+    },
+    {
+      "epoch": 0.3904,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00015287346495461315,
+      "loss": 1.2702,
+      "step": 19520
+    },
+    {
+      "epoch": 0.3905,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00015284383347223472,
+      "loss": 1.279,
+      "step": 19525
+    },
+    {
+      "epoch": 0.3906,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015281419555099546,
+      "loss": 1.2767,
+      "step": 19530
+    },
+    {
+      "epoch": 0.3907,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00015278455119450664,
+      "loss": 1.2645,
+      "step": 19535
+    },
+    {
+      "epoch": 0.3908,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00015275490040638038,
+      "loss": 1.2582,
+      "step": 19540
+    },
+    {
+      "epoch": 0.3909,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00015272524319022955,
+      "loss": 1.2832,
+      "step": 19545
+    },
+    {
+      "epoch": 0.391,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00015269557954966778,
+      "loss": 1.2829,
+      "step": 19550
+    },
+    {
+      "epoch": 0.3911,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00015266590948830947,
+      "loss": 1.2795,
+      "step": 19555
+    },
+    {
+      "epoch": 0.3912,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00015263623300976978,
+      "loss": 1.2586,
+      "step": 19560
+    },
+    {
+      "epoch": 0.3913,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00015260655011766484,
+      "loss": 1.2806,
+      "step": 19565
+    },
+    {
+      "epoch": 0.3914,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00015257686081561134,
+      "loss": 1.2851,
+      "step": 19570
+    },
+    {
+      "epoch": 0.3915,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00015254716510722677,
+      "loss": 1.2867,
+      "step": 19575
+    },
+    {
+      "epoch": 0.3916,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0001525174629961296,
+      "loss": 1.2749,
+      "step": 19580
+    },
+    {
+      "epoch": 0.3917,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00015248775448593882,
+      "loss": 1.2855,
+      "step": 19585
+    },
+    {
+      "epoch": 0.3918,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00015245803958027434,
+      "loss": 1.2719,
+      "step": 19590
+    },
+    {
+      "epoch": 0.3919,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00015242831828275692,
+      "loss": 1.2834,
+      "step": 19595
+    },
+    {
+      "epoch": 0.392,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00015239859059700794,
+      "loss": 1.2698,
+      "step": 19600
+    },
+    {
+      "epoch": 0.3921,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00015236885652664963,
+      "loss": 1.2659,
+      "step": 19605
+    },
+    {
+      "epoch": 0.3922,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00015233911607530497,
+      "loss": 1.2639,
+      "step": 19610
+    },
+    {
+      "epoch": 0.3923,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001523093692465978,
+      "loss": 1.2555,
+      "step": 19615
+    },
+    {
+      "epoch": 0.3924,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00015227961604415266,
+      "loss": 1.271,
+      "step": 19620
+    },
+    {
+      "epoch": 0.3925,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.0001522498564715949,
+      "loss": 1.2635,
+      "step": 19625
+    },
+    {
+      "epoch": 0.3926,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0001522200905325506,
+      "loss": 1.2573,
+      "step": 19630
+    },
+    {
+      "epoch": 0.3927,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00015219031823064666,
+      "loss": 1.2693,
+      "step": 19635
+    },
+    {
+      "epoch": 0.3928,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0001521605395695108,
+      "loss": 1.286,
+      "step": 19640
+    },
+    {
+      "epoch": 0.3929,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001521307545527714,
+      "loss": 1.2836,
+      "step": 19645
+    },
+    {
+      "epoch": 0.393,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00015210096318405767,
+      "loss": 1.2796,
+      "step": 19650
+    },
+    {
+      "epoch": 0.3931,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001520711654669996,
+      "loss": 1.2694,
+      "step": 19655
+    },
+    {
+      "epoch": 0.3932,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00015204136140522799,
+      "loss": 1.3075,
+      "step": 19660
+    },
+    {
+      "epoch": 0.3933,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0001520115510023743,
+      "loss": 1.2748,
+      "step": 19665
+    },
+    {
+      "epoch": 0.3934,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00015198173426207094,
+      "loss": 1.2779,
+      "step": 19670
+    },
+    {
+      "epoch": 0.3935,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00015195191118795096,
+      "loss": 1.2699,
+      "step": 19675
+    },
+    {
+      "epoch": 0.3936,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00015192208178364816,
+      "loss": 1.2786,
+      "step": 19680
+    },
+    {
+      "epoch": 0.3937,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00015189224605279718,
+      "loss": 1.2972,
+      "step": 19685
+    },
+    {
+      "epoch": 0.3938,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015186240399903342,
+      "loss": 1.275,
+      "step": 19690
+    },
+    {
+      "epoch": 0.3939,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00015183255562599307,
+      "loss": 1.2645,
+      "step": 19695
+    },
+    {
+      "epoch": 0.394,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00015180270093731303,
+      "loss": 1.2714,
+      "step": 19700
+    },
+    {
+      "epoch": 0.3941,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.000151772839936631,
+      "loss": 1.2754,
+      "step": 19705
+    },
+    {
+      "epoch": 0.3942,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001517429726275855,
+      "loss": 1.2703,
+      "step": 19710
+    },
+    {
+      "epoch": 0.3943,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00015171309901381572,
+      "loss": 1.2778,
+      "step": 19715
+    },
+    {
+      "epoch": 0.3944,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00015168321909896172,
+      "loss": 1.2697,
+      "step": 19720
+    },
+    {
+      "epoch": 0.3945,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0001516533328866642,
+      "loss": 1.2616,
+      "step": 19725
+    },
+    {
+      "epoch": 0.3946,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00015162344038056476,
+      "loss": 1.2825,
+      "step": 19730
+    },
+    {
+      "epoch": 0.3947,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00015159354158430572,
+      "loss": 1.269,
+      "step": 19735
+    },
+    {
+      "epoch": 0.3948,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00015156363650153012,
+      "loss": 1.2712,
+      "step": 19740
+    },
+    {
+      "epoch": 0.3949,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00015153372513588182,
+      "loss": 1.2748,
+      "step": 19745
+    },
+    {
+      "epoch": 0.395,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00015150380749100545,
+      "loss": 1.2639,
+      "step": 19750
+    },
+    {
+      "epoch": 0.3951,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00015147388357054632,
+      "loss": 1.2776,
+      "step": 19755
+    },
+    {
+      "epoch": 0.3952,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00015144395337815064,
+      "loss": 1.2909,
+      "step": 19760
+    },
+    {
+      "epoch": 0.3953,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001514140169174653,
+      "loss": 1.2736,
+      "step": 19765
+    },
+    {
+      "epoch": 0.3954,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00015138407419213796,
+      "loss": 1.256,
+      "step": 19770
+    },
+    {
+      "epoch": 0.3955,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00015135412520581702,
+      "loss": 1.2644,
+      "step": 19775
+    },
+    {
+      "epoch": 0.3956,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0001513241699621517,
+      "loss": 1.2993,
+      "step": 19780
+    },
+    {
+      "epoch": 0.3957,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00015129420846479196,
+      "loss": 1.2884,
+      "step": 19785
+    },
+    {
+      "epoch": 0.3958,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001512642407173885,
+      "loss": 1.2928,
+      "step": 19790
+    },
+    {
+      "epoch": 0.3959,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015123426672359285,
+      "loss": 1.2716,
+      "step": 19795
+    },
+    {
+      "epoch": 0.396,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00015120428648705717,
+      "loss": 1.2659,
+      "step": 19800
+    },
+    {
+      "epoch": 0.3961,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00015117430001143452,
+      "loss": 1.2821,
+      "step": 19805
+    },
+    {
+      "epoch": 0.3962,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001511443073003786,
+      "loss": 1.2718,
+      "step": 19810
+    },
+    {
+      "epoch": 0.3963,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.000151114308357544,
+      "loss": 1.2699,
+      "step": 19815
+    },
+    {
+      "epoch": 0.3964,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.000151084303186586,
+      "loss": 1.2674,
+      "step": 19820
+    },
+    {
+      "epoch": 0.3965,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001510542917911606,
+      "loss": 1.2797,
+      "step": 19825
+    },
+    {
+      "epoch": 0.3966,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.0001510242741749246,
+      "loss": 1.279,
+      "step": 19830
+    },
+    {
+      "epoch": 0.3967,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00015099425034153553,
+      "loss": 1.2748,
+      "step": 19835
+    },
+    {
+      "epoch": 0.3968,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00015096422029465178,
+      "loss": 1.2689,
+      "step": 19840
+    },
+    {
+      "epoch": 0.3969,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00015093418403793238,
+      "loss": 1.2702,
+      "step": 19845
+    },
+    {
+      "epoch": 0.397,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00015090414157503714,
+      "loss": 1.2732,
+      "step": 19850
+    },
+    {
+      "epoch": 0.3971,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00015087409290962667,
+      "loss": 1.2814,
+      "step": 19855
+    },
+    {
+      "epoch": 0.3972,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0001508440380453623,
+      "loss": 1.2558,
+      "step": 19860
+    },
+    {
+      "epoch": 0.3973,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0001508139769859061,
+      "loss": 1.2744,
+      "step": 19865
+    },
+    {
+      "epoch": 0.3974,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00015078390973492092,
+      "loss": 1.2776,
+      "step": 19870
+    },
+    {
+      "epoch": 0.3975,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00015075383629607042,
+      "loss": 1.274,
+      "step": 19875
+    },
+    {
+      "epoch": 0.3976,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00015072375667301893,
+      "loss": 1.2687,
+      "step": 19880
+    },
+    {
+      "epoch": 0.3977,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00015069367086943154,
+      "loss": 1.2741,
+      "step": 19885
+    },
+    {
+      "epoch": 0.3978,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001506635788889741,
+      "loss": 1.2842,
+      "step": 19890
+    },
+    {
+      "epoch": 0.3979,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00015063348073531324,
+      "loss": 1.2791,
+      "step": 19895
+    },
+    {
+      "epoch": 0.398,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00015060337641211637,
+      "loss": 1.2846,
+      "step": 19900
+    },
+    {
+      "epoch": 0.3981,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001505732659230516,
+      "loss": 1.2696,
+      "step": 19905
+    },
+    {
+      "epoch": 0.3982,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001505431492717878,
+      "loss": 1.2836,
+      "step": 19910
+    },
+    {
+      "epoch": 0.3983,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0001505130264619945,
+      "loss": 1.2753,
+      "step": 19915
+    },
+    {
+      "epoch": 0.3984,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001504828974973422,
+      "loss": 1.2818,
+      "step": 19920
+    },
+    {
+      "epoch": 0.3985,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00015045276238150192,
+      "loss": 1.2726,
+      "step": 19925
+    },
+    {
+      "epoch": 0.3986,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00015042262111814565,
+      "loss": 1.294,
+      "step": 19930
+    },
+    {
+      "epoch": 0.3987,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015039247371094588,
+      "loss": 1.2585,
+      "step": 19935
+    },
+    {
+      "epoch": 0.3988,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001503623201635761,
+      "loss": 1.2649,
+      "step": 19940
+    },
+    {
+      "epoch": 0.3989,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00015033216047971031,
+      "loss": 1.2659,
+      "step": 19945
+    },
+    {
+      "epoch": 0.399,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00015030199466302353,
+      "loss": 1.2677,
+      "step": 19950
+    },
+    {
+      "epoch": 0.3991,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00015027182271719122,
+      "loss": 1.2669,
+      "step": 19955
+    },
+    {
+      "epoch": 0.3992,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00015024164464588982,
+      "loss": 1.265,
+      "step": 19960
+    },
+    {
+      "epoch": 0.3993,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001502114604527964,
+      "loss": 1.2823,
+      "step": 19965
+    },
+    {
+      "epoch": 0.3994,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00015018127014158886,
+      "loss": 1.2656,
+      "step": 19970
+    },
+    {
+      "epoch": 0.3995,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00015015107371594573,
+      "loss": 1.265,
+      "step": 19975
+    },
+    {
+      "epoch": 0.3996,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00015012087117954642,
+      "loss": 1.2768,
+      "step": 19980
+    },
+    {
+      "epoch": 0.3997,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.000150090662536071,
+      "loss": 1.2822,
+      "step": 19985
+    },
+    {
+      "epoch": 0.3998,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0001500604477892003,
+      "loss": 1.2757,
+      "step": 19990
+    },
+    {
+      "epoch": 0.3999,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00015003022694261585,
+      "loss": 1.2718,
+      "step": 19995
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 1.2658,
+      "step": 20000
+    },
+    {
+      "epoch": 0.4001,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00014996976696503587,
+      "loss": 1.2824,
+      "step": 20005
+    },
+    {
+      "epoch": 0.4002,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00014993952784140717,
+      "loss": 1.2763,
+      "step": 20010
+    },
+    {
+      "epoch": 0.4003,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014990928263279848,
+      "loss": 1.2778,
+      "step": 20015
+    },
+    {
+      "epoch": 0.4004,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00014987903134289508,
+      "loss": 1.2835,
+      "step": 20020
+    },
+    {
+      "epoch": 0.4005,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00014984877397538303,
+      "loss": 1.2751,
+      "step": 20025
+    },
+    {
+      "epoch": 0.4006,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001498185105339491,
+      "loss": 1.265,
+      "step": 20030
+    },
+    {
+      "epoch": 0.4007,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00014978824102228076,
+      "loss": 1.2656,
+      "step": 20035
+    },
+    {
+      "epoch": 0.4008,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00014975796544406625,
+      "loss": 1.2968,
+      "step": 20040
+    },
+    {
+      "epoch": 0.4009,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001497276838029946,
+      "loss": 1.2806,
+      "step": 20045
+    },
+    {
+      "epoch": 0.401,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00014969739610275556,
+      "loss": 1.2912,
+      "step": 20050
+    },
+    {
+      "epoch": 0.4011,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001496671023470395,
+      "loss": 1.2642,
+      "step": 20055
+    },
+    {
+      "epoch": 0.4012,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001496368025395377,
+      "loss": 1.2723,
+      "step": 20060
+    },
+    {
+      "epoch": 0.4013,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014960649668394207,
+      "loss": 1.2812,
+      "step": 20065
+    },
+    {
+      "epoch": 0.4014,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00014957618478394529,
+      "loss": 1.2734,
+      "step": 20070
+    },
+    {
+      "epoch": 0.4015,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00014954586684324078,
+      "loss": 1.2632,
+      "step": 20075
+    },
+    {
+      "epoch": 0.4016,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00014951554286552266,
+      "loss": 1.2683,
+      "step": 20080
+    },
+    {
+      "epoch": 0.4017,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00014948521285448586,
+      "loss": 1.2889,
+      "step": 20085
+    },
+    {
+      "epoch": 0.4018,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00014945487681382598,
+      "loss": 1.2834,
+      "step": 20090
+    },
+    {
+      "epoch": 0.4019,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00014942453474723935,
+      "loss": 1.272,
+      "step": 20095
+    },
+    {
+      "epoch": 0.402,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0001493941866584231,
+      "loss": 1.2867,
+      "step": 20100
+    },
+    {
+      "epoch": 0.4021,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00014936383255107505,
+      "loss": 1.2818,
+      "step": 20105
+    },
+    {
+      "epoch": 0.4022,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001493334724288937,
+      "loss": 1.2804,
+      "step": 20110
+    },
+    {
+      "epoch": 0.4023,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001493031062955784,
+      "loss": 1.2695,
+      "step": 20115
+    },
+    {
+      "epoch": 0.4024,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00014927273415482915,
+      "loss": 1.2755,
+      "step": 20120
+    },
+    {
+      "epoch": 0.4025,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00014924235601034672,
+      "loss": 1.2818,
+      "step": 20125
+    },
+    {
+      "epoch": 0.4026,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00014921197186583255,
+      "loss": 1.2698,
+      "step": 20130
+    },
+    {
+      "epoch": 0.4027,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001491815817249889,
+      "loss": 1.279,
+      "step": 20135
+    },
+    {
+      "epoch": 0.4028,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001491511855915187,
+      "loss": 1.2763,
+      "step": 20140
+    },
+    {
+      "epoch": 0.4029,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014912078346912563,
+      "loss": 1.2794,
+      "step": 20145
+    },
+    {
+      "epoch": 0.403,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00014909037536151409,
+      "loss": 1.2784,
+      "step": 20150
+    },
+    {
+      "epoch": 0.4031,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0001490599612723892,
+      "loss": 1.2746,
+      "step": 20155
+    },
+    {
+      "epoch": 0.4032,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00014902954120545687,
+      "loss": 1.2765,
+      "step": 20160
+    },
+    {
+      "epoch": 0.4033,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00014899911516442365,
+      "loss": 1.2823,
+      "step": 20165
+    },
+    {
+      "epoch": 0.4034,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00014896868315299693,
+      "loss": 1.2888,
+      "step": 20170
+    },
+    {
+      "epoch": 0.4035,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00014893824517488464,
+      "loss": 1.264,
+      "step": 20175
+    },
+    {
+      "epoch": 0.4036,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00014890780123379564,
+      "loss": 1.2935,
+      "step": 20180
+    },
+    {
+      "epoch": 0.4037,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0001488773513334394,
+      "loss": 1.2719,
+      "step": 20185
+    },
+    {
+      "epoch": 0.4038,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.0001488468954775262,
+      "loss": 1.2809,
+      "step": 20190
+    },
+    {
+      "epoch": 0.4039,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00014881643366976692,
+      "loss": 1.2841,
+      "step": 20195
+    },
+    {
+      "epoch": 0.404,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0001487859659138733,
+      "loss": 1.2966,
+      "step": 20200
+    },
+    {
+      "epoch": 0.4041,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00014875549221355768,
+      "loss": 1.2667,
+      "step": 20205
+    },
+    {
+      "epoch": 0.4042,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00014872501257253323,
+      "loss": 1.2717,
+      "step": 20210
+    },
+    {
+      "epoch": 0.4043,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014869452699451383,
+      "loss": 1.277,
+      "step": 20215
+    },
+    {
+      "epoch": 0.4044,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.000148664035483214,
+      "loss": 1.2832,
+      "step": 20220
+    },
+    {
+      "epoch": 0.4045,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014863353804234905,
+      "loss": 1.2712,
+      "step": 20225
+    },
+    {
+      "epoch": 0.4046,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00014860303467563503,
+      "loss": 1.2658,
+      "step": 20230
+    },
+    {
+      "epoch": 0.4047,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00014857252538678865,
+      "loss": 1.2713,
+      "step": 20235
+    },
+    {
+      "epoch": 0.4048,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001485420101795274,
+      "loss": 1.2648,
+      "step": 20240
+    },
+    {
+      "epoch": 0.4049,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014851148905756947,
+      "loss": 1.2808,
+      "step": 20245
+    },
+    {
+      "epoch": 0.405,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00014848096202463372,
+      "loss": 1.2686,
+      "step": 20250
+    },
+    {
+      "epoch": 0.4051,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001484504290844398,
+      "loss": 1.2848,
+      "step": 20255
+    },
+    {
+      "epoch": 0.4052,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00014841989024070809,
+      "loss": 1.2881,
+      "step": 20260
+    },
+    {
+      "epoch": 0.4053,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00014838934549715963,
+      "loss": 1.2739,
+      "step": 20265
+    },
+    {
+      "epoch": 0.4054,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00014835879485751617,
+      "loss": 1.2799,
+      "step": 20270
+    },
+    {
+      "epoch": 0.4055,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00014832823832550024,
+      "loss": 1.2569,
+      "step": 20275
+    },
+    {
+      "epoch": 0.4056,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00014829767590483506,
+      "loss": 1.2755,
+      "step": 20280
+    },
+    {
+      "epoch": 0.4057,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001482671075992446,
+      "loss": 1.2592,
+      "step": 20285
+    },
+    {
+      "epoch": 0.4058,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00014823653341245353,
+      "loss": 1.2811,
+      "step": 20290
+    },
+    {
+      "epoch": 0.4059,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00014820595334818712,
+      "loss": 1.2851,
+      "step": 20295
+    },
+    {
+      "epoch": 0.406,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00014817536741017152,
+      "loss": 1.2796,
+      "step": 20300
+    },
+    {
+      "epoch": 0.4061,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00014814477560213358,
+      "loss": 1.2705,
+      "step": 20305
+    },
+    {
+      "epoch": 0.4062,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00014811417792780075,
+      "loss": 1.2792,
+      "step": 20310
+    },
+    {
+      "epoch": 0.4063,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00014808357439090127,
+      "loss": 1.2651,
+      "step": 20315
+    },
+    {
+      "epoch": 0.4064,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00014805296499516407,
+      "loss": 1.2526,
+      "step": 20320
+    },
+    {
+      "epoch": 0.4065,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001480223497443189,
+      "loss": 1.2763,
+      "step": 20325
+    },
+    {
+      "epoch": 0.4066,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00014799172864209608,
+      "loss": 1.2751,
+      "step": 20330
+    },
+    {
+      "epoch": 0.4067,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00014796110169222666,
+      "loss": 1.2868,
+      "step": 20335
+    },
+    {
+      "epoch": 0.4068,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0001479304688984425,
+      "loss": 1.2689,
+      "step": 20340
+    },
+    {
+      "epoch": 0.4069,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00014789983026447612,
+      "loss": 1.2709,
+      "step": 20345
+    },
+    {
+      "epoch": 0.407,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0001478691857940607,
+      "loss": 1.2836,
+      "step": 20350
+    },
+    {
+      "epoch": 0.4071,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00014783853549093018,
+      "loss": 1.2667,
+      "step": 20355
+    },
+    {
+      "epoch": 0.4072,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00014780787935881923,
+      "loss": 1.2747,
+      "step": 20360
+    },
+    {
+      "epoch": 0.4073,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0001477772174014632,
+      "loss": 1.2781,
+      "step": 20365
+    },
+    {
+      "epoch": 0.4074,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00014774654962259812,
+      "loss": 1.2658,
+      "step": 20370
+    },
+    {
+      "epoch": 0.4075,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00014771587602596084,
+      "loss": 1.2651,
+      "step": 20375
+    },
+    {
+      "epoch": 0.4076,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001476851966152888,
+      "loss": 1.2988,
+      "step": 20380
+    },
+    {
+      "epoch": 0.4077,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0001476545113943202,
+      "loss": 1.2803,
+      "step": 20385
+    },
+    {
+      "epoch": 0.4078,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.0001476238203667939,
+      "loss": 1.2813,
+      "step": 20390
+    },
+    {
+      "epoch": 0.4079,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001475931235364496,
+      "loss": 1.2615,
+      "step": 20395
+    },
+    {
+      "epoch": 0.408,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00014756242090702756,
+      "loss": 1.2823,
+      "step": 20400
+    },
+    {
+      "epoch": 0.4081,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00014753171248226875,
+      "loss": 1.2678,
+      "step": 20405
+    },
+    {
+      "epoch": 0.4082,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00014750099826591498,
+      "loss": 1.2794,
+      "step": 20410
+    },
+    {
+      "epoch": 0.4083,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00014747027826170867,
+      "loss": 1.2655,
+      "step": 20415
+    },
+    {
+      "epoch": 0.4084,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00014743955247339293,
+      "loss": 1.2605,
+      "step": 20420
+    },
+    {
+      "epoch": 0.4085,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001474088209047116,
+      "loss": 1.279,
+      "step": 20425
+    },
+    {
+      "epoch": 0.4086,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00014737808355940932,
+      "loss": 1.271,
+      "step": 20430
+    },
+    {
+      "epoch": 0.4087,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0001473473404412312,
+      "loss": 1.2739,
+      "step": 20435
+    },
+    {
+      "epoch": 0.4088,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00014731659155392332,
+      "loss": 1.2759,
+      "step": 20440
+    },
+    {
+      "epoch": 0.4089,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00014728583690123224,
+      "loss": 1.2734,
+      "step": 20445
+    },
+    {
+      "epoch": 0.409,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00014725507648690543,
+      "loss": 1.279,
+      "step": 20450
+    },
+    {
+      "epoch": 0.4091,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00014722431031469083,
+      "loss": 1.2834,
+      "step": 20455
+    },
+    {
+      "epoch": 0.4092,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001471935383883373,
+      "loss": 1.2843,
+      "step": 20460
+    },
+    {
+      "epoch": 0.4093,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00014716276071159422,
+      "loss": 1.2712,
+      "step": 20465
+    },
+    {
+      "epoch": 0.4094,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00014713197728821183,
+      "loss": 1.2809,
+      "step": 20470
+    },
+    {
+      "epoch": 0.4095,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.000147101188121941,
+      "loss": 1.2903,
+      "step": 20475
+    },
+    {
+      "epoch": 0.4096,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001470703932165333,
+      "loss": 1.2579,
+      "step": 20480
+    },
+    {
+      "epoch": 0.4097,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001470395925757409,
+      "loss": 1.2579,
+      "step": 20485
+    },
+    {
+      "epoch": 0.4098,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00014700878620331684,
+      "loss": 1.2765,
+      "step": 20490
+    },
+    {
+      "epoch": 0.4099,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001469779741030148,
+      "loss": 1.2757,
+      "step": 20495
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00014694715627858908,
+      "loss": 1.2831,
+      "step": 20500
+    },
+    {
+      "epoch": 0.4101,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001469163327337948,
+      "loss": 1.27,
+      "step": 20505
+    },
+    {
+      "epoch": 0.4102,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001468855034723877,
+      "loss": 1.2907,
+      "step": 20510
+    },
+    {
+      "epoch": 0.4103,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00014685466849812418,
+      "loss": 1.2714,
+      "step": 20515
+    },
+    {
+      "epoch": 0.4104,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00014682382781476146,
+      "loss": 1.2783,
+      "step": 20520
+    },
+    {
+      "epoch": 0.4105,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00014679298142605734,
+      "loss": 1.2941,
+      "step": 20525
+    },
+    {
+      "epoch": 0.4106,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001467621293357704,
+      "loss": 1.283,
+      "step": 20530
+    },
+    {
+      "epoch": 0.4107,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001467312715476598,
+      "loss": 1.2741,
+      "step": 20535
+    },
+    {
+      "epoch": 0.4108,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00014670040806548555,
+      "loss": 1.272,
+      "step": 20540
+    },
+    {
+      "epoch": 0.4109,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001466695388930082,
+      "loss": 1.2678,
+      "step": 20545
+    },
+    {
+      "epoch": 0.411,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00014663866403398913,
+      "loss": 1.2767,
+      "step": 20550
+    },
+    {
+      "epoch": 0.4111,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001466077834921903,
+      "loss": 1.2645,
+      "step": 20555
+    },
+    {
+      "epoch": 0.4112,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00014657689727137443,
+      "loss": 1.2724,
+      "step": 20560
+    },
+    {
+      "epoch": 0.4113,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.0001465460053753049,
+      "loss": 1.2648,
+      "step": 20565
+    },
+    {
+      "epoch": 0.4114,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00014651510780774583,
+      "loss": 1.2795,
+      "step": 20570
+    },
+    {
+      "epoch": 0.4115,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.000146484204572462,
+      "loss": 1.2733,
+      "step": 20575
+    },
+    {
+      "epoch": 0.4116,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001464532956732188,
+      "loss": 1.2653,
+      "step": 20580
+    },
+    {
+      "epoch": 0.4117,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001464223811137824,
+      "loss": 1.2685,
+      "step": 20585
+    },
+    {
+      "epoch": 0.4118,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001463914608979197,
+      "loss": 1.2818,
+      "step": 20590
+    },
+    {
+      "epoch": 0.4119,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00014636053502939823,
+      "loss": 1.268,
+      "step": 20595
+    },
+    {
+      "epoch": 0.412,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00014632960351198618,
+      "loss": 1.2722,
+      "step": 20600
+    },
+    {
+      "epoch": 0.4121,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00014629866634945248,
+      "loss": 1.2997,
+      "step": 20605
+    },
+    {
+      "epoch": 0.4122,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0001462677235455667,
+      "loss": 1.2734,
+      "step": 20610
+    },
+    {
+      "epoch": 0.4123,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00014623677510409918,
+      "loss": 1.2712,
+      "step": 20615
+    },
+    {
+      "epoch": 0.4124,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00014620582102882089,
+      "loss": 1.2783,
+      "step": 20620
+    },
+    {
+      "epoch": 0.4125,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00014617486132350343,
+      "loss": 1.2719,
+      "step": 20625
+    },
+    {
+      "epoch": 0.4126,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00014614389599191917,
+      "loss": 1.2681,
+      "step": 20630
+    },
+    {
+      "epoch": 0.4127,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00014611292503784117,
+      "loss": 1.2733,
+      "step": 20635
+    },
+    {
+      "epoch": 0.4128,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001460819484650431,
+      "loss": 1.2713,
+      "step": 20640
+    },
+    {
+      "epoch": 0.4129,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0001460509662772994,
+      "loss": 1.2691,
+      "step": 20645
+    },
+    {
+      "epoch": 0.413,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00014601997847838518,
+      "loss": 1.2679,
+      "step": 20650
+    },
+    {
+      "epoch": 0.4131,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00014598898507207615,
+      "loss": 1.2672,
+      "step": 20655
+    },
+    {
+      "epoch": 0.4132,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00014595798606214882,
+      "loss": 1.2611,
+      "step": 20660
+    },
+    {
+      "epoch": 0.4133,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00014592698145238028,
+      "loss": 1.2696,
+      "step": 20665
+    },
+    {
+      "epoch": 0.4134,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00014589597124654833,
+      "loss": 1.2734,
+      "step": 20670
+    },
+    {
+      "epoch": 0.4135,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00014586495544843152,
+      "loss": 1.2646,
+      "step": 20675
+    },
+    {
+      "epoch": 0.4136,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.000145833934061809,
+      "loss": 1.2734,
+      "step": 20680
+    },
+    {
+      "epoch": 0.4137,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00014580290709046066,
+      "loss": 1.271,
+      "step": 20685
+    },
+    {
+      "epoch": 0.4138,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.000145771874538167,
+      "loss": 1.2906,
+      "step": 20690
+    },
+    {
+      "epoch": 0.4139,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0001457408364087093,
+      "loss": 1.2771,
+      "step": 20695
+    },
+    {
+      "epoch": 0.414,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00014570979270586945,
+      "loss": 1.2488,
+      "step": 20700
+    },
+    {
+      "epoch": 0.4141,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00014567874343342997,
+      "loss": 1.2668,
+      "step": 20705
+    },
+    {
+      "epoch": 0.4142,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00014564768859517418,
+      "loss": 1.2771,
+      "step": 20710
+    },
+    {
+      "epoch": 0.4143,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00014561662819488597,
+      "loss": 1.27,
+      "step": 20715
+    },
+    {
+      "epoch": 0.4144,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00014558556223635003,
+      "loss": 1.2658,
+      "step": 20720
+    },
+    {
+      "epoch": 0.4145,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00014555449072335157,
+      "loss": 1.2819,
+      "step": 20725
+    },
+    {
+      "epoch": 0.4146,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00014552341365967658,
+      "loss": 1.2811,
+      "step": 20730
+    },
+    {
+      "epoch": 0.4147,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00014549233104911178,
+      "loss": 1.2737,
+      "step": 20735
+    },
+    {
+      "epoch": 0.4148,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.0001454612428954444,
+      "loss": 1.2614,
+      "step": 20740
+    },
+    {
+      "epoch": 0.4149,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00014543014920246247,
+      "loss": 1.2974,
+      "step": 20745
+    },
+    {
+      "epoch": 0.415,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00014539904997395468,
+      "loss": 1.2794,
+      "step": 20750
+    },
+    {
+      "epoch": 0.4151,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00014536794521371037,
+      "loss": 1.2668,
+      "step": 20755
+    },
+    {
+      "epoch": 0.4152,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00014533683492551952,
+      "loss": 1.2538,
+      "step": 20760
+    },
+    {
+      "epoch": 0.4153,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0001453057191131729,
+      "loss": 1.2723,
+      "step": 20765
+    },
+    {
+      "epoch": 0.4154,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0001452745977804618,
+      "loss": 1.2735,
+      "step": 20770
+    },
+    {
+      "epoch": 0.4155,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00014524347093117828,
+      "loss": 1.2711,
+      "step": 20775
+    },
+    {
+      "epoch": 0.4156,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00014521233856911508,
+      "loss": 1.2654,
+      "step": 20780
+    },
+    {
+      "epoch": 0.4157,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00014518120069806557,
+      "loss": 1.2746,
+      "step": 20785
+    },
+    {
+      "epoch": 0.4158,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00014515005732182383,
+      "loss": 1.2771,
+      "step": 20790
+    },
+    {
+      "epoch": 0.4159,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00014511890844418453,
+      "loss": 1.2641,
+      "step": 20795
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00014508775406894307,
+      "loss": 1.2809,
+      "step": 20800
+    },
+    {
+      "epoch": 0.4161,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0001450565941998956,
+      "loss": 1.2695,
+      "step": 20805
+    },
+    {
+      "epoch": 0.4162,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00014502542884083875,
+      "loss": 1.278,
+      "step": 20810
+    },
+    {
+      "epoch": 0.4163,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00014499425799557,
+      "loss": 1.2781,
+      "step": 20815
+    },
+    {
+      "epoch": 0.4164,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0001449630816678874,
+      "loss": 1.2838,
+      "step": 20820
+    },
+    {
+      "epoch": 0.4165,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00014493189986158965,
+      "loss": 1.2619,
+      "step": 20825
+    },
+    {
+      "epoch": 0.4166,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00014490071258047623,
+      "loss": 1.2799,
+      "step": 20830
+    },
+    {
+      "epoch": 0.4167,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001448695198283472,
+      "loss": 1.2659,
+      "step": 20835
+    },
+    {
+      "epoch": 0.4168,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00014483832160900326,
+      "loss": 1.2855,
+      "step": 20840
+    },
+    {
+      "epoch": 0.4169,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001448071179262458,
+      "loss": 1.2705,
+      "step": 20845
+    },
+    {
+      "epoch": 0.417,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00014477590878387696,
+      "loss": 1.272,
+      "step": 20850
+    },
+    {
+      "epoch": 0.4171,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001447446941856995,
+      "loss": 1.2724,
+      "step": 20855
+    },
+    {
+      "epoch": 0.4172,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00014471347413551672,
+      "loss": 1.2782,
+      "step": 20860
+    },
+    {
+      "epoch": 0.4173,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00014468224863713278,
+      "loss": 1.2777,
+      "step": 20865
+    },
+    {
+      "epoch": 0.4174,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00014465101769435234,
+      "loss": 1.2623,
+      "step": 20870
+    },
+    {
+      "epoch": 0.4175,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00014461978131098088,
+      "loss": 1.2688,
+      "step": 20875
+    },
+    {
+      "epoch": 0.4176,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00014458853949082443,
+      "loss": 1.2758,
+      "step": 20880
+    },
+    {
+      "epoch": 0.4177,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00014455729223768966,
+      "loss": 1.2559,
+      "step": 20885
+    },
+    {
+      "epoch": 0.4178,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00014452603955538397,
+      "loss": 1.2848,
+      "step": 20890
+    },
+    {
+      "epoch": 0.4179,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00014449478144771543,
+      "loss": 1.2835,
+      "step": 20895
+    },
+    {
+      "epoch": 0.418,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00014446351791849276,
+      "loss": 1.2853,
+      "step": 20900
+    },
+    {
+      "epoch": 0.4181,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.0001444322489715253,
+      "loss": 1.2675,
+      "step": 20905
+    },
+    {
+      "epoch": 0.4182,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00014440097461062307,
+      "loss": 1.2902,
+      "step": 20910
+    },
+    {
+      "epoch": 0.4183,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00014436969483959676,
+      "loss": 1.2904,
+      "step": 20915
+    },
+    {
+      "epoch": 0.4184,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00014433840966225772,
+      "loss": 1.2736,
+      "step": 20920
+    },
+    {
+      "epoch": 0.4185,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00014430711908241798,
+      "loss": 1.2804,
+      "step": 20925
+    },
+    {
+      "epoch": 0.4186,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001442758231038902,
+      "loss": 1.2605,
+      "step": 20930
+    },
+    {
+      "epoch": 0.4187,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001442445217304876,
+      "loss": 1.2839,
+      "step": 20935
+    },
+    {
+      "epoch": 0.4188,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00014421321496602428,
+      "loss": 1.2729,
+      "step": 20940
+    },
+    {
+      "epoch": 0.4189,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00014418190281431482,
+      "loss": 1.2758,
+      "step": 20945
+    },
+    {
+      "epoch": 0.419,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00014415058527917452,
+      "loss": 1.275,
+      "step": 20950
+    },
+    {
+      "epoch": 0.4191,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00014411926236441934,
+      "loss": 1.2835,
+      "step": 20955
+    },
+    {
+      "epoch": 0.4192,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00014408793407386588,
+      "loss": 1.282,
+      "step": 20960
+    },
+    {
+      "epoch": 0.4193,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00014405660041133132,
+      "loss": 1.2712,
+      "step": 20965
+    },
+    {
+      "epoch": 0.4194,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00014402526138063373,
+      "loss": 1.267,
+      "step": 20970
+    },
+    {
+      "epoch": 0.4195,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00014399391698559152,
+      "loss": 1.2658,
+      "step": 20975
+    },
+    {
+      "epoch": 0.4196,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.000143962567230024,
+      "loss": 1.28,
+      "step": 20980
+    },
+    {
+      "epoch": 0.4197,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.000143931212117751,
+      "loss": 1.269,
+      "step": 20985
+    },
+    {
+      "epoch": 0.4198,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00014389985165259308,
+      "loss": 1.2836,
+      "step": 20990
+    },
+    {
+      "epoch": 0.4199,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001438684858383714,
+      "loss": 1.2704,
+      "step": 20995
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00014383711467890774,
+      "loss": 1.2602,
+      "step": 21000
+    },
+    {
+      "epoch": 0.4201,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00014380573817802467,
+      "loss": 1.2723,
+      "step": 21005
+    },
+    {
+      "epoch": 0.4202,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00014377435633954527,
+      "loss": 1.2789,
+      "step": 21010
+    },
+    {
+      "epoch": 0.4203,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00014374296916729336,
+      "loss": 1.2553,
+      "step": 21015
+    },
+    {
+      "epoch": 0.4204,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001437115766650933,
+      "loss": 1.272,
+      "step": 21020
+    },
+    {
+      "epoch": 0.4205,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00014368017883677024,
+      "loss": 1.2664,
+      "step": 21025
+    },
+    {
+      "epoch": 0.4206,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001436487756861499,
+      "loss": 1.2751,
+      "step": 21030
+    },
+    {
+      "epoch": 0.4207,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001436173672170586,
+      "loss": 1.2618,
+      "step": 21035
+    },
+    {
+      "epoch": 0.4208,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00014358595343332342,
+      "loss": 1.2966,
+      "step": 21040
+    },
+    {
+      "epoch": 0.4209,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00014355453433877204,
+      "loss": 1.2617,
+      "step": 21045
+    },
+    {
+      "epoch": 0.421,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00014352310993723277,
+      "loss": 1.2944,
+      "step": 21050
+    },
+    {
+      "epoch": 0.4211,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00014349168023253456,
+      "loss": 1.2768,
+      "step": 21055
+    },
+    {
+      "epoch": 0.4212,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00014346024522850703,
+      "loss": 1.2619,
+      "step": 21060
+    },
+    {
+      "epoch": 0.4213,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00014342880492898048,
+      "loss": 1.2823,
+      "step": 21065
+    },
+    {
+      "epoch": 0.4214,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00014339735933778576,
+      "loss": 1.2771,
+      "step": 21070
+    },
+    {
+      "epoch": 0.4215,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00014336590845875446,
+      "loss": 1.2893,
+      "step": 21075
+    },
+    {
+      "epoch": 0.4216,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00014333445229571873,
+      "loss": 1.2846,
+      "step": 21080
+    },
+    {
+      "epoch": 0.4217,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00014330299085251144,
+      "loss": 1.2832,
+      "step": 21085
+    },
+    {
+      "epoch": 0.4218,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00014327152413296608,
+      "loss": 1.2833,
+      "step": 21090
+    },
+    {
+      "epoch": 0.4219,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00014324005214091676,
+      "loss": 1.2762,
+      "step": 21095
+    },
+    {
+      "epoch": 0.422,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00014320857488019824,
+      "loss": 1.2708,
+      "step": 21100
+    },
+    {
+      "epoch": 0.4221,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00014317709235464593,
+      "loss": 1.2595,
+      "step": 21105
+    },
+    {
+      "epoch": 0.4222,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001431456045680959,
+      "loss": 1.2825,
+      "step": 21110
+    },
+    {
+      "epoch": 0.4223,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00014311411152438482,
+      "loss": 1.2656,
+      "step": 21115
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00014308261322735005,
+      "loss": 1.2657,
+      "step": 21120
+    },
+    {
+      "epoch": 0.4225,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00014305110968082952,
+      "loss": 1.2672,
+      "step": 21125
+    },
+    {
+      "epoch": 0.4226,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014301960088866186,
+      "loss": 1.2794,
+      "step": 21130
+    },
+    {
+      "epoch": 0.4227,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00014298808685468635,
+      "loss": 1.2957,
+      "step": 21135
+    },
+    {
+      "epoch": 0.4228,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00014295656758274284,
+      "loss": 1.2683,
+      "step": 21140
+    },
+    {
+      "epoch": 0.4229,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00014292504307667186,
+      "loss": 1.2705,
+      "step": 21145
+    },
+    {
+      "epoch": 0.423,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001428935133403146,
+      "loss": 1.2706,
+      "step": 21150
+    },
+    {
+      "epoch": 0.4231,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00014286197837751286,
+      "loss": 1.2625,
+      "step": 21155
+    },
+    {
+      "epoch": 0.4232,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00014283043819210905,
+      "loss": 1.2695,
+      "step": 21160
+    },
+    {
+      "epoch": 0.4233,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00014279889278794627,
+      "loss": 1.29,
+      "step": 21165
+    },
+    {
+      "epoch": 0.4234,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00014276734216886821,
+      "loss": 1.2728,
+      "step": 21170
+    },
+    {
+      "epoch": 0.4235,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00014273578633871927,
+      "loss": 1.2778,
+      "step": 21175
+    },
+    {
+      "epoch": 0.4236,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00014270422530134432,
+      "loss": 1.2923,
+      "step": 21180
+    },
+    {
+      "epoch": 0.4237,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00014267265906058914,
+      "loss": 1.2922,
+      "step": 21185
+    },
+    {
+      "epoch": 0.4238,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001426410876202999,
+      "loss": 1.307,
+      "step": 21190
+    },
+    {
+      "epoch": 0.4239,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00014260951098432343,
+      "loss": 1.2665,
+      "step": 21195
+    },
+    {
+      "epoch": 0.424,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00014257792915650728,
+      "loss": 1.272,
+      "step": 21200
+    },
+    {
+      "epoch": 0.4241,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00014254634214069963,
+      "loss": 1.2705,
+      "step": 21205
+    },
+    {
+      "epoch": 0.4242,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00014251474994074928,
+      "loss": 1.2982,
+      "step": 21210
+    },
+    {
+      "epoch": 0.4243,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014248315256050557,
+      "loss": 1.2753,
+      "step": 21215
+    },
+    {
+      "epoch": 0.4244,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001424515500038186,
+      "loss": 1.2657,
+      "step": 21220
+    },
+    {
+      "epoch": 0.4245,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00014241994227453901,
+      "loss": 1.2871,
+      "step": 21225
+    },
+    {
+      "epoch": 0.4246,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00014238832937651816,
+      "loss": 1.2742,
+      "step": 21230
+    },
+    {
+      "epoch": 0.4247,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00014235671131360798,
+      "loss": 1.2588,
+      "step": 21235
+    },
+    {
+      "epoch": 0.4248,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00014232508808966098,
+      "loss": 1.2694,
+      "step": 21240
+    },
+    {
+      "epoch": 0.4249,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00014229345970853032,
+      "loss": 1.275,
+      "step": 21245
+    },
+    {
+      "epoch": 0.425,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00014226182617406996,
+      "loss": 1.2675,
+      "step": 21250
+    },
+    {
+      "epoch": 0.4251,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00014223018749013423,
+      "loss": 1.2839,
+      "step": 21255
+    },
+    {
+      "epoch": 0.4252,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.0001421985436605783,
+      "loss": 1.284,
+      "step": 21260
+    },
+    {
+      "epoch": 0.4253,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001421668946892578,
+      "loss": 1.2713,
+      "step": 21265
+    },
+    {
+      "epoch": 0.4254,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.0001421352405800291,
+      "loss": 1.2749,
+      "step": 21270
+    },
+    {
+      "epoch": 0.4255,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014210358133674912,
+      "loss": 1.2781,
+      "step": 21275
+    },
+    {
+      "epoch": 0.4256,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00014207191696327548,
+      "loss": 1.2684,
+      "step": 21280
+    },
+    {
+      "epoch": 0.4257,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00014204024746346637,
+      "loss": 1.2812,
+      "step": 21285
+    },
+    {
+      "epoch": 0.4258,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014200857284118066,
+      "loss": 1.2789,
+      "step": 21290
+    },
+    {
+      "epoch": 0.4259,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00014197689310027772,
+      "loss": 1.2816,
+      "step": 21295
+    },
+    {
+      "epoch": 0.426,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00014194520824461771,
+      "loss": 1.2795,
+      "step": 21300
+    },
+    {
+      "epoch": 0.4261,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00014191351827806133,
+      "loss": 1.274,
+      "step": 21305
+    },
+    {
+      "epoch": 0.4262,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00014188182320446985,
+      "loss": 1.2781,
+      "step": 21310
+    },
+    {
+      "epoch": 0.4263,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00014185012302770527,
+      "loss": 1.2628,
+      "step": 21315
+    },
+    {
+      "epoch": 0.4264,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00014181841775163013,
+      "loss": 1.2664,
+      "step": 21320
+    },
+    {
+      "epoch": 0.4265,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001417867073801077,
+      "loss": 1.2728,
+      "step": 21325
+    },
+    {
+      "epoch": 0.4266,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00014175499191700167,
+      "loss": 1.2858,
+      "step": 21330
+    },
+    {
+      "epoch": 0.4267,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00014172327136617656,
+      "loss": 1.2743,
+      "step": 21335
+    },
+    {
+      "epoch": 0.4268,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00014169154573149737,
+      "loss": 1.2814,
+      "step": 21340
+    },
+    {
+      "epoch": 0.4269,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0001416598150168298,
+      "loss": 1.2898,
+      "step": 21345
+    },
+    {
+      "epoch": 0.427,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00014162807922604012,
+      "loss": 1.2976,
+      "step": 21350
+    },
+    {
+      "epoch": 0.4271,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00014159633836299527,
+      "loss": 1.2692,
+      "step": 21355
+    },
+    {
+      "epoch": 0.4272,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001415645924315628,
+      "loss": 1.2726,
+      "step": 21360
+    },
+    {
+      "epoch": 0.4273,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001415328414356108,
+      "loss": 1.2786,
+      "step": 21365
+    },
+    {
+      "epoch": 0.4274,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00014150108537900805,
+      "loss": 1.2686,
+      "step": 21370
+    },
+    {
+      "epoch": 0.4275,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00014146932426562392,
+      "loss": 1.2681,
+      "step": 21375
+    },
+    {
+      "epoch": 0.4276,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00014143755809932845,
+      "loss": 1.2801,
+      "step": 21380
+    },
+    {
+      "epoch": 0.4277,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00014140578688399218,
+      "loss": 1.2713,
+      "step": 21385
+    },
+    {
+      "epoch": 0.4278,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00014137401062348638,
+      "loss": 1.2682,
+      "step": 21390
+    },
+    {
+      "epoch": 0.4279,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0001413422293216829,
+      "loss": 1.2783,
+      "step": 21395
+    },
+    {
+      "epoch": 0.428,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001413104429824542,
+      "loss": 1.2729,
+      "step": 21400
+    },
+    {
+      "epoch": 0.4281,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001412786516096733,
+      "loss": 1.2664,
+      "step": 21405
+    },
+    {
+      "epoch": 0.4282,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00014124685520721392,
+      "loss": 1.2757,
+      "step": 21410
+    },
+    {
+      "epoch": 0.4283,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00014121505377895038,
+      "loss": 1.2632,
+      "step": 21415
+    },
+    {
+      "epoch": 0.4284,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00014118324732875748,
+      "loss": 1.2825,
+      "step": 21420
+    },
+    {
+      "epoch": 0.4285,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00014115143586051088,
+      "loss": 1.2634,
+      "step": 21425
+    },
+    {
+      "epoch": 0.4286,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00014111961937808665,
+      "loss": 1.2694,
+      "step": 21430
+    },
+    {
+      "epoch": 0.4287,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001410877978853615,
+      "loss": 1.2868,
+      "step": 21435
+    },
+    {
+      "epoch": 0.4288,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001410559713862128,
+      "loss": 1.2804,
+      "step": 21440
+    },
+    {
+      "epoch": 0.4289,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00014102413988451856,
+      "loss": 1.2873,
+      "step": 21445
+    },
+    {
+      "epoch": 0.429,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00014099230338415728,
+      "loss": 1.2625,
+      "step": 21450
+    },
+    {
+      "epoch": 0.4291,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00014096046188900822,
+      "loss": 1.2779,
+      "step": 21455
+    },
+    {
+      "epoch": 0.4292,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00014092861540295108,
+      "loss": 1.2722,
+      "step": 21460
+    },
+    {
+      "epoch": 0.4293,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001408967639298663,
+      "loss": 1.2712,
+      "step": 21465
+    },
+    {
+      "epoch": 0.4294,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00014086490747363493,
+      "loss": 1.2784,
+      "step": 21470
+    },
+    {
+      "epoch": 0.4295,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00014083304603813848,
+      "loss": 1.2691,
+      "step": 21475
+    },
+    {
+      "epoch": 0.4296,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0001408011796272593,
+      "loss": 1.2654,
+      "step": 21480
+    },
+    {
+      "epoch": 0.4297,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001407693082448801,
+      "loss": 1.2825,
+      "step": 21485
+    },
+    {
+      "epoch": 0.4298,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00014073743189488435,
+      "loss": 1.2742,
+      "step": 21490
+    },
+    {
+      "epoch": 0.4299,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00014070555058115614,
+      "loss": 1.2808,
+      "step": 21495
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00014067366430758004,
+      "loss": 1.2933,
+      "step": 21500
+    },
+    {
+      "epoch": 0.4301,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00014064177307804133,
+      "loss": 1.28,
+      "step": 21505
+    },
+    {
+      "epoch": 0.4302,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00014060987689642581,
+      "loss": 1.2702,
+      "step": 21510
+    },
+    {
+      "epoch": 0.4303,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00014057797576662,
+      "loss": 1.2671,
+      "step": 21515
+    },
+    {
+      "epoch": 0.4304,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00014054606969251095,
+      "loss": 1.2824,
+      "step": 21520
+    },
+    {
+      "epoch": 0.4305,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00014051415867798628,
+      "loss": 1.2679,
+      "step": 21525
+    },
+    {
+      "epoch": 0.4306,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00014048224272693424,
+      "loss": 1.2827,
+      "step": 21530
+    },
+    {
+      "epoch": 0.4307,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00014045032184324373,
+      "loss": 1.2732,
+      "step": 21535
+    },
+    {
+      "epoch": 0.4308,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00014041839603080422,
+      "loss": 1.2766,
+      "step": 21540
+    },
+    {
+      "epoch": 0.4309,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00014038646529350579,
+      "loss": 1.2716,
+      "step": 21545
+    },
+    {
+      "epoch": 0.431,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00014035452963523902,
+      "loss": 1.262,
+      "step": 21550
+    },
+    {
+      "epoch": 0.4311,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00014032258905989522,
+      "loss": 1.2846,
+      "step": 21555
+    },
+    {
+      "epoch": 0.4312,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00014029064357136628,
+      "loss": 1.27,
+      "step": 21560
+    },
+    {
+      "epoch": 0.4313,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00014025869317354462,
+      "loss": 1.2674,
+      "step": 21565
+    },
+    {
+      "epoch": 0.4314,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00014022673787032332,
+      "loss": 1.266,
+      "step": 21570
+    },
+    {
+      "epoch": 0.4315,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00014019477766559604,
+      "loss": 1.2667,
+      "step": 21575
+    },
+    {
+      "epoch": 0.4316,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.000140162812563257,
+      "loss": 1.2634,
+      "step": 21580
+    },
+    {
+      "epoch": 0.4317,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00014013084256720107,
+      "loss": 1.2767,
+      "step": 21585
+    },
+    {
+      "epoch": 0.4318,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014009886768132375,
+      "loss": 1.2643,
+      "step": 21590
+    },
+    {
+      "epoch": 0.4319,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.000140066887909521,
+      "loss": 1.2622,
+      "step": 21595
+    },
+    {
+      "epoch": 0.432,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00014003490325568954,
+      "loss": 1.2848,
+      "step": 21600
+    },
+    {
+      "epoch": 0.4321,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00014000291372372647,
+      "loss": 1.2663,
+      "step": 21605
+    },
+    {
+      "epoch": 0.4322,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00013997091931752977,
+      "loss": 1.2849,
+      "step": 21610
+    },
+    {
+      "epoch": 0.4323,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00013993892004099777,
+      "loss": 1.2682,
+      "step": 21615
+    },
+    {
+      "epoch": 0.4324,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00013990691589802954,
+      "loss": 1.2755,
+      "step": 21620
+    },
+    {
+      "epoch": 0.4325,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00013987490689252463,
+      "loss": 1.2872,
+      "step": 21625
+    },
+    {
+      "epoch": 0.4326,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00013984289302838328,
+      "loss": 1.2625,
+      "step": 21630
+    },
+    {
+      "epoch": 0.4327,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00013981087430950628,
+      "loss": 1.2712,
+      "step": 21635
+    },
+    {
+      "epoch": 0.4328,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.000139778850739795,
+      "loss": 1.2814,
+      "step": 21640
+    },
+    {
+      "epoch": 0.4329,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001397468223231514,
+      "loss": 1.292,
+      "step": 21645
+    },
+    {
+      "epoch": 0.433,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00013971478906347806,
+      "loss": 1.2969,
+      "step": 21650
+    },
+    {
+      "epoch": 0.4331,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001396827509646782,
+      "loss": 1.2767,
+      "step": 21655
+    },
+    {
+      "epoch": 0.4332,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00013965070803065543,
+      "loss": 1.2709,
+      "step": 21660
+    },
+    {
+      "epoch": 0.4333,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00013961866026531417,
+      "loss": 1.2757,
+      "step": 21665
+    },
+    {
+      "epoch": 0.4334,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00013958660767255938,
+      "loss": 1.2673,
+      "step": 21670
+    },
+    {
+      "epoch": 0.4335,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00013955455025629651,
+      "loss": 1.2728,
+      "step": 21675
+    },
+    {
+      "epoch": 0.4336,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00013952248802043165,
+      "loss": 1.2734,
+      "step": 21680
+    },
+    {
+      "epoch": 0.4337,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00013949042096887153,
+      "loss": 1.2682,
+      "step": 21685
+    },
+    {
+      "epoch": 0.4338,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001394583491055234,
+      "loss": 1.2755,
+      "step": 21690
+    },
+    {
+      "epoch": 0.4339,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00013942627243429512,
+      "loss": 1.2897,
+      "step": 21695
+    },
+    {
+      "epoch": 0.434,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00013939419095909512,
+      "loss": 1.2655,
+      "step": 21700
+    },
+    {
+      "epoch": 0.4341,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00013936210468383246,
+      "loss": 1.2859,
+      "step": 21705
+    },
+    {
+      "epoch": 0.4342,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00013933001361241673,
+      "loss": 1.2684,
+      "step": 21710
+    },
+    {
+      "epoch": 0.4343,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00013929791774875815,
+      "loss": 1.268,
+      "step": 21715
+    },
+    {
+      "epoch": 0.4344,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00013926581709676751,
+      "loss": 1.2693,
+      "step": 21720
+    },
+    {
+      "epoch": 0.4345,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00013923371166035616,
+      "loss": 1.2829,
+      "step": 21725
+    },
+    {
+      "epoch": 0.4346,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00013920160144343603,
+      "loss": 1.2907,
+      "step": 21730
+    },
+    {
+      "epoch": 0.4347,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001391694864499197,
+      "loss": 1.2641,
+      "step": 21735
+    },
+    {
+      "epoch": 0.4348,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013913736668372026,
+      "loss": 1.274,
+      "step": 21740
+    },
+    {
+      "epoch": 0.4349,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013910524214875137,
+      "loss": 1.2842,
+      "step": 21745
+    },
+    {
+      "epoch": 0.435,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00013907311284892736,
+      "loss": 1.279,
+      "step": 21750
+    },
+    {
+      "epoch": 0.4351,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00013904097878816312,
+      "loss": 1.2714,
+      "step": 21755
+    },
+    {
+      "epoch": 0.4352,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00013900883997037397,
+      "loss": 1.2682,
+      "step": 21760
+    },
+    {
+      "epoch": 0.4353,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00013897669639947606,
+      "loss": 1.2751,
+      "step": 21765
+    },
+    {
+      "epoch": 0.4354,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00013894454807938586,
+      "loss": 1.2664,
+      "step": 21770
+    },
+    {
+      "epoch": 0.4355,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001389123950140206,
+      "loss": 1.2891,
+      "step": 21775
+    },
+    {
+      "epoch": 0.4356,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001388802372072981,
+      "loss": 1.2741,
+      "step": 21780
+    },
+    {
+      "epoch": 0.4357,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00013884807466313663,
+      "loss": 1.2719,
+      "step": 21785
+    },
+    {
+      "epoch": 0.4358,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00013881590738545508,
+      "loss": 1.2752,
+      "step": 21790
+    },
+    {
+      "epoch": 0.4359,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00013878373537817292,
+      "loss": 1.2749,
+      "step": 21795
+    },
+    {
+      "epoch": 0.436,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001387515586452103,
+      "loss": 1.2907,
+      "step": 21800
+    },
+    {
+      "epoch": 0.4361,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00013871937719048779,
+      "loss": 1.2703,
+      "step": 21805
+    },
+    {
+      "epoch": 0.4362,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00013868719101792665,
+      "loss": 1.2597,
+      "step": 21810
+    },
+    {
+      "epoch": 0.4363,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00013865500013144857,
+      "loss": 1.2728,
+      "step": 21815
+    },
+    {
+      "epoch": 0.4364,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.000138622804534976,
+      "loss": 1.2609,
+      "step": 21820
+    },
+    {
+      "epoch": 0.4365,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00013859060423243187,
+      "loss": 1.2766,
+      "step": 21825
+    },
+    {
+      "epoch": 0.4366,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00013855839922773968,
+      "loss": 1.2679,
+      "step": 21830
+    },
+    {
+      "epoch": 0.4367,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00013852618952482347,
+      "loss": 1.2756,
+      "step": 21835
+    },
+    {
+      "epoch": 0.4368,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00013849397512760795,
+      "loss": 1.2842,
+      "step": 21840
+    },
+    {
+      "epoch": 0.4369,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001384617560400183,
+      "loss": 1.2885,
+      "step": 21845
+    },
+    {
+      "epoch": 0.437,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00013842953226598037,
+      "loss": 1.264,
+      "step": 21850
+    },
+    {
+      "epoch": 0.4371,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00013839730380942053,
+      "loss": 1.2726,
+      "step": 21855
+    },
+    {
+      "epoch": 0.4372,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00013836507067426564,
+      "loss": 1.2737,
+      "step": 21860
+    },
+    {
+      "epoch": 0.4373,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00013833283286444328,
+      "loss": 1.2913,
+      "step": 21865
+    },
+    {
+      "epoch": 0.4374,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001383005903838815,
+      "loss": 1.2564,
+      "step": 21870
+    },
+    {
+      "epoch": 0.4375,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.000138268343236509,
+      "loss": 1.2674,
+      "step": 21875
+    },
+    {
+      "epoch": 0.4376,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001382360914262549,
+      "loss": 1.2864,
+      "step": 21880
+    },
+    {
+      "epoch": 0.4377,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001382038349570491,
+      "loss": 1.2676,
+      "step": 21885
+    },
+    {
+      "epoch": 0.4378,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00013817157383282184,
+      "loss": 1.2685,
+      "step": 21890
+    },
+    {
+      "epoch": 0.4379,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00013813930805750413,
+      "loss": 1.2569,
+      "step": 21895
+    },
+    {
+      "epoch": 0.438,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00013810703763502744,
+      "loss": 1.2594,
+      "step": 21900
+    },
+    {
+      "epoch": 0.4381,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00013807476256932376,
+      "loss": 1.2717,
+      "step": 21905
+    },
+    {
+      "epoch": 0.4382,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00013804248286432578,
+      "loss": 1.3036,
+      "step": 21910
+    },
+    {
+      "epoch": 0.4383,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013801019852396665,
+      "loss": 1.2928,
+      "step": 21915
+    },
+    {
+      "epoch": 0.4384,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00013797790955218014,
+      "loss": 1.2855,
+      "step": 21920
+    },
+    {
+      "epoch": 0.4385,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00013794561595290052,
+      "loss": 1.267,
+      "step": 21925
+    },
+    {
+      "epoch": 0.4386,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001379133177300627,
+      "loss": 1.279,
+      "step": 21930
+    },
+    {
+      "epoch": 0.4387,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00013788101488760215,
+      "loss": 1.2741,
+      "step": 21935
+    },
+    {
+      "epoch": 0.4388,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00013784870742945482,
+      "loss": 1.2706,
+      "step": 21940
+    },
+    {
+      "epoch": 0.4389,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00013781639535955732,
+      "loss": 1.2875,
+      "step": 21945
+    },
+    {
+      "epoch": 0.439,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00013778407868184672,
+      "loss": 1.2654,
+      "step": 21950
+    },
+    {
+      "epoch": 0.4391,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00013775175740026078,
+      "loss": 1.268,
+      "step": 21955
+    },
+    {
+      "epoch": 0.4392,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00013771943151873767,
+      "loss": 1.2911,
+      "step": 21960
+    },
+    {
+      "epoch": 0.4393,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00013768710104121627,
+      "loss": 1.2838,
+      "step": 21965
+    },
+    {
+      "epoch": 0.4394,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00013765476597163594,
+      "loss": 1.2691,
+      "step": 21970
+    },
+    {
+      "epoch": 0.4395,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00013762242631393655,
+      "loss": 1.2777,
+      "step": 21975
+    },
+    {
+      "epoch": 0.4396,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001375900820720587,
+      "loss": 1.2676,
+      "step": 21980
+    },
+    {
+      "epoch": 0.4397,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001375577332499433,
+      "loss": 1.274,
+      "step": 21985
+    },
+    {
+      "epoch": 0.4398,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001375253798515321,
+      "loss": 1.2638,
+      "step": 21990
+    },
+    {
+      "epoch": 0.4399,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00013749302188076717,
+      "loss": 1.2826,
+      "step": 21995
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00013746065934159123,
+      "loss": 1.2946,
+      "step": 22000
+    },
+    {
+      "epoch": 0.4401,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00013742829223794759,
+      "loss": 1.2878,
+      "step": 22005
+    },
+    {
+      "epoch": 0.4402,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00013739592057378003,
+      "loss": 1.2656,
+      "step": 22010
+    },
+    {
+      "epoch": 0.4403,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00013736354435303305,
+      "loss": 1.2709,
+      "step": 22015
+    },
+    {
+      "epoch": 0.4404,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0001373311635796515,
+      "loss": 1.2523,
+      "step": 22020
+    },
+    {
+      "epoch": 0.4405,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001372987782575809,
+      "loss": 1.2678,
+      "step": 22025
+    },
+    {
+      "epoch": 0.4406,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001372663883907673,
+      "loss": 1.2821,
+      "step": 22030
+    },
+    {
+      "epoch": 0.4407,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00013723399398315734,
+      "loss": 1.2728,
+      "step": 22035
+    },
+    {
+      "epoch": 0.4408,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00013720159503869815,
+      "loss": 1.2681,
+      "step": 22040
+    },
+    {
+      "epoch": 0.4409,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00013716919156133746,
+      "loss": 1.2783,
+      "step": 22045
+    },
+    {
+      "epoch": 0.441,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00013713678355502351,
+      "loss": 1.2785,
+      "step": 22050
+    },
+    {
+      "epoch": 0.4411,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001371043710237051,
+      "loss": 1.2712,
+      "step": 22055
+    },
+    {
+      "epoch": 0.4412,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00013707195397133165,
+      "loss": 1.2617,
+      "step": 22060
+    },
+    {
+      "epoch": 0.4413,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001370395324018531,
+      "loss": 1.2834,
+      "step": 22065
+    },
+    {
+      "epoch": 0.4414,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00013700710631921984,
+      "loss": 1.2729,
+      "step": 22070
+    },
+    {
+      "epoch": 0.4415,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00013697467572738295,
+      "loss": 1.267,
+      "step": 22075
+    },
+    {
+      "epoch": 0.4416,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00013694224063029396,
+      "loss": 1.2932,
+      "step": 22080
+    },
+    {
+      "epoch": 0.4417,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00013690980103190503,
+      "loss": 1.3023,
+      "step": 22085
+    },
+    {
+      "epoch": 0.4418,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00013687735693616876,
+      "loss": 1.2593,
+      "step": 22090
+    },
+    {
+      "epoch": 0.4419,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00013684490834703843,
+      "loss": 1.2705,
+      "step": 22095
+    },
+    {
+      "epoch": 0.442,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00013681245526846783,
+      "loss": 1.2797,
+      "step": 22100
+    },
+    {
+      "epoch": 0.4421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00013677999770441115,
+      "loss": 1.2945,
+      "step": 22105
+    },
+    {
+      "epoch": 0.4422,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00013674753565882334,
+      "loss": 1.2694,
+      "step": 22110
+    },
+    {
+      "epoch": 0.4423,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001367150691356598,
+      "loss": 1.2717,
+      "step": 22115
+    },
+    {
+      "epoch": 0.4424,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00013668259813887643,
+      "loss": 1.2837,
+      "step": 22120
+    },
+    {
+      "epoch": 0.4425,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00013665012267242974,
+      "loss": 1.273,
+      "step": 22125
+    },
+    {
+      "epoch": 0.4426,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00013661764274027678,
+      "loss": 1.2782,
+      "step": 22130
+    },
+    {
+      "epoch": 0.4427,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00013658515834637512,
+      "loss": 1.2633,
+      "step": 22135
+    },
+    {
+      "epoch": 0.4428,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001365526694946829,
+      "loss": 1.2809,
+      "step": 22140
+    },
+    {
+      "epoch": 0.4429,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001365201761891588,
+      "loss": 1.2818,
+      "step": 22145
+    },
+    {
+      "epoch": 0.443,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00013648767843376196,
+      "loss": 1.2646,
+      "step": 22150
+    },
+    {
+      "epoch": 0.4431,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001364551762324522,
+      "loss": 1.2875,
+      "step": 22155
+    },
+    {
+      "epoch": 0.4432,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00013642266958918984,
+      "loss": 1.2818,
+      "step": 22160
+    },
+    {
+      "epoch": 0.4433,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00013639015850793563,
+      "loss": 1.2715,
+      "step": 22165
+    },
+    {
+      "epoch": 0.4434,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.000136357642992651,
+      "loss": 1.2753,
+      "step": 22170
+    },
+    {
+      "epoch": 0.4435,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00013632512304729785,
+      "loss": 1.2633,
+      "step": 22175
+    },
+    {
+      "epoch": 0.4436,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00013629259867583863,
+      "loss": 1.2745,
+      "step": 22180
+    },
+    {
+      "epoch": 0.4437,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00013626006988223636,
+      "loss": 1.2682,
+      "step": 22185
+    },
+    {
+      "epoch": 0.4438,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00013622753667045457,
+      "loss": 1.2832,
+      "step": 22190
+    },
+    {
+      "epoch": 0.4439,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00013619499904445734,
+      "loss": 1.2657,
+      "step": 22195
+    },
+    {
+      "epoch": 0.444,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00013616245700820922,
+      "loss": 1.2559,
+      "step": 22200
+    },
+    {
+      "epoch": 0.4441,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00013612991056567543,
+      "loss": 1.2965,
+      "step": 22205
+    },
+    {
+      "epoch": 0.4442,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00013609735972082166,
+      "loss": 1.2728,
+      "step": 22210
+    },
+    {
+      "epoch": 0.4443,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00013606480447761409,
+      "loss": 1.2812,
+      "step": 22215
+    },
+    {
+      "epoch": 0.4444,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00013603224484001948,
+      "loss": 1.2856,
+      "step": 22220
+    },
+    {
+      "epoch": 0.4445,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00013599968081200514,
+      "loss": 1.2667,
+      "step": 22225
+    },
+    {
+      "epoch": 0.4446,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001359671123975389,
+      "loss": 1.2595,
+      "step": 22230
+    },
+    {
+      "epoch": 0.4447,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00013593453960058908,
+      "loss": 1.2664,
+      "step": 22235
+    },
+    {
+      "epoch": 0.4448,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00013590196242512463,
+      "loss": 1.2653,
+      "step": 22240
+    },
+    {
+      "epoch": 0.4449,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00013586938087511494,
+      "loss": 1.2823,
+      "step": 22245
+    },
+    {
+      "epoch": 0.445,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00013583679495453,
+      "loss": 1.2858,
+      "step": 22250
+    },
+    {
+      "epoch": 0.4451,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013580420466734037,
+      "loss": 1.2757,
+      "step": 22255
+    },
+    {
+      "epoch": 0.4452,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00013577161001751694,
+      "loss": 1.2756,
+      "step": 22260
+    },
+    {
+      "epoch": 0.4453,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00013573901100903134,
+      "loss": 1.2693,
+      "step": 22265
+    },
+    {
+      "epoch": 0.4454,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00013570640764585566,
+      "loss": 1.2684,
+      "step": 22270
+    },
+    {
+      "epoch": 0.4455,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00013567379993196252,
+      "loss": 1.266,
+      "step": 22275
+    },
+    {
+      "epoch": 0.4456,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00013564118787132506,
+      "loss": 1.281,
+      "step": 22280
+    },
+    {
+      "epoch": 0.4457,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.000135608571467917,
+      "loss": 1.2774,
+      "step": 22285
+    },
+    {
+      "epoch": 0.4458,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0001355759507257125,
+      "loss": 1.2692,
+      "step": 22290
+    },
+    {
+      "epoch": 0.4459,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001355433256486863,
+      "loss": 1.2759,
+      "step": 22295
+    },
+    {
+      "epoch": 0.446,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0001355106962408137,
+      "loss": 1.2706,
+      "step": 22300
+    },
+    {
+      "epoch": 0.4461,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0001354780625060705,
+      "loss": 1.2799,
+      "step": 22305
+    },
+    {
+      "epoch": 0.4462,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00013544542444843299,
+      "loss": 1.2793,
+      "step": 22310
+    },
+    {
+      "epoch": 0.4463,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.000135412782071878,
+      "loss": 1.2786,
+      "step": 22315
+    },
+    {
+      "epoch": 0.4464,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00013538013538038295,
+      "loss": 1.2624,
+      "step": 22320
+    },
+    {
+      "epoch": 0.4465,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00013534748437792573,
+      "loss": 1.2622,
+      "step": 22325
+    },
+    {
+      "epoch": 0.4466,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00013531482906848475,
+      "loss": 1.2735,
+      "step": 22330
+    },
+    {
+      "epoch": 0.4467,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.000135282169456039,
+      "loss": 1.278,
+      "step": 22335
+    },
+    {
+      "epoch": 0.4468,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00013524950554456784,
+      "loss": 1.2788,
+      "step": 22340
+    },
+    {
+      "epoch": 0.4469,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00013521683733805145,
+      "loss": 1.2821,
+      "step": 22345
+    },
+    {
+      "epoch": 0.447,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00013518416484047018,
+      "loss": 1.2604,
+      "step": 22350
+    },
+    {
+      "epoch": 0.4471,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001351514880558052,
+      "loss": 1.2742,
+      "step": 22355
+    },
+    {
+      "epoch": 0.4472,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00013511880698803801,
+      "loss": 1.2771,
+      "step": 22360
+    },
+    {
+      "epoch": 0.4473,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00013508612164115068,
+      "loss": 1.2801,
+      "step": 22365
+    },
+    {
+      "epoch": 0.4474,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001350534320191259,
+      "loss": 1.2854,
+      "step": 22370
+    },
+    {
+      "epoch": 0.4475,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00013502073812594675,
+      "loss": 1.2836,
+      "step": 22375
+    },
+    {
+      "epoch": 0.4476,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00013498803996559692,
+      "loss": 1.277,
+      "step": 22380
+    },
+    {
+      "epoch": 0.4477,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001349553375420605,
+      "loss": 1.2706,
+      "step": 22385
+    },
+    {
+      "epoch": 0.4478,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00013492263085932224,
+      "loss": 1.2832,
+      "step": 22390
+    },
+    {
+      "epoch": 0.4479,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00013488991992136734,
+      "loss": 1.2603,
+      "step": 22395
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00013485720473218154,
+      "loss": 1.2686,
+      "step": 22400
+    },
+    {
+      "epoch": 0.4481,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00013482448529575106,
+      "loss": 1.2764,
+      "step": 22405
+    },
+    {
+      "epoch": 0.4482,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001347917616160627,
+      "loss": 1.2679,
+      "step": 22410
+    },
+    {
+      "epoch": 0.4483,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001347590336971037,
+      "loss": 1.2694,
+      "step": 22415
+    },
+    {
+      "epoch": 0.4484,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001347263015428619,
+      "loss": 1.2681,
+      "step": 22420
+    },
+    {
+      "epoch": 0.4485,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00013469356515732558,
+      "loss": 1.2929,
+      "step": 22425
+    },
+    {
+      "epoch": 0.4486,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00013466082454448362,
+      "loss": 1.2724,
+      "step": 22430
+    },
+    {
+      "epoch": 0.4487,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001346280797083253,
+      "loss": 1.2895,
+      "step": 22435
+    },
+    {
+      "epoch": 0.4488,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001345953306528405,
+      "loss": 1.2684,
+      "step": 22440
+    },
+    {
+      "epoch": 0.4489,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00013456257738201957,
+      "loss": 1.2792,
+      "step": 22445
+    },
+    {
+      "epoch": 0.449,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00013452981989985348,
+      "loss": 1.2815,
+      "step": 22450
+    },
+    {
+      "epoch": 0.4491,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00013449705821033355,
+      "loss": 1.257,
+      "step": 22455
+    },
+    {
+      "epoch": 0.4492,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001344642923174517,
+      "loss": 1.2777,
+      "step": 22460
+    },
+    {
+      "epoch": 0.4493,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00013443152222520038,
+      "loss": 1.267,
+      "step": 22465
+    },
+    {
+      "epoch": 0.4494,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00013439874793757254,
+      "loss": 1.2722,
+      "step": 22470
+    },
+    {
+      "epoch": 0.4495,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00013436596945856164,
+      "loss": 1.2701,
+      "step": 22475
+    },
+    {
+      "epoch": 0.4496,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00013433318679216153,
+      "loss": 1.2652,
+      "step": 22480
+    },
+    {
+      "epoch": 0.4497,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00013430039994236678,
+      "loss": 1.2626,
+      "step": 22485
+    },
+    {
+      "epoch": 0.4498,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00013426760891317236,
+      "loss": 1.2684,
+      "step": 22490
+    },
+    {
+      "epoch": 0.4499,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00013423481370857375,
+      "loss": 1.2697,
+      "step": 22495
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00013420201433256689,
+      "loss": 1.3023,
+      "step": 22500
+    },
+    {
+      "epoch": 0.4501,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00013416921078914835,
+      "loss": 1.2647,
+      "step": 22505
+    },
+    {
+      "epoch": 0.4502,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0001341364030823151,
+      "loss": 1.2661,
+      "step": 22510
+    },
+    {
+      "epoch": 0.4503,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001341035912160647,
+      "loss": 1.2684,
+      "step": 22515
+    },
+    {
+      "epoch": 0.4504,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001340707751943952,
+      "loss": 1.2599,
+      "step": 22520
+    },
+    {
+      "epoch": 0.4505,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00013403795502130503,
+      "loss": 1.2674,
+      "step": 22525
+    },
+    {
+      "epoch": 0.4506,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001340051307007933,
+      "loss": 1.2804,
+      "step": 22530
+    },
+    {
+      "epoch": 0.4507,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00013397230223685956,
+      "loss": 1.2939,
+      "step": 22535
+    },
+    {
+      "epoch": 0.4508,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00013393946963350382,
+      "loss": 1.278,
+      "step": 22540
+    },
+    {
+      "epoch": 0.4509,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00013390663289472666,
+      "loss": 1.2585,
+      "step": 22545
+    },
+    {
+      "epoch": 0.451,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00013387379202452917,
+      "loss": 1.2773,
+      "step": 22550
+    },
+    {
+      "epoch": 0.4511,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00013384094702691281,
+      "loss": 1.2778,
+      "step": 22555
+    },
+    {
+      "epoch": 0.4512,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00013380809790587974,
+      "loss": 1.255,
+      "step": 22560
+    },
+    {
+      "epoch": 0.4513,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00013377524466543248,
+      "loss": 1.2729,
+      "step": 22565
+    },
+    {
+      "epoch": 0.4514,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00013374238730957412,
+      "loss": 1.2604,
+      "step": 22570
+    },
+    {
+      "epoch": 0.4515,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00013370952584230823,
+      "loss": 1.2808,
+      "step": 22575
+    },
+    {
+      "epoch": 0.4516,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00013367666026763882,
+      "loss": 1.3037,
+      "step": 22580
+    },
+    {
+      "epoch": 0.4517,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00013364379058957056,
+      "loss": 1.2764,
+      "step": 22585
+    },
+    {
+      "epoch": 0.4518,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00013361091681210845,
+      "loss": 1.2606,
+      "step": 22590
+    },
+    {
+      "epoch": 0.4519,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013357803893925807,
+      "loss": 1.2852,
+      "step": 22595
+    },
+    {
+      "epoch": 0.452,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00013354515697502553,
+      "loss": 1.2743,
+      "step": 22600
+    },
+    {
+      "epoch": 0.4521,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00013351227092341732,
+      "loss": 1.28,
+      "step": 22605
+    },
+    {
+      "epoch": 0.4522,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0001334793807884406,
+      "loss": 1.2639,
+      "step": 22610
+    },
+    {
+      "epoch": 0.4523,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00013344648657410282,
+      "loss": 1.2697,
+      "step": 22615
+    },
+    {
+      "epoch": 0.4524,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00013341358828441218,
+      "loss": 1.2808,
+      "step": 22620
+    },
+    {
+      "epoch": 0.4525,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001333806859233771,
+      "loss": 1.2574,
+      "step": 22625
+    },
+    {
+      "epoch": 0.4526,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00013334777949500673,
+      "loss": 1.2757,
+      "step": 22630
+    },
+    {
+      "epoch": 0.4527,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00013331486900331057,
+      "loss": 1.2769,
+      "step": 22635
+    },
+    {
+      "epoch": 0.4528,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00013328195445229868,
+      "loss": 1.2704,
+      "step": 22640
+    },
+    {
+      "epoch": 0.4529,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0001332490358459816,
+      "loss": 1.2719,
+      "step": 22645
+    },
+    {
+      "epoch": 0.453,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00013321611318837032,
+      "loss": 1.2788,
+      "step": 22650
+    },
+    {
+      "epoch": 0.4531,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00013318318648347646,
+      "loss": 1.2869,
+      "step": 22655
+    },
+    {
+      "epoch": 0.4532,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00013315025573531198,
+      "loss": 1.2869,
+      "step": 22660
+    },
+    {
+      "epoch": 0.4533,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00013311732094788937,
+      "loss": 1.2591,
+      "step": 22665
+    },
+    {
+      "epoch": 0.4534,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00013308438212522163,
+      "loss": 1.2826,
+      "step": 22670
+    },
+    {
+      "epoch": 0.4535,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001330514392713223,
+      "loss": 1.2835,
+      "step": 22675
+    },
+    {
+      "epoch": 0.4536,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00013301849239020537,
+      "loss": 1.2649,
+      "step": 22680
+    },
+    {
+      "epoch": 0.4537,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00013298554148588528,
+      "loss": 1.2912,
+      "step": 22685
+    },
+    {
+      "epoch": 0.4538,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00013295258656237702,
+      "loss": 1.2776,
+      "step": 22690
+    },
+    {
+      "epoch": 0.4539,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00013291962762369608,
+      "loss": 1.2659,
+      "step": 22695
+    },
+    {
+      "epoch": 0.454,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00013288666467385833,
+      "loss": 1.2775,
+      "step": 22700
+    },
+    {
+      "epoch": 0.4541,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00013285369771688025,
+      "loss": 1.2688,
+      "step": 22705
+    },
+    {
+      "epoch": 0.4542,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001328207267567788,
+      "loss": 1.2909,
+      "step": 22710
+    },
+    {
+      "epoch": 0.4543,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00013278775179757131,
+      "loss": 1.2654,
+      "step": 22715
+    },
+    {
+      "epoch": 0.4544,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0001327547728432757,
+      "loss": 1.2761,
+      "step": 22720
+    },
+    {
+      "epoch": 0.4545,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.0001327217898979104,
+      "loss": 1.2719,
+      "step": 22725
+    },
+    {
+      "epoch": 0.4546,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00013268880296549425,
+      "loss": 1.2875,
+      "step": 22730
+    },
+    {
+      "epoch": 0.4547,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00013265581205004661,
+      "loss": 1.2899,
+      "step": 22735
+    },
+    {
+      "epoch": 0.4548,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00013262281715558736,
+      "loss": 1.2727,
+      "step": 22740
+    },
+    {
+      "epoch": 0.4549,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00013258981828613678,
+      "loss": 1.279,
+      "step": 22745
+    },
+    {
+      "epoch": 0.455,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00013255681544571568,
+      "loss": 1.2615,
+      "step": 22750
+    },
+    {
+      "epoch": 0.4551,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001325238086383454,
+      "loss": 1.303,
+      "step": 22755
+    },
+    {
+      "epoch": 0.4552,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00013249079786804765,
+      "loss": 1.2625,
+      "step": 22760
+    },
+    {
+      "epoch": 0.4553,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00013245778313884478,
+      "loss": 1.2723,
+      "step": 22765
+    },
+    {
+      "epoch": 0.4554,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00013242476445475944,
+      "loss": 1.2898,
+      "step": 22770
+    },
+    {
+      "epoch": 0.4555,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00013239174181981495,
+      "loss": 1.2715,
+      "step": 22775
+    },
+    {
+      "epoch": 0.4556,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00013235871523803496,
+      "loss": 1.2648,
+      "step": 22780
+    },
+    {
+      "epoch": 0.4557,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001323256847134437,
+      "loss": 1.2694,
+      "step": 22785
+    },
+    {
+      "epoch": 0.4558,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00013229265025006576,
+      "loss": 1.2709,
+      "step": 22790
+    },
+    {
+      "epoch": 0.4559,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00013225961185192638,
+      "loss": 1.2842,
+      "step": 22795
+    },
+    {
+      "epoch": 0.456,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00013222656952305113,
+      "loss": 1.2729,
+      "step": 22800
+    },
+    {
+      "epoch": 0.4561,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00013219352326746613,
+      "loss": 1.2896,
+      "step": 22805
+    },
+    {
+      "epoch": 0.4562,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00013216047308919798,
+      "loss": 1.2772,
+      "step": 22810
+    },
+    {
+      "epoch": 0.4563,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00013212741899227374,
+      "loss": 1.2866,
+      "step": 22815
+    },
+    {
+      "epoch": 0.4564,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00013209436098072095,
+      "loss": 1.2767,
+      "step": 22820
+    },
+    {
+      "epoch": 0.4565,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00013206129905856765,
+      "loss": 1.2729,
+      "step": 22825
+    },
+    {
+      "epoch": 0.4566,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00013202823322984228,
+      "loss": 1.2902,
+      "step": 22830
+    },
+    {
+      "epoch": 0.4567,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00013199516349857385,
+      "loss": 1.2884,
+      "step": 22835
+    },
+    {
+      "epoch": 0.4568,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001319620898687918,
+      "loss": 1.2811,
+      "step": 22840
+    },
+    {
+      "epoch": 0.4569,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00013192901234452607,
+      "loss": 1.2804,
+      "step": 22845
+    },
+    {
+      "epoch": 0.457,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00013189593092980702,
+      "loss": 1.2647,
+      "step": 22850
+    },
+    {
+      "epoch": 0.4571,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00013186284562866554,
+      "loss": 1.2792,
+      "step": 22855
+    },
+    {
+      "epoch": 0.4572,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013182975644513296,
+      "loss": 1.2711,
+      "step": 22860
+    },
+    {
+      "epoch": 0.4573,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00013179666338324108,
+      "loss": 1.2863,
+      "step": 22865
+    },
+    {
+      "epoch": 0.4574,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00013176356644702225,
+      "loss": 1.2736,
+      "step": 22870
+    },
+    {
+      "epoch": 0.4575,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00013173046564050924,
+      "loss": 1.268,
+      "step": 22875
+    },
+    {
+      "epoch": 0.4576,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001316973609677352,
+      "loss": 1.2691,
+      "step": 22880
+    },
+    {
+      "epoch": 0.4577,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00013166425243273385,
+      "loss": 1.2831,
+      "step": 22885
+    },
+    {
+      "epoch": 0.4578,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001316311400395394,
+      "loss": 1.2691,
+      "step": 22890
+    },
+    {
+      "epoch": 0.4579,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001315980237921865,
+      "loss": 1.2689,
+      "step": 22895
+    },
+    {
+      "epoch": 0.458,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00013156490369471027,
+      "loss": 1.2825,
+      "step": 22900
+    },
+    {
+      "epoch": 0.4581,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00013153177975114625,
+      "loss": 1.2652,
+      "step": 22905
+    },
+    {
+      "epoch": 0.4582,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001314986519655305,
+      "loss": 1.2575,
+      "step": 22910
+    },
+    {
+      "epoch": 0.4583,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00013146552034189955,
+      "loss": 1.2591,
+      "step": 22915
+    },
+    {
+      "epoch": 0.4584,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001314323848842904,
+      "loss": 1.2596,
+      "step": 22920
+    },
+    {
+      "epoch": 0.4585,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00013139924559674052,
+      "loss": 1.2571,
+      "step": 22925
+    },
+    {
+      "epoch": 0.4586,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00013136610248328778,
+      "loss": 1.2745,
+      "step": 22930
+    },
+    {
+      "epoch": 0.4587,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.0001313329555479706,
+      "loss": 1.2704,
+      "step": 22935
+    },
+    {
+      "epoch": 0.4588,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00013129980479482782,
+      "loss": 1.31,
+      "step": 22940
+    },
+    {
+      "epoch": 0.4589,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00013126665022789879,
+      "loss": 1.2734,
+      "step": 22945
+    },
+    {
+      "epoch": 0.459,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00013123349185122327,
+      "loss": 1.2825,
+      "step": 22950
+    },
+    {
+      "epoch": 0.4591,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001312003296688415,
+      "loss": 1.2781,
+      "step": 22955
+    },
+    {
+      "epoch": 0.4592,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001311671636847942,
+      "loss": 1.2738,
+      "step": 22960
+    },
+    {
+      "epoch": 0.4593,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00013113399390312256,
+      "loss": 1.2506,
+      "step": 22965
+    },
+    {
+      "epoch": 0.4594,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001311008203278682,
+      "loss": 1.2873,
+      "step": 22970
+    },
+    {
+      "epoch": 0.4595,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001310676429630732,
+      "loss": 1.2692,
+      "step": 22975
+    },
+    {
+      "epoch": 0.4596,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00013103446181278013,
+      "loss": 1.28,
+      "step": 22980
+    },
+    {
+      "epoch": 0.4597,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00013100127688103205,
+      "loss": 1.2789,
+      "step": 22985
+    },
+    {
+      "epoch": 0.4598,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00013096808817187242,
+      "loss": 1.2754,
+      "step": 22990
+    },
+    {
+      "epoch": 0.4599,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001309348956893452,
+      "loss": 1.2722,
+      "step": 22995
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00013090169943749476,
+      "loss": 1.2703,
+      "step": 23000
+    },
+    {
+      "epoch": 0.4601,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00013086849942036595,
+      "loss": 1.2592,
+      "step": 23005
+    },
+    {
+      "epoch": 0.4602,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00013083529564200417,
+      "loss": 1.2771,
+      "step": 23010
+    },
+    {
+      "epoch": 0.4603,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00013080208810645514,
+      "loss": 1.272,
+      "step": 23015
+    },
+    {
+      "epoch": 0.4604,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001307688768177651,
+      "loss": 1.2826,
+      "step": 23020
+    },
+    {
+      "epoch": 0.4605,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00013073566177998074,
+      "loss": 1.2713,
+      "step": 23025
+    },
+    {
+      "epoch": 0.4606,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.0001307024429971492,
+      "loss": 1.2859,
+      "step": 23030
+    },
+    {
+      "epoch": 0.4607,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00013066922047331813,
+      "loss": 1.2692,
+      "step": 23035
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00013063599421253558,
+      "loss": 1.2674,
+      "step": 23040
+    },
+    {
+      "epoch": 0.4609,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001306027642188501,
+      "loss": 1.2934,
+      "step": 23045
+    },
+    {
+      "epoch": 0.461,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00013056953049631057,
+      "loss": 1.2871,
+      "step": 23050
+    },
+    {
+      "epoch": 0.4611,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001305362930489665,
+      "loss": 1.2868,
+      "step": 23055
+    },
+    {
+      "epoch": 0.4612,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001305030518808678,
+      "loss": 1.2842,
+      "step": 23060
+    },
+    {
+      "epoch": 0.4613,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00013046980699606467,
+      "loss": 1.2768,
+      "step": 23065
+    },
+    {
+      "epoch": 0.4614,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00013043655839860803,
+      "loss": 1.2723,
+      "step": 23070
+    },
+    {
+      "epoch": 0.4615,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00013040330609254903,
+      "loss": 1.2801,
+      "step": 23075
+    },
+    {
+      "epoch": 0.4616,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00013037005008193942,
+      "loss": 1.2768,
+      "step": 23080
+    },
+    {
+      "epoch": 0.4617,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00013033679037083138,
+      "loss": 1.2745,
+      "step": 23085
+    },
+    {
+      "epoch": 0.4618,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00013030352696327742,
+      "loss": 1.2675,
+      "step": 23090
+    },
+    {
+      "epoch": 0.4619,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0001302702598633306,
+      "loss": 1.2783,
+      "step": 23095
+    },
+    {
+      "epoch": 0.462,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00013023698907504446,
+      "loss": 1.2738,
+      "step": 23100
+    },
+    {
+      "epoch": 0.4621,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001302037146024729,
+      "loss": 1.2768,
+      "step": 23105
+    },
+    {
+      "epoch": 0.4622,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00013017043644967036,
+      "loss": 1.2779,
+      "step": 23110
+    },
+    {
+      "epoch": 0.4623,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00013013715462069166,
+      "loss": 1.2781,
+      "step": 23115
+    },
+    {
+      "epoch": 0.4624,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00013010386911959206,
+      "loss": 1.282,
+      "step": 23120
+    },
+    {
+      "epoch": 0.4625,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00013007057995042732,
+      "loss": 1.2819,
+      "step": 23125
+    },
+    {
+      "epoch": 0.4626,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001300372871172536,
+      "loss": 1.2838,
+      "step": 23130
+    },
+    {
+      "epoch": 0.4627,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00013000399062412763,
+      "loss": 1.2521,
+      "step": 23135
+    },
+    {
+      "epoch": 0.4628,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0001299706904751064,
+      "loss": 1.281,
+      "step": 23140
+    },
+    {
+      "epoch": 0.4629,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001299373866742474,
+      "loss": 1.2816,
+      "step": 23145
+    },
+    {
+      "epoch": 0.463,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00012990407922560868,
+      "loss": 1.2639,
+      "step": 23150
+    },
+    {
+      "epoch": 0.4631,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00012987076813324858,
+      "loss": 1.2782,
+      "step": 23155
+    },
+    {
+      "epoch": 0.4632,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00012983745340122604,
+      "loss": 1.2721,
+      "step": 23160
+    },
+    {
+      "epoch": 0.4633,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0001298041350336003,
+      "loss": 1.3018,
+      "step": 23165
+    },
+    {
+      "epoch": 0.4634,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00012977081303443107,
+      "loss": 1.2696,
+      "step": 23170
+    },
+    {
+      "epoch": 0.4635,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001297374874077786,
+      "loss": 1.2881,
+      "step": 23175
+    },
+    {
+      "epoch": 0.4636,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001297041581577035,
+      "loss": 1.2744,
+      "step": 23180
+    },
+    {
+      "epoch": 0.4637,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00012967082528826684,
+      "loss": 1.2721,
+      "step": 23185
+    },
+    {
+      "epoch": 0.4638,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00012963748880353011,
+      "loss": 1.2663,
+      "step": 23190
+    },
+    {
+      "epoch": 0.4639,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00012960414870755524,
+      "loss": 1.2676,
+      "step": 23195
+    },
+    {
+      "epoch": 0.464,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00012957080500440468,
+      "loss": 1.2651,
+      "step": 23200
+    },
+    {
+      "epoch": 0.4641,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00012953745769814123,
+      "loss": 1.2672,
+      "step": 23205
+    },
+    {
+      "epoch": 0.4642,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00012950410679282815,
+      "loss": 1.2693,
+      "step": 23210
+    },
+    {
+      "epoch": 0.4643,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00012947075229252915,
+      "loss": 1.2766,
+      "step": 23215
+    },
+    {
+      "epoch": 0.4644,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00012943739420130836,
+      "loss": 1.2748,
+      "step": 23220
+    },
+    {
+      "epoch": 0.4645,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001294040325232304,
+      "loss": 1.2801,
+      "step": 23225
+    },
+    {
+      "epoch": 0.4646,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001293706672623603,
+      "loss": 1.2666,
+      "step": 23230
+    },
+    {
+      "epoch": 0.4647,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00012933729842276343,
+      "loss": 1.2734,
+      "step": 23235
+    },
+    {
+      "epoch": 0.4648,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00012930392600850573,
+      "loss": 1.277,
+      "step": 23240
+    },
+    {
+      "epoch": 0.4649,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0001292705500236536,
+      "loss": 1.2852,
+      "step": 23245
+    },
+    {
+      "epoch": 0.465,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00012923717047227368,
+      "loss": 1.2783,
+      "step": 23250
+    },
+    {
+      "epoch": 0.4651,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00012920378735843327,
+      "loss": 1.2732,
+      "step": 23255
+    },
+    {
+      "epoch": 0.4652,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0001291704006861999,
+      "loss": 1.2577,
+      "step": 23260
+    },
+    {
+      "epoch": 0.4653,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00012913701045964173,
+      "loss": 1.2651,
+      "step": 23265
+    },
+    {
+      "epoch": 0.4654,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012910361668282719,
+      "loss": 1.2892,
+      "step": 23270
+    },
+    {
+      "epoch": 0.4655,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00012907021935982526,
+      "loss": 1.2654,
+      "step": 23275
+    },
+    {
+      "epoch": 0.4656,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00012903681849470527,
+      "loss": 1.269,
+      "step": 23280
+    },
+    {
+      "epoch": 0.4657,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00012900341409153703,
+      "loss": 1.2555,
+      "step": 23285
+    },
+    {
+      "epoch": 0.4658,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00012897000615439075,
+      "loss": 1.2923,
+      "step": 23290
+    },
+    {
+      "epoch": 0.4659,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001289365946873371,
+      "loss": 1.2649,
+      "step": 23295
+    },
+    {
+      "epoch": 0.466,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00012890317969444716,
+      "loss": 1.2848,
+      "step": 23300
+    },
+    {
+      "epoch": 0.4661,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00012886976117979247,
+      "loss": 1.2709,
+      "step": 23305
+    },
+    {
+      "epoch": 0.4662,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00012883633914744492,
+      "loss": 1.2779,
+      "step": 23310
+    },
+    {
+      "epoch": 0.4663,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00012880291360147693,
+      "loss": 1.279,
+      "step": 23315
+    },
+    {
+      "epoch": 0.4664,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00012876948454596128,
+      "loss": 1.2694,
+      "step": 23320
+    },
+    {
+      "epoch": 0.4665,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00012873605198497124,
+      "loss": 1.2558,
+      "step": 23325
+    },
+    {
+      "epoch": 0.4666,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00012870261592258037,
+      "loss": 1.2634,
+      "step": 23330
+    },
+    {
+      "epoch": 0.4667,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00012866917636286286,
+      "loss": 1.276,
+      "step": 23335
+    },
+    {
+      "epoch": 0.4668,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00012863573330989313,
+      "loss": 1.2633,
+      "step": 23340
+    },
+    {
+      "epoch": 0.4669,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0001286022867677462,
+      "loss": 1.2874,
+      "step": 23345
+    },
+    {
+      "epoch": 0.467,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00012856883674049736,
+      "loss": 1.2783,
+      "step": 23350
+    },
+    {
+      "epoch": 0.4671,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00012853538323222242,
+      "loss": 1.2658,
+      "step": 23355
+    },
+    {
+      "epoch": 0.4672,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001285019262469976,
+      "loss": 1.2911,
+      "step": 23360
+    },
+    {
+      "epoch": 0.4673,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00012846846578889955,
+      "loss": 1.2903,
+      "step": 23365
+    },
+    {
+      "epoch": 0.4674,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00012843500186200527,
+      "loss": 1.2771,
+      "step": 23370
+    },
+    {
+      "epoch": 0.4675,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012840153447039228,
+      "loss": 1.282,
+      "step": 23375
+    },
+    {
+      "epoch": 0.4676,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012836806361813844,
+      "loss": 1.2752,
+      "step": 23380
+    },
+    {
+      "epoch": 0.4677,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00012833458930932212,
+      "loss": 1.2804,
+      "step": 23385
+    },
+    {
+      "epoch": 0.4678,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00012830111154802203,
+      "loss": 1.2736,
+      "step": 23390
+    },
+    {
+      "epoch": 0.4679,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00012826763033831735,
+      "loss": 1.2727,
+      "step": 23395
+    },
+    {
+      "epoch": 0.468,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00012823414568428768,
+      "loss": 1.2894,
+      "step": 23400
+    },
+    {
+      "epoch": 0.4681,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00012820065759001293,
+      "loss": 1.2648,
+      "step": 23405
+    },
+    {
+      "epoch": 0.4682,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00012816716605957367,
+      "loss": 1.2697,
+      "step": 23410
+    },
+    {
+      "epoch": 0.4683,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012813367109705063,
+      "loss": 1.2866,
+      "step": 23415
+    },
+    {
+      "epoch": 0.4684,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00012810017270652513,
+      "loss": 1.2599,
+      "step": 23420
+    },
+    {
+      "epoch": 0.4685,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0001280666708920788,
+      "loss": 1.2657,
+      "step": 23425
+    },
+    {
+      "epoch": 0.4686,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00012803316565779377,
+      "loss": 1.2821,
+      "step": 23430
+    },
+    {
+      "epoch": 0.4687,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001279996570077525,
+      "loss": 1.262,
+      "step": 23435
+    },
+    {
+      "epoch": 0.4688,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.000127966144946038,
+      "loss": 1.2738,
+      "step": 23440
+    },
+    {
+      "epoch": 0.4689,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00012793262947673355,
+      "loss": 1.2712,
+      "step": 23445
+    },
+    {
+      "epoch": 0.469,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00012789911060392294,
+      "loss": 1.2792,
+      "step": 23450
+    },
+    {
+      "epoch": 0.4691,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00012786558833169031,
+      "loss": 1.2748,
+      "step": 23455
+    },
+    {
+      "epoch": 0.4692,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001278320626641203,
+      "loss": 1.2594,
+      "step": 23460
+    },
+    {
+      "epoch": 0.4693,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00012779853360529785,
+      "loss": 1.2753,
+      "step": 23465
+    },
+    {
+      "epoch": 0.4694,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00012776500115930842,
+      "loss": 1.2897,
+      "step": 23470
+    },
+    {
+      "epoch": 0.4695,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00012773146533023782,
+      "loss": 1.2899,
+      "step": 23475
+    },
+    {
+      "epoch": 0.4696,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00012769792612217224,
+      "loss": 1.2822,
+      "step": 23480
+    },
+    {
+      "epoch": 0.4697,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001276643835391984,
+      "loss": 1.2831,
+      "step": 23485
+    },
+    {
+      "epoch": 0.4698,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00012763083758540337,
+      "loss": 1.278,
+      "step": 23490
+    },
+    {
+      "epoch": 0.4699,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0001275972882648746,
+      "loss": 1.2758,
+      "step": 23495
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001275637355816999,
+      "loss": 1.2806,
+      "step": 23500
+    },
+    {
+      "epoch": 0.4701,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0001275301795399677,
+      "loss": 1.282,
+      "step": 23505
+    },
+    {
+      "epoch": 0.4702,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00012749662014376663,
+      "loss": 1.2574,
+      "step": 23510
+    },
+    {
+      "epoch": 0.4703,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00012746305739718577,
+      "loss": 1.2752,
+      "step": 23515
+    },
+    {
+      "epoch": 0.4704,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00012742949130431467,
+      "loss": 1.2793,
+      "step": 23520
+    },
+    {
+      "epoch": 0.4705,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00012739592186924328,
+      "loss": 1.2906,
+      "step": 23525
+    },
+    {
+      "epoch": 0.4706,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00012736234909606185,
+      "loss": 1.279,
+      "step": 23530
+    },
+    {
+      "epoch": 0.4707,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00012732877298886124,
+      "loss": 1.2904,
+      "step": 23535
+    },
+    {
+      "epoch": 0.4708,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00012729519355173254,
+      "loss": 1.2717,
+      "step": 23540
+    },
+    {
+      "epoch": 0.4709,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001272616107887673,
+      "loss": 1.2689,
+      "step": 23545
+    },
+    {
+      "epoch": 0.471,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00012722802470405744,
+      "loss": 1.2638,
+      "step": 23550
+    },
+    {
+      "epoch": 0.4711,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001271944353016954,
+      "loss": 1.2643,
+      "step": 23555
+    },
+    {
+      "epoch": 0.4712,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00012716084258577388,
+      "loss": 1.2669,
+      "step": 23560
+    },
+    {
+      "epoch": 0.4713,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0001271272465603861,
+      "loss": 1.2609,
+      "step": 23565
+    },
+    {
+      "epoch": 0.4714,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0001270936472296256,
+      "loss": 1.2789,
+      "step": 23570
+    },
+    {
+      "epoch": 0.4715,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00012706004459758636,
+      "loss": 1.3007,
+      "step": 23575
+    },
+    {
+      "epoch": 0.4716,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00012702643866836278,
+      "loss": 1.282,
+      "step": 23580
+    },
+    {
+      "epoch": 0.4717,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00012699282944604967,
+      "loss": 1.2685,
+      "step": 23585
+    },
+    {
+      "epoch": 0.4718,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00012695921693474212,
+      "loss": 1.2526,
+      "step": 23590
+    },
+    {
+      "epoch": 0.4719,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00012692560113853574,
+      "loss": 1.2733,
+      "step": 23595
+    },
+    {
+      "epoch": 0.472,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00012689198206152657,
+      "loss": 1.2782,
+      "step": 23600
+    },
+    {
+      "epoch": 0.4721,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00012685835970781097,
+      "loss": 1.2726,
+      "step": 23605
+    },
+    {
+      "epoch": 0.4722,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001268247340814857,
+      "loss": 1.274,
+      "step": 23610
+    },
+    {
+      "epoch": 0.4723,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00012679110518664795,
+      "loss": 1.2662,
+      "step": 23615
+    },
+    {
+      "epoch": 0.4724,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00012675747302739527,
+      "loss": 1.2823,
+      "step": 23620
+    },
+    {
+      "epoch": 0.4725,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00012672383760782568,
+      "loss": 1.256,
+      "step": 23625
+    },
+    {
+      "epoch": 0.4726,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00012669019893203759,
+      "loss": 1.2795,
+      "step": 23630
+    },
+    {
+      "epoch": 0.4727,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00012665655700412967,
+      "loss": 1.2819,
+      "step": 23635
+    },
+    {
+      "epoch": 0.4728,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00012662291182820114,
+      "loss": 1.287,
+      "step": 23640
+    },
+    {
+      "epoch": 0.4729,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00012658926340835156,
+      "loss": 1.2925,
+      "step": 23645
+    },
+    {
+      "epoch": 0.473,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00012655561174868088,
+      "loss": 1.2579,
+      "step": 23650
+    },
+    {
+      "epoch": 0.4731,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00012652195685328946,
+      "loss": 1.2744,
+      "step": 23655
+    },
+    {
+      "epoch": 0.4732,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00012648829872627807,
+      "loss": 1.2729,
+      "step": 23660
+    },
+    {
+      "epoch": 0.4733,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00012645463737174782,
+      "loss": 1.2766,
+      "step": 23665
+    },
+    {
+      "epoch": 0.4734,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00012642097279380027,
+      "loss": 1.2797,
+      "step": 23670
+    },
+    {
+      "epoch": 0.4735,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.0001263873049965373,
+      "loss": 1.2755,
+      "step": 23675
+    },
+    {
+      "epoch": 0.4736,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00012635363398406127,
+      "loss": 1.2678,
+      "step": 23680
+    },
+    {
+      "epoch": 0.4737,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00012631995976047488,
+      "loss": 1.2709,
+      "step": 23685
+    },
+    {
+      "epoch": 0.4738,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0001262862823298812,
+      "loss": 1.2785,
+      "step": 23690
+    },
+    {
+      "epoch": 0.4739,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00012625260169638378,
+      "loss": 1.2709,
+      "step": 23695
+    },
+    {
+      "epoch": 0.474,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00012621891786408648,
+      "loss": 1.2678,
+      "step": 23700
+    },
+    {
+      "epoch": 0.4741,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00012618523083709357,
+      "loss": 1.2915,
+      "step": 23705
+    },
+    {
+      "epoch": 0.4742,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00012615154061950968,
+      "loss": 1.2683,
+      "step": 23710
+    },
+    {
+      "epoch": 0.4743,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00012611784721543995,
+      "loss": 1.2758,
+      "step": 23715
+    },
+    {
+      "epoch": 0.4744,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00012608415062898972,
+      "loss": 1.2738,
+      "step": 23720
+    },
+    {
+      "epoch": 0.4745,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00012605045086426487,
+      "loss": 1.2732,
+      "step": 23725
+    },
+    {
+      "epoch": 0.4746,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00012601674792537157,
+      "loss": 1.2905,
+      "step": 23730
+    },
+    {
+      "epoch": 0.4747,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00012598304181641646,
+      "loss": 1.2822,
+      "step": 23735
+    },
+    {
+      "epoch": 0.4748,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00012594933254150655,
+      "loss": 1.2726,
+      "step": 23740
+    },
+    {
+      "epoch": 0.4749,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00012591562010474914,
+      "loss": 1.254,
+      "step": 23745
+    },
+    {
+      "epoch": 0.475,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00012588190451025207,
+      "loss": 1.2851,
+      "step": 23750
+    },
+    {
+      "epoch": 0.4751,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00012584818576212345,
+      "loss": 1.2681,
+      "step": 23755
+    },
+    {
+      "epoch": 0.4752,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001258144638644718,
+      "loss": 1.2716,
+      "step": 23760
+    },
+    {
+      "epoch": 0.4753,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000125780738821406,
+      "loss": 1.2745,
+      "step": 23765
+    },
+    {
+      "epoch": 0.4754,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001257470106370354,
+      "loss": 1.2913,
+      "step": 23770
+    },
+    {
+      "epoch": 0.4755,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00012571327931546963,
+      "loss": 1.2742,
+      "step": 23775
+    },
+    {
+      "epoch": 0.4756,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00012567954486081878,
+      "loss": 1.2833,
+      "step": 23780
+    },
+    {
+      "epoch": 0.4757,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0001256458072771933,
+      "loss": 1.2759,
+      "step": 23785
+    },
+    {
+      "epoch": 0.4758,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012561206656870396,
+      "loss": 1.294,
+      "step": 23790
+    },
+    {
+      "epoch": 0.4759,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00012557832273946205,
+      "loss": 1.2733,
+      "step": 23795
+    },
+    {
+      "epoch": 0.476,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00012554457579357905,
+      "loss": 1.2848,
+      "step": 23800
+    },
+    {
+      "epoch": 0.4761,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00012551082573516705,
+      "loss": 1.2883,
+      "step": 23805
+    },
+    {
+      "epoch": 0.4762,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00012547707256833823,
+      "loss": 1.2823,
+      "step": 23810
+    },
+    {
+      "epoch": 0.4763,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00012544331629720543,
+      "loss": 1.2726,
+      "step": 23815
+    },
+    {
+      "epoch": 0.4764,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00012540955692588173,
+      "loss": 1.2827,
+      "step": 23820
+    },
+    {
+      "epoch": 0.4765,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00012537579445848058,
+      "loss": 1.2745,
+      "step": 23825
+    },
+    {
+      "epoch": 0.4766,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00012534202889911584,
+      "loss": 1.2639,
+      "step": 23830
+    },
+    {
+      "epoch": 0.4767,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00012530826025190174,
+      "loss": 1.2795,
+      "step": 23835
+    },
+    {
+      "epoch": 0.4768,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00012527448852095295,
+      "loss": 1.2698,
+      "step": 23840
+    },
+    {
+      "epoch": 0.4769,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00012524071371038434,
+      "loss": 1.2721,
+      "step": 23845
+    },
+    {
+      "epoch": 0.477,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.0001252069358243114,
+      "loss": 1.2646,
+      "step": 23850
+    },
+    {
+      "epoch": 0.4771,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00012517315486684972,
+      "loss": 1.2829,
+      "step": 23855
+    },
+    {
+      "epoch": 0.4772,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0001251393708421155,
+      "loss": 1.2824,
+      "step": 23860
+    },
+    {
+      "epoch": 0.4773,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00012510558375422522,
+      "loss": 1.2638,
+      "step": 23865
+    },
+    {
+      "epoch": 0.4774,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.0001250717936072957,
+      "loss": 1.2756,
+      "step": 23870
+    },
+    {
+      "epoch": 0.4775,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00012503800040544416,
+      "loss": 1.278,
+      "step": 23875
+    },
+    {
+      "epoch": 0.4776,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00012500420415278822,
+      "loss": 1.2736,
+      "step": 23880
+    },
+    {
+      "epoch": 0.4777,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00012497040485344584,
+      "loss": 1.2722,
+      "step": 23885
+    },
+    {
+      "epoch": 0.4778,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001249366025115354,
+      "loss": 1.2706,
+      "step": 23890
+    },
+    {
+      "epoch": 0.4779,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0001249027971311756,
+      "loss": 1.2657,
+      "step": 23895
+    },
+    {
+      "epoch": 0.478,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001248689887164855,
+      "loss": 1.2826,
+      "step": 23900
+    },
+    {
+      "epoch": 0.4781,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00012483517727158454,
+      "loss": 1.2691,
+      "step": 23905
+    },
+    {
+      "epoch": 0.4782,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00012480136280059256,
+      "loss": 1.2576,
+      "step": 23910
+    },
+    {
+      "epoch": 0.4783,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00012476754530762977,
+      "loss": 1.2808,
+      "step": 23915
+    },
+    {
+      "epoch": 0.4784,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00012473372479681672,
+      "loss": 1.2968,
+      "step": 23920
+    },
+    {
+      "epoch": 0.4785,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00012469990127227432,
+      "loss": 1.2664,
+      "step": 23925
+    },
+    {
+      "epoch": 0.4786,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00012466607473812387,
+      "loss": 1.2718,
+      "step": 23930
+    },
+    {
+      "epoch": 0.4787,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.000124632245198487,
+      "loss": 1.2701,
+      "step": 23935
+    },
+    {
+      "epoch": 0.4788,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0001245984126574858,
+      "loss": 1.2767,
+      "step": 23940
+    },
+    {
+      "epoch": 0.4789,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00012456457711924266,
+      "loss": 1.2695,
+      "step": 23945
+    },
+    {
+      "epoch": 0.479,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00012453073858788026,
+      "loss": 1.2855,
+      "step": 23950
+    },
+    {
+      "epoch": 0.4791,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00012449689706752178,
+      "loss": 1.2832,
+      "step": 23955
+    },
+    {
+      "epoch": 0.4792,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00012446305256229073,
+      "loss": 1.2625,
+      "step": 23960
+    },
+    {
+      "epoch": 0.4793,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00012442920507631092,
+      "loss": 1.2675,
+      "step": 23965
+    },
+    {
+      "epoch": 0.4794,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00012439535461370658,
+      "loss": 1.3015,
+      "step": 23970
+    },
+    {
+      "epoch": 0.4795,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012436150117860225,
+      "loss": 1.2556,
+      "step": 23975
+    },
+    {
+      "epoch": 0.4796,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00012432764477512292,
+      "loss": 1.271,
+      "step": 23980
+    },
+    {
+      "epoch": 0.4797,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00012429378540739386,
+      "loss": 1.2614,
+      "step": 23985
+    },
+    {
+      "epoch": 0.4798,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012425992307954075,
+      "loss": 1.2809,
+      "step": 23990
+    },
+    {
+      "epoch": 0.4799,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001242260577956896,
+      "loss": 1.2771,
+      "step": 23995
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00012419218955996676,
+      "loss": 1.2835,
+      "step": 24000
+    },
+    {
+      "epoch": 0.4801,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00012415831837649905,
+      "loss": 1.2831,
+      "step": 24005
+    },
+    {
+      "epoch": 0.4802,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0001241244442494135,
+      "loss": 1.2985,
+      "step": 24010
+    },
+    {
+      "epoch": 0.4803,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0001240905671828376,
+      "loss": 1.2712,
+      "step": 24015
+    },
+    {
+      "epoch": 0.4804,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00012405668718089917,
+      "loss": 1.2634,
+      "step": 24020
+    },
+    {
+      "epoch": 0.4805,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00012402280424772639,
+      "loss": 1.2622,
+      "step": 24025
+    },
+    {
+      "epoch": 0.4806,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00012398891838744778,
+      "loss": 1.2904,
+      "step": 24030
+    },
+    {
+      "epoch": 0.4807,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0001239550296041922,
+      "loss": 1.285,
+      "step": 24035
+    },
+    {
+      "epoch": 0.4808,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00012392113790208895,
+      "loss": 1.2632,
+      "step": 24040
+    },
+    {
+      "epoch": 0.4809,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.0001238872432852676,
+      "loss": 1.2638,
+      "step": 24045
+    },
+    {
+      "epoch": 0.481,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001238533457578581,
+      "loss": 1.2721,
+      "step": 24050
+    },
+    {
+      "epoch": 0.4811,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001238194453239908,
+      "loss": 1.2738,
+      "step": 24055
+    },
+    {
+      "epoch": 0.4812,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001237855419877963,
+      "loss": 1.2818,
+      "step": 24060
+    },
+    {
+      "epoch": 0.4813,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0001237516357534057,
+      "loss": 1.2816,
+      "step": 24065
+    },
+    {
+      "epoch": 0.4814,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0001237177266249503,
+      "loss": 1.2788,
+      "step": 24070
+    },
+    {
+      "epoch": 0.4815,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001236838146065619,
+      "loss": 1.293,
+      "step": 24075
+    },
+    {
+      "epoch": 0.4816,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00012364989970237248,
+      "loss": 1.2822,
+      "step": 24080
+    },
+    {
+      "epoch": 0.4817,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00012361598191651453,
+      "loss": 1.2744,
+      "step": 24085
+    },
+    {
+      "epoch": 0.4818,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00012358206125312083,
+      "loss": 1.2717,
+      "step": 24090
+    },
+    {
+      "epoch": 0.4819,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00012354813771632447,
+      "loss": 1.2697,
+      "step": 24095
+    },
+    {
+      "epoch": 0.482,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.000123514211310259,
+      "loss": 1.2781,
+      "step": 24100
+    },
+    {
+      "epoch": 0.4821,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00012348028203905818,
+      "loss": 1.2736,
+      "step": 24105
+    },
+    {
+      "epoch": 0.4822,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00012344634990685624,
+      "loss": 1.2691,
+      "step": 24110
+    },
+    {
+      "epoch": 0.4823,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0001234124149177877,
+      "loss": 1.2695,
+      "step": 24115
+    },
+    {
+      "epoch": 0.4824,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00012337847707598738,
+      "loss": 1.2724,
+      "step": 24120
+    },
+    {
+      "epoch": 0.4825,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00012334453638559057,
+      "loss": 1.2868,
+      "step": 24125
+    },
+    {
+      "epoch": 0.4826,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001233105928507328,
+      "loss": 1.2821,
+      "step": 24130
+    },
+    {
+      "epoch": 0.4827,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00012327664647554998,
+      "loss": 1.2621,
+      "step": 24135
+    },
+    {
+      "epoch": 0.4828,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001232426972641784,
+      "loss": 1.2845,
+      "step": 24140
+    },
+    {
+      "epoch": 0.4829,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00012320874522075468,
+      "loss": 1.2679,
+      "step": 24145
+    },
+    {
+      "epoch": 0.483,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00012317479034941573,
+      "loss": 1.2742,
+      "step": 24150
+    },
+    {
+      "epoch": 0.4831,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001231408326542989,
+      "loss": 1.2699,
+      "step": 24155
+    },
+    {
+      "epoch": 0.4832,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00012310687213954182,
+      "loss": 1.2778,
+      "step": 24160
+    },
+    {
+      "epoch": 0.4833,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001230729088092824,
+      "loss": 1.2593,
+      "step": 24165
+    },
+    {
+      "epoch": 0.4834,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00012303894266765908,
+      "loss": 1.2643,
+      "step": 24170
+    },
+    {
+      "epoch": 0.4835,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00012300497371881046,
+      "loss": 1.2748,
+      "step": 24175
+    },
+    {
+      "epoch": 0.4836,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00012297100196687557,
+      "loss": 1.2614,
+      "step": 24180
+    },
+    {
+      "epoch": 0.4837,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00012293702741599378,
+      "loss": 1.2685,
+      "step": 24185
+    },
+    {
+      "epoch": 0.4838,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00012290305007030478,
+      "loss": 1.2682,
+      "step": 24190
+    },
+    {
+      "epoch": 0.4839,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00012286906993394856,
+      "loss": 1.2651,
+      "step": 24195
+    },
+    {
+      "epoch": 0.484,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00012283508701106557,
+      "loss": 1.2625,
+      "step": 24200
+    },
+    {
+      "epoch": 0.4841,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001228011013057965,
+      "loss": 1.2487,
+      "step": 24205
+    },
+    {
+      "epoch": 0.4842,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001227671128222824,
+      "loss": 1.2909,
+      "step": 24210
+    },
+    {
+      "epoch": 0.4843,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00012273312156466464,
+      "loss": 1.2826,
+      "step": 24215
+    },
+    {
+      "epoch": 0.4844,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00012269912753708502,
+      "loss": 1.2538,
+      "step": 24220
+    },
+    {
+      "epoch": 0.4845,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001226651307436855,
+      "loss": 1.2718,
+      "step": 24225
+    },
+    {
+      "epoch": 0.4846,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001226311311886086,
+      "loss": 1.2633,
+      "step": 24230
+    },
+    {
+      "epoch": 0.4847,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00012259712887599698,
+      "loss": 1.2708,
+      "step": 24235
+    },
+    {
+      "epoch": 0.4848,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00012256312380999376,
+      "loss": 1.2777,
+      "step": 24240
+    },
+    {
+      "epoch": 0.4849,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00012252911599474237,
+      "loss": 1.2673,
+      "step": 24245
+    },
+    {
+      "epoch": 0.485,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001224951054343865,
+      "loss": 1.2596,
+      "step": 24250
+    },
+    {
+      "epoch": 0.4851,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001224610921330703,
+      "loss": 1.2906,
+      "step": 24255
+    },
+    {
+      "epoch": 0.4852,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00012242707609493814,
+      "loss": 1.282,
+      "step": 24260
+    },
+    {
+      "epoch": 0.4853,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00012239305732413477,
+      "loss": 1.281,
+      "step": 24265
+    },
+    {
+      "epoch": 0.4854,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001223590358248053,
+      "loss": 1.2748,
+      "step": 24270
+    },
+    {
+      "epoch": 0.4855,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00012232501160109514,
+      "loss": 1.2654,
+      "step": 24275
+    },
+    {
+      "epoch": 0.4856,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00012229098465715006,
+      "loss": 1.2749,
+      "step": 24280
+    },
+    {
+      "epoch": 0.4857,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00012225695499711606,
+      "loss": 1.2864,
+      "step": 24285
+    },
+    {
+      "epoch": 0.4858,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00012222292262513965,
+      "loss": 1.2649,
+      "step": 24290
+    },
+    {
+      "epoch": 0.4859,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001221888875453675,
+      "loss": 1.2685,
+      "step": 24295
+    },
+    {
+      "epoch": 0.486,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00012215484976194676,
+      "loss": 1.285,
+      "step": 24300
+    },
+    {
+      "epoch": 0.4861,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00012212080927902474,
+      "loss": 1.2607,
+      "step": 24305
+    },
+    {
+      "epoch": 0.4862,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00012208676610074917,
+      "loss": 1.2726,
+      "step": 24310
+    },
+    {
+      "epoch": 0.4863,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012205272023126821,
+      "loss": 1.2821,
+      "step": 24315
+    },
+    {
+      "epoch": 0.4864,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00012201867167473015,
+      "loss": 1.2895,
+      "step": 24320
+    },
+    {
+      "epoch": 0.4865,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00012198462043528376,
+      "loss": 1.2866,
+      "step": 24325
+    },
+    {
+      "epoch": 0.4866,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012195056651707806,
+      "loss": 1.2725,
+      "step": 24330
+    },
+    {
+      "epoch": 0.4867,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00012191650992426238,
+      "loss": 1.2922,
+      "step": 24335
+    },
+    {
+      "epoch": 0.4868,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00012188245066098647,
+      "loss": 1.281,
+      "step": 24340
+    },
+    {
+      "epoch": 0.4869,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00012184838873140032,
+      "loss": 1.2622,
+      "step": 24345
+    },
+    {
+      "epoch": 0.487,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00012181432413965428,
+      "loss": 1.2779,
+      "step": 24350
+    },
+    {
+      "epoch": 0.4871,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.000121780256889899,
+      "loss": 1.2627,
+      "step": 24355
+    },
+    {
+      "epoch": 0.4872,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00012174618698628549,
+      "loss": 1.2855,
+      "step": 24360
+    },
+    {
+      "epoch": 0.4873,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00012171211443296505,
+      "loss": 1.2746,
+      "step": 24365
+    },
+    {
+      "epoch": 0.4874,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012167803923408934,
+      "loss": 1.2769,
+      "step": 24370
+    },
+    {
+      "epoch": 0.4875,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00012164396139381029,
+      "loss": 1.2869,
+      "step": 24375
+    },
+    {
+      "epoch": 0.4876,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00012160988091628022,
+      "loss": 1.2646,
+      "step": 24380
+    },
+    {
+      "epoch": 0.4877,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00012157579780565169,
+      "loss": 1.2936,
+      "step": 24385
+    },
+    {
+      "epoch": 0.4878,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00012154171206607764,
+      "loss": 1.2714,
+      "step": 24390
+    },
+    {
+      "epoch": 0.4879,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00012150762370171136,
+      "loss": 1.2978,
+      "step": 24395
+    },
+    {
+      "epoch": 0.488,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00012147353271670634,
+      "loss": 1.2857,
+      "step": 24400
+    },
+    {
+      "epoch": 0.4881,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00012143943911521647,
+      "loss": 1.2909,
+      "step": 24405
+    },
+    {
+      "epoch": 0.4882,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.000121405342901396,
+      "loss": 1.2787,
+      "step": 24410
+    },
+    {
+      "epoch": 0.4883,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00012137124407939943,
+      "loss": 1.2948,
+      "step": 24415
+    },
+    {
+      "epoch": 0.4884,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00012133714265338161,
+      "loss": 1.2628,
+      "step": 24420
+    },
+    {
+      "epoch": 0.4885,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00012130303862749767,
+      "loss": 1.2624,
+      "step": 24425
+    },
+    {
+      "epoch": 0.4886,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00012126893200590308,
+      "loss": 1.2621,
+      "step": 24430
+    },
+    {
+      "epoch": 0.4887,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00012123482279275365,
+      "loss": 1.2729,
+      "step": 24435
+    },
+    {
+      "epoch": 0.4888,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00012120071099220549,
+      "loss": 1.272,
+      "step": 24440
+    },
+    {
+      "epoch": 0.4889,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00012116659660841499,
+      "loss": 1.2517,
+      "step": 24445
+    },
+    {
+      "epoch": 0.489,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00012113247964553888,
+      "loss": 1.2723,
+      "step": 24450
+    },
+    {
+      "epoch": 0.4891,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00012109836010773424,
+      "loss": 1.2764,
+      "step": 24455
+    },
+    {
+      "epoch": 0.4892,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001210642379991584,
+      "loss": 1.265,
+      "step": 24460
+    },
+    {
+      "epoch": 0.4893,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00012103011332396908,
+      "loss": 1.267,
+      "step": 24465
+    },
+    {
+      "epoch": 0.4894,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00012099598608632428,
+      "loss": 1.2673,
+      "step": 24470
+    },
+    {
+      "epoch": 0.4895,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001209618562903822,
+      "loss": 1.2664,
+      "step": 24475
+    },
+    {
+      "epoch": 0.4896,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00012092772394030152,
+      "loss": 1.2622,
+      "step": 24480
+    },
+    {
+      "epoch": 0.4897,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00012089358904024117,
+      "loss": 1.2575,
+      "step": 24485
+    },
+    {
+      "epoch": 0.4898,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00012085945159436038,
+      "loss": 1.2692,
+      "step": 24490
+    },
+    {
+      "epoch": 0.4899,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00012082531160681869,
+      "loss": 1.2827,
+      "step": 24495
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00012079116908177593,
+      "loss": 1.2603,
+      "step": 24500
+    },
+    {
+      "epoch": 0.4901,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00012075702402339231,
+      "loss": 1.2639,
+      "step": 24505
+    },
+    {
+      "epoch": 0.4902,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00012072287643582825,
+      "loss": 1.2789,
+      "step": 24510
+    },
+    {
+      "epoch": 0.4903,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001206887263232446,
+      "loss": 1.2771,
+      "step": 24515
+    },
+    {
+      "epoch": 0.4904,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00012065457368980236,
+      "loss": 1.2917,
+      "step": 24520
+    },
+    {
+      "epoch": 0.4905,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012062041853966298,
+      "loss": 1.2826,
+      "step": 24525
+    },
+    {
+      "epoch": 0.4906,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00012058626087698814,
+      "loss": 1.2899,
+      "step": 24530
+    },
+    {
+      "epoch": 0.4907,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00012055210070593988,
+      "loss": 1.2752,
+      "step": 24535
+    },
+    {
+      "epoch": 0.4908,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00012051793803068046,
+      "loss": 1.27,
+      "step": 24540
+    },
+    {
+      "epoch": 0.4909,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00012048377285537256,
+      "loss": 1.2764,
+      "step": 24545
+    },
+    {
+      "epoch": 0.491,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00012044960518417903,
+      "loss": 1.2613,
+      "step": 24550
+    },
+    {
+      "epoch": 0.4911,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00012041543502126318,
+      "loss": 1.2655,
+      "step": 24555
+    },
+    {
+      "epoch": 0.4912,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001203812623707885,
+      "loss": 1.2585,
+      "step": 24560
+    },
+    {
+      "epoch": 0.4913,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00012034708723691881,
+      "loss": 1.2629,
+      "step": 24565
+    },
+    {
+      "epoch": 0.4914,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00012031290962381823,
+      "loss": 1.2709,
+      "step": 24570
+    },
+    {
+      "epoch": 0.4915,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00012027872953565125,
+      "loss": 1.2678,
+      "step": 24575
+    },
+    {
+      "epoch": 0.4916,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00012024454697658261,
+      "loss": 1.2746,
+      "step": 24580
+    },
+    {
+      "epoch": 0.4917,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00012021036195077731,
+      "loss": 1.2901,
+      "step": 24585
+    },
+    {
+      "epoch": 0.4918,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001201761744624007,
+      "loss": 1.2743,
+      "step": 24590
+    },
+    {
+      "epoch": 0.4919,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00012014198451561842,
+      "loss": 1.2691,
+      "step": 24595
+    },
+    {
+      "epoch": 0.492,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00012010779211459648,
+      "loss": 1.2974,
+      "step": 24600
+    },
+    {
+      "epoch": 0.4921,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00012007359726350105,
+      "loss": 1.2896,
+      "step": 24605
+    },
+    {
+      "epoch": 0.4922,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00012003939996649865,
+      "loss": 1.2781,
+      "step": 24610
+    },
+    {
+      "epoch": 0.4923,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00012000520022775617,
+      "loss": 1.2623,
+      "step": 24615
+    },
+    {
+      "epoch": 0.4924,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00011997099805144069,
+      "loss": 1.265,
+      "step": 24620
+    },
+    {
+      "epoch": 0.4925,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00011993679344171973,
+      "loss": 1.2836,
+      "step": 24625
+    },
+    {
+      "epoch": 0.4926,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00011990258640276094,
+      "loss": 1.292,
+      "step": 24630
+    },
+    {
+      "epoch": 0.4927,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00011986837693873237,
+      "loss": 1.2678,
+      "step": 24635
+    },
+    {
+      "epoch": 0.4928,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00011983416505380234,
+      "loss": 1.2817,
+      "step": 24640
+    },
+    {
+      "epoch": 0.4929,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00011979995075213946,
+      "loss": 1.2771,
+      "step": 24645
+    },
+    {
+      "epoch": 0.493,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00011976573403791262,
+      "loss": 1.2687,
+      "step": 24650
+    },
+    {
+      "epoch": 0.4931,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00011973151491529107,
+      "loss": 1.2717,
+      "step": 24655
+    },
+    {
+      "epoch": 0.4932,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00011969729338844429,
+      "loss": 1.2676,
+      "step": 24660
+    },
+    {
+      "epoch": 0.4933,
+      "grad_norm": 0.5,
+      "learning_rate": 0.000119663069461542,
+      "loss": 1.2724,
+      "step": 24665
+    },
+    {
+      "epoch": 0.4934,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001196288431387544,
+      "loss": 1.2703,
+      "step": 24670
+    },
+    {
+      "epoch": 0.4935,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00011959461442425177,
+      "loss": 1.2916,
+      "step": 24675
+    },
+    {
+      "epoch": 0.4936,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00011956038332220483,
+      "loss": 1.2628,
+      "step": 24680
+    },
+    {
+      "epoch": 0.4937,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00011952614983678452,
+      "loss": 1.2581,
+      "step": 24685
+    },
+    {
+      "epoch": 0.4938,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00011949191397216206,
+      "loss": 1.2956,
+      "step": 24690
+    },
+    {
+      "epoch": 0.4939,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00011945767573250903,
+      "loss": 1.2842,
+      "step": 24695
+    },
+    {
+      "epoch": 0.494,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0001194234351219972,
+      "loss": 1.2743,
+      "step": 24700
+    },
+    {
+      "epoch": 0.4941,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00011938919214479876,
+      "loss": 1.2664,
+      "step": 24705
+    },
+    {
+      "epoch": 0.4942,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00011935494680508606,
+      "loss": 1.2586,
+      "step": 24710
+    },
+    {
+      "epoch": 0.4943,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00011932069910703176,
+      "loss": 1.2738,
+      "step": 24715
+    },
+    {
+      "epoch": 0.4944,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001192864490548089,
+      "loss": 1.2682,
+      "step": 24720
+    },
+    {
+      "epoch": 0.4945,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00011925219665259075,
+      "loss": 1.2705,
+      "step": 24725
+    },
+    {
+      "epoch": 0.4946,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00011921794190455082,
+      "loss": 1.2694,
+      "step": 24730
+    },
+    {
+      "epoch": 0.4947,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011918368481486297,
+      "loss": 1.2778,
+      "step": 24735
+    },
+    {
+      "epoch": 0.4948,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00011914942538770131,
+      "loss": 1.2843,
+      "step": 24740
+    },
+    {
+      "epoch": 0.4949,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00011911516362724024,
+      "loss": 1.2686,
+      "step": 24745
+    },
+    {
+      "epoch": 0.495,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00011908089953765449,
+      "loss": 1.2947,
+      "step": 24750
+    },
+    {
+      "epoch": 0.4951,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011904663312311901,
+      "loss": 1.2784,
+      "step": 24755
+    },
+    {
+      "epoch": 0.4952,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00011901236438780902,
+      "loss": 1.2684,
+      "step": 24760
+    },
+    {
+      "epoch": 0.4953,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00011897809333590014,
+      "loss": 1.2702,
+      "step": 24765
+    },
+    {
+      "epoch": 0.4954,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00011894381997156813,
+      "loss": 1.2761,
+      "step": 24770
+    },
+    {
+      "epoch": 0.4955,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00011890954429898912,
+      "loss": 1.2794,
+      "step": 24775
+    },
+    {
+      "epoch": 0.4956,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00011887526632233954,
+      "loss": 1.2586,
+      "step": 24780
+    },
+    {
+      "epoch": 0.4957,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00011884098604579597,
+      "loss": 1.2722,
+      "step": 24785
+    },
+    {
+      "epoch": 0.4958,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00011880670347353539,
+      "loss": 1.2658,
+      "step": 24790
+    },
+    {
+      "epoch": 0.4959,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00011877241860973507,
+      "loss": 1.2724,
+      "step": 24795
+    },
+    {
+      "epoch": 0.496,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00011873813145857249,
+      "loss": 1.2703,
+      "step": 24800
+    },
+    {
+      "epoch": 0.4961,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001187038420242254,
+      "loss": 1.2806,
+      "step": 24805
+    },
+    {
+      "epoch": 0.4962,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.0001186695503108719,
+      "loss": 1.2662,
+      "step": 24810
+    },
+    {
+      "epoch": 0.4963,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011863525632269032,
+      "loss": 1.2722,
+      "step": 24815
+    },
+    {
+      "epoch": 0.4964,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001186009600638593,
+      "loss": 1.2643,
+      "step": 24820
+    },
+    {
+      "epoch": 0.4965,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00011856666153855776,
+      "loss": 1.2703,
+      "step": 24825
+    },
+    {
+      "epoch": 0.4966,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00011853236075096474,
+      "loss": 1.2915,
+      "step": 24830
+    },
+    {
+      "epoch": 0.4967,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00011849805770525983,
+      "loss": 1.2771,
+      "step": 24835
+    },
+    {
+      "epoch": 0.4968,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001184637524056227,
+      "loss": 1.2649,
+      "step": 24840
+    },
+    {
+      "epoch": 0.4969,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00011842944485623335,
+      "loss": 1.2713,
+      "step": 24845
+    },
+    {
+      "epoch": 0.497,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00011839513506127203,
+      "loss": 1.2656,
+      "step": 24850
+    },
+    {
+      "epoch": 0.4971,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001183608230249193,
+      "loss": 1.2612,
+      "step": 24855
+    },
+    {
+      "epoch": 0.4972,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00011832650875135598,
+      "loss": 1.2662,
+      "step": 24860
+    },
+    {
+      "epoch": 0.4973,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00011829219224476318,
+      "loss": 1.2838,
+      "step": 24865
+    },
+    {
+      "epoch": 0.4974,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00011825787350932222,
+      "loss": 1.2722,
+      "step": 24870
+    },
+    {
+      "epoch": 0.4975,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00011822355254921478,
+      "loss": 1.2834,
+      "step": 24875
+    },
+    {
+      "epoch": 0.4976,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00011818922936862269,
+      "loss": 1.2939,
+      "step": 24880
+    },
+    {
+      "epoch": 0.4977,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00011815490397172821,
+      "loss": 1.2738,
+      "step": 24885
+    },
+    {
+      "epoch": 0.4978,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00011812057636271374,
+      "loss": 1.2821,
+      "step": 24890
+    },
+    {
+      "epoch": 0.4979,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00011808624654576202,
+      "loss": 1.2649,
+      "step": 24895
+    },
+    {
+      "epoch": 0.498,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00011805191452505602,
+      "loss": 1.262,
+      "step": 24900
+    },
+    {
+      "epoch": 0.4981,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00011801758030477897,
+      "loss": 1.2829,
+      "step": 24905
+    },
+    {
+      "epoch": 0.4982,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00011798324388911444,
+      "loss": 1.2803,
+      "step": 24910
+    },
+    {
+      "epoch": 0.4983,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00011794890528224618,
+      "loss": 1.2757,
+      "step": 24915
+    },
+    {
+      "epoch": 0.4984,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00011791456448835825,
+      "loss": 1.2732,
+      "step": 24920
+    },
+    {
+      "epoch": 0.4985,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011788022151163495,
+      "loss": 1.2717,
+      "step": 24925
+    },
+    {
+      "epoch": 0.4986,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00011784587635626094,
+      "loss": 1.2653,
+      "step": 24930
+    },
+    {
+      "epoch": 0.4987,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.000117811529026421,
+      "loss": 1.2732,
+      "step": 24935
+    },
+    {
+      "epoch": 0.4988,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00011777717952630031,
+      "loss": 1.2862,
+      "step": 24940
+    },
+    {
+      "epoch": 0.4989,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00011774282786008422,
+      "loss": 1.2608,
+      "step": 24945
+    },
+    {
+      "epoch": 0.499,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00011770847403195834,
+      "loss": 1.259,
+      "step": 24950
+    },
+    {
+      "epoch": 0.4991,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00011767411804610864,
+      "loss": 1.273,
+      "step": 24955
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00011763975990672125,
+      "loss": 1.2771,
+      "step": 24960
+    },
+    {
+      "epoch": 0.4993,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00011760539961798262,
+      "loss": 1.2847,
+      "step": 24965
+    },
+    {
+      "epoch": 0.4994,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00011757103718407947,
+      "loss": 1.2726,
+      "step": 24970
+    },
+    {
+      "epoch": 0.4995,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00011753667260919872,
+      "loss": 1.2606,
+      "step": 24975
+    },
+    {
+      "epoch": 0.4996,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00011750230589752762,
+      "loss": 1.2672,
+      "step": 24980
+    },
+    {
+      "epoch": 0.4997,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00011746793705325363,
+      "loss": 1.259,
+      "step": 24985
+    },
+    {
+      "epoch": 0.4998,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00011743356608056449,
+      "loss": 1.2742,
+      "step": 24990
+    },
+    {
+      "epoch": 0.4999,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001173991929836482,
+      "loss": 1.2713,
+      "step": 24995
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00011736481776669306,
+      "loss": 1.2669,
+      "step": 25000
+    },
+    {
+      "epoch": 0.5001,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00011733044043388752,
+      "loss": 1.2688,
+      "step": 25005
+    },
+    {
+      "epoch": 0.5002,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00011729606098942039,
+      "loss": 1.2767,
+      "step": 25010
+    },
+    {
+      "epoch": 0.5003,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00011726167943748067,
+      "loss": 1.2707,
+      "step": 25015
+    },
+    {
+      "epoch": 0.5004,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00011722729578225769,
+      "loss": 1.2796,
+      "step": 25020
+    },
+    {
+      "epoch": 0.5005,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00011719291002794096,
+      "loss": 1.2913,
+      "step": 25025
+    },
+    {
+      "epoch": 0.5006,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001171585221787203,
+      "loss": 1.2629,
+      "step": 25030
+    },
+    {
+      "epoch": 0.5007,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011712413223878578,
+      "loss": 1.2978,
+      "step": 25035
+    },
+    {
+      "epoch": 0.5008,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011708974021232769,
+      "loss": 1.2896,
+      "step": 25040
+    },
+    {
+      "epoch": 0.5009,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00011705534610353657,
+      "loss": 1.2771,
+      "step": 25045
+    },
+    {
+      "epoch": 0.501,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00011702094991660326,
+      "loss": 1.2898,
+      "step": 25050
+    },
+    {
+      "epoch": 0.5011,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00011698655165571886,
+      "loss": 1.2775,
+      "step": 25055
+    },
+    {
+      "epoch": 0.5012,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00011695215132507464,
+      "loss": 1.2634,
+      "step": 25060
+    },
+    {
+      "epoch": 0.5013,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00011691774892886222,
+      "loss": 1.2736,
+      "step": 25065
+    },
+    {
+      "epoch": 0.5014,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00011688334447127338,
+      "loss": 1.2766,
+      "step": 25070
+    },
+    {
+      "epoch": 0.5015,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00011684893795650027,
+      "loss": 1.2693,
+      "step": 25075
+    },
+    {
+      "epoch": 0.5016,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00011681452938873516,
+      "loss": 1.2673,
+      "step": 25080
+    },
+    {
+      "epoch": 0.5017,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00011678011877217065,
+      "loss": 1.2792,
+      "step": 25085
+    },
+    {
+      "epoch": 0.5018,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00011674570611099955,
+      "loss": 1.2709,
+      "step": 25090
+    },
+    {
+      "epoch": 0.5019,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011671129140941499,
+      "loss": 1.272,
+      "step": 25095
+    },
+    {
+      "epoch": 0.502,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00011667687467161024,
+      "loss": 1.2742,
+      "step": 25100
+    },
+    {
+      "epoch": 0.5021,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00011664245590177892,
+      "loss": 1.276,
+      "step": 25105
+    },
+    {
+      "epoch": 0.5022,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0001166080351041148,
+      "loss": 1.2615,
+      "step": 25110
+    },
+    {
+      "epoch": 0.5023,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00011657361228281199,
+      "loss": 1.2639,
+      "step": 25115
+    },
+    {
+      "epoch": 0.5024,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00011653918744206478,
+      "loss": 1.2752,
+      "step": 25120
+    },
+    {
+      "epoch": 0.5025,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00011650476058606777,
+      "loss": 1.2749,
+      "step": 25125
+    },
+    {
+      "epoch": 0.5026,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00011647033171901573,
+      "loss": 1.2698,
+      "step": 25130
+    },
+    {
+      "epoch": 0.5027,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00011643590084510379,
+      "loss": 1.2784,
+      "step": 25135
+    },
+    {
+      "epoch": 0.5028,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00011640146796852711,
+      "loss": 1.2771,
+      "step": 25140
+    },
+    {
+      "epoch": 0.5029,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00011636703309348133,
+      "loss": 1.2743,
+      "step": 25145
+    },
+    {
+      "epoch": 0.503,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00011633259622416224,
+      "loss": 1.2627,
+      "step": 25150
+    },
+    {
+      "epoch": 0.5031,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00011629815736476581,
+      "loss": 1.2758,
+      "step": 25155
+    },
+    {
+      "epoch": 0.5032,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00011626371651948838,
+      "loss": 1.2736,
+      "step": 25160
+    },
+    {
+      "epoch": 0.5033,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00011622927369252638,
+      "loss": 1.2914,
+      "step": 25165
+    },
+    {
+      "epoch": 0.5034,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00011619482888807662,
+      "loss": 1.2736,
+      "step": 25170
+    },
+    {
+      "epoch": 0.5035,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00011616038211033613,
+      "loss": 1.266,
+      "step": 25175
+    },
+    {
+      "epoch": 0.5036,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00011612593336350208,
+      "loss": 1.2642,
+      "step": 25180
+    },
+    {
+      "epoch": 0.5037,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00011609148265177193,
+      "loss": 1.2661,
+      "step": 25185
+    },
+    {
+      "epoch": 0.5038,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00011605702997934345,
+      "loss": 1.2829,
+      "step": 25190
+    },
+    {
+      "epoch": 0.5039,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00011602257535041459,
+      "loss": 1.2674,
+      "step": 25195
+    },
+    {
+      "epoch": 0.504,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001159881187691835,
+      "loss": 1.2852,
+      "step": 25200
+    },
+    {
+      "epoch": 0.5041,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00011595366023984864,
+      "loss": 1.2599,
+      "step": 25205
+    },
+    {
+      "epoch": 0.5042,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00011591919976660868,
+      "loss": 1.2683,
+      "step": 25210
+    },
+    {
+      "epoch": 0.5043,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00011588473735366249,
+      "loss": 1.2602,
+      "step": 25215
+    },
+    {
+      "epoch": 0.5044,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001158502730052093,
+      "loss": 1.2659,
+      "step": 25220
+    },
+    {
+      "epoch": 0.5045,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00011581580672544838,
+      "loss": 1.2726,
+      "step": 25225
+    },
+    {
+      "epoch": 0.5046,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001157813385185794,
+      "loss": 1.2829,
+      "step": 25230
+    },
+    {
+      "epoch": 0.5047,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00011574686838880215,
+      "loss": 1.259,
+      "step": 25235
+    },
+    {
+      "epoch": 0.5048,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00011571239634031679,
+      "loss": 1.2503,
+      "step": 25240
+    },
+    {
+      "epoch": 0.5049,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00011567792237732358,
+      "loss": 1.2816,
+      "step": 25245
+    },
+    {
+      "epoch": 0.505,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0001156434465040231,
+      "loss": 1.2666,
+      "step": 25250
+    },
+    {
+      "epoch": 0.5051,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0001156089687246161,
+      "loss": 1.2566,
+      "step": 25255
+    },
+    {
+      "epoch": 0.5052,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011557448904330362,
+      "loss": 1.2725,
+      "step": 25260
+    },
+    {
+      "epoch": 0.5053,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0001155400074642869,
+      "loss": 1.2767,
+      "step": 25265
+    },
+    {
+      "epoch": 0.5054,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00011550552399176739,
+      "loss": 1.2726,
+      "step": 25270
+    },
+    {
+      "epoch": 0.5055,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011547103862994684,
+      "loss": 1.267,
+      "step": 25275
+    },
+    {
+      "epoch": 0.5056,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00011543655138302714,
+      "loss": 1.2876,
+      "step": 25280
+    },
+    {
+      "epoch": 0.5057,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00011540206225521046,
+      "loss": 1.2767,
+      "step": 25285
+    },
+    {
+      "epoch": 0.5058,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00011536757125069923,
+      "loss": 1.2717,
+      "step": 25290
+    },
+    {
+      "epoch": 0.5059,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00011533307837369607,
+      "loss": 1.2754,
+      "step": 25295
+    },
+    {
+      "epoch": 0.506,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00011529858362840382,
+      "loss": 1.2672,
+      "step": 25300
+    },
+    {
+      "epoch": 0.5061,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011526408701902556,
+      "loss": 1.286,
+      "step": 25305
+    },
+    {
+      "epoch": 0.5062,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00011522958854976458,
+      "loss": 1.291,
+      "step": 25310
+    },
+    {
+      "epoch": 0.5063,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00011519508822482446,
+      "loss": 1.2718,
+      "step": 25315
+    },
+    {
+      "epoch": 0.5064,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00011516058604840891,
+      "loss": 1.3027,
+      "step": 25320
+    },
+    {
+      "epoch": 0.5065,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00011512608202472194,
+      "loss": 1.276,
+      "step": 25325
+    },
+    {
+      "epoch": 0.5066,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011509157615796776,
+      "loss": 1.2728,
+      "step": 25330
+    },
+    {
+      "epoch": 0.5067,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011505706845235078,
+      "loss": 1.2849,
+      "step": 25335
+    },
+    {
+      "epoch": 0.5068,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00011502255891207572,
+      "loss": 1.2695,
+      "step": 25340
+    },
+    {
+      "epoch": 0.5069,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001149880475413474,
+      "loss": 1.2737,
+      "step": 25345
+    },
+    {
+      "epoch": 0.507,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00011495353434437098,
+      "loss": 1.2736,
+      "step": 25350
+    },
+    {
+      "epoch": 0.5071,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00011491901932535172,
+      "loss": 1.2827,
+      "step": 25355
+    },
+    {
+      "epoch": 0.5072,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00011488450248849522,
+      "loss": 1.2691,
+      "step": 25360
+    },
+    {
+      "epoch": 0.5073,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00011484998383800726,
+      "loss": 1.2919,
+      "step": 25365
+    },
+    {
+      "epoch": 0.5074,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00011481546337809381,
+      "loss": 1.2762,
+      "step": 25370
+    },
+    {
+      "epoch": 0.5075,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00011478094111296109,
+      "loss": 1.2783,
+      "step": 25375
+    },
+    {
+      "epoch": 0.5076,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001147464170468155,
+      "loss": 1.2706,
+      "step": 25380
+    },
+    {
+      "epoch": 0.5077,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00011471189118386375,
+      "loss": 1.2672,
+      "step": 25385
+    },
+    {
+      "epoch": 0.5078,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00011467736352831266,
+      "loss": 1.2851,
+      "step": 25390
+    },
+    {
+      "epoch": 0.5079,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00011464283408436938,
+      "loss": 1.2841,
+      "step": 25395
+    },
+    {
+      "epoch": 0.508,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00011460830285624118,
+      "loss": 1.2753,
+      "step": 25400
+    },
+    {
+      "epoch": 0.5081,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00011457376984813557,
+      "loss": 1.2698,
+      "step": 25405
+    },
+    {
+      "epoch": 0.5082,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00011453923506426032,
+      "loss": 1.2871,
+      "step": 25410
+    },
+    {
+      "epoch": 0.5083,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00011450469850882337,
+      "loss": 1.2846,
+      "step": 25415
+    },
+    {
+      "epoch": 0.5084,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00011447016018603292,
+      "loss": 1.2811,
+      "step": 25420
+    },
+    {
+      "epoch": 0.5085,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00011443562010009731,
+      "loss": 1.282,
+      "step": 25425
+    },
+    {
+      "epoch": 0.5086,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00011440107825522521,
+      "loss": 1.2831,
+      "step": 25430
+    },
+    {
+      "epoch": 0.5087,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00011436653465562542,
+      "loss": 1.2643,
+      "step": 25435
+    },
+    {
+      "epoch": 0.5088,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00011433198930550695,
+      "loss": 1.2749,
+      "step": 25440
+    },
+    {
+      "epoch": 0.5089,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00011429744220907903,
+      "loss": 1.274,
+      "step": 25445
+    },
+    {
+      "epoch": 0.509,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00011426289337055119,
+      "loss": 1.2656,
+      "step": 25450
+    },
+    {
+      "epoch": 0.5091,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00011422834279413301,
+      "loss": 1.2795,
+      "step": 25455
+    },
+    {
+      "epoch": 0.5092,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00011419379048403444,
+      "loss": 1.2772,
+      "step": 25460
+    },
+    {
+      "epoch": 0.5093,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00011415923644446557,
+      "loss": 1.2749,
+      "step": 25465
+    },
+    {
+      "epoch": 0.5094,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00011412468067963669,
+      "loss": 1.2821,
+      "step": 25470
+    },
+    {
+      "epoch": 0.5095,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00011409012319375827,
+      "loss": 1.2732,
+      "step": 25475
+    },
+    {
+      "epoch": 0.5096,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00011405556399104109,
+      "loss": 1.2647,
+      "step": 25480
+    },
+    {
+      "epoch": 0.5097,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00011402100307569612,
+      "loss": 1.2817,
+      "step": 25485
+    },
+    {
+      "epoch": 0.5098,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00011398644045193444,
+      "loss": 1.2725,
+      "step": 25490
+    },
+    {
+      "epoch": 0.5099,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00011395187612396738,
+      "loss": 1.2788,
+      "step": 25495
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00011391731009600654,
+      "loss": 1.2686,
+      "step": 25500
+    },
+    {
+      "epoch": 0.5101,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00011388274237226371,
+      "loss": 1.2753,
+      "step": 25505
+    },
+    {
+      "epoch": 0.5102,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00011384817295695083,
+      "loss": 1.2628,
+      "step": 25510
+    },
+    {
+      "epoch": 0.5103,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00011381360185428007,
+      "loss": 1.2684,
+      "step": 25515
+    },
+    {
+      "epoch": 0.5104,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0001137790290684638,
+      "loss": 1.2788,
+      "step": 25520
+    },
+    {
+      "epoch": 0.5105,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00011374445460371466,
+      "loss": 1.2794,
+      "step": 25525
+    },
+    {
+      "epoch": 0.5106,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00011370987846424546,
+      "loss": 1.274,
+      "step": 25530
+    },
+    {
+      "epoch": 0.5107,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001136753006542691,
+      "loss": 1.2626,
+      "step": 25535
+    },
+    {
+      "epoch": 0.5108,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00011364072117799885,
+      "loss": 1.2779,
+      "step": 25540
+    },
+    {
+      "epoch": 0.5109,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00011360614003964809,
+      "loss": 1.266,
+      "step": 25545
+    },
+    {
+      "epoch": 0.511,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00011357155724343045,
+      "loss": 1.2824,
+      "step": 25550
+    },
+    {
+      "epoch": 0.5111,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00011353697279355973,
+      "loss": 1.2789,
+      "step": 25555
+    },
+    {
+      "epoch": 0.5112,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00011350238669424993,
+      "loss": 1.2704,
+      "step": 25560
+    },
+    {
+      "epoch": 0.5113,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00011346779894971527,
+      "loss": 1.2655,
+      "step": 25565
+    },
+    {
+      "epoch": 0.5114,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011343320956417014,
+      "loss": 1.2746,
+      "step": 25570
+    },
+    {
+      "epoch": 0.5115,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00011339861854182922,
+      "loss": 1.2626,
+      "step": 25575
+    },
+    {
+      "epoch": 0.5116,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00011336402588690726,
+      "loss": 1.2667,
+      "step": 25580
+    },
+    {
+      "epoch": 0.5117,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00011332943160361926,
+      "loss": 1.2749,
+      "step": 25585
+    },
+    {
+      "epoch": 0.5118,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00011329483569618045,
+      "loss": 1.2718,
+      "step": 25590
+    },
+    {
+      "epoch": 0.5119,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011326023816880625,
+      "loss": 1.2644,
+      "step": 25595
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00011322563902571226,
+      "loss": 1.2784,
+      "step": 25600
+    },
+    {
+      "epoch": 0.5121,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00011319103827111426,
+      "loss": 1.2703,
+      "step": 25605
+    },
+    {
+      "epoch": 0.5122,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00011315643590922827,
+      "loss": 1.2775,
+      "step": 25610
+    },
+    {
+      "epoch": 0.5123,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00011312183194427046,
+      "loss": 1.2656,
+      "step": 25615
+    },
+    {
+      "epoch": 0.5124,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00011308722638045724,
+      "loss": 1.2859,
+      "step": 25620
+    },
+    {
+      "epoch": 0.5125,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00011305261922200519,
+      "loss": 1.2973,
+      "step": 25625
+    },
+    {
+      "epoch": 0.5126,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00011301801047313105,
+      "loss": 1.2861,
+      "step": 25630
+    },
+    {
+      "epoch": 0.5127,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00011298340013805184,
+      "loss": 1.2612,
+      "step": 25635
+    },
+    {
+      "epoch": 0.5128,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011294878822098469,
+      "loss": 1.2725,
+      "step": 25640
+    },
+    {
+      "epoch": 0.5129,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.000112914174726147,
+      "loss": 1.2903,
+      "step": 25645
+    },
+    {
+      "epoch": 0.513,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001128795596577563,
+      "loss": 1.2782,
+      "step": 25650
+    },
+    {
+      "epoch": 0.5131,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0001128449430200303,
+      "loss": 1.2785,
+      "step": 25655
+    },
+    {
+      "epoch": 0.5132,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00011281032481718697,
+      "loss": 1.2673,
+      "step": 25660
+    },
+    {
+      "epoch": 0.5133,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001127757050534444,
+      "loss": 1.2689,
+      "step": 25665
+    },
+    {
+      "epoch": 0.5134,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00011274108373302095,
+      "loss": 1.2661,
+      "step": 25670
+    },
+    {
+      "epoch": 0.5135,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00011270646086013505,
+      "loss": 1.2695,
+      "step": 25675
+    },
+    {
+      "epoch": 0.5136,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00011267183643900548,
+      "loss": 1.2713,
+      "step": 25680
+    },
+    {
+      "epoch": 0.5137,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00011263721047385105,
+      "loss": 1.2808,
+      "step": 25685
+    },
+    {
+      "epoch": 0.5138,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00011260258296889086,
+      "loss": 1.2706,
+      "step": 25690
+    },
+    {
+      "epoch": 0.5139,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00011256795392834419,
+      "loss": 1.2869,
+      "step": 25695
+    },
+    {
+      "epoch": 0.514,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00011253332335643043,
+      "loss": 1.2883,
+      "step": 25700
+    },
+    {
+      "epoch": 0.5141,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00011249869125736925,
+      "loss": 1.2661,
+      "step": 25705
+    },
+    {
+      "epoch": 0.5142,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00011246405763538046,
+      "loss": 1.2669,
+      "step": 25710
+    },
+    {
+      "epoch": 0.5143,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00011242942249468402,
+      "loss": 1.2787,
+      "step": 25715
+    },
+    {
+      "epoch": 0.5144,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00011239478583950018,
+      "loss": 1.2832,
+      "step": 25720
+    },
+    {
+      "epoch": 0.5145,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00011236014767404927,
+      "loss": 1.2748,
+      "step": 25725
+    },
+    {
+      "epoch": 0.5146,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00011232550800255188,
+      "loss": 1.2997,
+      "step": 25730
+    },
+    {
+      "epoch": 0.5147,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00011229086682922869,
+      "loss": 1.2734,
+      "step": 25735
+    },
+    {
+      "epoch": 0.5148,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00011225622415830068,
+      "loss": 1.2666,
+      "step": 25740
+    },
+    {
+      "epoch": 0.5149,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00011222157999398895,
+      "loss": 1.284,
+      "step": 25745
+    },
+    {
+      "epoch": 0.515,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00011218693434051475,
+      "loss": 1.2606,
+      "step": 25750
+    },
+    {
+      "epoch": 0.5151,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00011215228720209958,
+      "loss": 1.2805,
+      "step": 25755
+    },
+    {
+      "epoch": 0.5152,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00011211763858296507,
+      "loss": 1.2866,
+      "step": 25760
+    },
+    {
+      "epoch": 0.5153,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00011208298848733305,
+      "loss": 1.2588,
+      "step": 25765
+    },
+    {
+      "epoch": 0.5154,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00011204833691942553,
+      "loss": 1.2823,
+      "step": 25770
+    },
+    {
+      "epoch": 0.5155,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00011201368388346471,
+      "loss": 1.2759,
+      "step": 25775
+    },
+    {
+      "epoch": 0.5156,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00011197902938367298,
+      "loss": 1.2568,
+      "step": 25780
+    },
+    {
+      "epoch": 0.5157,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001119443734242728,
+      "loss": 1.2764,
+      "step": 25785
+    },
+    {
+      "epoch": 0.5158,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00011190971600948699,
+      "loss": 1.2744,
+      "step": 25790
+    },
+    {
+      "epoch": 0.5159,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001118750571435384,
+      "loss": 1.2658,
+      "step": 25795
+    },
+    {
+      "epoch": 0.516,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00011184039683065013,
+      "loss": 1.2732,
+      "step": 25800
+    },
+    {
+      "epoch": 0.5161,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00011180573507504537,
+      "loss": 1.273,
+      "step": 25805
+    },
+    {
+      "epoch": 0.5162,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00011177107188094764,
+      "loss": 1.2891,
+      "step": 25810
+    },
+    {
+      "epoch": 0.5163,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00011173640725258052,
+      "loss": 1.2646,
+      "step": 25815
+    },
+    {
+      "epoch": 0.5164,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00011170174119416776,
+      "loss": 1.2595,
+      "step": 25820
+    },
+    {
+      "epoch": 0.5165,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00011166707370993333,
+      "loss": 1.2844,
+      "step": 25825
+    },
+    {
+      "epoch": 0.5166,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00011163240480410135,
+      "loss": 1.2852,
+      "step": 25830
+    },
+    {
+      "epoch": 0.5167,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00011159773448089614,
+      "loss": 1.2692,
+      "step": 25835
+    },
+    {
+      "epoch": 0.5168,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00011156306274454218,
+      "loss": 1.2687,
+      "step": 25840
+    },
+    {
+      "epoch": 0.5169,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00011152838959926408,
+      "loss": 1.2831,
+      "step": 25845
+    },
+    {
+      "epoch": 0.517,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011149371504928668,
+      "loss": 1.2742,
+      "step": 25850
+    },
+    {
+      "epoch": 0.5171,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00011145903909883495,
+      "loss": 1.2886,
+      "step": 25855
+    },
+    {
+      "epoch": 0.5172,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00011142436175213409,
+      "loss": 1.2809,
+      "step": 25860
+    },
+    {
+      "epoch": 0.5173,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001113896830134094,
+      "loss": 1.2814,
+      "step": 25865
+    },
+    {
+      "epoch": 0.5174,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00011135500288688636,
+      "loss": 1.2741,
+      "step": 25870
+    },
+    {
+      "epoch": 0.5175,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.0001113203213767907,
+      "loss": 1.2808,
+      "step": 25875
+    },
+    {
+      "epoch": 0.5176,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00011128563848734816,
+      "loss": 1.2632,
+      "step": 25880
+    },
+    {
+      "epoch": 0.5177,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00011125095422278486,
+      "loss": 1.2807,
+      "step": 25885
+    },
+    {
+      "epoch": 0.5178,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0001112162685873269,
+      "loss": 1.2875,
+      "step": 25890
+    },
+    {
+      "epoch": 0.5179,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00011118158158520064,
+      "loss": 1.2746,
+      "step": 25895
+    },
+    {
+      "epoch": 0.518,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00011114689322063255,
+      "loss": 1.2692,
+      "step": 25900
+    },
+    {
+      "epoch": 0.5181,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00011111220349784937,
+      "loss": 1.2887,
+      "step": 25905
+    },
+    {
+      "epoch": 0.5182,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00011107751242107787,
+      "loss": 1.2774,
+      "step": 25910
+    },
+    {
+      "epoch": 0.5183,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00011104281999454511,
+      "loss": 1.2807,
+      "step": 25915
+    },
+    {
+      "epoch": 0.5184,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00011100812622247822,
+      "loss": 1.2692,
+      "step": 25920
+    },
+    {
+      "epoch": 0.5185,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00011097343110910452,
+      "loss": 1.2793,
+      "step": 25925
+    },
+    {
+      "epoch": 0.5186,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00011093873465865157,
+      "loss": 1.2578,
+      "step": 25930
+    },
+    {
+      "epoch": 0.5187,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00011090403687534697,
+      "loss": 1.2842,
+      "step": 25935
+    },
+    {
+      "epoch": 0.5188,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00011086933776341852,
+      "loss": 1.2734,
+      "step": 25940
+    },
+    {
+      "epoch": 0.5189,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00011083463732709425,
+      "loss": 1.2705,
+      "step": 25945
+    },
+    {
+      "epoch": 0.519,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001107999355706023,
+      "loss": 1.272,
+      "step": 25950
+    },
+    {
+      "epoch": 0.5191,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00011076523249817094,
+      "loss": 1.2797,
+      "step": 25955
+    },
+    {
+      "epoch": 0.5192,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00011073052811402867,
+      "loss": 1.2703,
+      "step": 25960
+    },
+    {
+      "epoch": 0.5193,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001106958224224041,
+      "loss": 1.2791,
+      "step": 25965
+    },
+    {
+      "epoch": 0.5194,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.000110661115427526,
+      "loss": 1.2714,
+      "step": 25970
+    },
+    {
+      "epoch": 0.5195,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00011062640713362333,
+      "loss": 1.2695,
+      "step": 25975
+    },
+    {
+      "epoch": 0.5196,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001105916975449252,
+      "loss": 1.2719,
+      "step": 25980
+    },
+    {
+      "epoch": 0.5197,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00011055698666566084,
+      "loss": 1.2858,
+      "step": 25985
+    },
+    {
+      "epoch": 0.5198,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00011052227450005967,
+      "loss": 1.2693,
+      "step": 25990
+    },
+    {
+      "epoch": 0.5199,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00011048756105235125,
+      "loss": 1.272,
+      "step": 25995
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00011045284632676536,
+      "loss": 1.2756,
+      "step": 26000
+    },
+    {
+      "epoch": 0.5201,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00011041813032753183,
+      "loss": 1.2762,
+      "step": 26005
+    },
+    {
+      "epoch": 0.5202,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011038341305888074,
+      "loss": 1.2713,
+      "step": 26010
+    },
+    {
+      "epoch": 0.5203,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00011034869452504226,
+      "loss": 1.2684,
+      "step": 26015
+    },
+    {
+      "epoch": 0.5204,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00011031397473024674,
+      "loss": 1.2583,
+      "step": 26020
+    },
+    {
+      "epoch": 0.5205,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00011027925367872469,
+      "loss": 1.2631,
+      "step": 26025
+    },
+    {
+      "epoch": 0.5206,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00011024453137470677,
+      "loss": 1.2764,
+      "step": 26030
+    },
+    {
+      "epoch": 0.5207,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00011020980782242376,
+      "loss": 1.27,
+      "step": 26035
+    },
+    {
+      "epoch": 0.5208,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00011017508302610664,
+      "loss": 1.2794,
+      "step": 26040
+    },
+    {
+      "epoch": 0.5209,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00011014035698998651,
+      "loss": 1.2647,
+      "step": 26045
+    },
+    {
+      "epoch": 0.521,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00011010562971829463,
+      "loss": 1.2775,
+      "step": 26050
+    },
+    {
+      "epoch": 0.5211,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00011007090121526245,
+      "loss": 1.29,
+      "step": 26055
+    },
+    {
+      "epoch": 0.5212,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00011003617148512149,
+      "loss": 1.2875,
+      "step": 26060
+    },
+    {
+      "epoch": 0.5213,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00011000144053210348,
+      "loss": 1.274,
+      "step": 26065
+    },
+    {
+      "epoch": 0.5214,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001099667083604403,
+      "loss": 1.2799,
+      "step": 26070
+    },
+    {
+      "epoch": 0.5215,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00010993197497436391,
+      "loss": 1.28,
+      "step": 26075
+    },
+    {
+      "epoch": 0.5216,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00010989724037810652,
+      "loss": 1.2564,
+      "step": 26080
+    },
+    {
+      "epoch": 0.5217,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00010986250457590039,
+      "loss": 1.2881,
+      "step": 26085
+    },
+    {
+      "epoch": 0.5218,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00010982776757197799,
+      "loss": 1.2608,
+      "step": 26090
+    },
+    {
+      "epoch": 0.5219,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00010979302937057192,
+      "loss": 1.2783,
+      "step": 26095
+    },
+    {
+      "epoch": 0.522,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00010975828997591495,
+      "loss": 1.2704,
+      "step": 26100
+    },
+    {
+      "epoch": 0.5221,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00010972354939223996,
+      "loss": 1.2732,
+      "step": 26105
+    },
+    {
+      "epoch": 0.5222,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00010968880762377993,
+      "loss": 1.2745,
+      "step": 26110
+    },
+    {
+      "epoch": 0.5223,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00010965406467476808,
+      "loss": 1.2622,
+      "step": 26115
+    },
+    {
+      "epoch": 0.5224,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00010961932054943778,
+      "loss": 1.2565,
+      "step": 26120
+    },
+    {
+      "epoch": 0.5225,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00010958457525202241,
+      "loss": 1.2713,
+      "step": 26125
+    },
+    {
+      "epoch": 0.5226,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00010954982878675563,
+      "loss": 1.2734,
+      "step": 26130
+    },
+    {
+      "epoch": 0.5227,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00010951508115787119,
+      "loss": 1.2945,
+      "step": 26135
+    },
+    {
+      "epoch": 0.5228,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00010948033236960294,
+      "loss": 1.2564,
+      "step": 26140
+    },
+    {
+      "epoch": 0.5229,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00010944558242618496,
+      "loss": 1.2614,
+      "step": 26145
+    },
+    {
+      "epoch": 0.523,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00010941083133185146,
+      "loss": 1.2803,
+      "step": 26150
+    },
+    {
+      "epoch": 0.5231,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00010937607909083667,
+      "loss": 1.2593,
+      "step": 26155
+    },
+    {
+      "epoch": 0.5232,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00010934132570737507,
+      "loss": 1.2782,
+      "step": 26160
+    },
+    {
+      "epoch": 0.5233,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00010930657118570126,
+      "loss": 1.2817,
+      "step": 26165
+    },
+    {
+      "epoch": 0.5234,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00010927181553005002,
+      "loss": 1.2655,
+      "step": 26170
+    },
+    {
+      "epoch": 0.5235,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00010923705874465618,
+      "loss": 1.2765,
+      "step": 26175
+    },
+    {
+      "epoch": 0.5236,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00010920230083375473,
+      "loss": 1.2665,
+      "step": 26180
+    },
+    {
+      "epoch": 0.5237,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00010916754180158082,
+      "loss": 1.2702,
+      "step": 26185
+    },
+    {
+      "epoch": 0.5238,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00010913278165236978,
+      "loss": 1.2815,
+      "step": 26190
+    },
+    {
+      "epoch": 0.5239,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00010909802039035701,
+      "loss": 1.2616,
+      "step": 26195
+    },
+    {
+      "epoch": 0.524,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00010906325801977804,
+      "loss": 1.2583,
+      "step": 26200
+    },
+    {
+      "epoch": 0.5241,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010902849454486856,
+      "loss": 1.2801,
+      "step": 26205
+    },
+    {
+      "epoch": 0.5242,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00010899372996986439,
+      "loss": 1.2645,
+      "step": 26210
+    },
+    {
+      "epoch": 0.5243,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00010895896429900154,
+      "loss": 1.2629,
+      "step": 26215
+    },
+    {
+      "epoch": 0.5244,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00010892419753651606,
+      "loss": 1.2879,
+      "step": 26220
+    },
+    {
+      "epoch": 0.5245,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00010888942968664417,
+      "loss": 1.2556,
+      "step": 26225
+    },
+    {
+      "epoch": 0.5246,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00010885466075362223,
+      "loss": 1.27,
+      "step": 26230
+    },
+    {
+      "epoch": 0.5247,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00010881989074168673,
+      "loss": 1.2845,
+      "step": 26235
+    },
+    {
+      "epoch": 0.5248,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00010878511965507434,
+      "loss": 1.2859,
+      "step": 26240
+    },
+    {
+      "epoch": 0.5249,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00010875034749802173,
+      "loss": 1.2875,
+      "step": 26245
+    },
+    {
+      "epoch": 0.525,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00010871557427476583,
+      "loss": 1.2745,
+      "step": 26250
+    },
+    {
+      "epoch": 0.5251,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00010868079998954364,
+      "loss": 1.2693,
+      "step": 26255
+    },
+    {
+      "epoch": 0.5252,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001086460246465923,
+      "loss": 1.2842,
+      "step": 26260
+    },
+    {
+      "epoch": 0.5253,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00010861124825014908,
+      "loss": 1.2649,
+      "step": 26265
+    },
+    {
+      "epoch": 0.5254,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00010857647080445139,
+      "loss": 1.275,
+      "step": 26270
+    },
+    {
+      "epoch": 0.5255,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00010854169231373676,
+      "loss": 1.2674,
+      "step": 26275
+    },
+    {
+      "epoch": 0.5256,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00010850691278224281,
+      "loss": 1.2698,
+      "step": 26280
+    },
+    {
+      "epoch": 0.5257,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00010847213221420736,
+      "loss": 1.2721,
+      "step": 26285
+    },
+    {
+      "epoch": 0.5258,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00010843735061386828,
+      "loss": 1.2624,
+      "step": 26290
+    },
+    {
+      "epoch": 0.5259,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00010840256798546364,
+      "loss": 1.2725,
+      "step": 26295
+    },
+    {
+      "epoch": 0.526,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00010836778433323158,
+      "loss": 1.2792,
+      "step": 26300
+    },
+    {
+      "epoch": 0.5261,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00010833299966141035,
+      "loss": 1.2707,
+      "step": 26305
+    },
+    {
+      "epoch": 0.5262,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0001082982139742384,
+      "loss": 1.2723,
+      "step": 26310
+    },
+    {
+      "epoch": 0.5263,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00010826342727595426,
+      "loss": 1.2748,
+      "step": 26315
+    },
+    {
+      "epoch": 0.5264,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00010822863957079656,
+      "loss": 1.2739,
+      "step": 26320
+    },
+    {
+      "epoch": 0.5265,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0001081938508630041,
+      "loss": 1.2628,
+      "step": 26325
+    },
+    {
+      "epoch": 0.5266,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00010815906115681578,
+      "loss": 1.2645,
+      "step": 26330
+    },
+    {
+      "epoch": 0.5267,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010812427045647058,
+      "loss": 1.2705,
+      "step": 26335
+    },
+    {
+      "epoch": 0.5268,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00010808947876620767,
+      "loss": 1.2742,
+      "step": 26340
+    },
+    {
+      "epoch": 0.5269,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00010805468609026632,
+      "loss": 1.2809,
+      "step": 26345
+    },
+    {
+      "epoch": 0.527,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00010801989243288589,
+      "loss": 1.2647,
+      "step": 26350
+    },
+    {
+      "epoch": 0.5271,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0001079850977983059,
+      "loss": 1.2641,
+      "step": 26355
+    },
+    {
+      "epoch": 0.5272,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010795030219076599,
+      "loss": 1.2738,
+      "step": 26360
+    },
+    {
+      "epoch": 0.5273,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00010791550561450586,
+      "loss": 1.2925,
+      "step": 26365
+    },
+    {
+      "epoch": 0.5274,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00010788070807376536,
+      "loss": 1.2695,
+      "step": 26370
+    },
+    {
+      "epoch": 0.5275,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001078459095727845,
+      "loss": 1.2844,
+      "step": 26375
+    },
+    {
+      "epoch": 0.5276,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00010781111011580336,
+      "loss": 1.2836,
+      "step": 26380
+    },
+    {
+      "epoch": 0.5277,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00010777630970706217,
+      "loss": 1.2595,
+      "step": 26385
+    },
+    {
+      "epoch": 0.5278,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00010774150835080119,
+      "loss": 1.2814,
+      "step": 26390
+    },
+    {
+      "epoch": 0.5279,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00010770670605126092,
+      "loss": 1.2758,
+      "step": 26395
+    },
+    {
+      "epoch": 0.528,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010767190281268187,
+      "loss": 1.2683,
+      "step": 26400
+    },
+    {
+      "epoch": 0.5281,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00010763709863930476,
+      "loss": 1.2876,
+      "step": 26405
+    },
+    {
+      "epoch": 0.5282,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00010760229353537033,
+      "loss": 1.2789,
+      "step": 26410
+    },
+    {
+      "epoch": 0.5283,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00010756748750511953,
+      "loss": 1.2769,
+      "step": 26415
+    },
+    {
+      "epoch": 0.5284,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00010753268055279329,
+      "loss": 1.2842,
+      "step": 26420
+    },
+    {
+      "epoch": 0.5285,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00010749787268263279,
+      "loss": 1.2608,
+      "step": 26425
+    },
+    {
+      "epoch": 0.5286,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00010746306389887924,
+      "loss": 1.288,
+      "step": 26430
+    },
+    {
+      "epoch": 0.5287,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.000107428254205774,
+      "loss": 1.2879,
+      "step": 26435
+    },
+    {
+      "epoch": 0.5288,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00010739344360755852,
+      "loss": 1.2691,
+      "step": 26440
+    },
+    {
+      "epoch": 0.5289,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00010735863210847433,
+      "loss": 1.2909,
+      "step": 26445
+    },
+    {
+      "epoch": 0.529,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00010732381971276318,
+      "loss": 1.2741,
+      "step": 26450
+    },
+    {
+      "epoch": 0.5291,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001072890064246668,
+      "loss": 1.2598,
+      "step": 26455
+    },
+    {
+      "epoch": 0.5292,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0001072541922484271,
+      "loss": 1.2721,
+      "step": 26460
+    },
+    {
+      "epoch": 0.5293,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001072193771882861,
+      "loss": 1.2533,
+      "step": 26465
+    },
+    {
+      "epoch": 0.5294,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00010718456124848583,
+      "loss": 1.2756,
+      "step": 26470
+    },
+    {
+      "epoch": 0.5295,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001071497444332686,
+      "loss": 1.2657,
+      "step": 26475
+    },
+    {
+      "epoch": 0.5296,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00010711492674687671,
+      "loss": 1.2725,
+      "step": 26480
+    },
+    {
+      "epoch": 0.5297,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00010708010819355256,
+      "loss": 1.2796,
+      "step": 26485
+    },
+    {
+      "epoch": 0.5298,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001070452887775387,
+      "loss": 1.2541,
+      "step": 26490
+    },
+    {
+      "epoch": 0.5299,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00010701046850307777,
+      "loss": 1.2694,
+      "step": 26495
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00010697564737441252,
+      "loss": 1.2556,
+      "step": 26500
+    },
+    {
+      "epoch": 0.5301,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00010694082539578585,
+      "loss": 1.2783,
+      "step": 26505
+    },
+    {
+      "epoch": 0.5302,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010690600257144061,
+      "loss": 1.2703,
+      "step": 26510
+    },
+    {
+      "epoch": 0.5303,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00010687117890561988,
+      "loss": 1.2715,
+      "step": 26515
+    },
+    {
+      "epoch": 0.5304,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00010683635440256687,
+      "loss": 1.2902,
+      "step": 26520
+    },
+    {
+      "epoch": 0.5305,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00010680152906652483,
+      "loss": 1.2597,
+      "step": 26525
+    },
+    {
+      "epoch": 0.5306,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00010676670290173709,
+      "loss": 1.2801,
+      "step": 26530
+    },
+    {
+      "epoch": 0.5307,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00010673187591244714,
+      "loss": 1.2746,
+      "step": 26535
+    },
+    {
+      "epoch": 0.5308,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00010669704810289851,
+      "loss": 1.2809,
+      "step": 26540
+    },
+    {
+      "epoch": 0.5309,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00010666221947733486,
+      "loss": 1.2711,
+      "step": 26545
+    },
+    {
+      "epoch": 0.531,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00010662739004000005,
+      "loss": 1.2702,
+      "step": 26550
+    },
+    {
+      "epoch": 0.5311,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001065925597951378,
+      "loss": 1.2776,
+      "step": 26555
+    },
+    {
+      "epoch": 0.5312,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00010655772874699217,
+      "loss": 1.2839,
+      "step": 26560
+    },
+    {
+      "epoch": 0.5313,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00010652289689980714,
+      "loss": 1.2785,
+      "step": 26565
+    },
+    {
+      "epoch": 0.5314,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00010648806425782695,
+      "loss": 1.2779,
+      "step": 26570
+    },
+    {
+      "epoch": 0.5315,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010645323082529581,
+      "loss": 1.2665,
+      "step": 26575
+    },
+    {
+      "epoch": 0.5316,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00010641839660645805,
+      "loss": 1.2547,
+      "step": 26580
+    },
+    {
+      "epoch": 0.5317,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010638356160555816,
+      "loss": 1.291,
+      "step": 26585
+    },
+    {
+      "epoch": 0.5318,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00010634872582684061,
+      "loss": 1.2939,
+      "step": 26590
+    },
+    {
+      "epoch": 0.5319,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00010631388927455013,
+      "loss": 1.2711,
+      "step": 26595
+    },
+    {
+      "epoch": 0.532,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00010627905195293135,
+      "loss": 1.2767,
+      "step": 26600
+    },
+    {
+      "epoch": 0.5321,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00010624421386622916,
+      "loss": 1.2721,
+      "step": 26605
+    },
+    {
+      "epoch": 0.5322,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00010620937501868841,
+      "loss": 1.2759,
+      "step": 26610
+    },
+    {
+      "epoch": 0.5323,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0001061745354145542,
+      "loss": 1.2932,
+      "step": 26615
+    },
+    {
+      "epoch": 0.5324,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00010613969505807156,
+      "loss": 1.2916,
+      "step": 26620
+    },
+    {
+      "epoch": 0.5325,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00010610485395348571,
+      "loss": 1.2707,
+      "step": 26625
+    },
+    {
+      "epoch": 0.5326,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00010607001210504191,
+      "loss": 1.2762,
+      "step": 26630
+    },
+    {
+      "epoch": 0.5327,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00010603516951698556,
+      "loss": 1.2804,
+      "step": 26635
+    },
+    {
+      "epoch": 0.5328,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00010600032619356209,
+      "loss": 1.2755,
+      "step": 26640
+    },
+    {
+      "epoch": 0.5329,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00010596548213901708,
+      "loss": 1.2915,
+      "step": 26645
+    },
+    {
+      "epoch": 0.533,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00010593063735759618,
+      "loss": 1.2717,
+      "step": 26650
+    },
+    {
+      "epoch": 0.5331,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001058957918535451,
+      "loss": 1.2713,
+      "step": 26655
+    },
+    {
+      "epoch": 0.5332,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00010586094563110964,
+      "loss": 1.263,
+      "step": 26660
+    },
+    {
+      "epoch": 0.5333,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00010582609869453577,
+      "loss": 1.2688,
+      "step": 26665
+    },
+    {
+      "epoch": 0.5334,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00010579125104806944,
+      "loss": 1.2716,
+      "step": 26670
+    },
+    {
+      "epoch": 0.5335,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00010575640269595675,
+      "loss": 1.2798,
+      "step": 26675
+    },
+    {
+      "epoch": 0.5336,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00010572155364244382,
+      "loss": 1.2594,
+      "step": 26680
+    },
+    {
+      "epoch": 0.5337,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00010568670389177696,
+      "loss": 1.2633,
+      "step": 26685
+    },
+    {
+      "epoch": 0.5338,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00010565185344820247,
+      "loss": 1.2686,
+      "step": 26690
+    },
+    {
+      "epoch": 0.5339,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00010561700231596678,
+      "loss": 1.2789,
+      "step": 26695
+    },
+    {
+      "epoch": 0.534,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00010558215049931638,
+      "loss": 1.2732,
+      "step": 26700
+    },
+    {
+      "epoch": 0.5341,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00010554729800249792,
+      "loss": 1.2851,
+      "step": 26705
+    },
+    {
+      "epoch": 0.5342,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00010551244482975798,
+      "loss": 1.2801,
+      "step": 26710
+    },
+    {
+      "epoch": 0.5343,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00010547759098534335,
+      "loss": 1.2853,
+      "step": 26715
+    },
+    {
+      "epoch": 0.5344,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00010544273647350092,
+      "loss": 1.2778,
+      "step": 26720
+    },
+    {
+      "epoch": 0.5345,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00010540788129847756,
+      "loss": 1.2726,
+      "step": 26725
+    },
+    {
+      "epoch": 0.5346,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00010537302546452022,
+      "loss": 1.2536,
+      "step": 26730
+    },
+    {
+      "epoch": 0.5347,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00010533816897587606,
+      "loss": 1.2706,
+      "step": 26735
+    },
+    {
+      "epoch": 0.5348,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00010530331183679218,
+      "loss": 1.2795,
+      "step": 26740
+    },
+    {
+      "epoch": 0.5349,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00010526845405151586,
+      "loss": 1.2817,
+      "step": 26745
+    },
+    {
+      "epoch": 0.535,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001052335956242944,
+      "loss": 1.296,
+      "step": 26750
+    },
+    {
+      "epoch": 0.5351,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00010519873655937516,
+      "loss": 1.2829,
+      "step": 26755
+    },
+    {
+      "epoch": 0.5352,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00010516387686100566,
+      "loss": 1.2796,
+      "step": 26760
+    },
+    {
+      "epoch": 0.5353,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00010512901653343344,
+      "loss": 1.2726,
+      "step": 26765
+    },
+    {
+      "epoch": 0.5354,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00010509415558090609,
+      "loss": 1.2788,
+      "step": 26770
+    },
+    {
+      "epoch": 0.5355,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00010505929400767134,
+      "loss": 1.2584,
+      "step": 26775
+    },
+    {
+      "epoch": 0.5356,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00010502443181797697,
+      "loss": 1.2978,
+      "step": 26780
+    },
+    {
+      "epoch": 0.5357,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00010498956901607083,
+      "loss": 1.2685,
+      "step": 26785
+    },
+    {
+      "epoch": 0.5358,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00010495470560620083,
+      "loss": 1.2724,
+      "step": 26790
+    },
+    {
+      "epoch": 0.5359,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00010491984159261496,
+      "loss": 1.2847,
+      "step": 26795
+    },
+    {
+      "epoch": 0.536,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00010488497697956135,
+      "loss": 1.2719,
+      "step": 26800
+    },
+    {
+      "epoch": 0.5361,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00010485011177128807,
+      "loss": 1.2662,
+      "step": 26805
+    },
+    {
+      "epoch": 0.5362,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00010481524597204342,
+      "loss": 1.275,
+      "step": 26810
+    },
+    {
+      "epoch": 0.5363,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00010478037958607568,
+      "loss": 1.2695,
+      "step": 26815
+    },
+    {
+      "epoch": 0.5364,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00010474551261763314,
+      "loss": 1.2928,
+      "step": 26820
+    },
+    {
+      "epoch": 0.5365,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010471064507096426,
+      "loss": 1.2857,
+      "step": 26825
+    },
+    {
+      "epoch": 0.5366,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010467577695031762,
+      "loss": 1.2781,
+      "step": 26830
+    },
+    {
+      "epoch": 0.5367,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00010464090825994173,
+      "loss": 1.2678,
+      "step": 26835
+    },
+    {
+      "epoch": 0.5368,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00010460603900408523,
+      "loss": 1.2744,
+      "step": 26840
+    },
+    {
+      "epoch": 0.5369,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00010457116918699688,
+      "loss": 1.2609,
+      "step": 26845
+    },
+    {
+      "epoch": 0.537,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00010453629881292538,
+      "loss": 1.2678,
+      "step": 26850
+    },
+    {
+      "epoch": 0.5371,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00010450142788611965,
+      "loss": 1.2759,
+      "step": 26855
+    },
+    {
+      "epoch": 0.5372,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00010446655641082862,
+      "loss": 1.2634,
+      "step": 26860
+    },
+    {
+      "epoch": 0.5373,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00010443168439130122,
+      "loss": 1.2635,
+      "step": 26865
+    },
+    {
+      "epoch": 0.5374,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001043968118317865,
+      "loss": 1.2627,
+      "step": 26870
+    },
+    {
+      "epoch": 0.5375,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00010436193873653361,
+      "loss": 1.267,
+      "step": 26875
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00010432706510979171,
+      "loss": 1.2727,
+      "step": 26880
+    },
+    {
+      "epoch": 0.5377,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00010429219095581007,
+      "loss": 1.2728,
+      "step": 26885
+    },
+    {
+      "epoch": 0.5378,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00010425731627883797,
+      "loss": 1.2685,
+      "step": 26890
+    },
+    {
+      "epoch": 0.5379,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0001042224410831248,
+      "loss": 1.2677,
+      "step": 26895
+    },
+    {
+      "epoch": 0.538,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00010418756537291996,
+      "loss": 1.2765,
+      "step": 26900
+    },
+    {
+      "epoch": 0.5381,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00010415268915247303,
+      "loss": 1.2666,
+      "step": 26905
+    },
+    {
+      "epoch": 0.5382,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00010411781242603352,
+      "loss": 1.2812,
+      "step": 26910
+    },
+    {
+      "epoch": 0.5383,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00010408293519785101,
+      "loss": 1.2707,
+      "step": 26915
+    },
+    {
+      "epoch": 0.5384,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00010404805747217526,
+      "loss": 1.2739,
+      "step": 26920
+    },
+    {
+      "epoch": 0.5385,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010401317925325598,
+      "loss": 1.2767,
+      "step": 26925
+    },
+    {
+      "epoch": 0.5386,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.000103978300545343,
+      "loss": 1.2698,
+      "step": 26930
+    },
+    {
+      "epoch": 0.5387,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00010394342135268613,
+      "loss": 1.267,
+      "step": 26935
+    },
+    {
+      "epoch": 0.5388,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00010390854167953537,
+      "loss": 1.2781,
+      "step": 26940
+    },
+    {
+      "epoch": 0.5389,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00010387366153014062,
+      "loss": 1.2775,
+      "step": 26945
+    },
+    {
+      "epoch": 0.539,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00010383878090875201,
+      "loss": 1.2732,
+      "step": 26950
+    },
+    {
+      "epoch": 0.5391,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00010380389981961958,
+      "loss": 1.2779,
+      "step": 26955
+    },
+    {
+      "epoch": 0.5392,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00010376901826699348,
+      "loss": 1.2731,
+      "step": 26960
+    },
+    {
+      "epoch": 0.5393,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00010373413625512394,
+      "loss": 1.2618,
+      "step": 26965
+    },
+    {
+      "epoch": 0.5394,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001036992537882612,
+      "loss": 1.287,
+      "step": 26970
+    },
+    {
+      "epoch": 0.5395,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00010366437087065564,
+      "loss": 1.2937,
+      "step": 26975
+    },
+    {
+      "epoch": 0.5396,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00010362948750655759,
+      "loss": 1.2818,
+      "step": 26980
+    },
+    {
+      "epoch": 0.5397,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001035946037002175,
+      "loss": 1.2738,
+      "step": 26985
+    },
+    {
+      "epoch": 0.5398,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00010355971945588585,
+      "loss": 1.2735,
+      "step": 26990
+    },
+    {
+      "epoch": 0.5399,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001035248347778132,
+      "loss": 1.2897,
+      "step": 26995
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00010348994967025012,
+      "loss": 1.2681,
+      "step": 27000
+    },
+    {
+      "epoch": 0.5401,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00010345506413744726,
+      "loss": 1.2764,
+      "step": 27005
+    },
+    {
+      "epoch": 0.5402,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0001034201781836553,
+      "loss": 1.2735,
+      "step": 27010
+    },
+    {
+      "epoch": 0.5403,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00010338529181312497,
+      "loss": 1.2786,
+      "step": 27015
+    },
+    {
+      "epoch": 0.5404,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00010335040503010716,
+      "loss": 1.2709,
+      "step": 27020
+    },
+    {
+      "epoch": 0.5405,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00010331551783885263,
+      "loss": 1.2624,
+      "step": 27025
+    },
+    {
+      "epoch": 0.5406,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00010328063024361232,
+      "loss": 1.2644,
+      "step": 27030
+    },
+    {
+      "epoch": 0.5407,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00010324574224863717,
+      "loss": 1.2822,
+      "step": 27035
+    },
+    {
+      "epoch": 0.5408,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00010321085385817817,
+      "loss": 1.2698,
+      "step": 27040
+    },
+    {
+      "epoch": 0.5409,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00010317596507648636,
+      "loss": 1.2811,
+      "step": 27045
+    },
+    {
+      "epoch": 0.541,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00010314107590781284,
+      "loss": 1.2769,
+      "step": 27050
+    },
+    {
+      "epoch": 0.5411,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00010310618635640876,
+      "loss": 1.2833,
+      "step": 27055
+    },
+    {
+      "epoch": 0.5412,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010307129642652528,
+      "loss": 1.2699,
+      "step": 27060
+    },
+    {
+      "epoch": 0.5413,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00010303640612241363,
+      "loss": 1.277,
+      "step": 27065
+    },
+    {
+      "epoch": 0.5414,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00010300151544832512,
+      "loss": 1.284,
+      "step": 27070
+    },
+    {
+      "epoch": 0.5415,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00010296662440851108,
+      "loss": 1.2778,
+      "step": 27075
+    },
+    {
+      "epoch": 0.5416,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00010293173300722285,
+      "loss": 1.2745,
+      "step": 27080
+    },
+    {
+      "epoch": 0.5417,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00010289684124871182,
+      "loss": 1.278,
+      "step": 27085
+    },
+    {
+      "epoch": 0.5418,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00010286194913722948,
+      "loss": 1.2717,
+      "step": 27090
+    },
+    {
+      "epoch": 0.5419,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010282705667702734,
+      "loss": 1.2743,
+      "step": 27095
+    },
+    {
+      "epoch": 0.542,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001027921638723569,
+      "loss": 1.2598,
+      "step": 27100
+    },
+    {
+      "epoch": 0.5421,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00010275727072746977,
+      "loss": 1.2935,
+      "step": 27105
+    },
+    {
+      "epoch": 0.5422,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00010272237724661753,
+      "loss": 1.2772,
+      "step": 27110
+    },
+    {
+      "epoch": 0.5423,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00010268748343405192,
+      "loss": 1.2554,
+      "step": 27115
+    },
+    {
+      "epoch": 0.5424,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0001026525892940246,
+      "loss": 1.2788,
+      "step": 27120
+    },
+    {
+      "epoch": 0.5425,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010261769483078733,
+      "loss": 1.2608,
+      "step": 27125
+    },
+    {
+      "epoch": 0.5426,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00010258280004859188,
+      "loss": 1.2694,
+      "step": 27130
+    },
+    {
+      "epoch": 0.5427,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00010254790495169006,
+      "loss": 1.265,
+      "step": 27135
+    },
+    {
+      "epoch": 0.5428,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00010251300954433376,
+      "loss": 1.2813,
+      "step": 27140
+    },
+    {
+      "epoch": 0.5429,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00010247811383077487,
+      "loss": 1.2837,
+      "step": 27145
+    },
+    {
+      "epoch": 0.543,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00010244321781526533,
+      "loss": 1.2682,
+      "step": 27150
+    },
+    {
+      "epoch": 0.5431,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0001024083215020571,
+      "loss": 1.2861,
+      "step": 27155
+    },
+    {
+      "epoch": 0.5432,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001023734248954022,
+      "loss": 1.2782,
+      "step": 27160
+    },
+    {
+      "epoch": 0.5433,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00010233852799955268,
+      "loss": 1.2691,
+      "step": 27165
+    },
+    {
+      "epoch": 0.5434,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00010230363081876064,
+      "loss": 1.2727,
+      "step": 27170
+    },
+    {
+      "epoch": 0.5435,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00010226873335727816,
+      "loss": 1.276,
+      "step": 27175
+    },
+    {
+      "epoch": 0.5436,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00010223383561935738,
+      "loss": 1.2719,
+      "step": 27180
+    },
+    {
+      "epoch": 0.5437,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00010219893760925052,
+      "loss": 1.2736,
+      "step": 27185
+    },
+    {
+      "epoch": 0.5438,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00010216403933120979,
+      "loss": 1.2799,
+      "step": 27190
+    },
+    {
+      "epoch": 0.5439,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001021291407894874,
+      "loss": 1.304,
+      "step": 27195
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0001020942419883357,
+      "loss": 1.2904,
+      "step": 27200
+    },
+    {
+      "epoch": 0.5441,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00010205934293200696,
+      "loss": 1.3064,
+      "step": 27205
+    },
+    {
+      "epoch": 0.5442,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00010202444362475352,
+      "loss": 1.2539,
+      "step": 27210
+    },
+    {
+      "epoch": 0.5443,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001019895440708278,
+      "loss": 1.2852,
+      "step": 27215
+    },
+    {
+      "epoch": 0.5444,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00010195464427448213,
+      "loss": 1.271,
+      "step": 27220
+    },
+    {
+      "epoch": 0.5445,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.000101919744239969,
+      "loss": 1.2671,
+      "step": 27225
+    },
+    {
+      "epoch": 0.5446,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00010188484397154084,
+      "loss": 1.2793,
+      "step": 27230
+    },
+    {
+      "epoch": 0.5447,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00010184994347345016,
+      "loss": 1.2655,
+      "step": 27235
+    },
+    {
+      "epoch": 0.5448,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00010181504274994948,
+      "loss": 1.2873,
+      "step": 27240
+    },
+    {
+      "epoch": 0.5449,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00010178014180529136,
+      "loss": 1.2622,
+      "step": 27245
+    },
+    {
+      "epoch": 0.545,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00010174524064372837,
+      "loss": 1.2742,
+      "step": 27250
+    },
+    {
+      "epoch": 0.5451,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00010171033926951305,
+      "loss": 1.2784,
+      "step": 27255
+    },
+    {
+      "epoch": 0.5452,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00010167543768689815,
+      "loss": 1.2706,
+      "step": 27260
+    },
+    {
+      "epoch": 0.5453,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00010164053590013622,
+      "loss": 1.2754,
+      "step": 27265
+    },
+    {
+      "epoch": 0.5454,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00010160563391347996,
+      "loss": 1.2676,
+      "step": 27270
+    },
+    {
+      "epoch": 0.5455,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00010157073173118208,
+      "loss": 1.2686,
+      "step": 27275
+    },
+    {
+      "epoch": 0.5456,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0001015358293574953,
+      "loss": 1.2646,
+      "step": 27280
+    },
+    {
+      "epoch": 0.5457,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010150092679667238,
+      "loss": 1.2843,
+      "step": 27285
+    },
+    {
+      "epoch": 0.5458,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00010146602405296607,
+      "loss": 1.2575,
+      "step": 27290
+    },
+    {
+      "epoch": 0.5459,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00010143112113062918,
+      "loss": 1.2572,
+      "step": 27295
+    },
+    {
+      "epoch": 0.546,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00010139621803391455,
+      "loss": 1.2712,
+      "step": 27300
+    },
+    {
+      "epoch": 0.5461,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00010136131476707496,
+      "loss": 1.2746,
+      "step": 27305
+    },
+    {
+      "epoch": 0.5462,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001013264113343633,
+      "loss": 1.2702,
+      "step": 27310
+    },
+    {
+      "epoch": 0.5463,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00010129150774003245,
+      "loss": 1.2841,
+      "step": 27315
+    },
+    {
+      "epoch": 0.5464,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00010125660398833528,
+      "loss": 1.2778,
+      "step": 27320
+    },
+    {
+      "epoch": 0.5465,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00010122170008352472,
+      "loss": 1.2752,
+      "step": 27325
+    },
+    {
+      "epoch": 0.5466,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00010118679602985373,
+      "loss": 1.284,
+      "step": 27330
+    },
+    {
+      "epoch": 0.5467,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00010115189183157523,
+      "loss": 1.2795,
+      "step": 27335
+    },
+    {
+      "epoch": 0.5468,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00010111698749294223,
+      "loss": 1.2752,
+      "step": 27340
+    },
+    {
+      "epoch": 0.5469,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00010108208301820767,
+      "loss": 1.2792,
+      "step": 27345
+    },
+    {
+      "epoch": 0.547,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00010104717841162458,
+      "loss": 1.2695,
+      "step": 27350
+    },
+    {
+      "epoch": 0.5471,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.000101012273677446,
+      "loss": 1.2728,
+      "step": 27355
+    },
+    {
+      "epoch": 0.5472,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00010097736881992492,
+      "loss": 1.2713,
+      "step": 27360
+    },
+    {
+      "epoch": 0.5473,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00010094246384331442,
+      "loss": 1.2689,
+      "step": 27365
+    },
+    {
+      "epoch": 0.5474,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00010090755875186753,
+      "loss": 1.2585,
+      "step": 27370
+    },
+    {
+      "epoch": 0.5475,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0001008726535498374,
+      "loss": 1.2688,
+      "step": 27375
+    },
+    {
+      "epoch": 0.5476,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00010083774824147708,
+      "loss": 1.2854,
+      "step": 27380
+    },
+    {
+      "epoch": 0.5477,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00010080284283103965,
+      "loss": 1.2853,
+      "step": 27385
+    },
+    {
+      "epoch": 0.5478,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00010076793732277829,
+      "loss": 1.2606,
+      "step": 27390
+    },
+    {
+      "epoch": 0.5479,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00010073303172094606,
+      "loss": 1.2812,
+      "step": 27395
+    },
+    {
+      "epoch": 0.548,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00010069812602979615,
+      "loss": 1.2714,
+      "step": 27400
+    },
+    {
+      "epoch": 0.5481,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00010066322025358173,
+      "loss": 1.276,
+      "step": 27405
+    },
+    {
+      "epoch": 0.5482,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010062831439655591,
+      "loss": 1.2968,
+      "step": 27410
+    },
+    {
+      "epoch": 0.5483,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00010059340846297189,
+      "loss": 1.2784,
+      "step": 27415
+    },
+    {
+      "epoch": 0.5484,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00010055850245708283,
+      "loss": 1.2796,
+      "step": 27420
+    },
+    {
+      "epoch": 0.5485,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00010052359638314195,
+      "loss": 1.2947,
+      "step": 27425
+    },
+    {
+      "epoch": 0.5486,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00010048869024540247,
+      "loss": 1.2686,
+      "step": 27430
+    },
+    {
+      "epoch": 0.5487,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00010045378404811756,
+      "loss": 1.2813,
+      "step": 27435
+    },
+    {
+      "epoch": 0.5488,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0001004188777955404,
+      "loss": 1.2692,
+      "step": 27440
+    },
+    {
+      "epoch": 0.5489,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00010038397149192426,
+      "loss": 1.2579,
+      "step": 27445
+    },
+    {
+      "epoch": 0.549,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010034906514152238,
+      "loss": 1.2684,
+      "step": 27450
+    },
+    {
+      "epoch": 0.5491,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00010031415874858797,
+      "loss": 1.3023,
+      "step": 27455
+    },
+    {
+      "epoch": 0.5492,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00010027925231737428,
+      "loss": 1.2709,
+      "step": 27460
+    },
+    {
+      "epoch": 0.5493,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010024434585213451,
+      "loss": 1.2865,
+      "step": 27465
+    },
+    {
+      "epoch": 0.5494,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00010020943935712192,
+      "loss": 1.2858,
+      "step": 27470
+    },
+    {
+      "epoch": 0.5495,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00010017453283658984,
+      "loss": 1.2808,
+      "step": 27475
+    },
+    {
+      "epoch": 0.5496,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00010013962629479146,
+      "loss": 1.2839,
+      "step": 27480
+    },
+    {
+      "epoch": 0.5497,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010010471973598002,
+      "loss": 1.2987,
+      "step": 27485
+    },
+    {
+      "epoch": 0.5498,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00010006981316440875,
+      "loss": 1.2925,
+      "step": 27490
+    },
+    {
+      "epoch": 0.5499,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00010003490658433101,
+      "loss": 1.2973,
+      "step": 27495
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001,
+      "loss": 1.2684,
+      "step": 27500
+    },
+    {
+      "epoch": 0.5501,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.996509341566904e-05,
+      "loss": 1.283,
+      "step": 27505
+    },
+    {
+      "epoch": 0.5502,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.993018683559126e-05,
+      "loss": 1.3075,
+      "step": 27510
+    },
+    {
+      "epoch": 0.5503,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.989528026402003e-05,
+      "loss": 1.2902,
+      "step": 27515
+    },
+    {
+      "epoch": 0.5504,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.986037370520857e-05,
+      "loss": 1.2746,
+      "step": 27520
+    },
+    {
+      "epoch": 0.5505,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.98254671634102e-05,
+      "loss": 1.2763,
+      "step": 27525
+    },
+    {
+      "epoch": 0.5506,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.979056064287806e-05,
+      "loss": 1.2681,
+      "step": 27530
+    },
+    {
+      "epoch": 0.5507,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 9.975565414786551e-05,
+      "loss": 1.2754,
+      "step": 27535
+    },
+    {
+      "epoch": 0.5508,
+      "grad_norm": 0.197265625,
+      "learning_rate": 9.972074768262576e-05,
+      "loss": 1.2898,
+      "step": 27540
+    },
+    {
+      "epoch": 0.5509,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.968584125141204e-05,
+      "loss": 1.259,
+      "step": 27545
+    },
+    {
+      "epoch": 0.551,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.965093485847767e-05,
+      "loss": 1.2684,
+      "step": 27550
+    },
+    {
+      "epoch": 0.5511,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.961602850807576e-05,
+      "loss": 1.2794,
+      "step": 27555
+    },
+    {
+      "epoch": 0.5512,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.958112220445963e-05,
+      "loss": 1.2676,
+      "step": 27560
+    },
+    {
+      "epoch": 0.5513,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.954621595188247e-05,
+      "loss": 1.2727,
+      "step": 27565
+    },
+    {
+      "epoch": 0.5514,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.951130975459757e-05,
+      "loss": 1.2852,
+      "step": 27570
+    },
+    {
+      "epoch": 0.5515,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.947640361685804e-05,
+      "loss": 1.2743,
+      "step": 27575
+    },
+    {
+      "epoch": 0.5516,
+      "grad_norm": 0.435546875,
+      "learning_rate": 9.94414975429172e-05,
+      "loss": 1.2765,
+      "step": 27580
+    },
+    {
+      "epoch": 0.5517,
+      "grad_norm": 0.240234375,
+      "learning_rate": 9.940659153702813e-05,
+      "loss": 1.2904,
+      "step": 27585
+    },
+    {
+      "epoch": 0.5518,
+      "grad_norm": 0.1796875,
+      "learning_rate": 9.937168560344412e-05,
+      "loss": 1.2751,
+      "step": 27590
+    },
+    {
+      "epoch": 0.5519,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 9.933677974641831e-05,
+      "loss": 1.2732,
+      "step": 27595
+    },
+    {
+      "epoch": 0.552,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.930187397020386e-05,
+      "loss": 1.2708,
+      "step": 27600
+    },
+    {
+      "epoch": 0.5521,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.926696827905395e-05,
+      "loss": 1.2657,
+      "step": 27605
+    },
+    {
+      "epoch": 0.5522,
+      "grad_norm": 0.205078125,
+      "learning_rate": 9.923206267722173e-05,
+      "loss": 1.2824,
+      "step": 27610
+    },
+    {
+      "epoch": 0.5523,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.919715716896036e-05,
+      "loss": 1.2688,
+      "step": 27615
+    },
+    {
+      "epoch": 0.5524,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.916225175852293e-05,
+      "loss": 1.2695,
+      "step": 27620
+    },
+    {
+      "epoch": 0.5525,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.912734645016263e-05,
+      "loss": 1.2797,
+      "step": 27625
+    },
+    {
+      "epoch": 0.5526,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.909244124813247e-05,
+      "loss": 1.2808,
+      "step": 27630
+    },
+    {
+      "epoch": 0.5527,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.90575361566856e-05,
+      "loss": 1.2752,
+      "step": 27635
+    },
+    {
+      "epoch": 0.5528,
+      "grad_norm": 0.427734375,
+      "learning_rate": 9.902263118007513e-05,
+      "loss": 1.2732,
+      "step": 27640
+    },
+    {
+      "epoch": 0.5529,
+      "grad_norm": 0.23046875,
+      "learning_rate": 9.898772632255403e-05,
+      "loss": 1.2976,
+      "step": 27645
+    },
+    {
+      "epoch": 0.553,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.895282158837545e-05,
+      "loss": 1.2811,
+      "step": 27650
+    },
+    {
+      "epoch": 0.5531,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.891791698179235e-05,
+      "loss": 1.2803,
+      "step": 27655
+    },
+    {
+      "epoch": 0.5532,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.888301250705779e-05,
+      "loss": 1.2704,
+      "step": 27660
+    },
+    {
+      "epoch": 0.5533,
+      "grad_norm": 0.201171875,
+      "learning_rate": 9.884810816842475e-05,
+      "loss": 1.2681,
+      "step": 27665
+    },
+    {
+      "epoch": 0.5534,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.881320397014629e-05,
+      "loss": 1.2805,
+      "step": 27670
+    },
+    {
+      "epoch": 0.5535,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.877829991647528e-05,
+      "loss": 1.2782,
+      "step": 27675
+    },
+    {
+      "epoch": 0.5536,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.874339601166473e-05,
+      "loss": 1.2651,
+      "step": 27680
+    },
+    {
+      "epoch": 0.5537,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 9.87084922599676e-05,
+      "loss": 1.2786,
+      "step": 27685
+    },
+    {
+      "epoch": 0.5538,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.867358866563673e-05,
+      "loss": 1.274,
+      "step": 27690
+    },
+    {
+      "epoch": 0.5539,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.863868523292509e-05,
+      "loss": 1.2739,
+      "step": 27695
+    },
+    {
+      "epoch": 0.554,
+      "grad_norm": 0.484375,
+      "learning_rate": 9.860378196608549e-05,
+      "loss": 1.2856,
+      "step": 27700
+    },
+    {
+      "epoch": 0.5541,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.856887886937083e-05,
+      "loss": 1.2738,
+      "step": 27705
+    },
+    {
+      "epoch": 0.5542,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.853397594703394e-05,
+      "loss": 1.288,
+      "step": 27710
+    },
+    {
+      "epoch": 0.5543,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.849907320332766e-05,
+      "loss": 1.2678,
+      "step": 27715
+    },
+    {
+      "epoch": 0.5544,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.846417064250471e-05,
+      "loss": 1.2867,
+      "step": 27720
+    },
+    {
+      "epoch": 0.5545,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.842926826881796e-05,
+      "loss": 1.2786,
+      "step": 27725
+    },
+    {
+      "epoch": 0.5546,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 9.839436608652007e-05,
+      "loss": 1.2897,
+      "step": 27730
+    },
+    {
+      "epoch": 0.5547,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 9.83594640998638e-05,
+      "loss": 1.2875,
+      "step": 27735
+    },
+    {
+      "epoch": 0.5548,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.832456231310189e-05,
+      "loss": 1.2741,
+      "step": 27740
+    },
+    {
+      "epoch": 0.5549,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.828966073048693e-05,
+      "loss": 1.2753,
+      "step": 27745
+    },
+    {
+      "epoch": 0.555,
+      "grad_norm": 0.228515625,
+      "learning_rate": 9.825475935627165e-05,
+      "loss": 1.2856,
+      "step": 27750
+    },
+    {
+      "epoch": 0.5551,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.821985819470863e-05,
+      "loss": 1.2703,
+      "step": 27755
+    },
+    {
+      "epoch": 0.5552,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.818495725005054e-05,
+      "loss": 1.2832,
+      "step": 27760
+    },
+    {
+      "epoch": 0.5553,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.815005652654985e-05,
+      "loss": 1.2648,
+      "step": 27765
+    },
+    {
+      "epoch": 0.5554,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.81151560284592e-05,
+      "loss": 1.271,
+      "step": 27770
+    },
+    {
+      "epoch": 0.5555,
+      "grad_norm": 0.482421875,
+      "learning_rate": 9.808025576003104e-05,
+      "loss": 1.2808,
+      "step": 27775
+    },
+    {
+      "epoch": 0.5556,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.804535572551789e-05,
+      "loss": 1.2725,
+      "step": 27780
+    },
+    {
+      "epoch": 0.5557,
+      "grad_norm": 0.224609375,
+      "learning_rate": 9.801045592917226e-05,
+      "loss": 1.2743,
+      "step": 27785
+    },
+    {
+      "epoch": 0.5558,
+      "grad_norm": 0.166015625,
+      "learning_rate": 9.797555637524649e-05,
+      "loss": 1.2737,
+      "step": 27790
+    },
+    {
+      "epoch": 0.5559,
+      "grad_norm": 0.1953125,
+      "learning_rate": 9.794065706799306e-05,
+      "loss": 1.2795,
+      "step": 27795
+    },
+    {
+      "epoch": 0.556,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.790575801166432e-05,
+      "loss": 1.268,
+      "step": 27800
+    },
+    {
+      "epoch": 0.5561,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.78708592105126e-05,
+      "loss": 1.2615,
+      "step": 27805
+    },
+    {
+      "epoch": 0.5562,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.783596066879022e-05,
+      "loss": 1.2842,
+      "step": 27810
+    },
+    {
+      "epoch": 0.5563,
+      "grad_norm": 0.130859375,
+      "learning_rate": 9.78010623907495e-05,
+      "loss": 1.2789,
+      "step": 27815
+    },
+    {
+      "epoch": 0.5564,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.776616438064264e-05,
+      "loss": 1.2717,
+      "step": 27820
+    },
+    {
+      "epoch": 0.5565,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.773126664272186e-05,
+      "loss": 1.2589,
+      "step": 27825
+    },
+    {
+      "epoch": 0.5566,
+      "grad_norm": 0.484375,
+      "learning_rate": 9.76963691812394e-05,
+      "loss": 1.2697,
+      "step": 27830
+    },
+    {
+      "epoch": 0.5567,
+      "grad_norm": 0.2265625,
+      "learning_rate": 9.766147200044732e-05,
+      "loss": 1.2905,
+      "step": 27835
+    },
+    {
+      "epoch": 0.5568,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.762657510459783e-05,
+      "loss": 1.2819,
+      "step": 27840
+    },
+    {
+      "epoch": 0.5569,
+      "grad_norm": 0.40625,
+      "learning_rate": 9.759167849794292e-05,
+      "loss": 1.2849,
+      "step": 27845
+    },
+    {
+      "epoch": 0.557,
+      "grad_norm": 0.1640625,
+      "learning_rate": 9.755678218473469e-05,
+      "loss": 1.2619,
+      "step": 27850
+    },
+    {
+      "epoch": 0.5571,
+      "grad_norm": 0.2421875,
+      "learning_rate": 9.752188616922518e-05,
+      "loss": 1.2696,
+      "step": 27855
+    },
+    {
+      "epoch": 0.5572,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.748699045566626e-05,
+      "loss": 1.2615,
+      "step": 27860
+    },
+    {
+      "epoch": 0.5573,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.745209504830996e-05,
+      "loss": 1.2869,
+      "step": 27865
+    },
+    {
+      "epoch": 0.5574,
+      "grad_norm": 0.197265625,
+      "learning_rate": 9.741719995140814e-05,
+      "loss": 1.2659,
+      "step": 27870
+    },
+    {
+      "epoch": 0.5575,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.73823051692127e-05,
+      "loss": 1.2579,
+      "step": 27875
+    },
+    {
+      "epoch": 0.5576,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.734741070597539e-05,
+      "loss": 1.2649,
+      "step": 27880
+    },
+    {
+      "epoch": 0.5577,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.73125165659481e-05,
+      "loss": 1.2758,
+      "step": 27885
+    },
+    {
+      "epoch": 0.5578,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.727762275338246e-05,
+      "loss": 1.2773,
+      "step": 27890
+    },
+    {
+      "epoch": 0.5579,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.724272927253025e-05,
+      "loss": 1.2636,
+      "step": 27895
+    },
+    {
+      "epoch": 0.558,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.720783612764314e-05,
+      "loss": 1.272,
+      "step": 27900
+    },
+    {
+      "epoch": 0.5581,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.717294332297268e-05,
+      "loss": 1.2678,
+      "step": 27905
+    },
+    {
+      "epoch": 0.5582,
+      "grad_norm": 0.46484375,
+      "learning_rate": 9.713805086277054e-05,
+      "loss": 1.2754,
+      "step": 27910
+    },
+    {
+      "epoch": 0.5583,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.710315875128819e-05,
+      "loss": 1.2745,
+      "step": 27915
+    },
+    {
+      "epoch": 0.5584,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.706826699277718e-05,
+      "loss": 1.2628,
+      "step": 27920
+    },
+    {
+      "epoch": 0.5585,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.703337559148892e-05,
+      "loss": 1.2696,
+      "step": 27925
+    },
+    {
+      "epoch": 0.5586,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.699848455167489e-05,
+      "loss": 1.2605,
+      "step": 27930
+    },
+    {
+      "epoch": 0.5587,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.696359387758636e-05,
+      "loss": 1.2777,
+      "step": 27935
+    },
+    {
+      "epoch": 0.5588,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.692870357347473e-05,
+      "loss": 1.2588,
+      "step": 27940
+    },
+    {
+      "epoch": 0.5589,
+      "grad_norm": 0.236328125,
+      "learning_rate": 9.689381364359129e-05,
+      "loss": 1.2733,
+      "step": 27945
+    },
+    {
+      "epoch": 0.559,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.685892409218717e-05,
+      "loss": 1.2697,
+      "step": 27950
+    },
+    {
+      "epoch": 0.5591,
+      "grad_norm": 0.46875,
+      "learning_rate": 9.682403492351369e-05,
+      "loss": 1.2785,
+      "step": 27955
+    },
+    {
+      "epoch": 0.5592,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.678914614182185e-05,
+      "loss": 1.2578,
+      "step": 27960
+    },
+    {
+      "epoch": 0.5593,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 9.675425775136286e-05,
+      "loss": 1.286,
+      "step": 27965
+    },
+    {
+      "epoch": 0.5594,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.671936975638768e-05,
+      "loss": 1.3026,
+      "step": 27970
+    },
+    {
+      "epoch": 0.5595,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.668448216114739e-05,
+      "loss": 1.2644,
+      "step": 27975
+    },
+    {
+      "epoch": 0.5596,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.664959496989285e-05,
+      "loss": 1.2783,
+      "step": 27980
+    },
+    {
+      "epoch": 0.5597,
+      "grad_norm": 0.1953125,
+      "learning_rate": 9.661470818687503e-05,
+      "loss": 1.2758,
+      "step": 27985
+    },
+    {
+      "epoch": 0.5598,
+      "grad_norm": 0.486328125,
+      "learning_rate": 9.657982181634475e-05,
+      "loss": 1.28,
+      "step": 27990
+    },
+    {
+      "epoch": 0.5599,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.654493586255278e-05,
+      "loss": 1.257,
+      "step": 27995
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.651005032974994e-05,
+      "loss": 1.2597,
+      "step": 28000
+    },
+    {
+      "epoch": 0.5601,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.647516522218683e-05,
+      "loss": 1.2686,
+      "step": 28005
+    },
+    {
+      "epoch": 0.5602,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.644028054411416e-05,
+      "loss": 1.2888,
+      "step": 28010
+    },
+    {
+      "epoch": 0.5603,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 9.64053962997825e-05,
+      "loss": 1.2809,
+      "step": 28015
+    },
+    {
+      "epoch": 0.5604,
+      "grad_norm": 0.376953125,
+      "learning_rate": 9.637051249344243e-05,
+      "loss": 1.275,
+      "step": 28020
+    },
+    {
+      "epoch": 0.5605,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.633562912934436e-05,
+      "loss": 1.2731,
+      "step": 28025
+    },
+    {
+      "epoch": 0.5606,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.630074621173883e-05,
+      "loss": 1.2662,
+      "step": 28030
+    },
+    {
+      "epoch": 0.5607,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.62658637448761e-05,
+      "loss": 1.2733,
+      "step": 28035
+    },
+    {
+      "epoch": 0.5608,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.623098173300654e-05,
+      "loss": 1.2633,
+      "step": 28040
+    },
+    {
+      "epoch": 0.5609,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.619610018038048e-05,
+      "loss": 1.272,
+      "step": 28045
+    },
+    {
+      "epoch": 0.561,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.616121909124801e-05,
+      "loss": 1.2659,
+      "step": 28050
+    },
+    {
+      "epoch": 0.5611,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.612633846985941e-05,
+      "loss": 1.2802,
+      "step": 28055
+    },
+    {
+      "epoch": 0.5612,
+      "grad_norm": 0.140625,
+      "learning_rate": 9.609145832046465e-05,
+      "loss": 1.2625,
+      "step": 28060
+    },
+    {
+      "epoch": 0.5613,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.605657864731388e-05,
+      "loss": 1.2728,
+      "step": 28065
+    },
+    {
+      "epoch": 0.5614,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.602169945465702e-05,
+      "loss": 1.2829,
+      "step": 28070
+    },
+    {
+      "epoch": 0.5615,
+      "grad_norm": 0.146484375,
+      "learning_rate": 9.598682074674405e-05,
+      "loss": 1.2896,
+      "step": 28075
+    },
+    {
+      "epoch": 0.5616,
+      "grad_norm": 0.455078125,
+      "learning_rate": 9.595194252782477e-05,
+      "loss": 1.2998,
+      "step": 28080
+    },
+    {
+      "epoch": 0.5617,
+      "grad_norm": 0.19140625,
+      "learning_rate": 9.591706480214901e-05,
+      "loss": 1.2781,
+      "step": 28085
+    },
+    {
+      "epoch": 0.5618,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.588218757396655e-05,
+      "loss": 1.2708,
+      "step": 28090
+    },
+    {
+      "epoch": 0.5619,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.584731084752699e-05,
+      "loss": 1.2718,
+      "step": 28095
+    },
+    {
+      "epoch": 0.562,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.581243462708006e-05,
+      "loss": 1.2622,
+      "step": 28100
+    },
+    {
+      "epoch": 0.5621,
+      "grad_norm": 0.365234375,
+      "learning_rate": 9.577755891687523e-05,
+      "loss": 1.3006,
+      "step": 28105
+    },
+    {
+      "epoch": 0.5622,
+      "grad_norm": 0.46875,
+      "learning_rate": 9.574268372116205e-05,
+      "loss": 1.2687,
+      "step": 28110
+    },
+    {
+      "epoch": 0.5623,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.570780904418993e-05,
+      "loss": 1.2806,
+      "step": 28115
+    },
+    {
+      "epoch": 0.5624,
+      "grad_norm": 0.392578125,
+      "learning_rate": 9.567293489020831e-05,
+      "loss": 1.2795,
+      "step": 28120
+    },
+    {
+      "epoch": 0.5625,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 9.563806126346642e-05,
+      "loss": 1.2717,
+      "step": 28125
+    },
+    {
+      "epoch": 0.5626,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.560318816821353e-05,
+      "loss": 1.2764,
+      "step": 28130
+    },
+    {
+      "epoch": 0.5627,
+      "grad_norm": 0.490234375,
+      "learning_rate": 9.556831560869882e-05,
+      "loss": 1.2815,
+      "step": 28135
+    },
+    {
+      "epoch": 0.5628,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.55334435891714e-05,
+      "loss": 1.276,
+      "step": 28140
+    },
+    {
+      "epoch": 0.5629,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.549857211388037e-05,
+      "loss": 1.2846,
+      "step": 28145
+    },
+    {
+      "epoch": 0.563,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 9.546370118707463e-05,
+      "loss": 1.274,
+      "step": 28150
+    },
+    {
+      "epoch": 0.5631,
+      "grad_norm": 0.19140625,
+      "learning_rate": 9.542883081300316e-05,
+      "loss": 1.2784,
+      "step": 28155
+    },
+    {
+      "epoch": 0.5632,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.539396099591476e-05,
+      "loss": 1.2693,
+      "step": 28160
+    },
+    {
+      "epoch": 0.5633,
+      "grad_norm": 0.2265625,
+      "learning_rate": 9.53590917400583e-05,
+      "loss": 1.288,
+      "step": 28165
+    },
+    {
+      "epoch": 0.5634,
+      "grad_norm": 0.234375,
+      "learning_rate": 9.532422304968243e-05,
+      "loss": 1.2823,
+      "step": 28170
+    },
+    {
+      "epoch": 0.5635,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.528935492903575e-05,
+      "loss": 1.2887,
+      "step": 28175
+    },
+    {
+      "epoch": 0.5636,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 9.525448738236691e-05,
+      "loss": 1.2777,
+      "step": 28180
+    },
+    {
+      "epoch": 0.5637,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.521962041392436e-05,
+      "loss": 1.2731,
+      "step": 28185
+    },
+    {
+      "epoch": 0.5638,
+      "grad_norm": 0.48828125,
+      "learning_rate": 9.518475402795661e-05,
+      "loss": 1.2707,
+      "step": 28190
+    },
+    {
+      "epoch": 0.5639,
+      "grad_norm": 0.25,
+      "learning_rate": 9.514988822871193e-05,
+      "loss": 1.2858,
+      "step": 28195
+    },
+    {
+      "epoch": 0.564,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.511502302043868e-05,
+      "loss": 1.2653,
+      "step": 28200
+    },
+    {
+      "epoch": 0.5641,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.508015840738503e-05,
+      "loss": 1.2701,
+      "step": 28205
+    },
+    {
+      "epoch": 0.5642,
+      "grad_norm": 0.1484375,
+      "learning_rate": 9.504529439379921e-05,
+      "loss": 1.2606,
+      "step": 28210
+    },
+    {
+      "epoch": 0.5643,
+      "grad_norm": 0.427734375,
+      "learning_rate": 9.501043098392924e-05,
+      "loss": 1.2726,
+      "step": 28215
+    },
+    {
+      "epoch": 0.5644,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.497556818202306e-05,
+      "loss": 1.2679,
+      "step": 28220
+    },
+    {
+      "epoch": 0.5645,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.494070599232868e-05,
+      "loss": 1.2916,
+      "step": 28225
+    },
+    {
+      "epoch": 0.5646,
+      "grad_norm": 0.498046875,
+      "learning_rate": 9.490584441909392e-05,
+      "loss": 1.2814,
+      "step": 28230
+    },
+    {
+      "epoch": 0.5647,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.48709834665666e-05,
+      "loss": 1.2648,
+      "step": 28235
+    },
+    {
+      "epoch": 0.5648,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.483612313899435e-05,
+      "loss": 1.2664,
+      "step": 28240
+    },
+    {
+      "epoch": 0.5649,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.480126344062487e-05,
+      "loss": 1.2692,
+      "step": 28245
+    },
+    {
+      "epoch": 0.565,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.476640437570562e-05,
+      "loss": 1.2732,
+      "step": 28250
+    },
+    {
+      "epoch": 0.5651,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.473154594848415e-05,
+      "loss": 1.2751,
+      "step": 28255
+    },
+    {
+      "epoch": 0.5652,
+      "grad_norm": 0.140625,
+      "learning_rate": 9.469668816320784e-05,
+      "loss": 1.2757,
+      "step": 28260
+    },
+    {
+      "epoch": 0.5653,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 9.466183102412395e-05,
+      "loss": 1.2575,
+      "step": 28265
+    },
+    {
+      "epoch": 0.5654,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.462697453547979e-05,
+      "loss": 1.2705,
+      "step": 28270
+    },
+    {
+      "epoch": 0.5655,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.459211870152245e-05,
+      "loss": 1.277,
+      "step": 28275
+    },
+    {
+      "epoch": 0.5656,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.455726352649911e-05,
+      "loss": 1.2706,
+      "step": 28280
+    },
+    {
+      "epoch": 0.5657,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.452240901465663e-05,
+      "loss": 1.2846,
+      "step": 28285
+    },
+    {
+      "epoch": 0.5658,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 9.448755517024206e-05,
+      "loss": 1.2905,
+      "step": 28290
+    },
+    {
+      "epoch": 0.5659,
+      "grad_norm": 0.13671875,
+      "learning_rate": 9.445270199750212e-05,
+      "loss": 1.2856,
+      "step": 28295
+    },
+    {
+      "epoch": 0.566,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.441784950068362e-05,
+      "loss": 1.2703,
+      "step": 28300
+    },
+    {
+      "epoch": 0.5661,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.438299768403327e-05,
+      "loss": 1.2791,
+      "step": 28305
+    },
+    {
+      "epoch": 0.5662,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.434814655179755e-05,
+      "loss": 1.2942,
+      "step": 28310
+    },
+    {
+      "epoch": 0.5663,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 9.43132961082231e-05,
+      "loss": 1.263,
+      "step": 28315
+    },
+    {
+      "epoch": 0.5664,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.427844635755619e-05,
+      "loss": 1.2759,
+      "step": 28320
+    },
+    {
+      "epoch": 0.5665,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.424359730404329e-05,
+      "loss": 1.282,
+      "step": 28325
+    },
+    {
+      "epoch": 0.5666,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.420874895193056e-05,
+      "loss": 1.2858,
+      "step": 28330
+    },
+    {
+      "epoch": 0.5667,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.417390130546426e-05,
+      "loss": 1.2736,
+      "step": 28335
+    },
+    {
+      "epoch": 0.5668,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.413905436889035e-05,
+      "loss": 1.2748,
+      "step": 28340
+    },
+    {
+      "epoch": 0.5669,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 9.410420814645493e-05,
+      "loss": 1.2648,
+      "step": 28345
+    },
+    {
+      "epoch": 0.567,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.406936264240386e-05,
+      "loss": 1.2796,
+      "step": 28350
+    },
+    {
+      "epoch": 0.5671,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.403451786098294e-05,
+      "loss": 1.279,
+      "step": 28355
+    },
+    {
+      "epoch": 0.5672,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.399967380643796e-05,
+      "loss": 1.2836,
+      "step": 28360
+    },
+    {
+      "epoch": 0.5673,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.396483048301448e-05,
+      "loss": 1.2617,
+      "step": 28365
+    },
+    {
+      "epoch": 0.5674,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.392998789495811e-05,
+      "loss": 1.2842,
+      "step": 28370
+    },
+    {
+      "epoch": 0.5675,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.38951460465143e-05,
+      "loss": 1.2918,
+      "step": 28375
+    },
+    {
+      "epoch": 0.5676,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.386030494192846e-05,
+      "loss": 1.2644,
+      "step": 28380
+    },
+    {
+      "epoch": 0.5677,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.382546458544582e-05,
+      "loss": 1.2729,
+      "step": 28385
+    },
+    {
+      "epoch": 0.5678,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.37906249813116e-05,
+      "loss": 1.2792,
+      "step": 28390
+    },
+    {
+      "epoch": 0.5679,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.375578613377089e-05,
+      "loss": 1.2678,
+      "step": 28395
+    },
+    {
+      "epoch": 0.568,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.372094804706867e-05,
+      "loss": 1.2844,
+      "step": 28400
+    },
+    {
+      "epoch": 0.5681,
+      "grad_norm": 0.48828125,
+      "learning_rate": 9.368611072544992e-05,
+      "loss": 1.2695,
+      "step": 28405
+    },
+    {
+      "epoch": 0.5682,
+      "grad_norm": 0.439453125,
+      "learning_rate": 9.36512741731594e-05,
+      "loss": 1.2642,
+      "step": 28410
+    },
+    {
+      "epoch": 0.5683,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.361643839444188e-05,
+      "loss": 1.2643,
+      "step": 28415
+    },
+    {
+      "epoch": 0.5684,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.358160339354194e-05,
+      "loss": 1.2856,
+      "step": 28420
+    },
+    {
+      "epoch": 0.5685,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.354676917470422e-05,
+      "loss": 1.2713,
+      "step": 28425
+    },
+    {
+      "epoch": 0.5686,
+      "grad_norm": 0.201171875,
+      "learning_rate": 9.351193574217306e-05,
+      "loss": 1.2674,
+      "step": 28430
+    },
+    {
+      "epoch": 0.5687,
+      "grad_norm": 0.451171875,
+      "learning_rate": 9.347710310019288e-05,
+      "loss": 1.2765,
+      "step": 28435
+    },
+    {
+      "epoch": 0.5688,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.344227125300788e-05,
+      "loss": 1.2686,
+      "step": 28440
+    },
+    {
+      "epoch": 0.5689,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.340744020486222e-05,
+      "loss": 1.2742,
+      "step": 28445
+    },
+    {
+      "epoch": 0.569,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.337260996000002e-05,
+      "loss": 1.2656,
+      "step": 28450
+    },
+    {
+      "epoch": 0.5691,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.333778052266513e-05,
+      "loss": 1.2748,
+      "step": 28455
+    },
+    {
+      "epoch": 0.5692,
+      "grad_norm": 0.162109375,
+      "learning_rate": 9.330295189710152e-05,
+      "loss": 1.2733,
+      "step": 28460
+    },
+    {
+      "epoch": 0.5693,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.32681240875529e-05,
+      "loss": 1.2794,
+      "step": 28465
+    },
+    {
+      "epoch": 0.5694,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.323329709826294e-05,
+      "loss": 1.2768,
+      "step": 28470
+    },
+    {
+      "epoch": 0.5695,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.319847093347522e-05,
+      "loss": 1.2861,
+      "step": 28475
+    },
+    {
+      "epoch": 0.5696,
+      "grad_norm": 0.224609375,
+      "learning_rate": 9.316364559743314e-05,
+      "loss": 1.2675,
+      "step": 28480
+    },
+    {
+      "epoch": 0.5697,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.312882109438013e-05,
+      "loss": 1.2607,
+      "step": 28485
+    },
+    {
+      "epoch": 0.5698,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.309399742855942e-05,
+      "loss": 1.2672,
+      "step": 28490
+    },
+    {
+      "epoch": 0.5699,
+      "grad_norm": 0.41015625,
+      "learning_rate": 9.30591746042142e-05,
+      "loss": 1.2726,
+      "step": 28495
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.302435262558747e-05,
+      "loss": 1.2763,
+      "step": 28500
+    },
+    {
+      "epoch": 0.5701,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 9.298953149692225e-05,
+      "loss": 1.2711,
+      "step": 28505
+    },
+    {
+      "epoch": 0.5702,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 9.295471122246131e-05,
+      "loss": 1.2639,
+      "step": 28510
+    },
+    {
+      "epoch": 0.5703,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.291989180644747e-05,
+      "loss": 1.2837,
+      "step": 28515
+    },
+    {
+      "epoch": 0.5704,
+      "grad_norm": 0.169921875,
+      "learning_rate": 9.288507325312335e-05,
+      "loss": 1.2631,
+      "step": 28520
+    },
+    {
+      "epoch": 0.5705,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.285025556673141e-05,
+      "loss": 1.263,
+      "step": 28525
+    },
+    {
+      "epoch": 0.5706,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.281543875151419e-05,
+      "loss": 1.2748,
+      "step": 28530
+    },
+    {
+      "epoch": 0.5707,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.278062281171393e-05,
+      "loss": 1.2872,
+      "step": 28535
+    },
+    {
+      "epoch": 0.5708,
+      "grad_norm": 0.236328125,
+      "learning_rate": 9.274580775157294e-05,
+      "loss": 1.2746,
+      "step": 28540
+    },
+    {
+      "epoch": 0.5709,
+      "grad_norm": 0.158203125,
+      "learning_rate": 9.271099357533321e-05,
+      "loss": 1.2769,
+      "step": 28545
+    },
+    {
+      "epoch": 0.571,
+      "grad_norm": 0.41015625,
+      "learning_rate": 9.267618028723686e-05,
+      "loss": 1.3029,
+      "step": 28550
+    },
+    {
+      "epoch": 0.5711,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.264136789152567e-05,
+      "loss": 1.3,
+      "step": 28555
+    },
+    {
+      "epoch": 0.5712,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 9.26065563924415e-05,
+      "loss": 1.2771,
+      "step": 28560
+    },
+    {
+      "epoch": 0.5713,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.257174579422605e-05,
+      "loss": 1.2562,
+      "step": 28565
+    },
+    {
+      "epoch": 0.5714,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.253693610112078e-05,
+      "loss": 1.2716,
+      "step": 28570
+    },
+    {
+      "epoch": 0.5715,
+      "grad_norm": 0.19140625,
+      "learning_rate": 9.250212731736726e-05,
+      "loss": 1.2729,
+      "step": 28575
+    },
+    {
+      "epoch": 0.5716,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.246731944720675e-05,
+      "loss": 1.2604,
+      "step": 28580
+    },
+    {
+      "epoch": 0.5717,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.243251249488052e-05,
+      "loss": 1.2534,
+      "step": 28585
+    },
+    {
+      "epoch": 0.5718,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.239770646462968e-05,
+      "loss": 1.2782,
+      "step": 28590
+    },
+    {
+      "epoch": 0.5719,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.236290136069528e-05,
+      "loss": 1.2626,
+      "step": 28595
+    },
+    {
+      "epoch": 0.572,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.232809718731814e-05,
+      "loss": 1.2867,
+      "step": 28600
+    },
+    {
+      "epoch": 0.5721,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.229329394873911e-05,
+      "loss": 1.2649,
+      "step": 28605
+    },
+    {
+      "epoch": 0.5722,
+      "grad_norm": 0.244140625,
+      "learning_rate": 9.225849164919885e-05,
+      "loss": 1.2739,
+      "step": 28610
+    },
+    {
+      "epoch": 0.5723,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.222369029293787e-05,
+      "loss": 1.2883,
+      "step": 28615
+    },
+    {
+      "epoch": 0.5724,
+      "grad_norm": 0.138671875,
+      "learning_rate": 9.218888988419668e-05,
+      "loss": 1.2667,
+      "step": 28620
+    },
+    {
+      "epoch": 0.5725,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.215409042721552e-05,
+      "loss": 1.2677,
+      "step": 28625
+    },
+    {
+      "epoch": 0.5726,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.211929192623467e-05,
+      "loss": 1.2714,
+      "step": 28630
+    },
+    {
+      "epoch": 0.5727,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.208449438549415e-05,
+      "loss": 1.2827,
+      "step": 28635
+    },
+    {
+      "epoch": 0.5728,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.204969780923403e-05,
+      "loss": 1.2656,
+      "step": 28640
+    },
+    {
+      "epoch": 0.5729,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.201490220169408e-05,
+      "loss": 1.264,
+      "step": 28645
+    },
+    {
+      "epoch": 0.573,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.198010756711412e-05,
+      "loss": 1.2724,
+      "step": 28650
+    },
+    {
+      "epoch": 0.5731,
+      "grad_norm": 0.1328125,
+      "learning_rate": 9.194531390973371e-05,
+      "loss": 1.2742,
+      "step": 28655
+    },
+    {
+      "epoch": 0.5732,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.191052123379234e-05,
+      "loss": 1.292,
+      "step": 28660
+    },
+    {
+      "epoch": 0.5733,
+      "grad_norm": 0.23046875,
+      "learning_rate": 9.187572954352947e-05,
+      "loss": 1.2769,
+      "step": 28665
+    },
+    {
+      "epoch": 0.5734,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.184093884318425e-05,
+      "loss": 1.2785,
+      "step": 28670
+    },
+    {
+      "epoch": 0.5735,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.180614913699592e-05,
+      "loss": 1.2735,
+      "step": 28675
+    },
+    {
+      "epoch": 0.5736,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.177136042920344e-05,
+      "loss": 1.2873,
+      "step": 28680
+    },
+    {
+      "epoch": 0.5737,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.173657272404576e-05,
+      "loss": 1.2744,
+      "step": 28685
+    },
+    {
+      "epoch": 0.5738,
+      "grad_norm": 0.158203125,
+      "learning_rate": 9.17017860257616e-05,
+      "loss": 1.2766,
+      "step": 28690
+    },
+    {
+      "epoch": 0.5739,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.166700033858969e-05,
+      "loss": 1.2742,
+      "step": 28695
+    },
+    {
+      "epoch": 0.574,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.163221566676847e-05,
+      "loss": 1.291,
+      "step": 28700
+    },
+    {
+      "epoch": 0.5741,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.159743201453638e-05,
+      "loss": 1.2863,
+      "step": 28705
+    },
+    {
+      "epoch": 0.5742,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.156264938613174e-05,
+      "loss": 1.282,
+      "step": 28710
+    },
+    {
+      "epoch": 0.5743,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.152786778579267e-05,
+      "loss": 1.2779,
+      "step": 28715
+    },
+    {
+      "epoch": 0.5744,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.14930872177572e-05,
+      "loss": 1.268,
+      "step": 28720
+    },
+    {
+      "epoch": 0.5745,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.145830768626327e-05,
+      "loss": 1.2698,
+      "step": 28725
+    },
+    {
+      "epoch": 0.5746,
+      "grad_norm": 0.228515625,
+      "learning_rate": 9.142352919554862e-05,
+      "loss": 1.2819,
+      "step": 28730
+    },
+    {
+      "epoch": 0.5747,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.138875174985091e-05,
+      "loss": 1.2645,
+      "step": 28735
+    },
+    {
+      "epoch": 0.5748,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.135397535340773e-05,
+      "loss": 1.2866,
+      "step": 28740
+    },
+    {
+      "epoch": 0.5749,
+      "grad_norm": 0.146484375,
+      "learning_rate": 9.131920001045638e-05,
+      "loss": 1.2926,
+      "step": 28745
+    },
+    {
+      "epoch": 0.575,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.128442572523417e-05,
+      "loss": 1.2686,
+      "step": 28750
+    },
+    {
+      "epoch": 0.5751,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.12496525019783e-05,
+      "loss": 1.2712,
+      "step": 28755
+    },
+    {
+      "epoch": 0.5752,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.121488034492569e-05,
+      "loss": 1.2835,
+      "step": 28760
+    },
+    {
+      "epoch": 0.5753,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.11801092583133e-05,
+      "loss": 1.2735,
+      "step": 28765
+    },
+    {
+      "epoch": 0.5754,
+      "grad_norm": 0.15625,
+      "learning_rate": 9.114533924637778e-05,
+      "loss": 1.2636,
+      "step": 28770
+    },
+    {
+      "epoch": 0.5755,
+      "grad_norm": 0.234375,
+      "learning_rate": 9.111057031335585e-05,
+      "loss": 1.2707,
+      "step": 28775
+    },
+    {
+      "epoch": 0.5756,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.107580246348395e-05,
+      "loss": 1.2621,
+      "step": 28780
+    },
+    {
+      "epoch": 0.5757,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.104103570099848e-05,
+      "loss": 1.2703,
+      "step": 28785
+    },
+    {
+      "epoch": 0.5758,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.100627003013562e-05,
+      "loss": 1.2713,
+      "step": 28790
+    },
+    {
+      "epoch": 0.5759,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.097150545513145e-05,
+      "loss": 1.2574,
+      "step": 28795
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 9.093674198022201e-05,
+      "loss": 1.2599,
+      "step": 28800
+    },
+    {
+      "epoch": 0.5761,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.090197960964301e-05,
+      "loss": 1.2815,
+      "step": 28805
+    },
+    {
+      "epoch": 0.5762,
+      "grad_norm": 0.22265625,
+      "learning_rate": 9.086721834763024e-05,
+      "loss": 1.2906,
+      "step": 28810
+    },
+    {
+      "epoch": 0.5763,
+      "grad_norm": 0.15234375,
+      "learning_rate": 9.083245819841918e-05,
+      "loss": 1.2877,
+      "step": 28815
+    },
+    {
+      "epoch": 0.5764,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.07976991662453e-05,
+      "loss": 1.2722,
+      "step": 28820
+    },
+    {
+      "epoch": 0.5765,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.076294125534383e-05,
+      "loss": 1.2716,
+      "step": 28825
+    },
+    {
+      "epoch": 0.5766,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.072818446994999e-05,
+      "loss": 1.2859,
+      "step": 28830
+    },
+    {
+      "epoch": 0.5767,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.069342881429876e-05,
+      "loss": 1.2749,
+      "step": 28835
+    },
+    {
+      "epoch": 0.5768,
+      "grad_norm": 0.470703125,
+      "learning_rate": 9.065867429262496e-05,
+      "loss": 1.2611,
+      "step": 28840
+    },
+    {
+      "epoch": 0.5769,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.062392090916337e-05,
+      "loss": 1.2622,
+      "step": 28845
+    },
+    {
+      "epoch": 0.577,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.058916866814858e-05,
+      "loss": 1.2689,
+      "step": 28850
+    },
+    {
+      "epoch": 0.5771,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.055441757381506e-05,
+      "loss": 1.2787,
+      "step": 28855
+    },
+    {
+      "epoch": 0.5772,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.051966763039707e-05,
+      "loss": 1.2794,
+      "step": 28860
+    },
+    {
+      "epoch": 0.5773,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.048491884212884e-05,
+      "loss": 1.2851,
+      "step": 28865
+    },
+    {
+      "epoch": 0.5774,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.045017121324438e-05,
+      "loss": 1.2833,
+      "step": 28870
+    },
+    {
+      "epoch": 0.5775,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.04154247479776e-05,
+      "loss": 1.2607,
+      "step": 28875
+    },
+    {
+      "epoch": 0.5776,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.038067945056227e-05,
+      "loss": 1.2854,
+      "step": 28880
+    },
+    {
+      "epoch": 0.5777,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.034593532523193e-05,
+      "loss": 1.2759,
+      "step": 28885
+    },
+    {
+      "epoch": 0.5778,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.03111923762201e-05,
+      "loss": 1.272,
+      "step": 28890
+    },
+    {
+      "epoch": 0.5779,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.027645060776006e-05,
+      "loss": 1.2632,
+      "step": 28895
+    },
+    {
+      "epoch": 0.578,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.024171002408506e-05,
+      "loss": 1.2841,
+      "step": 28900
+    },
+    {
+      "epoch": 0.5781,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.020697062942807e-05,
+      "loss": 1.285,
+      "step": 28905
+    },
+    {
+      "epoch": 0.5782,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.017223242802204e-05,
+      "loss": 1.2787,
+      "step": 28910
+    },
+    {
+      "epoch": 0.5783,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.013749542409963e-05,
+      "loss": 1.2748,
+      "step": 28915
+    },
+    {
+      "epoch": 0.5784,
+      "grad_norm": 0.46875,
+      "learning_rate": 9.01027596218935e-05,
+      "loss": 1.285,
+      "step": 28920
+    },
+    {
+      "epoch": 0.5785,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.006802502563612e-05,
+      "loss": 1.2765,
+      "step": 28925
+    },
+    {
+      "epoch": 0.5786,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.003329163955972e-05,
+      "loss": 1.2708,
+      "step": 28930
+    },
+    {
+      "epoch": 0.5787,
+      "grad_norm": 0.248046875,
+      "learning_rate": 8.999855946789653e-05,
+      "loss": 1.2786,
+      "step": 28935
+    },
+    {
+      "epoch": 0.5788,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.99638285148785e-05,
+      "loss": 1.2686,
+      "step": 28940
+    },
+    {
+      "epoch": 0.5789,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 8.992909878473758e-05,
+      "loss": 1.2737,
+      "step": 28945
+    },
+    {
+      "epoch": 0.579,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.989437028170537e-05,
+      "loss": 1.2636,
+      "step": 28950
+    },
+    {
+      "epoch": 0.5791,
+      "grad_norm": 0.2734375,
+      "learning_rate": 8.985964301001353e-05,
+      "loss": 1.2696,
+      "step": 28955
+    },
+    {
+      "epoch": 0.5792,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.982491697389338e-05,
+      "loss": 1.2733,
+      "step": 28960
+    },
+    {
+      "epoch": 0.5793,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.979019217757625e-05,
+      "loss": 1.2643,
+      "step": 28965
+    },
+    {
+      "epoch": 0.5794,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.975546862529328e-05,
+      "loss": 1.2752,
+      "step": 28970
+    },
+    {
+      "epoch": 0.5795,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.972074632127533e-05,
+      "loss": 1.2818,
+      "step": 28975
+    },
+    {
+      "epoch": 0.5796,
+      "grad_norm": 0.421875,
+      "learning_rate": 8.96860252697533e-05,
+      "loss": 1.279,
+      "step": 28980
+    },
+    {
+      "epoch": 0.5797,
+      "grad_norm": 0.142578125,
+      "learning_rate": 8.965130547495776e-05,
+      "loss": 1.2807,
+      "step": 28985
+    },
+    {
+      "epoch": 0.5798,
+      "grad_norm": 0.25,
+      "learning_rate": 8.961658694111929e-05,
+      "loss": 1.2696,
+      "step": 28990
+    },
+    {
+      "epoch": 0.5799,
+      "grad_norm": 0.41015625,
+      "learning_rate": 8.958186967246816e-05,
+      "loss": 1.2749,
+      "step": 28995
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.954715367323468e-05,
+      "loss": 1.2768,
+      "step": 29000
+    },
+    {
+      "epoch": 0.5801,
+      "grad_norm": 0.291015625,
+      "learning_rate": 8.951243894764876e-05,
+      "loss": 1.2807,
+      "step": 29005
+    },
+    {
+      "epoch": 0.5802,
+      "grad_norm": 0.318359375,
+      "learning_rate": 8.947772549994035e-05,
+      "loss": 1.2683,
+      "step": 29010
+    },
+    {
+      "epoch": 0.5803,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.944301333433922e-05,
+      "loss": 1.2778,
+      "step": 29015
+    },
+    {
+      "epoch": 0.5804,
+      "grad_norm": 0.150390625,
+      "learning_rate": 8.940830245507483e-05,
+      "loss": 1.2678,
+      "step": 29020
+    },
+    {
+      "epoch": 0.5805,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.93735928663767e-05,
+      "loss": 1.2859,
+      "step": 29025
+    },
+    {
+      "epoch": 0.5806,
+      "grad_norm": 0.205078125,
+      "learning_rate": 8.933888457247402e-05,
+      "loss": 1.2644,
+      "step": 29030
+    },
+    {
+      "epoch": 0.5807,
+      "grad_norm": 0.140625,
+      "learning_rate": 8.930417757759592e-05,
+      "loss": 1.2785,
+      "step": 29035
+    },
+    {
+      "epoch": 0.5808,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.926947188597134e-05,
+      "loss": 1.269,
+      "step": 29040
+    },
+    {
+      "epoch": 0.5809,
+      "grad_norm": 0.212890625,
+      "learning_rate": 8.923476750182908e-05,
+      "loss": 1.288,
+      "step": 29045
+    },
+    {
+      "epoch": 0.581,
+      "grad_norm": 0.291015625,
+      "learning_rate": 8.920006442939772e-05,
+      "loss": 1.2821,
+      "step": 29050
+    },
+    {
+      "epoch": 0.5811,
+      "grad_norm": 0.208984375,
+      "learning_rate": 8.916536267290578e-05,
+      "loss": 1.2729,
+      "step": 29055
+    },
+    {
+      "epoch": 0.5812,
+      "grad_norm": 0.28125,
+      "learning_rate": 8.913066223658151e-05,
+      "loss": 1.2816,
+      "step": 29060
+    },
+    {
+      "epoch": 0.5813,
+      "grad_norm": 0.36328125,
+      "learning_rate": 8.909596312465306e-05,
+      "loss": 1.2674,
+      "step": 29065
+    },
+    {
+      "epoch": 0.5814,
+      "grad_norm": 0.26171875,
+      "learning_rate": 8.906126534134848e-05,
+      "loss": 1.2714,
+      "step": 29070
+    },
+    {
+      "epoch": 0.5815,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.902656889089548e-05,
+      "loss": 1.2943,
+      "step": 29075
+    },
+    {
+      "epoch": 0.5816,
+      "grad_norm": 0.203125,
+      "learning_rate": 8.89918737775218e-05,
+      "loss": 1.2849,
+      "step": 29080
+    },
+    {
+      "epoch": 0.5817,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.895718000545489e-05,
+      "loss": 1.2697,
+      "step": 29085
+    },
+    {
+      "epoch": 0.5818,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.892248757892214e-05,
+      "loss": 1.2761,
+      "step": 29090
+    },
+    {
+      "epoch": 0.5819,
+      "grad_norm": 0.3046875,
+      "learning_rate": 8.888779650215068e-05,
+      "loss": 1.2713,
+      "step": 29095
+    },
+    {
+      "epoch": 0.582,
+      "grad_norm": 0.310546875,
+      "learning_rate": 8.885310677936746e-05,
+      "loss": 1.2833,
+      "step": 29100
+    },
+    {
+      "epoch": 0.5821,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 8.88184184147994e-05,
+      "loss": 1.2727,
+      "step": 29105
+    },
+    {
+      "epoch": 0.5822,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.878373141267311e-05,
+      "loss": 1.2643,
+      "step": 29110
+    },
+    {
+      "epoch": 0.5823,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.874904577721518e-05,
+      "loss": 1.2608,
+      "step": 29115
+    },
+    {
+      "epoch": 0.5824,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.871436151265184e-05,
+      "loss": 1.2777,
+      "step": 29120
+    },
+    {
+      "epoch": 0.5825,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.867967862320934e-05,
+      "loss": 1.2735,
+      "step": 29125
+    },
+    {
+      "epoch": 0.5826,
+      "grad_norm": 0.435546875,
+      "learning_rate": 8.864499711311362e-05,
+      "loss": 1.2651,
+      "step": 29130
+    },
+    {
+      "epoch": 0.5827,
+      "grad_norm": 0.279296875,
+      "learning_rate": 8.861031698659063e-05,
+      "loss": 1.2741,
+      "step": 29135
+    },
+    {
+      "epoch": 0.5828,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 8.857563824786596e-05,
+      "loss": 1.2586,
+      "step": 29140
+    },
+    {
+      "epoch": 0.5829,
+      "grad_norm": 0.328125,
+      "learning_rate": 8.854096090116508e-05,
+      "loss": 1.2841,
+      "step": 29145
+    },
+    {
+      "epoch": 0.583,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.850628495071336e-05,
+      "loss": 1.2647,
+      "step": 29150
+    },
+    {
+      "epoch": 0.5831,
+      "grad_norm": 0.142578125,
+      "learning_rate": 8.847161040073594e-05,
+      "loss": 1.2638,
+      "step": 29155
+    },
+    {
+      "epoch": 0.5832,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 8.843693725545786e-05,
+      "loss": 1.3008,
+      "step": 29160
+    },
+    {
+      "epoch": 0.5833,
+      "grad_norm": 0.470703125,
+      "learning_rate": 8.840226551910387e-05,
+      "loss": 1.2868,
+      "step": 29165
+    },
+    {
+      "epoch": 0.5834,
+      "grad_norm": 0.1796875,
+      "learning_rate": 8.836759519589867e-05,
+      "loss": 1.2779,
+      "step": 29170
+    },
+    {
+      "epoch": 0.5835,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.833292629006668e-05,
+      "loss": 1.2804,
+      "step": 29175
+    },
+    {
+      "epoch": 0.5836,
+      "grad_norm": 0.294921875,
+      "learning_rate": 8.829825880583226e-05,
+      "loss": 1.2866,
+      "step": 29180
+    },
+    {
+      "epoch": 0.5837,
+      "grad_norm": 0.435546875,
+      "learning_rate": 8.826359274741953e-05,
+      "loss": 1.2848,
+      "step": 29185
+    },
+    {
+      "epoch": 0.5838,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.822892811905237e-05,
+      "loss": 1.2719,
+      "step": 29190
+    },
+    {
+      "epoch": 0.5839,
+      "grad_norm": 0.25390625,
+      "learning_rate": 8.819426492495464e-05,
+      "loss": 1.2705,
+      "step": 29195
+    },
+    {
+      "epoch": 0.584,
+      "grad_norm": 0.447265625,
+      "learning_rate": 8.81596031693499e-05,
+      "loss": 1.265,
+      "step": 29200
+    },
+    {
+      "epoch": 0.5841,
+      "grad_norm": 0.248046875,
+      "learning_rate": 8.812494285646163e-05,
+      "loss": 1.2703,
+      "step": 29205
+    },
+    {
+      "epoch": 0.5842,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.809028399051302e-05,
+      "loss": 1.2733,
+      "step": 29210
+    },
+    {
+      "epoch": 0.5843,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.805562657572723e-05,
+      "loss": 1.2652,
+      "step": 29215
+    },
+    {
+      "epoch": 0.5844,
+      "grad_norm": 0.3046875,
+      "learning_rate": 8.802097061632705e-05,
+      "loss": 1.2813,
+      "step": 29220
+    },
+    {
+      "epoch": 0.5845,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.79863161165353e-05,
+      "loss": 1.2676,
+      "step": 29225
+    },
+    {
+      "epoch": 0.5846,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.79516630805745e-05,
+      "loss": 1.2764,
+      "step": 29230
+    },
+    {
+      "epoch": 0.5847,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.791701151266696e-05,
+      "loss": 1.2778,
+      "step": 29235
+    },
+    {
+      "epoch": 0.5848,
+      "grad_norm": 0.302734375,
+      "learning_rate": 8.788236141703498e-05,
+      "loss": 1.285,
+      "step": 29240
+    },
+    {
+      "epoch": 0.5849,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 8.784771279790044e-05,
+      "loss": 1.2758,
+      "step": 29245
+    },
+    {
+      "epoch": 0.585,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.781306565948528e-05,
+      "loss": 1.2775,
+      "step": 29250
+    },
+    {
+      "epoch": 0.5851,
+      "grad_norm": 0.359375,
+      "learning_rate": 8.777842000601105e-05,
+      "loss": 1.2759,
+      "step": 29255
+    },
+    {
+      "epoch": 0.5852,
+      "grad_norm": 0.2734375,
+      "learning_rate": 8.774377584169933e-05,
+      "loss": 1.2873,
+      "step": 29260
+    },
+    {
+      "epoch": 0.5853,
+      "grad_norm": 0.162109375,
+      "learning_rate": 8.77091331707713e-05,
+      "loss": 1.2694,
+      "step": 29265
+    },
+    {
+      "epoch": 0.5854,
+      "grad_norm": 0.162109375,
+      "learning_rate": 8.767449199744814e-05,
+      "loss": 1.2619,
+      "step": 29270
+    },
+    {
+      "epoch": 0.5855,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.763985232595075e-05,
+      "loss": 1.2814,
+      "step": 29275
+    },
+    {
+      "epoch": 0.5856,
+      "grad_norm": 0.17578125,
+      "learning_rate": 8.760521416049983e-05,
+      "loss": 1.2705,
+      "step": 29280
+    },
+    {
+      "epoch": 0.5857,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.7570577505316e-05,
+      "loss": 1.2601,
+      "step": 29285
+    },
+    {
+      "epoch": 0.5858,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.753594236461957e-05,
+      "loss": 1.2895,
+      "step": 29290
+    },
+    {
+      "epoch": 0.5859,
+      "grad_norm": 0.25,
+      "learning_rate": 8.750130874263077e-05,
+      "loss": 1.2645,
+      "step": 29295
+    },
+    {
+      "epoch": 0.586,
+      "grad_norm": 0.25390625,
+      "learning_rate": 8.746667664356956e-05,
+      "loss": 1.2697,
+      "step": 29300
+    },
+    {
+      "epoch": 0.5861,
+      "grad_norm": 0.38671875,
+      "learning_rate": 8.743204607165583e-05,
+      "loss": 1.2806,
+      "step": 29305
+    },
+    {
+      "epoch": 0.5862,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.739741703110913e-05,
+      "loss": 1.2725,
+      "step": 29310
+    },
+    {
+      "epoch": 0.5863,
+      "grad_norm": 0.28125,
+      "learning_rate": 8.736278952614898e-05,
+      "loss": 1.2676,
+      "step": 29315
+    },
+    {
+      "epoch": 0.5864,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.732816356099456e-05,
+      "loss": 1.2847,
+      "step": 29320
+    },
+    {
+      "epoch": 0.5865,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 8.729353913986496e-05,
+      "loss": 1.2595,
+      "step": 29325
+    },
+    {
+      "epoch": 0.5866,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 8.72589162669791e-05,
+      "loss": 1.2753,
+      "step": 29330
+    },
+    {
+      "epoch": 0.5867,
+      "grad_norm": 0.462890625,
+      "learning_rate": 8.722429494655561e-05,
+      "loss": 1.2758,
+      "step": 29335
+    },
+    {
+      "epoch": 0.5868,
+      "grad_norm": 0.447265625,
+      "learning_rate": 8.718967518281307e-05,
+      "loss": 1.275,
+      "step": 29340
+    },
+    {
+      "epoch": 0.5869,
+      "grad_norm": 0.455078125,
+      "learning_rate": 8.715505697996971e-05,
+      "loss": 1.2731,
+      "step": 29345
+    },
+    {
+      "epoch": 0.587,
+      "grad_norm": 0.21875,
+      "learning_rate": 8.712044034224374e-05,
+      "loss": 1.2767,
+      "step": 29350
+    },
+    {
+      "epoch": 0.5871,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 8.708582527385301e-05,
+      "loss": 1.2814,
+      "step": 29355
+    },
+    {
+      "epoch": 0.5872,
+      "grad_norm": 0.404296875,
+      "learning_rate": 8.705121177901532e-05,
+      "loss": 1.2647,
+      "step": 29360
+    },
+    {
+      "epoch": 0.5873,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.70165998619482e-05,
+      "loss": 1.2843,
+      "step": 29365
+    },
+    {
+      "epoch": 0.5874,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.698198952686896e-05,
+      "loss": 1.2673,
+      "step": 29370
+    },
+    {
+      "epoch": 0.5875,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.694738077799488e-05,
+      "loss": 1.2708,
+      "step": 29375
+    },
+    {
+      "epoch": 0.5876,
+      "grad_norm": 0.18359375,
+      "learning_rate": 8.691277361954279e-05,
+      "loss": 1.2715,
+      "step": 29380
+    },
+    {
+      "epoch": 0.5877,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.687816805572956e-05,
+      "loss": 1.2595,
+      "step": 29385
+    },
+    {
+      "epoch": 0.5878,
+      "grad_norm": 0.2421875,
+      "learning_rate": 8.684356409077176e-05,
+      "loss": 1.2763,
+      "step": 29390
+    },
+    {
+      "epoch": 0.5879,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.680896172888576e-05,
+      "loss": 1.2774,
+      "step": 29395
+    },
+    {
+      "epoch": 0.588,
+      "grad_norm": 0.30859375,
+      "learning_rate": 8.677436097428775e-05,
+      "loss": 1.2849,
+      "step": 29400
+    },
+    {
+      "epoch": 0.5881,
+      "grad_norm": 0.298828125,
+      "learning_rate": 8.673976183119376e-05,
+      "loss": 1.2647,
+      "step": 29405
+    },
+    {
+      "epoch": 0.5882,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 8.670516430381958e-05,
+      "loss": 1.2714,
+      "step": 29410
+    },
+    {
+      "epoch": 0.5883,
+      "grad_norm": 0.158203125,
+      "learning_rate": 8.667056839638075e-05,
+      "loss": 1.282,
+      "step": 29415
+    },
+    {
+      "epoch": 0.5884,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.663597411309279e-05,
+      "loss": 1.2731,
+      "step": 29420
+    },
+    {
+      "epoch": 0.5885,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 8.660138145817079e-05,
+      "loss": 1.2802,
+      "step": 29425
+    },
+    {
+      "epoch": 0.5886,
+      "grad_norm": 0.240234375,
+      "learning_rate": 8.656679043582986e-05,
+      "loss": 1.2656,
+      "step": 29430
+    },
+    {
+      "epoch": 0.5887,
+      "grad_norm": 0.265625,
+      "learning_rate": 8.653220105028474e-05,
+      "loss": 1.2771,
+      "step": 29435
+    },
+    {
+      "epoch": 0.5888,
+      "grad_norm": 0.380859375,
+      "learning_rate": 8.64976133057501e-05,
+      "loss": 1.2866,
+      "step": 29440
+    },
+    {
+      "epoch": 0.5889,
+      "grad_norm": 0.28125,
+      "learning_rate": 8.646302720644027e-05,
+      "loss": 1.2695,
+      "step": 29445
+    },
+    {
+      "epoch": 0.589,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.642844275656957e-05,
+      "loss": 1.2808,
+      "step": 29450
+    },
+    {
+      "epoch": 0.5891,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.639385996035194e-05,
+      "loss": 1.2768,
+      "step": 29455
+    },
+    {
+      "epoch": 0.5892,
+      "grad_norm": 0.1875,
+      "learning_rate": 8.635927882200116e-05,
+      "loss": 1.2911,
+      "step": 29460
+    },
+    {
+      "epoch": 0.5893,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.632469934573094e-05,
+      "loss": 1.2708,
+      "step": 29465
+    },
+    {
+      "epoch": 0.5894,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 8.629012153575458e-05,
+      "loss": 1.2682,
+      "step": 29470
+    },
+    {
+      "epoch": 0.5895,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.625554539628535e-05,
+      "loss": 1.2718,
+      "step": 29475
+    },
+    {
+      "epoch": 0.5896,
+      "grad_norm": 1.5234375,
+      "learning_rate": 8.62209709315362e-05,
+      "loss": 1.2815,
+      "step": 29480
+    },
+    {
+      "epoch": 0.5897,
+      "grad_norm": 0.375,
+      "learning_rate": 8.618639814571996e-05,
+      "loss": 1.2691,
+      "step": 29485
+    },
+    {
+      "epoch": 0.5898,
+      "grad_norm": 0.318359375,
+      "learning_rate": 8.615182704304918e-05,
+      "loss": 1.2703,
+      "step": 29490
+    },
+    {
+      "epoch": 0.5899,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 8.611725762773631e-05,
+      "loss": 1.2661,
+      "step": 29495
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.608268990399349e-05,
+      "loss": 1.2563,
+      "step": 29500
+    },
+    {
+      "epoch": 0.5901,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 8.604812387603265e-05,
+      "loss": 1.2782,
+      "step": 29505
+    },
+    {
+      "epoch": 0.5902,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.601355954806561e-05,
+      "loss": 1.2633,
+      "step": 29510
+    },
+    {
+      "epoch": 0.5903,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 8.597899692430389e-05,
+      "loss": 1.2685,
+      "step": 29515
+    },
+    {
+      "epoch": 0.5904,
+      "grad_norm": 0.486328125,
+      "learning_rate": 8.594443600895892e-05,
+      "loss": 1.2774,
+      "step": 29520
+    },
+    {
+      "epoch": 0.5905,
+      "grad_norm": 0.263671875,
+      "learning_rate": 8.590987680624174e-05,
+      "loss": 1.2616,
+      "step": 29525
+    },
+    {
+      "epoch": 0.5906,
+      "grad_norm": 0.328125,
+      "learning_rate": 8.587531932036335e-05,
+      "loss": 1.2648,
+      "step": 29530
+    },
+    {
+      "epoch": 0.5907,
+      "grad_norm": 0.21875,
+      "learning_rate": 8.584076355553444e-05,
+      "loss": 1.268,
+      "step": 29535
+    },
+    {
+      "epoch": 0.5908,
+      "grad_norm": 0.388671875,
+      "learning_rate": 8.580620951596557e-05,
+      "loss": 1.2725,
+      "step": 29540
+    },
+    {
+      "epoch": 0.5909,
+      "grad_norm": 0.458984375,
+      "learning_rate": 8.577165720586703e-05,
+      "loss": 1.2649,
+      "step": 29545
+    },
+    {
+      "epoch": 0.591,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.573710662944885e-05,
+      "loss": 1.273,
+      "step": 29550
+    },
+    {
+      "epoch": 0.5911,
+      "grad_norm": 0.23828125,
+      "learning_rate": 8.570255779092098e-05,
+      "loss": 1.2839,
+      "step": 29555
+    },
+    {
+      "epoch": 0.5912,
+      "grad_norm": 0.44921875,
+      "learning_rate": 8.566801069449305e-05,
+      "loss": 1.2781,
+      "step": 29560
+    },
+    {
+      "epoch": 0.5913,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.56334653443746e-05,
+      "loss": 1.2913,
+      "step": 29565
+    },
+    {
+      "epoch": 0.5914,
+      "grad_norm": 0.45703125,
+      "learning_rate": 8.559892174477479e-05,
+      "loss": 1.293,
+      "step": 29570
+    },
+    {
+      "epoch": 0.5915,
+      "grad_norm": 0.875,
+      "learning_rate": 8.55643798999027e-05,
+      "loss": 1.2904,
+      "step": 29575
+    },
+    {
+      "epoch": 0.5916,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.55298398139671e-05,
+      "loss": 1.2825,
+      "step": 29580
+    },
+    {
+      "epoch": 0.5917,
+      "grad_norm": 0.134765625,
+      "learning_rate": 8.549530149117664e-05,
+      "loss": 1.2682,
+      "step": 29585
+    },
+    {
+      "epoch": 0.5918,
+      "grad_norm": 0.44921875,
+      "learning_rate": 8.546076493573972e-05,
+      "loss": 1.2733,
+      "step": 29590
+    },
+    {
+      "epoch": 0.5919,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.542623015186445e-05,
+      "loss": 1.2728,
+      "step": 29595
+    },
+    {
+      "epoch": 0.592,
+      "grad_norm": 0.21875,
+      "learning_rate": 8.539169714375885e-05,
+      "loss": 1.2735,
+      "step": 29600
+    },
+    {
+      "epoch": 0.5921,
+      "grad_norm": 0.375,
+      "learning_rate": 8.535716591563062e-05,
+      "loss": 1.266,
+      "step": 29605
+    },
+    {
+      "epoch": 0.5922,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.532263647168735e-05,
+      "loss": 1.2849,
+      "step": 29610
+    },
+    {
+      "epoch": 0.5923,
+      "grad_norm": 0.267578125,
+      "learning_rate": 8.528810881613626e-05,
+      "loss": 1.2725,
+      "step": 29615
+    },
+    {
+      "epoch": 0.5924,
+      "grad_norm": 0.322265625,
+      "learning_rate": 8.525358295318454e-05,
+      "loss": 1.2927,
+      "step": 29620
+    },
+    {
+      "epoch": 0.5925,
+      "grad_norm": 0.49609375,
+      "learning_rate": 8.521905888703893e-05,
+      "loss": 1.2772,
+      "step": 29625
+    },
+    {
+      "epoch": 0.5926,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.51845366219062e-05,
+      "loss": 1.2705,
+      "step": 29630
+    },
+    {
+      "epoch": 0.5927,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.515001616199279e-05,
+      "loss": 1.2831,
+      "step": 29635
+    },
+    {
+      "epoch": 0.5928,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.511549751150479e-05,
+      "loss": 1.2656,
+      "step": 29640
+    },
+    {
+      "epoch": 0.5929,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.508098067464832e-05,
+      "loss": 1.2768,
+      "step": 29645
+    },
+    {
+      "epoch": 0.593,
+      "grad_norm": 0.212890625,
+      "learning_rate": 8.504646565562906e-05,
+      "loss": 1.2691,
+      "step": 29650
+    },
+    {
+      "epoch": 0.5931,
+      "grad_norm": 0.326171875,
+      "learning_rate": 8.501195245865263e-05,
+      "loss": 1.2895,
+      "step": 29655
+    },
+    {
+      "epoch": 0.5932,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.497744108792429e-05,
+      "loss": 1.2854,
+      "step": 29660
+    },
+    {
+      "epoch": 0.5933,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 8.494293154764924e-05,
+      "loss": 1.262,
+      "step": 29665
+    },
+    {
+      "epoch": 0.5934,
+      "grad_norm": 0.119140625,
+      "learning_rate": 8.490842384203225e-05,
+      "loss": 1.2709,
+      "step": 29670
+    },
+    {
+      "epoch": 0.5935,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.487391797527808e-05,
+      "loss": 1.2623,
+      "step": 29675
+    },
+    {
+      "epoch": 0.5936,
+      "grad_norm": 0.3671875,
+      "learning_rate": 8.483941395159114e-05,
+      "loss": 1.2718,
+      "step": 29680
+    },
+    {
+      "epoch": 0.5937,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.480491177517557e-05,
+      "loss": 1.2699,
+      "step": 29685
+    },
+    {
+      "epoch": 0.5938,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.477041145023546e-05,
+      "loss": 1.271,
+      "step": 29690
+    },
+    {
+      "epoch": 0.5939,
+      "grad_norm": 0.361328125,
+      "learning_rate": 8.473591298097448e-05,
+      "loss": 1.2701,
+      "step": 29695
+    },
+    {
+      "epoch": 0.594,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 8.47014163715962e-05,
+      "loss": 1.2724,
+      "step": 29700
+    },
+    {
+      "epoch": 0.5941,
+      "grad_norm": 0.306640625,
+      "learning_rate": 8.466692162630392e-05,
+      "loss": 1.2785,
+      "step": 29705
+    },
+    {
+      "epoch": 0.5942,
+      "grad_norm": 0.490234375,
+      "learning_rate": 8.463242874930079e-05,
+      "loss": 1.2874,
+      "step": 29710
+    },
+    {
+      "epoch": 0.5943,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 8.459793774478957e-05,
+      "loss": 1.2764,
+      "step": 29715
+    },
+    {
+      "epoch": 0.5944,
+      "grad_norm": 0.17578125,
+      "learning_rate": 8.456344861697289e-05,
+      "loss": 1.2947,
+      "step": 29720
+    },
+    {
+      "epoch": 0.5945,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 8.452896137005321e-05,
+      "loss": 1.2711,
+      "step": 29725
+    },
+    {
+      "epoch": 0.5946,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.449447600823262e-05,
+      "loss": 1.2738,
+      "step": 29730
+    },
+    {
+      "epoch": 0.5947,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.445999253571315e-05,
+      "loss": 1.2772,
+      "step": 29735
+    },
+    {
+      "epoch": 0.5948,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.442551095669639e-05,
+      "loss": 1.2827,
+      "step": 29740
+    },
+    {
+      "epoch": 0.5949,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.439103127538392e-05,
+      "loss": 1.2819,
+      "step": 29745
+    },
+    {
+      "epoch": 0.595,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.435655349597689e-05,
+      "loss": 1.2878,
+      "step": 29750
+    },
+    {
+      "epoch": 0.5951,
+      "grad_norm": 0.28515625,
+      "learning_rate": 8.432207762267644e-05,
+      "loss": 1.2861,
+      "step": 29755
+    },
+    {
+      "epoch": 0.5952,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.428760365968326e-05,
+      "loss": 1.2844,
+      "step": 29760
+    },
+    {
+      "epoch": 0.5953,
+      "grad_norm": 0.498046875,
+      "learning_rate": 8.425313161119787e-05,
+      "loss": 1.2727,
+      "step": 29765
+    },
+    {
+      "epoch": 0.5954,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.421866148142066e-05,
+      "loss": 1.2662,
+      "step": 29770
+    },
+    {
+      "epoch": 0.5955,
+      "grad_norm": 0.24609375,
+      "learning_rate": 8.418419327455164e-05,
+      "loss": 1.2806,
+      "step": 29775
+    },
+    {
+      "epoch": 0.5956,
+      "grad_norm": 0.390625,
+      "learning_rate": 8.414972699479075e-05,
+      "loss": 1.2894,
+      "step": 29780
+    },
+    {
+      "epoch": 0.5957,
+      "grad_norm": 0.408203125,
+      "learning_rate": 8.41152626463375e-05,
+      "loss": 1.2941,
+      "step": 29785
+    },
+    {
+      "epoch": 0.5958,
+      "grad_norm": 0.14453125,
+      "learning_rate": 8.408080023339133e-05,
+      "loss": 1.2639,
+      "step": 29790
+    },
+    {
+      "epoch": 0.5959,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.404633976015134e-05,
+      "loss": 1.2676,
+      "step": 29795
+    },
+    {
+      "epoch": 0.596,
+      "grad_norm": 0.474609375,
+      "learning_rate": 8.401188123081653e-05,
+      "loss": 1.2871,
+      "step": 29800
+    },
+    {
+      "epoch": 0.5961,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.397742464958547e-05,
+      "loss": 1.2727,
+      "step": 29805
+    },
+    {
+      "epoch": 0.5962,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.394297002065658e-05,
+      "loss": 1.2732,
+      "step": 29810
+    },
+    {
+      "epoch": 0.5963,
+      "grad_norm": 0.298828125,
+      "learning_rate": 8.390851734822808e-05,
+      "loss": 1.2687,
+      "step": 29815
+    },
+    {
+      "epoch": 0.5964,
+      "grad_norm": 0.322265625,
+      "learning_rate": 8.387406663649795e-05,
+      "loss": 1.2742,
+      "step": 29820
+    },
+    {
+      "epoch": 0.5965,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 8.383961788966391e-05,
+      "loss": 1.2668,
+      "step": 29825
+    },
+    {
+      "epoch": 0.5966,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.380517111192337e-05,
+      "loss": 1.2742,
+      "step": 29830
+    },
+    {
+      "epoch": 0.5967,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 8.377072630747364e-05,
+      "loss": 1.2757,
+      "step": 29835
+    },
+    {
+      "epoch": 0.5968,
+      "grad_norm": 0.2109375,
+      "learning_rate": 8.373628348051165e-05,
+      "loss": 1.2769,
+      "step": 29840
+    },
+    {
+      "epoch": 0.5969,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.37018426352342e-05,
+      "loss": 1.2729,
+      "step": 29845
+    },
+    {
+      "epoch": 0.597,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 8.366740377583781e-05,
+      "loss": 1.2663,
+      "step": 29850
+    },
+    {
+      "epoch": 0.5971,
+      "grad_norm": 0.26171875,
+      "learning_rate": 8.363296690651868e-05,
+      "loss": 1.2731,
+      "step": 29855
+    },
+    {
+      "epoch": 0.5972,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.359853203147291e-05,
+      "loss": 1.2778,
+      "step": 29860
+    },
+    {
+      "epoch": 0.5973,
+      "grad_norm": 0.30859375,
+      "learning_rate": 8.356409915489625e-05,
+      "loss": 1.2618,
+      "step": 29865
+    },
+    {
+      "epoch": 0.5974,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.352966828098428e-05,
+      "loss": 1.2779,
+      "step": 29870
+    },
+    {
+      "epoch": 0.5975,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.349523941393224e-05,
+      "loss": 1.2875,
+      "step": 29875
+    },
+    {
+      "epoch": 0.5976,
+      "grad_norm": 0.150390625,
+      "learning_rate": 8.346081255793525e-05,
+      "loss": 1.2742,
+      "step": 29880
+    },
+    {
+      "epoch": 0.5977,
+      "grad_norm": 0.205078125,
+      "learning_rate": 8.342638771718802e-05,
+      "loss": 1.2569,
+      "step": 29885
+    },
+    {
+      "epoch": 0.5978,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.339196489588523e-05,
+      "loss": 1.2864,
+      "step": 29890
+    },
+    {
+      "epoch": 0.5979,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.335754409822114e-05,
+      "loss": 1.2679,
+      "step": 29895
+    },
+    {
+      "epoch": 0.598,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.332312532838978e-05,
+      "loss": 1.2714,
+      "step": 29900
+    },
+    {
+      "epoch": 0.5981,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.328870859058506e-05,
+      "loss": 1.2784,
+      "step": 29905
+    },
+    {
+      "epoch": 0.5982,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.325429388900046e-05,
+      "loss": 1.2692,
+      "step": 29910
+    },
+    {
+      "epoch": 0.5983,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 8.321988122782937e-05,
+      "loss": 1.2786,
+      "step": 29915
+    },
+    {
+      "epoch": 0.5984,
+      "grad_norm": 0.388671875,
+      "learning_rate": 8.318547061126485e-05,
+      "loss": 1.2836,
+      "step": 29920
+    },
+    {
+      "epoch": 0.5985,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 8.315106204349976e-05,
+      "loss": 1.2823,
+      "step": 29925
+    },
+    {
+      "epoch": 0.5986,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.311665552872662e-05,
+      "loss": 1.2686,
+      "step": 29930
+    },
+    {
+      "epoch": 0.5987,
+      "grad_norm": 0.3046875,
+      "learning_rate": 8.30822510711378e-05,
+      "loss": 1.2818,
+      "step": 29935
+    },
+    {
+      "epoch": 0.5988,
+      "grad_norm": 0.5,
+      "learning_rate": 8.30478486749254e-05,
+      "loss": 1.262,
+      "step": 29940
+    },
+    {
+      "epoch": 0.5989,
+      "grad_norm": 0.17578125,
+      "learning_rate": 8.301344834428116e-05,
+      "loss": 1.2702,
+      "step": 29945
+    },
+    {
+      "epoch": 0.599,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.297905008339677e-05,
+      "loss": 1.2863,
+      "step": 29950
+    },
+    {
+      "epoch": 0.5991,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.294465389646345e-05,
+      "loss": 1.2678,
+      "step": 29955
+    },
+    {
+      "epoch": 0.5992,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.291025978767235e-05,
+      "loss": 1.2843,
+      "step": 29960
+    },
+    {
+      "epoch": 0.5993,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.287586776121423e-05,
+      "loss": 1.2784,
+      "step": 29965
+    },
+    {
+      "epoch": 0.5994,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 8.284147782127971e-05,
+      "loss": 1.2779,
+      "step": 29970
+    },
+    {
+      "epoch": 0.5995,
+      "grad_norm": 0.30078125,
+      "learning_rate": 8.280708997205904e-05,
+      "loss": 1.2623,
+      "step": 29975
+    },
+    {
+      "epoch": 0.5996,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.277270421774235e-05,
+      "loss": 1.2833,
+      "step": 29980
+    },
+    {
+      "epoch": 0.5997,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.273832056251937e-05,
+      "loss": 1.2771,
+      "step": 29985
+    },
+    {
+      "epoch": 0.5998,
+      "grad_norm": 0.20703125,
+      "learning_rate": 8.270393901057964e-05,
+      "loss": 1.2686,
+      "step": 29990
+    },
+    {
+      "epoch": 0.5999,
+      "grad_norm": 0.4921875,
+      "learning_rate": 8.266955956611253e-05,
+      "loss": 1.2721,
+      "step": 29995
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 8.263518223330697e-05,
+      "loss": 1.2715,
+      "step": 30000
+    },
+    {
+      "epoch": 0.6001,
+      "grad_norm": 0.28125,
+      "learning_rate": 8.26008070163518e-05,
+      "loss": 1.2839,
+      "step": 30005
+    },
+    {
+      "epoch": 0.6002,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.256643391943551e-05,
+      "loss": 1.2832,
+      "step": 30010
+    },
+    {
+      "epoch": 0.6003,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.25320629467464e-05,
+      "loss": 1.2633,
+      "step": 30015
+    },
+    {
+      "epoch": 0.6004,
+      "grad_norm": 0.49609375,
+      "learning_rate": 8.249769410247239e-05,
+      "loss": 1.276,
+      "step": 30020
+    },
+    {
+      "epoch": 0.6005,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 8.24633273908013e-05,
+      "loss": 1.2861,
+      "step": 30025
+    },
+    {
+      "epoch": 0.6006,
+      "grad_norm": 0.392578125,
+      "learning_rate": 8.242896281592057e-05,
+      "loss": 1.2746,
+      "step": 30030
+    },
+    {
+      "epoch": 0.6007,
+      "grad_norm": 0.267578125,
+      "learning_rate": 8.239460038201739e-05,
+      "loss": 1.2675,
+      "step": 30035
+    },
+    {
+      "epoch": 0.6008,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.236024009327879e-05,
+      "loss": 1.2706,
+      "step": 30040
+    },
+    {
+      "epoch": 0.6009,
+      "grad_norm": 0.26171875,
+      "learning_rate": 8.23258819538914e-05,
+      "loss": 1.2693,
+      "step": 30045
+    },
+    {
+      "epoch": 0.601,
+      "grad_norm": 0.40234375,
+      "learning_rate": 8.229152596804168e-05,
+      "loss": 1.2657,
+      "step": 30050
+    },
+    {
+      "epoch": 0.6011,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.225717213991579e-05,
+      "loss": 1.2837,
+      "step": 30055
+    },
+    {
+      "epoch": 0.6012,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.222282047369971e-05,
+      "loss": 1.2668,
+      "step": 30060
+    },
+    {
+      "epoch": 0.6013,
+      "grad_norm": 0.27734375,
+      "learning_rate": 8.218847097357898e-05,
+      "loss": 1.2716,
+      "step": 30065
+    },
+    {
+      "epoch": 0.6014,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.215412364373907e-05,
+      "loss": 1.2898,
+      "step": 30070
+    },
+    {
+      "epoch": 0.6015,
+      "grad_norm": 0.28515625,
+      "learning_rate": 8.211977848836506e-05,
+      "loss": 1.2881,
+      "step": 30075
+    },
+    {
+      "epoch": 0.6016,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 8.208543551164178e-05,
+      "loss": 1.2747,
+      "step": 30080
+    },
+    {
+      "epoch": 0.6017,
+      "grad_norm": 0.49609375,
+      "learning_rate": 8.205109471775387e-05,
+      "loss": 1.2784,
+      "step": 30085
+    },
+    {
+      "epoch": 0.6018,
+      "grad_norm": 0.154296875,
+      "learning_rate": 8.201675611088558e-05,
+      "loss": 1.2684,
+      "step": 30090
+    },
+    {
+      "epoch": 0.6019,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.198241969522107e-05,
+      "loss": 1.2571,
+      "step": 30095
+    },
+    {
+      "epoch": 0.602,
+      "grad_norm": 0.20703125,
+      "learning_rate": 8.194808547494401e-05,
+      "loss": 1.2581,
+      "step": 30100
+    },
+    {
+      "epoch": 0.6021,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.191375345423799e-05,
+      "loss": 1.2676,
+      "step": 30105
+    },
+    {
+      "epoch": 0.6022,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.187942363728625e-05,
+      "loss": 1.2708,
+      "step": 30110
+    },
+    {
+      "epoch": 0.6023,
+      "grad_norm": 0.314453125,
+      "learning_rate": 8.184509602827181e-05,
+      "loss": 1.2727,
+      "step": 30115
+    },
+    {
+      "epoch": 0.6024,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.181077063137733e-05,
+      "loss": 1.278,
+      "step": 30120
+    },
+    {
+      "epoch": 0.6025,
+      "grad_norm": 0.451171875,
+      "learning_rate": 8.177644745078526e-05,
+      "loss": 1.2786,
+      "step": 30125
+    },
+    {
+      "epoch": 0.6026,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.174212649067781e-05,
+      "loss": 1.2766,
+      "step": 30130
+    },
+    {
+      "epoch": 0.6027,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.170780775523684e-05,
+      "loss": 1.2665,
+      "step": 30135
+    },
+    {
+      "epoch": 0.6028,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 8.167349124864405e-05,
+      "loss": 1.2686,
+      "step": 30140
+    },
+    {
+      "epoch": 0.6029,
+      "grad_norm": 0.15625,
+      "learning_rate": 8.163917697508072e-05,
+      "loss": 1.276,
+      "step": 30145
+    },
+    {
+      "epoch": 0.603,
+      "grad_norm": 0.328125,
+      "learning_rate": 8.160486493872798e-05,
+      "loss": 1.2791,
+      "step": 30150
+    },
+    {
+      "epoch": 0.6031,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 8.157055514376666e-05,
+      "loss": 1.2658,
+      "step": 30155
+    },
+    {
+      "epoch": 0.6032,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.153624759437732e-05,
+      "loss": 1.2785,
+      "step": 30160
+    },
+    {
+      "epoch": 0.6033,
+      "grad_norm": 0.24609375,
+      "learning_rate": 8.15019422947402e-05,
+      "loss": 1.2839,
+      "step": 30165
+    },
+    {
+      "epoch": 0.6034,
+      "grad_norm": 0.3046875,
+      "learning_rate": 8.146763924903527e-05,
+      "loss": 1.2633,
+      "step": 30170
+    },
+    {
+      "epoch": 0.6035,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 8.14333384614423e-05,
+      "loss": 1.2754,
+      "step": 30175
+    },
+    {
+      "epoch": 0.6036,
+      "grad_norm": 0.203125,
+      "learning_rate": 8.139903993614068e-05,
+      "loss": 1.2818,
+      "step": 30180
+    },
+    {
+      "epoch": 0.6037,
+      "grad_norm": 0.3671875,
+      "learning_rate": 8.136474367730969e-05,
+      "loss": 1.2663,
+      "step": 30185
+    },
+    {
+      "epoch": 0.6038,
+      "grad_norm": 0.189453125,
+      "learning_rate": 8.133044968912811e-05,
+      "loss": 1.2754,
+      "step": 30190
+    },
+    {
+      "epoch": 0.6039,
+      "grad_norm": 0.146484375,
+      "learning_rate": 8.129615797577461e-05,
+      "loss": 1.2667,
+      "step": 30195
+    },
+    {
+      "epoch": 0.604,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.126186854142752e-05,
+      "loss": 1.2599,
+      "step": 30200
+    },
+    {
+      "epoch": 0.6041,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.122758139026495e-05,
+      "loss": 1.2673,
+      "step": 30205
+    },
+    {
+      "epoch": 0.6042,
+      "grad_norm": 0.171875,
+      "learning_rate": 8.119329652646463e-05,
+      "loss": 1.2734,
+      "step": 30210
+    },
+    {
+      "epoch": 0.6043,
+      "grad_norm": 0.490234375,
+      "learning_rate": 8.115901395420407e-05,
+      "loss": 1.2747,
+      "step": 30215
+    },
+    {
+      "epoch": 0.6044,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 8.11247336776605e-05,
+      "loss": 1.2683,
+      "step": 30220
+    },
+    {
+      "epoch": 0.6045,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.109045570101086e-05,
+      "loss": 1.286,
+      "step": 30225
+    },
+    {
+      "epoch": 0.6046,
+      "grad_norm": 0.451171875,
+      "learning_rate": 8.105618002843189e-05,
+      "loss": 1.2788,
+      "step": 30230
+    },
+    {
+      "epoch": 0.6047,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.102190666409987e-05,
+      "loss": 1.2616,
+      "step": 30235
+    },
+    {
+      "epoch": 0.6048,
+      "grad_norm": 0.259765625,
+      "learning_rate": 8.0987635612191e-05,
+      "loss": 1.2922,
+      "step": 30240
+    },
+    {
+      "epoch": 0.6049,
+      "grad_norm": 0.392578125,
+      "learning_rate": 8.095336687688102e-05,
+      "loss": 1.2682,
+      "step": 30245
+    },
+    {
+      "epoch": 0.605,
+      "grad_norm": 0.314453125,
+      "learning_rate": 8.091910046234552e-05,
+      "loss": 1.2801,
+      "step": 30250
+    },
+    {
+      "epoch": 0.6051,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 8.088483637275979e-05,
+      "loss": 1.2744,
+      "step": 30255
+    },
+    {
+      "epoch": 0.6052,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.085057461229872e-05,
+      "loss": 1.2693,
+      "step": 30260
+    },
+    {
+      "epoch": 0.6053,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 8.081631518513704e-05,
+      "loss": 1.2826,
+      "step": 30265
+    },
+    {
+      "epoch": 0.6054,
+      "grad_norm": 0.294921875,
+      "learning_rate": 8.078205809544917e-05,
+      "loss": 1.2832,
+      "step": 30270
+    },
+    {
+      "epoch": 0.6055,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 8.074780334740928e-05,
+      "loss": 1.2706,
+      "step": 30275
+    },
+    {
+      "epoch": 0.6056,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 8.071355094519109e-05,
+      "loss": 1.277,
+      "step": 30280
+    },
+    {
+      "epoch": 0.6057,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.067930089296827e-05,
+      "loss": 1.2672,
+      "step": 30285
+    },
+    {
+      "epoch": 0.6058,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 8.064505319491398e-05,
+      "loss": 1.2711,
+      "step": 30290
+    },
+    {
+      "epoch": 0.6059,
+      "grad_norm": 0.25,
+      "learning_rate": 8.061080785520126e-05,
+      "loss": 1.269,
+      "step": 30295
+    },
+    {
+      "epoch": 0.606,
+      "grad_norm": 0.140625,
+      "learning_rate": 8.057656487800282e-05,
+      "loss": 1.2721,
+      "step": 30300
+    },
+    {
+      "epoch": 0.6061,
+      "grad_norm": 0.423828125,
+      "learning_rate": 8.0542324267491e-05,
+      "loss": 1.2617,
+      "step": 30305
+    },
+    {
+      "epoch": 0.6062,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.050808602783795e-05,
+      "loss": 1.2859,
+      "step": 30310
+    },
+    {
+      "epoch": 0.6063,
+      "grad_norm": 0.146484375,
+      "learning_rate": 8.047385016321552e-05,
+      "loss": 1.2709,
+      "step": 30315
+    },
+    {
+      "epoch": 0.6064,
+      "grad_norm": 0.322265625,
+      "learning_rate": 8.04396166777952e-05,
+      "loss": 1.2664,
+      "step": 30320
+    },
+    {
+      "epoch": 0.6065,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.040538557574822e-05,
+      "loss": 1.2712,
+      "step": 30325
+    },
+    {
+      "epoch": 0.6066,
+      "grad_norm": 0.32421875,
+      "learning_rate": 8.037115686124564e-05,
+      "loss": 1.2958,
+      "step": 30330
+    },
+    {
+      "epoch": 0.6067,
+      "grad_norm": 0.291015625,
+      "learning_rate": 8.033693053845801e-05,
+      "loss": 1.274,
+      "step": 30335
+    },
+    {
+      "epoch": 0.6068,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.030270661155574e-05,
+      "loss": 1.2759,
+      "step": 30340
+    },
+    {
+      "epoch": 0.6069,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.026848508470897e-05,
+      "loss": 1.2774,
+      "step": 30345
+    },
+    {
+      "epoch": 0.607,
+      "grad_norm": 0.4609375,
+      "learning_rate": 8.023426596208739e-05,
+      "loss": 1.2823,
+      "step": 30350
+    },
+    {
+      "epoch": 0.6071,
+      "grad_norm": 0.22265625,
+      "learning_rate": 8.020004924786059e-05,
+      "loss": 1.2689,
+      "step": 30355
+    },
+    {
+      "epoch": 0.6072,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.016583494619769e-05,
+      "loss": 1.2752,
+      "step": 30360
+    },
+    {
+      "epoch": 0.6073,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.013162306126765e-05,
+      "loss": 1.2697,
+      "step": 30365
+    },
+    {
+      "epoch": 0.6074,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.009741359723906e-05,
+      "loss": 1.2632,
+      "step": 30370
+    },
+    {
+      "epoch": 0.6075,
+      "grad_norm": 0.22265625,
+      "learning_rate": 8.00632065582803e-05,
+      "loss": 1.2734,
+      "step": 30375
+    },
+    {
+      "epoch": 0.6076,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 8.002900194855932e-05,
+      "loss": 1.2886,
+      "step": 30380
+    },
+    {
+      "epoch": 0.6077,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.999479977224384e-05,
+      "loss": 1.281,
+      "step": 30385
+    },
+    {
+      "epoch": 0.6078,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.996060003350139e-05,
+      "loss": 1.279,
+      "step": 30390
+    },
+    {
+      "epoch": 0.6079,
+      "grad_norm": 0.2421875,
+      "learning_rate": 7.992640273649898e-05,
+      "loss": 1.2679,
+      "step": 30395
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.989220788540355e-05,
+      "loss": 1.2811,
+      "step": 30400
+    },
+    {
+      "epoch": 0.6081,
+      "grad_norm": 0.46484375,
+      "learning_rate": 7.985801548438157e-05,
+      "loss": 1.2727,
+      "step": 30405
+    },
+    {
+      "epoch": 0.6082,
+      "grad_norm": 0.31640625,
+      "learning_rate": 7.982382553759931e-05,
+      "loss": 1.2751,
+      "step": 30410
+    },
+    {
+      "epoch": 0.6083,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.97896380492227e-05,
+      "loss": 1.2694,
+      "step": 30415
+    },
+    {
+      "epoch": 0.6084,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 7.975545302341743e-05,
+      "loss": 1.2783,
+      "step": 30420
+    },
+    {
+      "epoch": 0.6085,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.972127046434878e-05,
+      "loss": 1.2708,
+      "step": 30425
+    },
+    {
+      "epoch": 0.6086,
+      "grad_norm": 0.28125,
+      "learning_rate": 7.96870903761818e-05,
+      "loss": 1.2701,
+      "step": 30430
+    },
+    {
+      "epoch": 0.6087,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.965291276308124e-05,
+      "loss": 1.2767,
+      "step": 30435
+    },
+    {
+      "epoch": 0.6088,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 7.961873762921153e-05,
+      "loss": 1.2726,
+      "step": 30440
+    },
+    {
+      "epoch": 0.6089,
+      "grad_norm": 0.16015625,
+      "learning_rate": 7.958456497873685e-05,
+      "loss": 1.2694,
+      "step": 30445
+    },
+    {
+      "epoch": 0.609,
+      "grad_norm": 0.466796875,
+      "learning_rate": 7.955039481582097e-05,
+      "loss": 1.2829,
+      "step": 30450
+    },
+    {
+      "epoch": 0.6091,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 7.951622714462746e-05,
+      "loss": 1.2854,
+      "step": 30455
+    },
+    {
+      "epoch": 0.6092,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.948206196931954e-05,
+      "loss": 1.2872,
+      "step": 30460
+    },
+    {
+      "epoch": 0.6093,
+      "grad_norm": 0.322265625,
+      "learning_rate": 7.944789929406016e-05,
+      "loss": 1.2703,
+      "step": 30465
+    },
+    {
+      "epoch": 0.6094,
+      "grad_norm": 0.306640625,
+      "learning_rate": 7.941373912301189e-05,
+      "loss": 1.2785,
+      "step": 30470
+    },
+    {
+      "epoch": 0.6095,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.937958146033705e-05,
+      "loss": 1.2786,
+      "step": 30475
+    },
+    {
+      "epoch": 0.6096,
+      "grad_norm": 0.4921875,
+      "learning_rate": 7.934542631019768e-05,
+      "loss": 1.2893,
+      "step": 30480
+    },
+    {
+      "epoch": 0.6097,
+      "grad_norm": 0.29296875,
+      "learning_rate": 7.931127367675543e-05,
+      "loss": 1.2794,
+      "step": 30485
+    },
+    {
+      "epoch": 0.6098,
+      "grad_norm": 0.19921875,
+      "learning_rate": 7.927712356417176e-05,
+      "loss": 1.2784,
+      "step": 30490
+    },
+    {
+      "epoch": 0.6099,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.92429759766077e-05,
+      "loss": 1.2752,
+      "step": 30495
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.920883091822408e-05,
+      "loss": 1.2763,
+      "step": 30500
+    },
+    {
+      "epoch": 0.6101,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.917468839318132e-05,
+      "loss": 1.2698,
+      "step": 30505
+    },
+    {
+      "epoch": 0.6102,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.914054840563963e-05,
+      "loss": 1.2757,
+      "step": 30510
+    },
+    {
+      "epoch": 0.6103,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.910641095975886e-05,
+      "loss": 1.2695,
+      "step": 30515
+    },
+    {
+      "epoch": 0.6104,
+      "grad_norm": 0.330078125,
+      "learning_rate": 7.907227605969849e-05,
+      "loss": 1.2765,
+      "step": 30520
+    },
+    {
+      "epoch": 0.6105,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.903814370961784e-05,
+      "loss": 1.281,
+      "step": 30525
+    },
+    {
+      "epoch": 0.6106,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.900401391367576e-05,
+      "loss": 1.2677,
+      "step": 30530
+    },
+    {
+      "epoch": 0.6107,
+      "grad_norm": 0.361328125,
+      "learning_rate": 7.896988667603093e-05,
+      "loss": 1.2759,
+      "step": 30535
+    },
+    {
+      "epoch": 0.6108,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.893576200084159e-05,
+      "loss": 1.285,
+      "step": 30540
+    },
+    {
+      "epoch": 0.6109,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.89016398922658e-05,
+      "loss": 1.2771,
+      "step": 30545
+    },
+    {
+      "epoch": 0.611,
+      "grad_norm": 0.48046875,
+      "learning_rate": 7.886752035446114e-05,
+      "loss": 1.278,
+      "step": 30550
+    },
+    {
+      "epoch": 0.6111,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 7.883340339158505e-05,
+      "loss": 1.2741,
+      "step": 30555
+    },
+    {
+      "epoch": 0.6112,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.879928900779456e-05,
+      "loss": 1.2655,
+      "step": 30560
+    },
+    {
+      "epoch": 0.6113,
+      "grad_norm": 0.296875,
+      "learning_rate": 7.876517720724636e-05,
+      "loss": 1.2844,
+      "step": 30565
+    },
+    {
+      "epoch": 0.6114,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.873106799409695e-05,
+      "loss": 1.2727,
+      "step": 30570
+    },
+    {
+      "epoch": 0.6115,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.869696137250235e-05,
+      "loss": 1.2824,
+      "step": 30575
+    },
+    {
+      "epoch": 0.6116,
+      "grad_norm": 0.3046875,
+      "learning_rate": 7.866285734661841e-05,
+      "loss": 1.2822,
+      "step": 30580
+    },
+    {
+      "epoch": 0.6117,
+      "grad_norm": 0.1953125,
+      "learning_rate": 7.862875592060056e-05,
+      "loss": 1.2976,
+      "step": 30585
+    },
+    {
+      "epoch": 0.6118,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.8594657098604e-05,
+      "loss": 1.2749,
+      "step": 30590
+    },
+    {
+      "epoch": 0.6119,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 7.856056088478352e-05,
+      "loss": 1.2725,
+      "step": 30595
+    },
+    {
+      "epoch": 0.612,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.852646728329368e-05,
+      "loss": 1.2616,
+      "step": 30600
+    },
+    {
+      "epoch": 0.6121,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 7.849237629828869e-05,
+      "loss": 1.2636,
+      "step": 30605
+    },
+    {
+      "epoch": 0.6122,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 7.845828793392236e-05,
+      "loss": 1.2854,
+      "step": 30610
+    },
+    {
+      "epoch": 0.6123,
+      "grad_norm": 0.197265625,
+      "learning_rate": 7.842420219434833e-05,
+      "loss": 1.286,
+      "step": 30615
+    },
+    {
+      "epoch": 0.6124,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.83901190837198e-05,
+      "loss": 1.2744,
+      "step": 30620
+    },
+    {
+      "epoch": 0.6125,
+      "grad_norm": 0.416015625,
+      "learning_rate": 7.835603860618972e-05,
+      "loss": 1.2797,
+      "step": 30625
+    },
+    {
+      "epoch": 0.6126,
+      "grad_norm": 0.322265625,
+      "learning_rate": 7.832196076591067e-05,
+      "loss": 1.2583,
+      "step": 30630
+    },
+    {
+      "epoch": 0.6127,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.828788556703498e-05,
+      "loss": 1.2685,
+      "step": 30635
+    },
+    {
+      "epoch": 0.6128,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.825381301371452e-05,
+      "loss": 1.2659,
+      "step": 30640
+    },
+    {
+      "epoch": 0.6129,
+      "grad_norm": 0.283203125,
+      "learning_rate": 7.821974311010102e-05,
+      "loss": 1.2686,
+      "step": 30645
+    },
+    {
+      "epoch": 0.613,
+      "grad_norm": 0.265625,
+      "learning_rate": 7.818567586034577e-05,
+      "loss": 1.2791,
+      "step": 30650
+    },
+    {
+      "epoch": 0.6131,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.81516112685997e-05,
+      "loss": 1.2638,
+      "step": 30655
+    },
+    {
+      "epoch": 0.6132,
+      "grad_norm": 0.48046875,
+      "learning_rate": 7.811754933901358e-05,
+      "loss": 1.2731,
+      "step": 30660
+    },
+    {
+      "epoch": 0.6133,
+      "grad_norm": 0.30859375,
+      "learning_rate": 7.808349007573763e-05,
+      "loss": 1.2683,
+      "step": 30665
+    },
+    {
+      "epoch": 0.6134,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 7.804943348292197e-05,
+      "loss": 1.2797,
+      "step": 30670
+    },
+    {
+      "epoch": 0.6135,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.801537956471624e-05,
+      "loss": 1.2781,
+      "step": 30675
+    },
+    {
+      "epoch": 0.6136,
+      "grad_norm": 0.228515625,
+      "learning_rate": 7.798132832526986e-05,
+      "loss": 1.2847,
+      "step": 30680
+    },
+    {
+      "epoch": 0.6137,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 7.79472797687318e-05,
+      "loss": 1.2699,
+      "step": 30685
+    },
+    {
+      "epoch": 0.6138,
+      "grad_norm": 0.236328125,
+      "learning_rate": 7.791323389925084e-05,
+      "loss": 1.2734,
+      "step": 30690
+    },
+    {
+      "epoch": 0.6139,
+      "grad_norm": 0.326171875,
+      "learning_rate": 7.787919072097531e-05,
+      "loss": 1.2817,
+      "step": 30695
+    },
+    {
+      "epoch": 0.614,
+      "grad_norm": 0.625,
+      "learning_rate": 7.784515023805328e-05,
+      "loss": 1.2855,
+      "step": 30700
+    },
+    {
+      "epoch": 0.6141,
+      "grad_norm": 0.30078125,
+      "learning_rate": 7.781111245463252e-05,
+      "loss": 1.2814,
+      "step": 30705
+    },
+    {
+      "epoch": 0.6142,
+      "grad_norm": 0.265625,
+      "learning_rate": 7.777707737486037e-05,
+      "loss": 1.2652,
+      "step": 30710
+    },
+    {
+      "epoch": 0.6143,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.774304500288394e-05,
+      "loss": 1.2728,
+      "step": 30715
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 0.3671875,
+      "learning_rate": 7.770901534284995e-05,
+      "loss": 1.2711,
+      "step": 30720
+    },
+    {
+      "epoch": 0.6145,
+      "grad_norm": 0.208984375,
+      "learning_rate": 7.767498839890488e-05,
+      "loss": 1.2636,
+      "step": 30725
+    },
+    {
+      "epoch": 0.6146,
+      "grad_norm": 0.39453125,
+      "learning_rate": 7.76409641751947e-05,
+      "loss": 1.2777,
+      "step": 30730
+    },
+    {
+      "epoch": 0.6147,
+      "grad_norm": 0.306640625,
+      "learning_rate": 7.760694267586525e-05,
+      "loss": 1.2816,
+      "step": 30735
+    },
+    {
+      "epoch": 0.6148,
+      "grad_norm": 0.484375,
+      "learning_rate": 7.75729239050619e-05,
+      "loss": 1.2839,
+      "step": 30740
+    },
+    {
+      "epoch": 0.6149,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.753890786692972e-05,
+      "loss": 1.2725,
+      "step": 30745
+    },
+    {
+      "epoch": 0.615,
+      "grad_norm": 0.212890625,
+      "learning_rate": 7.750489456561352e-05,
+      "loss": 1.2657,
+      "step": 30750
+    },
+    {
+      "epoch": 0.6151,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 7.747088400525766e-05,
+      "loss": 1.2663,
+      "step": 30755
+    },
+    {
+      "epoch": 0.6152,
+      "grad_norm": 0.376953125,
+      "learning_rate": 7.743687619000626e-05,
+      "loss": 1.2789,
+      "step": 30760
+    },
+    {
+      "epoch": 0.6153,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.740287112400303e-05,
+      "loss": 1.2612,
+      "step": 30765
+    },
+    {
+      "epoch": 0.6154,
+      "grad_norm": 0.4765625,
+      "learning_rate": 7.736886881139142e-05,
+      "loss": 1.2679,
+      "step": 30770
+    },
+    {
+      "epoch": 0.6155,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.733486925631447e-05,
+      "loss": 1.2782,
+      "step": 30775
+    },
+    {
+      "epoch": 0.6156,
+      "grad_norm": 0.37109375,
+      "learning_rate": 7.730087246291502e-05,
+      "loss": 1.2618,
+      "step": 30780
+    },
+    {
+      "epoch": 0.6157,
+      "grad_norm": 0.255859375,
+      "learning_rate": 7.726687843533538e-05,
+      "loss": 1.2723,
+      "step": 30785
+    },
+    {
+      "epoch": 0.6158,
+      "grad_norm": 0.2109375,
+      "learning_rate": 7.723288717771761e-05,
+      "loss": 1.2746,
+      "step": 30790
+    },
+    {
+      "epoch": 0.6159,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.719889869420353e-05,
+      "loss": 1.28,
+      "step": 30795
+    },
+    {
+      "epoch": 0.616,
+      "grad_norm": 0.37109375,
+      "learning_rate": 7.716491298893442e-05,
+      "loss": 1.2758,
+      "step": 30800
+    },
+    {
+      "epoch": 0.6161,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.713093006605145e-05,
+      "loss": 1.2816,
+      "step": 30805
+    },
+    {
+      "epoch": 0.6162,
+      "grad_norm": 0.388671875,
+      "learning_rate": 7.709694992969526e-05,
+      "loss": 1.2609,
+      "step": 30810
+    },
+    {
+      "epoch": 0.6163,
+      "grad_norm": 0.40625,
+      "learning_rate": 7.706297258400624e-05,
+      "loss": 1.2587,
+      "step": 30815
+    },
+    {
+      "epoch": 0.6164,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 7.702899803312443e-05,
+      "loss": 1.2735,
+      "step": 30820
+    },
+    {
+      "epoch": 0.6165,
+      "grad_norm": 0.294921875,
+      "learning_rate": 7.699502628118958e-05,
+      "loss": 1.2619,
+      "step": 30825
+    },
+    {
+      "epoch": 0.6166,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 7.696105733234098e-05,
+      "loss": 1.2728,
+      "step": 30830
+    },
+    {
+      "epoch": 0.6167,
+      "grad_norm": 0.158203125,
+      "learning_rate": 7.692709119071762e-05,
+      "loss": 1.262,
+      "step": 30835
+    },
+    {
+      "epoch": 0.6168,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 7.689312786045823e-05,
+      "loss": 1.2727,
+      "step": 30840
+    },
+    {
+      "epoch": 0.6169,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.685916734570112e-05,
+      "loss": 1.2632,
+      "step": 30845
+    },
+    {
+      "epoch": 0.617,
+      "grad_norm": 0.193359375,
+      "learning_rate": 7.682520965058428e-05,
+      "loss": 1.2685,
+      "step": 30850
+    },
+    {
+      "epoch": 0.6171,
+      "grad_norm": 0.283203125,
+      "learning_rate": 7.679125477924534e-05,
+      "loss": 1.2695,
+      "step": 30855
+    },
+    {
+      "epoch": 0.6172,
+      "grad_norm": 0.390625,
+      "learning_rate": 7.67573027358216e-05,
+      "loss": 1.2821,
+      "step": 30860
+    },
+    {
+      "epoch": 0.6173,
+      "grad_norm": 0.75,
+      "learning_rate": 7.672335352445002e-05,
+      "loss": 1.2562,
+      "step": 30865
+    },
+    {
+      "epoch": 0.6174,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 7.668940714926725e-05,
+      "loss": 1.2769,
+      "step": 30870
+    },
+    {
+      "epoch": 0.6175,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 7.66554636144095e-05,
+      "loss": 1.2921,
+      "step": 30875
+    },
+    {
+      "epoch": 0.6176,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 7.662152292401264e-05,
+      "loss": 1.2763,
+      "step": 30880
+    },
+    {
+      "epoch": 0.6177,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.658758508221234e-05,
+      "loss": 1.2923,
+      "step": 30885
+    },
+    {
+      "epoch": 0.6178,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 7.655365009314374e-05,
+      "loss": 1.2679,
+      "step": 30890
+    },
+    {
+      "epoch": 0.6179,
+      "grad_norm": 0.3203125,
+      "learning_rate": 7.651971796094183e-05,
+      "loss": 1.2697,
+      "step": 30895
+    },
+    {
+      "epoch": 0.618,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.6485788689741e-05,
+      "loss": 1.2655,
+      "step": 30900
+    },
+    {
+      "epoch": 0.6181,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.645186228367554e-05,
+      "loss": 1.2812,
+      "step": 30905
+    },
+    {
+      "epoch": 0.6182,
+      "grad_norm": 0.283203125,
+      "learning_rate": 7.641793874687918e-05,
+      "loss": 1.2777,
+      "step": 30910
+    },
+    {
+      "epoch": 0.6183,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.638401808348548e-05,
+      "loss": 1.285,
+      "step": 30915
+    },
+    {
+      "epoch": 0.6184,
+      "grad_norm": 0.435546875,
+      "learning_rate": 7.635010029762756e-05,
+      "loss": 1.2931,
+      "step": 30920
+    },
+    {
+      "epoch": 0.6185,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.631618539343814e-05,
+      "loss": 1.2757,
+      "step": 30925
+    },
+    {
+      "epoch": 0.6186,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 7.628227337504972e-05,
+      "loss": 1.2677,
+      "step": 30930
+    },
+    {
+      "epoch": 0.6187,
+      "grad_norm": 0.275390625,
+      "learning_rate": 7.62483642465943e-05,
+      "loss": 1.2905,
+      "step": 30935
+    },
+    {
+      "epoch": 0.6188,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.621445801220371e-05,
+      "loss": 1.269,
+      "step": 30940
+    },
+    {
+      "epoch": 0.6189,
+      "grad_norm": 0.146484375,
+      "learning_rate": 7.618055467600922e-05,
+      "loss": 1.2693,
+      "step": 30945
+    },
+    {
+      "epoch": 0.619,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 7.614665424214193e-05,
+      "loss": 1.2873,
+      "step": 30950
+    },
+    {
+      "epoch": 0.6191,
+      "grad_norm": 0.44140625,
+      "learning_rate": 7.611275671473245e-05,
+      "loss": 1.2877,
+      "step": 30955
+    },
+    {
+      "epoch": 0.6192,
+      "grad_norm": 0.416015625,
+      "learning_rate": 7.607886209791107e-05,
+      "loss": 1.2781,
+      "step": 30960
+    },
+    {
+      "epoch": 0.6193,
+      "grad_norm": 0.298828125,
+      "learning_rate": 7.604497039580785e-05,
+      "loss": 1.2702,
+      "step": 30965
+    },
+    {
+      "epoch": 0.6194,
+      "grad_norm": 0.875,
+      "learning_rate": 7.601108161255226e-05,
+      "loss": 1.2722,
+      "step": 30970
+    },
+    {
+      "epoch": 0.6195,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.597719575227364e-05,
+      "loss": 1.273,
+      "step": 30975
+    },
+    {
+      "epoch": 0.6196,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 7.594331281910082e-05,
+      "loss": 1.2799,
+      "step": 30980
+    },
+    {
+      "epoch": 0.6197,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.590943281716241e-05,
+      "loss": 1.2873,
+      "step": 30985
+    },
+    {
+      "epoch": 0.6198,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 7.587555575058649e-05,
+      "loss": 1.2807,
+      "step": 30990
+    },
+    {
+      "epoch": 0.6199,
+      "grad_norm": 0.283203125,
+      "learning_rate": 7.584168162350098e-05,
+      "loss": 1.2611,
+      "step": 30995
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.47265625,
+      "learning_rate": 7.580781044003324e-05,
+      "loss": 1.2933,
+      "step": 31000
+    },
+    {
+      "epoch": 0.6201,
+      "grad_norm": 0.38671875,
+      "learning_rate": 7.577394220431042e-05,
+      "loss": 1.2611,
+      "step": 31005
+    },
+    {
+      "epoch": 0.6202,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 7.574007692045928e-05,
+      "loss": 1.2622,
+      "step": 31010
+    },
+    {
+      "epoch": 0.6203,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.570621459260615e-05,
+      "loss": 1.2741,
+      "step": 31015
+    },
+    {
+      "epoch": 0.6204,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.567235522487712e-05,
+      "loss": 1.2813,
+      "step": 31020
+    },
+    {
+      "epoch": 0.6205,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 7.563849882139776e-05,
+      "loss": 1.2748,
+      "step": 31025
+    },
+    {
+      "epoch": 0.6206,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.560464538629344e-05,
+      "loss": 1.2729,
+      "step": 31030
+    },
+    {
+      "epoch": 0.6207,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.557079492368909e-05,
+      "loss": 1.2787,
+      "step": 31035
+    },
+    {
+      "epoch": 0.6208,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 7.553694743770928e-05,
+      "loss": 1.2706,
+      "step": 31040
+    },
+    {
+      "epoch": 0.6209,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.550310293247823e-05,
+      "loss": 1.2698,
+      "step": 31045
+    },
+    {
+      "epoch": 0.621,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.546926141211974e-05,
+      "loss": 1.2788,
+      "step": 31050
+    },
+    {
+      "epoch": 0.6211,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.543542288075739e-05,
+      "loss": 1.2689,
+      "step": 31055
+    },
+    {
+      "epoch": 0.6212,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.54015873425142e-05,
+      "loss": 1.2846,
+      "step": 31060
+    },
+    {
+      "epoch": 0.6213,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.536775480151303e-05,
+      "loss": 1.2879,
+      "step": 31065
+    },
+    {
+      "epoch": 0.6214,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.533392526187617e-05,
+      "loss": 1.2844,
+      "step": 31070
+    },
+    {
+      "epoch": 0.6215,
+      "grad_norm": 0.13671875,
+      "learning_rate": 7.530009872772572e-05,
+      "loss": 1.2733,
+      "step": 31075
+    },
+    {
+      "epoch": 0.6216,
+      "grad_norm": 0.185546875,
+      "learning_rate": 7.526627520318329e-05,
+      "loss": 1.2747,
+      "step": 31080
+    },
+    {
+      "epoch": 0.6217,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.523245469237026e-05,
+      "loss": 1.2726,
+      "step": 31085
+    },
+    {
+      "epoch": 0.6218,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.519863719940748e-05,
+      "loss": 1.2654,
+      "step": 31090
+    },
+    {
+      "epoch": 0.6219,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.516482272841549e-05,
+      "loss": 1.2794,
+      "step": 31095
+    },
+    {
+      "epoch": 0.622,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.513101128351454e-05,
+      "loss": 1.2857,
+      "step": 31100
+    },
+    {
+      "epoch": 0.6221,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.50972028688244e-05,
+      "loss": 1.2736,
+      "step": 31105
+    },
+    {
+      "epoch": 0.6222,
+      "grad_norm": 0.24609375,
+      "learning_rate": 7.506339748846461e-05,
+      "loss": 1.2817,
+      "step": 31110
+    },
+    {
+      "epoch": 0.6223,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.502959514655414e-05,
+      "loss": 1.2702,
+      "step": 31115
+    },
+    {
+      "epoch": 0.6224,
+      "grad_norm": 0.255859375,
+      "learning_rate": 7.499579584721179e-05,
+      "loss": 1.2855,
+      "step": 31120
+    },
+    {
+      "epoch": 0.6225,
+      "grad_norm": 0.31640625,
+      "learning_rate": 7.496199959455584e-05,
+      "loss": 1.2857,
+      "step": 31125
+    },
+    {
+      "epoch": 0.6226,
+      "grad_norm": 0.25,
+      "learning_rate": 7.492820639270434e-05,
+      "loss": 1.273,
+      "step": 31130
+    },
+    {
+      "epoch": 0.6227,
+      "grad_norm": 0.396484375,
+      "learning_rate": 7.489441624577485e-05,
+      "loss": 1.2766,
+      "step": 31135
+    },
+    {
+      "epoch": 0.6228,
+      "grad_norm": 0.49609375,
+      "learning_rate": 7.486062915788452e-05,
+      "loss": 1.2727,
+      "step": 31140
+    },
+    {
+      "epoch": 0.6229,
+      "grad_norm": 0.21875,
+      "learning_rate": 7.48268451331503e-05,
+      "loss": 1.2731,
+      "step": 31145
+    },
+    {
+      "epoch": 0.623,
+      "grad_norm": 0.265625,
+      "learning_rate": 7.479306417568864e-05,
+      "loss": 1.2647,
+      "step": 31150
+    },
+    {
+      "epoch": 0.6231,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.475928628961566e-05,
+      "loss": 1.2716,
+      "step": 31155
+    },
+    {
+      "epoch": 0.6232,
+      "grad_norm": 0.390625,
+      "learning_rate": 7.472551147904708e-05,
+      "loss": 1.2637,
+      "step": 31160
+    },
+    {
+      "epoch": 0.6233,
+      "grad_norm": 0.466796875,
+      "learning_rate": 7.469173974809826e-05,
+      "loss": 1.2729,
+      "step": 31165
+    },
+    {
+      "epoch": 0.6234,
+      "grad_norm": 0.19921875,
+      "learning_rate": 7.465797110088417e-05,
+      "loss": 1.2567,
+      "step": 31170
+    },
+    {
+      "epoch": 0.6235,
+      "grad_norm": 0.3515625,
+      "learning_rate": 7.462420554151944e-05,
+      "loss": 1.2711,
+      "step": 31175
+    },
+    {
+      "epoch": 0.6236,
+      "grad_norm": 0.322265625,
+      "learning_rate": 7.459044307411832e-05,
+      "loss": 1.2757,
+      "step": 31180
+    },
+    {
+      "epoch": 0.6237,
+      "grad_norm": 0.330078125,
+      "learning_rate": 7.45566837027946e-05,
+      "loss": 1.2734,
+      "step": 31185
+    },
+    {
+      "epoch": 0.6238,
+      "grad_norm": 0.25390625,
+      "learning_rate": 7.45229274316618e-05,
+      "loss": 1.2595,
+      "step": 31190
+    },
+    {
+      "epoch": 0.6239,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.448917426483299e-05,
+      "loss": 1.2736,
+      "step": 31195
+    },
+    {
+      "epoch": 0.624,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.445542420642097e-05,
+      "loss": 1.2874,
+      "step": 31200
+    },
+    {
+      "epoch": 0.6241,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.442167726053797e-05,
+      "loss": 1.2986,
+      "step": 31205
+    },
+    {
+      "epoch": 0.6242,
+      "grad_norm": 0.3515625,
+      "learning_rate": 7.438793343129605e-05,
+      "loss": 1.2914,
+      "step": 31210
+    },
+    {
+      "epoch": 0.6243,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.435419272280672e-05,
+      "loss": 1.2664,
+      "step": 31215
+    },
+    {
+      "epoch": 0.6244,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.432045513918122e-05,
+      "loss": 1.2595,
+      "step": 31220
+    },
+    {
+      "epoch": 0.6245,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.42867206845304e-05,
+      "loss": 1.2649,
+      "step": 31225
+    },
+    {
+      "epoch": 0.6246,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.425298936296463e-05,
+      "loss": 1.273,
+      "step": 31230
+    },
+    {
+      "epoch": 0.6247,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 7.421926117859403e-05,
+      "loss": 1.2693,
+      "step": 31235
+    },
+    {
+      "epoch": 0.6248,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 7.418553613552825e-05,
+      "loss": 1.2652,
+      "step": 31240
+    },
+    {
+      "epoch": 0.6249,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 7.415181423787659e-05,
+      "loss": 1.2818,
+      "step": 31245
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.411809548974792e-05,
+      "loss": 1.2627,
+      "step": 31250
+    },
+    {
+      "epoch": 0.6251,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 7.408437989525085e-05,
+      "loss": 1.2607,
+      "step": 31255
+    },
+    {
+      "epoch": 0.6252,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.405066745849346e-05,
+      "loss": 1.2844,
+      "step": 31260
+    },
+    {
+      "epoch": 0.6253,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.401695818358353e-05,
+      "loss": 1.2714,
+      "step": 31265
+    },
+    {
+      "epoch": 0.6254,
+      "grad_norm": 0.490234375,
+      "learning_rate": 7.398325207462846e-05,
+      "loss": 1.2645,
+      "step": 31270
+    },
+    {
+      "epoch": 0.6255,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.394954913573517e-05,
+      "loss": 1.2763,
+      "step": 31275
+    },
+    {
+      "epoch": 0.6256,
+      "grad_norm": 0.357421875,
+      "learning_rate": 7.391584937101033e-05,
+      "loss": 1.2768,
+      "step": 31280
+    },
+    {
+      "epoch": 0.6257,
+      "grad_norm": 0.357421875,
+      "learning_rate": 7.38821527845601e-05,
+      "loss": 1.2687,
+      "step": 31285
+    },
+    {
+      "epoch": 0.6258,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.384845938049031e-05,
+      "loss": 1.2732,
+      "step": 31290
+    },
+    {
+      "epoch": 0.6259,
+      "grad_norm": 0.27734375,
+      "learning_rate": 7.381476916290644e-05,
+      "loss": 1.2804,
+      "step": 31295
+    },
+    {
+      "epoch": 0.626,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 7.378108213591355e-05,
+      "loss": 1.2675,
+      "step": 31300
+    },
+    {
+      "epoch": 0.6261,
+      "grad_norm": 0.28515625,
+      "learning_rate": 7.374739830361621e-05,
+      "loss": 1.2812,
+      "step": 31305
+    },
+    {
+      "epoch": 0.6262,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.37137176701188e-05,
+      "loss": 1.2687,
+      "step": 31310
+    },
+    {
+      "epoch": 0.6263,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.368004023952517e-05,
+      "loss": 1.2977,
+      "step": 31315
+    },
+    {
+      "epoch": 0.6264,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 7.364636601593875e-05,
+      "loss": 1.2725,
+      "step": 31320
+    },
+    {
+      "epoch": 0.6265,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.361269500346274e-05,
+      "loss": 1.2695,
+      "step": 31325
+    },
+    {
+      "epoch": 0.6266,
+      "grad_norm": 0.4453125,
+      "learning_rate": 7.357902720619976e-05,
+      "loss": 1.2744,
+      "step": 31330
+    },
+    {
+      "epoch": 0.6267,
+      "grad_norm": 0.310546875,
+      "learning_rate": 7.354536262825219e-05,
+      "loss": 1.2718,
+      "step": 31335
+    },
+    {
+      "epoch": 0.6268,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.351170127372191e-05,
+      "loss": 1.2814,
+      "step": 31340
+    },
+    {
+      "epoch": 0.6269,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.347804314671055e-05,
+      "loss": 1.2838,
+      "step": 31345
+    },
+    {
+      "epoch": 0.627,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.344438825131911e-05,
+      "loss": 1.2747,
+      "step": 31350
+    },
+    {
+      "epoch": 0.6271,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.341073659164848e-05,
+      "loss": 1.2631,
+      "step": 31355
+    },
+    {
+      "epoch": 0.6272,
+      "grad_norm": 0.310546875,
+      "learning_rate": 7.33770881717989e-05,
+      "loss": 1.261,
+      "step": 31360
+    },
+    {
+      "epoch": 0.6273,
+      "grad_norm": 0.251953125,
+      "learning_rate": 7.334344299587035e-05,
+      "loss": 1.2672,
+      "step": 31365
+    },
+    {
+      "epoch": 0.6274,
+      "grad_norm": 0.125,
+      "learning_rate": 7.330980106796246e-05,
+      "loss": 1.261,
+      "step": 31370
+    },
+    {
+      "epoch": 0.6275,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.327616239217431e-05,
+      "loss": 1.2717,
+      "step": 31375
+    },
+    {
+      "epoch": 0.6276,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 7.324252697260474e-05,
+      "loss": 1.2769,
+      "step": 31380
+    },
+    {
+      "epoch": 0.6277,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.320889481335207e-05,
+      "loss": 1.2649,
+      "step": 31385
+    },
+    {
+      "epoch": 0.6278,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 7.317526591851433e-05,
+      "loss": 1.2733,
+      "step": 31390
+    },
+    {
+      "epoch": 0.6279,
+      "grad_norm": 0.2734375,
+      "learning_rate": 7.314164029218904e-05,
+      "loss": 1.2792,
+      "step": 31395
+    },
+    {
+      "epoch": 0.628,
+      "grad_norm": 0.20703125,
+      "learning_rate": 7.310801793847344e-05,
+      "loss": 1.2828,
+      "step": 31400
+    },
+    {
+      "epoch": 0.6281,
+      "grad_norm": 0.330078125,
+      "learning_rate": 7.307439886146428e-05,
+      "loss": 1.2897,
+      "step": 31405
+    },
+    {
+      "epoch": 0.6282,
+      "grad_norm": 0.26953125,
+      "learning_rate": 7.30407830652579e-05,
+      "loss": 1.2712,
+      "step": 31410
+    },
+    {
+      "epoch": 0.6283,
+      "grad_norm": 0.4140625,
+      "learning_rate": 7.300717055395039e-05,
+      "loss": 1.2816,
+      "step": 31415
+    },
+    {
+      "epoch": 0.6284,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.297356133163721e-05,
+      "loss": 1.2878,
+      "step": 31420
+    },
+    {
+      "epoch": 0.6285,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.293995540241366e-05,
+      "loss": 1.2739,
+      "step": 31425
+    },
+    {
+      "epoch": 0.6286,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.290635277037442e-05,
+      "loss": 1.2907,
+      "step": 31430
+    },
+    {
+      "epoch": 0.6287,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.287275343961392e-05,
+      "loss": 1.2777,
+      "step": 31435
+    },
+    {
+      "epoch": 0.6288,
+      "grad_norm": 0.28515625,
+      "learning_rate": 7.283915741422612e-05,
+      "loss": 1.257,
+      "step": 31440
+    },
+    {
+      "epoch": 0.6289,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.280556469830464e-05,
+      "loss": 1.2678,
+      "step": 31445
+    },
+    {
+      "epoch": 0.629,
+      "grad_norm": 0.41796875,
+      "learning_rate": 7.277197529594257e-05,
+      "loss": 1.2691,
+      "step": 31450
+    },
+    {
+      "epoch": 0.6291,
+      "grad_norm": 0.2890625,
+      "learning_rate": 7.273838921123272e-05,
+      "loss": 1.2788,
+      "step": 31455
+    },
+    {
+      "epoch": 0.6292,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.270480644826749e-05,
+      "loss": 1.2698,
+      "step": 31460
+    },
+    {
+      "epoch": 0.6293,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 7.267122701113876e-05,
+      "loss": 1.2657,
+      "step": 31465
+    },
+    {
+      "epoch": 0.6294,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 7.263765090393817e-05,
+      "loss": 1.267,
+      "step": 31470
+    },
+    {
+      "epoch": 0.6295,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 7.260407813075676e-05,
+      "loss": 1.2785,
+      "step": 31475
+    },
+    {
+      "epoch": 0.6296,
+      "grad_norm": 0.296875,
+      "learning_rate": 7.257050869568535e-05,
+      "loss": 1.2783,
+      "step": 31480
+    },
+    {
+      "epoch": 0.6297,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.253694260281425e-05,
+      "loss": 1.2701,
+      "step": 31485
+    },
+    {
+      "epoch": 0.6298,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.250337985623342e-05,
+      "loss": 1.2871,
+      "step": 31490
+    },
+    {
+      "epoch": 0.6299,
+      "grad_norm": 0.36328125,
+      "learning_rate": 7.246982046003234e-05,
+      "loss": 1.2854,
+      "step": 31495
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.243626441830009e-05,
+      "loss": 1.2752,
+      "step": 31500
+    },
+    {
+      "epoch": 0.6301,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 7.240271173512546e-05,
+      "loss": 1.287,
+      "step": 31505
+    },
+    {
+      "epoch": 0.6302,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.236916241459663e-05,
+      "loss": 1.2761,
+      "step": 31510
+    },
+    {
+      "epoch": 0.6303,
+      "grad_norm": 0.255859375,
+      "learning_rate": 7.233561646080161e-05,
+      "loss": 1.2727,
+      "step": 31515
+    },
+    {
+      "epoch": 0.6304,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 7.230207387782776e-05,
+      "loss": 1.278,
+      "step": 31520
+    },
+    {
+      "epoch": 0.6305,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.226853466976222e-05,
+      "loss": 1.2849,
+      "step": 31525
+    },
+    {
+      "epoch": 0.6306,
+      "grad_norm": 0.451171875,
+      "learning_rate": 7.22349988406916e-05,
+      "loss": 1.2811,
+      "step": 31530
+    },
+    {
+      "epoch": 0.6307,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 7.220146639470218e-05,
+      "loss": 1.2766,
+      "step": 31535
+    },
+    {
+      "epoch": 0.6308,
+      "grad_norm": 0.30078125,
+      "learning_rate": 7.216793733587976e-05,
+      "loss": 1.2766,
+      "step": 31540
+    },
+    {
+      "epoch": 0.6309,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.21344116683097e-05,
+      "loss": 1.2888,
+      "step": 31545
+    },
+    {
+      "epoch": 0.631,
+      "grad_norm": 0.490234375,
+      "learning_rate": 7.210088939607708e-05,
+      "loss": 1.277,
+      "step": 31550
+    },
+    {
+      "epoch": 0.6311,
+      "grad_norm": 0.462890625,
+      "learning_rate": 7.206737052326645e-05,
+      "loss": 1.271,
+      "step": 31555
+    },
+    {
+      "epoch": 0.6312,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.203385505396203e-05,
+      "loss": 1.269,
+      "step": 31560
+    },
+    {
+      "epoch": 0.6313,
+      "grad_norm": 0.234375,
+      "learning_rate": 7.20003429922475e-05,
+      "loss": 1.2552,
+      "step": 31565
+    },
+    {
+      "epoch": 0.6314,
+      "grad_norm": 0.43359375,
+      "learning_rate": 7.196683434220625e-05,
+      "loss": 1.2868,
+      "step": 31570
+    },
+    {
+      "epoch": 0.6315,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.193332910792124e-05,
+      "loss": 1.2963,
+      "step": 31575
+    },
+    {
+      "epoch": 0.6316,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.18998272934749e-05,
+      "loss": 1.2747,
+      "step": 31580
+    },
+    {
+      "epoch": 0.6317,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 7.186632890294941e-05,
+      "loss": 1.2761,
+      "step": 31585
+    },
+    {
+      "epoch": 0.6318,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.183283394042634e-05,
+      "loss": 1.2776,
+      "step": 31590
+    },
+    {
+      "epoch": 0.6319,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 7.179934240998706e-05,
+      "loss": 1.278,
+      "step": 31595
+    },
+    {
+      "epoch": 0.632,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.176585431571235e-05,
+      "loss": 1.2907,
+      "step": 31600
+    },
+    {
+      "epoch": 0.6321,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 7.173236966168268e-05,
+      "loss": 1.2713,
+      "step": 31605
+    },
+    {
+      "epoch": 0.6322,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.169888845197798e-05,
+      "loss": 1.2728,
+      "step": 31610
+    },
+    {
+      "epoch": 0.6323,
+      "grad_norm": 0.466796875,
+      "learning_rate": 7.166541069067792e-05,
+      "loss": 1.2765,
+      "step": 31615
+    },
+    {
+      "epoch": 0.6324,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.163193638186158e-05,
+      "loss": 1.2857,
+      "step": 31620
+    },
+    {
+      "epoch": 0.6325,
+      "grad_norm": 0.21875,
+      "learning_rate": 7.159846552960774e-05,
+      "loss": 1.2658,
+      "step": 31625
+    },
+    {
+      "epoch": 0.6326,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.156499813799476e-05,
+      "loss": 1.2606,
+      "step": 31630
+    },
+    {
+      "epoch": 0.6327,
+      "grad_norm": 0.275390625,
+      "learning_rate": 7.153153421110048e-05,
+      "loss": 1.2697,
+      "step": 31635
+    },
+    {
+      "epoch": 0.6328,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.149807375300239e-05,
+      "loss": 1.2749,
+      "step": 31640
+    },
+    {
+      "epoch": 0.6329,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 7.146461676777756e-05,
+      "loss": 1.271,
+      "step": 31645
+    },
+    {
+      "epoch": 0.633,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 7.143116325950265e-05,
+      "loss": 1.2737,
+      "step": 31650
+    },
+    {
+      "epoch": 0.6331,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.139771323225381e-05,
+      "loss": 1.2693,
+      "step": 31655
+    },
+    {
+      "epoch": 0.6332,
+      "grad_norm": 0.298828125,
+      "learning_rate": 7.136426669010689e-05,
+      "loss": 1.2726,
+      "step": 31660
+    },
+    {
+      "epoch": 0.6333,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.13308236371372e-05,
+      "loss": 1.2737,
+      "step": 31665
+    },
+    {
+      "epoch": 0.6334,
+      "grad_norm": 0.263671875,
+      "learning_rate": 7.129738407741964e-05,
+      "loss": 1.269,
+      "step": 31670
+    },
+    {
+      "epoch": 0.6335,
+      "grad_norm": 0.265625,
+      "learning_rate": 7.126394801502882e-05,
+      "loss": 1.2694,
+      "step": 31675
+    },
+    {
+      "epoch": 0.6336,
+      "grad_norm": 0.478515625,
+      "learning_rate": 7.123051545403874e-05,
+      "loss": 1.2698,
+      "step": 31680
+    },
+    {
+      "epoch": 0.6337,
+      "grad_norm": 0.234375,
+      "learning_rate": 7.119708639852312e-05,
+      "loss": 1.2654,
+      "step": 31685
+    },
+    {
+      "epoch": 0.6338,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 7.11636608525551e-05,
+      "loss": 1.2932,
+      "step": 31690
+    },
+    {
+      "epoch": 0.6339,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.113023882020757e-05,
+      "loss": 1.2646,
+      "step": 31695
+    },
+    {
+      "epoch": 0.634,
+      "grad_norm": 0.3828125,
+      "learning_rate": 7.109682030555283e-05,
+      "loss": 1.2729,
+      "step": 31700
+    },
+    {
+      "epoch": 0.6341,
+      "grad_norm": 0.29296875,
+      "learning_rate": 7.106340531266292e-05,
+      "loss": 1.2831,
+      "step": 31705
+    },
+    {
+      "epoch": 0.6342,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.102999384560927e-05,
+      "loss": 1.283,
+      "step": 31710
+    },
+    {
+      "epoch": 0.6343,
+      "grad_norm": 0.39453125,
+      "learning_rate": 7.099658590846299e-05,
+      "loss": 1.2825,
+      "step": 31715
+    },
+    {
+      "epoch": 0.6344,
+      "grad_norm": 0.21875,
+      "learning_rate": 7.096318150529477e-05,
+      "loss": 1.2808,
+      "step": 31720
+    },
+    {
+      "epoch": 0.6345,
+      "grad_norm": 0.287109375,
+      "learning_rate": 7.092978064017475e-05,
+      "loss": 1.2952,
+      "step": 31725
+    },
+    {
+      "epoch": 0.6346,
+      "grad_norm": 0.162109375,
+      "learning_rate": 7.089638331717284e-05,
+      "loss": 1.2592,
+      "step": 31730
+    },
+    {
+      "epoch": 0.6347,
+      "grad_norm": 0.5,
+      "learning_rate": 7.08629895403583e-05,
+      "loss": 1.2638,
+      "step": 31735
+    },
+    {
+      "epoch": 0.6348,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 7.082959931380011e-05,
+      "loss": 1.2717,
+      "step": 31740
+    },
+    {
+      "epoch": 0.6349,
+      "grad_norm": 0.193359375,
+      "learning_rate": 7.079621264156675e-05,
+      "loss": 1.2752,
+      "step": 31745
+    },
+    {
+      "epoch": 0.635,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 7.076282952772633e-05,
+      "loss": 1.2701,
+      "step": 31750
+    },
+    {
+      "epoch": 0.6351,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.072944997634646e-05,
+      "loss": 1.2606,
+      "step": 31755
+    },
+    {
+      "epoch": 0.6352,
+      "grad_norm": 0.482421875,
+      "learning_rate": 7.069607399149428e-05,
+      "loss": 1.2695,
+      "step": 31760
+    },
+    {
+      "epoch": 0.6353,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.06627015772366e-05,
+      "loss": 1.2704,
+      "step": 31765
+    },
+    {
+      "epoch": 0.6354,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.062933273763975e-05,
+      "loss": 1.2908,
+      "step": 31770
+    },
+    {
+      "epoch": 0.6355,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.059596747676962e-05,
+      "loss": 1.2792,
+      "step": 31775
+    },
+    {
+      "epoch": 0.6356,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.056260579869165e-05,
+      "loss": 1.2786,
+      "step": 31780
+    },
+    {
+      "epoch": 0.6357,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.052924770747087e-05,
+      "loss": 1.2804,
+      "step": 31785
+    },
+    {
+      "epoch": 0.6358,
+      "grad_norm": 0.474609375,
+      "learning_rate": 7.049589320717186e-05,
+      "loss": 1.2774,
+      "step": 31790
+    },
+    {
+      "epoch": 0.6359,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.04625423018588e-05,
+      "loss": 1.2724,
+      "step": 31795
+    },
+    {
+      "epoch": 0.636,
+      "grad_norm": 0.224609375,
+      "learning_rate": 7.042919499559537e-05,
+      "loss": 1.2816,
+      "step": 31800
+    },
+    {
+      "epoch": 0.6361,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.039585129244477e-05,
+      "loss": 1.2749,
+      "step": 31805
+    },
+    {
+      "epoch": 0.6362,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.036251119646992e-05,
+      "loss": 1.269,
+      "step": 31810
+    },
+    {
+      "epoch": 0.6363,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.032917471173318e-05,
+      "loss": 1.2725,
+      "step": 31815
+    },
+    {
+      "epoch": 0.6364,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.029584184229653e-05,
+      "loss": 1.2774,
+      "step": 31820
+    },
+    {
+      "epoch": 0.6365,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 7.026251259222141e-05,
+      "loss": 1.2743,
+      "step": 31825
+    },
+    {
+      "epoch": 0.6366,
+      "grad_norm": 0.29296875,
+      "learning_rate": 7.022918696556896e-05,
+      "loss": 1.288,
+      "step": 31830
+    },
+    {
+      "epoch": 0.6367,
+      "grad_norm": 0.46875,
+      "learning_rate": 7.019586496639974e-05,
+      "loss": 1.2677,
+      "step": 31835
+    },
+    {
+      "epoch": 0.6368,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 7.016254659877398e-05,
+      "loss": 1.2761,
+      "step": 31840
+    },
+    {
+      "epoch": 0.6369,
+      "grad_norm": 0.296875,
+      "learning_rate": 7.012923186675144e-05,
+      "loss": 1.2759,
+      "step": 31845
+    },
+    {
+      "epoch": 0.637,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.009592077439134e-05,
+      "loss": 1.2722,
+      "step": 31850
+    },
+    {
+      "epoch": 0.6371,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.00626133257526e-05,
+      "loss": 1.2763,
+      "step": 31855
+    },
+    {
+      "epoch": 0.6372,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 7.002930952489362e-05,
+      "loss": 1.2673,
+      "step": 31860
+    },
+    {
+      "epoch": 0.6373,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 6.999600937587239e-05,
+      "loss": 1.2697,
+      "step": 31865
+    },
+    {
+      "epoch": 0.6374,
+      "grad_norm": 0.369140625,
+      "learning_rate": 6.996271288274636e-05,
+      "loss": 1.2762,
+      "step": 31870
+    },
+    {
+      "epoch": 0.6375,
+      "grad_norm": 0.16015625,
+      "learning_rate": 6.992942004957271e-05,
+      "loss": 1.2823,
+      "step": 31875
+    },
+    {
+      "epoch": 0.6376,
+      "grad_norm": 0.259765625,
+      "learning_rate": 6.989613088040796e-05,
+      "loss": 1.2649,
+      "step": 31880
+    },
+    {
+      "epoch": 0.6377,
+      "grad_norm": 0.28515625,
+      "learning_rate": 6.986284537930838e-05,
+      "loss": 1.2685,
+      "step": 31885
+    },
+    {
+      "epoch": 0.6378,
+      "grad_norm": 0.197265625,
+      "learning_rate": 6.982956355032968e-05,
+      "loss": 1.288,
+      "step": 31890
+    },
+    {
+      "epoch": 0.6379,
+      "grad_norm": 0.1640625,
+      "learning_rate": 6.979628539752711e-05,
+      "loss": 1.2709,
+      "step": 31895
+    },
+    {
+      "epoch": 0.638,
+      "grad_norm": 0.451171875,
+      "learning_rate": 6.976301092495556e-05,
+      "loss": 1.2778,
+      "step": 31900
+    },
+    {
+      "epoch": 0.6381,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.972974013666942e-05,
+      "loss": 1.289,
+      "step": 31905
+    },
+    {
+      "epoch": 0.6382,
+      "grad_norm": 0.275390625,
+      "learning_rate": 6.969647303672262e-05,
+      "loss": 1.2799,
+      "step": 31910
+    },
+    {
+      "epoch": 0.6383,
+      "grad_norm": 0.19921875,
+      "learning_rate": 6.966320962916864e-05,
+      "loss": 1.2658,
+      "step": 31915
+    },
+    {
+      "epoch": 0.6384,
+      "grad_norm": 0.19921875,
+      "learning_rate": 6.962994991806059e-05,
+      "loss": 1.2655,
+      "step": 31920
+    },
+    {
+      "epoch": 0.6385,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.959669390745097e-05,
+      "loss": 1.2681,
+      "step": 31925
+    },
+    {
+      "epoch": 0.6386,
+      "grad_norm": 0.16015625,
+      "learning_rate": 6.956344160139201e-05,
+      "loss": 1.2903,
+      "step": 31930
+    },
+    {
+      "epoch": 0.6387,
+      "grad_norm": 0.201171875,
+      "learning_rate": 6.953019300393538e-05,
+      "loss": 1.2945,
+      "step": 31935
+    },
+    {
+      "epoch": 0.6388,
+      "grad_norm": 0.271484375,
+      "learning_rate": 6.949694811913225e-05,
+      "loss": 1.2802,
+      "step": 31940
+    },
+    {
+      "epoch": 0.6389,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.946370695103353e-05,
+      "loss": 1.2731,
+      "step": 31945
+    },
+    {
+      "epoch": 0.639,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.943046950368944e-05,
+      "loss": 1.2894,
+      "step": 31950
+    },
+    {
+      "epoch": 0.6391,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.939723578114993e-05,
+      "loss": 1.2602,
+      "step": 31955
+    },
+    {
+      "epoch": 0.6392,
+      "grad_norm": 0.44140625,
+      "learning_rate": 6.93640057874644e-05,
+      "loss": 1.2633,
+      "step": 31960
+    },
+    {
+      "epoch": 0.6393,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.93307795266819e-05,
+      "loss": 1.2746,
+      "step": 31965
+    },
+    {
+      "epoch": 0.6394,
+      "grad_norm": 0.275390625,
+      "learning_rate": 6.929755700285081e-05,
+      "loss": 1.2609,
+      "step": 31970
+    },
+    {
+      "epoch": 0.6395,
+      "grad_norm": 0.376953125,
+      "learning_rate": 6.92643382200193e-05,
+      "loss": 1.2744,
+      "step": 31975
+    },
+    {
+      "epoch": 0.6396,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.923112318223496e-05,
+      "loss": 1.2736,
+      "step": 31980
+    },
+    {
+      "epoch": 0.6397,
+      "grad_norm": 0.326171875,
+      "learning_rate": 6.91979118935449e-05,
+      "loss": 1.2628,
+      "step": 31985
+    },
+    {
+      "epoch": 0.6398,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 6.916470435799587e-05,
+      "loss": 1.2829,
+      "step": 31990
+    },
+    {
+      "epoch": 0.6399,
+      "grad_norm": 0.380859375,
+      "learning_rate": 6.913150057963404e-05,
+      "loss": 1.2763,
+      "step": 31995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 6.909830056250527e-05,
+      "loss": 1.2694,
+      "step": 32000
+    },
+    {
+      "epoch": 0.6401,
+      "grad_norm": 0.271484375,
+      "learning_rate": 6.90651043106548e-05,
+      "loss": 1.2773,
+      "step": 32005
+    },
+    {
+      "epoch": 0.6402,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.90319118281276e-05,
+      "loss": 1.2706,
+      "step": 32010
+    },
+    {
+      "epoch": 0.6403,
+      "grad_norm": 0.2265625,
+      "learning_rate": 6.899872311896795e-05,
+      "loss": 1.2654,
+      "step": 32015
+    },
+    {
+      "epoch": 0.6404,
+      "grad_norm": 0.28515625,
+      "learning_rate": 6.896553818721989e-05,
+      "loss": 1.2764,
+      "step": 32020
+    },
+    {
+      "epoch": 0.6405,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.893235703692685e-05,
+      "loss": 1.2906,
+      "step": 32025
+    },
+    {
+      "epoch": 0.6406,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.889917967213185e-05,
+      "loss": 1.2723,
+      "step": 32030
+    },
+    {
+      "epoch": 0.6407,
+      "grad_norm": 0.49609375,
+      "learning_rate": 6.88660060968775e-05,
+      "loss": 1.27,
+      "step": 32035
+    },
+    {
+      "epoch": 0.6408,
+      "grad_norm": 0.283203125,
+      "learning_rate": 6.883283631520582e-05,
+      "loss": 1.2724,
+      "step": 32040
+    },
+    {
+      "epoch": 0.6409,
+      "grad_norm": 0.80078125,
+      "learning_rate": 6.879967033115853e-05,
+      "loss": 1.2728,
+      "step": 32045
+    },
+    {
+      "epoch": 0.641,
+      "grad_norm": 0.1484375,
+      "learning_rate": 6.876650814877674e-05,
+      "loss": 1.2905,
+      "step": 32050
+    },
+    {
+      "epoch": 0.6411,
+      "grad_norm": 0.27734375,
+      "learning_rate": 6.873334977210122e-05,
+      "loss": 1.2809,
+      "step": 32055
+    },
+    {
+      "epoch": 0.6412,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 6.870019520517217e-05,
+      "loss": 1.2656,
+      "step": 32060
+    },
+    {
+      "epoch": 0.6413,
+      "grad_norm": 0.310546875,
+      "learning_rate": 6.866704445202943e-05,
+      "loss": 1.267,
+      "step": 32065
+    },
+    {
+      "epoch": 0.6414,
+      "grad_norm": 0.279296875,
+      "learning_rate": 6.863389751671225e-05,
+      "loss": 1.3025,
+      "step": 32070
+    },
+    {
+      "epoch": 0.6415,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.860075440325951e-05,
+      "loss": 1.2817,
+      "step": 32075
+    },
+    {
+      "epoch": 0.6416,
+      "grad_norm": 0.322265625,
+      "learning_rate": 6.856761511570963e-05,
+      "loss": 1.2915,
+      "step": 32080
+    },
+    {
+      "epoch": 0.6417,
+      "grad_norm": 0.796875,
+      "learning_rate": 6.853447965810046e-05,
+      "loss": 1.2826,
+      "step": 32085
+    },
+    {
+      "epoch": 0.6418,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.850134803446954e-05,
+      "loss": 1.2718,
+      "step": 32090
+    },
+    {
+      "epoch": 0.6419,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 6.846822024885379e-05,
+      "loss": 1.2904,
+      "step": 32095
+    },
+    {
+      "epoch": 0.642,
+      "grad_norm": 0.49609375,
+      "learning_rate": 6.843509630528977e-05,
+      "loss": 1.283,
+      "step": 32100
+    },
+    {
+      "epoch": 0.6421,
+      "grad_norm": 0.14453125,
+      "learning_rate": 6.840197620781349e-05,
+      "loss": 1.2646,
+      "step": 32105
+    },
+    {
+      "epoch": 0.6422,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 6.836885996046061e-05,
+      "loss": 1.271,
+      "step": 32110
+    },
+    {
+      "epoch": 0.6423,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.833574756726618e-05,
+      "loss": 1.274,
+      "step": 32115
+    },
+    {
+      "epoch": 0.6424,
+      "grad_norm": 0.35546875,
+      "learning_rate": 6.830263903226483e-05,
+      "loss": 1.2777,
+      "step": 32120
+    },
+    {
+      "epoch": 0.6425,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.826953435949081e-05,
+      "loss": 1.2684,
+      "step": 32125
+    },
+    {
+      "epoch": 0.6426,
+      "grad_norm": 0.208984375,
+      "learning_rate": 6.823643355297773e-05,
+      "loss": 1.2776,
+      "step": 32130
+    },
+    {
+      "epoch": 0.6427,
+      "grad_norm": 0.205078125,
+      "learning_rate": 6.820333661675893e-05,
+      "loss": 1.26,
+      "step": 32135
+    },
+    {
+      "epoch": 0.6428,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 6.817024355486706e-05,
+      "loss": 1.2718,
+      "step": 32140
+    },
+    {
+      "epoch": 0.6429,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 6.81371543713345e-05,
+      "loss": 1.2703,
+      "step": 32145
+    },
+    {
+      "epoch": 0.643,
+      "grad_norm": 0.7890625,
+      "learning_rate": 6.8104069070193e-05,
+      "loss": 1.2774,
+      "step": 32150
+    },
+    {
+      "epoch": 0.6431,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.807098765547397e-05,
+      "loss": 1.2829,
+      "step": 32155
+    },
+    {
+      "epoch": 0.6432,
+      "grad_norm": 0.322265625,
+      "learning_rate": 6.803791013120822e-05,
+      "loss": 1.2744,
+      "step": 32160
+    },
+    {
+      "epoch": 0.6433,
+      "grad_norm": 0.197265625,
+      "learning_rate": 6.800483650142617e-05,
+      "loss": 1.28,
+      "step": 32165
+    },
+    {
+      "epoch": 0.6434,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.797176677015775e-05,
+      "loss": 1.2817,
+      "step": 32170
+    },
+    {
+      "epoch": 0.6435,
+      "grad_norm": 0.453125,
+      "learning_rate": 6.793870094143238e-05,
+      "loss": 1.2645,
+      "step": 32175
+    },
+    {
+      "epoch": 0.6436,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.790563901927907e-05,
+      "loss": 1.2604,
+      "step": 32180
+    },
+    {
+      "epoch": 0.6437,
+      "grad_norm": 0.2734375,
+      "learning_rate": 6.787258100772627e-05,
+      "loss": 1.2794,
+      "step": 32185
+    },
+    {
+      "epoch": 0.6438,
+      "grad_norm": 0.404296875,
+      "learning_rate": 6.783952691080203e-05,
+      "loss": 1.2804,
+      "step": 32190
+    },
+    {
+      "epoch": 0.6439,
+      "grad_norm": 0.30859375,
+      "learning_rate": 6.780647673253391e-05,
+      "loss": 1.2625,
+      "step": 32195
+    },
+    {
+      "epoch": 0.644,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.77734304769489e-05,
+      "loss": 1.2613,
+      "step": 32200
+    },
+    {
+      "epoch": 0.6441,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.774038814807369e-05,
+      "loss": 1.2641,
+      "step": 32205
+    },
+    {
+      "epoch": 0.6442,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 6.770734974993426e-05,
+      "loss": 1.2768,
+      "step": 32210
+    },
+    {
+      "epoch": 0.6443,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 6.767431528655635e-05,
+      "loss": 1.2659,
+      "step": 32215
+    },
+    {
+      "epoch": 0.6444,
+      "grad_norm": 0.369140625,
+      "learning_rate": 6.764128476196505e-05,
+      "loss": 1.2683,
+      "step": 32220
+    },
+    {
+      "epoch": 0.6445,
+      "grad_norm": 0.30078125,
+      "learning_rate": 6.760825818018508e-05,
+      "loss": 1.2656,
+      "step": 32225
+    },
+    {
+      "epoch": 0.6446,
+      "grad_norm": 0.271484375,
+      "learning_rate": 6.757523554524056e-05,
+      "loss": 1.2854,
+      "step": 32230
+    },
+    {
+      "epoch": 0.6447,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.754221686115525e-05,
+      "loss": 1.2826,
+      "step": 32235
+    },
+    {
+      "epoch": 0.6448,
+      "grad_norm": 0.46484375,
+      "learning_rate": 6.750920213195238e-05,
+      "loss": 1.2677,
+      "step": 32240
+    },
+    {
+      "epoch": 0.6449,
+      "grad_norm": 0.296875,
+      "learning_rate": 6.747619136165463e-05,
+      "loss": 1.2663,
+      "step": 32245
+    },
+    {
+      "epoch": 0.645,
+      "grad_norm": 0.447265625,
+      "learning_rate": 6.744318455428436e-05,
+      "loss": 1.2758,
+      "step": 32250
+    },
+    {
+      "epoch": 0.6451,
+      "grad_norm": 0.37109375,
+      "learning_rate": 6.741018171386326e-05,
+      "loss": 1.2698,
+      "step": 32255
+    },
+    {
+      "epoch": 0.6452,
+      "grad_norm": 0.22265625,
+      "learning_rate": 6.737718284441267e-05,
+      "loss": 1.2668,
+      "step": 32260
+    },
+    {
+      "epoch": 0.6453,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.734418794995337e-05,
+      "loss": 1.2853,
+      "step": 32265
+    },
+    {
+      "epoch": 0.6454,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.731119703450577e-05,
+      "loss": 1.2971,
+      "step": 32270
+    },
+    {
+      "epoch": 0.6455,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 6.727821010208962e-05,
+      "loss": 1.2712,
+      "step": 32275
+    },
+    {
+      "epoch": 0.6456,
+      "grad_norm": 0.80078125,
+      "learning_rate": 6.724522715672432e-05,
+      "loss": 1.2661,
+      "step": 32280
+    },
+    {
+      "epoch": 0.6457,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 6.721224820242875e-05,
+      "loss": 1.2663,
+      "step": 32285
+    },
+    {
+      "epoch": 0.6458,
+      "grad_norm": 0.28515625,
+      "learning_rate": 6.717927324322124e-05,
+      "loss": 1.2664,
+      "step": 32290
+    },
+    {
+      "epoch": 0.6459,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.714630228311978e-05,
+      "loss": 1.2679,
+      "step": 32295
+    },
+    {
+      "epoch": 0.646,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.711333532614168e-05,
+      "loss": 1.2596,
+      "step": 32300
+    },
+    {
+      "epoch": 0.6461,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.708037237630395e-05,
+      "loss": 1.2873,
+      "step": 32305
+    },
+    {
+      "epoch": 0.6462,
+      "grad_norm": 0.298828125,
+      "learning_rate": 6.704741343762295e-05,
+      "loss": 1.2838,
+      "step": 32310
+    },
+    {
+      "epoch": 0.6463,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.701445851411472e-05,
+      "loss": 1.2923,
+      "step": 32315
+    },
+    {
+      "epoch": 0.6464,
+      "grad_norm": 0.248046875,
+      "learning_rate": 6.698150760979463e-05,
+      "loss": 1.2756,
+      "step": 32320
+    },
+    {
+      "epoch": 0.6465,
+      "grad_norm": 0.41796875,
+      "learning_rate": 6.69485607286777e-05,
+      "loss": 1.2694,
+      "step": 32325
+    },
+    {
+      "epoch": 0.6466,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.69156178747784e-05,
+      "loss": 1.2797,
+      "step": 32330
+    },
+    {
+      "epoch": 0.6467,
+      "grad_norm": 0.15625,
+      "learning_rate": 6.688267905211067e-05,
+      "loss": 1.2658,
+      "step": 32335
+    },
+    {
+      "epoch": 0.6468,
+      "grad_norm": 0.23828125,
+      "learning_rate": 6.684974426468808e-05,
+      "loss": 1.2883,
+      "step": 32340
+    },
+    {
+      "epoch": 0.6469,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.681681351652356e-05,
+      "loss": 1.2628,
+      "step": 32345
+    },
+    {
+      "epoch": 0.647,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.67838868116297e-05,
+      "loss": 1.2757,
+      "step": 32350
+    },
+    {
+      "epoch": 0.6471,
+      "grad_norm": 0.28125,
+      "learning_rate": 6.675096415401842e-05,
+      "loss": 1.2681,
+      "step": 32355
+    },
+    {
+      "epoch": 0.6472,
+      "grad_norm": 0.42578125,
+      "learning_rate": 6.671804554770135e-05,
+      "loss": 1.2638,
+      "step": 32360
+    },
+    {
+      "epoch": 0.6473,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.668513099668944e-05,
+      "loss": 1.2854,
+      "step": 32365
+    },
+    {
+      "epoch": 0.6474,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.66522205049933e-05,
+      "loss": 1.2705,
+      "step": 32370
+    },
+    {
+      "epoch": 0.6475,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.661931407662292e-05,
+      "loss": 1.2775,
+      "step": 32375
+    },
+    {
+      "epoch": 0.6476,
+      "grad_norm": 0.302734375,
+      "learning_rate": 6.658641171558785e-05,
+      "loss": 1.2749,
+      "step": 32380
+    },
+    {
+      "epoch": 0.6477,
+      "grad_norm": 0.197265625,
+      "learning_rate": 6.65535134258972e-05,
+      "loss": 1.2712,
+      "step": 32385
+    },
+    {
+      "epoch": 0.6478,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.652061921155943e-05,
+      "loss": 1.2941,
+      "step": 32390
+    },
+    {
+      "epoch": 0.6479,
+      "grad_norm": 0.27734375,
+      "learning_rate": 6.648772907658272e-05,
+      "loss": 1.2762,
+      "step": 32395
+    },
+    {
+      "epoch": 0.648,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.64548430249745e-05,
+      "loss": 1.273,
+      "step": 32400
+    },
+    {
+      "epoch": 0.6481,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.642196106074194e-05,
+      "loss": 1.2702,
+      "step": 32405
+    },
+    {
+      "epoch": 0.6482,
+      "grad_norm": 0.22265625,
+      "learning_rate": 6.638908318789156e-05,
+      "loss": 1.275,
+      "step": 32410
+    },
+    {
+      "epoch": 0.6483,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.635620941042945e-05,
+      "loss": 1.2821,
+      "step": 32415
+    },
+    {
+      "epoch": 0.6484,
+      "grad_norm": 0.28515625,
+      "learning_rate": 6.63233397323612e-05,
+      "loss": 1.2756,
+      "step": 32420
+    },
+    {
+      "epoch": 0.6485,
+      "grad_norm": 0.87890625,
+      "learning_rate": 6.62904741576918e-05,
+      "loss": 1.2992,
+      "step": 32425
+    },
+    {
+      "epoch": 0.6486,
+      "grad_norm": 0.359375,
+      "learning_rate": 6.62576126904259e-05,
+      "loss": 1.2688,
+      "step": 32430
+    },
+    {
+      "epoch": 0.6487,
+      "grad_norm": 0.470703125,
+      "learning_rate": 6.622475533456751e-05,
+      "loss": 1.281,
+      "step": 32435
+    },
+    {
+      "epoch": 0.6488,
+      "grad_norm": 0.15625,
+      "learning_rate": 6.61919020941203e-05,
+      "loss": 1.2744,
+      "step": 32440
+    },
+    {
+      "epoch": 0.6489,
+      "grad_norm": 0.322265625,
+      "learning_rate": 6.61590529730872e-05,
+      "loss": 1.2816,
+      "step": 32445
+    },
+    {
+      "epoch": 0.649,
+      "grad_norm": 0.298828125,
+      "learning_rate": 6.612620797547087e-05,
+      "loss": 1.2761,
+      "step": 32450
+    },
+    {
+      "epoch": 0.6491,
+      "grad_norm": 0.3203125,
+      "learning_rate": 6.609336710527332e-05,
+      "loss": 1.3014,
+      "step": 32455
+    },
+    {
+      "epoch": 0.6492,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.606053036649619e-05,
+      "loss": 1.2634,
+      "step": 32460
+    },
+    {
+      "epoch": 0.6493,
+      "grad_norm": 0.3671875,
+      "learning_rate": 6.602769776314049e-05,
+      "loss": 1.288,
+      "step": 32465
+    },
+    {
+      "epoch": 0.6494,
+      "grad_norm": 0.3671875,
+      "learning_rate": 6.599486929920673e-05,
+      "loss": 1.2627,
+      "step": 32470
+    },
+    {
+      "epoch": 0.6495,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 6.5962044978695e-05,
+      "loss": 1.2841,
+      "step": 32475
+    },
+    {
+      "epoch": 0.6496,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.592922480560483e-05,
+      "loss": 1.2851,
+      "step": 32480
+    },
+    {
+      "epoch": 0.6497,
+      "grad_norm": 0.98046875,
+      "learning_rate": 6.589640878393531e-05,
+      "loss": 1.2666,
+      "step": 32485
+    },
+    {
+      "epoch": 0.6498,
+      "grad_norm": 0.25390625,
+      "learning_rate": 6.58635969176849e-05,
+      "loss": 1.2954,
+      "step": 32490
+    },
+    {
+      "epoch": 0.6499,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.583078921085167e-05,
+      "loss": 1.2742,
+      "step": 32495
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.921875,
+      "learning_rate": 6.579798566743314e-05,
+      "loss": 1.2806,
+      "step": 32500
+    },
+    {
+      "epoch": 0.6501,
+      "grad_norm": 0.275390625,
+      "learning_rate": 6.57651862914263e-05,
+      "loss": 1.2702,
+      "step": 32505
+    },
+    {
+      "epoch": 0.6502,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 6.573239108682768e-05,
+      "loss": 1.281,
+      "step": 32510
+    },
+    {
+      "epoch": 0.6503,
+      "grad_norm": 0.40234375,
+      "learning_rate": 6.569960005763323e-05,
+      "loss": 1.2537,
+      "step": 32515
+    },
+    {
+      "epoch": 0.6504,
+      "grad_norm": 0.431640625,
+      "learning_rate": 6.566681320783849e-05,
+      "loss": 1.2823,
+      "step": 32520
+    },
+    {
+      "epoch": 0.6505,
+      "grad_norm": 0.46875,
+      "learning_rate": 6.56340305414384e-05,
+      "loss": 1.2763,
+      "step": 32525
+    },
+    {
+      "epoch": 0.6506,
+      "grad_norm": 0.16015625,
+      "learning_rate": 6.560125206242746e-05,
+      "loss": 1.2696,
+      "step": 32530
+    },
+    {
+      "epoch": 0.6507,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.55684777747996e-05,
+      "loss": 1.2747,
+      "step": 32535
+    },
+    {
+      "epoch": 0.6508,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.55357076825483e-05,
+      "loss": 1.2739,
+      "step": 32540
+    },
+    {
+      "epoch": 0.6509,
+      "grad_norm": 0.8671875,
+      "learning_rate": 6.550294178966647e-05,
+      "loss": 1.3022,
+      "step": 32545
+    },
+    {
+      "epoch": 0.651,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.547018010014654e-05,
+      "loss": 1.2794,
+      "step": 32550
+    },
+    {
+      "epoch": 0.6511,
+      "grad_norm": 0.287109375,
+      "learning_rate": 6.543742261798045e-05,
+      "loss": 1.2683,
+      "step": 32555
+    },
+    {
+      "epoch": 0.6512,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 6.540466934715953e-05,
+      "loss": 1.2812,
+      "step": 32560
+    },
+    {
+      "epoch": 0.6513,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.537192029167474e-05,
+      "loss": 1.2554,
+      "step": 32565
+    },
+    {
+      "epoch": 0.6514,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.53391754555164e-05,
+      "loss": 1.2816,
+      "step": 32570
+    },
+    {
+      "epoch": 0.6515,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.530643484267443e-05,
+      "loss": 1.2872,
+      "step": 32575
+    },
+    {
+      "epoch": 0.6516,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 6.52736984571381e-05,
+      "loss": 1.2768,
+      "step": 32580
+    },
+    {
+      "epoch": 0.6517,
+      "grad_norm": 0.498046875,
+      "learning_rate": 6.52409663028963e-05,
+      "loss": 1.278,
+      "step": 32585
+    },
+    {
+      "epoch": 0.6518,
+      "grad_norm": 0.75,
+      "learning_rate": 6.520823838393731e-05,
+      "loss": 1.2682,
+      "step": 32590
+    },
+    {
+      "epoch": 0.6519,
+      "grad_norm": 0.208984375,
+      "learning_rate": 6.517551470424895e-05,
+      "loss": 1.2712,
+      "step": 32595
+    },
+    {
+      "epoch": 0.652,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.51427952678185e-05,
+      "loss": 1.2701,
+      "step": 32600
+    },
+    {
+      "epoch": 0.6521,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 6.511008007863268e-05,
+      "loss": 1.2774,
+      "step": 32605
+    },
+    {
+      "epoch": 0.6522,
+      "grad_norm": 0.359375,
+      "learning_rate": 6.507736914067781e-05,
+      "loss": 1.2842,
+      "step": 32610
+    },
+    {
+      "epoch": 0.6523,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.504466245793955e-05,
+      "loss": 1.2664,
+      "step": 32615
+    },
+    {
+      "epoch": 0.6524,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.501196003440314e-05,
+      "loss": 1.2817,
+      "step": 32620
+    },
+    {
+      "epoch": 0.6525,
+      "grad_norm": 0.2109375,
+      "learning_rate": 6.497926187405326e-05,
+      "loss": 1.2742,
+      "step": 32625
+    },
+    {
+      "epoch": 0.6526,
+      "grad_norm": 0.478515625,
+      "learning_rate": 6.494656798087412e-05,
+      "loss": 1.2658,
+      "step": 32630
+    },
+    {
+      "epoch": 0.6527,
+      "grad_norm": 0.13671875,
+      "learning_rate": 6.49138783588493e-05,
+      "loss": 1.2651,
+      "step": 32635
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.488119301196201e-05,
+      "loss": 1.2741,
+      "step": 32640
+    },
+    {
+      "epoch": 0.6529,
+      "grad_norm": 0.115234375,
+      "learning_rate": 6.484851194419484e-05,
+      "loss": 1.2617,
+      "step": 32645
+    },
+    {
+      "epoch": 0.653,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 6.481583515952983e-05,
+      "loss": 1.2657,
+      "step": 32650
+    },
+    {
+      "epoch": 0.6531,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 6.478316266194861e-05,
+      "loss": 1.2779,
+      "step": 32655
+    },
+    {
+      "epoch": 0.6532,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 6.475049445543215e-05,
+      "loss": 1.2833,
+      "step": 32660
+    },
+    {
+      "epoch": 0.6533,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.471783054396105e-05,
+      "loss": 1.2589,
+      "step": 32665
+    },
+    {
+      "epoch": 0.6534,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.468517093151525e-05,
+      "loss": 1.2736,
+      "step": 32670
+    },
+    {
+      "epoch": 0.6535,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.465251562207431e-05,
+      "loss": 1.2762,
+      "step": 32675
+    },
+    {
+      "epoch": 0.6536,
+      "grad_norm": 0.306640625,
+      "learning_rate": 6.461986461961706e-05,
+      "loss": 1.258,
+      "step": 32680
+    },
+    {
+      "epoch": 0.6537,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.458721792812204e-05,
+      "loss": 1.2897,
+      "step": 32685
+    },
+    {
+      "epoch": 0.6538,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.455457555156705e-05,
+      "loss": 1.2723,
+      "step": 32690
+    },
+    {
+      "epoch": 0.6539,
+      "grad_norm": 0.267578125,
+      "learning_rate": 6.452193749392952e-05,
+      "loss": 1.278,
+      "step": 32695
+    },
+    {
+      "epoch": 0.654,
+      "grad_norm": 0.421875,
+      "learning_rate": 6.448930375918631e-05,
+      "loss": 1.2743,
+      "step": 32700
+    },
+    {
+      "epoch": 0.6541,
+      "grad_norm": 0.388671875,
+      "learning_rate": 6.44566743513137e-05,
+      "loss": 1.2642,
+      "step": 32705
+    },
+    {
+      "epoch": 0.6542,
+      "grad_norm": 0.193359375,
+      "learning_rate": 6.442404927428751e-05,
+      "loss": 1.2725,
+      "step": 32710
+    },
+    {
+      "epoch": 0.6543,
+      "grad_norm": 0.43359375,
+      "learning_rate": 6.4391428532083e-05,
+      "loss": 1.2818,
+      "step": 32715
+    },
+    {
+      "epoch": 0.6544,
+      "grad_norm": 0.421875,
+      "learning_rate": 6.435881212867493e-05,
+      "loss": 1.272,
+      "step": 32720
+    },
+    {
+      "epoch": 0.6545,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.432620006803746e-05,
+      "loss": 1.2771,
+      "step": 32725
+    },
+    {
+      "epoch": 0.6546,
+      "grad_norm": 0.439453125,
+      "learning_rate": 6.429359235414436e-05,
+      "loss": 1.2692,
+      "step": 32730
+    },
+    {
+      "epoch": 0.6547,
+      "grad_norm": 0.232421875,
+      "learning_rate": 6.426098899096868e-05,
+      "loss": 1.2782,
+      "step": 32735
+    },
+    {
+      "epoch": 0.6548,
+      "grad_norm": 0.201171875,
+      "learning_rate": 6.422838998248307e-05,
+      "loss": 1.2606,
+      "step": 32740
+    },
+    {
+      "epoch": 0.6549,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 6.419579533265968e-05,
+      "loss": 1.2635,
+      "step": 32745
+    },
+    {
+      "epoch": 0.655,
+      "grad_norm": 0.15234375,
+      "learning_rate": 6.416320504546997e-05,
+      "loss": 1.2786,
+      "step": 32750
+    },
+    {
+      "epoch": 0.6551,
+      "grad_norm": 0.4609375,
+      "learning_rate": 6.413061912488508e-05,
+      "loss": 1.2646,
+      "step": 32755
+    },
+    {
+      "epoch": 0.6552,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 6.409803757487538e-05,
+      "loss": 1.2565,
+      "step": 32760
+    },
+    {
+      "epoch": 0.6553,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.406546039941094e-05,
+      "loss": 1.2643,
+      "step": 32765
+    },
+    {
+      "epoch": 0.6554,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 6.403288760246112e-05,
+      "loss": 1.2752,
+      "step": 32770
+    },
+    {
+      "epoch": 0.6555,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.40003191879949e-05,
+      "loss": 1.2823,
+      "step": 32775
+    },
+    {
+      "epoch": 0.6556,
+      "grad_norm": 0.35546875,
+      "learning_rate": 6.396775515998055e-05,
+      "loss": 1.2949,
+      "step": 32780
+    },
+    {
+      "epoch": 0.6557,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 6.393519552238591e-05,
+      "loss": 1.274,
+      "step": 32785
+    },
+    {
+      "epoch": 0.6558,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 6.390264027917836e-05,
+      "loss": 1.2849,
+      "step": 32790
+    },
+    {
+      "epoch": 0.6559,
+      "grad_norm": 0.13671875,
+      "learning_rate": 6.387008943432455e-05,
+      "loss": 1.2795,
+      "step": 32795
+    },
+    {
+      "epoch": 0.656,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.383754299179079e-05,
+      "loss": 1.2654,
+      "step": 32800
+    },
+    {
+      "epoch": 0.6561,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.380500095554268e-05,
+      "loss": 1.2623,
+      "step": 32805
+    },
+    {
+      "epoch": 0.6562,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 6.377246332954544e-05,
+      "loss": 1.2751,
+      "step": 32810
+    },
+    {
+      "epoch": 0.6563,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.373993011776367e-05,
+      "loss": 1.2677,
+      "step": 32815
+    },
+    {
+      "epoch": 0.6564,
+      "grad_norm": 0.310546875,
+      "learning_rate": 6.370740132416138e-05,
+      "loss": 1.2755,
+      "step": 32820
+    },
+    {
+      "epoch": 0.6565,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.367487695270217e-05,
+      "loss": 1.2617,
+      "step": 32825
+    },
+    {
+      "epoch": 0.6566,
+      "grad_norm": 0.427734375,
+      "learning_rate": 6.364235700734903e-05,
+      "loss": 1.2527,
+      "step": 32830
+    },
+    {
+      "epoch": 0.6567,
+      "grad_norm": 0.291015625,
+      "learning_rate": 6.360984149206439e-05,
+      "loss": 1.282,
+      "step": 32835
+    },
+    {
+      "epoch": 0.6568,
+      "grad_norm": 0.98046875,
+      "learning_rate": 6.357733041081018e-05,
+      "loss": 1.2799,
+      "step": 32840
+    },
+    {
+      "epoch": 0.6569,
+      "grad_norm": 0.31640625,
+      "learning_rate": 6.35448237675478e-05,
+      "loss": 1.2913,
+      "step": 32845
+    },
+    {
+      "epoch": 0.657,
+      "grad_norm": 0.23828125,
+      "learning_rate": 6.351232156623803e-05,
+      "loss": 1.311,
+      "step": 32850
+    },
+    {
+      "epoch": 0.6571,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 6.347982381084123e-05,
+      "loss": 1.2794,
+      "step": 32855
+    },
+    {
+      "epoch": 0.6572,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.344733050531713e-05,
+      "loss": 1.2771,
+      "step": 32860
+    },
+    {
+      "epoch": 0.6573,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 6.341484165362487e-05,
+      "loss": 1.2768,
+      "step": 32865
+    },
+    {
+      "epoch": 0.6574,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.338235725972325e-05,
+      "loss": 1.2737,
+      "step": 32870
+    },
+    {
+      "epoch": 0.6575,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.334987732757029e-05,
+      "loss": 1.2582,
+      "step": 32875
+    },
+    {
+      "epoch": 0.6576,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.33174018611236e-05,
+      "loss": 1.2833,
+      "step": 32880
+    },
+    {
+      "epoch": 0.6577,
+      "grad_norm": 0.435546875,
+      "learning_rate": 6.328493086434023e-05,
+      "loss": 1.2673,
+      "step": 32885
+    },
+    {
+      "epoch": 0.6578,
+      "grad_norm": 0.263671875,
+      "learning_rate": 6.325246434117668e-05,
+      "loss": 1.2758,
+      "step": 32890
+    },
+    {
+      "epoch": 0.6579,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.322000229558887e-05,
+      "loss": 1.2694,
+      "step": 32895
+    },
+    {
+      "epoch": 0.658,
+      "grad_norm": 0.291015625,
+      "learning_rate": 6.318754473153221e-05,
+      "loss": 1.2681,
+      "step": 32900
+    },
+    {
+      "epoch": 0.6581,
+      "grad_norm": 0.279296875,
+      "learning_rate": 6.315509165296159e-05,
+      "loss": 1.2877,
+      "step": 32905
+    },
+    {
+      "epoch": 0.6582,
+      "grad_norm": 0.265625,
+      "learning_rate": 6.312264306383124e-05,
+      "loss": 1.276,
+      "step": 32910
+    },
+    {
+      "epoch": 0.6583,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.309019896809503e-05,
+      "loss": 1.2986,
+      "step": 32915
+    },
+    {
+      "epoch": 0.6584,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.305775936970606e-05,
+      "loss": 1.2882,
+      "step": 32920
+    },
+    {
+      "epoch": 0.6585,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.302532427261709e-05,
+      "loss": 1.2721,
+      "step": 32925
+    },
+    {
+      "epoch": 0.6586,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.299289368078016e-05,
+      "loss": 1.2861,
+      "step": 32930
+    },
+    {
+      "epoch": 0.6587,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 6.296046759814694e-05,
+      "loss": 1.2789,
+      "step": 32935
+    },
+    {
+      "epoch": 0.6588,
+      "grad_norm": 0.185546875,
+      "learning_rate": 6.292804602866834e-05,
+      "loss": 1.2771,
+      "step": 32940
+    },
+    {
+      "epoch": 0.6589,
+      "grad_norm": 0.306640625,
+      "learning_rate": 6.289562897629492e-05,
+      "loss": 1.2715,
+      "step": 32945
+    },
+    {
+      "epoch": 0.659,
+      "grad_norm": 0.25390625,
+      "learning_rate": 6.286321644497655e-05,
+      "loss": 1.2838,
+      "step": 32950
+    },
+    {
+      "epoch": 0.6591,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 6.283080843866256e-05,
+      "loss": 1.2729,
+      "step": 32955
+    },
+    {
+      "epoch": 0.6592,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.27984049613019e-05,
+      "loss": 1.2808,
+      "step": 32960
+    },
+    {
+      "epoch": 0.6593,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 6.276600601684267e-05,
+      "loss": 1.2822,
+      "step": 32965
+    },
+    {
+      "epoch": 0.6594,
+      "grad_norm": 0.154296875,
+      "learning_rate": 6.27336116092327e-05,
+      "loss": 1.2835,
+      "step": 32970
+    },
+    {
+      "epoch": 0.6595,
+      "grad_norm": 0.47265625,
+      "learning_rate": 6.27012217424191e-05,
+      "loss": 1.2685,
+      "step": 32975
+    },
+    {
+      "epoch": 0.6596,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 6.266883642034853e-05,
+      "loss": 1.2945,
+      "step": 32980
+    },
+    {
+      "epoch": 0.6597,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 6.263645564696696e-05,
+      "loss": 1.2634,
+      "step": 32985
+    },
+    {
+      "epoch": 0.6598,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.260407942621998e-05,
+      "loss": 1.2711,
+      "step": 32990
+    },
+    {
+      "epoch": 0.6599,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 6.257170776205245e-05,
+      "loss": 1.2596,
+      "step": 32995
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.25393406584088e-05,
+      "loss": 1.2771,
+      "step": 33000
+    },
+    {
+      "epoch": 0.6601,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 6.25069781192329e-05,
+      "loss": 1.2612,
+      "step": 33005
+    },
+    {
+      "epoch": 0.6602,
+      "grad_norm": 0.412109375,
+      "learning_rate": 6.247462014846792e-05,
+      "loss": 1.2578,
+      "step": 33010
+    },
+    {
+      "epoch": 0.6603,
+      "grad_norm": 0.3046875,
+      "learning_rate": 6.24422667500567e-05,
+      "loss": 1.2753,
+      "step": 33015
+    },
+    {
+      "epoch": 0.6604,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.240991792794133e-05,
+      "loss": 1.2857,
+      "step": 33020
+    },
+    {
+      "epoch": 0.6605,
+      "grad_norm": 0.13671875,
+      "learning_rate": 6.237757368606345e-05,
+      "loss": 1.2688,
+      "step": 33025
+    },
+    {
+      "epoch": 0.6606,
+      "grad_norm": 0.302734375,
+      "learning_rate": 6.234523402836407e-05,
+      "loss": 1.2708,
+      "step": 33030
+    },
+    {
+      "epoch": 0.6607,
+      "grad_norm": 0.78125,
+      "learning_rate": 6.231289895878375e-05,
+      "loss": 1.2734,
+      "step": 33035
+    },
+    {
+      "epoch": 0.6608,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 6.228056848126236e-05,
+      "loss": 1.2589,
+      "step": 33040
+    },
+    {
+      "epoch": 0.6609,
+      "grad_norm": 0.388671875,
+      "learning_rate": 6.224824259973925e-05,
+      "loss": 1.2822,
+      "step": 33045
+    },
+    {
+      "epoch": 0.661,
+      "grad_norm": 0.326171875,
+      "learning_rate": 6.22159213181533e-05,
+      "loss": 1.2773,
+      "step": 33050
+    },
+    {
+      "epoch": 0.6611,
+      "grad_norm": 0.25,
+      "learning_rate": 6.21836046404427e-05,
+      "loss": 1.2667,
+      "step": 33055
+    },
+    {
+      "epoch": 0.6612,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.215129257054522e-05,
+      "loss": 1.2873,
+      "step": 33060
+    },
+    {
+      "epoch": 0.6613,
+      "grad_norm": 0.2421875,
+      "learning_rate": 6.211898511239787e-05,
+      "loss": 1.2652,
+      "step": 33065
+    },
+    {
+      "epoch": 0.6614,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.20866822699373e-05,
+      "loss": 1.2881,
+      "step": 33070
+    },
+    {
+      "epoch": 0.6615,
+      "grad_norm": 0.423828125,
+      "learning_rate": 6.205438404709947e-05,
+      "loss": 1.274,
+      "step": 33075
+    },
+    {
+      "epoch": 0.6616,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 6.20220904478199e-05,
+      "loss": 1.284,
+      "step": 33080
+    },
+    {
+      "epoch": 0.6617,
+      "grad_norm": 0.625,
+      "learning_rate": 6.198980147603339e-05,
+      "loss": 1.2751,
+      "step": 33085
+    },
+    {
+      "epoch": 0.6618,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.195751713567427e-05,
+      "loss": 1.267,
+      "step": 33090
+    },
+    {
+      "epoch": 0.6619,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.192523743067626e-05,
+      "loss": 1.2722,
+      "step": 33095
+    },
+    {
+      "epoch": 0.662,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.18929623649726e-05,
+      "loss": 1.2763,
+      "step": 33100
+    },
+    {
+      "epoch": 0.6621,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 6.18606919424959e-05,
+      "loss": 1.2863,
+      "step": 33105
+    },
+    {
+      "epoch": 0.6622,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 6.182842616717817e-05,
+      "loss": 1.2597,
+      "step": 33110
+    },
+    {
+      "epoch": 0.6623,
+      "grad_norm": 0.232421875,
+      "learning_rate": 6.179616504295092e-05,
+      "loss": 1.2797,
+      "step": 33115
+    },
+    {
+      "epoch": 0.6624,
+      "grad_norm": 0.83203125,
+      "learning_rate": 6.176390857374507e-05,
+      "loss": 1.2617,
+      "step": 33120
+    },
+    {
+      "epoch": 0.6625,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.173165676349103e-05,
+      "loss": 1.2698,
+      "step": 33125
+    },
+    {
+      "epoch": 0.6626,
+      "grad_norm": 0.46484375,
+      "learning_rate": 6.169940961611854e-05,
+      "loss": 1.2743,
+      "step": 33130
+    },
+    {
+      "epoch": 0.6627,
+      "grad_norm": 0.314453125,
+      "learning_rate": 6.166716713555674e-05,
+      "loss": 1.2991,
+      "step": 33135
+    },
+    {
+      "epoch": 0.6628,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.163492932573438e-05,
+      "loss": 1.2662,
+      "step": 33140
+    },
+    {
+      "epoch": 0.6629,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.160269619057951e-05,
+      "loss": 1.2959,
+      "step": 33145
+    },
+    {
+      "epoch": 0.663,
+      "grad_norm": 0.37109375,
+      "learning_rate": 6.157046773401964e-05,
+      "loss": 1.2802,
+      "step": 33150
+    },
+    {
+      "epoch": 0.6631,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.153824395998168e-05,
+      "loss": 1.2649,
+      "step": 33155
+    },
+    {
+      "epoch": 0.6632,
+      "grad_norm": 0.220703125,
+      "learning_rate": 6.150602487239206e-05,
+      "loss": 1.2743,
+      "step": 33160
+    },
+    {
+      "epoch": 0.6633,
+      "grad_norm": 0.30859375,
+      "learning_rate": 6.147381047517655e-05,
+      "loss": 1.2773,
+      "step": 33165
+    },
+    {
+      "epoch": 0.6634,
+      "grad_norm": 0.451171875,
+      "learning_rate": 6.144160077226036e-05,
+      "loss": 1.2681,
+      "step": 33170
+    },
+    {
+      "epoch": 0.6635,
+      "grad_norm": 0.21875,
+      "learning_rate": 6.140939576756817e-05,
+      "loss": 1.2619,
+      "step": 33175
+    },
+    {
+      "epoch": 0.6636,
+      "grad_norm": 0.25390625,
+      "learning_rate": 6.1377195465024e-05,
+      "loss": 1.2702,
+      "step": 33180
+    },
+    {
+      "epoch": 0.6637,
+      "grad_norm": 0.205078125,
+      "learning_rate": 6.134499986855144e-05,
+      "loss": 1.2687,
+      "step": 33185
+    },
+    {
+      "epoch": 0.6638,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.131280898207339e-05,
+      "loss": 1.273,
+      "step": 33190
+    },
+    {
+      "epoch": 0.6639,
+      "grad_norm": 0.2421875,
+      "learning_rate": 6.128062280951224e-05,
+      "loss": 1.2714,
+      "step": 33195
+    },
+    {
+      "epoch": 0.664,
+      "grad_norm": 0.294921875,
+      "learning_rate": 6.12484413547897e-05,
+      "loss": 1.2748,
+      "step": 33200
+    },
+    {
+      "epoch": 0.6641,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.121626462182707e-05,
+      "loss": 1.2664,
+      "step": 33205
+    },
+    {
+      "epoch": 0.6642,
+      "grad_norm": 0.265625,
+      "learning_rate": 6.118409261454494e-05,
+      "loss": 1.2847,
+      "step": 33210
+    },
+    {
+      "epoch": 0.6643,
+      "grad_norm": 0.224609375,
+      "learning_rate": 6.11519253368634e-05,
+      "loss": 1.2825,
+      "step": 33215
+    },
+    {
+      "epoch": 0.6644,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 6.111976279270192e-05,
+      "loss": 1.2736,
+      "step": 33220
+    },
+    {
+      "epoch": 0.6645,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.108760498597938e-05,
+      "loss": 1.2875,
+      "step": 33225
+    },
+    {
+      "epoch": 0.6646,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 6.105545192061416e-05,
+      "loss": 1.2719,
+      "step": 33230
+    },
+    {
+      "epoch": 0.6647,
+      "grad_norm": 0.462890625,
+      "learning_rate": 6.1023303600523975e-05,
+      "loss": 1.2649,
+      "step": 33235
+    },
+    {
+      "epoch": 0.6648,
+      "grad_norm": 0.255859375,
+      "learning_rate": 6.099116002962604e-05,
+      "loss": 1.2641,
+      "step": 33240
+    },
+    {
+      "epoch": 0.6649,
+      "grad_norm": 0.255859375,
+      "learning_rate": 6.09590212118369e-05,
+      "loss": 1.2772,
+      "step": 33245
+    },
+    {
+      "epoch": 0.665,
+      "grad_norm": 0.380859375,
+      "learning_rate": 6.092688715107264e-05,
+      "loss": 1.2732,
+      "step": 33250
+    },
+    {
+      "epoch": 0.6651,
+      "grad_norm": 0.421875,
+      "learning_rate": 6.089475785124863e-05,
+      "loss": 1.2678,
+      "step": 33255
+    },
+    {
+      "epoch": 0.6652,
+      "grad_norm": 0.353515625,
+      "learning_rate": 6.086263331627976e-05,
+      "loss": 1.2769,
+      "step": 33260
+    },
+    {
+      "epoch": 0.6653,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.0830513550080335e-05,
+      "loss": 1.2726,
+      "step": 33265
+    },
+    {
+      "epoch": 0.6654,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.0798398556563976e-05,
+      "loss": 1.2695,
+      "step": 33270
+    },
+    {
+      "epoch": 0.6655,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.076628833964388e-05,
+      "loss": 1.2724,
+      "step": 33275
+    },
+    {
+      "epoch": 0.6656,
+      "grad_norm": 0.2890625,
+      "learning_rate": 6.073418290323251e-05,
+      "loss": 1.2817,
+      "step": 33280
+    },
+    {
+      "epoch": 0.6657,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.070208225124185e-05,
+      "loss": 1.2734,
+      "step": 33285
+    },
+    {
+      "epoch": 0.6658,
+      "grad_norm": 0.166015625,
+      "learning_rate": 6.066998638758326e-05,
+      "loss": 1.2793,
+      "step": 33290
+    },
+    {
+      "epoch": 0.6659,
+      "grad_norm": 0.16015625,
+      "learning_rate": 6.063789531616757e-05,
+      "loss": 1.2896,
+      "step": 33295
+    },
+    {
+      "epoch": 0.666,
+      "grad_norm": 0.3671875,
+      "learning_rate": 6.0605809040904894e-05,
+      "loss": 1.2645,
+      "step": 33300
+    },
+    {
+      "epoch": 0.6661,
+      "grad_norm": 0.82421875,
+      "learning_rate": 6.05737275657049e-05,
+      "loss": 1.2746,
+      "step": 33305
+    },
+    {
+      "epoch": 0.6662,
+      "grad_norm": 0.419921875,
+      "learning_rate": 6.054165089447663e-05,
+      "loss": 1.2701,
+      "step": 33310
+    },
+    {
+      "epoch": 0.6663,
+      "grad_norm": 0.162109375,
+      "learning_rate": 6.0509579031128485e-05,
+      "loss": 1.296,
+      "step": 33315
+    },
+    {
+      "epoch": 0.6664,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.047751197956838e-05,
+      "loss": 1.2722,
+      "step": 33320
+    },
+    {
+      "epoch": 0.6665,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 6.0445449743703516e-05,
+      "loss": 1.2847,
+      "step": 33325
+    },
+    {
+      "epoch": 0.6666,
+      "grad_norm": 0.193359375,
+      "learning_rate": 6.0413392327440635e-05,
+      "loss": 1.2635,
+      "step": 33330
+    },
+    {
+      "epoch": 0.6667,
+      "grad_norm": 0.7734375,
+      "learning_rate": 6.03813397346858e-05,
+      "loss": 1.2796,
+      "step": 33335
+    },
+    {
+      "epoch": 0.6668,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 6.034929196934459e-05,
+      "loss": 1.28,
+      "step": 33340
+    },
+    {
+      "epoch": 0.6669,
+      "grad_norm": 0.26953125,
+      "learning_rate": 6.031724903532183e-05,
+      "loss": 1.263,
+      "step": 33345
+    },
+    {
+      "epoch": 0.667,
+      "grad_norm": 0.142578125,
+      "learning_rate": 6.0285210936521955e-05,
+      "loss": 1.2695,
+      "step": 33350
+    },
+    {
+      "epoch": 0.6671,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.025317767684864e-05,
+      "loss": 1.2816,
+      "step": 33355
+    },
+    {
+      "epoch": 0.6672,
+      "grad_norm": 0.392578125,
+      "learning_rate": 6.022114926020504e-05,
+      "loss": 1.278,
+      "step": 33360
+    },
+    {
+      "epoch": 0.6673,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.018912569049376e-05,
+      "loss": 1.2859,
+      "step": 33365
+    },
+    {
+      "epoch": 0.6674,
+      "grad_norm": 0.357421875,
+      "learning_rate": 6.015710697161674e-05,
+      "loss": 1.2709,
+      "step": 33370
+    },
+    {
+      "epoch": 0.6675,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.012509310747538e-05,
+      "loss": 1.2781,
+      "step": 33375
+    },
+    {
+      "epoch": 0.6676,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.009308410197047e-05,
+      "loss": 1.2859,
+      "step": 33380
+    },
+    {
+      "epoch": 0.6677,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 6.006107995900224e-05,
+      "loss": 1.2834,
+      "step": 33385
+    },
+    {
+      "epoch": 0.6678,
+      "grad_norm": 0.427734375,
+      "learning_rate": 6.002908068247024e-05,
+      "loss": 1.2785,
+      "step": 33390
+    },
+    {
+      "epoch": 0.6679,
+      "grad_norm": 0.32421875,
+      "learning_rate": 5.999708627627354e-05,
+      "loss": 1.2778,
+      "step": 33395
+    },
+    {
+      "epoch": 0.668,
+      "grad_norm": 0.1796875,
+      "learning_rate": 5.9965096744310526e-05,
+      "loss": 1.2721,
+      "step": 33400
+    },
+    {
+      "epoch": 0.6681,
+      "grad_norm": 0.150390625,
+      "learning_rate": 5.9933112090479006e-05,
+      "loss": 1.2624,
+      "step": 33405
+    },
+    {
+      "epoch": 0.6682,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 5.990113231867629e-05,
+      "loss": 1.2744,
+      "step": 33410
+    },
+    {
+      "epoch": 0.6683,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.9869157432798926e-05,
+      "loss": 1.2786,
+      "step": 33415
+    },
+    {
+      "epoch": 0.6684,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.9837187436743016e-05,
+      "loss": 1.2841,
+      "step": 33420
+    },
+    {
+      "epoch": 0.6685,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.9805222334404e-05,
+      "loss": 1.2726,
+      "step": 33425
+    },
+    {
+      "epoch": 0.6686,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.977326212967671e-05,
+      "loss": 1.2878,
+      "step": 33430
+    },
+    {
+      "epoch": 0.6687,
+      "grad_norm": 0.2890625,
+      "learning_rate": 5.974130682645538e-05,
+      "loss": 1.2778,
+      "step": 33435
+    },
+    {
+      "epoch": 0.6688,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 5.9709356428633746e-05,
+      "loss": 1.2692,
+      "step": 33440
+    },
+    {
+      "epoch": 0.6689,
+      "grad_norm": 0.376953125,
+      "learning_rate": 5.967741094010479e-05,
+      "loss": 1.2781,
+      "step": 33445
+    },
+    {
+      "epoch": 0.669,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 5.964547036476099e-05,
+      "loss": 1.2889,
+      "step": 33450
+    },
+    {
+      "epoch": 0.6691,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.961353470649426e-05,
+      "loss": 1.2729,
+      "step": 33455
+    },
+    {
+      "epoch": 0.6692,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 5.9581603969195766e-05,
+      "loss": 1.2816,
+      "step": 33460
+    },
+    {
+      "epoch": 0.6693,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 5.954967815675627e-05,
+      "loss": 1.2654,
+      "step": 33465
+    },
+    {
+      "epoch": 0.6694,
+      "grad_norm": 0.8125,
+      "learning_rate": 5.951775727306577e-05,
+      "loss": 1.2703,
+      "step": 33470
+    },
+    {
+      "epoch": 0.6695,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 5.9485841322013755e-05,
+      "loss": 1.2571,
+      "step": 33475
+    },
+    {
+      "epoch": 0.6696,
+      "grad_norm": 0.470703125,
+      "learning_rate": 5.94539303074891e-05,
+      "loss": 1.2691,
+      "step": 33480
+    },
+    {
+      "epoch": 0.6697,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.9422024233380013e-05,
+      "loss": 1.271,
+      "step": 33485
+    },
+    {
+      "epoch": 0.6698,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 5.9390123103574215e-05,
+      "loss": 1.2944,
+      "step": 33490
+    },
+    {
+      "epoch": 0.6699,
+      "grad_norm": 0.271484375,
+      "learning_rate": 5.935822692195869e-05,
+      "loss": 1.2764,
+      "step": 33495
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.193359375,
+      "learning_rate": 5.9326335692419995e-05,
+      "loss": 1.2764,
+      "step": 33500
+    },
+    {
+      "epoch": 0.6701,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.929444941884388e-05,
+      "loss": 1.2669,
+      "step": 33505
+    },
+    {
+      "epoch": 0.6702,
+      "grad_norm": 0.197265625,
+      "learning_rate": 5.9262568105115654e-05,
+      "loss": 1.2891,
+      "step": 33510
+    },
+    {
+      "epoch": 0.6703,
+      "grad_norm": 1.15625,
+      "learning_rate": 5.92306917551199e-05,
+      "loss": 1.263,
+      "step": 33515
+    },
+    {
+      "epoch": 0.6704,
+      "grad_norm": 0.46875,
+      "learning_rate": 5.9198820372740726e-05,
+      "loss": 1.282,
+      "step": 33520
+    },
+    {
+      "epoch": 0.6705,
+      "grad_norm": 0.92578125,
+      "learning_rate": 5.916695396186154e-05,
+      "loss": 1.2713,
+      "step": 33525
+    },
+    {
+      "epoch": 0.6706,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.913509252636511e-05,
+      "loss": 1.2798,
+      "step": 33530
+    },
+    {
+      "epoch": 0.6707,
+      "grad_norm": 0.2109375,
+      "learning_rate": 5.910323607013373e-05,
+      "loss": 1.2708,
+      "step": 33535
+    },
+    {
+      "epoch": 0.6708,
+      "grad_norm": 0.42578125,
+      "learning_rate": 5.907138459704895e-05,
+      "loss": 1.2814,
+      "step": 33540
+    },
+    {
+      "epoch": 0.6709,
+      "grad_norm": 0.271484375,
+      "learning_rate": 5.903953811099183e-05,
+      "loss": 1.2649,
+      "step": 33545
+    },
+    {
+      "epoch": 0.671,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.900769661584272e-05,
+      "loss": 1.2694,
+      "step": 33550
+    },
+    {
+      "epoch": 0.6711,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 5.8975860115481487e-05,
+      "loss": 1.2662,
+      "step": 33555
+    },
+    {
+      "epoch": 0.6712,
+      "grad_norm": 0.287109375,
+      "learning_rate": 5.8944028613787206e-05,
+      "loss": 1.2709,
+      "step": 33560
+    },
+    {
+      "epoch": 0.6713,
+      "grad_norm": 0.201171875,
+      "learning_rate": 5.8912202114638524e-05,
+      "loss": 1.293,
+      "step": 33565
+    },
+    {
+      "epoch": 0.6714,
+      "grad_norm": 0.423828125,
+      "learning_rate": 5.88803806219134e-05,
+      "loss": 1.2856,
+      "step": 33570
+    },
+    {
+      "epoch": 0.6715,
+      "grad_norm": 0.19140625,
+      "learning_rate": 5.884856413948913e-05,
+      "loss": 1.2813,
+      "step": 33575
+    },
+    {
+      "epoch": 0.6716,
+      "grad_norm": 0.287109375,
+      "learning_rate": 5.881675267124254e-05,
+      "loss": 1.2793,
+      "step": 33580
+    },
+    {
+      "epoch": 0.6717,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 5.8784946221049666e-05,
+      "loss": 1.2822,
+      "step": 33585
+    },
+    {
+      "epoch": 0.6718,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.8753144792786096e-05,
+      "loss": 1.2875,
+      "step": 33590
+    },
+    {
+      "epoch": 0.6719,
+      "grad_norm": 0.376953125,
+      "learning_rate": 5.87213483903267e-05,
+      "loss": 1.2677,
+      "step": 33595
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 0.19921875,
+      "learning_rate": 5.868955701754584e-05,
+      "loss": 1.2642,
+      "step": 33600
+    },
+    {
+      "epoch": 0.6721,
+      "grad_norm": 0.423828125,
+      "learning_rate": 5.86577706783171e-05,
+      "loss": 1.2766,
+      "step": 33605
+    },
+    {
+      "epoch": 0.6722,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.862598937651365e-05,
+      "loss": 1.2864,
+      "step": 33610
+    },
+    {
+      "epoch": 0.6723,
+      "grad_norm": 0.220703125,
+      "learning_rate": 5.8594213116007855e-05,
+      "loss": 1.2845,
+      "step": 33615
+    },
+    {
+      "epoch": 0.6724,
+      "grad_norm": 0.1953125,
+      "learning_rate": 5.856244190067159e-05,
+      "loss": 1.2726,
+      "step": 33620
+    },
+    {
+      "epoch": 0.6725,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.853067573437612e-05,
+      "loss": 1.2669,
+      "step": 33625
+    },
+    {
+      "epoch": 0.6726,
+      "grad_norm": 0.220703125,
+      "learning_rate": 5.849891462099198e-05,
+      "loss": 1.2787,
+      "step": 33630
+    },
+    {
+      "epoch": 0.6727,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.8467158564389234e-05,
+      "loss": 1.2727,
+      "step": 33635
+    },
+    {
+      "epoch": 0.6728,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 5.8435407568437216e-05,
+      "loss": 1.2714,
+      "step": 33640
+    },
+    {
+      "epoch": 0.6729,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.8403661637004736e-05,
+      "loss": 1.2862,
+      "step": 33645
+    },
+    {
+      "epoch": 0.673,
+      "grad_norm": 0.3046875,
+      "learning_rate": 5.83719207739599e-05,
+      "loss": 1.2781,
+      "step": 33650
+    },
+    {
+      "epoch": 0.6731,
+      "grad_norm": 0.216796875,
+      "learning_rate": 5.834018498317024e-05,
+      "loss": 1.271,
+      "step": 33655
+    },
+    {
+      "epoch": 0.6732,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.8308454268502675e-05,
+      "loss": 1.2541,
+      "step": 33660
+    },
+    {
+      "epoch": 0.6733,
+      "grad_norm": 0.142578125,
+      "learning_rate": 5.82767286338235e-05,
+      "loss": 1.2579,
+      "step": 33665
+    },
+    {
+      "epoch": 0.6734,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.8245008082998364e-05,
+      "loss": 1.2857,
+      "step": 33670
+    },
+    {
+      "epoch": 0.6735,
+      "grad_norm": 0.43359375,
+      "learning_rate": 5.82132926198923e-05,
+      "loss": 1.2745,
+      "step": 33675
+    },
+    {
+      "epoch": 0.6736,
+      "grad_norm": 0.8359375,
+      "learning_rate": 5.818158224836987e-05,
+      "loss": 1.2876,
+      "step": 33680
+    },
+    {
+      "epoch": 0.6737,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.814987697229471e-05,
+      "loss": 1.26,
+      "step": 33685
+    },
+    {
+      "epoch": 0.6738,
+      "grad_norm": 0.263671875,
+      "learning_rate": 5.8118176795530176e-05,
+      "loss": 1.2587,
+      "step": 33690
+    },
+    {
+      "epoch": 0.6739,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.808648172193868e-05,
+      "loss": 1.2751,
+      "step": 33695
+    },
+    {
+      "epoch": 0.674,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.805479175538229e-05,
+      "loss": 1.268,
+      "step": 33700
+    },
+    {
+      "epoch": 0.6741,
+      "grad_norm": 0.328125,
+      "learning_rate": 5.802310689972233e-05,
+      "loss": 1.2753,
+      "step": 33705
+    },
+    {
+      "epoch": 0.6742,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 5.799142715881938e-05,
+      "loss": 1.2611,
+      "step": 33710
+    },
+    {
+      "epoch": 0.6743,
+      "grad_norm": 0.166015625,
+      "learning_rate": 5.795975253653364e-05,
+      "loss": 1.2675,
+      "step": 33715
+    },
+    {
+      "epoch": 0.6744,
+      "grad_norm": 0.216796875,
+      "learning_rate": 5.792808303672454e-05,
+      "loss": 1.2584,
+      "step": 33720
+    },
+    {
+      "epoch": 0.6745,
+      "grad_norm": 0.447265625,
+      "learning_rate": 5.789641866325091e-05,
+      "loss": 1.2733,
+      "step": 33725
+    },
+    {
+      "epoch": 0.6746,
+      "grad_norm": 0.298828125,
+      "learning_rate": 5.786475941997094e-05,
+      "loss": 1.278,
+      "step": 33730
+    },
+    {
+      "epoch": 0.6747,
+      "grad_norm": 0.396484375,
+      "learning_rate": 5.783310531074223e-05,
+      "loss": 1.2749,
+      "step": 33735
+    },
+    {
+      "epoch": 0.6748,
+      "grad_norm": 0.490234375,
+      "learning_rate": 5.780145633942173e-05,
+      "loss": 1.2648,
+      "step": 33740
+    },
+    {
+      "epoch": 0.6749,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 5.7769812509865773e-05,
+      "loss": 1.2797,
+      "step": 33745
+    },
+    {
+      "epoch": 0.675,
+      "grad_norm": 0.193359375,
+      "learning_rate": 5.773817382593008e-05,
+      "loss": 1.2868,
+      "step": 33750
+    },
+    {
+      "epoch": 0.6751,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.770654029146969e-05,
+      "loss": 1.2811,
+      "step": 33755
+    },
+    {
+      "epoch": 0.6752,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.7674911910339094e-05,
+      "loss": 1.266,
+      "step": 33760
+    },
+    {
+      "epoch": 0.6753,
+      "grad_norm": 0.181640625,
+      "learning_rate": 5.764328868639208e-05,
+      "loss": 1.2694,
+      "step": 33765
+    },
+    {
+      "epoch": 0.6754,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.7611670623481864e-05,
+      "loss": 1.2682,
+      "step": 33770
+    },
+    {
+      "epoch": 0.6755,
+      "grad_norm": 0.3203125,
+      "learning_rate": 5.758005772546097e-05,
+      "loss": 1.2715,
+      "step": 33775
+    },
+    {
+      "epoch": 0.6756,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.754844999618144e-05,
+      "loss": 1.2574,
+      "step": 33780
+    },
+    {
+      "epoch": 0.6757,
+      "grad_norm": 0.375,
+      "learning_rate": 5.7516847439494435e-05,
+      "loss": 1.2642,
+      "step": 33785
+    },
+    {
+      "epoch": 0.6758,
+      "grad_norm": 0.90234375,
+      "learning_rate": 5.748525005925074e-05,
+      "loss": 1.2855,
+      "step": 33790
+    },
+    {
+      "epoch": 0.6759,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.7453657859300415e-05,
+      "loss": 1.2875,
+      "step": 33795
+    },
+    {
+      "epoch": 0.676,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.7422070843492734e-05,
+      "loss": 1.2768,
+      "step": 33800
+    },
+    {
+      "epoch": 0.6761,
+      "grad_norm": 0.126953125,
+      "learning_rate": 5.739048901567665e-05,
+      "loss": 1.2817,
+      "step": 33805
+    },
+    {
+      "epoch": 0.6762,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.735891237970015e-05,
+      "loss": 1.2819,
+      "step": 33810
+    },
+    {
+      "epoch": 0.6763,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 5.732734093941087e-05,
+      "loss": 1.2687,
+      "step": 33815
+    },
+    {
+      "epoch": 0.6764,
+      "grad_norm": 0.2265625,
+      "learning_rate": 5.7295774698655655e-05,
+      "loss": 1.276,
+      "step": 33820
+    },
+    {
+      "epoch": 0.6765,
+      "grad_norm": 0.271484375,
+      "learning_rate": 5.7264213661280765e-05,
+      "loss": 1.2823,
+      "step": 33825
+    },
+    {
+      "epoch": 0.6766,
+      "grad_norm": 0.318359375,
+      "learning_rate": 5.723265783113181e-05,
+      "loss": 1.2696,
+      "step": 33830
+    },
+    {
+      "epoch": 0.6767,
+      "grad_norm": 0.2421875,
+      "learning_rate": 5.720110721205376e-05,
+      "loss": 1.2666,
+      "step": 33835
+    },
+    {
+      "epoch": 0.6768,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.716956180789098e-05,
+      "loss": 1.2681,
+      "step": 33840
+    },
+    {
+      "epoch": 0.6769,
+      "grad_norm": 0.9140625,
+      "learning_rate": 5.713802162248718e-05,
+      "loss": 1.279,
+      "step": 33845
+    },
+    {
+      "epoch": 0.677,
+      "grad_norm": 0.1796875,
+      "learning_rate": 5.710648665968543e-05,
+      "loss": 1.2878,
+      "step": 33850
+    },
+    {
+      "epoch": 0.6771,
+      "grad_norm": 0.158203125,
+      "learning_rate": 5.707495692332816e-05,
+      "loss": 1.2917,
+      "step": 33855
+    },
+    {
+      "epoch": 0.6772,
+      "grad_norm": 0.32421875,
+      "learning_rate": 5.704343241725719e-05,
+      "loss": 1.265,
+      "step": 33860
+    },
+    {
+      "epoch": 0.6773,
+      "grad_norm": 0.2578125,
+      "learning_rate": 5.701191314531364e-05,
+      "loss": 1.2891,
+      "step": 33865
+    },
+    {
+      "epoch": 0.6774,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.6980399111338156e-05,
+      "loss": 1.2916,
+      "step": 33870
+    },
+    {
+      "epoch": 0.6775,
+      "grad_norm": 0.44140625,
+      "learning_rate": 5.694889031917047e-05,
+      "loss": 1.2804,
+      "step": 33875
+    },
+    {
+      "epoch": 0.6776,
+      "grad_norm": 0.287109375,
+      "learning_rate": 5.691738677265e-05,
+      "loss": 1.271,
+      "step": 33880
+    },
+    {
+      "epoch": 0.6777,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 5.6885888475615204e-05,
+      "loss": 1.264,
+      "step": 33885
+    },
+    {
+      "epoch": 0.6778,
+      "grad_norm": 0.38671875,
+      "learning_rate": 5.6854395431904094e-05,
+      "loss": 1.2737,
+      "step": 33890
+    },
+    {
+      "epoch": 0.6779,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 5.68229076453541e-05,
+      "loss": 1.2829,
+      "step": 33895
+    },
+    {
+      "epoch": 0.678,
+      "grad_norm": 0.373046875,
+      "learning_rate": 5.679142511980175e-05,
+      "loss": 1.2823,
+      "step": 33900
+    },
+    {
+      "epoch": 0.6781,
+      "grad_norm": 0.236328125,
+      "learning_rate": 5.675994785908329e-05,
+      "loss": 1.272,
+      "step": 33905
+    },
+    {
+      "epoch": 0.6782,
+      "grad_norm": 0.373046875,
+      "learning_rate": 5.6728475867033925e-05,
+      "loss": 1.2967,
+      "step": 33910
+    },
+    {
+      "epoch": 0.6783,
+      "grad_norm": 1.0,
+      "learning_rate": 5.669700914748857e-05,
+      "loss": 1.2835,
+      "step": 33915
+    },
+    {
+      "epoch": 0.6784,
+      "grad_norm": 0.365234375,
+      "learning_rate": 5.666554770428129e-05,
+      "loss": 1.2769,
+      "step": 33920
+    },
+    {
+      "epoch": 0.6785,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 5.663409154124557e-05,
+      "loss": 1.2758,
+      "step": 33925
+    },
+    {
+      "epoch": 0.6786,
+      "grad_norm": 0.287109375,
+      "learning_rate": 5.6602640662214256e-05,
+      "loss": 1.2672,
+      "step": 33930
+    },
+    {
+      "epoch": 0.6787,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 5.657119507101954e-05,
+      "loss": 1.2789,
+      "step": 33935
+    },
+    {
+      "epoch": 0.6788,
+      "grad_norm": 0.318359375,
+      "learning_rate": 5.653975477149298e-05,
+      "loss": 1.2624,
+      "step": 33940
+    },
+    {
+      "epoch": 0.6789,
+      "grad_norm": 0.193359375,
+      "learning_rate": 5.6508319767465465e-05,
+      "loss": 1.2599,
+      "step": 33945
+    },
+    {
+      "epoch": 0.679,
+      "grad_norm": 0.2890625,
+      "learning_rate": 5.647689006276726e-05,
+      "loss": 1.2649,
+      "step": 33950
+    },
+    {
+      "epoch": 0.6791,
+      "grad_norm": 0.306640625,
+      "learning_rate": 5.6445465661227994e-05,
+      "loss": 1.2828,
+      "step": 33955
+    },
+    {
+      "epoch": 0.6792,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 5.641404656667661e-05,
+      "loss": 1.2734,
+      "step": 33960
+    },
+    {
+      "epoch": 0.6793,
+      "grad_norm": 0.251953125,
+      "learning_rate": 5.6382632782941405e-05,
+      "loss": 1.2656,
+      "step": 33965
+    },
+    {
+      "epoch": 0.6794,
+      "grad_norm": 0.244140625,
+      "learning_rate": 5.635122431385016e-05,
+      "loss": 1.2643,
+      "step": 33970
+    },
+    {
+      "epoch": 0.6795,
+      "grad_norm": 0.3203125,
+      "learning_rate": 5.63198211632298e-05,
+      "loss": 1.2682,
+      "step": 33975
+    },
+    {
+      "epoch": 0.6796,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 5.6288423334906735e-05,
+      "loss": 1.2788,
+      "step": 33980
+    },
+    {
+      "epoch": 0.6797,
+      "grad_norm": 0.23046875,
+      "learning_rate": 5.6257030832706695e-05,
+      "loss": 1.2832,
+      "step": 33985
+    },
+    {
+      "epoch": 0.6798,
+      "grad_norm": 0.84765625,
+      "learning_rate": 5.622564366045472e-05,
+      "loss": 1.2844,
+      "step": 33990
+    },
+    {
+      "epoch": 0.6799,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.619426182197536e-05,
+      "loss": 1.2814,
+      "step": 33995
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2734375,
+      "learning_rate": 5.616288532109225e-05,
+      "loss": 1.2862,
+      "step": 34000
+    },
+    {
+      "epoch": 0.6801,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.6131514161628626e-05,
+      "loss": 1.2662,
+      "step": 34005
+    },
+    {
+      "epoch": 0.6802,
+      "grad_norm": 0.208984375,
+      "learning_rate": 5.610014834740693e-05,
+      "loss": 1.2763,
+      "step": 34010
+    },
+    {
+      "epoch": 0.6803,
+      "grad_norm": 0.2109375,
+      "learning_rate": 5.6068787882249005e-05,
+      "loss": 1.2653,
+      "step": 34015
+    },
+    {
+      "epoch": 0.6804,
+      "grad_norm": 0.31640625,
+      "learning_rate": 5.6037432769976064e-05,
+      "loss": 1.2715,
+      "step": 34020
+    },
+    {
+      "epoch": 0.6805,
+      "grad_norm": 0.296875,
+      "learning_rate": 5.6006083014408484e-05,
+      "loss": 1.261,
+      "step": 34025
+    },
+    {
+      "epoch": 0.6806,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 5.5974738619366295e-05,
+      "loss": 1.2682,
+      "step": 34030
+    },
+    {
+      "epoch": 0.6807,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.594339958866867e-05,
+      "loss": 1.2784,
+      "step": 34035
+    },
+    {
+      "epoch": 0.6808,
+      "grad_norm": 0.279296875,
+      "learning_rate": 5.591206592613416e-05,
+      "loss": 1.2807,
+      "step": 34040
+    },
+    {
+      "epoch": 0.6809,
+      "grad_norm": 0.1875,
+      "learning_rate": 5.588073763558068e-05,
+      "loss": 1.2734,
+      "step": 34045
+    },
+    {
+      "epoch": 0.681,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 5.584941472082549e-05,
+      "loss": 1.2982,
+      "step": 34050
+    },
+    {
+      "epoch": 0.6811,
+      "grad_norm": 0.2734375,
+      "learning_rate": 5.5818097185685206e-05,
+      "loss": 1.2769,
+      "step": 34055
+    },
+    {
+      "epoch": 0.6812,
+      "grad_norm": 0.490234375,
+      "learning_rate": 5.578678503397574e-05,
+      "loss": 1.2864,
+      "step": 34060
+    },
+    {
+      "epoch": 0.6813,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 5.575547826951242e-05,
+      "loss": 1.2657,
+      "step": 34065
+    },
+    {
+      "epoch": 0.6814,
+      "grad_norm": 0.38671875,
+      "learning_rate": 5.572417689610987e-05,
+      "loss": 1.2852,
+      "step": 34070
+    },
+    {
+      "epoch": 0.6815,
+      "grad_norm": 0.75390625,
+      "learning_rate": 5.5692880917582046e-05,
+      "loss": 1.2657,
+      "step": 34075
+    },
+    {
+      "epoch": 0.6816,
+      "grad_norm": 0.296875,
+      "learning_rate": 5.566159033774225e-05,
+      "loss": 1.2794,
+      "step": 34080
+    },
+    {
+      "epoch": 0.6817,
+      "grad_norm": 0.197265625,
+      "learning_rate": 5.5630305160403275e-05,
+      "loss": 1.2779,
+      "step": 34085
+    },
+    {
+      "epoch": 0.6818,
+      "grad_norm": 0.150390625,
+      "learning_rate": 5.5599025389376935e-05,
+      "loss": 1.2826,
+      "step": 34090
+    },
+    {
+      "epoch": 0.6819,
+      "grad_norm": 0.138671875,
+      "learning_rate": 5.556775102847475e-05,
+      "loss": 1.2624,
+      "step": 34095
+    },
+    {
+      "epoch": 0.682,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.553648208150728e-05,
+      "loss": 1.278,
+      "step": 34100
+    },
+    {
+      "epoch": 0.6821,
+      "grad_norm": 0.453125,
+      "learning_rate": 5.5505218552284565e-05,
+      "loss": 1.2689,
+      "step": 34105
+    },
+    {
+      "epoch": 0.6822,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 5.547396044461608e-05,
+      "loss": 1.2768,
+      "step": 34110
+    },
+    {
+      "epoch": 0.6823,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 5.544270776231038e-05,
+      "loss": 1.2661,
+      "step": 34115
+    },
+    {
+      "epoch": 0.6824,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.541146050917561e-05,
+      "loss": 1.2761,
+      "step": 34120
+    },
+    {
+      "epoch": 0.6825,
+      "grad_norm": 0.76953125,
+      "learning_rate": 5.5380218689019125e-05,
+      "loss": 1.2815,
+      "step": 34125
+    },
+    {
+      "epoch": 0.6826,
+      "grad_norm": 0.203125,
+      "learning_rate": 5.5348982305647643e-05,
+      "loss": 1.2739,
+      "step": 34130
+    },
+    {
+      "epoch": 0.6827,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 5.5317751362867234e-05,
+      "loss": 1.268,
+      "step": 34135
+    },
+    {
+      "epoch": 0.6828,
+      "grad_norm": 0.197265625,
+      "learning_rate": 5.5286525864483285e-05,
+      "loss": 1.2863,
+      "step": 34140
+    },
+    {
+      "epoch": 0.6829,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 5.525530581430054e-05,
+      "loss": 1.2735,
+      "step": 34145
+    },
+    {
+      "epoch": 0.683,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 5.522409121612304e-05,
+      "loss": 1.2777,
+      "step": 34150
+    },
+    {
+      "epoch": 0.6831,
+      "grad_norm": 0.26953125,
+      "learning_rate": 5.519288207375422e-05,
+      "loss": 1.2784,
+      "step": 34155
+    },
+    {
+      "epoch": 0.6832,
+      "grad_norm": 0.302734375,
+      "learning_rate": 5.5161678390996796e-05,
+      "loss": 1.278,
+      "step": 34160
+    },
+    {
+      "epoch": 0.6833,
+      "grad_norm": 0.291015625,
+      "learning_rate": 5.513048017165284e-05,
+      "loss": 1.2645,
+      "step": 34165
+    },
+    {
+      "epoch": 0.6834,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.509928741952379e-05,
+      "loss": 1.2842,
+      "step": 34170
+    },
+    {
+      "epoch": 0.6835,
+      "grad_norm": 0.474609375,
+      "learning_rate": 5.506810013841036e-05,
+      "loss": 1.2772,
+      "step": 34175
+    },
+    {
+      "epoch": 0.6836,
+      "grad_norm": 0.236328125,
+      "learning_rate": 5.50369183321126e-05,
+      "loss": 1.2803,
+      "step": 34180
+    },
+    {
+      "epoch": 0.6837,
+      "grad_norm": 0.283203125,
+      "learning_rate": 5.5005742004430025e-05,
+      "loss": 1.274,
+      "step": 34185
+    },
+    {
+      "epoch": 0.6838,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.497457115916127e-05,
+      "loss": 1.2913,
+      "step": 34190
+    },
+    {
+      "epoch": 0.6839,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.494340580010441e-05,
+      "loss": 1.2844,
+      "step": 34195
+    },
+    {
+      "epoch": 0.684,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 5.491224593105695e-05,
+      "loss": 1.2986,
+      "step": 34200
+    },
+    {
+      "epoch": 0.6841,
+      "grad_norm": 0.412109375,
+      "learning_rate": 5.488109155581549e-05,
+      "loss": 1.2725,
+      "step": 34205
+    },
+    {
+      "epoch": 0.6842,
+      "grad_norm": 0.47265625,
+      "learning_rate": 5.484994267817624e-05,
+      "loss": 1.2838,
+      "step": 34210
+    },
+    {
+      "epoch": 0.6843,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 5.481879930193443e-05,
+      "loss": 1.2694,
+      "step": 34215
+    },
+    {
+      "epoch": 0.6844,
+      "grad_norm": 0.166015625,
+      "learning_rate": 5.478766143088492e-05,
+      "loss": 1.2796,
+      "step": 34220
+    },
+    {
+      "epoch": 0.6845,
+      "grad_norm": 0.298828125,
+      "learning_rate": 5.475652906882173e-05,
+      "loss": 1.2581,
+      "step": 34225
+    },
+    {
+      "epoch": 0.6846,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.4725402219538236e-05,
+      "loss": 1.2779,
+      "step": 34230
+    },
+    {
+      "epoch": 0.6847,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.469428088682717e-05,
+      "loss": 1.268,
+      "step": 34235
+    },
+    {
+      "epoch": 0.6848,
+      "grad_norm": 0.5,
+      "learning_rate": 5.466316507448048e-05,
+      "loss": 1.2755,
+      "step": 34240
+    },
+    {
+      "epoch": 0.6849,
+      "grad_norm": 0.19140625,
+      "learning_rate": 5.4632054786289656e-05,
+      "loss": 1.2821,
+      "step": 34245
+    },
+    {
+      "epoch": 0.685,
+      "grad_norm": 0.73828125,
+      "learning_rate": 5.4600950026045326e-05,
+      "loss": 1.2656,
+      "step": 34250
+    },
+    {
+      "epoch": 0.6851,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 5.4569850797537536e-05,
+      "loss": 1.2804,
+      "step": 34255
+    },
+    {
+      "epoch": 0.6852,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.4538757104555615e-05,
+      "loss": 1.2718,
+      "step": 34260
+    },
+    {
+      "epoch": 0.6853,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.4507668950888245e-05,
+      "loss": 1.2633,
+      "step": 34265
+    },
+    {
+      "epoch": 0.6854,
+      "grad_norm": 0.359375,
+      "learning_rate": 5.447658634032338e-05,
+      "loss": 1.28,
+      "step": 34270
+    },
+    {
+      "epoch": 0.6855,
+      "grad_norm": 0.4140625,
+      "learning_rate": 5.4445509276648466e-05,
+      "loss": 1.2677,
+      "step": 34275
+    },
+    {
+      "epoch": 0.6856,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 5.441443776365003e-05,
+      "loss": 1.2748,
+      "step": 34280
+    },
+    {
+      "epoch": 0.6857,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.438337180511406e-05,
+      "loss": 1.2639,
+      "step": 34285
+    },
+    {
+      "epoch": 0.6858,
+      "grad_norm": 0.361328125,
+      "learning_rate": 5.435231140482587e-05,
+      "loss": 1.2621,
+      "step": 34290
+    },
+    {
+      "epoch": 0.6859,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.4321256566570036e-05,
+      "loss": 1.2786,
+      "step": 34295
+    },
+    {
+      "epoch": 0.686,
+      "grad_norm": 0.88671875,
+      "learning_rate": 5.4290207294130615e-05,
+      "loss": 1.2753,
+      "step": 34300
+    },
+    {
+      "epoch": 0.6861,
+      "grad_norm": 0.35546875,
+      "learning_rate": 5.42591635912907e-05,
+      "loss": 1.2869,
+      "step": 34305
+    },
+    {
+      "epoch": 0.6862,
+      "grad_norm": 0.2890625,
+      "learning_rate": 5.4228125461833024e-05,
+      "loss": 1.2987,
+      "step": 34310
+    },
+    {
+      "epoch": 0.6863,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.419709290953936e-05,
+      "loss": 1.2873,
+      "step": 34315
+    },
+    {
+      "epoch": 0.6864,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 5.416606593819101e-05,
+      "loss": 1.2635,
+      "step": 34320
+    },
+    {
+      "epoch": 0.6865,
+      "grad_norm": 0.28125,
+      "learning_rate": 5.4135044551568546e-05,
+      "loss": 1.2677,
+      "step": 34325
+    },
+    {
+      "epoch": 0.6866,
+      "grad_norm": 0.1328125,
+      "learning_rate": 5.4104028753451696e-05,
+      "loss": 1.2656,
+      "step": 34330
+    },
+    {
+      "epoch": 0.6867,
+      "grad_norm": 0.275390625,
+      "learning_rate": 5.407301854761977e-05,
+      "loss": 1.2693,
+      "step": 34335
+    },
+    {
+      "epoch": 0.6868,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.404201393785122e-05,
+      "loss": 1.2876,
+      "step": 34340
+    },
+    {
+      "epoch": 0.6869,
+      "grad_norm": 0.51171875,
+      "learning_rate": 5.401101492792386e-05,
+      "loss": 1.266,
+      "step": 34345
+    },
+    {
+      "epoch": 0.687,
+      "grad_norm": 0.1640625,
+      "learning_rate": 5.398002152161484e-05,
+      "loss": 1.281,
+      "step": 34350
+    },
+    {
+      "epoch": 0.6871,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 5.394903372270062e-05,
+      "loss": 1.2857,
+      "step": 34355
+    },
+    {
+      "epoch": 0.6872,
+      "grad_norm": 0.326171875,
+      "learning_rate": 5.3918051534956926e-05,
+      "loss": 1.2764,
+      "step": 34360
+    },
+    {
+      "epoch": 0.6873,
+      "grad_norm": 0.234375,
+      "learning_rate": 5.388707496215888e-05,
+      "loss": 1.2602,
+      "step": 34365
+    },
+    {
+      "epoch": 0.6874,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.3856104008080876e-05,
+      "loss": 1.2527,
+      "step": 34370
+    },
+    {
+      "epoch": 0.6875,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.382513867649663e-05,
+      "loss": 1.2566,
+      "step": 34375
+    },
+    {
+      "epoch": 0.6876,
+      "grad_norm": 0.203125,
+      "learning_rate": 5.3794178971179165e-05,
+      "loss": 1.2825,
+      "step": 34380
+    },
+    {
+      "epoch": 0.6877,
+      "grad_norm": 0.296875,
+      "learning_rate": 5.3763224895900846e-05,
+      "loss": 1.2832,
+      "step": 34385
+    },
+    {
+      "epoch": 0.6878,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.373227645443332e-05,
+      "loss": 1.2661,
+      "step": 34390
+    },
+    {
+      "epoch": 0.6879,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 5.3701333650547525e-05,
+      "loss": 1.274,
+      "step": 34395
+    },
+    {
+      "epoch": 0.688,
+      "grad_norm": 0.3125,
+      "learning_rate": 5.3670396488013854e-05,
+      "loss": 1.2656,
+      "step": 34400
+    },
+    {
+      "epoch": 0.6881,
+      "grad_norm": 0.220703125,
+      "learning_rate": 5.3639464970601775e-05,
+      "loss": 1.2599,
+      "step": 34405
+    },
+    {
+      "epoch": 0.6882,
+      "grad_norm": 0.275390625,
+      "learning_rate": 5.360853910208028e-05,
+      "loss": 1.2852,
+      "step": 34410
+    },
+    {
+      "epoch": 0.6883,
+      "grad_norm": 0.298828125,
+      "learning_rate": 5.357761888621764e-05,
+      "loss": 1.2747,
+      "step": 34415
+    },
+    {
+      "epoch": 0.6884,
+      "grad_norm": 0.212890625,
+      "learning_rate": 5.3546704326781236e-05,
+      "loss": 1.263,
+      "step": 34420
+    },
+    {
+      "epoch": 0.6885,
+      "grad_norm": 0.51171875,
+      "learning_rate": 5.351579542753807e-05,
+      "loss": 1.2689,
+      "step": 34425
+    },
+    {
+      "epoch": 0.6886,
+      "grad_norm": 0.447265625,
+      "learning_rate": 5.348489219225416e-05,
+      "loss": 1.293,
+      "step": 34430
+    },
+    {
+      "epoch": 0.6887,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.345399462469509e-05,
+      "loss": 1.2671,
+      "step": 34435
+    },
+    {
+      "epoch": 0.6888,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.3423102728625574e-05,
+      "loss": 1.2773,
+      "step": 34440
+    },
+    {
+      "epoch": 0.6889,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 5.3392216507809714e-05,
+      "loss": 1.2859,
+      "step": 34445
+    },
+    {
+      "epoch": 0.689,
+      "grad_norm": 0.361328125,
+      "learning_rate": 5.33613359660109e-05,
+      "loss": 1.2643,
+      "step": 34450
+    },
+    {
+      "epoch": 0.6891,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 5.33304611069918e-05,
+      "loss": 1.2746,
+      "step": 34455
+    },
+    {
+      "epoch": 0.6892,
+      "grad_norm": 0.162109375,
+      "learning_rate": 5.329959193451448e-05,
+      "loss": 1.2664,
+      "step": 34460
+    },
+    {
+      "epoch": 0.6893,
+      "grad_norm": 0.380859375,
+      "learning_rate": 5.326872845234021e-05,
+      "loss": 1.2634,
+      "step": 34465
+    },
+    {
+      "epoch": 0.6894,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 5.3237870664229636e-05,
+      "loss": 1.2616,
+      "step": 34470
+    },
+    {
+      "epoch": 0.6895,
+      "grad_norm": 1.34375,
+      "learning_rate": 5.320701857394268e-05,
+      "loss": 1.2696,
+      "step": 34475
+    },
+    {
+      "epoch": 0.6896,
+      "grad_norm": 0.498046875,
+      "learning_rate": 5.317617218523856e-05,
+      "loss": 1.2871,
+      "step": 34480
+    },
+    {
+      "epoch": 0.6897,
+      "grad_norm": 0.451171875,
+      "learning_rate": 5.3145331501875796e-05,
+      "loss": 1.2729,
+      "step": 34485
+    },
+    {
+      "epoch": 0.6898,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.311449652761235e-05,
+      "loss": 1.2746,
+      "step": 34490
+    },
+    {
+      "epoch": 0.6899,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.3083667266205194e-05,
+      "loss": 1.2854,
+      "step": 34495
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.305284372141095e-05,
+      "loss": 1.2623,
+      "step": 34500
+    },
+    {
+      "epoch": 0.6901,
+      "grad_norm": 0.7890625,
+      "learning_rate": 5.302202589698525e-05,
+      "loss": 1.275,
+      "step": 34505
+    },
+    {
+      "epoch": 0.6902,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.299121379668316e-05,
+      "loss": 1.2577,
+      "step": 34510
+    },
+    {
+      "epoch": 0.6903,
+      "grad_norm": 0.28125,
+      "learning_rate": 5.296040742425916e-05,
+      "loss": 1.2704,
+      "step": 34515
+    },
+    {
+      "epoch": 0.6904,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 5.292960678346675e-05,
+      "loss": 1.2726,
+      "step": 34520
+    },
+    {
+      "epoch": 0.6905,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.2898811878059e-05,
+      "loss": 1.2677,
+      "step": 34525
+    },
+    {
+      "epoch": 0.6906,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.286802271178815e-05,
+      "loss": 1.261,
+      "step": 34530
+    },
+    {
+      "epoch": 0.6907,
+      "grad_norm": 0.310546875,
+      "learning_rate": 5.2837239288405784e-05,
+      "loss": 1.2728,
+      "step": 34535
+    },
+    {
+      "epoch": 0.6908,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.2806461611662735e-05,
+      "loss": 1.2714,
+      "step": 34540
+    },
+    {
+      "epoch": 0.6909,
+      "grad_norm": 0.181640625,
+      "learning_rate": 5.277568968530919e-05,
+      "loss": 1.2559,
+      "step": 34545
+    },
+    {
+      "epoch": 0.691,
+      "grad_norm": 0.87890625,
+      "learning_rate": 5.274492351309461e-05,
+      "loss": 1.2667,
+      "step": 34550
+    },
+    {
+      "epoch": 0.6911,
+      "grad_norm": 0.30078125,
+      "learning_rate": 5.271416309876776e-05,
+      "loss": 1.266,
+      "step": 34555
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 5.26834084460767e-05,
+      "loss": 1.28,
+      "step": 34560
+    },
+    {
+      "epoch": 0.6913,
+      "grad_norm": 0.240234375,
+      "learning_rate": 5.265265955876879e-05,
+      "loss": 1.2755,
+      "step": 34565
+    },
+    {
+      "epoch": 0.6914,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.2621916440590715e-05,
+      "loss": 1.2795,
+      "step": 34570
+    },
+    {
+      "epoch": 0.6915,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 5.259117909528839e-05,
+      "loss": 1.2875,
+      "step": 34575
+    },
+    {
+      "epoch": 0.6916,
+      "grad_norm": 0.294921875,
+      "learning_rate": 5.256044752660709e-05,
+      "loss": 1.2695,
+      "step": 34580
+    },
+    {
+      "epoch": 0.6917,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.2529721738291315e-05,
+      "loss": 1.2847,
+      "step": 34585
+    },
+    {
+      "epoch": 0.6918,
+      "grad_norm": 0.890625,
+      "learning_rate": 5.2499001734085044e-05,
+      "loss": 1.2767,
+      "step": 34590
+    },
+    {
+      "epoch": 0.6919,
+      "grad_norm": 0.44921875,
+      "learning_rate": 5.2468287517731276e-05,
+      "loss": 1.2783,
+      "step": 34595
+    },
+    {
+      "epoch": 0.692,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.243757909297247e-05,
+      "loss": 1.264,
+      "step": 34600
+    },
+    {
+      "epoch": 0.6921,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.2406876463550445e-05,
+      "loss": 1.2815,
+      "step": 34605
+    },
+    {
+      "epoch": 0.6922,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 5.237617963320608e-05,
+      "loss": 1.2689,
+      "step": 34610
+    },
+    {
+      "epoch": 0.6923,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.234548860567985e-05,
+      "loss": 1.2659,
+      "step": 34615
+    },
+    {
+      "epoch": 0.6924,
+      "grad_norm": 0.78125,
+      "learning_rate": 5.2314803384711195e-05,
+      "loss": 1.2699,
+      "step": 34620
+    },
+    {
+      "epoch": 0.6925,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.2284123974039154e-05,
+      "loss": 1.2589,
+      "step": 34625
+    },
+    {
+      "epoch": 0.6926,
+      "grad_norm": 0.224609375,
+      "learning_rate": 5.225345037740186e-05,
+      "loss": 1.2557,
+      "step": 34630
+    },
+    {
+      "epoch": 0.6927,
+      "grad_norm": 0.466796875,
+      "learning_rate": 5.222278259853681e-05,
+      "loss": 1.258,
+      "step": 34635
+    },
+    {
+      "epoch": 0.6928,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 5.2192120641180786e-05,
+      "loss": 1.2718,
+      "step": 34640
+    },
+    {
+      "epoch": 0.6929,
+      "grad_norm": 0.1796875,
+      "learning_rate": 5.216146450906984e-05,
+      "loss": 1.2565,
+      "step": 34645
+    },
+    {
+      "epoch": 0.693,
+      "grad_norm": 0.19140625,
+      "learning_rate": 5.213081420593933e-05,
+      "loss": 1.2747,
+      "step": 34650
+    },
+    {
+      "epoch": 0.6931,
+      "grad_norm": 0.412109375,
+      "learning_rate": 5.210016973552391e-05,
+      "loss": 1.2768,
+      "step": 34655
+    },
+    {
+      "epoch": 0.6932,
+      "grad_norm": 0.162109375,
+      "learning_rate": 5.20695311015575e-05,
+      "loss": 1.2807,
+      "step": 34660
+    },
+    {
+      "epoch": 0.6933,
+      "grad_norm": 0.2578125,
+      "learning_rate": 5.2038898307773354e-05,
+      "loss": 1.2879,
+      "step": 34665
+    },
+    {
+      "epoch": 0.6934,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.200827135790396e-05,
+      "loss": 1.2785,
+      "step": 34670
+    },
+    {
+      "epoch": 0.6935,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.197765025568109e-05,
+      "loss": 1.2834,
+      "step": 34675
+    },
+    {
+      "epoch": 0.6936,
+      "grad_norm": 0.2734375,
+      "learning_rate": 5.194703500483593e-05,
+      "loss": 1.2858,
+      "step": 34680
+    },
+    {
+      "epoch": 0.6937,
+      "grad_norm": 0.373046875,
+      "learning_rate": 5.1916425609098775e-05,
+      "loss": 1.2722,
+      "step": 34685
+    },
+    {
+      "epoch": 0.6938,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.188582207219931e-05,
+      "loss": 1.2756,
+      "step": 34690
+    },
+    {
+      "epoch": 0.6939,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.1855224397866476e-05,
+      "loss": 1.2618,
+      "step": 34695
+    },
+    {
+      "epoch": 0.694,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.182463258982846e-05,
+      "loss": 1.2661,
+      "step": 34700
+    },
+    {
+      "epoch": 0.6941,
+      "grad_norm": 0.80859375,
+      "learning_rate": 5.179404665181291e-05,
+      "loss": 1.2743,
+      "step": 34705
+    },
+    {
+      "epoch": 0.6942,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.1763466587546485e-05,
+      "loss": 1.2842,
+      "step": 34710
+    },
+    {
+      "epoch": 0.6943,
+      "grad_norm": 0.4921875,
+      "learning_rate": 5.1732892400755376e-05,
+      "loss": 1.2727,
+      "step": 34715
+    },
+    {
+      "epoch": 0.6944,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.170232409516496e-05,
+      "loss": 1.2703,
+      "step": 34720
+    },
+    {
+      "epoch": 0.6945,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 5.1671761674499765e-05,
+      "loss": 1.2768,
+      "step": 34725
+    },
+    {
+      "epoch": 0.6946,
+      "grad_norm": 0.203125,
+      "learning_rate": 5.1641205142483894e-05,
+      "loss": 1.2947,
+      "step": 34730
+    },
+    {
+      "epoch": 0.6947,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.16106545028404e-05,
+      "loss": 1.2641,
+      "step": 34735
+    },
+    {
+      "epoch": 0.6948,
+      "grad_norm": 0.25,
+      "learning_rate": 5.158010975929193e-05,
+      "loss": 1.2711,
+      "step": 34740
+    },
+    {
+      "epoch": 0.6949,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 5.1549570915560206e-05,
+      "loss": 1.2733,
+      "step": 34745
+    },
+    {
+      "epoch": 0.695,
+      "grad_norm": 0.271484375,
+      "learning_rate": 5.15190379753663e-05,
+      "loss": 1.2611,
+      "step": 34750
+    },
+    {
+      "epoch": 0.6951,
+      "grad_norm": 0.375,
+      "learning_rate": 5.148851094243057e-05,
+      "loss": 1.2934,
+      "step": 34755
+    },
+    {
+      "epoch": 0.6952,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.145798982047261e-05,
+      "loss": 1.2955,
+      "step": 34760
+    },
+    {
+      "epoch": 0.6953,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 5.1427474613211356e-05,
+      "loss": 1.2659,
+      "step": 34765
+    },
+    {
+      "epoch": 0.6954,
+      "grad_norm": 0.376953125,
+      "learning_rate": 5.1396965324364986e-05,
+      "loss": 1.2889,
+      "step": 34770
+    },
+    {
+      "epoch": 0.6955,
+      "grad_norm": 0.8046875,
+      "learning_rate": 5.1366461957650954e-05,
+      "loss": 1.2774,
+      "step": 34775
+    },
+    {
+      "epoch": 0.6956,
+      "grad_norm": 0.36328125,
+      "learning_rate": 5.133596451678603e-05,
+      "loss": 1.2756,
+      "step": 34780
+    },
+    {
+      "epoch": 0.6957,
+      "grad_norm": 0.162109375,
+      "learning_rate": 5.13054730054862e-05,
+      "loss": 1.2644,
+      "step": 34785
+    },
+    {
+      "epoch": 0.6958,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.127498742746675e-05,
+      "loss": 1.2759,
+      "step": 34790
+    },
+    {
+      "epoch": 0.6959,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.1244507786442356e-05,
+      "loss": 1.2915,
+      "step": 34795
+    },
+    {
+      "epoch": 0.696,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.121403408612672e-05,
+      "loss": 1.2719,
+      "step": 34800
+    },
+    {
+      "epoch": 0.6961,
+      "grad_norm": 0.455078125,
+      "learning_rate": 5.1183566330233124e-05,
+      "loss": 1.2888,
+      "step": 34805
+    },
+    {
+      "epoch": 0.6962,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.115310452247386e-05,
+      "loss": 1.2699,
+      "step": 34810
+    },
+    {
+      "epoch": 0.6963,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 5.112264866656059e-05,
+      "loss": 1.2819,
+      "step": 34815
+    },
+    {
+      "epoch": 0.6964,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 5.1092198766204415e-05,
+      "loss": 1.2777,
+      "step": 34820
+    },
+    {
+      "epoch": 0.6965,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.1061754825115374e-05,
+      "loss": 1.2691,
+      "step": 34825
+    },
+    {
+      "epoch": 0.6966,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.103131684700314e-05,
+      "loss": 1.3094,
+      "step": 34830
+    },
+    {
+      "epoch": 0.6967,
+      "grad_norm": 0.240234375,
+      "learning_rate": 5.100088483557634e-05,
+      "loss": 1.2936,
+      "step": 34835
+    },
+    {
+      "epoch": 0.6968,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.097045879454313e-05,
+      "loss": 1.2702,
+      "step": 34840
+    },
+    {
+      "epoch": 0.6969,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.0940038727610796e-05,
+      "loss": 1.2669,
+      "step": 34845
+    },
+    {
+      "epoch": 0.697,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.090962463848592e-05,
+      "loss": 1.2977,
+      "step": 34850
+    },
+    {
+      "epoch": 0.6971,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.087921653087437e-05,
+      "loss": 1.2654,
+      "step": 34855
+    },
+    {
+      "epoch": 0.6972,
+      "grad_norm": 0.30859375,
+      "learning_rate": 5.0848814408481305e-05,
+      "loss": 1.2595,
+      "step": 34860
+    },
+    {
+      "epoch": 0.6973,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 5.0818418275011104e-05,
+      "loss": 1.2667,
+      "step": 34865
+    },
+    {
+      "epoch": 0.6974,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.0788028134167456e-05,
+      "loss": 1.2566,
+      "step": 34870
+    },
+    {
+      "epoch": 0.6975,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.07576439896533e-05,
+      "loss": 1.2778,
+      "step": 34875
+    },
+    {
+      "epoch": 0.6976,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 5.072726584517086e-05,
+      "loss": 1.2856,
+      "step": 34880
+    },
+    {
+      "epoch": 0.6977,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 5.069689370442161e-05,
+      "loss": 1.2716,
+      "step": 34885
+    },
+    {
+      "epoch": 0.6978,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 5.066652757110628e-05,
+      "loss": 1.282,
+      "step": 34890
+    },
+    {
+      "epoch": 0.6979,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.0636167448924987e-05,
+      "loss": 1.2827,
+      "step": 34895
+    },
+    {
+      "epoch": 0.698,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.0605813341576924e-05,
+      "loss": 1.2789,
+      "step": 34900
+    },
+    {
+      "epoch": 0.6981,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.057546525276068e-05,
+      "loss": 1.2901,
+      "step": 34905
+    },
+    {
+      "epoch": 0.6982,
+      "grad_norm": 0.296875,
+      "learning_rate": 5.054512318617406e-05,
+      "loss": 1.2765,
+      "step": 34910
+    },
+    {
+      "epoch": 0.6983,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.051478714551414e-05,
+      "loss": 1.2835,
+      "step": 34915
+    },
+    {
+      "epoch": 0.6984,
+      "grad_norm": 0.29296875,
+      "learning_rate": 5.048445713447738e-05,
+      "loss": 1.2852,
+      "step": 34920
+    },
+    {
+      "epoch": 0.6985,
+      "grad_norm": 0.189453125,
+      "learning_rate": 5.045413315675924e-05,
+      "loss": 1.2821,
+      "step": 34925
+    },
+    {
+      "epoch": 0.6986,
+      "grad_norm": 0.234375,
+      "learning_rate": 5.0423815216054724e-05,
+      "loss": 1.2695,
+      "step": 34930
+    },
+    {
+      "epoch": 0.6987,
+      "grad_norm": 0.494140625,
+      "learning_rate": 5.039350331605794e-05,
+      "loss": 1.2856,
+      "step": 34935
+    },
+    {
+      "epoch": 0.6988,
+      "grad_norm": 0.25,
+      "learning_rate": 5.036319746046232e-05,
+      "loss": 1.2654,
+      "step": 34940
+    },
+    {
+      "epoch": 0.6989,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 5.033289765296054e-05,
+      "loss": 1.2752,
+      "step": 34945
+    },
+    {
+      "epoch": 0.699,
+      "grad_norm": 0.265625,
+      "learning_rate": 5.0302603897244474e-05,
+      "loss": 1.2685,
+      "step": 34950
+    },
+    {
+      "epoch": 0.6991,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.0272316197005396e-05,
+      "loss": 1.2726,
+      "step": 34955
+    },
+    {
+      "epoch": 0.6992,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.024203455593375e-05,
+      "loss": 1.2755,
+      "step": 34960
+    },
+    {
+      "epoch": 0.6993,
+      "grad_norm": 0.322265625,
+      "learning_rate": 5.021175897771927e-05,
+      "loss": 1.2712,
+      "step": 34965
+    },
+    {
+      "epoch": 0.6994,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.018148946605092e-05,
+      "loss": 1.2666,
+      "step": 34970
+    },
+    {
+      "epoch": 0.6995,
+      "grad_norm": 0.189453125,
+      "learning_rate": 5.015122602461698e-05,
+      "loss": 1.2712,
+      "step": 34975
+    },
+    {
+      "epoch": 0.6996,
+      "grad_norm": 0.197265625,
+      "learning_rate": 5.012096865710494e-05,
+      "loss": 1.2599,
+      "step": 34980
+    },
+    {
+      "epoch": 0.6997,
+      "grad_norm": 0.240234375,
+      "learning_rate": 5.0090717367201554e-05,
+      "loss": 1.2776,
+      "step": 34985
+    },
+    {
+      "epoch": 0.6998,
+      "grad_norm": 0.71875,
+      "learning_rate": 5.006047215859289e-05,
+      "loss": 1.2664,
+      "step": 34990
+    },
+    {
+      "epoch": 0.6999,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.003023303496419e-05,
+      "loss": 1.2647,
+      "step": 34995
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.361328125,
+      "learning_rate": 5.000000000000002e-05,
+      "loss": 1.282,
+      "step": 35000
+    },
+    {
+      "epoch": 0.7001,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 4.996977305738415e-05,
+      "loss": 1.2715,
+      "step": 35005
+    },
+    {
+      "epoch": 0.7002,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 4.9939552210799755e-05,
+      "loss": 1.2573,
+      "step": 35010
+    },
+    {
+      "epoch": 0.7003,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 4.990933746392899e-05,
+      "loss": 1.2778,
+      "step": 35015
+    },
+    {
+      "epoch": 0.7004,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.98791288204536e-05,
+      "loss": 1.2697,
+      "step": 35020
+    },
+    {
+      "epoch": 0.7005,
+      "grad_norm": 0.162109375,
+      "learning_rate": 4.9848926284054255e-05,
+      "loss": 1.267,
+      "step": 35025
+    },
+    {
+      "epoch": 0.7006,
+      "grad_norm": 0.3203125,
+      "learning_rate": 4.981872985841115e-05,
+      "loss": 1.2601,
+      "step": 35030
+    },
+    {
+      "epoch": 0.7007,
+      "grad_norm": 0.1953125,
+      "learning_rate": 4.978853954720364e-05,
+      "loss": 1.2586,
+      "step": 35035
+    },
+    {
+      "epoch": 0.7008,
+      "grad_norm": 0.275390625,
+      "learning_rate": 4.97583553541102e-05,
+      "loss": 1.2762,
+      "step": 35040
+    },
+    {
+      "epoch": 0.7009,
+      "grad_norm": 0.23828125,
+      "learning_rate": 4.97281772828088e-05,
+      "loss": 1.2701,
+      "step": 35045
+    },
+    {
+      "epoch": 0.701,
+      "grad_norm": 0.32421875,
+      "learning_rate": 4.969800533697649e-05,
+      "loss": 1.2737,
+      "step": 35050
+    },
+    {
+      "epoch": 0.7011,
+      "grad_norm": 0.23828125,
+      "learning_rate": 4.966783952028967e-05,
+      "loss": 1.268,
+      "step": 35055
+    },
+    {
+      "epoch": 0.7012,
+      "grad_norm": 0.19140625,
+      "learning_rate": 4.9637679836423924e-05,
+      "loss": 1.2917,
+      "step": 35060
+    },
+    {
+      "epoch": 0.7013,
+      "grad_norm": 0.212890625,
+      "learning_rate": 4.960752628905412e-05,
+      "loss": 1.2711,
+      "step": 35065
+    },
+    {
+      "epoch": 0.7014,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.957737888185439e-05,
+      "loss": 1.2864,
+      "step": 35070
+    },
+    {
+      "epoch": 0.7015,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 4.9547237618498085e-05,
+      "loss": 1.2727,
+      "step": 35075
+    },
+    {
+      "epoch": 0.7016,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.9517102502657845e-05,
+      "loss": 1.2732,
+      "step": 35080
+    },
+    {
+      "epoch": 0.7017,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 4.9486973538005535e-05,
+      "loss": 1.2676,
+      "step": 35085
+    },
+    {
+      "epoch": 0.7018,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.945685072821227e-05,
+      "loss": 1.2652,
+      "step": 35090
+    },
+    {
+      "epoch": 0.7019,
+      "grad_norm": 0.1640625,
+      "learning_rate": 4.9426734076948436e-05,
+      "loss": 1.2885,
+      "step": 35095
+    },
+    {
+      "epoch": 0.702,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 4.939662358788364e-05,
+      "loss": 1.2767,
+      "step": 35100
+    },
+    {
+      "epoch": 0.7021,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 4.9366519264686725e-05,
+      "loss": 1.2837,
+      "step": 35105
+    },
+    {
+      "epoch": 0.7022,
+      "grad_norm": 0.24609375,
+      "learning_rate": 4.933642111102594e-05,
+      "loss": 1.2736,
+      "step": 35110
+    },
+    {
+      "epoch": 0.7023,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.9306329130568474e-05,
+      "loss": 1.29,
+      "step": 35115
+    },
+    {
+      "epoch": 0.7024,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 4.927624332698109e-05,
+      "loss": 1.2665,
+      "step": 35120
+    },
+    {
+      "epoch": 0.7025,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 4.924616370392961e-05,
+      "loss": 1.2979,
+      "step": 35125
+    },
+    {
+      "epoch": 0.7026,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.921609026507907e-05,
+      "loss": 1.2815,
+      "step": 35130
+    },
+    {
+      "epoch": 0.7027,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 4.918602301409395e-05,
+      "loss": 1.28,
+      "step": 35135
+    },
+    {
+      "epoch": 0.7028,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.915596195463773e-05,
+      "loss": 1.258,
+      "step": 35140
+    },
+    {
+      "epoch": 0.7029,
+      "grad_norm": 0.28515625,
+      "learning_rate": 4.912590709037335e-05,
+      "loss": 1.2704,
+      "step": 35145
+    },
+    {
+      "epoch": 0.703,
+      "grad_norm": 0.259765625,
+      "learning_rate": 4.909585842496287e-05,
+      "loss": 1.2727,
+      "step": 35150
+    },
+    {
+      "epoch": 0.7031,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.906581596206764e-05,
+      "loss": 1.2767,
+      "step": 35155
+    },
+    {
+      "epoch": 0.7032,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 4.9035779705348226e-05,
+      "loss": 1.2765,
+      "step": 35160
+    },
+    {
+      "epoch": 0.7033,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.900574965846447e-05,
+      "loss": 1.2747,
+      "step": 35165
+    },
+    {
+      "epoch": 0.7034,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.8975725825075435e-05,
+      "loss": 1.2834,
+      "step": 35170
+    },
+    {
+      "epoch": 0.7035,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 4.894570820883944e-05,
+      "loss": 1.2602,
+      "step": 35175
+    },
+    {
+      "epoch": 0.7036,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.8915696813414026e-05,
+      "loss": 1.2728,
+      "step": 35180
+    },
+    {
+      "epoch": 0.7037,
+      "grad_norm": 0.23046875,
+      "learning_rate": 4.888569164245601e-05,
+      "loss": 1.2751,
+      "step": 35185
+    },
+    {
+      "epoch": 0.7038,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.885569269962142e-05,
+      "loss": 1.2931,
+      "step": 35190
+    },
+    {
+      "epoch": 0.7039,
+      "grad_norm": 0.3828125,
+      "learning_rate": 4.8825699988565485e-05,
+      "loss": 1.2816,
+      "step": 35195
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 0.36328125,
+      "learning_rate": 4.8795713512942865e-05,
+      "loss": 1.2784,
+      "step": 35200
+    },
+    {
+      "epoch": 0.7041,
+      "grad_norm": 0.259765625,
+      "learning_rate": 4.8765733276407156e-05,
+      "loss": 1.2817,
+      "step": 35205
+    },
+    {
+      "epoch": 0.7042,
+      "grad_norm": 0.275390625,
+      "learning_rate": 4.8735759282611516e-05,
+      "loss": 1.2672,
+      "step": 35210
+    },
+    {
+      "epoch": 0.7043,
+      "grad_norm": 0.20703125,
+      "learning_rate": 4.870579153520807e-05,
+      "loss": 1.2655,
+      "step": 35215
+    },
+    {
+      "epoch": 0.7044,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.867583003784829e-05,
+      "loss": 1.2696,
+      "step": 35220
+    },
+    {
+      "epoch": 0.7045,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.864587479418302e-05,
+      "loss": 1.2848,
+      "step": 35225
+    },
+    {
+      "epoch": 0.7046,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.861592580786205e-05,
+      "loss": 1.2809,
+      "step": 35230
+    },
+    {
+      "epoch": 0.7047,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.858598308253473e-05,
+      "loss": 1.2724,
+      "step": 35235
+    },
+    {
+      "epoch": 0.7048,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 4.8556046621849346e-05,
+      "loss": 1.2697,
+      "step": 35240
+    },
+    {
+      "epoch": 0.7049,
+      "grad_norm": 0.310546875,
+      "learning_rate": 4.852611642945368e-05,
+      "loss": 1.2898,
+      "step": 35245
+    },
+    {
+      "epoch": 0.705,
+      "grad_norm": 0.212890625,
+      "learning_rate": 4.8496192508994576e-05,
+      "loss": 1.2704,
+      "step": 35250
+    },
+    {
+      "epoch": 0.7051,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.84662748641182e-05,
+      "loss": 1.2837,
+      "step": 35255
+    },
+    {
+      "epoch": 0.7052,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.8436363498469906e-05,
+      "loss": 1.2854,
+      "step": 35260
+    },
+    {
+      "epoch": 0.7053,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.840645841569431e-05,
+      "loss": 1.2634,
+      "step": 35265
+    },
+    {
+      "epoch": 0.7054,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.837655961943526e-05,
+      "loss": 1.2759,
+      "step": 35270
+    },
+    {
+      "epoch": 0.7055,
+      "grad_norm": 0.3203125,
+      "learning_rate": 4.834666711333582e-05,
+      "loss": 1.2697,
+      "step": 35275
+    },
+    {
+      "epoch": 0.7056,
+      "grad_norm": 0.388671875,
+      "learning_rate": 4.8316780901038314e-05,
+      "loss": 1.2798,
+      "step": 35280
+    },
+    {
+      "epoch": 0.7057,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.828690098618429e-05,
+      "loss": 1.2992,
+      "step": 35285
+    },
+    {
+      "epoch": 0.7058,
+      "grad_norm": 0.283203125,
+      "learning_rate": 4.825702737241452e-05,
+      "loss": 1.2741,
+      "step": 35290
+    },
+    {
+      "epoch": 0.7059,
+      "grad_norm": 0.294921875,
+      "learning_rate": 4.822716006336897e-05,
+      "loss": 1.2667,
+      "step": 35295
+    },
+    {
+      "epoch": 0.706,
+      "grad_norm": 0.298828125,
+      "learning_rate": 4.8197299062686995e-05,
+      "loss": 1.2794,
+      "step": 35300
+    },
+    {
+      "epoch": 0.7061,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.816744437400697e-05,
+      "loss": 1.2568,
+      "step": 35305
+    },
+    {
+      "epoch": 0.7062,
+      "grad_norm": 0.388671875,
+      "learning_rate": 4.8137596000966614e-05,
+      "loss": 1.2733,
+      "step": 35310
+    },
+    {
+      "epoch": 0.7063,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.810775394720286e-05,
+      "loss": 1.2643,
+      "step": 35315
+    },
+    {
+      "epoch": 0.7064,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 4.807791821635186e-05,
+      "loss": 1.2781,
+      "step": 35320
+    },
+    {
+      "epoch": 0.7065,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 4.8048088812049096e-05,
+      "loss": 1.2782,
+      "step": 35325
+    },
+    {
+      "epoch": 0.7066,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 4.8018265737929044e-05,
+      "loss": 1.2884,
+      "step": 35330
+    },
+    {
+      "epoch": 0.7067,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.798844899762568e-05,
+      "loss": 1.2597,
+      "step": 35335
+    },
+    {
+      "epoch": 0.7068,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.7958638594772064e-05,
+      "loss": 1.2685,
+      "step": 35340
+    },
+    {
+      "epoch": 0.7069,
+      "grad_norm": 0.25390625,
+      "learning_rate": 4.792883453300042e-05,
+      "loss": 1.2715,
+      "step": 35345
+    },
+    {
+      "epoch": 0.707,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 4.78990368159424e-05,
+      "loss": 1.2606,
+      "step": 35350
+    },
+    {
+      "epoch": 0.7071,
+      "grad_norm": 0.224609375,
+      "learning_rate": 4.786924544722864e-05,
+      "loss": 1.2679,
+      "step": 35355
+    },
+    {
+      "epoch": 0.7072,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.783946043048923e-05,
+      "loss": 1.2692,
+      "step": 35360
+    },
+    {
+      "epoch": 0.7073,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.780968176935333e-05,
+      "loss": 1.282,
+      "step": 35365
+    },
+    {
+      "epoch": 0.7074,
+      "grad_norm": 0.1328125,
+      "learning_rate": 4.7779909467449414e-05,
+      "loss": 1.2664,
+      "step": 35370
+    },
+    {
+      "epoch": 0.7075,
+      "grad_norm": 0.255859375,
+      "learning_rate": 4.7750143528405126e-05,
+      "loss": 1.3035,
+      "step": 35375
+    },
+    {
+      "epoch": 0.7076,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 4.7720383955847345e-05,
+      "loss": 1.2724,
+      "step": 35380
+    },
+    {
+      "epoch": 0.7077,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.769063075340222e-05,
+      "loss": 1.2845,
+      "step": 35385
+    },
+    {
+      "epoch": 0.7078,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.766088392469506e-05,
+      "loss": 1.2648,
+      "step": 35390
+    },
+    {
+      "epoch": 0.7079,
+      "grad_norm": 0.404296875,
+      "learning_rate": 4.763114347335043e-05,
+      "loss": 1.2724,
+      "step": 35395
+    },
+    {
+      "epoch": 0.708,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 4.7601409402992106e-05,
+      "loss": 1.2671,
+      "step": 35400
+    },
+    {
+      "epoch": 0.7081,
+      "grad_norm": 0.3203125,
+      "learning_rate": 4.757168171724311e-05,
+      "loss": 1.2502,
+      "step": 35405
+    },
+    {
+      "epoch": 0.7082,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.7541960419725626e-05,
+      "loss": 1.2763,
+      "step": 35410
+    },
+    {
+      "epoch": 0.7083,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.7512245514061225e-05,
+      "loss": 1.2709,
+      "step": 35415
+    },
+    {
+      "epoch": 0.7084,
+      "grad_norm": 0.26953125,
+      "learning_rate": 4.748253700387042e-05,
+      "loss": 1.286,
+      "step": 35420
+    },
+    {
+      "epoch": 0.7085,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.745283489277325e-05,
+      "loss": 1.2709,
+      "step": 35425
+    },
+    {
+      "epoch": 0.7086,
+      "grad_norm": 0.291015625,
+      "learning_rate": 4.742313918438872e-05,
+      "loss": 1.2765,
+      "step": 35430
+    },
+    {
+      "epoch": 0.7087,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 4.739344988233516e-05,
+      "loss": 1.2603,
+      "step": 35435
+    },
+    {
+      "epoch": 0.7088,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 4.736376699023023e-05,
+      "loss": 1.2903,
+      "step": 35440
+    },
+    {
+      "epoch": 0.7089,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.7334090511690554e-05,
+      "loss": 1.2645,
+      "step": 35445
+    },
+    {
+      "epoch": 0.709,
+      "grad_norm": 0.287109375,
+      "learning_rate": 4.7304420450332244e-05,
+      "loss": 1.2768,
+      "step": 35450
+    },
+    {
+      "epoch": 0.7091,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 4.7274756809770446e-05,
+      "loss": 1.275,
+      "step": 35455
+    },
+    {
+      "epoch": 0.7092,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.724509959361961e-05,
+      "loss": 1.275,
+      "step": 35460
+    },
+    {
+      "epoch": 0.7093,
+      "grad_norm": 0.2421875,
+      "learning_rate": 4.721544880549337e-05,
+      "loss": 1.2885,
+      "step": 35465
+    },
+    {
+      "epoch": 0.7094,
+      "grad_norm": 0.18359375,
+      "learning_rate": 4.7185804449004565e-05,
+      "loss": 1.2818,
+      "step": 35470
+    },
+    {
+      "epoch": 0.7095,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.71561665277653e-05,
+      "loss": 1.261,
+      "step": 35475
+    },
+    {
+      "epoch": 0.7096,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.712653504538683e-05,
+      "loss": 1.2629,
+      "step": 35480
+    },
+    {
+      "epoch": 0.7097,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.70969100054797e-05,
+      "loss": 1.2805,
+      "step": 35485
+    },
+    {
+      "epoch": 0.7098,
+      "grad_norm": 0.236328125,
+      "learning_rate": 4.706729141165361e-05,
+      "loss": 1.2983,
+      "step": 35490
+    },
+    {
+      "epoch": 0.7099,
+      "grad_norm": 0.26171875,
+      "learning_rate": 4.7037679267517495e-05,
+      "loss": 1.2816,
+      "step": 35495
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 4.700807357667952e-05,
+      "loss": 1.2694,
+      "step": 35500
+    },
+    {
+      "epoch": 0.7101,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 4.697847434274704e-05,
+      "loss": 1.2729,
+      "step": 35505
+    },
+    {
+      "epoch": 0.7102,
+      "grad_norm": 0.25390625,
+      "learning_rate": 4.694888156932658e-05,
+      "loss": 1.2655,
+      "step": 35510
+    },
+    {
+      "epoch": 0.7103,
+      "grad_norm": 0.23828125,
+      "learning_rate": 4.6919295260024054e-05,
+      "loss": 1.2693,
+      "step": 35515
+    },
+    {
+      "epoch": 0.7104,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.688971541844436e-05,
+      "loss": 1.2761,
+      "step": 35520
+    },
+    {
+      "epoch": 0.7105,
+      "grad_norm": 0.1953125,
+      "learning_rate": 4.68601420481917e-05,
+      "loss": 1.284,
+      "step": 35525
+    },
+    {
+      "epoch": 0.7106,
+      "grad_norm": 0.279296875,
+      "learning_rate": 4.6830575152869616e-05,
+      "loss": 1.282,
+      "step": 35530
+    },
+    {
+      "epoch": 0.7107,
+      "grad_norm": 0.212890625,
+      "learning_rate": 4.6801014736080596e-05,
+      "loss": 1.2641,
+      "step": 35535
+    },
+    {
+      "epoch": 0.7108,
+      "grad_norm": 0.314453125,
+      "learning_rate": 4.6771460801426635e-05,
+      "loss": 1.2685,
+      "step": 35540
+    },
+    {
+      "epoch": 0.7109,
+      "grad_norm": 0.28125,
+      "learning_rate": 4.674191335250865e-05,
+      "loss": 1.269,
+      "step": 35545
+    },
+    {
+      "epoch": 0.711,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.6712372392927e-05,
+      "loss": 1.2738,
+      "step": 35550
+    },
+    {
+      "epoch": 0.7111,
+      "grad_norm": 0.181640625,
+      "learning_rate": 4.668283792628114e-05,
+      "loss": 1.2727,
+      "step": 35555
+    },
+    {
+      "epoch": 0.7112,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.665330995616974e-05,
+      "loss": 1.2732,
+      "step": 35560
+    },
+    {
+      "epoch": 0.7113,
+      "grad_norm": 0.2734375,
+      "learning_rate": 4.6623788486190725e-05,
+      "loss": 1.2638,
+      "step": 35565
+    },
+    {
+      "epoch": 0.7114,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.6594273519941154e-05,
+      "loss": 1.274,
+      "step": 35570
+    },
+    {
+      "epoch": 0.7115,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.656476506101737e-05,
+      "loss": 1.2732,
+      "step": 35575
+    },
+    {
+      "epoch": 0.7116,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.653526311301488e-05,
+      "loss": 1.289,
+      "step": 35580
+    },
+    {
+      "epoch": 0.7117,
+      "grad_norm": 0.302734375,
+      "learning_rate": 4.65057676795284e-05,
+      "loss": 1.2745,
+      "step": 35585
+    },
+    {
+      "epoch": 0.7118,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.647627876415186e-05,
+      "loss": 1.2702,
+      "step": 35590
+    },
+    {
+      "epoch": 0.7119,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 4.6446796370478394e-05,
+      "loss": 1.2776,
+      "step": 35595
+    },
+    {
+      "epoch": 0.712,
+      "grad_norm": 0.259765625,
+      "learning_rate": 4.6417320502100316e-05,
+      "loss": 1.2572,
+      "step": 35600
+    },
+    {
+      "epoch": 0.7121,
+      "grad_norm": 0.296875,
+      "learning_rate": 4.6387851162609275e-05,
+      "loss": 1.2761,
+      "step": 35605
+    },
+    {
+      "epoch": 0.7122,
+      "grad_norm": 0.291015625,
+      "learning_rate": 4.6358388355595904e-05,
+      "loss": 1.2736,
+      "step": 35610
+    },
+    {
+      "epoch": 0.7123,
+      "grad_norm": 0.2890625,
+      "learning_rate": 4.632893208465021e-05,
+      "loss": 1.2831,
+      "step": 35615
+    },
+    {
+      "epoch": 0.7124,
+      "grad_norm": 0.365234375,
+      "learning_rate": 4.629948235336133e-05,
+      "loss": 1.2758,
+      "step": 35620
+    },
+    {
+      "epoch": 0.7125,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.6270039165317605e-05,
+      "loss": 1.2856,
+      "step": 35625
+    },
+    {
+      "epoch": 0.7126,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.62406025241067e-05,
+      "loss": 1.2572,
+      "step": 35630
+    },
+    {
+      "epoch": 0.7127,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 4.621117243331523e-05,
+      "loss": 1.2815,
+      "step": 35635
+    },
+    {
+      "epoch": 0.7128,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.6181748896529273e-05,
+      "loss": 1.2834,
+      "step": 35640
+    },
+    {
+      "epoch": 0.7129,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.615233191733398e-05,
+      "loss": 1.2599,
+      "step": 35645
+    },
+    {
+      "epoch": 0.713,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 4.612292149931369e-05,
+      "loss": 1.2827,
+      "step": 35650
+    },
+    {
+      "epoch": 0.7131,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.6093517646052034e-05,
+      "loss": 1.2644,
+      "step": 35655
+    },
+    {
+      "epoch": 0.7132,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.6064120361131656e-05,
+      "loss": 1.2619,
+      "step": 35660
+    },
+    {
+      "epoch": 0.7133,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.603472964813466e-05,
+      "loss": 1.2716,
+      "step": 35665
+    },
+    {
+      "epoch": 0.7134,
+      "grad_norm": 0.314453125,
+      "learning_rate": 4.600534551064215e-05,
+      "loss": 1.2741,
+      "step": 35670
+    },
+    {
+      "epoch": 0.7135,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.59759679522345e-05,
+      "loss": 1.273,
+      "step": 35675
+    },
+    {
+      "epoch": 0.7136,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.5946596976491295e-05,
+      "loss": 1.264,
+      "step": 35680
+    },
+    {
+      "epoch": 0.7137,
+      "grad_norm": 0.2890625,
+      "learning_rate": 4.591723258699127e-05,
+      "loss": 1.2778,
+      "step": 35685
+    },
+    {
+      "epoch": 0.7138,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.588787478731242e-05,
+      "loss": 1.2754,
+      "step": 35690
+    },
+    {
+      "epoch": 0.7139,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 4.5858523581031884e-05,
+      "loss": 1.2611,
+      "step": 35695
+    },
+    {
+      "epoch": 0.714,
+      "grad_norm": 0.3046875,
+      "learning_rate": 4.582917897172603e-05,
+      "loss": 1.2691,
+      "step": 35700
+    },
+    {
+      "epoch": 0.7141,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.579984096297038e-05,
+      "loss": 1.2884,
+      "step": 35705
+    },
+    {
+      "epoch": 0.7142,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.577050955833973e-05,
+      "loss": 1.2722,
+      "step": 35710
+    },
+    {
+      "epoch": 0.7143,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.574118476140794e-05,
+      "loss": 1.2557,
+      "step": 35715
+    },
+    {
+      "epoch": 0.7144,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.5711866575748276e-05,
+      "loss": 1.2662,
+      "step": 35720
+    },
+    {
+      "epoch": 0.7145,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 4.568255500493292e-05,
+      "loss": 1.271,
+      "step": 35725
+    },
+    {
+      "epoch": 0.7146,
+      "grad_norm": 0.193359375,
+      "learning_rate": 4.565325005253356e-05,
+      "loss": 1.2736,
+      "step": 35730
+    },
+    {
+      "epoch": 0.7147,
+      "grad_norm": 0.1640625,
+      "learning_rate": 4.5623951722120736e-05,
+      "loss": 1.2941,
+      "step": 35735
+    },
+    {
+      "epoch": 0.7148,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.559466001726451e-05,
+      "loss": 1.2903,
+      "step": 35740
+    },
+    {
+      "epoch": 0.7149,
+      "grad_norm": 0.294921875,
+      "learning_rate": 4.5565374941533965e-05,
+      "loss": 1.2741,
+      "step": 35745
+    },
+    {
+      "epoch": 0.715,
+      "grad_norm": 0.2109375,
+      "learning_rate": 4.5536096498497295e-05,
+      "loss": 1.2654,
+      "step": 35750
+    },
+    {
+      "epoch": 0.7151,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 4.5506824691722126e-05,
+      "loss": 1.2587,
+      "step": 35755
+    },
+    {
+      "epoch": 0.7152,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.5477559524774994e-05,
+      "loss": 1.2822,
+      "step": 35760
+    },
+    {
+      "epoch": 0.7153,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.5448301001221895e-05,
+      "loss": 1.2782,
+      "step": 35765
+    },
+    {
+      "epoch": 0.7154,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.541904912462784e-05,
+      "loss": 1.2757,
+      "step": 35770
+    },
+    {
+      "epoch": 0.7155,
+      "grad_norm": 0.36328125,
+      "learning_rate": 4.5389803898557106e-05,
+      "loss": 1.2779,
+      "step": 35775
+    },
+    {
+      "epoch": 0.7156,
+      "grad_norm": 0.625,
+      "learning_rate": 4.5360565326573104e-05,
+      "loss": 1.2691,
+      "step": 35780
+    },
+    {
+      "epoch": 0.7157,
+      "grad_norm": 0.2734375,
+      "learning_rate": 4.5331333412238475e-05,
+      "loss": 1.2645,
+      "step": 35785
+    },
+    {
+      "epoch": 0.7158,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.530210815911504e-05,
+      "loss": 1.262,
+      "step": 35790
+    },
+    {
+      "epoch": 0.7159,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 4.527288957076382e-05,
+      "loss": 1.2733,
+      "step": 35795
+    },
+    {
+      "epoch": 0.716,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.524367765074499e-05,
+      "loss": 1.2776,
+      "step": 35800
+    },
+    {
+      "epoch": 0.7161,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.5214472402617944e-05,
+      "loss": 1.2702,
+      "step": 35805
+    },
+    {
+      "epoch": 0.7162,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.518527382994127e-05,
+      "loss": 1.256,
+      "step": 35810
+    },
+    {
+      "epoch": 0.7163,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.515608193627265e-05,
+      "loss": 1.2751,
+      "step": 35815
+    },
+    {
+      "epoch": 0.7164,
+      "grad_norm": 0.3046875,
+      "learning_rate": 4.512689672516918e-05,
+      "loss": 1.2639,
+      "step": 35820
+    },
+    {
+      "epoch": 0.7165,
+      "grad_norm": 0.201171875,
+      "learning_rate": 4.5097718200186814e-05,
+      "loss": 1.2948,
+      "step": 35825
+    },
+    {
+      "epoch": 0.7166,
+      "grad_norm": 0.49609375,
+      "learning_rate": 4.506854636488103e-05,
+      "loss": 1.2694,
+      "step": 35830
+    },
+    {
+      "epoch": 0.7167,
+      "grad_norm": 0.25390625,
+      "learning_rate": 4.50393812228062e-05,
+      "loss": 1.2615,
+      "step": 35835
+    },
+    {
+      "epoch": 0.7168,
+      "grad_norm": 0.2578125,
+      "learning_rate": 4.501022277751602e-05,
+      "loss": 1.27,
+      "step": 35840
+    },
+    {
+      "epoch": 0.7169,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.498107103256346e-05,
+      "loss": 1.2858,
+      "step": 35845
+    },
+    {
+      "epoch": 0.717,
+      "grad_norm": 0.359375,
+      "learning_rate": 4.495192599150044e-05,
+      "loss": 1.2695,
+      "step": 35850
+    },
+    {
+      "epoch": 0.7171,
+      "grad_norm": 0.322265625,
+      "learning_rate": 4.4922787657878294e-05,
+      "loss": 1.2581,
+      "step": 35855
+    },
+    {
+      "epoch": 0.7172,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 4.48936560352474e-05,
+      "loss": 1.2617,
+      "step": 35860
+    },
+    {
+      "epoch": 0.7173,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.4864531127157374e-05,
+      "loss": 1.2908,
+      "step": 35865
+    },
+    {
+      "epoch": 0.7174,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.483541293715698e-05,
+      "loss": 1.2814,
+      "step": 35870
+    },
+    {
+      "epoch": 0.7175,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.480630146879419e-05,
+      "loss": 1.2833,
+      "step": 35875
+    },
+    {
+      "epoch": 0.7176,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.4777196725616146e-05,
+      "loss": 1.2767,
+      "step": 35880
+    },
+    {
+      "epoch": 0.7177,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.474809871116916e-05,
+      "loss": 1.2686,
+      "step": 35885
+    },
+    {
+      "epoch": 0.7178,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 4.471900742899876e-05,
+      "loss": 1.2537,
+      "step": 35890
+    },
+    {
+      "epoch": 0.7179,
+      "grad_norm": 0.283203125,
+      "learning_rate": 4.4689922882649626e-05,
+      "loss": 1.2755,
+      "step": 35895
+    },
+    {
+      "epoch": 0.718,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.46608450756656e-05,
+      "loss": 1.2773,
+      "step": 35900
+    },
+    {
+      "epoch": 0.7181,
+      "grad_norm": 0.296875,
+      "learning_rate": 4.463177401158975e-05,
+      "loss": 1.2744,
+      "step": 35905
+    },
+    {
+      "epoch": 0.7182,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.460270969396429e-05,
+      "loss": 1.2591,
+      "step": 35910
+    },
+    {
+      "epoch": 0.7183,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.457365212633058e-05,
+      "loss": 1.2958,
+      "step": 35915
+    },
+    {
+      "epoch": 0.7184,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.45446013122293e-05,
+      "loss": 1.2671,
+      "step": 35920
+    },
+    {
+      "epoch": 0.7185,
+      "grad_norm": 0.388671875,
+      "learning_rate": 4.451555725520009e-05,
+      "loss": 1.2883,
+      "step": 35925
+    },
+    {
+      "epoch": 0.7186,
+      "grad_norm": 0.314453125,
+      "learning_rate": 4.44865199587819e-05,
+      "loss": 1.2746,
+      "step": 35930
+    },
+    {
+      "epoch": 0.7187,
+      "grad_norm": 0.26953125,
+      "learning_rate": 4.4457489426512947e-05,
+      "loss": 1.2741,
+      "step": 35935
+    },
+    {
+      "epoch": 0.7188,
+      "grad_norm": 0.255859375,
+      "learning_rate": 4.4428465661930343e-05,
+      "loss": 1.2786,
+      "step": 35940
+    },
+    {
+      "epoch": 0.7189,
+      "grad_norm": 0.27734375,
+      "learning_rate": 4.43994486685707e-05,
+      "loss": 1.2751,
+      "step": 35945
+    },
+    {
+      "epoch": 0.719,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.437043844996952e-05,
+      "loss": 1.2683,
+      "step": 35950
+    },
+    {
+      "epoch": 0.7191,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 4.43414350096617e-05,
+      "loss": 1.2763,
+      "step": 35955
+    },
+    {
+      "epoch": 0.7192,
+      "grad_norm": 0.2734375,
+      "learning_rate": 4.431243835118124e-05,
+      "loss": 1.2906,
+      "step": 35960
+    },
+    {
+      "epoch": 0.7193,
+      "grad_norm": 0.26953125,
+      "learning_rate": 4.428344847806116e-05,
+      "loss": 1.2675,
+      "step": 35965
+    },
+    {
+      "epoch": 0.7194,
+      "grad_norm": 0.16015625,
+      "learning_rate": 4.425446539383393e-05,
+      "loss": 1.2676,
+      "step": 35970
+    },
+    {
+      "epoch": 0.7195,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.4225489102030995e-05,
+      "loss": 1.2899,
+      "step": 35975
+    },
+    {
+      "epoch": 0.7196,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.419651960618302e-05,
+      "loss": 1.2767,
+      "step": 35980
+    },
+    {
+      "epoch": 0.7197,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.4167556909819874e-05,
+      "loss": 1.2724,
+      "step": 35985
+    },
+    {
+      "epoch": 0.7198,
+      "grad_norm": 0.26171875,
+      "learning_rate": 4.413860101647055e-05,
+      "loss": 1.2813,
+      "step": 35990
+    },
+    {
+      "epoch": 0.7199,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.4109651929663256e-05,
+      "loss": 1.2701,
+      "step": 35995
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.361328125,
+      "learning_rate": 4.4080709652925336e-05,
+      "loss": 1.2709,
+      "step": 36000
+    },
+    {
+      "epoch": 0.7201,
+      "grad_norm": 0.357421875,
+      "learning_rate": 4.4051774189783315e-05,
+      "loss": 1.277,
+      "step": 36005
+    },
+    {
+      "epoch": 0.7202,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 4.4022845543762915e-05,
+      "loss": 1.2815,
+      "step": 36010
+    },
+    {
+      "epoch": 0.7203,
+      "grad_norm": 0.146484375,
+      "learning_rate": 4.399392371838897e-05,
+      "loss": 1.2772,
+      "step": 36015
+    },
+    {
+      "epoch": 0.7204,
+      "grad_norm": 0.21875,
+      "learning_rate": 4.396500871718555e-05,
+      "loss": 1.2837,
+      "step": 36020
+    },
+    {
+      "epoch": 0.7205,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.393610054367585e-05,
+      "loss": 1.26,
+      "step": 36025
+    },
+    {
+      "epoch": 0.7206,
+      "grad_norm": 0.138671875,
+      "learning_rate": 4.39071992013822e-05,
+      "loss": 1.2657,
+      "step": 36030
+    },
+    {
+      "epoch": 0.7207,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.387830469382624e-05,
+      "loss": 1.2768,
+      "step": 36035
+    },
+    {
+      "epoch": 0.7208,
+      "grad_norm": 0.25390625,
+      "learning_rate": 4.3849417024528564e-05,
+      "loss": 1.2667,
+      "step": 36040
+    },
+    {
+      "epoch": 0.7209,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.382053619700912e-05,
+      "loss": 1.2655,
+      "step": 36045
+    },
+    {
+      "epoch": 0.721,
+      "grad_norm": 0.3125,
+      "learning_rate": 4.379166221478697e-05,
+      "loss": 1.2717,
+      "step": 36050
+    },
+    {
+      "epoch": 0.7211,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.3762795081380215e-05,
+      "loss": 1.2725,
+      "step": 36055
+    },
+    {
+      "epoch": 0.7212,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.3733934800306366e-05,
+      "loss": 1.2892,
+      "step": 36060
+    },
+    {
+      "epoch": 0.7213,
+      "grad_norm": 0.140625,
+      "learning_rate": 4.37050813750818e-05,
+      "loss": 1.2781,
+      "step": 36065
+    },
+    {
+      "epoch": 0.7214,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.367623480922236e-05,
+      "loss": 1.2893,
+      "step": 36070
+    },
+    {
+      "epoch": 0.7215,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.364739510624286e-05,
+      "loss": 1.2586,
+      "step": 36075
+    },
+    {
+      "epoch": 0.7216,
+      "grad_norm": 0.404296875,
+      "learning_rate": 4.361856226965733e-05,
+      "loss": 1.2616,
+      "step": 36080
+    },
+    {
+      "epoch": 0.7217,
+      "grad_norm": 0.380859375,
+      "learning_rate": 4.3589736302978954e-05,
+      "loss": 1.2673,
+      "step": 36085
+    },
+    {
+      "epoch": 0.7218,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.356091720972011e-05,
+      "loss": 1.2742,
+      "step": 36090
+    },
+    {
+      "epoch": 0.7219,
+      "grad_norm": 0.298828125,
+      "learning_rate": 4.3532104993392306e-05,
+      "loss": 1.276,
+      "step": 36095
+    },
+    {
+      "epoch": 0.722,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 4.350329965750621e-05,
+      "loss": 1.2781,
+      "step": 36100
+    },
+    {
+      "epoch": 0.7221,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 4.347450120557169e-05,
+      "loss": 1.2803,
+      "step": 36105
+    },
+    {
+      "epoch": 0.7222,
+      "grad_norm": 0.32421875,
+      "learning_rate": 4.3445709641097745e-05,
+      "loss": 1.2666,
+      "step": 36110
+    },
+    {
+      "epoch": 0.7223,
+      "grad_norm": 0.125,
+      "learning_rate": 4.341692496759252e-05,
+      "loss": 1.2721,
+      "step": 36115
+    },
+    {
+      "epoch": 0.7224,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.3388147188563325e-05,
+      "loss": 1.2836,
+      "step": 36120
+    },
+    {
+      "epoch": 0.7225,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 4.335937630751674e-05,
+      "loss": 1.2593,
+      "step": 36125
+    },
+    {
+      "epoch": 0.7226,
+      "grad_norm": 0.251953125,
+      "learning_rate": 4.333061232795826e-05,
+      "loss": 1.2872,
+      "step": 36130
+    },
+    {
+      "epoch": 0.7227,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 4.3301855253392864e-05,
+      "loss": 1.2726,
+      "step": 36135
+    },
+    {
+      "epoch": 0.7228,
+      "grad_norm": 0.2421875,
+      "learning_rate": 4.327310508732437e-05,
+      "loss": 1.2658,
+      "step": 36140
+    },
+    {
+      "epoch": 0.7229,
+      "grad_norm": 0.279296875,
+      "learning_rate": 4.324436183325593e-05,
+      "loss": 1.2732,
+      "step": 36145
+    },
+    {
+      "epoch": 0.723,
+      "grad_norm": 0.361328125,
+      "learning_rate": 4.32156254946899e-05,
+      "loss": 1.2826,
+      "step": 36150
+    },
+    {
+      "epoch": 0.7231,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 4.3186896075127595e-05,
+      "loss": 1.2763,
+      "step": 36155
+    },
+    {
+      "epoch": 0.7232,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 4.315817357806974e-05,
+      "loss": 1.2616,
+      "step": 36160
+    },
+    {
+      "epoch": 0.7233,
+      "grad_norm": 0.2890625,
+      "learning_rate": 4.3129458007015946e-05,
+      "loss": 1.2573,
+      "step": 36165
+    },
+    {
+      "epoch": 0.7234,
+      "grad_norm": 0.232421875,
+      "learning_rate": 4.310074936546521e-05,
+      "loss": 1.2709,
+      "step": 36170
+    },
+    {
+      "epoch": 0.7235,
+      "grad_norm": 0.208984375,
+      "learning_rate": 4.307204765691558e-05,
+      "loss": 1.2711,
+      "step": 36175
+    },
+    {
+      "epoch": 0.7236,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.304335288486426e-05,
+      "loss": 1.2811,
+      "step": 36180
+    },
+    {
+      "epoch": 0.7237,
+      "grad_norm": 0.19140625,
+      "learning_rate": 4.301466505280762e-05,
+      "loss": 1.2748,
+      "step": 36185
+    },
+    {
+      "epoch": 0.7238,
+      "grad_norm": 0.21484375,
+      "learning_rate": 4.29859841642412e-05,
+      "loss": 1.2978,
+      "step": 36190
+    },
+    {
+      "epoch": 0.7239,
+      "grad_norm": 0.298828125,
+      "learning_rate": 4.295731022265966e-05,
+      "loss": 1.2697,
+      "step": 36195
+    },
+    {
+      "epoch": 0.724,
+      "grad_norm": 0.279296875,
+      "learning_rate": 4.2928643231556844e-05,
+      "loss": 1.2766,
+      "step": 36200
+    },
+    {
+      "epoch": 0.7241,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.289998319442573e-05,
+      "loss": 1.2761,
+      "step": 36205
+    },
+    {
+      "epoch": 0.7242,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 4.287133011475847e-05,
+      "loss": 1.2759,
+      "step": 36210
+    },
+    {
+      "epoch": 0.7243,
+      "grad_norm": 0.166015625,
+      "learning_rate": 4.2842683996046327e-05,
+      "loss": 1.2767,
+      "step": 36215
+    },
+    {
+      "epoch": 0.7244,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 4.2814044841779745e-05,
+      "loss": 1.2664,
+      "step": 36220
+    },
+    {
+      "epoch": 0.7245,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.27854126554484e-05,
+      "loss": 1.2846,
+      "step": 36225
+    },
+    {
+      "epoch": 0.7246,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 4.2756787440540936e-05,
+      "loss": 1.2854,
+      "step": 36230
+    },
+    {
+      "epoch": 0.7247,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 4.2728169200545286e-05,
+      "loss": 1.2679,
+      "step": 36235
+    },
+    {
+      "epoch": 0.7248,
+      "grad_norm": 0.1875,
+      "learning_rate": 4.26995579389485e-05,
+      "loss": 1.2744,
+      "step": 36240
+    },
+    {
+      "epoch": 0.7249,
+      "grad_norm": 0.28515625,
+      "learning_rate": 4.267095365923672e-05,
+      "loss": 1.2784,
+      "step": 36245
+    },
+    {
+      "epoch": 0.725,
+      "grad_norm": 0.2578125,
+      "learning_rate": 4.264235636489542e-05,
+      "loss": 1.2682,
+      "step": 36250
+    },
+    {
+      "epoch": 0.7251,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.261376605940894e-05,
+      "loss": 1.2766,
+      "step": 36255
+    },
+    {
+      "epoch": 0.7252,
+      "grad_norm": 0.37109375,
+      "learning_rate": 4.2585182746261035e-05,
+      "loss": 1.2789,
+      "step": 36260
+    },
+    {
+      "epoch": 0.7253,
+      "grad_norm": 0.177734375,
+      "learning_rate": 4.2556606428934443e-05,
+      "loss": 1.2592,
+      "step": 36265
+    },
+    {
+      "epoch": 0.7254,
+      "grad_norm": 0.21484375,
+      "learning_rate": 4.252803711091112e-05,
+      "loss": 1.2803,
+      "step": 36270
+    },
+    {
+      "epoch": 0.7255,
+      "grad_norm": 0.314453125,
+      "learning_rate": 4.249947479567218e-05,
+      "loss": 1.2753,
+      "step": 36275
+    },
+    {
+      "epoch": 0.7256,
+      "grad_norm": 0.30078125,
+      "learning_rate": 4.2470919486697744e-05,
+      "loss": 1.2693,
+      "step": 36280
+    },
+    {
+      "epoch": 0.7257,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 4.244237118746731e-05,
+      "loss": 1.2747,
+      "step": 36285
+    },
+    {
+      "epoch": 0.7258,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 4.2413829901459344e-05,
+      "loss": 1.2866,
+      "step": 36290
+    },
+    {
+      "epoch": 0.7259,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 4.238529563215153e-05,
+      "loss": 1.2693,
+      "step": 36295
+    },
+    {
+      "epoch": 0.726,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.235676838302068e-05,
+      "loss": 1.2547,
+      "step": 36300
+    },
+    {
+      "epoch": 0.7261,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 4.232824815754276e-05,
+      "loss": 1.2843,
+      "step": 36305
+    },
+    {
+      "epoch": 0.7262,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 4.229973495919286e-05,
+      "loss": 1.2761,
+      "step": 36310
+    },
+    {
+      "epoch": 0.7263,
+      "grad_norm": 0.275390625,
+      "learning_rate": 4.227122879144523e-05,
+      "loss": 1.2726,
+      "step": 36315
+    },
+    {
+      "epoch": 0.7264,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.224272965777326e-05,
+      "loss": 1.2774,
+      "step": 36320
+    },
+    {
+      "epoch": 0.7265,
+      "grad_norm": 0.380859375,
+      "learning_rate": 4.221423756164948e-05,
+      "loss": 1.2751,
+      "step": 36325
+    },
+    {
+      "epoch": 0.7266,
+      "grad_norm": 1.265625,
+      "learning_rate": 4.2185752506545585e-05,
+      "loss": 1.2826,
+      "step": 36330
+    },
+    {
+      "epoch": 0.7267,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 4.215727449593233e-05,
+      "loss": 1.2642,
+      "step": 36335
+    },
+    {
+      "epoch": 0.7268,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.212880353327979e-05,
+      "loss": 1.2623,
+      "step": 36340
+    },
+    {
+      "epoch": 0.7269,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.210033962205694e-05,
+      "loss": 1.2686,
+      "step": 36345
+    },
+    {
+      "epoch": 0.727,
+      "grad_norm": 0.23828125,
+      "learning_rate": 4.207188276573214e-05,
+      "loss": 1.2817,
+      "step": 36350
+    },
+    {
+      "epoch": 0.7271,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 4.204343296777265e-05,
+      "loss": 1.2736,
+      "step": 36355
+    },
+    {
+      "epoch": 0.7272,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.201499023164508e-05,
+      "loss": 1.2738,
+      "step": 36360
+    },
+    {
+      "epoch": 0.7273,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.1986554560815096e-05,
+      "loss": 1.2635,
+      "step": 36365
+    },
+    {
+      "epoch": 0.7274,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.195812595874739e-05,
+      "loss": 1.2778,
+      "step": 36370
+    },
+    {
+      "epoch": 0.7275,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.1929704428906026e-05,
+      "loss": 1.2762,
+      "step": 36375
+    },
+    {
+      "epoch": 0.7276,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.190128997475402e-05,
+      "loss": 1.2923,
+      "step": 36380
+    },
+    {
+      "epoch": 0.7277,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.1872882599753605e-05,
+      "loss": 1.2623,
+      "step": 36385
+    },
+    {
+      "epoch": 0.7278,
+      "grad_norm": 0.2734375,
+      "learning_rate": 4.184448230736613e-05,
+      "loss": 1.2777,
+      "step": 36390
+    },
+    {
+      "epoch": 0.7279,
+      "grad_norm": 0.185546875,
+      "learning_rate": 4.181608910105207e-05,
+      "loss": 1.2682,
+      "step": 36395
+    },
+    {
+      "epoch": 0.728,
+      "grad_norm": 0.3828125,
+      "learning_rate": 4.1787702984271074e-05,
+      "loss": 1.269,
+      "step": 36400
+    },
+    {
+      "epoch": 0.7281,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.175932396048188e-05,
+      "loss": 1.2768,
+      "step": 36405
+    },
+    {
+      "epoch": 0.7282,
+      "grad_norm": 0.384765625,
+      "learning_rate": 4.173095203314241e-05,
+      "loss": 1.2729,
+      "step": 36410
+    },
+    {
+      "epoch": 0.7283,
+      "grad_norm": 0.244140625,
+      "learning_rate": 4.170258720570968e-05,
+      "loss": 1.2687,
+      "step": 36415
+    },
+    {
+      "epoch": 0.7284,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 4.167422948163986e-05,
+      "loss": 1.271,
+      "step": 36420
+    },
+    {
+      "epoch": 0.7285,
+      "grad_norm": 0.251953125,
+      "learning_rate": 4.1645878864388266e-05,
+      "loss": 1.2474,
+      "step": 36425
+    },
+    {
+      "epoch": 0.7286,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.161753535740932e-05,
+      "loss": 1.2699,
+      "step": 36430
+    },
+    {
+      "epoch": 0.7287,
+      "grad_norm": 0.1953125,
+      "learning_rate": 4.158919896415656e-05,
+      "loss": 1.273,
+      "step": 36435
+    },
+    {
+      "epoch": 0.7288,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.15608696880828e-05,
+      "loss": 1.2561,
+      "step": 36440
+    },
+    {
+      "epoch": 0.7289,
+      "grad_norm": 0.28515625,
+      "learning_rate": 4.153254753263974e-05,
+      "loss": 1.2717,
+      "step": 36445
+    },
+    {
+      "epoch": 0.729,
+      "grad_norm": 0.25,
+      "learning_rate": 4.150423250127845e-05,
+      "loss": 1.2642,
+      "step": 36450
+    },
+    {
+      "epoch": 0.7291,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.1475924597449024e-05,
+      "loss": 1.2686,
+      "step": 36455
+    },
+    {
+      "epoch": 0.7292,
+      "grad_norm": 0.197265625,
+      "learning_rate": 4.144762382460059e-05,
+      "loss": 1.257,
+      "step": 36460
+    },
+    {
+      "epoch": 0.7293,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.141933018618165e-05,
+      "loss": 1.2794,
+      "step": 36465
+    },
+    {
+      "epoch": 0.7294,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.1391043685639576e-05,
+      "loss": 1.2684,
+      "step": 36470
+    },
+    {
+      "epoch": 0.7295,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.1362764326421064e-05,
+      "loss": 1.2698,
+      "step": 36475
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 4.133449211197188e-05,
+      "loss": 1.2618,
+      "step": 36480
+    },
+    {
+      "epoch": 0.7297,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.130622704573685e-05,
+      "loss": 1.2791,
+      "step": 36485
+    },
+    {
+      "epoch": 0.7298,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.1277969131160045e-05,
+      "loss": 1.2613,
+      "step": 36490
+    },
+    {
+      "epoch": 0.7299,
+      "grad_norm": 0.291015625,
+      "learning_rate": 4.1249718371684564e-05,
+      "loss": 1.2807,
+      "step": 36495
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.25390625,
+      "learning_rate": 4.12214747707527e-05,
+      "loss": 1.2763,
+      "step": 36500
+    },
+    {
+      "epoch": 0.7301,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.1193238331805826e-05,
+      "loss": 1.2862,
+      "step": 36505
+    },
+    {
+      "epoch": 0.7302,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.11650090582845e-05,
+      "loss": 1.2745,
+      "step": 36510
+    },
+    {
+      "epoch": 0.7303,
+      "grad_norm": 0.373046875,
+      "learning_rate": 4.1136786953628334e-05,
+      "loss": 1.2946,
+      "step": 36515
+    },
+    {
+      "epoch": 0.7304,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.110857202127615e-05,
+      "loss": 1.313,
+      "step": 36520
+    },
+    {
+      "epoch": 0.7305,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 4.1080364264665774e-05,
+      "loss": 1.2719,
+      "step": 36525
+    },
+    {
+      "epoch": 0.7306,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 4.1052163687234366e-05,
+      "loss": 1.271,
+      "step": 36530
+    },
+    {
+      "epoch": 0.7307,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.1023970292417935e-05,
+      "loss": 1.2663,
+      "step": 36535
+    },
+    {
+      "epoch": 0.7308,
+      "grad_norm": 0.318359375,
+      "learning_rate": 4.099578408365191e-05,
+      "loss": 1.2722,
+      "step": 36540
+    },
+    {
+      "epoch": 0.7309,
+      "grad_norm": 0.310546875,
+      "learning_rate": 4.096760506437057e-05,
+      "loss": 1.2871,
+      "step": 36545
+    },
+    {
+      "epoch": 0.731,
+      "grad_norm": 0.1953125,
+      "learning_rate": 4.093943323800745e-05,
+      "loss": 1.267,
+      "step": 36550
+    },
+    {
+      "epoch": 0.7311,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.0911268607995325e-05,
+      "loss": 1.2776,
+      "step": 36555
+    },
+    {
+      "epoch": 0.7312,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.08831111777658e-05,
+      "loss": 1.2819,
+      "step": 36560
+    },
+    {
+      "epoch": 0.7313,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.08549609507499e-05,
+      "loss": 1.2609,
+      "step": 36565
+    },
+    {
+      "epoch": 0.7314,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.08268179303776e-05,
+      "loss": 1.2623,
+      "step": 36570
+    },
+    {
+      "epoch": 0.7315,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.0798682120078044e-05,
+      "loss": 1.3035,
+      "step": 36575
+    },
+    {
+      "epoch": 0.7316,
+      "grad_norm": 0.158203125,
+      "learning_rate": 4.077055352327953e-05,
+      "loss": 1.2993,
+      "step": 36580
+    },
+    {
+      "epoch": 0.7317,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.074243214340934e-05,
+      "loss": 1.2792,
+      "step": 36585
+    },
+    {
+      "epoch": 0.7318,
+      "grad_norm": 0.248046875,
+      "learning_rate": 4.071431798389408e-05,
+      "loss": 1.2828,
+      "step": 36590
+    },
+    {
+      "epoch": 0.7319,
+      "grad_norm": 0.177734375,
+      "learning_rate": 4.068621104815934e-05,
+      "loss": 1.2826,
+      "step": 36595
+    },
+    {
+      "epoch": 0.732,
+      "grad_norm": 0.232421875,
+      "learning_rate": 4.065811133962987e-05,
+      "loss": 1.2878,
+      "step": 36600
+    },
+    {
+      "epoch": 0.7321,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.063001886172952e-05,
+      "loss": 1.2718,
+      "step": 36605
+    },
+    {
+      "epoch": 0.7322,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 4.0601933617881294e-05,
+      "loss": 1.2825,
+      "step": 36610
+    },
+    {
+      "epoch": 0.7323,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.057385561150727e-05,
+      "loss": 1.2727,
+      "step": 36615
+    },
+    {
+      "epoch": 0.7324,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.05457848460287e-05,
+      "loss": 1.2601,
+      "step": 36620
+    },
+    {
+      "epoch": 0.7325,
+      "grad_norm": 0.203125,
+      "learning_rate": 4.0517721324865884e-05,
+      "loss": 1.2783,
+      "step": 36625
+    },
+    {
+      "epoch": 0.7326,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.048966505143831e-05,
+      "loss": 1.2664,
+      "step": 36630
+    },
+    {
+      "epoch": 0.7327,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.0461616029164526e-05,
+      "loss": 1.2741,
+      "step": 36635
+    },
+    {
+      "epoch": 0.7328,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.0433574261462206e-05,
+      "loss": 1.2936,
+      "step": 36640
+    },
+    {
+      "epoch": 0.7329,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.040553975174823e-05,
+      "loss": 1.2666,
+      "step": 36645
+    },
+    {
+      "epoch": 0.733,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.037751250343841e-05,
+      "loss": 1.2674,
+      "step": 36650
+    },
+    {
+      "epoch": 0.7331,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.0349492519947904e-05,
+      "loss": 1.2671,
+      "step": 36655
+    },
+    {
+      "epoch": 0.7332,
+      "grad_norm": 0.205078125,
+      "learning_rate": 4.032147980469072e-05,
+      "loss": 1.2698,
+      "step": 36660
+    },
+    {
+      "epoch": 0.7333,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 4.0293474361080244e-05,
+      "loss": 1.2978,
+      "step": 36665
+    },
+    {
+      "epoch": 0.7334,
+      "grad_norm": 0.310546875,
+      "learning_rate": 4.026547619252883e-05,
+      "loss": 1.2585,
+      "step": 36670
+    },
+    {
+      "epoch": 0.7335,
+      "grad_norm": 0.189453125,
+      "learning_rate": 4.023748530244789e-05,
+      "loss": 1.2886,
+      "step": 36675
+    },
+    {
+      "epoch": 0.7336,
+      "grad_norm": 0.130859375,
+      "learning_rate": 4.020950169424815e-05,
+      "loss": 1.2744,
+      "step": 36680
+    },
+    {
+      "epoch": 0.7337,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.018152537133919e-05,
+      "loss": 1.2744,
+      "step": 36685
+    },
+    {
+      "epoch": 0.7338,
+      "grad_norm": 0.28125,
+      "learning_rate": 4.015355633712996e-05,
+      "loss": 1.2876,
+      "step": 36690
+    },
+    {
+      "epoch": 0.7339,
+      "grad_norm": 0.19921875,
+      "learning_rate": 4.012559459502835e-05,
+      "loss": 1.2804,
+      "step": 36695
+    },
+    {
+      "epoch": 0.734,
+      "grad_norm": 0.1640625,
+      "learning_rate": 4.009764014844143e-05,
+      "loss": 1.2675,
+      "step": 36700
+    },
+    {
+      "epoch": 0.7341,
+      "grad_norm": 0.255859375,
+      "learning_rate": 4.006969300077534e-05,
+      "loss": 1.2627,
+      "step": 36705
+    },
+    {
+      "epoch": 0.7342,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.004175315543538e-05,
+      "loss": 1.269,
+      "step": 36710
+    },
+    {
+      "epoch": 0.7343,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.001382061582593e-05,
+      "loss": 1.2807,
+      "step": 36715
+    },
+    {
+      "epoch": 0.7344,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.9985895385350456e-05,
+      "loss": 1.2802,
+      "step": 36720
+    },
+    {
+      "epoch": 0.7345,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.9957977467411615e-05,
+      "loss": 1.2683,
+      "step": 36725
+    },
+    {
+      "epoch": 0.7346,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.9930066865411075e-05,
+      "loss": 1.2649,
+      "step": 36730
+    },
+    {
+      "epoch": 0.7347,
+      "grad_norm": 0.1640625,
+      "learning_rate": 3.990216358274969e-05,
+      "loss": 1.2728,
+      "step": 36735
+    },
+    {
+      "epoch": 0.7348,
+      "grad_norm": 1.3125,
+      "learning_rate": 3.987426762282733e-05,
+      "loss": 1.2657,
+      "step": 36740
+    },
+    {
+      "epoch": 0.7349,
+      "grad_norm": 0.25,
+      "learning_rate": 3.9846378989043156e-05,
+      "loss": 1.2562,
+      "step": 36745
+    },
+    {
+      "epoch": 0.735,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.981849768479517e-05,
+      "loss": 1.2688,
+      "step": 36750
+    },
+    {
+      "epoch": 0.7351,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.979062371348075e-05,
+      "loss": 1.299,
+      "step": 36755
+    },
+    {
+      "epoch": 0.7352,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.976275707849616e-05,
+      "loss": 1.2511,
+      "step": 36760
+    },
+    {
+      "epoch": 0.7353,
+      "grad_norm": 0.1875,
+      "learning_rate": 3.973489778323688e-05,
+      "loss": 1.2794,
+      "step": 36765
+    },
+    {
+      "epoch": 0.7354,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.9707045831097555e-05,
+      "loss": 1.2655,
+      "step": 36770
+    },
+    {
+      "epoch": 0.7355,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.967920122547175e-05,
+      "loss": 1.266,
+      "step": 36775
+    },
+    {
+      "epoch": 0.7356,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 3.9651363969752344e-05,
+      "loss": 1.2767,
+      "step": 36780
+    },
+    {
+      "epoch": 0.7357,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 3.962353406733117e-05,
+      "loss": 1.2684,
+      "step": 36785
+    },
+    {
+      "epoch": 0.7358,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.9595711521599224e-05,
+      "loss": 1.2791,
+      "step": 36790
+    },
+    {
+      "epoch": 0.7359,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 3.956789633594661e-05,
+      "loss": 1.2815,
+      "step": 36795
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.954008851376252e-05,
+      "loss": 1.2738,
+      "step": 36800
+    },
+    {
+      "epoch": 0.7361,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 3.9512288058435256e-05,
+      "loss": 1.2552,
+      "step": 36805
+    },
+    {
+      "epoch": 0.7362,
+      "grad_norm": 0.298828125,
+      "learning_rate": 3.948449497335219e-05,
+      "loss": 1.2786,
+      "step": 36810
+    },
+    {
+      "epoch": 0.7363,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.945670926189987e-05,
+      "loss": 1.2803,
+      "step": 36815
+    },
+    {
+      "epoch": 0.7364,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.942893092746387e-05,
+      "loss": 1.2762,
+      "step": 36820
+    },
+    {
+      "epoch": 0.7365,
+      "grad_norm": 0.177734375,
+      "learning_rate": 3.940115997342891e-05,
+      "loss": 1.2774,
+      "step": 36825
+    },
+    {
+      "epoch": 0.7366,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.9373396403178786e-05,
+      "loss": 1.2639,
+      "step": 36830
+    },
+    {
+      "epoch": 0.7367,
+      "grad_norm": 0.330078125,
+      "learning_rate": 3.9345640220096417e-05,
+      "loss": 1.2698,
+      "step": 36835
+    },
+    {
+      "epoch": 0.7368,
+      "grad_norm": 0.271484375,
+      "learning_rate": 3.931789142756377e-05,
+      "loss": 1.2745,
+      "step": 36840
+    },
+    {
+      "epoch": 0.7369,
+      "grad_norm": 0.28515625,
+      "learning_rate": 3.9290150028962044e-05,
+      "loss": 1.3118,
+      "step": 36845
+    },
+    {
+      "epoch": 0.737,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.9262416027671356e-05,
+      "loss": 1.2741,
+      "step": 36850
+    },
+    {
+      "epoch": 0.7371,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 3.9234689427071006e-05,
+      "loss": 1.2548,
+      "step": 36855
+    },
+    {
+      "epoch": 0.7372,
+      "grad_norm": 0.373046875,
+      "learning_rate": 3.920697023053949e-05,
+      "loss": 1.2549,
+      "step": 36860
+    },
+    {
+      "epoch": 0.7373,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.917925844145418e-05,
+      "loss": 1.2782,
+      "step": 36865
+    },
+    {
+      "epoch": 0.7374,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.915155406319181e-05,
+      "loss": 1.2817,
+      "step": 36870
+    },
+    {
+      "epoch": 0.7375,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.9123857099127936e-05,
+      "loss": 1.2573,
+      "step": 36875
+    },
+    {
+      "epoch": 0.7376,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 3.9096167552637454e-05,
+      "loss": 1.2781,
+      "step": 36880
+    },
+    {
+      "epoch": 0.7377,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.9068485427094205e-05,
+      "loss": 1.2966,
+      "step": 36885
+    },
+    {
+      "epoch": 0.7378,
+      "grad_norm": 0.24609375,
+      "learning_rate": 3.904081072587119e-05,
+      "loss": 1.2569,
+      "step": 36890
+    },
+    {
+      "epoch": 0.7379,
+      "grad_norm": 0.474609375,
+      "learning_rate": 3.9013143452340475e-05,
+      "loss": 1.2872,
+      "step": 36895
+    },
+    {
+      "epoch": 0.738,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.8985483609873244e-05,
+      "loss": 1.2622,
+      "step": 36900
+    },
+    {
+      "epoch": 0.7381,
+      "grad_norm": 0.287109375,
+      "learning_rate": 3.895783120183976e-05,
+      "loss": 1.2675,
+      "step": 36905
+    },
+    {
+      "epoch": 0.7382,
+      "grad_norm": 0.30078125,
+      "learning_rate": 3.893018623160938e-05,
+      "loss": 1.2765,
+      "step": 36910
+    },
+    {
+      "epoch": 0.7383,
+      "grad_norm": 0.2421875,
+      "learning_rate": 3.890254870255055e-05,
+      "loss": 1.2599,
+      "step": 36915
+    },
+    {
+      "epoch": 0.7384,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.887491861803085e-05,
+      "loss": 1.2683,
+      "step": 36920
+    },
+    {
+      "epoch": 0.7385,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.8847295981416896e-05,
+      "loss": 1.2836,
+      "step": 36925
+    },
+    {
+      "epoch": 0.7386,
+      "grad_norm": 0.322265625,
+      "learning_rate": 3.88196807960744e-05,
+      "loss": 1.3053,
+      "step": 36930
+    },
+    {
+      "epoch": 0.7387,
+      "grad_norm": 0.18359375,
+      "learning_rate": 3.879207306536829e-05,
+      "loss": 1.2854,
+      "step": 36935
+    },
+    {
+      "epoch": 0.7388,
+      "grad_norm": 0.1640625,
+      "learning_rate": 3.876447279266238e-05,
+      "loss": 1.2676,
+      "step": 36940
+    },
+    {
+      "epoch": 0.7389,
+      "grad_norm": 0.244140625,
+      "learning_rate": 3.8736879981319695e-05,
+      "loss": 1.2867,
+      "step": 36945
+    },
+    {
+      "epoch": 0.739,
+      "grad_norm": 0.384765625,
+      "learning_rate": 3.8709294634702376e-05,
+      "loss": 1.2712,
+      "step": 36950
+    },
+    {
+      "epoch": 0.7391,
+      "grad_norm": 0.375,
+      "learning_rate": 3.868171675617155e-05,
+      "loss": 1.2764,
+      "step": 36955
+    },
+    {
+      "epoch": 0.7392,
+      "grad_norm": 0.275390625,
+      "learning_rate": 3.8654146349087606e-05,
+      "loss": 1.2747,
+      "step": 36960
+    },
+    {
+      "epoch": 0.7393,
+      "grad_norm": 0.216796875,
+      "learning_rate": 3.862658341680977e-05,
+      "loss": 1.2644,
+      "step": 36965
+    },
+    {
+      "epoch": 0.7394,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 3.859902796269663e-05,
+      "loss": 1.2808,
+      "step": 36970
+    },
+    {
+      "epoch": 0.7395,
+      "grad_norm": 0.181640625,
+      "learning_rate": 3.857147999010567e-05,
+      "loss": 1.2807,
+      "step": 36975
+    },
+    {
+      "epoch": 0.7396,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.854393950239355e-05,
+      "loss": 1.2673,
+      "step": 36980
+    },
+    {
+      "epoch": 0.7397,
+      "grad_norm": 0.326171875,
+      "learning_rate": 3.851640650291603e-05,
+      "loss": 1.2594,
+      "step": 36985
+    },
+    {
+      "epoch": 0.7398,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.8488880995027786e-05,
+      "loss": 1.264,
+      "step": 36990
+    },
+    {
+      "epoch": 0.7399,
+      "grad_norm": 1.1015625,
+      "learning_rate": 3.846136298208285e-05,
+      "loss": 1.275,
+      "step": 36995
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.843385246743417e-05,
+      "loss": 1.2745,
+      "step": 37000
+    },
+    {
+      "epoch": 0.7401,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.840634945443382e-05,
+      "loss": 1.2706,
+      "step": 37005
+    },
+    {
+      "epoch": 0.7402,
+      "grad_norm": 0.240234375,
+      "learning_rate": 3.837885394643296e-05,
+      "loss": 1.2731,
+      "step": 37010
+    },
+    {
+      "epoch": 0.7403,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.835136594678183e-05,
+      "loss": 1.293,
+      "step": 37015
+    },
+    {
+      "epoch": 0.7404,
+      "grad_norm": 0.25,
+      "learning_rate": 3.832388545882976e-05,
+      "loss": 1.2596,
+      "step": 37020
+    },
+    {
+      "epoch": 0.7405,
+      "grad_norm": 0.25,
+      "learning_rate": 3.829641248592515e-05,
+      "loss": 1.2746,
+      "step": 37025
+    },
+    {
+      "epoch": 0.7406,
+      "grad_norm": 0.25390625,
+      "learning_rate": 3.8268947031415514e-05,
+      "loss": 1.2717,
+      "step": 37030
+    },
+    {
+      "epoch": 0.7407,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.824148909864744e-05,
+      "loss": 1.2517,
+      "step": 37035
+    },
+    {
+      "epoch": 0.7408,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 3.821403869096658e-05,
+      "loss": 1.2594,
+      "step": 37040
+    },
+    {
+      "epoch": 0.7409,
+      "grad_norm": 0.205078125,
+      "learning_rate": 3.818659581171766e-05,
+      "loss": 1.2792,
+      "step": 37045
+    },
+    {
+      "epoch": 0.741,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.8159160464244606e-05,
+      "loss": 1.2614,
+      "step": 37050
+    },
+    {
+      "epoch": 0.7411,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.81317326518902e-05,
+      "loss": 1.2777,
+      "step": 37055
+    },
+    {
+      "epoch": 0.7412,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.8104312377996564e-05,
+      "loss": 1.2806,
+      "step": 37060
+    },
+    {
+      "epoch": 0.7413,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 3.807689964590466e-05,
+      "loss": 1.2724,
+      "step": 37065
+    },
+    {
+      "epoch": 0.7414,
+      "grad_norm": 0.287109375,
+      "learning_rate": 3.8049494458954725e-05,
+      "loss": 1.2657,
+      "step": 37070
+    },
+    {
+      "epoch": 0.7415,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.802209682048602e-05,
+      "loss": 1.2636,
+      "step": 37075
+    },
+    {
+      "epoch": 0.7416,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.799470673383674e-05,
+      "loss": 1.2784,
+      "step": 37080
+    },
+    {
+      "epoch": 0.7417,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.796732420234443e-05,
+      "loss": 1.2804,
+      "step": 37085
+    },
+    {
+      "epoch": 0.7418,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.793994922934544e-05,
+      "loss": 1.288,
+      "step": 37090
+    },
+    {
+      "epoch": 0.7419,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.791258181817542e-05,
+      "loss": 1.275,
+      "step": 37095
+    },
+    {
+      "epoch": 0.742,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 3.788522197216897e-05,
+      "loss": 1.2831,
+      "step": 37100
+    },
+    {
+      "epoch": 0.7421,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.785786969465981e-05,
+      "loss": 1.272,
+      "step": 37105
+    },
+    {
+      "epoch": 0.7422,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.783052498898073e-05,
+      "loss": 1.2752,
+      "step": 37110
+    },
+    {
+      "epoch": 0.7423,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.7803187858463605e-05,
+      "loss": 1.2867,
+      "step": 37115
+    },
+    {
+      "epoch": 0.7424,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.777585830643937e-05,
+      "loss": 1.2738,
+      "step": 37120
+    },
+    {
+      "epoch": 0.7425,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.774853633623806e-05,
+      "loss": 1.2612,
+      "step": 37125
+    },
+    {
+      "epoch": 0.7426,
+      "grad_norm": 0.3125,
+      "learning_rate": 3.7721221951188765e-05,
+      "loss": 1.2697,
+      "step": 37130
+    },
+    {
+      "epoch": 0.7427,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.769391515461966e-05,
+      "loss": 1.2646,
+      "step": 37135
+    },
+    {
+      "epoch": 0.7428,
+      "grad_norm": 0.142578125,
+      "learning_rate": 3.766661594985801e-05,
+      "loss": 1.272,
+      "step": 37140
+    },
+    {
+      "epoch": 0.7429,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.7639324340230085e-05,
+      "loss": 1.2866,
+      "step": 37145
+    },
+    {
+      "epoch": 0.743,
+      "grad_norm": 0.28515625,
+      "learning_rate": 3.7612040329061405e-05,
+      "loss": 1.2884,
+      "step": 37150
+    },
+    {
+      "epoch": 0.7431,
+      "grad_norm": 0.2890625,
+      "learning_rate": 3.7584763919676294e-05,
+      "loss": 1.2765,
+      "step": 37155
+    },
+    {
+      "epoch": 0.7432,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.755749511539845e-05,
+      "loss": 1.2775,
+      "step": 37160
+    },
+    {
+      "epoch": 0.7433,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 3.753023391955037e-05,
+      "loss": 1.2711,
+      "step": 37165
+    },
+    {
+      "epoch": 0.7434,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.7502980335453774e-05,
+      "loss": 1.2687,
+      "step": 37170
+    },
+    {
+      "epoch": 0.7435,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.747573436642951e-05,
+      "loss": 1.2896,
+      "step": 37175
+    },
+    {
+      "epoch": 0.7436,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.7448496015797295e-05,
+      "loss": 1.2659,
+      "step": 37180
+    },
+    {
+      "epoch": 0.7437,
+      "grad_norm": 0.466796875,
+      "learning_rate": 3.742126528687614e-05,
+      "loss": 1.2864,
+      "step": 37185
+    },
+    {
+      "epoch": 0.7438,
+      "grad_norm": 0.240234375,
+      "learning_rate": 3.739404218298398e-05,
+      "loss": 1.2693,
+      "step": 37190
+    },
+    {
+      "epoch": 0.7439,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 3.736682670743787e-05,
+      "loss": 1.2789,
+      "step": 37195
+    },
+    {
+      "epoch": 0.744,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.733961886355398e-05,
+      "loss": 1.276,
+      "step": 37200
+    },
+    {
+      "epoch": 0.7441,
+      "grad_norm": 0.189453125,
+      "learning_rate": 3.7312418654647406e-05,
+      "loss": 1.2808,
+      "step": 37205
+    },
+    {
+      "epoch": 0.7442,
+      "grad_norm": 0.1953125,
+      "learning_rate": 3.728522608403249e-05,
+      "loss": 1.2691,
+      "step": 37210
+    },
+    {
+      "epoch": 0.7443,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.725804115502254e-05,
+      "loss": 1.2815,
+      "step": 37215
+    },
+    {
+      "epoch": 0.7444,
+      "grad_norm": 0.46484375,
+      "learning_rate": 3.7230863870929964e-05,
+      "loss": 1.2687,
+      "step": 37220
+    },
+    {
+      "epoch": 0.7445,
+      "grad_norm": 0.36328125,
+      "learning_rate": 3.720369423506622e-05,
+      "loss": 1.2766,
+      "step": 37225
+    },
+    {
+      "epoch": 0.7446,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.717653225074186e-05,
+      "loss": 1.2742,
+      "step": 37230
+    },
+    {
+      "epoch": 0.7447,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.714937792126647e-05,
+      "loss": 1.2885,
+      "step": 37235
+    },
+    {
+      "epoch": 0.7448,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 3.712223124994875e-05,
+      "loss": 1.2851,
+      "step": 37240
+    },
+    {
+      "epoch": 0.7449,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 3.709509224009641e-05,
+      "loss": 1.2898,
+      "step": 37245
+    },
+    {
+      "epoch": 0.745,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.7067960895016275e-05,
+      "loss": 1.2794,
+      "step": 37250
+    },
+    {
+      "epoch": 0.7451,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.704083721801421e-05,
+      "loss": 1.292,
+      "step": 37255
+    },
+    {
+      "epoch": 0.7452,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.701372121239512e-05,
+      "loss": 1.2769,
+      "step": 37260
+    },
+    {
+      "epoch": 0.7453,
+      "grad_norm": 0.203125,
+      "learning_rate": 3.698661288146311e-05,
+      "loss": 1.2606,
+      "step": 37265
+    },
+    {
+      "epoch": 0.7454,
+      "grad_norm": 0.236328125,
+      "learning_rate": 3.695951222852112e-05,
+      "loss": 1.2836,
+      "step": 37270
+    },
+    {
+      "epoch": 0.7455,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.6932419256871406e-05,
+      "loss": 1.2754,
+      "step": 37275
+    },
+    {
+      "epoch": 0.7456,
+      "grad_norm": 0.224609375,
+      "learning_rate": 3.690533396981504e-05,
+      "loss": 1.2554,
+      "step": 37280
+    },
+    {
+      "epoch": 0.7457,
+      "grad_norm": 0.158203125,
+      "learning_rate": 3.687825637065236e-05,
+      "loss": 1.2687,
+      "step": 37285
+    },
+    {
+      "epoch": 0.7458,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.685118646268272e-05,
+      "loss": 1.2577,
+      "step": 37290
+    },
+    {
+      "epoch": 0.7459,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.682412424920438e-05,
+      "loss": 1.2701,
+      "step": 37295
+    },
+    {
+      "epoch": 0.746,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.679706973351491e-05,
+      "loss": 1.2732,
+      "step": 37300
+    },
+    {
+      "epoch": 0.7461,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.677002291891078e-05,
+      "loss": 1.293,
+      "step": 37305
+    },
+    {
+      "epoch": 0.7462,
+      "grad_norm": 0.240234375,
+      "learning_rate": 3.674298380868756e-05,
+      "loss": 1.2689,
+      "step": 37310
+    },
+    {
+      "epoch": 0.7463,
+      "grad_norm": 0.298828125,
+      "learning_rate": 3.6715952406139885e-05,
+      "loss": 1.2572,
+      "step": 37315
+    },
+    {
+      "epoch": 0.7464,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.668892871456144e-05,
+      "loss": 1.2938,
+      "step": 37320
+    },
+    {
+      "epoch": 0.7465,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.6661912737245e-05,
+      "loss": 1.2914,
+      "step": 37325
+    },
+    {
+      "epoch": 0.7466,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.6634904477482354e-05,
+      "loss": 1.2732,
+      "step": 37330
+    },
+    {
+      "epoch": 0.7467,
+      "grad_norm": 0.4765625,
+      "learning_rate": 3.6607903938564405e-05,
+      "loss": 1.2774,
+      "step": 37335
+    },
+    {
+      "epoch": 0.7468,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.6580911123781056e-05,
+      "loss": 1.2772,
+      "step": 37340
+    },
+    {
+      "epoch": 0.7469,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.6553926036421326e-05,
+      "loss": 1.2749,
+      "step": 37345
+    },
+    {
+      "epoch": 0.747,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.6526948679773257e-05,
+      "loss": 1.2709,
+      "step": 37350
+    },
+    {
+      "epoch": 0.7471,
+      "grad_norm": 0.13671875,
+      "learning_rate": 3.649997905712396e-05,
+      "loss": 1.2635,
+      "step": 37355
+    },
+    {
+      "epoch": 0.7472,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.647301717175956e-05,
+      "loss": 1.2881,
+      "step": 37360
+    },
+    {
+      "epoch": 0.7473,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.6446063026965384e-05,
+      "loss": 1.2788,
+      "step": 37365
+    },
+    {
+      "epoch": 0.7474,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 3.641911662602559e-05,
+      "loss": 1.2748,
+      "step": 37370
+    },
+    {
+      "epoch": 0.7475,
+      "grad_norm": 0.220703125,
+      "learning_rate": 3.6392177972223594e-05,
+      "loss": 1.263,
+      "step": 37375
+    },
+    {
+      "epoch": 0.7476,
+      "grad_norm": 0.3046875,
+      "learning_rate": 3.6365247068841814e-05,
+      "loss": 1.2764,
+      "step": 37380
+    },
+    {
+      "epoch": 0.7477,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.633832391916159e-05,
+      "loss": 1.2749,
+      "step": 37385
+    },
+    {
+      "epoch": 0.7478,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.631140852646355e-05,
+      "loss": 1.2795,
+      "step": 37390
+    },
+    {
+      "epoch": 0.7479,
+      "grad_norm": 0.373046875,
+      "learning_rate": 3.628450089402713e-05,
+      "loss": 1.2722,
+      "step": 37395
+    },
+    {
+      "epoch": 0.748,
+      "grad_norm": 0.16015625,
+      "learning_rate": 3.6257601025131026e-05,
+      "loss": 1.2629,
+      "step": 37400
+    },
+    {
+      "epoch": 0.7481,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.62307089230529e-05,
+      "loss": 1.2843,
+      "step": 37405
+    },
+    {
+      "epoch": 0.7482,
+      "grad_norm": 0.283203125,
+      "learning_rate": 3.6203824591069456e-05,
+      "loss": 1.2761,
+      "step": 37410
+    },
+    {
+      "epoch": 0.7483,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.6176948032456473e-05,
+      "loss": 1.2641,
+      "step": 37415
+    },
+    {
+      "epoch": 0.7484,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 3.615007925048878e-05,
+      "loss": 1.2882,
+      "step": 37420
+    },
+    {
+      "epoch": 0.7485,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.612321824844024e-05,
+      "loss": 1.2847,
+      "step": 37425
+    },
+    {
+      "epoch": 0.7486,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.60963650295838e-05,
+      "loss": 1.2731,
+      "step": 37430
+    },
+    {
+      "epoch": 0.7487,
+      "grad_norm": 0.474609375,
+      "learning_rate": 3.606951959719145e-05,
+      "loss": 1.2759,
+      "step": 37435
+    },
+    {
+      "epoch": 0.7488,
+      "grad_norm": 0.234375,
+      "learning_rate": 3.604268195453421e-05,
+      "loss": 1.2885,
+      "step": 37440
+    },
+    {
+      "epoch": 0.7489,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 3.601585210488218e-05,
+      "loss": 1.2675,
+      "step": 37445
+    },
+    {
+      "epoch": 0.749,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 3.5989030051504434e-05,
+      "loss": 1.2618,
+      "step": 37450
+    },
+    {
+      "epoch": 0.7491,
+      "grad_norm": 0.1953125,
+      "learning_rate": 3.59622157976693e-05,
+      "loss": 1.2879,
+      "step": 37455
+    },
+    {
+      "epoch": 0.7492,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.5935409346643835e-05,
+      "loss": 1.28,
+      "step": 37460
+    },
+    {
+      "epoch": 0.7493,
+      "grad_norm": 0.23828125,
+      "learning_rate": 3.590861070169449e-05,
+      "loss": 1.2841,
+      "step": 37465
+    },
+    {
+      "epoch": 0.7494,
+      "grad_norm": 0.451171875,
+      "learning_rate": 3.5881819866086484e-05,
+      "loss": 1.2735,
+      "step": 37470
+    },
+    {
+      "epoch": 0.7495,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.585503684308421e-05,
+      "loss": 1.2665,
+      "step": 37475
+    },
+    {
+      "epoch": 0.7496,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 3.582826163595119e-05,
+      "loss": 1.2742,
+      "step": 37480
+    },
+    {
+      "epoch": 0.7497,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.580149424794976e-05,
+      "loss": 1.2863,
+      "step": 37485
+    },
+    {
+      "epoch": 0.7498,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.577473468234156e-05,
+      "loss": 1.2823,
+      "step": 37490
+    },
+    {
+      "epoch": 0.7499,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.574798294238713e-05,
+      "loss": 1.2681,
+      "step": 37495
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.5721239031346066e-05,
+      "loss": 1.261,
+      "step": 37500
+    },
+    {
+      "epoch": 0.7501,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.569450295247706e-05,
+      "loss": 1.2751,
+      "step": 37505
+    },
+    {
+      "epoch": 0.7502,
+      "grad_norm": 0.29296875,
+      "learning_rate": 3.56677747090378e-05,
+      "loss": 1.2804,
+      "step": 37510
+    },
+    {
+      "epoch": 0.7503,
+      "grad_norm": 0.2734375,
+      "learning_rate": 3.564105430428506e-05,
+      "loss": 1.2843,
+      "step": 37515
+    },
+    {
+      "epoch": 0.7504,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.561434174147463e-05,
+      "loss": 1.2677,
+      "step": 37520
+    },
+    {
+      "epoch": 0.7505,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.558763702386135e-05,
+      "loss": 1.2857,
+      "step": 37525
+    },
+    {
+      "epoch": 0.7506,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.556094015469913e-05,
+      "loss": 1.2615,
+      "step": 37530
+    },
+    {
+      "epoch": 0.7507,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.553425113724088e-05,
+      "loss": 1.2639,
+      "step": 37535
+    },
+    {
+      "epoch": 0.7508,
+      "grad_norm": 0.353515625,
+      "learning_rate": 3.5507569974738574e-05,
+      "loss": 1.2727,
+      "step": 37540
+    },
+    {
+      "epoch": 0.7509,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.548089667044325e-05,
+      "loss": 1.2754,
+      "step": 37545
+    },
+    {
+      "epoch": 0.751,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.545423122760493e-05,
+      "loss": 1.2847,
+      "step": 37550
+    },
+    {
+      "epoch": 0.7511,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.542757364947281e-05,
+      "loss": 1.2684,
+      "step": 37555
+    },
+    {
+      "epoch": 0.7512,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.540092393929494e-05,
+      "loss": 1.2811,
+      "step": 37560
+    },
+    {
+      "epoch": 0.7513,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.537428210031849e-05,
+      "loss": 1.2686,
+      "step": 37565
+    },
+    {
+      "epoch": 0.7514,
+      "grad_norm": 0.30078125,
+      "learning_rate": 3.534764813578982e-05,
+      "loss": 1.2679,
+      "step": 37570
+    },
+    {
+      "epoch": 0.7515,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.5321022048954035e-05,
+      "loss": 1.265,
+      "step": 37575
+    },
+    {
+      "epoch": 0.7516,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 3.52944038430556e-05,
+      "loss": 1.2634,
+      "step": 37580
+    },
+    {
+      "epoch": 0.7517,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.52677935213377e-05,
+      "loss": 1.295,
+      "step": 37585
+    },
+    {
+      "epoch": 0.7518,
+      "grad_norm": 0.26171875,
+      "learning_rate": 3.524119108704286e-05,
+      "loss": 1.287,
+      "step": 37590
+    },
+    {
+      "epoch": 0.7519,
+      "grad_norm": 0.1875,
+      "learning_rate": 3.521459654341244e-05,
+      "loss": 1.2799,
+      "step": 37595
+    },
+    {
+      "epoch": 0.752,
+      "grad_norm": 0.2265625,
+      "learning_rate": 3.518800989368691e-05,
+      "loss": 1.2757,
+      "step": 37600
+    },
+    {
+      "epoch": 0.7521,
+      "grad_norm": 0.3828125,
+      "learning_rate": 3.516143114110582e-05,
+      "loss": 1.2694,
+      "step": 37605
+    },
+    {
+      "epoch": 0.7522,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.51348602889076e-05,
+      "loss": 1.2675,
+      "step": 37610
+    },
+    {
+      "epoch": 0.7523,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.510829734032993e-05,
+      "loss": 1.2628,
+      "step": 37615
+    },
+    {
+      "epoch": 0.7524,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.50817422986094e-05,
+      "loss": 1.2749,
+      "step": 37620
+    },
+    {
+      "epoch": 0.7525,
+      "grad_norm": 0.306640625,
+      "learning_rate": 3.5055195166981645e-05,
+      "loss": 1.2764,
+      "step": 37625
+    },
+    {
+      "epoch": 0.7526,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.5028655948681355e-05,
+      "loss": 1.282,
+      "step": 37630
+    },
+    {
+      "epoch": 0.7527,
+      "grad_norm": 0.16015625,
+      "learning_rate": 3.500212464694227e-05,
+      "loss": 1.2745,
+      "step": 37635
+    },
+    {
+      "epoch": 0.7528,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.497560126499709e-05,
+      "loss": 1.2745,
+      "step": 37640
+    },
+    {
+      "epoch": 0.7529,
+      "grad_norm": 0.1875,
+      "learning_rate": 3.494908580607774e-05,
+      "loss": 1.2752,
+      "step": 37645
+    },
+    {
+      "epoch": 0.753,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.492257827341492e-05,
+      "loss": 1.2701,
+      "step": 37650
+    },
+    {
+      "epoch": 0.7531,
+      "grad_norm": 0.458984375,
+      "learning_rate": 3.489607867023854e-05,
+      "loss": 1.272,
+      "step": 37655
+    },
+    {
+      "epoch": 0.7532,
+      "grad_norm": 0.267578125,
+      "learning_rate": 3.4869586999777495e-05,
+      "loss": 1.2694,
+      "step": 37660
+    },
+    {
+      "epoch": 0.7533,
+      "grad_norm": 0.330078125,
+      "learning_rate": 3.484310326525967e-05,
+      "loss": 1.261,
+      "step": 37665
+    },
+    {
+      "epoch": 0.7534,
+      "grad_norm": 0.38671875,
+      "learning_rate": 3.481662746991214e-05,
+      "loss": 1.2716,
+      "step": 37670
+    },
+    {
+      "epoch": 0.7535,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 3.479015961696077e-05,
+      "loss": 1.2771,
+      "step": 37675
+    },
+    {
+      "epoch": 0.7536,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 3.4763699709630716e-05,
+      "loss": 1.3042,
+      "step": 37680
+    },
+    {
+      "epoch": 0.7537,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.4737247751145896e-05,
+      "loss": 1.2752,
+      "step": 37685
+    },
+    {
+      "epoch": 0.7538,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 3.4710803744729515e-05,
+      "loss": 1.2669,
+      "step": 37690
+    },
+    {
+      "epoch": 0.7539,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 3.468436769360368e-05,
+      "loss": 1.2805,
+      "step": 37695
+    },
+    {
+      "epoch": 0.754,
+      "grad_norm": 0.203125,
+      "learning_rate": 3.465793960098945e-05,
+      "loss": 1.273,
+      "step": 37700
+    },
+    {
+      "epoch": 0.7541,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.463151947010712e-05,
+      "loss": 1.2754,
+      "step": 37705
+    },
+    {
+      "epoch": 0.7542,
+      "grad_norm": 0.283203125,
+      "learning_rate": 3.460510730417585e-05,
+      "loss": 1.2873,
+      "step": 37710
+    },
+    {
+      "epoch": 0.7543,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 3.4578703106413904e-05,
+      "loss": 1.2701,
+      "step": 37715
+    },
+    {
+      "epoch": 0.7544,
+      "grad_norm": 0.37109375,
+      "learning_rate": 3.455230688003852e-05,
+      "loss": 1.2676,
+      "step": 37720
+    },
+    {
+      "epoch": 0.7545,
+      "grad_norm": 0.875,
+      "learning_rate": 3.452591862826603e-05,
+      "loss": 1.2793,
+      "step": 37725
+    },
+    {
+      "epoch": 0.7546,
+      "grad_norm": 0.2578125,
+      "learning_rate": 3.4499538354311755e-05,
+      "loss": 1.2769,
+      "step": 37730
+    },
+    {
+      "epoch": 0.7547,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 3.447316606139004e-05,
+      "loss": 1.2701,
+      "step": 37735
+    },
+    {
+      "epoch": 0.7548,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 3.444680175271428e-05,
+      "loss": 1.2762,
+      "step": 37740
+    },
+    {
+      "epoch": 0.7549,
+      "grad_norm": 0.263671875,
+      "learning_rate": 3.442044543149688e-05,
+      "loss": 1.2724,
+      "step": 37745
+    },
+    {
+      "epoch": 0.755,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.439409710094929e-05,
+      "loss": 1.2634,
+      "step": 37750
+    },
+    {
+      "epoch": 0.7551,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.4367756764281955e-05,
+      "loss": 1.2815,
+      "step": 37755
+    },
+    {
+      "epoch": 0.7552,
+      "grad_norm": 0.38671875,
+      "learning_rate": 3.4341424424704375e-05,
+      "loss": 1.2697,
+      "step": 37760
+    },
+    {
+      "epoch": 0.7553,
+      "grad_norm": 0.294921875,
+      "learning_rate": 3.4315100085425034e-05,
+      "loss": 1.2607,
+      "step": 37765
+    },
+    {
+      "epoch": 0.7554,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.4288783749651564e-05,
+      "loss": 1.2753,
+      "step": 37770
+    },
+    {
+      "epoch": 0.7555,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.426247542059041e-05,
+      "loss": 1.2888,
+      "step": 37775
+    },
+    {
+      "epoch": 0.7556,
+      "grad_norm": 0.16796875,
+      "learning_rate": 3.423617510144727e-05,
+      "loss": 1.2946,
+      "step": 37780
+    },
+    {
+      "epoch": 0.7557,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.4209882795426716e-05,
+      "loss": 1.2914,
+      "step": 37785
+    },
+    {
+      "epoch": 0.7558,
+      "grad_norm": 0.15625,
+      "learning_rate": 3.418359850573234e-05,
+      "loss": 1.2928,
+      "step": 37790
+    },
+    {
+      "epoch": 0.7559,
+      "grad_norm": 0.4453125,
+      "learning_rate": 3.415732223556689e-05,
+      "loss": 1.2908,
+      "step": 37795
+    },
+    {
+      "epoch": 0.756,
+      "grad_norm": 0.451171875,
+      "learning_rate": 3.413105398813195e-05,
+      "loss": 1.2678,
+      "step": 37800
+    },
+    {
+      "epoch": 0.7561,
+      "grad_norm": 0.189453125,
+      "learning_rate": 3.4104793766628304e-05,
+      "loss": 1.2715,
+      "step": 37805
+    },
+    {
+      "epoch": 0.7562,
+      "grad_norm": 0.294921875,
+      "learning_rate": 3.4078541574255664e-05,
+      "loss": 1.2605,
+      "step": 37810
+    },
+    {
+      "epoch": 0.7563,
+      "grad_norm": 0.25,
+      "learning_rate": 3.4052297414212777e-05,
+      "loss": 1.2672,
+      "step": 37815
+    },
+    {
+      "epoch": 0.7564,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 3.40260612896974e-05,
+      "loss": 1.275,
+      "step": 37820
+    },
+    {
+      "epoch": 0.7565,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.3999833203906326e-05,
+      "loss": 1.2659,
+      "step": 37825
+    },
+    {
+      "epoch": 0.7566,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 3.397361316003539e-05,
+      "loss": 1.2687,
+      "step": 37830
+    },
+    {
+      "epoch": 0.7567,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.394740116127941e-05,
+      "loss": 1.2683,
+      "step": 37835
+    },
+    {
+      "epoch": 0.7568,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 3.3921197210832235e-05,
+      "loss": 1.262,
+      "step": 37840
+    },
+    {
+      "epoch": 0.7569,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.389500131188674e-05,
+      "loss": 1.2757,
+      "step": 37845
+    },
+    {
+      "epoch": 0.757,
+      "grad_norm": 0.27734375,
+      "learning_rate": 3.386881346763483e-05,
+      "loss": 1.2743,
+      "step": 37850
+    },
+    {
+      "epoch": 0.7571,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.3842633681267356e-05,
+      "loss": 1.2827,
+      "step": 37855
+    },
+    {
+      "epoch": 0.7572,
+      "grad_norm": 0.3125,
+      "learning_rate": 3.3816461955974365e-05,
+      "loss": 1.2809,
+      "step": 37860
+    },
+    {
+      "epoch": 0.7573,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 3.379029829494469e-05,
+      "loss": 1.2646,
+      "step": 37865
+    },
+    {
+      "epoch": 0.7574,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 3.376414270136633e-05,
+      "loss": 1.2713,
+      "step": 37870
+    },
+    {
+      "epoch": 0.7575,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 3.373799517842627e-05,
+      "loss": 1.2613,
+      "step": 37875
+    },
+    {
+      "epoch": 0.7576,
+      "grad_norm": 0.212890625,
+      "learning_rate": 3.371185572931048e-05,
+      "loss": 1.2868,
+      "step": 37880
+    },
+    {
+      "epoch": 0.7577,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.3685724357204054e-05,
+      "loss": 1.2822,
+      "step": 37885
+    },
+    {
+      "epoch": 0.7578,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.3659601065290893e-05,
+      "loss": 1.2653,
+      "step": 37890
+    },
+    {
+      "epoch": 0.7579,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.363348585675414e-05,
+      "loss": 1.2819,
+      "step": 37895
+    },
+    {
+      "epoch": 0.758,
+      "grad_norm": 0.22265625,
+      "learning_rate": 3.360737873477584e-05,
+      "loss": 1.2797,
+      "step": 37900
+    },
+    {
+      "epoch": 0.7581,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.358127970253704e-05,
+      "loss": 1.293,
+      "step": 37905
+    },
+    {
+      "epoch": 0.7582,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.355518876321787e-05,
+      "loss": 1.2755,
+      "step": 37910
+    },
+    {
+      "epoch": 0.7583,
+      "grad_norm": 0.16796875,
+      "learning_rate": 3.352910591999734e-05,
+      "loss": 1.2548,
+      "step": 37915
+    },
+    {
+      "epoch": 0.7584,
+      "grad_norm": 0.46875,
+      "learning_rate": 3.3503031176053656e-05,
+      "loss": 1.2619,
+      "step": 37920
+    },
+    {
+      "epoch": 0.7585,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 3.347696453456393e-05,
+      "loss": 1.2771,
+      "step": 37925
+    },
+    {
+      "epoch": 0.7586,
+      "grad_norm": 0.28125,
+      "learning_rate": 3.3450905998704275e-05,
+      "loss": 1.2821,
+      "step": 37930
+    },
+    {
+      "epoch": 0.7587,
+      "grad_norm": 0.15234375,
+      "learning_rate": 3.342485557164986e-05,
+      "loss": 1.283,
+      "step": 37935
+    },
+    {
+      "epoch": 0.7588,
+      "grad_norm": 0.28515625,
+      "learning_rate": 3.339881325657484e-05,
+      "loss": 1.2734,
+      "step": 37940
+    },
+    {
+      "epoch": 0.7589,
+      "grad_norm": 0.21484375,
+      "learning_rate": 3.3372779056652426e-05,
+      "loss": 1.2658,
+      "step": 37945
+    },
+    {
+      "epoch": 0.759,
+      "grad_norm": 0.458984375,
+      "learning_rate": 3.334675297505476e-05,
+      "loss": 1.2915,
+      "step": 37950
+    },
+    {
+      "epoch": 0.7591,
+      "grad_norm": 0.203125,
+      "learning_rate": 3.3320735014953076e-05,
+      "loss": 1.2878,
+      "step": 37955
+    },
+    {
+      "epoch": 0.7592,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 3.3294725179517574e-05,
+      "loss": 1.2746,
+      "step": 37960
+    },
+    {
+      "epoch": 0.7593,
+      "grad_norm": 0.203125,
+      "learning_rate": 3.326872347191746e-05,
+      "loss": 1.2702,
+      "step": 37965
+    },
+    {
+      "epoch": 0.7594,
+      "grad_norm": 0.24609375,
+      "learning_rate": 3.3242729895320946e-05,
+      "loss": 1.2673,
+      "step": 37970
+    },
+    {
+      "epoch": 0.7595,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 3.3216744452895354e-05,
+      "loss": 1.2734,
+      "step": 37975
+    },
+    {
+      "epoch": 0.7596,
+      "grad_norm": 0.322265625,
+      "learning_rate": 3.319076714780682e-05,
+      "loss": 1.2923,
+      "step": 37980
+    },
+    {
+      "epoch": 0.7597,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.316479798322072e-05,
+      "loss": 1.2829,
+      "step": 37985
+    },
+    {
+      "epoch": 0.7598,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.313883696230119e-05,
+      "loss": 1.2587,
+      "step": 37990
+    },
+    {
+      "epoch": 0.7599,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 3.311288408821159e-05,
+      "loss": 1.2668,
+      "step": 37995
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 3.308693936411421e-05,
+      "loss": 1.2675,
+      "step": 38000
+    },
+    {
+      "epoch": 0.7601,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 3.306100279317024e-05,
+      "loss": 1.2807,
+      "step": 38005
+    },
+    {
+      "epoch": 0.7602,
+      "grad_norm": 0.224609375,
+      "learning_rate": 3.303507437854009e-05,
+      "loss": 1.283,
+      "step": 38010
+    },
+    {
+      "epoch": 0.7603,
+      "grad_norm": 0.2265625,
+      "learning_rate": 3.3009154123382936e-05,
+      "loss": 1.2777,
+      "step": 38015
+    },
+    {
+      "epoch": 0.7604,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.2983242030857174e-05,
+      "loss": 1.2756,
+      "step": 38020
+    },
+    {
+      "epoch": 0.7605,
+      "grad_norm": 0.197265625,
+      "learning_rate": 3.2957338104120096e-05,
+      "loss": 1.2743,
+      "step": 38025
+    },
+    {
+      "epoch": 0.7606,
+      "grad_norm": 0.3828125,
+      "learning_rate": 3.2931442346328004e-05,
+      "loss": 1.2725,
+      "step": 38030
+    },
+    {
+      "epoch": 0.7607,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.290555476063622e-05,
+      "loss": 1.2785,
+      "step": 38035
+    },
+    {
+      "epoch": 0.7608,
+      "grad_norm": 0.49609375,
+      "learning_rate": 3.287967535019908e-05,
+      "loss": 1.2743,
+      "step": 38040
+    },
+    {
+      "epoch": 0.7609,
+      "grad_norm": 0.181640625,
+      "learning_rate": 3.285380411816988e-05,
+      "loss": 1.293,
+      "step": 38045
+    },
+    {
+      "epoch": 0.761,
+      "grad_norm": 0.283203125,
+      "learning_rate": 3.2827941067700996e-05,
+      "loss": 1.2664,
+      "step": 38050
+    },
+    {
+      "epoch": 0.7611,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 3.2802086201943724e-05,
+      "loss": 1.2603,
+      "step": 38055
+    },
+    {
+      "epoch": 0.7612,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 3.277623952404842e-05,
+      "loss": 1.2807,
+      "step": 38060
+    },
+    {
+      "epoch": 0.7613,
+      "grad_norm": 0.326171875,
+      "learning_rate": 3.275040103716441e-05,
+      "loss": 1.2873,
+      "step": 38065
+    },
+    {
+      "epoch": 0.7614,
+      "grad_norm": 0.19921875,
+      "learning_rate": 3.272457074444003e-05,
+      "loss": 1.2802,
+      "step": 38070
+    },
+    {
+      "epoch": 0.7615,
+      "grad_norm": 1.1328125,
+      "learning_rate": 3.269874864902269e-05,
+      "loss": 1.2737,
+      "step": 38075
+    },
+    {
+      "epoch": 0.7616,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.2672934754058616e-05,
+      "loss": 1.2673,
+      "step": 38080
+    },
+    {
+      "epoch": 0.7617,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.264712906269328e-05,
+      "loss": 1.2793,
+      "step": 38085
+    },
+    {
+      "epoch": 0.7618,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.2621331578070934e-05,
+      "loss": 1.2692,
+      "step": 38090
+    },
+    {
+      "epoch": 0.7619,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 3.2595542303334924e-05,
+      "loss": 1.2544,
+      "step": 38095
+    },
+    {
+      "epoch": 0.762,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.2569761241627696e-05,
+      "loss": 1.2824,
+      "step": 38100
+    },
+    {
+      "epoch": 0.7621,
+      "grad_norm": 0.25390625,
+      "learning_rate": 3.254398839609044e-05,
+      "loss": 1.2887,
+      "step": 38105
+    },
+    {
+      "epoch": 0.7622,
+      "grad_norm": 0.388671875,
+      "learning_rate": 3.251822376986363e-05,
+      "loss": 1.2712,
+      "step": 38110
+    },
+    {
+      "epoch": 0.7623,
+      "grad_norm": 0.4453125,
+      "learning_rate": 3.249246736608655e-05,
+      "loss": 1.2781,
+      "step": 38115
+    },
+    {
+      "epoch": 0.7624,
+      "grad_norm": 0.248046875,
+      "learning_rate": 3.246671918789755e-05,
+      "loss": 1.2956,
+      "step": 38120
+    },
+    {
+      "epoch": 0.7625,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 3.244097923843398e-05,
+      "loss": 1.2679,
+      "step": 38125
+    },
+    {
+      "epoch": 0.7626,
+      "grad_norm": 0.234375,
+      "learning_rate": 3.2415247520832146e-05,
+      "loss": 1.2767,
+      "step": 38130
+    },
+    {
+      "epoch": 0.7627,
+      "grad_norm": 0.18359375,
+      "learning_rate": 3.23895240382274e-05,
+      "loss": 1.253,
+      "step": 38135
+    },
+    {
+      "epoch": 0.7628,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.236380879375408e-05,
+      "loss": 1.2696,
+      "step": 38140
+    },
+    {
+      "epoch": 0.7629,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.233810179054548e-05,
+      "loss": 1.2768,
+      "step": 38145
+    },
+    {
+      "epoch": 0.763,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.231240303173394e-05,
+      "loss": 1.2654,
+      "step": 38150
+    },
+    {
+      "epoch": 0.7631,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.2286712520450765e-05,
+      "loss": 1.282,
+      "step": 38155
+    },
+    {
+      "epoch": 0.7632,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 3.226103025982628e-05,
+      "loss": 1.2651,
+      "step": 38160
+    },
+    {
+      "epoch": 0.7633,
+      "grad_norm": 0.365234375,
+      "learning_rate": 3.223535625298979e-05,
+      "loss": 1.2771,
+      "step": 38165
+    },
+    {
+      "epoch": 0.7634,
+      "grad_norm": 0.28515625,
+      "learning_rate": 3.220969050306955e-05,
+      "loss": 1.2455,
+      "step": 38170
+    },
+    {
+      "epoch": 0.7635,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 3.218403301319296e-05,
+      "loss": 1.2706,
+      "step": 38175
+    },
+    {
+      "epoch": 0.7636,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 3.21583837864862e-05,
+      "loss": 1.2751,
+      "step": 38180
+    },
+    {
+      "epoch": 0.7637,
+      "grad_norm": 0.318359375,
+      "learning_rate": 3.213274282607457e-05,
+      "loss": 1.2675,
+      "step": 38185
+    },
+    {
+      "epoch": 0.7638,
+      "grad_norm": 0.44921875,
+      "learning_rate": 3.210711013508242e-05,
+      "loss": 1.2838,
+      "step": 38190
+    },
+    {
+      "epoch": 0.7639,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.208148571663289e-05,
+      "loss": 1.2747,
+      "step": 38195
+    },
+    {
+      "epoch": 0.764,
+      "grad_norm": 0.27734375,
+      "learning_rate": 3.205586957384838e-05,
+      "loss": 1.2805,
+      "step": 38200
+    },
+    {
+      "epoch": 0.7641,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.2030261709849996e-05,
+      "loss": 1.2879,
+      "step": 38205
+    },
+    {
+      "epoch": 0.7642,
+      "grad_norm": 0.2890625,
+      "learning_rate": 3.200466212775808e-05,
+      "loss": 1.2784,
+      "step": 38210
+    },
+    {
+      "epoch": 0.7643,
+      "grad_norm": 0.232421875,
+      "learning_rate": 3.197907083069184e-05,
+      "loss": 1.2651,
+      "step": 38215
+    },
+    {
+      "epoch": 0.7644,
+      "grad_norm": 0.25,
+      "learning_rate": 3.195348782176948e-05,
+      "loss": 1.2893,
+      "step": 38220
+    },
+    {
+      "epoch": 0.7645,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.192791310410822e-05,
+      "loss": 1.2727,
+      "step": 38225
+    },
+    {
+      "epoch": 0.7646,
+      "grad_norm": 0.2890625,
+      "learning_rate": 3.190234668082427e-05,
+      "loss": 1.27,
+      "step": 38230
+    },
+    {
+      "epoch": 0.7647,
+      "grad_norm": 0.283203125,
+      "learning_rate": 3.187678855503282e-05,
+      "loss": 1.2766,
+      "step": 38235
+    },
+    {
+      "epoch": 0.7648,
+      "grad_norm": 0.451171875,
+      "learning_rate": 3.1851238729848034e-05,
+      "loss": 1.2738,
+      "step": 38240
+    },
+    {
+      "epoch": 0.7649,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.1825697208383096e-05,
+      "loss": 1.2838,
+      "step": 38245
+    },
+    {
+      "epoch": 0.765,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.1800163993750166e-05,
+      "loss": 1.2716,
+      "step": 38250
+    },
+    {
+      "epoch": 0.7651,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.1774639089060363e-05,
+      "loss": 1.2679,
+      "step": 38255
+    },
+    {
+      "epoch": 0.7652,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 3.174912249742382e-05,
+      "loss": 1.2667,
+      "step": 38260
+    },
+    {
+      "epoch": 0.7653,
+      "grad_norm": 0.26953125,
+      "learning_rate": 3.172361422194974e-05,
+      "loss": 1.2701,
+      "step": 38265
+    },
+    {
+      "epoch": 0.7654,
+      "grad_norm": 0.326171875,
+      "learning_rate": 3.1698114265746124e-05,
+      "loss": 1.2695,
+      "step": 38270
+    },
+    {
+      "epoch": 0.7655,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.16726226319201e-05,
+      "loss": 1.265,
+      "step": 38275
+    },
+    {
+      "epoch": 0.7656,
+      "grad_norm": 0.326171875,
+      "learning_rate": 3.164713932357776e-05,
+      "loss": 1.2739,
+      "step": 38280
+    },
+    {
+      "epoch": 0.7657,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 3.162166434382412e-05,
+      "loss": 1.2757,
+      "step": 38285
+    },
+    {
+      "epoch": 0.7658,
+      "grad_norm": 0.23046875,
+      "learning_rate": 3.159619769576333e-05,
+      "loss": 1.2917,
+      "step": 38290
+    },
+    {
+      "epoch": 0.7659,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.157073938249829e-05,
+      "loss": 1.2739,
+      "step": 38295
+    },
+    {
+      "epoch": 0.766,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.154528940713113e-05,
+      "loss": 1.2868,
+      "step": 38300
+    },
+    {
+      "epoch": 0.7661,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.15198477727628e-05,
+      "loss": 1.2835,
+      "step": 38305
+    },
+    {
+      "epoch": 0.7662,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.1494414482493304e-05,
+      "loss": 1.2725,
+      "step": 38310
+    },
+    {
+      "epoch": 0.7663,
+      "grad_norm": 0.287109375,
+      "learning_rate": 3.146898953942163e-05,
+      "loss": 1.2709,
+      "step": 38315
+    },
+    {
+      "epoch": 0.7664,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.144357294664565e-05,
+      "loss": 1.2732,
+      "step": 38320
+    },
+    {
+      "epoch": 0.7665,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.141816470726238e-05,
+      "loss": 1.2752,
+      "step": 38325
+    },
+    {
+      "epoch": 0.7666,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 3.1392764824367704e-05,
+      "loss": 1.2669,
+      "step": 38330
+    },
+    {
+      "epoch": 0.7667,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.1367373301056536e-05,
+      "loss": 1.2771,
+      "step": 38335
+    },
+    {
+      "epoch": 0.7668,
+      "grad_norm": 0.30078125,
+      "learning_rate": 3.134199014042274e-05,
+      "loss": 1.2764,
+      "step": 38340
+    },
+    {
+      "epoch": 0.7669,
+      "grad_norm": 0.212890625,
+      "learning_rate": 3.1316615345559185e-05,
+      "loss": 1.2796,
+      "step": 38345
+    },
+    {
+      "epoch": 0.767,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 3.129124891955771e-05,
+      "loss": 1.2669,
+      "step": 38350
+    },
+    {
+      "epoch": 0.7671,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.126589086550914e-05,
+      "loss": 1.272,
+      "step": 38355
+    },
+    {
+      "epoch": 0.7672,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.124054118650327e-05,
+      "loss": 1.2841,
+      "step": 38360
+    },
+    {
+      "epoch": 0.7673,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.12151998856289e-05,
+      "loss": 1.2599,
+      "step": 38365
+    },
+    {
+      "epoch": 0.7674,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 3.1189866965973766e-05,
+      "loss": 1.2721,
+      "step": 38370
+    },
+    {
+      "epoch": 0.7675,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.116454243062459e-05,
+      "loss": 1.2843,
+      "step": 38375
+    },
+    {
+      "epoch": 0.7676,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.113922628266718e-05,
+      "loss": 1.2631,
+      "step": 38380
+    },
+    {
+      "epoch": 0.7677,
+      "grad_norm": 0.373046875,
+      "learning_rate": 3.111391852518611e-05,
+      "loss": 1.274,
+      "step": 38385
+    },
+    {
+      "epoch": 0.7678,
+      "grad_norm": 0.283203125,
+      "learning_rate": 3.108861916126518e-05,
+      "loss": 1.2567,
+      "step": 38390
+    },
+    {
+      "epoch": 0.7679,
+      "grad_norm": 0.3125,
+      "learning_rate": 3.1063328193986904e-05,
+      "loss": 1.2784,
+      "step": 38395
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 3.103804562643302e-05,
+      "loss": 1.2791,
+      "step": 38400
+    },
+    {
+      "epoch": 0.7681,
+      "grad_norm": 0.265625,
+      "learning_rate": 3.101277146168412e-05,
+      "loss": 1.2745,
+      "step": 38405
+    },
+    {
+      "epoch": 0.7682,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 3.098750570281969e-05,
+      "loss": 1.2784,
+      "step": 38410
+    },
+    {
+      "epoch": 0.7683,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 3.096224835291839e-05,
+      "loss": 1.2709,
+      "step": 38415
+    },
+    {
+      "epoch": 0.7684,
+      "grad_norm": 0.2734375,
+      "learning_rate": 3.093699941505771e-05,
+      "loss": 1.2641,
+      "step": 38420
+    },
+    {
+      "epoch": 0.7685,
+      "grad_norm": 0.32421875,
+      "learning_rate": 3.0911758892314166e-05,
+      "loss": 1.2708,
+      "step": 38425
+    },
+    {
+      "epoch": 0.7686,
+      "grad_norm": 1.2734375,
+      "learning_rate": 3.0886526787763234e-05,
+      "loss": 1.2691,
+      "step": 38430
+    },
+    {
+      "epoch": 0.7687,
+      "grad_norm": 0.203125,
+      "learning_rate": 3.086130310447937e-05,
+      "loss": 1.2727,
+      "step": 38435
+    },
+    {
+      "epoch": 0.7688,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.0836087845536e-05,
+      "loss": 1.2791,
+      "step": 38440
+    },
+    {
+      "epoch": 0.7689,
+      "grad_norm": 0.28125,
+      "learning_rate": 3.081088101400552e-05,
+      "loss": 1.2886,
+      "step": 38445
+    },
+    {
+      "epoch": 0.769,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 3.078568261295933e-05,
+      "loss": 1.258,
+      "step": 38450
+    },
+    {
+      "epoch": 0.7691,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.0760492645467765e-05,
+      "loss": 1.2571,
+      "step": 38455
+    },
+    {
+      "epoch": 0.7692,
+      "grad_norm": 0.4609375,
+      "learning_rate": 3.073531111460013e-05,
+      "loss": 1.2651,
+      "step": 38460
+    },
+    {
+      "epoch": 0.7693,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 3.071013802342475e-05,
+      "loss": 1.2681,
+      "step": 38465
+    },
+    {
+      "epoch": 0.7694,
+      "grad_norm": 0.44921875,
+      "learning_rate": 3.068497337500886e-05,
+      "loss": 1.2737,
+      "step": 38470
+    },
+    {
+      "epoch": 0.7695,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.0659817172418693e-05,
+      "loss": 1.2627,
+      "step": 38475
+    },
+    {
+      "epoch": 0.7696,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.063466941871952e-05,
+      "loss": 1.269,
+      "step": 38480
+    },
+    {
+      "epoch": 0.7697,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 3.060953011697545e-05,
+      "loss": 1.2879,
+      "step": 38485
+    },
+    {
+      "epoch": 0.7698,
+      "grad_norm": 1.2421875,
+      "learning_rate": 3.058439927024962e-05,
+      "loss": 1.2733,
+      "step": 38490
+    },
+    {
+      "epoch": 0.7699,
+      "grad_norm": 0.203125,
+      "learning_rate": 3.0559276881604236e-05,
+      "loss": 1.2704,
+      "step": 38495
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.462890625,
+      "learning_rate": 3.053416295410026e-05,
+      "loss": 1.2778,
+      "step": 38500
+    },
+    {
+      "epoch": 0.7701,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 3.0509057490797888e-05,
+      "loss": 1.2709,
+      "step": 38505
+    },
+    {
+      "epoch": 0.7702,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.0483960494756016e-05,
+      "loss": 1.2704,
+      "step": 38510
+    },
+    {
+      "epoch": 0.7703,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.045887196903271e-05,
+      "loss": 1.2824,
+      "step": 38515
+    },
+    {
+      "epoch": 0.7704,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 3.0433791916684916e-05,
+      "loss": 1.2816,
+      "step": 38520
+    },
+    {
+      "epoch": 0.7705,
+      "grad_norm": 0.4921875,
+      "learning_rate": 3.0408720340768572e-05,
+      "loss": 1.2746,
+      "step": 38525
+    },
+    {
+      "epoch": 0.7706,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 3.038365724433858e-05,
+      "loss": 1.272,
+      "step": 38530
+    },
+    {
+      "epoch": 0.7707,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.035860263044873e-05,
+      "loss": 1.2893,
+      "step": 38535
+    },
+    {
+      "epoch": 0.7708,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 3.0333556502151926e-05,
+      "loss": 1.2944,
+      "step": 38540
+    },
+    {
+      "epoch": 0.7709,
+      "grad_norm": 0.36328125,
+      "learning_rate": 3.0308518862499957e-05,
+      "loss": 1.2609,
+      "step": 38545
+    },
+    {
+      "epoch": 0.771,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.0283489714543556e-05,
+      "loss": 1.2668,
+      "step": 38550
+    },
+    {
+      "epoch": 0.7711,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 3.0258469061332463e-05,
+      "loss": 1.266,
+      "step": 38555
+    },
+    {
+      "epoch": 0.7712,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.023345690591537e-05,
+      "loss": 1.2767,
+      "step": 38560
+    },
+    {
+      "epoch": 0.7713,
+      "grad_norm": 0.30859375,
+      "learning_rate": 3.0208453251339885e-05,
+      "loss": 1.2804,
+      "step": 38565
+    },
+    {
+      "epoch": 0.7714,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.018345810065275e-05,
+      "loss": 1.2923,
+      "step": 38570
+    },
+    {
+      "epoch": 0.7715,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 3.0158471456899428e-05,
+      "loss": 1.2813,
+      "step": 38575
+    },
+    {
+      "epoch": 0.7716,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.0133493323124505e-05,
+      "loss": 1.2699,
+      "step": 38580
+    },
+    {
+      "epoch": 0.7717,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.0108523702371505e-05,
+      "loss": 1.2747,
+      "step": 38585
+    },
+    {
+      "epoch": 0.7718,
+      "grad_norm": 0.140625,
+      "learning_rate": 3.0083562597682847e-05,
+      "loss": 1.2592,
+      "step": 38590
+    },
+    {
+      "epoch": 0.7719,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.0058610012100074e-05,
+      "loss": 1.283,
+      "step": 38595
+    },
+    {
+      "epoch": 0.772,
+      "grad_norm": 0.2890625,
+      "learning_rate": 3.0033665948663448e-05,
+      "loss": 1.2763,
+      "step": 38600
+    },
+    {
+      "epoch": 0.7721,
+      "grad_norm": 0.205078125,
+      "learning_rate": 3.0008730410412466e-05,
+      "loss": 1.2656,
+      "step": 38605
+    },
+    {
+      "epoch": 0.7722,
+      "grad_norm": 0.23046875,
+      "learning_rate": 2.9983803400385312e-05,
+      "loss": 1.2797,
+      "step": 38610
+    },
+    {
+      "epoch": 0.7723,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.9958884921619367e-05,
+      "loss": 1.285,
+      "step": 38615
+    },
+    {
+      "epoch": 0.7724,
+      "grad_norm": 0.451171875,
+      "learning_rate": 2.993397497715086e-05,
+      "loss": 1.2751,
+      "step": 38620
+    },
+    {
+      "epoch": 0.7725,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.9909073570014912e-05,
+      "loss": 1.2773,
+      "step": 38625
+    },
+    {
+      "epoch": 0.7726,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.9884180703245767e-05,
+      "loss": 1.2632,
+      "step": 38630
+    },
+    {
+      "epoch": 0.7727,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.9859296379876523e-05,
+      "loss": 1.2591,
+      "step": 38635
+    },
+    {
+      "epoch": 0.7728,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.9834420602939263e-05,
+      "loss": 1.2819,
+      "step": 38640
+    },
+    {
+      "epoch": 0.7729,
+      "grad_norm": 0.3125,
+      "learning_rate": 2.9809553375465004e-05,
+      "loss": 1.2696,
+      "step": 38645
+    },
+    {
+      "epoch": 0.773,
+      "grad_norm": 0.5,
+      "learning_rate": 2.9784694700483762e-05,
+      "loss": 1.2775,
+      "step": 38650
+    },
+    {
+      "epoch": 0.7731,
+      "grad_norm": 0.3046875,
+      "learning_rate": 2.9759844581024486e-05,
+      "loss": 1.2737,
+      "step": 38655
+    },
+    {
+      "epoch": 0.7732,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.9735003020115092e-05,
+      "loss": 1.2668,
+      "step": 38660
+    },
+    {
+      "epoch": 0.7733,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.9710170020782435e-05,
+      "loss": 1.2898,
+      "step": 38665
+    },
+    {
+      "epoch": 0.7734,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.968534558605236e-05,
+      "loss": 1.2666,
+      "step": 38670
+    },
+    {
+      "epoch": 0.7735,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.9660529718949627e-05,
+      "loss": 1.2765,
+      "step": 38675
+    },
+    {
+      "epoch": 0.7736,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.9635722422497993e-05,
+      "loss": 1.2664,
+      "step": 38680
+    },
+    {
+      "epoch": 0.7737,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.961092369972014e-05,
+      "loss": 1.2774,
+      "step": 38685
+    },
+    {
+      "epoch": 0.7738,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.9586133553637683e-05,
+      "loss": 1.2721,
+      "step": 38690
+    },
+    {
+      "epoch": 0.7739,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.9561351987271334e-05,
+      "loss": 1.2721,
+      "step": 38695
+    },
+    {
+      "epoch": 0.774,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.953657900364053e-05,
+      "loss": 1.2798,
+      "step": 38700
+    },
+    {
+      "epoch": 0.7741,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.9511814605763855e-05,
+      "loss": 1.2766,
+      "step": 38705
+    },
+    {
+      "epoch": 0.7742,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.9487058796658783e-05,
+      "loss": 1.2792,
+      "step": 38710
+    },
+    {
+      "epoch": 0.7743,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 2.9462311579341663e-05,
+      "loss": 1.2542,
+      "step": 38715
+    },
+    {
+      "epoch": 0.7744,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.9437572956827964e-05,
+      "loss": 1.2719,
+      "step": 38720
+    },
+    {
+      "epoch": 0.7745,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.94128429321319e-05,
+      "loss": 1.2976,
+      "step": 38725
+    },
+    {
+      "epoch": 0.7746,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 2.938812150826684e-05,
+      "loss": 1.2735,
+      "step": 38730
+    },
+    {
+      "epoch": 0.7747,
+      "grad_norm": 0.37890625,
+      "learning_rate": 2.9363408688245e-05,
+      "loss": 1.258,
+      "step": 38735
+    },
+    {
+      "epoch": 0.7748,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 2.933870447507753e-05,
+      "loss": 1.269,
+      "step": 38740
+    },
+    {
+      "epoch": 0.7749,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.931400887177459e-05,
+      "loss": 1.2681,
+      "step": 38745
+    },
+    {
+      "epoch": 0.775,
+      "grad_norm": 0.205078125,
+      "learning_rate": 2.9289321881345254e-05,
+      "loss": 1.2784,
+      "step": 38750
+    },
+    {
+      "epoch": 0.7751,
+      "grad_norm": 0.1328125,
+      "learning_rate": 2.926464350679756e-05,
+      "loss": 1.2735,
+      "step": 38755
+    },
+    {
+      "epoch": 0.7752,
+      "grad_norm": 0.30859375,
+      "learning_rate": 2.9239973751138495e-05,
+      "loss": 1.2778,
+      "step": 38760
+    },
+    {
+      "epoch": 0.7753,
+      "grad_norm": 0.361328125,
+      "learning_rate": 2.921531261737398e-05,
+      "loss": 1.2677,
+      "step": 38765
+    },
+    {
+      "epoch": 0.7754,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.9190660108508917e-05,
+      "loss": 1.2772,
+      "step": 38770
+    },
+    {
+      "epoch": 0.7755,
+      "grad_norm": 0.23828125,
+      "learning_rate": 2.9166016227547133e-05,
+      "loss": 1.2814,
+      "step": 38775
+    },
+    {
+      "epoch": 0.7756,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.9141380977491373e-05,
+      "loss": 1.2741,
+      "step": 38780
+    },
+    {
+      "epoch": 0.7757,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.911675436134347e-05,
+      "loss": 1.2678,
+      "step": 38785
+    },
+    {
+      "epoch": 0.7758,
+      "grad_norm": 0.5,
+      "learning_rate": 2.9092136382103973e-05,
+      "loss": 1.273,
+      "step": 38790
+    },
+    {
+      "epoch": 0.7759,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.9067527042772636e-05,
+      "loss": 1.2703,
+      "step": 38795
+    },
+    {
+      "epoch": 0.776,
+      "grad_norm": 1.0,
+      "learning_rate": 2.904292634634793e-05,
+      "loss": 1.2742,
+      "step": 38800
+    },
+    {
+      "epoch": 0.7761,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 2.9018334295827388e-05,
+      "loss": 1.2758,
+      "step": 38805
+    },
+    {
+      "epoch": 0.7762,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.899375089420756e-05,
+      "loss": 1.2732,
+      "step": 38810
+    },
+    {
+      "epoch": 0.7763,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.8969176144483744e-05,
+      "loss": 1.265,
+      "step": 38815
+    },
+    {
+      "epoch": 0.7764,
+      "grad_norm": 0.296875,
+      "learning_rate": 2.894461004965038e-05,
+      "loss": 1.2774,
+      "step": 38820
+    },
+    {
+      "epoch": 0.7765,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.8920052612700754e-05,
+      "loss": 1.2635,
+      "step": 38825
+    },
+    {
+      "epoch": 0.7766,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.8895503836627103e-05,
+      "loss": 1.2729,
+      "step": 38830
+    },
+    {
+      "epoch": 0.7767,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.887096372442063e-05,
+      "loss": 1.275,
+      "step": 38835
+    },
+    {
+      "epoch": 0.7768,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.8846432279071467e-05,
+      "loss": 1.261,
+      "step": 38840
+    },
+    {
+      "epoch": 0.7769,
+      "grad_norm": 0.306640625,
+      "learning_rate": 2.88219095035687e-05,
+      "loss": 1.2952,
+      "step": 38845
+    },
+    {
+      "epoch": 0.777,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.879739540090036e-05,
+      "loss": 1.2929,
+      "step": 38850
+    },
+    {
+      "epoch": 0.7771,
+      "grad_norm": 0.3046875,
+      "learning_rate": 2.877288997405341e-05,
+      "loss": 1.2705,
+      "step": 38855
+    },
+    {
+      "epoch": 0.7772,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 2.874839322601375e-05,
+      "loss": 1.2867,
+      "step": 38860
+    },
+    {
+      "epoch": 0.7773,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.872390515976625e-05,
+      "loss": 1.2805,
+      "step": 38865
+    },
+    {
+      "epoch": 0.7774,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.869942577829471e-05,
+      "loss": 1.2621,
+      "step": 38870
+    },
+    {
+      "epoch": 0.7775,
+      "grad_norm": 0.326171875,
+      "learning_rate": 2.8674955084581857e-05,
+      "loss": 1.2788,
+      "step": 38875
+    },
+    {
+      "epoch": 0.7776,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 2.865049308160934e-05,
+      "loss": 1.2795,
+      "step": 38880
+    },
+    {
+      "epoch": 0.7777,
+      "grad_norm": 0.326171875,
+      "learning_rate": 2.8626039772357882e-05,
+      "loss": 1.2816,
+      "step": 38885
+    },
+    {
+      "epoch": 0.7778,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.860159515980695e-05,
+      "loss": 1.2831,
+      "step": 38890
+    },
+    {
+      "epoch": 0.7779,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.8577159246935037e-05,
+      "loss": 1.2772,
+      "step": 38895
+    },
+    {
+      "epoch": 0.778,
+      "grad_norm": 0.13671875,
+      "learning_rate": 2.8552732036719687e-05,
+      "loss": 1.2672,
+      "step": 38900
+    },
+    {
+      "epoch": 0.7781,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.852831353213715e-05,
+      "loss": 1.2741,
+      "step": 38905
+    },
+    {
+      "epoch": 0.7782,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.8503903736162875e-05,
+      "loss": 1.2646,
+      "step": 38910
+    },
+    {
+      "epoch": 0.7783,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.8479502651770995e-05,
+      "loss": 1.2762,
+      "step": 38915
+    },
+    {
+      "epoch": 0.7784,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.8455110281934803e-05,
+      "loss": 1.2849,
+      "step": 38920
+    },
+    {
+      "epoch": 0.7785,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.8430726629626413e-05,
+      "loss": 1.2819,
+      "step": 38925
+    },
+    {
+      "epoch": 0.7786,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 2.840635169781688e-05,
+      "loss": 1.2941,
+      "step": 38930
+    },
+    {
+      "epoch": 0.7787,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.838198548947627e-05,
+      "loss": 1.2793,
+      "step": 38935
+    },
+    {
+      "epoch": 0.7788,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.835762800757341e-05,
+      "loss": 1.2635,
+      "step": 38940
+    },
+    {
+      "epoch": 0.7789,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 2.8333279255076306e-05,
+      "loss": 1.2682,
+      "step": 38945
+    },
+    {
+      "epoch": 0.779,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.8308939234951726e-05,
+      "loss": 1.2852,
+      "step": 38950
+    },
+    {
+      "epoch": 0.7791,
+      "grad_norm": 0.30859375,
+      "learning_rate": 2.8284607950165442e-05,
+      "loss": 1.2848,
+      "step": 38955
+    },
+    {
+      "epoch": 0.7792,
+      "grad_norm": 0.384765625,
+      "learning_rate": 2.826028540368215e-05,
+      "loss": 1.2724,
+      "step": 38960
+    },
+    {
+      "epoch": 0.7793,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.823597159846547e-05,
+      "loss": 1.2782,
+      "step": 38965
+    },
+    {
+      "epoch": 0.7794,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.8211666537477933e-05,
+      "loss": 1.2633,
+      "step": 38970
+    },
+    {
+      "epoch": 0.7795,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.8187370223681132e-05,
+      "loss": 1.2633,
+      "step": 38975
+    },
+    {
+      "epoch": 0.7796,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 2.816308266003541e-05,
+      "loss": 1.2568,
+      "step": 38980
+    },
+    {
+      "epoch": 0.7797,
+      "grad_norm": 0.44140625,
+      "learning_rate": 2.813880384950016e-05,
+      "loss": 1.28,
+      "step": 38985
+    },
+    {
+      "epoch": 0.7798,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 2.8114533795033683e-05,
+      "loss": 1.2573,
+      "step": 38990
+    },
+    {
+      "epoch": 0.7799,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 2.8090272499593173e-05,
+      "loss": 1.2746,
+      "step": 38995
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.8066019966134904e-05,
+      "loss": 1.2753,
+      "step": 39000
+    },
+    {
+      "epoch": 0.7801,
+      "grad_norm": 0.44921875,
+      "learning_rate": 2.8041776197613844e-05,
+      "loss": 1.2848,
+      "step": 39005
+    },
+    {
+      "epoch": 0.7802,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.8017541196984142e-05,
+      "loss": 1.2749,
+      "step": 39010
+    },
+    {
+      "epoch": 0.7803,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.7993314967198635e-05,
+      "loss": 1.2691,
+      "step": 39015
+    },
+    {
+      "epoch": 0.7804,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.7969097511209308e-05,
+      "loss": 1.277,
+      "step": 39020
+    },
+    {
+      "epoch": 0.7805,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.7944888831966987e-05,
+      "loss": 1.265,
+      "step": 39025
+    },
+    {
+      "epoch": 0.7806,
+      "grad_norm": 0.46484375,
+      "learning_rate": 2.7920688932421335e-05,
+      "loss": 1.2695,
+      "step": 39030
+    },
+    {
+      "epoch": 0.7807,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.7896497815521128e-05,
+      "loss": 1.275,
+      "step": 39035
+    },
+    {
+      "epoch": 0.7808,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 2.7872315484213952e-05,
+      "loss": 1.2768,
+      "step": 39040
+    },
+    {
+      "epoch": 0.7809,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.7848141941446347e-05,
+      "loss": 1.2755,
+      "step": 39045
+    },
+    {
+      "epoch": 0.781,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.7823977190163786e-05,
+      "loss": 1.2676,
+      "step": 39050
+    },
+    {
+      "epoch": 0.7811,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.7799821233310674e-05,
+      "loss": 1.2767,
+      "step": 39055
+    },
+    {
+      "epoch": 0.7812,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.7775674073830337e-05,
+      "loss": 1.2857,
+      "step": 39060
+    },
+    {
+      "epoch": 0.7813,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.775153571466502e-05,
+      "loss": 1.265,
+      "step": 39065
+    },
+    {
+      "epoch": 0.7814,
+      "grad_norm": 0.287109375,
+      "learning_rate": 2.772740615875594e-05,
+      "loss": 1.2706,
+      "step": 39070
+    },
+    {
+      "epoch": 0.7815,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.770328540904319e-05,
+      "loss": 1.2739,
+      "step": 39075
+    },
+    {
+      "epoch": 0.7816,
+      "grad_norm": 0.244140625,
+      "learning_rate": 2.7679173468465812e-05,
+      "loss": 1.2799,
+      "step": 39080
+    },
+    {
+      "epoch": 0.7817,
+      "grad_norm": 0.3046875,
+      "learning_rate": 2.7655070339961776e-05,
+      "loss": 1.271,
+      "step": 39085
+    },
+    {
+      "epoch": 0.7818,
+      "grad_norm": 0.384765625,
+      "learning_rate": 2.7630976026467968e-05,
+      "loss": 1.2602,
+      "step": 39090
+    },
+    {
+      "epoch": 0.7819,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.7606890530920195e-05,
+      "loss": 1.2597,
+      "step": 39095
+    },
+    {
+      "epoch": 0.782,
+      "grad_norm": 0.44921875,
+      "learning_rate": 2.7582813856253275e-05,
+      "loss": 1.2774,
+      "step": 39100
+    },
+    {
+      "epoch": 0.7821,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.755874600540078e-05,
+      "loss": 1.2688,
+      "step": 39105
+    },
+    {
+      "epoch": 0.7822,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.753468698129533e-05,
+      "loss": 1.2658,
+      "step": 39110
+    },
+    {
+      "epoch": 0.7823,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.7510636786868514e-05,
+      "loss": 1.2551,
+      "step": 39115
+    },
+    {
+      "epoch": 0.7824,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.7486595425050665e-05,
+      "loss": 1.274,
+      "step": 39120
+    },
+    {
+      "epoch": 0.7825,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.746256289877126e-05,
+      "loss": 1.2849,
+      "step": 39125
+    },
+    {
+      "epoch": 0.7826,
+      "grad_norm": 0.291015625,
+      "learning_rate": 2.743853921095848e-05,
+      "loss": 1.2676,
+      "step": 39130
+    },
+    {
+      "epoch": 0.7827,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.741452436453963e-05,
+      "loss": 1.2672,
+      "step": 39135
+    },
+    {
+      "epoch": 0.7828,
+      "grad_norm": 0.2421875,
+      "learning_rate": 2.7390518362440808e-05,
+      "loss": 1.2703,
+      "step": 39140
+    },
+    {
+      "epoch": 0.7829,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.736652120758708e-05,
+      "loss": 1.2766,
+      "step": 39145
+    },
+    {
+      "epoch": 0.783,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.734253290290242e-05,
+      "loss": 1.2632,
+      "step": 39150
+    },
+    {
+      "epoch": 0.7831,
+      "grad_norm": 0.2890625,
+      "learning_rate": 2.7318553451309726e-05,
+      "loss": 1.2832,
+      "step": 39155
+    },
+    {
+      "epoch": 0.7832,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.7294582855730832e-05,
+      "loss": 1.2667,
+      "step": 39160
+    },
+    {
+      "epoch": 0.7833,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.727062111908647e-05,
+      "loss": 1.2706,
+      "step": 39165
+    },
+    {
+      "epoch": 0.7834,
+      "grad_norm": 0.275390625,
+      "learning_rate": 2.7246668244296323e-05,
+      "loss": 1.2737,
+      "step": 39170
+    },
+    {
+      "epoch": 0.7835,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.722272423427896e-05,
+      "loss": 1.28,
+      "step": 39175
+    },
+    {
+      "epoch": 0.7836,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.7198789091951902e-05,
+      "loss": 1.2642,
+      "step": 39180
+    },
+    {
+      "epoch": 0.7837,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.717486282023153e-05,
+      "loss": 1.2743,
+      "step": 39185
+    },
+    {
+      "epoch": 0.7838,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.715094542203327e-05,
+      "loss": 1.2926,
+      "step": 39190
+    },
+    {
+      "epoch": 0.7839,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.7127036900271317e-05,
+      "loss": 1.2841,
+      "step": 39195
+    },
+    {
+      "epoch": 0.784,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.7103137257858868e-05,
+      "loss": 1.2745,
+      "step": 39200
+    },
+    {
+      "epoch": 0.7841,
+      "grad_norm": 0.275390625,
+      "learning_rate": 2.707924649770802e-05,
+      "loss": 1.2708,
+      "step": 39205
+    },
+    {
+      "epoch": 0.7842,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.7055364622729773e-05,
+      "loss": 1.2784,
+      "step": 39210
+    },
+    {
+      "epoch": 0.7843,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.7031491635834137e-05,
+      "loss": 1.28,
+      "step": 39215
+    },
+    {
+      "epoch": 0.7844,
+      "grad_norm": 0.162109375,
+      "learning_rate": 2.700762753992985e-05,
+      "loss": 1.2788,
+      "step": 39220
+    },
+    {
+      "epoch": 0.7845,
+      "grad_norm": 0.298828125,
+      "learning_rate": 2.698377233792476e-05,
+      "loss": 1.2629,
+      "step": 39225
+    },
+    {
+      "epoch": 0.7846,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.6959926032725535e-05,
+      "loss": 1.2745,
+      "step": 39230
+    },
+    {
+      "epoch": 0.7847,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.6936088627237765e-05,
+      "loss": 1.2913,
+      "step": 39235
+    },
+    {
+      "epoch": 0.7848,
+      "grad_norm": 0.142578125,
+      "learning_rate": 2.6912260124366006e-05,
+      "loss": 1.2602,
+      "step": 39240
+    },
+    {
+      "epoch": 0.7849,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.688844052701359e-05,
+      "loss": 1.2646,
+      "step": 39245
+    },
+    {
+      "epoch": 0.785,
+      "grad_norm": 0.15625,
+      "learning_rate": 2.6864629838082956e-05,
+      "loss": 1.2633,
+      "step": 39250
+    },
+    {
+      "epoch": 0.7851,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.6840828060475332e-05,
+      "loss": 1.2703,
+      "step": 39255
+    },
+    {
+      "epoch": 0.7852,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.681703519709089e-05,
+      "loss": 1.2819,
+      "step": 39260
+    },
+    {
+      "epoch": 0.7853,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 2.679325125082872e-05,
+      "loss": 1.271,
+      "step": 39265
+    },
+    {
+      "epoch": 0.7854,
+      "grad_norm": 0.37109375,
+      "learning_rate": 2.676947622458683e-05,
+      "loss": 1.2697,
+      "step": 39270
+    },
+    {
+      "epoch": 0.7855,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.6745710121262136e-05,
+      "loss": 1.2884,
+      "step": 39275
+    },
+    {
+      "epoch": 0.7856,
+      "grad_norm": 0.275390625,
+      "learning_rate": 2.672195294375045e-05,
+      "loss": 1.2794,
+      "step": 39280
+    },
+    {
+      "epoch": 0.7857,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.6698204694946527e-05,
+      "loss": 1.2831,
+      "step": 39285
+    },
+    {
+      "epoch": 0.7858,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.6674465377744017e-05,
+      "loss": 1.2781,
+      "step": 39290
+    },
+    {
+      "epoch": 0.7859,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.6650734995035477e-05,
+      "loss": 1.269,
+      "step": 39295
+    },
+    {
+      "epoch": 0.786,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.6627013549712355e-05,
+      "loss": 1.2594,
+      "step": 39300
+    },
+    {
+      "epoch": 0.7861,
+      "grad_norm": 0.240234375,
+      "learning_rate": 2.660330104466513e-05,
+      "loss": 1.2565,
+      "step": 39305
+    },
+    {
+      "epoch": 0.7862,
+      "grad_norm": 0.154296875,
+      "learning_rate": 2.657959748278297e-05,
+      "loss": 1.2642,
+      "step": 39310
+    },
+    {
+      "epoch": 0.7863,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.655590286695422e-05,
+      "loss": 1.2737,
+      "step": 39315
+    },
+    {
+      "epoch": 0.7864,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.6532217200065858e-05,
+      "loss": 1.2848,
+      "step": 39320
+    },
+    {
+      "epoch": 0.7865,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.6508540485004006e-05,
+      "loss": 1.2701,
+      "step": 39325
+    },
+    {
+      "epoch": 0.7866,
+      "grad_norm": 0.455078125,
+      "learning_rate": 2.6484872724653608e-05,
+      "loss": 1.275,
+      "step": 39330
+    },
+    {
+      "epoch": 0.7867,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.646121392189841e-05,
+      "loss": 1.2836,
+      "step": 39335
+    },
+    {
+      "epoch": 0.7868,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.6437564079621267e-05,
+      "loss": 1.281,
+      "step": 39340
+    },
+    {
+      "epoch": 0.7869,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.6413923200703794e-05,
+      "loss": 1.2718,
+      "step": 39345
+    },
+    {
+      "epoch": 0.787,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.639029128802657e-05,
+      "loss": 1.2778,
+      "step": 39350
+    },
+    {
+      "epoch": 0.7871,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.636666834446907e-05,
+      "loss": 1.2831,
+      "step": 39355
+    },
+    {
+      "epoch": 0.7872,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.634305437290968e-05,
+      "loss": 1.2562,
+      "step": 39360
+    },
+    {
+      "epoch": 0.7873,
+      "grad_norm": 0.271484375,
+      "learning_rate": 2.631944937622569e-05,
+      "loss": 1.259,
+      "step": 39365
+    },
+    {
+      "epoch": 0.7874,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.6295853357293298e-05,
+      "loss": 1.2617,
+      "step": 39370
+    },
+    {
+      "epoch": 0.7875,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 2.6272266318987603e-05,
+      "loss": 1.2978,
+      "step": 39375
+    },
+    {
+      "epoch": 0.7876,
+      "grad_norm": 0.2109375,
+      "learning_rate": 2.624868826418262e-05,
+      "loss": 1.2683,
+      "step": 39380
+    },
+    {
+      "epoch": 0.7877,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 2.6225119195751258e-05,
+      "loss": 1.3008,
+      "step": 39385
+    },
+    {
+      "epoch": 0.7878,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.6201559116565345e-05,
+      "loss": 1.2812,
+      "step": 39390
+    },
+    {
+      "epoch": 0.7879,
+      "grad_norm": 0.24609375,
+      "learning_rate": 2.6178008029495592e-05,
+      "loss": 1.2695,
+      "step": 39395
+    },
+    {
+      "epoch": 0.788,
+      "grad_norm": 0.45703125,
+      "learning_rate": 2.615446593741161e-05,
+      "loss": 1.2668,
+      "step": 39400
+    },
+    {
+      "epoch": 0.7881,
+      "grad_norm": 0.240234375,
+      "learning_rate": 2.613093284318201e-05,
+      "loss": 1.2856,
+      "step": 39405
+    },
+    {
+      "epoch": 0.7882,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.6107408749674122e-05,
+      "loss": 1.2672,
+      "step": 39410
+    },
+    {
+      "epoch": 0.7883,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.6083893659754356e-05,
+      "loss": 1.2787,
+      "step": 39415
+    },
+    {
+      "epoch": 0.7884,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.606038757628798e-05,
+      "loss": 1.2675,
+      "step": 39420
+    },
+    {
+      "epoch": 0.7885,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.603689050213902e-05,
+      "loss": 1.2974,
+      "step": 39425
+    },
+    {
+      "epoch": 0.7886,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.6013402440170676e-05,
+      "loss": 1.2732,
+      "step": 39430
+    },
+    {
+      "epoch": 0.7887,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.5989923393244742e-05,
+      "loss": 1.2789,
+      "step": 39435
+    },
+    {
+      "epoch": 0.7888,
+      "grad_norm": 0.466796875,
+      "learning_rate": 2.5966453364222186e-05,
+      "loss": 1.2844,
+      "step": 39440
+    },
+    {
+      "epoch": 0.7889,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.5942992355962727e-05,
+      "loss": 1.2824,
+      "step": 39445
+    },
+    {
+      "epoch": 0.789,
+      "grad_norm": 0.162109375,
+      "learning_rate": 2.5919540371325e-05,
+      "loss": 1.2634,
+      "step": 39450
+    },
+    {
+      "epoch": 0.7891,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.5896097413166564e-05,
+      "loss": 1.2893,
+      "step": 39455
+    },
+    {
+      "epoch": 0.7892,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.5872663484343884e-05,
+      "loss": 1.2771,
+      "step": 39460
+    },
+    {
+      "epoch": 0.7893,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.584923858771231e-05,
+      "loss": 1.2917,
+      "step": 39465
+    },
+    {
+      "epoch": 0.7894,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 2.582582272612609e-05,
+      "loss": 1.2688,
+      "step": 39470
+    },
+    {
+      "epoch": 0.7895,
+      "grad_norm": 0.482421875,
+      "learning_rate": 2.580241590243837e-05,
+      "loss": 1.2763,
+      "step": 39475
+    },
+    {
+      "epoch": 0.7896,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.5779018119501208e-05,
+      "loss": 1.2618,
+      "step": 39480
+    },
+    {
+      "epoch": 0.7897,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 2.575562938016556e-05,
+      "loss": 1.2929,
+      "step": 39485
+    },
+    {
+      "epoch": 0.7898,
+      "grad_norm": 0.318359375,
+      "learning_rate": 2.573224968728123e-05,
+      "loss": 1.2658,
+      "step": 39490
+    },
+    {
+      "epoch": 0.7899,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.5708879043697054e-05,
+      "loss": 1.2668,
+      "step": 39495
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.5685517452260567e-05,
+      "loss": 1.2817,
+      "step": 39500
+    },
+    {
+      "epoch": 0.7901,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.566216491581841e-05,
+      "loss": 1.2803,
+      "step": 39505
+    },
+    {
+      "epoch": 0.7902,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.5638821437215944e-05,
+      "loss": 1.2758,
+      "step": 39510
+    },
+    {
+      "epoch": 0.7903,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.561548701929749e-05,
+      "loss": 1.2739,
+      "step": 39515
+    },
+    {
+      "epoch": 0.7904,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.5592161664906368e-05,
+      "loss": 1.2569,
+      "step": 39520
+    },
+    {
+      "epoch": 0.7905,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.5568845376884587e-05,
+      "loss": 1.2804,
+      "step": 39525
+    },
+    {
+      "epoch": 0.7906,
+      "grad_norm": 0.373046875,
+      "learning_rate": 2.554553815807328e-05,
+      "loss": 1.2831,
+      "step": 39530
+    },
+    {
+      "epoch": 0.7907,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.5522240011312247e-05,
+      "loss": 1.2668,
+      "step": 39535
+    },
+    {
+      "epoch": 0.7908,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.549895093944039e-05,
+      "loss": 1.2818,
+      "step": 39540
+    },
+    {
+      "epoch": 0.7909,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 2.547567094529537e-05,
+      "loss": 1.2622,
+      "step": 39545
+    },
+    {
+      "epoch": 0.791,
+      "grad_norm": 0.498046875,
+      "learning_rate": 2.5452400031713785e-05,
+      "loss": 1.2628,
+      "step": 39550
+    },
+    {
+      "epoch": 0.7911,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.542913820153113e-05,
+      "loss": 1.2654,
+      "step": 39555
+    },
+    {
+      "epoch": 0.7912,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.540588545758179e-05,
+      "loss": 1.2778,
+      "step": 39560
+    },
+    {
+      "epoch": 0.7913,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.5382641802699035e-05,
+      "loss": 1.2855,
+      "step": 39565
+    },
+    {
+      "epoch": 0.7914,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.535940723971505e-05,
+      "loss": 1.2785,
+      "step": 39570
+    },
+    {
+      "epoch": 0.7915,
+      "grad_norm": 0.376953125,
+      "learning_rate": 2.5336181771460876e-05,
+      "loss": 1.2645,
+      "step": 39575
+    },
+    {
+      "epoch": 0.7916,
+      "grad_norm": 0.2421875,
+      "learning_rate": 2.5312965400766474e-05,
+      "loss": 1.2793,
+      "step": 39580
+    },
+    {
+      "epoch": 0.7917,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.5289758130460683e-05,
+      "loss": 1.2847,
+      "step": 39585
+    },
+    {
+      "epoch": 0.7918,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.5266559963371216e-05,
+      "loss": 1.2872,
+      "step": 39590
+    },
+    {
+      "epoch": 0.7919,
+      "grad_norm": 0.1640625,
+      "learning_rate": 2.5243370902324792e-05,
+      "loss": 1.269,
+      "step": 39595
+    },
+    {
+      "epoch": 0.792,
+      "grad_norm": 0.25,
+      "learning_rate": 2.5220190950146827e-05,
+      "loss": 1.2734,
+      "step": 39600
+    },
+    {
+      "epoch": 0.7921,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.5197020109661772e-05,
+      "loss": 1.2613,
+      "step": 39605
+    },
+    {
+      "epoch": 0.7922,
+      "grad_norm": 0.283203125,
+      "learning_rate": 2.5173858383692906e-05,
+      "loss": 1.2651,
+      "step": 39610
+    },
+    {
+      "epoch": 0.7923,
+      "grad_norm": 0.32421875,
+      "learning_rate": 2.51507057750624e-05,
+      "loss": 1.2717,
+      "step": 39615
+    },
+    {
+      "epoch": 0.7924,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.512756228659141e-05,
+      "loss": 1.2806,
+      "step": 39620
+    },
+    {
+      "epoch": 0.7925,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 2.5104427921099782e-05,
+      "loss": 1.2903,
+      "step": 39625
+    },
+    {
+      "epoch": 0.7926,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 2.508130268140646e-05,
+      "loss": 1.292,
+      "step": 39630
+    },
+    {
+      "epoch": 0.7927,
+      "grad_norm": 0.380859375,
+      "learning_rate": 2.5058186570329156e-05,
+      "loss": 1.2771,
+      "step": 39635
+    },
+    {
+      "epoch": 0.7928,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 2.5035079590684497e-05,
+      "loss": 1.2812,
+      "step": 39640
+    },
+    {
+      "epoch": 0.7929,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.5011981745288015e-05,
+      "loss": 1.2666,
+      "step": 39645
+    },
+    {
+      "epoch": 0.793,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.4988893036954043e-05,
+      "loss": 1.2777,
+      "step": 39650
+    },
+    {
+      "epoch": 0.7931,
+      "grad_norm": 1.15625,
+      "learning_rate": 2.496581346849596e-05,
+      "loss": 1.2718,
+      "step": 39655
+    },
+    {
+      "epoch": 0.7932,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.4942743042725892e-05,
+      "loss": 1.2772,
+      "step": 39660
+    },
+    {
+      "epoch": 0.7933,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.4919681762454918e-05,
+      "loss": 1.2563,
+      "step": 39665
+    },
+    {
+      "epoch": 0.7934,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.4896629630492973e-05,
+      "loss": 1.2679,
+      "step": 39670
+    },
+    {
+      "epoch": 0.7935,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.4873586649648894e-05,
+      "loss": 1.2818,
+      "step": 39675
+    },
+    {
+      "epoch": 0.7936,
+      "grad_norm": 0.28125,
+      "learning_rate": 2.48505528227304e-05,
+      "loss": 1.2677,
+      "step": 39680
+    },
+    {
+      "epoch": 0.7937,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.48275281525441e-05,
+      "loss": 1.263,
+      "step": 39685
+    },
+    {
+      "epoch": 0.7938,
+      "grad_norm": 0.326171875,
+      "learning_rate": 2.480451264189546e-05,
+      "loss": 1.2754,
+      "step": 39690
+    },
+    {
+      "epoch": 0.7939,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 2.4781506293588873e-05,
+      "loss": 1.2787,
+      "step": 39695
+    },
+    {
+      "epoch": 0.794,
+      "grad_norm": 0.2421875,
+      "learning_rate": 2.4758509110427575e-05,
+      "loss": 1.2793,
+      "step": 39700
+    },
+    {
+      "epoch": 0.7941,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 2.4735521095213687e-05,
+      "loss": 1.2735,
+      "step": 39705
+    },
+    {
+      "epoch": 0.7942,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.4712542250748304e-05,
+      "loss": 1.2632,
+      "step": 39710
+    },
+    {
+      "epoch": 0.7943,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.4689572579831222e-05,
+      "loss": 1.281,
+      "step": 39715
+    },
+    {
+      "epoch": 0.7944,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 2.4666612085261342e-05,
+      "loss": 1.266,
+      "step": 39720
+    },
+    {
+      "epoch": 0.7945,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.464366076983623e-05,
+      "loss": 1.2688,
+      "step": 39725
+    },
+    {
+      "epoch": 0.7946,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.4620718636352457e-05,
+      "loss": 1.2642,
+      "step": 39730
+    },
+    {
+      "epoch": 0.7947,
+      "grad_norm": 0.384765625,
+      "learning_rate": 2.4597785687605513e-05,
+      "loss": 1.2717,
+      "step": 39735
+    },
+    {
+      "epoch": 0.7948,
+      "grad_norm": 0.443359375,
+      "learning_rate": 2.4574861926389615e-05,
+      "loss": 1.2683,
+      "step": 39740
+    },
+    {
+      "epoch": 0.7949,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.4551947355498027e-05,
+      "loss": 1.2783,
+      "step": 39745
+    },
+    {
+      "epoch": 0.795,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.45290419777228e-05,
+      "loss": 1.2731,
+      "step": 39750
+    },
+    {
+      "epoch": 0.7951,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.4506145795854873e-05,
+      "loss": 1.2735,
+      "step": 39755
+    },
+    {
+      "epoch": 0.7952,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 2.4483258812684096e-05,
+      "loss": 1.2854,
+      "step": 39760
+    },
+    {
+      "epoch": 0.7953,
+      "grad_norm": 1.1640625,
+      "learning_rate": 2.4460381030999158e-05,
+      "loss": 1.2688,
+      "step": 39765
+    },
+    {
+      "epoch": 0.7954,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.443751245358765e-05,
+      "loss": 1.2759,
+      "step": 39770
+    },
+    {
+      "epoch": 0.7955,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.441465308323605e-05,
+      "loss": 1.2626,
+      "step": 39775
+    },
+    {
+      "epoch": 0.7956,
+      "grad_norm": 0.37890625,
+      "learning_rate": 2.43918029227297e-05,
+      "loss": 1.2702,
+      "step": 39780
+    },
+    {
+      "epoch": 0.7957,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.436896197485282e-05,
+      "loss": 1.2797,
+      "step": 39785
+    },
+    {
+      "epoch": 0.7958,
+      "grad_norm": 0.234375,
+      "learning_rate": 2.43461302423885e-05,
+      "loss": 1.2724,
+      "step": 39790
+    },
+    {
+      "epoch": 0.7959,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.4323307728118738e-05,
+      "loss": 1.2689,
+      "step": 39795
+    },
+    {
+      "epoch": 0.796,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 2.4300494434824373e-05,
+      "loss": 1.2651,
+      "step": 39800
+    },
+    {
+      "epoch": 0.7961,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.4277690365285112e-05,
+      "loss": 1.2845,
+      "step": 39805
+    },
+    {
+      "epoch": 0.7962,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.425489552227964e-05,
+      "loss": 1.2735,
+      "step": 39810
+    },
+    {
+      "epoch": 0.7963,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.4232109908585377e-05,
+      "loss": 1.2711,
+      "step": 39815
+    },
+    {
+      "epoch": 0.7964,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.420933352697865e-05,
+      "loss": 1.2711,
+      "step": 39820
+    },
+    {
+      "epoch": 0.7965,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.4186566380234798e-05,
+      "loss": 1.2846,
+      "step": 39825
+    },
+    {
+      "epoch": 0.7966,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.4163808471127812e-05,
+      "loss": 1.293,
+      "step": 39830
+    },
+    {
+      "epoch": 0.7967,
+      "grad_norm": 0.4609375,
+      "learning_rate": 2.4141059802430777e-05,
+      "loss": 1.2617,
+      "step": 39835
+    },
+    {
+      "epoch": 0.7968,
+      "grad_norm": 0.287109375,
+      "learning_rate": 2.411832037691545e-05,
+      "loss": 1.2812,
+      "step": 39840
+    },
+    {
+      "epoch": 0.7969,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.4095590197352635e-05,
+      "loss": 1.2626,
+      "step": 39845
+    },
+    {
+      "epoch": 0.797,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.407286926651192e-05,
+      "loss": 1.2727,
+      "step": 39850
+    },
+    {
+      "epoch": 0.7971,
+      "grad_norm": 0.443359375,
+      "learning_rate": 2.405015758716177e-05,
+      "loss": 1.2688,
+      "step": 39855
+    },
+    {
+      "epoch": 0.7972,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.4027455162069567e-05,
+      "loss": 1.2695,
+      "step": 39860
+    },
+    {
+      "epoch": 0.7973,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.4004761994001435e-05,
+      "loss": 1.2822,
+      "step": 39865
+    },
+    {
+      "epoch": 0.7974,
+      "grad_norm": 0.298828125,
+      "learning_rate": 2.3982078085722582e-05,
+      "loss": 1.2545,
+      "step": 39870
+    },
+    {
+      "epoch": 0.7975,
+      "grad_norm": 0.302734375,
+      "learning_rate": 2.3959403439996907e-05,
+      "loss": 1.2802,
+      "step": 39875
+    },
+    {
+      "epoch": 0.7976,
+      "grad_norm": 0.240234375,
+      "learning_rate": 2.3936738059587282e-05,
+      "loss": 1.2778,
+      "step": 39880
+    },
+    {
+      "epoch": 0.7977,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 2.3914081947255397e-05,
+      "loss": 1.2865,
+      "step": 39885
+    },
+    {
+      "epoch": 0.7978,
+      "grad_norm": 0.28125,
+      "learning_rate": 2.389143510576184e-05,
+      "loss": 1.2689,
+      "step": 39890
+    },
+    {
+      "epoch": 0.7979,
+      "grad_norm": 0.2890625,
+      "learning_rate": 2.3868797537866016e-05,
+      "loss": 1.2828,
+      "step": 39895
+    },
+    {
+      "epoch": 0.798,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.3846169246326343e-05,
+      "loss": 1.2887,
+      "step": 39900
+    },
+    {
+      "epoch": 0.7981,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.3823550233899915e-05,
+      "loss": 1.265,
+      "step": 39905
+    },
+    {
+      "epoch": 0.7982,
+      "grad_norm": 0.25,
+      "learning_rate": 2.3800940503342828e-05,
+      "loss": 1.2751,
+      "step": 39910
+    },
+    {
+      "epoch": 0.7983,
+      "grad_norm": 0.279296875,
+      "learning_rate": 2.3778340057409998e-05,
+      "loss": 1.277,
+      "step": 39915
+    },
+    {
+      "epoch": 0.7984,
+      "grad_norm": 0.28515625,
+      "learning_rate": 2.37557488988552e-05,
+      "loss": 1.2662,
+      "step": 39920
+    },
+    {
+      "epoch": 0.7985,
+      "grad_norm": 0.36328125,
+      "learning_rate": 2.373316703043119e-05,
+      "loss": 1.2799,
+      "step": 39925
+    },
+    {
+      "epoch": 0.7986,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.3710594454889378e-05,
+      "loss": 1.2669,
+      "step": 39930
+    },
+    {
+      "epoch": 0.7987,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.3688031174980275e-05,
+      "loss": 1.2736,
+      "step": 39935
+    },
+    {
+      "epoch": 0.7988,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 2.3665477193453034e-05,
+      "loss": 1.2689,
+      "step": 39940
+    },
+    {
+      "epoch": 0.7989,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 2.3642932513055884e-05,
+      "loss": 1.2772,
+      "step": 39945
+    },
+    {
+      "epoch": 0.799,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.362039713653581e-05,
+      "loss": 1.2587,
+      "step": 39950
+    },
+    {
+      "epoch": 0.7991,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.359787106663861e-05,
+      "loss": 1.2741,
+      "step": 39955
+    },
+    {
+      "epoch": 0.7992,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.35753543061091e-05,
+      "loss": 1.2875,
+      "step": 39960
+    },
+    {
+      "epoch": 0.7993,
+      "grad_norm": 0.203125,
+      "learning_rate": 2.3552846857690846e-05,
+      "loss": 1.2787,
+      "step": 39965
+    },
+    {
+      "epoch": 0.7994,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.3530348724126307e-05,
+      "loss": 1.2744,
+      "step": 39970
+    },
+    {
+      "epoch": 0.7995,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.3507859908156827e-05,
+      "loss": 1.2647,
+      "step": 39975
+    },
+    {
+      "epoch": 0.7996,
+      "grad_norm": 0.23828125,
+      "learning_rate": 2.3485380412522585e-05,
+      "loss": 1.2966,
+      "step": 39980
+    },
+    {
+      "epoch": 0.7997,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.3462910239962654e-05,
+      "loss": 1.2631,
+      "step": 39985
+    },
+    {
+      "epoch": 0.7998,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.3440449393214948e-05,
+      "loss": 1.2697,
+      "step": 39990
+    },
+    {
+      "epoch": 0.7999,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.341799787501625e-05,
+      "loss": 1.26,
+      "step": 39995
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.339555568810221e-05,
+      "loss": 1.2801,
+      "step": 40000
+    },
+    {
+      "epoch": 0.8001,
+      "grad_norm": 0.30078125,
+      "learning_rate": 2.337312283520735e-05,
+      "loss": 1.2733,
+      "step": 40005
+    },
+    {
+      "epoch": 0.8002,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.3350699319065026e-05,
+      "loss": 1.26,
+      "step": 40010
+    },
+    {
+      "epoch": 0.8003,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.3328285142407503e-05,
+      "loss": 1.2676,
+      "step": 40015
+    },
+    {
+      "epoch": 0.8004,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.3305880307965833e-05,
+      "loss": 1.2737,
+      "step": 40020
+    },
+    {
+      "epoch": 0.8005,
+      "grad_norm": 0.203125,
+      "learning_rate": 2.328348481847006e-05,
+      "loss": 1.2694,
+      "step": 40025
+    },
+    {
+      "epoch": 0.8006,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 2.3261098676648907e-05,
+      "loss": 1.27,
+      "step": 40030
+    },
+    {
+      "epoch": 0.8007,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.323872188523013e-05,
+      "loss": 1.2779,
+      "step": 40035
+    },
+    {
+      "epoch": 0.8008,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 2.321635444694028e-05,
+      "loss": 1.261,
+      "step": 40040
+    },
+    {
+      "epoch": 0.8009,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.319399636450468e-05,
+      "loss": 1.278,
+      "step": 40045
+    },
+    {
+      "epoch": 0.801,
+      "grad_norm": 0.2109375,
+      "learning_rate": 2.3171647640647687e-05,
+      "loss": 1.2711,
+      "step": 40050
+    },
+    {
+      "epoch": 0.8011,
+      "grad_norm": 0.27734375,
+      "learning_rate": 2.3149308278092342e-05,
+      "loss": 1.2741,
+      "step": 40055
+    },
+    {
+      "epoch": 0.8012,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.3126978279560684e-05,
+      "loss": 1.2814,
+      "step": 40060
+    },
+    {
+      "epoch": 0.8013,
+      "grad_norm": 0.2890625,
+      "learning_rate": 2.3104657647773554e-05,
+      "loss": 1.2737,
+      "step": 40065
+    },
+    {
+      "epoch": 0.8014,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.3082346385450637e-05,
+      "loss": 1.2724,
+      "step": 40070
+    },
+    {
+      "epoch": 0.8015,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.3060044495310505e-05,
+      "loss": 1.277,
+      "step": 40075
+    },
+    {
+      "epoch": 0.8016,
+      "grad_norm": 0.15234375,
+      "learning_rate": 2.3037751980070555e-05,
+      "loss": 1.2984,
+      "step": 40080
+    },
+    {
+      "epoch": 0.8017,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.3015468842447086e-05,
+      "loss": 1.279,
+      "step": 40085
+    },
+    {
+      "epoch": 0.8018,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 2.2993195085155205e-05,
+      "loss": 1.2717,
+      "step": 40090
+    },
+    {
+      "epoch": 0.8019,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.2970930710908935e-05,
+      "loss": 1.2679,
+      "step": 40095
+    },
+    {
+      "epoch": 0.802,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.2948675722421086e-05,
+      "loss": 1.2793,
+      "step": 40100
+    },
+    {
+      "epoch": 0.8021,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.2926430122403386e-05,
+      "loss": 1.2681,
+      "step": 40105
+    },
+    {
+      "epoch": 0.8022,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.2904193913566363e-05,
+      "loss": 1.28,
+      "step": 40110
+    },
+    {
+      "epoch": 0.8023,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.2881967098619506e-05,
+      "loss": 1.2818,
+      "step": 40115
+    },
+    {
+      "epoch": 0.8024,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.2859749680270982e-05,
+      "loss": 1.2703,
+      "step": 40120
+    },
+    {
+      "epoch": 0.8025,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.2837541661228025e-05,
+      "loss": 1.2672,
+      "step": 40125
+    },
+    {
+      "epoch": 0.8026,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.2815343044196525e-05,
+      "loss": 1.265,
+      "step": 40130
+    },
+    {
+      "epoch": 0.8027,
+      "grad_norm": 0.29296875,
+      "learning_rate": 2.279315383188132e-05,
+      "loss": 1.2726,
+      "step": 40135
+    },
+    {
+      "epoch": 0.8028,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.277097402698619e-05,
+      "loss": 1.2797,
+      "step": 40140
+    },
+    {
+      "epoch": 0.8029,
+      "grad_norm": 0.205078125,
+      "learning_rate": 2.2748803632213557e-05,
+      "loss": 1.2617,
+      "step": 40145
+    },
+    {
+      "epoch": 0.803,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 2.2726642650264895e-05,
+      "loss": 1.2624,
+      "step": 40150
+    },
+    {
+      "epoch": 0.8031,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.270449108384044e-05,
+      "loss": 1.271,
+      "step": 40155
+    },
+    {
+      "epoch": 0.8032,
+      "grad_norm": 0.375,
+      "learning_rate": 2.2682348935639274e-05,
+      "loss": 1.2828,
+      "step": 40160
+    },
+    {
+      "epoch": 0.8033,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.2660216208359365e-05,
+      "loss": 1.281,
+      "step": 40165
+    },
+    {
+      "epoch": 0.8034,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.2638092904697518e-05,
+      "loss": 1.2674,
+      "step": 40170
+    },
+    {
+      "epoch": 0.8035,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.2615979027349387e-05,
+      "loss": 1.2574,
+      "step": 40175
+    },
+    {
+      "epoch": 0.8036,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.259387457900948e-05,
+      "loss": 1.2604,
+      "step": 40180
+    },
+    {
+      "epoch": 0.8037,
+      "grad_norm": 0.287109375,
+      "learning_rate": 2.2571779562371153e-05,
+      "loss": 1.2692,
+      "step": 40185
+    },
+    {
+      "epoch": 0.8038,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 2.254969398012663e-05,
+      "loss": 1.2754,
+      "step": 40190
+    },
+    {
+      "epoch": 0.8039,
+      "grad_norm": 0.291015625,
+      "learning_rate": 2.2527617834966954e-05,
+      "loss": 1.2621,
+      "step": 40195
+    },
+    {
+      "epoch": 0.804,
+      "grad_norm": 0.224609375,
+      "learning_rate": 2.2505551129582047e-05,
+      "loss": 1.2734,
+      "step": 40200
+    },
+    {
+      "epoch": 0.8041,
+      "grad_norm": 0.205078125,
+      "learning_rate": 2.2483493866660676e-05,
+      "loss": 1.2932,
+      "step": 40205
+    },
+    {
+      "epoch": 0.8042,
+      "grad_norm": 0.32421875,
+      "learning_rate": 2.246144604889042e-05,
+      "loss": 1.2953,
+      "step": 40210
+    },
+    {
+      "epoch": 0.8043,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.2439407678957812e-05,
+      "loss": 1.2726,
+      "step": 40215
+    },
+    {
+      "epoch": 0.8044,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.241737875954808e-05,
+      "loss": 1.274,
+      "step": 40220
+    },
+    {
+      "epoch": 0.8045,
+      "grad_norm": 0.294921875,
+      "learning_rate": 2.2395359293345396e-05,
+      "loss": 1.281,
+      "step": 40225
+    },
+    {
+      "epoch": 0.8046,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.237334928303283e-05,
+      "loss": 1.2661,
+      "step": 40230
+    },
+    {
+      "epoch": 0.8047,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.235134873129213e-05,
+      "loss": 1.2783,
+      "step": 40235
+    },
+    {
+      "epoch": 0.8048,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.2329357640804117e-05,
+      "loss": 1.2628,
+      "step": 40240
+    },
+    {
+      "epoch": 0.8049,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.2307376014248216e-05,
+      "loss": 1.2711,
+      "step": 40245
+    },
+    {
+      "epoch": 0.805,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.2285403854302912e-05,
+      "loss": 1.2639,
+      "step": 40250
+    },
+    {
+      "epoch": 0.8051,
+      "grad_norm": 0.240234375,
+      "learning_rate": 2.2263441163645403e-05,
+      "loss": 1.2854,
+      "step": 40255
+    },
+    {
+      "epoch": 0.8052,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.22414879449518e-05,
+      "loss": 1.2779,
+      "step": 40260
+    },
+    {
+      "epoch": 0.8053,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 2.2219544200897025e-05,
+      "loss": 1.267,
+      "step": 40265
+    },
+    {
+      "epoch": 0.8054,
+      "grad_norm": 0.375,
+      "learning_rate": 2.219760993415485e-05,
+      "loss": 1.2602,
+      "step": 40270
+    },
+    {
+      "epoch": 0.8055,
+      "grad_norm": 0.298828125,
+      "learning_rate": 2.2175685147397906e-05,
+      "loss": 1.2707,
+      "step": 40275
+    },
+    {
+      "epoch": 0.8056,
+      "grad_norm": 0.236328125,
+      "learning_rate": 2.2153769843297667e-05,
+      "loss": 1.26,
+      "step": 40280
+    },
+    {
+      "epoch": 0.8057,
+      "grad_norm": 0.37890625,
+      "learning_rate": 2.213186402452443e-05,
+      "loss": 1.2771,
+      "step": 40285
+    },
+    {
+      "epoch": 0.8058,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.210996769374737e-05,
+      "loss": 1.2789,
+      "step": 40290
+    },
+    {
+      "epoch": 0.8059,
+      "grad_norm": 0.240234375,
+      "learning_rate": 2.2088080853634473e-05,
+      "loss": 1.2597,
+      "step": 40295
+    },
+    {
+      "epoch": 0.806,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.2066203506852566e-05,
+      "loss": 1.2613,
+      "step": 40300
+    },
+    {
+      "epoch": 0.8061,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.204433565606743e-05,
+      "loss": 1.2636,
+      "step": 40305
+    },
+    {
+      "epoch": 0.8062,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.202247730394349e-05,
+      "loss": 1.2745,
+      "step": 40310
+    },
+    {
+      "epoch": 0.8063,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.200062845314417e-05,
+      "loss": 1.2615,
+      "step": 40315
+    },
+    {
+      "epoch": 0.8064,
+      "grad_norm": 0.296875,
+      "learning_rate": 2.1978789106331665e-05,
+      "loss": 1.2799,
+      "step": 40320
+    },
+    {
+      "epoch": 0.8065,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.195695926616702e-05,
+      "loss": 1.2644,
+      "step": 40325
+    },
+    {
+      "epoch": 0.8066,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.1935138935310206e-05,
+      "loss": 1.269,
+      "step": 40330
+    },
+    {
+      "epoch": 0.8067,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.1913328116419873e-05,
+      "loss": 1.2756,
+      "step": 40335
+    },
+    {
+      "epoch": 0.8068,
+      "grad_norm": 0.302734375,
+      "learning_rate": 2.1891526812153672e-05,
+      "loss": 1.2864,
+      "step": 40340
+    },
+    {
+      "epoch": 0.8069,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.1869735025168026e-05,
+      "loss": 1.2748,
+      "step": 40345
+    },
+    {
+      "epoch": 0.807,
+      "grad_norm": 0.279296875,
+      "learning_rate": 2.1847952758118117e-05,
+      "loss": 1.2662,
+      "step": 40350
+    },
+    {
+      "epoch": 0.8071,
+      "grad_norm": 0.498046875,
+      "learning_rate": 2.1826180013658172e-05,
+      "loss": 1.2704,
+      "step": 40355
+    },
+    {
+      "epoch": 0.8072,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.1804416794440995e-05,
+      "loss": 1.2791,
+      "step": 40360
+    },
+    {
+      "epoch": 0.8073,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.178266310311847e-05,
+      "loss": 1.2716,
+      "step": 40365
+    },
+    {
+      "epoch": 0.8074,
+      "grad_norm": 0.365234375,
+      "learning_rate": 2.1760918942341192e-05,
+      "loss": 1.2772,
+      "step": 40370
+    },
+    {
+      "epoch": 0.8075,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.173918431475861e-05,
+      "loss": 1.282,
+      "step": 40375
+    },
+    {
+      "epoch": 0.8076,
+      "grad_norm": 0.875,
+      "learning_rate": 2.1717459223019033e-05,
+      "loss": 1.2654,
+      "step": 40380
+    },
+    {
+      "epoch": 0.8077,
+      "grad_norm": 0.19140625,
+      "learning_rate": 2.1695743669769596e-05,
+      "loss": 1.26,
+      "step": 40385
+    },
+    {
+      "epoch": 0.8078,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.1674037657656266e-05,
+      "loss": 1.2744,
+      "step": 40390
+    },
+    {
+      "epoch": 0.8079,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.1652341189323866e-05,
+      "loss": 1.282,
+      "step": 40395
+    },
+    {
+      "epoch": 0.808,
+      "grad_norm": 0.234375,
+      "learning_rate": 2.163065426741603e-05,
+      "loss": 1.2651,
+      "step": 40400
+    },
+    {
+      "epoch": 0.8081,
+      "grad_norm": 0.232421875,
+      "learning_rate": 2.160897689457526e-05,
+      "loss": 1.2763,
+      "step": 40405
+    },
+    {
+      "epoch": 0.8082,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.1587309073442863e-05,
+      "loss": 1.2714,
+      "step": 40410
+    },
+    {
+      "epoch": 0.8083,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.1565650806658975e-05,
+      "loss": 1.2756,
+      "step": 40415
+    },
+    {
+      "epoch": 0.8084,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.154400209686268e-05,
+      "loss": 1.2816,
+      "step": 40420
+    },
+    {
+      "epoch": 0.8085,
+      "grad_norm": 1.359375,
+      "learning_rate": 2.1522362946691698e-05,
+      "loss": 1.2618,
+      "step": 40425
+    },
+    {
+      "epoch": 0.8086,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.1500733358782786e-05,
+      "loss": 1.2613,
+      "step": 40430
+    },
+    {
+      "epoch": 0.8087,
+      "grad_norm": 0.2421875,
+      "learning_rate": 2.1479113335771383e-05,
+      "loss": 1.2701,
+      "step": 40435
+    },
+    {
+      "epoch": 0.8088,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 2.1457502880291812e-05,
+      "loss": 1.2898,
+      "step": 40440
+    },
+    {
+      "epoch": 0.8089,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.1435901994977326e-05,
+      "loss": 1.2834,
+      "step": 40445
+    },
+    {
+      "epoch": 0.809,
+      "grad_norm": 0.232421875,
+      "learning_rate": 2.1414310682459802e-05,
+      "loss": 1.2783,
+      "step": 40450
+    },
+    {
+      "epoch": 0.8091,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 2.1392728945370222e-05,
+      "loss": 1.2725,
+      "step": 40455
+    },
+    {
+      "epoch": 0.8092,
+      "grad_norm": 0.36328125,
+      "learning_rate": 2.137115678633811e-05,
+      "loss": 1.2639,
+      "step": 40460
+    },
+    {
+      "epoch": 0.8093,
+      "grad_norm": 0.375,
+      "learning_rate": 2.1349594207992064e-05,
+      "loss": 1.2844,
+      "step": 40465
+    },
+    {
+      "epoch": 0.8094,
+      "grad_norm": 0.486328125,
+      "learning_rate": 2.1328041212959403e-05,
+      "loss": 1.2677,
+      "step": 40470
+    },
+    {
+      "epoch": 0.8095,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.1306497803866277e-05,
+      "loss": 1.2785,
+      "step": 40475
+    },
+    {
+      "epoch": 0.8096,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.128496398333768e-05,
+      "loss": 1.285,
+      "step": 40480
+    },
+    {
+      "epoch": 0.8097,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.126343975399747e-05,
+      "loss": 1.2789,
+      "step": 40485
+    },
+    {
+      "epoch": 0.8098,
+      "grad_norm": 0.2890625,
+      "learning_rate": 2.1241925118468287e-05,
+      "loss": 1.2699,
+      "step": 40490
+    },
+    {
+      "epoch": 0.8099,
+      "grad_norm": 0.498046875,
+      "learning_rate": 2.1220420079371628e-05,
+      "loss": 1.2742,
+      "step": 40495
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.119892463932781e-05,
+      "loss": 1.2886,
+      "step": 40500
+    },
+    {
+      "epoch": 0.8101,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 2.1177438800956007e-05,
+      "loss": 1.278,
+      "step": 40505
+    },
+    {
+      "epoch": 0.8102,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.115596256687419e-05,
+      "loss": 1.2766,
+      "step": 40510
+    },
+    {
+      "epoch": 0.8103,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.113449593969915e-05,
+      "loss": 1.272,
+      "step": 40515
+    },
+    {
+      "epoch": 0.8104,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.1113038922046602e-05,
+      "loss": 1.2882,
+      "step": 40520
+    },
+    {
+      "epoch": 0.8105,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.1091591516530952e-05,
+      "loss": 1.2682,
+      "step": 40525
+    },
+    {
+      "epoch": 0.8106,
+      "grad_norm": 0.283203125,
+      "learning_rate": 2.107015372576552e-05,
+      "loss": 1.2698,
+      "step": 40530
+    },
+    {
+      "epoch": 0.8107,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.1048725552362435e-05,
+      "loss": 1.2642,
+      "step": 40535
+    },
+    {
+      "epoch": 0.8108,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.1027306998932627e-05,
+      "loss": 1.2861,
+      "step": 40540
+    },
+    {
+      "epoch": 0.8109,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.100589806808597e-05,
+      "loss": 1.2646,
+      "step": 40545
+    },
+    {
+      "epoch": 0.811,
+      "grad_norm": 0.294921875,
+      "learning_rate": 2.098449876243096e-05,
+      "loss": 1.2739,
+      "step": 40550
+    },
+    {
+      "epoch": 0.8111,
+      "grad_norm": 0.314453125,
+      "learning_rate": 2.096310908457513e-05,
+      "loss": 1.273,
+      "step": 40555
+    },
+    {
+      "epoch": 0.8112,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.09417290371247e-05,
+      "loss": 1.2715,
+      "step": 40560
+    },
+    {
+      "epoch": 0.8113,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.0920358622684788e-05,
+      "loss": 1.2828,
+      "step": 40565
+    },
+    {
+      "epoch": 0.8114,
+      "grad_norm": 0.5,
+      "learning_rate": 2.0898997843859334e-05,
+      "loss": 1.266,
+      "step": 40570
+    },
+    {
+      "epoch": 0.8115,
+      "grad_norm": 0.3828125,
+      "learning_rate": 2.0877646703250996e-05,
+      "loss": 1.2746,
+      "step": 40575
+    },
+    {
+      "epoch": 0.8116,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.0856305203461436e-05,
+      "loss": 1.2753,
+      "step": 40580
+    },
+    {
+      "epoch": 0.8117,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 2.0834973347091014e-05,
+      "loss": 1.272,
+      "step": 40585
+    },
+    {
+      "epoch": 0.8118,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.0813651136738954e-05,
+      "loss": 1.2529,
+      "step": 40590
+    },
+    {
+      "epoch": 0.8119,
+      "grad_norm": 0.375,
+      "learning_rate": 2.07923385750033e-05,
+      "loss": 1.2599,
+      "step": 40595
+    },
+    {
+      "epoch": 0.812,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.0771035664480942e-05,
+      "loss": 1.2778,
+      "step": 40600
+    },
+    {
+      "epoch": 0.8121,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.0749742407767546e-05,
+      "loss": 1.2937,
+      "step": 40605
+    },
+    {
+      "epoch": 0.8122,
+      "grad_norm": 0.30078125,
+      "learning_rate": 2.0728458807457662e-05,
+      "loss": 1.2643,
+      "step": 40610
+    },
+    {
+      "epoch": 0.8123,
+      "grad_norm": 0.4921875,
+      "learning_rate": 2.0707184866144603e-05,
+      "loss": 1.2746,
+      "step": 40615
+    },
+    {
+      "epoch": 0.8124,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.068592058642056e-05,
+      "loss": 1.2731,
+      "step": 40620
+    },
+    {
+      "epoch": 0.8125,
+      "grad_norm": 0.283203125,
+      "learning_rate": 2.0664665970876496e-05,
+      "loss": 1.2703,
+      "step": 40625
+    },
+    {
+      "epoch": 0.8126,
+      "grad_norm": 0.37890625,
+      "learning_rate": 2.0643421022102217e-05,
+      "loss": 1.2727,
+      "step": 40630
+    },
+    {
+      "epoch": 0.8127,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.0622185742686416e-05,
+      "loss": 1.2674,
+      "step": 40635
+    },
+    {
+      "epoch": 0.8128,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.0600960135216462e-05,
+      "loss": 1.2725,
+      "step": 40640
+    },
+    {
+      "epoch": 0.8129,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.0579744202278718e-05,
+      "loss": 1.2698,
+      "step": 40645
+    },
+    {
+      "epoch": 0.813,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.0558537946458177e-05,
+      "loss": 1.2943,
+      "step": 40650
+    },
+    {
+      "epoch": 0.8131,
+      "grad_norm": 0.275390625,
+      "learning_rate": 2.0537341370338857e-05,
+      "loss": 1.2779,
+      "step": 40655
+    },
+    {
+      "epoch": 0.8132,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.051615447650347e-05,
+      "loss": 1.2609,
+      "step": 40660
+    },
+    {
+      "epoch": 0.8133,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.049497726753351e-05,
+      "loss": 1.2635,
+      "step": 40665
+    },
+    {
+      "epoch": 0.8134,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.0473809746009444e-05,
+      "loss": 1.2818,
+      "step": 40670
+    },
+    {
+      "epoch": 0.8135,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.045265191451041e-05,
+      "loss": 1.2825,
+      "step": 40675
+    },
+    {
+      "epoch": 0.8136,
+      "grad_norm": 0.30859375,
+      "learning_rate": 2.0431503775614457e-05,
+      "loss": 1.2719,
+      "step": 40680
+    },
+    {
+      "epoch": 0.8137,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.0410365331898416e-05,
+      "loss": 1.2612,
+      "step": 40685
+    },
+    {
+      "epoch": 0.8138,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.0389236585937945e-05,
+      "loss": 1.2829,
+      "step": 40690
+    },
+    {
+      "epoch": 0.8139,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.0368117540307496e-05,
+      "loss": 1.2814,
+      "step": 40695
+    },
+    {
+      "epoch": 0.814,
+      "grad_norm": 0.228515625,
+      "learning_rate": 2.0347008197580374e-05,
+      "loss": 1.2933,
+      "step": 40700
+    },
+    {
+      "epoch": 0.8141,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.03259085603287e-05,
+      "loss": 1.2695,
+      "step": 40705
+    },
+    {
+      "epoch": 0.8142,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.030481863112339e-05,
+      "loss": 1.2652,
+      "step": 40710
+    },
+    {
+      "epoch": 0.8143,
+      "grad_norm": 0.291015625,
+      "learning_rate": 2.028373841253419e-05,
+      "loss": 1.2931,
+      "step": 40715
+    },
+    {
+      "epoch": 0.8144,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.026266790712965e-05,
+      "loss": 1.2569,
+      "step": 40720
+    },
+    {
+      "epoch": 0.8145,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.024160711747717e-05,
+      "loss": 1.2709,
+      "step": 40725
+    },
+    {
+      "epoch": 0.8146,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.0220556046142893e-05,
+      "loss": 1.2806,
+      "step": 40730
+    },
+    {
+      "epoch": 0.8147,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 2.019951469569191e-05,
+      "loss": 1.274,
+      "step": 40735
+    },
+    {
+      "epoch": 0.8148,
+      "grad_norm": 0.203125,
+      "learning_rate": 2.017848306868797e-05,
+      "loss": 1.2819,
+      "step": 40740
+    },
+    {
+      "epoch": 0.8149,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.0157461167693758e-05,
+      "loss": 1.28,
+      "step": 40745
+    },
+    {
+      "epoch": 0.815,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.013644899527074e-05,
+      "loss": 1.2648,
+      "step": 40750
+    },
+    {
+      "epoch": 0.8151,
+      "grad_norm": 0.37890625,
+      "learning_rate": 2.01154465539791e-05,
+      "loss": 1.2567,
+      "step": 40755
+    },
+    {
+      "epoch": 0.8152,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.009445384637805e-05,
+      "loss": 1.2671,
+      "step": 40760
+    },
+    {
+      "epoch": 0.8153,
+      "grad_norm": 0.318359375,
+      "learning_rate": 2.0073470875025358e-05,
+      "loss": 1.2663,
+      "step": 40765
+    },
+    {
+      "epoch": 0.8154,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 2.005249764247783e-05,
+      "loss": 1.2871,
+      "step": 40770
+    },
+    {
+      "epoch": 0.8155,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 2.0031534151290943e-05,
+      "loss": 1.2591,
+      "step": 40775
+    },
+    {
+      "epoch": 0.8156,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.0010580404019062e-05,
+      "loss": 1.2698,
+      "step": 40780
+    },
+    {
+      "epoch": 0.8157,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9989636403215328e-05,
+      "loss": 1.2727,
+      "step": 40785
+    },
+    {
+      "epoch": 0.8158,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.9968702151431696e-05,
+      "loss": 1.2697,
+      "step": 40790
+    },
+    {
+      "epoch": 0.8159,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9947777651218946e-05,
+      "loss": 1.2856,
+      "step": 40795
+    },
+    {
+      "epoch": 0.816,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.9926862905126665e-05,
+      "loss": 1.2955,
+      "step": 40800
+    },
+    {
+      "epoch": 0.8161,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.9905957915703244e-05,
+      "loss": 1.2623,
+      "step": 40805
+    },
+    {
+      "epoch": 0.8162,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.9885062685495904e-05,
+      "loss": 1.2616,
+      "step": 40810
+    },
+    {
+      "epoch": 0.8163,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.9864177217050674e-05,
+      "loss": 1.2659,
+      "step": 40815
+    },
+    {
+      "epoch": 0.8164,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.9843301512912327e-05,
+      "loss": 1.2708,
+      "step": 40820
+    },
+    {
+      "epoch": 0.8165,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9822435575624608e-05,
+      "loss": 1.2604,
+      "step": 40825
+    },
+    {
+      "epoch": 0.8166,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 1.9801579407729866e-05,
+      "loss": 1.2825,
+      "step": 40830
+    },
+    {
+      "epoch": 0.8167,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.9780733011769447e-05,
+      "loss": 1.2697,
+      "step": 40835
+    },
+    {
+      "epoch": 0.8168,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 1.9759896390283362e-05,
+      "loss": 1.2608,
+      "step": 40840
+    },
+    {
+      "epoch": 0.8169,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.9739069545810485e-05,
+      "loss": 1.2725,
+      "step": 40845
+    },
+    {
+      "epoch": 0.817,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.9718252480888566e-05,
+      "loss": 1.2692,
+      "step": 40850
+    },
+    {
+      "epoch": 0.8171,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.969744519805402e-05,
+      "loss": 1.283,
+      "step": 40855
+    },
+    {
+      "epoch": 0.8172,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.9676647699842242e-05,
+      "loss": 1.2643,
+      "step": 40860
+    },
+    {
+      "epoch": 0.8173,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.965585998878724e-05,
+      "loss": 1.2645,
+      "step": 40865
+    },
+    {
+      "epoch": 0.8174,
+      "grad_norm": 0.18359375,
+      "learning_rate": 1.963508206742202e-05,
+      "loss": 1.2928,
+      "step": 40870
+    },
+    {
+      "epoch": 0.8175,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9614313938278272e-05,
+      "loss": 1.2691,
+      "step": 40875
+    },
+    {
+      "epoch": 0.8176,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.9593555603886538e-05,
+      "loss": 1.2809,
+      "step": 40880
+    },
+    {
+      "epoch": 0.8177,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.9572807066776143e-05,
+      "loss": 1.2704,
+      "step": 40885
+    },
+    {
+      "epoch": 0.8178,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.955206832947526e-05,
+      "loss": 1.2808,
+      "step": 40890
+    },
+    {
+      "epoch": 0.8179,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.9531339394510827e-05,
+      "loss": 1.2718,
+      "step": 40895
+    },
+    {
+      "epoch": 0.818,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.9510620264408596e-05,
+      "loss": 1.2768,
+      "step": 40900
+    },
+    {
+      "epoch": 0.8181,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 1.9489910941693133e-05,
+      "loss": 1.2701,
+      "step": 40905
+    },
+    {
+      "epoch": 0.8182,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.946921142888781e-05,
+      "loss": 1.2645,
+      "step": 40910
+    },
+    {
+      "epoch": 0.8183,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.9448521728514802e-05,
+      "loss": 1.2849,
+      "step": 40915
+    },
+    {
+      "epoch": 0.8184,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9427841843095063e-05,
+      "loss": 1.2647,
+      "step": 40920
+    },
+    {
+      "epoch": 0.8185,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 1.9407171775148436e-05,
+      "loss": 1.2655,
+      "step": 40925
+    },
+    {
+      "epoch": 0.8186,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.938651152719344e-05,
+      "loss": 1.272,
+      "step": 40930
+    },
+    {
+      "epoch": 0.8187,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9365861101747485e-05,
+      "loss": 1.2736,
+      "step": 40935
+    },
+    {
+      "epoch": 0.8188,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.9345220501326777e-05,
+      "loss": 1.2688,
+      "step": 40940
+    },
+    {
+      "epoch": 0.8189,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.9324589728446262e-05,
+      "loss": 1.2648,
+      "step": 40945
+    },
+    {
+      "epoch": 0.819,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.930396878561983e-05,
+      "loss": 1.2732,
+      "step": 40950
+    },
+    {
+      "epoch": 0.8191,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.928335767535997e-05,
+      "loss": 1.2762,
+      "step": 40955
+    },
+    {
+      "epoch": 0.8192,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9262756400178162e-05,
+      "loss": 1.2662,
+      "step": 40960
+    },
+    {
+      "epoch": 0.8193,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.9242164962584618e-05,
+      "loss": 1.2782,
+      "step": 40965
+    },
+    {
+      "epoch": 0.8194,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 1.922158336508825e-05,
+      "loss": 1.2634,
+      "step": 40970
+    },
+    {
+      "epoch": 0.8195,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 1.9201011610196973e-05,
+      "loss": 1.2746,
+      "step": 40975
+    },
+    {
+      "epoch": 0.8196,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.918044970041729e-05,
+      "loss": 1.2618,
+      "step": 40980
+    },
+    {
+      "epoch": 0.8197,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.91598976382547e-05,
+      "loss": 1.283,
+      "step": 40985
+    },
+    {
+      "epoch": 0.8198,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9139355426213347e-05,
+      "loss": 1.2865,
+      "step": 40990
+    },
+    {
+      "epoch": 0.8199,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.9118823066796276e-05,
+      "loss": 1.2809,
+      "step": 40995
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9098300562505266e-05,
+      "loss": 1.2795,
+      "step": 41000
+    },
+    {
+      "epoch": 0.8201,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9077787915840927e-05,
+      "loss": 1.2757,
+      "step": 41005
+    },
+    {
+      "epoch": 0.8202,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9057285129302683e-05,
+      "loss": 1.2713,
+      "step": 41010
+    },
+    {
+      "epoch": 0.8203,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 1.903679220538871e-05,
+      "loss": 1.2712,
+      "step": 41015
+    },
+    {
+      "epoch": 0.8204,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 1.9016309146596023e-05,
+      "loss": 1.2715,
+      "step": 41020
+    },
+    {
+      "epoch": 0.8205,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8995835955420416e-05,
+      "loss": 1.2722,
+      "step": 41025
+    },
+    {
+      "epoch": 0.8206,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.897537263435648e-05,
+      "loss": 1.2764,
+      "step": 41030
+    },
+    {
+      "epoch": 0.8207,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.895491918589759e-05,
+      "loss": 1.2804,
+      "step": 41035
+    },
+    {
+      "epoch": 0.8208,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.8934475612536017e-05,
+      "loss": 1.274,
+      "step": 41040
+    },
+    {
+      "epoch": 0.8209,
+      "grad_norm": 0.18359375,
+      "learning_rate": 1.891404191676265e-05,
+      "loss": 1.2895,
+      "step": 41045
+    },
+    {
+      "epoch": 0.821,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.8893618101067355e-05,
+      "loss": 1.269,
+      "step": 41050
+    },
+    {
+      "epoch": 0.8211,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.8873204167938653e-05,
+      "loss": 1.2757,
+      "step": 41055
+    },
+    {
+      "epoch": 0.8212,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.885280011986391e-05,
+      "loss": 1.2624,
+      "step": 41060
+    },
+    {
+      "epoch": 0.8213,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.883240595932938e-05,
+      "loss": 1.2903,
+      "step": 41065
+    },
+    {
+      "epoch": 0.8214,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.8812021688819915e-05,
+      "loss": 1.2792,
+      "step": 41070
+    },
+    {
+      "epoch": 0.8215,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.879164731081937e-05,
+      "loss": 1.2708,
+      "step": 41075
+    },
+    {
+      "epoch": 0.8216,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.877128282781028e-05,
+      "loss": 1.2789,
+      "step": 41080
+    },
+    {
+      "epoch": 0.8217,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8750928242273968e-05,
+      "loss": 1.267,
+      "step": 41085
+    },
+    {
+      "epoch": 0.8218,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8730583556690605e-05,
+      "loss": 1.2821,
+      "step": 41090
+    },
+    {
+      "epoch": 0.8219,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.8710248773539118e-05,
+      "loss": 1.2787,
+      "step": 41095
+    },
+    {
+      "epoch": 0.822,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.8689923895297245e-05,
+      "loss": 1.2762,
+      "step": 41100
+    },
+    {
+      "epoch": 0.8221,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.8669608924441496e-05,
+      "loss": 1.2658,
+      "step": 41105
+    },
+    {
+      "epoch": 0.8222,
+      "grad_norm": 0.228515625,
+      "learning_rate": 1.8649303863447198e-05,
+      "loss": 1.2796,
+      "step": 41110
+    },
+    {
+      "epoch": 0.8223,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.8629008714788464e-05,
+      "loss": 1.2853,
+      "step": 41115
+    },
+    {
+      "epoch": 0.8224,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8608723480938206e-05,
+      "loss": 1.257,
+      "step": 41120
+    },
+    {
+      "epoch": 0.8225,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.858844816436809e-05,
+      "loss": 1.2681,
+      "step": 41125
+    },
+    {
+      "epoch": 0.8226,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.8568182767548626e-05,
+      "loss": 1.2697,
+      "step": 41130
+    },
+    {
+      "epoch": 0.8227,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 1.854792729294905e-05,
+      "loss": 1.2863,
+      "step": 41135
+    },
+    {
+      "epoch": 0.8228,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.852768174303752e-05,
+      "loss": 1.2854,
+      "step": 41140
+    },
+    {
+      "epoch": 0.8229,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.8507446120280814e-05,
+      "loss": 1.2603,
+      "step": 41145
+    },
+    {
+      "epoch": 0.823,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 1.848722042714457e-05,
+      "loss": 1.2819,
+      "step": 41150
+    },
+    {
+      "epoch": 0.8231,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.8467004666093325e-05,
+      "loss": 1.279,
+      "step": 41155
+    },
+    {
+      "epoch": 0.8232,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8446798839590186e-05,
+      "loss": 1.2747,
+      "step": 41160
+    },
+    {
+      "epoch": 0.8233,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8426602950097283e-05,
+      "loss": 1.2769,
+      "step": 41165
+    },
+    {
+      "epoch": 0.8234,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.8406417000075325e-05,
+      "loss": 1.2715,
+      "step": 41170
+    },
+    {
+      "epoch": 0.8235,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.838624099198397e-05,
+      "loss": 1.2742,
+      "step": 41175
+    },
+    {
+      "epoch": 0.8236,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.8366074928281607e-05,
+      "loss": 1.2698,
+      "step": 41180
+    },
+    {
+      "epoch": 0.8237,
+      "grad_norm": 0.228515625,
+      "learning_rate": 1.834591881142538e-05,
+      "loss": 1.2639,
+      "step": 41185
+    },
+    {
+      "epoch": 0.8238,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.8325772643871265e-05,
+      "loss": 1.2701,
+      "step": 41190
+    },
+    {
+      "epoch": 0.8239,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8305636428074014e-05,
+      "loss": 1.2633,
+      "step": 41195
+    },
+    {
+      "epoch": 0.824,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.8285510166487152e-05,
+      "loss": 1.2764,
+      "step": 41200
+    },
+    {
+      "epoch": 0.8241,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.826539386156302e-05,
+      "loss": 1.2952,
+      "step": 41205
+    },
+    {
+      "epoch": 0.8242,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.8245287515752708e-05,
+      "loss": 1.2899,
+      "step": 41210
+    },
+    {
+      "epoch": 0.8243,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.8225191131506126e-05,
+      "loss": 1.2954,
+      "step": 41215
+    },
+    {
+      "epoch": 0.8244,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.8205104711271957e-05,
+      "loss": 1.2662,
+      "step": 41220
+    },
+    {
+      "epoch": 0.8245,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 1.818502825749764e-05,
+      "loss": 1.2944,
+      "step": 41225
+    },
+    {
+      "epoch": 0.8246,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.816496177262952e-05,
+      "loss": 1.2651,
+      "step": 41230
+    },
+    {
+      "epoch": 0.8247,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.8144905259112542e-05,
+      "loss": 1.2664,
+      "step": 41235
+    },
+    {
+      "epoch": 0.8248,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.812485871939056e-05,
+      "loss": 1.2679,
+      "step": 41240
+    },
+    {
+      "epoch": 0.8249,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.81048221559062e-05,
+      "loss": 1.2806,
+      "step": 41245
+    },
+    {
+      "epoch": 0.825,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 1.808479557110081e-05,
+      "loss": 1.269,
+      "step": 41250
+    },
+    {
+      "epoch": 0.8251,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.8064778967414662e-05,
+      "loss": 1.2807,
+      "step": 41255
+    },
+    {
+      "epoch": 0.8252,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.804477234728661e-05,
+      "loss": 1.2709,
+      "step": 41260
+    },
+    {
+      "epoch": 0.8253,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.8024775713154473e-05,
+      "loss": 1.2801,
+      "step": 41265
+    },
+    {
+      "epoch": 0.8254,
+      "grad_norm": 0.240234375,
+      "learning_rate": 1.8004789067454764e-05,
+      "loss": 1.2729,
+      "step": 41270
+    },
+    {
+      "epoch": 0.8255,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.7984812412622787e-05,
+      "loss": 1.2849,
+      "step": 41275
+    },
+    {
+      "epoch": 0.8256,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7964845751092664e-05,
+      "loss": 1.2804,
+      "step": 41280
+    },
+    {
+      "epoch": 0.8257,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.794488908529719e-05,
+      "loss": 1.271,
+      "step": 41285
+    },
+    {
+      "epoch": 0.8258,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.792494241766811e-05,
+      "loss": 1.2734,
+      "step": 41290
+    },
+    {
+      "epoch": 0.8259,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.790500575063584e-05,
+      "loss": 1.2678,
+      "step": 41295
+    },
+    {
+      "epoch": 0.826,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.78850790866296e-05,
+      "loss": 1.2796,
+      "step": 41300
+    },
+    {
+      "epoch": 0.8261,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 1.7865162428077386e-05,
+      "loss": 1.2693,
+      "step": 41305
+    },
+    {
+      "epoch": 0.8262,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 1.7845255777406e-05,
+      "loss": 1.2637,
+      "step": 41310
+    },
+    {
+      "epoch": 0.8263,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.7825359137040988e-05,
+      "loss": 1.2856,
+      "step": 41315
+    },
+    {
+      "epoch": 0.8264,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7805472509406696e-05,
+      "loss": 1.2681,
+      "step": 41320
+    },
+    {
+      "epoch": 0.8265,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.7785595896926265e-05,
+      "loss": 1.2599,
+      "step": 41325
+    },
+    {
+      "epoch": 0.8266,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 1.7765729302021596e-05,
+      "loss": 1.2842,
+      "step": 41330
+    },
+    {
+      "epoch": 0.8267,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.7745872727113356e-05,
+      "loss": 1.27,
+      "step": 41335
+    },
+    {
+      "epoch": 0.8268,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.7726026174621003e-05,
+      "loss": 1.2744,
+      "step": 41340
+    },
+    {
+      "epoch": 0.8269,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.7706189646962847e-05,
+      "loss": 1.2779,
+      "step": 41345
+    },
+    {
+      "epoch": 0.827,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.7686363146555805e-05,
+      "loss": 1.2741,
+      "step": 41350
+    },
+    {
+      "epoch": 0.8271,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 1.7666546675815778e-05,
+      "loss": 1.2686,
+      "step": 41355
+    },
+    {
+      "epoch": 0.8272,
+      "grad_norm": 0.232421875,
+      "learning_rate": 1.7646740237157256e-05,
+      "loss": 1.2685,
+      "step": 41360
+    },
+    {
+      "epoch": 0.8273,
+      "grad_norm": 0.23828125,
+      "learning_rate": 1.7626943832993647e-05,
+      "loss": 1.2745,
+      "step": 41365
+    },
+    {
+      "epoch": 0.8274,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 1.760715746573709e-05,
+      "loss": 1.2683,
+      "step": 41370
+    },
+    {
+      "epoch": 0.8275,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.7587381137798432e-05,
+      "loss": 1.2701,
+      "step": 41375
+    },
+    {
+      "epoch": 0.8276,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 1.7567614851587443e-05,
+      "loss": 1.2966,
+      "step": 41380
+    },
+    {
+      "epoch": 0.8277,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.754785860951249e-05,
+      "loss": 1.2689,
+      "step": 41385
+    },
+    {
+      "epoch": 0.8278,
+      "grad_norm": 0.205078125,
+      "learning_rate": 1.752811241398089e-05,
+      "loss": 1.2721,
+      "step": 41390
+    },
+    {
+      "epoch": 0.8279,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.750837626739863e-05,
+      "loss": 1.2875,
+      "step": 41395
+    },
+    {
+      "epoch": 0.828,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 1.7488650172170496e-05,
+      "loss": 1.2783,
+      "step": 41400
+    },
+    {
+      "epoch": 0.8281,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.7468934130700044e-05,
+      "loss": 1.2696,
+      "step": 41405
+    },
+    {
+      "epoch": 0.8282,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.744922814538964e-05,
+      "loss": 1.2802,
+      "step": 41410
+    },
+    {
+      "epoch": 0.8283,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 1.7429532218640377e-05,
+      "loss": 1.262,
+      "step": 41415
+    },
+    {
+      "epoch": 0.8284,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.7409846352852143e-05,
+      "loss": 1.2727,
+      "step": 41420
+    },
+    {
+      "epoch": 0.8285,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.7390170550423625e-05,
+      "loss": 1.2715,
+      "step": 41425
+    },
+    {
+      "epoch": 0.8286,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.737050481375223e-05,
+      "loss": 1.2719,
+      "step": 41430
+    },
+    {
+      "epoch": 0.8287,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7350849145234184e-05,
+      "loss": 1.2642,
+      "step": 41435
+    },
+    {
+      "epoch": 0.8288,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 1.7331203547264453e-05,
+      "loss": 1.2614,
+      "step": 41440
+    },
+    {
+      "epoch": 0.8289,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.7311568022236845e-05,
+      "loss": 1.2802,
+      "step": 41445
+    },
+    {
+      "epoch": 0.829,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 1.7291942572543807e-05,
+      "loss": 1.253,
+      "step": 41450
+    },
+    {
+      "epoch": 0.8291,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.7272327200576742e-05,
+      "loss": 1.2747,
+      "step": 41455
+    },
+    {
+      "epoch": 0.8292,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7252721908725632e-05,
+      "loss": 1.274,
+      "step": 41460
+    },
+    {
+      "epoch": 0.8293,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 1.7233126699379343e-05,
+      "loss": 1.2718,
+      "step": 41465
+    },
+    {
+      "epoch": 0.8294,
+      "grad_norm": 0.24609375,
+      "learning_rate": 1.721354157492555e-05,
+      "loss": 1.2703,
+      "step": 41470
+    },
+    {
+      "epoch": 0.8295,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.719396653775056e-05,
+      "loss": 1.2817,
+      "step": 41475
+    },
+    {
+      "epoch": 0.8296,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 1.7174401590239587e-05,
+      "loss": 1.255,
+      "step": 41480
+    },
+    {
+      "epoch": 0.8297,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.715484673477654e-05,
+      "loss": 1.2746,
+      "step": 41485
+    },
+    {
+      "epoch": 0.8298,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 1.7135301973744124e-05,
+      "loss": 1.2794,
+      "step": 41490
+    },
+    {
+      "epoch": 0.8299,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7115767309523812e-05,
+      "loss": 1.2824,
+      "step": 41495
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.7096242744495837e-05,
+      "loss": 1.2777,
+      "step": 41500
+    },
+    {
+      "epoch": 0.8301,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.70767282810392e-05,
+      "loss": 1.2665,
+      "step": 41505
+    },
+    {
+      "epoch": 0.8302,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7057223921531707e-05,
+      "loss": 1.274,
+      "step": 41510
+    },
+    {
+      "epoch": 0.8303,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.7037729668349877e-05,
+      "loss": 1.2619,
+      "step": 41515
+    },
+    {
+      "epoch": 0.8304,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.7018245523869036e-05,
+      "loss": 1.272,
+      "step": 41520
+    },
+    {
+      "epoch": 0.8305,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.6998771490463262e-05,
+      "loss": 1.2673,
+      "step": 41525
+    },
+    {
+      "epoch": 0.8306,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.697930757050542e-05,
+      "loss": 1.2672,
+      "step": 41530
+    },
+    {
+      "epoch": 0.8307,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.6959853766367118e-05,
+      "loss": 1.2879,
+      "step": 41535
+    },
+    {
+      "epoch": 0.8308,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.6940410080418723e-05,
+      "loss": 1.2845,
+      "step": 41540
+    },
+    {
+      "epoch": 0.8309,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 1.6920976515029463e-05,
+      "loss": 1.2843,
+      "step": 41545
+    },
+    {
+      "epoch": 0.831,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.690155307256719e-05,
+      "loss": 1.287,
+      "step": 41550
+    },
+    {
+      "epoch": 0.8311,
+      "grad_norm": 0.185546875,
+      "learning_rate": 1.6882139755398574e-05,
+      "loss": 1.2673,
+      "step": 41555
+    },
+    {
+      "epoch": 0.8312,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.686273656588917e-05,
+      "loss": 1.2869,
+      "step": 41560
+    },
+    {
+      "epoch": 0.8313,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.6843343506403075e-05,
+      "loss": 1.2686,
+      "step": 41565
+    },
+    {
+      "epoch": 0.8314,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.6823960579303377e-05,
+      "loss": 1.2637,
+      "step": 41570
+    },
+    {
+      "epoch": 0.8315,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.680458778695174e-05,
+      "loss": 1.2777,
+      "step": 41575
+    },
+    {
+      "epoch": 0.8316,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.678522513170875e-05,
+      "loss": 1.2627,
+      "step": 41580
+    },
+    {
+      "epoch": 0.8317,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6765872615933677e-05,
+      "loss": 1.2677,
+      "step": 41585
+    },
+    {
+      "epoch": 0.8318,
+      "grad_norm": 0.16015625,
+      "learning_rate": 1.6746530241984504e-05,
+      "loss": 1.2646,
+      "step": 41590
+    },
+    {
+      "epoch": 0.8319,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.6727198012218114e-05,
+      "loss": 1.2693,
+      "step": 41595
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6707875928990058e-05,
+      "loss": 1.2904,
+      "step": 41600
+    },
+    {
+      "epoch": 0.8321,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.668856399465466e-05,
+      "loss": 1.2734,
+      "step": 41605
+    },
+    {
+      "epoch": 0.8322,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.6669262211565028e-05,
+      "loss": 1.2651,
+      "step": 41610
+    },
+    {
+      "epoch": 0.8323,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.6649970582073028e-05,
+      "loss": 1.2869,
+      "step": 41615
+    },
+    {
+      "epoch": 0.8324,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.6630689108529284e-05,
+      "loss": 1.2823,
+      "step": 41620
+    },
+    {
+      "epoch": 0.8325,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 1.661141779328319e-05,
+      "loss": 1.2686,
+      "step": 41625
+    },
+    {
+      "epoch": 0.8326,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.6592156638682886e-05,
+      "loss": 1.2899,
+      "step": 41630
+    },
+    {
+      "epoch": 0.8327,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.6572905647075298e-05,
+      "loss": 1.2659,
+      "step": 41635
+    },
+    {
+      "epoch": 0.8328,
+      "grad_norm": 0.2265625,
+      "learning_rate": 1.6553664820806102e-05,
+      "loss": 1.2799,
+      "step": 41640
+    },
+    {
+      "epoch": 0.8329,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.6534434162219727e-05,
+      "loss": 1.2855,
+      "step": 41645
+    },
+    {
+      "epoch": 0.833,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.6515213673659357e-05,
+      "loss": 1.2698,
+      "step": 41650
+    },
+    {
+      "epoch": 0.8331,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.649600335746695e-05,
+      "loss": 1.2816,
+      "step": 41655
+    },
+    {
+      "epoch": 0.8332,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.6476803215983294e-05,
+      "loss": 1.2638,
+      "step": 41660
+    },
+    {
+      "epoch": 0.8333,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6457613251547754e-05,
+      "loss": 1.2643,
+      "step": 41665
+    },
+    {
+      "epoch": 0.8334,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.643843346649866e-05,
+      "loss": 1.29,
+      "step": 41670
+    },
+    {
+      "epoch": 0.8335,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6419263863172997e-05,
+      "loss": 1.2776,
+      "step": 41675
+    },
+    {
+      "epoch": 0.8336,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.640010444390646e-05,
+      "loss": 1.2735,
+      "step": 41680
+    },
+    {
+      "epoch": 0.8337,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.6380955211033656e-05,
+      "loss": 1.2706,
+      "step": 41685
+    },
+    {
+      "epoch": 0.8338,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 1.6361816166887768e-05,
+      "loss": 1.2693,
+      "step": 41690
+    },
+    {
+      "epoch": 0.8339,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.634268731380091e-05,
+      "loss": 1.2755,
+      "step": 41695
+    },
+    {
+      "epoch": 0.834,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.632356865410384e-05,
+      "loss": 1.2732,
+      "step": 41700
+    },
+    {
+      "epoch": 0.8341,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.6304460190126103e-05,
+      "loss": 1.2901,
+      "step": 41705
+    },
+    {
+      "epoch": 0.8342,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.628536192419603e-05,
+      "loss": 1.2768,
+      "step": 41710
+    },
+    {
+      "epoch": 0.8343,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.6266273858640656e-05,
+      "loss": 1.2797,
+      "step": 41715
+    },
+    {
+      "epoch": 0.8344,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.6247195995785837e-05,
+      "loss": 1.2944,
+      "step": 41720
+    },
+    {
+      "epoch": 0.8345,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.622812833795613e-05,
+      "loss": 1.2739,
+      "step": 41725
+    },
+    {
+      "epoch": 0.8346,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6209070887474876e-05,
+      "loss": 1.2793,
+      "step": 41730
+    },
+    {
+      "epoch": 0.8347,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.6190023646664175e-05,
+      "loss": 1.2859,
+      "step": 41735
+    },
+    {
+      "epoch": 0.8348,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.6170986617844863e-05,
+      "loss": 1.2655,
+      "step": 41740
+    },
+    {
+      "epoch": 0.8349,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6151959803336535e-05,
+      "loss": 1.2696,
+      "step": 41745
+    },
+    {
+      "epoch": 0.835,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.6132943205457606e-05,
+      "loss": 1.2721,
+      "step": 41750
+    },
+    {
+      "epoch": 0.8351,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 1.611393682652511e-05,
+      "loss": 1.2919,
+      "step": 41755
+    },
+    {
+      "epoch": 0.8352,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.6094940668855006e-05,
+      "loss": 1.2733,
+      "step": 41760
+    },
+    {
+      "epoch": 0.8353,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 1.6075954734761845e-05,
+      "loss": 1.2804,
+      "step": 41765
+    },
+    {
+      "epoch": 0.8354,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.6056979026559004e-05,
+      "loss": 1.2838,
+      "step": 41770
+    },
+    {
+      "epoch": 0.8355,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.6038013546558696e-05,
+      "loss": 1.2856,
+      "step": 41775
+    },
+    {
+      "epoch": 0.8356,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.601905829707171e-05,
+      "loss": 1.283,
+      "step": 41780
+    },
+    {
+      "epoch": 0.8357,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.600011328040777e-05,
+      "loss": 1.272,
+      "step": 41785
+    },
+    {
+      "epoch": 0.8358,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.598117849887518e-05,
+      "loss": 1.279,
+      "step": 41790
+    },
+    {
+      "epoch": 0.8359,
+      "grad_norm": 0.205078125,
+      "learning_rate": 1.596225395478116e-05,
+      "loss": 1.2713,
+      "step": 41795
+    },
+    {
+      "epoch": 0.836,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.5943339650431576e-05,
+      "loss": 1.2698,
+      "step": 41800
+    },
+    {
+      "epoch": 0.8361,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.592443558813109e-05,
+      "loss": 1.28,
+      "step": 41805
+    },
+    {
+      "epoch": 0.8362,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.5905541770183096e-05,
+      "loss": 1.271,
+      "step": 41810
+    },
+    {
+      "epoch": 0.8363,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.588665819888976e-05,
+      "loss": 1.291,
+      "step": 41815
+    },
+    {
+      "epoch": 0.8364,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.586778487655197e-05,
+      "loss": 1.2684,
+      "step": 41820
+    },
+    {
+      "epoch": 0.8365,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.5848921805469397e-05,
+      "loss": 1.2736,
+      "step": 41825
+    },
+    {
+      "epoch": 0.8366,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 1.5830068987940438e-05,
+      "loss": 1.2581,
+      "step": 41830
+    },
+    {
+      "epoch": 0.8367,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.581122642626226e-05,
+      "loss": 1.2826,
+      "step": 41835
+    },
+    {
+      "epoch": 0.8368,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.579239412273078e-05,
+      "loss": 1.271,
+      "step": 41840
+    },
+    {
+      "epoch": 0.8369,
+      "grad_norm": 0.240234375,
+      "learning_rate": 1.577357207964062e-05,
+      "loss": 1.2827,
+      "step": 41845
+    },
+    {
+      "epoch": 0.837,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.5754760299285252e-05,
+      "loss": 1.2776,
+      "step": 41850
+    },
+    {
+      "epoch": 0.8371,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.5735958783956794e-05,
+      "loss": 1.2643,
+      "step": 41855
+    },
+    {
+      "epoch": 0.8372,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.5717167535946144e-05,
+      "loss": 1.2816,
+      "step": 41860
+    },
+    {
+      "epoch": 0.8373,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.569838655754298e-05,
+      "loss": 1.2738,
+      "step": 41865
+    },
+    {
+      "epoch": 0.8374,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.567961585103567e-05,
+      "loss": 1.2866,
+      "step": 41870
+    },
+    {
+      "epoch": 0.8375,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.566085541871145e-05,
+      "loss": 1.2706,
+      "step": 41875
+    },
+    {
+      "epoch": 0.8376,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.564210526285612e-05,
+      "loss": 1.2899,
+      "step": 41880
+    },
+    {
+      "epoch": 0.8377,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.5623365385754408e-05,
+      "loss": 1.275,
+      "step": 41885
+    },
+    {
+      "epoch": 0.8378,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 1.560463578968967e-05,
+      "loss": 1.2758,
+      "step": 41890
+    },
+    {
+      "epoch": 0.8379,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.5585916476944073e-05,
+      "loss": 1.271,
+      "step": 41895
+    },
+    {
+      "epoch": 0.838,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.5567207449798515e-05,
+      "loss": 1.2659,
+      "step": 41900
+    },
+    {
+      "epoch": 0.8381,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.5548508710532572e-05,
+      "loss": 1.2749,
+      "step": 41905
+    },
+    {
+      "epoch": 0.8382,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.5529820261424698e-05,
+      "loss": 1.2764,
+      "step": 41910
+    },
+    {
+      "epoch": 0.8383,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.551114210475201e-05,
+      "loss": 1.2575,
+      "step": 41915
+    },
+    {
+      "epoch": 0.8384,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5492474242790366e-05,
+      "loss": 1.2711,
+      "step": 41920
+    },
+    {
+      "epoch": 0.8385,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.547381667781439e-05,
+      "loss": 1.2593,
+      "step": 41925
+    },
+    {
+      "epoch": 0.8386,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.545516941209747e-05,
+      "loss": 1.2713,
+      "step": 41930
+    },
+    {
+      "epoch": 0.8387,
+      "grad_norm": 0.158203125,
+      "learning_rate": 1.54365324479117e-05,
+      "loss": 1.2788,
+      "step": 41935
+    },
+    {
+      "epoch": 0.8388,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 1.541790578752794e-05,
+      "loss": 1.2883,
+      "step": 41940
+    },
+    {
+      "epoch": 0.8389,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.539928943321579e-05,
+      "loss": 1.2641,
+      "step": 41945
+    },
+    {
+      "epoch": 0.839,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.538068338724361e-05,
+      "loss": 1.2725,
+      "step": 41950
+    },
+    {
+      "epoch": 0.8391,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.5362087651878475e-05,
+      "loss": 1.2749,
+      "step": 41955
+    },
+    {
+      "epoch": 0.8392,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.5343502229386207e-05,
+      "loss": 1.2614,
+      "step": 41960
+    },
+    {
+      "epoch": 0.8393,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.532492712203145e-05,
+      "loss": 1.2619,
+      "step": 41965
+    },
+    {
+      "epoch": 0.8394,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.530636233207743e-05,
+      "loss": 1.2807,
+      "step": 41970
+    },
+    {
+      "epoch": 0.8395,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 1.528780786178631e-05,
+      "loss": 1.262,
+      "step": 41975
+    },
+    {
+      "epoch": 0.8396,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.526926371341878e-05,
+      "loss": 1.2697,
+      "step": 41980
+    },
+    {
+      "epoch": 0.8397,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.5250729889234482e-05,
+      "loss": 1.2578,
+      "step": 41985
+    },
+    {
+      "epoch": 0.8398,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.5232206391491699e-05,
+      "loss": 1.2767,
+      "step": 41990
+    },
+    {
+      "epoch": 0.8399,
+      "grad_norm": 0.234375,
+      "learning_rate": 1.521369322244739e-05,
+      "loss": 1.2772,
+      "step": 41995
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 1.5195190384357404e-05,
+      "loss": 1.2715,
+      "step": 42000
+    },
+    {
+      "epoch": 0.8401,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.5176697879476232e-05,
+      "loss": 1.2708,
+      "step": 42005
+    },
+    {
+      "epoch": 0.8402,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.5158215710057123e-05,
+      "loss": 1.2698,
+      "step": 42010
+    },
+    {
+      "epoch": 0.8403,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.5139743878352075e-05,
+      "loss": 1.2748,
+      "step": 42015
+    },
+    {
+      "epoch": 0.8404,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.5121282386611824e-05,
+      "loss": 1.2842,
+      "step": 42020
+    },
+    {
+      "epoch": 0.8405,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5102831237085857e-05,
+      "loss": 1.2719,
+      "step": 42025
+    },
+    {
+      "epoch": 0.8406,
+      "grad_norm": 0.228515625,
+      "learning_rate": 1.5084390432022377e-05,
+      "loss": 1.2875,
+      "step": 42030
+    },
+    {
+      "epoch": 0.8407,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.5065959973668353e-05,
+      "loss": 1.2761,
+      "step": 42035
+    },
+    {
+      "epoch": 0.8408,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5047539864269478e-05,
+      "loss": 1.2865,
+      "step": 42040
+    },
+    {
+      "epoch": 0.8409,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5029130106070167e-05,
+      "loss": 1.2775,
+      "step": 42045
+    },
+    {
+      "epoch": 0.841,
+      "grad_norm": 0.2265625,
+      "learning_rate": 1.5010730701313625e-05,
+      "loss": 1.268,
+      "step": 42050
+    },
+    {
+      "epoch": 0.8411,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.4992341652241737e-05,
+      "loss": 1.2636,
+      "step": 42055
+    },
+    {
+      "epoch": 0.8412,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 1.4973962961095134e-05,
+      "loss": 1.2778,
+      "step": 42060
+    },
+    {
+      "epoch": 0.8413,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.495559463011329e-05,
+      "loss": 1.2754,
+      "step": 42065
+    },
+    {
+      "epoch": 0.8414,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.4937236661534226e-05,
+      "loss": 1.2687,
+      "step": 42070
+    },
+    {
+      "epoch": 0.8415,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.4918889057594876e-05,
+      "loss": 1.2711,
+      "step": 42075
+    },
+    {
+      "epoch": 0.8416,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.4900551820530828e-05,
+      "loss": 1.2842,
+      "step": 42080
+    },
+    {
+      "epoch": 0.8417,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.4882224952576373e-05,
+      "loss": 1.2829,
+      "step": 42085
+    },
+    {
+      "epoch": 0.8418,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.486390845596466e-05,
+      "loss": 1.2713,
+      "step": 42090
+    },
+    {
+      "epoch": 0.8419,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.484560233292741e-05,
+      "loss": 1.2712,
+      "step": 42095
+    },
+    {
+      "epoch": 0.842,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.4827306585695234e-05,
+      "loss": 1.254,
+      "step": 42100
+    },
+    {
+      "epoch": 0.8421,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4809021216497399e-05,
+      "loss": 1.2684,
+      "step": 42105
+    },
+    {
+      "epoch": 0.8422,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4790746227561925e-05,
+      "loss": 1.2746,
+      "step": 42110
+    },
+    {
+      "epoch": 0.8423,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.4772481621115541e-05,
+      "loss": 1.2928,
+      "step": 42115
+    },
+    {
+      "epoch": 0.8424,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.4754227399383757e-05,
+      "loss": 1.2778,
+      "step": 42120
+    },
+    {
+      "epoch": 0.8425,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.4735983564590783e-05,
+      "loss": 1.2883,
+      "step": 42125
+    },
+    {
+      "epoch": 0.8426,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.4717750118959584e-05,
+      "loss": 1.272,
+      "step": 42130
+    },
+    {
+      "epoch": 0.8427,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.4699527064711838e-05,
+      "loss": 1.2764,
+      "step": 42135
+    },
+    {
+      "epoch": 0.8428,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.468131440406798e-05,
+      "loss": 1.2914,
+      "step": 42140
+    },
+    {
+      "epoch": 0.8429,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.466311213924716e-05,
+      "loss": 1.2711,
+      "step": 42145
+    },
+    {
+      "epoch": 0.843,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.4644920272467244e-05,
+      "loss": 1.2702,
+      "step": 42150
+    },
+    {
+      "epoch": 0.8431,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.462673880594494e-05,
+      "loss": 1.2868,
+      "step": 42155
+    },
+    {
+      "epoch": 0.8432,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4608567741895495e-05,
+      "loss": 1.2764,
+      "step": 42160
+    },
+    {
+      "epoch": 0.8433,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.4590407082533097e-05,
+      "loss": 1.2741,
+      "step": 42165
+    },
+    {
+      "epoch": 0.8434,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4572256830070497e-05,
+      "loss": 1.2796,
+      "step": 42170
+    },
+    {
+      "epoch": 0.8435,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.4554116986719257e-05,
+      "loss": 1.2694,
+      "step": 42175
+    },
+    {
+      "epoch": 0.8436,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.4535987554689712e-05,
+      "loss": 1.261,
+      "step": 42180
+    },
+    {
+      "epoch": 0.8437,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.4517868536190803e-05,
+      "loss": 1.2692,
+      "step": 42185
+    },
+    {
+      "epoch": 0.8438,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.4499759933430346e-05,
+      "loss": 1.2801,
+      "step": 42190
+    },
+    {
+      "epoch": 0.8439,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.4481661748614784e-05,
+      "loss": 1.2875,
+      "step": 42195
+    },
+    {
+      "epoch": 0.844,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.4463573983949341e-05,
+      "loss": 1.2747,
+      "step": 42200
+    },
+    {
+      "epoch": 0.8441,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.4445496641637967e-05,
+      "loss": 1.2653,
+      "step": 42205
+    },
+    {
+      "epoch": 0.8442,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.4427429723883257e-05,
+      "loss": 1.2632,
+      "step": 42210
+    },
+    {
+      "epoch": 0.8443,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.4409373232886702e-05,
+      "loss": 1.2607,
+      "step": 42215
+    },
+    {
+      "epoch": 0.8444,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.439132717084839e-05,
+      "loss": 1.2756,
+      "step": 42220
+    },
+    {
+      "epoch": 0.8445,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.4373291539967182e-05,
+      "loss": 1.2627,
+      "step": 42225
+    },
+    {
+      "epoch": 0.8446,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.4355266342440677e-05,
+      "loss": 1.2703,
+      "step": 42230
+    },
+    {
+      "epoch": 0.8447,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.4337251580465172e-05,
+      "loss": 1.2822,
+      "step": 42235
+    },
+    {
+      "epoch": 0.8448,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.4319247256235714e-05,
+      "loss": 1.2684,
+      "step": 42240
+    },
+    {
+      "epoch": 0.8449,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.4301253371946089e-05,
+      "loss": 1.2804,
+      "step": 42245
+    },
+    {
+      "epoch": 0.845,
+      "grad_norm": 0.25,
+      "learning_rate": 1.4283269929788779e-05,
+      "loss": 1.267,
+      "step": 42250
+    },
+    {
+      "epoch": 0.8451,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.426529693195503e-05,
+      "loss": 1.2696,
+      "step": 42255
+    },
+    {
+      "epoch": 0.8452,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.4247334380634792e-05,
+      "loss": 1.2807,
+      "step": 42260
+    },
+    {
+      "epoch": 0.8453,
+      "grad_norm": 0.189453125,
+      "learning_rate": 1.4229382278016712e-05,
+      "loss": 1.2917,
+      "step": 42265
+    },
+    {
+      "epoch": 0.8454,
+      "grad_norm": 0.22265625,
+      "learning_rate": 1.4211440626288286e-05,
+      "loss": 1.2836,
+      "step": 42270
+    },
+    {
+      "epoch": 0.8455,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.4193509427635543e-05,
+      "loss": 1.2784,
+      "step": 42275
+    },
+    {
+      "epoch": 0.8456,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4175588684243446e-05,
+      "loss": 1.2682,
+      "step": 42280
+    },
+    {
+      "epoch": 0.8457,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.4157678398295481e-05,
+      "loss": 1.2723,
+      "step": 42285
+    },
+    {
+      "epoch": 0.8458,
+      "grad_norm": 0.193359375,
+      "learning_rate": 1.4139778571974049e-05,
+      "loss": 1.283,
+      "step": 42290
+    },
+    {
+      "epoch": 0.8459,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 1.412188920746017e-05,
+      "loss": 1.2845,
+      "step": 42295
+    },
+    {
+      "epoch": 0.846,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.4104010306933557e-05,
+      "loss": 1.2721,
+      "step": 42300
+    },
+    {
+      "epoch": 0.8461,
+      "grad_norm": 0.140625,
+      "learning_rate": 1.4086141872572789e-05,
+      "loss": 1.2634,
+      "step": 42305
+    },
+    {
+      "epoch": 0.8462,
+      "grad_norm": 0.5,
+      "learning_rate": 1.406828390655497e-05,
+      "loss": 1.279,
+      "step": 42310
+    },
+    {
+      "epoch": 0.8463,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4050436411056123e-05,
+      "loss": 1.2784,
+      "step": 42315
+    },
+    {
+      "epoch": 0.8464,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.40325993882509e-05,
+      "loss": 1.2695,
+      "step": 42320
+    },
+    {
+      "epoch": 0.8465,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.4014772840312663e-05,
+      "loss": 1.2663,
+      "step": 42325
+    },
+    {
+      "epoch": 0.8466,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.3996956769413538e-05,
+      "loss": 1.2701,
+      "step": 42330
+    },
+    {
+      "epoch": 0.8467,
+      "grad_norm": 0.228515625,
+      "learning_rate": 1.3979151177724347e-05,
+      "loss": 1.2743,
+      "step": 42335
+    },
+    {
+      "epoch": 0.8468,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.3961356067414666e-05,
+      "loss": 1.2697,
+      "step": 42340
+    },
+    {
+      "epoch": 0.8469,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.3943571440652769e-05,
+      "loss": 1.261,
+      "step": 42345
+    },
+    {
+      "epoch": 0.847,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.3925797299605647e-05,
+      "loss": 1.2893,
+      "step": 42350
+    },
+    {
+      "epoch": 0.8471,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.3908033646439033e-05,
+      "loss": 1.2634,
+      "step": 42355
+    },
+    {
+      "epoch": 0.8472,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.3890280483317374e-05,
+      "loss": 1.2572,
+      "step": 42360
+    },
+    {
+      "epoch": 0.8473,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.3872537812403829e-05,
+      "loss": 1.2811,
+      "step": 42365
+    },
+    {
+      "epoch": 0.8474,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.3854805635860336e-05,
+      "loss": 1.2661,
+      "step": 42370
+    },
+    {
+      "epoch": 0.8475,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.3837083955847418e-05,
+      "loss": 1.2875,
+      "step": 42375
+    },
+    {
+      "epoch": 0.8476,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.3819372774524508e-05,
+      "loss": 1.2812,
+      "step": 42380
+    },
+    {
+      "epoch": 0.8477,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.38016720940496e-05,
+      "loss": 1.2754,
+      "step": 42385
+    },
+    {
+      "epoch": 0.8478,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.3783981916579446e-05,
+      "loss": 1.2745,
+      "step": 42390
+    },
+    {
+      "epoch": 0.8479,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.3766302244269624e-05,
+      "loss": 1.2714,
+      "step": 42395
+    },
+    {
+      "epoch": 0.848,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.3748633079274253e-05,
+      "loss": 1.278,
+      "step": 42400
+    },
+    {
+      "epoch": 0.8481,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.3730974423746335e-05,
+      "loss": 1.2749,
+      "step": 42405
+    },
+    {
+      "epoch": 0.8482,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.3713326279837501e-05,
+      "loss": 1.2815,
+      "step": 42410
+    },
+    {
+      "epoch": 0.8483,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.3695688649698124e-05,
+      "loss": 1.279,
+      "step": 42415
+    },
+    {
+      "epoch": 0.8484,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.3678061535477304e-05,
+      "loss": 1.2832,
+      "step": 42420
+    },
+    {
+      "epoch": 0.8485,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.3660444939322836e-05,
+      "loss": 1.262,
+      "step": 42425
+    },
+    {
+      "epoch": 0.8486,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.3642838863381257e-05,
+      "loss": 1.2618,
+      "step": 42430
+    },
+    {
+      "epoch": 0.8487,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 1.362524330979782e-05,
+      "loss": 1.2931,
+      "step": 42435
+    },
+    {
+      "epoch": 0.8488,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.3607658280716473e-05,
+      "loss": 1.2689,
+      "step": 42440
+    },
+    {
+      "epoch": 0.8489,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.3590083778279916e-05,
+      "loss": 1.2608,
+      "step": 42445
+    },
+    {
+      "epoch": 0.849,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.3572519804629536e-05,
+      "loss": 1.2836,
+      "step": 42450
+    },
+    {
+      "epoch": 0.8491,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.3554966361905464e-05,
+      "loss": 1.2547,
+      "step": 42455
+    },
+    {
+      "epoch": 0.8492,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.3537423452246523e-05,
+      "loss": 1.2851,
+      "step": 42460
+    },
+    {
+      "epoch": 0.8493,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 1.3519891077790236e-05,
+      "loss": 1.2902,
+      "step": 42465
+    },
+    {
+      "epoch": 0.8494,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.3502369240672941e-05,
+      "loss": 1.2688,
+      "step": 42470
+    },
+    {
+      "epoch": 0.8495,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.3484857943029572e-05,
+      "loss": 1.2859,
+      "step": 42475
+    },
+    {
+      "epoch": 0.8496,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.34673571869938e-05,
+      "loss": 1.289,
+      "step": 42480
+    },
+    {
+      "epoch": 0.8497,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.3449866974698122e-05,
+      "loss": 1.2681,
+      "step": 42485
+    },
+    {
+      "epoch": 0.8498,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 1.3432387308273575e-05,
+      "loss": 1.2698,
+      "step": 42490
+    },
+    {
+      "epoch": 0.8499,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.3414918189850089e-05,
+      "loss": 1.2813,
+      "step": 42495
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.339745962155613e-05,
+      "loss": 1.2676,
+      "step": 42500
+    },
+    {
+      "epoch": 0.8501,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.338001160551906e-05,
+      "loss": 1.2667,
+      "step": 42505
+    },
+    {
+      "epoch": 0.8502,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.3362574143864814e-05,
+      "loss": 1.2654,
+      "step": 42510
+    },
+    {
+      "epoch": 0.8503,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.3345147238718126e-05,
+      "loss": 1.277,
+      "step": 42515
+    },
+    {
+      "epoch": 0.8504,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.3327730892202383e-05,
+      "loss": 1.2754,
+      "step": 42520
+    },
+    {
+      "epoch": 0.8505,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.3310325106439726e-05,
+      "loss": 1.2779,
+      "step": 42525
+    },
+    {
+      "epoch": 0.8506,
+      "grad_norm": 0.19921875,
+      "learning_rate": 1.3292929883550998e-05,
+      "loss": 1.2737,
+      "step": 42530
+    },
+    {
+      "epoch": 0.8507,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.327554522565576e-05,
+      "loss": 1.2744,
+      "step": 42535
+    },
+    {
+      "epoch": 0.8508,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.3258171134872265e-05,
+      "loss": 1.2825,
+      "step": 42540
+    },
+    {
+      "epoch": 0.8509,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.3240807613317507e-05,
+      "loss": 1.278,
+      "step": 42545
+    },
+    {
+      "epoch": 0.851,
+      "grad_norm": 0.19921875,
+      "learning_rate": 1.3223454663107172e-05,
+      "loss": 1.2682,
+      "step": 42550
+    },
+    {
+      "epoch": 0.8511,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.3206112286355632e-05,
+      "loss": 1.2844,
+      "step": 42555
+    },
+    {
+      "epoch": 0.8512,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.3188780485176088e-05,
+      "loss": 1.2727,
+      "step": 42560
+    },
+    {
+      "epoch": 0.8513,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.3171459261680297e-05,
+      "loss": 1.2743,
+      "step": 42565
+    },
+    {
+      "epoch": 0.8514,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.3154148617978812e-05,
+      "loss": 1.2844,
+      "step": 42570
+    },
+    {
+      "epoch": 0.8515,
+      "grad_norm": 0.21484375,
+      "learning_rate": 1.3136848556180892e-05,
+      "loss": 1.2899,
+      "step": 42575
+    },
+    {
+      "epoch": 0.8516,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.3119559078394461e-05,
+      "loss": 1.268,
+      "step": 42580
+    },
+    {
+      "epoch": 0.8517,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 1.3102280186726269e-05,
+      "loss": 1.2789,
+      "step": 42585
+    },
+    {
+      "epoch": 0.8518,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.3085011883281606e-05,
+      "loss": 1.2873,
+      "step": 42590
+    },
+    {
+      "epoch": 0.8519,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3067754170164614e-05,
+      "loss": 1.2737,
+      "step": 42595
+    },
+    {
+      "epoch": 0.852,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.30505070494781e-05,
+      "loss": 1.2758,
+      "step": 42600
+    },
+    {
+      "epoch": 0.8521,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.3033270523323549e-05,
+      "loss": 1.2643,
+      "step": 42605
+    },
+    {
+      "epoch": 0.8522,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3016044593801202e-05,
+      "loss": 1.2963,
+      "step": 42610
+    },
+    {
+      "epoch": 0.8523,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.2998829263009938e-05,
+      "loss": 1.2785,
+      "step": 42615
+    },
+    {
+      "epoch": 0.8524,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.2981624533047432e-05,
+      "loss": 1.2655,
+      "step": 42620
+    },
+    {
+      "epoch": 0.8525,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.296443040601003e-05,
+      "loss": 1.2783,
+      "step": 42625
+    },
+    {
+      "epoch": 0.8526,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.294724688399278e-05,
+      "loss": 1.2857,
+      "step": 42630
+    },
+    {
+      "epoch": 0.8527,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.293007396908944e-05,
+      "loss": 1.2847,
+      "step": 42635
+    },
+    {
+      "epoch": 0.8528,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.2912911663392469e-05,
+      "loss": 1.27,
+      "step": 42640
+    },
+    {
+      "epoch": 0.8529,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2895759968993048e-05,
+      "loss": 1.2709,
+      "step": 42645
+    },
+    {
+      "epoch": 0.853,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.2878618887981064e-05,
+      "loss": 1.2761,
+      "step": 42650
+    },
+    {
+      "epoch": 0.8531,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.28614884224451e-05,
+      "loss": 1.2791,
+      "step": 42655
+    },
+    {
+      "epoch": 0.8532,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.2844368574472454e-05,
+      "loss": 1.2729,
+      "step": 42660
+    },
+    {
+      "epoch": 0.8533,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.2827259346149122e-05,
+      "loss": 1.289,
+      "step": 42665
+    },
+    {
+      "epoch": 0.8534,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.2810160739559796e-05,
+      "loss": 1.2656,
+      "step": 42670
+    },
+    {
+      "epoch": 0.8535,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.279307275678795e-05,
+      "loss": 1.2614,
+      "step": 42675
+    },
+    {
+      "epoch": 0.8536,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.2775995399915631e-05,
+      "loss": 1.2643,
+      "step": 42680
+    },
+    {
+      "epoch": 0.8537,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.2758928671023718e-05,
+      "loss": 1.281,
+      "step": 42685
+    },
+    {
+      "epoch": 0.8538,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 1.2741872572191682e-05,
+      "loss": 1.2719,
+      "step": 42690
+    },
+    {
+      "epoch": 0.8539,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.2724827105497816e-05,
+      "loss": 1.2806,
+      "step": 42695
+    },
+    {
+      "epoch": 0.854,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.2707792273019048e-05,
+      "loss": 1.2732,
+      "step": 42700
+    },
+    {
+      "epoch": 0.8541,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.2690768076830972e-05,
+      "loss": 1.2992,
+      "step": 42705
+    },
+    {
+      "epoch": 0.8542,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2673754519008008e-05,
+      "loss": 1.2747,
+      "step": 42710
+    },
+    {
+      "epoch": 0.8543,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.2656751601623118e-05,
+      "loss": 1.2882,
+      "step": 42715
+    },
+    {
+      "epoch": 0.8544,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.2639759326748135e-05,
+      "loss": 1.2655,
+      "step": 42720
+    },
+    {
+      "epoch": 0.8545,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.262277769645348e-05,
+      "loss": 1.2557,
+      "step": 42725
+    },
+    {
+      "epoch": 0.8546,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.260580671280832e-05,
+      "loss": 1.2672,
+      "step": 42730
+    },
+    {
+      "epoch": 0.8547,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2588846377880525e-05,
+      "loss": 1.2624,
+      "step": 42735
+    },
+    {
+      "epoch": 0.8548,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.257189669373664e-05,
+      "loss": 1.2685,
+      "step": 42740
+    },
+    {
+      "epoch": 0.8549,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 1.2554957662441957e-05,
+      "loss": 1.2671,
+      "step": 42745
+    },
+    {
+      "epoch": 0.855,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.2538029286060426e-05,
+      "loss": 1.2766,
+      "step": 42750
+    },
+    {
+      "epoch": 0.8551,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.2521111566654731e-05,
+      "loss": 1.2796,
+      "step": 42755
+    },
+    {
+      "epoch": 0.8552,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.2504204506286243e-05,
+      "loss": 1.2783,
+      "step": 42760
+    },
+    {
+      "epoch": 0.8553,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.2487308107015027e-05,
+      "loss": 1.2792,
+      "step": 42765
+    },
+    {
+      "epoch": 0.8554,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.2470422370899838e-05,
+      "loss": 1.2729,
+      "step": 42770
+    },
+    {
+      "epoch": 0.8555,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 1.2453547299998225e-05,
+      "loss": 1.269,
+      "step": 42775
+    },
+    {
+      "epoch": 0.8556,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.243668289636628e-05,
+      "loss": 1.2653,
+      "step": 42780
+    },
+    {
+      "epoch": 0.8557,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.241982916205895e-05,
+      "loss": 1.29,
+      "step": 42785
+    },
+    {
+      "epoch": 0.8558,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.2402986099129765e-05,
+      "loss": 1.265,
+      "step": 42790
+    },
+    {
+      "epoch": 0.8559,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.2386153709630988e-05,
+      "loss": 1.287,
+      "step": 42795
+    },
+    {
+      "epoch": 0.856,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2369331995613665e-05,
+      "loss": 1.2815,
+      "step": 42800
+    },
+    {
+      "epoch": 0.8561,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.2352520959127379e-05,
+      "loss": 1.2737,
+      "step": 42805
+    },
+    {
+      "epoch": 0.8562,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2335720602220569e-05,
+      "loss": 1.2809,
+      "step": 42810
+    },
+    {
+      "epoch": 0.8563,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.2318930926940298e-05,
+      "loss": 1.265,
+      "step": 42815
+    },
+    {
+      "epoch": 0.8564,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.2302151935332329e-05,
+      "loss": 1.2775,
+      "step": 42820
+    },
+    {
+      "epoch": 0.8565,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 1.228538362944115e-05,
+      "loss": 1.2758,
+      "step": 42825
+    },
+    {
+      "epoch": 0.8566,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.2268626011309858e-05,
+      "loss": 1.2684,
+      "step": 42830
+    },
+    {
+      "epoch": 0.8567,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.22518790829804e-05,
+      "loss": 1.262,
+      "step": 42835
+    },
+    {
+      "epoch": 0.8568,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.2235142846493308e-05,
+      "loss": 1.2723,
+      "step": 42840
+    },
+    {
+      "epoch": 0.8569,
+      "grad_norm": 0.2265625,
+      "learning_rate": 1.2218417303887842e-05,
+      "loss": 1.292,
+      "step": 42845
+    },
+    {
+      "epoch": 0.857,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.2201702457201947e-05,
+      "loss": 1.274,
+      "step": 42850
+    },
+    {
+      "epoch": 0.8571,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.2184998308472295e-05,
+      "loss": 1.2818,
+      "step": 42855
+    },
+    {
+      "epoch": 0.8572,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.2168304859734226e-05,
+      "loss": 1.2647,
+      "step": 42860
+    },
+    {
+      "epoch": 0.8573,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.2151622113021787e-05,
+      "loss": 1.2778,
+      "step": 42865
+    },
+    {
+      "epoch": 0.8574,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.2134950070367723e-05,
+      "loss": 1.2734,
+      "step": 42870
+    },
+    {
+      "epoch": 0.8575,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.2118288733803473e-05,
+      "loss": 1.2659,
+      "step": 42875
+    },
+    {
+      "epoch": 0.8576,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.2101638105359169e-05,
+      "loss": 1.2821,
+      "step": 42880
+    },
+    {
+      "epoch": 0.8577,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.2084998187063613e-05,
+      "loss": 1.2711,
+      "step": 42885
+    },
+    {
+      "epoch": 0.8578,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.206836898094439e-05,
+      "loss": 1.2992,
+      "step": 42890
+    },
+    {
+      "epoch": 0.8579,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 1.2051750489027647e-05,
+      "loss": 1.2738,
+      "step": 42895
+    },
+    {
+      "epoch": 0.858,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2035142713338366e-05,
+      "loss": 1.2739,
+      "step": 42900
+    },
+    {
+      "epoch": 0.8581,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.2018545655900081e-05,
+      "loss": 1.2781,
+      "step": 42905
+    },
+    {
+      "epoch": 0.8582,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.2001959318735156e-05,
+      "loss": 1.2724,
+      "step": 42910
+    },
+    {
+      "epoch": 0.8583,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.1985383703864583e-05,
+      "loss": 1.2788,
+      "step": 42915
+    },
+    {
+      "epoch": 0.8584,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.196881881330798e-05,
+      "loss": 1.2815,
+      "step": 42920
+    },
+    {
+      "epoch": 0.8585,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.19522646490838e-05,
+      "loss": 1.2615,
+      "step": 42925
+    },
+    {
+      "epoch": 0.8586,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.1935721213209105e-05,
+      "loss": 1.2659,
+      "step": 42930
+    },
+    {
+      "epoch": 0.8587,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.191918850769964e-05,
+      "loss": 1.2612,
+      "step": 42935
+    },
+    {
+      "epoch": 0.8588,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.1902666534569883e-05,
+      "loss": 1.2898,
+      "step": 42940
+    },
+    {
+      "epoch": 0.8589,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.188615529583299e-05,
+      "loss": 1.2651,
+      "step": 42945
+    },
+    {
+      "epoch": 0.859,
+      "grad_norm": 0.244140625,
+      "learning_rate": 1.1869654793500784e-05,
+      "loss": 1.2816,
+      "step": 42950
+    },
+    {
+      "epoch": 0.8591,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.1853165029583823e-05,
+      "loss": 1.2727,
+      "step": 42955
+    },
+    {
+      "epoch": 0.8592,
+      "grad_norm": 0.228515625,
+      "learning_rate": 1.1836686006091313e-05,
+      "loss": 1.262,
+      "step": 42960
+    },
+    {
+      "epoch": 0.8593,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.1820217725031191e-05,
+      "loss": 1.2717,
+      "step": 42965
+    },
+    {
+      "epoch": 0.8594,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.1803760188410073e-05,
+      "loss": 1.2782,
+      "step": 42970
+    },
+    {
+      "epoch": 0.8595,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.1787313398233235e-05,
+      "loss": 1.2656,
+      "step": 42975
+    },
+    {
+      "epoch": 0.8596,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.1770877356504683e-05,
+      "loss": 1.272,
+      "step": 42980
+    },
+    {
+      "epoch": 0.8597,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.1754452065227084e-05,
+      "loss": 1.2694,
+      "step": 42985
+    },
+    {
+      "epoch": 0.8598,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.1738037526401857e-05,
+      "loss": 1.2776,
+      "step": 42990
+    },
+    {
+      "epoch": 0.8599,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.1721633742028992e-05,
+      "loss": 1.2656,
+      "step": 42995
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.1705240714107302e-05,
+      "loss": 1.28,
+      "step": 43000
+    },
+    {
+      "epoch": 0.8601,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.168885844463422e-05,
+      "loss": 1.2683,
+      "step": 43005
+    },
+    {
+      "epoch": 0.8602,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.1672486935605831e-05,
+      "loss": 1.2704,
+      "step": 43010
+    },
+    {
+      "epoch": 0.8603,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.1656126189017014e-05,
+      "loss": 1.275,
+      "step": 43015
+    },
+    {
+      "epoch": 0.8604,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.1639776206861196e-05,
+      "loss": 1.2633,
+      "step": 43020
+    },
+    {
+      "epoch": 0.8605,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.1623436991130654e-05,
+      "loss": 1.2816,
+      "step": 43025
+    },
+    {
+      "epoch": 0.8606,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.1607108543816248e-05,
+      "loss": 1.2728,
+      "step": 43030
+    },
+    {
+      "epoch": 0.8607,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.159079086690753e-05,
+      "loss": 1.2767,
+      "step": 43035
+    },
+    {
+      "epoch": 0.8608,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.1574483962392767e-05,
+      "loss": 1.2865,
+      "step": 43040
+    },
+    {
+      "epoch": 0.8609,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.1558187832258926e-05,
+      "loss": 1.2665,
+      "step": 43045
+    },
+    {
+      "epoch": 0.861,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 1.1541902478491606e-05,
+      "loss": 1.2698,
+      "step": 43050
+    },
+    {
+      "epoch": 0.8611,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.1525627903075165e-05,
+      "loss": 1.2844,
+      "step": 43055
+    },
+    {
+      "epoch": 0.8612,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.1509364107992583e-05,
+      "loss": 1.2852,
+      "step": 43060
+    },
+    {
+      "epoch": 0.8613,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.1493111095225562e-05,
+      "loss": 1.2882,
+      "step": 43065
+    },
+    {
+      "epoch": 0.8614,
+      "grad_norm": 0.193359375,
+      "learning_rate": 1.1476868866754486e-05,
+      "loss": 1.2671,
+      "step": 43070
+    },
+    {
+      "epoch": 0.8615,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.1460637424558407e-05,
+      "loss": 1.2605,
+      "step": 43075
+    },
+    {
+      "epoch": 0.8616,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.1444416770615118e-05,
+      "loss": 1.2557,
+      "step": 43080
+    },
+    {
+      "epoch": 0.8617,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.1428206906900995e-05,
+      "loss": 1.269,
+      "step": 43085
+    },
+    {
+      "epoch": 0.8618,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.1412007835391236e-05,
+      "loss": 1.2779,
+      "step": 43090
+    },
+    {
+      "epoch": 0.8619,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.1395819558059572e-05,
+      "loss": 1.2868,
+      "step": 43095
+    },
+    {
+      "epoch": 0.862,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.1379642076878527e-05,
+      "loss": 1.2734,
+      "step": 43100
+    },
+    {
+      "epoch": 0.8621,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.1363475393819311e-05,
+      "loss": 1.2721,
+      "step": 43105
+    },
+    {
+      "epoch": 0.8622,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.1347319510851717e-05,
+      "loss": 1.2934,
+      "step": 43110
+    },
+    {
+      "epoch": 0.8623,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.1331174429944347e-05,
+      "loss": 1.2693,
+      "step": 43115
+    },
+    {
+      "epoch": 0.8624,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 1.1315040153064416e-05,
+      "loss": 1.277,
+      "step": 43120
+    },
+    {
+      "epoch": 0.8625,
+      "grad_norm": 0.185546875,
+      "learning_rate": 1.129891668217783e-05,
+      "loss": 1.2604,
+      "step": 43125
+    },
+    {
+      "epoch": 0.8626,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1282804019249182e-05,
+      "loss": 1.2812,
+      "step": 43130
+    },
+    {
+      "epoch": 0.8627,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.1266702166241772e-05,
+      "loss": 1.2854,
+      "step": 43135
+    },
+    {
+      "epoch": 0.8628,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.1250611125117527e-05,
+      "loss": 1.2705,
+      "step": 43140
+    },
+    {
+      "epoch": 0.8629,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.1234530897837126e-05,
+      "loss": 1.2715,
+      "step": 43145
+    },
+    {
+      "epoch": 0.863,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.1218461486359877e-05,
+      "loss": 1.2689,
+      "step": 43150
+    },
+    {
+      "epoch": 0.8631,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.1202402892643781e-05,
+      "loss": 1.2791,
+      "step": 43155
+    },
+    {
+      "epoch": 0.8632,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.1186355118645554e-05,
+      "loss": 1.2774,
+      "step": 43160
+    },
+    {
+      "epoch": 0.8633,
+      "grad_norm": 0.185546875,
+      "learning_rate": 1.1170318166320548e-05,
+      "loss": 1.2734,
+      "step": 43165
+    },
+    {
+      "epoch": 0.8634,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.1154292037622838e-05,
+      "loss": 1.2822,
+      "step": 43170
+    },
+    {
+      "epoch": 0.8635,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.1138276734505104e-05,
+      "loss": 1.2847,
+      "step": 43175
+    },
+    {
+      "epoch": 0.8636,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.1122272258918865e-05,
+      "loss": 1.2881,
+      "step": 43180
+    },
+    {
+      "epoch": 0.8637,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.1106278612814125e-05,
+      "loss": 1.2658,
+      "step": 43185
+    },
+    {
+      "epoch": 0.8638,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 1.1090295798139672e-05,
+      "loss": 1.3042,
+      "step": 43190
+    },
+    {
+      "epoch": 0.8639,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.1074323816843024e-05,
+      "loss": 1.2643,
+      "step": 43195
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.1058362670870249e-05,
+      "loss": 1.272,
+      "step": 43200
+    },
+    {
+      "epoch": 0.8641,
+      "grad_norm": 0.140625,
+      "learning_rate": 1.1042412362166222e-05,
+      "loss": 1.2719,
+      "step": 43205
+    },
+    {
+      "epoch": 0.8642,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.1026472892674378e-05,
+      "loss": 1.2691,
+      "step": 43210
+    },
+    {
+      "epoch": 0.8643,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.1010544264336942e-05,
+      "loss": 1.2743,
+      "step": 43215
+    },
+    {
+      "epoch": 0.8644,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.099462647909475e-05,
+      "loss": 1.2756,
+      "step": 43220
+    },
+    {
+      "epoch": 0.8645,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.0978719538887349e-05,
+      "loss": 1.2641,
+      "step": 43225
+    },
+    {
+      "epoch": 0.8646,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.0962823445652959e-05,
+      "loss": 1.2885,
+      "step": 43230
+    },
+    {
+      "epoch": 0.8647,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.0946938201328416e-05,
+      "loss": 1.2863,
+      "step": 43235
+    },
+    {
+      "epoch": 0.8648,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 1.093106380784934e-05,
+      "loss": 1.2711,
+      "step": 43240
+    },
+    {
+      "epoch": 0.8649,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.0915200267149972e-05,
+      "loss": 1.2708,
+      "step": 43245
+    },
+    {
+      "epoch": 0.865,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.0899347581163221e-05,
+      "loss": 1.301,
+      "step": 43250
+    },
+    {
+      "epoch": 0.8651,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.08835057518207e-05,
+      "loss": 1.2851,
+      "step": 43255
+    },
+    {
+      "epoch": 0.8652,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.0867674781052684e-05,
+      "loss": 1.2848,
+      "step": 43260
+    },
+    {
+      "epoch": 0.8653,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.0851854670788108e-05,
+      "loss": 1.2738,
+      "step": 43265
+    },
+    {
+      "epoch": 0.8654,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.0836045422954666e-05,
+      "loss": 1.2666,
+      "step": 43270
+    },
+    {
+      "epoch": 0.8655,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 1.0820247039478604e-05,
+      "loss": 1.27,
+      "step": 43275
+    },
+    {
+      "epoch": 0.8656,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.0804459522284926e-05,
+      "loss": 1.2779,
+      "step": 43280
+    },
+    {
+      "epoch": 0.8657,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.0788682873297307e-05,
+      "loss": 1.2686,
+      "step": 43285
+    },
+    {
+      "epoch": 0.8658,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.0772917094438051e-05,
+      "loss": 1.274,
+      "step": 43290
+    },
+    {
+      "epoch": 0.8659,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.0757162187628222e-05,
+      "loss": 1.288,
+      "step": 43295
+    },
+    {
+      "epoch": 0.866,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.0741418154787442e-05,
+      "loss": 1.2643,
+      "step": 43300
+    },
+    {
+      "epoch": 0.8661,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.0725684997834162e-05,
+      "loss": 1.2817,
+      "step": 43305
+    },
+    {
+      "epoch": 0.8662,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.0709962718685318e-05,
+      "loss": 1.2537,
+      "step": 43310
+    },
+    {
+      "epoch": 0.8663,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.0694251319256687e-05,
+      "loss": 1.2717,
+      "step": 43315
+    },
+    {
+      "epoch": 0.8664,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.067855080146266e-05,
+      "loss": 1.2583,
+      "step": 43320
+    },
+    {
+      "epoch": 0.8665,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.0662861167216243e-05,
+      "loss": 1.2736,
+      "step": 43325
+    },
+    {
+      "epoch": 0.8666,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.0647182418429225e-05,
+      "loss": 1.2771,
+      "step": 43330
+    },
+    {
+      "epoch": 0.8667,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.063151455701199e-05,
+      "loss": 1.2749,
+      "step": 43335
+    },
+    {
+      "epoch": 0.8668,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.0615857584873623e-05,
+      "loss": 1.2778,
+      "step": 43340
+    },
+    {
+      "epoch": 0.8669,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.0600211503921887e-05,
+      "loss": 1.2697,
+      "step": 43345
+    },
+    {
+      "epoch": 0.867,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.0584576316063188e-05,
+      "loss": 1.3016,
+      "step": 43350
+    },
+    {
+      "epoch": 0.8671,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.0568952023202638e-05,
+      "loss": 1.2673,
+      "step": 43355
+    },
+    {
+      "epoch": 0.8672,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.0553338627244025e-05,
+      "loss": 1.2786,
+      "step": 43360
+    },
+    {
+      "epoch": 0.8673,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.053773613008977e-05,
+      "loss": 1.263,
+      "step": 43365
+    },
+    {
+      "epoch": 0.8674,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.0522144533640998e-05,
+      "loss": 1.2849,
+      "step": 43370
+    },
+    {
+      "epoch": 0.8675,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.0506563839797501e-05,
+      "loss": 1.2568,
+      "step": 43375
+    },
+    {
+      "epoch": 0.8676,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 1.0490994050457748e-05,
+      "loss": 1.2809,
+      "step": 43380
+    },
+    {
+      "epoch": 0.8677,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.0475435167518843e-05,
+      "loss": 1.2761,
+      "step": 43385
+    },
+    {
+      "epoch": 0.8678,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 1.0459887192876594e-05,
+      "loss": 1.2788,
+      "step": 43390
+    },
+    {
+      "epoch": 0.8679,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 1.0444350128425529e-05,
+      "loss": 1.2878,
+      "step": 43395
+    },
+    {
+      "epoch": 0.868,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.042882397605871e-05,
+      "loss": 1.2683,
+      "step": 43400
+    },
+    {
+      "epoch": 0.8681,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.0413308737668005e-05,
+      "loss": 1.2816,
+      "step": 43405
+    },
+    {
+      "epoch": 0.8682,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.0397804415143909e-05,
+      "loss": 1.2753,
+      "step": 43410
+    },
+    {
+      "epoch": 0.8683,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.0382311010375512e-05,
+      "loss": 1.2684,
+      "step": 43415
+    },
+    {
+      "epoch": 0.8684,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.0366828525250726e-05,
+      "loss": 1.2653,
+      "step": 43420
+    },
+    {
+      "epoch": 0.8685,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.0351356961655945e-05,
+      "loss": 1.272,
+      "step": 43425
+    },
+    {
+      "epoch": 0.8686,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.0335896321476413e-05,
+      "loss": 1.2754,
+      "step": 43430
+    },
+    {
+      "epoch": 0.8687,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.0320446606595934e-05,
+      "loss": 1.2742,
+      "step": 43435
+    },
+    {
+      "epoch": 0.8688,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.0305007818897006e-05,
+      "loss": 1.2834,
+      "step": 43440
+    },
+    {
+      "epoch": 0.8689,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.028957996026081e-05,
+      "loss": 1.261,
+      "step": 43445
+    },
+    {
+      "epoch": 0.869,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.0274163032567163e-05,
+      "loss": 1.2766,
+      "step": 43450
+    },
+    {
+      "epoch": 0.8691,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.0258757037694589e-05,
+      "loss": 1.2738,
+      "step": 43455
+    },
+    {
+      "epoch": 0.8692,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 1.0243361977520249e-05,
+      "loss": 1.2669,
+      "step": 43460
+    },
+    {
+      "epoch": 0.8693,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 1.022797785392e-05,
+      "loss": 1.2813,
+      "step": 43465
+    },
+    {
+      "epoch": 0.8694,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.0212604668768343e-05,
+      "loss": 1.2794,
+      "step": 43470
+    },
+    {
+      "epoch": 0.8695,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.0197242423938446e-05,
+      "loss": 1.2691,
+      "step": 43475
+    },
+    {
+      "epoch": 0.8696,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.0181891121302145e-05,
+      "loss": 1.2551,
+      "step": 43480
+    },
+    {
+      "epoch": 0.8697,
+      "grad_norm": 0.158203125,
+      "learning_rate": 1.0166550762729998e-05,
+      "loss": 1.2524,
+      "step": 43485
+    },
+    {
+      "epoch": 0.8698,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.0151221350091134e-05,
+      "loss": 1.2819,
+      "step": 43490
+    },
+    {
+      "epoch": 0.8699,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.0135902885253401e-05,
+      "loss": 1.2785,
+      "step": 43495
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.0120595370083318e-05,
+      "loss": 1.2656,
+      "step": 43500
+    },
+    {
+      "epoch": 0.8701,
+      "grad_norm": 0.1484375,
+      "learning_rate": 1.0105298806446028e-05,
+      "loss": 1.2762,
+      "step": 43505
+    },
+    {
+      "epoch": 0.8702,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.009001319620545e-05,
+      "loss": 1.272,
+      "step": 43510
+    },
+    {
+      "epoch": 0.8703,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.0074738541223993e-05,
+      "loss": 1.2625,
+      "step": 43515
+    },
+    {
+      "epoch": 0.8704,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.0059474843362892e-05,
+      "loss": 1.2491,
+      "step": 43520
+    },
+    {
+      "epoch": 0.8705,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.0044222104481971e-05,
+      "loss": 1.2701,
+      "step": 43525
+    },
+    {
+      "epoch": 0.8706,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 1.0028980326439707e-05,
+      "loss": 1.2891,
+      "step": 43530
+    },
+    {
+      "epoch": 0.8707,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.0013749511093307e-05,
+      "loss": 1.2839,
+      "step": 43535
+    },
+    {
+      "epoch": 0.8708,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 9.998529660298539e-06,
+      "loss": 1.2659,
+      "step": 43540
+    },
+    {
+      "epoch": 0.8709,
+      "grad_norm": 0.466796875,
+      "learning_rate": 9.983320775909933e-06,
+      "loss": 1.2729,
+      "step": 43545
+    },
+    {
+      "epoch": 0.871,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.968122859780648e-06,
+      "loss": 1.2621,
+      "step": 43550
+    },
+    {
+      "epoch": 0.8711,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 9.952935913762506e-06,
+      "loss": 1.2597,
+      "step": 43555
+    },
+    {
+      "epoch": 0.8712,
+      "grad_norm": 0.1875,
+      "learning_rate": 9.937759939705971e-06,
+      "loss": 1.2783,
+      "step": 43560
+    },
+    {
+      "epoch": 0.8713,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.922594939460194e-06,
+      "loss": 1.2721,
+      "step": 43565
+    },
+    {
+      "epoch": 0.8714,
+      "grad_norm": 0.376953125,
+      "learning_rate": 9.907440914873e-06,
+      "loss": 1.2746,
+      "step": 43570
+    },
+    {
+      "epoch": 0.8715,
+      "grad_norm": 0.451171875,
+      "learning_rate": 9.892297867790845e-06,
+      "loss": 1.2825,
+      "step": 43575
+    },
+    {
+      "epoch": 0.8716,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.877165800058874e-06,
+      "loss": 1.2772,
+      "step": 43580
+    },
+    {
+      "epoch": 0.8717,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.86204471352088e-06,
+      "loss": 1.2722,
+      "step": 43585
+    },
+    {
+      "epoch": 0.8718,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.84693461001932e-06,
+      "loss": 1.2743,
+      "step": 43590
+    },
+    {
+      "epoch": 0.8719,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.831835491395292e-06,
+      "loss": 1.2704,
+      "step": 43595
+    },
+    {
+      "epoch": 0.872,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.816747359488632e-06,
+      "loss": 1.2702,
+      "step": 43600
+    },
+    {
+      "epoch": 0.8721,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.801670216137727e-06,
+      "loss": 1.2761,
+      "step": 43605
+    },
+    {
+      "epoch": 0.8722,
+      "grad_norm": 0.236328125,
+      "learning_rate": 9.786604063179728e-06,
+      "loss": 1.2606,
+      "step": 43610
+    },
+    {
+      "epoch": 0.8723,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.771548902450357e-06,
+      "loss": 1.2797,
+      "step": 43615
+    },
+    {
+      "epoch": 0.8724,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.756504735784067e-06,
+      "loss": 1.2695,
+      "step": 43620
+    },
+    {
+      "epoch": 0.8725,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.74147156501396e-06,
+      "loss": 1.2734,
+      "step": 43625
+    },
+    {
+      "epoch": 0.8726,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 9.726449391971714e-06,
+      "loss": 1.2964,
+      "step": 43630
+    },
+    {
+      "epoch": 0.8727,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.711438218487834e-06,
+      "loss": 1.2872,
+      "step": 43635
+    },
+    {
+      "epoch": 0.8728,
+      "grad_norm": 0.4921875,
+      "learning_rate": 9.696438046391288e-06,
+      "loss": 1.2718,
+      "step": 43640
+    },
+    {
+      "epoch": 0.8729,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.681448877509858e-06,
+      "loss": 1.2649,
+      "step": 43645
+    },
+    {
+      "epoch": 0.873,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.666470713669918e-06,
+      "loss": 1.2897,
+      "step": 43650
+    },
+    {
+      "epoch": 0.8731,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.651503556696516e-06,
+      "loss": 1.2739,
+      "step": 43655
+    },
+    {
+      "epoch": 0.8732,
+      "grad_norm": 0.2265625,
+      "learning_rate": 9.636547408413355e-06,
+      "loss": 1.2745,
+      "step": 43660
+    },
+    {
+      "epoch": 0.8733,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.621602270642781e-06,
+      "loss": 1.2543,
+      "step": 43665
+    },
+    {
+      "epoch": 0.8734,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.606668145205833e-06,
+      "loss": 1.2889,
+      "step": 43670
+    },
+    {
+      "epoch": 0.8735,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.591745033922173e-06,
+      "loss": 1.2713,
+      "step": 43675
+    },
+    {
+      "epoch": 0.8736,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.576832938610137e-06,
+      "loss": 1.282,
+      "step": 43680
+    },
+    {
+      "epoch": 0.8737,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.561931861086737e-06,
+      "loss": 1.2777,
+      "step": 43685
+    },
+    {
+      "epoch": 0.8738,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.5470418031676e-06,
+      "loss": 1.2797,
+      "step": 43690
+    },
+    {
+      "epoch": 0.8739,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.532162766667042e-06,
+      "loss": 1.2656,
+      "step": 43695
+    },
+    {
+      "epoch": 0.874,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.517294753398064e-06,
+      "loss": 1.2835,
+      "step": 43700
+    },
+    {
+      "epoch": 0.8741,
+      "grad_norm": 0.1953125,
+      "learning_rate": 9.502437765172212e-06,
+      "loss": 1.262,
+      "step": 43705
+    },
+    {
+      "epoch": 0.8742,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.487591803799856e-06,
+      "loss": 1.2688,
+      "step": 43710
+    },
+    {
+      "epoch": 0.8743,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.47275687108986e-06,
+      "loss": 1.2759,
+      "step": 43715
+    },
+    {
+      "epoch": 0.8744,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.457932968849825e-06,
+      "loss": 1.2789,
+      "step": 43720
+    },
+    {
+      "epoch": 0.8745,
+      "grad_norm": 0.1640625,
+      "learning_rate": 9.443120098886061e-06,
+      "loss": 1.2944,
+      "step": 43725
+    },
+    {
+      "epoch": 0.8746,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 9.428318263003378e-06,
+      "loss": 1.2777,
+      "step": 43730
+    },
+    {
+      "epoch": 0.8747,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 9.4135274630054e-06,
+      "loss": 1.2795,
+      "step": 43735
+    },
+    {
+      "epoch": 0.8748,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.398747700694322e-06,
+      "loss": 1.2926,
+      "step": 43740
+    },
+    {
+      "epoch": 0.8749,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 9.383978977871021e-06,
+      "loss": 1.2686,
+      "step": 43745
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 0.240234375,
+      "learning_rate": 9.369221296335006e-06,
+      "loss": 1.2834,
+      "step": 43750
+    },
+    {
+      "epoch": 0.8751,
+      "grad_norm": 0.185546875,
+      "learning_rate": 9.354474657884472e-06,
+      "loss": 1.2618,
+      "step": 43755
+    },
+    {
+      "epoch": 0.8752,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 9.339739064316233e-06,
+      "loss": 1.2649,
+      "step": 43760
+    },
+    {
+      "epoch": 0.8753,
+      "grad_norm": 0.48828125,
+      "learning_rate": 9.32501451742579e-06,
+      "loss": 1.2606,
+      "step": 43765
+    },
+    {
+      "epoch": 0.8754,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.310301019007285e-06,
+      "loss": 1.2792,
+      "step": 43770
+    },
+    {
+      "epoch": 0.8755,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.295598570853514e-06,
+      "loss": 1.2683,
+      "step": 43775
+    },
+    {
+      "epoch": 0.8756,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 9.280907174755915e-06,
+      "loss": 1.3017,
+      "step": 43780
+    },
+    {
+      "epoch": 0.8757,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.266226832504598e-06,
+      "loss": 1.2703,
+      "step": 43785
+    },
+    {
+      "epoch": 0.8758,
+      "grad_norm": 0.16015625,
+      "learning_rate": 9.251557545888312e-06,
+      "loss": 1.2769,
+      "step": 43790
+    },
+    {
+      "epoch": 0.8759,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.236899316694459e-06,
+      "loss": 1.2726,
+      "step": 43795
+    },
+    {
+      "epoch": 0.876,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.222252146709142e-06,
+      "loss": 1.2924,
+      "step": 43800
+    },
+    {
+      "epoch": 0.8761,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.207616037717025e-06,
+      "loss": 1.2975,
+      "step": 43805
+    },
+    {
+      "epoch": 0.8762,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.192990991501482e-06,
+      "loss": 1.286,
+      "step": 43810
+    },
+    {
+      "epoch": 0.8763,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 9.178377009844563e-06,
+      "loss": 1.2941,
+      "step": 43815
+    },
+    {
+      "epoch": 0.8764,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.163774094526889e-06,
+      "loss": 1.2732,
+      "step": 43820
+    },
+    {
+      "epoch": 0.8765,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 9.149182247327837e-06,
+      "loss": 1.2702,
+      "step": 43825
+    },
+    {
+      "epoch": 0.8766,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.134601470025306e-06,
+      "loss": 1.2565,
+      "step": 43830
+    },
+    {
+      "epoch": 0.8767,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.120031764395987e-06,
+      "loss": 1.2803,
+      "step": 43835
+    },
+    {
+      "epoch": 0.8768,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.105473132215125e-06,
+      "loss": 1.283,
+      "step": 43840
+    },
+    {
+      "epoch": 0.8769,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 9.09092557525666e-06,
+      "loss": 1.2771,
+      "step": 43845
+    },
+    {
+      "epoch": 0.877,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.076389095293148e-06,
+      "loss": 1.2872,
+      "step": 43850
+    },
+    {
+      "epoch": 0.8771,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.061863694095828e-06,
+      "loss": 1.2826,
+      "step": 43855
+    },
+    {
+      "epoch": 0.8772,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.047349373434566e-06,
+      "loss": 1.2719,
+      "step": 43860
+    },
+    {
+      "epoch": 0.8773,
+      "grad_norm": 0.228515625,
+      "learning_rate": 9.0328461350779e-06,
+      "loss": 1.2869,
+      "step": 43865
+    },
+    {
+      "epoch": 0.8774,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.018353980792993e-06,
+      "loss": 1.2772,
+      "step": 43870
+    },
+    {
+      "epoch": 0.8775,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.00387291234569e-06,
+      "loss": 1.2639,
+      "step": 43875
+    },
+    {
+      "epoch": 0.8776,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.989402931500434e-06,
+      "loss": 1.2777,
+      "step": 43880
+    },
+    {
+      "epoch": 0.8777,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 8.974944040020362e-06,
+      "loss": 1.2758,
+      "step": 43885
+    },
+    {
+      "epoch": 0.8778,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.960496239667282e-06,
+      "loss": 1.2683,
+      "step": 43890
+    },
+    {
+      "epoch": 0.8779,
+      "grad_norm": 0.158203125,
+      "learning_rate": 8.946059532201567e-06,
+      "loss": 1.2739,
+      "step": 43895
+    },
+    {
+      "epoch": 0.878,
+      "grad_norm": 0.392578125,
+      "learning_rate": 8.931633919382298e-06,
+      "loss": 1.2693,
+      "step": 43900
+    },
+    {
+      "epoch": 0.8781,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 8.917219402967202e-06,
+      "loss": 1.2606,
+      "step": 43905
+    },
+    {
+      "epoch": 0.8782,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.90281598471262e-06,
+      "loss": 1.2905,
+      "step": 43910
+    },
+    {
+      "epoch": 0.8783,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.888423666373614e-06,
+      "loss": 1.2584,
+      "step": 43915
+    },
+    {
+      "epoch": 0.8784,
+      "grad_norm": 0.23046875,
+      "learning_rate": 8.87404244970378e-06,
+      "loss": 1.2853,
+      "step": 43920
+    },
+    {
+      "epoch": 0.8785,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.85967233645547e-06,
+      "loss": 1.2652,
+      "step": 43925
+    },
+    {
+      "epoch": 0.8786,
+      "grad_norm": 0.23828125,
+      "learning_rate": 8.845313328379634e-06,
+      "loss": 1.277,
+      "step": 43930
+    },
+    {
+      "epoch": 0.8787,
+      "grad_norm": 0.25,
+      "learning_rate": 8.830965427225868e-06,
+      "loss": 1.2709,
+      "step": 43935
+    },
+    {
+      "epoch": 0.8788,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 8.816628634742441e-06,
+      "loss": 1.3019,
+      "step": 43940
+    },
+    {
+      "epoch": 0.8789,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.80230295267619e-06,
+      "loss": 1.284,
+      "step": 43945
+    },
+    {
+      "epoch": 0.879,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.787988382772705e-06,
+      "loss": 1.2642,
+      "step": 43950
+    },
+    {
+      "epoch": 0.8791,
+      "grad_norm": 0.353515625,
+      "learning_rate": 8.77368492677616e-06,
+      "loss": 1.2758,
+      "step": 43955
+    },
+    {
+      "epoch": 0.8792,
+      "grad_norm": 0.3828125,
+      "learning_rate": 8.759392586429393e-06,
+      "loss": 1.2816,
+      "step": 43960
+    },
+    {
+      "epoch": 0.8793,
+      "grad_norm": 0.232421875,
+      "learning_rate": 8.745111363473868e-06,
+      "loss": 1.2768,
+      "step": 43965
+    },
+    {
+      "epoch": 0.8794,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 8.730841259649725e-06,
+      "loss": 1.2736,
+      "step": 43970
+    },
+    {
+      "epoch": 0.8795,
+      "grad_norm": 0.212890625,
+      "learning_rate": 8.716582276695728e-06,
+      "loss": 1.2697,
+      "step": 43975
+    },
+    {
+      "epoch": 0.8796,
+      "grad_norm": 0.1640625,
+      "learning_rate": 8.702334416349278e-06,
+      "loss": 1.2723,
+      "step": 43980
+    },
+    {
+      "epoch": 0.8797,
+      "grad_norm": 0.27734375,
+      "learning_rate": 8.688097680346453e-06,
+      "loss": 1.2648,
+      "step": 43985
+    },
+    {
+      "epoch": 0.8798,
+      "grad_norm": 0.2734375,
+      "learning_rate": 8.67387207042194e-06,
+      "loss": 1.2695,
+      "step": 43990
+    },
+    {
+      "epoch": 0.8799,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 8.6596575883091e-06,
+      "loss": 1.2863,
+      "step": 43995
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.645454235739903e-06,
+      "loss": 1.27,
+      "step": 44000
+    },
+    {
+      "epoch": 0.8801,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.63126201444503e-06,
+      "loss": 1.28,
+      "step": 44005
+    },
+    {
+      "epoch": 0.8802,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.617080926153698e-06,
+      "loss": 1.2825,
+      "step": 44010
+    },
+    {
+      "epoch": 0.8803,
+      "grad_norm": 0.384765625,
+      "learning_rate": 8.602910972593892e-06,
+      "loss": 1.2874,
+      "step": 44015
+    },
+    {
+      "epoch": 0.8804,
+      "grad_norm": 0.140625,
+      "learning_rate": 8.588752155492119e-06,
+      "loss": 1.2722,
+      "step": 44020
+    },
+    {
+      "epoch": 0.8805,
+      "grad_norm": 0.40234375,
+      "learning_rate": 8.574604476573621e-06,
+      "loss": 1.2818,
+      "step": 44025
+    },
+    {
+      "epoch": 0.8806,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.560467937562277e-06,
+      "loss": 1.2851,
+      "step": 44030
+    },
+    {
+      "epoch": 0.8807,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.546342540180508e-06,
+      "loss": 1.2726,
+      "step": 44035
+    },
+    {
+      "epoch": 0.8808,
+      "grad_norm": 0.498046875,
+      "learning_rate": 8.532228286149501e-06,
+      "loss": 1.2814,
+      "step": 44040
+    },
+    {
+      "epoch": 0.8809,
+      "grad_norm": 0.1640625,
+      "learning_rate": 8.51812517718904e-06,
+      "loss": 1.2531,
+      "step": 44045
+    },
+    {
+      "epoch": 0.881,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.504033215017527e-06,
+      "loss": 1.281,
+      "step": 44050
+    },
+    {
+      "epoch": 0.8811,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.489952401352019e-06,
+      "loss": 1.2822,
+      "step": 44055
+    },
+    {
+      "epoch": 0.8812,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.475882737908248e-06,
+      "loss": 1.2708,
+      "step": 44060
+    },
+    {
+      "epoch": 0.8813,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 8.46182422640054e-06,
+      "loss": 1.2772,
+      "step": 44065
+    },
+    {
+      "epoch": 0.8814,
+      "grad_norm": 0.158203125,
+      "learning_rate": 8.447776868541879e-06,
+      "loss": 1.2725,
+      "step": 44070
+    },
+    {
+      "epoch": 0.8815,
+      "grad_norm": 0.19140625,
+      "learning_rate": 8.433740666043898e-06,
+      "loss": 1.2741,
+      "step": 44075
+    },
+    {
+      "epoch": 0.8816,
+      "grad_norm": 0.359375,
+      "learning_rate": 8.419715620616874e-06,
+      "loss": 1.2652,
+      "step": 44080
+    },
+    {
+      "epoch": 0.8817,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.405701733969706e-06,
+      "loss": 1.2603,
+      "step": 44085
+    },
+    {
+      "epoch": 0.8818,
+      "grad_norm": 0.3203125,
+      "learning_rate": 8.39169900780995e-06,
+      "loss": 1.2807,
+      "step": 44090
+    },
+    {
+      "epoch": 0.8819,
+      "grad_norm": 0.205078125,
+      "learning_rate": 8.377707443843786e-06,
+      "loss": 1.2662,
+      "step": 44095
+    },
+    {
+      "epoch": 0.882,
+      "grad_norm": 0.37109375,
+      "learning_rate": 8.363727043776038e-06,
+      "loss": 1.2637,
+      "step": 44100
+    },
+    {
+      "epoch": 0.8821,
+      "grad_norm": 0.17578125,
+      "learning_rate": 8.34975780931021e-06,
+      "loss": 1.2722,
+      "step": 44105
+    },
+    {
+      "epoch": 0.8822,
+      "grad_norm": 0.2890625,
+      "learning_rate": 8.335799742148387e-06,
+      "loss": 1.2781,
+      "step": 44110
+    },
+    {
+      "epoch": 0.8823,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.321852843991295e-06,
+      "loss": 1.2828,
+      "step": 44115
+    },
+    {
+      "epoch": 0.8824,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.307917116538378e-06,
+      "loss": 1.2922,
+      "step": 44120
+    },
+    {
+      "epoch": 0.8825,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.293992561487596e-06,
+      "loss": 1.2672,
+      "step": 44125
+    },
+    {
+      "epoch": 0.8826,
+      "grad_norm": 0.373046875,
+      "learning_rate": 8.280079180535672e-06,
+      "loss": 1.2801,
+      "step": 44130
+    },
+    {
+      "epoch": 0.8827,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.26617697537786e-06,
+      "loss": 1.304,
+      "step": 44135
+    },
+    {
+      "epoch": 0.8828,
+      "grad_norm": 0.208984375,
+      "learning_rate": 8.252285947708139e-06,
+      "loss": 1.2634,
+      "step": 44140
+    },
+    {
+      "epoch": 0.8829,
+      "grad_norm": 0.328125,
+      "learning_rate": 8.238406099219077e-06,
+      "loss": 1.2788,
+      "step": 44145
+    },
+    {
+      "epoch": 0.883,
+      "grad_norm": 0.18359375,
+      "learning_rate": 8.224537431601886e-06,
+      "loss": 1.2731,
+      "step": 44150
+    },
+    {
+      "epoch": 0.8831,
+      "grad_norm": 0.421875,
+      "learning_rate": 8.21067994654644e-06,
+      "loss": 1.2684,
+      "step": 44155
+    },
+    {
+      "epoch": 0.8832,
+      "grad_norm": 0.1953125,
+      "learning_rate": 8.196833645741186e-06,
+      "loss": 1.2711,
+      "step": 44160
+    },
+    {
+      "epoch": 0.8833,
+      "grad_norm": 0.22265625,
+      "learning_rate": 8.182998530873298e-06,
+      "loss": 1.2787,
+      "step": 44165
+    },
+    {
+      "epoch": 0.8834,
+      "grad_norm": 0.443359375,
+      "learning_rate": 8.169174603628538e-06,
+      "loss": 1.2819,
+      "step": 44170
+    },
+    {
+      "epoch": 0.8835,
+      "grad_norm": 0.171875,
+      "learning_rate": 8.15536186569129e-06,
+      "loss": 1.2697,
+      "step": 44175
+    },
+    {
+      "epoch": 0.8836,
+      "grad_norm": 0.400390625,
+      "learning_rate": 8.1415603187446e-06,
+      "loss": 1.2698,
+      "step": 44180
+    },
+    {
+      "epoch": 0.8837,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.127769964470156e-06,
+      "loss": 1.265,
+      "step": 44185
+    },
+    {
+      "epoch": 0.8838,
+      "grad_norm": 0.28515625,
+      "learning_rate": 8.113990804548244e-06,
+      "loss": 1.2775,
+      "step": 44190
+    },
+    {
+      "epoch": 0.8839,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.100222840657878e-06,
+      "loss": 1.2768,
+      "step": 44195
+    },
+    {
+      "epoch": 0.884,
+      "grad_norm": 0.2109375,
+      "learning_rate": 8.086466074476563e-06,
+      "loss": 1.2764,
+      "step": 44200
+    },
+    {
+      "epoch": 0.8841,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.072720507680565e-06,
+      "loss": 1.2576,
+      "step": 44205
+    },
+    {
+      "epoch": 0.8842,
+      "grad_norm": 0.275390625,
+      "learning_rate": 8.058986141944724e-06,
+      "loss": 1.2581,
+      "step": 44210
+    },
+    {
+      "epoch": 0.8843,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.045262978942513e-06,
+      "loss": 1.2689,
+      "step": 44215
+    },
+    {
+      "epoch": 0.8844,
+      "grad_norm": 0.255859375,
+      "learning_rate": 8.031551020346128e-06,
+      "loss": 1.2715,
+      "step": 44220
+    },
+    {
+      "epoch": 0.8845,
+      "grad_norm": 0.189453125,
+      "learning_rate": 8.017850267826232e-06,
+      "loss": 1.2805,
+      "step": 44225
+    },
+    {
+      "epoch": 0.8846,
+      "grad_norm": 0.1875,
+      "learning_rate": 8.004160723052312e-06,
+      "loss": 1.2944,
+      "step": 44230
+    },
+    {
+      "epoch": 0.8847,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.990482387692311e-06,
+      "loss": 1.29,
+      "step": 44235
+    },
+    {
+      "epoch": 0.8848,
+      "grad_norm": 0.23046875,
+      "learning_rate": 7.976815263412963e-06,
+      "loss": 1.284,
+      "step": 44240
+    },
+    {
+      "epoch": 0.8849,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.963159351879556e-06,
+      "loss": 1.2795,
+      "step": 44245
+    },
+    {
+      "epoch": 0.885,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 7.949514654755962e-06,
+      "loss": 1.2754,
+      "step": 44250
+    },
+    {
+      "epoch": 0.8851,
+      "grad_norm": 0.486328125,
+      "learning_rate": 7.935881173704819e-06,
+      "loss": 1.2708,
+      "step": 44255
+    },
+    {
+      "epoch": 0.8852,
+      "grad_norm": 0.2109375,
+      "learning_rate": 7.922258910387282e-06,
+      "loss": 1.269,
+      "step": 44260
+    },
+    {
+      "epoch": 0.8853,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.908647866463203e-06,
+      "loss": 1.2577,
+      "step": 44265
+    },
+    {
+      "epoch": 0.8854,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.895048043591036e-06,
+      "loss": 1.2704,
+      "step": 44270
+    },
+    {
+      "epoch": 0.8855,
+      "grad_norm": 0.212890625,
+      "learning_rate": 7.881459443427886e-06,
+      "loss": 1.2869,
+      "step": 44275
+    },
+    {
+      "epoch": 0.8856,
+      "grad_norm": 0.361328125,
+      "learning_rate": 7.867882067629472e-06,
+      "loss": 1.281,
+      "step": 44280
+    },
+    {
+      "epoch": 0.8857,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 7.854315917850163e-06,
+      "loss": 1.2917,
+      "step": 44285
+    },
+    {
+      "epoch": 0.8858,
+      "grad_norm": 0.29296875,
+      "learning_rate": 7.840760995742946e-06,
+      "loss": 1.2766,
+      "step": 44290
+    },
+    {
+      "epoch": 0.8859,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.827217302959467e-06,
+      "loss": 1.2722,
+      "step": 44295
+    },
+    {
+      "epoch": 0.886,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.81368484114996e-06,
+      "loss": 1.2671,
+      "step": 44300
+    },
+    {
+      "epoch": 0.8861,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 7.800163611963318e-06,
+      "loss": 1.2828,
+      "step": 44305
+    },
+    {
+      "epoch": 0.8862,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.786653617047079e-06,
+      "loss": 1.2851,
+      "step": 44310
+    },
+    {
+      "epoch": 0.8863,
+      "grad_norm": 0.296875,
+      "learning_rate": 7.77315485804736e-06,
+      "loss": 1.2824,
+      "step": 44315
+    },
+    {
+      "epoch": 0.8864,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.75966733660901e-06,
+      "loss": 1.2683,
+      "step": 44320
+    },
+    {
+      "epoch": 0.8865,
+      "grad_norm": 0.48046875,
+      "learning_rate": 7.746191054375362e-06,
+      "loss": 1.2575,
+      "step": 44325
+    },
+    {
+      "epoch": 0.8866,
+      "grad_norm": 0.45703125,
+      "learning_rate": 7.73272601298851e-06,
+      "loss": 1.2692,
+      "step": 44330
+    },
+    {
+      "epoch": 0.8867,
+      "grad_norm": 0.251953125,
+      "learning_rate": 7.719272214089145e-06,
+      "loss": 1.2583,
+      "step": 44335
+    },
+    {
+      "epoch": 0.8868,
+      "grad_norm": 0.318359375,
+      "learning_rate": 7.7058296593165e-06,
+      "loss": 1.2604,
+      "step": 44340
+    },
+    {
+      "epoch": 0.8869,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 7.692398350308594e-06,
+      "loss": 1.2657,
+      "step": 44345
+    },
+    {
+      "epoch": 0.887,
+      "grad_norm": 0.240234375,
+      "learning_rate": 7.67897828870191e-06,
+      "loss": 1.2647,
+      "step": 44350
+    },
+    {
+      "epoch": 0.8871,
+      "grad_norm": 0.19921875,
+      "learning_rate": 7.665569476131706e-06,
+      "loss": 1.276,
+      "step": 44355
+    },
+    {
+      "epoch": 0.8872,
+      "grad_norm": 0.306640625,
+      "learning_rate": 7.652171914231776e-06,
+      "loss": 1.2764,
+      "step": 44360
+    },
+    {
+      "epoch": 0.8873,
+      "grad_norm": 0.224609375,
+      "learning_rate": 7.638785604634579e-06,
+      "loss": 1.2747,
+      "step": 44365
+    },
+    {
+      "epoch": 0.8874,
+      "grad_norm": 0.318359375,
+      "learning_rate": 7.625410548971191e-06,
+      "loss": 1.2654,
+      "step": 44370
+    },
+    {
+      "epoch": 0.8875,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.612046748871327e-06,
+      "loss": 1.2737,
+      "step": 44375
+    },
+    {
+      "epoch": 0.8876,
+      "grad_norm": 0.28125,
+      "learning_rate": 7.59869420596333e-06,
+      "loss": 1.2611,
+      "step": 44380
+    },
+    {
+      "epoch": 0.8877,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.585352921874156e-06,
+      "loss": 1.2806,
+      "step": 44385
+    },
+    {
+      "epoch": 0.8878,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 7.572022898229403e-06,
+      "loss": 1.274,
+      "step": 44390
+    },
+    {
+      "epoch": 0.8879,
+      "grad_norm": 0.16015625,
+      "learning_rate": 7.558704136653305e-06,
+      "loss": 1.2804,
+      "step": 44395
+    },
+    {
+      "epoch": 0.888,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.545396638768698e-06,
+      "loss": 1.2734,
+      "step": 44400
+    },
+    {
+      "epoch": 0.8881,
+      "grad_norm": 0.291015625,
+      "learning_rate": 7.5321004061970405e-06,
+      "loss": 1.2725,
+      "step": 44405
+    },
+    {
+      "epoch": 0.8882,
+      "grad_norm": 0.314453125,
+      "learning_rate": 7.518815440558513e-06,
+      "loss": 1.2799,
+      "step": 44410
+    },
+    {
+      "epoch": 0.8883,
+      "grad_norm": 0.3125,
+      "learning_rate": 7.505541743471756e-06,
+      "loss": 1.2739,
+      "step": 44415
+    },
+    {
+      "epoch": 0.8884,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.492279316554207e-06,
+      "loss": 1.2807,
+      "step": 44420
+    },
+    {
+      "epoch": 0.8885,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.479028161421797e-06,
+      "loss": 1.2731,
+      "step": 44425
+    },
+    {
+      "epoch": 0.8886,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.465788279689156e-06,
+      "loss": 1.2763,
+      "step": 44430
+    },
+    {
+      "epoch": 0.8887,
+      "grad_norm": 0.28125,
+      "learning_rate": 7.45255967296955e-06,
+      "loss": 1.2796,
+      "step": 44435
+    },
+    {
+      "epoch": 0.8888,
+      "grad_norm": 0.365234375,
+      "learning_rate": 7.439342342874789e-06,
+      "loss": 1.2753,
+      "step": 44440
+    },
+    {
+      "epoch": 0.8889,
+      "grad_norm": 0.3046875,
+      "learning_rate": 7.426136291015417e-06,
+      "loss": 1.2813,
+      "step": 44445
+    },
+    {
+      "epoch": 0.889,
+      "grad_norm": 0.197265625,
+      "learning_rate": 7.412941519000527e-06,
+      "loss": 1.2891,
+      "step": 44450
+    },
+    {
+      "epoch": 0.8891,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.399758028437864e-06,
+      "loss": 1.2873,
+      "step": 44455
+    },
+    {
+      "epoch": 0.8892,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.386585820933811e-06,
+      "loss": 1.2942,
+      "step": 44460
+    },
+    {
+      "epoch": 0.8893,
+      "grad_norm": 0.236328125,
+      "learning_rate": 7.3734248980933395e-06,
+      "loss": 1.2725,
+      "step": 44465
+    },
+    {
+      "epoch": 0.8894,
+      "grad_norm": 0.1953125,
+      "learning_rate": 7.360275261520078e-06,
+      "loss": 1.2779,
+      "step": 44470
+    },
+    {
+      "epoch": 0.8895,
+      "grad_norm": 0.27734375,
+      "learning_rate": 7.347136912816277e-06,
+      "loss": 1.2886,
+      "step": 44475
+    },
+    {
+      "epoch": 0.8896,
+      "grad_norm": 0.201171875,
+      "learning_rate": 7.3340098535827905e-06,
+      "loss": 1.2615,
+      "step": 44480
+    },
+    {
+      "epoch": 0.8897,
+      "grad_norm": 0.30859375,
+      "learning_rate": 7.320894085419116e-06,
+      "loss": 1.2714,
+      "step": 44485
+    },
+    {
+      "epoch": 0.8898,
+      "grad_norm": 0.455078125,
+      "learning_rate": 7.3077896099233765e-06,
+      "loss": 1.266,
+      "step": 44490
+    },
+    {
+      "epoch": 0.8899,
+      "grad_norm": 0.314453125,
+      "learning_rate": 7.2946964286923046e-06,
+      "loss": 1.2808,
+      "step": 44495
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.451171875,
+      "learning_rate": 7.281614543321269e-06,
+      "loss": 1.2734,
+      "step": 44500
+    },
+    {
+      "epoch": 0.8901,
+      "grad_norm": 0.283203125,
+      "learning_rate": 7.268543955404239e-06,
+      "loss": 1.2828,
+      "step": 44505
+    },
+    {
+      "epoch": 0.8902,
+      "grad_norm": 0.46484375,
+      "learning_rate": 7.255484666533874e-06,
+      "loss": 1.2769,
+      "step": 44510
+    },
+    {
+      "epoch": 0.8903,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.242436678301367e-06,
+      "loss": 1.2786,
+      "step": 44515
+    },
+    {
+      "epoch": 0.8904,
+      "grad_norm": 0.1328125,
+      "learning_rate": 7.2293999922965705e-06,
+      "loss": 1.2752,
+      "step": 44520
+    },
+    {
+      "epoch": 0.8905,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.216374610108012e-06,
+      "loss": 1.2778,
+      "step": 44525
+    },
+    {
+      "epoch": 0.8906,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.203360533322734e-06,
+      "loss": 1.2708,
+      "step": 44530
+    },
+    {
+      "epoch": 0.8907,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.190357763526523e-06,
+      "loss": 1.2754,
+      "step": 44535
+    },
+    {
+      "epoch": 0.8908,
+      "grad_norm": 0.19921875,
+      "learning_rate": 7.177366302303667e-06,
+      "loss": 1.2637,
+      "step": 44540
+    },
+    {
+      "epoch": 0.8909,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 7.164386151237179e-06,
+      "loss": 1.2683,
+      "step": 44545
+    },
+    {
+      "epoch": 0.891,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 7.151417311908648e-06,
+      "loss": 1.29,
+      "step": 44550
+    },
+    {
+      "epoch": 0.8911,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.138459785898266e-06,
+      "loss": 1.2726,
+      "step": 44555
+    },
+    {
+      "epoch": 0.8912,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 7.125513574784903e-06,
+      "loss": 1.2949,
+      "step": 44560
+    },
+    {
+      "epoch": 0.8913,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 7.112578680145954e-06,
+      "loss": 1.2862,
+      "step": 44565
+    },
+    {
+      "epoch": 0.8914,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.099655103557556e-06,
+      "loss": 1.269,
+      "step": 44570
+    },
+    {
+      "epoch": 0.8915,
+      "grad_norm": 0.22265625,
+      "learning_rate": 7.086742846594385e-06,
+      "loss": 1.2669,
+      "step": 44575
+    },
+    {
+      "epoch": 0.8916,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 7.07384191082977e-06,
+      "loss": 1.2615,
+      "step": 44580
+    },
+    {
+      "epoch": 0.8917,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 7.060952297835633e-06,
+      "loss": 1.2752,
+      "step": 44585
+    },
+    {
+      "epoch": 0.8918,
+      "grad_norm": 0.24609375,
+      "learning_rate": 7.048074009182548e-06,
+      "loss": 1.2813,
+      "step": 44590
+    },
+    {
+      "epoch": 0.8919,
+      "grad_norm": 0.3984375,
+      "learning_rate": 7.035207046439673e-06,
+      "loss": 1.2958,
+      "step": 44595
+    },
+    {
+      "epoch": 0.892,
+      "grad_norm": 0.37890625,
+      "learning_rate": 7.022351411174866e-06,
+      "loss": 1.2648,
+      "step": 44600
+    },
+    {
+      "epoch": 0.8921,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 7.009507104954493e-06,
+      "loss": 1.2734,
+      "step": 44605
+    },
+    {
+      "epoch": 0.8922,
+      "grad_norm": 0.3828125,
+      "learning_rate": 6.996674129343606e-06,
+      "loss": 1.2882,
+      "step": 44610
+    },
+    {
+      "epoch": 0.8923,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 6.9838524859058616e-06,
+      "loss": 1.259,
+      "step": 44615
+    },
+    {
+      "epoch": 0.8924,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.971042176203535e-06,
+      "loss": 1.2831,
+      "step": 44620
+    },
+    {
+      "epoch": 0.8925,
+      "grad_norm": 0.19140625,
+      "learning_rate": 6.958243201797554e-06,
+      "loss": 1.2727,
+      "step": 44625
+    },
+    {
+      "epoch": 0.8926,
+      "grad_norm": 0.267578125,
+      "learning_rate": 6.945455564247394e-06,
+      "loss": 1.2671,
+      "step": 44630
+    },
+    {
+      "epoch": 0.8927,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.932679265111231e-06,
+      "loss": 1.2795,
+      "step": 44635
+    },
+    {
+      "epoch": 0.8928,
+      "grad_norm": 0.494140625,
+      "learning_rate": 6.919914305945774e-06,
+      "loss": 1.2685,
+      "step": 44640
+    },
+    {
+      "epoch": 0.8929,
+      "grad_norm": 0.84375,
+      "learning_rate": 6.907160688306425e-06,
+      "loss": 1.2648,
+      "step": 44645
+    },
+    {
+      "epoch": 0.893,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.894418413747183e-06,
+      "loss": 1.2707,
+      "step": 44650
+    },
+    {
+      "epoch": 0.8931,
+      "grad_norm": 0.2578125,
+      "learning_rate": 6.881687483820609e-06,
+      "loss": 1.2883,
+      "step": 44655
+    },
+    {
+      "epoch": 0.8932,
+      "grad_norm": 0.251953125,
+      "learning_rate": 6.868967900077972e-06,
+      "loss": 1.2625,
+      "step": 44660
+    },
+    {
+      "epoch": 0.8933,
+      "grad_norm": 0.41796875,
+      "learning_rate": 6.856259664069098e-06,
+      "loss": 1.298,
+      "step": 44665
+    },
+    {
+      "epoch": 0.8934,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 6.8435627773424495e-06,
+      "loss": 1.2789,
+      "step": 44670
+    },
+    {
+      "epoch": 0.8935,
+      "grad_norm": 0.431640625,
+      "learning_rate": 6.830877241445111e-06,
+      "loss": 1.2756,
+      "step": 44675
+    },
+    {
+      "epoch": 0.8936,
+      "grad_norm": 0.2109375,
+      "learning_rate": 6.818203057922757e-06,
+      "loss": 1.2686,
+      "step": 44680
+    },
+    {
+      "epoch": 0.8937,
+      "grad_norm": 0.15625,
+      "learning_rate": 6.805540228319718e-06,
+      "loss": 1.2605,
+      "step": 44685
+    },
+    {
+      "epoch": 0.8938,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.7928887541789055e-06,
+      "loss": 1.2844,
+      "step": 44690
+    },
+    {
+      "epoch": 0.8939,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 6.780248637041875e-06,
+      "loss": 1.2654,
+      "step": 44695
+    },
+    {
+      "epoch": 0.894,
+      "grad_norm": 0.419921875,
+      "learning_rate": 6.767619878448783e-06,
+      "loss": 1.2586,
+      "step": 44700
+    },
+    {
+      "epoch": 0.8941,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 6.755002479938411e-06,
+      "loss": 1.2887,
+      "step": 44705
+    },
+    {
+      "epoch": 0.8942,
+      "grad_norm": 0.416015625,
+      "learning_rate": 6.742396443048138e-06,
+      "loss": 1.261,
+      "step": 44710
+    },
+    {
+      "epoch": 0.8943,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 6.729801769313981e-06,
+      "loss": 1.2712,
+      "step": 44715
+    },
+    {
+      "epoch": 0.8944,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.717218460270536e-06,
+      "loss": 1.2705,
+      "step": 44720
+    },
+    {
+      "epoch": 0.8945,
+      "grad_norm": 0.146484375,
+      "learning_rate": 6.704646517451107e-06,
+      "loss": 1.2665,
+      "step": 44725
+    },
+    {
+      "epoch": 0.8946,
+      "grad_norm": 0.158203125,
+      "learning_rate": 6.692085942387483e-06,
+      "loss": 1.2743,
+      "step": 44730
+    },
+    {
+      "epoch": 0.8947,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 6.679536736610137e-06,
+      "loss": 1.2684,
+      "step": 44735
+    },
+    {
+      "epoch": 0.8948,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.666998901648203e-06,
+      "loss": 1.2731,
+      "step": 44740
+    },
+    {
+      "epoch": 0.8949,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 6.654472439029313e-06,
+      "loss": 1.2567,
+      "step": 44745
+    },
+    {
+      "epoch": 0.895,
+      "grad_norm": 0.185546875,
+      "learning_rate": 6.6419573502798374e-06,
+      "loss": 1.2765,
+      "step": 44750
+    },
+    {
+      "epoch": 0.8951,
+      "grad_norm": 1.1875,
+      "learning_rate": 6.629453636924643e-06,
+      "loss": 1.2828,
+      "step": 44755
+    },
+    {
+      "epoch": 0.8952,
+      "grad_norm": 0.400390625,
+      "learning_rate": 6.616961300487324e-06,
+      "loss": 1.2759,
+      "step": 44760
+    },
+    {
+      "epoch": 0.8953,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.604480342490004e-06,
+      "loss": 1.2837,
+      "step": 44765
+    },
+    {
+      "epoch": 0.8954,
+      "grad_norm": 0.419921875,
+      "learning_rate": 6.592010764453449e-06,
+      "loss": 1.2539,
+      "step": 44770
+    },
+    {
+      "epoch": 0.8955,
+      "grad_norm": 0.263671875,
+      "learning_rate": 6.579552567897051e-06,
+      "loss": 1.2877,
+      "step": 44775
+    },
+    {
+      "epoch": 0.8956,
+      "grad_norm": 0.30859375,
+      "learning_rate": 6.5671057543387985e-06,
+      "loss": 1.2774,
+      "step": 44780
+    },
+    {
+      "epoch": 0.8957,
+      "grad_norm": 0.197265625,
+      "learning_rate": 6.554670325295298e-06,
+      "loss": 1.2889,
+      "step": 44785
+    },
+    {
+      "epoch": 0.8958,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 6.542246282281772e-06,
+      "loss": 1.272,
+      "step": 44790
+    },
+    {
+      "epoch": 0.8959,
+      "grad_norm": 0.4140625,
+      "learning_rate": 6.529833626812043e-06,
+      "loss": 1.288,
+      "step": 44795
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.517432360398556e-06,
+      "loss": 1.2823,
+      "step": 44800
+    },
+    {
+      "epoch": 0.8961,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.5050424845523815e-06,
+      "loss": 1.2866,
+      "step": 44805
+    },
+    {
+      "epoch": 0.8962,
+      "grad_norm": 0.2109375,
+      "learning_rate": 6.492664000783166e-06,
+      "loss": 1.2747,
+      "step": 44810
+    },
+    {
+      "epoch": 0.8963,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.480296910599237e-06,
+      "loss": 1.2862,
+      "step": 44815
+    },
+    {
+      "epoch": 0.8964,
+      "grad_norm": 0.279296875,
+      "learning_rate": 6.467941215507433e-06,
+      "loss": 1.2893,
+      "step": 44820
+    },
+    {
+      "epoch": 0.8965,
+      "grad_norm": 0.3125,
+      "learning_rate": 6.455596917013273e-06,
+      "loss": 1.2812,
+      "step": 44825
+    },
+    {
+      "epoch": 0.8966,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.443264016620887e-06,
+      "loss": 1.2658,
+      "step": 44830
+    },
+    {
+      "epoch": 0.8967,
+      "grad_norm": 0.875,
+      "learning_rate": 6.430942515832983e-06,
+      "loss": 1.2678,
+      "step": 44835
+    },
+    {
+      "epoch": 0.8968,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.418632416150927e-06,
+      "loss": 1.2977,
+      "step": 44840
+    },
+    {
+      "epoch": 0.8969,
+      "grad_norm": 0.23828125,
+      "learning_rate": 6.406333719074619e-06,
+      "loss": 1.3,
+      "step": 44845
+    },
+    {
+      "epoch": 0.897,
+      "grad_norm": 0.40625,
+      "learning_rate": 6.394046426102674e-06,
+      "loss": 1.2662,
+      "step": 44850
+    },
+    {
+      "epoch": 0.8971,
+      "grad_norm": 0.267578125,
+      "learning_rate": 6.381770538732224e-06,
+      "loss": 1.2672,
+      "step": 44855
+    },
+    {
+      "epoch": 0.8972,
+      "grad_norm": 0.314453125,
+      "learning_rate": 6.3695060584590625e-06,
+      "loss": 1.2643,
+      "step": 44860
+    },
+    {
+      "epoch": 0.8973,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 6.357252986777595e-06,
+      "loss": 1.2695,
+      "step": 44865
+    },
+    {
+      "epoch": 0.8974,
+      "grad_norm": 0.265625,
+      "learning_rate": 6.345011325180772e-06,
+      "loss": 1.2736,
+      "step": 44870
+    },
+    {
+      "epoch": 0.8975,
+      "grad_norm": 0.2265625,
+      "learning_rate": 6.332781075160243e-06,
+      "loss": 1.265,
+      "step": 44875
+    },
+    {
+      "epoch": 0.8976,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.320562238206218e-06,
+      "loss": 1.2902,
+      "step": 44880
+    },
+    {
+      "epoch": 0.8977,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.308354815807527e-06,
+      "loss": 1.2527,
+      "step": 44885
+    },
+    {
+      "epoch": 0.8978,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.296158809451602e-06,
+      "loss": 1.2837,
+      "step": 44890
+    },
+    {
+      "epoch": 0.8979,
+      "grad_norm": 0.298828125,
+      "learning_rate": 6.283974220624489e-06,
+      "loss": 1.2693,
+      "step": 44895
+    },
+    {
+      "epoch": 0.898,
+      "grad_norm": 0.2734375,
+      "learning_rate": 6.2718010508108545e-06,
+      "loss": 1.3028,
+      "step": 44900
+    },
+    {
+      "epoch": 0.8981,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.259639301493947e-06,
+      "loss": 1.2604,
+      "step": 44905
+    },
+    {
+      "epoch": 0.8982,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.2474889741556575e-06,
+      "loss": 1.2716,
+      "step": 44910
+    },
+    {
+      "epoch": 0.8983,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 6.235350070276447e-06,
+      "loss": 1.2818,
+      "step": 44915
+    },
+    {
+      "epoch": 0.8984,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.223222591335409e-06,
+      "loss": 1.2759,
+      "step": 44920
+    },
+    {
+      "epoch": 0.8985,
+      "grad_norm": 0.20703125,
+      "learning_rate": 6.21110653881023e-06,
+      "loss": 1.264,
+      "step": 44925
+    },
+    {
+      "epoch": 0.8986,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 6.1990019141772605e-06,
+      "loss": 1.2785,
+      "step": 44930
+    },
+    {
+      "epoch": 0.8987,
+      "grad_norm": 0.322265625,
+      "learning_rate": 6.186908718911344e-06,
+      "loss": 1.2709,
+      "step": 44935
+    },
+    {
+      "epoch": 0.8988,
+      "grad_norm": 0.18359375,
+      "learning_rate": 6.174826954486068e-06,
+      "loss": 1.2793,
+      "step": 44940
+    },
+    {
+      "epoch": 0.8989,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 6.1627566223735e-06,
+      "loss": 1.2698,
+      "step": 44945
+    },
+    {
+      "epoch": 0.899,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 6.1506977240444074e-06,
+      "loss": 1.2746,
+      "step": 44950
+    },
+    {
+      "epoch": 0.8991,
+      "grad_norm": 0.435546875,
+      "learning_rate": 6.138650260968137e-06,
+      "loss": 1.2721,
+      "step": 44955
+    },
+    {
+      "epoch": 0.8992,
+      "grad_norm": 0.30078125,
+      "learning_rate": 6.126614234612593e-06,
+      "loss": 1.2955,
+      "step": 44960
+    },
+    {
+      "epoch": 0.8993,
+      "grad_norm": 0.50390625,
+      "learning_rate": 6.1145896464443685e-06,
+      "loss": 1.2724,
+      "step": 44965
+    },
+    {
+      "epoch": 0.8994,
+      "grad_norm": 0.83984375,
+      "learning_rate": 6.102576497928614e-06,
+      "loss": 1.2608,
+      "step": 44970
+    },
+    {
+      "epoch": 0.8995,
+      "grad_norm": 0.236328125,
+      "learning_rate": 6.090574790529091e-06,
+      "loss": 1.2686,
+      "step": 44975
+    },
+    {
+      "epoch": 0.8996,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 6.078584525708176e-06,
+      "loss": 1.2744,
+      "step": 44980
+    },
+    {
+      "epoch": 0.8997,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.066605704926831e-06,
+      "loss": 1.2696,
+      "step": 44985
+    },
+    {
+      "epoch": 0.8998,
+      "grad_norm": 0.1640625,
+      "learning_rate": 6.054638329644657e-06,
+      "loss": 1.2694,
+      "step": 44990
+    },
+    {
+      "epoch": 0.8999,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 6.042682401319844e-06,
+      "loss": 1.2744,
+      "step": 44995
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.294921875,
+      "learning_rate": 6.030737921409169e-06,
+      "loss": 1.2738,
+      "step": 45000
+    },
+    {
+      "epoch": 0.9001,
+      "grad_norm": 0.322265625,
+      "learning_rate": 6.018804891368035e-06,
+      "loss": 1.2865,
+      "step": 45005
+    },
+    {
+      "epoch": 0.9002,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 6.006883312650457e-06,
+      "loss": 1.2552,
+      "step": 45010
+    },
+    {
+      "epoch": 0.9003,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.994973186709041e-06,
+      "loss": 1.2723,
+      "step": 45015
+    },
+    {
+      "epoch": 0.9004,
+      "grad_norm": 0.3671875,
+      "learning_rate": 5.98307451499498e-06,
+      "loss": 1.2807,
+      "step": 45020
+    },
+    {
+      "epoch": 0.9005,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 5.971187298958103e-06,
+      "loss": 1.2737,
+      "step": 45025
+    },
+    {
+      "epoch": 0.9006,
+      "grad_norm": 0.2890625,
+      "learning_rate": 5.9593115400468636e-06,
+      "loss": 1.2645,
+      "step": 45030
+    },
+    {
+      "epoch": 0.9007,
+      "grad_norm": 0.4765625,
+      "learning_rate": 5.947447239708215e-06,
+      "loss": 1.2696,
+      "step": 45035
+    },
+    {
+      "epoch": 0.9008,
+      "grad_norm": 0.8359375,
+      "learning_rate": 5.935594399387856e-06,
+      "loss": 1.2876,
+      "step": 45040
+    },
+    {
+      "epoch": 0.9009,
+      "grad_norm": 0.216796875,
+      "learning_rate": 5.923753020529999e-06,
+      "loss": 1.2654,
+      "step": 45045
+    },
+    {
+      "epoch": 0.901,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.911923104577455e-06,
+      "loss": 1.271,
+      "step": 45050
+    },
+    {
+      "epoch": 0.9011,
+      "grad_norm": 0.416015625,
+      "learning_rate": 5.900104652971694e-06,
+      "loss": 1.2745,
+      "step": 45055
+    },
+    {
+      "epoch": 0.9012,
+      "grad_norm": 0.306640625,
+      "learning_rate": 5.888297667152731e-06,
+      "loss": 1.2719,
+      "step": 45060
+    },
+    {
+      "epoch": 0.9013,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.8765021485592376e-06,
+      "loss": 1.2824,
+      "step": 45065
+    },
+    {
+      "epoch": 0.9014,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.864718098628441e-06,
+      "loss": 1.2951,
+      "step": 45070
+    },
+    {
+      "epoch": 0.9015,
+      "grad_norm": 0.416015625,
+      "learning_rate": 5.852945518796205e-06,
+      "loss": 1.2652,
+      "step": 45075
+    },
+    {
+      "epoch": 0.9016,
+      "grad_norm": 0.296875,
+      "learning_rate": 5.8411844104969916e-06,
+      "loss": 1.2836,
+      "step": 45080
+    },
+    {
+      "epoch": 0.9017,
+      "grad_norm": 0.2265625,
+      "learning_rate": 5.829434775163833e-06,
+      "loss": 1.2866,
+      "step": 45085
+    },
+    {
+      "epoch": 0.9018,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 5.8176966142283965e-06,
+      "loss": 1.2658,
+      "step": 45090
+    },
+    {
+      "epoch": 0.9019,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.805969929120947e-06,
+      "loss": 1.2721,
+      "step": 45095
+    },
+    {
+      "epoch": 0.902,
+      "grad_norm": 0.208984375,
+      "learning_rate": 5.7942547212703315e-06,
+      "loss": 1.2724,
+      "step": 45100
+    },
+    {
+      "epoch": 0.9021,
+      "grad_norm": 0.322265625,
+      "learning_rate": 5.78255099210403e-06,
+      "loss": 1.2726,
+      "step": 45105
+    },
+    {
+      "epoch": 0.9022,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.770858743048091e-06,
+      "loss": 1.2671,
+      "step": 45110
+    },
+    {
+      "epoch": 0.9023,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 5.759177975527186e-06,
+      "loss": 1.2788,
+      "step": 45115
+    },
+    {
+      "epoch": 0.9024,
+      "grad_norm": 0.32421875,
+      "learning_rate": 5.747508690964598e-06,
+      "loss": 1.2708,
+      "step": 45120
+    },
+    {
+      "epoch": 0.9025,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 5.735850890782157e-06,
+      "loss": 1.2897,
+      "step": 45125
+    },
+    {
+      "epoch": 0.9026,
+      "grad_norm": 0.37109375,
+      "learning_rate": 5.724204576400371e-06,
+      "loss": 1.2637,
+      "step": 45130
+    },
+    {
+      "epoch": 0.9027,
+      "grad_norm": 0.439453125,
+      "learning_rate": 5.7125697492382835e-06,
+      "loss": 1.2718,
+      "step": 45135
+    },
+    {
+      "epoch": 0.9028,
+      "grad_norm": 0.271484375,
+      "learning_rate": 5.700946410713548e-06,
+      "loss": 1.283,
+      "step": 45140
+    },
+    {
+      "epoch": 0.9029,
+      "grad_norm": 0.31640625,
+      "learning_rate": 5.6893345622424874e-06,
+      "loss": 1.2785,
+      "step": 45145
+    },
+    {
+      "epoch": 0.903,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.6777342052399045e-06,
+      "loss": 1.2791,
+      "step": 45150
+    },
+    {
+      "epoch": 0.9031,
+      "grad_norm": 0.310546875,
+      "learning_rate": 5.666145341119322e-06,
+      "loss": 1.261,
+      "step": 45155
+    },
+    {
+      "epoch": 0.9032,
+      "grad_norm": 0.29296875,
+      "learning_rate": 5.654567971292757e-06,
+      "loss": 1.261,
+      "step": 45160
+    },
+    {
+      "epoch": 0.9033,
+      "grad_norm": 0.228515625,
+      "learning_rate": 5.643002097170924e-06,
+      "loss": 1.2709,
+      "step": 45165
+    },
+    {
+      "epoch": 0.9034,
+      "grad_norm": 0.490234375,
+      "learning_rate": 5.6314477201630745e-06,
+      "loss": 1.2789,
+      "step": 45170
+    },
+    {
+      "epoch": 0.9035,
+      "grad_norm": 0.255859375,
+      "learning_rate": 5.619904841677059e-06,
+      "loss": 1.2644,
+      "step": 45175
+    },
+    {
+      "epoch": 0.9036,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.608373463119354e-06,
+      "loss": 1.2735,
+      "step": 45180
+    },
+    {
+      "epoch": 0.9037,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 5.5968535858950345e-06,
+      "loss": 1.2799,
+      "step": 45185
+    },
+    {
+      "epoch": 0.9038,
+      "grad_norm": 0.318359375,
+      "learning_rate": 5.585345211407733e-06,
+      "loss": 1.2723,
+      "step": 45190
+    },
+    {
+      "epoch": 0.9039,
+      "grad_norm": 0.412109375,
+      "learning_rate": 5.573848341059739e-06,
+      "loss": 1.2927,
+      "step": 45195
+    },
+    {
+      "epoch": 0.904,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.562362976251901e-06,
+      "loss": 1.288,
+      "step": 45200
+    },
+    {
+      "epoch": 0.9041,
+      "grad_norm": 0.396484375,
+      "learning_rate": 5.550889118383673e-06,
+      "loss": 1.2711,
+      "step": 45205
+    },
+    {
+      "epoch": 0.9042,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.5394267688531066e-06,
+      "loss": 1.2737,
+      "step": 45210
+    },
+    {
+      "epoch": 0.9043,
+      "grad_norm": 0.365234375,
+      "learning_rate": 5.52797592905685e-06,
+      "loss": 1.2781,
+      "step": 45215
+    },
+    {
+      "epoch": 0.9044,
+      "grad_norm": 0.23046875,
+      "learning_rate": 5.516536600390188e-06,
+      "loss": 1.2786,
+      "step": 45220
+    },
+    {
+      "epoch": 0.9045,
+      "grad_norm": 0.28125,
+      "learning_rate": 5.505108784246926e-06,
+      "loss": 1.2634,
+      "step": 45225
+    },
+    {
+      "epoch": 0.9046,
+      "grad_norm": 0.5,
+      "learning_rate": 5.49369248201953e-06,
+      "loss": 1.2717,
+      "step": 45230
+    },
+    {
+      "epoch": 0.9047,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 5.482287695099031e-06,
+      "loss": 1.2605,
+      "step": 45235
+    },
+    {
+      "epoch": 0.9048,
+      "grad_norm": 0.275390625,
+      "learning_rate": 5.470894424875062e-06,
+      "loss": 1.2806,
+      "step": 45240
+    },
+    {
+      "epoch": 0.9049,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.4595126727359e-06,
+      "loss": 1.2782,
+      "step": 45245
+    },
+    {
+      "epoch": 0.905,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.448142440068316e-06,
+      "loss": 1.283,
+      "step": 45250
+    },
+    {
+      "epoch": 0.9051,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.436783728257788e-06,
+      "loss": 1.2958,
+      "step": 45255
+    },
+    {
+      "epoch": 0.9052,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 5.425436538688322e-06,
+      "loss": 1.2603,
+      "step": 45260
+    },
+    {
+      "epoch": 0.9053,
+      "grad_norm": 0.232421875,
+      "learning_rate": 5.414100872742534e-06,
+      "loss": 1.29,
+      "step": 45265
+    },
+    {
+      "epoch": 0.9054,
+      "grad_norm": 0.283203125,
+      "learning_rate": 5.402776731801662e-06,
+      "loss": 1.2658,
+      "step": 45270
+    },
+    {
+      "epoch": 0.9055,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.39146411724547e-06,
+      "loss": 1.2683,
+      "step": 45275
+    },
+    {
+      "epoch": 0.9056,
+      "grad_norm": 0.25390625,
+      "learning_rate": 5.380163030452412e-06,
+      "loss": 1.2673,
+      "step": 45280
+    },
+    {
+      "epoch": 0.9057,
+      "grad_norm": 0.8125,
+      "learning_rate": 5.368873472799474e-06,
+      "loss": 1.2796,
+      "step": 45285
+    },
+    {
+      "epoch": 0.9058,
+      "grad_norm": 0.1875,
+      "learning_rate": 5.357595445662267e-06,
+      "loss": 1.2819,
+      "step": 45290
+    },
+    {
+      "epoch": 0.9059,
+      "grad_norm": 0.189453125,
+      "learning_rate": 5.346328950414969e-06,
+      "loss": 1.2807,
+      "step": 45295
+    },
+    {
+      "epoch": 0.906,
+      "grad_norm": 0.283203125,
+      "learning_rate": 5.335073988430372e-06,
+      "loss": 1.2671,
+      "step": 45300
+    },
+    {
+      "epoch": 0.9061,
+      "grad_norm": 0.384765625,
+      "learning_rate": 5.3238305610798565e-06,
+      "loss": 1.2654,
+      "step": 45305
+    },
+    {
+      "epoch": 0.9062,
+      "grad_norm": 0.271484375,
+      "learning_rate": 5.312598669733404e-06,
+      "loss": 1.2575,
+      "step": 45310
+    },
+    {
+      "epoch": 0.9063,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.301378315759598e-06,
+      "loss": 1.2823,
+      "step": 45315
+    },
+    {
+      "epoch": 0.9064,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.290169500525577e-06,
+      "loss": 1.2824,
+      "step": 45320
+    },
+    {
+      "epoch": 0.9065,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.278972225397127e-06,
+      "loss": 1.2798,
+      "step": 45325
+    },
+    {
+      "epoch": 0.9066,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 5.267786491738569e-06,
+      "loss": 1.2619,
+      "step": 45330
+    },
+    {
+      "epoch": 0.9067,
+      "grad_norm": 0.51171875,
+      "learning_rate": 5.256612300912911e-06,
+      "loss": 1.2561,
+      "step": 45335
+    },
+    {
+      "epoch": 0.9068,
+      "grad_norm": 0.28515625,
+      "learning_rate": 5.245449654281631e-06,
+      "loss": 1.2663,
+      "step": 45340
+    },
+    {
+      "epoch": 0.9069,
+      "grad_norm": 0.279296875,
+      "learning_rate": 5.2342985532049084e-06,
+      "loss": 1.2673,
+      "step": 45345
+    },
+    {
+      "epoch": 0.907,
+      "grad_norm": 0.3828125,
+      "learning_rate": 5.223158999041444e-06,
+      "loss": 1.2803,
+      "step": 45350
+    },
+    {
+      "epoch": 0.9071,
+      "grad_norm": 0.2734375,
+      "learning_rate": 5.212030993148553e-06,
+      "loss": 1.2725,
+      "step": 45355
+    },
+    {
+      "epoch": 0.9072,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.200914536882185e-06,
+      "loss": 1.2812,
+      "step": 45360
+    },
+    {
+      "epoch": 0.9073,
+      "grad_norm": 0.51171875,
+      "learning_rate": 5.189809631596798e-06,
+      "loss": 1.2889,
+      "step": 45365
+    },
+    {
+      "epoch": 0.9074,
+      "grad_norm": 0.298828125,
+      "learning_rate": 5.178716278645535e-06,
+      "loss": 1.2687,
+      "step": 45370
+    },
+    {
+      "epoch": 0.9075,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.167634479380068e-06,
+      "loss": 1.2846,
+      "step": 45375
+    },
+    {
+      "epoch": 0.9076,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.1565642351506845e-06,
+      "loss": 1.2786,
+      "step": 45380
+    },
+    {
+      "epoch": 0.9077,
+      "grad_norm": 0.240234375,
+      "learning_rate": 5.145505547306251e-06,
+      "loss": 1.2744,
+      "step": 45385
+    },
+    {
+      "epoch": 0.9078,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.134458417194254e-06,
+      "loss": 1.2934,
+      "step": 45390
+    },
+    {
+      "epoch": 0.9079,
+      "grad_norm": 0.224609375,
+      "learning_rate": 5.1234228461607304e-06,
+      "loss": 1.2776,
+      "step": 45395
+    },
+    {
+      "epoch": 0.908,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 5.1123988355503475e-06,
+      "loss": 1.2762,
+      "step": 45400
+    },
+    {
+      "epoch": 0.9081,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.101386386706342e-06,
+      "loss": 1.2633,
+      "step": 45405
+    },
+    {
+      "epoch": 0.9082,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.0903855009705514e-06,
+      "loss": 1.266,
+      "step": 45410
+    },
+    {
+      "epoch": 0.9083,
+      "grad_norm": 0.498046875,
+      "learning_rate": 5.079396179683383e-06,
+      "loss": 1.2751,
+      "step": 45415
+    },
+    {
+      "epoch": 0.9084,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.068418424183874e-06,
+      "loss": 1.2793,
+      "step": 45420
+    },
+    {
+      "epoch": 0.9085,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.057452235809624e-06,
+      "loss": 1.2725,
+      "step": 45425
+    },
+    {
+      "epoch": 0.9086,
+      "grad_norm": 0.80859375,
+      "learning_rate": 5.046497615896806e-06,
+      "loss": 1.268,
+      "step": 45430
+    },
+    {
+      "epoch": 0.9087,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.035554565780265e-06,
+      "loss": 1.2644,
+      "step": 45435
+    },
+    {
+      "epoch": 0.9088,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.024623086793323e-06,
+      "loss": 1.2774,
+      "step": 45440
+    },
+    {
+      "epoch": 0.9089,
+      "grad_norm": 0.310546875,
+      "learning_rate": 5.013703180267959e-06,
+      "loss": 1.281,
+      "step": 45445
+    },
+    {
+      "epoch": 0.909,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.002794847534764e-06,
+      "loss": 1.2667,
+      "step": 45450
+    },
+    {
+      "epoch": 0.9091,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.991898089922819e-06,
+      "loss": 1.2728,
+      "step": 45455
+    },
+    {
+      "epoch": 0.9092,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.981012908759941e-06,
+      "loss": 1.278,
+      "step": 45460
+    },
+    {
+      "epoch": 0.9093,
+      "grad_norm": 0.310546875,
+      "learning_rate": 4.97013930537239e-06,
+      "loss": 1.2702,
+      "step": 45465
+    },
+    {
+      "epoch": 0.9094,
+      "grad_norm": 0.27734375,
+      "learning_rate": 4.959277281085129e-06,
+      "loss": 1.2789,
+      "step": 45470
+    },
+    {
+      "epoch": 0.9095,
+      "grad_norm": 0.19921875,
+      "learning_rate": 4.948426837221631e-06,
+      "loss": 1.2694,
+      "step": 45475
+    },
+    {
+      "epoch": 0.9096,
+      "grad_norm": 0.326171875,
+      "learning_rate": 4.937587975103996e-06,
+      "loss": 1.2617,
+      "step": 45480
+    },
+    {
+      "epoch": 0.9097,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.926760696052934e-06,
+      "loss": 1.2702,
+      "step": 45485
+    },
+    {
+      "epoch": 0.9098,
+      "grad_norm": 0.291015625,
+      "learning_rate": 4.915945001387667e-06,
+      "loss": 1.2739,
+      "step": 45490
+    },
+    {
+      "epoch": 0.9099,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.905140892426097e-06,
+      "loss": 1.2834,
+      "step": 45495
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 4.8943483704846475e-06,
+      "loss": 1.2743,
+      "step": 45500
+    },
+    {
+      "epoch": 0.9101,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.8835674368783666e-06,
+      "loss": 1.2797,
+      "step": 45505
+    },
+    {
+      "epoch": 0.9102,
+      "grad_norm": 0.259765625,
+      "learning_rate": 4.872798092920872e-06,
+      "loss": 1.289,
+      "step": 45510
+    },
+    {
+      "epoch": 0.9103,
+      "grad_norm": 0.2421875,
+      "learning_rate": 4.862040339924378e-06,
+      "loss": 1.2582,
+      "step": 45515
+    },
+    {
+      "epoch": 0.9104,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.8512941791996726e-06,
+      "loss": 1.2648,
+      "step": 45520
+    },
+    {
+      "epoch": 0.9105,
+      "grad_norm": 0.166015625,
+      "learning_rate": 4.840559612056183e-06,
+      "loss": 1.3005,
+      "step": 45525
+    },
+    {
+      "epoch": 0.9106,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 4.829836639801843e-06,
+      "loss": 1.2695,
+      "step": 45530
+    },
+    {
+      "epoch": 0.9107,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 4.819125263743229e-06,
+      "loss": 1.2684,
+      "step": 45535
+    },
+    {
+      "epoch": 0.9108,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.808425485185486e-06,
+      "loss": 1.2732,
+      "step": 45540
+    },
+    {
+      "epoch": 0.9109,
+      "grad_norm": 0.298828125,
+      "learning_rate": 4.797737305432337e-06,
+      "loss": 1.2846,
+      "step": 45545
+    },
+    {
+      "epoch": 0.911,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.7870607257861415e-06,
+      "loss": 1.2643,
+      "step": 45550
+    },
+    {
+      "epoch": 0.9111,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 4.776395747547757e-06,
+      "loss": 1.2846,
+      "step": 45555
+    },
+    {
+      "epoch": 0.9112,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.765742372016735e-06,
+      "loss": 1.283,
+      "step": 45560
+    },
+    {
+      "epoch": 0.9113,
+      "grad_norm": 0.357421875,
+      "learning_rate": 4.755100600491102e-06,
+      "loss": 1.2986,
+      "step": 45565
+    },
+    {
+      "epoch": 0.9114,
+      "grad_norm": 0.28515625,
+      "learning_rate": 4.744470434267567e-06,
+      "loss": 1.2573,
+      "step": 45570
+    },
+    {
+      "epoch": 0.9115,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 4.733851874641382e-06,
+      "loss": 1.2821,
+      "step": 45575
+    },
+    {
+      "epoch": 0.9116,
+      "grad_norm": 0.236328125,
+      "learning_rate": 4.723244922906356e-06,
+      "loss": 1.266,
+      "step": 45580
+    },
+    {
+      "epoch": 0.9117,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 4.712649580354933e-06,
+      "loss": 1.2782,
+      "step": 45585
+    },
+    {
+      "epoch": 0.9118,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.702065848278126e-06,
+      "loss": 1.2737,
+      "step": 45590
+    },
+    {
+      "epoch": 0.9119,
+      "grad_norm": 0.390625,
+      "learning_rate": 4.691493727965512e-06,
+      "loss": 1.2657,
+      "step": 45595
+    },
+    {
+      "epoch": 0.912,
+      "grad_norm": 0.240234375,
+      "learning_rate": 4.680933220705308e-06,
+      "loss": 1.2815,
+      "step": 45600
+    },
+    {
+      "epoch": 0.9121,
+      "grad_norm": 0.23046875,
+      "learning_rate": 4.670384327784239e-06,
+      "loss": 1.2528,
+      "step": 45605
+    },
+    {
+      "epoch": 0.9122,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.659847050487687e-06,
+      "loss": 1.2808,
+      "step": 45610
+    },
+    {
+      "epoch": 0.9123,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.64932139009957e-06,
+      "loss": 1.2744,
+      "step": 45615
+    },
+    {
+      "epoch": 0.9124,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 4.638807347902408e-06,
+      "loss": 1.2813,
+      "step": 45620
+    },
+    {
+      "epoch": 0.9125,
+      "grad_norm": 0.26953125,
+      "learning_rate": 4.628304925177318e-06,
+      "loss": 1.2862,
+      "step": 45625
+    },
+    {
+      "epoch": 0.9126,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.617814123203967e-06,
+      "loss": 1.2747,
+      "step": 45630
+    },
+    {
+      "epoch": 0.9127,
+      "grad_norm": 0.37109375,
+      "learning_rate": 4.607334943260655e-06,
+      "loss": 1.2785,
+      "step": 45635
+    },
+    {
+      "epoch": 0.9128,
+      "grad_norm": 0.201171875,
+      "learning_rate": 4.596867386624215e-06,
+      "loss": 1.2671,
+      "step": 45640
+    },
+    {
+      "epoch": 0.9129,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.586411454570083e-06,
+      "loss": 1.2723,
+      "step": 45645
+    },
+    {
+      "epoch": 0.913,
+      "grad_norm": 0.203125,
+      "learning_rate": 4.575967148372317e-06,
+      "loss": 1.2751,
+      "step": 45650
+    },
+    {
+      "epoch": 0.9131,
+      "grad_norm": 0.3828125,
+      "learning_rate": 4.5655344693034895e-06,
+      "loss": 1.2626,
+      "step": 45655
+    },
+    {
+      "epoch": 0.9132,
+      "grad_norm": 0.28125,
+      "learning_rate": 4.555113418634805e-06,
+      "loss": 1.2711,
+      "step": 45660
+    },
+    {
+      "epoch": 0.9133,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.544703997636046e-06,
+      "loss": 1.2832,
+      "step": 45665
+    },
+    {
+      "epoch": 0.9134,
+      "grad_norm": 0.25,
+      "learning_rate": 4.534306207575545e-06,
+      "loss": 1.2703,
+      "step": 45670
+    },
+    {
+      "epoch": 0.9135,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.523920049720265e-06,
+      "loss": 1.2906,
+      "step": 45675
+    },
+    {
+      "epoch": 0.9136,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.513545525335705e-06,
+      "loss": 1.2791,
+      "step": 45680
+    },
+    {
+      "epoch": 0.9137,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.5031826356859876e-06,
+      "loss": 1.2693,
+      "step": 45685
+    },
+    {
+      "epoch": 0.9138,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 4.492831382033791e-06,
+      "loss": 1.261,
+      "step": 45690
+    },
+    {
+      "epoch": 0.9139,
+      "grad_norm": 0.361328125,
+      "learning_rate": 4.482491765640395e-06,
+      "loss": 1.2816,
+      "step": 45695
+    },
+    {
+      "epoch": 0.914,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.4721637877656375e-06,
+      "loss": 1.2896,
+      "step": 45700
+    },
+    {
+      "epoch": 0.9141,
+      "grad_norm": 0.271484375,
+      "learning_rate": 4.461847449667955e-06,
+      "loss": 1.2848,
+      "step": 45705
+    },
+    {
+      "epoch": 0.9142,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.451542752604365e-06,
+      "loss": 1.2717,
+      "step": 45710
+    },
+    {
+      "epoch": 0.9143,
+      "grad_norm": 0.228515625,
+      "learning_rate": 4.4412496978304515e-06,
+      "loss": 1.2947,
+      "step": 45715
+    },
+    {
+      "epoch": 0.9144,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.4309682866004125e-06,
+      "loss": 1.2597,
+      "step": 45720
+    },
+    {
+      "epoch": 0.9145,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.420698520166988e-06,
+      "loss": 1.281,
+      "step": 45725
+    },
+    {
+      "epoch": 0.9146,
+      "grad_norm": 0.36328125,
+      "learning_rate": 4.410440399781534e-06,
+      "loss": 1.2729,
+      "step": 45730
+    },
+    {
+      "epoch": 0.9147,
+      "grad_norm": 0.322265625,
+      "learning_rate": 4.400193926693952e-06,
+      "loss": 1.2709,
+      "step": 45735
+    },
+    {
+      "epoch": 0.9148,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 4.389959102152774e-06,
+      "loss": 1.2796,
+      "step": 45740
+    },
+    {
+      "epoch": 0.9149,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 4.379735927405038e-06,
+      "loss": 1.2764,
+      "step": 45745
+    },
+    {
+      "epoch": 0.915,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.369524403696457e-06,
+      "loss": 1.273,
+      "step": 45750
+    },
+    {
+      "epoch": 0.9151,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.3593245322712474e-06,
+      "loss": 1.2698,
+      "step": 45755
+    },
+    {
+      "epoch": 0.9152,
+      "grad_norm": 0.259765625,
+      "learning_rate": 4.349136314372204e-06,
+      "loss": 1.2709,
+      "step": 45760
+    },
+    {
+      "epoch": 0.9153,
+      "grad_norm": 0.263671875,
+      "learning_rate": 4.338959751240801e-06,
+      "loss": 1.2861,
+      "step": 45765
+    },
+    {
+      "epoch": 0.9154,
+      "grad_norm": 0.306640625,
+      "learning_rate": 4.328794844116946e-06,
+      "loss": 1.2704,
+      "step": 45770
+    },
+    {
+      "epoch": 0.9155,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.318641594239259e-06,
+      "loss": 1.2573,
+      "step": 45775
+    },
+    {
+      "epoch": 0.9156,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.308500002844862e-06,
+      "loss": 1.2703,
+      "step": 45780
+    },
+    {
+      "epoch": 0.9157,
+      "grad_norm": 0.302734375,
+      "learning_rate": 4.2983700711694665e-06,
+      "loss": 1.2993,
+      "step": 45785
+    },
+    {
+      "epoch": 0.9158,
+      "grad_norm": 0.625,
+      "learning_rate": 4.288251800447385e-06,
+      "loss": 1.2638,
+      "step": 45790
+    },
+    {
+      "epoch": 0.9159,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 4.278145191911509e-06,
+      "loss": 1.2642,
+      "step": 45795
+    },
+    {
+      "epoch": 0.916,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.268050246793276e-06,
+      "loss": 1.2652,
+      "step": 45800
+    },
+    {
+      "epoch": 0.9161,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.257966966322735e-06,
+      "loss": 1.268,
+      "step": 45805
+    },
+    {
+      "epoch": 0.9162,
+      "grad_norm": 0.294921875,
+      "learning_rate": 4.247895351728504e-06,
+      "loss": 1.2861,
+      "step": 45810
+    },
+    {
+      "epoch": 0.9163,
+      "grad_norm": 0.287109375,
+      "learning_rate": 4.237835404237778e-06,
+      "loss": 1.2828,
+      "step": 45815
+    },
+    {
+      "epoch": 0.9164,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 4.227787125076332e-06,
+      "loss": 1.2728,
+      "step": 45820
+    },
+    {
+      "epoch": 0.9165,
+      "grad_norm": 0.384765625,
+      "learning_rate": 4.217750515468522e-06,
+      "loss": 1.2645,
+      "step": 45825
+    },
+    {
+      "epoch": 0.9166,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.207725576637256e-06,
+      "loss": 1.2719,
+      "step": 45830
+    },
+    {
+      "epoch": 0.9167,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.197712309804058e-06,
+      "loss": 1.2652,
+      "step": 45835
+    },
+    {
+      "epoch": 0.9168,
+      "grad_norm": 0.306640625,
+      "learning_rate": 4.187710716189042e-06,
+      "loss": 1.2722,
+      "step": 45840
+    },
+    {
+      "epoch": 0.9169,
+      "grad_norm": 0.138671875,
+      "learning_rate": 4.177720797010831e-06,
+      "loss": 1.2877,
+      "step": 45845
+    },
+    {
+      "epoch": 0.917,
+      "grad_norm": 0.158203125,
+      "learning_rate": 4.167742553486675e-06,
+      "loss": 1.2706,
+      "step": 45850
+    },
+    {
+      "epoch": 0.9171,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.1577759868324125e-06,
+      "loss": 1.2794,
+      "step": 45855
+    },
+    {
+      "epoch": 0.9172,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.147821098262405e-06,
+      "loss": 1.2756,
+      "step": 45860
+    },
+    {
+      "epoch": 0.9173,
+      "grad_norm": 0.390625,
+      "learning_rate": 4.137877888989672e-06,
+      "loss": 1.2722,
+      "step": 45865
+    },
+    {
+      "epoch": 0.9174,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 4.127946360225721e-06,
+      "loss": 1.2743,
+      "step": 45870
+    },
+    {
+      "epoch": 0.9175,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.118026513180695e-06,
+      "loss": 1.2649,
+      "step": 45875
+    },
+    {
+      "epoch": 0.9176,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.108118349063306e-06,
+      "loss": 1.2704,
+      "step": 45880
+    },
+    {
+      "epoch": 0.9177,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.09822186908082e-06,
+      "loss": 1.2786,
+      "step": 45885
+    },
+    {
+      "epoch": 0.9178,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 4.088337074439097e-06,
+      "loss": 1.2858,
+      "step": 45890
+    },
+    {
+      "epoch": 0.9179,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.078463966342571e-06,
+      "loss": 1.2726,
+      "step": 45895
+    },
+    {
+      "epoch": 0.918,
+      "grad_norm": 0.287109375,
+      "learning_rate": 4.068602545994249e-06,
+      "loss": 1.2725,
+      "step": 45900
+    },
+    {
+      "epoch": 0.9181,
+      "grad_norm": 0.205078125,
+      "learning_rate": 4.0587528145957235e-06,
+      "loss": 1.2735,
+      "step": 45905
+    },
+    {
+      "epoch": 0.9182,
+      "grad_norm": 0.20703125,
+      "learning_rate": 4.048914773347134e-06,
+      "loss": 1.2746,
+      "step": 45910
+    },
+    {
+      "epoch": 0.9183,
+      "grad_norm": 0.75,
+      "learning_rate": 4.039088423447235e-06,
+      "loss": 1.263,
+      "step": 45915
+    },
+    {
+      "epoch": 0.9184,
+      "grad_norm": 0.27734375,
+      "learning_rate": 4.029273766093333e-06,
+      "loss": 1.2958,
+      "step": 45920
+    },
+    {
+      "epoch": 0.9185,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.019470802481307e-06,
+      "loss": 1.2809,
+      "step": 45925
+    },
+    {
+      "epoch": 0.9186,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 4.009679533805633e-06,
+      "loss": 1.2821,
+      "step": 45930
+    },
+    {
+      "epoch": 0.9187,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.999899961259335e-06,
+      "loss": 1.27,
+      "step": 45935
+    },
+    {
+      "epoch": 0.9188,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 3.990132086034026e-06,
+      "loss": 1.276,
+      "step": 45940
+    },
+    {
+      "epoch": 0.9189,
+      "grad_norm": 0.32421875,
+      "learning_rate": 3.9803759093199e-06,
+      "loss": 1.2845,
+      "step": 45945
+    },
+    {
+      "epoch": 0.919,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.970631432305694e-06,
+      "loss": 1.282,
+      "step": 45950
+    },
+    {
+      "epoch": 0.9191,
+      "grad_norm": 0.375,
+      "learning_rate": 3.96089865617878e-06,
+      "loss": 1.2726,
+      "step": 45955
+    },
+    {
+      "epoch": 0.9192,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.951177582125021e-06,
+      "loss": 1.2839,
+      "step": 45960
+    },
+    {
+      "epoch": 0.9193,
+      "grad_norm": 0.3046875,
+      "learning_rate": 3.941468211328947e-06,
+      "loss": 1.2687,
+      "step": 45965
+    },
+    {
+      "epoch": 0.9194,
+      "grad_norm": 0.22265625,
+      "learning_rate": 3.931770544973601e-06,
+      "loss": 1.2607,
+      "step": 45970
+    },
+    {
+      "epoch": 0.9195,
+      "grad_norm": 0.263671875,
+      "learning_rate": 3.922084584240582e-06,
+      "loss": 1.2684,
+      "step": 45975
+    },
+    {
+      "epoch": 0.9196,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 3.912410330310156e-06,
+      "loss": 1.2913,
+      "step": 45980
+    },
+    {
+      "epoch": 0.9197,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.902747784361038e-06,
+      "loss": 1.281,
+      "step": 45985
+    },
+    {
+      "epoch": 0.9198,
+      "grad_norm": 0.185546875,
+      "learning_rate": 3.893096947570618e-06,
+      "loss": 1.2803,
+      "step": 45990
+    },
+    {
+      "epoch": 0.9199,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.883457821114811e-06,
+      "loss": 1.2729,
+      "step": 45995
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.279296875,
+      "learning_rate": 3.873830406168111e-06,
+      "loss": 1.2687,
+      "step": 46000
+    },
+    {
+      "epoch": 0.9201,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 3.864214703903601e-06,
+      "loss": 1.2778,
+      "step": 46005
+    },
+    {
+      "epoch": 0.9202,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 3.8546107154929235e-06,
+      "loss": 1.2577,
+      "step": 46010
+    },
+    {
+      "epoch": 0.9203,
+      "grad_norm": 0.216796875,
+      "learning_rate": 3.845018442106285e-06,
+      "loss": 1.2617,
+      "step": 46015
+    },
+    {
+      "epoch": 0.9204,
+      "grad_norm": 0.23828125,
+      "learning_rate": 3.835437884912474e-06,
+      "loss": 1.2764,
+      "step": 46020
+    },
+    {
+      "epoch": 0.9205,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.825869045078867e-06,
+      "loss": 1.2727,
+      "step": 46025
+    },
+    {
+      "epoch": 0.9206,
+      "grad_norm": 0.19921875,
+      "learning_rate": 3.816311923771387e-06,
+      "loss": 1.2887,
+      "step": 46030
+    },
+    {
+      "epoch": 0.9207,
+      "grad_norm": 0.306640625,
+      "learning_rate": 3.806766522154548e-06,
+      "loss": 1.291,
+      "step": 46035
+    },
+    {
+      "epoch": 0.9208,
+      "grad_norm": 0.26171875,
+      "learning_rate": 3.797232841391407e-06,
+      "loss": 1.2813,
+      "step": 46040
+    },
+    {
+      "epoch": 0.9209,
+      "grad_norm": 0.265625,
+      "learning_rate": 3.787710882643658e-06,
+      "loss": 1.2752,
+      "step": 46045
+    },
+    {
+      "epoch": 0.921,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.7782006470714616e-06,
+      "loss": 1.2639,
+      "step": 46050
+    },
+    {
+      "epoch": 0.9211,
+      "grad_norm": 0.177734375,
+      "learning_rate": 3.7687021358336683e-06,
+      "loss": 1.2867,
+      "step": 46055
+    },
+    {
+      "epoch": 0.9212,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 3.759215350087619e-06,
+      "loss": 1.2739,
+      "step": 46060
+    },
+    {
+      "epoch": 0.9213,
+      "grad_norm": 0.310546875,
+      "learning_rate": 3.7497402909892342e-06,
+      "loss": 1.2701,
+      "step": 46065
+    },
+    {
+      "epoch": 0.9214,
+      "grad_norm": 0.2421875,
+      "learning_rate": 3.7402769596930563e-06,
+      "loss": 1.2892,
+      "step": 46070
+    },
+    {
+      "epoch": 0.9215,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.730825357352119e-06,
+      "loss": 1.2759,
+      "step": 46075
+    },
+    {
+      "epoch": 0.9216,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.721385485118123e-06,
+      "loss": 1.2726,
+      "step": 46080
+    },
+    {
+      "epoch": 0.9217,
+      "grad_norm": 0.376953125,
+      "learning_rate": 3.711957344141237e-06,
+      "loss": 1.2673,
+      "step": 46085
+    },
+    {
+      "epoch": 0.9218,
+      "grad_norm": 0.3046875,
+      "learning_rate": 3.7025409355702976e-06,
+      "loss": 1.2602,
+      "step": 46090
+    },
+    {
+      "epoch": 0.9219,
+      "grad_norm": 0.3125,
+      "learning_rate": 3.693136260552632e-06,
+      "loss": 1.2649,
+      "step": 46095
+    },
+    {
+      "epoch": 0.922,
+      "grad_norm": 0.375,
+      "learning_rate": 3.68374332023419e-06,
+      "loss": 1.2977,
+      "step": 46100
+    },
+    {
+      "epoch": 0.9221,
+      "grad_norm": 0.294921875,
+      "learning_rate": 3.6743621157594555e-06,
+      "loss": 1.2651,
+      "step": 46105
+    },
+    {
+      "epoch": 0.9222,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 3.664992648271526e-06,
+      "loss": 1.287,
+      "step": 46110
+    },
+    {
+      "epoch": 0.9223,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.6556349189120097e-06,
+      "loss": 1.2781,
+      "step": 46115
+    },
+    {
+      "epoch": 0.9224,
+      "grad_norm": 0.216796875,
+      "learning_rate": 3.6462889288211512e-06,
+      "loss": 1.2713,
+      "step": 46120
+    },
+    {
+      "epoch": 0.9225,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.6369546791377052e-06,
+      "loss": 1.2948,
+      "step": 46125
+    },
+    {
+      "epoch": 0.9226,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.627632170999029e-06,
+      "loss": 1.2864,
+      "step": 46130
+    },
+    {
+      "epoch": 0.9227,
+      "grad_norm": 0.30078125,
+      "learning_rate": 3.6183214055410586e-06,
+      "loss": 1.2676,
+      "step": 46135
+    },
+    {
+      "epoch": 0.9228,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 3.6090223838982417e-06,
+      "loss": 1.2788,
+      "step": 46140
+    },
+    {
+      "epoch": 0.9229,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.5997351072036945e-06,
+      "loss": 1.2734,
+      "step": 46145
+    },
+    {
+      "epoch": 0.923,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.5904595765890005e-06,
+      "loss": 1.2708,
+      "step": 46150
+    },
+    {
+      "epoch": 0.9231,
+      "grad_norm": 0.203125,
+      "learning_rate": 3.5811957931843554e-06,
+      "loss": 1.2811,
+      "step": 46155
+    },
+    {
+      "epoch": 0.9232,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 3.5719437581185454e-06,
+      "loss": 1.2657,
+      "step": 46160
+    },
+    {
+      "epoch": 0.9233,
+      "grad_norm": 0.30078125,
+      "learning_rate": 3.562703472518869e-06,
+      "loss": 1.2649,
+      "step": 46165
+    },
+    {
+      "epoch": 0.9234,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 3.553474937511281e-06,
+      "loss": 1.2652,
+      "step": 46170
+    },
+    {
+      "epoch": 0.9235,
+      "grad_norm": 0.15625,
+      "learning_rate": 3.5442581542201923e-06,
+      "loss": 1.2846,
+      "step": 46175
+    },
+    {
+      "epoch": 0.9236,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 3.5350531237686724e-06,
+      "loss": 1.2727,
+      "step": 46180
+    },
+    {
+      "epoch": 0.9237,
+      "grad_norm": 0.326171875,
+      "learning_rate": 3.5258598472783233e-06,
+      "loss": 1.2745,
+      "step": 46185
+    },
+    {
+      "epoch": 0.9238,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.516678325869316e-06,
+      "loss": 1.2694,
+      "step": 46190
+    },
+    {
+      "epoch": 0.9239,
+      "grad_norm": 0.369140625,
+      "learning_rate": 3.5075085606604e-06,
+      "loss": 1.2681,
+      "step": 46195
+    },
+    {
+      "epoch": 0.924,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.4983505527688586e-06,
+      "loss": 1.287,
+      "step": 46200
+    },
+    {
+      "epoch": 0.9241,
+      "grad_norm": 0.3125,
+      "learning_rate": 3.489204303310578e-06,
+      "loss": 1.2721,
+      "step": 46205
+    },
+    {
+      "epoch": 0.9242,
+      "grad_norm": 0.291015625,
+      "learning_rate": 3.480069813400022e-06,
+      "loss": 1.2795,
+      "step": 46210
+    },
+    {
+      "epoch": 0.9243,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 3.470947084150167e-06,
+      "loss": 1.2689,
+      "step": 46215
+    },
+    {
+      "epoch": 0.9244,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.461836116672612e-06,
+      "loss": 1.267,
+      "step": 46220
+    },
+    {
+      "epoch": 0.9245,
+      "grad_norm": 0.37109375,
+      "learning_rate": 3.452736912077503e-06,
+      "loss": 1.2675,
+      "step": 46225
+    },
+    {
+      "epoch": 0.9246,
+      "grad_norm": 0.18359375,
+      "learning_rate": 3.443649471473531e-06,
+      "loss": 1.2754,
+      "step": 46230
+    },
+    {
+      "epoch": 0.9247,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.434573795967988e-06,
+      "loss": 1.288,
+      "step": 46235
+    },
+    {
+      "epoch": 0.9248,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 3.425509886666711e-06,
+      "loss": 1.2677,
+      "step": 46240
+    },
+    {
+      "epoch": 0.9249,
+      "grad_norm": 0.140625,
+      "learning_rate": 3.4164577446741174e-06,
+      "loss": 1.2691,
+      "step": 46245
+    },
+    {
+      "epoch": 0.925,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.40741737109318e-06,
+      "loss": 1.2666,
+      "step": 46250
+    },
+    {
+      "epoch": 0.9251,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.3983887670254177e-06,
+      "loss": 1.2606,
+      "step": 46255
+    },
+    {
+      "epoch": 0.9252,
+      "grad_norm": 0.265625,
+      "learning_rate": 3.389371933570995e-06,
+      "loss": 1.2858,
+      "step": 46260
+    },
+    {
+      "epoch": 0.9253,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.380366871828522e-06,
+      "loss": 1.2767,
+      "step": 46265
+    },
+    {
+      "epoch": 0.9254,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 3.3713735828952987e-06,
+      "loss": 1.2796,
+      "step": 46270
+    },
+    {
+      "epoch": 0.9255,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.3623920678670597e-06,
+      "loss": 1.2807,
+      "step": 46275
+    },
+    {
+      "epoch": 0.9256,
+      "grad_norm": 0.453125,
+      "learning_rate": 3.3534223278382405e-06,
+      "loss": 1.2786,
+      "step": 46280
+    },
+    {
+      "epoch": 0.9257,
+      "grad_norm": 0.23046875,
+      "learning_rate": 3.344464363901756e-06,
+      "loss": 1.272,
+      "step": 46285
+    },
+    {
+      "epoch": 0.9258,
+      "grad_norm": 0.2109375,
+      "learning_rate": 3.3355181771490772e-06,
+      "loss": 1.2694,
+      "step": 46290
+    },
+    {
+      "epoch": 0.9259,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.3265837686703106e-06,
+      "loss": 1.273,
+      "step": 46295
+    },
+    {
+      "epoch": 0.926,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.3176611395540626e-06,
+      "loss": 1.2757,
+      "step": 46300
+    },
+    {
+      "epoch": 0.9261,
+      "grad_norm": 0.158203125,
+      "learning_rate": 3.3087502908875413e-06,
+      "loss": 1.2683,
+      "step": 46305
+    },
+    {
+      "epoch": 0.9262,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.2998512237565005e-06,
+      "loss": 1.2724,
+      "step": 46310
+    },
+    {
+      "epoch": 0.9263,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.290963939245262e-06,
+      "loss": 1.2722,
+      "step": 46315
+    },
+    {
+      "epoch": 0.9264,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.2820884384367146e-06,
+      "loss": 1.2694,
+      "step": 46320
+    },
+    {
+      "epoch": 0.9265,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 3.273224722412327e-06,
+      "loss": 1.2783,
+      "step": 46325
+    },
+    {
+      "epoch": 0.9266,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.2643727922520906e-06,
+      "loss": 1.2764,
+      "step": 46330
+    },
+    {
+      "epoch": 0.9267,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.2555326490346095e-06,
+      "loss": 1.2789,
+      "step": 46335
+    },
+    {
+      "epoch": 0.9268,
+      "grad_norm": 0.181640625,
+      "learning_rate": 3.246704293837011e-06,
+      "loss": 1.2655,
+      "step": 46340
+    },
+    {
+      "epoch": 0.9269,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.2378877277350116e-06,
+      "loss": 1.2735,
+      "step": 46345
+    },
+    {
+      "epoch": 0.927,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 3.2290829518028862e-06,
+      "loss": 1.2829,
+      "step": 46350
+    },
+    {
+      "epoch": 0.9271,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 3.2202899671134546e-06,
+      "loss": 1.2877,
+      "step": 46355
+    },
+    {
+      "epoch": 0.9272,
+      "grad_norm": 0.326171875,
+      "learning_rate": 3.2115087747381366e-06,
+      "loss": 1.2759,
+      "step": 46360
+    },
+    {
+      "epoch": 0.9273,
+      "grad_norm": 0.2578125,
+      "learning_rate": 3.2027393757468773e-06,
+      "loss": 1.2756,
+      "step": 46365
+    },
+    {
+      "epoch": 0.9274,
+      "grad_norm": 0.26171875,
+      "learning_rate": 3.19398177120821e-06,
+      "loss": 1.2746,
+      "step": 46370
+    },
+    {
+      "epoch": 0.9275,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 3.1852359621892367e-06,
+      "loss": 1.2771,
+      "step": 46375
+    },
+    {
+      "epoch": 0.9276,
+      "grad_norm": 0.3203125,
+      "learning_rate": 3.1765019497555616e-06,
+      "loss": 1.265,
+      "step": 46380
+    },
+    {
+      "epoch": 0.9277,
+      "grad_norm": 0.38671875,
+      "learning_rate": 3.1677797349714544e-06,
+      "loss": 1.2697,
+      "step": 46385
+    },
+    {
+      "epoch": 0.9278,
+      "grad_norm": 0.27734375,
+      "learning_rate": 3.1590693188996323e-06,
+      "loss": 1.2846,
+      "step": 46390
+    },
+    {
+      "epoch": 0.9279,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.1503707026014906e-06,
+      "loss": 1.2846,
+      "step": 46395
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.1416838871368924e-06,
+      "loss": 1.268,
+      "step": 46400
+    },
+    {
+      "epoch": 0.9281,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.1330088735643025e-06,
+      "loss": 1.2574,
+      "step": 46405
+    },
+    {
+      "epoch": 0.9282,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.124345662940764e-06,
+      "loss": 1.266,
+      "step": 46410
+    },
+    {
+      "epoch": 0.9283,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 3.115694256321855e-06,
+      "loss": 1.269,
+      "step": 46415
+    },
+    {
+      "epoch": 0.9284,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.1070546547617095e-06,
+      "loss": 1.2757,
+      "step": 46420
+    },
+    {
+      "epoch": 0.9285,
+      "grad_norm": 0.625,
+      "learning_rate": 3.0984268593130528e-06,
+      "loss": 1.2777,
+      "step": 46425
+    },
+    {
+      "epoch": 0.9286,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.0898108710271434e-06,
+      "loss": 1.267,
+      "step": 46430
+    },
+    {
+      "epoch": 0.9287,
+      "grad_norm": 0.310546875,
+      "learning_rate": 3.081206690953831e-06,
+      "loss": 1.2741,
+      "step": 46435
+    },
+    {
+      "epoch": 0.9288,
+      "grad_norm": 0.296875,
+      "learning_rate": 3.072614320141487e-06,
+      "loss": 1.27,
+      "step": 46440
+    },
+    {
+      "epoch": 0.9289,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.064033759637064e-06,
+      "loss": 1.28,
+      "step": 46445
+    },
+    {
+      "epoch": 0.929,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.0554650104861136e-06,
+      "loss": 1.2701,
+      "step": 46450
+    },
+    {
+      "epoch": 0.9291,
+      "grad_norm": 0.279296875,
+      "learning_rate": 3.046908073732668e-06,
+      "loss": 1.2874,
+      "step": 46455
+    },
+    {
+      "epoch": 0.9292,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.0383629504194046e-06,
+      "loss": 1.2804,
+      "step": 46460
+    },
+    {
+      "epoch": 0.9293,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.0298296415874894e-06,
+      "loss": 1.2631,
+      "step": 46465
+    },
+    {
+      "epoch": 0.9294,
+      "grad_norm": 0.48046875,
+      "learning_rate": 3.0213081482766805e-06,
+      "loss": 1.2862,
+      "step": 46470
+    },
+    {
+      "epoch": 0.9295,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.012798471525324e-06,
+      "loss": 1.2793,
+      "step": 46475
+    },
+    {
+      "epoch": 0.9296,
+      "grad_norm": 0.271484375,
+      "learning_rate": 3.0043006123702697e-06,
+      "loss": 1.2723,
+      "step": 46480
+    },
+    {
+      "epoch": 0.9297,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.9958145718469777e-06,
+      "loss": 1.2807,
+      "step": 46485
+    },
+    {
+      "epoch": 0.9298,
+      "grad_norm": 0.25,
+      "learning_rate": 2.9873403509894203e-06,
+      "loss": 1.2622,
+      "step": 46490
+    },
+    {
+      "epoch": 0.9299,
+      "grad_norm": 0.271484375,
+      "learning_rate": 2.978877950830172e-06,
+      "loss": 1.2818,
+      "step": 46495
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 2.970427372400353e-06,
+      "loss": 1.2949,
+      "step": 46500
+    },
+    {
+      "epoch": 0.9301,
+      "grad_norm": 0.314453125,
+      "learning_rate": 2.9619886167296384e-06,
+      "loss": 1.2842,
+      "step": 46505
+    },
+    {
+      "epoch": 0.9302,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 2.953561684846262e-06,
+      "loss": 1.271,
+      "step": 46510
+    },
+    {
+      "epoch": 0.9303,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.9451465777770247e-06,
+      "loss": 1.2626,
+      "step": 46515
+    },
+    {
+      "epoch": 0.9304,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 2.936743296547273e-06,
+      "loss": 1.279,
+      "step": 46520
+    },
+    {
+      "epoch": 0.9305,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.928351842180921e-06,
+      "loss": 1.2771,
+      "step": 46525
+    },
+    {
+      "epoch": 0.9306,
+      "grad_norm": 0.3203125,
+      "learning_rate": 2.9199722157004616e-06,
+      "loss": 1.2767,
+      "step": 46530
+    },
+    {
+      "epoch": 0.9307,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.9116044181269007e-06,
+      "loss": 1.2683,
+      "step": 46535
+    },
+    {
+      "epoch": 0.9308,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.9032484504798452e-06,
+      "loss": 1.2664,
+      "step": 46540
+    },
+    {
+      "epoch": 0.9309,
+      "grad_norm": 0.361328125,
+      "learning_rate": 2.8949043137774355e-06,
+      "loss": 1.2719,
+      "step": 46545
+    },
+    {
+      "epoch": 0.931,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.8865720090364034e-06,
+      "loss": 1.262,
+      "step": 46550
+    },
+    {
+      "epoch": 0.9311,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.878251537271981e-06,
+      "loss": 1.2695,
+      "step": 46555
+    },
+    {
+      "epoch": 0.9312,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.8699428994980017e-06,
+      "loss": 1.266,
+      "step": 46560
+    },
+    {
+      "epoch": 0.9313,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.8616460967268667e-06,
+      "loss": 1.2771,
+      "step": 46565
+    },
+    {
+      "epoch": 0.9314,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.8533611299694783e-06,
+      "loss": 1.2654,
+      "step": 46570
+    },
+    {
+      "epoch": 0.9315,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.845088000235396e-06,
+      "loss": 1.2633,
+      "step": 46575
+    },
+    {
+      "epoch": 0.9316,
+      "grad_norm": 0.28125,
+      "learning_rate": 2.836826708532603e-06,
+      "loss": 1.269,
+      "step": 46580
+    },
+    {
+      "epoch": 0.9317,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.8285772558677705e-06,
+      "loss": 1.2779,
+      "step": 46585
+    },
+    {
+      "epoch": 0.9318,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.8203396432460506e-06,
+      "loss": 1.2697,
+      "step": 46590
+    },
+    {
+      "epoch": 0.9319,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.8121138716711404e-06,
+      "loss": 1.2767,
+      "step": 46595
+    },
+    {
+      "epoch": 0.932,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.8038999421453826e-06,
+      "loss": 1.2887,
+      "step": 46600
+    },
+    {
+      "epoch": 0.9321,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.7956978556695767e-06,
+      "loss": 1.2886,
+      "step": 46605
+    },
+    {
+      "epoch": 0.9322,
+      "grad_norm": 0.298828125,
+      "learning_rate": 2.7875076132431344e-06,
+      "loss": 1.2889,
+      "step": 46610
+    },
+    {
+      "epoch": 0.9323,
+      "grad_norm": 0.326171875,
+      "learning_rate": 2.7793292158640126e-06,
+      "loss": 1.2776,
+      "step": 46615
+    },
+    {
+      "epoch": 0.9324,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 2.771162664528726e-06,
+      "loss": 1.2586,
+      "step": 46620
+    },
+    {
+      "epoch": 0.9325,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.7630079602323442e-06,
+      "loss": 1.269,
+      "step": 46625
+    },
+    {
+      "epoch": 0.9326,
+      "grad_norm": 0.302734375,
+      "learning_rate": 2.7548651039684846e-06,
+      "loss": 1.2829,
+      "step": 46630
+    },
+    {
+      "epoch": 0.9327,
+      "grad_norm": 0.130859375,
+      "learning_rate": 2.746734096729342e-06,
+      "loss": 1.277,
+      "step": 46635
+    },
+    {
+      "epoch": 0.9328,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.738614939505646e-06,
+      "loss": 1.29,
+      "step": 46640
+    },
+    {
+      "epoch": 0.9329,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.7305076332867054e-06,
+      "loss": 1.2721,
+      "step": 46645
+    },
+    {
+      "epoch": 0.933,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.7224121790603517e-06,
+      "loss": 1.2641,
+      "step": 46650
+    },
+    {
+      "epoch": 0.9331,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.714328577812997e-06,
+      "loss": 1.2756,
+      "step": 46655
+    },
+    {
+      "epoch": 0.9332,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.7062568305295965e-06,
+      "loss": 1.2831,
+      "step": 46660
+    },
+    {
+      "epoch": 0.9333,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 2.6981969381936978e-06,
+      "loss": 1.2695,
+      "step": 46665
+    },
+    {
+      "epoch": 0.9334,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.690148901787337e-06,
+      "loss": 1.263,
+      "step": 46670
+    },
+    {
+      "epoch": 0.9335,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.6821127222911857e-06,
+      "loss": 1.2595,
+      "step": 46675
+    },
+    {
+      "epoch": 0.9336,
+      "grad_norm": 0.32421875,
+      "learning_rate": 2.6740884006843825e-06,
+      "loss": 1.2651,
+      "step": 46680
+    },
+    {
+      "epoch": 0.9337,
+      "grad_norm": 0.310546875,
+      "learning_rate": 2.66607593794469e-06,
+      "loss": 1.2772,
+      "step": 46685
+    },
+    {
+      "epoch": 0.9338,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.6580753350484046e-06,
+      "loss": 1.2714,
+      "step": 46690
+    },
+    {
+      "epoch": 0.9339,
+      "grad_norm": 0.2421875,
+      "learning_rate": 2.650086592970358e-06,
+      "loss": 1.2787,
+      "step": 46695
+    },
+    {
+      "epoch": 0.934,
+      "grad_norm": 0.373046875,
+      "learning_rate": 2.6421097126839712e-06,
+      "loss": 1.2879,
+      "step": 46700
+    },
+    {
+      "epoch": 0.9341,
+      "grad_norm": 0.265625,
+      "learning_rate": 2.6341446951612005e-06,
+      "loss": 1.2653,
+      "step": 46705
+    },
+    {
+      "epoch": 0.9342,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.6261915413725578e-06,
+      "loss": 1.2523,
+      "step": 46710
+    },
+    {
+      "epoch": 0.9343,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 2.618250252287113e-06,
+      "loss": 1.2708,
+      "step": 46715
+    },
+    {
+      "epoch": 0.9344,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.610320828872481e-06,
+      "loss": 1.2632,
+      "step": 46720
+    },
+    {
+      "epoch": 0.9345,
+      "grad_norm": 0.2734375,
+      "learning_rate": 2.6024032720948443e-06,
+      "loss": 1.2812,
+      "step": 46725
+    },
+    {
+      "epoch": 0.9346,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.59449758291892e-06,
+      "loss": 1.2771,
+      "step": 46730
+    },
+    {
+      "epoch": 0.9347,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.5866037623080153e-06,
+      "loss": 1.2741,
+      "step": 46735
+    },
+    {
+      "epoch": 0.9348,
+      "grad_norm": 0.203125,
+      "learning_rate": 2.5787218112239496e-06,
+      "loss": 1.2812,
+      "step": 46740
+    },
+    {
+      "epoch": 0.9349,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.570851730627122e-06,
+      "loss": 1.2759,
+      "step": 46745
+    },
+    {
+      "epoch": 0.935,
+      "grad_norm": 0.296875,
+      "learning_rate": 2.5629935214764865e-06,
+      "loss": 1.2749,
+      "step": 46750
+    },
+    {
+      "epoch": 0.9351,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.5551471847295228e-06,
+      "loss": 1.2715,
+      "step": 46755
+    },
+    {
+      "epoch": 0.9352,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.5473127213422763e-06,
+      "loss": 1.2866,
+      "step": 46760
+    },
+    {
+      "epoch": 0.9353,
+      "grad_norm": 0.27734375,
+      "learning_rate": 2.5394901322694067e-06,
+      "loss": 1.2767,
+      "step": 46765
+    },
+    {
+      "epoch": 0.9354,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.531679418464006e-06,
+      "loss": 1.2758,
+      "step": 46770
+    },
+    {
+      "epoch": 0.9355,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.5238805808778242e-06,
+      "loss": 1.2739,
+      "step": 46775
+    },
+    {
+      "epoch": 0.9356,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.516093620461124e-06,
+      "loss": 1.2605,
+      "step": 46780
+    },
+    {
+      "epoch": 0.9357,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.508318538162702e-06,
+      "loss": 1.2616,
+      "step": 46785
+    },
+    {
+      "epoch": 0.9358,
+      "grad_norm": 0.275390625,
+      "learning_rate": 2.5005553349299547e-06,
+      "loss": 1.275,
+      "step": 46790
+    },
+    {
+      "epoch": 0.9359,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.4928040117087827e-06,
+      "loss": 1.2502,
+      "step": 46795
+    },
+    {
+      "epoch": 0.936,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.4850645694436736e-06,
+      "loss": 1.2586,
+      "step": 46800
+    },
+    {
+      "epoch": 0.9361,
+      "grad_norm": 0.13671875,
+      "learning_rate": 2.4773370090776626e-06,
+      "loss": 1.2824,
+      "step": 46805
+    },
+    {
+      "epoch": 0.9362,
+      "grad_norm": 0.384765625,
+      "learning_rate": 2.4696213315523074e-06,
+      "loss": 1.2761,
+      "step": 46810
+    },
+    {
+      "epoch": 0.9363,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.4619175378077565e-06,
+      "loss": 1.2799,
+      "step": 46815
+    },
+    {
+      "epoch": 0.9364,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.4542256287826914e-06,
+      "loss": 1.2732,
+      "step": 46820
+    },
+    {
+      "epoch": 0.9365,
+      "grad_norm": 0.486328125,
+      "learning_rate": 2.446545605414341e-06,
+      "loss": 1.2703,
+      "step": 46825
+    },
+    {
+      "epoch": 0.9366,
+      "grad_norm": 0.25,
+      "learning_rate": 2.4388774686385007e-06,
+      "loss": 1.2723,
+      "step": 46830
+    },
+    {
+      "epoch": 0.9367,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.4312212193895125e-06,
+      "loss": 1.2655,
+      "step": 46835
+    },
+    {
+      "epoch": 0.9368,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.4235768586002517e-06,
+      "loss": 1.2673,
+      "step": 46840
+    },
+    {
+      "epoch": 0.9369,
+      "grad_norm": 0.291015625,
+      "learning_rate": 2.415944387202174e-06,
+      "loss": 1.2855,
+      "step": 46845
+    },
+    {
+      "epoch": 0.937,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 2.4083238061252567e-06,
+      "loss": 1.2754,
+      "step": 46850
+    },
+    {
+      "epoch": 0.9371,
+      "grad_norm": 0.13671875,
+      "learning_rate": 2.40071511629808e-06,
+      "loss": 1.2749,
+      "step": 46855
+    },
+    {
+      "epoch": 0.9372,
+      "grad_norm": 0.23828125,
+      "learning_rate": 2.3931183186477026e-06,
+      "loss": 1.2711,
+      "step": 46860
+    },
+    {
+      "epoch": 0.9373,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 2.385533414099783e-06,
+      "loss": 1.2677,
+      "step": 46865
+    },
+    {
+      "epoch": 0.9374,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 2.3779604035785273e-06,
+      "loss": 1.2821,
+      "step": 46870
+    },
+    {
+      "epoch": 0.9375,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.3703992880066638e-06,
+      "loss": 1.2687,
+      "step": 46875
+    },
+    {
+      "epoch": 0.9376,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.3628500683055222e-06,
+      "loss": 1.2631,
+      "step": 46880
+    },
+    {
+      "epoch": 0.9377,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.355312745394922e-06,
+      "loss": 1.2734,
+      "step": 46885
+    },
+    {
+      "epoch": 0.9378,
+      "grad_norm": 0.44921875,
+      "learning_rate": 2.3477873201932734e-06,
+      "loss": 1.2792,
+      "step": 46890
+    },
+    {
+      "epoch": 0.9379,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.3402737936175425e-06,
+      "loss": 1.279,
+      "step": 46895
+    },
+    {
+      "epoch": 0.938,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.332772166583208e-06,
+      "loss": 1.277,
+      "step": 46900
+    },
+    {
+      "epoch": 0.9381,
+      "grad_norm": 0.314453125,
+      "learning_rate": 2.325282440004339e-06,
+      "loss": 1.2852,
+      "step": 46905
+    },
+    {
+      "epoch": 0.9382,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.3178046147935175e-06,
+      "loss": 1.2896,
+      "step": 46910
+    },
+    {
+      "epoch": 0.9383,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.3103386918619018e-06,
+      "loss": 1.2772,
+      "step": 46915
+    },
+    {
+      "epoch": 0.9384,
+      "grad_norm": 0.365234375,
+      "learning_rate": 2.3028846721191876e-06,
+      "loss": 1.3008,
+      "step": 46920
+    },
+    {
+      "epoch": 0.9385,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.295442556473637e-06,
+      "loss": 1.2568,
+      "step": 46925
+    },
+    {
+      "epoch": 0.9386,
+      "grad_norm": 0.287109375,
+      "learning_rate": 2.288012345832047e-06,
+      "loss": 1.2695,
+      "step": 46930
+    },
+    {
+      "epoch": 0.9387,
+      "grad_norm": 0.2109375,
+      "learning_rate": 2.2805940410997484e-06,
+      "loss": 1.275,
+      "step": 46935
+    },
+    {
+      "epoch": 0.9388,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.273187643180652e-06,
+      "loss": 1.2606,
+      "step": 46940
+    },
+    {
+      "epoch": 0.9389,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.2657931529772136e-06,
+      "loss": 1.2585,
+      "step": 46945
+    },
+    {
+      "epoch": 0.939,
+      "grad_norm": 0.224609375,
+      "learning_rate": 2.2584105713904125e-06,
+      "loss": 1.2692,
+      "step": 46950
+    },
+    {
+      "epoch": 0.9391,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.2510398993198067e-06,
+      "loss": 1.2666,
+      "step": 46955
+    },
+    {
+      "epoch": 0.9392,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.2436811376634893e-06,
+      "loss": 1.2822,
+      "step": 46960
+    },
+    {
+      "epoch": 0.9393,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.2363342873180757e-06,
+      "loss": 1.2654,
+      "step": 46965
+    },
+    {
+      "epoch": 0.9394,
+      "grad_norm": 0.3203125,
+      "learning_rate": 2.2289993491788064e-06,
+      "loss": 1.2628,
+      "step": 46970
+    },
+    {
+      "epoch": 0.9395,
+      "grad_norm": 0.36328125,
+      "learning_rate": 2.2216763241393767e-06,
+      "loss": 1.2687,
+      "step": 46975
+    },
+    {
+      "epoch": 0.9396,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.2143652130921176e-06,
+      "loss": 1.2905,
+      "step": 46980
+    },
+    {
+      "epoch": 0.9397,
+      "grad_norm": 0.326171875,
+      "learning_rate": 2.2070660169278166e-06,
+      "loss": 1.2706,
+      "step": 46985
+    },
+    {
+      "epoch": 0.9398,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.1997787365358958e-06,
+      "loss": 1.2798,
+      "step": 46990
+    },
+    {
+      "epoch": 0.9399,
+      "grad_norm": 0.283203125,
+      "learning_rate": 2.192503372804278e-06,
+      "loss": 1.2655,
+      "step": 46995
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2734375,
+      "learning_rate": 2.1852399266194314e-06,
+      "loss": 1.2895,
+      "step": 47000
+    },
+    {
+      "epoch": 0.9401,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.177988398866415e-06,
+      "loss": 1.2878,
+      "step": 47005
+    },
+    {
+      "epoch": 0.9402,
+      "grad_norm": 0.322265625,
+      "learning_rate": 2.1707487904287672e-06,
+      "loss": 1.2559,
+      "step": 47010
+    },
+    {
+      "epoch": 0.9403,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.163521102188648e-06,
+      "loss": 1.2643,
+      "step": 47015
+    },
+    {
+      "epoch": 0.9404,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.156305335026698e-06,
+      "loss": 1.2734,
+      "step": 47020
+    },
+    {
+      "epoch": 0.9405,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 2.1491014898221582e-06,
+      "loss": 1.2662,
+      "step": 47025
+    },
+    {
+      "epoch": 0.9406,
+      "grad_norm": 0.4609375,
+      "learning_rate": 2.141909567452793e-06,
+      "loss": 1.2768,
+      "step": 47030
+    },
+    {
+      "epoch": 0.9407,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.134729568794902e-06,
+      "loss": 1.2587,
+      "step": 47035
+    },
+    {
+      "epoch": 0.9408,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.1275614947233624e-06,
+      "loss": 1.2819,
+      "step": 47040
+    },
+    {
+      "epoch": 0.9409,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 2.120405346111576e-06,
+      "loss": 1.2738,
+      "step": 47045
+    },
+    {
+      "epoch": 0.941,
+      "grad_norm": 0.3203125,
+      "learning_rate": 2.1132611238315003e-06,
+      "loss": 1.2622,
+      "step": 47050
+    },
+    {
+      "epoch": 0.9411,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.1061288287536285e-06,
+      "loss": 1.2677,
+      "step": 47055
+    },
+    {
+      "epoch": 0.9412,
+      "grad_norm": 0.23828125,
+      "learning_rate": 2.0990084617470206e-06,
+      "loss": 1.2655,
+      "step": 47060
+    },
+    {
+      "epoch": 0.9413,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 2.0919000236792607e-06,
+      "loss": 1.2614,
+      "step": 47065
+    },
+    {
+      "epoch": 0.9414,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.084803515416511e-06,
+      "loss": 1.2565,
+      "step": 47070
+    },
+    {
+      "epoch": 0.9415,
+      "grad_norm": 0.443359375,
+      "learning_rate": 2.0777189378234143e-06,
+      "loss": 1.2741,
+      "step": 47075
+    },
+    {
+      "epoch": 0.9416,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 2.0706462917632673e-06,
+      "loss": 1.2713,
+      "step": 47080
+    },
+    {
+      "epoch": 0.9417,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.0635855780978044e-06,
+      "loss": 1.2636,
+      "step": 47085
+    },
+    {
+      "epoch": 0.9418,
+      "grad_norm": 0.2734375,
+      "learning_rate": 2.0565367976873584e-06,
+      "loss": 1.2603,
+      "step": 47090
+    },
+    {
+      "epoch": 0.9419,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.049499951390832e-06,
+      "loss": 1.2742,
+      "step": 47095
+    },
+    {
+      "epoch": 0.942,
+      "grad_norm": 0.22265625,
+      "learning_rate": 2.0424750400655947e-06,
+      "loss": 1.2795,
+      "step": 47100
+    },
+    {
+      "epoch": 0.9421,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.0354620645676504e-06,
+      "loss": 1.279,
+      "step": 47105
+    },
+    {
+      "epoch": 0.9422,
+      "grad_norm": 0.244140625,
+      "learning_rate": 2.0284610257514937e-06,
+      "loss": 1.2654,
+      "step": 47110
+    },
+    {
+      "epoch": 0.9423,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.021471924470175e-06,
+      "loss": 1.2645,
+      "step": 47115
+    },
+    {
+      "epoch": 0.9424,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.014494761575314e-06,
+      "loss": 1.2835,
+      "step": 47120
+    },
+    {
+      "epoch": 0.9425,
+      "grad_norm": 0.27734375,
+      "learning_rate": 2.0075295379170412e-06,
+      "loss": 1.2766,
+      "step": 47125
+    },
+    {
+      "epoch": 0.9426,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.0005762543440445e-06,
+      "loss": 1.2874,
+      "step": 47130
+    },
+    {
+      "epoch": 0.9427,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.993634911703579e-06,
+      "loss": 1.2842,
+      "step": 47135
+    },
+    {
+      "epoch": 0.9428,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.986705510841402e-06,
+      "loss": 1.2836,
+      "step": 47140
+    },
+    {
+      "epoch": 0.9429,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.979788052601861e-06,
+      "loss": 1.2772,
+      "step": 47145
+    },
+    {
+      "epoch": 0.943,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.9728825378278246e-06,
+      "loss": 1.2899,
+      "step": 47150
+    },
+    {
+      "epoch": 0.9431,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.965988967360688e-06,
+      "loss": 1.281,
+      "step": 47155
+    },
+    {
+      "epoch": 0.9432,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.9591073420404337e-06,
+      "loss": 1.269,
+      "step": 47160
+    },
+    {
+      "epoch": 0.9433,
+      "grad_norm": 0.2421875,
+      "learning_rate": 1.9522376627055583e-06,
+      "loss": 1.2765,
+      "step": 47165
+    },
+    {
+      "epoch": 0.9434,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.945379930193125e-06,
+      "loss": 1.282,
+      "step": 47170
+    },
+    {
+      "epoch": 0.9435,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9385341453386995e-06,
+      "loss": 1.2617,
+      "step": 47175
+    },
+    {
+      "epoch": 0.9436,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.931700308976436e-06,
+      "loss": 1.2721,
+      "step": 47180
+    },
+    {
+      "epoch": 0.9437,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.924878421939036e-06,
+      "loss": 1.2712,
+      "step": 47185
+    },
+    {
+      "epoch": 0.9438,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.918068485057689e-06,
+      "loss": 1.2754,
+      "step": 47190
+    },
+    {
+      "epoch": 0.9439,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.911270499162199e-06,
+      "loss": 1.285,
+      "step": 47195
+    },
+    {
+      "epoch": 0.944,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.904484465080847e-06,
+      "loss": 1.2731,
+      "step": 47200
+    },
+    {
+      "epoch": 0.9441,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.8977103836405053e-06,
+      "loss": 1.2714,
+      "step": 47205
+    },
+    {
+      "epoch": 0.9442,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8909482556666024e-06,
+      "loss": 1.2643,
+      "step": 47210
+    },
+    {
+      "epoch": 0.9443,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 1.8841980819830351e-06,
+      "loss": 1.274,
+      "step": 47215
+    },
+    {
+      "epoch": 0.9444,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.8774598634123232e-06,
+      "loss": 1.2753,
+      "step": 47220
+    },
+    {
+      "epoch": 0.9445,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.8707336007754873e-06,
+      "loss": 1.2733,
+      "step": 47225
+    },
+    {
+      "epoch": 0.9446,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8640192948921054e-06,
+      "loss": 1.2907,
+      "step": 47230
+    },
+    {
+      "epoch": 0.9447,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8573169465802898e-06,
+      "loss": 1.278,
+      "step": 47235
+    },
+    {
+      "epoch": 0.9448,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.8506265566567094e-06,
+      "loss": 1.2919,
+      "step": 47240
+    },
+    {
+      "epoch": 0.9449,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 1.8439481259365675e-06,
+      "loss": 1.2507,
+      "step": 47245
+    },
+    {
+      "epoch": 0.945,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 1.8372816552336026e-06,
+      "loss": 1.269,
+      "step": 47250
+    },
+    {
+      "epoch": 0.9451,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.8306271453601199e-06,
+      "loss": 1.2757,
+      "step": 47255
+    },
+    {
+      "epoch": 0.9452,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8239845971269266e-06,
+      "loss": 1.2711,
+      "step": 47260
+    },
+    {
+      "epoch": 0.9453,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8173540113434194e-06,
+      "loss": 1.2771,
+      "step": 47265
+    },
+    {
+      "epoch": 0.9454,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.8107353888175083e-06,
+      "loss": 1.2793,
+      "step": 47270
+    },
+    {
+      "epoch": 0.9455,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.8041287303556364e-06,
+      "loss": 1.2683,
+      "step": 47275
+    },
+    {
+      "epoch": 0.9456,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7975340367628268e-06,
+      "loss": 1.2699,
+      "step": 47280
+    },
+    {
+      "epoch": 0.9457,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.7909513088426255e-06,
+      "loss": 1.2792,
+      "step": 47285
+    },
+    {
+      "epoch": 0.9458,
+      "grad_norm": 0.240234375,
+      "learning_rate": 1.7843805473970798e-06,
+      "loss": 1.273,
+      "step": 47290
+    },
+    {
+      "epoch": 0.9459,
+      "grad_norm": 0.201171875,
+      "learning_rate": 1.7778217532268714e-06,
+      "loss": 1.2788,
+      "step": 47295
+    },
+    {
+      "epoch": 0.946,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 1.771274927131139e-06,
+      "loss": 1.287,
+      "step": 47300
+    },
+    {
+      "epoch": 0.9461,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7647400699075888e-06,
+      "loss": 1.2593,
+      "step": 47305
+    },
+    {
+      "epoch": 0.9462,
+      "grad_norm": 0.16015625,
+      "learning_rate": 1.7582171823524951e-06,
+      "loss": 1.278,
+      "step": 47310
+    },
+    {
+      "epoch": 0.9463,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.751706265260611e-06,
+      "loss": 1.2755,
+      "step": 47315
+    },
+    {
+      "epoch": 0.9464,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.7452073194253238e-06,
+      "loss": 1.2725,
+      "step": 47320
+    },
+    {
+      "epoch": 0.9465,
+      "grad_norm": 0.25,
+      "learning_rate": 1.7387203456384782e-06,
+      "loss": 1.2748,
+      "step": 47325
+    },
+    {
+      "epoch": 0.9466,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7322453446905085e-06,
+      "loss": 1.2583,
+      "step": 47330
+    },
+    {
+      "epoch": 0.9467,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 1.7257823173703503e-06,
+      "loss": 1.2766,
+      "step": 47335
+    },
+    {
+      "epoch": 0.9468,
+      "grad_norm": 0.23828125,
+      "learning_rate": 1.719331264465529e-06,
+      "loss": 1.2651,
+      "step": 47340
+    },
+    {
+      "epoch": 0.9469,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.712892186762083e-06,
+      "loss": 1.273,
+      "step": 47345
+    },
+    {
+      "epoch": 0.947,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 1.706465085044584e-06,
+      "loss": 1.2713,
+      "step": 47350
+    },
+    {
+      "epoch": 0.9471,
+      "grad_norm": 0.154296875,
+      "learning_rate": 1.7000499600961505e-06,
+      "loss": 1.2897,
+      "step": 47355
+    },
+    {
+      "epoch": 0.9472,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.6936468126984572e-06,
+      "loss": 1.2656,
+      "step": 47360
+    },
+    {
+      "epoch": 0.9473,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.6872556436317022e-06,
+      "loss": 1.277,
+      "step": 47365
+    },
+    {
+      "epoch": 0.9474,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.680876453674629e-06,
+      "loss": 1.2911,
+      "step": 47370
+    },
+    {
+      "epoch": 0.9475,
+      "grad_norm": 0.23828125,
+      "learning_rate": 1.6745092436045494e-06,
+      "loss": 1.2622,
+      "step": 47375
+    },
+    {
+      "epoch": 0.9476,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6681540141972429e-06,
+      "loss": 1.2731,
+      "step": 47380
+    },
+    {
+      "epoch": 0.9477,
+      "grad_norm": 0.22265625,
+      "learning_rate": 1.661810766227112e-06,
+      "loss": 1.2723,
+      "step": 47385
+    },
+    {
+      "epoch": 0.9478,
+      "grad_norm": 0.158203125,
+      "learning_rate": 1.6554795004670388e-06,
+      "loss": 1.2683,
+      "step": 47390
+    },
+    {
+      "epoch": 0.9479,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 1.6491602176884724e-06,
+      "loss": 1.2772,
+      "step": 47395
+    },
+    {
+      "epoch": 0.948,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.6428529186614195e-06,
+      "loss": 1.2704,
+      "step": 47400
+    },
+    {
+      "epoch": 0.9481,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.636557604154365e-06,
+      "loss": 1.2842,
+      "step": 47405
+    },
+    {
+      "epoch": 0.9482,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.6302742749344291e-06,
+      "loss": 1.2649,
+      "step": 47410
+    },
+    {
+      "epoch": 0.9483,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6240029317671658e-06,
+      "loss": 1.282,
+      "step": 47415
+    },
+    {
+      "epoch": 0.9484,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6177435754167415e-06,
+      "loss": 1.2816,
+      "step": 47420
+    },
+    {
+      "epoch": 0.9485,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.611496206645835e-06,
+      "loss": 1.2619,
+      "step": 47425
+    },
+    {
+      "epoch": 0.9486,
+      "grad_norm": 0.234375,
+      "learning_rate": 1.605260826215682e-06,
+      "loss": 1.2679,
+      "step": 47430
+    },
+    {
+      "epoch": 0.9487,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.5990374348860305e-06,
+      "loss": 1.2831,
+      "step": 47435
+    },
+    {
+      "epoch": 0.9488,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.5928260334151845e-06,
+      "loss": 1.2701,
+      "step": 47440
+    },
+    {
+      "epoch": 0.9489,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.5866266225599834e-06,
+      "loss": 1.2704,
+      "step": 47445
+    },
+    {
+      "epoch": 0.949,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.580439203075812e-06,
+      "loss": 1.2906,
+      "step": 47450
+    },
+    {
+      "epoch": 0.9491,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.574263775716578e-06,
+      "loss": 1.2816,
+      "step": 47455
+    },
+    {
+      "epoch": 0.9492,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.5681003412347572e-06,
+      "loss": 1.2772,
+      "step": 47460
+    },
+    {
+      "epoch": 0.9493,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.561948900381327e-06,
+      "loss": 1.2671,
+      "step": 47465
+    },
+    {
+      "epoch": 0.9494,
+      "grad_norm": 0.24609375,
+      "learning_rate": 1.555809453905821e-06,
+      "loss": 1.2658,
+      "step": 47470
+    },
+    {
+      "epoch": 0.9495,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.5496820025563409e-06,
+      "loss": 1.2835,
+      "step": 47475
+    },
+    {
+      "epoch": 0.9496,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.543566547079467e-06,
+      "loss": 1.2803,
+      "step": 47480
+    },
+    {
+      "epoch": 0.9497,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.5374630882203588e-06,
+      "loss": 1.2503,
+      "step": 47485
+    },
+    {
+      "epoch": 0.9498,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.5313716267226996e-06,
+      "loss": 1.3024,
+      "step": 47490
+    },
+    {
+      "epoch": 0.9499,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.5252921633287177e-06,
+      "loss": 1.2774,
+      "step": 47495
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 1.5192246987791981e-06,
+      "loss": 1.2676,
+      "step": 47500
+    },
+    {
+      "epoch": 0.9501,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.5131692338134052e-06,
+      "loss": 1.2692,
+      "step": 47505
+    },
+    {
+      "epoch": 0.9502,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.5071257691692153e-06,
+      "loss": 1.2862,
+      "step": 47510
+    },
+    {
+      "epoch": 0.9503,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.501094305582984e-06,
+      "loss": 1.2892,
+      "step": 47515
+    },
+    {
+      "epoch": 0.9504,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.4950748437896234e-06,
+      "loss": 1.2778,
+      "step": 47520
+    },
+    {
+      "epoch": 0.9505,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.4890673845226133e-06,
+      "loss": 1.2671,
+      "step": 47525
+    },
+    {
+      "epoch": 0.9506,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.4830719285139127e-06,
+      "loss": 1.2686,
+      "step": 47530
+    },
+    {
+      "epoch": 0.9507,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.4770884764940706e-06,
+      "loss": 1.2759,
+      "step": 47535
+    },
+    {
+      "epoch": 0.9508,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.4711170291921484e-06,
+      "loss": 1.2679,
+      "step": 47540
+    },
+    {
+      "epoch": 0.9509,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.4651575873357416e-06,
+      "loss": 1.2697,
+      "step": 47545
+    },
+    {
+      "epoch": 0.951,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.4592101516509914e-06,
+      "loss": 1.2766,
+      "step": 47550
+    },
+    {
+      "epoch": 0.9511,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.4532747228625854e-06,
+      "loss": 1.2709,
+      "step": 47555
+    },
+    {
+      "epoch": 0.9512,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.4473513016937223e-06,
+      "loss": 1.2572,
+      "step": 47560
+    },
+    {
+      "epoch": 0.9513,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.4414398888661695e-06,
+      "loss": 1.274,
+      "step": 47565
+    },
+    {
+      "epoch": 0.9514,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.4355404851001952e-06,
+      "loss": 1.2651,
+      "step": 47570
+    },
+    {
+      "epoch": 0.9515,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 1.4296530911146466e-06,
+      "loss": 1.2708,
+      "step": 47575
+    },
+    {
+      "epoch": 0.9516,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.423777707626872e-06,
+      "loss": 1.2937,
+      "step": 47580
+    },
+    {
+      "epoch": 0.9517,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.4179143353527547e-06,
+      "loss": 1.2864,
+      "step": 47585
+    },
+    {
+      "epoch": 0.9518,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.412062975006767e-06,
+      "loss": 1.2776,
+      "step": 47590
+    },
+    {
+      "epoch": 0.9519,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.4062236273018392e-06,
+      "loss": 1.27,
+      "step": 47595
+    },
+    {
+      "epoch": 0.952,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.400396292949513e-06,
+      "loss": 1.2682,
+      "step": 47600
+    },
+    {
+      "epoch": 0.9521,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.394580972659798e-06,
+      "loss": 1.2741,
+      "step": 47605
+    },
+    {
+      "epoch": 0.9522,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.3887776671412943e-06,
+      "loss": 1.268,
+      "step": 47610
+    },
+    {
+      "epoch": 0.9523,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.3829863771011253e-06,
+      "loss": 1.2755,
+      "step": 47615
+    },
+    {
+      "epoch": 0.9524,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.377207103244904e-06,
+      "loss": 1.273,
+      "step": 47620
+    },
+    {
+      "epoch": 0.9525,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.3714398462768563e-06,
+      "loss": 1.2848,
+      "step": 47625
+    },
+    {
+      "epoch": 0.9526,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3656846068996976e-06,
+      "loss": 1.2739,
+      "step": 47630
+    },
+    {
+      "epoch": 0.9527,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 1.359941385814667e-06,
+      "loss": 1.2684,
+      "step": 47635
+    },
+    {
+      "epoch": 0.9528,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.3542101837215826e-06,
+      "loss": 1.2766,
+      "step": 47640
+    },
+    {
+      "epoch": 0.9529,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.3484910013187524e-06,
+      "loss": 1.2601,
+      "step": 47645
+    },
+    {
+      "epoch": 0.953,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3427838393030633e-06,
+      "loss": 1.2777,
+      "step": 47650
+    },
+    {
+      "epoch": 0.9531,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.3370886983698927e-06,
+      "loss": 1.3115,
+      "step": 47655
+    },
+    {
+      "epoch": 0.9532,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.3314055792131964e-06,
+      "loss": 1.2611,
+      "step": 47660
+    },
+    {
+      "epoch": 0.9533,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 1.3257344825254315e-06,
+      "loss": 1.2887,
+      "step": 47665
+    },
+    {
+      "epoch": 0.9534,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.320075408997612e-06,
+      "loss": 1.2689,
+      "step": 47670
+    },
+    {
+      "epoch": 0.9535,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.3144283593192752e-06,
+      "loss": 1.266,
+      "step": 47675
+    },
+    {
+      "epoch": 0.9536,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.308793334178493e-06,
+      "loss": 1.2868,
+      "step": 47680
+    },
+    {
+      "epoch": 0.9537,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.303170334261883e-06,
+      "loss": 1.259,
+      "step": 47685
+    },
+    {
+      "epoch": 0.9538,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.2975593602545965e-06,
+      "loss": 1.2813,
+      "step": 47690
+    },
+    {
+      "epoch": 0.9539,
+      "grad_norm": 0.228515625,
+      "learning_rate": 1.2919604128402874e-06,
+      "loss": 1.2656,
+      "step": 47695
+    },
+    {
+      "epoch": 0.954,
+      "grad_norm": 0.142578125,
+      "learning_rate": 1.2863734927012095e-06,
+      "loss": 1.2738,
+      "step": 47700
+    },
+    {
+      "epoch": 0.9541,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.280798600518085e-06,
+      "loss": 1.2708,
+      "step": 47705
+    },
+    {
+      "epoch": 0.9542,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.275235736970193e-06,
+      "loss": 1.2713,
+      "step": 47710
+    },
+    {
+      "epoch": 0.9543,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.2696849027353796e-06,
+      "loss": 1.2734,
+      "step": 47715
+    },
+    {
+      "epoch": 0.9544,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.26414609848996e-06,
+      "loss": 1.2636,
+      "step": 47720
+    },
+    {
+      "epoch": 0.9545,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.2586193249088608e-06,
+      "loss": 1.2817,
+      "step": 47725
+    },
+    {
+      "epoch": 0.9546,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.2531045826654653e-06,
+      "loss": 1.2806,
+      "step": 47730
+    },
+    {
+      "epoch": 0.9547,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.2476018724317585e-06,
+      "loss": 1.2795,
+      "step": 47735
+    },
+    {
+      "epoch": 0.9548,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.2421111948782149e-06,
+      "loss": 1.2747,
+      "step": 47740
+    },
+    {
+      "epoch": 0.9549,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.236632550673844e-06,
+      "loss": 1.2583,
+      "step": 47745
+    },
+    {
+      "epoch": 0.955,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.231165940486234e-06,
+      "loss": 1.2659,
+      "step": 47750
+    },
+    {
+      "epoch": 0.9551,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.225711364981441e-06,
+      "loss": 1.2814,
+      "step": 47755
+    },
+    {
+      "epoch": 0.9552,
+      "grad_norm": 0.18359375,
+      "learning_rate": 1.2202688248241112e-06,
+      "loss": 1.2633,
+      "step": 47760
+    },
+    {
+      "epoch": 0.9553,
+      "grad_norm": 0.193359375,
+      "learning_rate": 1.2148383206773916e-06,
+      "loss": 1.262,
+      "step": 47765
+    },
+    {
+      "epoch": 0.9554,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2094198532029755e-06,
+      "loss": 1.2783,
+      "step": 47770
+    },
+    {
+      "epoch": 0.9555,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.20401342306109e-06,
+      "loss": 1.2601,
+      "step": 47775
+    },
+    {
+      "epoch": 0.9556,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.1986190309104861e-06,
+      "loss": 1.2758,
+      "step": 47780
+    },
+    {
+      "epoch": 0.9557,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 1.193236677408449e-06,
+      "loss": 1.2695,
+      "step": 47785
+    },
+    {
+      "epoch": 0.9558,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.1878663632108322e-06,
+      "loss": 1.2729,
+      "step": 47790
+    },
+    {
+      "epoch": 0.9559,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.1825080889719563e-06,
+      "loss": 1.2806,
+      "step": 47795
+    },
+    {
+      "epoch": 0.956,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.1771618553447216e-06,
+      "loss": 1.2738,
+      "step": 47800
+    },
+    {
+      "epoch": 0.9561,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.1718276629805625e-06,
+      "loss": 1.273,
+      "step": 47805
+    },
+    {
+      "epoch": 0.9562,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.1665055125294033e-06,
+      "loss": 1.2817,
+      "step": 47810
+    },
+    {
+      "epoch": 0.9563,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.16119540463977e-06,
+      "loss": 1.269,
+      "step": 47815
+    },
+    {
+      "epoch": 0.9564,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.155897339958667e-06,
+      "loss": 1.2682,
+      "step": 47820
+    },
+    {
+      "epoch": 0.9565,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.1506113191316447e-06,
+      "loss": 1.289,
+      "step": 47825
+    },
+    {
+      "epoch": 0.9566,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.1453373428027992e-06,
+      "loss": 1.2705,
+      "step": 47830
+    },
+    {
+      "epoch": 0.9567,
+      "grad_norm": 0.2421875,
+      "learning_rate": 1.1400754116147271e-06,
+      "loss": 1.2643,
+      "step": 47835
+    },
+    {
+      "epoch": 0.9568,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.134825526208605e-06,
+      "loss": 1.2705,
+      "step": 47840
+    },
+    {
+      "epoch": 0.9569,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.1295876872240873e-06,
+      "loss": 1.2687,
+      "step": 47845
+    },
+    {
+      "epoch": 0.957,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.1243618952994195e-06,
+      "loss": 1.271,
+      "step": 47850
+    },
+    {
+      "epoch": 0.9571,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.1191481510713253e-06,
+      "loss": 1.2755,
+      "step": 47855
+    },
+    {
+      "epoch": 0.9572,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.1139464551750856e-06,
+      "loss": 1.2779,
+      "step": 47860
+    },
+    {
+      "epoch": 0.9573,
+      "grad_norm": 0.185546875,
+      "learning_rate": 1.1087568082445264e-06,
+      "loss": 1.2746,
+      "step": 47865
+    },
+    {
+      "epoch": 0.9574,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.103579210911976e-06,
+      "loss": 1.2579,
+      "step": 47870
+    },
+    {
+      "epoch": 0.9575,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.0984136638083177e-06,
+      "loss": 1.2608,
+      "step": 47875
+    },
+    {
+      "epoch": 0.9576,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.0932601675629595e-06,
+      "loss": 1.274,
+      "step": 47880
+    },
+    {
+      "epoch": 0.9577,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 1.0881187228038215e-06,
+      "loss": 1.2644,
+      "step": 47885
+    },
+    {
+      "epoch": 0.9578,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.0829893301573913e-06,
+      "loss": 1.292,
+      "step": 47890
+    },
+    {
+      "epoch": 0.9579,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.0778719902486689e-06,
+      "loss": 1.3094,
+      "step": 47895
+    },
+    {
+      "epoch": 0.958,
+      "grad_norm": 0.189453125,
+      "learning_rate": 1.0727667037011668e-06,
+      "loss": 1.2797,
+      "step": 47900
+    },
+    {
+      "epoch": 0.9581,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.0676734711369762e-06,
+      "loss": 1.2612,
+      "step": 47905
+    },
+    {
+      "epoch": 0.9582,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 1.0625922931766785e-06,
+      "loss": 1.2728,
+      "step": 47910
+    },
+    {
+      "epoch": 0.9583,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.0575231704393895e-06,
+      "loss": 1.2781,
+      "step": 47915
+    },
+    {
+      "epoch": 0.9584,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.052466103542793e-06,
+      "loss": 1.2838,
+      "step": 47920
+    },
+    {
+      "epoch": 0.9585,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 1.0474210931030514e-06,
+      "loss": 1.2802,
+      "step": 47925
+    },
+    {
+      "epoch": 0.9586,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.0423881397349068e-06,
+      "loss": 1.2797,
+      "step": 47930
+    },
+    {
+      "epoch": 0.9587,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 1.0373672440515902e-06,
+      "loss": 1.2592,
+      "step": 47935
+    },
+    {
+      "epoch": 0.9588,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.0323584066648795e-06,
+      "loss": 1.2851,
+      "step": 47940
+    },
+    {
+      "epoch": 0.9589,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.0273616281851083e-06,
+      "loss": 1.2636,
+      "step": 47945
+    },
+    {
+      "epoch": 0.959,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.0223769092211012e-06,
+      "loss": 1.2801,
+      "step": 47950
+    },
+    {
+      "epoch": 0.9591,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 1.0174042503802493e-06,
+      "loss": 1.2833,
+      "step": 47955
+    },
+    {
+      "epoch": 0.9592,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.0124436522684243e-06,
+      "loss": 1.2729,
+      "step": 47960
+    },
+    {
+      "epoch": 0.9593,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.007495115490087e-06,
+      "loss": 1.2816,
+      "step": 47965
+    },
+    {
+      "epoch": 0.9594,
+      "grad_norm": 0.189453125,
+      "learning_rate": 1.002558640648199e-06,
+      "loss": 1.2704,
+      "step": 47970
+    },
+    {
+      "epoch": 0.9595,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.976342283442463e-07,
+      "loss": 1.2767,
+      "step": 47975
+    },
+    {
+      "epoch": 0.9596,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.927218791782599e-07,
+      "loss": 1.2751,
+      "step": 47980
+    },
+    {
+      "epoch": 0.9597,
+      "grad_norm": 0.392578125,
+      "learning_rate": 9.878215937487834e-07,
+      "loss": 1.2757,
+      "step": 47985
+    },
+    {
+      "epoch": 0.9598,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.829333726529056e-07,
+      "loss": 1.2741,
+      "step": 47990
+    },
+    {
+      "epoch": 0.9599,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.78057216486261e-07,
+      "loss": 1.2723,
+      "step": 47995
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 9.731931258429638e-07,
+      "loss": 1.2731,
+      "step": 48000
+    },
+    {
+      "epoch": 0.9601,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.683411013157174e-07,
+      "loss": 1.2699,
+      "step": 48005
+    },
+    {
+      "epoch": 0.9602,
+      "grad_norm": 0.166015625,
+      "learning_rate": 9.635011434957152e-07,
+      "loss": 1.2788,
+      "step": 48010
+    },
+    {
+      "epoch": 0.9603,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.58673252972675e-07,
+      "loss": 1.2734,
+      "step": 48015
+    },
+    {
+      "epoch": 0.9604,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.538574303348813e-07,
+      "loss": 1.2655,
+      "step": 48020
+    },
+    {
+      "epoch": 0.9605,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.490536761691204e-07,
+      "loss": 1.278,
+      "step": 48025
+    },
+    {
+      "epoch": 0.9606,
+      "grad_norm": 0.365234375,
+      "learning_rate": 9.442619910607131e-07,
+      "loss": 1.2694,
+      "step": 48030
+    },
+    {
+      "epoch": 0.9607,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.394823755935145e-07,
+      "loss": 1.2701,
+      "step": 48035
+    },
+    {
+      "epoch": 0.9608,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 9.347148303499142e-07,
+      "loss": 1.2757,
+      "step": 48040
+    },
+    {
+      "epoch": 0.9609,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.299593559108033e-07,
+      "loss": 1.2997,
+      "step": 48045
+    },
+    {
+      "epoch": 0.961,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.252159528556403e-07,
+      "loss": 1.2652,
+      "step": 48050
+    },
+    {
+      "epoch": 0.9611,
+      "grad_norm": 0.248046875,
+      "learning_rate": 9.204846217623853e-07,
+      "loss": 1.2829,
+      "step": 48055
+    },
+    {
+      "epoch": 0.9612,
+      "grad_norm": 0.486328125,
+      "learning_rate": 9.157653632075436e-07,
+      "loss": 1.274,
+      "step": 48060
+    },
+    {
+      "epoch": 0.9613,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.110581777661331e-07,
+      "loss": 1.2564,
+      "step": 48065
+    },
+    {
+      "epoch": 0.9614,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 9.063630660117173e-07,
+      "loss": 1.2724,
+      "step": 48070
+    },
+    {
+      "epoch": 0.9615,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.016800285163718e-07,
+      "loss": 1.2673,
+      "step": 48075
+    },
+    {
+      "epoch": 0.9616,
+      "grad_norm": 0.458984375,
+      "learning_rate": 8.970090658507291e-07,
+      "loss": 1.2654,
+      "step": 48080
+    },
+    {
+      "epoch": 0.9617,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.923501785839117e-07,
+      "loss": 1.2916,
+      "step": 48085
+    },
+    {
+      "epoch": 0.9618,
+      "grad_norm": 0.357421875,
+      "learning_rate": 8.877033672835988e-07,
+      "loss": 1.2702,
+      "step": 48090
+    },
+    {
+      "epoch": 0.9619,
+      "grad_norm": 0.19921875,
+      "learning_rate": 8.830686325160042e-07,
+      "loss": 1.2864,
+      "step": 48095
+    },
+    {
+      "epoch": 0.962,
+      "grad_norm": 0.203125,
+      "learning_rate": 8.784459748458318e-07,
+      "loss": 1.2842,
+      "step": 48100
+    },
+    {
+      "epoch": 0.9621,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 8.73835394836342e-07,
+      "loss": 1.2819,
+      "step": 48105
+    },
+    {
+      "epoch": 0.9622,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.692368930493521e-07,
+      "loss": 1.2605,
+      "step": 48110
+    },
+    {
+      "epoch": 0.9623,
+      "grad_norm": 0.1796875,
+      "learning_rate": 8.646504700451252e-07,
+      "loss": 1.2617,
+      "step": 48115
+    },
+    {
+      "epoch": 0.9624,
+      "grad_norm": 0.365234375,
+      "learning_rate": 8.600761263825474e-07,
+      "loss": 1.2711,
+      "step": 48120
+    },
+    {
+      "epoch": 0.9625,
+      "grad_norm": 0.259765625,
+      "learning_rate": 8.555138626189618e-07,
+      "loss": 1.2666,
+      "step": 48125
+    },
+    {
+      "epoch": 0.9626,
+      "grad_norm": 0.283203125,
+      "learning_rate": 8.509636793102682e-07,
+      "loss": 1.266,
+      "step": 48130
+    },
+    {
+      "epoch": 0.9627,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 8.46425577010912e-07,
+      "loss": 1.261,
+      "step": 48135
+    },
+    {
+      "epoch": 0.9628,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.418995562738285e-07,
+      "loss": 1.2718,
+      "step": 48140
+    },
+    {
+      "epoch": 0.9629,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 8.373856176505101e-07,
+      "loss": 1.2739,
+      "step": 48145
+    },
+    {
+      "epoch": 0.963,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.328837616909613e-07,
+      "loss": 1.2836,
+      "step": 48150
+    },
+    {
+      "epoch": 0.9631,
+      "grad_norm": 0.326171875,
+      "learning_rate": 8.283939889437209e-07,
+      "loss": 1.2534,
+      "step": 48155
+    },
+    {
+      "epoch": 0.9632,
+      "grad_norm": 0.470703125,
+      "learning_rate": 8.239162999558403e-07,
+      "loss": 1.2687,
+      "step": 48160
+    },
+    {
+      "epoch": 0.9633,
+      "grad_norm": 0.306640625,
+      "learning_rate": 8.194506952729386e-07,
+      "loss": 1.2713,
+      "step": 48165
+    },
+    {
+      "epoch": 0.9634,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.14997175439125e-07,
+      "loss": 1.2841,
+      "step": 48170
+    },
+    {
+      "epoch": 0.9635,
+      "grad_norm": 0.3125,
+      "learning_rate": 8.105557409970432e-07,
+      "loss": 1.2769,
+      "step": 48175
+    },
+    {
+      "epoch": 0.9636,
+      "grad_norm": 0.373046875,
+      "learning_rate": 8.061263924878604e-07,
+      "loss": 1.2781,
+      "step": 48180
+    },
+    {
+      "epoch": 0.9637,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 8.017091304513003e-07,
+      "loss": 1.2708,
+      "step": 48185
+    },
+    {
+      "epoch": 0.9638,
+      "grad_norm": 0.296875,
+      "learning_rate": 7.973039554255768e-07,
+      "loss": 1.2782,
+      "step": 48190
+    },
+    {
+      "epoch": 0.9639,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 7.929108679474607e-07,
+      "loss": 1.29,
+      "step": 48195
+    },
+    {
+      "epoch": 0.964,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 7.885298685522235e-07,
+      "loss": 1.272,
+      "step": 48200
+    },
+    {
+      "epoch": 0.9641,
+      "grad_norm": 0.41015625,
+      "learning_rate": 7.841609577736719e-07,
+      "loss": 1.2624,
+      "step": 48205
+    },
+    {
+      "epoch": 0.9642,
+      "grad_norm": 0.275390625,
+      "learning_rate": 7.798041361441688e-07,
+      "loss": 1.2675,
+      "step": 48210
+    },
+    {
+      "epoch": 0.9643,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.754594041945562e-07,
+      "loss": 1.2827,
+      "step": 48215
+    },
+    {
+      "epoch": 0.9644,
+      "grad_norm": 0.390625,
+      "learning_rate": 7.711267624542329e-07,
+      "loss": 1.2765,
+      "step": 48220
+    },
+    {
+      "epoch": 0.9645,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 7.668062114511321e-07,
+      "loss": 1.2648,
+      "step": 48225
+    },
+    {
+      "epoch": 0.9646,
+      "grad_norm": 0.498046875,
+      "learning_rate": 7.624977517116772e-07,
+      "loss": 1.2645,
+      "step": 48230
+    },
+    {
+      "epoch": 0.9647,
+      "grad_norm": 0.3046875,
+      "learning_rate": 7.582013837608592e-07,
+      "loss": 1.273,
+      "step": 48235
+    },
+    {
+      "epoch": 0.9648,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 7.539171081221596e-07,
+      "loss": 1.2585,
+      "step": 48240
+    },
+    {
+      "epoch": 0.9649,
+      "grad_norm": 0.3046875,
+      "learning_rate": 7.496449253176274e-07,
+      "loss": 1.2697,
+      "step": 48245
+    },
+    {
+      "epoch": 0.965,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 7.453848358678017e-07,
+      "loss": 1.2647,
+      "step": 48250
+    },
+    {
+      "epoch": 0.9651,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.411368402917563e-07,
+      "loss": 1.259,
+      "step": 48255
+    },
+    {
+      "epoch": 0.9652,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.369009391070992e-07,
+      "loss": 1.2821,
+      "step": 48260
+    },
+    {
+      "epoch": 0.9653,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 7.326771328299731e-07,
+      "loss": 1.2725,
+      "step": 48265
+    },
+    {
+      "epoch": 0.9654,
+      "grad_norm": 0.232421875,
+      "learning_rate": 7.284654219750331e-07,
+      "loss": 1.2769,
+      "step": 48270
+    },
+    {
+      "epoch": 0.9655,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.242658070554464e-07,
+      "loss": 1.2749,
+      "step": 48275
+    },
+    {
+      "epoch": 0.9656,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.200782885829482e-07,
+      "loss": 1.2881,
+      "step": 48280
+    },
+    {
+      "epoch": 0.9657,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.159028670677526e-07,
+      "loss": 1.2849,
+      "step": 48285
+    },
+    {
+      "epoch": 0.9658,
+      "grad_norm": 0.212890625,
+      "learning_rate": 7.117395430186414e-07,
+      "loss": 1.2811,
+      "step": 48290
+    },
+    {
+      "epoch": 0.9659,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.075883169428754e-07,
+      "loss": 1.2606,
+      "step": 48295
+    },
+    {
+      "epoch": 0.966,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 7.034491893463058e-07,
+      "loss": 1.2574,
+      "step": 48300
+    },
+    {
+      "epoch": 0.9661,
+      "grad_norm": 0.240234375,
+      "learning_rate": 6.9932216073324e-07,
+      "loss": 1.2762,
+      "step": 48305
+    },
+    {
+      "epoch": 0.9662,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.952072316065761e-07,
+      "loss": 1.2877,
+      "step": 48310
+    },
+    {
+      "epoch": 0.9663,
+      "grad_norm": 0.353515625,
+      "learning_rate": 6.911044024676683e-07,
+      "loss": 1.272,
+      "step": 48315
+    },
+    {
+      "epoch": 0.9664,
+      "grad_norm": 0.478515625,
+      "learning_rate": 6.870136738164612e-07,
+      "loss": 1.2754,
+      "step": 48320
+    },
+    {
+      "epoch": 0.9665,
+      "grad_norm": 0.326171875,
+      "learning_rate": 6.829350461514006e-07,
+      "loss": 1.2701,
+      "step": 48325
+    },
+    {
+      "epoch": 0.9666,
+      "grad_norm": 0.37109375,
+      "learning_rate": 6.788685199694222e-07,
+      "loss": 1.2856,
+      "step": 48330
+    },
+    {
+      "epoch": 0.9667,
+      "grad_norm": 0.251953125,
+      "learning_rate": 6.748140957660631e-07,
+      "loss": 1.2874,
+      "step": 48335
+    },
+    {
+      "epoch": 0.9668,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 6.707717740353059e-07,
+      "loss": 1.2757,
+      "step": 48340
+    },
+    {
+      "epoch": 0.9669,
+      "grad_norm": 0.1640625,
+      "learning_rate": 6.66741555269712e-07,
+      "loss": 1.2723,
+      "step": 48345
+    },
+    {
+      "epoch": 0.967,
+      "grad_norm": 0.265625,
+      "learning_rate": 6.627234399603555e-07,
+      "loss": 1.2743,
+      "step": 48350
+    },
+    {
+      "epoch": 0.9671,
+      "grad_norm": 0.21484375,
+      "learning_rate": 6.587174285968223e-07,
+      "loss": 1.2812,
+      "step": 48355
+    },
+    {
+      "epoch": 0.9672,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.547235216672443e-07,
+      "loss": 1.2774,
+      "step": 48360
+    },
+    {
+      "epoch": 0.9673,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 6.507417196582544e-07,
+      "loss": 1.2735,
+      "step": 48365
+    },
+    {
+      "epoch": 0.9674,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 6.4677202305502e-07,
+      "loss": 1.2679,
+      "step": 48370
+    },
+    {
+      "epoch": 0.9675,
+      "grad_norm": 0.193359375,
+      "learning_rate": 6.428144323412544e-07,
+      "loss": 1.2702,
+      "step": 48375
+    },
+    {
+      "epoch": 0.9676,
+      "grad_norm": 0.259765625,
+      "learning_rate": 6.388689479991605e-07,
+      "loss": 1.2773,
+      "step": 48380
+    },
+    {
+      "epoch": 0.9677,
+      "grad_norm": 0.41796875,
+      "learning_rate": 6.349355705094984e-07,
+      "loss": 1.2784,
+      "step": 48385
+    },
+    {
+      "epoch": 0.9678,
+      "grad_norm": 0.416015625,
+      "learning_rate": 6.310143003515179e-07,
+      "loss": 1.2705,
+      "step": 48390
+    },
+    {
+      "epoch": 0.9679,
+      "grad_norm": 0.201171875,
+      "learning_rate": 6.271051380030368e-07,
+      "loss": 1.2705,
+      "step": 48395
+    },
+    {
+      "epoch": 0.968,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 6.232080839403631e-07,
+      "loss": 1.2698,
+      "step": 48400
+    },
+    {
+      "epoch": 0.9681,
+      "grad_norm": 0.46875,
+      "learning_rate": 6.193231386383391e-07,
+      "loss": 1.2819,
+      "step": 48405
+    },
+    {
+      "epoch": 0.9682,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 6.154503025703417e-07,
+      "loss": 1.2708,
+      "step": 48410
+    },
+    {
+      "epoch": 0.9683,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.115895762082602e-07,
+      "loss": 1.2692,
+      "step": 48415
+    },
+    {
+      "epoch": 0.9684,
+      "grad_norm": 0.28515625,
+      "learning_rate": 6.07740960022507e-07,
+      "loss": 1.263,
+      "step": 48420
+    },
+    {
+      "epoch": 0.9685,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.039044544820404e-07,
+      "loss": 1.265,
+      "step": 48425
+    },
+    {
+      "epoch": 0.9686,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 6.000800600542977e-07,
+      "loss": 1.285,
+      "step": 48430
+    },
+    {
+      "epoch": 0.9687,
+      "grad_norm": 0.48828125,
+      "learning_rate": 5.96267777205295e-07,
+      "loss": 1.2835,
+      "step": 48435
+    },
+    {
+      "epoch": 0.9688,
+      "grad_norm": 0.2734375,
+      "learning_rate": 5.924676063995382e-07,
+      "loss": 1.2793,
+      "step": 48440
+    },
+    {
+      "epoch": 0.9689,
+      "grad_norm": 0.212890625,
+      "learning_rate": 5.886795481000795e-07,
+      "loss": 1.2667,
+      "step": 48445
+    },
+    {
+      "epoch": 0.969,
+      "grad_norm": 0.462890625,
+      "learning_rate": 5.849036027684606e-07,
+      "loss": 1.2975,
+      "step": 48450
+    },
+    {
+      "epoch": 0.9691,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 5.811397708647803e-07,
+      "loss": 1.2763,
+      "step": 48455
+    },
+    {
+      "epoch": 0.9692,
+      "grad_norm": 0.26953125,
+      "learning_rate": 5.773880528476494e-07,
+      "loss": 1.2685,
+      "step": 48460
+    },
+    {
+      "epoch": 0.9693,
+      "grad_norm": 0.23046875,
+      "learning_rate": 5.736484491742134e-07,
+      "loss": 1.2714,
+      "step": 48465
+    },
+    {
+      "epoch": 0.9694,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 5.699209603001076e-07,
+      "loss": 1.2707,
+      "step": 48470
+    },
+    {
+      "epoch": 0.9695,
+      "grad_norm": 0.294921875,
+      "learning_rate": 5.662055866795357e-07,
+      "loss": 1.2672,
+      "step": 48475
+    },
+    {
+      "epoch": 0.9696,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.62502328765202e-07,
+      "loss": 1.2754,
+      "step": 48480
+    },
+    {
+      "epoch": 0.9697,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.588111870083346e-07,
+      "loss": 1.278,
+      "step": 48485
+    },
+    {
+      "epoch": 0.9698,
+      "grad_norm": 0.287109375,
+      "learning_rate": 5.551321618586736e-07,
+      "loss": 1.257,
+      "step": 48490
+    },
+    {
+      "epoch": 0.9699,
+      "grad_norm": 0.203125,
+      "learning_rate": 5.514652537645271e-07,
+      "loss": 1.2584,
+      "step": 48495
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.478104631726711e-07,
+      "loss": 1.2738,
+      "step": 48500
+    },
+    {
+      "epoch": 0.9701,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 5.441677905284381e-07,
+      "loss": 1.2619,
+      "step": 48505
+    },
+    {
+      "epoch": 0.9702,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 5.405372362756734e-07,
+      "loss": 1.2804,
+      "step": 48510
+    },
+    {
+      "epoch": 0.9703,
+      "grad_norm": 0.298828125,
+      "learning_rate": 5.369188008567672e-07,
+      "loss": 1.2655,
+      "step": 48515
+    },
+    {
+      "epoch": 0.9704,
+      "grad_norm": 0.314453125,
+      "learning_rate": 5.333124847125892e-07,
+      "loss": 1.2652,
+      "step": 48520
+    },
+    {
+      "epoch": 0.9705,
+      "grad_norm": 0.27734375,
+      "learning_rate": 5.297182882825879e-07,
+      "loss": 1.2722,
+      "step": 48525
+    },
+    {
+      "epoch": 0.9706,
+      "grad_norm": 0.259765625,
+      "learning_rate": 5.261362120046686e-07,
+      "loss": 1.2821,
+      "step": 48530
+    },
+    {
+      "epoch": 0.9707,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.225662563153266e-07,
+      "loss": 1.2659,
+      "step": 48535
+    },
+    {
+      "epoch": 0.9708,
+      "grad_norm": 0.466796875,
+      "learning_rate": 5.190084216495361e-07,
+      "loss": 1.2825,
+      "step": 48540
+    },
+    {
+      "epoch": 0.9709,
+      "grad_norm": 0.234375,
+      "learning_rate": 5.154627084408059e-07,
+      "loss": 1.2673,
+      "step": 48545
+    },
+    {
+      "epoch": 0.971,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 5.119291171211793e-07,
+      "loss": 1.2655,
+      "step": 48550
+    },
+    {
+      "epoch": 0.9711,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 5.084076481212119e-07,
+      "loss": 1.2702,
+      "step": 48555
+    },
+    {
+      "epoch": 0.9712,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 5.048983018699827e-07,
+      "loss": 1.2771,
+      "step": 48560
+    },
+    {
+      "epoch": 0.9713,
+      "grad_norm": 0.208984375,
+      "learning_rate": 5.01401078795094e-07,
+      "loss": 1.2878,
+      "step": 48565
+    },
+    {
+      "epoch": 0.9714,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.979159793226718e-07,
+      "loss": 1.2736,
+      "step": 48570
+    },
+    {
+      "epoch": 0.9715,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.944430038773762e-07,
+      "loss": 1.2809,
+      "step": 48575
+    },
+    {
+      "epoch": 0.9716,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.909821528823577e-07,
+      "loss": 1.2665,
+      "step": 48580
+    },
+    {
+      "epoch": 0.9717,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.875334267593234e-07,
+      "loss": 1.2697,
+      "step": 48585
+    },
+    {
+      "epoch": 0.9718,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.840968259284817e-07,
+      "loss": 1.2741,
+      "step": 48590
+    },
+    {
+      "epoch": 0.9719,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.806723508085864e-07,
+      "loss": 1.2689,
+      "step": 48595
+    },
+    {
+      "epoch": 0.972,
+      "grad_norm": 0.388671875,
+      "learning_rate": 4.772600018168816e-07,
+      "loss": 1.2851,
+      "step": 48600
+    },
+    {
+      "epoch": 0.9721,
+      "grad_norm": 0.2421875,
+      "learning_rate": 4.738597793691679e-07,
+      "loss": 1.2879,
+      "step": 48605
+    },
+    {
+      "epoch": 0.9722,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.704716838797363e-07,
+      "loss": 1.2632,
+      "step": 48610
+    },
+    {
+      "epoch": 0.9723,
+      "grad_norm": 0.32421875,
+      "learning_rate": 4.670957157614453e-07,
+      "loss": 1.2747,
+      "step": 48615
+    },
+    {
+      "epoch": 0.9724,
+      "grad_norm": 0.29296875,
+      "learning_rate": 4.6373187542561035e-07,
+      "loss": 1.2666,
+      "step": 48620
+    },
+    {
+      "epoch": 0.9725,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.6038016328211476e-07,
+      "loss": 1.2711,
+      "step": 48625
+    },
+    {
+      "epoch": 0.9726,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.570405797393762e-07,
+      "loss": 1.2698,
+      "step": 48630
+    },
+    {
+      "epoch": 0.9727,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.5371312520429144e-07,
+      "loss": 1.2734,
+      "step": 48635
+    },
+    {
+      "epoch": 0.9728,
+      "grad_norm": 0.1796875,
+      "learning_rate": 4.503978000823028e-07,
+      "loss": 1.278,
+      "step": 48640
+    },
+    {
+      "epoch": 0.9729,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 4.4709460477737607e-07,
+      "loss": 1.2688,
+      "step": 48645
+    },
+    {
+      "epoch": 0.973,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.438035396920004e-07,
+      "loss": 1.2732,
+      "step": 48650
+    },
+    {
+      "epoch": 0.9731,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 4.405246052271772e-07,
+      "loss": 1.2721,
+      "step": 48655
+    },
+    {
+      "epoch": 0.9732,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 4.372578017824314e-07,
+      "loss": 1.3022,
+      "step": 48660
+    },
+    {
+      "epoch": 0.9733,
+      "grad_norm": 0.2578125,
+      "learning_rate": 4.3400312975581117e-07,
+      "loss": 1.2604,
+      "step": 48665
+    },
+    {
+      "epoch": 0.9734,
+      "grad_norm": 0.318359375,
+      "learning_rate": 4.307605895439104e-07,
+      "loss": 1.2581,
+      "step": 48670
+    },
+    {
+      "epoch": 0.9735,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.275301815417909e-07,
+      "loss": 1.2576,
+      "step": 48675
+    },
+    {
+      "epoch": 0.9736,
+      "grad_norm": 0.15625,
+      "learning_rate": 4.2431190614309335e-07,
+      "loss": 1.274,
+      "step": 48680
+    },
+    {
+      "epoch": 0.9737,
+      "grad_norm": 0.41015625,
+      "learning_rate": 4.2110576373993736e-07,
+      "loss": 1.288,
+      "step": 48685
+    },
+    {
+      "epoch": 0.9738,
+      "grad_norm": 0.375,
+      "learning_rate": 4.179117547229883e-07,
+      "loss": 1.2618,
+      "step": 48690
+    },
+    {
+      "epoch": 0.9739,
+      "grad_norm": 0.326171875,
+      "learning_rate": 4.1472987948143473e-07,
+      "loss": 1.2689,
+      "step": 48695
+    },
+    {
+      "epoch": 0.974,
+      "grad_norm": 0.294921875,
+      "learning_rate": 4.115601384029666e-07,
+      "loss": 1.2872,
+      "step": 48700
+    },
+    {
+      "epoch": 0.9741,
+      "grad_norm": 0.365234375,
+      "learning_rate": 4.084025318738083e-07,
+      "loss": 1.269,
+      "step": 48705
+    },
+    {
+      "epoch": 0.9742,
+      "grad_norm": 0.28125,
+      "learning_rate": 4.0525706027870756e-07,
+      "loss": 1.2718,
+      "step": 48710
+    },
+    {
+      "epoch": 0.9743,
+      "grad_norm": 0.298828125,
+      "learning_rate": 4.021237240009468e-07,
+      "loss": 1.2664,
+      "step": 48715
+    },
+    {
+      "epoch": 0.9744,
+      "grad_norm": 0.322265625,
+      "learning_rate": 3.9900252342228717e-07,
+      "loss": 1.2861,
+      "step": 48720
+    },
+    {
+      "epoch": 0.9745,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 3.958934589230467e-07,
+      "loss": 1.2683,
+      "step": 48725
+    },
+    {
+      "epoch": 0.9746,
+      "grad_norm": 0.27734375,
+      "learning_rate": 3.9279653088205584e-07,
+      "loss": 1.2997,
+      "step": 48730
+    },
+    {
+      "epoch": 0.9747,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.8971173967666807e-07,
+      "loss": 1.2761,
+      "step": 48735
+    },
+    {
+      "epoch": 0.9748,
+      "grad_norm": 0.4609375,
+      "learning_rate": 3.866390856827495e-07,
+      "loss": 1.271,
+      "step": 48740
+    },
+    {
+      "epoch": 0.9749,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.835785692747118e-07,
+      "loss": 1.2652,
+      "step": 48745
+    },
+    {
+      "epoch": 0.975,
+      "grad_norm": 0.2265625,
+      "learning_rate": 3.805301908254455e-07,
+      "loss": 1.29,
+      "step": 48750
+    },
+    {
+      "epoch": 0.9751,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 3.774939507063979e-07,
+      "loss": 1.2624,
+      "step": 48755
+    },
+    {
+      "epoch": 0.9752,
+      "grad_norm": 0.326171875,
+      "learning_rate": 3.744698492875398e-07,
+      "loss": 1.2758,
+      "step": 48760
+    },
+    {
+      "epoch": 0.9753,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.7145788693732086e-07,
+      "loss": 1.2609,
+      "step": 48765
+    },
+    {
+      "epoch": 0.9754,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 3.6845806402275863e-07,
+      "loss": 1.2693,
+      "step": 48770
+    },
+    {
+      "epoch": 0.9755,
+      "grad_norm": 0.173828125,
+      "learning_rate": 3.654703809093607e-07,
+      "loss": 1.2789,
+      "step": 48775
+    },
+    {
+      "epoch": 0.9756,
+      "grad_norm": 0.38671875,
+      "learning_rate": 3.6249483796116924e-07,
+      "loss": 1.2719,
+      "step": 48780
+    },
+    {
+      "epoch": 0.9757,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.595314355407609e-07,
+      "loss": 1.2718,
+      "step": 48785
+    },
+    {
+      "epoch": 0.9758,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 3.565801740092023e-07,
+      "loss": 1.2764,
+      "step": 48790
+    },
+    {
+      "epoch": 0.9759,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.536410537260948e-07,
+      "loss": 1.2645,
+      "step": 48795
+    },
+    {
+      "epoch": 0.976,
+      "grad_norm": 0.32421875,
+      "learning_rate": 3.50714075049563e-07,
+      "loss": 1.2793,
+      "step": 48800
+    },
+    {
+      "epoch": 0.9761,
+      "grad_norm": 0.2265625,
+      "learning_rate": 3.4779923833626606e-07,
+      "loss": 1.2778,
+      "step": 48805
+    },
+    {
+      "epoch": 0.9762,
+      "grad_norm": 0.3203125,
+      "learning_rate": 3.4489654394134205e-07,
+      "loss": 1.265,
+      "step": 48810
+    },
+    {
+      "epoch": 0.9763,
+      "grad_norm": 0.25390625,
+      "learning_rate": 3.4200599221848595e-07,
+      "loss": 1.2569,
+      "step": 48815
+    },
+    {
+      "epoch": 0.9764,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.3912758351991593e-07,
+      "loss": 1.2809,
+      "step": 48820
+    },
+    {
+      "epoch": 0.9765,
+      "grad_norm": 0.31640625,
+      "learning_rate": 3.362613181963403e-07,
+      "loss": 1.2718,
+      "step": 48825
+    },
+    {
+      "epoch": 0.9766,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.3340719659701313e-07,
+      "loss": 1.2839,
+      "step": 48830
+    },
+    {
+      "epoch": 0.9767,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.305652190696895e-07,
+      "loss": 1.2552,
+      "step": 48835
+    },
+    {
+      "epoch": 0.9768,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.277353859606813e-07,
+      "loss": 1.2866,
+      "step": 48840
+    },
+    {
+      "epoch": 0.9769,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.249176976147683e-07,
+      "loss": 1.2796,
+      "step": 48845
+    },
+    {
+      "epoch": 0.977,
+      "grad_norm": 0.37109375,
+      "learning_rate": 3.2211215437528694e-07,
+      "loss": 1.2611,
+      "step": 48850
+    },
+    {
+      "epoch": 0.9771,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.1931875658408604e-07,
+      "loss": 1.2702,
+      "step": 48855
+    },
+    {
+      "epoch": 0.9772,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 3.1653750458152666e-07,
+      "loss": 1.2697,
+      "step": 48860
+    },
+    {
+      "epoch": 0.9773,
+      "grad_norm": 0.236328125,
+      "learning_rate": 3.137683987065043e-07,
+      "loss": 1.2654,
+      "step": 48865
+    },
+    {
+      "epoch": 0.9774,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.1101143929641585e-07,
+      "loss": 1.2602,
+      "step": 48870
+    },
+    {
+      "epoch": 0.9775,
+      "grad_norm": 0.1875,
+      "learning_rate": 3.0826662668720364e-07,
+      "loss": 1.2682,
+      "step": 48875
+    },
+    {
+      "epoch": 0.9776,
+      "grad_norm": 0.2734375,
+      "learning_rate": 3.0553396121330013e-07,
+      "loss": 1.2738,
+      "step": 48880
+    },
+    {
+      "epoch": 0.9777,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.0281344320768347e-07,
+      "loss": 1.2659,
+      "step": 48885
+    },
+    {
+      "epoch": 0.9778,
+      "grad_norm": 0.16015625,
+      "learning_rate": 3.001050730018218e-07,
+      "loss": 1.2633,
+      "step": 48890
+    },
+    {
+      "epoch": 0.9779,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.974088509257511e-07,
+      "loss": 1.2574,
+      "step": 48895
+    },
+    {
+      "epoch": 0.978,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.947247773079753e-07,
+      "loss": 1.2665,
+      "step": 48900
+    },
+    {
+      "epoch": 0.9781,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.9205285247555505e-07,
+      "loss": 1.2764,
+      "step": 48905
+    },
+    {
+      "epoch": 0.9782,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 2.893930767540298e-07,
+      "loss": 1.2718,
+      "step": 48910
+    },
+    {
+      "epoch": 0.9783,
+      "grad_norm": 0.3125,
+      "learning_rate": 2.867454504675182e-07,
+      "loss": 1.2539,
+      "step": 48915
+    },
+    {
+      "epoch": 0.9784,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.841099739386066e-07,
+      "loss": 1.2699,
+      "step": 48920
+    },
+    {
+      "epoch": 0.9785,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.81486647488427e-07,
+      "loss": 1.2747,
+      "step": 48925
+    },
+    {
+      "epoch": 0.9786,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.7887547143662373e-07,
+      "loss": 1.2603,
+      "step": 48930
+    },
+    {
+      "epoch": 0.9787,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.762764461013423e-07,
+      "loss": 1.28,
+      "step": 48935
+    },
+    {
+      "epoch": 0.9788,
+      "grad_norm": 0.296875,
+      "learning_rate": 2.73689571799296e-07,
+      "loss": 1.2533,
+      "step": 48940
+    },
+    {
+      "epoch": 0.9789,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.7111484884567717e-07,
+      "loss": 1.2712,
+      "step": 48945
+    },
+    {
+      "epoch": 0.979,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.685522775541904e-07,
+      "loss": 1.2739,
+      "step": 48950
+    },
+    {
+      "epoch": 0.9791,
+      "grad_norm": 0.234375,
+      "learning_rate": 2.660018582370971e-07,
+      "loss": 1.2764,
+      "step": 48955
+    },
+    {
+      "epoch": 0.9792,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.6346359120514863e-07,
+      "loss": 1.2678,
+      "step": 48960
+    },
+    {
+      "epoch": 0.9793,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.609374767676309e-07,
+      "loss": 1.272,
+      "step": 48965
+    },
+    {
+      "epoch": 0.9794,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.584235152323422e-07,
+      "loss": 1.2764,
+      "step": 48970
+    },
+    {
+      "epoch": 0.9795,
+      "grad_norm": 0.25,
+      "learning_rate": 2.5592170690560414e-07,
+      "loss": 1.257,
+      "step": 48975
+    },
+    {
+      "epoch": 0.9796,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.534320520922506e-07,
+      "loss": 1.2661,
+      "step": 48980
+    },
+    {
+      "epoch": 0.9797,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.5095455109562795e-07,
+      "loss": 1.2852,
+      "step": 48985
+    },
+    {
+      "epoch": 0.9798,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.484892042176279e-07,
+      "loss": 1.272,
+      "step": 48990
+    },
+    {
+      "epoch": 0.9799,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.4603601175864356e-07,
+      "loss": 1.278,
+      "step": 48995
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.4359497401758024e-07,
+      "loss": 1.2729,
+      "step": 49000
+    },
+    {
+      "epoch": 0.9801,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.4116609129187786e-07,
+      "loss": 1.2696,
+      "step": 49005
+    },
+    {
+      "epoch": 0.9802,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.387493638774774e-07,
+      "loss": 1.2746,
+      "step": 49010
+    },
+    {
+      "epoch": 0.9803,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.363447920688655e-07,
+      "loss": 1.2905,
+      "step": 49015
+    },
+    {
+      "epoch": 0.9804,
+      "grad_norm": 0.291015625,
+      "learning_rate": 2.339523761590301e-07,
+      "loss": 1.2812,
+      "step": 49020
+    },
+    {
+      "epoch": 0.9805,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.315721164394713e-07,
+      "loss": 1.2779,
+      "step": 49025
+    },
+    {
+      "epoch": 0.9806,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.2920401320022378e-07,
+      "loss": 1.2685,
+      "step": 49030
+    },
+    {
+      "epoch": 0.9807,
+      "grad_norm": 0.287109375,
+      "learning_rate": 2.2684806672982338e-07,
+      "loss": 1.2762,
+      "step": 49035
+    },
+    {
+      "epoch": 0.9808,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.2450427731534053e-07,
+      "loss": 1.2615,
+      "step": 49040
+    },
+    {
+      "epoch": 0.9809,
+      "grad_norm": 0.498046875,
+      "learning_rate": 2.2217264524236892e-07,
+      "loss": 1.2689,
+      "step": 49045
+    },
+    {
+      "epoch": 0.981,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.1985317079500356e-07,
+      "loss": 1.2673,
+      "step": 49050
+    },
+    {
+      "epoch": 0.9811,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 2.175458542558517e-07,
+      "loss": 1.2664,
+      "step": 49055
+    },
+    {
+      "epoch": 0.9812,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 2.1525069590607737e-07,
+      "loss": 1.3007,
+      "step": 49060
+    },
+    {
+      "epoch": 0.9813,
+      "grad_norm": 0.44921875,
+      "learning_rate": 2.1296769602532352e-07,
+      "loss": 1.2683,
+      "step": 49065
+    },
+    {
+      "epoch": 0.9814,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.106968548917676e-07,
+      "loss": 1.2865,
+      "step": 49070
+    },
+    {
+      "epoch": 0.9815,
+      "grad_norm": 0.23828125,
+      "learning_rate": 2.0843817278209942e-07,
+      "loss": 1.2736,
+      "step": 49075
+    },
+    {
+      "epoch": 0.9816,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.0619164997155438e-07,
+      "loss": 1.2743,
+      "step": 49080
+    },
+    {
+      "epoch": 0.9817,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.0395728673383574e-07,
+      "loss": 1.2704,
+      "step": 49085
+    },
+    {
+      "epoch": 0.9818,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.017350833412146e-07,
+      "loss": 1.2988,
+      "step": 49090
+    },
+    {
+      "epoch": 0.9819,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 1.9952504006446325e-07,
+      "loss": 1.2649,
+      "step": 49095
+    },
+    {
+      "epoch": 0.982,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.973271571728441e-07,
+      "loss": 1.276,
+      "step": 49100
+    },
+    {
+      "epoch": 0.9821,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 1.9514143493417625e-07,
+      "loss": 1.2879,
+      "step": 49105
+    },
+    {
+      "epoch": 0.9822,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.9296787361480216e-07,
+      "loss": 1.2673,
+      "step": 49110
+    },
+    {
+      "epoch": 0.9823,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.908064734795323e-07,
+      "loss": 1.2655,
+      "step": 49115
+    },
+    {
+      "epoch": 0.9824,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.8865723479173368e-07,
+      "loss": 1.2659,
+      "step": 49120
+    },
+    {
+      "epoch": 0.9825,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.86520157813308e-07,
+      "loss": 1.2799,
+      "step": 49125
+    },
+    {
+      "epoch": 0.9826,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.8439524280462472e-07,
+      "loss": 1.2892,
+      "step": 49130
+    },
+    {
+      "epoch": 0.9827,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8228249002461006e-07,
+      "loss": 1.2861,
+      "step": 49135
+    },
+    {
+      "epoch": 0.9828,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 1.8018189973069143e-07,
+      "loss": 1.2747,
+      "step": 49140
+    },
+    {
+      "epoch": 0.9829,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7809347217881966e-07,
+      "loss": 1.2803,
+      "step": 49145
+    },
+    {
+      "epoch": 0.983,
+      "grad_norm": 0.240234375,
+      "learning_rate": 1.7601720762346897e-07,
+      "loss": 1.2721,
+      "step": 49150
+    },
+    {
+      "epoch": 0.9831,
+      "grad_norm": 0.19921875,
+      "learning_rate": 1.7395310631762585e-07,
+      "loss": 1.2719,
+      "step": 49155
+    },
+    {
+      "epoch": 0.9832,
+      "grad_norm": 0.2265625,
+      "learning_rate": 1.7190116851280026e-07,
+      "loss": 1.2783,
+      "step": 49160
+    },
+    {
+      "epoch": 0.9833,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.698613944589922e-07,
+      "loss": 1.2721,
+      "step": 49165
+    },
+    {
+      "epoch": 0.9834,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.678337844047695e-07,
+      "loss": 1.2672,
+      "step": 49170
+    },
+    {
+      "epoch": 0.9835,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.6581833859716788e-07,
+      "loss": 1.2673,
+      "step": 49175
+    },
+    {
+      "epoch": 0.9836,
+      "grad_norm": 0.23828125,
+      "learning_rate": 1.6381505728176872e-07,
+      "loss": 1.268,
+      "step": 49180
+    },
+    {
+      "epoch": 0.9837,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.618239407026767e-07,
+      "loss": 1.2814,
+      "step": 49185
+    },
+    {
+      "epoch": 0.9838,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.598449891024978e-07,
+      "loss": 1.2836,
+      "step": 49190
+    },
+    {
+      "epoch": 0.9839,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.578782027223502e-07,
+      "loss": 1.2735,
+      "step": 49195
+    },
+    {
+      "epoch": 0.984,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.5592358180189782e-07,
+      "loss": 1.2662,
+      "step": 49200
+    },
+    {
+      "epoch": 0.9841,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.5398112657929453e-07,
+      "loss": 1.2845,
+      "step": 49205
+    },
+    {
+      "epoch": 0.9842,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.520508372912288e-07,
+      "loss": 1.2782,
+      "step": 49210
+    },
+    {
+      "epoch": 0.9843,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5013271417290143e-07,
+      "loss": 1.2726,
+      "step": 49215
+    },
+    {
+      "epoch": 0.9844,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.4822675745801429e-07,
+      "loss": 1.2727,
+      "step": 49220
+    },
+    {
+      "epoch": 0.9845,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.4633296737882607e-07,
+      "loss": 1.2626,
+      "step": 49225
+    },
+    {
+      "epoch": 0.9846,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4445134416607442e-07,
+      "loss": 1.2564,
+      "step": 49230
+    },
+    {
+      "epoch": 0.9847,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.425818880490315e-07,
+      "loss": 1.2738,
+      "step": 49235
+    },
+    {
+      "epoch": 0.9848,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.4072459925548177e-07,
+      "loss": 1.2811,
+      "step": 49240
+    },
+    {
+      "epoch": 0.9849,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.3887947801173307e-07,
+      "loss": 1.2673,
+      "step": 49245
+    },
+    {
+      "epoch": 0.985,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.3704652454261668e-07,
+      "loss": 1.2732,
+      "step": 49250
+    },
+    {
+      "epoch": 0.9851,
+      "grad_norm": 0.189453125,
+      "learning_rate": 1.3522573907145398e-07,
+      "loss": 1.2763,
+      "step": 49255
+    },
+    {
+      "epoch": 0.9852,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.33417121820123e-07,
+      "loss": 1.2795,
+      "step": 49260
+    },
+    {
+      "epoch": 0.9853,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.3162067300898084e-07,
+      "loss": 1.2936,
+      "step": 49265
+    },
+    {
+      "epoch": 0.9854,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.2983639285693017e-07,
+      "loss": 1.2683,
+      "step": 49270
+    },
+    {
+      "epoch": 0.9855,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.2806428158138596e-07,
+      "loss": 1.2655,
+      "step": 49275
+    },
+    {
+      "epoch": 0.9856,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.2630433939825327e-07,
+      "loss": 1.269,
+      "step": 49280
+    },
+    {
+      "epoch": 0.9857,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 1.2455656652198277e-07,
+      "loss": 1.277,
+      "step": 49285
+    },
+    {
+      "epoch": 0.9858,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.2282096316554858e-07,
+      "loss": 1.2799,
+      "step": 49290
+    },
+    {
+      "epoch": 0.9859,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.2109752954042597e-07,
+      "loss": 1.276,
+      "step": 49295
+    },
+    {
+      "epoch": 0.986,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.193862658566025e-07,
+      "loss": 1.2641,
+      "step": 49300
+    },
+    {
+      "epoch": 0.9861,
+      "grad_norm": 0.75,
+      "learning_rate": 1.1768717232257809e-07,
+      "loss": 1.2824,
+      "step": 49305
+    },
+    {
+      "epoch": 0.9862,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.160002491454093e-07,
+      "loss": 1.2636,
+      "step": 49310
+    },
+    {
+      "epoch": 0.9863,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.1432549653063174e-07,
+      "loss": 1.2987,
+      "step": 49315
+    },
+    {
+      "epoch": 0.9864,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.1266291468229328e-07,
+      "loss": 1.2684,
+      "step": 49320
+    },
+    {
+      "epoch": 0.9865,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.1101250380300965e-07,
+      "loss": 1.2574,
+      "step": 49325
+    },
+    {
+      "epoch": 0.9866,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.0937426409384221e-07,
+      "loss": 1.2743,
+      "step": 49330
+    },
+    {
+      "epoch": 0.9867,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.0774819575442019e-07,
+      "loss": 1.2614,
+      "step": 49335
+    },
+    {
+      "epoch": 0.9868,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.0613429898287398e-07,
+      "loss": 1.2668,
+      "step": 49340
+    },
+    {
+      "epoch": 0.9869,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.0453257397585737e-07,
+      "loss": 1.2772,
+      "step": 49345
+    },
+    {
+      "epoch": 0.987,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.0294302092853647e-07,
+      "loss": 1.2742,
+      "step": 49350
+    },
+    {
+      "epoch": 0.9871,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.013656400345786e-07,
+      "loss": 1.2648,
+      "step": 49355
+    },
+    {
+      "epoch": 0.9872,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.980043148619667e-08,
+      "loss": 1.2745,
+      "step": 49360
+    },
+    {
+      "epoch": 0.9873,
+      "grad_norm": 0.1875,
+      "learning_rate": 9.824739547410477e-08,
+      "loss": 1.2676,
+      "step": 49365
+    },
+    {
+      "epoch": 0.9874,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.670653218752934e-08,
+      "loss": 1.2799,
+      "step": 49370
+    },
+    {
+      "epoch": 0.9875,
+      "grad_norm": 0.25,
+      "learning_rate": 9.517784181422019e-08,
+      "loss": 1.2743,
+      "step": 49375
+    },
+    {
+      "epoch": 0.9876,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 9.366132454046162e-08,
+      "loss": 1.2932,
+      "step": 49380
+    },
+    {
+      "epoch": 0.9877,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.215698055100585e-08,
+      "loss": 1.2608,
+      "step": 49385
+    },
+    {
+      "epoch": 0.9878,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.066481002918403e-08,
+      "loss": 1.2644,
+      "step": 49390
+    },
+    {
+      "epoch": 0.9879,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.918481315678406e-08,
+      "loss": 1.2677,
+      "step": 49395
+    },
+    {
+      "epoch": 0.988,
+      "grad_norm": 0.380859375,
+      "learning_rate": 8.771699011416168e-08,
+      "loss": 1.2689,
+      "step": 49400
+    },
+    {
+      "epoch": 0.9881,
+      "grad_norm": 0.24609375,
+      "learning_rate": 8.626134108016271e-08,
+      "loss": 1.2625,
+      "step": 49405
+    },
+    {
+      "epoch": 0.9882,
+      "grad_norm": 0.189453125,
+      "learning_rate": 8.481786623214528e-08,
+      "loss": 1.2818,
+      "step": 49410
+    },
+    {
+      "epoch": 0.9883,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 8.33865657459909e-08,
+      "loss": 1.2697,
+      "step": 49415
+    },
+    {
+      "epoch": 0.9884,
+      "grad_norm": 0.275390625,
+      "learning_rate": 8.196743979610455e-08,
+      "loss": 1.2704,
+      "step": 49420
+    },
+    {
+      "epoch": 0.9885,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 8.056048855540343e-08,
+      "loss": 1.2593,
+      "step": 49425
+    },
+    {
+      "epoch": 0.9886,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.91657121953171e-08,
+      "loss": 1.2591,
+      "step": 49430
+    },
+    {
+      "epoch": 0.9887,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.778311088579849e-08,
+      "loss": 1.2733,
+      "step": 49435
+    },
+    {
+      "epoch": 0.9888,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.641268479531283e-08,
+      "loss": 1.282,
+      "step": 49440
+    },
+    {
+      "epoch": 0.9889,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.505443409083768e-08,
+      "loss": 1.2773,
+      "step": 49445
+    },
+    {
+      "epoch": 0.989,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.370835893788508e-08,
+      "loss": 1.262,
+      "step": 49450
+    },
+    {
+      "epoch": 0.9891,
+      "grad_norm": 0.322265625,
+      "learning_rate": 7.237445950044608e-08,
+      "loss": 1.2808,
+      "step": 49455
+    },
+    {
+      "epoch": 0.9892,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.105273594107953e-08,
+      "loss": 1.2682,
+      "step": 49460
+    },
+    {
+      "epoch": 0.9893,
+      "grad_norm": 0.3125,
+      "learning_rate": 6.974318842081217e-08,
+      "loss": 1.2806,
+      "step": 49465
+    },
+    {
+      "epoch": 0.9894,
+      "grad_norm": 0.291015625,
+      "learning_rate": 6.844581709921638e-08,
+      "loss": 1.2674,
+      "step": 49470
+    },
+    {
+      "epoch": 0.9895,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.71606221343768e-08,
+      "loss": 1.2866,
+      "step": 49475
+    },
+    {
+      "epoch": 0.9896,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.588760368287928e-08,
+      "loss": 1.2875,
+      "step": 49480
+    },
+    {
+      "epoch": 0.9897,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.462676189985528e-08,
+      "loss": 1.2915,
+      "step": 49485
+    },
+    {
+      "epoch": 0.9898,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.337809693891527e-08,
+      "loss": 1.2901,
+      "step": 49490
+    },
+    {
+      "epoch": 0.9899,
+      "grad_norm": 0.435546875,
+      "learning_rate": 6.214160895222643e-08,
+      "loss": 1.2897,
+      "step": 49495
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.203125,
+      "learning_rate": 6.09172980904238e-08,
+      "loss": 1.284,
+      "step": 49500
+    },
+    {
+      "epoch": 0.9901,
+      "grad_norm": 0.2890625,
+      "learning_rate": 5.970516450271025e-08,
+      "loss": 1.2789,
+      "step": 49505
+    },
+    {
+      "epoch": 0.9902,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 5.850520833676765e-08,
+      "loss": 1.2749,
+      "step": 49510
+    },
+    {
+      "epoch": 0.9903,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 5.7317429738812376e-08,
+      "loss": 1.2561,
+      "step": 49515
+    },
+    {
+      "epoch": 0.9904,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.6141828853573106e-08,
+      "loss": 1.2634,
+      "step": 49520
+    },
+    {
+      "epoch": 0.9905,
+      "grad_norm": 0.166015625,
+      "learning_rate": 5.497840582429081e-08,
+      "loss": 1.2634,
+      "step": 49525
+    },
+    {
+      "epoch": 0.9906,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 5.382716079271877e-08,
+      "loss": 1.2775,
+      "step": 49530
+    },
+    {
+      "epoch": 0.9907,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.268809389913365e-08,
+      "loss": 1.2689,
+      "step": 49535
+    },
+    {
+      "epoch": 0.9908,
+      "grad_norm": 0.28515625,
+      "learning_rate": 5.1561205282335547e-08,
+      "loss": 1.2951,
+      "step": 49540
+    },
+    {
+      "epoch": 0.9909,
+      "grad_norm": 0.41015625,
+      "learning_rate": 5.0446495079636836e-08,
+      "loss": 1.2824,
+      "step": 49545
+    },
+    {
+      "epoch": 0.991,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.934396342684e-08,
+      "loss": 1.2812,
+      "step": 49550
+    },
+    {
+      "epoch": 0.9911,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 4.825361045831534e-08,
+      "loss": 1.282,
+      "step": 49555
+    },
+    {
+      "epoch": 0.9912,
+      "grad_norm": 0.275390625,
+      "learning_rate": 4.717543630688992e-08,
+      "loss": 1.2656,
+      "step": 49560
+    },
+    {
+      "epoch": 0.9913,
+      "grad_norm": 0.32421875,
+      "learning_rate": 4.610944110394755e-08,
+      "loss": 1.2651,
+      "step": 49565
+    },
+    {
+      "epoch": 0.9914,
+      "grad_norm": 0.181640625,
+      "learning_rate": 4.5055624979384316e-08,
+      "loss": 1.2714,
+      "step": 49570
+    },
+    {
+      "epoch": 0.9915,
+      "grad_norm": 0.314453125,
+      "learning_rate": 4.401398806159751e-08,
+      "loss": 1.2699,
+      "step": 49575
+    },
+    {
+      "epoch": 0.9916,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 4.298453047749673e-08,
+      "loss": 1.2611,
+      "step": 49580
+    },
+    {
+      "epoch": 0.9917,
+      "grad_norm": 0.2578125,
+      "learning_rate": 4.196725235253718e-08,
+      "loss": 1.2762,
+      "step": 49585
+    },
+    {
+      "epoch": 0.9918,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 4.096215381066415e-08,
+      "loss": 1.2709,
+      "step": 49590
+    },
+    {
+      "epoch": 0.9919,
+      "grad_norm": 0.314453125,
+      "learning_rate": 3.996923497434635e-08,
+      "loss": 1.2681,
+      "step": 49595
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 3.898849596456478e-08,
+      "loss": 1.2734,
+      "step": 49600
+    },
+    {
+      "epoch": 0.9921,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 3.8019936900812735e-08,
+      "loss": 1.2685,
+      "step": 49605
+    },
+    {
+      "epoch": 0.9922,
+      "grad_norm": 0.44921875,
+      "learning_rate": 3.7063557901129144e-08,
+      "loss": 1.2785,
+      "step": 49610
+    },
+    {
+      "epoch": 0.9923,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.61193590820208e-08,
+      "loss": 1.2754,
+      "step": 49615
+    },
+    {
+      "epoch": 0.9924,
+      "grad_norm": 0.29296875,
+      "learning_rate": 3.518734055855122e-08,
+      "loss": 1.2721,
+      "step": 49620
+    },
+    {
+      "epoch": 0.9925,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.4267502444274015e-08,
+      "loss": 1.2579,
+      "step": 49625
+    },
+    {
+      "epoch": 0.9926,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.33598448512773e-08,
+      "loss": 1.2708,
+      "step": 49630
+    },
+    {
+      "epoch": 0.9927,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.246436789015039e-08,
+      "loss": 1.266,
+      "step": 49635
+    },
+    {
+      "epoch": 0.9928,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 3.1581071670006015e-08,
+      "loss": 1.2711,
+      "step": 49640
+    },
+    {
+      "epoch": 0.9929,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.070995629846918e-08,
+      "loss": 1.2871,
+      "step": 49645
+    },
+    {
+      "epoch": 0.993,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.985102188168831e-08,
+      "loss": 1.2671,
+      "step": 49650
+    },
+    {
+      "epoch": 0.9931,
+      "grad_norm": 0.44921875,
+      "learning_rate": 2.900426852431304e-08,
+      "loss": 1.2757,
+      "step": 49655
+    },
+    {
+      "epoch": 0.9932,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 2.8169696329527483e-08,
+      "loss": 1.284,
+      "step": 49660
+    },
+    {
+      "epoch": 0.9933,
+      "grad_norm": 0.2734375,
+      "learning_rate": 2.7347305399016977e-08,
+      "loss": 1.2779,
+      "step": 49665
+    },
+    {
+      "epoch": 0.9934,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 2.6537095832990245e-08,
+      "loss": 1.2715,
+      "step": 49670
+    },
+    {
+      "epoch": 0.9935,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.573906773016832e-08,
+      "loss": 1.2785,
+      "step": 49675
+    },
+    {
+      "epoch": 0.9936,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.4953221187784537e-08,
+      "loss": 1.2859,
+      "step": 49680
+    },
+    {
+      "epoch": 0.9937,
+      "grad_norm": 0.314453125,
+      "learning_rate": 2.417955630159563e-08,
+      "loss": 1.2845,
+      "step": 49685
+    },
+    {
+      "epoch": 0.9938,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.3418073165870636e-08,
+      "loss": 1.2803,
+      "step": 49690
+    },
+    {
+      "epoch": 0.9939,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.266877187339089e-08,
+      "loss": 1.2685,
+      "step": 49695
+    },
+    {
+      "epoch": 0.994,
+      "grad_norm": 0.291015625,
+      "learning_rate": 2.193165251545004e-08,
+      "loss": 1.2958,
+      "step": 49700
+    },
+    {
+      "epoch": 0.9941,
+      "grad_norm": 0.283203125,
+      "learning_rate": 2.1206715181876226e-08,
+      "loss": 1.2768,
+      "step": 49705
+    },
+    {
+      "epoch": 0.9942,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 2.0493959960998787e-08,
+      "loss": 1.2639,
+      "step": 49710
+    },
+    {
+      "epoch": 0.9943,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.9793386939659376e-08,
+      "loss": 1.2694,
+      "step": 49715
+    },
+    {
+      "epoch": 0.9944,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.9104996203223037e-08,
+      "loss": 1.2716,
+      "step": 49720
+    },
+    {
+      "epoch": 0.9945,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.842878783557822e-08,
+      "loss": 1.271,
+      "step": 49725
+    },
+    {
+      "epoch": 0.9946,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.7764761919103477e-08,
+      "loss": 1.2875,
+      "step": 49730
+    },
+    {
+      "epoch": 0.9947,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.7112918534711865e-08,
+      "loss": 1.2713,
+      "step": 49735
+    },
+    {
+      "epoch": 0.9948,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 1.6473257761828732e-08,
+      "loss": 1.2699,
+      "step": 49740
+    },
+    {
+      "epoch": 0.9949,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.584577967840284e-08,
+      "loss": 1.2644,
+      "step": 49745
+    },
+    {
+      "epoch": 0.995,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.5230484360873044e-08,
+      "loss": 1.2683,
+      "step": 49750
+    },
+    {
+      "epoch": 0.9951,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.4627371884234909e-08,
+      "loss": 1.2741,
+      "step": 49755
+    },
+    {
+      "epoch": 0.9952,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.4036442321962995e-08,
+      "loss": 1.2817,
+      "step": 49760
+    },
+    {
+      "epoch": 0.9953,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3457695746055265e-08,
+      "loss": 1.2606,
+      "step": 49765
+    },
+    {
+      "epoch": 0.9954,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.2891132227033087e-08,
+      "loss": 1.272,
+      "step": 49770
+    },
+    {
+      "epoch": 0.9955,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.2336751833941229e-08,
+      "loss": 1.2604,
+      "step": 49775
+    },
+    {
+      "epoch": 0.9956,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.1794554634314558e-08,
+      "loss": 1.2638,
+      "step": 49780
+    },
+    {
+      "epoch": 0.9957,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.126454069423355e-08,
+      "loss": 1.2804,
+      "step": 49785
+    },
+    {
+      "epoch": 0.9958,
+      "grad_norm": 0.201171875,
+      "learning_rate": 1.074671007825767e-08,
+      "loss": 1.2707,
+      "step": 49790
+    },
+    {
+      "epoch": 0.9959,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.0241062849503103e-08,
+      "loss": 1.2677,
+      "step": 49795
+    },
+    {
+      "epoch": 0.996,
+      "grad_norm": 0.169921875,
+      "learning_rate": 9.747599069576119e-09,
+      "loss": 1.2716,
+      "step": 49800
+    },
+    {
+      "epoch": 0.9961,
+      "grad_norm": 0.365234375,
+      "learning_rate": 9.2663187986064e-09,
+      "loss": 1.2695,
+      "step": 49805
+    },
+    {
+      "epoch": 0.9962,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.797222095224823e-09,
+      "loss": 1.2924,
+      "step": 49810
+    },
+    {
+      "epoch": 0.9963,
+      "grad_norm": 0.314453125,
+      "learning_rate": 8.340309016585668e-09,
+      "loss": 1.2707,
+      "step": 49815
+    },
+    {
+      "epoch": 0.9964,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.895579618388827e-09,
+      "loss": 1.2869,
+      "step": 49820
+    },
+    {
+      "epoch": 0.9965,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.463033954802078e-09,
+      "loss": 1.282,
+      "step": 49825
+    },
+    {
+      "epoch": 0.9966,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 7.042672078527713e-09,
+      "loss": 1.2831,
+      "step": 49830
+    },
+    {
+      "epoch": 0.9967,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.634494040802519e-09,
+      "loss": 1.2799,
+      "step": 49835
+    },
+    {
+      "epoch": 0.9968,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.238499891353389e-09,
+      "loss": 1.276,
+      "step": 49840
+    },
+    {
+      "epoch": 0.9969,
+      "grad_norm": 0.193359375,
+      "learning_rate": 5.854689678419512e-09,
+      "loss": 1.2754,
+      "step": 49845
+    },
+    {
+      "epoch": 0.997,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 5.483063448785686e-09,
+      "loss": 1.2853,
+      "step": 49850
+    },
+    {
+      "epoch": 0.9971,
+      "grad_norm": 0.205078125,
+      "learning_rate": 5.123621247726806e-09,
+      "loss": 1.2739,
+      "step": 49855
+    },
+    {
+      "epoch": 0.9972,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 4.776363119030069e-09,
+      "loss": 1.2894,
+      "step": 49860
+    },
+    {
+      "epoch": 0.9973,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.4412891050171765e-09,
+      "loss": 1.2865,
+      "step": 49865
+    },
+    {
+      "epoch": 0.9974,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 4.1183992465221315e-09,
+      "loss": 1.2855,
+      "step": 49870
+    },
+    {
+      "epoch": 0.9975,
+      "grad_norm": 0.369140625,
+      "learning_rate": 3.807693582869032e-09,
+      "loss": 1.2729,
+      "step": 49875
+    },
+    {
+      "epoch": 0.9976,
+      "grad_norm": 0.27734375,
+      "learning_rate": 3.509172151938689e-09,
+      "loss": 1.2885,
+      "step": 49880
+    },
+    {
+      "epoch": 0.9977,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.222834990090906e-09,
+      "loss": 1.2905,
+      "step": 49885
+    },
+    {
+      "epoch": 0.9978,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 2.948682132208891e-09,
+      "loss": 1.2617,
+      "step": 49890
+    },
+    {
+      "epoch": 0.9979,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.6867136117214587e-09,
+      "loss": 1.268,
+      "step": 49895
+    },
+    {
+      "epoch": 0.998,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.4369294605253166e-09,
+      "loss": 1.2668,
+      "step": 49900
+    },
+    {
+      "epoch": 0.9981,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.1993297090627806e-09,
+      "loss": 1.2553,
+      "step": 49905
+    },
+    {
+      "epoch": 0.9982,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.973914386288467e-09,
+      "loss": 1.2899,
+      "step": 49910
+    },
+    {
+      "epoch": 0.9983,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.7606835196692927e-09,
+      "loss": 1.2764,
+      "step": 49915
+    },
+    {
+      "epoch": 0.9984,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.5596371351733752e-09,
+      "loss": 1.2766,
+      "step": 49920
+    },
+    {
+      "epoch": 0.9985,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.3707752573255405e-09,
+      "loss": 1.2689,
+      "step": 49925
+    },
+    {
+      "epoch": 0.9986,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.1940979091074056e-09,
+      "loss": 1.2919,
+      "step": 49930
+    },
+    {
+      "epoch": 0.9987,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.029605112068399e-09,
+      "loss": 1.2687,
+      "step": 49935
+    },
+    {
+      "epoch": 0.9988,
+      "grad_norm": 0.322265625,
+      "learning_rate": 8.772968862369446e-10,
+      "loss": 1.2645,
+      "step": 49940
+    },
+    {
+      "epoch": 0.9989,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.371732501759709e-10,
+      "loss": 1.2608,
+      "step": 49945
+    },
+    {
+      "epoch": 0.999,
+      "grad_norm": 1.0,
+      "learning_rate": 6.092342209607083e-10,
+      "loss": 1.2779,
+      "step": 49950
+    },
+    {
+      "epoch": 0.9991,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.934798141786879e-10,
+      "loss": 1.2804,
+      "step": 49955
+    },
+    {
+      "epoch": 0.9992,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.899100439408443e-10,
+      "loss": 1.2743,
+      "step": 49960
+    },
+    {
+      "epoch": 0.9993,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.985249228593112e-10,
+      "loss": 1.2691,
+      "step": 49965
+    },
+    {
+      "epoch": 0.9994,
+      "grad_norm": 0.29296875,
+      "learning_rate": 2.1932446206962553e-10,
+      "loss": 1.2677,
+      "step": 49970
+    },
+    {
+      "epoch": 0.9995,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.5230867123072756e-10,
+      "loss": 1.2702,
+      "step": 49975
+    },
+    {
+      "epoch": 0.9996,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.74775584916543e-11,
+      "loss": 1.2681,
+      "step": 49980
+    },
+    {
+      "epoch": 0.9997,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.483113054705058e-11,
+      "loss": 1.2783,
+      "step": 49985
+    },
+    {
+      "epoch": 0.9998,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.4369392592760164e-11,
+      "loss": 1.2747,
+      "step": 49990
+    },
+    {
+      "epoch": 0.9999,
+      "grad_norm": 0.4296875,
+      "learning_rate": 6.092348336927955e-12,
+      "loss": 1.2678,
+      "step": 49995
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0,
+      "loss": 1.2771,
+      "step": 50000
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.324538230895996,
+      "eval_runtime": 0.0947,
+      "eval_samples_per_second": 52.819,
+      "eval_steps_per_second": 31.691,
+      "step": 50000
+    },
+    {
+      "epoch": 1.0,
+      "step": 50000,
+      "total_flos": 6.321413148031058e+17,
+      "train_loss": 1.2974469694900512,
+      "train_runtime": 20909.5717,
+      "train_samples_per_second": 19.13,
+      "train_steps_per_second": 2.391
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 50000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "total_flos": 6.321413148031058e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}