diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,35045 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 25000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.9375,
+      "learning_rate": 8e-08,
+      "loss": 1.9137,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.5,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 1.9266,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.9375,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 1.9972,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.75,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 1.9461,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.75,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 1.8741,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.3125,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.9313,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.53125,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 1.8874,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.03125,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 1.8434,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.0,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 1.7749,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.0,
+      "learning_rate": 3.6e-06,
+      "loss": 1.6824,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.5496,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.4e-06,
+      "loss": 1.445,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 1.3503,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.84375,
+      "learning_rate": 5.2e-06,
+      "loss": 1.2192,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 1.09,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.34375,
+      "learning_rate": 6e-06,
+      "loss": 0.9502,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.453125,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 0.7846,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 0.6436,
+      "step": 85
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3828125,
+      "learning_rate": 7.2e-06,
+      "loss": 0.5717,
+      "step": 90
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.6e-06,
+      "loss": 0.5297,
+      "step": 95
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.453125,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.4972,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 0.4808,
+      "step": 105
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.34375,
+      "learning_rate": 8.8e-06,
+      "loss": 0.4422,
+      "step": 110
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.2e-06,
+      "loss": 0.4178,
+      "step": 115
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 0.3877,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1e-05,
+      "loss": 0.3634,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.04e-05,
+      "loss": 0.345,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.08e-05,
+      "loss": 0.3376,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 0.3279,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.16e-05,
+      "loss": 0.3223,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.2e-05,
+      "loss": 0.3151,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.24e-05,
+      "loss": 0.3096,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 0.3046,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.32e-05,
+      "loss": 0.3008,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 0.2993,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 0.3011,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.44e-05,
+      "loss": 0.2967,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.48e-05,
+      "loss": 0.2978,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.52e-05,
+      "loss": 0.2971,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.56e-05,
+      "loss": 0.2926,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.2993,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.6400000000000002e-05,
+      "loss": 0.295,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 0.2936,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.625,
+      "learning_rate": 1.7199999999999998e-05,
+      "loss": 0.2918,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.76e-05,
+      "loss": 0.2951,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.8e-05,
+      "loss": 0.2925,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.84e-05,
+      "loss": 0.2984,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5,
+      "learning_rate": 1.88e-05,
+      "loss": 0.2959,
+      "step": 235
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 0.2924,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 0.2897,
+      "step": 245
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.37109375,
+      "learning_rate": 2e-05,
+      "loss": 0.2961,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.04e-05,
+      "loss": 0.2902,
+      "step": 255
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.08e-05,
+      "loss": 0.2924,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.12e-05,
+      "loss": 0.2905,
+      "step": 265
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.16e-05,
+      "loss": 0.2912,
+      "step": 270
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 0.2933,
+      "step": 275
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.2400000000000002e-05,
+      "loss": 0.2934,
+      "step": 280
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.2800000000000002e-05,
+      "loss": 0.2921,
+      "step": 285
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.32e-05,
+      "loss": 0.2903,
+      "step": 290
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.365234375,
+      "learning_rate": 2.36e-05,
+      "loss": 0.2911,
+      "step": 295
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.4e-05,
+      "loss": 0.2875,
+      "step": 300
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.44e-05,
+      "loss": 0.2922,
+      "step": 305
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4921875,
+      "learning_rate": 2.48e-05,
+      "loss": 0.2933,
+      "step": 310
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.5200000000000003e-05,
+      "loss": 0.2893,
+      "step": 315
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.5600000000000002e-05,
+      "loss": 0.2919,
+      "step": 320
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 0.2937,
+      "step": 325
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.37109375,
+      "learning_rate": 2.64e-05,
+      "loss": 0.2898,
+      "step": 330
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.6800000000000004e-05,
+      "loss": 0.29,
+      "step": 335
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.283203125,
+      "learning_rate": 2.7200000000000004e-05,
+      "loss": 0.2903,
+      "step": 340
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.302734375,
+      "learning_rate": 2.7600000000000003e-05,
+      "loss": 0.291,
+      "step": 345
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 0.291,
+      "step": 350
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.46484375,
+      "learning_rate": 2.84e-05,
+      "loss": 0.2911,
+      "step": 355
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.88e-05,
+      "loss": 0.2894,
+      "step": 360
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.310546875,
+      "learning_rate": 2.9199999999999998e-05,
+      "loss": 0.2903,
+      "step": 365
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.294921875,
+      "learning_rate": 2.96e-05,
+      "loss": 0.2876,
+      "step": 370
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.466796875,
+      "learning_rate": 3e-05,
+      "loss": 0.2907,
+      "step": 375
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.04e-05,
+      "loss": 0.291,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.22265625,
+      "learning_rate": 3.08e-05,
+      "loss": 0.2884,
+      "step": 385
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.32421875,
+      "learning_rate": 3.12e-05,
+      "loss": 0.2919,
+      "step": 390
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3203125,
+      "learning_rate": 3.16e-05,
+      "loss": 0.2918,
+      "step": 395
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.289,
+      "step": 400
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.240234375,
+      "learning_rate": 3.24e-05,
+      "loss": 0.2891,
+      "step": 405
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 3.2800000000000004e-05,
+      "loss": 0.2876,
+      "step": 410
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.32e-05,
+      "loss": 0.2905,
+      "step": 415
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.37890625,
+      "learning_rate": 3.3600000000000004e-05,
+      "loss": 0.2873,
+      "step": 420
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.302734375,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 0.2865,
+      "step": 425
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.4399999999999996e-05,
+      "loss": 0.2875,
+      "step": 430
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.353515625,
+      "learning_rate": 3.48e-05,
+      "loss": 0.2879,
+      "step": 435
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.52e-05,
+      "loss": 0.2914,
+      "step": 440
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.56e-05,
+      "loss": 0.293,
+      "step": 445
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.376953125,
+      "learning_rate": 3.6e-05,
+      "loss": 0.2903,
+      "step": 450
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.275390625,
+      "learning_rate": 3.6400000000000004e-05,
+      "loss": 0.2918,
+      "step": 455
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.68e-05,
+      "loss": 0.289,
+      "step": 460
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.72e-05,
+      "loss": 0.2909,
+      "step": 465
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.76e-05,
+      "loss": 0.2882,
+      "step": 470
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.296875,
+      "learning_rate": 3.8e-05,
+      "loss": 0.2919,
+      "step": 475
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.20703125,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 0.2923,
+      "step": 480
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.330078125,
+      "learning_rate": 3.88e-05,
+      "loss": 0.2887,
+      "step": 485
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.9200000000000004e-05,
+      "loss": 0.2875,
+      "step": 490
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2265625,
+      "learning_rate": 3.960000000000001e-05,
+      "loss": 0.2904,
+      "step": 495
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.26953125,
+      "learning_rate": 4e-05,
+      "loss": 0.2892,
+      "step": 500
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3203125,
+      "learning_rate": 4.0400000000000006e-05,
+      "loss": 0.2887,
+      "step": 505
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.259765625,
+      "learning_rate": 4.08e-05,
+      "loss": 0.2907,
+      "step": 510
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.25,
+      "learning_rate": 4.12e-05,
+      "loss": 0.2882,
+      "step": 515
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.16e-05,
+      "loss": 0.2916,
+      "step": 520
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.2e-05,
+      "loss": 0.2898,
+      "step": 525
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.24e-05,
+      "loss": 0.2911,
+      "step": 530
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.263671875,
+      "learning_rate": 4.2800000000000004e-05,
+      "loss": 0.288,
+      "step": 535
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.32e-05,
+      "loss": 0.2916,
+      "step": 540
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.388671875,
+      "learning_rate": 4.36e-05,
+      "loss": 0.2885,
+      "step": 545
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.2867,
+      "step": 550
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 4.44e-05,
+      "loss": 0.2903,
+      "step": 555
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.4800000000000005e-05,
+      "loss": 0.2866,
+      "step": 560
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 4.52e-05,
+      "loss": 0.2873,
+      "step": 565
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.18359375,
+      "learning_rate": 4.5600000000000004e-05,
+      "loss": 0.2872,
+      "step": 570
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.2877,
+      "step": 575
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.64e-05,
+      "loss": 0.2892,
+      "step": 580
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.189453125,
+      "learning_rate": 4.6800000000000006e-05,
+      "loss": 0.2874,
+      "step": 585
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 4.72e-05,
+      "loss": 0.2881,
+      "step": 590
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 4.76e-05,
+      "loss": 0.2885,
+      "step": 595
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1484375,
+      "learning_rate": 4.8e-05,
+      "loss": 0.2864,
+      "step": 600
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.185546875,
+      "learning_rate": 4.8400000000000004e-05,
+      "loss": 0.2877,
+      "step": 605
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.30859375,
+      "learning_rate": 4.88e-05,
+      "loss": 0.2885,
+      "step": 610
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.25390625,
+      "learning_rate": 4.92e-05,
+      "loss": 0.2898,
+      "step": 615
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 4.96e-05,
+      "loss": 0.2887,
+      "step": 620
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 5e-05,
+      "loss": 0.288,
+      "step": 625
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 5.0400000000000005e-05,
+      "loss": 0.2882,
+      "step": 630
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.154296875,
+      "learning_rate": 5.08e-05,
+      "loss": 0.2872,
+      "step": 635
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1953125,
+      "learning_rate": 5.1200000000000004e-05,
+      "loss": 0.2873,
+      "step": 640
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.16e-05,
+      "loss": 0.2898,
+      "step": 645
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2109375,
+      "learning_rate": 5.2000000000000004e-05,
+      "loss": 0.2876,
+      "step": 650
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1953125,
+      "learning_rate": 5.2400000000000007e-05,
+      "loss": 0.286,
+      "step": 655
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.380859375,
+      "learning_rate": 5.28e-05,
+      "loss": 0.2864,
+      "step": 660
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.3200000000000006e-05,
+      "loss": 0.2896,
+      "step": 665
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.27734375,
+      "learning_rate": 5.360000000000001e-05,
+      "loss": 0.2898,
+      "step": 670
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.27734375,
+      "learning_rate": 5.4000000000000005e-05,
+      "loss": 0.2927,
+      "step": 675
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.28515625,
+      "learning_rate": 5.440000000000001e-05,
+      "loss": 0.2906,
+      "step": 680
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 5.4800000000000004e-05,
+      "loss": 0.2911,
+      "step": 685
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.189453125,
+      "learning_rate": 5.520000000000001e-05,
+      "loss": 0.2885,
+      "step": 690
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.228515625,
+      "learning_rate": 5.560000000000001e-05,
+      "loss": 0.2919,
+      "step": 695
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 5.6000000000000006e-05,
+      "loss": 0.2913,
+      "step": 700
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13671875,
+      "learning_rate": 5.6399999999999995e-05,
+      "loss": 0.2878,
+      "step": 705
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 5.68e-05,
+      "loss": 0.291,
+      "step": 710
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.158203125,
+      "learning_rate": 5.72e-05,
+      "loss": 0.2893,
+      "step": 715
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 5.76e-05,
+      "loss": 0.2902,
+      "step": 720
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 5.8e-05,
+      "loss": 0.2897,
+      "step": 725
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.390625,
+      "learning_rate": 5.8399999999999997e-05,
+      "loss": 0.2885,
+      "step": 730
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.24609375,
+      "learning_rate": 5.88e-05,
+      "loss": 0.2868,
+      "step": 735
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.92e-05,
+      "loss": 0.2905,
+      "step": 740
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 5.96e-05,
+      "loss": 0.2888,
+      "step": 745
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 6e-05,
+      "loss": 0.286,
+      "step": 750
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.04e-05,
+      "loss": 0.2884,
+      "step": 755
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.224609375,
+      "learning_rate": 6.08e-05,
+      "loss": 0.2904,
+      "step": 760
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.26953125,
+      "learning_rate": 6.12e-05,
+      "loss": 0.2903,
+      "step": 765
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.255859375,
+      "learning_rate": 6.16e-05,
+      "loss": 0.2896,
+      "step": 770
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 6.2e-05,
+      "loss": 0.2885,
+      "step": 775
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1484375,
+      "learning_rate": 6.24e-05,
+      "loss": 0.2884,
+      "step": 780
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 6.280000000000001e-05,
+      "loss": 0.2891,
+      "step": 785
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13671875,
+      "learning_rate": 6.32e-05,
+      "loss": 0.2854,
+      "step": 790
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 6.36e-05,
+      "loss": 0.2889,
+      "step": 795
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 0.2929,
+      "step": 800
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.440000000000001e-05,
+      "loss": 0.288,
+      "step": 805
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 6.48e-05,
+      "loss": 0.2893,
+      "step": 810
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3671875,
+      "learning_rate": 6.52e-05,
+      "loss": 0.2887,
+      "step": 815
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.560000000000001e-05,
+      "loss": 0.2868,
+      "step": 820
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 6.6e-05,
+      "loss": 0.2876,
+      "step": 825
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 6.64e-05,
+      "loss": 0.2876,
+      "step": 830
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.680000000000001e-05,
+      "loss": 0.2899,
+      "step": 835
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 6.720000000000001e-05,
+      "loss": 0.2909,
+      "step": 840
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 6.76e-05,
+      "loss": 0.2897,
+      "step": 845
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13671875,
+      "learning_rate": 6.800000000000001e-05,
+      "loss": 0.2895,
+      "step": 850
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 6.840000000000001e-05,
+      "loss": 0.29,
+      "step": 855
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.197265625,
+      "learning_rate": 6.879999999999999e-05,
+      "loss": 0.2905,
+      "step": 860
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 6.92e-05,
+      "loss": 0.288,
+      "step": 865
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 6.96e-05,
+      "loss": 0.286,
+      "step": 870
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.15625,
+      "learning_rate": 7e-05,
+      "loss": 0.2894,
+      "step": 875
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.04e-05,
+      "loss": 0.2884,
+      "step": 880
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 7.08e-05,
+      "loss": 0.289,
+      "step": 885
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 7.12e-05,
+      "loss": 0.2882,
+      "step": 890
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 7.16e-05,
+      "loss": 0.2892,
+      "step": 895
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.2e-05,
+      "loss": 0.289,
+      "step": 900
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.15625,
+      "learning_rate": 7.24e-05,
+      "loss": 0.2885,
+      "step": 905
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 7.280000000000001e-05,
+      "loss": 0.2885,
+      "step": 910
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 7.32e-05,
+      "loss": 0.2874,
+      "step": 915
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.36e-05,
+      "loss": 0.2873,
+      "step": 920
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 7.4e-05,
+      "loss": 0.2862,
+      "step": 925
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1015625,
+      "learning_rate": 7.44e-05,
+      "loss": 0.2883,
+      "step": 930
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 7.48e-05,
+      "loss": 0.2871,
+      "step": 935
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.142578125,
+      "learning_rate": 7.52e-05,
+      "loss": 0.2889,
+      "step": 940
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.123046875,
+      "learning_rate": 7.560000000000001e-05,
+      "loss": 0.2871,
+      "step": 945
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 7.6e-05,
+      "loss": 0.2889,
+      "step": 950
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 7.64e-05,
+      "loss": 0.2877,
+      "step": 955
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.197265625,
+      "learning_rate": 7.680000000000001e-05,
+      "loss": 0.2877,
+      "step": 960
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.72e-05,
+      "loss": 0.2883,
+      "step": 965
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.111328125,
+      "learning_rate": 7.76e-05,
+      "loss": 0.2882,
+      "step": 970
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 7.800000000000001e-05,
+      "loss": 0.2864,
+      "step": 975
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.840000000000001e-05,
+      "loss": 0.2899,
+      "step": 980
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.119140625,
+      "learning_rate": 7.88e-05,
+      "loss": 0.2851,
+      "step": 985
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 7.920000000000001e-05,
+      "loss": 0.287,
+      "step": 990
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 7.960000000000001e-05,
+      "loss": 0.2868,
+      "step": 995
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 8e-05,
+      "loss": 0.2873,
+      "step": 1000
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 8.04e-05,
+      "loss": 0.2875,
+      "step": 1005
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.23828125,
+      "learning_rate": 8.080000000000001e-05,
+      "loss": 0.2877,
+      "step": 1010
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.111328125,
+      "learning_rate": 8.120000000000001e-05,
+      "loss": 0.2859,
+      "step": 1015
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 8.16e-05,
+      "loss": 0.2879,
+      "step": 1020
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.171875,
+      "learning_rate": 8.2e-05,
+      "loss": 0.2874,
+      "step": 1025
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.24e-05,
+      "loss": 0.2888,
+      "step": 1030
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.109375,
+      "learning_rate": 8.28e-05,
+      "loss": 0.2903,
+      "step": 1035
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 8.32e-05,
+      "loss": 0.2878,
+      "step": 1040
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 8.36e-05,
+      "loss": 0.2879,
+      "step": 1045
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 8.4e-05,
+      "loss": 0.2876,
+      "step": 1050
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.134765625,
+      "learning_rate": 8.44e-05,
+      "loss": 0.2923,
+      "step": 1055
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.138671875,
+      "learning_rate": 8.48e-05,
+      "loss": 0.2883,
+      "step": 1060
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.103515625,
+      "learning_rate": 8.52e-05,
+      "loss": 0.2866,
+      "step": 1065
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 8.560000000000001e-05,
+      "loss": 0.2869,
+      "step": 1070
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 8.6e-05,
+      "loss": 0.2893,
+      "step": 1075
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.09765625,
+      "learning_rate": 8.64e-05,
+      "loss": 0.2878,
+      "step": 1080
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 8.680000000000001e-05,
+      "loss": 0.2895,
+      "step": 1085
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.111328125,
+      "learning_rate": 8.72e-05,
+      "loss": 0.2865,
+      "step": 1090
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1484375,
+      "learning_rate": 8.76e-05,
+      "loss": 0.2886,
+      "step": 1095
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.28515625,
+      "learning_rate": 8.800000000000001e-05,
+      "loss": 0.29,
+      "step": 1100
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.087890625,
+      "learning_rate": 8.840000000000001e-05,
+      "loss": 0.2888,
+      "step": 1105
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 8.88e-05,
+      "loss": 0.2884,
+      "step": 1110
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 8.92e-05,
+      "loss": 0.287,
+      "step": 1115
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 8.960000000000001e-05,
+      "loss": 0.2887,
+      "step": 1120
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.125,
+      "learning_rate": 9e-05,
+      "loss": 0.2881,
+      "step": 1125
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.04e-05,
+      "loss": 0.2894,
+      "step": 1130
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.119140625,
+      "learning_rate": 9.080000000000001e-05,
+      "loss": 0.2895,
+      "step": 1135
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 9.120000000000001e-05,
+      "loss": 0.2903,
+      "step": 1140
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.16e-05,
+      "loss": 0.2882,
+      "step": 1145
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.200000000000001e-05,
+      "loss": 0.2898,
+      "step": 1150
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 9.240000000000001e-05,
+      "loss": 0.2873,
+      "step": 1155
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.205078125,
+      "learning_rate": 9.28e-05,
+      "loss": 0.2902,
+      "step": 1160
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 9.320000000000002e-05,
+      "loss": 0.2851,
+      "step": 1165
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1640625,
+      "learning_rate": 9.360000000000001e-05,
+      "loss": 0.2889,
+      "step": 1170
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 9.4e-05,
+      "loss": 0.2905,
+      "step": 1175
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 9.44e-05,
+      "loss": 0.2866,
+      "step": 1180
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 9.48e-05,
+      "loss": 0.2895,
+      "step": 1185
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.150390625,
+      "learning_rate": 9.52e-05,
+      "loss": 0.288,
+      "step": 1190
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 9.56e-05,
+      "loss": 0.2873,
+      "step": 1195
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 9.6e-05,
+      "loss": 0.2885,
+      "step": 1200
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.130859375,
+      "learning_rate": 9.64e-05,
+      "loss": 0.2878,
+      "step": 1205
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 9.680000000000001e-05,
+      "loss": 0.2874,
+      "step": 1210
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 9.72e-05,
+      "loss": 0.2878,
+      "step": 1215
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.09375,
+      "learning_rate": 9.76e-05,
+      "loss": 0.2887,
+      "step": 1220
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.162109375,
+      "learning_rate": 9.8e-05,
+      "loss": 0.2875,
+      "step": 1225
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 9.84e-05,
+      "loss": 0.2875,
+      "step": 1230
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 9.88e-05,
+      "loss": 0.2888,
+      "step": 1235
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 9.92e-05,
+      "loss": 0.2872,
+      "step": 1240
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 9.960000000000001e-05,
+      "loss": 0.2889,
+      "step": 1245
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0001,
+      "loss": 0.2882,
+      "step": 1250
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0001004,
+      "loss": 0.2871,
+      "step": 1255
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00010080000000000001,
+      "loss": 0.2893,
+      "step": 1260
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00010120000000000001,
+      "loss": 0.2897,
+      "step": 1265
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0001016,
+      "loss": 0.2882,
+      "step": 1270
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.00010200000000000001,
+      "loss": 0.29,
+      "step": 1275
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00010240000000000001,
+      "loss": 0.2874,
+      "step": 1280
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0001028,
+      "loss": 0.291,
+      "step": 1285
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0001032,
+      "loss": 0.2911,
+      "step": 1290
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00010360000000000001,
+      "loss": 0.2884,
+      "step": 1295
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00010400000000000001,
+      "loss": 0.2878,
+      "step": 1300
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0001044,
+      "loss": 0.2849,
+      "step": 1305
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00010480000000000001,
+      "loss": 0.2899,
+      "step": 1310
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00010520000000000001,
+      "loss": 0.2889,
+      "step": 1315
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0001056,
+      "loss": 0.2883,
+      "step": 1320
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00010600000000000002,
+      "loss": 0.2883,
+      "step": 1325
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00010640000000000001,
+      "loss": 0.2859,
+      "step": 1330
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00010680000000000001,
+      "loss": 0.2953,
+      "step": 1335
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00010720000000000002,
+      "loss": 0.2893,
+      "step": 1340
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.00010760000000000001,
+      "loss": 0.2893,
+      "step": 1345
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.00010800000000000001,
+      "loss": 0.2894,
+      "step": 1350
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00010840000000000002,
+      "loss": 0.2893,
+      "step": 1355
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00010880000000000002,
+      "loss": 0.2869,
+      "step": 1360
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00010920000000000001,
+      "loss": 0.2887,
+      "step": 1365
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.00010960000000000001,
+      "loss": 0.288,
+      "step": 1370
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00011000000000000002,
+      "loss": 0.2875,
+      "step": 1375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00011040000000000001,
+      "loss": 0.2901,
+      "step": 1380
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00011080000000000001,
+      "loss": 0.2885,
+      "step": 1385
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00011120000000000002,
+      "loss": 0.2915,
+      "step": 1390
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00011160000000000002,
+      "loss": 0.2955,
+      "step": 1395
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00011200000000000001,
+      "loss": 0.2918,
+      "step": 1400
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00011240000000000002,
+      "loss": 0.2901,
+      "step": 1405
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00011279999999999999,
+      "loss": 0.2896,
+      "step": 1410
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0001132,
+      "loss": 0.2889,
+      "step": 1415
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0001136,
+      "loss": 0.2916,
+      "step": 1420
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00011399999999999999,
+      "loss": 0.2918,
+      "step": 1425
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001144,
+      "loss": 0.2924,
+      "step": 1430
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001148,
+      "loss": 0.2903,
+      "step": 1435
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001152,
+      "loss": 0.2915,
+      "step": 1440
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00011559999999999999,
+      "loss": 0.2905,
+      "step": 1445
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.000116,
+      "loss": 0.2891,
+      "step": 1450
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001164,
+      "loss": 0.2892,
+      "step": 1455
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00011679999999999999,
+      "loss": 0.2898,
+      "step": 1460
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001172,
+      "loss": 0.287,
+      "step": 1465
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0001176,
+      "loss": 0.2881,
+      "step": 1470
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.000118,
+      "loss": 0.2954,
+      "step": 1475
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001184,
+      "loss": 0.2902,
+      "step": 1480
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0001188,
+      "loss": 0.2874,
+      "step": 1485
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001192,
+      "loss": 0.2877,
+      "step": 1490
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00011960000000000001,
+      "loss": 0.2927,
+      "step": 1495
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00012,
+      "loss": 0.2876,
+      "step": 1500
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0001204,
+      "loss": 0.2872,
+      "step": 1505
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0001208,
+      "loss": 0.2868,
+      "step": 1510
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0001212,
+      "loss": 0.288,
+      "step": 1515
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001216,
+      "loss": 0.2921,
+      "step": 1520
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.000122,
+      "loss": 0.2885,
+      "step": 1525
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0001224,
+      "loss": 0.29,
+      "step": 1530
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001228,
+      "loss": 0.2899,
+      "step": 1535
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0001232,
+      "loss": 0.2913,
+      "step": 1540
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0001236,
+      "loss": 0.2876,
+      "step": 1545
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.000124,
+      "loss": 0.2923,
+      "step": 1550
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00012440000000000002,
+      "loss": 0.2901,
+      "step": 1555
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0001248,
+      "loss": 0.2879,
+      "step": 1560
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001252,
+      "loss": 0.2889,
+      "step": 1565
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00012560000000000002,
+      "loss": 0.2866,
+      "step": 1570
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.000126,
+      "loss": 0.2894,
+      "step": 1575
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0001264,
+      "loss": 0.2859,
+      "step": 1580
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00012680000000000002,
+      "loss": 0.2912,
+      "step": 1585
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0001272,
+      "loss": 0.2888,
+      "step": 1590
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001276,
+      "loss": 0.2874,
+      "step": 1595
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.00012800000000000002,
+      "loss": 0.2895,
+      "step": 1600
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001284,
+      "loss": 0.2877,
+      "step": 1605
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00012880000000000001,
+      "loss": 0.2882,
+      "step": 1610
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.00012920000000000002,
+      "loss": 0.2885,
+      "step": 1615
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0001296,
+      "loss": 0.2886,
+      "step": 1620
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00013000000000000002,
+      "loss": 0.2887,
+      "step": 1625
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0001304,
+      "loss": 0.29,
+      "step": 1630
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001308,
+      "loss": 0.2901,
+      "step": 1635
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.00013120000000000002,
+      "loss": 0.2873,
+      "step": 1640
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001316,
+      "loss": 0.2897,
+      "step": 1645
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.000132,
+      "loss": 0.2889,
+      "step": 1650
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00013240000000000002,
+      "loss": 0.2923,
+      "step": 1655
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001328,
+      "loss": 0.289,
+      "step": 1660
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001332,
+      "loss": 0.2866,
+      "step": 1665
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.00013360000000000002,
+      "loss": 0.286,
+      "step": 1670
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.000134,
+      "loss": 0.2893,
+      "step": 1675
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00013440000000000001,
+      "loss": 0.2865,
+      "step": 1680
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.00013480000000000002,
+      "loss": 0.2928,
+      "step": 1685
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0001352,
+      "loss": 0.2867,
+      "step": 1690
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00013560000000000002,
+      "loss": 0.2896,
+      "step": 1695
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00013600000000000003,
+      "loss": 0.2898,
+      "step": 1700
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001364,
+      "loss": 0.2876,
+      "step": 1705
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00013680000000000002,
+      "loss": 0.2865,
+      "step": 1710
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00013720000000000003,
+      "loss": 0.2864,
+      "step": 1715
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.00013759999999999998,
+      "loss": 0.2914,
+      "step": 1720
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.000138,
+      "loss": 0.2928,
+      "step": 1725
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0001384,
+      "loss": 0.2863,
+      "step": 1730
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.00013879999999999999,
+      "loss": 0.2863,
+      "step": 1735
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001392,
+      "loss": 0.2919,
+      "step": 1740
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0001396,
+      "loss": 0.2885,
+      "step": 1745
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00014,
+      "loss": 0.2885,
+      "step": 1750
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0001404,
+      "loss": 0.2879,
+      "step": 1755
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0001408,
+      "loss": 0.2889,
+      "step": 1760
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0001412,
+      "loss": 0.2879,
+      "step": 1765
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0001416,
+      "loss": 0.2865,
+      "step": 1770
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.000142,
+      "loss": 0.2893,
+      "step": 1775
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0001424,
+      "loss": 0.2878,
+      "step": 1780
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0001428,
+      "loss": 0.2879,
+      "step": 1785
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0001432,
+      "loss": 0.2864,
+      "step": 1790
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0001436,
+      "loss": 0.288,
+      "step": 1795
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.000144,
+      "loss": 0.286,
+      "step": 1800
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0001444,
+      "loss": 0.2876,
+      "step": 1805
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0001448,
+      "loss": 0.2865,
+      "step": 1810
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0001452,
+      "loss": 0.2883,
+      "step": 1815
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00014560000000000002,
+      "loss": 0.2883,
+      "step": 1820
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.000146,
+      "loss": 0.287,
+      "step": 1825
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0001464,
+      "loss": 0.2862,
+      "step": 1830
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.00014680000000000002,
+      "loss": 0.2876,
+      "step": 1835
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0001472,
+      "loss": 0.2892,
+      "step": 1840
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0001476,
+      "loss": 0.2894,
+      "step": 1845
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.000148,
+      "loss": 0.2927,
+      "step": 1850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0001484,
+      "loss": 0.2915,
+      "step": 1855
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001488,
+      "loss": 0.2902,
+      "step": 1860
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0001492,
+      "loss": 0.2904,
+      "step": 1865
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0001496,
+      "loss": 0.2891,
+      "step": 1870
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.2895,
+      "step": 1875
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0001504,
+      "loss": 0.2882,
+      "step": 1880
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0001508,
+      "loss": 0.2874,
+      "step": 1885
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00015120000000000002,
+      "loss": 0.2868,
+      "step": 1890
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0001516,
+      "loss": 0.2869,
+      "step": 1895
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.000152,
+      "loss": 0.2902,
+      "step": 1900
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00015240000000000002,
+      "loss": 0.2906,
+      "step": 1905
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0001528,
+      "loss": 0.2912,
+      "step": 1910
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0001532,
+      "loss": 0.29,
+      "step": 1915
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00015360000000000002,
+      "loss": 0.2896,
+      "step": 1920
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.000154,
+      "loss": 0.2922,
+      "step": 1925
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0001544,
+      "loss": 0.29,
+      "step": 1930
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.00015480000000000002,
+      "loss": 0.2881,
+      "step": 1935
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0001552,
+      "loss": 0.2874,
+      "step": 1940
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00015560000000000001,
+      "loss": 0.2888,
+      "step": 1945
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.00015600000000000002,
+      "loss": 0.2891,
+      "step": 1950
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0001564,
+      "loss": 0.2875,
+      "step": 1955
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00015680000000000002,
+      "loss": 0.2902,
+      "step": 1960
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00015720000000000003,
+      "loss": 0.2888,
+      "step": 1965
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001576,
+      "loss": 0.2901,
+      "step": 1970
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.00015800000000000002,
+      "loss": 0.2892,
+      "step": 1975
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00015840000000000003,
+      "loss": 0.2886,
+      "step": 1980
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0001588,
+      "loss": 0.2884,
+      "step": 1985
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00015920000000000002,
+      "loss": 0.2874,
+      "step": 1990
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0001596,
+      "loss": 0.2878,
+      "step": 1995
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00016,
+      "loss": 0.2896,
+      "step": 2000
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.00016040000000000002,
+      "loss": 0.287,
+      "step": 2005
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0001608,
+      "loss": 0.2867,
+      "step": 2010
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.00016120000000000002,
+      "loss": 0.2862,
+      "step": 2015
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.00016160000000000002,
+      "loss": 0.2881,
+      "step": 2020
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.000162,
+      "loss": 0.2899,
+      "step": 2025
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00016240000000000002,
+      "loss": 0.2903,
+      "step": 2030
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0001628,
+      "loss": 0.2889,
+      "step": 2035
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0001632,
+      "loss": 0.2854,
+      "step": 2040
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0001636,
+      "loss": 0.2883,
+      "step": 2045
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.000164,
+      "loss": 0.287,
+      "step": 2050
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0001644,
+      "loss": 0.2924,
+      "step": 2055
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0001648,
+      "loss": 0.2884,
+      "step": 2060
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001652,
+      "loss": 0.2907,
+      "step": 2065
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0001656,
+      "loss": 0.2918,
+      "step": 2070
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.000166,
+      "loss": 0.2886,
+      "step": 2075
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001664,
+      "loss": 0.2881,
+      "step": 2080
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0001668,
+      "loss": 0.2897,
+      "step": 2085
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0001672,
+      "loss": 0.2887,
+      "step": 2090
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0001676,
+      "loss": 0.287,
+      "step": 2095
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.000168,
+      "loss": 0.2898,
+      "step": 2100
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0001684,
+      "loss": 0.2876,
+      "step": 2105
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0001688,
+      "loss": 0.2856,
+      "step": 2110
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0001692,
+      "loss": 0.2882,
+      "step": 2115
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0001696,
+      "loss": 0.2898,
+      "step": 2120
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00017,
+      "loss": 0.2878,
+      "step": 2125
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0001704,
+      "loss": 0.2877,
+      "step": 2130
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001708,
+      "loss": 0.2893,
+      "step": 2135
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00017120000000000001,
+      "loss": 0.2875,
+      "step": 2140
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0001716,
+      "loss": 0.289,
+      "step": 2145
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.000172,
+      "loss": 0.2878,
+      "step": 2150
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00017240000000000002,
+      "loss": 0.2893,
+      "step": 2155
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0001728,
+      "loss": 0.2896,
+      "step": 2160
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0001732,
+      "loss": 0.289,
+      "step": 2165
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00017360000000000002,
+      "loss": 0.286,
+      "step": 2170
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.000174,
+      "loss": 0.2869,
+      "step": 2175
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001744,
+      "loss": 0.2918,
+      "step": 2180
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00017480000000000002,
+      "loss": 0.2905,
+      "step": 2185
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0001752,
+      "loss": 0.2888,
+      "step": 2190
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0001756,
+      "loss": 0.2874,
+      "step": 2195
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00017600000000000002,
+      "loss": 0.2888,
+      "step": 2200
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0001764,
+      "loss": 0.2892,
+      "step": 2205
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.00017680000000000001,
+      "loss": 0.2887,
+      "step": 2210
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0001772,
+      "loss": 0.2891,
+      "step": 2215
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0001776,
+      "loss": 0.289,
+      "step": 2220
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.00017800000000000002,
+      "loss": 0.2873,
+      "step": 2225
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001784,
+      "loss": 0.2886,
+      "step": 2230
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001788,
+      "loss": 0.289,
+      "step": 2235
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00017920000000000002,
+      "loss": 0.2887,
+      "step": 2240
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001796,
+      "loss": 0.2881,
+      "step": 2245
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.00018,
+      "loss": 0.2904,
+      "step": 2250
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00018040000000000002,
+      "loss": 0.2898,
+      "step": 2255
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0001808,
+      "loss": 0.2905,
+      "step": 2260
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0001812,
+      "loss": 0.2945,
+      "step": 2265
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00018160000000000002,
+      "loss": 0.3762,
+      "step": 2270
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.000182,
+      "loss": 0.3179,
+      "step": 2275
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00018240000000000002,
+      "loss": 0.2914,
+      "step": 2280
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00018280000000000003,
+      "loss": 0.2931,
+      "step": 2285
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001832,
+      "loss": 0.2906,
+      "step": 2290
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00018360000000000002,
+      "loss": 0.3082,
+      "step": 2295
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00018400000000000003,
+      "loss": 0.2977,
+      "step": 2300
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001844,
+      "loss": 0.2927,
+      "step": 2305
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00018480000000000002,
+      "loss": 0.2943,
+      "step": 2310
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00018520000000000003,
+      "loss": 0.292,
+      "step": 2315
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001856,
+      "loss": 0.2995,
+      "step": 2320
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00018600000000000002,
+      "loss": 0.3,
+      "step": 2325
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018640000000000003,
+      "loss": 0.2955,
+      "step": 2330
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00018680000000000001,
+      "loss": 0.2931,
+      "step": 2335
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00018720000000000002,
+      "loss": 0.2937,
+      "step": 2340
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001876,
+      "loss": 0.2977,
+      "step": 2345
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.000188,
+      "loss": 0.2928,
+      "step": 2350
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0001884,
+      "loss": 0.2911,
+      "step": 2355
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001888,
+      "loss": 0.2922,
+      "step": 2360
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001892,
+      "loss": 0.2936,
+      "step": 2365
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001896,
+      "loss": 0.2938,
+      "step": 2370
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00019,
+      "loss": 0.29,
+      "step": 2375
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001904,
+      "loss": 0.2875,
+      "step": 2380
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0001908,
+      "loss": 0.2915,
+      "step": 2385
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0001912,
+      "loss": 0.2939,
+      "step": 2390
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001916,
+      "loss": 0.2943,
+      "step": 2395
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.000192,
+      "loss": 0.2897,
+      "step": 2400
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00019240000000000001,
+      "loss": 0.2894,
+      "step": 2405
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0001928,
+      "loss": 0.29,
+      "step": 2410
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0001932,
+      "loss": 0.2894,
+      "step": 2415
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00019360000000000002,
+      "loss": 0.2927,
+      "step": 2420
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.000194,
+      "loss": 0.2896,
+      "step": 2425
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001944,
+      "loss": 0.2883,
+      "step": 2430
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0001948,
+      "loss": 0.2899,
+      "step": 2435
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0001952,
+      "loss": 0.2882,
+      "step": 2440
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0001956,
+      "loss": 0.2887,
+      "step": 2445
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.000196,
+      "loss": 0.2903,
+      "step": 2450
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0001964,
+      "loss": 0.2883,
+      "step": 2455
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0001968,
+      "loss": 0.2887,
+      "step": 2460
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001972,
+      "loss": 0.2916,
+      "step": 2465
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0001976,
+      "loss": 0.2909,
+      "step": 2470
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.00019800000000000002,
+      "loss": 0.2902,
+      "step": 2475
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001984,
+      "loss": 0.2927,
+      "step": 2480
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0001988,
+      "loss": 0.2898,
+      "step": 2485
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.00019920000000000002,
+      "loss": 0.2906,
+      "step": 2490
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0001996,
+      "loss": 0.2886,
+      "step": 2495
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0002,
+      "loss": 0.2859,
+      "step": 2500
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0001999999756306074,
+      "loss": 0.2882,
+      "step": 2505
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.00019999990252244151,
+      "loss": 0.2901,
+      "step": 2510
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.00019999978067553796,
+      "loss": 0.2879,
+      "step": 2515
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00019999961008995606,
+      "loss": 0.288,
+      "step": 2520
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019999939076577905,
+      "loss": 0.2879,
+      "step": 2525
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00019999912270311375,
+      "loss": 0.2995,
+      "step": 2530
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0001999988059020909,
+      "loss": 0.2872,
+      "step": 2535
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.00019999844036286483,
+      "loss": 0.291,
+      "step": 2540
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001999980260856137,
+      "loss": 0.2935,
+      "step": 2545
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00019999756307053948,
+      "loss": 0.2899,
+      "step": 2550
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001999970513178678,
+      "loss": 0.2878,
+      "step": 2555
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001999964908278481,
+      "loss": 0.2904,
+      "step": 2560
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00019999588160075348,
+      "loss": 0.288,
+      "step": 2565
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.000199995223636881,
+      "loss": 0.2901,
+      "step": 2570
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00019999451693655123,
+      "loss": 0.292,
+      "step": 2575
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00019999376150010866,
+      "loss": 0.2869,
+      "step": 2580
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00019999295732792146,
+      "loss": 0.2878,
+      "step": 2585
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019999210442038162,
+      "loss": 0.2926,
+      "step": 2590
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019999120277790477,
+      "loss": 0.2938,
+      "step": 2595
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019999025240093044,
+      "loss": 0.2919,
+      "step": 2600
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00019998925328992175,
+      "loss": 0.2899,
+      "step": 2605
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0001999882054453657,
+      "loss": 0.2899,
+      "step": 2610
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.000199987108867773,
+      "loss": 0.2948,
+      "step": 2615
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00019998596355767805,
+      "loss": 0.2948,
+      "step": 2620
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00019998476951563915,
+      "loss": 0.2911,
+      "step": 2625
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019998352674223816,
+      "loss": 0.288,
+      "step": 2630
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001999822352380809,
+      "loss": 0.2905,
+      "step": 2635
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019998089500379676,
+      "loss": 0.2939,
+      "step": 2640
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.000199979506040039,
+      "loss": 0.291,
+      "step": 2645
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00019997806834748456,
+      "loss": 0.2939,
+      "step": 2650
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00019997658192683412,
+      "loss": 0.2907,
+      "step": 2655
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 18.875,
+      "learning_rate": 0.00019997504677881224,
+      "loss": 0.2944,
+      "step": 2660
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.000199973462904167,
+      "loss": 0.2935,
+      "step": 2665
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019997183030367048,
+      "loss": 0.2941,
+      "step": 2670
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019997014897811833,
+      "loss": 0.2929,
+      "step": 2675
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019996841892833,
+      "loss": 0.2931,
+      "step": 2680
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019996664015514871,
+      "loss": 0.2947,
+      "step": 2685
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00019996481265944146,
+      "loss": 0.2894,
+      "step": 2690
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019996293644209887,
+      "loss": 0.2929,
+      "step": 2695
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019996101150403543,
+      "loss": 0.3026,
+      "step": 2700
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00019995903784618936,
+      "loss": 0.2991,
+      "step": 2705
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001999570154695225,
+      "loss": 0.2998,
+      "step": 2710
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00019995494437502064,
+      "loss": 0.2907,
+      "step": 2715
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001999528245636931,
+      "loss": 0.2971,
+      "step": 2720
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00019995065603657316,
+      "loss": 0.2897,
+      "step": 2725
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0001999484387947177,
+      "loss": 0.2909,
+      "step": 2730
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001999461728392073,
+      "loss": 0.2922,
+      "step": 2735
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019994385817114646,
+      "loss": 0.2956,
+      "step": 2740
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00019994149479166324,
+      "loss": 0.2929,
+      "step": 2745
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001999390827019096,
+      "loss": 0.2957,
+      "step": 2750
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001999366219030611,
+      "loss": 0.2912,
+      "step": 2755
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001999341123963171,
+      "loss": 0.2913,
+      "step": 2760
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0001999315541829008,
+      "loss": 0.2927,
+      "step": 2765
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00019992894726405893,
+      "loss": 0.2927,
+      "step": 2770
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0001999262916410621,
+      "loss": 0.2924,
+      "step": 2775
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019992358731520468,
+      "loss": 0.2896,
+      "step": 2780
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0001999208342878047,
+      "loss": 0.287,
+      "step": 2785
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00019991803256020393,
+      "loss": 0.3274,
+      "step": 2790
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00019991518213376787,
+      "loss": 0.2898,
+      "step": 2795
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00019991228300988585,
+      "loss": 0.2926,
+      "step": 2800
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019990933518997084,
+      "loss": 0.2929,
+      "step": 2805
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019990633867545955,
+      "loss": 0.2906,
+      "step": 2810
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0001999032934678125,
+      "loss": 0.2898,
+      "step": 2815
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.00019990019956851382,
+      "loss": 0.289,
+      "step": 2820
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00019989705697907149,
+      "loss": 0.2877,
+      "step": 2825
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00019989386570101714,
+      "loss": 0.2937,
+      "step": 2830
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00019989062573590616,
+      "loss": 0.3376,
+      "step": 2835
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0001998873370853177,
+      "loss": 0.293,
+      "step": 2840
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001998839997508546,
+      "loss": 0.2889,
+      "step": 2845
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0001998806137341434,
+      "loss": 0.2976,
+      "step": 2850
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019987717903683448,
+      "loss": 0.2978,
+      "step": 2855
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00019987369566060176,
+      "loss": 0.2912,
+      "step": 2860
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.00019987016360714307,
+      "loss": 0.2876,
+      "step": 2865
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.00019986658287817987,
+      "loss": 0.2879,
+      "step": 2870
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0001998629534754574,
+      "loss": 0.2902,
+      "step": 2875
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019985927540074454,
+      "loss": 0.2874,
+      "step": 2880
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.00019985554865583393,
+      "loss": 0.2868,
+      "step": 2885
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.000199851773242542,
+      "loss": 0.2892,
+      "step": 2890
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019984794916270876,
+      "loss": 0.2866,
+      "step": 2895
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019984407641819812,
+      "loss": 0.2893,
+      "step": 2900
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019984015501089752,
+      "loss": 0.2892,
+      "step": 2905
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019983618494271826,
+      "loss": 0.289,
+      "step": 2910
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00019983216621559524,
+      "loss": 0.2873,
+      "step": 2915
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019982809883148722,
+      "loss": 0.2899,
+      "step": 2920
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00019982398279237655,
+      "loss": 0.2894,
+      "step": 2925
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.00019981981810026934,
+      "loss": 0.2868,
+      "step": 2930
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.00019981560475719538,
+      "loss": 0.2875,
+      "step": 2935
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.00019981134276520827,
+      "loss": 0.2882,
+      "step": 2940
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.00019980703212638522,
+      "loss": 0.2868,
+      "step": 2945
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00019980267284282717,
+      "loss": 0.2906,
+      "step": 2950
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019979826491665881,
+      "loss": 0.2916,
+      "step": 2955
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00019979380835002846,
+      "loss": 0.2886,
+      "step": 2960
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00019978930314510824,
+      "loss": 0.2882,
+      "step": 2965
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019978474930409395,
+      "loss": 0.2872,
+      "step": 2970
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.000199780146829205,
+      "loss": 0.2895,
+      "step": 2975
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019977549572268468,
+      "loss": 0.29,
+      "step": 2980
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00019977079598679977,
+      "loss": 0.2871,
+      "step": 2985
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.000199766047623841,
+      "loss": 0.2909,
+      "step": 2990
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00019976125063612252,
+      "loss": 0.2885,
+      "step": 2995
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00019975640502598244,
+      "loss": 0.291,
+      "step": 3000
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019975151079578237,
+      "loss": 0.2893,
+      "step": 3005
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00019974656794790775,
+      "loss": 0.2893,
+      "step": 3010
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.00019974157648476766,
+      "loss": 0.2888,
+      "step": 3015
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019973653640879485,
+      "loss": 0.2888,
+      "step": 3020
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00019973144772244582,
+      "loss": 0.2883,
+      "step": 3025
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0001997263104282007,
+      "loss": 0.2871,
+      "step": 3030
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00019972112452856339,
+      "loss": 0.2873,
+      "step": 3035
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0001997158900260614,
+      "loss": 0.2868,
+      "step": 3040
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019971060692324598,
+      "loss": 0.2872,
+      "step": 3045
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.00019970527522269205,
+      "loss": 0.2882,
+      "step": 3050
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0001996998949269982,
+      "loss": 0.2894,
+      "step": 3055
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00019969446603878673,
+      "loss": 0.2865,
+      "step": 3060
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0001996889885607036,
+      "loss": 0.2881,
+      "step": 3065
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.00019968346249541846,
+      "loss": 0.2874,
+      "step": 3070
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.00019967788784562473,
+      "loss": 0.286,
+      "step": 3075
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.00019967226461403933,
+      "loss": 0.2902,
+      "step": 3080
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.00019966659280340297,
+      "loss": 0.2897,
+      "step": 3085
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0001996608724164801,
+      "loss": 0.2873,
+      "step": 3090
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.00019965510345605866,
+      "loss": 0.2864,
+      "step": 3095
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00019964928592495045,
+      "loss": 0.2863,
+      "step": 3100
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00019964341982599078,
+      "loss": 0.288,
+      "step": 3105
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.00019963750516203884,
+      "loss": 0.2888,
+      "step": 3110
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.00019963154193597727,
+      "loss": 0.2871,
+      "step": 3115
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0001996255301507125,
+      "loss": 0.2881,
+      "step": 3120
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00019961946980917456,
+      "loss": 0.2869,
+      "step": 3125
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00019961336091431727,
+      "loss": 0.2886,
+      "step": 3130
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00019960720346911797,
+      "loss": 0.2895,
+      "step": 3135
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.00019960099747657774,
+      "loss": 0.2893,
+      "step": 3140
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019959474293972129,
+      "loss": 0.2905,
+      "step": 3145
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00019958843986159704,
+      "loss": 0.2903,
+      "step": 3150
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.000199582088245277,
+      "loss": 0.2911,
+      "step": 3155
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.00019957568809385694,
+      "loss": 0.2891,
+      "step": 3160
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001995692394104561,
+      "loss": 0.2889,
+      "step": 3165
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.00019956274219821757,
+      "loss": 0.2906,
+      "step": 3170
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00019955619646030802,
+      "loss": 0.2881,
+      "step": 3175
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001995496021999177,
+      "loss": 0.2901,
+      "step": 3180
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00019954295942026064,
+      "loss": 0.2876,
+      "step": 3185
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0001995362681245744,
+      "loss": 0.2891,
+      "step": 3190
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.00019952952831612026,
+      "loss": 0.2881,
+      "step": 3195
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0001995227399981831,
+      "loss": 0.2898,
+      "step": 3200
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.00019951590317407152,
+      "loss": 0.2878,
+      "step": 3205
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00019950901784711764,
+      "loss": 0.2875,
+      "step": 3210
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.00019950208402067733,
+      "loss": 0.289,
+      "step": 3215
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.00019949510169813003,
+      "loss": 0.2914,
+      "step": 3220
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.00019948807088287883,
+      "loss": 0.29,
+      "step": 3225
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.00019948099157835047,
+      "loss": 0.2897,
+      "step": 3230
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00019947386378799532,
+      "loss": 0.2879,
+      "step": 3235
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00019946668751528744,
+      "loss": 0.2925,
+      "step": 3240
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00019945946276372434,
+      "loss": 0.2879,
+      "step": 3245
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00019945218953682734,
+      "loss": 0.29,
+      "step": 3250
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00019944486783814134,
+      "loss": 0.2896,
+      "step": 3255
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0001994374976712348,
+      "loss": 0.2888,
+      "step": 3260
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0001994300790396999,
+      "loss": 0.2908,
+      "step": 3265
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00019942261194715236,
+      "loss": 0.2898,
+      "step": 3270
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00019941509639723155,
+      "loss": 0.2947,
+      "step": 3275
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00019940753239360047,
+      "loss": 0.2922,
+      "step": 3280
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001993999199399457,
+      "loss": 0.2903,
+      "step": 3285
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0001993922590399775,
+      "loss": 0.2883,
+      "step": 3290
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00019938454969742968,
+      "loss": 0.2889,
+      "step": 3295
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019937679191605963,
+      "loss": 0.2895,
+      "step": 3300
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019936898569964848,
+      "loss": 0.2891,
+      "step": 3305
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00019936113105200085,
+      "loss": 0.2902,
+      "step": 3310
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.000199353227976945,
+      "loss": 0.2887,
+      "step": 3315
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00019934527647833276,
+      "loss": 0.2892,
+      "step": 3320
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00019933727656003963,
+      "loss": 0.2924,
+      "step": 3325
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.00019932922822596473,
+      "loss": 0.2896,
+      "step": 3330
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.00019932113148003058,
+      "loss": 0.2892,
+      "step": 3335
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00019931298632618356,
+      "loss": 0.2896,
+      "step": 3340
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.00019930479276839344,
+      "loss": 0.2911,
+      "step": 3345
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0001992965508106537,
+      "loss": 0.2881,
+      "step": 3350
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.00019928826045698136,
+      "loss": 0.2869,
+      "step": 3355
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019927992171141708,
+      "loss": 0.2895,
+      "step": 3360
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.000199271534578025,
+      "loss": 0.2896,
+      "step": 3365
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0001992630990608929,
+      "loss": 0.2867,
+      "step": 3370
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019925461516413223,
+      "loss": 0.2873,
+      "step": 3375
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.00019924608289187786,
+      "loss": 0.2884,
+      "step": 3380
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00019923750224828832,
+      "loss": 0.2894,
+      "step": 3385
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00019922887323754577,
+      "loss": 0.2893,
+      "step": 3390
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.00019922019586385585,
+      "loss": 0.2884,
+      "step": 3395
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0001992114701314478,
+      "loss": 0.2876,
+      "step": 3400
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.00019920269604457446,
+      "loss": 0.2888,
+      "step": 3405
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00019919387360751217,
+      "loss": 0.2897,
+      "step": 3410
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0001991850028245609,
+      "loss": 0.2871,
+      "step": 3415
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019917608370004417,
+      "loss": 0.2902,
+      "step": 3420
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00019916711623830903,
+      "loss": 0.2877,
+      "step": 3425
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.00019915810044372618,
+      "loss": 0.2896,
+      "step": 3430
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.00019914903632068973,
+      "loss": 0.2864,
+      "step": 3435
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019913992387361745,
+      "loss": 0.2886,
+      "step": 3440
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00019913076310695068,
+      "loss": 0.2907,
+      "step": 3445
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.00019912155402515417,
+      "loss": 0.288,
+      "step": 3450
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0001991122966327164,
+      "loss": 0.2891,
+      "step": 3455
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0001991029909341493,
+      "loss": 0.2896,
+      "step": 3460
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.00019909363693398828,
+      "loss": 0.2867,
+      "step": 3465
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00019908423463679248,
+      "loss": 0.2879,
+      "step": 3470
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00019907478404714436,
+      "loss": 0.2889,
+      "step": 3475
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019906528516965008,
+      "loss": 0.2899,
+      "step": 3480
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0001990557380089393,
+      "loss": 0.2876,
+      "step": 3485
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019904614256966512,
+      "loss": 0.2894,
+      "step": 3490
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0001990364988565043,
+      "loss": 0.2872,
+      "step": 3495
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019902680687415705,
+      "loss": 0.2861,
+      "step": 3500
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.00019901706662734712,
+      "loss": 0.2894,
+      "step": 3505
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00019900727812082177,
+      "loss": 0.2878,
+      "step": 3510
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0001989974413593518,
+      "loss": 0.2866,
+      "step": 3515
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00019898755634773158,
+      "loss": 0.2873,
+      "step": 3520
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0001989776230907789,
+      "loss": 0.286,
+      "step": 3525
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0001989676415933351,
+      "loss": 0.2889,
+      "step": 3530
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0001989576118602651,
+      "loss": 0.288,
+      "step": 3535
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.00019894753389645723,
+      "loss": 0.2875,
+      "step": 3540
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.00019893740770682335,
+      "loss": 0.2886,
+      "step": 3545
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00019892723329629887,
+      "loss": 0.2876,
+      "step": 3550
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.00019891701066984262,
+      "loss": 0.2883,
+      "step": 3555
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00019890673983243706,
+      "loss": 0.2889,
+      "step": 3560
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00019889642078908804,
+      "loss": 0.2875,
+      "step": 3565
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0001988860535448249,
+      "loss": 0.288,
+      "step": 3570
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0001988756381047006,
+      "loss": 0.2863,
+      "step": 3575
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0001988651744737914,
+      "loss": 0.2872,
+      "step": 3580
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0001988546626571972,
+      "loss": 0.2865,
+      "step": 3585
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.00019884410266004135,
+      "loss": 0.289,
+      "step": 3590
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00019883349448747062,
+      "loss": 0.2894,
+      "step": 3595
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0001988228381446553,
+      "loss": 0.2894,
+      "step": 3600
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0001988121336367892,
+      "loss": 0.2896,
+      "step": 3605
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00019880138096908952,
+      "loss": 0.2858,
+      "step": 3610
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019879058014679704,
+      "loss": 0.2882,
+      "step": 3615
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001987797311751759,
+      "loss": 0.2893,
+      "step": 3620
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019876883405951377,
+      "loss": 0.2885,
+      "step": 3625
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001987578888051218,
+      "loss": 0.2897,
+      "step": 3630
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00019874689541733457,
+      "loss": 0.2887,
+      "step": 3635
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019873585390151003,
+      "loss": 0.2877,
+      "step": 3640
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.00019872476426302982,
+      "loss": 0.2885,
+      "step": 3645
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0001987136265072988,
+      "loss": 0.2864,
+      "step": 3650
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0001987024406397454,
+      "loss": 0.2893,
+      "step": 3655
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00019869120666582153,
+      "loss": 0.2871,
+      "step": 3660
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0001986799245910024,
+      "loss": 0.2879,
+      "step": 3665
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0001986685944207868,
+      "loss": 0.2913,
+      "step": 3670
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00019865721616069696,
+      "loss": 0.2871,
+      "step": 3675
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00019864578981627844,
+      "loss": 0.2903,
+      "step": 3680
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0001986343153931003,
+      "loss": 0.2891,
+      "step": 3685
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.00019862279289675509,
+      "loss": 0.2886,
+      "step": 3690
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0001986112223328587,
+      "loss": 0.2892,
+      "step": 3695
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0001985996037070505,
+      "loss": 0.2872,
+      "step": 3700
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019858793702499323,
+      "loss": 0.2881,
+      "step": 3705
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00019857622229237313,
+      "loss": 0.2876,
+      "step": 3710
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.00019856445951489982,
+      "loss": 0.2872,
+      "step": 3715
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.00019855264869830629,
+      "loss": 0.2873,
+      "step": 3720
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00019854078984834903,
+      "loss": 0.2906,
+      "step": 3725
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.00019852888297080786,
+      "loss": 0.2884,
+      "step": 3730
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0001985169280714861,
+      "loss": 0.2862,
+      "step": 3735
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00019850492515621038,
+      "loss": 0.288,
+      "step": 3740
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.00019849287423083078,
+      "loss": 0.2876,
+      "step": 3745
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.00019848077530122083,
+      "loss": 0.2889,
+      "step": 3750
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0001984686283732773,
+      "loss": 0.2861,
+      "step": 3755
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00019845643345292054,
+      "loss": 0.2863,
+      "step": 3760
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0001984441905460942,
+      "loss": 0.2873,
+      "step": 3765
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00019843189965876526,
+      "loss": 0.2898,
+      "step": 3770
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0001984195607969242,
+      "loss": 0.2879,
+      "step": 3775
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00019840717396658484,
+      "loss": 0.2882,
+      "step": 3780
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.00019839473917378434,
+      "loss": 0.2877,
+      "step": 3785
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00019838225642458327,
+      "loss": 0.2878,
+      "step": 3790
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.00019836972572506557,
+      "loss": 0.2861,
+      "step": 3795
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.00019835714708133862,
+      "loss": 0.288,
+      "step": 3800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00019834452049953297,
+      "loss": 0.2875,
+      "step": 3805
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.00019833184598580276,
+      "loss": 0.2874,
+      "step": 3810
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00019831912354632535,
+      "loss": 0.2891,
+      "step": 3815
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00019830635318730154,
+      "loss": 0.2847,
+      "step": 3820
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.00019829353491495545,
+      "loss": 0.2885,
+      "step": 3825
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00019828066873553448,
+      "loss": 0.289,
+      "step": 3830
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0001982677546553095,
+      "loss": 0.2887,
+      "step": 3835
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.00019825479268057467,
+      "loss": 0.2895,
+      "step": 3840
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00019824178281764753,
+      "loss": 0.2861,
+      "step": 3845
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0001982287250728689,
+      "loss": 0.2881,
+      "step": 3850
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0001982156194526029,
+      "loss": 0.2876,
+      "step": 3855
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0001982024659632372,
+      "loss": 0.2892,
+      "step": 3860
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00019818926461118253,
+      "loss": 0.2877,
+      "step": 3865
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00019817601540287306,
+      "loss": 0.2867,
+      "step": 3870
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00019816271834476642,
+      "loss": 0.2884,
+      "step": 3875
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0001981493734433433,
+      "loss": 0.2861,
+      "step": 3880
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.00019813598070510792,
+      "loss": 0.2873,
+      "step": 3885
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.00019812254013658768,
+      "loss": 0.2859,
+      "step": 3890
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001981090517443334,
+      "loss": 0.2878,
+      "step": 3895
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.00019809551553491916,
+      "loss": 0.2862,
+      "step": 3900
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00019808193151494232,
+      "loss": 0.2892,
+      "step": 3905
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.00019806829969102357,
+      "loss": 0.2884,
+      "step": 3910
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.00019805462006980689,
+      "loss": 0.2873,
+      "step": 3915
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0001980408926579596,
+      "loss": 0.2863,
+      "step": 3920
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00019802711746217218,
+      "loss": 0.2861,
+      "step": 3925
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.00019801329448915862,
+      "loss": 0.2885,
+      "step": 3930
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00019799942374565597,
+      "loss": 0.2879,
+      "step": 3935
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0001979855052384247,
+      "loss": 0.2892,
+      "step": 3940
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.00019797153897424852,
+      "loss": 0.288,
+      "step": 3945
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0001979575249599344,
+      "loss": 0.2848,
+      "step": 3950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019794346320231265,
+      "loss": 0.2882,
+      "step": 3955
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00019792935370823675,
+      "loss": 0.288,
+      "step": 3960
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019791519648458352,
+      "loss": 0.2878,
+      "step": 3965
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.000197900991538253,
+      "loss": 0.2873,
+      "step": 3970
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0001978867388761685,
+      "loss": 0.289,
+      "step": 3975
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.00019787243850527664,
+      "loss": 0.2911,
+      "step": 3980
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00019785809043254722,
+      "loss": 0.2875,
+      "step": 3985
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001978436946649733,
+      "loss": 0.2891,
+      "step": 3990
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00019782925120957124,
+      "loss": 0.2879,
+      "step": 3995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00019781476007338058,
+      "loss": 0.2875,
+      "step": 4000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0001978002212634641,
+      "loss": 0.286,
+      "step": 4005
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0001977856347869079,
+      "loss": 0.2873,
+      "step": 4010
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.00019777100065082118,
+      "loss": 0.2874,
+      "step": 4015
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.00019775631886233654,
+      "loss": 0.2862,
+      "step": 4020
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0001977415894286096,
+      "loss": 0.2877,
+      "step": 4025
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.00019772681235681936,
+      "loss": 0.2871,
+      "step": 4030
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.000197711987654168,
+      "loss": 0.2861,
+      "step": 4035
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.00019769711532788083,
+      "loss": 0.2872,
+      "step": 4040
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0001976821953852065,
+      "loss": 0.2872,
+      "step": 4045
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0001976672278334168,
+      "loss": 0.2872,
+      "step": 4050
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.00019765221267980675,
+      "loss": 0.2881,
+      "step": 4055
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.00019763714993169452,
+      "loss": 0.2889,
+      "step": 4060
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0001976220395964215,
+      "loss": 0.2885,
+      "step": 4065
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.00019760688168135232,
+      "loss": 0.2894,
+      "step": 4070
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.00019759167619387476,
+      "loss": 0.2887,
+      "step": 4075
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.00019757642314139977,
+      "loss": 0.2863,
+      "step": 4080
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0001975611225313615,
+      "loss": 0.2891,
+      "step": 4085
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00019754577437121733,
+      "loss": 0.2863,
+      "step": 4090
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.00019753037866844771,
+      "loss": 0.288,
+      "step": 4095
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.00019751493543055632,
+      "loss": 0.2862,
+      "step": 4100
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00019749944466507008,
+      "loss": 0.2874,
+      "step": 4105
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0001974839063795389,
+      "loss": 0.2888,
+      "step": 4110
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.00019746832058153602,
+      "loss": 0.2866,
+      "step": 4115
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.00019745268727865774,
+      "loss": 0.287,
+      "step": 4120
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.00019743700647852354,
+      "loss": 0.2855,
+      "step": 4125
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.00019742127818877606,
+      "loss": 0.2863,
+      "step": 4130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00019740550241708108,
+      "loss": 0.2875,
+      "step": 4135
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0001973896791711275,
+      "loss": 0.2869,
+      "step": 4140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.00019737380845862745,
+      "loss": 0.2873,
+      "step": 4145
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00019735789028731604,
+      "loss": 0.2876,
+      "step": 4150
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00019734192466495162,
+      "loss": 0.2873,
+      "step": 4155
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0001973259115993156,
+      "loss": 0.2868,
+      "step": 4160
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00019730985109821266,
+      "loss": 0.2871,
+      "step": 4165
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0001972937431694704,
+      "loss": 0.2881,
+      "step": 4170
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019727758782093967,
+      "loss": 0.2882,
+      "step": 4175
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00019726138506049438,
+      "loss": 0.2878,
+      "step": 4180
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.00019724513489603155,
+      "loss": 0.2869,
+      "step": 4185
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.00019722883733547128,
+      "loss": 0.2855,
+      "step": 4190
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.00019721249238675688,
+      "loss": 0.2872,
+      "step": 4195
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00019719610005785465,
+      "loss": 0.2861,
+      "step": 4200
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00019717966035675397,
+      "loss": 0.2867,
+      "step": 4205
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0001971631732914674,
+      "loss": 0.2889,
+      "step": 4210
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.00019714663887003054,
+      "loss": 0.2873,
+      "step": 4215
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.000197130057100502,
+      "loss": 0.2866,
+      "step": 4220
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.00019711342799096361,
+      "loss": 0.2884,
+      "step": 4225
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019709675154952017,
+      "loss": 0.2897,
+      "step": 4230
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.00019708002778429955,
+      "loss": 0.2886,
+      "step": 4235
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00019706325670345275,
+      "loss": 0.2893,
+      "step": 4240
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00019704643831515374,
+      "loss": 0.2894,
+      "step": 4245
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00019702957262759965,
+      "loss": 0.2899,
+      "step": 4250
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0001970126596490106,
+      "loss": 0.2901,
+      "step": 4255
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.00019699569938762973,
+      "loss": 0.29,
+      "step": 4260
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00019697869185172331,
+      "loss": 0.2874,
+      "step": 4265
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0001969616370495806,
+      "loss": 0.2886,
+      "step": 4270
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0001969445349895139,
+      "loss": 0.2876,
+      "step": 4275
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00019692738567985853,
+      "loss": 0.2876,
+      "step": 4280
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00019691018912897286,
+      "loss": 0.2889,
+      "step": 4285
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0001968929453452383,
+      "loss": 0.2872,
+      "step": 4290
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00019687565433705926,
+      "loss": 0.285,
+      "step": 4295
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001968583161128631,
+      "loss": 0.2864,
+      "step": 4300
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00019684093068110038,
+      "loss": 0.2873,
+      "step": 4305
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00019682349805024446,
+      "loss": 0.2897,
+      "step": 4310
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.00019680601822879182,
+      "loss": 0.2862,
+      "step": 4315
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00019678849122526187,
+      "loss": 0.2864,
+      "step": 4320
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.00019677091704819715,
+      "loss": 0.2872,
+      "step": 4325
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00019675329570616298,
+      "loss": 0.2861,
+      "step": 4330
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00019673562720774792,
+      "loss": 0.2869,
+      "step": 4335
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0001967179115615633,
+      "loss": 0.2895,
+      "step": 4340
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.00019670014877624353,
+      "loss": 0.2868,
+      "step": 4345
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.00019668233886044597,
+      "loss": 0.2873,
+      "step": 4350
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.00019666448182285094,
+      "loss": 0.287,
+      "step": 4355
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019664657767216176,
+      "loss": 0.2888,
+      "step": 4360
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0001966286264171047,
+      "loss": 0.2871,
+      "step": 4365
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019661062806642903,
+      "loss": 0.2869,
+      "step": 4370
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00019659258262890683,
+      "loss": 0.2869,
+      "step": 4375
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00019657449011333328,
+      "loss": 0.2874,
+      "step": 4380
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00019655635052852647,
+      "loss": 0.2874,
+      "step": 4385
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0001965381638833274,
+      "loss": 0.285,
+      "step": 4390
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0001965199301866,
+      "loss": 0.2871,
+      "step": 4395
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00019650164944723115,
+      "loss": 0.2881,
+      "step": 4400
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00019648332167413067,
+      "loss": 0.2859,
+      "step": 4405
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.00019646494687623135,
+      "loss": 0.2864,
+      "step": 4410
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.00019644652506248874,
+      "loss": 0.2845,
+      "step": 4415
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.00019642805624188147,
+      "loss": 0.2872,
+      "step": 4420
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00019640954042341103,
+      "loss": 0.2877,
+      "step": 4425
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00019639097761610174,
+      "loss": 0.2901,
+      "step": 4430
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.000196372367829001,
+      "loss": 0.2857,
+      "step": 4435
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.00019635371107117888,
+      "loss": 0.2882,
+      "step": 4440
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0001963350073517285,
+      "loss": 0.2861,
+      "step": 4445
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00019631625667976583,
+      "loss": 0.2888,
+      "step": 4450
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0001962974590644297,
+      "loss": 0.2885,
+      "step": 4455
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.00019627861451488189,
+      "loss": 0.2857,
+      "step": 4460
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00019625972304030697,
+      "loss": 0.2857,
+      "step": 4465
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0001962407846499124,
+      "loss": 0.2871,
+      "step": 4470
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019622179935292855,
+      "loss": 0.2899,
+      "step": 4475
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0001962027671586086,
+      "loss": 0.2866,
+      "step": 4480
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00019618368807622862,
+      "loss": 0.2871,
+      "step": 4485
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.00019616456211508752,
+      "loss": 0.2847,
+      "step": 4490
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001961453892845071,
+      "loss": 0.2862,
+      "step": 4495
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0001961261695938319,
+      "loss": 0.2864,
+      "step": 4500
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0001961069030524294,
+      "loss": 0.288,
+      "step": 4505
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019608758966968988,
+      "loss": 0.2891,
+      "step": 4510
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0001960682294550264,
+      "loss": 0.2861,
+      "step": 4515
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.00019604882241787498,
+      "loss": 0.2897,
+      "step": 4520
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0001960293685676943,
+      "loss": 0.286,
+      "step": 4525
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.000196009867913966,
+      "loss": 0.2861,
+      "step": 4530
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019599032046619438,
+      "loss": 0.2869,
+      "step": 4535
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.00019597072623390668,
+      "loss": 0.2867,
+      "step": 4540
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0001959510852266529,
+      "loss": 0.2869,
+      "step": 4545
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.00019593139745400576,
+      "loss": 0.287,
+      "step": 4550
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0001959116629255609,
+      "loss": 0.2872,
+      "step": 4555
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0001958918816509367,
+      "loss": 0.2878,
+      "step": 4560
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.00019587205363977427,
+      "loss": 0.2863,
+      "step": 4565
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0001958521789017376,
+      "loss": 0.2877,
+      "step": 4570
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.00019583225744651333,
+      "loss": 0.2866,
+      "step": 4575
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.00019581228928381098,
+      "loss": 0.2867,
+      "step": 4580
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.00019579227442336278,
+      "loss": 0.2872,
+      "step": 4585
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00019577221287492367,
+      "loss": 0.2878,
+      "step": 4590
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.00019575210464827149,
+      "loss": 0.2859,
+      "step": 4595
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00019573194975320673,
+      "loss": 0.2862,
+      "step": 4600
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.00019571174819955263,
+      "loss": 0.2868,
+      "step": 4605
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00019569149999715515,
+      "loss": 0.2865,
+      "step": 4610
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.00019567120515588308,
+      "loss": 0.2868,
+      "step": 4615
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0001956508636856278,
+      "loss": 0.2856,
+      "step": 4620
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00019563047559630357,
+      "loss": 0.2848,
+      "step": 4625
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.00019561004089784723,
+      "loss": 0.2854,
+      "step": 4630
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.00019558955960021849,
+      "loss": 0.2859,
+      "step": 4635
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00019556903171339963,
+      "loss": 0.2871,
+      "step": 4640
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.00019554845724739566,
+      "loss": 0.2858,
+      "step": 4645
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.00019552783621223436,
+      "loss": 0.2869,
+      "step": 4650
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.00019550716861796623,
+      "loss": 0.2867,
+      "step": 4655
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00019548645447466431,
+      "loss": 0.2869,
+      "step": 4660
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00019546569379242444,
+      "loss": 0.2879,
+      "step": 4665
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.00019544488658136523,
+      "loss": 0.2867,
+      "step": 4670
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0001954240328516277,
+      "loss": 0.2867,
+      "step": 4675
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0001954031326133758,
+      "loss": 0.2862,
+      "step": 4680
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.00019538218587679605,
+      "loss": 0.2871,
+      "step": 4685
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0001953611926520976,
+      "loss": 0.2875,
+      "step": 4690
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.00019534015294951233,
+      "loss": 0.2865,
+      "step": 4695
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0001953190667792947,
+      "loss": 0.2875,
+      "step": 4700
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00019529793415172192,
+      "loss": 0.2922,
+      "step": 4705
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00019527675507709366,
+      "loss": 0.2868,
+      "step": 4710
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00019525552956573244,
+      "loss": 0.2868,
+      "step": 4715
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00019523425762798329,
+      "loss": 0.2878,
+      "step": 4720
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.00019521293927421388,
+      "loss": 0.2875,
+      "step": 4725
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.00019519157451481454,
+      "loss": 0.2875,
+      "step": 4730
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0001951701633601982,
+      "loss": 0.2871,
+      "step": 4735
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.00019514870582080032,
+      "loss": 0.2887,
+      "step": 4740
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.00019512720190707913,
+      "loss": 0.2857,
+      "step": 4745
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.00019510565162951537,
+      "loss": 0.2869,
+      "step": 4750
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00019508405499861232,
+      "loss": 0.288,
+      "step": 4755
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.00019506241202489602,
+      "loss": 0.288,
+      "step": 4760
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.00019504072271891488,
+      "loss": 0.2879,
+      "step": 4765
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00019501898709124008,
+      "loss": 0.2879,
+      "step": 4770
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.00019499720515246525,
+      "loss": 0.2865,
+      "step": 4775
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.00019497537691320668,
+      "loss": 0.2887,
+      "step": 4780
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0001949535023841032,
+      "loss": 0.2869,
+      "step": 4785
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00019493158157581615,
+      "loss": 0.2851,
+      "step": 4790
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.00019490961449902946,
+      "loss": 0.2872,
+      "step": 4795
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.00019488760116444966,
+      "loss": 0.2859,
+      "step": 4800
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.00019486554158280574,
+      "loss": 0.289,
+      "step": 4805
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.00019484343576484933,
+      "loss": 0.2868,
+      "step": 4810
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.00019482128372135446,
+      "loss": 0.2865,
+      "step": 4815
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.00019479908546311781,
+      "loss": 0.2878,
+      "step": 4820
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0001947768410009586,
+      "loss": 0.2891,
+      "step": 4825
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.00019475455034571838,
+      "loss": 0.2869,
+      "step": 4830
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.00019473221350826142,
+      "loss": 0.2887,
+      "step": 4835
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.00019470983049947444,
+      "loss": 0.2883,
+      "step": 4840
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0001946874013302666,
+      "loss": 0.2853,
+      "step": 4845
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.00019466492601156966,
+      "loss": 0.2875,
+      "step": 4850
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00019464240455433775,
+      "loss": 0.2869,
+      "step": 4855
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.00019461983696954758,
+      "loss": 0.288,
+      "step": 4860
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.00019459722326819838,
+      "loss": 0.2867,
+      "step": 4865
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0001945745634613117,
+      "loss": 0.2859,
+      "step": 4870
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0001945518575599317,
+      "loss": 0.2878,
+      "step": 4875
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.00019452910557512496,
+      "loss": 0.2875,
+      "step": 4880
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.00019450630751798048,
+      "loss": 0.2842,
+      "step": 4885
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00019448346339960982,
+      "loss": 0.2873,
+      "step": 4890
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0001944605732311469,
+      "loss": 0.2881,
+      "step": 4895
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00019443763702374812,
+      "loss": 0.2879,
+      "step": 4900
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.00019441465478859228,
+      "loss": 0.2861,
+      "step": 4905
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.00019439162653688065,
+      "loss": 0.2874,
+      "step": 4910
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.00019436855227983695,
+      "loss": 0.2869,
+      "step": 4915
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.00019434543202870725,
+      "loss": 0.2883,
+      "step": 4920
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0001943222657947601,
+      "loss": 0.2866,
+      "step": 4925
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.00019429905358928646,
+      "loss": 0.2866,
+      "step": 4930
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.00019427579542359965,
+      "loss": 0.287,
+      "step": 4935
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00019425249130903543,
+      "loss": 0.2873,
+      "step": 4940
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0001942291412569519,
+      "loss": 0.2858,
+      "step": 4945
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.00019420574527872968,
+      "loss": 0.2865,
+      "step": 4950
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0001941823033857716,
+      "loss": 0.2861,
+      "step": 4955
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.00019415881558950302,
+      "loss": 0.2871,
+      "step": 4960
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0001941352819013716,
+      "loss": 0.2875,
+      "step": 4965
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.00019411170233284727,
+      "loss": 0.2867,
+      "step": 4970
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.00019408807689542257,
+      "loss": 0.2864,
+      "step": 4975
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.00019406440560061216,
+      "loss": 0.2879,
+      "step": 4980
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.00019404068845995317,
+      "loss": 0.2857,
+      "step": 4985
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.00019401692548500502,
+      "loss": 0.286,
+      "step": 4990
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.00019399311668734956,
+      "loss": 0.2883,
+      "step": 4995
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.00019396926207859084,
+      "loss": 0.2871,
+      "step": 5000
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.00019394536167035534,
+      "loss": 0.2858,
+      "step": 5005
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00019392141547429183,
+      "loss": 0.2876,
+      "step": 5010
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.00019389742350207141,
+      "loss": 0.288,
+      "step": 5015
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00019387338576538744,
+      "loss": 0.2861,
+      "step": 5020
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0001938493022759556,
+      "loss": 0.2876,
+      "step": 5025
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.00019382517304551396,
+      "loss": 0.2859,
+      "step": 5030
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.00019380099808582278,
+      "loss": 0.2876,
+      "step": 5035
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0001937767774086646,
+      "loss": 0.2876,
+      "step": 5040
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0001937525110258444,
+      "loss": 0.2874,
+      "step": 5045
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.00019372819894918915,
+      "loss": 0.289,
+      "step": 5050
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0001937038411905484,
+      "loss": 0.2886,
+      "step": 5055
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0001936794377617938,
+      "loss": 0.2857,
+      "step": 5060
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00019365498867481923,
+      "loss": 0.2879,
+      "step": 5065
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00019363049394154094,
+      "loss": 0.2883,
+      "step": 5070
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00019360595357389735,
+      "loss": 0.2858,
+      "step": 5075
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.00019358136758384912,
+      "loss": 0.2852,
+      "step": 5080
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.00019355673598337914,
+      "loss": 0.2864,
+      "step": 5085
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.00019353205878449258,
+      "loss": 0.2878,
+      "step": 5090
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.00019350733599921683,
+      "loss": 0.286,
+      "step": 5095
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.00019348256763960145,
+      "loss": 0.2875,
+      "step": 5100
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.00019345775371771824,
+      "loss": 0.2862,
+      "step": 5105
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.00019343289424566122,
+      "loss": 0.2857,
+      "step": 5110
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.00019340798923554657,
+      "loss": 0.2874,
+      "step": 5115
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.00019338303869951269,
+      "loss": 0.2862,
+      "step": 5120
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.00019335804264972018,
+      "loss": 0.2855,
+      "step": 5125
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0001933330010983518,
+      "loss": 0.2874,
+      "step": 5130
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.00019330791405761252,
+      "loss": 0.2873,
+      "step": 5135
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00019328278153972947,
+      "loss": 0.2897,
+      "step": 5140
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00019325760355695188,
+      "loss": 0.2867,
+      "step": 5145
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.00019323238012155123,
+      "loss": 0.2872,
+      "step": 5150
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0001932071112458211,
+      "loss": 0.288,
+      "step": 5155
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.00019318179694207725,
+      "loss": 0.2866,
+      "step": 5160
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.00019315643722265757,
+      "loss": 0.2869,
+      "step": 5165
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.00019313103209992204,
+      "loss": 0.2876,
+      "step": 5170
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.00019310558158625285,
+      "loss": 0.2861,
+      "step": 5175
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00019308008569405422,
+      "loss": 0.2875,
+      "step": 5180
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00019305454443575262,
+      "loss": 0.2877,
+      "step": 5185
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0001930289578237965,
+      "loss": 0.2888,
+      "step": 5190
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0001930033258706564,
+      "loss": 0.2874,
+      "step": 5195
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00019297764858882514,
+      "loss": 0.2885,
+      "step": 5200
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.00019295192599081746,
+      "loss": 0.2869,
+      "step": 5205
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00019292615808917026,
+      "loss": 0.2876,
+      "step": 5210
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00019290034489644246,
+      "loss": 0.2866,
+      "step": 5215
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.00019287448642521513,
+      "loss": 0.2865,
+      "step": 5220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00019284858268809137,
+      "loss": 0.2882,
+      "step": 5225
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.00019282263369769633,
+      "loss": 0.2868,
+      "step": 5230
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.00019279663946667727,
+      "loss": 0.2856,
+      "step": 5235
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.00019277060000770342,
+      "loss": 0.2878,
+      "step": 5240
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00019274451533346615,
+      "loss": 0.2872,
+      "step": 5245
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.00019271838545667876,
+      "loss": 0.2888,
+      "step": 5250
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019269221039007665,
+      "loss": 0.2866,
+      "step": 5255
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0001926659901464172,
+      "loss": 0.29,
+      "step": 5260
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00019263972473847993,
+      "loss": 0.2885,
+      "step": 5265
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00019261341417906621,
+      "loss": 0.2896,
+      "step": 5270
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0001925870584809995,
+      "loss": 0.286,
+      "step": 5275
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.00019256065765712522,
+      "loss": 0.2869,
+      "step": 5280
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00019253421172031086,
+      "loss": 0.2883,
+      "step": 5285
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0001925077206834458,
+      "loss": 0.2876,
+      "step": 5290
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0001924811845594415,
+      "loss": 0.2876,
+      "step": 5295
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019245460336123134,
+      "loss": 0.2869,
+      "step": 5300
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0001924279771017706,
+      "loss": 0.2874,
+      "step": 5305
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0001924013057940367,
+      "loss": 0.2869,
+      "step": 5310
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.00019237458945102882,
+      "loss": 0.2855,
+      "step": 5315
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00019234782808576824,
+      "loss": 0.2879,
+      "step": 5320
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00019232102171129811,
+      "loss": 0.2866,
+      "step": 5325
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0001922941703406835,
+      "loss": 0.2879,
+      "step": 5330
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0001922672739870115,
+      "loss": 0.2884,
+      "step": 5335
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00019224033266339102,
+      "loss": 0.2867,
+      "step": 5340
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00019221334638295294,
+      "loss": 0.2879,
+      "step": 5345
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.00019218631515885006,
+      "loss": 0.2845,
+      "step": 5350
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00019215923900425707,
+      "loss": 0.2876,
+      "step": 5355
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.00019213211793237057,
+      "loss": 0.2861,
+      "step": 5360
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.00019210495195640895,
+      "loss": 0.2858,
+      "step": 5365
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.00019207774108961272,
+      "loss": 0.2864,
+      "step": 5370
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.00019205048534524406,
+      "loss": 0.2868,
+      "step": 5375
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.00019202318473658705,
+      "loss": 0.2867,
+      "step": 5380
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.00019199583927694772,
+      "loss": 0.2871,
+      "step": 5385
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.00019196844897965393,
+      "loss": 0.2877,
+      "step": 5390
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0001919410138580553,
+      "loss": 0.2873,
+      "step": 5395
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00019191353392552344,
+      "loss": 0.2863,
+      "step": 5400
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.00019188600919545174,
+      "loss": 0.289,
+      "step": 5405
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0001918584396812554,
+      "loss": 0.2894,
+      "step": 5410
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.00019183082539637146,
+      "loss": 0.2878,
+      "step": 5415
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0001918031663542588,
+      "loss": 0.2861,
+      "step": 5420
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00019177546256839812,
+      "loss": 0.2887,
+      "step": 5425
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.00019174771405229186,
+      "loss": 0.2864,
+      "step": 5430
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.00019171992081946435,
+      "loss": 0.2857,
+      "step": 5435
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00019169208288346166,
+      "loss": 0.2907,
+      "step": 5440
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00019166420025785164,
+      "loss": 0.2874,
+      "step": 5445
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.00019163627295622397,
+      "loss": 0.2869,
+      "step": 5450
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00019160830099219006,
+      "loss": 0.2875,
+      "step": 5455
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00019158028437938317,
+      "loss": 0.2853,
+      "step": 5460
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.00019155222313145816,
+      "loss": 0.2861,
+      "step": 5465
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00019152411726209176,
+      "loss": 0.2862,
+      "step": 5470
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0001914959667849825,
+      "loss": 0.2855,
+      "step": 5475
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0001914677717138505,
+      "loss": 0.2864,
+      "step": 5480
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.00019143953206243776,
+      "loss": 0.2867,
+      "step": 5485
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0001914112478445079,
+      "loss": 0.2875,
+      "step": 5490
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0001913829190738463,
+      "loss": 0.2868,
+      "step": 5495
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0001913545457642601,
+      "loss": 0.2881,
+      "step": 5500
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00019132612792957808,
+      "loss": 0.2861,
+      "step": 5505
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.00019129766558365074,
+      "loss": 0.2897,
+      "step": 5510
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0001912691587403503,
+      "loss": 0.2874,
+      "step": 5515
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.00019124060741357063,
+      "loss": 0.2886,
+      "step": 5520
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0001912120116172273,
+      "loss": 0.2876,
+      "step": 5525
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0001911833713652576,
+      "loss": 0.286,
+      "step": 5530
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00019115468667162038,
+      "loss": 0.2891,
+      "step": 5535
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00019112595755029624,
+      "loss": 0.2869,
+      "step": 5540
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0001910971840152874,
+      "loss": 0.2866,
+      "step": 5545
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.00019106836608061772,
+      "loss": 0.2861,
+      "step": 5550
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00019103950376033276,
+      "loss": 0.2877,
+      "step": 5555
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.00019101059706849957,
+      "loss": 0.2867,
+      "step": 5560
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.000190981646019207,
+      "loss": 0.2872,
+      "step": 5565
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00019095265062656544,
+      "loss": 0.2872,
+      "step": 5570
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00019092361090470688,
+      "loss": 0.2852,
+      "step": 5575
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.00019089452686778488,
+      "loss": 0.2865,
+      "step": 5580
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0001908653985299747,
+      "loss": 0.286,
+      "step": 5585
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.00019083622590547312,
+      "loss": 0.2874,
+      "step": 5590
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00019080700900849851,
+      "loss": 0.2865,
+      "step": 5595
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.00019077774785329087,
+      "loss": 0.285,
+      "step": 5600
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0001907484424541117,
+      "loss": 0.2871,
+      "step": 5605
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.00019071909282524413,
+      "loss": 0.2854,
+      "step": 5610
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0001906896989809927,
+      "loss": 0.2856,
+      "step": 5615
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.00019066026093568378,
+      "loss": 0.2876,
+      "step": 5620
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.000190630778703665,
+      "loss": 0.2857,
+      "step": 5625
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0001906012522993057,
+      "loss": 0.288,
+      "step": 5630
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00019057168173699664,
+      "loss": 0.288,
+      "step": 5635
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0001905420670311502,
+      "loss": 0.2873,
+      "step": 5640
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00019051240819620014,
+      "loss": 0.2856,
+      "step": 5645
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00019048270524660196,
+      "loss": 0.287,
+      "step": 5650
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00019045295819683242,
+      "loss": 0.2861,
+      "step": 5655
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.00019042316706138987,
+      "loss": 0.2866,
+      "step": 5660
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.00019039333185479418,
+      "loss": 0.2892,
+      "step": 5665
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00019036345259158667,
+      "loss": 0.2874,
+      "step": 5670
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0001903335292863301,
+      "loss": 0.2878,
+      "step": 5675
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00019030356195360874,
+      "loss": 0.2866,
+      "step": 5680
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0001902735506080283,
+      "loss": 0.2862,
+      "step": 5685
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.00019024349526421594,
+      "loss": 0.2875,
+      "step": 5690
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.00019021339593682028,
+      "loss": 0.288,
+      "step": 5695
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0001901832526405114,
+      "loss": 0.2858,
+      "step": 5700
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0001901530653899807,
+      "loss": 0.289,
+      "step": 5705
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.00019012283419994115,
+      "loss": 0.2852,
+      "step": 5710
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.000190092559085127,
+      "loss": 0.2871,
+      "step": 5715
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00019006224006029406,
+      "loss": 0.2861,
+      "step": 5720
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.00019003187714021938,
+      "loss": 0.2845,
+      "step": 5725
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.00019000147033970148,
+      "loss": 0.287,
+      "step": 5730
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0001899710196735603,
+      "loss": 0.2881,
+      "step": 5735
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0001899405251566371,
+      "loss": 0.2852,
+      "step": 5740
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00018990998680379456,
+      "loss": 0.2897,
+      "step": 5745
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0001898794046299167,
+      "loss": 0.2864,
+      "step": 5750
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.00018984877864990888,
+      "loss": 0.2878,
+      "step": 5755
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00018981810887869785,
+      "loss": 0.2865,
+      "step": 5760
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0001897873953312317,
+      "loss": 0.2858,
+      "step": 5765
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.00018975663802247976,
+      "loss": 0.2876,
+      "step": 5770
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.00018972583696743285,
+      "loss": 0.2878,
+      "step": 5775
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.000189694992181103,
+      "loss": 0.2868,
+      "step": 5780
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.00018966410367852362,
+      "loss": 0.2874,
+      "step": 5785
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0001896331714747493,
+      "loss": 0.2861,
+      "step": 5790
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0001896021955848561,
+      "loss": 0.2867,
+      "step": 5795
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0001895711760239413,
+      "loss": 0.287,
+      "step": 5800
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0001895401128071234,
+      "loss": 0.2889,
+      "step": 5805
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00018950900594954227,
+      "loss": 0.2865,
+      "step": 5810
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00018947785546635904,
+      "loss": 0.2872,
+      "step": 5815
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.000189446661372756,
+      "loss": 0.2869,
+      "step": 5820
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0001894154236839368,
+      "loss": 0.2873,
+      "step": 5825
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0001893841424151264,
+      "loss": 0.286,
+      "step": 5830
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.00018935281758157078,
+      "loss": 0.2889,
+      "step": 5835
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0001893214491985374,
+      "loss": 0.2891,
+      "step": 5840
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0001892900372813147,
+      "loss": 0.2866,
+      "step": 5845
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.00018925858184521256,
+      "loss": 0.2886,
+      "step": 5850
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.00018922708290556198,
+      "loss": 0.2853,
+      "step": 5855
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0001891955404777151,
+      "loss": 0.286,
+      "step": 5860
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00018916395457704534,
+      "loss": 0.2886,
+      "step": 5865
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.00018913232521894732,
+      "loss": 0.2872,
+      "step": 5870
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0001891006524188368,
+      "loss": 0.2884,
+      "step": 5875
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.00018906893619215066,
+      "loss": 0.2868,
+      "step": 5880
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.00018903717655434707,
+      "loss": 0.2868,
+      "step": 5885
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00018900537352090524,
+      "loss": 0.287,
+      "step": 5890
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.00018897352710732564,
+      "loss": 0.2881,
+      "step": 5895
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.00018894163732912977,
+      "loss": 0.2873,
+      "step": 5900
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00018890970420186033,
+      "loss": 0.2859,
+      "step": 5905
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.00018887772774108116,
+      "loss": 0.2881,
+      "step": 5910
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00018884570796237718,
+      "loss": 0.2846,
+      "step": 5915
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.00018881364488135448,
+      "loss": 0.2878,
+      "step": 5920
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00018878153851364013,
+      "loss": 0.2891,
+      "step": 5925
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.00018874938887488248,
+      "loss": 0.2868,
+      "step": 5930
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0001887171959807508,
+      "loss": 0.2859,
+      "step": 5935
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0001886849598469356,
+      "loss": 0.2871,
+      "step": 5940
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.00018865268048914828,
+      "loss": 0.2864,
+      "step": 5945
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.00018862035792312147,
+      "loss": 0.2888,
+      "step": 5950
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.00018858799216460881,
+      "loss": 0.2872,
+      "step": 5955
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.00018855558322938493,
+      "loss": 0.2879,
+      "step": 5960
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.00018852313113324552,
+      "loss": 0.289,
+      "step": 5965
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00018849063589200743,
+      "loss": 0.2863,
+      "step": 5970
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0001884580975215084,
+      "loss": 0.2878,
+      "step": 5975
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.00018842551603760724,
+      "loss": 0.287,
+      "step": 5980
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00018839289145618378,
+      "loss": 0.2878,
+      "step": 5985
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.00018836022379313883,
+      "loss": 0.283,
+      "step": 5990
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.00018832751306439418,
+      "loss": 0.2872,
+      "step": 5995
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00018829475928589271,
+      "loss": 0.2876,
+      "step": 6000
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00018826196247359817,
+      "loss": 0.2877,
+      "step": 6005
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00018822912264349534,
+      "loss": 0.289,
+      "step": 6010
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.00018819623981158995,
+      "loss": 0.2864,
+      "step": 6015
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0001881633139939087,
+      "loss": 0.2872,
+      "step": 6020
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0001881303452064992,
+      "loss": 0.2863,
+      "step": 6025
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00018809733346543013,
+      "loss": 0.2854,
+      "step": 6030
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00018806427878679093,
+      "loss": 0.2851,
+      "step": 6035
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00018803118118669202,
+      "loss": 0.2867,
+      "step": 6040
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.00018799804068126485,
+      "loss": 0.2865,
+      "step": 6045
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00018796485728666165,
+      "loss": 0.2863,
+      "step": 6050
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.00018793163101905563,
+      "loss": 0.2839,
+      "step": 6055
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.00018789836189464086,
+      "loss": 0.2884,
+      "step": 6060
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001878650499296323,
+      "loss": 0.2855,
+      "step": 6065
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.00018783169514026578,
+      "loss": 0.2879,
+      "step": 6070
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.00018779829754279805,
+      "loss": 0.2873,
+      "step": 6075
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00018776485715350671,
+      "loss": 0.2853,
+      "step": 6080
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00018773137398869015,
+      "loss": 0.2876,
+      "step": 6085
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0001876978480646677,
+      "loss": 0.2887,
+      "step": 6090
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.00018766427939777945,
+      "loss": 0.2875,
+      "step": 6095
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00018763066800438636,
+      "loss": 0.2872,
+      "step": 6100
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.00018759701390087027,
+      "loss": 0.2887,
+      "step": 6105
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00018756331710363374,
+      "loss": 0.2874,
+      "step": 6110
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00018752957762910018,
+      "loss": 0.2864,
+      "step": 6115
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0001874957954937138,
+      "loss": 0.2867,
+      "step": 6120
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.00018746197071393958,
+      "loss": 0.2876,
+      "step": 6125
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.00018742810330626337,
+      "loss": 0.2876,
+      "step": 6130
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0001873941932871917,
+      "loss": 0.2885,
+      "step": 6135
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.00018736024067325188,
+      "loss": 0.2867,
+      "step": 6140
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018732624548099204,
+      "loss": 0.2878,
+      "step": 6145
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.00018729220772698097,
+      "loss": 0.2878,
+      "step": 6150
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00018725812742780834,
+      "loss": 0.2883,
+      "step": 6155
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0001872240046000844,
+      "loss": 0.2876,
+      "step": 6160
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0001871898392604402,
+      "loss": 0.287,
+      "step": 6165
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.00018715563142552758,
+      "loss": 0.2866,
+      "step": 6170
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.00018712138111201895,
+      "loss": 0.2878,
+      "step": 6175
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.00018708708833660754,
+      "loss": 0.2856,
+      "step": 6180
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00018705275311600722,
+      "loss": 0.2887,
+      "step": 6185
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0001870183754669526,
+      "loss": 0.2867,
+      "step": 6190
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0001869839554061988,
+      "loss": 0.2864,
+      "step": 6195
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0001869494929505219,
+      "loss": 0.2859,
+      "step": 6200
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0001869149881167184,
+      "loss": 0.2881,
+      "step": 6205
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.00018688044092160551,
+      "loss": 0.2868,
+      "step": 6210
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00018684585138202122,
+      "loss": 0.2874,
+      "step": 6215
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00018681121951482393,
+      "loss": 0.2853,
+      "step": 6220
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.00018677654533689287,
+      "loss": 0.2884,
+      "step": 6225
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00018674182886512774,
+      "loss": 0.2854,
+      "step": 6230
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.000186707070116449,
+      "loss": 0.2871,
+      "step": 6235
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00018667226910779765,
+      "loss": 0.2885,
+      "step": 6240
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00018663742585613518,
+      "loss": 0.2868,
+      "step": 6245
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.00018660254037844388,
+      "loss": 0.2863,
+      "step": 6250
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.00018656761269172643,
+      "loss": 0.2863,
+      "step": 6255
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.00018653264281300622,
+      "loss": 0.2881,
+      "step": 6260
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00018649763075932708,
+      "loss": 0.2876,
+      "step": 6265
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0001864625765477535,
+      "loss": 0.2884,
+      "step": 6270
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0001864274801953705,
+      "loss": 0.2856,
+      "step": 6275
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.00018639234171928353,
+      "loss": 0.2878,
+      "step": 6280
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.00018635716113661873,
+      "loss": 0.2858,
+      "step": 6285
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0001863219384645227,
+      "loss": 0.2878,
+      "step": 6290
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0001862866737201625,
+      "loss": 0.2885,
+      "step": 6295
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.00018625136692072575,
+      "loss": 0.2857,
+      "step": 6300
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.00018621601808342056,
+      "loss": 0.2875,
+      "step": 6305
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0001861806272254755,
+      "loss": 0.2853,
+      "step": 6310
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0001861451943641397,
+      "loss": 0.2854,
+      "step": 6315
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.00018610971951668265,
+      "loss": 0.2872,
+      "step": 6320
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0001860742027003944,
+      "loss": 0.2876,
+      "step": 6325
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.00018603864393258534,
+      "loss": 0.2862,
+      "step": 6330
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.00018600304323058647,
+      "loss": 0.2867,
+      "step": 6335
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0001859674006117491,
+      "loss": 0.2866,
+      "step": 6340
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.00018593171609344503,
+      "loss": 0.2869,
+      "step": 6345
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.00018589598969306645,
+      "loss": 0.2882,
+      "step": 6350
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.00018586022142802597,
+      "loss": 0.2861,
+      "step": 6355
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0001858244113157566,
+      "loss": 0.2874,
+      "step": 6360
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00018578855937371173,
+      "loss": 0.2858,
+      "step": 6365
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00018575266561936523,
+      "loss": 0.2879,
+      "step": 6370
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00018571673007021123,
+      "loss": 0.288,
+      "step": 6375
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0001856807527437643,
+      "loss": 0.2878,
+      "step": 6380
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.00018564473365755935,
+      "loss": 0.2883,
+      "step": 6385
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0001856086728291516,
+      "loss": 0.2888,
+      "step": 6390
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.00018557257027611675,
+      "loss": 0.2858,
+      "step": 6395
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.00018553642601605068,
+      "loss": 0.2865,
+      "step": 6400
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.00018550024006656966,
+      "loss": 0.2875,
+      "step": 6405
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0001854640124453103,
+      "loss": 0.2871,
+      "step": 6410
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0001854277431699295,
+      "loss": 0.2861,
+      "step": 6415
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0001853914322581045,
+      "loss": 0.2873,
+      "step": 6420
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00018535507972753274,
+      "loss": 0.2886,
+      "step": 6425
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.00018531868559593204,
+      "loss": 0.2858,
+      "step": 6430
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.00018528224988104044,
+      "loss": 0.2875,
+      "step": 6435
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.00018524577260061627,
+      "loss": 0.2873,
+      "step": 6440
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0001852092537724381,
+      "loss": 0.2877,
+      "step": 6445
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.00018517269341430476,
+      "loss": 0.2878,
+      "step": 6450
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.00018513609154403534,
+      "loss": 0.2874,
+      "step": 6455
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.00018509944817946922,
+      "loss": 0.2856,
+      "step": 6460
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.00018506276333846579,
+      "loss": 0.2886,
+      "step": 6465
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00018502603703890488,
+      "loss": 0.2866,
+      "step": 6470
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.00018498926929868642,
+      "loss": 0.2851,
+      "step": 6475
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.00018495246013573054,
+      "loss": 0.2869,
+      "step": 6480
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00018491560956797765,
+      "loss": 0.2874,
+      "step": 6485
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0001848787176133882,
+      "loss": 0.2894,
+      "step": 6490
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0001848417842899429,
+      "loss": 0.2883,
+      "step": 6495
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0001848048096156426,
+      "loss": 0.2867,
+      "step": 6500
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00018476779360850832,
+      "loss": 0.2883,
+      "step": 6505
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0001847307362865812,
+      "loss": 0.2878,
+      "step": 6510
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00018469363766792255,
+      "loss": 0.2855,
+      "step": 6515
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0001846564977706138,
+      "loss": 0.2858,
+      "step": 6520
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00018461931661275643,
+      "loss": 0.2868,
+      "step": 6525
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.00018458209421247208,
+      "loss": 0.2838,
+      "step": 6530
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00018454483058790255,
+      "loss": 0.2862,
+      "step": 6535
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00018450752575720967,
+      "loss": 0.2868,
+      "step": 6540
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00018447017973857532,
+      "loss": 0.2854,
+      "step": 6545
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00018443279255020152,
+      "loss": 0.2882,
+      "step": 6550
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00018439536421031033,
+      "loss": 0.2856,
+      "step": 6555
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0001843578947371439,
+      "loss": 0.2883,
+      "step": 6560
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00018432038414896434,
+      "loss": 0.2871,
+      "step": 6565
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0001842828324640539,
+      "loss": 0.288,
+      "step": 6570
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00018424523970071477,
+      "loss": 0.286,
+      "step": 6575
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.00018420760587726923,
+      "loss": 0.2855,
+      "step": 6580
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.00018416993101205958,
+      "loss": 0.2877,
+      "step": 6585
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.00018413221512344805,
+      "loss": 0.2853,
+      "step": 6590
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.00018409445822981693,
+      "loss": 0.2875,
+      "step": 6595
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.00018405666034956844,
+      "loss": 0.2887,
+      "step": 6600
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.00018401882150112484,
+      "loss": 0.2875,
+      "step": 6605
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0001839809417029283,
+      "loss": 0.2876,
+      "step": 6610
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.000183943020973441,
+      "loss": 0.2862,
+      "step": 6615
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.000183905059331145,
+      "loss": 0.285,
+      "step": 6620
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.00018386705679454242,
+      "loss": 0.287,
+      "step": 6625
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00018382901338215516,
+      "loss": 0.2855,
+      "step": 6630
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00018379092911252514,
+      "loss": 0.2864,
+      "step": 6635
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0001837528040042142,
+      "loss": 0.285,
+      "step": 6640
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.000183714638075804,
+      "loss": 0.2861,
+      "step": 6645
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00018367643134589617,
+      "loss": 0.2886,
+      "step": 6650
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.00018363818383311225,
+      "loss": 0.2875,
+      "step": 6655
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.00018359989555609353,
+      "loss": 0.2876,
+      "step": 6660
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00018356156653350137,
+      "loss": 0.2881,
+      "step": 6665
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.00018352319678401676,
+      "loss": 0.286,
+      "step": 6670
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.00018348478632634066,
+      "loss": 0.2871,
+      "step": 6675
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.00018344633517919392,
+      "loss": 0.2868,
+      "step": 6680
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.00018340784336131713,
+      "loss": 0.2864,
+      "step": 6685
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.00018336931089147073,
+      "loss": 0.2875,
+      "step": 6690
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.000183330737788435,
+      "loss": 0.2873,
+      "step": 6695
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.00018329212407100994,
+      "loss": 0.2882,
+      "step": 6700
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0001832534697580155,
+      "loss": 0.2855,
+      "step": 6705
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.00018321477486829126,
+      "loss": 0.2872,
+      "step": 6710
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.00018317603942069664,
+      "loss": 0.2876,
+      "step": 6715
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.00018313726343411086,
+      "loss": 0.287,
+      "step": 6720
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.00018309844692743283,
+      "loss": 0.2859,
+      "step": 6725
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00018305958991958127,
+      "loss": 0.2879,
+      "step": 6730
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0001830206924294946,
+      "loss": 0.2871,
+      "step": 6735
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00018298175447613096,
+      "loss": 0.2853,
+      "step": 6740
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.00018294277607846832,
+      "loss": 0.2862,
+      "step": 6745
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.00018290375725550417,
+      "loss": 0.289,
+      "step": 6750
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.00018286469802625589,
+      "loss": 0.2866,
+      "step": 6755
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.00018282559840976042,
+      "loss": 0.2859,
+      "step": 6760
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00018278645842507448,
+      "loss": 0.2866,
+      "step": 6765
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00018274727809127438,
+      "loss": 0.2865,
+      "step": 6770
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.00018270805742745617,
+      "loss": 0.2857,
+      "step": 6775
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.00018266879645273556,
+      "loss": 0.2867,
+      "step": 6780
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0001826294951862478,
+      "loss": 0.2874,
+      "step": 6785
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.00018259015364714787,
+      "loss": 0.2868,
+      "step": 6790
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.00018255077185461038,
+      "loss": 0.2856,
+      "step": 6795
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00018251134982782952,
+      "loss": 0.2881,
+      "step": 6800
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0001824718875860191,
+      "loss": 0.2871,
+      "step": 6805
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0001824323851484126,
+      "loss": 0.2869,
+      "step": 6810
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.00018239284253426295,
+      "loss": 0.2869,
+      "step": 6815
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.00018235325976284275,
+      "loss": 0.2858,
+      "step": 6820
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0001823136368534442,
+      "loss": 0.2883,
+      "step": 6825
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.000182273973825379,
+      "loss": 0.2855,
+      "step": 6830
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.00018223427069797844,
+      "loss": 0.2851,
+      "step": 6835
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0001821945274905933,
+      "loss": 0.2862,
+      "step": 6840
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.00018215474422259402,
+      "loss": 0.2861,
+      "step": 6845
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.00018211492091337042,
+      "loss": 0.2873,
+      "step": 6850
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0001820750575823319,
+      "loss": 0.2862,
+      "step": 6855
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0001820351542489074,
+      "loss": 0.2862,
+      "step": 6860
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.00018199521093254523,
+      "loss": 0.2879,
+      "step": 6865
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0001819552276527134,
+      "loss": 0.2871,
+      "step": 6870
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0001819152044288992,
+      "loss": 0.2857,
+      "step": 6875
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.00018187514128060946,
+      "loss": 0.2874,
+      "step": 6880
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0001818350382273705,
+      "loss": 0.2851,
+      "step": 6885
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00018179489528872807,
+      "loss": 0.2874,
+      "step": 6890
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0001817547124842473,
+      "loss": 0.287,
+      "step": 6895
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.00018171448983351284,
+      "loss": 0.2871,
+      "step": 6900
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.00018167422735612877,
+      "loss": 0.2893,
+      "step": 6905
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.00018163392507171842,
+      "loss": 0.2861,
+      "step": 6910
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.00018159358299992467,
+      "loss": 0.2848,
+      "step": 6915
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.00018155320116040982,
+      "loss": 0.2854,
+      "step": 6920
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00018151277957285543,
+      "loss": 0.2859,
+      "step": 6925
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.00018147231825696252,
+      "loss": 0.2847,
+      "step": 6930
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0001814318172324514,
+      "loss": 0.2895,
+      "step": 6935
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.00018139127651906184,
+      "loss": 0.2867,
+      "step": 6940
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0001813506961365528,
+      "loss": 0.2895,
+      "step": 6945
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00018131007610470276,
+      "loss": 0.2874,
+      "step": 6950
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0001812694164433094,
+      "loss": 0.2866,
+      "step": 6955
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0001812287171721897,
+      "loss": 0.2863,
+      "step": 6960
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0001811879783111801,
+      "loss": 0.2872,
+      "step": 6965
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00018114719988013612,
+      "loss": 0.2863,
+      "step": 6970
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.00018110638189893267,
+      "loss": 0.286,
+      "step": 6975
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.000181065524387464,
+      "loss": 0.2906,
+      "step": 6980
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00018102462736564355,
+      "loss": 0.2868,
+      "step": 6985
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.00018098369085340398,
+      "loss": 0.2894,
+      "step": 6990
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.00018094271487069735,
+      "loss": 0.286,
+      "step": 6995
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00018090169943749476,
+      "loss": 0.2863,
+      "step": 7000
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.00018086064457378665,
+      "loss": 0.2872,
+      "step": 7005
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.00018081955029958274,
+      "loss": 0.288,
+      "step": 7010
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00018077841663491175,
+      "loss": 0.2853,
+      "step": 7015
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.00018073724359982186,
+      "loss": 0.2863,
+      "step": 7020
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.00018069603121438022,
+      "loss": 0.286,
+      "step": 7025
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00018065477949867327,
+      "loss": 0.2855,
+      "step": 7030
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0001806134884728066,
+      "loss": 0.2868,
+      "step": 7035
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.00018057215815690494,
+      "loss": 0.286,
+      "step": 7040
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0001805307885711122,
+      "loss": 0.2848,
+      "step": 7045
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0001804893797355914,
+      "loss": 0.2881,
+      "step": 7050
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00018044793167052477,
+      "loss": 0.2854,
+      "step": 7055
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.00018040644439611348,
+      "loss": 0.2894,
+      "step": 7060
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.00018036491793257798,
+      "loss": 0.2867,
+      "step": 7065
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0001803233523001578,
+      "loss": 0.2862,
+      "step": 7070
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00018028174751911146,
+      "loss": 0.2885,
+      "step": 7075
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0001802401036097167,
+      "loss": 0.2859,
+      "step": 7080
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.00018019842059227012,
+      "loss": 0.286,
+      "step": 7085
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.00018015669848708767,
+      "loss": 0.2864,
+      "step": 7090
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00018011493731450413,
+      "loss": 0.2847,
+      "step": 7095
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.00018007313709487334,
+      "loss": 0.2863,
+      "step": 7100
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0001800312978485683,
+      "loss": 0.291,
+      "step": 7105
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.00017998941959598095,
+      "loss": 0.2865,
+      "step": 7110
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0001799475023575222,
+      "loss": 0.2867,
+      "step": 7115
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.00017990554615362198,
+      "loss": 0.2874,
+      "step": 7120
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00017986355100472928,
+      "loss": 0.2863,
+      "step": 7125
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.00017982151693131203,
+      "loss": 0.2872,
+      "step": 7130
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0001797794439538571,
+      "loss": 0.2865,
+      "step": 7135
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.00017973733209287036,
+      "loss": 0.287,
+      "step": 7140
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00017969518136887663,
+      "loss": 0.287,
+      "step": 7145
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00017965299180241963,
+      "loss": 0.287,
+      "step": 7150
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.00017961076341406208,
+      "loss": 0.2891,
+      "step": 7155
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00017956849622438554,
+      "loss": 0.2878,
+      "step": 7160
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.00017952619025399057,
+      "loss": 0.2862,
+      "step": 7165
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.00017948384552349657,
+      "loss": 0.2887,
+      "step": 7170
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.00017944146205354182,
+      "loss": 0.2862,
+      "step": 7175
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00017939903986478355,
+      "loss": 0.286,
+      "step": 7180
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0001793565789778978,
+      "loss": 0.2884,
+      "step": 7185
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.00017931407941357947,
+      "loss": 0.285,
+      "step": 7190
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00017927154119254236,
+      "loss": 0.2861,
+      "step": 7195
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.00017922896433551907,
+      "loss": 0.289,
+      "step": 7200
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00017918634886326108,
+      "loss": 0.2874,
+      "step": 7205
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0001791436947965386,
+      "loss": 0.2863,
+      "step": 7210
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0001791010021561407,
+      "loss": 0.2856,
+      "step": 7215
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0001790582709628753,
+      "loss": 0.2862,
+      "step": 7220
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.00017901550123756906,
+      "loss": 0.2866,
+      "step": 7225
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.00017897269300106737,
+      "loss": 0.2866,
+      "step": 7230
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0001789298462742345,
+      "loss": 0.2874,
+      "step": 7235
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.00017888696107795342,
+      "loss": 0.2865,
+      "step": 7240
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.00017884403743312582,
+      "loss": 0.2879,
+      "step": 7245
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00017880107536067218,
+      "loss": 0.2864,
+      "step": 7250
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00017875807488153175,
+      "loss": 0.2887,
+      "step": 7255
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.00017871503601666233,
+      "loss": 0.2873,
+      "step": 7260
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0001786719587870406,
+      "loss": 0.2861,
+      "step": 7265
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.00017862884321366188,
+      "loss": 0.288,
+      "step": 7270
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0001785856893175402,
+      "loss": 0.2865,
+      "step": 7275
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00017854249711970818,
+      "loss": 0.2882,
+      "step": 7280
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00017849926664121726,
+      "loss": 0.2887,
+      "step": 7285
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.00017845599790313735,
+      "loss": 0.2876,
+      "step": 7290
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.00017841269092655715,
+      "loss": 0.2865,
+      "step": 7295
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.000178369345732584,
+      "loss": 0.2837,
+      "step": 7300
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.00017832596234234376,
+      "loss": 0.2872,
+      "step": 7305
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.000178282540776981,
+      "loss": 0.2863,
+      "step": 7310
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0001782390810576588,
+      "loss": 0.2862,
+      "step": 7315
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.000178195583205559,
+      "loss": 0.2864,
+      "step": 7320
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00017815204724188187,
+      "loss": 0.2845,
+      "step": 7325
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0001781084731878463,
+      "loss": 0.2872,
+      "step": 7330
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.00017806486106468981,
+      "loss": 0.2856,
+      "step": 7335
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00017802121089366836,
+      "loss": 0.2886,
+      "step": 7340
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00017797752269605653,
+      "loss": 0.2874,
+      "step": 7345
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.00017793379649314744,
+      "loss": 0.2848,
+      "step": 7350
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00017789003230625266,
+      "loss": 0.2853,
+      "step": 7355
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.00017784623015670238,
+      "loss": 0.2863,
+      "step": 7360
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.00017780239006584515,
+      "loss": 0.2863,
+      "step": 7365
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0001777585120550482,
+      "loss": 0.2876,
+      "step": 7370
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0001777145961456971,
+      "loss": 0.2875,
+      "step": 7375
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00017767064235919592,
+      "loss": 0.2872,
+      "step": 7380
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0001776266507169672,
+      "loss": 0.2865,
+      "step": 7385
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.00017758262124045195,
+      "loss": 0.2872,
+      "step": 7390
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0001775385539511096,
+      "loss": 0.2849,
+      "step": 7395
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.00017749444887041799,
+      "loss": 0.2883,
+      "step": 7400
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.00017745030601987337,
+      "loss": 0.288,
+      "step": 7405
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.00017740612542099053,
+      "loss": 0.2879,
+      "step": 7410
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0001773619070953025,
+      "loss": 0.2869,
+      "step": 7415
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.00017731765106436073,
+      "loss": 0.2863,
+      "step": 7420
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00017727335734973512,
+      "loss": 0.2872,
+      "step": 7425
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.00017722902597301383,
+      "loss": 0.2883,
+      "step": 7430
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0001771846569558035,
+      "loss": 0.2866,
+      "step": 7435
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.00017714025031972903,
+      "loss": 0.2844,
+      "step": 7440
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00017709580608643363,
+      "loss": 0.2887,
+      "step": 7445
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.00017705132427757895,
+      "loss": 0.2878,
+      "step": 7450
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0001770068049148448,
+      "loss": 0.2871,
+      "step": 7455
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00017696224801992945,
+      "loss": 0.2878,
+      "step": 7460
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.00017691765361454938,
+      "loss": 0.2864,
+      "step": 7465
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00017687302172043933,
+      "loss": 0.2863,
+      "step": 7470
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.00017682835235935236,
+      "loss": 0.2874,
+      "step": 7475
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00017678364555305978,
+      "loss": 0.288,
+      "step": 7480
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0001767389013233511,
+      "loss": 0.2866,
+      "step": 7485
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00017669411969203417,
+      "loss": 0.2882,
+      "step": 7490
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.00017664930068093498,
+      "loss": 0.2867,
+      "step": 7495
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0001766044443118978,
+      "loss": 0.2855,
+      "step": 7500
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.00017655955060678506,
+      "loss": 0.2882,
+      "step": 7505
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.00017651461958747745,
+      "loss": 0.2889,
+      "step": 7510
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0001764696512758737,
+      "loss": 0.2872,
+      "step": 7515
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0001764246456938909,
+      "loss": 0.2878,
+      "step": 7520
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.00017637960286346425,
+      "loss": 0.2887,
+      "step": 7525
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.000176334522806547,
+      "loss": 0.2866,
+      "step": 7530
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.00017628940554511061,
+      "loss": 0.2875,
+      "step": 7535
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0001762442511011448,
+      "loss": 0.2873,
+      "step": 7540
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0001761990594966572,
+      "loss": 0.288,
+      "step": 7545
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0001761538307536737,
+      "loss": 0.2867,
+      "step": 7550
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0001761085648942382,
+      "loss": 0.2874,
+      "step": 7555
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.00017606326194041273,
+      "loss": 0.286,
+      "step": 7560
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00017601792191427741,
+      "loss": 0.2869,
+      "step": 7565
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.00017597254483793048,
+      "loss": 0.2878,
+      "step": 7570
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.00017592713073348807,
+      "loss": 0.2872,
+      "step": 7575
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.00017588167962308458,
+      "loss": 0.2862,
+      "step": 7580
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0001758361915288722,
+      "loss": 0.2872,
+      "step": 7585
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.00017579066647302133,
+      "loss": 0.2864,
+      "step": 7590
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00017574510447772039,
+      "loss": 0.2852,
+      "step": 7595
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.00017569950556517566,
+      "loss": 0.2872,
+      "step": 7600
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0001756538697576115,
+      "loss": 0.2857,
+      "step": 7605
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.00017560819707727033,
+      "loss": 0.2868,
+      "step": 7610
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.00017556248754641235,
+      "loss": 0.2846,
+      "step": 7615
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.00017551674118731591,
+      "loss": 0.2868,
+      "step": 7620
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.00017547095802227723,
+      "loss": 0.2882,
+      "step": 7625
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00017542513807361037,
+      "loss": 0.2873,
+      "step": 7630
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.00017537928136364755,
+      "loss": 0.286,
+      "step": 7635
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0001753333879147387,
+      "loss": 0.2868,
+      "step": 7640
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.00017528745774925172,
+      "loss": 0.2869,
+      "step": 7645
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00017524149088957245,
+      "loss": 0.2848,
+      "step": 7650
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.00017519548735810456,
+      "loss": 0.2867,
+      "step": 7655
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.00017514944717726962,
+      "loss": 0.285,
+      "step": 7660
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.00017510337036950703,
+      "loss": 0.2869,
+      "step": 7665
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.00017505725695727412,
+      "loss": 0.2867,
+      "step": 7670
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00017501110696304596,
+      "loss": 0.287,
+      "step": 7675
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.00017496492040931552,
+      "loss": 0.2868,
+      "step": 7680
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00017491869731859353,
+      "loss": 0.2877,
+      "step": 7685
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0001748724377134086,
+      "loss": 0.2892,
+      "step": 7690
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.00017482614161630714,
+      "loss": 0.2893,
+      "step": 7695
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0001747798090498532,
+      "loss": 0.2872,
+      "step": 7700
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.00017473344003662877,
+      "loss": 0.2861,
+      "step": 7705
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.00017468703459923356,
+      "loss": 0.2866,
+      "step": 7710
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00017464059276028497,
+      "loss": 0.2864,
+      "step": 7715
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.00017459411454241822,
+      "loss": 0.2878,
+      "step": 7720
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.00017454759996828623,
+      "loss": 0.2874,
+      "step": 7725
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.00017450104906055963,
+      "loss": 0.2851,
+      "step": 7730
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.00017445446184192676,
+      "loss": 0.2861,
+      "step": 7735
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.00017440783833509366,
+      "loss": 0.2876,
+      "step": 7740
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0001743611785627841,
+      "loss": 0.2859,
+      "step": 7745
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00017431448254773944,
+      "loss": 0.287,
+      "step": 7750
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.00017426775031271875,
+      "loss": 0.2892,
+      "step": 7755
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.00017422098188049883,
+      "loss": 0.2863,
+      "step": 7760
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00017417417727387394,
+      "loss": 0.2873,
+      "step": 7765
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0001741273365156561,
+      "loss": 0.2876,
+      "step": 7770
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.000174080459628675,
+      "loss": 0.2853,
+      "step": 7775
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.00017403354663577783,
+      "loss": 0.2874,
+      "step": 7780
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.00017398659755982936,
+      "loss": 0.2872,
+      "step": 7785
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.00017393961242371205,
+      "loss": 0.2856,
+      "step": 7790
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0001738925912503259,
+      "loss": 0.2899,
+      "step": 7795
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00017384553406258842,
+      "loss": 0.2865,
+      "step": 7800
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.00017379844088343468,
+      "loss": 0.286,
+      "step": 7805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0001737513117358174,
+      "loss": 0.2863,
+      "step": 7810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.00017370414664270674,
+      "loss": 0.2846,
+      "step": 7815
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.00017365694562709034,
+      "loss": 0.2878,
+      "step": 7820
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.00017360970871197346,
+      "loss": 0.2888,
+      "step": 7825
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00017356243592037875,
+      "loss": 0.2889,
+      "step": 7830
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00017351512727534644,
+      "loss": 0.2864,
+      "step": 7835
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.00017346778279993415,
+      "loss": 0.2879,
+      "step": 7840
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00017342040251721702,
+      "loss": 0.2862,
+      "step": 7845
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.00017337298645028764,
+      "loss": 0.2865,
+      "step": 7850
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00017332553462225602,
+      "loss": 0.2891,
+      "step": 7855
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00017327804705624957,
+      "loss": 0.2865,
+      "step": 7860
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00017323052377541317,
+      "loss": 0.2883,
+      "step": 7865
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0001731829648029091,
+      "loss": 0.2862,
+      "step": 7870
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.00017313537016191706,
+      "loss": 0.2881,
+      "step": 7875
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.00017308773987563406,
+      "loss": 0.2856,
+      "step": 7880
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017304007396727448,
+      "loss": 0.2861,
+      "step": 7885
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.00017299237246007015,
+      "loss": 0.2889,
+      "step": 7890
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.00017294463537727024,
+      "loss": 0.2866,
+      "step": 7895
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.00017289686274214118,
+      "loss": 0.2886,
+      "step": 7900
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.00017284905457796675,
+      "loss": 0.2886,
+      "step": 7905
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.00017280121090804812,
+      "loss": 0.2872,
+      "step": 7910
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00017275333175570368,
+      "loss": 0.2874,
+      "step": 7915
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001727054171442692,
+      "loss": 0.2873,
+      "step": 7920
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0001726574670970976,
+      "loss": 0.2859,
+      "step": 7925
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.00017260948163755918,
+      "loss": 0.2851,
+      "step": 7930
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00017256146078904153,
+      "loss": 0.2873,
+      "step": 7935
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.00017251340457494934,
+      "loss": 0.2875,
+      "step": 7940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0001724653130187047,
+      "loss": 0.2857,
+      "step": 7945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.00017241718614374678,
+      "loss": 0.2897,
+      "step": 7950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00017236902397353205,
+      "loss": 0.2868,
+      "step": 7955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00017232082653153422,
+      "loss": 0.2872,
+      "step": 7960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0001722725938412441,
+      "loss": 0.2855,
+      "step": 7965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0001722243259261697,
+      "loss": 0.2879,
+      "step": 7970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.00017217602280983623,
+      "loss": 0.2896,
+      "step": 7975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.00017212768451578607,
+      "loss": 0.2875,
+      "step": 7980
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.00017207931106757868,
+      "loss": 0.2868,
+      "step": 7985
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0001720309024887907,
+      "loss": 0.2899,
+      "step": 7990
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001719824588030159,
+      "loss": 0.2879,
+      "step": 7995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0001719339800338651,
+      "loss": 0.2893,
+      "step": 8000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00017188546620496635,
+      "loss": 0.286,
+      "step": 8005
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.00017183691733996462,
+      "loss": 0.2882,
+      "step": 8010
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.00017178833346252206,
+      "loss": 0.288,
+      "step": 8015
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.00017173971459631787,
+      "loss": 0.2877,
+      "step": 8020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0001716910607650483,
+      "loss": 0.2863,
+      "step": 8025
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0001716423719924266,
+      "loss": 0.2861,
+      "step": 8030
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.00017159364830218312,
+      "loss": 0.2862,
+      "step": 8035
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.00017154488971806518,
+      "loss": 0.2878,
+      "step": 8040
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.00017149609626383717,
+      "loss": 0.2862,
+      "step": 8045
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.00017144726796328034,
+      "loss": 0.2885,
+      "step": 8050
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0001713984048401931,
+      "loss": 0.2862,
+      "step": 8055
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.00017134950691839065,
+      "loss": 0.2874,
+      "step": 8060
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0001713005742217053,
+      "loss": 0.2869,
+      "step": 8065
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.00017125160677398626,
+      "loss": 0.2891,
+      "step": 8070
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.00017120260459909967,
+      "loss": 0.2858,
+      "step": 8075
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00017115356772092857,
+      "loss": 0.2872,
+      "step": 8080
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00017110449616337289,
+      "loss": 0.2879,
+      "step": 8085
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00017105538995034963,
+      "loss": 0.285,
+      "step": 8090
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0001710062491057925,
+      "loss": 0.2886,
+      "step": 8095
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0001709570736536521,
+      "loss": 0.286,
+      "step": 8100
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.000170907863617896,
+      "loss": 0.2859,
+      "step": 8105
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.00017085861902250865,
+      "loss": 0.2871,
+      "step": 8110
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0001708093398914911,
+      "loss": 0.2861,
+      "step": 8115
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.00017076002624886154,
+      "loss": 0.2877,
+      "step": 8120
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00017071067811865476,
+      "loss": 0.2872,
+      "step": 8125
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0001706612955249225,
+      "loss": 0.2895,
+      "step": 8130
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.00017061187849173317,
+      "loss": 0.2869,
+      "step": 8135
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0001705624270431721,
+      "loss": 0.2873,
+      "step": 8140
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.00017051294120334125,
+      "loss": 0.2888,
+      "step": 8145
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00017046342099635948,
+      "loss": 0.2875,
+      "step": 8150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0001704138664463623,
+      "loss": 0.2886,
+      "step": 8155
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00017036427757750205,
+      "loss": 0.2875,
+      "step": 8160
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00017031465441394767,
+      "loss": 0.2859,
+      "step": 8165
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00017026499697988493,
+      "loss": 0.2872,
+      "step": 8170
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.00017021530529951625,
+      "loss": 0.2879,
+      "step": 8175
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.00017016557939706075,
+      "loss": 0.2869,
+      "step": 8180
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00017011581929675425,
+      "loss": 0.2863,
+      "step": 8185
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00017006602502284918,
+      "loss": 0.2865,
+      "step": 8190
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00017001619659961467,
+      "loss": 0.287,
+      "step": 8195
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.00016996633405133655,
+      "loss": 0.2857,
+      "step": 8200
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.00016991643740231716,
+      "loss": 0.2895,
+      "step": 8205
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00016986650667687552,
+      "loss": 0.2842,
+      "step": 8210
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.00016981654189934727,
+      "loss": 0.2866,
+      "step": 8215
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.00016976654309408464,
+      "loss": 0.2885,
+      "step": 8220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.00016971651028545648,
+      "loss": 0.2886,
+      "step": 8225
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.00016966644349784808,
+      "loss": 0.2865,
+      "step": 8230
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.00016961634275566146,
+      "loss": 0.2858,
+      "step": 8235
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0001695662080833151,
+      "loss": 0.2881,
+      "step": 8240
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.000169516039505244,
+      "loss": 0.2868,
+      "step": 8245
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.00016946583704589973,
+      "loss": 0.2877,
+      "step": 8250
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0001694156007297504,
+      "loss": 0.2867,
+      "step": 8255
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0001693653305812805,
+      "loss": 0.2871,
+      "step": 8260
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00016931502662499118,
+      "loss": 0.2879,
+      "step": 8265
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0001692646888853999,
+      "loss": 0.2846,
+      "step": 8270
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0001692143173870407,
+      "loss": 0.2878,
+      "step": 8275
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.000169163912154464,
+      "loss": 0.2863,
+      "step": 8280
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0001691134732122368,
+      "loss": 0.2869,
+      "step": 8285
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.00016906300058494228,
+      "loss": 0.2874,
+      "step": 8290
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.00016901249429718032,
+      "loss": 0.2858,
+      "step": 8295
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.000168961954373567,
+      "loss": 0.2871,
+      "step": 8300
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00016891138083873487,
+      "loss": 0.2879,
+      "step": 8305
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.00016886077371733283,
+      "loss": 0.2861,
+      "step": 8310
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0001688101330340263,
+      "loss": 0.2858,
+      "step": 8315
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00016875945881349676,
+      "loss": 0.2867,
+      "step": 8320
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0001687087510804423,
+      "loss": 0.2869,
+      "step": 8325
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.00016865800985957726,
+      "loss": 0.2871,
+      "step": 8330
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0001686072351756323,
+      "loss": 0.2842,
+      "step": 8335
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.00016855642705335437,
+      "loss": 0.2861,
+      "step": 8340
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0001685055855175067,
+      "loss": 0.286,
+      "step": 8345
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00016845471059286887,
+      "loss": 0.2866,
+      "step": 8350
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00016840380230423668,
+      "loss": 0.2873,
+      "step": 8355
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00016835286067642228,
+      "loss": 0.2859,
+      "step": 8360
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.00016830188573425389,
+      "loss": 0.2882,
+      "step": 8365
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0001682508775025762,
+      "loss": 0.2867,
+      "step": 8370
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.00016819983600624986,
+      "loss": 0.287,
+      "step": 8375
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.000168148761270152,
+      "loss": 0.2867,
+      "step": 8380
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00016809765331917575,
+      "loss": 0.2867,
+      "step": 8385
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.00016804651217823053,
+      "loss": 0.2846,
+      "step": 8390
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.00016799533787224192,
+      "loss": 0.285,
+      "step": 8395
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.00016794413042615168,
+      "loss": 0.2853,
+      "step": 8400
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.00016789288986491762,
+      "loss": 0.2876,
+      "step": 8405
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.00016784161621351382,
+      "loss": 0.2869,
+      "step": 8410
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.00016779030949693044,
+      "loss": 0.2862,
+      "step": 8415
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.00016773896974017373,
+      "loss": 0.2861,
+      "step": 8420
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.00016768759696826608,
+      "loss": 0.2873,
+      "step": 8425
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.00016763619120624594,
+      "loss": 0.286,
+      "step": 8430
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00016758475247916787,
+      "loss": 0.2856,
+      "step": 8435
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.00016753328081210245,
+      "loss": 0.2845,
+      "step": 8440
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.00016748177623013638,
+      "loss": 0.2885,
+      "step": 8445
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.00016743023875837233,
+      "loss": 0.2869,
+      "step": 8450
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0001673786684219291,
+      "loss": 0.2848,
+      "step": 8455
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.00016732706524594137,
+      "loss": 0.289,
+      "step": 8460
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.00016727542925555996,
+      "loss": 0.287,
+      "step": 8465
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.00016722376047595164,
+      "loss": 0.2862,
+      "step": 8470
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.00016717205893229903,
+      "loss": 0.2894,
+      "step": 8475
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00016712032464980095,
+      "loss": 0.2887,
+      "step": 8480
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.000167068557653672,
+      "loss": 0.2864,
+      "step": 8485
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.00016701675796914286,
+      "loss": 0.289,
+      "step": 8490
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.00016696492562145996,
+      "loss": 0.2867,
+      "step": 8495
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.00016691306063588583,
+      "loss": 0.2855,
+      "step": 8500
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00016686116303769882,
+      "loss": 0.2868,
+      "step": 8505
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00016680923285219317,
+      "loss": 0.2865,
+      "step": 8510
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.00016675727010467906,
+      "loss": 0.2898,
+      "step": 8515
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.00016670527482048246,
+      "loss": 0.2868,
+      "step": 8520
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.00016665324702494524,
+      "loss": 0.2877,
+      "step": 8525
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.00016660118674342517,
+      "loss": 0.2877,
+      "step": 8530
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.00016654909400129575,
+      "loss": 0.2875,
+      "step": 8535
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.00016649696882394633,
+      "loss": 0.287,
+      "step": 8540
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.00016644481123678217,
+      "loss": 0.2848,
+      "step": 8545
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.00016639262126522418,
+      "loss": 0.2855,
+      "step": 8550
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00016634039893470912,
+      "loss": 0.2858,
+      "step": 8555
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00016628814427068953,
+      "loss": 0.2865,
+      "step": 8560
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.00016623585729863368,
+      "loss": 0.2863,
+      "step": 8565
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.00016618353804402568,
+      "loss": 0.2871,
+      "step": 8570
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00016613118653236518,
+      "loss": 0.2843,
+      "step": 8575
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00016607880278916777,
+      "loss": 0.2851,
+      "step": 8580
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00016602638683996463,
+      "loss": 0.2871,
+      "step": 8585
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00016597393871030264,
+      "loss": 0.2878,
+      "step": 8590
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00016592145842574433,
+      "loss": 0.2851,
+      "step": 8595
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.00016586894601186805,
+      "loss": 0.2869,
+      "step": 8600
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.00016581640149426768,
+      "loss": 0.2861,
+      "step": 8605
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.00016576382489855274,
+      "loss": 0.2873,
+      "step": 8610
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00016571121625034847,
+      "loss": 0.288,
+      "step": 8615
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.00016565857557529566,
+      "loss": 0.2871,
+      "step": 8620
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00016560590289905073,
+      "loss": 0.2884,
+      "step": 8625
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.00016555319824728575,
+      "loss": 0.2878,
+      "step": 8630
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00016550046164568827,
+      "loss": 0.2855,
+      "step": 8635
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.00016544769311996148,
+      "loss": 0.2857,
+      "step": 8640
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.00016539489269582416,
+      "loss": 0.2874,
+      "step": 8645
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00016534206039901057,
+      "loss": 0.2862,
+      "step": 8650
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0001652891962552705,
+      "loss": 0.2867,
+      "step": 8655
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00016523630029036931,
+      "loss": 0.2876,
+      "step": 8660
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0001651833725300879,
+      "loss": 0.2855,
+      "step": 8665
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.00016513041300022255,
+      "loss": 0.2893,
+      "step": 8670
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0001650774217265851,
+      "loss": 0.2863,
+      "step": 8675
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.00016502439873500289,
+      "loss": 0.2874,
+      "step": 8680
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.00016497134405131866,
+      "loss": 0.2864,
+      "step": 8685
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00016491825770139062,
+      "loss": 0.2848,
+      "step": 8690
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.00016486513971109243,
+      "loss": 0.2876,
+      "step": 8695
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0001648119901063131,
+      "loss": 0.2881,
+      "step": 8700
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.00016475880891295716,
+      "loss": 0.2895,
+      "step": 8705
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00016470559615694446,
+      "loss": 0.2882,
+      "step": 8710
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0001646523518642102,
+      "loss": 0.2873,
+      "step": 8715
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0001645990760607051,
+      "loss": 0.2862,
+      "step": 8720
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.00016454576877239507,
+      "loss": 0.2863,
+      "step": 8725
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.00016449243002526144,
+      "loss": 0.2861,
+      "step": 8730
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0001644390598453009,
+      "loss": 0.2884,
+      "step": 8735
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0001643856582585254,
+      "loss": 0.2874,
+      "step": 8740
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0001643322252909622,
+      "loss": 0.2889,
+      "step": 8745
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.00016427876096865394,
+      "loss": 0.2855,
+      "step": 8750
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00016422526531765846,
+      "loss": 0.2856,
+      "step": 8755
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00016417173836404887,
+      "loss": 0.288,
+      "step": 8760
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.00016411818013391355,
+      "loss": 0.2879,
+      "step": 8765
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.00016406459065335615,
+      "loss": 0.286,
+      "step": 8770
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.00016401096994849557,
+      "loss": 0.2868,
+      "step": 8775
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0001639573180454658,
+      "loss": 0.2872,
+      "step": 8780
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00016390363497041622,
+      "loss": 0.2849,
+      "step": 8785
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.00016384992074951123,
+      "loss": 0.2863,
+      "step": 8790
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00016379617540893056,
+      "loss": 0.2851,
+      "step": 8795
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.000163742398974869,
+      "loss": 0.2887,
+      "step": 8800
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0001636885914735365,
+      "loss": 0.2867,
+      "step": 8805
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00016363475293115824,
+      "loss": 0.2862,
+      "step": 8810
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.00016358088337397442,
+      "loss": 0.2848,
+      "step": 8815
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.00016352698282824044,
+      "loss": 0.2856,
+      "step": 8820
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.00016347305132022677,
+      "loss": 0.2874,
+      "step": 8825
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00016341908887621895,
+      "loss": 0.2856,
+      "step": 8830
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00016336509552251766,
+      "loss": 0.2867,
+      "step": 8835
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00016331107128543857,
+      "loss": 0.2876,
+      "step": 8840
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00016325701619131246,
+      "loss": 0.2863,
+      "step": 8845
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0001632029302664851,
+      "loss": 0.2867,
+      "step": 8850
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.00016314881353731732,
+      "loss": 0.2858,
+      "step": 8855
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00016309466603018496,
+      "loss": 0.2862,
+      "step": 8860
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0001630404877714789,
+      "loss": 0.289,
+      "step": 8865
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.00016298627878760487,
+      "loss": 0.2875,
+      "step": 8870
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00016293203910498376,
+      "loss": 0.286,
+      "step": 8875
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0001628777687500513,
+      "loss": 0.2853,
+      "step": 8880
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.00016282346774925817,
+      "loss": 0.2875,
+      "step": 8885
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.00016276913612907007,
+      "loss": 0.2859,
+      "step": 8890
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.00016271477391596753,
+      "loss": 0.2858,
+      "step": 8895
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.00016266038113644607,
+      "loss": 0.2878,
+      "step": 8900
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.00016260595781701604,
+      "loss": 0.2865,
+      "step": 8905
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.00016255150398420271,
+      "loss": 0.2853,
+      "step": 8910
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.00016249701966454625,
+      "loss": 0.2878,
+      "step": 8915
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.00016244250488460158,
+      "loss": 0.2881,
+      "step": 8920
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.00016238795967093864,
+      "loss": 0.2865,
+      "step": 8925
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00016233338405014202,
+      "loss": 0.2857,
+      "step": 8930
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.00016227877804881127,
+      "loss": 0.2857,
+      "step": 8935
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.00016222414169356065,
+      "loss": 0.2863,
+      "step": 8940
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0001621694750110193,
+      "loss": 0.287,
+      "step": 8945
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00016211477802783103,
+      "loss": 0.285,
+      "step": 8950
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.00016206005077065458,
+      "loss": 0.288,
+      "step": 8955
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.00016200529326616328,
+      "loss": 0.2897,
+      "step": 8960
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00016195050554104528,
+      "loss": 0.2867,
+      "step": 8965
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00016189568762200348,
+      "loss": 0.287,
+      "step": 8970
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0001618408395357554,
+      "loss": 0.2857,
+      "step": 8975
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.00016178596130903344,
+      "loss": 0.2871,
+      "step": 8980
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.00016173105296858452,
+      "loss": 0.2854,
+      "step": 8985
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.00016167611454117025,
+      "loss": 0.2868,
+      "step": 8990
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00016162114605356703,
+      "loss": 0.2857,
+      "step": 8995
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0001615661475325658,
+      "loss": 0.2865,
+      "step": 9000
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.00016151111900497225,
+      "loss": 0.288,
+      "step": 9005
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.00016145606049760644,
+      "loss": 0.2857,
+      "step": 9010
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.00016140097203730337,
+      "loss": 0.2875,
+      "step": 9015
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.00016134585365091243,
+      "loss": 0.2878,
+      "step": 9020
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.00016129070536529766,
+      "loss": 0.2864,
+      "step": 9025
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.00016123552720733765,
+      "loss": 0.2875,
+      "step": 9030
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00016118031920392558,
+      "loss": 0.2872,
+      "step": 9035
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.00016112508138196917,
+      "loss": 0.2871,
+      "step": 9040
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00016106981376839066,
+      "loss": 0.2871,
+      "step": 9045
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0001610145163901268,
+      "loss": 0.2857,
+      "step": 9050
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0001609591892741288,
+      "loss": 0.2874,
+      "step": 9055
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.00016090383244736256,
+      "loss": 0.2899,
+      "step": 9060
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0001608484459368082,
+      "loss": 0.287,
+      "step": 9065
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.00016079302976946055,
+      "loss": 0.2875,
+      "step": 9070
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.00016073758397232868,
+      "loss": 0.286,
+      "step": 9075
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00016068210857243624,
+      "loss": 0.287,
+      "step": 9080
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00016062660359682124,
+      "loss": 0.2862,
+      "step": 9085
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.00016057106907253616,
+      "loss": 0.287,
+      "step": 9090
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0001605155050266478,
+      "loss": 0.2871,
+      "step": 9095
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0001604599114862375,
+      "loss": 0.2866,
+      "step": 9100
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00016040428847840079,
+      "loss": 0.2893,
+      "step": 9105
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.00016034863603024767,
+      "loss": 0.2865,
+      "step": 9110
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.00016029295416890248,
+      "loss": 0.2882,
+      "step": 9115
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00016023724292150385,
+      "loss": 0.2863,
+      "step": 9120
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.00016018150231520486,
+      "loss": 0.2862,
+      "step": 9125
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0001601257323771727,
+      "loss": 0.2872,
+      "step": 9130
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00016006993313458896,
+      "loss": 0.2871,
+      "step": 9135
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.00016001410461464956,
+      "loss": 0.2869,
+      "step": 9140
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.00015995824684456465,
+      "loss": 0.2856,
+      "step": 9145
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0001599023598515586,
+      "loss": 0.2865,
+      "step": 9150
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.00015984644366287006,
+      "loss": 0.2888,
+      "step": 9155
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0001597904983057519,
+      "loss": 0.2868,
+      "step": 9160
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00015973452380747122,
+      "loss": 0.2867,
+      "step": 9165
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.00015967852019530926,
+      "loss": 0.2882,
+      "step": 9170
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0001596224874965616,
+      "loss": 0.2871,
+      "step": 9175
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0001595664257385378,
+      "loss": 0.2866,
+      "step": 9180
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.00015951033494856175,
+      "loss": 0.2871,
+      "step": 9185
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.00015945421515397133,
+      "loss": 0.2881,
+      "step": 9190
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0001593980663821187,
+      "loss": 0.2871,
+      "step": 9195
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.00015934188866037016,
+      "loss": 0.2885,
+      "step": 9200
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00015928568201610595,
+      "loss": 0.2852,
+      "step": 9205
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00015922944647672052,
+      "loss": 0.2887,
+      "step": 9210
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0001591731820696224,
+      "loss": 0.2842,
+      "step": 9215
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0001591168888222342,
+      "loss": 0.2873,
+      "step": 9220
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.00015906056676199255,
+      "loss": 0.2864,
+      "step": 9225
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.00015900421591634814,
+      "loss": 0.2871,
+      "step": 9230
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.00015894783631276567,
+      "loss": 0.2857,
+      "step": 9235
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.00015889142797872387,
+      "loss": 0.286,
+      "step": 9240
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.00015883499094171554,
+      "loss": 0.2843,
+      "step": 9245
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00015877852522924732,
+      "loss": 0.2871,
+      "step": 9250
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.00015872203086883996,
+      "loss": 0.2862,
+      "step": 9255
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00015866550788802813,
+      "loss": 0.2852,
+      "step": 9260
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.00015860895631436043,
+      "loss": 0.2895,
+      "step": 9265
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0001585523761753994,
+      "loss": 0.2874,
+      "step": 9270
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00015849576749872157,
+      "loss": 0.2865,
+      "step": 9275
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00015843913031191723,
+      "loss": 0.2874,
+      "step": 9280
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0001583824646425907,
+      "loss": 0.2848,
+      "step": 9285
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.00015832577051836015,
+      "loss": 0.2877,
+      "step": 9290
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.00015826904796685762,
+      "loss": 0.2871,
+      "step": 9295
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00015821229701572896,
+      "loss": 0.2881,
+      "step": 9300
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00015815551769263387,
+      "loss": 0.2856,
+      "step": 9305
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.000158098710025246,
+      "loss": 0.2876,
+      "step": 9310
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0001580418740412526,
+      "loss": 0.2861,
+      "step": 9315
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.00015798500976835493,
+      "loss": 0.2865,
+      "step": 9320
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0001579281172342679,
+      "loss": 0.2884,
+      "step": 9325
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00015787119646672025,
+      "loss": 0.2867,
+      "step": 9330
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.00015781424749345446,
+      "loss": 0.2875,
+      "step": 9335
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.00015775727034222675,
+      "loss": 0.2873,
+      "step": 9340
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.00015770026504080718,
+      "loss": 0.286,
+      "step": 9345
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00015764323161697935,
+      "loss": 0.2883,
+      "step": 9350
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0001575861700985407,
+      "loss": 0.2867,
+      "step": 9355
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00015752908051330227,
+      "loss": 0.2863,
+      "step": 9360
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00015747196288908887,
+      "loss": 0.2858,
+      "step": 9365
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.000157414817253739,
+      "loss": 0.2866,
+      "step": 9370
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0001573576436351046,
+      "loss": 0.2868,
+      "step": 9375
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.00015730044206105156,
+      "loss": 0.2846,
+      "step": 9380
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0001572432125594591,
+      "loss": 0.2859,
+      "step": 9385
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.00015718595515822027,
+      "loss": 0.2859,
+      "step": 9390
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.00015712866988524155,
+      "loss": 0.287,
+      "step": 9395
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0001570713567684432,
+      "loss": 0.2871,
+      "step": 9400
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.00015701401583575884,
+      "loss": 0.2873,
+      "step": 9405
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00015695664711513576,
+      "loss": 0.2864,
+      "step": 9410
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0001568992506345348,
+      "loss": 0.2881,
+      "step": 9415
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0001568418264219303,
+      "loss": 0.2863,
+      "step": 9420
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.00015678437450531013,
+      "loss": 0.2871,
+      "step": 9425
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00015672689491267567,
+      "loss": 0.2864,
+      "step": 9430
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0001566693876720417,
+      "loss": 0.2865,
+      "step": 9435
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00015661185281143667,
+      "loss": 0.2869,
+      "step": 9440
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0001565542903589023,
+      "loss": 0.2871,
+      "step": 9445
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0001564967003424938,
+      "loss": 0.2866,
+      "step": 9450
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.00015643908279027992,
+      "loss": 0.2881,
+      "step": 9455
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.00015638143773034267,
+      "loss": 0.2878,
+      "step": 9460
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.00015632376519077767,
+      "loss": 0.2884,
+      "step": 9465
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0001562660651996937,
+      "loss": 0.2866,
+      "step": 9470
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00015620833778521307,
+      "loss": 0.2857,
+      "step": 9475
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.00015615058297547145,
+      "loss": 0.2882,
+      "step": 9480
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0001560928007986178,
+      "loss": 0.2882,
+      "step": 9485
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00015603499128281448,
+      "loss": 0.2868,
+      "step": 9490
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.00015597715445623712,
+      "loss": 0.288,
+      "step": 9495
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0001559192903470747,
+      "loss": 0.2859,
+      "step": 9500
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00015586139898352946,
+      "loss": 0.2865,
+      "step": 9505
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.000155803480393817,
+      "loss": 0.2873,
+      "step": 9510
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.00015574553460616608,
+      "loss": 0.2859,
+      "step": 9515
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.00015568756164881882,
+      "loss": 0.2863,
+      "step": 9520
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0001556295615500305,
+      "loss": 0.2863,
+      "step": 9525
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.00015557153433806966,
+      "loss": 0.2857,
+      "step": 9530
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0001555134800412181,
+      "loss": 0.286,
+      "step": 9535
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00015545539868777074,
+      "loss": 0.2873,
+      "step": 9540
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00015539729030603574,
+      "loss": 0.2873,
+      "step": 9545
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.00015533915492433443,
+      "loss": 0.2853,
+      "step": 9550
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.00015528099257100127,
+      "loss": 0.2856,
+      "step": 9555
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.00015522280327438388,
+      "loss": 0.2868,
+      "step": 9560
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.00015516458706284303,
+      "loss": 0.2845,
+      "step": 9565
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0001551063439647526,
+      "loss": 0.2853,
+      "step": 9570
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00015504807400849958,
+      "loss": 0.2857,
+      "step": 9575
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.000154989777222484,
+      "loss": 0.2884,
+      "step": 9580
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.000154931453635119,
+      "loss": 0.2877,
+      "step": 9585
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.00015487310327483086,
+      "loss": 0.286,
+      "step": 9590
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00015481472617005876,
+      "loss": 0.2878,
+      "step": 9595
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.00015475632234925504,
+      "loss": 0.289,
+      "step": 9600
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.00015469789184088497,
+      "loss": 0.2853,
+      "step": 9605
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.00015463943467342693,
+      "loss": 0.2861,
+      "step": 9610
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00015458095087537218,
+      "loss": 0.2862,
+      "step": 9615
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00015452244047522502,
+      "loss": 0.2864,
+      "step": 9620
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.00015446390350150273,
+      "loss": 0.287,
+      "step": 9625
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00015440533998273547,
+      "loss": 0.2856,
+      "step": 9630
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0001543467499474665,
+      "loss": 0.2856,
+      "step": 9635
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.00015428813342425177,
+      "loss": 0.2888,
+      "step": 9640
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0001542294904416603,
+      "loss": 0.2865,
+      "step": 9645
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.000154170821028274,
+      "loss": 0.2859,
+      "step": 9650
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.00015411212521268758,
+      "loss": 0.2868,
+      "step": 9655
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00015405340302350871,
+      "loss": 0.2863,
+      "step": 9660
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00015399465448935788,
+      "loss": 0.2864,
+      "step": 9665
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.00015393587963886835,
+      "loss": 0.2867,
+      "step": 9670
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0001538770785006863,
+      "loss": 0.2889,
+      "step": 9675
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.00015381825110347074,
+      "loss": 0.2883,
+      "step": 9680
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.00015375939747589335,
+      "loss": 0.2839,
+      "step": 9685
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0001537005176466387,
+      "loss": 0.2889,
+      "step": 9690
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0001536416116444041,
+      "loss": 0.2856,
+      "step": 9695
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.00015358267949789966,
+      "loss": 0.2863,
+      "step": 9700
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.00015352372123584814,
+      "loss": 0.2852,
+      "step": 9705
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00015346473688698513,
+      "loss": 0.2866,
+      "step": 9710
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.00015340572648005888,
+      "loss": 0.2868,
+      "step": 9715
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0001533466900438303,
+      "loss": 0.2868,
+      "step": 9720
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.000153287627607073,
+      "loss": 0.287,
+      "step": 9725
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0001532285391985734,
+      "loss": 0.2864,
+      "step": 9730
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0001531694248471304,
+      "loss": 0.2862,
+      "step": 9735
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00015311028458155567,
+      "loss": 0.286,
+      "step": 9740
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0001530511184306734,
+      "loss": 0.2859,
+      "step": 9745
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0001529919264233205,
+      "loss": 0.2868,
+      "step": 9750
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.00015293270858834644,
+      "loss": 0.287,
+      "step": 9755
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.00015287346495461315,
+      "loss": 0.2871,
+      "step": 9760
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.00015281419555099546,
+      "loss": 0.2857,
+      "step": 9765
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.00015275490040638038,
+      "loss": 0.2865,
+      "step": 9770
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.00015269557954966778,
+      "loss": 0.2854,
+      "step": 9775
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00015263623300976978,
+      "loss": 0.2855,
+      "step": 9780
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.00015257686081561134,
+      "loss": 0.2873,
+      "step": 9785
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0001525174629961296,
+      "loss": 0.286,
+      "step": 9790
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.00015245803958027434,
+      "loss": 0.2876,
+      "step": 9795
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.00015239859059700794,
+      "loss": 0.2852,
+      "step": 9800
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00015233911607530497,
+      "loss": 0.2876,
+      "step": 9805
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.00015227961604415266,
+      "loss": 0.2873,
+      "step": 9810
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0001522200905325506,
+      "loss": 0.287,
+      "step": 9815
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0001521605395695108,
+      "loss": 0.2859,
+      "step": 9820
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.00015210096318405767,
+      "loss": 0.2867,
+      "step": 9825
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00015204136140522799,
+      "loss": 0.2878,
+      "step": 9830
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.00015198173426207094,
+      "loss": 0.2884,
+      "step": 9835
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.00015192208178364816,
+      "loss": 0.2865,
+      "step": 9840
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.00015186240399903342,
+      "loss": 0.2868,
+      "step": 9845
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00015180270093731303,
+      "loss": 0.2864,
+      "step": 9850
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0001517429726275855,
+      "loss": 0.2882,
+      "step": 9855
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00015168321909896172,
+      "loss": 0.2873,
+      "step": 9860
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.00015162344038056476,
+      "loss": 0.2873,
+      "step": 9865
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00015156363650153012,
+      "loss": 0.2876,
+      "step": 9870
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00015150380749100545,
+      "loss": 0.2878,
+      "step": 9875
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00015144395337815064,
+      "loss": 0.287,
+      "step": 9880
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00015138407419213796,
+      "loss": 0.2877,
+      "step": 9885
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0001513241699621517,
+      "loss": 0.2876,
+      "step": 9890
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0001512642407173885,
+      "loss": 0.2875,
+      "step": 9895
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.00015120428648705717,
+      "loss": 0.2866,
+      "step": 9900
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0001511443073003786,
+      "loss": 0.2868,
+      "step": 9905
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.000151084303186586,
+      "loss": 0.2861,
+      "step": 9910
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0001510242741749246,
+      "loss": 0.2859,
+      "step": 9915
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.00015096422029465178,
+      "loss": 0.2872,
+      "step": 9920
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.00015090414157503714,
+      "loss": 0.2883,
+      "step": 9925
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0001508440380453623,
+      "loss": 0.2877,
+      "step": 9930
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00015078390973492092,
+      "loss": 0.285,
+      "step": 9935
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.00015072375667301893,
+      "loss": 0.2891,
+      "step": 9940
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0001506635788889741,
+      "loss": 0.2864,
+      "step": 9945
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00015060337641211637,
+      "loss": 0.2873,
+      "step": 9950
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0001505431492717878,
+      "loss": 0.2854,
+      "step": 9955
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0001504828974973422,
+      "loss": 0.2867,
+      "step": 9960
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.00015042262111814565,
+      "loss": 0.2858,
+      "step": 9965
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0001503623201635761,
+      "loss": 0.2871,
+      "step": 9970
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.00015030199466302353,
+      "loss": 0.2874,
+      "step": 9975
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00015024164464588982,
+      "loss": 0.2867,
+      "step": 9980
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.00015018127014158886,
+      "loss": 0.2856,
+      "step": 9985
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00015012087117954642,
+      "loss": 0.289,
+      "step": 9990
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0001500604477892003,
+      "loss": 0.2879,
+      "step": 9995
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.2868,
+      "step": 10000
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00014993952784140717,
+      "loss": 0.2878,
+      "step": 10005
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.00014987903134289508,
+      "loss": 0.2874,
+      "step": 10010
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0001498185105339491,
+      "loss": 0.2869,
+      "step": 10015
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00014975796544406625,
+      "loss": 0.2865,
+      "step": 10020
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00014969739610275556,
+      "loss": 0.2875,
+      "step": 10025
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0001496368025395377,
+      "loss": 0.2853,
+      "step": 10030
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.00014957618478394529,
+      "loss": 0.2863,
+      "step": 10035
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.00014951554286552266,
+      "loss": 0.2888,
+      "step": 10040
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.00014945487681382598,
+      "loss": 0.2888,
+      "step": 10045
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0001493941866584231,
+      "loss": 0.2877,
+      "step": 10050
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0001493334724288937,
+      "loss": 0.2849,
+      "step": 10055
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.00014927273415482915,
+      "loss": 0.2854,
+      "step": 10060
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.00014921197186583255,
+      "loss": 0.287,
+      "step": 10065
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0001491511855915187,
+      "loss": 0.286,
+      "step": 10070
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00014909037536151409,
+      "loss": 0.2853,
+      "step": 10075
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00014902954120545687,
+      "loss": 0.2874,
+      "step": 10080
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.00014896868315299693,
+      "loss": 0.2879,
+      "step": 10085
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00014890780123379564,
+      "loss": 0.2861,
+      "step": 10090
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0001488468954775262,
+      "loss": 0.2883,
+      "step": 10095
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0001487859659138733,
+      "loss": 0.2857,
+      "step": 10100
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.00014872501257253323,
+      "loss": 0.2866,
+      "step": 10105
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.000148664035483214,
+      "loss": 0.2862,
+      "step": 10110
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.00014860303467563503,
+      "loss": 0.2871,
+      "step": 10115
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0001485420101795274,
+      "loss": 0.2862,
+      "step": 10120
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.00014848096202463372,
+      "loss": 0.2862,
+      "step": 10125
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.00014841989024070809,
+      "loss": 0.2885,
+      "step": 10130
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00014835879485751617,
+      "loss": 0.285,
+      "step": 10135
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.00014829767590483506,
+      "loss": 0.2872,
+      "step": 10140
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.00014823653341245353,
+      "loss": 0.2884,
+      "step": 10145
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00014817536741017152,
+      "loss": 0.2884,
+      "step": 10150
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.00014811417792780075,
+      "loss": 0.2845,
+      "step": 10155
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00014805296499516407,
+      "loss": 0.2871,
+      "step": 10160
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.00014799172864209608,
+      "loss": 0.2874,
+      "step": 10165
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0001479304688984425,
+      "loss": 0.2856,
+      "step": 10170
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0001478691857940607,
+      "loss": 0.2873,
+      "step": 10175
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00014780787935881923,
+      "loss": 0.2866,
+      "step": 10180
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.00014774654962259812,
+      "loss": 0.2869,
+      "step": 10185
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0001476851966152888,
+      "loss": 0.2879,
+      "step": 10190
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0001476238203667939,
+      "loss": 0.2862,
+      "step": 10195
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.00014756242090702756,
+      "loss": 0.2856,
+      "step": 10200
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.00014750099826591498,
+      "loss": 0.2867,
+      "step": 10205
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00014743955247339293,
+      "loss": 0.2859,
+      "step": 10210
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.00014737808355940932,
+      "loss": 0.2857,
+      "step": 10215
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.00014731659155392332,
+      "loss": 0.2868,
+      "step": 10220
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.00014725507648690543,
+      "loss": 0.2841,
+      "step": 10225
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0001471935383883373,
+      "loss": 0.2865,
+      "step": 10230
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.00014713197728821183,
+      "loss": 0.2851,
+      "step": 10235
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0001470703932165333,
+      "loss": 0.2846,
+      "step": 10240
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00014700878620331684,
+      "loss": 0.2877,
+      "step": 10245
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.00014694715627858908,
+      "loss": 0.2883,
+      "step": 10250
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0001468855034723877,
+      "loss": 0.2874,
+      "step": 10255
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00014682382781476146,
+      "loss": 0.2869,
+      "step": 10260
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0001467621293357704,
+      "loss": 0.2864,
+      "step": 10265
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.00014670040806548555,
+      "loss": 0.2867,
+      "step": 10270
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.00014663866403398913,
+      "loss": 0.288,
+      "step": 10275
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.00014657689727137443,
+      "loss": 0.2854,
+      "step": 10280
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.00014651510780774583,
+      "loss": 0.2849,
+      "step": 10285
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0001464532956732188,
+      "loss": 0.2862,
+      "step": 10290
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0001463914608979197,
+      "loss": 0.2856,
+      "step": 10295
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.00014632960351198618,
+      "loss": 0.2875,
+      "step": 10300
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0001462677235455667,
+      "loss": 0.2849,
+      "step": 10305
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.00014620582102882089,
+      "loss": 0.2888,
+      "step": 10310
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.00014614389599191917,
+      "loss": 0.287,
+      "step": 10315
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0001460819484650431,
+      "loss": 0.2867,
+      "step": 10320
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00014601997847838518,
+      "loss": 0.2866,
+      "step": 10325
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00014595798606214882,
+      "loss": 0.2864,
+      "step": 10330
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00014589597124654833,
+      "loss": 0.2855,
+      "step": 10335
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.000145833934061809,
+      "loss": 0.2854,
+      "step": 10340
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.000145771874538167,
+      "loss": 0.2855,
+      "step": 10345
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00014570979270586945,
+      "loss": 0.287,
+      "step": 10350
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00014564768859517418,
+      "loss": 0.2868,
+      "step": 10355
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00014558556223635003,
+      "loss": 0.2869,
+      "step": 10360
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.00014552341365967658,
+      "loss": 0.2876,
+      "step": 10365
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0001454612428954444,
+      "loss": 0.2857,
+      "step": 10370
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.00014539904997395468,
+      "loss": 0.2864,
+      "step": 10375
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00014533683492551952,
+      "loss": 0.2857,
+      "step": 10380
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0001452745977804618,
+      "loss": 0.2852,
+      "step": 10385
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.00014521233856911508,
+      "loss": 0.2847,
+      "step": 10390
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.00014515005732182383,
+      "loss": 0.2872,
+      "step": 10395
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00014508775406894307,
+      "loss": 0.2851,
+      "step": 10400
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.00014502542884083875,
+      "loss": 0.2871,
+      "step": 10405
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0001449630816678874,
+      "loss": 0.2854,
+      "step": 10410
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.00014490071258047623,
+      "loss": 0.2894,
+      "step": 10415
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.00014483832160900326,
+      "loss": 0.287,
+      "step": 10420
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.00014477590878387696,
+      "loss": 0.2862,
+      "step": 10425
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.00014471347413551672,
+      "loss": 0.2856,
+      "step": 10430
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.00014465101769435234,
+      "loss": 0.2872,
+      "step": 10435
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00014458853949082443,
+      "loss": 0.287,
+      "step": 10440
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.00014452603955538397,
+      "loss": 0.2859,
+      "step": 10445
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00014446351791849276,
+      "loss": 0.2879,
+      "step": 10450
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.00014440097461062307,
+      "loss": 0.2866,
+      "step": 10455
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.00014433840966225772,
+      "loss": 0.2867,
+      "step": 10460
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0001442758231038902,
+      "loss": 0.2882,
+      "step": 10465
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.00014421321496602428,
+      "loss": 0.2862,
+      "step": 10470
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.00014415058527917452,
+      "loss": 0.286,
+      "step": 10475
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.00014408793407386588,
+      "loss": 0.2865,
+      "step": 10480
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.00014402526138063373,
+      "loss": 0.2869,
+      "step": 10485
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.000143962567230024,
+      "loss": 0.2884,
+      "step": 10490
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.00014389985165259308,
+      "loss": 0.2861,
+      "step": 10495
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00014383711467890774,
+      "loss": 0.2875,
+      "step": 10500
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00014377435633954527,
+      "loss": 0.2849,
+      "step": 10505
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0001437115766650933,
+      "loss": 0.2857,
+      "step": 10510
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0001436487756861499,
+      "loss": 0.2857,
+      "step": 10515
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.00014358595343332342,
+      "loss": 0.2871,
+      "step": 10520
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.00014352310993723277,
+      "loss": 0.2869,
+      "step": 10525
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.00014346024522850703,
+      "loss": 0.2872,
+      "step": 10530
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.00014339735933778576,
+      "loss": 0.2875,
+      "step": 10535
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00014333445229571873,
+      "loss": 0.2859,
+      "step": 10540
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.00014327152413296608,
+      "loss": 0.2843,
+      "step": 10545
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00014320857488019824,
+      "loss": 0.2869,
+      "step": 10550
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0001431456045680959,
+      "loss": 0.2847,
+      "step": 10555
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.00014308261322735005,
+      "loss": 0.283,
+      "step": 10560
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.00014301960088866186,
+      "loss": 0.2869,
+      "step": 10565
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00014295656758274284,
+      "loss": 0.2869,
+      "step": 10570
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0001428935133403146,
+      "loss": 0.2875,
+      "step": 10575
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.00014283043819210905,
+      "loss": 0.2879,
+      "step": 10580
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.00014276734216886821,
+      "loss": 0.2857,
+      "step": 10585
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.00014270422530134432,
+      "loss": 0.2867,
+      "step": 10590
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0001426410876202999,
+      "loss": 0.2866,
+      "step": 10595
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.00014257792915650728,
+      "loss": 0.2849,
+      "step": 10600
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00014251474994074928,
+      "loss": 0.2868,
+      "step": 10605
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0001424515500038186,
+      "loss": 0.2856,
+      "step": 10610
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00014238832937651816,
+      "loss": 0.2864,
+      "step": 10615
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.00014232508808966098,
+      "loss": 0.2885,
+      "step": 10620
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00014226182617406996,
+      "loss": 0.2871,
+      "step": 10625
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0001421985436605783,
+      "loss": 0.2865,
+      "step": 10630
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0001421352405800291,
+      "loss": 0.286,
+      "step": 10635
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.00014207191696327548,
+      "loss": 0.2871,
+      "step": 10640
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00014200857284118066,
+      "loss": 0.2857,
+      "step": 10645
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.00014194520824461771,
+      "loss": 0.2878,
+      "step": 10650
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.00014188182320446985,
+      "loss": 0.2892,
+      "step": 10655
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00014181841775163013,
+      "loss": 0.2882,
+      "step": 10660
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.00014175499191700167,
+      "loss": 0.2883,
+      "step": 10665
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.00014169154573149737,
+      "loss": 0.2864,
+      "step": 10670
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.00014162807922604012,
+      "loss": 0.2867,
+      "step": 10675
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0001415645924315628,
+      "loss": 0.2863,
+      "step": 10680
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00014150108537900805,
+      "loss": 0.2874,
+      "step": 10685
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.00014143755809932845,
+      "loss": 0.2865,
+      "step": 10690
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.00014137401062348638,
+      "loss": 0.2875,
+      "step": 10695
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0001413104429824542,
+      "loss": 0.2877,
+      "step": 10700
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.00014124685520721392,
+      "loss": 0.2868,
+      "step": 10705
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.00014118324732875748,
+      "loss": 0.286,
+      "step": 10710
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.00014111961937808665,
+      "loss": 0.2866,
+      "step": 10715
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0001410559713862128,
+      "loss": 0.2868,
+      "step": 10720
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.00014099230338415728,
+      "loss": 0.2893,
+      "step": 10725
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.00014092861540295108,
+      "loss": 0.2862,
+      "step": 10730
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.00014086490747363493,
+      "loss": 0.2852,
+      "step": 10735
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0001408011796272593,
+      "loss": 0.2853,
+      "step": 10740
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.00014073743189488435,
+      "loss": 0.2886,
+      "step": 10745
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.00014067366430758004,
+      "loss": 0.2852,
+      "step": 10750
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00014060987689642581,
+      "loss": 0.2853,
+      "step": 10755
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.00014054606969251095,
+      "loss": 0.2884,
+      "step": 10760
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00014048224272693424,
+      "loss": 0.2873,
+      "step": 10765
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.00014041839603080422,
+      "loss": 0.2858,
+      "step": 10770
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.00014035452963523902,
+      "loss": 0.2853,
+      "step": 10775
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00014029064357136628,
+      "loss": 0.2869,
+      "step": 10780
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00014022673787032332,
+      "loss": 0.2849,
+      "step": 10785
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.000140162812563257,
+      "loss": 0.2866,
+      "step": 10790
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00014009886768132375,
+      "loss": 0.2857,
+      "step": 10795
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.00014003490325568954,
+      "loss": 0.289,
+      "step": 10800
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.00013997091931752977,
+      "loss": 0.2863,
+      "step": 10805
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.00013990691589802954,
+      "loss": 0.2855,
+      "step": 10810
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.00013984289302838328,
+      "loss": 0.2862,
+      "step": 10815
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.000139778850739795,
+      "loss": 0.2904,
+      "step": 10820
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.00013971478906347806,
+      "loss": 0.2856,
+      "step": 10825
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.00013965070803065543,
+      "loss": 0.2854,
+      "step": 10830
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00013958660767255938,
+      "loss": 0.2847,
+      "step": 10835
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.00013952248802043165,
+      "loss": 0.2869,
+      "step": 10840
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0001394583491055234,
+      "loss": 0.2871,
+      "step": 10845
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.00013939419095909512,
+      "loss": 0.2879,
+      "step": 10850
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.00013933001361241673,
+      "loss": 0.285,
+      "step": 10855
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.00013926581709676751,
+      "loss": 0.2852,
+      "step": 10860
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00013920160144343603,
+      "loss": 0.2873,
+      "step": 10865
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.00013913736668372026,
+      "loss": 0.2868,
+      "step": 10870
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00013907311284892736,
+      "loss": 0.2867,
+      "step": 10875
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.00013900883997037397,
+      "loss": 0.2868,
+      "step": 10880
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.00013894454807938586,
+      "loss": 0.286,
+      "step": 10885
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0001388802372072981,
+      "loss": 0.2862,
+      "step": 10890
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.00013881590738545508,
+      "loss": 0.2867,
+      "step": 10895
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0001387515586452103,
+      "loss": 0.2877,
+      "step": 10900
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.00013868719101792665,
+      "loss": 0.2876,
+      "step": 10905
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.000138622804534976,
+      "loss": 0.2871,
+      "step": 10910
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.00013855839922773968,
+      "loss": 0.2861,
+      "step": 10915
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.00013849397512760795,
+      "loss": 0.287,
+      "step": 10920
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.00013842953226598037,
+      "loss": 0.2868,
+      "step": 10925
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.00013836507067426564,
+      "loss": 0.2859,
+      "step": 10930
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0001383005903838815,
+      "loss": 0.2878,
+      "step": 10935
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0001382360914262549,
+      "loss": 0.2865,
+      "step": 10940
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.00013817157383282184,
+      "loss": 0.2878,
+      "step": 10945
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00013810703763502744,
+      "loss": 0.2858,
+      "step": 10950
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.00013804248286432578,
+      "loss": 0.2855,
+      "step": 10955
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00013797790955218014,
+      "loss": 0.2846,
+      "step": 10960
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0001379133177300627,
+      "loss": 0.2853,
+      "step": 10965
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.00013784870742945482,
+      "loss": 0.2869,
+      "step": 10970
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00013778407868184672,
+      "loss": 0.2871,
+      "step": 10975
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.00013771943151873767,
+      "loss": 0.2858,
+      "step": 10980
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.00013765476597163594,
+      "loss": 0.2851,
+      "step": 10985
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0001375900820720587,
+      "loss": 0.2868,
+      "step": 10990
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0001375253798515321,
+      "loss": 0.2892,
+      "step": 10995
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00013746065934159123,
+      "loss": 0.2854,
+      "step": 11000
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.00013739592057378003,
+      "loss": 0.2875,
+      "step": 11005
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0001373311635796515,
+      "loss": 0.2864,
+      "step": 11010
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0001372663883907673,
+      "loss": 0.2878,
+      "step": 11015
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.00013720159503869815,
+      "loss": 0.2876,
+      "step": 11020
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.00013713678355502351,
+      "loss": 0.2875,
+      "step": 11025
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00013707195397133165,
+      "loss": 0.2864,
+      "step": 11030
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.00013700710631921984,
+      "loss": 0.2878,
+      "step": 11035
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00013694224063029396,
+      "loss": 0.2882,
+      "step": 11040
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.00013687735693616876,
+      "loss": 0.2867,
+      "step": 11045
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.00013681245526846783,
+      "loss": 0.2853,
+      "step": 11050
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00013674753565882334,
+      "loss": 0.2861,
+      "step": 11055
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.00013668259813887643,
+      "loss": 0.288,
+      "step": 11060
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.00013661764274027678,
+      "loss": 0.287,
+      "step": 11065
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0001365526694946829,
+      "loss": 0.2875,
+      "step": 11070
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.00013648767843376196,
+      "loss": 0.2864,
+      "step": 11075
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00013642266958918984,
+      "loss": 0.2855,
+      "step": 11080
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.000136357642992651,
+      "loss": 0.2875,
+      "step": 11085
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.00013629259867583863,
+      "loss": 0.2868,
+      "step": 11090
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00013622753667045457,
+      "loss": 0.2878,
+      "step": 11095
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00013616245700820922,
+      "loss": 0.2866,
+      "step": 11100
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00013609735972082166,
+      "loss": 0.2875,
+      "step": 11105
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.00013603224484001948,
+      "loss": 0.2861,
+      "step": 11110
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0001359671123975389,
+      "loss": 0.2869,
+      "step": 11115
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00013590196242512463,
+      "loss": 0.2857,
+      "step": 11120
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00013583679495453,
+      "loss": 0.286,
+      "step": 11125
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.00013577161001751694,
+      "loss": 0.2855,
+      "step": 11130
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.00013570640764585566,
+      "loss": 0.2873,
+      "step": 11135
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.00013564118787132506,
+      "loss": 0.2867,
+      "step": 11140
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0001355759507257125,
+      "loss": 0.2872,
+      "step": 11145
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0001355106962408137,
+      "loss": 0.2884,
+      "step": 11150
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.00013544542444843299,
+      "loss": 0.2857,
+      "step": 11155
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.00013538013538038295,
+      "loss": 0.2859,
+      "step": 11160
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.00013531482906848475,
+      "loss": 0.2849,
+      "step": 11165
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.00013524950554456784,
+      "loss": 0.2852,
+      "step": 11170
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00013518416484047018,
+      "loss": 0.2876,
+      "step": 11175
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00013511880698803801,
+      "loss": 0.2858,
+      "step": 11180
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0001350534320191259,
+      "loss": 0.2882,
+      "step": 11185
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00013498803996559692,
+      "loss": 0.2873,
+      "step": 11190
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00013492263085932224,
+      "loss": 0.2855,
+      "step": 11195
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.00013485720473218154,
+      "loss": 0.2879,
+      "step": 11200
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0001347917616160627,
+      "loss": 0.2877,
+      "step": 11205
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0001347263015428619,
+      "loss": 0.2857,
+      "step": 11210
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00013466082454448362,
+      "loss": 0.2883,
+      "step": 11215
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0001345953306528405,
+      "loss": 0.2865,
+      "step": 11220
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.00013452981989985348,
+      "loss": 0.2869,
+      "step": 11225
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0001344642923174517,
+      "loss": 0.287,
+      "step": 11230
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.00013439874793757254,
+      "loss": 0.2877,
+      "step": 11235
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.00013433318679216153,
+      "loss": 0.2871,
+      "step": 11240
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.00013426760891317236,
+      "loss": 0.286,
+      "step": 11245
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00013420201433256689,
+      "loss": 0.2859,
+      "step": 11250
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0001341364030823151,
+      "loss": 0.2867,
+      "step": 11255
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0001340707751943952,
+      "loss": 0.2861,
+      "step": 11260
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0001340051307007933,
+      "loss": 0.2852,
+      "step": 11265
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.00013393946963350382,
+      "loss": 0.2885,
+      "step": 11270
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.00013387379202452917,
+      "loss": 0.2853,
+      "step": 11275
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.00013380809790587974,
+      "loss": 0.2846,
+      "step": 11280
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.00013374238730957412,
+      "loss": 0.2845,
+      "step": 11285
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00013367666026763882,
+      "loss": 0.2867,
+      "step": 11290
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00013361091681210845,
+      "loss": 0.2858,
+      "step": 11295
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.00013354515697502553,
+      "loss": 0.2854,
+      "step": 11300
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0001334793807884406,
+      "loss": 0.2874,
+      "step": 11305
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00013341358828441218,
+      "loss": 0.2871,
+      "step": 11310
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00013334777949500673,
+      "loss": 0.2869,
+      "step": 11315
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00013328195445229868,
+      "loss": 0.2862,
+      "step": 11320
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.00013321611318837032,
+      "loss": 0.2879,
+      "step": 11325
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.00013315025573531198,
+      "loss": 0.2864,
+      "step": 11330
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.00013308438212522163,
+      "loss": 0.2863,
+      "step": 11335
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.00013301849239020537,
+      "loss": 0.2867,
+      "step": 11340
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00013295258656237702,
+      "loss": 0.2847,
+      "step": 11345
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.00013288666467385833,
+      "loss": 0.2855,
+      "step": 11350
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0001328207267567788,
+      "loss": 0.2876,
+      "step": 11355
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0001327547728432757,
+      "loss": 0.2874,
+      "step": 11360
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00013268880296549425,
+      "loss": 0.2864,
+      "step": 11365
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.00013262281715558736,
+      "loss": 0.2854,
+      "step": 11370
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.00013255681544571568,
+      "loss": 0.2874,
+      "step": 11375
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.00013249079786804765,
+      "loss": 0.2892,
+      "step": 11380
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00013242476445475944,
+      "loss": 0.2866,
+      "step": 11385
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.00013235871523803496,
+      "loss": 0.2846,
+      "step": 11390
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.00013229265025006576,
+      "loss": 0.2865,
+      "step": 11395
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.00013222656952305113,
+      "loss": 0.2857,
+      "step": 11400
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.00013216047308919798,
+      "loss": 0.2854,
+      "step": 11405
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00013209436098072095,
+      "loss": 0.2878,
+      "step": 11410
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.00013202823322984228,
+      "loss": 0.2852,
+      "step": 11415
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0001319620898687918,
+      "loss": 0.2855,
+      "step": 11420
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00013189593092980702,
+      "loss": 0.2869,
+      "step": 11425
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00013182975644513296,
+      "loss": 0.2855,
+      "step": 11430
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.00013176356644702225,
+      "loss": 0.2867,
+      "step": 11435
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0001316973609677352,
+      "loss": 0.2862,
+      "step": 11440
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0001316311400395394,
+      "loss": 0.2869,
+      "step": 11445
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.00013156490369471027,
+      "loss": 0.2864,
+      "step": 11450
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0001314986519655305,
+      "loss": 0.2869,
+      "step": 11455
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0001314323848842904,
+      "loss": 0.2874,
+      "step": 11460
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00013136610248328778,
+      "loss": 0.2879,
+      "step": 11465
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.00013129980479482782,
+      "loss": 0.2864,
+      "step": 11470
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00013123349185122327,
+      "loss": 0.2871,
+      "step": 11475
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0001311671636847942,
+      "loss": 0.2878,
+      "step": 11480
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0001311008203278682,
+      "loss": 0.2867,
+      "step": 11485
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.00013103446181278013,
+      "loss": 0.2845,
+      "step": 11490
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.00013096808817187242,
+      "loss": 0.2865,
+      "step": 11495
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00013090169943749476,
+      "loss": 0.2883,
+      "step": 11500
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.00013083529564200417,
+      "loss": 0.2884,
+      "step": 11505
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0001307688768177651,
+      "loss": 0.2854,
+      "step": 11510
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0001307024429971492,
+      "loss": 0.2859,
+      "step": 11515
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00013063599421253558,
+      "loss": 0.2876,
+      "step": 11520
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00013056953049631057,
+      "loss": 0.2875,
+      "step": 11525
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0001305030518808678,
+      "loss": 0.2853,
+      "step": 11530
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00013043655839860803,
+      "loss": 0.2875,
+      "step": 11535
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.00013037005008193942,
+      "loss": 0.2865,
+      "step": 11540
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.00013030352696327742,
+      "loss": 0.2872,
+      "step": 11545
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.00013023698907504446,
+      "loss": 0.2861,
+      "step": 11550
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.00013017043644967036,
+      "loss": 0.2863,
+      "step": 11555
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.00013010386911959206,
+      "loss": 0.2876,
+      "step": 11560
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0001300372871172536,
+      "loss": 0.2865,
+      "step": 11565
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0001299706904751064,
+      "loss": 0.2867,
+      "step": 11570
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.00012990407922560868,
+      "loss": 0.2883,
+      "step": 11575
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.00012983745340122604,
+      "loss": 0.287,
+      "step": 11580
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00012977081303443107,
+      "loss": 0.2887,
+      "step": 11585
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0001297041581577035,
+      "loss": 0.2867,
+      "step": 11590
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.00012963748880353011,
+      "loss": 0.2854,
+      "step": 11595
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.00012957080500440468,
+      "loss": 0.2871,
+      "step": 11600
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00012950410679282815,
+      "loss": 0.2874,
+      "step": 11605
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.00012943739420130836,
+      "loss": 0.2882,
+      "step": 11610
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0001293706672623603,
+      "loss": 0.2872,
+      "step": 11615
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.00012930392600850573,
+      "loss": 0.2861,
+      "step": 11620
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.00012923717047227368,
+      "loss": 0.2849,
+      "step": 11625
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0001291704006861999,
+      "loss": 0.2853,
+      "step": 11630
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.00012910361668282719,
+      "loss": 0.2867,
+      "step": 11635
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.00012903681849470527,
+      "loss": 0.2868,
+      "step": 11640
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.00012897000615439075,
+      "loss": 0.2893,
+      "step": 11645
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.00012890317969444716,
+      "loss": 0.2862,
+      "step": 11650
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00012883633914744492,
+      "loss": 0.2841,
+      "step": 11655
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00012876948454596128,
+      "loss": 0.287,
+      "step": 11660
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.00012870261592258037,
+      "loss": 0.2874,
+      "step": 11665
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.00012863573330989313,
+      "loss": 0.2867,
+      "step": 11670
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.00012856883674049736,
+      "loss": 0.2857,
+      "step": 11675
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0001285019262469976,
+      "loss": 0.2861,
+      "step": 11680
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.00012843500186200527,
+      "loss": 0.2863,
+      "step": 11685
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.00012836806361813844,
+      "loss": 0.2864,
+      "step": 11690
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.00012830111154802203,
+      "loss": 0.2864,
+      "step": 11695
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.00012823414568428768,
+      "loss": 0.2858,
+      "step": 11700
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.00012816716605957367,
+      "loss": 0.2877,
+      "step": 11705
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00012810017270652513,
+      "loss": 0.287,
+      "step": 11710
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.00012803316565779377,
+      "loss": 0.2858,
+      "step": 11715
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.000127966144946038,
+      "loss": 0.2857,
+      "step": 11720
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00012789911060392294,
+      "loss": 0.2866,
+      "step": 11725
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0001278320626641203,
+      "loss": 0.2866,
+      "step": 11730
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00012776500115930842,
+      "loss": 0.2865,
+      "step": 11735
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.00012769792612217224,
+      "loss": 0.2849,
+      "step": 11740
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.00012763083758540337,
+      "loss": 0.2879,
+      "step": 11745
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0001275637355816999,
+      "loss": 0.2865,
+      "step": 11750
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.00012749662014376663,
+      "loss": 0.2877,
+      "step": 11755
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.00012742949130431467,
+      "loss": 0.286,
+      "step": 11760
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00012736234909606185,
+      "loss": 0.2863,
+      "step": 11765
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.00012729519355173254,
+      "loss": 0.2882,
+      "step": 11770
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00012722802470405744,
+      "loss": 0.2872,
+      "step": 11775
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.00012716084258577388,
+      "loss": 0.2857,
+      "step": 11780
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0001270936472296256,
+      "loss": 0.2855,
+      "step": 11785
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.00012702643866836278,
+      "loss": 0.2858,
+      "step": 11790
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.00012695921693474212,
+      "loss": 0.286,
+      "step": 11795
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.00012689198206152657,
+      "loss": 0.2879,
+      "step": 11800
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0001268247340814857,
+      "loss": 0.2855,
+      "step": 11805
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.00012675747302739527,
+      "loss": 0.288,
+      "step": 11810
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.00012669019893203759,
+      "loss": 0.2855,
+      "step": 11815
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.00012662291182820114,
+      "loss": 0.2862,
+      "step": 11820
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.00012655561174868088,
+      "loss": 0.2874,
+      "step": 11825
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00012648829872627807,
+      "loss": 0.2871,
+      "step": 11830
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00012642097279380027,
+      "loss": 0.2872,
+      "step": 11835
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00012635363398406127,
+      "loss": 0.2873,
+      "step": 11840
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0001262862823298812,
+      "loss": 0.2854,
+      "step": 11845
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.00012621891786408648,
+      "loss": 0.2881,
+      "step": 11850
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00012615154061950968,
+      "loss": 0.2848,
+      "step": 11855
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.00012608415062898972,
+      "loss": 0.2857,
+      "step": 11860
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00012601674792537157,
+      "loss": 0.2854,
+      "step": 11865
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.00012594933254150655,
+      "loss": 0.2885,
+      "step": 11870
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.00012588190451025207,
+      "loss": 0.2853,
+      "step": 11875
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0001258144638644718,
+      "loss": 0.2872,
+      "step": 11880
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0001257470106370354,
+      "loss": 0.2867,
+      "step": 11885
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.00012567954486081878,
+      "loss": 0.286,
+      "step": 11890
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.00012561206656870396,
+      "loss": 0.2872,
+      "step": 11895
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00012554457579357905,
+      "loss": 0.2872,
+      "step": 11900
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.00012547707256833823,
+      "loss": 0.2878,
+      "step": 11905
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.00012540955692588173,
+      "loss": 0.286,
+      "step": 11910
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00012534202889911584,
+      "loss": 0.286,
+      "step": 11915
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00012527448852095295,
+      "loss": 0.2855,
+      "step": 11920
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0001252069358243114,
+      "loss": 0.2869,
+      "step": 11925
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0001251393708421155,
+      "loss": 0.2867,
+      "step": 11930
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0001250717936072957,
+      "loss": 0.2877,
+      "step": 11935
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.00012500420415278822,
+      "loss": 0.2851,
+      "step": 11940
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0001249366025115354,
+      "loss": 0.2869,
+      "step": 11945
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0001248689887164855,
+      "loss": 0.287,
+      "step": 11950
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00012480136280059256,
+      "loss": 0.2866,
+      "step": 11955
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00012473372479681672,
+      "loss": 0.2883,
+      "step": 11960
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.00012466607473812387,
+      "loss": 0.2857,
+      "step": 11965
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0001245984126574858,
+      "loss": 0.287,
+      "step": 11970
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.00012453073858788026,
+      "loss": 0.2862,
+      "step": 11975
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00012446305256229073,
+      "loss": 0.2866,
+      "step": 11980
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00012439535461370658,
+      "loss": 0.2883,
+      "step": 11985
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00012432764477512292,
+      "loss": 0.2889,
+      "step": 11990
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00012425992307954075,
+      "loss": 0.2877,
+      "step": 11995
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00012419218955996676,
+      "loss": 0.287,
+      "step": 12000
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0001241244442494135,
+      "loss": 0.287,
+      "step": 12005
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.00012405668718089917,
+      "loss": 0.2854,
+      "step": 12010
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00012398891838744778,
+      "loss": 0.2863,
+      "step": 12015
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00012392113790208895,
+      "loss": 0.286,
+      "step": 12020
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0001238533457578581,
+      "loss": 0.2878,
+      "step": 12025
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0001237855419877963,
+      "loss": 0.2865,
+      "step": 12030
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0001237177266249503,
+      "loss": 0.2871,
+      "step": 12035
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.00012364989970237248,
+      "loss": 0.2862,
+      "step": 12040
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.00012358206125312083,
+      "loss": 0.2874,
+      "step": 12045
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.000123514211310259,
+      "loss": 0.2871,
+      "step": 12050
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00012344634990685624,
+      "loss": 0.2863,
+      "step": 12055
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00012337847707598738,
+      "loss": 0.287,
+      "step": 12060
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0001233105928507328,
+      "loss": 0.2868,
+      "step": 12065
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0001232426972641784,
+      "loss": 0.2862,
+      "step": 12070
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.00012317479034941573,
+      "loss": 0.288,
+      "step": 12075
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00012310687213954182,
+      "loss": 0.2878,
+      "step": 12080
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.00012303894266765908,
+      "loss": 0.2833,
+      "step": 12085
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.00012297100196687557,
+      "loss": 0.2868,
+      "step": 12090
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.00012290305007030478,
+      "loss": 0.2858,
+      "step": 12095
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.00012283508701106557,
+      "loss": 0.2869,
+      "step": 12100
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0001227671128222824,
+      "loss": 0.2869,
+      "step": 12105
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.00012269912753708502,
+      "loss": 0.2861,
+      "step": 12110
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0001226311311886086,
+      "loss": 0.2863,
+      "step": 12115
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00012256312380999376,
+      "loss": 0.2859,
+      "step": 12120
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0001224951054343865,
+      "loss": 0.2882,
+      "step": 12125
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.00012242707609493814,
+      "loss": 0.2857,
+      "step": 12130
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0001223590358248053,
+      "loss": 0.2845,
+      "step": 12135
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.00012229098465715006,
+      "loss": 0.2886,
+      "step": 12140
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.00012222292262513965,
+      "loss": 0.2856,
+      "step": 12145
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.00012215484976194676,
+      "loss": 0.2861,
+      "step": 12150
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.00012208676610074917,
+      "loss": 0.2857,
+      "step": 12155
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00012201867167473015,
+      "loss": 0.2862,
+      "step": 12160
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00012195056651707806,
+      "loss": 0.2879,
+      "step": 12165
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.00012188245066098647,
+      "loss": 0.2889,
+      "step": 12170
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.00012181432413965428,
+      "loss": 0.2888,
+      "step": 12175
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.00012174618698628549,
+      "loss": 0.2864,
+      "step": 12180
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00012167803923408934,
+      "loss": 0.2856,
+      "step": 12185
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.00012160988091628022,
+      "loss": 0.2865,
+      "step": 12190
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 0.00012154171206607764,
+      "loss": 0.2866,
+      "step": 12195
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.00012147353271670634,
+      "loss": 0.2872,
+      "step": 12200
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.000121405342901396,
+      "loss": 0.2854,
+      "step": 12205
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00012133714265338161,
+      "loss": 0.2852,
+      "step": 12210
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.00012126893200590308,
+      "loss": 0.2866,
+      "step": 12215
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00012120071099220549,
+      "loss": 0.2879,
+      "step": 12220
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.00012113247964553888,
+      "loss": 0.2865,
+      "step": 12225
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0001210642379991584,
+      "loss": 0.287,
+      "step": 12230
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.00012099598608632428,
+      "loss": 0.2857,
+      "step": 12235
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.00012092772394030152,
+      "loss": 0.2862,
+      "step": 12240
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.00012085945159436038,
+      "loss": 0.2866,
+      "step": 12245
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.00012079116908177593,
+      "loss": 0.287,
+      "step": 12250
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.00012072287643582825,
+      "loss": 0.2854,
+      "step": 12255
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.00012065457368980236,
+      "loss": 0.2872,
+      "step": 12260
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 0.00012058626087698814,
+      "loss": 0.2852,
+      "step": 12265
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.00012051793803068046,
+      "loss": 0.2878,
+      "step": 12270
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.00012044960518417903,
+      "loss": 0.2865,
+      "step": 12275
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0001203812623707885,
+      "loss": 0.285,
+      "step": 12280
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00012031290962381823,
+      "loss": 0.2869,
+      "step": 12285
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.00012024454697658261,
+      "loss": 0.2856,
+      "step": 12290
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0001201761744624007,
+      "loss": 0.2851,
+      "step": 12295
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.00012010779211459648,
+      "loss": 0.2866,
+      "step": 12300
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.00012003939996649865,
+      "loss": 0.2844,
+      "step": 12305
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00011997099805144069,
+      "loss": 0.2874,
+      "step": 12310
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00011990258640276094,
+      "loss": 0.2861,
+      "step": 12315
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.00011983416505380234,
+      "loss": 0.286,
+      "step": 12320
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.00011976573403791262,
+      "loss": 0.2863,
+      "step": 12325
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.00011969729338844429,
+      "loss": 0.2858,
+      "step": 12330
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0001196288431387544,
+      "loss": 0.2876,
+      "step": 12335
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.00011956038332220483,
+      "loss": 0.2883,
+      "step": 12340
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.00011949191397216206,
+      "loss": 0.2876,
+      "step": 12345
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0001194234351219972,
+      "loss": 0.2872,
+      "step": 12350
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00011935494680508606,
+      "loss": 0.2871,
+      "step": 12355
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0001192864490548089,
+      "loss": 0.2868,
+      "step": 12360
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.00011921794190455082,
+      "loss": 0.285,
+      "step": 12365
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.00011914942538770131,
+      "loss": 0.2864,
+      "step": 12370
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.00011908089953765449,
+      "loss": 0.2862,
+      "step": 12375
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00011901236438780902,
+      "loss": 0.2855,
+      "step": 12380
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.00011894381997156813,
+      "loss": 0.2863,
+      "step": 12385
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.00011887526632233954,
+      "loss": 0.2878,
+      "step": 12390
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00011880670347353539,
+      "loss": 0.2873,
+      "step": 12395
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00011873813145857249,
+      "loss": 0.2864,
+      "step": 12400
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0001186695503108719,
+      "loss": 0.2863,
+      "step": 12405
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0001186009600638593,
+      "loss": 0.2874,
+      "step": 12410
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.00011853236075096474,
+      "loss": 0.2875,
+      "step": 12415
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0001184637524056227,
+      "loss": 0.2872,
+      "step": 12420
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00011839513506127203,
+      "loss": 0.2878,
+      "step": 12425
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.00011832650875135598,
+      "loss": 0.2856,
+      "step": 12430
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.00011825787350932222,
+      "loss": 0.2863,
+      "step": 12435
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00011818922936862269,
+      "loss": 0.2878,
+      "step": 12440
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.00011812057636271374,
+      "loss": 0.2859,
+      "step": 12445
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.00011805191452505602,
+      "loss": 0.2862,
+      "step": 12450
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.00011798324388911444,
+      "loss": 0.2882,
+      "step": 12455
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.00011791456448835825,
+      "loss": 0.2868,
+      "step": 12460
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00011784587635626094,
+      "loss": 0.2849,
+      "step": 12465
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00011777717952630031,
+      "loss": 0.2853,
+      "step": 12470
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.00011770847403195834,
+      "loss": 0.2874,
+      "step": 12475
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.00011763975990672125,
+      "loss": 0.2866,
+      "step": 12480
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.00011757103718407947,
+      "loss": 0.2857,
+      "step": 12485
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00011750230589752762,
+      "loss": 0.2879,
+      "step": 12490
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.00011743356608056449,
+      "loss": 0.2864,
+      "step": 12495
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00011736481776669306,
+      "loss": 0.2893,
+      "step": 12500
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.00011729606098942039,
+      "loss": 0.2865,
+      "step": 12505
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.00011722729578225769,
+      "loss": 0.2864,
+      "step": 12510
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0001171585221787203,
+      "loss": 0.287,
+      "step": 12515
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00011708974021232769,
+      "loss": 0.2894,
+      "step": 12520
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00011702094991660326,
+      "loss": 0.2858,
+      "step": 12525
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.00011695215132507464,
+      "loss": 0.2867,
+      "step": 12530
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.00011688334447127338,
+      "loss": 0.2852,
+      "step": 12535
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00011681452938873516,
+      "loss": 0.2881,
+      "step": 12540
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.00011674570611099955,
+      "loss": 0.2864,
+      "step": 12545
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.00011667687467161024,
+      "loss": 0.287,
+      "step": 12550
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0001166080351041148,
+      "loss": 0.2861,
+      "step": 12555
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.00011653918744206478,
+      "loss": 0.2864,
+      "step": 12560
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.00011647033171901573,
+      "loss": 0.2872,
+      "step": 12565
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00011640146796852711,
+      "loss": 0.2865,
+      "step": 12570
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00011633259622416224,
+      "loss": 0.2865,
+      "step": 12575
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00011626371651948838,
+      "loss": 0.2856,
+      "step": 12580
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00011619482888807662,
+      "loss": 0.2851,
+      "step": 12585
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00011612593336350208,
+      "loss": 0.2867,
+      "step": 12590
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.00011605702997934345,
+      "loss": 0.285,
+      "step": 12595
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0001159881187691835,
+      "loss": 0.285,
+      "step": 12600
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.00011591919976660868,
+      "loss": 0.2877,
+      "step": 12605
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0001158502730052093,
+      "loss": 0.2859,
+      "step": 12610
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0001157813385185794,
+      "loss": 0.2845,
+      "step": 12615
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.00011571239634031679,
+      "loss": 0.2881,
+      "step": 12620
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0001156434465040231,
+      "loss": 0.2855,
+      "step": 12625
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00011557448904330362,
+      "loss": 0.2865,
+      "step": 12630
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.00011550552399176739,
+      "loss": 0.2869,
+      "step": 12635
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.00011543655138302714,
+      "loss": 0.285,
+      "step": 12640
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.00011536757125069923,
+      "loss": 0.2857,
+      "step": 12645
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.00011529858362840382,
+      "loss": 0.2854,
+      "step": 12650
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.00011522958854976458,
+      "loss": 0.2857,
+      "step": 12655
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.00011516058604840891,
+      "loss": 0.286,
+      "step": 12660
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.00011509157615796776,
+      "loss": 0.2869,
+      "step": 12665
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.00011502255891207572,
+      "loss": 0.2882,
+      "step": 12670
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.00011495353434437098,
+      "loss": 0.2868,
+      "step": 12675
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.00011488450248849522,
+      "loss": 0.288,
+      "step": 12680
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00011481546337809381,
+      "loss": 0.285,
+      "step": 12685
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0001147464170468155,
+      "loss": 0.2846,
+      "step": 12690
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00011467736352831266,
+      "loss": 0.2884,
+      "step": 12695
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.00011460830285624118,
+      "loss": 0.2853,
+      "step": 12700
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.00011453923506426032,
+      "loss": 0.286,
+      "step": 12705
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.00011447016018603292,
+      "loss": 0.2886,
+      "step": 12710
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.00011440107825522521,
+      "loss": 0.2872,
+      "step": 12715
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.00011433198930550695,
+      "loss": 0.2867,
+      "step": 12720
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00011426289337055119,
+      "loss": 0.287,
+      "step": 12725
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.00011419379048403444,
+      "loss": 0.2862,
+      "step": 12730
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.00011412468067963669,
+      "loss": 0.2872,
+      "step": 12735
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.00011405556399104109,
+      "loss": 0.2859,
+      "step": 12740
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.00011398644045193444,
+      "loss": 0.2869,
+      "step": 12745
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00011391731009600654,
+      "loss": 0.2862,
+      "step": 12750
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00011384817295695083,
+      "loss": 0.2887,
+      "step": 12755
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0001137790290684638,
+      "loss": 0.2867,
+      "step": 12760
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.00011370987846424546,
+      "loss": 0.2878,
+      "step": 12765
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.00011364072117799885,
+      "loss": 0.2875,
+      "step": 12770
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 0.00011357155724343045,
+      "loss": 0.2871,
+      "step": 12775
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.00011350238669424993,
+      "loss": 0.2877,
+      "step": 12780
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.00011343320956417014,
+      "loss": 0.2865,
+      "step": 12785
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.00011336402588690726,
+      "loss": 0.2865,
+      "step": 12790
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.00011329483569618045,
+      "loss": 0.2853,
+      "step": 12795
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.00011322563902571226,
+      "loss": 0.2857,
+      "step": 12800
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.00011315643590922827,
+      "loss": 0.2882,
+      "step": 12805
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.00011308722638045724,
+      "loss": 0.2883,
+      "step": 12810
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.00011301801047313105,
+      "loss": 0.2846,
+      "step": 12815
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.00011294878822098469,
+      "loss": 0.2861,
+      "step": 12820
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0001128795596577563,
+      "loss": 0.2862,
+      "step": 12825
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.00011281032481718697,
+      "loss": 0.2871,
+      "step": 12830
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.00011274108373302095,
+      "loss": 0.2877,
+      "step": 12835
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.00011267183643900548,
+      "loss": 0.2894,
+      "step": 12840
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.00011260258296889086,
+      "loss": 0.2864,
+      "step": 12845
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.00011253332335643043,
+      "loss": 0.2867,
+      "step": 12850
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.00011246405763538046,
+      "loss": 0.2857,
+      "step": 12855
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.00011239478583950018,
+      "loss": 0.2868,
+      "step": 12860
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.00011232550800255188,
+      "loss": 0.2874,
+      "step": 12865
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.00011225622415830068,
+      "loss": 0.2858,
+      "step": 12870
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.00011218693434051475,
+      "loss": 0.2861,
+      "step": 12875
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.00011211763858296507,
+      "loss": 0.2862,
+      "step": 12880
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.00011204833691942553,
+      "loss": 0.286,
+      "step": 12885
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00011197902938367298,
+      "loss": 0.2874,
+      "step": 12890
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.00011190971600948699,
+      "loss": 0.2858,
+      "step": 12895
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.00011184039683065013,
+      "loss": 0.287,
+      "step": 12900
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.00011177107188094764,
+      "loss": 0.2855,
+      "step": 12905
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.00011170174119416776,
+      "loss": 0.2853,
+      "step": 12910
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.00011163240480410135,
+      "loss": 0.2855,
+      "step": 12915
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 0.00011156306274454218,
+      "loss": 0.2855,
+      "step": 12920
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.00011149371504928668,
+      "loss": 0.2866,
+      "step": 12925
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00011142436175213409,
+      "loss": 0.2856,
+      "step": 12930
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.00011135500288688636,
+      "loss": 0.2867,
+      "step": 12935
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.00011128563848734816,
+      "loss": 0.2861,
+      "step": 12940
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0001112162685873269,
+      "loss": 0.2862,
+      "step": 12945
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.00011114689322063255,
+      "loss": 0.2874,
+      "step": 12950
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.00011107751242107787,
+      "loss": 0.2873,
+      "step": 12955
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.00011100812622247822,
+      "loss": 0.2863,
+      "step": 12960
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.00011093873465865157,
+      "loss": 0.2852,
+      "step": 12965
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.00011086933776341852,
+      "loss": 0.2852,
+      "step": 12970
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0001107999355706023,
+      "loss": 0.2848,
+      "step": 12975
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.00011073052811402867,
+      "loss": 0.2869,
+      "step": 12980
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.000110661115427526,
+      "loss": 0.2858,
+      "step": 12985
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0001105916975449252,
+      "loss": 0.2867,
+      "step": 12990
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.00011052227450005967,
+      "loss": 0.288,
+      "step": 12995
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.00011045284632676536,
+      "loss": 0.2853,
+      "step": 13000
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.00011038341305888074,
+      "loss": 0.2866,
+      "step": 13005
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00011031397473024674,
+      "loss": 0.2877,
+      "step": 13010
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.00011024453137470677,
+      "loss": 0.2892,
+      "step": 13015
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.00011017508302610664,
+      "loss": 0.2884,
+      "step": 13020
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.00011010562971829463,
+      "loss": 0.2877,
+      "step": 13025
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.00011003617148512149,
+      "loss": 0.286,
+      "step": 13030
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0001099667083604403,
+      "loss": 0.2853,
+      "step": 13035
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.00010989724037810652,
+      "loss": 0.2888,
+      "step": 13040
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.00010982776757197799,
+      "loss": 0.2873,
+      "step": 13045
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.00010975828997591495,
+      "loss": 0.2884,
+      "step": 13050
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.00010968880762377993,
+      "loss": 0.2875,
+      "step": 13055
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.00010961932054943778,
+      "loss": 0.286,
+      "step": 13060
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.00010954982878675563,
+      "loss": 0.287,
+      "step": 13065
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00010948033236960294,
+      "loss": 0.2869,
+      "step": 13070
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 0.00010941083133185146,
+      "loss": 0.2853,
+      "step": 13075
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.00010934132570737507,
+      "loss": 0.2859,
+      "step": 13080
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.00010927181553005002,
+      "loss": 0.2848,
+      "step": 13085
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.00010920230083375473,
+      "loss": 0.2858,
+      "step": 13090
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.00010913278165236978,
+      "loss": 0.2858,
+      "step": 13095
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.00010906325801977804,
+      "loss": 0.2864,
+      "step": 13100
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00010899372996986439,
+      "loss": 0.2887,
+      "step": 13105
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.00010892419753651606,
+      "loss": 0.2886,
+      "step": 13110
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.00010885466075362223,
+      "loss": 0.286,
+      "step": 13115
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.00010878511965507434,
+      "loss": 0.2873,
+      "step": 13120
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.00010871557427476583,
+      "loss": 0.2866,
+      "step": 13125
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0001086460246465923,
+      "loss": 0.2849,
+      "step": 13130
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.00010857647080445139,
+      "loss": 0.2863,
+      "step": 13135
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.00010850691278224281,
+      "loss": 0.2887,
+      "step": 13140
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.00010843735061386828,
+      "loss": 0.2854,
+      "step": 13145
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.00010836778433323158,
+      "loss": 0.2857,
+      "step": 13150
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0001082982139742384,
+      "loss": 0.2871,
+      "step": 13155
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.00010822863957079656,
+      "loss": 0.2865,
+      "step": 13160
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.00010815906115681578,
+      "loss": 0.2864,
+      "step": 13165
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.00010808947876620767,
+      "loss": 0.2861,
+      "step": 13170
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.00010801989243288589,
+      "loss": 0.2875,
+      "step": 13175
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.00010795030219076599,
+      "loss": 0.285,
+      "step": 13180
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.00010788070807376536,
+      "loss": 0.2865,
+      "step": 13185
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.00010781111011580336,
+      "loss": 0.2845,
+      "step": 13190
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.00010774150835080119,
+      "loss": 0.2856,
+      "step": 13195
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.00010767190281268187,
+      "loss": 0.2855,
+      "step": 13200
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.00010760229353537033,
+      "loss": 0.2867,
+      "step": 13205
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.00010753268055279329,
+      "loss": 0.2855,
+      "step": 13210
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.00010746306389887924,
+      "loss": 0.2879,
+      "step": 13215
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.00010739344360755852,
+      "loss": 0.2851,
+      "step": 13220
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.00010732381971276318,
+      "loss": 0.288,
+      "step": 13225
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0001072541922484271,
+      "loss": 0.2864,
+      "step": 13230
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.00010718456124848583,
+      "loss": 0.2849,
+      "step": 13235
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.00010711492674687671,
+      "loss": 0.2856,
+      "step": 13240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0001070452887775387,
+      "loss": 0.2864,
+      "step": 13245
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.00010697564737441252,
+      "loss": 0.2848,
+      "step": 13250
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.00010690600257144061,
+      "loss": 0.2879,
+      "step": 13255
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.00010683635440256687,
+      "loss": 0.286,
+      "step": 13260
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.00010676670290173709,
+      "loss": 0.2866,
+      "step": 13265
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.00010669704810289851,
+      "loss": 0.287,
+      "step": 13270
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00010662739004000005,
+      "loss": 0.2889,
+      "step": 13275
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.00010655772874699217,
+      "loss": 0.2844,
+      "step": 13280
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.00010648806425782695,
+      "loss": 0.2847,
+      "step": 13285
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00010641839660645805,
+      "loss": 0.2873,
+      "step": 13290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.00010634872582684061,
+      "loss": 0.2882,
+      "step": 13295
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.00010627905195293135,
+      "loss": 0.288,
+      "step": 13300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.00010620937501868841,
+      "loss": 0.2864,
+      "step": 13305
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.00010613969505807156,
+      "loss": 0.2856,
+      "step": 13310
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00010607001210504191,
+      "loss": 0.288,
+      "step": 13315
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.00010600032619356209,
+      "loss": 0.2857,
+      "step": 13320
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.00010593063735759618,
+      "loss": 0.2882,
+      "step": 13325
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00010586094563110964,
+      "loss": 0.2869,
+      "step": 13330
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.00010579125104806944,
+      "loss": 0.2856,
+      "step": 13335
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.00010572155364244382,
+      "loss": 0.2853,
+      "step": 13340
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.00010565185344820247,
+      "loss": 0.2872,
+      "step": 13345
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.00010558215049931638,
+      "loss": 0.287,
+      "step": 13350
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.00010551244482975798,
+      "loss": 0.2868,
+      "step": 13355
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.00010544273647350092,
+      "loss": 0.2863,
+      "step": 13360
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.00010537302546452022,
+      "loss": 0.2862,
+      "step": 13365
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.00010530331183679218,
+      "loss": 0.2867,
+      "step": 13370
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0001052335956242944,
+      "loss": 0.2853,
+      "step": 13375
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.00010516387686100566,
+      "loss": 0.2838,
+      "step": 13380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.00010509415558090609,
+      "loss": 0.2878,
+      "step": 13385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.00010502443181797697,
+      "loss": 0.2864,
+      "step": 13390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.00010495470560620083,
+      "loss": 0.2851,
+      "step": 13395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.00010488497697956135,
+      "loss": 0.2895,
+      "step": 13400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.00010481524597204342,
+      "loss": 0.2859,
+      "step": 13405
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.00010474551261763314,
+      "loss": 0.2867,
+      "step": 13410
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.00010467577695031762,
+      "loss": 0.2859,
+      "step": 13415
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.00010460603900408523,
+      "loss": 0.2852,
+      "step": 13420
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.00010453629881292538,
+      "loss": 0.2855,
+      "step": 13425
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.00010446655641082862,
+      "loss": 0.2843,
+      "step": 13430
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0001043968118317865,
+      "loss": 0.285,
+      "step": 13435
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.00010432706510979171,
+      "loss": 0.2863,
+      "step": 13440
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.00010425731627883797,
+      "loss": 0.2862,
+      "step": 13445
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.00010418756537291996,
+      "loss": 0.285,
+      "step": 13450
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.00010411781242603352,
+      "loss": 0.2861,
+      "step": 13455
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.00010404805747217526,
+      "loss": 0.2861,
+      "step": 13460
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.000103978300545343,
+      "loss": 0.2875,
+      "step": 13465
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.00010390854167953537,
+      "loss": 0.2872,
+      "step": 13470
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.00010383878090875201,
+      "loss": 0.2868,
+      "step": 13475
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.00010376901826699348,
+      "loss": 0.2862,
+      "step": 13480
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0001036992537882612,
+      "loss": 0.2871,
+      "step": 13485
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.00010362948750655759,
+      "loss": 0.287,
+      "step": 13490
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.00010355971945588585,
+      "loss": 0.2874,
+      "step": 13495
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.00010348994967025012,
+      "loss": 0.287,
+      "step": 13500
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0001034201781836553,
+      "loss": 0.2896,
+      "step": 13505
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.00010335040503010716,
+      "loss": 0.288,
+      "step": 13510
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.00010328063024361232,
+      "loss": 0.2876,
+      "step": 13515
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.00010321085385817817,
+      "loss": 0.2853,
+      "step": 13520
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.00010314107590781284,
+      "loss": 0.2871,
+      "step": 13525
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.00010307129642652528,
+      "loss": 0.2859,
+      "step": 13530
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.00010300151544832512,
+      "loss": 0.2842,
+      "step": 13535
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.00010293173300722285,
+      "loss": 0.2848,
+      "step": 13540
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.00010286194913722948,
+      "loss": 0.287,
+      "step": 13545
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0001027921638723569,
+      "loss": 0.2877,
+      "step": 13550
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.00010272237724661753,
+      "loss": 0.2861,
+      "step": 13555
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0001026525892940246,
+      "loss": 0.2884,
+      "step": 13560
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00010258280004859188,
+      "loss": 0.2864,
+      "step": 13565
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.00010251300954433376,
+      "loss": 0.288,
+      "step": 13570
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.00010244321781526533,
+      "loss": 0.285,
+      "step": 13575
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0001023734248954022,
+      "loss": 0.2872,
+      "step": 13580
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.00010230363081876064,
+      "loss": 0.2883,
+      "step": 13585
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.00010223383561935738,
+      "loss": 0.286,
+      "step": 13590
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.00010216403933120979,
+      "loss": 0.2866,
+      "step": 13595
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0001020942419883357,
+      "loss": 0.2861,
+      "step": 13600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00010202444362475352,
+      "loss": 0.2869,
+      "step": 13605
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.00010195464427448213,
+      "loss": 0.288,
+      "step": 13610
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.00010188484397154084,
+      "loss": 0.2872,
+      "step": 13615
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.00010181504274994948,
+      "loss": 0.2866,
+      "step": 13620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.00010174524064372837,
+      "loss": 0.2853,
+      "step": 13625
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00010167543768689815,
+      "loss": 0.2872,
+      "step": 13630
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.00010160563391347996,
+      "loss": 0.2892,
+      "step": 13635
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0001015358293574953,
+      "loss": 0.2862,
+      "step": 13640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.00010146602405296607,
+      "loss": 0.2833,
+      "step": 13645
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.00010139621803391455,
+      "loss": 0.2867,
+      "step": 13650
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0001013264113343633,
+      "loss": 0.2867,
+      "step": 13655
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.00010125660398833528,
+      "loss": 0.2853,
+      "step": 13660
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.00010118679602985373,
+      "loss": 0.2872,
+      "step": 13665
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00010111698749294223,
+      "loss": 0.2865,
+      "step": 13670
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.00010104717841162458,
+      "loss": 0.2864,
+      "step": 13675
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00010097736881992492,
+      "loss": 0.2858,
+      "step": 13680
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.00010090755875186753,
+      "loss": 0.288,
+      "step": 13685
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.00010083774824147708,
+      "loss": 0.2867,
+      "step": 13690
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.00010076793732277829,
+      "loss": 0.2874,
+      "step": 13695
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.00010069812602979615,
+      "loss": 0.2866,
+      "step": 13700
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00010062831439655591,
+      "loss": 0.2847,
+      "step": 13705
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00010055850245708283,
+      "loss": 0.2864,
+      "step": 13710
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.00010048869024540247,
+      "loss": 0.2889,
+      "step": 13715
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0001004188777955404,
+      "loss": 0.2866,
+      "step": 13720
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.00010034906514152238,
+      "loss": 0.2849,
+      "step": 13725
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.00010027925231737428,
+      "loss": 0.2873,
+      "step": 13730
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.00010020943935712192,
+      "loss": 0.2865,
+      "step": 13735
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00010013962629479146,
+      "loss": 0.2889,
+      "step": 13740
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.00010006981316440875,
+      "loss": 0.2872,
+      "step": 13745
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0001,
+      "loss": 0.2863,
+      "step": 13750
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 9.993018683559126e-05,
+      "loss": 0.2866,
+      "step": 13755
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 9.986037370520857e-05,
+      "loss": 0.2862,
+      "step": 13760
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 9.979056064287806e-05,
+      "loss": 0.2877,
+      "step": 13765
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 9.972074768262576e-05,
+      "loss": 0.2856,
+      "step": 13770
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 9.965093485847767e-05,
+      "loss": 0.2843,
+      "step": 13775
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.056640625,
+      "learning_rate": 9.958112220445963e-05,
+      "loss": 0.2886,
+      "step": 13780
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 9.951130975459757e-05,
+      "loss": 0.2875,
+      "step": 13785
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 9.94414975429172e-05,
+      "loss": 0.2872,
+      "step": 13790
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.041015625,
+      "learning_rate": 9.937168560344412e-05,
+      "loss": 0.2869,
+      "step": 13795
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 9.930187397020386e-05,
+      "loss": 0.2872,
+      "step": 13800
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 9.923206267722173e-05,
+      "loss": 0.2859,
+      "step": 13805
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 9.916225175852293e-05,
+      "loss": 0.2869,
+      "step": 13810
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 9.909244124813247e-05,
+      "loss": 0.2874,
+      "step": 13815
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 9.902263118007513e-05,
+      "loss": 0.2861,
+      "step": 13820
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 9.895282158837545e-05,
+      "loss": 0.2884,
+      "step": 13825
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.052734375,
+      "learning_rate": 9.888301250705779e-05,
+      "loss": 0.2867,
+      "step": 13830
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 9.881320397014629e-05,
+      "loss": 0.2868,
+      "step": 13835
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.05078125,
+      "learning_rate": 9.874339601166473e-05,
+      "loss": 0.2853,
+      "step": 13840
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 9.867358866563673e-05,
+      "loss": 0.2869,
+      "step": 13845
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 9.860378196608549e-05,
+      "loss": 0.2885,
+      "step": 13850
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 9.853397594703394e-05,
+      "loss": 0.285,
+      "step": 13855
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.052734375,
+      "learning_rate": 9.846417064250471e-05,
+      "loss": 0.287,
+      "step": 13860
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 9.839436608652007e-05,
+      "loss": 0.2851,
+      "step": 13865
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 9.832456231310189e-05,
+      "loss": 0.2858,
+      "step": 13870
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 9.825475935627165e-05,
+      "loss": 0.2861,
+      "step": 13875
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 9.818495725005054e-05,
+      "loss": 0.2858,
+      "step": 13880
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 9.81151560284592e-05,
+      "loss": 0.2879,
+      "step": 13885
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 9.804535572551789e-05,
+      "loss": 0.2861,
+      "step": 13890
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 9.797555637524649e-05,
+      "loss": 0.2863,
+      "step": 13895
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 9.790575801166432e-05,
+      "loss": 0.286,
+      "step": 13900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 9.783596066879022e-05,
+      "loss": 0.2872,
+      "step": 13905
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 9.776616438064264e-05,
+      "loss": 0.287,
+      "step": 13910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 9.76963691812394e-05,
+      "loss": 0.2857,
+      "step": 13915
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 9.762657510459783e-05,
+      "loss": 0.2848,
+      "step": 13920
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 9.755678218473469e-05,
+      "loss": 0.286,
+      "step": 13925
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 9.748699045566626e-05,
+      "loss": 0.2874,
+      "step": 13930
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 9.741719995140814e-05,
+      "loss": 0.2876,
+      "step": 13935
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 9.734741070597539e-05,
+      "loss": 0.2856,
+      "step": 13940
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 9.727762275338246e-05,
+      "loss": 0.2872,
+      "step": 13945
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 9.720783612764314e-05,
+      "loss": 0.2871,
+      "step": 13950
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.052734375,
+      "learning_rate": 9.713805086277054e-05,
+      "loss": 0.288,
+      "step": 13955
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 9.706826699277718e-05,
+      "loss": 0.2862,
+      "step": 13960
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 9.699848455167489e-05,
+      "loss": 0.286,
+      "step": 13965
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 9.692870357347473e-05,
+      "loss": 0.2854,
+      "step": 13970
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 9.685892409218717e-05,
+      "loss": 0.2879,
+      "step": 13975
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.048828125,
+      "learning_rate": 9.678914614182185e-05,
+      "loss": 0.2871,
+      "step": 13980
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.05078125,
+      "learning_rate": 9.671936975638768e-05,
+      "loss": 0.2858,
+      "step": 13985
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 9.664959496989285e-05,
+      "loss": 0.2873,
+      "step": 13990
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 9.657982181634475e-05,
+      "loss": 0.2854,
+      "step": 13995
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 9.651005032974994e-05,
+      "loss": 0.2849,
+      "step": 14000
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 9.644028054411416e-05,
+      "loss": 0.2868,
+      "step": 14005
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 9.637051249344243e-05,
+      "loss": 0.2851,
+      "step": 14010
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 9.630074621173883e-05,
+      "loss": 0.2857,
+      "step": 14015
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 9.623098173300654e-05,
+      "loss": 0.2869,
+      "step": 14020
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.048828125,
+      "learning_rate": 9.616121909124801e-05,
+      "loss": 0.2876,
+      "step": 14025
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 9.609145832046465e-05,
+      "loss": 0.2873,
+      "step": 14030
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.060546875,
+      "learning_rate": 9.602169945465702e-05,
+      "loss": 0.2866,
+      "step": 14035
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 9.595194252782477e-05,
+      "loss": 0.2867,
+      "step": 14040
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 9.588218757396655e-05,
+      "loss": 0.2862,
+      "step": 14045
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.044921875,
+      "learning_rate": 9.581243462708006e-05,
+      "loss": 0.2862,
+      "step": 14050
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 9.574268372116205e-05,
+      "loss": 0.2865,
+      "step": 14055
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 9.567293489020831e-05,
+      "loss": 0.2857,
+      "step": 14060
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 9.560318816821353e-05,
+      "loss": 0.287,
+      "step": 14065
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 9.55334435891714e-05,
+      "loss": 0.287,
+      "step": 14070
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 9.546370118707463e-05,
+      "loss": 0.2894,
+      "step": 14075
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 9.539396099591476e-05,
+      "loss": 0.2861,
+      "step": 14080
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 9.532422304968243e-05,
+      "loss": 0.2851,
+      "step": 14085
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 9.525448738236691e-05,
+      "loss": 0.2861,
+      "step": 14090
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.046875,
+      "learning_rate": 9.518475402795661e-05,
+      "loss": 0.2856,
+      "step": 14095
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.046875,
+      "learning_rate": 9.511502302043868e-05,
+      "loss": 0.2842,
+      "step": 14100
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 9.504529439379921e-05,
+      "loss": 0.2853,
+      "step": 14105
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 9.497556818202306e-05,
+      "loss": 0.2861,
+      "step": 14110
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 9.490584441909392e-05,
+      "loss": 0.2867,
+      "step": 14115
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 9.483612313899435e-05,
+      "loss": 0.2859,
+      "step": 14120
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.060546875,
+      "learning_rate": 9.476640437570562e-05,
+      "loss": 0.2853,
+      "step": 14125
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 9.469668816320784e-05,
+      "loss": 0.286,
+      "step": 14130
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 9.462697453547979e-05,
+      "loss": 0.2862,
+      "step": 14135
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 9.455726352649911e-05,
+      "loss": 0.2862,
+      "step": 14140
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 9.448755517024206e-05,
+      "loss": 0.2881,
+      "step": 14145
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 9.441784950068362e-05,
+      "loss": 0.2869,
+      "step": 14150
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 9.434814655179755e-05,
+      "loss": 0.2852,
+      "step": 14155
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0546875,
+      "learning_rate": 9.427844635755619e-05,
+      "loss": 0.2854,
+      "step": 14160
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 9.420874895193056e-05,
+      "loss": 0.2895,
+      "step": 14165
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 9.413905436889035e-05,
+      "loss": 0.2882,
+      "step": 14170
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 9.406936264240386e-05,
+      "loss": 0.2854,
+      "step": 14175
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 9.399967380643796e-05,
+      "loss": 0.2842,
+      "step": 14180
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 9.392998789495811e-05,
+      "loss": 0.2858,
+      "step": 14185
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 9.386030494192846e-05,
+      "loss": 0.2866,
+      "step": 14190
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.052734375,
+      "learning_rate": 9.37906249813116e-05,
+      "loss": 0.2874,
+      "step": 14195
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.046875,
+      "learning_rate": 9.372094804706867e-05,
+      "loss": 0.2866,
+      "step": 14200
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 9.36512741731594e-05,
+      "loss": 0.2874,
+      "step": 14205
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 9.358160339354194e-05,
+      "loss": 0.2874,
+      "step": 14210
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 9.351193574217306e-05,
+      "loss": 0.2874,
+      "step": 14215
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 9.344227125300788e-05,
+      "loss": 0.2849,
+      "step": 14220
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 9.337260996000002e-05,
+      "loss": 0.2859,
+      "step": 14225
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 9.330295189710152e-05,
+      "loss": 0.2882,
+      "step": 14230
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 9.323329709826294e-05,
+      "loss": 0.2882,
+      "step": 14235
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0361328125,
+      "learning_rate": 9.316364559743314e-05,
+      "loss": 0.2859,
+      "step": 14240
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 9.309399742855942e-05,
+      "loss": 0.2861,
+      "step": 14245
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.05859375,
+      "learning_rate": 9.302435262558747e-05,
+      "loss": 0.2869,
+      "step": 14250
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 9.295471122246131e-05,
+      "loss": 0.2865,
+      "step": 14255
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 9.288507325312335e-05,
+      "loss": 0.2868,
+      "step": 14260
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 9.281543875151419e-05,
+      "loss": 0.2854,
+      "step": 14265
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 9.274580775157294e-05,
+      "loss": 0.289,
+      "step": 14270
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 9.267618028723686e-05,
+      "loss": 0.2893,
+      "step": 14275
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 9.26065563924415e-05,
+      "loss": 0.2873,
+      "step": 14280
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 9.253693610112078e-05,
+      "loss": 0.2879,
+      "step": 14285
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 9.246731944720675e-05,
+      "loss": 0.2867,
+      "step": 14290
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 9.239770646462968e-05,
+      "loss": 0.2869,
+      "step": 14295
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 9.232809718731814e-05,
+      "loss": 0.286,
+      "step": 14300
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 9.225849164919885e-05,
+      "loss": 0.2867,
+      "step": 14305
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 9.218888988419668e-05,
+      "loss": 0.2868,
+      "step": 14310
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 9.211929192623467e-05,
+      "loss": 0.2864,
+      "step": 14315
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 9.204969780923403e-05,
+      "loss": 0.2855,
+      "step": 14320
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 9.198010756711412e-05,
+      "loss": 0.2875,
+      "step": 14325
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 9.191052123379234e-05,
+      "loss": 0.2868,
+      "step": 14330
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 9.184093884318425e-05,
+      "loss": 0.2846,
+      "step": 14335
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0546875,
+      "learning_rate": 9.177136042920344e-05,
+      "loss": 0.286,
+      "step": 14340
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 9.17017860257616e-05,
+      "loss": 0.2868,
+      "step": 14345
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 9.163221566676847e-05,
+      "loss": 0.2866,
+      "step": 14350
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 9.156264938613174e-05,
+      "loss": 0.2876,
+      "step": 14355
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 9.14930872177572e-05,
+      "loss": 0.284,
+      "step": 14360
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 9.142352919554862e-05,
+      "loss": 0.2856,
+      "step": 14365
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 9.135397535340773e-05,
+      "loss": 0.2854,
+      "step": 14370
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 9.128442572523417e-05,
+      "loss": 0.2868,
+      "step": 14375
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 9.121488034492569e-05,
+      "loss": 0.2854,
+      "step": 14380
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 9.114533924637778e-05,
+      "loss": 0.2861,
+      "step": 14385
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 9.107580246348395e-05,
+      "loss": 0.2858,
+      "step": 14390
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 9.100627003013562e-05,
+      "loss": 0.2865,
+      "step": 14395
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 9.093674198022201e-05,
+      "loss": 0.2861,
+      "step": 14400
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.05859375,
+      "learning_rate": 9.086721834763024e-05,
+      "loss": 0.2883,
+      "step": 14405
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 9.07976991662453e-05,
+      "loss": 0.2856,
+      "step": 14410
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 9.072818446994999e-05,
+      "loss": 0.2873,
+      "step": 14415
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 9.065867429262496e-05,
+      "loss": 0.2878,
+      "step": 14420
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 9.058916866814858e-05,
+      "loss": 0.2852,
+      "step": 14425
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 9.051966763039707e-05,
+      "loss": 0.2864,
+      "step": 14430
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 9.045017121324438e-05,
+      "loss": 0.2849,
+      "step": 14435
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 9.038067945056227e-05,
+      "loss": 0.2859,
+      "step": 14440
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 9.03111923762201e-05,
+      "loss": 0.2861,
+      "step": 14445
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 9.024171002408506e-05,
+      "loss": 0.2864,
+      "step": 14450
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.035400390625,
+      "learning_rate": 9.017223242802204e-05,
+      "loss": 0.2883,
+      "step": 14455
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 9.01027596218935e-05,
+      "loss": 0.2873,
+      "step": 14460
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 9.003329163955972e-05,
+      "loss": 0.2859,
+      "step": 14465
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.056640625,
+      "learning_rate": 8.99638285148785e-05,
+      "loss": 0.2888,
+      "step": 14470
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.044921875,
+      "learning_rate": 8.989437028170537e-05,
+      "loss": 0.2862,
+      "step": 14475
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.048828125,
+      "learning_rate": 8.982491697389338e-05,
+      "loss": 0.2879,
+      "step": 14480
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 8.975546862529328e-05,
+      "loss": 0.2862,
+      "step": 14485
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 8.96860252697533e-05,
+      "loss": 0.2862,
+      "step": 14490
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 8.961658694111929e-05,
+      "loss": 0.2863,
+      "step": 14495
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.048828125,
+      "learning_rate": 8.954715367323468e-05,
+      "loss": 0.2874,
+      "step": 14500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.052734375,
+      "learning_rate": 8.947772549994035e-05,
+      "loss": 0.2865,
+      "step": 14505
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 8.940830245507483e-05,
+      "loss": 0.2867,
+      "step": 14510
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 8.933888457247402e-05,
+      "loss": 0.2867,
+      "step": 14515
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 8.926947188597134e-05,
+      "loss": 0.2877,
+      "step": 14520
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 8.920006442939772e-05,
+      "loss": 0.287,
+      "step": 14525
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.044921875,
+      "learning_rate": 8.913066223658151e-05,
+      "loss": 0.2863,
+      "step": 14530
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 8.906126534134848e-05,
+      "loss": 0.2863,
+      "step": 14535
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 8.89918737775218e-05,
+      "loss": 0.2869,
+      "step": 14540
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.056640625,
+      "learning_rate": 8.892248757892214e-05,
+      "loss": 0.287,
+      "step": 14545
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 8.885310677936746e-05,
+      "loss": 0.2873,
+      "step": 14550
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 8.878373141267311e-05,
+      "loss": 0.2862,
+      "step": 14555
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 8.871436151265184e-05,
+      "loss": 0.2859,
+      "step": 14560
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 8.864499711311362e-05,
+      "loss": 0.2866,
+      "step": 14565
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 8.857563824786596e-05,
+      "loss": 0.2871,
+      "step": 14570
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 8.850628495071336e-05,
+      "loss": 0.2871,
+      "step": 14575
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 8.843693725545786e-05,
+      "loss": 0.2858,
+      "step": 14580
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 8.836759519589867e-05,
+      "loss": 0.2876,
+      "step": 14585
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.056640625,
+      "learning_rate": 8.829825880583226e-05,
+      "loss": 0.2858,
+      "step": 14590
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07421875,
+      "learning_rate": 8.822892811905237e-05,
+      "loss": 0.2874,
+      "step": 14595
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 8.81596031693499e-05,
+      "loss": 0.2877,
+      "step": 14600
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 8.809028399051302e-05,
+      "loss": 0.2876,
+      "step": 14605
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.05078125,
+      "learning_rate": 8.802097061632705e-05,
+      "loss": 0.2864,
+      "step": 14610
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.060546875,
+      "learning_rate": 8.79516630805745e-05,
+      "loss": 0.2877,
+      "step": 14615
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.046875,
+      "learning_rate": 8.788236141703498e-05,
+      "loss": 0.2853,
+      "step": 14620
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 8.781306565948528e-05,
+      "loss": 0.2859,
+      "step": 14625
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 8.774377584169933e-05,
+      "loss": 0.2855,
+      "step": 14630
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 8.767449199744814e-05,
+      "loss": 0.2863,
+      "step": 14635
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 8.760521416049983e-05,
+      "loss": 0.2869,
+      "step": 14640
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 8.753594236461957e-05,
+      "loss": 0.2878,
+      "step": 14645
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 8.746667664356956e-05,
+      "loss": 0.2875,
+      "step": 14650
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 8.739741703110913e-05,
+      "loss": 0.2844,
+      "step": 14655
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.060546875,
+      "learning_rate": 8.732816356099456e-05,
+      "loss": 0.2867,
+      "step": 14660
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 8.72589162669791e-05,
+      "loss": 0.2862,
+      "step": 14665
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 8.718967518281307e-05,
+      "loss": 0.2854,
+      "step": 14670
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 8.712044034224374e-05,
+      "loss": 0.2854,
+      "step": 14675
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 8.705121177901532e-05,
+      "loss": 0.285,
+      "step": 14680
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 8.698198952686896e-05,
+      "loss": 0.2873,
+      "step": 14685
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 8.691277361954279e-05,
+      "loss": 0.2848,
+      "step": 14690
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 8.684356409077176e-05,
+      "loss": 0.2868,
+      "step": 14695
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 8.677436097428775e-05,
+      "loss": 0.2858,
+      "step": 14700
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 8.670516430381958e-05,
+      "loss": 0.2849,
+      "step": 14705
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0546875,
+      "learning_rate": 8.663597411309279e-05,
+      "loss": 0.2875,
+      "step": 14710
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 8.656679043582986e-05,
+      "loss": 0.2872,
+      "step": 14715
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 8.64976133057501e-05,
+      "loss": 0.2892,
+      "step": 14720
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 8.642844275656957e-05,
+      "loss": 0.2872,
+      "step": 14725
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 8.635927882200116e-05,
+      "loss": 0.286,
+      "step": 14730
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 8.629012153575458e-05,
+      "loss": 0.2839,
+      "step": 14735
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.05078125,
+      "learning_rate": 8.62209709315362e-05,
+      "loss": 0.2862,
+      "step": 14740
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 8.615182704304918e-05,
+      "loss": 0.2863,
+      "step": 14745
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.037353515625,
+      "learning_rate": 8.608268990399349e-05,
+      "loss": 0.2857,
+      "step": 14750
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 8.601355954806561e-05,
+      "loss": 0.2855,
+      "step": 14755
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 8.594443600895892e-05,
+      "loss": 0.2843,
+      "step": 14760
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 8.587531932036335e-05,
+      "loss": 0.288,
+      "step": 14765
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 8.580620951596557e-05,
+      "loss": 0.2894,
+      "step": 14770
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 8.573710662944885e-05,
+      "loss": 0.2844,
+      "step": 14775
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 8.566801069449305e-05,
+      "loss": 0.2874,
+      "step": 14780
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 8.559892174477479e-05,
+      "loss": 0.2862,
+      "step": 14785
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 8.55298398139671e-05,
+      "loss": 0.2858,
+      "step": 14790
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 8.546076493573972e-05,
+      "loss": 0.287,
+      "step": 14795
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 8.539169714375885e-05,
+      "loss": 0.2882,
+      "step": 14800
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.048828125,
+      "learning_rate": 8.532263647168735e-05,
+      "loss": 0.2872,
+      "step": 14805
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.05078125,
+      "learning_rate": 8.525358295318454e-05,
+      "loss": 0.2857,
+      "step": 14810
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.041015625,
+      "learning_rate": 8.51845366219062e-05,
+      "loss": 0.2871,
+      "step": 14815
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 8.511549751150479e-05,
+      "loss": 0.2859,
+      "step": 14820
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 8.504646565562906e-05,
+      "loss": 0.2868,
+      "step": 14825
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 8.497744108792429e-05,
+      "loss": 0.2874,
+      "step": 14830
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 8.490842384203225e-05,
+      "loss": 0.2861,
+      "step": 14835
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 8.483941395159114e-05,
+      "loss": 0.2863,
+      "step": 14840
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 8.477041145023546e-05,
+      "loss": 0.2857,
+      "step": 14845
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 8.47014163715962e-05,
+      "loss": 0.2883,
+      "step": 14850
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 8.463242874930079e-05,
+      "loss": 0.2869,
+      "step": 14855
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 8.456344861697289e-05,
+      "loss": 0.2877,
+      "step": 14860
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 8.449447600823262e-05,
+      "loss": 0.2865,
+      "step": 14865
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 8.442551095669639e-05,
+      "loss": 0.2855,
+      "step": 14870
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 8.435655349597689e-05,
+      "loss": 0.2875,
+      "step": 14875
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 8.428760365968326e-05,
+      "loss": 0.2845,
+      "step": 14880
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 8.421866148142066e-05,
+      "loss": 0.2885,
+      "step": 14885
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 8.414972699479075e-05,
+      "loss": 0.2867,
+      "step": 14890
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 8.408080023339133e-05,
+      "loss": 0.2879,
+      "step": 14895
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 8.401188123081653e-05,
+      "loss": 0.2856,
+      "step": 14900
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.05078125,
+      "learning_rate": 8.394297002065658e-05,
+      "loss": 0.2877,
+      "step": 14905
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 8.387406663649795e-05,
+      "loss": 0.2875,
+      "step": 14910
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 8.380517111192337e-05,
+      "loss": 0.288,
+      "step": 14915
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 8.373628348051165e-05,
+      "loss": 0.2861,
+      "step": 14920
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 8.366740377583781e-05,
+      "loss": 0.2864,
+      "step": 14925
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 8.359853203147291e-05,
+      "loss": 0.2867,
+      "step": 14930
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 8.352966828098428e-05,
+      "loss": 0.2863,
+      "step": 14935
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 8.346081255793525e-05,
+      "loss": 0.2858,
+      "step": 14940
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 8.339196489588523e-05,
+      "loss": 0.2853,
+      "step": 14945
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 8.332312532838978e-05,
+      "loss": 0.2861,
+      "step": 14950
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0390625,
+      "learning_rate": 8.325429388900046e-05,
+      "loss": 0.287,
+      "step": 14955
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 8.318547061126485e-05,
+      "loss": 0.2883,
+      "step": 14960
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0546875,
+      "learning_rate": 8.311665552872662e-05,
+      "loss": 0.2872,
+      "step": 14965
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 8.30478486749254e-05,
+      "loss": 0.2851,
+      "step": 14970
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 8.297905008339677e-05,
+      "loss": 0.2866,
+      "step": 14975
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 8.291025978767235e-05,
+      "loss": 0.2866,
+      "step": 14980
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.041015625,
+      "learning_rate": 8.284147782127971e-05,
+      "loss": 0.286,
+      "step": 14985
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 8.277270421774235e-05,
+      "loss": 0.2866,
+      "step": 14990
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 8.270393901057964e-05,
+      "loss": 0.2863,
+      "step": 14995
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 8.263518223330697e-05,
+      "loss": 0.2883,
+      "step": 15000
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 8.256643391943551e-05,
+      "loss": 0.2866,
+      "step": 15005
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 8.249769410247239e-05,
+      "loss": 0.2866,
+      "step": 15010
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 8.242896281592057e-05,
+      "loss": 0.2863,
+      "step": 15015
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 8.236024009327879e-05,
+      "loss": 0.2858,
+      "step": 15020
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 8.229152596804168e-05,
+      "loss": 0.2868,
+      "step": 15025
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.046875,
+      "learning_rate": 8.222282047369971e-05,
+      "loss": 0.287,
+      "step": 15030
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 8.215412364373907e-05,
+      "loss": 0.2869,
+      "step": 15035
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 8.208543551164178e-05,
+      "loss": 0.2856,
+      "step": 15040
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 8.201675611088558e-05,
+      "loss": 0.2862,
+      "step": 15045
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 8.194808547494401e-05,
+      "loss": 0.2872,
+      "step": 15050
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 8.187942363728625e-05,
+      "loss": 0.2833,
+      "step": 15055
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 8.181077063137733e-05,
+      "loss": 0.2867,
+      "step": 15060
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 8.174212649067781e-05,
+      "loss": 0.2863,
+      "step": 15065
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 8.167349124864405e-05,
+      "loss": 0.287,
+      "step": 15070
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.041015625,
+      "learning_rate": 8.160486493872798e-05,
+      "loss": 0.287,
+      "step": 15075
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.04296875,
+      "learning_rate": 8.153624759437732e-05,
+      "loss": 0.2892,
+      "step": 15080
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.044921875,
+      "learning_rate": 8.146763924903527e-05,
+      "loss": 0.288,
+      "step": 15085
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 8.139903993614068e-05,
+      "loss": 0.2858,
+      "step": 15090
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 8.133044968912811e-05,
+      "loss": 0.2855,
+      "step": 15095
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 8.126186854142752e-05,
+      "loss": 0.2881,
+      "step": 15100
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 8.119329652646463e-05,
+      "loss": 0.2849,
+      "step": 15105
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 8.11247336776605e-05,
+      "loss": 0.2867,
+      "step": 15110
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 8.105618002843189e-05,
+      "loss": 0.2869,
+      "step": 15115
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 8.0987635612191e-05,
+      "loss": 0.2862,
+      "step": 15120
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 8.091910046234552e-05,
+      "loss": 0.2878,
+      "step": 15125
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 8.085057461229872e-05,
+      "loss": 0.2869,
+      "step": 15130
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.048828125,
+      "learning_rate": 8.078205809544917e-05,
+      "loss": 0.2862,
+      "step": 15135
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 8.071355094519109e-05,
+      "loss": 0.2859,
+      "step": 15140
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 8.064505319491398e-05,
+      "loss": 0.2849,
+      "step": 15145
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 8.057656487800282e-05,
+      "loss": 0.2868,
+      "step": 15150
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 8.050808602783795e-05,
+      "loss": 0.2856,
+      "step": 15155
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 8.04396166777952e-05,
+      "loss": 0.2855,
+      "step": 15160
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 8.037115686124564e-05,
+      "loss": 0.2865,
+      "step": 15165
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 8.030270661155574e-05,
+      "loss": 0.2865,
+      "step": 15170
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 8.023426596208739e-05,
+      "loss": 0.2865,
+      "step": 15175
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.05859375,
+      "learning_rate": 8.016583494619769e-05,
+      "loss": 0.2863,
+      "step": 15180
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 8.009741359723906e-05,
+      "loss": 0.2873,
+      "step": 15185
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 8.002900194855932e-05,
+      "loss": 0.2858,
+      "step": 15190
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 7.996060003350139e-05,
+      "loss": 0.2844,
+      "step": 15195
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 7.989220788540355e-05,
+      "loss": 0.2876,
+      "step": 15200
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 7.982382553759931e-05,
+      "loss": 0.2874,
+      "step": 15205
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 7.975545302341743e-05,
+      "loss": 0.287,
+      "step": 15210
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 7.96870903761818e-05,
+      "loss": 0.2853,
+      "step": 15215
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 7.961873762921153e-05,
+      "loss": 0.2858,
+      "step": 15220
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 7.955039481582097e-05,
+      "loss": 0.2887,
+      "step": 15225
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 7.948206196931954e-05,
+      "loss": 0.2876,
+      "step": 15230
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 7.941373912301189e-05,
+      "loss": 0.2863,
+      "step": 15235
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 7.934542631019768e-05,
+      "loss": 0.2847,
+      "step": 15240
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 7.927712356417176e-05,
+      "loss": 0.2865,
+      "step": 15245
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 7.920883091822408e-05,
+      "loss": 0.2854,
+      "step": 15250
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.041015625,
+      "learning_rate": 7.914054840563963e-05,
+      "loss": 0.2845,
+      "step": 15255
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 7.907227605969849e-05,
+      "loss": 0.2875,
+      "step": 15260
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 7.900401391367576e-05,
+      "loss": 0.2851,
+      "step": 15265
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.052734375,
+      "learning_rate": 7.893576200084159e-05,
+      "loss": 0.2871,
+      "step": 15270
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 7.886752035446114e-05,
+      "loss": 0.2866,
+      "step": 15275
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.044921875,
+      "learning_rate": 7.879928900779456e-05,
+      "loss": 0.2868,
+      "step": 15280
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0390625,
+      "learning_rate": 7.873106799409695e-05,
+      "loss": 0.2868,
+      "step": 15285
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 7.866285734661841e-05,
+      "loss": 0.2871,
+      "step": 15290
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 7.8594657098604e-05,
+      "loss": 0.2866,
+      "step": 15295
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 7.852646728329368e-05,
+      "loss": 0.2852,
+      "step": 15300
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 7.845828793392236e-05,
+      "loss": 0.2883,
+      "step": 15305
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 7.83901190837198e-05,
+      "loss": 0.2873,
+      "step": 15310
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 7.832196076591067e-05,
+      "loss": 0.2864,
+      "step": 15315
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 7.825381301371452e-05,
+      "loss": 0.2863,
+      "step": 15320
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 7.818567586034577e-05,
+      "loss": 0.2864,
+      "step": 15325
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 7.811754933901358e-05,
+      "loss": 0.2858,
+      "step": 15330
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 7.804943348292197e-05,
+      "loss": 0.2851,
+      "step": 15335
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 7.798132832526986e-05,
+      "loss": 0.2867,
+      "step": 15340
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 7.791323389925084e-05,
+      "loss": 0.2894,
+      "step": 15345
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.048828125,
+      "learning_rate": 7.784515023805328e-05,
+      "loss": 0.2875,
+      "step": 15350
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 7.777707737486037e-05,
+      "loss": 0.2891,
+      "step": 15355
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 7.770901534284995e-05,
+      "loss": 0.2889,
+      "step": 15360
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 7.76409641751947e-05,
+      "loss": 0.2856,
+      "step": 15365
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 7.75729239050619e-05,
+      "loss": 0.2872,
+      "step": 15370
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 7.750489456561352e-05,
+      "loss": 0.2851,
+      "step": 15375
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 7.743687619000626e-05,
+      "loss": 0.288,
+      "step": 15380
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 7.736886881139142e-05,
+      "loss": 0.2865,
+      "step": 15385
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 7.730087246291502e-05,
+      "loss": 0.2871,
+      "step": 15390
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 7.723288717771761e-05,
+      "loss": 0.2865,
+      "step": 15395
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 7.716491298893442e-05,
+      "loss": 0.2867,
+      "step": 15400
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 7.709694992969526e-05,
+      "loss": 0.2867,
+      "step": 15405
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.041015625,
+      "learning_rate": 7.702899803312443e-05,
+      "loss": 0.2852,
+      "step": 15410
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 7.696105733234098e-05,
+      "loss": 0.2869,
+      "step": 15415
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 7.689312786045823e-05,
+      "loss": 0.2864,
+      "step": 15420
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.038330078125,
+      "learning_rate": 7.682520965058428e-05,
+      "loss": 0.2851,
+      "step": 15425
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 7.67573027358216e-05,
+      "loss": 0.2849,
+      "step": 15430
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 7.668940714926725e-05,
+      "loss": 0.2872,
+      "step": 15435
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 7.662152292401264e-05,
+      "loss": 0.2881,
+      "step": 15440
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 7.655365009314374e-05,
+      "loss": 0.2857,
+      "step": 15445
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 7.6485788689741e-05,
+      "loss": 0.2872,
+      "step": 15450
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 7.641793874687918e-05,
+      "loss": 0.2866,
+      "step": 15455
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 7.635010029762756e-05,
+      "loss": 0.2867,
+      "step": 15460
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 7.628227337504972e-05,
+      "loss": 0.2864,
+      "step": 15465
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 7.621445801220371e-05,
+      "loss": 0.2881,
+      "step": 15470
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 7.614665424214193e-05,
+      "loss": 0.2869,
+      "step": 15475
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 7.607886209791107e-05,
+      "loss": 0.2858,
+      "step": 15480
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 7.601108161255226e-05,
+      "loss": 0.2865,
+      "step": 15485
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 7.594331281910082e-05,
+      "loss": 0.2854,
+      "step": 15490
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.041015625,
+      "learning_rate": 7.587555575058649e-05,
+      "loss": 0.2879,
+      "step": 15495
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 7.580781044003324e-05,
+      "loss": 0.287,
+      "step": 15500
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 7.574007692045928e-05,
+      "loss": 0.2855,
+      "step": 15505
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 7.567235522487712e-05,
+      "loss": 0.2871,
+      "step": 15510
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 7.560464538629344e-05,
+      "loss": 0.2861,
+      "step": 15515
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0546875,
+      "learning_rate": 7.553694743770928e-05,
+      "loss": 0.2869,
+      "step": 15520
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 7.546926141211974e-05,
+      "loss": 0.2859,
+      "step": 15525
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 7.54015873425142e-05,
+      "loss": 0.285,
+      "step": 15530
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.048828125,
+      "learning_rate": 7.533392526187617e-05,
+      "loss": 0.2852,
+      "step": 15535
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 7.526627520318329e-05,
+      "loss": 0.2849,
+      "step": 15540
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 7.519863719940748e-05,
+      "loss": 0.2852,
+      "step": 15545
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 7.513101128351454e-05,
+      "loss": 0.2853,
+      "step": 15550
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.03759765625,
+      "learning_rate": 7.506339748846461e-05,
+      "loss": 0.2858,
+      "step": 15555
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 7.499579584721179e-05,
+      "loss": 0.2839,
+      "step": 15560
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 7.492820639270434e-05,
+      "loss": 0.2866,
+      "step": 15565
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 7.486062915788452e-05,
+      "loss": 0.2876,
+      "step": 15570
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 7.479306417568864e-05,
+      "loss": 0.286,
+      "step": 15575
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 7.472551147904708e-05,
+      "loss": 0.2879,
+      "step": 15580
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 7.465797110088417e-05,
+      "loss": 0.2861,
+      "step": 15585
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 7.459044307411832e-05,
+      "loss": 0.2857,
+      "step": 15590
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 7.45229274316618e-05,
+      "loss": 0.2879,
+      "step": 15595
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 7.445542420642097e-05,
+      "loss": 0.2874,
+      "step": 15600
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 7.438793343129605e-05,
+      "loss": 0.2868,
+      "step": 15605
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 7.432045513918122e-05,
+      "loss": 0.2856,
+      "step": 15610
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 7.425298936296463e-05,
+      "loss": 0.2883,
+      "step": 15615
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 7.418553613552825e-05,
+      "loss": 0.2889,
+      "step": 15620
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 7.411809548974792e-05,
+      "loss": 0.2882,
+      "step": 15625
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.044921875,
+      "learning_rate": 7.405066745849346e-05,
+      "loss": 0.2864,
+      "step": 15630
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 7.398325207462846e-05,
+      "loss": 0.2862,
+      "step": 15635
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 7.391584937101033e-05,
+      "loss": 0.2856,
+      "step": 15640
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 7.384845938049031e-05,
+      "loss": 0.2857,
+      "step": 15645
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 7.378108213591355e-05,
+      "loss": 0.2861,
+      "step": 15650
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 7.37137176701188e-05,
+      "loss": 0.286,
+      "step": 15655
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 7.364636601593875e-05,
+      "loss": 0.289,
+      "step": 15660
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 7.357902720619976e-05,
+      "loss": 0.2868,
+      "step": 15665
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 7.351170127372191e-05,
+      "loss": 0.286,
+      "step": 15670
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.052734375,
+      "learning_rate": 7.344438825131911e-05,
+      "loss": 0.2858,
+      "step": 15675
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 7.33770881717989e-05,
+      "loss": 0.2863,
+      "step": 15680
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 7.330980106796246e-05,
+      "loss": 0.2874,
+      "step": 15685
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 7.324252697260474e-05,
+      "loss": 0.2865,
+      "step": 15690
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 7.317526591851433e-05,
+      "loss": 0.2879,
+      "step": 15695
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.048828125,
+      "learning_rate": 7.310801793847344e-05,
+      "loss": 0.2879,
+      "step": 15700
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.056640625,
+      "learning_rate": 7.30407830652579e-05,
+      "loss": 0.2884,
+      "step": 15705
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 7.297356133163721e-05,
+      "loss": 0.2874,
+      "step": 15710
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.068359375,
+      "learning_rate": 7.290635277037442e-05,
+      "loss": 0.287,
+      "step": 15715
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 7.283915741422612e-05,
+      "loss": 0.2864,
+      "step": 15720
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 7.277197529594257e-05,
+      "loss": 0.2869,
+      "step": 15725
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 7.270480644826749e-05,
+      "loss": 0.2873,
+      "step": 15730
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 7.263765090393817e-05,
+      "loss": 0.2857,
+      "step": 15735
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 7.257050869568535e-05,
+      "loss": 0.2863,
+      "step": 15740
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 7.250337985623342e-05,
+      "loss": 0.2852,
+      "step": 15745
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 7.243626441830009e-05,
+      "loss": 0.2875,
+      "step": 15750
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 7.236916241459663e-05,
+      "loss": 0.2857,
+      "step": 15755
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.04296875,
+      "learning_rate": 7.230207387782776e-05,
+      "loss": 0.2866,
+      "step": 15760
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 7.22349988406916e-05,
+      "loss": 0.2872,
+      "step": 15765
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 7.216793733587976e-05,
+      "loss": 0.2865,
+      "step": 15770
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.052734375,
+      "learning_rate": 7.210088939607708e-05,
+      "loss": 0.2846,
+      "step": 15775
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 7.203385505396203e-05,
+      "loss": 0.286,
+      "step": 15780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 7.196683434220625e-05,
+      "loss": 0.287,
+      "step": 15785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 7.18998272934749e-05,
+      "loss": 0.2857,
+      "step": 15790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 7.183283394042634e-05,
+      "loss": 0.2867,
+      "step": 15795
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 7.176585431571235e-05,
+      "loss": 0.286,
+      "step": 15800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 7.169888845197798e-05,
+      "loss": 0.2857,
+      "step": 15805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 7.163193638186158e-05,
+      "loss": 0.2855,
+      "step": 15810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 7.156499813799476e-05,
+      "loss": 0.2859,
+      "step": 15815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 7.149807375300239e-05,
+      "loss": 0.2848,
+      "step": 15820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 7.143116325950265e-05,
+      "loss": 0.2871,
+      "step": 15825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 7.136426669010689e-05,
+      "loss": 0.2866,
+      "step": 15830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 7.129738407741964e-05,
+      "loss": 0.2864,
+      "step": 15835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 7.123051545403874e-05,
+      "loss": 0.2849,
+      "step": 15840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.03515625,
+      "learning_rate": 7.11636608525551e-05,
+      "loss": 0.2882,
+      "step": 15845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 7.109682030555283e-05,
+      "loss": 0.2884,
+      "step": 15850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 7.102999384560927e-05,
+      "loss": 0.2874,
+      "step": 15855
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 7.096318150529477e-05,
+      "loss": 0.2862,
+      "step": 15860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 7.089638331717284e-05,
+      "loss": 0.2848,
+      "step": 15865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.052734375,
+      "learning_rate": 7.082959931380011e-05,
+      "loss": 0.2875,
+      "step": 15870
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 7.076282952772633e-05,
+      "loss": 0.2865,
+      "step": 15875
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 7.069607399149428e-05,
+      "loss": 0.2861,
+      "step": 15880
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 7.062933273763975e-05,
+      "loss": 0.2882,
+      "step": 15885
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 7.056260579869165e-05,
+      "loss": 0.2877,
+      "step": 15890
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 7.049589320717186e-05,
+      "loss": 0.2857,
+      "step": 15895
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 7.042919499559537e-05,
+      "loss": 0.2883,
+      "step": 15900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 7.036251119646992e-05,
+      "loss": 0.2867,
+      "step": 15905
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 7.029584184229653e-05,
+      "loss": 0.2879,
+      "step": 15910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.048828125,
+      "learning_rate": 7.022918696556896e-05,
+      "loss": 0.2856,
+      "step": 15915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 7.016254659877398e-05,
+      "loss": 0.2851,
+      "step": 15920
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 7.009592077439134e-05,
+      "loss": 0.2875,
+      "step": 15925
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 7.002930952489362e-05,
+      "loss": 0.2845,
+      "step": 15930
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0546875,
+      "learning_rate": 6.996271288274636e-05,
+      "loss": 0.2878,
+      "step": 15935
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 6.989613088040796e-05,
+      "loss": 0.2884,
+      "step": 15940
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.046875,
+      "learning_rate": 6.982956355032968e-05,
+      "loss": 0.2859,
+      "step": 15945
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 6.976301092495556e-05,
+      "loss": 0.2875,
+      "step": 15950
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 6.969647303672262e-05,
+      "loss": 0.2859,
+      "step": 15955
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 6.962994991806059e-05,
+      "loss": 0.2865,
+      "step": 15960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 6.956344160139201e-05,
+      "loss": 0.2875,
+      "step": 15965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 6.949694811913225e-05,
+      "loss": 0.286,
+      "step": 15970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 6.943046950368944e-05,
+      "loss": 0.2873,
+      "step": 15975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 6.93640057874644e-05,
+      "loss": 0.2869,
+      "step": 15980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 6.929755700285081e-05,
+      "loss": 0.2883,
+      "step": 15985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 6.923112318223496e-05,
+      "loss": 0.286,
+      "step": 15990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 6.916470435799587e-05,
+      "loss": 0.2852,
+      "step": 15995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 6.909830056250527e-05,
+      "loss": 0.2848,
+      "step": 16000
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 6.90319118281276e-05,
+      "loss": 0.2867,
+      "step": 16005
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 6.896553818721989e-05,
+      "loss": 0.2868,
+      "step": 16010
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 6.889917967213185e-05,
+      "loss": 0.2875,
+      "step": 16015
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 6.883283631520582e-05,
+      "loss": 0.2856,
+      "step": 16020
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 6.876650814877674e-05,
+      "loss": 0.286,
+      "step": 16025
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 6.870019520517217e-05,
+      "loss": 0.2867,
+      "step": 16030
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.046875,
+      "learning_rate": 6.863389751671225e-05,
+      "loss": 0.287,
+      "step": 16035
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 6.856761511570963e-05,
+      "loss": 0.2859,
+      "step": 16040
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 6.850134803446954e-05,
+      "loss": 0.2871,
+      "step": 16045
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 6.843509630528977e-05,
+      "loss": 0.2873,
+      "step": 16050
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 6.836885996046061e-05,
+      "loss": 0.2863,
+      "step": 16055
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 6.830263903226483e-05,
+      "loss": 0.2861,
+      "step": 16060
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 6.823643355297773e-05,
+      "loss": 0.2855,
+      "step": 16065
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 6.817024355486706e-05,
+      "loss": 0.2853,
+      "step": 16070
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 6.8104069070193e-05,
+      "loss": 0.2848,
+      "step": 16075
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 6.803791013120822e-05,
+      "loss": 0.2857,
+      "step": 16080
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 6.797176677015775e-05,
+      "loss": 0.287,
+      "step": 16085
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 6.790563901927907e-05,
+      "loss": 0.287,
+      "step": 16090
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 6.783952691080203e-05,
+      "loss": 0.2855,
+      "step": 16095
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 6.77734304769489e-05,
+      "loss": 0.2876,
+      "step": 16100
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 6.770734974993426e-05,
+      "loss": 0.2871,
+      "step": 16105
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 6.764128476196505e-05,
+      "loss": 0.2859,
+      "step": 16110
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 6.757523554524056e-05,
+      "loss": 0.2874,
+      "step": 16115
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 6.750920213195238e-05,
+      "loss": 0.2859,
+      "step": 16120
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 6.744318455428436e-05,
+      "loss": 0.2866,
+      "step": 16125
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 6.737718284441267e-05,
+      "loss": 0.2859,
+      "step": 16130
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 6.731119703450577e-05,
+      "loss": 0.2867,
+      "step": 16135
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 6.724522715672432e-05,
+      "loss": 0.2868,
+      "step": 16140
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 6.717927324322124e-05,
+      "loss": 0.2852,
+      "step": 16145
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 6.711333532614168e-05,
+      "loss": 0.2867,
+      "step": 16150
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 6.704741343762295e-05,
+      "loss": 0.2859,
+      "step": 16155
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 6.698150760979463e-05,
+      "loss": 0.2858,
+      "step": 16160
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 6.69156178747784e-05,
+      "loss": 0.2858,
+      "step": 16165
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 6.684974426468808e-05,
+      "loss": 0.2878,
+      "step": 16170
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 6.67838868116297e-05,
+      "loss": 0.2865,
+      "step": 16175
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 6.671804554770135e-05,
+      "loss": 0.2858,
+      "step": 16180
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 6.66522205049933e-05,
+      "loss": 0.2856,
+      "step": 16185
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 6.658641171558785e-05,
+      "loss": 0.2867,
+      "step": 16190
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 6.652061921155943e-05,
+      "loss": 0.2878,
+      "step": 16195
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 6.64548430249745e-05,
+      "loss": 0.2866,
+      "step": 16200
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 6.638908318789156e-05,
+      "loss": 0.2879,
+      "step": 16205
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 6.63233397323612e-05,
+      "loss": 0.287,
+      "step": 16210
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 6.62576126904259e-05,
+      "loss": 0.286,
+      "step": 16215
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.03759765625,
+      "learning_rate": 6.61919020941203e-05,
+      "loss": 0.2855,
+      "step": 16220
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 6.612620797547087e-05,
+      "loss": 0.2869,
+      "step": 16225
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 6.606053036649619e-05,
+      "loss": 0.2866,
+      "step": 16230
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.041015625,
+      "learning_rate": 6.599486929920673e-05,
+      "loss": 0.2864,
+      "step": 16235
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 6.592922480560483e-05,
+      "loss": 0.2867,
+      "step": 16240
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.052734375,
+      "learning_rate": 6.58635969176849e-05,
+      "loss": 0.2863,
+      "step": 16245
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.044921875,
+      "learning_rate": 6.579798566743314e-05,
+      "loss": 0.286,
+      "step": 16250
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 6.573239108682768e-05,
+      "loss": 0.2887,
+      "step": 16255
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 6.566681320783849e-05,
+      "loss": 0.2863,
+      "step": 16260
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 6.560125206242746e-05,
+      "loss": 0.2859,
+      "step": 16265
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 6.55357076825483e-05,
+      "loss": 0.2861,
+      "step": 16270
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.05859375,
+      "learning_rate": 6.547018010014654e-05,
+      "loss": 0.2885,
+      "step": 16275
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 6.540466934715953e-05,
+      "loss": 0.2861,
+      "step": 16280
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 6.53391754555164e-05,
+      "loss": 0.2852,
+      "step": 16285
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 6.52736984571381e-05,
+      "loss": 0.2868,
+      "step": 16290
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 6.520823838393731e-05,
+      "loss": 0.2869,
+      "step": 16295
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 6.51427952678185e-05,
+      "loss": 0.286,
+      "step": 16300
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 6.507736914067781e-05,
+      "loss": 0.2858,
+      "step": 16305
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 6.501196003440314e-05,
+      "loss": 0.287,
+      "step": 16310
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 6.494656798087412e-05,
+      "loss": 0.2867,
+      "step": 16315
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 6.488119301196201e-05,
+      "loss": 0.2871,
+      "step": 16320
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 6.481583515952983e-05,
+      "loss": 0.2866,
+      "step": 16325
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 6.475049445543215e-05,
+      "loss": 0.2857,
+      "step": 16330
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 6.468517093151525e-05,
+      "loss": 0.2859,
+      "step": 16335
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0546875,
+      "learning_rate": 6.461986461961706e-05,
+      "loss": 0.2852,
+      "step": 16340
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.044921875,
+      "learning_rate": 6.455457555156705e-05,
+      "loss": 0.2858,
+      "step": 16345
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 6.448930375918631e-05,
+      "loss": 0.2872,
+      "step": 16350
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 6.442404927428751e-05,
+      "loss": 0.2867,
+      "step": 16355
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 6.435881212867493e-05,
+      "loss": 0.2865,
+      "step": 16360
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 6.429359235414436e-05,
+      "loss": 0.2862,
+      "step": 16365
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 6.422838998248307e-05,
+      "loss": 0.285,
+      "step": 16370
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 6.416320504546997e-05,
+      "loss": 0.2867,
+      "step": 16375
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.046875,
+      "learning_rate": 6.409803757487538e-05,
+      "loss": 0.2853,
+      "step": 16380
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 6.403288760246112e-05,
+      "loss": 0.2876,
+      "step": 16385
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 6.396775515998055e-05,
+      "loss": 0.2876,
+      "step": 16390
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.041015625,
+      "learning_rate": 6.390264027917836e-05,
+      "loss": 0.2867,
+      "step": 16395
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 6.383754299179079e-05,
+      "loss": 0.2869,
+      "step": 16400
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 6.377246332954544e-05,
+      "loss": 0.2871,
+      "step": 16405
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 6.370740132416138e-05,
+      "loss": 0.2883,
+      "step": 16410
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 6.364235700734903e-05,
+      "loss": 0.2849,
+      "step": 16415
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.048828125,
+      "learning_rate": 6.357733041081018e-05,
+      "loss": 0.2865,
+      "step": 16420
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 6.351232156623803e-05,
+      "loss": 0.2874,
+      "step": 16425
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 6.344733050531713e-05,
+      "loss": 0.2873,
+      "step": 16430
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.05859375,
+      "learning_rate": 6.338235725972325e-05,
+      "loss": 0.2885,
+      "step": 16435
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.052734375,
+      "learning_rate": 6.33174018611236e-05,
+      "loss": 0.2849,
+      "step": 16440
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 6.325246434117668e-05,
+      "loss": 0.2881,
+      "step": 16445
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 6.318754473153221e-05,
+      "loss": 0.2864,
+      "step": 16450
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.046875,
+      "learning_rate": 6.312264306383124e-05,
+      "loss": 0.2881,
+      "step": 16455
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 6.305775936970606e-05,
+      "loss": 0.2865,
+      "step": 16460
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 6.299289368078016e-05,
+      "loss": 0.2871,
+      "step": 16465
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 6.292804602866834e-05,
+      "loss": 0.2872,
+      "step": 16470
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 6.286321644497655e-05,
+      "loss": 0.2871,
+      "step": 16475
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 6.27984049613019e-05,
+      "loss": 0.2873,
+      "step": 16480
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 6.27336116092327e-05,
+      "loss": 0.2882,
+      "step": 16485
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 6.266883642034853e-05,
+      "loss": 0.2886,
+      "step": 16490
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 6.260407942621998e-05,
+      "loss": 0.2854,
+      "step": 16495
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 6.25393406584088e-05,
+      "loss": 0.2878,
+      "step": 16500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 6.247462014846792e-05,
+      "loss": 0.2863,
+      "step": 16505
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0625,
+      "learning_rate": 6.240991792794133e-05,
+      "loss": 0.2848,
+      "step": 16510
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.04296875,
+      "learning_rate": 6.234523402836407e-05,
+      "loss": 0.2852,
+      "step": 16515
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 6.228056848126236e-05,
+      "loss": 0.2857,
+      "step": 16520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 6.22159213181533e-05,
+      "loss": 0.287,
+      "step": 16525
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 6.215129257054522e-05,
+      "loss": 0.2876,
+      "step": 16530
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0390625,
+      "learning_rate": 6.20866822699373e-05,
+      "loss": 0.2866,
+      "step": 16535
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 6.20220904478199e-05,
+      "loss": 0.2861,
+      "step": 16540
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 6.195751713567427e-05,
+      "loss": 0.2868,
+      "step": 16545
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 6.18929623649726e-05,
+      "loss": 0.2861,
+      "step": 16550
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 6.182842616717817e-05,
+      "loss": 0.2873,
+      "step": 16555
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 6.176390857374507e-05,
+      "loss": 0.2864,
+      "step": 16560
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 6.169940961611854e-05,
+      "loss": 0.2856,
+      "step": 16565
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 6.163492932573438e-05,
+      "loss": 0.2885,
+      "step": 16570
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 6.157046773401964e-05,
+      "loss": 0.286,
+      "step": 16575
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.05078125,
+      "learning_rate": 6.150602487239206e-05,
+      "loss": 0.2866,
+      "step": 16580
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.052734375,
+      "learning_rate": 6.144160077226036e-05,
+      "loss": 0.2854,
+      "step": 16585
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 6.1377195465024e-05,
+      "loss": 0.2849,
+      "step": 16590
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 6.131280898207339e-05,
+      "loss": 0.2869,
+      "step": 16595
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 6.12484413547897e-05,
+      "loss": 0.2862,
+      "step": 16600
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 6.118409261454494e-05,
+      "loss": 0.286,
+      "step": 16605
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 6.111976279270192e-05,
+      "loss": 0.2872,
+      "step": 16610
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 6.105545192061416e-05,
+      "loss": 0.2879,
+      "step": 16615
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 6.099116002962604e-05,
+      "loss": 0.2877,
+      "step": 16620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 6.092688715107264e-05,
+      "loss": 0.2863,
+      "step": 16625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 6.086263331627976e-05,
+      "loss": 0.2862,
+      "step": 16630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.064453125,
+      "learning_rate": 6.0798398556563976e-05,
+      "loss": 0.2865,
+      "step": 16635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 6.073418290323251e-05,
+      "loss": 0.286,
+      "step": 16640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 6.066998638758326e-05,
+      "loss": 0.2852,
+      "step": 16645
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 6.0605809040904894e-05,
+      "loss": 0.2869,
+      "step": 16650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 6.054165089447663e-05,
+      "loss": 0.2871,
+      "step": 16655
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 6.047751197956838e-05,
+      "loss": 0.2865,
+      "step": 16660
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 6.0413392327440635e-05,
+      "loss": 0.2869,
+      "step": 16665
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 6.034929196934459e-05,
+      "loss": 0.2874,
+      "step": 16670
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.041015625,
+      "learning_rate": 6.0285210936521955e-05,
+      "loss": 0.2871,
+      "step": 16675
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 6.022114926020504e-05,
+      "loss": 0.2847,
+      "step": 16680
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 6.015710697161674e-05,
+      "loss": 0.2856,
+      "step": 16685
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.05078125,
+      "learning_rate": 6.009308410197047e-05,
+      "loss": 0.2865,
+      "step": 16690
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 6.002908068247024e-05,
+      "loss": 0.2847,
+      "step": 16695
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 5.9965096744310526e-05,
+      "loss": 0.2876,
+      "step": 16700
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 5.990113231867629e-05,
+      "loss": 0.2887,
+      "step": 16705
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 5.9837187436743016e-05,
+      "loss": 0.2866,
+      "step": 16710
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 5.977326212967671e-05,
+      "loss": 0.2852,
+      "step": 16715
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 5.9709356428633746e-05,
+      "loss": 0.2856,
+      "step": 16720
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 5.964547036476099e-05,
+      "loss": 0.286,
+      "step": 16725
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 5.9581603969195766e-05,
+      "loss": 0.2867,
+      "step": 16730
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 5.951775727306577e-05,
+      "loss": 0.288,
+      "step": 16735
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 5.94539303074891e-05,
+      "loss": 0.2859,
+      "step": 16740
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 5.9390123103574215e-05,
+      "loss": 0.2871,
+      "step": 16745
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.044921875,
+      "learning_rate": 5.9326335692419995e-05,
+      "loss": 0.2872,
+      "step": 16750
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 5.9262568105115654e-05,
+      "loss": 0.2861,
+      "step": 16755
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 5.9198820372740726e-05,
+      "loss": 0.2859,
+      "step": 16760
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.052734375,
+      "learning_rate": 5.913509252636511e-05,
+      "loss": 0.2848,
+      "step": 16765
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 5.907138459704895e-05,
+      "loss": 0.2856,
+      "step": 16770
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 5.900769661584272e-05,
+      "loss": 0.2867,
+      "step": 16775
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 5.8944028613787206e-05,
+      "loss": 0.2881,
+      "step": 16780
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.06640625,
+      "learning_rate": 5.88803806219134e-05,
+      "loss": 0.2873,
+      "step": 16785
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 5.881675267124254e-05,
+      "loss": 0.2876,
+      "step": 16790
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 5.8753144792786096e-05,
+      "loss": 0.2864,
+      "step": 16795
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 5.868955701754584e-05,
+      "loss": 0.2859,
+      "step": 16800
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 5.862598937651365e-05,
+      "loss": 0.2862,
+      "step": 16805
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 5.856244190067159e-05,
+      "loss": 0.2849,
+      "step": 16810
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 5.849891462099198e-05,
+      "loss": 0.287,
+      "step": 16815
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 5.8435407568437216e-05,
+      "loss": 0.286,
+      "step": 16820
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 5.83719207739599e-05,
+      "loss": 0.2869,
+      "step": 16825
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 5.8308454268502675e-05,
+      "loss": 0.2842,
+      "step": 16830
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 5.8245008082998364e-05,
+      "loss": 0.2869,
+      "step": 16835
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 5.818158224836987e-05,
+      "loss": 0.2862,
+      "step": 16840
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 5.8118176795530176e-05,
+      "loss": 0.2872,
+      "step": 16845
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 5.805479175538229e-05,
+      "loss": 0.2848,
+      "step": 16850
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.05859375,
+      "learning_rate": 5.799142715881938e-05,
+      "loss": 0.2873,
+      "step": 16855
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0625,
+      "learning_rate": 5.792808303672454e-05,
+      "loss": 0.2857,
+      "step": 16860
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 5.786475941997094e-05,
+      "loss": 0.2852,
+      "step": 16865
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 5.780145633942173e-05,
+      "loss": 0.2854,
+      "step": 16870
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 5.773817382593008e-05,
+      "loss": 0.2871,
+      "step": 16875
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 5.7674911910339094e-05,
+      "loss": 0.2869,
+      "step": 16880
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 5.7611670623481864e-05,
+      "loss": 0.2868,
+      "step": 16885
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.046875,
+      "learning_rate": 5.754844999618144e-05,
+      "loss": 0.2872,
+      "step": 16890
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 5.748525005925074e-05,
+      "loss": 0.2857,
+      "step": 16895
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.052734375,
+      "learning_rate": 5.7422070843492734e-05,
+      "loss": 0.2871,
+      "step": 16900
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 5.735891237970015e-05,
+      "loss": 0.2868,
+      "step": 16905
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 5.7295774698655655e-05,
+      "loss": 0.2855,
+      "step": 16910
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 5.723265783113181e-05,
+      "loss": 0.2865,
+      "step": 16915
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 5.716956180789098e-05,
+      "loss": 0.2867,
+      "step": 16920
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 5.710648665968543e-05,
+      "loss": 0.2872,
+      "step": 16925
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 5.704343241725719e-05,
+      "loss": 0.2865,
+      "step": 16930
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.056640625,
+      "learning_rate": 5.6980399111338156e-05,
+      "loss": 0.2866,
+      "step": 16935
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 5.691738677265e-05,
+      "loss": 0.2838,
+      "step": 16940
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 5.6854395431904094e-05,
+      "loss": 0.2857,
+      "step": 16945
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 5.679142511980175e-05,
+      "loss": 0.2855,
+      "step": 16950
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 5.6728475867033925e-05,
+      "loss": 0.287,
+      "step": 16955
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 5.666554770428129e-05,
+      "loss": 0.2847,
+      "step": 16960
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 5.6602640662214256e-05,
+      "loss": 0.286,
+      "step": 16965
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 5.653975477149298e-05,
+      "loss": 0.2858,
+      "step": 16970
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 5.647689006276726e-05,
+      "loss": 0.2864,
+      "step": 16975
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 5.641404656667661e-05,
+      "loss": 0.2864,
+      "step": 16980
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 5.635122431385016e-05,
+      "loss": 0.2876,
+      "step": 16985
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.046875,
+      "learning_rate": 5.6288423334906735e-05,
+      "loss": 0.2856,
+      "step": 16990
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 5.622564366045472e-05,
+      "loss": 0.2869,
+      "step": 16995
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 5.616288532109225e-05,
+      "loss": 0.2856,
+      "step": 17000
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 5.610014834740693e-05,
+      "loss": 0.2859,
+      "step": 17005
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 5.6037432769976064e-05,
+      "loss": 0.2851,
+      "step": 17010
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 5.5974738619366295e-05,
+      "loss": 0.2864,
+      "step": 17015
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 5.591206592613416e-05,
+      "loss": 0.2862,
+      "step": 17020
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 5.584941472082549e-05,
+      "loss": 0.2881,
+      "step": 17025
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 5.578678503397574e-05,
+      "loss": 0.2877,
+      "step": 17030
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 5.572417689610987e-05,
+      "loss": 0.2875,
+      "step": 17035
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 5.566159033774225e-05,
+      "loss": 0.2867,
+      "step": 17040
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 5.5599025389376935e-05,
+      "loss": 0.2862,
+      "step": 17045
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 5.553648208150728e-05,
+      "loss": 0.2882,
+      "step": 17050
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 5.547396044461608e-05,
+      "loss": 0.2872,
+      "step": 17055
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 5.541146050917561e-05,
+      "loss": 0.2849,
+      "step": 17060
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 5.5348982305647643e-05,
+      "loss": 0.2852,
+      "step": 17065
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 5.5286525864483285e-05,
+      "loss": 0.2864,
+      "step": 17070
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 5.522409121612304e-05,
+      "loss": 0.2864,
+      "step": 17075
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 5.5161678390996796e-05,
+      "loss": 0.2848,
+      "step": 17080
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 5.509928741952379e-05,
+      "loss": 0.2878,
+      "step": 17085
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 5.50369183321126e-05,
+      "loss": 0.2864,
+      "step": 17090
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 5.497457115916127e-05,
+      "loss": 0.2878,
+      "step": 17095
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 5.491224593105695e-05,
+      "loss": 0.2882,
+      "step": 17100
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 5.484994267817624e-05,
+      "loss": 0.2855,
+      "step": 17105
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 5.478766143088492e-05,
+      "loss": 0.2859,
+      "step": 17110
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 5.4725402219538236e-05,
+      "loss": 0.2847,
+      "step": 17115
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 5.466316507448048e-05,
+      "loss": 0.2861,
+      "step": 17120
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 5.4600950026045326e-05,
+      "loss": 0.2872,
+      "step": 17125
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 5.4538757104555615e-05,
+      "loss": 0.2871,
+      "step": 17130
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 5.447658634032338e-05,
+      "loss": 0.2871,
+      "step": 17135
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 5.441443776365003e-05,
+      "loss": 0.2861,
+      "step": 17140
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 5.435231140482587e-05,
+      "loss": 0.286,
+      "step": 17145
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 5.4290207294130615e-05,
+      "loss": 0.2855,
+      "step": 17150
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 5.4228125461833024e-05,
+      "loss": 0.2867,
+      "step": 17155
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 5.416606593819101e-05,
+      "loss": 0.2859,
+      "step": 17160
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 5.4104028753451696e-05,
+      "loss": 0.2873,
+      "step": 17165
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 5.404201393785122e-05,
+      "loss": 0.2844,
+      "step": 17170
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.046875,
+      "learning_rate": 5.398002152161484e-05,
+      "loss": 0.286,
+      "step": 17175
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 5.3918051534956926e-05,
+      "loss": 0.2842,
+      "step": 17180
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 5.3856104008080876e-05,
+      "loss": 0.2868,
+      "step": 17185
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 5.3794178971179165e-05,
+      "loss": 0.286,
+      "step": 17190
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 5.373227645443332e-05,
+      "loss": 0.2852,
+      "step": 17195
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 5.3670396488013854e-05,
+      "loss": 0.2878,
+      "step": 17200
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 5.360853910208028e-05,
+      "loss": 0.2866,
+      "step": 17205
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 5.3546704326781236e-05,
+      "loss": 0.2844,
+      "step": 17210
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 5.348489219225416e-05,
+      "loss": 0.2873,
+      "step": 17215
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 5.3423102728625574e-05,
+      "loss": 0.2894,
+      "step": 17220
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 5.33613359660109e-05,
+      "loss": 0.2865,
+      "step": 17225
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 5.329959193451448e-05,
+      "loss": 0.2866,
+      "step": 17230
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 5.3237870664229636e-05,
+      "loss": 0.2863,
+      "step": 17235
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 5.317617218523856e-05,
+      "loss": 0.2874,
+      "step": 17240
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 5.311449652761235e-05,
+      "loss": 0.2872,
+      "step": 17245
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 5.305284372141095e-05,
+      "loss": 0.2871,
+      "step": 17250
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 5.299121379668316e-05,
+      "loss": 0.2868,
+      "step": 17255
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.05078125,
+      "learning_rate": 5.292960678346675e-05,
+      "loss": 0.2868,
+      "step": 17260
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 5.286802271178815e-05,
+      "loss": 0.2868,
+      "step": 17265
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 5.2806461611662735e-05,
+      "loss": 0.2872,
+      "step": 17270
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 5.274492351309461e-05,
+      "loss": 0.2858,
+      "step": 17275
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 5.26834084460767e-05,
+      "loss": 0.287,
+      "step": 17280
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 5.2621916440590715e-05,
+      "loss": 0.2859,
+      "step": 17285
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 5.256044752660709e-05,
+      "loss": 0.2861,
+      "step": 17290
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 5.2499001734085044e-05,
+      "loss": 0.287,
+      "step": 17295
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 5.243757909297247e-05,
+      "loss": 0.2869,
+      "step": 17300
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 5.237617963320608e-05,
+      "loss": 0.2866,
+      "step": 17305
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 5.2314803384711195e-05,
+      "loss": 0.286,
+      "step": 17310
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 5.225345037740186e-05,
+      "loss": 0.2869,
+      "step": 17315
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 5.2192120641180786e-05,
+      "loss": 0.2856,
+      "step": 17320
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 5.213081420593933e-05,
+      "loss": 0.2875,
+      "step": 17325
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 5.20695311015575e-05,
+      "loss": 0.2862,
+      "step": 17330
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 5.200827135790396e-05,
+      "loss": 0.2866,
+      "step": 17335
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 5.194703500483593e-05,
+      "loss": 0.2865,
+      "step": 17340
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 5.188582207219931e-05,
+      "loss": 0.2858,
+      "step": 17345
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 5.182463258982846e-05,
+      "loss": 0.2851,
+      "step": 17350
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 5.1763466587546485e-05,
+      "loss": 0.287,
+      "step": 17355
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 5.170232409516496e-05,
+      "loss": 0.2872,
+      "step": 17360
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 5.1641205142483894e-05,
+      "loss": 0.2851,
+      "step": 17365
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 5.158010975929193e-05,
+      "loss": 0.2857,
+      "step": 17370
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 5.15190379753663e-05,
+      "loss": 0.2871,
+      "step": 17375
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 5.145798982047261e-05,
+      "loss": 0.2858,
+      "step": 17380
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 5.1396965324364986e-05,
+      "loss": 0.287,
+      "step": 17385
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 5.133596451678603e-05,
+      "loss": 0.285,
+      "step": 17390
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 5.127498742746675e-05,
+      "loss": 0.287,
+      "step": 17395
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0546875,
+      "learning_rate": 5.121403408612672e-05,
+      "loss": 0.288,
+      "step": 17400
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.05859375,
+      "learning_rate": 5.115310452247386e-05,
+      "loss": 0.2853,
+      "step": 17405
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 5.1092198766204415e-05,
+      "loss": 0.2868,
+      "step": 17410
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 5.103131684700314e-05,
+      "loss": 0.2872,
+      "step": 17415
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 5.097045879454313e-05,
+      "loss": 0.2857,
+      "step": 17420
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 5.090962463848592e-05,
+      "loss": 0.2882,
+      "step": 17425
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 5.0848814408481305e-05,
+      "loss": 0.2878,
+      "step": 17430
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 5.0788028134167456e-05,
+      "loss": 0.2871,
+      "step": 17435
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.048828125,
+      "learning_rate": 5.072726584517086e-05,
+      "loss": 0.286,
+      "step": 17440
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 5.066652757110628e-05,
+      "loss": 0.2881,
+      "step": 17445
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 5.0605813341576924e-05,
+      "loss": 0.2855,
+      "step": 17450
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 5.054512318617406e-05,
+      "loss": 0.2865,
+      "step": 17455
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 5.048445713447738e-05,
+      "loss": 0.2872,
+      "step": 17460
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 5.0423815216054724e-05,
+      "loss": 0.2859,
+      "step": 17465
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 5.036319746046232e-05,
+      "loss": 0.2845,
+      "step": 17470
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.056640625,
+      "learning_rate": 5.0302603897244474e-05,
+      "loss": 0.2863,
+      "step": 17475
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 5.024203455593375e-05,
+      "loss": 0.2852,
+      "step": 17480
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 5.018148946605092e-05,
+      "loss": 0.2863,
+      "step": 17485
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 5.012096865710494e-05,
+      "loss": 0.2859,
+      "step": 17490
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 5.006047215859289e-05,
+      "loss": 0.2845,
+      "step": 17495
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0625,
+      "learning_rate": 5.000000000000002e-05,
+      "loss": 0.2876,
+      "step": 17500
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 4.9939552210799755e-05,
+      "loss": 0.2875,
+      "step": 17505
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.05078125,
+      "learning_rate": 4.98791288204536e-05,
+      "loss": 0.2873,
+      "step": 17510
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 4.981872985841115e-05,
+      "loss": 0.2862,
+      "step": 17515
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 4.97583553541102e-05,
+      "loss": 0.2864,
+      "step": 17520
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 4.969800533697649e-05,
+      "loss": 0.2844,
+      "step": 17525
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 4.9637679836423924e-05,
+      "loss": 0.2869,
+      "step": 17530
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 4.957737888185439e-05,
+      "loss": 0.2862,
+      "step": 17535
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 4.9517102502657845e-05,
+      "loss": 0.2856,
+      "step": 17540
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.05078125,
+      "learning_rate": 4.945685072821227e-05,
+      "loss": 0.2853,
+      "step": 17545
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 4.939662358788364e-05,
+      "loss": 0.2856,
+      "step": 17550
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 4.933642111102594e-05,
+      "loss": 0.2871,
+      "step": 17555
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 4.927624332698109e-05,
+      "loss": 0.2868,
+      "step": 17560
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 4.921609026507907e-05,
+      "loss": 0.2849,
+      "step": 17565
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 4.915596195463773e-05,
+      "loss": 0.2853,
+      "step": 17570
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 4.909585842496287e-05,
+      "loss": 0.288,
+      "step": 17575
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.060546875,
+      "learning_rate": 4.9035779705348226e-05,
+      "loss": 0.2848,
+      "step": 17580
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 4.8975725825075435e-05,
+      "loss": 0.285,
+      "step": 17585
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 4.8915696813414026e-05,
+      "loss": 0.2871,
+      "step": 17590
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 4.885569269962142e-05,
+      "loss": 0.2869,
+      "step": 17595
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 4.8795713512942865e-05,
+      "loss": 0.2869,
+      "step": 17600
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 4.8735759282611516e-05,
+      "loss": 0.2871,
+      "step": 17605
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.048828125,
+      "learning_rate": 4.867583003784829e-05,
+      "loss": 0.2865,
+      "step": 17610
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 4.861592580786205e-05,
+      "loss": 0.2854,
+      "step": 17615
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 4.8556046621849346e-05,
+      "loss": 0.2855,
+      "step": 17620
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 4.8496192508994576e-05,
+      "loss": 0.2874,
+      "step": 17625
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 4.8436363498469906e-05,
+      "loss": 0.2881,
+      "step": 17630
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 4.837655961943526e-05,
+      "loss": 0.2861,
+      "step": 17635
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0546875,
+      "learning_rate": 4.8316780901038314e-05,
+      "loss": 0.288,
+      "step": 17640
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0390625,
+      "learning_rate": 4.825702737241452e-05,
+      "loss": 0.286,
+      "step": 17645
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 4.8197299062686995e-05,
+      "loss": 0.2879,
+      "step": 17650
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.035400390625,
+      "learning_rate": 4.8137596000966614e-05,
+      "loss": 0.2865,
+      "step": 17655
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 4.807791821635186e-05,
+      "loss": 0.2875,
+      "step": 17660
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 4.8018265737929044e-05,
+      "loss": 0.2852,
+      "step": 17665
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 4.7958638594772064e-05,
+      "loss": 0.2872,
+      "step": 17670
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.056640625,
+      "learning_rate": 4.78990368159424e-05,
+      "loss": 0.2852,
+      "step": 17675
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.048828125,
+      "learning_rate": 4.783946043048923e-05,
+      "loss": 0.2864,
+      "step": 17680
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 4.7779909467449414e-05,
+      "loss": 0.2864,
+      "step": 17685
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 4.7720383955847345e-05,
+      "loss": 0.2866,
+      "step": 17690
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 4.766088392469506e-05,
+      "loss": 0.287,
+      "step": 17695
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.06640625,
+      "learning_rate": 4.7601409402992106e-05,
+      "loss": 0.2874,
+      "step": 17700
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.052734375,
+      "learning_rate": 4.7541960419725626e-05,
+      "loss": 0.2876,
+      "step": 17705
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 4.748253700387042e-05,
+      "loss": 0.2867,
+      "step": 17710
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 4.742313918438872e-05,
+      "loss": 0.2871,
+      "step": 17715
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 4.736376699023023e-05,
+      "loss": 0.2864,
+      "step": 17720
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 4.7304420450332244e-05,
+      "loss": 0.2861,
+      "step": 17725
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.076171875,
+      "learning_rate": 4.724509959361961e-05,
+      "loss": 0.2878,
+      "step": 17730
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 4.7185804449004565e-05,
+      "loss": 0.2853,
+      "step": 17735
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0546875,
+      "learning_rate": 4.712653504538683e-05,
+      "loss": 0.2882,
+      "step": 17740
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 4.706729141165361e-05,
+      "loss": 0.2879,
+      "step": 17745
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 4.700807357667952e-05,
+      "loss": 0.2862,
+      "step": 17750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.048828125,
+      "learning_rate": 4.694888156932658e-05,
+      "loss": 0.2867,
+      "step": 17755
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 4.688971541844436e-05,
+      "loss": 0.2857,
+      "step": 17760
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 4.6830575152869616e-05,
+      "loss": 0.2872,
+      "step": 17765
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 4.6771460801426635e-05,
+      "loss": 0.2865,
+      "step": 17770
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 4.6712372392927e-05,
+      "loss": 0.2864,
+      "step": 17775
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.056640625,
+      "learning_rate": 4.665330995616974e-05,
+      "loss": 0.2872,
+      "step": 17780
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 4.6594273519941154e-05,
+      "loss": 0.2868,
+      "step": 17785
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 4.653526311301488e-05,
+      "loss": 0.2863,
+      "step": 17790
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 4.647627876415186e-05,
+      "loss": 0.2873,
+      "step": 17795
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.048828125,
+      "learning_rate": 4.6417320502100316e-05,
+      "loss": 0.2871,
+      "step": 17800
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.041015625,
+      "learning_rate": 4.6358388355595904e-05,
+      "loss": 0.2872,
+      "step": 17805
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 4.629948235336133e-05,
+      "loss": 0.2879,
+      "step": 17810
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.04296875,
+      "learning_rate": 4.62406025241067e-05,
+      "loss": 0.2866,
+      "step": 17815
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 4.6181748896529273e-05,
+      "loss": 0.2857,
+      "step": 17820
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 4.612292149931369e-05,
+      "loss": 0.2881,
+      "step": 17825
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 4.6064120361131656e-05,
+      "loss": 0.2866,
+      "step": 17830
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 4.600534551064215e-05,
+      "loss": 0.2863,
+      "step": 17835
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.04296875,
+      "learning_rate": 4.5946596976491295e-05,
+      "loss": 0.2855,
+      "step": 17840
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 4.588787478731242e-05,
+      "loss": 0.2874,
+      "step": 17845
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 4.582917897172603e-05,
+      "loss": 0.2854,
+      "step": 17850
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 4.577050955833973e-05,
+      "loss": 0.2869,
+      "step": 17855
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 4.5711866575748276e-05,
+      "loss": 0.286,
+      "step": 17860
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 4.565325005253356e-05,
+      "loss": 0.2862,
+      "step": 17865
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 4.559466001726451e-05,
+      "loss": 0.2861,
+      "step": 17870
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 4.5536096498497295e-05,
+      "loss": 0.2865,
+      "step": 17875
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 4.5477559524774994e-05,
+      "loss": 0.2877,
+      "step": 17880
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 4.541904912462784e-05,
+      "loss": 0.2871,
+      "step": 17885
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 4.5360565326573104e-05,
+      "loss": 0.286,
+      "step": 17890
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 4.530210815911504e-05,
+      "loss": 0.2885,
+      "step": 17895
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.04296875,
+      "learning_rate": 4.524367765074499e-05,
+      "loss": 0.2862,
+      "step": 17900
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 4.518527382994127e-05,
+      "loss": 0.2886,
+      "step": 17905
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 4.512689672516918e-05,
+      "loss": 0.2858,
+      "step": 17910
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 4.506854636488103e-05,
+      "loss": 0.2879,
+      "step": 17915
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 4.501022277751602e-05,
+      "loss": 0.2859,
+      "step": 17920
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 4.495192599150044e-05,
+      "loss": 0.2849,
+      "step": 17925
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 4.48936560352474e-05,
+      "loss": 0.2878,
+      "step": 17930
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 4.483541293715698e-05,
+      "loss": 0.2872,
+      "step": 17935
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 4.4777196725616146e-05,
+      "loss": 0.2865,
+      "step": 17940
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 4.471900742899876e-05,
+      "loss": 0.2851,
+      "step": 17945
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.044921875,
+      "learning_rate": 4.46608450756656e-05,
+      "loss": 0.2856,
+      "step": 17950
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 4.460270969396429e-05,
+      "loss": 0.2873,
+      "step": 17955
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 4.45446013122293e-05,
+      "loss": 0.288,
+      "step": 17960
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.05078125,
+      "learning_rate": 4.44865199587819e-05,
+      "loss": 0.2858,
+      "step": 17965
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 4.4428465661930343e-05,
+      "loss": 0.2853,
+      "step": 17970
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 4.437043844996952e-05,
+      "loss": 0.2876,
+      "step": 17975
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 4.431243835118124e-05,
+      "loss": 0.2854,
+      "step": 17980
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 4.425446539383393e-05,
+      "loss": 0.2852,
+      "step": 17985
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 4.419651960618302e-05,
+      "loss": 0.2866,
+      "step": 17990
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.05078125,
+      "learning_rate": 4.413860101647055e-05,
+      "loss": 0.2857,
+      "step": 17995
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 4.4080709652925336e-05,
+      "loss": 0.2858,
+      "step": 18000
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.060546875,
+      "learning_rate": 4.4022845543762915e-05,
+      "loss": 0.287,
+      "step": 18005
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 4.396500871718555e-05,
+      "loss": 0.2873,
+      "step": 18010
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 4.39071992013822e-05,
+      "loss": 0.2871,
+      "step": 18015
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 4.3849417024528564e-05,
+      "loss": 0.2861,
+      "step": 18020
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0546875,
+      "learning_rate": 4.379166221478697e-05,
+      "loss": 0.2865,
+      "step": 18025
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 4.3733934800306366e-05,
+      "loss": 0.2882,
+      "step": 18030
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 4.367623480922236e-05,
+      "loss": 0.2871,
+      "step": 18035
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 4.361856226965733e-05,
+      "loss": 0.2861,
+      "step": 18040
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 4.356091720972011e-05,
+      "loss": 0.2866,
+      "step": 18045
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0546875,
+      "learning_rate": 4.350329965750621e-05,
+      "loss": 0.2864,
+      "step": 18050
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 4.3445709641097745e-05,
+      "loss": 0.2862,
+      "step": 18055
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 4.3388147188563325e-05,
+      "loss": 0.2851,
+      "step": 18060
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 4.333061232795826e-05,
+      "loss": 0.2879,
+      "step": 18065
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 4.327310508732437e-05,
+      "loss": 0.2855,
+      "step": 18070
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 4.32156254946899e-05,
+      "loss": 0.2867,
+      "step": 18075
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 4.315817357806974e-05,
+      "loss": 0.2867,
+      "step": 18080
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.05078125,
+      "learning_rate": 4.310074936546521e-05,
+      "loss": 0.2867,
+      "step": 18085
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 4.304335288486426e-05,
+      "loss": 0.2849,
+      "step": 18090
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 4.29859841642412e-05,
+      "loss": 0.2863,
+      "step": 18095
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 4.2928643231556844e-05,
+      "loss": 0.2864,
+      "step": 18100
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 4.287133011475847e-05,
+      "loss": 0.2873,
+      "step": 18105
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.04296875,
+      "learning_rate": 4.2814044841779745e-05,
+      "loss": 0.2859,
+      "step": 18110
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 4.2756787440540936e-05,
+      "loss": 0.2861,
+      "step": 18115
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 4.26995579389485e-05,
+      "loss": 0.2877,
+      "step": 18120
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.04296875,
+      "learning_rate": 4.264235636489542e-05,
+      "loss": 0.2868,
+      "step": 18125
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.068359375,
+      "learning_rate": 4.2585182746261035e-05,
+      "loss": 0.286,
+      "step": 18130
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 4.252803711091112e-05,
+      "loss": 0.2869,
+      "step": 18135
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 4.2470919486697744e-05,
+      "loss": 0.2873,
+      "step": 18140
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 4.2413829901459344e-05,
+      "loss": 0.2849,
+      "step": 18145
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 4.235676838302068e-05,
+      "loss": 0.2839,
+      "step": 18150
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 4.229973495919286e-05,
+      "loss": 0.2859,
+      "step": 18155
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 4.224272965777326e-05,
+      "loss": 0.2848,
+      "step": 18160
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 4.2185752506545585e-05,
+      "loss": 0.2863,
+      "step": 18165
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 4.212880353327979e-05,
+      "loss": 0.2878,
+      "step": 18170
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 4.207188276573214e-05,
+      "loss": 0.2863,
+      "step": 18175
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 4.201499023164508e-05,
+      "loss": 0.2854,
+      "step": 18180
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 4.195812595874739e-05,
+      "loss": 0.2859,
+      "step": 18185
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 4.190128997475402e-05,
+      "loss": 0.2883,
+      "step": 18190
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 4.184448230736613e-05,
+      "loss": 0.2884,
+      "step": 18195
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 4.1787702984271074e-05,
+      "loss": 0.2864,
+      "step": 18200
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 4.173095203314241e-05,
+      "loss": 0.2894,
+      "step": 18205
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 4.167422948163986e-05,
+      "loss": 0.287,
+      "step": 18210
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.046875,
+      "learning_rate": 4.161753535740932e-05,
+      "loss": 0.2879,
+      "step": 18215
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 4.15608696880828e-05,
+      "loss": 0.2884,
+      "step": 18220
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 4.150423250127845e-05,
+      "loss": 0.2857,
+      "step": 18225
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 4.144762382460059e-05,
+      "loss": 0.2871,
+      "step": 18230
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 4.1391043685639576e-05,
+      "loss": 0.2859,
+      "step": 18235
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 4.133449211197188e-05,
+      "loss": 0.2877,
+      "step": 18240
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 4.1277969131160045e-05,
+      "loss": 0.2859,
+      "step": 18245
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.05078125,
+      "learning_rate": 4.12214747707527e-05,
+      "loss": 0.2869,
+      "step": 18250
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 4.11650090582845e-05,
+      "loss": 0.2872,
+      "step": 18255
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 4.110857202127615e-05,
+      "loss": 0.2864,
+      "step": 18260
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 4.1052163687234366e-05,
+      "loss": 0.287,
+      "step": 18265
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 4.099578408365191e-05,
+      "loss": 0.287,
+      "step": 18270
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 4.093943323800745e-05,
+      "loss": 0.286,
+      "step": 18275
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 4.08831111777658e-05,
+      "loss": 0.287,
+      "step": 18280
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 4.08268179303776e-05,
+      "loss": 0.2867,
+      "step": 18285
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.060546875,
+      "learning_rate": 4.077055352327953e-05,
+      "loss": 0.2864,
+      "step": 18290
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 4.071431798389408e-05,
+      "loss": 0.2869,
+      "step": 18295
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 4.065811133962987e-05,
+      "loss": 0.2868,
+      "step": 18300
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 4.0601933617881294e-05,
+      "loss": 0.2861,
+      "step": 18305
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 4.05457848460287e-05,
+      "loss": 0.2866,
+      "step": 18310
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 4.048966505143831e-05,
+      "loss": 0.2865,
+      "step": 18315
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 4.0433574261462206e-05,
+      "loss": 0.2869,
+      "step": 18320
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.044921875,
+      "learning_rate": 4.037751250343841e-05,
+      "loss": 0.2865,
+      "step": 18325
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 4.032147980469072e-05,
+      "loss": 0.287,
+      "step": 18330
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 4.026547619252883e-05,
+      "loss": 0.2867,
+      "step": 18335
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 4.020950169424815e-05,
+      "loss": 0.2861,
+      "step": 18340
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 4.015355633712996e-05,
+      "loss": 0.2847,
+      "step": 18345
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 4.009764014844143e-05,
+      "loss": 0.287,
+      "step": 18350
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 4.004175315543538e-05,
+      "loss": 0.2878,
+      "step": 18355
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 3.9985895385350456e-05,
+      "loss": 0.2868,
+      "step": 18360
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 3.9930066865411075e-05,
+      "loss": 0.2868,
+      "step": 18365
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 3.987426762282733e-05,
+      "loss": 0.2878,
+      "step": 18370
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 3.981849768479517e-05,
+      "loss": 0.2874,
+      "step": 18375
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 3.976275707849616e-05,
+      "loss": 0.2854,
+      "step": 18380
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 3.9707045831097555e-05,
+      "loss": 0.2871,
+      "step": 18385
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 3.9651363969752344e-05,
+      "loss": 0.2854,
+      "step": 18390
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 3.9595711521599224e-05,
+      "loss": 0.287,
+      "step": 18395
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 3.954008851376252e-05,
+      "loss": 0.2863,
+      "step": 18400
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 3.948449497335219e-05,
+      "loss": 0.2878,
+      "step": 18405
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 3.942893092746387e-05,
+      "loss": 0.2849,
+      "step": 18410
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 3.9373396403178786e-05,
+      "loss": 0.2875,
+      "step": 18415
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 3.931789142756377e-05,
+      "loss": 0.2872,
+      "step": 18420
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 3.9262416027671356e-05,
+      "loss": 0.2867,
+      "step": 18425
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 3.920697023053949e-05,
+      "loss": 0.2858,
+      "step": 18430
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 3.915155406319181e-05,
+      "loss": 0.2868,
+      "step": 18435
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 3.9096167552637454e-05,
+      "loss": 0.2856,
+      "step": 18440
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 3.904081072587119e-05,
+      "loss": 0.2863,
+      "step": 18445
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 3.8985483609873244e-05,
+      "loss": 0.2868,
+      "step": 18450
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 3.893018623160938e-05,
+      "loss": 0.2875,
+      "step": 18455
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.046875,
+      "learning_rate": 3.887491861803085e-05,
+      "loss": 0.287,
+      "step": 18460
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 3.88196807960744e-05,
+      "loss": 0.2852,
+      "step": 18465
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 3.876447279266238e-05,
+      "loss": 0.2859,
+      "step": 18470
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 3.8709294634702376e-05,
+      "loss": 0.2872,
+      "step": 18475
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 3.8654146349087606e-05,
+      "loss": 0.2879,
+      "step": 18480
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 3.859902796269663e-05,
+      "loss": 0.2872,
+      "step": 18485
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 3.854393950239355e-05,
+      "loss": 0.2855,
+      "step": 18490
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 3.8488880995027786e-05,
+      "loss": 0.2858,
+      "step": 18495
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 3.843385246743417e-05,
+      "loss": 0.2867,
+      "step": 18500
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 3.837885394643296e-05,
+      "loss": 0.286,
+      "step": 18505
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 3.832388545882976e-05,
+      "loss": 0.2866,
+      "step": 18510
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 3.8268947031415514e-05,
+      "loss": 0.2864,
+      "step": 18515
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 3.821403869096658e-05,
+      "loss": 0.2863,
+      "step": 18520
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0546875,
+      "learning_rate": 3.8159160464244606e-05,
+      "loss": 0.2866,
+      "step": 18525
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 3.8104312377996564e-05,
+      "loss": 0.2872,
+      "step": 18530
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 3.8049494458954725e-05,
+      "loss": 0.2864,
+      "step": 18535
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 3.799470673383674e-05,
+      "loss": 0.2868,
+      "step": 18540
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 3.793994922934544e-05,
+      "loss": 0.2872,
+      "step": 18545
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 3.788522197216897e-05,
+      "loss": 0.2859,
+      "step": 18550
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 3.783052498898073e-05,
+      "loss": 0.2855,
+      "step": 18555
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 3.777585830643937e-05,
+      "loss": 0.2871,
+      "step": 18560
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.048828125,
+      "learning_rate": 3.7721221951188765e-05,
+      "loss": 0.2874,
+      "step": 18565
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 3.766661594985801e-05,
+      "loss": 0.287,
+      "step": 18570
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.041015625,
+      "learning_rate": 3.7612040329061405e-05,
+      "loss": 0.2865,
+      "step": 18575
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 3.755749511539845e-05,
+      "loss": 0.2862,
+      "step": 18580
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 3.7502980335453774e-05,
+      "loss": 0.2858,
+      "step": 18585
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 3.7448496015797295e-05,
+      "loss": 0.2868,
+      "step": 18590
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.044921875,
+      "learning_rate": 3.739404218298398e-05,
+      "loss": 0.2863,
+      "step": 18595
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 3.733961886355398e-05,
+      "loss": 0.2865,
+      "step": 18600
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.056640625,
+      "learning_rate": 3.728522608403249e-05,
+      "loss": 0.2848,
+      "step": 18605
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 3.7230863870929964e-05,
+      "loss": 0.2873,
+      "step": 18610
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 3.717653225074186e-05,
+      "loss": 0.2865,
+      "step": 18615
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 3.712223124994875e-05,
+      "loss": 0.2869,
+      "step": 18620
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 3.7067960895016275e-05,
+      "loss": 0.2854,
+      "step": 18625
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 3.701372121239512e-05,
+      "loss": 0.2861,
+      "step": 18630
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 3.695951222852112e-05,
+      "loss": 0.2859,
+      "step": 18635
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 3.690533396981504e-05,
+      "loss": 0.2854,
+      "step": 18640
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 3.685118646268272e-05,
+      "loss": 0.2871,
+      "step": 18645
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 3.679706973351491e-05,
+      "loss": 0.287,
+      "step": 18650
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 3.674298380868756e-05,
+      "loss": 0.2875,
+      "step": 18655
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 3.668892871456144e-05,
+      "loss": 0.286,
+      "step": 18660
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 3.6634904477482354e-05,
+      "loss": 0.2872,
+      "step": 18665
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 3.6580911123781056e-05,
+      "loss": 0.2846,
+      "step": 18670
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 3.6526948679773257e-05,
+      "loss": 0.2847,
+      "step": 18675
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 3.647301717175956e-05,
+      "loss": 0.2859,
+      "step": 18680
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 3.641911662602559e-05,
+      "loss": 0.2854,
+      "step": 18685
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 3.6365247068841814e-05,
+      "loss": 0.2885,
+      "step": 18690
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 3.631140852646355e-05,
+      "loss": 0.2868,
+      "step": 18695
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 3.6257601025131026e-05,
+      "loss": 0.2871,
+      "step": 18700
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 3.6203824591069456e-05,
+      "loss": 0.2863,
+      "step": 18705
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 3.615007925048878e-05,
+      "loss": 0.2872,
+      "step": 18710
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 3.60963650295838e-05,
+      "loss": 0.2868,
+      "step": 18715
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 3.604268195453421e-05,
+      "loss": 0.2867,
+      "step": 18720
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 3.5989030051504434e-05,
+      "loss": 0.2872,
+      "step": 18725
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 3.5935409346643835e-05,
+      "loss": 0.2853,
+      "step": 18730
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 3.5881819866086484e-05,
+      "loss": 0.2858,
+      "step": 18735
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 3.582826163595119e-05,
+      "loss": 0.2863,
+      "step": 18740
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 3.577473468234156e-05,
+      "loss": 0.2875,
+      "step": 18745
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 3.5721239031346066e-05,
+      "loss": 0.2853,
+      "step": 18750
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 3.56677747090378e-05,
+      "loss": 0.2876,
+      "step": 18755
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 3.561434174147463e-05,
+      "loss": 0.2869,
+      "step": 18760
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 3.556094015469913e-05,
+      "loss": 0.2833,
+      "step": 18765
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 3.5507569974738574e-05,
+      "loss": 0.287,
+      "step": 18770
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 3.545423122760493e-05,
+      "loss": 0.2861,
+      "step": 18775
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 3.540092393929494e-05,
+      "loss": 0.2862,
+      "step": 18780
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 3.534764813578982e-05,
+      "loss": 0.2856,
+      "step": 18785
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 3.52944038430556e-05,
+      "loss": 0.2858,
+      "step": 18790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.048828125,
+      "learning_rate": 3.524119108704286e-05,
+      "loss": 0.2863,
+      "step": 18795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 3.518800989368691e-05,
+      "loss": 0.2842,
+      "step": 18800
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 3.51348602889076e-05,
+      "loss": 0.2877,
+      "step": 18805
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 3.50817422986094e-05,
+      "loss": 0.2865,
+      "step": 18810
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 3.5028655948681355e-05,
+      "loss": 0.287,
+      "step": 18815
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.044921875,
+      "learning_rate": 3.497560126499709e-05,
+      "loss": 0.2853,
+      "step": 18820
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.044921875,
+      "learning_rate": 3.492257827341492e-05,
+      "loss": 0.2874,
+      "step": 18825
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 3.4869586999777495e-05,
+      "loss": 0.286,
+      "step": 18830
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 3.481662746991214e-05,
+      "loss": 0.2879,
+      "step": 18835
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 3.4763699709630716e-05,
+      "loss": 0.2858,
+      "step": 18840
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 3.4710803744729515e-05,
+      "loss": 0.2887,
+      "step": 18845
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 3.465793960098945e-05,
+      "loss": 0.2848,
+      "step": 18850
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 3.460510730417585e-05,
+      "loss": 0.2877,
+      "step": 18855
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 3.455230688003852e-05,
+      "loss": 0.2858,
+      "step": 18860
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 3.4499538354311755e-05,
+      "loss": 0.2873,
+      "step": 18865
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 3.444680175271428e-05,
+      "loss": 0.286,
+      "step": 18870
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 3.439409710094929e-05,
+      "loss": 0.2854,
+      "step": 18875
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 3.4341424424704375e-05,
+      "loss": 0.2856,
+      "step": 18880
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 3.4288783749651564e-05,
+      "loss": 0.2858,
+      "step": 18885
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 3.423617510144727e-05,
+      "loss": 0.2842,
+      "step": 18890
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 3.418359850573234e-05,
+      "loss": 0.2875,
+      "step": 18895
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 3.413105398813195e-05,
+      "loss": 0.2859,
+      "step": 18900
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 3.4078541574255664e-05,
+      "loss": 0.2859,
+      "step": 18905
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 3.40260612896974e-05,
+      "loss": 0.2865,
+      "step": 18910
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 3.397361316003539e-05,
+      "loss": 0.2873,
+      "step": 18915
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 3.3921197210832235e-05,
+      "loss": 0.2867,
+      "step": 18920
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 3.386881346763483e-05,
+      "loss": 0.2883,
+      "step": 18925
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 3.3816461955974365e-05,
+      "loss": 0.2864,
+      "step": 18930
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 3.376414270136633e-05,
+      "loss": 0.2863,
+      "step": 18935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 3.371185572931048e-05,
+      "loss": 0.286,
+      "step": 18940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0546875,
+      "learning_rate": 3.3659601065290893e-05,
+      "loss": 0.2884,
+      "step": 18945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.05078125,
+      "learning_rate": 3.360737873477584e-05,
+      "loss": 0.286,
+      "step": 18950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 3.355518876321787e-05,
+      "loss": 0.2874,
+      "step": 18955
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 3.3503031176053656e-05,
+      "loss": 0.2866,
+      "step": 18960
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0390625,
+      "learning_rate": 3.3450905998704275e-05,
+      "loss": 0.2862,
+      "step": 18965
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 3.339881325657484e-05,
+      "loss": 0.2862,
+      "step": 18970
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.056640625,
+      "learning_rate": 3.334675297505476e-05,
+      "loss": 0.286,
+      "step": 18975
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 3.3294725179517574e-05,
+      "loss": 0.2868,
+      "step": 18980
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.04296875,
+      "learning_rate": 3.3242729895320946e-05,
+      "loss": 0.2865,
+      "step": 18985
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 3.319076714780682e-05,
+      "loss": 0.2871,
+      "step": 18990
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 3.313883696230119e-05,
+      "loss": 0.2871,
+      "step": 18995
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 3.308693936411421e-05,
+      "loss": 0.2865,
+      "step": 19000
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 3.303507437854009e-05,
+      "loss": 0.286,
+      "step": 19005
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 3.2983242030857174e-05,
+      "loss": 0.2866,
+      "step": 19010
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 3.2931442346328004e-05,
+      "loss": 0.2874,
+      "step": 19015
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 3.287967535019908e-05,
+      "loss": 0.2859,
+      "step": 19020
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 3.2827941067700996e-05,
+      "loss": 0.2865,
+      "step": 19025
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.048828125,
+      "learning_rate": 3.277623952404842e-05,
+      "loss": 0.2874,
+      "step": 19030
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.056640625,
+      "learning_rate": 3.272457074444003e-05,
+      "loss": 0.2855,
+      "step": 19035
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 3.2672934754058616e-05,
+      "loss": 0.288,
+      "step": 19040
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 3.2621331578070934e-05,
+      "loss": 0.2871,
+      "step": 19045
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 3.2569761241627696e-05,
+      "loss": 0.2889,
+      "step": 19050
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 3.251822376986363e-05,
+      "loss": 0.2876,
+      "step": 19055
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 3.246671918789755e-05,
+      "loss": 0.2863,
+      "step": 19060
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 3.2415247520832146e-05,
+      "loss": 0.2861,
+      "step": 19065
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 3.236380879375408e-05,
+      "loss": 0.2861,
+      "step": 19070
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.037109375,
+      "learning_rate": 3.231240303173394e-05,
+      "loss": 0.2876,
+      "step": 19075
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 3.226103025982628e-05,
+      "loss": 0.2863,
+      "step": 19080
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 3.220969050306955e-05,
+      "loss": 0.2872,
+      "step": 19085
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 3.21583837864862e-05,
+      "loss": 0.2863,
+      "step": 19090
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 3.210711013508242e-05,
+      "loss": 0.2861,
+      "step": 19095
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 3.205586957384838e-05,
+      "loss": 0.2869,
+      "step": 19100
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 3.200466212775808e-05,
+      "loss": 0.2852,
+      "step": 19105
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.037109375,
+      "learning_rate": 3.195348782176948e-05,
+      "loss": 0.2882,
+      "step": 19110
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 3.190234668082427e-05,
+      "loss": 0.2881,
+      "step": 19115
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 3.1851238729848034e-05,
+      "loss": 0.2857,
+      "step": 19120
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 3.1800163993750166e-05,
+      "loss": 0.2864,
+      "step": 19125
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 3.174912249742382e-05,
+      "loss": 0.2868,
+      "step": 19130
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 3.1698114265746124e-05,
+      "loss": 0.288,
+      "step": 19135
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 3.164713932357776e-05,
+      "loss": 0.2854,
+      "step": 19140
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 3.159619769576333e-05,
+      "loss": 0.2857,
+      "step": 19145
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 3.154528940713113e-05,
+      "loss": 0.2866,
+      "step": 19150
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 3.1494414482493304e-05,
+      "loss": 0.2848,
+      "step": 19155
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 3.144357294664565e-05,
+      "loss": 0.2869,
+      "step": 19160
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 3.1392764824367704e-05,
+      "loss": 0.2867,
+      "step": 19165
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 3.134199014042274e-05,
+      "loss": 0.285,
+      "step": 19170
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 3.129124891955771e-05,
+      "loss": 0.2856,
+      "step": 19175
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 3.124054118650327e-05,
+      "loss": 0.289,
+      "step": 19180
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 3.1189866965973766e-05,
+      "loss": 0.2886,
+      "step": 19185
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.052734375,
+      "learning_rate": 3.113922628266718e-05,
+      "loss": 0.2863,
+      "step": 19190
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 3.108861916126518e-05,
+      "loss": 0.2869,
+      "step": 19195
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 3.103804562643302e-05,
+      "loss": 0.2872,
+      "step": 19200
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 3.098750570281969e-05,
+      "loss": 0.2873,
+      "step": 19205
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.05078125,
+      "learning_rate": 3.093699941505771e-05,
+      "loss": 0.2874,
+      "step": 19210
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 3.0886526787763234e-05,
+      "loss": 0.2877,
+      "step": 19215
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 3.0836087845536e-05,
+      "loss": 0.2852,
+      "step": 19220
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 3.078568261295933e-05,
+      "loss": 0.2864,
+      "step": 19225
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 3.073531111460013e-05,
+      "loss": 0.2864,
+      "step": 19230
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 3.068497337500886e-05,
+      "loss": 0.2877,
+      "step": 19235
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 3.063466941871952e-05,
+      "loss": 0.2859,
+      "step": 19240
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.052734375,
+      "learning_rate": 3.058439927024962e-05,
+      "loss": 0.2863,
+      "step": 19245
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.048828125,
+      "learning_rate": 3.053416295410026e-05,
+      "loss": 0.2847,
+      "step": 19250
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 3.0483960494756016e-05,
+      "loss": 0.2875,
+      "step": 19255
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 3.0433791916684916e-05,
+      "loss": 0.2868,
+      "step": 19260
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 3.038365724433858e-05,
+      "loss": 0.2869,
+      "step": 19265
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.05078125,
+      "learning_rate": 3.0333556502151926e-05,
+      "loss": 0.2867,
+      "step": 19270
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 3.0283489714543556e-05,
+      "loss": 0.2843,
+      "step": 19275
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 3.023345690591537e-05,
+      "loss": 0.2866,
+      "step": 19280
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 3.018345810065275e-05,
+      "loss": 0.2862,
+      "step": 19285
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 3.0133493323124505e-05,
+      "loss": 0.2848,
+      "step": 19290
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 3.0083562597682847e-05,
+      "loss": 0.2872,
+      "step": 19295
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 3.0033665948663448e-05,
+      "loss": 0.2847,
+      "step": 19300
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 2.9983803400385312e-05,
+      "loss": 0.2858,
+      "step": 19305
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 2.993397497715086e-05,
+      "loss": 0.2871,
+      "step": 19310
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 2.9884180703245767e-05,
+      "loss": 0.2858,
+      "step": 19315
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 2.9834420602939263e-05,
+      "loss": 0.289,
+      "step": 19320
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 2.9784694700483762e-05,
+      "loss": 0.2864,
+      "step": 19325
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 2.9735003020115092e-05,
+      "loss": 0.2871,
+      "step": 19330
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 2.968534558605236e-05,
+      "loss": 0.2864,
+      "step": 19335
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 2.9635722422497993e-05,
+      "loss": 0.2866,
+      "step": 19340
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 2.9586133553637683e-05,
+      "loss": 0.2861,
+      "step": 19345
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 2.953657900364053e-05,
+      "loss": 0.2875,
+      "step": 19350
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 2.9487058796658783e-05,
+      "loss": 0.2872,
+      "step": 19355
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 2.9437572956827964e-05,
+      "loss": 0.2875,
+      "step": 19360
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 2.938812150826684e-05,
+      "loss": 0.2854,
+      "step": 19365
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 2.933870447507753e-05,
+      "loss": 0.2869,
+      "step": 19370
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 2.9289321881345254e-05,
+      "loss": 0.2844,
+      "step": 19375
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 2.9239973751138495e-05,
+      "loss": 0.2857,
+      "step": 19380
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 2.9190660108508917e-05,
+      "loss": 0.2877,
+      "step": 19385
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 2.9141380977491373e-05,
+      "loss": 0.2855,
+      "step": 19390
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 2.9092136382103973e-05,
+      "loss": 0.2892,
+      "step": 19395
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 2.904292634634793e-05,
+      "loss": 0.287,
+      "step": 19400
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 2.899375089420756e-05,
+      "loss": 0.2871,
+      "step": 19405
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 2.894461004965038e-05,
+      "loss": 0.2871,
+      "step": 19410
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 2.8895503836627103e-05,
+      "loss": 0.287,
+      "step": 19415
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 2.8846432279071467e-05,
+      "loss": 0.2859,
+      "step": 19420
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 2.879739540090036e-05,
+      "loss": 0.2866,
+      "step": 19425
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.04296875,
+      "learning_rate": 2.874839322601375e-05,
+      "loss": 0.2864,
+      "step": 19430
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 2.869942577829471e-05,
+      "loss": 0.2856,
+      "step": 19435
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 2.865049308160934e-05,
+      "loss": 0.285,
+      "step": 19440
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 2.860159515980695e-05,
+      "loss": 0.2863,
+      "step": 19445
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 2.8552732036719687e-05,
+      "loss": 0.2862,
+      "step": 19450
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 2.8503903736162875e-05,
+      "loss": 0.285,
+      "step": 19455
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 2.8455110281934803e-05,
+      "loss": 0.2865,
+      "step": 19460
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 2.840635169781688e-05,
+      "loss": 0.2866,
+      "step": 19465
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 2.835762800757341e-05,
+      "loss": 0.2853,
+      "step": 19470
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 2.8308939234951726e-05,
+      "loss": 0.2862,
+      "step": 19475
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 2.826028540368215e-05,
+      "loss": 0.2872,
+      "step": 19480
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 2.8211666537477933e-05,
+      "loss": 0.2873,
+      "step": 19485
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 2.816308266003541e-05,
+      "loss": 0.2857,
+      "step": 19490
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 2.8114533795033683e-05,
+      "loss": 0.2876,
+      "step": 19495
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 2.8066019966134904e-05,
+      "loss": 0.2873,
+      "step": 19500
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 2.8017541196984142e-05,
+      "loss": 0.2861,
+      "step": 19505
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.05859375,
+      "learning_rate": 2.7969097511209308e-05,
+      "loss": 0.2863,
+      "step": 19510
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.052734375,
+      "learning_rate": 2.7920688932421335e-05,
+      "loss": 0.2877,
+      "step": 19515
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 2.7872315484213952e-05,
+      "loss": 0.2873,
+      "step": 19520
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 2.7823977190163786e-05,
+      "loss": 0.2863,
+      "step": 19525
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 2.7775674073830337e-05,
+      "loss": 0.2846,
+      "step": 19530
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.04296875,
+      "learning_rate": 2.772740615875594e-05,
+      "loss": 0.2846,
+      "step": 19535
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 2.7679173468465812e-05,
+      "loss": 0.2888,
+      "step": 19540
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 2.7630976026467968e-05,
+      "loss": 0.2879,
+      "step": 19545
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 2.7582813856253275e-05,
+      "loss": 0.2874,
+      "step": 19550
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 2.753468698129533e-05,
+      "loss": 0.2865,
+      "step": 19555
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 2.7486595425050665e-05,
+      "loss": 0.2857,
+      "step": 19560
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.041015625,
+      "learning_rate": 2.743853921095848e-05,
+      "loss": 0.2858,
+      "step": 19565
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 2.7390518362440808e-05,
+      "loss": 0.2845,
+      "step": 19570
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.052734375,
+      "learning_rate": 2.734253290290242e-05,
+      "loss": 0.2866,
+      "step": 19575
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 2.7294582855730832e-05,
+      "loss": 0.2863,
+      "step": 19580
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.046875,
+      "learning_rate": 2.7246668244296323e-05,
+      "loss": 0.2861,
+      "step": 19585
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.7198789091951902e-05,
+      "loss": 0.2857,
+      "step": 19590
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 2.715094542203327e-05,
+      "loss": 0.2862,
+      "step": 19595
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.052734375,
+      "learning_rate": 2.7103137257858868e-05,
+      "loss": 0.2867,
+      "step": 19600
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 2.7055364622729773e-05,
+      "loss": 0.2865,
+      "step": 19605
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 2.700762753992985e-05,
+      "loss": 0.2871,
+      "step": 19610
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 2.6959926032725535e-05,
+      "loss": 0.2859,
+      "step": 19615
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 2.6912260124366006e-05,
+      "loss": 0.2875,
+      "step": 19620
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 2.6864629838082956e-05,
+      "loss": 0.2863,
+      "step": 19625
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 2.681703519709089e-05,
+      "loss": 0.2869,
+      "step": 19630
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 2.676947622458683e-05,
+      "loss": 0.2869,
+      "step": 19635
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.04296875,
+      "learning_rate": 2.672195294375045e-05,
+      "loss": 0.286,
+      "step": 19640
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 2.6674465377744017e-05,
+      "loss": 0.2848,
+      "step": 19645
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 2.6627013549712355e-05,
+      "loss": 0.286,
+      "step": 19650
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 2.657959748278297e-05,
+      "loss": 0.2868,
+      "step": 19655
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0546875,
+      "learning_rate": 2.6532217200065858e-05,
+      "loss": 0.2867,
+      "step": 19660
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 2.6484872724653608e-05,
+      "loss": 0.2862,
+      "step": 19665
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 2.6437564079621267e-05,
+      "loss": 0.2855,
+      "step": 19670
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 2.639029128802657e-05,
+      "loss": 0.2883,
+      "step": 19675
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 2.634305437290968e-05,
+      "loss": 0.2872,
+      "step": 19680
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 2.6295853357293298e-05,
+      "loss": 0.2883,
+      "step": 19685
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 2.624868826418262e-05,
+      "loss": 0.2851,
+      "step": 19690
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.044921875,
+      "learning_rate": 2.6201559116565345e-05,
+      "loss": 0.2877,
+      "step": 19695
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 2.615446593741161e-05,
+      "loss": 0.2866,
+      "step": 19700
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 2.6107408749674122e-05,
+      "loss": 0.2877,
+      "step": 19705
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 2.606038757628798e-05,
+      "loss": 0.2855,
+      "step": 19710
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 2.6013402440170676e-05,
+      "loss": 0.2881,
+      "step": 19715
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 2.5966453364222186e-05,
+      "loss": 0.2868,
+      "step": 19720
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 2.5919540371325e-05,
+      "loss": 0.2865,
+      "step": 19725
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 2.5872663484343884e-05,
+      "loss": 0.2851,
+      "step": 19730
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 2.582582272612609e-05,
+      "loss": 0.2858,
+      "step": 19735
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 2.5779018119501208e-05,
+      "loss": 0.2855,
+      "step": 19740
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 2.573224968728123e-05,
+      "loss": 0.2844,
+      "step": 19745
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 2.5685517452260567e-05,
+      "loss": 0.2869,
+      "step": 19750
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 2.5638821437215944e-05,
+      "loss": 0.2859,
+      "step": 19755
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0625,
+      "learning_rate": 2.5592161664906368e-05,
+      "loss": 0.286,
+      "step": 19760
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 2.554553815807328e-05,
+      "loss": 0.2858,
+      "step": 19765
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 2.549895093944039e-05,
+      "loss": 0.2867,
+      "step": 19770
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06640625,
+      "learning_rate": 2.5452400031713785e-05,
+      "loss": 0.2886,
+      "step": 19775
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 2.540588545758179e-05,
+      "loss": 0.2868,
+      "step": 19780
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 2.535940723971505e-05,
+      "loss": 0.2849,
+      "step": 19785
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 2.5312965400766474e-05,
+      "loss": 0.287,
+      "step": 19790
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 2.5266559963371216e-05,
+      "loss": 0.2874,
+      "step": 19795
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 2.5220190950146827e-05,
+      "loss": 0.2863,
+      "step": 19800
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 2.5173858383692906e-05,
+      "loss": 0.2871,
+      "step": 19805
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 2.512756228659141e-05,
+      "loss": 0.2871,
+      "step": 19810
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 2.508130268140646e-05,
+      "loss": 0.2869,
+      "step": 19815
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0390625,
+      "learning_rate": 2.5035079590684497e-05,
+      "loss": 0.2864,
+      "step": 19820
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 2.4988893036954043e-05,
+      "loss": 0.2865,
+      "step": 19825
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 2.4942743042725892e-05,
+      "loss": 0.2855,
+      "step": 19830
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 2.4896629630492973e-05,
+      "loss": 0.2885,
+      "step": 19835
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 2.48505528227304e-05,
+      "loss": 0.2881,
+      "step": 19840
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 2.480451264189546e-05,
+      "loss": 0.2864,
+      "step": 19845
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.046875,
+      "learning_rate": 2.4758509110427575e-05,
+      "loss": 0.2893,
+      "step": 19850
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 2.4712542250748304e-05,
+      "loss": 0.2865,
+      "step": 19855
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 2.4666612085261342e-05,
+      "loss": 0.2868,
+      "step": 19860
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 2.4620718636352457e-05,
+      "loss": 0.2859,
+      "step": 19865
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 2.4574861926389615e-05,
+      "loss": 0.286,
+      "step": 19870
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 2.45290419777228e-05,
+      "loss": 0.2854,
+      "step": 19875
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 2.4483258812684096e-05,
+      "loss": 0.2857,
+      "step": 19880
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 2.443751245358765e-05,
+      "loss": 0.2862,
+      "step": 19885
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 2.43918029227297e-05,
+      "loss": 0.2853,
+      "step": 19890
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 2.43461302423885e-05,
+      "loss": 0.2869,
+      "step": 19895
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 2.4300494434824373e-05,
+      "loss": 0.2858,
+      "step": 19900
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 2.425489552227964e-05,
+      "loss": 0.2882,
+      "step": 19905
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 2.420933352697865e-05,
+      "loss": 0.2867,
+      "step": 19910
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 2.4163808471127812e-05,
+      "loss": 0.2882,
+      "step": 19915
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 2.411832037691545e-05,
+      "loss": 0.2859,
+      "step": 19920
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 2.407286926651192e-05,
+      "loss": 0.2862,
+      "step": 19925
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 2.4027455162069567e-05,
+      "loss": 0.287,
+      "step": 19930
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 2.3982078085722582e-05,
+      "loss": 0.2857,
+      "step": 19935
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 2.3936738059587282e-05,
+      "loss": 0.288,
+      "step": 19940
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 2.389143510576184e-05,
+      "loss": 0.2869,
+      "step": 19945
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 2.3846169246326343e-05,
+      "loss": 0.2864,
+      "step": 19950
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.046875,
+      "learning_rate": 2.3800940503342828e-05,
+      "loss": 0.2882,
+      "step": 19955
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.05078125,
+      "learning_rate": 2.37557488988552e-05,
+      "loss": 0.2858,
+      "step": 19960
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 2.3710594454889378e-05,
+      "loss": 0.2871,
+      "step": 19965
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.046875,
+      "learning_rate": 2.3665477193453034e-05,
+      "loss": 0.2859,
+      "step": 19970
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0703125,
+      "learning_rate": 2.362039713653581e-05,
+      "loss": 0.287,
+      "step": 19975
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.038330078125,
+      "learning_rate": 2.35753543061091e-05,
+      "loss": 0.2862,
+      "step": 19980
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 2.3530348724126307e-05,
+      "loss": 0.2878,
+      "step": 19985
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 2.3485380412522585e-05,
+      "loss": 0.2842,
+      "step": 19990
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 2.3440449393214948e-05,
+      "loss": 0.2845,
+      "step": 19995
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 2.339555568810221e-05,
+      "loss": 0.2859,
+      "step": 20000
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 2.3350699319065026e-05,
+      "loss": 0.2856,
+      "step": 20005
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 2.3305880307965833e-05,
+      "loss": 0.2855,
+      "step": 20010
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 2.3261098676648907e-05,
+      "loss": 0.2858,
+      "step": 20015
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 2.321635444694028e-05,
+      "loss": 0.2854,
+      "step": 20020
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 2.3171647640647687e-05,
+      "loss": 0.2864,
+      "step": 20025
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 2.3126978279560684e-05,
+      "loss": 0.2867,
+      "step": 20030
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 2.3082346385450637e-05,
+      "loss": 0.2867,
+      "step": 20035
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 2.3037751980070555e-05,
+      "loss": 0.2869,
+      "step": 20040
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 2.2993195085155205e-05,
+      "loss": 0.2864,
+      "step": 20045
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 2.2948675722421086e-05,
+      "loss": 0.2857,
+      "step": 20050
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.038330078125,
+      "learning_rate": 2.2904193913566363e-05,
+      "loss": 0.2869,
+      "step": 20055
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 2.2859749680270982e-05,
+      "loss": 0.2859,
+      "step": 20060
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 2.2815343044196525e-05,
+      "loss": 0.2865,
+      "step": 20065
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 2.277097402698619e-05,
+      "loss": 0.2865,
+      "step": 20070
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 2.2726642650264895e-05,
+      "loss": 0.2886,
+      "step": 20075
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 2.2682348935639274e-05,
+      "loss": 0.2871,
+      "step": 20080
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 2.2638092904697518e-05,
+      "loss": 0.2842,
+      "step": 20085
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 2.259387457900948e-05,
+      "loss": 0.2872,
+      "step": 20090
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 2.254969398012663e-05,
+      "loss": 0.2875,
+      "step": 20095
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 2.2505551129582047e-05,
+      "loss": 0.2868,
+      "step": 20100
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 2.246144604889042e-05,
+      "loss": 0.2878,
+      "step": 20105
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 2.241737875954808e-05,
+      "loss": 0.2868,
+      "step": 20110
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 2.237334928303283e-05,
+      "loss": 0.2871,
+      "step": 20115
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 2.2329357640804117e-05,
+      "loss": 0.2841,
+      "step": 20120
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0390625,
+      "learning_rate": 2.2285403854302912e-05,
+      "loss": 0.2855,
+      "step": 20125
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 2.22414879449518e-05,
+      "loss": 0.2857,
+      "step": 20130
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 2.219760993415485e-05,
+      "loss": 0.2864,
+      "step": 20135
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 2.2153769843297667e-05,
+      "loss": 0.2864,
+      "step": 20140
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 2.210996769374737e-05,
+      "loss": 0.287,
+      "step": 20145
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 2.2066203506852566e-05,
+      "loss": 0.285,
+      "step": 20150
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 2.202247730394349e-05,
+      "loss": 0.289,
+      "step": 20155
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 2.1978789106331665e-05,
+      "loss": 0.2865,
+      "step": 20160
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 2.1935138935310206e-05,
+      "loss": 0.286,
+      "step": 20165
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.048828125,
+      "learning_rate": 2.1891526812153672e-05,
+      "loss": 0.2891,
+      "step": 20170
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 2.1847952758118117e-05,
+      "loss": 0.2873,
+      "step": 20175
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 2.1804416794440995e-05,
+      "loss": 0.2864,
+      "step": 20180
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 2.1760918942341192e-05,
+      "loss": 0.2864,
+      "step": 20185
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 2.1717459223019033e-05,
+      "loss": 0.2844,
+      "step": 20190
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 2.1674037657656266e-05,
+      "loss": 0.2886,
+      "step": 20195
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 2.163065426741603e-05,
+      "loss": 0.2868,
+      "step": 20200
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 2.1587309073442863e-05,
+      "loss": 0.2861,
+      "step": 20205
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 2.154400209686268e-05,
+      "loss": 0.287,
+      "step": 20210
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 2.1500733358782786e-05,
+      "loss": 0.2861,
+      "step": 20215
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 2.1457502880291812e-05,
+      "loss": 0.2847,
+      "step": 20220
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 2.1414310682459802e-05,
+      "loss": 0.2869,
+      "step": 20225
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.052734375,
+      "learning_rate": 2.137115678633811e-05,
+      "loss": 0.2861,
+      "step": 20230
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 2.1328041212959403e-05,
+      "loss": 0.287,
+      "step": 20235
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 2.128496398333768e-05,
+      "loss": 0.2861,
+      "step": 20240
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 2.1241925118468287e-05,
+      "loss": 0.2873,
+      "step": 20245
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 2.119892463932781e-05,
+      "loss": 0.2866,
+      "step": 20250
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 2.115596256687419e-05,
+      "loss": 0.2859,
+      "step": 20255
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 2.1113038922046602e-05,
+      "loss": 0.2854,
+      "step": 20260
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 2.107015372576552e-05,
+      "loss": 0.2864,
+      "step": 20265
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0546875,
+      "learning_rate": 2.1027306998932627e-05,
+      "loss": 0.2847,
+      "step": 20270
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 2.098449876243096e-05,
+      "loss": 0.2857,
+      "step": 20275
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 2.09417290371247e-05,
+      "loss": 0.2861,
+      "step": 20280
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 2.0898997843859334e-05,
+      "loss": 0.2884,
+      "step": 20285
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 2.0856305203461436e-05,
+      "loss": 0.2857,
+      "step": 20290
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 2.0813651136738954e-05,
+      "loss": 0.2868,
+      "step": 20295
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 2.0771035664480942e-05,
+      "loss": 0.2865,
+      "step": 20300
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 2.0728458807457662e-05,
+      "loss": 0.2851,
+      "step": 20305
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 2.068592058642056e-05,
+      "loss": 0.2863,
+      "step": 20310
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.046875,
+      "learning_rate": 2.0643421022102217e-05,
+      "loss": 0.2866,
+      "step": 20315
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 2.0600960135216462e-05,
+      "loss": 0.2873,
+      "step": 20320
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 2.0558537946458177e-05,
+      "loss": 0.2861,
+      "step": 20325
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 2.051615447650347e-05,
+      "loss": 0.2864,
+      "step": 20330
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 2.0473809746009444e-05,
+      "loss": 0.2871,
+      "step": 20335
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.05078125,
+      "learning_rate": 2.0431503775614457e-05,
+      "loss": 0.2877,
+      "step": 20340
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 2.0389236585937945e-05,
+      "loss": 0.287,
+      "step": 20345
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 2.0347008197580374e-05,
+      "loss": 0.2851,
+      "step": 20350
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 2.030481863112339e-05,
+      "loss": 0.2873,
+      "step": 20355
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 2.026266790712965e-05,
+      "loss": 0.2861,
+      "step": 20360
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 2.0220556046142893e-05,
+      "loss": 0.2857,
+      "step": 20365
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 2.017848306868797e-05,
+      "loss": 0.2847,
+      "step": 20370
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 2.013644899527074e-05,
+      "loss": 0.2878,
+      "step": 20375
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 2.009445384637805e-05,
+      "loss": 0.2868,
+      "step": 20380
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 2.005249764247783e-05,
+      "loss": 0.2853,
+      "step": 20385
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 2.0010580404019062e-05,
+      "loss": 0.2862,
+      "step": 20390
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 1.9968702151431696e-05,
+      "loss": 0.2889,
+      "step": 20395
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 1.9926862905126665e-05,
+      "loss": 0.287,
+      "step": 20400
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 1.9885062685495904e-05,
+      "loss": 0.2879,
+      "step": 20405
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 1.9843301512912327e-05,
+      "loss": 0.2865,
+      "step": 20410
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.9801579407729866e-05,
+      "loss": 0.2874,
+      "step": 20415
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.9759896390283362e-05,
+      "loss": 0.2867,
+      "step": 20420
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.9718252480888566e-05,
+      "loss": 0.2872,
+      "step": 20425
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 1.9676647699842242e-05,
+      "loss": 0.2854,
+      "step": 20430
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 1.963508206742202e-05,
+      "loss": 0.2857,
+      "step": 20435
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.9593555603886538e-05,
+      "loss": 0.2871,
+      "step": 20440
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 1.955206832947526e-05,
+      "loss": 0.2844,
+      "step": 20445
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 1.9510620264408596e-05,
+      "loss": 0.2891,
+      "step": 20450
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 1.946921142888781e-05,
+      "loss": 0.2885,
+      "step": 20455
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 1.9427841843095063e-05,
+      "loss": 0.2878,
+      "step": 20460
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.938651152719344e-05,
+      "loss": 0.2864,
+      "step": 20465
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 1.9345220501326777e-05,
+      "loss": 0.2875,
+      "step": 20470
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 1.930396878561983e-05,
+      "loss": 0.2883,
+      "step": 20475
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 1.9262756400178162e-05,
+      "loss": 0.2857,
+      "step": 20480
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 1.922158336508825e-05,
+      "loss": 0.2864,
+      "step": 20485
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.918044970041729e-05,
+      "loss": 0.2878,
+      "step": 20490
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 1.9139355426213347e-05,
+      "loss": 0.2852,
+      "step": 20495
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.9098300562505266e-05,
+      "loss": 0.2847,
+      "step": 20500
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.041015625,
+      "learning_rate": 1.9057285129302683e-05,
+      "loss": 0.2854,
+      "step": 20505
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 1.9016309146596023e-05,
+      "loss": 0.2866,
+      "step": 20510
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.05078125,
+      "learning_rate": 1.897537263435648e-05,
+      "loss": 0.2872,
+      "step": 20515
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 1.8934475612536017e-05,
+      "loss": 0.2858,
+      "step": 20520
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.8893618101067355e-05,
+      "loss": 0.2869,
+      "step": 20525
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.885280011986391e-05,
+      "loss": 0.2869,
+      "step": 20530
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.8812021688819915e-05,
+      "loss": 0.2858,
+      "step": 20535
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 1.877128282781028e-05,
+      "loss": 0.2864,
+      "step": 20540
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.8730583556690605e-05,
+      "loss": 0.2856,
+      "step": 20545
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.8689923895297245e-05,
+      "loss": 0.2862,
+      "step": 20550
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.8649303863447198e-05,
+      "loss": 0.2856,
+      "step": 20555
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.8608723480938206e-05,
+      "loss": 0.2873,
+      "step": 20560
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 1.8568182767548626e-05,
+      "loss": 0.2864,
+      "step": 20565
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.041015625,
+      "learning_rate": 1.852768174303752e-05,
+      "loss": 0.2863,
+      "step": 20570
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 1.848722042714457e-05,
+      "loss": 0.2854,
+      "step": 20575
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.8446798839590186e-05,
+      "loss": 0.2859,
+      "step": 20580
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 1.8406417000075325e-05,
+      "loss": 0.2866,
+      "step": 20585
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.8366074928281607e-05,
+      "loss": 0.2861,
+      "step": 20590
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.8325772643871265e-05,
+      "loss": 0.2865,
+      "step": 20595
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 1.8285510166487152e-05,
+      "loss": 0.2879,
+      "step": 20600
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.8245287515752708e-05,
+      "loss": 0.2867,
+      "step": 20605
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 1.8205104711271957e-05,
+      "loss": 0.2864,
+      "step": 20610
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.816496177262952e-05,
+      "loss": 0.2887,
+      "step": 20615
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 1.812485871939056e-05,
+      "loss": 0.2857,
+      "step": 20620
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 1.808479557110081e-05,
+      "loss": 0.2862,
+      "step": 20625
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.804477234728661e-05,
+      "loss": 0.2867,
+      "step": 20630
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.04296875,
+      "learning_rate": 1.8004789067454764e-05,
+      "loss": 0.286,
+      "step": 20635
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 1.7964845751092664e-05,
+      "loss": 0.2873,
+      "step": 20640
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 1.792494241766811e-05,
+      "loss": 0.2868,
+      "step": 20645
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0390625,
+      "learning_rate": 1.78850790866296e-05,
+      "loss": 0.286,
+      "step": 20650
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.7845255777406e-05,
+      "loss": 0.2842,
+      "step": 20655
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 1.7805472509406696e-05,
+      "loss": 0.2859,
+      "step": 20660
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 1.7765729302021596e-05,
+      "loss": 0.2867,
+      "step": 20665
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.7726026174621003e-05,
+      "loss": 0.2877,
+      "step": 20670
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 1.7686363146555805e-05,
+      "loss": 0.2882,
+      "step": 20675
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 1.7646740237157256e-05,
+      "loss": 0.2873,
+      "step": 20680
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.760715746573709e-05,
+      "loss": 0.2866,
+      "step": 20685
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 1.7567614851587443e-05,
+      "loss": 0.2862,
+      "step": 20690
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 1.752811241398089e-05,
+      "loss": 0.2881,
+      "step": 20695
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.7488650172170496e-05,
+      "loss": 0.2851,
+      "step": 20700
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 1.744922814538964e-05,
+      "loss": 0.2838,
+      "step": 20705
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 1.7409846352852143e-05,
+      "loss": 0.2869,
+      "step": 20710
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 1.737050481375223e-05,
+      "loss": 0.2863,
+      "step": 20715
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 1.7331203547264453e-05,
+      "loss": 0.2861,
+      "step": 20720
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.7291942572543807e-05,
+      "loss": 0.2858,
+      "step": 20725
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.7252721908725632e-05,
+      "loss": 0.2873,
+      "step": 20730
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 1.721354157492555e-05,
+      "loss": 0.2872,
+      "step": 20735
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 1.7174401590239587e-05,
+      "loss": 0.2858,
+      "step": 20740
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 1.7135301973744124e-05,
+      "loss": 0.2871,
+      "step": 20745
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 1.7096242744495837e-05,
+      "loss": 0.2872,
+      "step": 20750
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 1.7057223921531707e-05,
+      "loss": 0.2883,
+      "step": 20755
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 1.7018245523869036e-05,
+      "loss": 0.2866,
+      "step": 20760
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.697930757050542e-05,
+      "loss": 0.2879,
+      "step": 20765
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.6940410080418723e-05,
+      "loss": 0.2862,
+      "step": 20770
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 1.690155307256719e-05,
+      "loss": 0.2884,
+      "step": 20775
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0625,
+      "learning_rate": 1.686273656588917e-05,
+      "loss": 0.2859,
+      "step": 20780
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.6823960579303377e-05,
+      "loss": 0.286,
+      "step": 20785
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 1.678522513170875e-05,
+      "loss": 0.2867,
+      "step": 20790
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.6746530241984504e-05,
+      "loss": 0.2868,
+      "step": 20795
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 1.6707875928990058e-05,
+      "loss": 0.2873,
+      "step": 20800
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 1.6669262211565028e-05,
+      "loss": 0.2861,
+      "step": 20805
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0546875,
+      "learning_rate": 1.6630689108529284e-05,
+      "loss": 0.2877,
+      "step": 20810
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.6592156638682886e-05,
+      "loss": 0.2869,
+      "step": 20815
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.6553664820806102e-05,
+      "loss": 0.2861,
+      "step": 20820
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.6515213673659357e-05,
+      "loss": 0.2852,
+      "step": 20825
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 1.6476803215983294e-05,
+      "loss": 0.2854,
+      "step": 20830
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.643843346649866e-05,
+      "loss": 0.2852,
+      "step": 20835
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 1.640010444390646e-05,
+      "loss": 0.2847,
+      "step": 20840
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.6361816166887768e-05,
+      "loss": 0.2856,
+      "step": 20845
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.632356865410384e-05,
+      "loss": 0.2868,
+      "step": 20850
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.628536192419603e-05,
+      "loss": 0.2868,
+      "step": 20855
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 1.6247195995785837e-05,
+      "loss": 0.2852,
+      "step": 20860
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 1.6209070887474876e-05,
+      "loss": 0.2864,
+      "step": 20865
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 1.6170986617844863e-05,
+      "loss": 0.2863,
+      "step": 20870
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 1.6132943205457606e-05,
+      "loss": 0.2869,
+      "step": 20875
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 1.6094940668855006e-05,
+      "loss": 0.2854,
+      "step": 20880
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 1.6056979026559004e-05,
+      "loss": 0.2857,
+      "step": 20885
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.601905829707171e-05,
+      "loss": 0.2861,
+      "step": 20890
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 1.598117849887518e-05,
+      "loss": 0.2874,
+      "step": 20895
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.060546875,
+      "learning_rate": 1.5943339650431576e-05,
+      "loss": 0.2867,
+      "step": 20900
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 1.5905541770183096e-05,
+      "loss": 0.2866,
+      "step": 20905
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.586778487655197e-05,
+      "loss": 0.2862,
+      "step": 20910
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.5830068987940438e-05,
+      "loss": 0.2865,
+      "step": 20915
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 1.579239412273078e-05,
+      "loss": 0.2872,
+      "step": 20920
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.5754760299285252e-05,
+      "loss": 0.2852,
+      "step": 20925
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 1.5717167535946144e-05,
+      "loss": 0.286,
+      "step": 20930
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.567961585103567e-05,
+      "loss": 0.2869,
+      "step": 20935
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.064453125,
+      "learning_rate": 1.564210526285612e-05,
+      "loss": 0.2851,
+      "step": 20940
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.560463578968967e-05,
+      "loss": 0.2871,
+      "step": 20945
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 1.5567207449798515e-05,
+      "loss": 0.287,
+      "step": 20950
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 1.5529820261424698e-05,
+      "loss": 0.2873,
+      "step": 20955
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 1.5492474242790366e-05,
+      "loss": 0.2869,
+      "step": 20960
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.046875,
+      "learning_rate": 1.545516941209747e-05,
+      "loss": 0.2881,
+      "step": 20965
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 1.541790578752794e-05,
+      "loss": 0.2854,
+      "step": 20970
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.538068338724361e-05,
+      "loss": 0.284,
+      "step": 20975
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 1.5343502229386207e-05,
+      "loss": 0.2875,
+      "step": 20980
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.530636233207743e-05,
+      "loss": 0.2871,
+      "step": 20985
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.526926371341878e-05,
+      "loss": 0.2849,
+      "step": 20990
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 1.5232206391491699e-05,
+      "loss": 0.2866,
+      "step": 20995
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.048828125,
+      "learning_rate": 1.5195190384357404e-05,
+      "loss": 0.2866,
+      "step": 21000
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.5158215710057123e-05,
+      "loss": 0.2864,
+      "step": 21005
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 1.5121282386611824e-05,
+      "loss": 0.2881,
+      "step": 21010
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 1.5084390432022377e-05,
+      "loss": 0.2868,
+      "step": 21015
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.056640625,
+      "learning_rate": 1.5047539864269478e-05,
+      "loss": 0.2846,
+      "step": 21020
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 1.5010730701313625e-05,
+      "loss": 0.2863,
+      "step": 21025
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.04296875,
+      "learning_rate": 1.4973962961095134e-05,
+      "loss": 0.2875,
+      "step": 21030
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 1.4937236661534226e-05,
+      "loss": 0.2854,
+      "step": 21035
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 1.4900551820530828e-05,
+      "loss": 0.2882,
+      "step": 21040
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 1.486390845596466e-05,
+      "loss": 0.2865,
+      "step": 21045
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.4827306585695234e-05,
+      "loss": 0.2854,
+      "step": 21050
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 1.4790746227561925e-05,
+      "loss": 0.2857,
+      "step": 21055
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0703125,
+      "learning_rate": 1.4754227399383757e-05,
+      "loss": 0.2858,
+      "step": 21060
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.4717750118959584e-05,
+      "loss": 0.2892,
+      "step": 21065
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 1.468131440406798e-05,
+      "loss": 0.285,
+      "step": 21070
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.052734375,
+      "learning_rate": 1.4644920272467244e-05,
+      "loss": 0.2869,
+      "step": 21075
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.4608567741895495e-05,
+      "loss": 0.2861,
+      "step": 21080
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.4572256830070497e-05,
+      "loss": 0.2861,
+      "step": 21085
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 1.4535987554689712e-05,
+      "loss": 0.2874,
+      "step": 21090
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.4499759933430346e-05,
+      "loss": 0.2868,
+      "step": 21095
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 1.4463573983949341e-05,
+      "loss": 0.2857,
+      "step": 21100
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 1.4427429723883257e-05,
+      "loss": 0.2864,
+      "step": 21105
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 1.439132717084839e-05,
+      "loss": 0.2866,
+      "step": 21110
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.4355266342440677e-05,
+      "loss": 0.287,
+      "step": 21115
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.4319247256235714e-05,
+      "loss": 0.2869,
+      "step": 21120
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 1.4283269929788779e-05,
+      "loss": 0.2866,
+      "step": 21125
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.4247334380634792e-05,
+      "loss": 0.2868,
+      "step": 21130
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 1.4211440626288286e-05,
+      "loss": 0.2881,
+      "step": 21135
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 1.4175588684243446e-05,
+      "loss": 0.2872,
+      "step": 21140
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 1.4139778571974049e-05,
+      "loss": 0.2861,
+      "step": 21145
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 1.4104010306933557e-05,
+      "loss": 0.2882,
+      "step": 21150
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 1.406828390655497e-05,
+      "loss": 0.2854,
+      "step": 21155
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.40325993882509e-05,
+      "loss": 0.2881,
+      "step": 21160
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 1.3996956769413538e-05,
+      "loss": 0.2864,
+      "step": 21165
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 1.3961356067414666e-05,
+      "loss": 0.2863,
+      "step": 21170
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.3925797299605647e-05,
+      "loss": 0.286,
+      "step": 21175
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.3890280483317374e-05,
+      "loss": 0.2868,
+      "step": 21180
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.3854805635860336e-05,
+      "loss": 0.2862,
+      "step": 21185
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 1.3819372774524508e-05,
+      "loss": 0.2865,
+      "step": 21190
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 1.3783981916579446e-05,
+      "loss": 0.2874,
+      "step": 21195
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 1.3748633079274253e-05,
+      "loss": 0.2868,
+      "step": 21200
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 1.3713326279837501e-05,
+      "loss": 0.2871,
+      "step": 21205
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.3678061535477304e-05,
+      "loss": 0.2874,
+      "step": 21210
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 1.3642838863381257e-05,
+      "loss": 0.2875,
+      "step": 21215
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 1.3607658280716473e-05,
+      "loss": 0.2873,
+      "step": 21220
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 1.3572519804629536e-05,
+      "loss": 0.2865,
+      "step": 21225
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.3537423452246523e-05,
+      "loss": 0.2872,
+      "step": 21230
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 1.3502369240672941e-05,
+      "loss": 0.288,
+      "step": 21235
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 1.34673571869938e-05,
+      "loss": 0.2867,
+      "step": 21240
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.3432387308273575e-05,
+      "loss": 0.2852,
+      "step": 21245
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.339745962155613e-05,
+      "loss": 0.2854,
+      "step": 21250
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.052734375,
+      "learning_rate": 1.3362574143864814e-05,
+      "loss": 0.2849,
+      "step": 21255
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 1.3327730892202383e-05,
+      "loss": 0.2871,
+      "step": 21260
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 1.3292929883550998e-05,
+      "loss": 0.2885,
+      "step": 21265
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.046875,
+      "learning_rate": 1.3258171134872265e-05,
+      "loss": 0.286,
+      "step": 21270
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.3223454663107172e-05,
+      "loss": 0.2863,
+      "step": 21275
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 1.3188780485176088e-05,
+      "loss": 0.2876,
+      "step": 21280
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 1.3154148617978812e-05,
+      "loss": 0.2872,
+      "step": 21285
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 1.3119559078394461e-05,
+      "loss": 0.2868,
+      "step": 21290
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 1.3085011883281606e-05,
+      "loss": 0.2867,
+      "step": 21295
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 1.30505070494781e-05,
+      "loss": 0.286,
+      "step": 21300
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.3016044593801202e-05,
+      "loss": 0.2849,
+      "step": 21305
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.2981624533047432e-05,
+      "loss": 0.2859,
+      "step": 21310
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 1.294724688399278e-05,
+      "loss": 0.2867,
+      "step": 21315
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 1.2912911663392469e-05,
+      "loss": 0.2849,
+      "step": 21320
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 1.2878618887981064e-05,
+      "loss": 0.2858,
+      "step": 21325
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 1.2844368574472454e-05,
+      "loss": 0.2852,
+      "step": 21330
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 1.2810160739559796e-05,
+      "loss": 0.2853,
+      "step": 21335
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 1.2775995399915631e-05,
+      "loss": 0.2859,
+      "step": 21340
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 1.2741872572191682e-05,
+      "loss": 0.2874,
+      "step": 21345
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 1.2707792273019048e-05,
+      "loss": 0.2889,
+      "step": 21350
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 1.2673754519008008e-05,
+      "loss": 0.2866,
+      "step": 21355
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 1.2639759326748135e-05,
+      "loss": 0.2861,
+      "step": 21360
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 1.260580671280832e-05,
+      "loss": 0.2878,
+      "step": 21365
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.257189669373664e-05,
+      "loss": 0.2854,
+      "step": 21370
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0703125,
+      "learning_rate": 1.2538029286060426e-05,
+      "loss": 0.2867,
+      "step": 21375
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 1.2504204506286243e-05,
+      "loss": 0.2887,
+      "step": 21380
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.2470422370899838e-05,
+      "loss": 0.2855,
+      "step": 21385
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 1.243668289636628e-05,
+      "loss": 0.2863,
+      "step": 21390
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 1.2402986099129765e-05,
+      "loss": 0.287,
+      "step": 21395
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 1.2369331995613665e-05,
+      "loss": 0.2861,
+      "step": 21400
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 1.2335720602220569e-05,
+      "loss": 0.2878,
+      "step": 21405
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.2302151935332329e-05,
+      "loss": 0.2868,
+      "step": 21410
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 1.2268626011309858e-05,
+      "loss": 0.2859,
+      "step": 21415
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 1.2235142846493308e-05,
+      "loss": 0.2864,
+      "step": 21420
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.2201702457201947e-05,
+      "loss": 0.2855,
+      "step": 21425
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.2168304859734226e-05,
+      "loss": 0.2867,
+      "step": 21430
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 1.2134950070367723e-05,
+      "loss": 0.2862,
+      "step": 21435
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.2101638105359169e-05,
+      "loss": 0.2854,
+      "step": 21440
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 1.206836898094439e-05,
+      "loss": 0.2862,
+      "step": 21445
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 1.2035142713338366e-05,
+      "loss": 0.2873,
+      "step": 21450
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 1.2001959318735156e-05,
+      "loss": 0.2876,
+      "step": 21455
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.196881881330798e-05,
+      "loss": 0.2878,
+      "step": 21460
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.1935721213209105e-05,
+      "loss": 0.2867,
+      "step": 21465
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 1.1902666534569883e-05,
+      "loss": 0.2865,
+      "step": 21470
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 1.1869654793500784e-05,
+      "loss": 0.2879,
+      "step": 21475
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 1.1836686006091313e-05,
+      "loss": 0.2875,
+      "step": 21480
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 1.1803760188410073e-05,
+      "loss": 0.2859,
+      "step": 21485
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 1.1770877356504683e-05,
+      "loss": 0.286,
+      "step": 21490
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 1.1738037526401857e-05,
+      "loss": 0.2854,
+      "step": 21495
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.1705240714107302e-05,
+      "loss": 0.2867,
+      "step": 21500
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.1672486935605831e-05,
+      "loss": 0.2854,
+      "step": 21505
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 1.1639776206861196e-05,
+      "loss": 0.2881,
+      "step": 21510
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 1.1607108543816248e-05,
+      "loss": 0.2873,
+      "step": 21515
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 1.1574483962392767e-05,
+      "loss": 0.2881,
+      "step": 21520
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 1.1541902478491606e-05,
+      "loss": 0.2874,
+      "step": 21525
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 1.1509364107992583e-05,
+      "loss": 0.285,
+      "step": 21530
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.048828125,
+      "learning_rate": 1.1476868866754486e-05,
+      "loss": 0.2846,
+      "step": 21535
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 1.1444416770615118e-05,
+      "loss": 0.2854,
+      "step": 21540
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 1.1412007835391236e-05,
+      "loss": 0.2865,
+      "step": 21545
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 1.1379642076878527e-05,
+      "loss": 0.2859,
+      "step": 21550
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 1.1347319510851717e-05,
+      "loss": 0.2861,
+      "step": 21555
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 1.1315040153064416e-05,
+      "loss": 0.2864,
+      "step": 21560
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 1.1282804019249182e-05,
+      "loss": 0.2864,
+      "step": 21565
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.046875,
+      "learning_rate": 1.1250611125117527e-05,
+      "loss": 0.2869,
+      "step": 21570
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 1.1218461486359877e-05,
+      "loss": 0.285,
+      "step": 21575
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 1.1186355118645554e-05,
+      "loss": 0.2871,
+      "step": 21580
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 1.1154292037622838e-05,
+      "loss": 0.2848,
+      "step": 21585
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.1122272258918865e-05,
+      "loss": 0.2861,
+      "step": 21590
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 1.1090295798139672e-05,
+      "loss": 0.2849,
+      "step": 21595
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.1058362670870249e-05,
+      "loss": 0.2864,
+      "step": 21600
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 1.1026472892674378e-05,
+      "loss": 0.2877,
+      "step": 21605
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 1.099462647909475e-05,
+      "loss": 0.2854,
+      "step": 21610
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.0962823445652959e-05,
+      "loss": 0.285,
+      "step": 21615
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.04296875,
+      "learning_rate": 1.093106380784934e-05,
+      "loss": 0.2845,
+      "step": 21620
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.0899347581163221e-05,
+      "loss": 0.2871,
+      "step": 21625
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.0867674781052684e-05,
+      "loss": 0.2866,
+      "step": 21630
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.0836045422954666e-05,
+      "loss": 0.2861,
+      "step": 21635
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.0804459522284926e-05,
+      "loss": 0.2872,
+      "step": 21640
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.0772917094438051e-05,
+      "loss": 0.286,
+      "step": 21645
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 1.0741418154787442e-05,
+      "loss": 0.2867,
+      "step": 21650
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 1.0709962718685318e-05,
+      "loss": 0.2875,
+      "step": 21655
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 1.067855080146266e-05,
+      "loss": 0.2857,
+      "step": 21660
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.0647182418429225e-05,
+      "loss": 0.2854,
+      "step": 21665
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 1.0615857584873623e-05,
+      "loss": 0.2859,
+      "step": 21670
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 1.0584576316063188e-05,
+      "loss": 0.2887,
+      "step": 21675
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 1.0553338627244025e-05,
+      "loss": 0.2864,
+      "step": 21680
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 1.0522144533640998e-05,
+      "loss": 0.2865,
+      "step": 21685
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 1.0490994050457748e-05,
+      "loss": 0.2867,
+      "step": 21690
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 1.0459887192876594e-05,
+      "loss": 0.2848,
+      "step": 21695
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.046875,
+      "learning_rate": 1.042882397605871e-05,
+      "loss": 0.2846,
+      "step": 21700
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 1.0397804415143909e-05,
+      "loss": 0.2882,
+      "step": 21705
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.0366828525250726e-05,
+      "loss": 0.2857,
+      "step": 21710
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 1.0335896321476413e-05,
+      "loss": 0.2864,
+      "step": 21715
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.0305007818897006e-05,
+      "loss": 0.2863,
+      "step": 21720
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.0274163032567163e-05,
+      "loss": 0.285,
+      "step": 21725
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.0243361977520249e-05,
+      "loss": 0.2872,
+      "step": 21730
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 1.0212604668768343e-05,
+      "loss": 0.2861,
+      "step": 21735
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.0181891121302145e-05,
+      "loss": 0.2863,
+      "step": 21740
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.0151221350091134e-05,
+      "loss": 0.287,
+      "step": 21745
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 1.0120595370083318e-05,
+      "loss": 0.2856,
+      "step": 21750
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.046875,
+      "learning_rate": 1.009001319620545e-05,
+      "loss": 0.2859,
+      "step": 21755
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 1.0059474843362892e-05,
+      "loss": 0.2876,
+      "step": 21760
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 1.0028980326439707e-05,
+      "loss": 0.2868,
+      "step": 21765
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 9.998529660298539e-06,
+      "loss": 0.2862,
+      "step": 21770
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 9.968122859780648e-06,
+      "loss": 0.2873,
+      "step": 21775
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.04296875,
+      "learning_rate": 9.937759939705971e-06,
+      "loss": 0.2861,
+      "step": 21780
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 9.907440914873e-06,
+      "loss": 0.2859,
+      "step": 21785
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 9.877165800058874e-06,
+      "loss": 0.2865,
+      "step": 21790
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 9.84693461001932e-06,
+      "loss": 0.2865,
+      "step": 21795
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 9.816747359488632e-06,
+      "loss": 0.2871,
+      "step": 21800
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 9.786604063179728e-06,
+      "loss": 0.2855,
+      "step": 21805
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.052734375,
+      "learning_rate": 9.756504735784067e-06,
+      "loss": 0.2868,
+      "step": 21810
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 9.726449391971714e-06,
+      "loss": 0.2858,
+      "step": 21815
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 9.696438046391288e-06,
+      "loss": 0.2867,
+      "step": 21820
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 9.666470713669918e-06,
+      "loss": 0.2857,
+      "step": 21825
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 9.636547408413355e-06,
+      "loss": 0.2848,
+      "step": 21830
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 9.606668145205833e-06,
+      "loss": 0.2873,
+      "step": 21835
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 9.576832938610137e-06,
+      "loss": 0.2879,
+      "step": 21840
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 9.5470418031676e-06,
+      "loss": 0.2863,
+      "step": 21845
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 9.517294753398064e-06,
+      "loss": 0.287,
+      "step": 21850
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 9.487591803799856e-06,
+      "loss": 0.2868,
+      "step": 21855
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 9.457932968849825e-06,
+      "loss": 0.2863,
+      "step": 21860
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 9.428318263003378e-06,
+      "loss": 0.2861,
+      "step": 21865
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 9.398747700694322e-06,
+      "loss": 0.2877,
+      "step": 21870
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 9.369221296335006e-06,
+      "loss": 0.2858,
+      "step": 21875
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 9.339739064316233e-06,
+      "loss": 0.2885,
+      "step": 21880
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 9.310301019007285e-06,
+      "loss": 0.2878,
+      "step": 21885
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 9.280907174755915e-06,
+      "loss": 0.2858,
+      "step": 21890
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 9.251557545888312e-06,
+      "loss": 0.2866,
+      "step": 21895
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 9.222252146709142e-06,
+      "loss": 0.2875,
+      "step": 21900
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 9.192990991501482e-06,
+      "loss": 0.2863,
+      "step": 21905
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 9.163774094526889e-06,
+      "loss": 0.2856,
+      "step": 21910
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 9.134601470025306e-06,
+      "loss": 0.2867,
+      "step": 21915
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 9.105473132215125e-06,
+      "loss": 0.2858,
+      "step": 21920
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.046875,
+      "learning_rate": 9.076389095293148e-06,
+      "loss": 0.2871,
+      "step": 21925
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 9.047349373434566e-06,
+      "loss": 0.286,
+      "step": 21930
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 9.018353980792993e-06,
+      "loss": 0.2876,
+      "step": 21935
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 8.989402931500434e-06,
+      "loss": 0.287,
+      "step": 21940
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 8.960496239667282e-06,
+      "loss": 0.2869,
+      "step": 21945
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 8.931633919382298e-06,
+      "loss": 0.2871,
+      "step": 21950
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 8.90281598471262e-06,
+      "loss": 0.2868,
+      "step": 21955
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 8.87404244970378e-06,
+      "loss": 0.2859,
+      "step": 21960
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 8.845313328379634e-06,
+      "loss": 0.2864,
+      "step": 21965
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 8.816628634742441e-06,
+      "loss": 0.2871,
+      "step": 21970
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 8.787988382772705e-06,
+      "loss": 0.2876,
+      "step": 21975
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 8.759392586429393e-06,
+      "loss": 0.2845,
+      "step": 21980
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 8.730841259649725e-06,
+      "loss": 0.2848,
+      "step": 21985
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 8.702334416349278e-06,
+      "loss": 0.2866,
+      "step": 21990
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 8.67387207042194e-06,
+      "loss": 0.2862,
+      "step": 21995
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 8.645454235739903e-06,
+      "loss": 0.2867,
+      "step": 22000
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.046875,
+      "learning_rate": 8.617080926153698e-06,
+      "loss": 0.2856,
+      "step": 22005
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 8.588752155492119e-06,
+      "loss": 0.287,
+      "step": 22010
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 8.560467937562277e-06,
+      "loss": 0.2864,
+      "step": 22015
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 8.532228286149501e-06,
+      "loss": 0.2867,
+      "step": 22020
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 8.504033215017527e-06,
+      "loss": 0.2864,
+      "step": 22025
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 8.475882737908248e-06,
+      "loss": 0.285,
+      "step": 22030
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 8.447776868541879e-06,
+      "loss": 0.2858,
+      "step": 22035
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 8.419715620616874e-06,
+      "loss": 0.2875,
+      "step": 22040
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 8.39169900780995e-06,
+      "loss": 0.2868,
+      "step": 22045
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 8.363727043776038e-06,
+      "loss": 0.2865,
+      "step": 22050
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 8.335799742148387e-06,
+      "loss": 0.2865,
+      "step": 22055
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 8.307917116538378e-06,
+      "loss": 0.2861,
+      "step": 22060
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 8.280079180535672e-06,
+      "loss": 0.2863,
+      "step": 22065
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 8.252285947708139e-06,
+      "loss": 0.2868,
+      "step": 22070
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 8.224537431601886e-06,
+      "loss": 0.2867,
+      "step": 22075
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 8.196833645741186e-06,
+      "loss": 0.2849,
+      "step": 22080
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 8.169174603628538e-06,
+      "loss": 0.2871,
+      "step": 22085
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 8.1415603187446e-06,
+      "loss": 0.286,
+      "step": 22090
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 8.113990804548244e-06,
+      "loss": 0.2866,
+      "step": 22095
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 8.086466074476563e-06,
+      "loss": 0.2869,
+      "step": 22100
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 8.058986141944724e-06,
+      "loss": 0.287,
+      "step": 22105
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 8.031551020346128e-06,
+      "loss": 0.2852,
+      "step": 22110
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.048828125,
+      "learning_rate": 8.004160723052312e-06,
+      "loss": 0.2867,
+      "step": 22115
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 7.976815263412963e-06,
+      "loss": 0.2866,
+      "step": 22120
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 7.949514654755962e-06,
+      "loss": 0.2868,
+      "step": 22125
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 7.922258910387282e-06,
+      "loss": 0.2862,
+      "step": 22130
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0390625,
+      "learning_rate": 7.895048043591036e-06,
+      "loss": 0.2856,
+      "step": 22135
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 7.867882067629472e-06,
+      "loss": 0.2874,
+      "step": 22140
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 7.840760995742946e-06,
+      "loss": 0.2855,
+      "step": 22145
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 7.81368484114996e-06,
+      "loss": 0.2858,
+      "step": 22150
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 7.786653617047079e-06,
+      "loss": 0.2872,
+      "step": 22155
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.048828125,
+      "learning_rate": 7.75966733660901e-06,
+      "loss": 0.2876,
+      "step": 22160
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 7.73272601298851e-06,
+      "loss": 0.2867,
+      "step": 22165
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 7.7058296593165e-06,
+      "loss": 0.2862,
+      "step": 22170
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0625,
+      "learning_rate": 7.67897828870191e-06,
+      "loss": 0.2852,
+      "step": 22175
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 7.652171914231776e-06,
+      "loss": 0.2847,
+      "step": 22180
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 7.625410548971191e-06,
+      "loss": 0.2851,
+      "step": 22185
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.044921875,
+      "learning_rate": 7.59869420596333e-06,
+      "loss": 0.2869,
+      "step": 22190
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 7.572022898229403e-06,
+      "loss": 0.2858,
+      "step": 22195
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 7.545396638768698e-06,
+      "loss": 0.2854,
+      "step": 22200
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 7.518815440558513e-06,
+      "loss": 0.2864,
+      "step": 22205
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 7.492279316554207e-06,
+      "loss": 0.2872,
+      "step": 22210
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.05078125,
+      "learning_rate": 7.465788279689156e-06,
+      "loss": 0.2869,
+      "step": 22215
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 7.439342342874789e-06,
+      "loss": 0.2875,
+      "step": 22220
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 7.412941519000527e-06,
+      "loss": 0.2864,
+      "step": 22225
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 7.386585820933811e-06,
+      "loss": 0.2862,
+      "step": 22230
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 7.360275261520078e-06,
+      "loss": 0.2857,
+      "step": 22235
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.046875,
+      "learning_rate": 7.3340098535827905e-06,
+      "loss": 0.2859,
+      "step": 22240
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 7.3077896099233765e-06,
+      "loss": 0.2873,
+      "step": 22245
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.05078125,
+      "learning_rate": 7.281614543321269e-06,
+      "loss": 0.2876,
+      "step": 22250
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 7.255484666533874e-06,
+      "loss": 0.2862,
+      "step": 22255
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.046875,
+      "learning_rate": 7.2293999922965705e-06,
+      "loss": 0.2861,
+      "step": 22260
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 7.203360533322734e-06,
+      "loss": 0.2885,
+      "step": 22265
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 7.177366302303667e-06,
+      "loss": 0.2865,
+      "step": 22270
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 7.151417311908648e-06,
+      "loss": 0.2851,
+      "step": 22275
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 7.125513574784903e-06,
+      "loss": 0.2868,
+      "step": 22280
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 7.099655103557556e-06,
+      "loss": 0.2864,
+      "step": 22285
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 7.07384191082977e-06,
+      "loss": 0.2868,
+      "step": 22290
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 7.048074009182548e-06,
+      "loss": 0.2864,
+      "step": 22295
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 7.022351411174866e-06,
+      "loss": 0.2848,
+      "step": 22300
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 6.996674129343606e-06,
+      "loss": 0.2865,
+      "step": 22305
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 6.971042176203535e-06,
+      "loss": 0.2842,
+      "step": 22310
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 6.945455564247394e-06,
+      "loss": 0.2857,
+      "step": 22315
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 6.919914305945774e-06,
+      "loss": 0.2863,
+      "step": 22320
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 6.894418413747183e-06,
+      "loss": 0.2865,
+      "step": 22325
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.041015625,
+      "learning_rate": 6.868967900077972e-06,
+      "loss": 0.2853,
+      "step": 22330
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 6.8435627773424495e-06,
+      "loss": 0.2878,
+      "step": 22335
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 6.818203057922757e-06,
+      "loss": 0.2869,
+      "step": 22340
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 6.7928887541789055e-06,
+      "loss": 0.286,
+      "step": 22345
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 6.767619878448783e-06,
+      "loss": 0.2867,
+      "step": 22350
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 6.742396443048138e-06,
+      "loss": 0.2849,
+      "step": 22355
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 6.717218460270536e-06,
+      "loss": 0.2875,
+      "step": 22360
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 6.692085942387483e-06,
+      "loss": 0.2853,
+      "step": 22365
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.044921875,
+      "learning_rate": 6.666998901648203e-06,
+      "loss": 0.287,
+      "step": 22370
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.05078125,
+      "learning_rate": 6.6419573502798374e-06,
+      "loss": 0.2856,
+      "step": 22375
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 6.616961300487324e-06,
+      "loss": 0.2862,
+      "step": 22380
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 6.592010764453449e-06,
+      "loss": 0.2871,
+      "step": 22385
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 6.5671057543387985e-06,
+      "loss": 0.2866,
+      "step": 22390
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 6.542246282281772e-06,
+      "loss": 0.2856,
+      "step": 22395
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 6.517432360398556e-06,
+      "loss": 0.2873,
+      "step": 22400
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 6.492664000783166e-06,
+      "loss": 0.2858,
+      "step": 22405
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 6.467941215507433e-06,
+      "loss": 0.2856,
+      "step": 22410
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 6.443264016620887e-06,
+      "loss": 0.2864,
+      "step": 22415
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 6.418632416150927e-06,
+      "loss": 0.2861,
+      "step": 22420
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 6.394046426102674e-06,
+      "loss": 0.284,
+      "step": 22425
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.05859375,
+      "learning_rate": 6.3695060584590625e-06,
+      "loss": 0.2867,
+      "step": 22430
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 6.345011325180772e-06,
+      "loss": 0.2841,
+      "step": 22435
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 6.320562238206218e-06,
+      "loss": 0.2861,
+      "step": 22440
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 6.296158809451602e-06,
+      "loss": 0.2858,
+      "step": 22445
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0625,
+      "learning_rate": 6.2718010508108545e-06,
+      "loss": 0.2855,
+      "step": 22450
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 6.2474889741556575e-06,
+      "loss": 0.2856,
+      "step": 22455
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 6.223222591335409e-06,
+      "loss": 0.287,
+      "step": 22460
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.041015625,
+      "learning_rate": 6.1990019141772605e-06,
+      "loss": 0.2859,
+      "step": 22465
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 6.174826954486068e-06,
+      "loss": 0.2865,
+      "step": 22470
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 6.1506977240444074e-06,
+      "loss": 0.2861,
+      "step": 22475
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0546875,
+      "learning_rate": 6.126614234612593e-06,
+      "loss": 0.2856,
+      "step": 22480
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 6.102576497928614e-06,
+      "loss": 0.2865,
+      "step": 22485
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 6.078584525708176e-06,
+      "loss": 0.2855,
+      "step": 22490
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 6.054638329644657e-06,
+      "loss": 0.2852,
+      "step": 22495
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 6.030737921409169e-06,
+      "loss": 0.2857,
+      "step": 22500
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 6.006883312650457e-06,
+      "loss": 0.2874,
+      "step": 22505
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 5.98307451499498e-06,
+      "loss": 0.2887,
+      "step": 22510
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 5.9593115400468636e-06,
+      "loss": 0.286,
+      "step": 22515
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 5.935594399387856e-06,
+      "loss": 0.2846,
+      "step": 22520
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 5.911923104577455e-06,
+      "loss": 0.2867,
+      "step": 22525
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 5.888297667152731e-06,
+      "loss": 0.2846,
+      "step": 22530
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 5.864718098628441e-06,
+      "loss": 0.2858,
+      "step": 22535
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 5.8411844104969916e-06,
+      "loss": 0.2861,
+      "step": 22540
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 5.8176966142283965e-06,
+      "loss": 0.2877,
+      "step": 22545
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.036865234375,
+      "learning_rate": 5.7942547212703315e-06,
+      "loss": 0.286,
+      "step": 22550
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 5.770858743048091e-06,
+      "loss": 0.2885,
+      "step": 22555
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 5.747508690964598e-06,
+      "loss": 0.2884,
+      "step": 22560
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 5.724204576400371e-06,
+      "loss": 0.2855,
+      "step": 22565
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 5.700946410713548e-06,
+      "loss": 0.2873,
+      "step": 22570
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 5.6777342052399045e-06,
+      "loss": 0.2872,
+      "step": 22575
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 5.654567971292757e-06,
+      "loss": 0.2881,
+      "step": 22580
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 5.6314477201630745e-06,
+      "loss": 0.2863,
+      "step": 22585
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 5.608373463119354e-06,
+      "loss": 0.286,
+      "step": 22590
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 5.585345211407733e-06,
+      "loss": 0.2868,
+      "step": 22595
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 5.562362976251901e-06,
+      "loss": 0.2876,
+      "step": 22600
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 5.5394267688531066e-06,
+      "loss": 0.2851,
+      "step": 22605
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 5.516536600390188e-06,
+      "loss": 0.2849,
+      "step": 22610
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.056640625,
+      "learning_rate": 5.49369248201953e-06,
+      "loss": 0.2864,
+      "step": 22615
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 5.470894424875062e-06,
+      "loss": 0.2871,
+      "step": 22620
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 5.448142440068316e-06,
+      "loss": 0.2866,
+      "step": 22625
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 5.425436538688322e-06,
+      "loss": 0.2868,
+      "step": 22630
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 5.402776731801662e-06,
+      "loss": 0.2863,
+      "step": 22635
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0546875,
+      "learning_rate": 5.380163030452412e-06,
+      "loss": 0.287,
+      "step": 22640
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 5.357595445662267e-06,
+      "loss": 0.2876,
+      "step": 22645
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 5.335073988430372e-06,
+      "loss": 0.2852,
+      "step": 22650
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 5.312598669733404e-06,
+      "loss": 0.2858,
+      "step": 22655
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 5.290169500525577e-06,
+      "loss": 0.2864,
+      "step": 22660
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 5.267786491738569e-06,
+      "loss": 0.287,
+      "step": 22665
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 5.245449654281631e-06,
+      "loss": 0.2868,
+      "step": 22670
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 5.223158999041444e-06,
+      "loss": 0.286,
+      "step": 22675
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 5.200914536882185e-06,
+      "loss": 0.2867,
+      "step": 22680
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 5.178716278645535e-06,
+      "loss": 0.2841,
+      "step": 22685
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 5.1565642351506845e-06,
+      "loss": 0.2868,
+      "step": 22690
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 5.134458417194254e-06,
+      "loss": 0.285,
+      "step": 22695
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.04296875,
+      "learning_rate": 5.1123988355503475e-06,
+      "loss": 0.286,
+      "step": 22700
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 5.0903855009705514e-06,
+      "loss": 0.2873,
+      "step": 22705
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 5.068418424183874e-06,
+      "loss": 0.2857,
+      "step": 22710
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 5.046497615896806e-06,
+      "loss": 0.2857,
+      "step": 22715
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 5.024623086793323e-06,
+      "loss": 0.2872,
+      "step": 22720
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 5.002794847534764e-06,
+      "loss": 0.2865,
+      "step": 22725
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 4.981012908759941e-06,
+      "loss": 0.2863,
+      "step": 22730
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 4.959277281085129e-06,
+      "loss": 0.2864,
+      "step": 22735
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 4.937587975103996e-06,
+      "loss": 0.2881,
+      "step": 22740
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 4.915945001387667e-06,
+      "loss": 0.2865,
+      "step": 22745
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 4.8943483704846475e-06,
+      "loss": 0.2864,
+      "step": 22750
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 4.872798092920872e-06,
+      "loss": 0.2859,
+      "step": 22755
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 4.8512941791996726e-06,
+      "loss": 0.2858,
+      "step": 22760
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.04296875,
+      "learning_rate": 4.829836639801843e-06,
+      "loss": 0.2861,
+      "step": 22765
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 4.808425485185486e-06,
+      "loss": 0.2881,
+      "step": 22770
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.048828125,
+      "learning_rate": 4.7870607257861415e-06,
+      "loss": 0.2875,
+      "step": 22775
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 4.765742372016735e-06,
+      "loss": 0.2871,
+      "step": 22780
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 4.744470434267567e-06,
+      "loss": 0.286,
+      "step": 22785
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 4.723244922906356e-06,
+      "loss": 0.2883,
+      "step": 22790
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 4.702065848278126e-06,
+      "loss": 0.2872,
+      "step": 22795
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 4.680933220705308e-06,
+      "loss": 0.2865,
+      "step": 22800
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 4.659847050487687e-06,
+      "loss": 0.2839,
+      "step": 22805
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 4.638807347902408e-06,
+      "loss": 0.2863,
+      "step": 22810
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 4.617814123203967e-06,
+      "loss": 0.2884,
+      "step": 22815
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.044921875,
+      "learning_rate": 4.596867386624215e-06,
+      "loss": 0.2865,
+      "step": 22820
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 4.575967148372317e-06,
+      "loss": 0.2864,
+      "step": 22825
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 4.555113418634805e-06,
+      "loss": 0.2865,
+      "step": 22830
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 4.534306207575545e-06,
+      "loss": 0.2841,
+      "step": 22835
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.060546875,
+      "learning_rate": 4.513545525335705e-06,
+      "loss": 0.2856,
+      "step": 22840
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 4.492831382033791e-06,
+      "loss": 0.286,
+      "step": 22845
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 4.4721637877656375e-06,
+      "loss": 0.2869,
+      "step": 22850
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 4.451542752604365e-06,
+      "loss": 0.2873,
+      "step": 22855
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 4.4309682866004125e-06,
+      "loss": 0.2858,
+      "step": 22860
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 4.410440399781534e-06,
+      "loss": 0.2853,
+      "step": 22865
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 4.389959102152774e-06,
+      "loss": 0.2869,
+      "step": 22870
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 4.369524403696457e-06,
+      "loss": 0.2856,
+      "step": 22875
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 4.349136314372204e-06,
+      "loss": 0.2859,
+      "step": 22880
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 4.328794844116946e-06,
+      "loss": 0.2858,
+      "step": 22885
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 4.308500002844862e-06,
+      "loss": 0.2859,
+      "step": 22890
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 4.288251800447385e-06,
+      "loss": 0.2872,
+      "step": 22895
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 4.268050246793276e-06,
+      "loss": 0.2863,
+      "step": 22900
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 4.247895351728504e-06,
+      "loss": 0.2867,
+      "step": 22905
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 4.227787125076332e-06,
+      "loss": 0.2861,
+      "step": 22910
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 4.207725576637256e-06,
+      "loss": 0.2858,
+      "step": 22915
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 4.187710716189042e-06,
+      "loss": 0.2869,
+      "step": 22920
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 4.167742553486675e-06,
+      "loss": 0.2856,
+      "step": 22925
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.05078125,
+      "learning_rate": 4.147821098262405e-06,
+      "loss": 0.2877,
+      "step": 22930
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 4.127946360225721e-06,
+      "loss": 0.2858,
+      "step": 22935
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 4.108118349063306e-06,
+      "loss": 0.2869,
+      "step": 22940
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 4.088337074439097e-06,
+      "loss": 0.2849,
+      "step": 22945
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 4.068602545994249e-06,
+      "loss": 0.285,
+      "step": 22950
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 4.048914773347134e-06,
+      "loss": 0.2859,
+      "step": 22955
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 4.029273766093333e-06,
+      "loss": 0.2866,
+      "step": 22960
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 4.009679533805633e-06,
+      "loss": 0.2852,
+      "step": 22965
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 3.990132086034026e-06,
+      "loss": 0.2869,
+      "step": 22970
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.04296875,
+      "learning_rate": 3.970631432305694e-06,
+      "loss": 0.2861,
+      "step": 22975
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 3.951177582125021e-06,
+      "loss": 0.2863,
+      "step": 22980
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.056640625,
+      "learning_rate": 3.931770544973601e-06,
+      "loss": 0.2875,
+      "step": 22985
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 3.912410330310156e-06,
+      "loss": 0.2867,
+      "step": 22990
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 3.893096947570618e-06,
+      "loss": 0.2869,
+      "step": 22995
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 3.873830406168111e-06,
+      "loss": 0.287,
+      "step": 23000
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 3.8546107154929235e-06,
+      "loss": 0.2854,
+      "step": 23005
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 3.835437884912474e-06,
+      "loss": 0.2853,
+      "step": 23010
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 3.816311923771387e-06,
+      "loss": 0.2865,
+      "step": 23015
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 3.797232841391407e-06,
+      "loss": 0.286,
+      "step": 23020
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 3.7782006470714616e-06,
+      "loss": 0.287,
+      "step": 23025
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 3.759215350087619e-06,
+      "loss": 0.2869,
+      "step": 23030
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 3.7402769596930563e-06,
+      "loss": 0.2856,
+      "step": 23035
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 3.721385485118123e-06,
+      "loss": 0.287,
+      "step": 23040
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 3.7025409355702976e-06,
+      "loss": 0.2852,
+      "step": 23045
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 3.68374332023419e-06,
+      "loss": 0.2849,
+      "step": 23050
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 3.664992648271526e-06,
+      "loss": 0.2862,
+      "step": 23055
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 3.6462889288211512e-06,
+      "loss": 0.2853,
+      "step": 23060
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 3.627632170999029e-06,
+      "loss": 0.2867,
+      "step": 23065
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 3.6090223838982417e-06,
+      "loss": 0.2851,
+      "step": 23070
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.046875,
+      "learning_rate": 3.5904595765890005e-06,
+      "loss": 0.2866,
+      "step": 23075
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 3.5719437581185454e-06,
+      "loss": 0.286,
+      "step": 23080
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 3.553474937511281e-06,
+      "loss": 0.2865,
+      "step": 23085
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 3.5350531237686724e-06,
+      "loss": 0.2866,
+      "step": 23090
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 3.516678325869316e-06,
+      "loss": 0.2844,
+      "step": 23095
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 3.4983505527688586e-06,
+      "loss": 0.2841,
+      "step": 23100
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 3.480069813400022e-06,
+      "loss": 0.2862,
+      "step": 23105
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 3.461836116672612e-06,
+      "loss": 0.2862,
+      "step": 23110
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 3.443649471473531e-06,
+      "loss": 0.2871,
+      "step": 23115
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 3.425509886666711e-06,
+      "loss": 0.2878,
+      "step": 23120
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 3.40741737109318e-06,
+      "loss": 0.286,
+      "step": 23125
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 3.389371933570995e-06,
+      "loss": 0.2872,
+      "step": 23130
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.052734375,
+      "learning_rate": 3.3713735828952987e-06,
+      "loss": 0.2862,
+      "step": 23135
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 3.3534223278382405e-06,
+      "loss": 0.2866,
+      "step": 23140
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 3.3355181771490772e-06,
+      "loss": 0.2859,
+      "step": 23145
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.052734375,
+      "learning_rate": 3.3176611395540626e-06,
+      "loss": 0.2876,
+      "step": 23150
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 3.2998512237565005e-06,
+      "loss": 0.2861,
+      "step": 23155
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 3.2820884384367146e-06,
+      "loss": 0.2868,
+      "step": 23160
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 3.2643727922520906e-06,
+      "loss": 0.2864,
+      "step": 23165
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 3.246704293837011e-06,
+      "loss": 0.2844,
+      "step": 23170
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 3.2290829518028862e-06,
+      "loss": 0.2865,
+      "step": 23175
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 3.2115087747381366e-06,
+      "loss": 0.286,
+      "step": 23180
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.052734375,
+      "learning_rate": 3.19398177120821e-06,
+      "loss": 0.2873,
+      "step": 23185
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 3.1765019497555616e-06,
+      "loss": 0.2877,
+      "step": 23190
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 3.1590693188996323e-06,
+      "loss": 0.2877,
+      "step": 23195
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 3.1416838871368924e-06,
+      "loss": 0.2863,
+      "step": 23200
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 3.124345662940764e-06,
+      "loss": 0.286,
+      "step": 23205
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.060546875,
+      "learning_rate": 3.1070546547617095e-06,
+      "loss": 0.2867,
+      "step": 23210
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 3.0898108710271434e-06,
+      "loss": 0.2875,
+      "step": 23215
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 3.072614320141487e-06,
+      "loss": 0.286,
+      "step": 23220
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.052734375,
+      "learning_rate": 3.0554650104861136e-06,
+      "loss": 0.2865,
+      "step": 23225
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 3.0383629504194046e-06,
+      "loss": 0.2859,
+      "step": 23230
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 3.0213081482766805e-06,
+      "loss": 0.2872,
+      "step": 23235
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 3.0043006123702697e-06,
+      "loss": 0.2858,
+      "step": 23240
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.048828125,
+      "learning_rate": 2.9873403509894203e-06,
+      "loss": 0.2867,
+      "step": 23245
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 2.970427372400353e-06,
+      "loss": 0.288,
+      "step": 23250
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 2.953561684846262e-06,
+      "loss": 0.2866,
+      "step": 23255
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 2.936743296547273e-06,
+      "loss": 0.286,
+      "step": 23260
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 2.9199722157004616e-06,
+      "loss": 0.2867,
+      "step": 23265
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0390625,
+      "learning_rate": 2.9032484504798452e-06,
+      "loss": 0.2867,
+      "step": 23270
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 2.8865720090364034e-06,
+      "loss": 0.2864,
+      "step": 23275
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.044921875,
+      "learning_rate": 2.8699428994980017e-06,
+      "loss": 0.2854,
+      "step": 23280
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 2.8533611299694783e-06,
+      "loss": 0.2866,
+      "step": 23285
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 2.836826708532603e-06,
+      "loss": 0.2868,
+      "step": 23290
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 2.8203396432460506e-06,
+      "loss": 0.2866,
+      "step": 23295
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 2.8038999421453826e-06,
+      "loss": 0.2857,
+      "step": 23300
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 2.7875076132431344e-06,
+      "loss": 0.2875,
+      "step": 23305
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 2.771162664528726e-06,
+      "loss": 0.2861,
+      "step": 23310
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 2.7548651039684846e-06,
+      "loss": 0.2849,
+      "step": 23315
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 2.738614939505646e-06,
+      "loss": 0.2855,
+      "step": 23320
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 2.7224121790603517e-06,
+      "loss": 0.2848,
+      "step": 23325
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.037109375,
+      "learning_rate": 2.7062568305295965e-06,
+      "loss": 0.2853,
+      "step": 23330
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.046875,
+      "learning_rate": 2.690148901787337e-06,
+      "loss": 0.2866,
+      "step": 23335
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.041015625,
+      "learning_rate": 2.6740884006843825e-06,
+      "loss": 0.2862,
+      "step": 23340
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.048828125,
+      "learning_rate": 2.6580753350484046e-06,
+      "loss": 0.2866,
+      "step": 23345
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.052734375,
+      "learning_rate": 2.6421097126839712e-06,
+      "loss": 0.2869,
+      "step": 23350
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 2.6261915413725578e-06,
+      "loss": 0.2862,
+      "step": 23355
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 2.610320828872481e-06,
+      "loss": 0.2862,
+      "step": 23360
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0546875,
+      "learning_rate": 2.59449758291892e-06,
+      "loss": 0.2867,
+      "step": 23365
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 2.5787218112239496e-06,
+      "loss": 0.2868,
+      "step": 23370
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 2.5629935214764865e-06,
+      "loss": 0.2861,
+      "step": 23375
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 2.5473127213422763e-06,
+      "loss": 0.286,
+      "step": 23380
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 2.531679418464006e-06,
+      "loss": 0.2856,
+      "step": 23385
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.048828125,
+      "learning_rate": 2.516093620461124e-06,
+      "loss": 0.2859,
+      "step": 23390
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 2.5005553349299547e-06,
+      "loss": 0.2851,
+      "step": 23395
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 2.4850645694436736e-06,
+      "loss": 0.2879,
+      "step": 23400
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 2.4696213315523074e-06,
+      "loss": 0.2867,
+      "step": 23405
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 2.4542256287826914e-06,
+      "loss": 0.2868,
+      "step": 23410
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 2.4388774686385007e-06,
+      "loss": 0.2867,
+      "step": 23415
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.05859375,
+      "learning_rate": 2.4235768586002517e-06,
+      "loss": 0.2864,
+      "step": 23420
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 2.4083238061252567e-06,
+      "loss": 0.2882,
+      "step": 23425
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 2.3931183186477026e-06,
+      "loss": 0.2857,
+      "step": 23430
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 2.3779604035785273e-06,
+      "loss": 0.2867,
+      "step": 23435
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 2.3628500683055222e-06,
+      "loss": 0.2856,
+      "step": 23440
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 2.3477873201932734e-06,
+      "loss": 0.2858,
+      "step": 23445
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 2.332772166583208e-06,
+      "loss": 0.2867,
+      "step": 23450
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 2.3178046147935175e-06,
+      "loss": 0.2871,
+      "step": 23455
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 2.3028846721191876e-06,
+      "loss": 0.2866,
+      "step": 23460
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 2.288012345832047e-06,
+      "loss": 0.2883,
+      "step": 23465
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 2.273187643180652e-06,
+      "loss": 0.2872,
+      "step": 23470
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 2.2584105713904125e-06,
+      "loss": 0.2884,
+      "step": 23475
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 2.2436811376634893e-06,
+      "loss": 0.2855,
+      "step": 23480
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 2.2289993491788064e-06,
+      "loss": 0.2855,
+      "step": 23485
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 2.2143652130921176e-06,
+      "loss": 0.2861,
+      "step": 23490
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 2.1997787365358958e-06,
+      "loss": 0.2866,
+      "step": 23495
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 2.1852399266194314e-06,
+      "loss": 0.286,
+      "step": 23500
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 2.1707487904287672e-06,
+      "loss": 0.2869,
+      "step": 23505
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.048828125,
+      "learning_rate": 2.156305335026698e-06,
+      "loss": 0.2871,
+      "step": 23510
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 2.141909567452793e-06,
+      "loss": 0.2877,
+      "step": 23515
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 2.1275614947233624e-06,
+      "loss": 0.2864,
+      "step": 23520
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 2.1132611238315003e-06,
+      "loss": 0.2854,
+      "step": 23525
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 2.0990084617470206e-06,
+      "loss": 0.2857,
+      "step": 23530
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 2.084803515416511e-06,
+      "loss": 0.2871,
+      "step": 23535
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 2.0706462917632673e-06,
+      "loss": 0.2856,
+      "step": 23540
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 2.0565367976873584e-06,
+      "loss": 0.2873,
+      "step": 23545
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 2.0424750400655947e-06,
+      "loss": 0.2858,
+      "step": 23550
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 2.0284610257514937e-06,
+      "loss": 0.2867,
+      "step": 23555
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 2.014494761575314e-06,
+      "loss": 0.2875,
+      "step": 23560
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 2.0005762543440445e-06,
+      "loss": 0.2866,
+      "step": 23565
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.986705510841402e-06,
+      "loss": 0.2858,
+      "step": 23570
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 1.9728825378278246e-06,
+      "loss": 0.2849,
+      "step": 23575
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.9591073420404337e-06,
+      "loss": 0.2879,
+      "step": 23580
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.056640625,
+      "learning_rate": 1.945379930193125e-06,
+      "loss": 0.286,
+      "step": 23585
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 1.931700308976436e-06,
+      "loss": 0.2867,
+      "step": 23590
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.037109375,
+      "learning_rate": 1.918068485057689e-06,
+      "loss": 0.2849,
+      "step": 23595
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 1.904484465080847e-06,
+      "loss": 0.2864,
+      "step": 23600
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 1.8909482556666024e-06,
+      "loss": 0.286,
+      "step": 23605
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.8774598634123232e-06,
+      "loss": 0.2856,
+      "step": 23610
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.8640192948921054e-06,
+      "loss": 0.2869,
+      "step": 23615
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.8506265566567094e-06,
+      "loss": 0.2878,
+      "step": 23620
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 1.8372816552336026e-06,
+      "loss": 0.2886,
+      "step": 23625
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 1.8239845971269266e-06,
+      "loss": 0.2857,
+      "step": 23630
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.8107353888175083e-06,
+      "loss": 0.2874,
+      "step": 23635
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.7975340367628268e-06,
+      "loss": 0.285,
+      "step": 23640
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 1.7843805473970798e-06,
+      "loss": 0.2882,
+      "step": 23645
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 1.771274927131139e-06,
+      "loss": 0.2848,
+      "step": 23650
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 1.7582171823524951e-06,
+      "loss": 0.2857,
+      "step": 23655
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 1.7452073194253238e-06,
+      "loss": 0.2871,
+      "step": 23660
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.7322453446905085e-06,
+      "loss": 0.2888,
+      "step": 23665
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 1.719331264465529e-06,
+      "loss": 0.287,
+      "step": 23670
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.04296875,
+      "learning_rate": 1.706465085044584e-06,
+      "loss": 0.2848,
+      "step": 23675
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 1.6936468126984572e-06,
+      "loss": 0.2862,
+      "step": 23680
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.048828125,
+      "learning_rate": 1.680876453674629e-06,
+      "loss": 0.2868,
+      "step": 23685
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.048828125,
+      "learning_rate": 1.6681540141972429e-06,
+      "loss": 0.2848,
+      "step": 23690
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 1.6554795004670388e-06,
+      "loss": 0.2884,
+      "step": 23695
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.6428529186614195e-06,
+      "loss": 0.2868,
+      "step": 23700
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 1.6302742749344291e-06,
+      "loss": 0.2884,
+      "step": 23705
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 1.6177435754167415e-06,
+      "loss": 0.2875,
+      "step": 23710
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 1.605260826215682e-06,
+      "loss": 0.2874,
+      "step": 23715
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 1.5928260334151845e-06,
+      "loss": 0.2879,
+      "step": 23720
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.580439203075812e-06,
+      "loss": 0.2863,
+      "step": 23725
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.5681003412347572e-06,
+      "loss": 0.2862,
+      "step": 23730
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.052734375,
+      "learning_rate": 1.555809453905821e-06,
+      "loss": 0.2865,
+      "step": 23735
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 1.543566547079467e-06,
+      "loss": 0.2872,
+      "step": 23740
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 1.5313716267226996e-06,
+      "loss": 0.288,
+      "step": 23745
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.03759765625,
+      "learning_rate": 1.5192246987791981e-06,
+      "loss": 0.2867,
+      "step": 23750
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.5071257691692153e-06,
+      "loss": 0.2851,
+      "step": 23755
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.4950748437896234e-06,
+      "loss": 0.2877,
+      "step": 23760
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 1.4830719285139127e-06,
+      "loss": 0.2869,
+      "step": 23765
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.4711170291921484e-06,
+      "loss": 0.2867,
+      "step": 23770
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 1.4592101516509914e-06,
+      "loss": 0.2855,
+      "step": 23775
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 1.4473513016937223e-06,
+      "loss": 0.2859,
+      "step": 23780
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 1.4355404851001952e-06,
+      "loss": 0.2891,
+      "step": 23785
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.056640625,
+      "learning_rate": 1.423777707626872e-06,
+      "loss": 0.2855,
+      "step": 23790
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 1.412062975006767e-06,
+      "loss": 0.2861,
+      "step": 23795
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.052734375,
+      "learning_rate": 1.400396292949513e-06,
+      "loss": 0.286,
+      "step": 23800
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 1.3887776671412943e-06,
+      "loss": 0.287,
+      "step": 23805
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 1.377207103244904e-06,
+      "loss": 0.2877,
+      "step": 23810
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.3656846068996976e-06,
+      "loss": 0.2867,
+      "step": 23815
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.3542101837215826e-06,
+      "loss": 0.2849,
+      "step": 23820
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 1.3427838393030633e-06,
+      "loss": 0.2869,
+      "step": 23825
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 1.3314055792131964e-06,
+      "loss": 0.2862,
+      "step": 23830
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 1.320075408997612e-06,
+      "loss": 0.2887,
+      "step": 23835
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.308793334178493e-06,
+      "loss": 0.2859,
+      "step": 23840
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 1.2975593602545965e-06,
+      "loss": 0.2865,
+      "step": 23845
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.2863734927012095e-06,
+      "loss": 0.2854,
+      "step": 23850
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 1.275235736970193e-06,
+      "loss": 0.2848,
+      "step": 23855
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 1.26414609848996e-06,
+      "loss": 0.2854,
+      "step": 23860
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 1.2531045826654653e-06,
+      "loss": 0.2851,
+      "step": 23865
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 1.2421111948782149e-06,
+      "loss": 0.2871,
+      "step": 23870
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 1.231165940486234e-06,
+      "loss": 0.2859,
+      "step": 23875
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 1.2202688248241112e-06,
+      "loss": 0.2855,
+      "step": 23880
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 1.2094198532029755e-06,
+      "loss": 0.2881,
+      "step": 23885
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.1986190309104861e-06,
+      "loss": 0.2879,
+      "step": 23890
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 1.1878663632108322e-06,
+      "loss": 0.285,
+      "step": 23895
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.041015625,
+      "learning_rate": 1.1771618553447216e-06,
+      "loss": 0.2869,
+      "step": 23900
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 1.1665055125294033e-06,
+      "loss": 0.286,
+      "step": 23905
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 1.155897339958667e-06,
+      "loss": 0.2864,
+      "step": 23910
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05078125,
+      "learning_rate": 1.1453373428027992e-06,
+      "loss": 0.285,
+      "step": 23915
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05078125,
+      "learning_rate": 1.134825526208605e-06,
+      "loss": 0.288,
+      "step": 23920
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 1.1243618952994195e-06,
+      "loss": 0.2874,
+      "step": 23925
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 1.1139464551750856e-06,
+      "loss": 0.2886,
+      "step": 23930
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 1.103579210911976e-06,
+      "loss": 0.2856,
+      "step": 23935
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 1.0932601675629595e-06,
+      "loss": 0.2874,
+      "step": 23940
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.037353515625,
+      "learning_rate": 1.0829893301573913e-06,
+      "loss": 0.2871,
+      "step": 23945
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 1.0727667037011668e-06,
+      "loss": 0.2869,
+      "step": 23950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.0625922931766785e-06,
+      "loss": 0.2864,
+      "step": 23955
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 1.052466103542793e-06,
+      "loss": 0.2865,
+      "step": 23960
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 1.0423881397349068e-06,
+      "loss": 0.286,
+      "step": 23965
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.0323584066648795e-06,
+      "loss": 0.286,
+      "step": 23970
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 1.0223769092211012e-06,
+      "loss": 0.2853,
+      "step": 23975
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 1.0124436522684243e-06,
+      "loss": 0.2845,
+      "step": 23980
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.002558640648199e-06,
+      "loss": 0.2865,
+      "step": 23985
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 9.927218791782599e-07,
+      "loss": 0.2856,
+      "step": 23990
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 9.829333726529056e-07,
+      "loss": 0.2859,
+      "step": 23995
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 9.731931258429638e-07,
+      "loss": 0.2872,
+      "step": 24000
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 9.635011434957152e-07,
+      "loss": 0.2864,
+      "step": 24005
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 9.538574303348813e-07,
+      "loss": 0.2868,
+      "step": 24010
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 9.442619910607131e-07,
+      "loss": 0.2851,
+      "step": 24015
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.044921875,
+      "learning_rate": 9.347148303499142e-07,
+      "loss": 0.2851,
+      "step": 24020
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 9.252159528556403e-07,
+      "loss": 0.2861,
+      "step": 24025
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 9.157653632075436e-07,
+      "loss": 0.2862,
+      "step": 24030
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 9.063630660117173e-07,
+      "loss": 0.2848,
+      "step": 24035
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 8.970090658507291e-07,
+      "loss": 0.2872,
+      "step": 24040
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 8.877033672835988e-07,
+      "loss": 0.2883,
+      "step": 24045
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.041015625,
+      "learning_rate": 8.784459748458318e-07,
+      "loss": 0.2866,
+      "step": 24050
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 8.692368930493521e-07,
+      "loss": 0.2881,
+      "step": 24055
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 8.600761263825474e-07,
+      "loss": 0.2854,
+      "step": 24060
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 8.509636793102682e-07,
+      "loss": 0.2853,
+      "step": 24065
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 8.418995562738285e-07,
+      "loss": 0.2862,
+      "step": 24070
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 8.328837616909613e-07,
+      "loss": 0.2864,
+      "step": 24075
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 8.239162999558403e-07,
+      "loss": 0.2876,
+      "step": 24080
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0625,
+      "learning_rate": 8.14997175439125e-07,
+      "loss": 0.2888,
+      "step": 24085
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 8.061263924878604e-07,
+      "loss": 0.288,
+      "step": 24090
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 7.973039554255768e-07,
+      "loss": 0.2869,
+      "step": 24095
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 7.885298685522235e-07,
+      "loss": 0.2861,
+      "step": 24100
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 7.798041361441688e-07,
+      "loss": 0.2877,
+      "step": 24105
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 7.711267624542329e-07,
+      "loss": 0.2862,
+      "step": 24110
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 7.624977517116772e-07,
+      "loss": 0.2863,
+      "step": 24115
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 7.539171081221596e-07,
+      "loss": 0.2861,
+      "step": 24120
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 7.453848358678017e-07,
+      "loss": 0.2878,
+      "step": 24125
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 7.369009391070992e-07,
+      "loss": 0.2869,
+      "step": 24130
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 7.284654219750331e-07,
+      "loss": 0.2857,
+      "step": 24135
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 7.200782885829482e-07,
+      "loss": 0.2867,
+      "step": 24140
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 7.117395430186414e-07,
+      "loss": 0.2875,
+      "step": 24145
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 7.034491893463058e-07,
+      "loss": 0.2862,
+      "step": 24150
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 6.952072316065761e-07,
+      "loss": 0.2853,
+      "step": 24155
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 6.870136738164612e-07,
+      "loss": 0.2849,
+      "step": 24160
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 6.788685199694222e-07,
+      "loss": 0.2866,
+      "step": 24165
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 6.707717740353059e-07,
+      "loss": 0.2859,
+      "step": 24170
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.046875,
+      "learning_rate": 6.627234399603555e-07,
+      "loss": 0.2842,
+      "step": 24175
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 6.547235216672443e-07,
+      "loss": 0.2858,
+      "step": 24180
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.044921875,
+      "learning_rate": 6.4677202305502e-07,
+      "loss": 0.2851,
+      "step": 24185
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 6.388689479991605e-07,
+      "loss": 0.2852,
+      "step": 24190
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 6.310143003515179e-07,
+      "loss": 0.2867,
+      "step": 24195
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 6.232080839403631e-07,
+      "loss": 0.2865,
+      "step": 24200
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 6.154503025703417e-07,
+      "loss": 0.2873,
+      "step": 24205
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 6.07740960022507e-07,
+      "loss": 0.289,
+      "step": 24210
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 6.000800600542977e-07,
+      "loss": 0.2863,
+      "step": 24215
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 5.924676063995382e-07,
+      "loss": 0.2859,
+      "step": 24220
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 5.849036027684606e-07,
+      "loss": 0.2872,
+      "step": 24225
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 5.773880528476494e-07,
+      "loss": 0.2859,
+      "step": 24230
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 5.699209603001076e-07,
+      "loss": 0.2867,
+      "step": 24235
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 5.62502328765202e-07,
+      "loss": 0.2873,
+      "step": 24240
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 5.551321618586736e-07,
+      "loss": 0.2852,
+      "step": 24245
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 5.478104631726711e-07,
+      "loss": 0.2855,
+      "step": 24250
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 5.405372362756734e-07,
+      "loss": 0.287,
+      "step": 24255
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 5.333124847125892e-07,
+      "loss": 0.2874,
+      "step": 24260
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 5.261362120046686e-07,
+      "loss": 0.2868,
+      "step": 24265
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 5.190084216495361e-07,
+      "loss": 0.2858,
+      "step": 24270
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 5.119291171211793e-07,
+      "loss": 0.2842,
+      "step": 24275
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 5.048983018699827e-07,
+      "loss": 0.2869,
+      "step": 24280
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.048828125,
+      "learning_rate": 4.979159793226718e-07,
+      "loss": 0.2863,
+      "step": 24285
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.056640625,
+      "learning_rate": 4.909821528823577e-07,
+      "loss": 0.2852,
+      "step": 24290
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 4.840968259284817e-07,
+      "loss": 0.2859,
+      "step": 24295
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 4.772600018168816e-07,
+      "loss": 0.2851,
+      "step": 24300
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 4.704716838797363e-07,
+      "loss": 0.2849,
+      "step": 24305
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 4.6373187542561035e-07,
+      "loss": 0.2859,
+      "step": 24310
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 4.570405797393762e-07,
+      "loss": 0.2861,
+      "step": 24315
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 4.503978000823028e-07,
+      "loss": 0.2868,
+      "step": 24320
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.04296875,
+      "learning_rate": 4.438035396920004e-07,
+      "loss": 0.2868,
+      "step": 24325
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 4.372578017824314e-07,
+      "loss": 0.2872,
+      "step": 24330
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 4.307605895439104e-07,
+      "loss": 0.2865,
+      "step": 24335
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 4.2431190614309335e-07,
+      "loss": 0.2877,
+      "step": 24340
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05859375,
+      "learning_rate": 4.179117547229883e-07,
+      "loss": 0.286,
+      "step": 24345
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05078125,
+      "learning_rate": 4.115601384029666e-07,
+      "loss": 0.2862,
+      "step": 24350
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 4.0525706027870756e-07,
+      "loss": 0.286,
+      "step": 24355
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 3.9900252342228717e-07,
+      "loss": 0.2869,
+      "step": 24360
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 3.9279653088205584e-07,
+      "loss": 0.2865,
+      "step": 24365
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 3.866390856827495e-07,
+      "loss": 0.2857,
+      "step": 24370
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 3.805301908254455e-07,
+      "loss": 0.2859,
+      "step": 24375
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 3.744698492875398e-07,
+      "loss": 0.2856,
+      "step": 24380
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 3.6845806402275863e-07,
+      "loss": 0.2866,
+      "step": 24385
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 3.6249483796116924e-07,
+      "loss": 0.2857,
+      "step": 24390
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 3.565801740092023e-07,
+      "loss": 0.2852,
+      "step": 24395
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 3.50714075049563e-07,
+      "loss": 0.2861,
+      "step": 24400
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 3.4489654394134205e-07,
+      "loss": 0.2868,
+      "step": 24405
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 3.3912758351991593e-07,
+      "loss": 0.2865,
+      "step": 24410
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 3.3340719659701313e-07,
+      "loss": 0.2864,
+      "step": 24415
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 3.277353859606813e-07,
+      "loss": 0.286,
+      "step": 24420
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 3.2211215437528694e-07,
+      "loss": 0.2868,
+      "step": 24425
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 3.1653750458152666e-07,
+      "loss": 0.2851,
+      "step": 24430
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 3.1101143929641585e-07,
+      "loss": 0.2863,
+      "step": 24435
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 3.0553396121330013e-07,
+      "loss": 0.2872,
+      "step": 24440
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 3.001050730018218e-07,
+      "loss": 0.2864,
+      "step": 24445
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 2.947247773079753e-07,
+      "loss": 0.2867,
+      "step": 24450
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 2.893930767540298e-07,
+      "loss": 0.2863,
+      "step": 24455
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 2.841099739386066e-07,
+      "loss": 0.2868,
+      "step": 24460
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.048828125,
+      "learning_rate": 2.7887547143662373e-07,
+      "loss": 0.2848,
+      "step": 24465
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 2.73689571799296e-07,
+      "loss": 0.2862,
+      "step": 24470
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 2.685522775541904e-07,
+      "loss": 0.2865,
+      "step": 24475
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 2.6346359120514863e-07,
+      "loss": 0.2858,
+      "step": 24480
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 2.584235152323422e-07,
+      "loss": 0.287,
+      "step": 24485
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 2.534320520922506e-07,
+      "loss": 0.2881,
+      "step": 24490
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 2.484892042176279e-07,
+      "loss": 0.2873,
+      "step": 24495
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 2.4359497401758024e-07,
+      "loss": 0.2868,
+      "step": 24500
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 2.387493638774774e-07,
+      "loss": 0.2874,
+      "step": 24505
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 2.339523761590301e-07,
+      "loss": 0.285,
+      "step": 24510
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 2.2920401320022378e-07,
+      "loss": 0.2857,
+      "step": 24515
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.046875,
+      "learning_rate": 2.2450427731534053e-07,
+      "loss": 0.2865,
+      "step": 24520
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0546875,
+      "learning_rate": 2.1985317079500356e-07,
+      "loss": 0.2861,
+      "step": 24525
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 2.1525069590607737e-07,
+      "loss": 0.2867,
+      "step": 24530
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 2.106968548917676e-07,
+      "loss": 0.2869,
+      "step": 24535
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 2.0619164997155438e-07,
+      "loss": 0.2886,
+      "step": 24540
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.052734375,
+      "learning_rate": 2.017350833412146e-07,
+      "loss": 0.2856,
+      "step": 24545
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 1.973271571728441e-07,
+      "loss": 0.2866,
+      "step": 24550
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 1.9296787361480216e-07,
+      "loss": 0.2846,
+      "step": 24555
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.8865723479173368e-07,
+      "loss": 0.2868,
+      "step": 24560
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.8439524280462472e-07,
+      "loss": 0.2861,
+      "step": 24565
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 1.8018189973069143e-07,
+      "loss": 0.2869,
+      "step": 24570
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 1.7601720762346897e-07,
+      "loss": 0.2885,
+      "step": 24575
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 1.7190116851280026e-07,
+      "loss": 0.2856,
+      "step": 24580
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.056640625,
+      "learning_rate": 1.678337844047695e-07,
+      "loss": 0.289,
+      "step": 24585
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 1.6381505728176872e-07,
+      "loss": 0.2868,
+      "step": 24590
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.598449891024978e-07,
+      "loss": 0.2869,
+      "step": 24595
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.056640625,
+      "learning_rate": 1.5592358180189782e-07,
+      "loss": 0.2851,
+      "step": 24600
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 1.520508372912288e-07,
+      "loss": 0.2844,
+      "step": 24605
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 1.4822675745801429e-07,
+      "loss": 0.2864,
+      "step": 24610
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.4445134416607442e-07,
+      "loss": 0.2877,
+      "step": 24615
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 1.4072459925548177e-07,
+      "loss": 0.2859,
+      "step": 24620
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05078125,
+      "learning_rate": 1.3704652454261668e-07,
+      "loss": 0.286,
+      "step": 24625
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.037353515625,
+      "learning_rate": 1.33417121820123e-07,
+      "loss": 0.2855,
+      "step": 24630
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 1.2983639285693017e-07,
+      "loss": 0.2885,
+      "step": 24635
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 1.2630433939825327e-07,
+      "loss": 0.2873,
+      "step": 24640
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 1.2282096316554858e-07,
+      "loss": 0.2864,
+      "step": 24645
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 1.193862658566025e-07,
+      "loss": 0.2866,
+      "step": 24650
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 1.160002491454093e-07,
+      "loss": 0.2872,
+      "step": 24655
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.1266291468229328e-07,
+      "loss": 0.2872,
+      "step": 24660
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.0937426409384221e-07,
+      "loss": 0.2856,
+      "step": 24665
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 1.0613429898287398e-07,
+      "loss": 0.2875,
+      "step": 24670
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 1.0294302092853647e-07,
+      "loss": 0.2854,
+      "step": 24675
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 9.980043148619667e-08,
+      "loss": 0.2848,
+      "step": 24680
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 9.670653218752934e-08,
+      "loss": 0.2861,
+      "step": 24685
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 9.366132454046162e-08,
+      "loss": 0.2852,
+      "step": 24690
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 9.066481002918403e-08,
+      "loss": 0.2871,
+      "step": 24695
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 8.771699011416168e-08,
+      "loss": 0.2837,
+      "step": 24700
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 8.481786623214528e-08,
+      "loss": 0.2861,
+      "step": 24705
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 8.196743979610455e-08,
+      "loss": 0.2868,
+      "step": 24710
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 7.91657121953171e-08,
+      "loss": 0.2873,
+      "step": 24715
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 7.641268479531283e-08,
+      "loss": 0.2843,
+      "step": 24720
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 7.370835893788508e-08,
+      "loss": 0.2869,
+      "step": 24725
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 7.105273594107953e-08,
+      "loss": 0.2858,
+      "step": 24730
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05078125,
+      "learning_rate": 6.844581709921638e-08,
+      "loss": 0.2867,
+      "step": 24735
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 6.588760368287928e-08,
+      "loss": 0.2862,
+      "step": 24740
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 6.337809693891527e-08,
+      "loss": 0.2869,
+      "step": 24745
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.052734375,
+      "learning_rate": 6.09172980904238e-08,
+      "loss": 0.285,
+      "step": 24750
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 5.850520833676765e-08,
+      "loss": 0.285,
+      "step": 24755
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 5.6141828853573106e-08,
+      "loss": 0.2874,
+      "step": 24760
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 5.382716079271877e-08,
+      "loss": 0.2873,
+      "step": 24765
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 5.1561205282335547e-08,
+      "loss": 0.2875,
+      "step": 24770
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 4.934396342684e-08,
+      "loss": 0.2867,
+      "step": 24775
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 4.717543630688992e-08,
+      "loss": 0.2878,
+      "step": 24780
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 4.5055624979384316e-08,
+      "loss": 0.2887,
+      "step": 24785
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 4.298453047749673e-08,
+      "loss": 0.2867,
+      "step": 24790
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 4.096215381066415e-08,
+      "loss": 0.2869,
+      "step": 24795
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 3.898849596456478e-08,
+      "loss": 0.2864,
+      "step": 24800
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 3.7063557901129144e-08,
+      "loss": 0.2874,
+      "step": 24805
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.044921875,
+      "learning_rate": 3.518734055855122e-08,
+      "loss": 0.2864,
+      "step": 24810
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 3.33598448512773e-08,
+      "loss": 0.286,
+      "step": 24815
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 3.1581071670006015e-08,
+      "loss": 0.2888,
+      "step": 24820
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 2.985102188168831e-08,
+      "loss": 0.2868,
+      "step": 24825
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 2.8169696329527483e-08,
+      "loss": 0.2874,
+      "step": 24830
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 2.6537095832990245e-08,
+      "loss": 0.2867,
+      "step": 24835
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 2.4953221187784537e-08,
+      "loss": 0.288,
+      "step": 24840
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 2.3418073165870636e-08,
+      "loss": 0.286,
+      "step": 24845
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 2.193165251545004e-08,
+      "loss": 0.2847,
+      "step": 24850
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 2.0493959960998787e-08,
+      "loss": 0.2865,
+      "step": 24855
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 1.9104996203223037e-08,
+      "loss": 0.286,
+      "step": 24860
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 1.7764761919103477e-08,
+      "loss": 0.2878,
+      "step": 24865
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 1.6473257761828732e-08,
+      "loss": 0.2856,
+      "step": 24870
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 1.5230484360873044e-08,
+      "loss": 0.2873,
+      "step": 24875
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 1.4036442321962995e-08,
+      "loss": 0.285,
+      "step": 24880
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.2891132227033087e-08,
+      "loss": 0.2864,
+      "step": 24885
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 1.1794554634314558e-08,
+      "loss": 0.2885,
+      "step": 24890
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.074671007825767e-08,
+      "loss": 0.2846,
+      "step": 24895
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 9.747599069576119e-09,
+      "loss": 0.288,
+      "step": 24900
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 8.797222095224823e-09,
+      "loss": 0.2859,
+      "step": 24905
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 7.895579618388827e-09,
+      "loss": 0.2869,
+      "step": 24910
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.046875,
+      "learning_rate": 7.042672078527713e-09,
+      "loss": 0.2846,
+      "step": 24915
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 6.238499891353389e-09,
+      "loss": 0.2866,
+      "step": 24920
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.036865234375,
+      "learning_rate": 5.483063448785686e-09,
+      "loss": 0.2843,
+      "step": 24925
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 4.776363119030069e-09,
+      "loss": 0.2877,
+      "step": 24930
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 4.1183992465221315e-09,
+      "loss": 0.2858,
+      "step": 24935
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.052734375,
+      "learning_rate": 3.509172151938689e-09,
+      "loss": 0.2857,
+      "step": 24940
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 2.948682132208891e-09,
+      "loss": 0.2861,
+      "step": 24945
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 2.4369294605253166e-09,
+      "loss": 0.2863,
+      "step": 24950
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 1.973914386288467e-09,
+      "loss": 0.2856,
+      "step": 24955
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.5596371351733752e-09,
+      "loss": 0.2869,
+      "step": 24960
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 1.1940979091074056e-09,
+      "loss": 0.287,
+      "step": 24965
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 8.772968862369446e-10,
+      "loss": 0.2868,
+      "step": 24970
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 6.092342209607083e-10,
+      "loss": 0.2866,
+      "step": 24975
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 3.899100439408443e-10,
+      "loss": 0.2878,
+      "step": 24980
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 2.1932446206962553e-10,
+      "loss": 0.2859,
+      "step": 24985
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 9.74775584916543e-11,
+      "loss": 0.2862,
+      "step": 24990
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 2.4369392592760164e-11,
+      "loss": 0.2882,
+      "step": 24995
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.0,
+      "loss": 0.2862,
+      "step": 25000
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.29339778423309326,
+      "eval_runtime": 0.0984,
+      "eval_samples_per_second": 50.8,
+      "eval_steps_per_second": 20.32,
+      "step": 25000
+    },
+    {
+      "epoch": 1.0,
+      "step": 25000,
+      "total_flos": 9.808003579259126e+17,
+      "train_loss": 0.291647286939621,
+      "train_runtime": 13916.3013,
+      "train_samples_per_second": 28.743,
+      "train_steps_per_second": 1.796
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 25000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "total_flos": 9.808003579259126e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}