{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.072,
  "eval_steps": 500,
  "global_step": 4800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0032,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0001,
      "loss": 0.3678,
      "step": 5
    },
    {
      "epoch": 0.0064,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.0001,
      "loss": 0.2519,
      "step": 10
    },
    {
      "epoch": 0.0096,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.0001,
      "loss": 0.136,
      "step": 15
    },
    {
      "epoch": 0.0128,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0001,
      "loss": 0.08,
      "step": 20
    },
    {
      "epoch": 0.016,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0001,
      "loss": 0.0344,
      "step": 25
    },
    {
      "epoch": 0.0192,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.0782,
      "step": 30
    },
    {
      "epoch": 0.0224,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0001,
      "loss": 0.144,
      "step": 35
    },
    {
      "epoch": 0.0256,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0001,
      "loss": 0.1175,
      "step": 40
    },
    {
      "epoch": 0.0288,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.0001,
      "loss": 0.0657,
      "step": 45
    },
    {
      "epoch": 0.032,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0001,
      "loss": 0.0472,
      "step": 50
    },
    {
      "epoch": 0.0352,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0001,
      "loss": 0.0764,
      "step": 55
    },
    {
      "epoch": 0.0384,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.0001,
      "loss": 0.0299,
      "step": 60
    },
    {
      "epoch": 0.0416,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.0001,
      "loss": 0.0111,
      "step": 65
    },
    {
      "epoch": 0.0448,
      "grad_norm": 0.03125,
      "learning_rate": 0.0001,
      "loss": 0.0184,
      "step": 70
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.00201416015625,
      "learning_rate": 0.0001,
      "loss": 0.0033,
      "step": 75
    },
    {
      "epoch": 0.0512,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.0001,
      "loss": 0.0134,
      "step": 80
    },
    {
      "epoch": 0.0544,
      "grad_norm": 0.0018463134765625,
      "learning_rate": 0.0001,
      "loss": 0.0023,
      "step": 85
    },
    {
      "epoch": 0.0576,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.0001,
      "loss": 0.0021,
      "step": 90
    },
    {
      "epoch": 0.0608,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.0001,
      "loss": 0.0031,
      "step": 95
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.0001,
      "loss": 0.0032,
      "step": 100
    },
    {
      "epoch": 0.0672,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0001,
      "loss": 0.0457,
      "step": 105
    },
    {
      "epoch": 0.0704,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.0001,
      "loss": 0.0152,
      "step": 110
    },
    {
      "epoch": 0.0736,
      "grad_norm": 0.00982666015625,
      "learning_rate": 0.0001,
      "loss": 0.0095,
      "step": 115
    },
    {
      "epoch": 0.0768,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.0001,
      "loss": 0.0049,
      "step": 120
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.002166748046875,
      "learning_rate": 0.0001,
      "loss": 0.0032,
      "step": 125
    },
    {
      "epoch": 0.0832,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.0001,
      "loss": 0.0061,
      "step": 130
    },
    {
      "epoch": 0.0864,
      "grad_norm": 0.0078125,
      "learning_rate": 0.0001,
      "loss": 0.0011,
      "step": 135
    },
    {
      "epoch": 0.0896,
      "grad_norm": 0.00086212158203125,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 140
    },
    {
      "epoch": 0.0928,
      "grad_norm": 0.0006256103515625,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 145
    },
    {
      "epoch": 0.096,
      "grad_norm": 0.00164794921875,
      "learning_rate": 0.0001,
      "loss": 0.0054,
      "step": 150
    },
    {
      "epoch": 0.0992,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.0001,
      "loss": 0.0264,
      "step": 155
    },
    {
      "epoch": 0.1024,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.0001,
      "loss": 0.0108,
      "step": 160
    },
    {
      "epoch": 0.1056,
      "grad_norm": 0.007354736328125,
      "learning_rate": 0.0001,
      "loss": 0.0035,
      "step": 165
    },
    {
      "epoch": 0.1088,
      "grad_norm": 0.0103759765625,
      "learning_rate": 0.0001,
      "loss": 0.004,
      "step": 170
    },
    {
      "epoch": 0.112,
      "grad_norm": 0.0013580322265625,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 175
    },
    {
      "epoch": 0.1152,
      "grad_norm": 0.001434326171875,
      "learning_rate": 0.0001,
      "loss": 0.0015,
      "step": 180
    },
    {
      "epoch": 0.1184,
      "grad_norm": 0.00102996826171875,
      "learning_rate": 0.0001,
      "loss": 0.0104,
      "step": 185
    },
    {
      "epoch": 0.1216,
      "grad_norm": 0.001708984375,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 190
    },
    {
      "epoch": 0.1248,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 195
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.0014190673828125,
      "learning_rate": 0.0001,
      "loss": 0.001,
      "step": 200
    },
    {
      "epoch": 0.1312,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.0001,
      "loss": 0.0165,
      "step": 205
    },
    {
      "epoch": 0.1344,
      "grad_norm": 0.00836181640625,
      "learning_rate": 0.0001,
      "loss": 0.0063,
      "step": 210
    },
    {
      "epoch": 0.1376,
      "grad_norm": 0.0111083984375,
      "learning_rate": 0.0001,
      "loss": 0.0018,
      "step": 215
    },
    {
      "epoch": 0.1408,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0001,
      "loss": 0.0048,
      "step": 220
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.00061798095703125,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 225
    },
    {
      "epoch": 0.1472,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.0001,
      "loss": 0.0016,
      "step": 230
    },
    {
      "epoch": 0.1504,
      "grad_norm": 0.00077056884765625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 235
    },
    {
      "epoch": 0.1536,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.0001,
      "loss": 0.0015,
      "step": 240
    },
    {
      "epoch": 0.1568,
      "grad_norm": 0.00579833984375,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 245
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.00121307373046875,
      "learning_rate": 0.0001,
      "loss": 0.0022,
      "step": 250
    },
    {
      "epoch": 0.1632,
      "grad_norm": 0.0115966796875,
      "learning_rate": 0.0001,
      "loss": 0.0077,
      "step": 255
    },
    {
      "epoch": 0.1664,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.0001,
      "loss": 0.0033,
      "step": 260
    },
    {
      "epoch": 0.1696,
      "grad_norm": 0.000514984130859375,
      "learning_rate": 0.0001,
      "loss": 0.001,
      "step": 265
    },
    {
      "epoch": 0.1728,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.0001,
      "loss": 0.0025,
      "step": 270
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.00579833984375,
      "learning_rate": 0.0001,
      "loss": 0.0014,
      "step": 275
    },
    {
      "epoch": 0.1792,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.0001,
      "loss": 0.0014,
      "step": 280
    },
    {
      "epoch": 0.1824,
      "grad_norm": 0.000598907470703125,
      "learning_rate": 0.0001,
      "loss": 0.0007,
      "step": 285
    },
    {
      "epoch": 0.1856,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.0001,
      "loss": 0.0021,
      "step": 290
    },
    {
      "epoch": 0.1888,
      "grad_norm": 0.0008087158203125,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 295
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.0029296875,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 300
    },
    {
      "epoch": 0.1952,
      "grad_norm": 0.00750732421875,
      "learning_rate": 0.0001,
      "loss": 0.0038,
      "step": 305
    },
    {
      "epoch": 0.1984,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.0001,
      "loss": 0.0023,
      "step": 310
    },
    {
      "epoch": 0.2016,
      "grad_norm": 0.0128173828125,
      "learning_rate": 0.0001,
      "loss": 0.0025,
      "step": 315
    },
    {
      "epoch": 0.2048,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.0001,
      "loss": 0.0014,
      "step": 320
    },
    {
      "epoch": 0.208,
      "grad_norm": 0.0096435546875,
      "learning_rate": 0.0001,
      "loss": 0.0011,
      "step": 325
    },
    {
      "epoch": 0.2112,
      "grad_norm": 0.01458740234375,
      "learning_rate": 0.0001,
      "loss": 0.0015,
      "step": 330
    },
    {
      "epoch": 0.2144,
      "grad_norm": 0.004150390625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 335
    },
    {
      "epoch": 0.2176,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 340
    },
    {
      "epoch": 0.2208,
      "grad_norm": 0.0010986328125,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 345
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.001220703125,
      "learning_rate": 0.0001,
      "loss": 0.0007,
      "step": 350
    },
    {
      "epoch": 0.2272,
      "grad_norm": 0.01348876953125,
      "learning_rate": 0.0001,
      "loss": 0.0051,
      "step": 355
    },
    {
      "epoch": 0.2304,
      "grad_norm": 0.01025390625,
      "learning_rate": 0.0001,
      "loss": 0.0024,
      "step": 360
    },
    {
      "epoch": 0.2336,
      "grad_norm": 0.0037994384765625,
      "learning_rate": 0.0001,
      "loss": 0.001,
      "step": 365
    },
    {
      "epoch": 0.2368,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.0001,
      "loss": 0.0023,
      "step": 370
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.0001373291015625,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 375
    },
    {
      "epoch": 0.2432,
      "grad_norm": 0.006500244140625,
      "learning_rate": 0.0001,
      "loss": 0.0025,
      "step": 380
    },
    {
      "epoch": 0.2464,
      "grad_norm": 0.00020503997802734375,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 385
    },
    {
      "epoch": 0.2496,
      "grad_norm": 0.00022983551025390625,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 390
    },
    {
      "epoch": 0.2528,
      "grad_norm": 0.00018787384033203125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 395
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.000614166259765625,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 400
    },
    {
      "epoch": 0.2592,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.0001,
      "loss": 0.0102,
      "step": 405
    },
    {
      "epoch": 0.2624,
      "grad_norm": 0.00946044921875,
      "learning_rate": 0.0001,
      "loss": 0.0018,
      "step": 410
    },
    {
      "epoch": 0.2656,
      "grad_norm": 0.00098419189453125,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 415
    },
    {
      "epoch": 0.2688,
      "grad_norm": 0.01025390625,
      "learning_rate": 0.0001,
      "loss": 0.0012,
      "step": 420
    },
    {
      "epoch": 0.272,
      "grad_norm": 0.000278472900390625,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 425
    },
    {
      "epoch": 0.2752,
      "grad_norm": 0.006866455078125,
      "learning_rate": 0.0001,
      "loss": 0.0013,
      "step": 430
    },
    {
      "epoch": 0.2784,
      "grad_norm": 0.0003032684326171875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 435
    },
    {
      "epoch": 0.2816,
      "grad_norm": 0.01123046875,
      "learning_rate": 0.0001,
      "loss": 0.0072,
      "step": 440
    },
    {
      "epoch": 0.2848,
      "grad_norm": 0.00022602081298828125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 445
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.000621795654296875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 450
    },
    {
      "epoch": 0.2912,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.0001,
      "loss": 0.0156,
      "step": 455
    },
    {
      "epoch": 0.2944,
      "grad_norm": 0.0038604736328125,
      "learning_rate": 0.0001,
      "loss": 0.0016,
      "step": 460
    },
    {
      "epoch": 0.2976,
      "grad_norm": 0.00179290771484375,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 465
    },
    {
      "epoch": 0.3008,
      "grad_norm": 0.01519775390625,
      "learning_rate": 0.0001,
      "loss": 0.0008,
      "step": 470
    },
    {
      "epoch": 0.304,
      "grad_norm": 0.0004405975341796875,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 475
    },
    {
      "epoch": 0.3072,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 480
    },
    {
      "epoch": 0.3104,
      "grad_norm": 0.000926971435546875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 485
    },
    {
      "epoch": 0.3136,
      "grad_norm": 0.0003108978271484375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 490
    },
    {
      "epoch": 0.3168,
      "grad_norm": 0.00020503997802734375,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 495
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.0005950927734375,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 500
    },
    {
      "epoch": 0.3232,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.0001,
      "loss": 0.0051,
      "step": 505
    },
    {
      "epoch": 0.3264,
      "grad_norm": 0.011962890625,
      "learning_rate": 0.0001,
      "loss": 0.0038,
      "step": 510
    },
    {
      "epoch": 0.3296,
      "grad_norm": 0.014404296875,
      "learning_rate": 0.0001,
      "loss": 0.0018,
      "step": 515
    },
    {
      "epoch": 0.3328,
      "grad_norm": 0.0038299560546875,
      "learning_rate": 0.0001,
      "loss": 0.0009,
      "step": 520
    },
    {
      "epoch": 0.336,
      "grad_norm": 0.0002880096435546875,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 525
    },
    {
      "epoch": 0.3392,
      "grad_norm": 0.007049560546875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 530
    },
    {
      "epoch": 0.3424,
      "grad_norm": 0.01397705078125,
      "learning_rate": 0.0001,
      "loss": 0.0008,
      "step": 535
    },
    {
      "epoch": 0.3456,
      "grad_norm": 0.00147247314453125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 540
    },
    {
      "epoch": 0.3488,
      "grad_norm": 0.00238037109375,
      "learning_rate": 0.0001,
      "loss": 0.0064,
      "step": 545
    },
    {
      "epoch": 0.352,
      "grad_norm": 0.001953125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 550
    },
    {
      "epoch": 0.3552,
      "grad_norm": 0.0047607421875,
      "learning_rate": 0.0001,
      "loss": 0.0067,
      "step": 555
    },
    {
      "epoch": 0.3584,
      "grad_norm": 0.0040283203125,
      "learning_rate": 0.0001,
      "loss": 0.0008,
      "step": 560
    },
    {
      "epoch": 0.3616,
      "grad_norm": 0.0001983642578125,
      "learning_rate": 0.0001,
      "loss": 0.0016,
      "step": 565
    },
    {
      "epoch": 0.3648,
      "grad_norm": 0.0072021484375,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 570
    },
    {
      "epoch": 0.368,
      "grad_norm": 6.246566772460938e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 575
    },
    {
      "epoch": 0.3712,
      "grad_norm": 0.00811767578125,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 580
    },
    {
      "epoch": 0.3744,
      "grad_norm": 0.00020694732666015625,
      "learning_rate": 0.0001,
      "loss": 0.0006,
      "step": 585
    },
    {
      "epoch": 0.3776,
      "grad_norm": 0.0098876953125,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 590
    },
    {
      "epoch": 0.3808,
      "grad_norm": 0.00012874603271484375,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 595
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.0003490447998046875,
      "learning_rate": 0.0001,
      "loss": 0.0025,
      "step": 600
    },
    {
      "epoch": 0.3872,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.0001,
      "loss": 0.0029,
      "step": 605
    },
    {
      "epoch": 0.3904,
      "grad_norm": 0.010986328125,
      "learning_rate": 0.0001,
      "loss": 0.0032,
      "step": 610
    },
    {
      "epoch": 0.3936,
      "grad_norm": 0.000507354736328125,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 615
    },
    {
      "epoch": 0.3968,
      "grad_norm": 0.0081787109375,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 620
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.00032806396484375,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 625
    },
    {
      "epoch": 0.4032,
      "grad_norm": 0.0032806396484375,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 630
    },
    {
      "epoch": 0.4064,
      "grad_norm": 0.000125885009765625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 635
    },
    {
      "epoch": 0.4096,
      "grad_norm": 0.0002956390380859375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 640
    },
    {
      "epoch": 0.4128,
      "grad_norm": 0.00010919570922851562,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 645
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.0001983642578125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 650
    },
    {
      "epoch": 0.4192,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.0001,
      "loss": 0.0037,
      "step": 655
    },
    {
      "epoch": 0.4224,
      "grad_norm": 0.01007080078125,
      "learning_rate": 0.0001,
      "loss": 0.0009,
      "step": 660
    },
    {
      "epoch": 0.4256,
      "grad_norm": 0.000640869140625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 665
    },
    {
      "epoch": 0.4288,
      "grad_norm": 0.009033203125,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 670
    },
    {
      "epoch": 0.432,
      "grad_norm": 6.580352783203125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0022,
      "step": 675
    },
    {
      "epoch": 0.4352,
      "grad_norm": 0.004730224609375,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 680
    },
    {
      "epoch": 0.4384,
      "grad_norm": 0.0024871826171875,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 685
    },
    {
      "epoch": 0.4416,
      "grad_norm": 0.00958251953125,
      "learning_rate": 0.0001,
      "loss": 0.0051,
      "step": 690
    },
    {
      "epoch": 0.4448,
      "grad_norm": 0.00011110305786132812,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 695
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.00019931793212890625,
      "learning_rate": 0.0001,
      "loss": 0.0009,
      "step": 700
    },
    {
      "epoch": 0.4512,
      "grad_norm": 0.00946044921875,
      "learning_rate": 0.0001,
      "loss": 0.0031,
      "step": 705
    },
    {
      "epoch": 0.4544,
      "grad_norm": 0.00885009765625,
      "learning_rate": 0.0001,
      "loss": 0.0065,
      "step": 710
    },
    {
      "epoch": 0.4576,
      "grad_norm": 0.0101318359375,
      "learning_rate": 0.0001,
      "loss": 0.0009,
      "step": 715
    },
    {
      "epoch": 0.4608,
      "grad_norm": 0.00579833984375,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 720
    },
    {
      "epoch": 0.464,
      "grad_norm": 4.649162292480469e-05,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 725
    },
    {
      "epoch": 0.4672,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 730
    },
    {
      "epoch": 0.4704,
      "grad_norm": 0.00010061264038085938,
      "learning_rate": 0.0001,
      "loss": 0.0104,
      "step": 735
    },
    {
      "epoch": 0.4736,
      "grad_norm": 0.0101318359375,
      "learning_rate": 0.0001,
      "loss": 0.0008,
      "step": 740
    },
    {
      "epoch": 0.4768,
      "grad_norm": 0.00017452239990234375,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 745
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.0003948211669921875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 750
    },
    {
      "epoch": 0.4832,
      "grad_norm": 0.0024566650390625,
      "learning_rate": 0.0001,
      "loss": 0.0016,
      "step": 755
    },
    {
      "epoch": 0.4864,
      "grad_norm": 0.00628662109375,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 760
    },
    {
      "epoch": 0.4896,
      "grad_norm": 0.000637054443359375,
      "learning_rate": 0.0001,
      "loss": 0.0012,
      "step": 765
    },
    {
      "epoch": 0.4928,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0001,
      "loss": 0.0007,
      "step": 770
    },
    {
      "epoch": 0.496,
      "grad_norm": 3.743171691894531e-05,
      "learning_rate": 0.0001,
      "loss": 0.0016,
      "step": 775
    },
    {
      "epoch": 0.4992,
      "grad_norm": 0.0018310546875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 780
    },
    {
      "epoch": 0.5024,
      "grad_norm": 0.00066375732421875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 785
    },
    {
      "epoch": 0.5056,
      "grad_norm": 0.00020503997802734375,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 790
    },
    {
      "epoch": 0.5088,
      "grad_norm": 0.0002899169921875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 795
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.00012159347534179688,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 800
    },
    {
      "epoch": 0.5152,
      "grad_norm": 0.00151824951171875,
      "learning_rate": 0.0001,
      "loss": 0.0013,
      "step": 805
    },
    {
      "epoch": 0.5184,
      "grad_norm": 0.01953125,
      "learning_rate": 0.0001,
      "loss": 0.0023,
      "step": 810
    },
    {
      "epoch": 0.5216,
      "grad_norm": 0.0002727508544921875,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 815
    },
    {
      "epoch": 0.5248,
      "grad_norm": 0.00087738037109375,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 820
    },
    {
      "epoch": 0.528,
      "grad_norm": 4.100799560546875e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 825
    },
    {
      "epoch": 0.5312,
      "grad_norm": 0.00110626220703125,
      "learning_rate": 0.0001,
      "loss": 0.0007,
      "step": 830
    },
    {
      "epoch": 0.5344,
      "grad_norm": 0.000926971435546875,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 835
    },
    {
      "epoch": 0.5376,
      "grad_norm": 0.01214599609375,
      "learning_rate": 0.0001,
      "loss": 0.0059,
      "step": 840
    },
    {
      "epoch": 0.5408,
      "grad_norm": 0.000423431396484375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 845
    },
    {
      "epoch": 0.544,
      "grad_norm": 0.00445556640625,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 850
    },
    {
      "epoch": 0.5472,
      "grad_norm": 0.00958251953125,
      "learning_rate": 0.0001,
      "loss": 0.0028,
      "step": 855
    },
    {
      "epoch": 0.5504,
      "grad_norm": 0.0155029296875,
      "learning_rate": 0.0001,
      "loss": 0.0017,
      "step": 860
    },
    {
      "epoch": 0.5536,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.0001,
      "loss": 0.0017,
      "step": 865
    },
    {
      "epoch": 0.5568,
      "grad_norm": 0.0128173828125,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 870
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.000263214111328125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 875
    },
    {
      "epoch": 0.5632,
      "grad_norm": 0.0016937255859375,
      "learning_rate": 0.0001,
      "loss": 0.001,
      "step": 880
    },
    {
      "epoch": 0.5664,
      "grad_norm": 0.0004329681396484375,
      "learning_rate": 0.0001,
      "loss": 0.0009,
      "step": 885
    },
    {
      "epoch": 0.5696,
      "grad_norm": 8.726119995117188e-05,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 890
    },
    {
      "epoch": 0.5728,
      "grad_norm": 0.0001392364501953125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 895
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.0032806396484375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 900
    },
    {
      "epoch": 0.5792,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.0001,
      "loss": 0.0023,
      "step": 905
    },
    {
      "epoch": 0.5824,
      "grad_norm": 0.0130615234375,
      "learning_rate": 0.0001,
      "loss": 0.0029,
      "step": 910
    },
    {
      "epoch": 0.5856,
      "grad_norm": 0.00136566162109375,
      "learning_rate": 0.0001,
      "loss": 0.001,
      "step": 915
    },
    {
      "epoch": 0.5888,
      "grad_norm": 0.0028076171875,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 920
    },
    {
      "epoch": 0.592,
      "grad_norm": 0.000263214111328125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 925
    },
    {
      "epoch": 0.5952,
      "grad_norm": 0.004302978515625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 930
    },
    {
      "epoch": 0.5984,
      "grad_norm": 0.0001239776611328125,
      "learning_rate": 0.0001,
      "loss": 0.0006,
      "step": 935
    },
    {
      "epoch": 0.6016,
      "grad_norm": 8.821487426757812e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 940
    },
    {
      "epoch": 0.6048,
      "grad_norm": 9.822845458984375e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 945
    },
    {
      "epoch": 0.608,
      "grad_norm": 0.00013446807861328125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 950
    },
    {
      "epoch": 0.6112,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.0001,
      "loss": 0.0028,
      "step": 955
    },
    {
      "epoch": 0.6144,
      "grad_norm": 0.0020599365234375,
      "learning_rate": 0.0001,
      "loss": 0.0009,
      "step": 960
    },
    {
      "epoch": 0.6176,
      "grad_norm": 0.00193023681640625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 965
    },
    {
      "epoch": 0.6208,
      "grad_norm": 0.00020885467529296875,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 970
    },
    {
      "epoch": 0.624,
      "grad_norm": 3.600120544433594e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 975
    },
    {
      "epoch": 0.6272,
      "grad_norm": 0.000385284423828125,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 980
    },
    {
      "epoch": 0.6304,
      "grad_norm": 0.01385498046875,
      "learning_rate": 0.0001,
      "loss": 0.0006,
      "step": 985
    },
    {
      "epoch": 0.6336,
      "grad_norm": 0.00010728836059570312,
      "learning_rate": 0.0001,
      "loss": 0.0045,
      "step": 990
    },
    {
      "epoch": 0.6368,
      "grad_norm": 7.2479248046875e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 995
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.00011730194091796875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1000
    },
    {
      "epoch": 0.6432,
      "grad_norm": 0.0142822265625,
      "learning_rate": 0.0001,
      "loss": 0.0033,
      "step": 1005
    },
    {
      "epoch": 0.6464,
      "grad_norm": 0.0048828125,
      "learning_rate": 0.0001,
      "loss": 0.0011,
      "step": 1010
    },
    {
      "epoch": 0.6496,
      "grad_norm": 0.0040283203125,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 1015
    },
    {
      "epoch": 0.6528,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.0001,
      "loss": 0.0015,
      "step": 1020
    },
    {
      "epoch": 0.656,
      "grad_norm": 0.00604248046875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1025
    },
    {
      "epoch": 0.6592,
      "grad_norm": 0.00054168701171875,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 1030
    },
    {
      "epoch": 0.6624,
      "grad_norm": 0.001129150390625,
      "learning_rate": 0.0001,
      "loss": 0.0096,
      "step": 1035
    },
    {
      "epoch": 0.6656,
      "grad_norm": 0.014404296875,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 1040
    },
    {
      "epoch": 0.6688,
      "grad_norm": 0.002716064453125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1045
    },
    {
      "epoch": 0.672,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1050
    },
    {
      "epoch": 0.6752,
      "grad_norm": 0.005126953125,
      "learning_rate": 0.0001,
      "loss": 0.0017,
      "step": 1055
    },
    {
      "epoch": 0.6784,
      "grad_norm": 0.00110626220703125,
      "learning_rate": 0.0001,
      "loss": 0.0006,
      "step": 1060
    },
    {
      "epoch": 0.6816,
      "grad_norm": 7.867813110351562e-05,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 1065
    },
    {
      "epoch": 0.6848,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.0001,
      "loss": 0.0011,
      "step": 1070
    },
    {
      "epoch": 0.688,
      "grad_norm": 0.0002841949462890625,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1075
    },
    {
      "epoch": 0.6912,
      "grad_norm": 0.0002117156982421875,
      "learning_rate": 0.0001,
      "loss": 0.0007,
      "step": 1080
    },
    {
      "epoch": 0.6944,
      "grad_norm": 0.00010156631469726562,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1085
    },
    {
      "epoch": 0.6976,
      "grad_norm": 0.0006256103515625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 1090
    },
    {
      "epoch": 0.7008,
      "grad_norm": 9.870529174804688e-05,
      "learning_rate": 0.0001,
      "loss": 0.0009,
      "step": 1095
    },
    {
      "epoch": 0.704,
      "grad_norm": 0.000858306884765625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 1100
    },
    {
      "epoch": 0.7072,
      "grad_norm": 0.0023040771484375,
      "learning_rate": 0.0001,
      "loss": 0.0012,
      "step": 1105
    },
    {
      "epoch": 0.7104,
      "grad_norm": 0.013916015625,
      "learning_rate": 0.0001,
      "loss": 0.0009,
      "step": 1110
    },
    {
      "epoch": 0.7136,
      "grad_norm": 0.000110626220703125,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 1115
    },
    {
      "epoch": 0.7168,
      "grad_norm": 0.0002536773681640625,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1120
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.0001163482666015625,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1125
    },
    {
      "epoch": 0.7232,
      "grad_norm": 0.00014495849609375,
      "learning_rate": 0.0001,
      "loss": 0.0091,
      "step": 1130
    },
    {
      "epoch": 0.7264,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 1135
    },
    {
      "epoch": 0.7296,
      "grad_norm": 9.5367431640625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1140
    },
    {
      "epoch": 0.7328,
      "grad_norm": 8.249282836914062e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1145
    },
    {
      "epoch": 0.736,
      "grad_norm": 0.001007080078125,
      "learning_rate": 0.0001,
      "loss": 0.0011,
      "step": 1150
    },
    {
      "epoch": 0.7392,
      "grad_norm": 0.01226806640625,
      "learning_rate": 0.0001,
      "loss": 0.0019,
      "step": 1155
    },
    {
      "epoch": 0.7424,
      "grad_norm": 0.005828857421875,
      "learning_rate": 0.0001,
      "loss": 0.0008,
      "step": 1160
    },
    {
      "epoch": 0.7456,
      "grad_norm": 5.6743621826171875e-05,
      "learning_rate": 0.0001,
      "loss": 0.003,
      "step": 1165
    },
    {
      "epoch": 0.7488,
      "grad_norm": 0.0002899169921875,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 1170
    },
    {
      "epoch": 0.752,
      "grad_norm": 0.0003147125244140625,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1175
    },
    {
      "epoch": 0.7552,
      "grad_norm": 0.015625,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 1180
    },
    {
      "epoch": 0.7584,
      "grad_norm": 0.000274658203125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1185
    },
    {
      "epoch": 0.7616,
      "grad_norm": 0.00093841552734375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1190
    },
    {
      "epoch": 0.7648,
      "grad_norm": 0.000335693359375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1195
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.0003681182861328125,
      "learning_rate": 0.0001,
      "loss": 0.0007,
      "step": 1200
    },
    {
      "epoch": 0.7712,
      "grad_norm": 0.0115966796875,
      "learning_rate": 0.0001,
      "loss": 0.0011,
      "step": 1205
    },
    {
      "epoch": 0.7744,
      "grad_norm": 0.00811767578125,
      "learning_rate": 0.0001,
      "loss": 0.0008,
      "step": 1210
    },
    {
      "epoch": 0.7776,
      "grad_norm": 0.00017547607421875,
      "learning_rate": 0.0001,
      "loss": 0.0016,
      "step": 1215
    },
    {
      "epoch": 0.7808,
      "grad_norm": 0.000782012939453125,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 1220
    },
    {
      "epoch": 0.784,
      "grad_norm": 6.198883056640625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 1225
    },
    {
      "epoch": 0.7872,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 1230
    },
    {
      "epoch": 0.7904,
      "grad_norm": 0.00180816650390625,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 1235
    },
    {
      "epoch": 0.7936,
      "grad_norm": 0.015380859375,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 1240
    },
    {
      "epoch": 0.7968,
      "grad_norm": 0.001373291015625,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1245
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.003997802734375,
      "learning_rate": 0.0001,
      "loss": 0.0006,
      "step": 1250
    },
    {
      "epoch": 0.8032,
      "grad_norm": 0.01226806640625,
      "learning_rate": 0.0001,
      "loss": 0.0087,
      "step": 1255
    },
    {
      "epoch": 0.8064,
      "grad_norm": 0.00616455078125,
      "learning_rate": 0.0001,
      "loss": 0.0017,
      "step": 1260
    },
    {
      "epoch": 0.8096,
      "grad_norm": 0.00077056884765625,
      "learning_rate": 0.0001,
      "loss": 0.0006,
      "step": 1265
    },
    {
      "epoch": 0.8128,
      "grad_norm": 0.00128936767578125,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 1270
    },
    {
      "epoch": 0.816,
      "grad_norm": 9.679794311523438e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1275
    },
    {
      "epoch": 0.8192,
      "grad_norm": 0.0008087158203125,
      "learning_rate": 0.0001,
      "loss": 0.0015,
      "step": 1280
    },
    {
      "epoch": 0.8224,
      "grad_norm": 0.00019359588623046875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1285
    },
    {
      "epoch": 0.8256,
      "grad_norm": 0.000812530517578125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1290
    },
    {
      "epoch": 0.8288,
      "grad_norm": 0.0006256103515625,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1295
    },
    {
      "epoch": 0.832,
      "grad_norm": 0.00067901611328125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1300
    },
    {
      "epoch": 0.8352,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.0001,
      "loss": 0.0045,
      "step": 1305
    },
    {
      "epoch": 0.8384,
      "grad_norm": 0.000347137451171875,
      "learning_rate": 0.0001,
      "loss": 0.0007,
      "step": 1310
    },
    {
      "epoch": 0.8416,
      "grad_norm": 0.00016117095947265625,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 1315
    },
    {
      "epoch": 0.8448,
      "grad_norm": 0.0023956298828125,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 1320
    },
    {
      "epoch": 0.848,
      "grad_norm": 4.6253204345703125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1325
    },
    {
      "epoch": 0.8512,
      "grad_norm": 0.00543212890625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 1330
    },
    {
      "epoch": 0.8544,
      "grad_norm": 7.009506225585938e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1335
    },
    {
      "epoch": 0.8576,
      "grad_norm": 0.00101470947265625,
      "learning_rate": 0.0001,
      "loss": 0.0048,
      "step": 1340
    },
    {
      "epoch": 0.8608,
      "grad_norm": 0.00011491775512695312,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1345
    },
    {
      "epoch": 0.864,
      "grad_norm": 0.0004730224609375,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 1350
    },
    {
      "epoch": 0.8672,
      "grad_norm": 0.00885009765625,
      "learning_rate": 0.0001,
      "loss": 0.0045,
      "step": 1355
    },
    {
      "epoch": 0.8704,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.0001,
      "loss": 0.0035,
      "step": 1360
    },
    {
      "epoch": 0.8736,
      "grad_norm": 0.0002651214599609375,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1365
    },
    {
      "epoch": 0.8768,
      "grad_norm": 0.002777099609375,
      "learning_rate": 0.0001,
      "loss": 0.0008,
      "step": 1370
    },
    {
      "epoch": 0.88,
      "grad_norm": 7.390975952148438e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1375
    },
    {
      "epoch": 0.8832,
      "grad_norm": 0.00023937225341796875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1380
    },
    {
      "epoch": 0.8864,
      "grad_norm": 0.01202392578125,
      "learning_rate": 0.0001,
      "loss": 0.0009,
      "step": 1385
    },
    {
      "epoch": 0.8896,
      "grad_norm": 0.00543212890625,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1390
    },
    {
      "epoch": 0.8928,
      "grad_norm": 8.487701416015625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0047,
      "step": 1395
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.007110595703125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1400
    },
    {
      "epoch": 0.8992,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.0001,
      "loss": 0.0038,
      "step": 1405
    },
    {
      "epoch": 0.9024,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.0001,
      "loss": 0.0007,
      "step": 1410
    },
    {
      "epoch": 0.9056,
      "grad_norm": 9.918212890625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0006,
      "step": 1415
    },
    {
      "epoch": 0.9088,
      "grad_norm": 0.0013580322265625,
      "learning_rate": 0.0001,
      "loss": 0.0008,
      "step": 1420
    },
    {
      "epoch": 0.912,
      "grad_norm": 0.000720977783203125,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 1425
    },
    {
      "epoch": 0.9152,
      "grad_norm": 0.0030670166015625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 1430
    },
    {
      "epoch": 0.9184,
      "grad_norm": 0.00010013580322265625,
      "learning_rate": 0.0001,
      "loss": 0.0088,
      "step": 1435
    },
    {
      "epoch": 0.9216,
      "grad_norm": 0.00153350830078125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1440
    },
    {
      "epoch": 0.9248,
      "grad_norm": 6.079673767089844e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1445
    },
    {
      "epoch": 0.928,
      "grad_norm": 0.0001621246337890625,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1450
    },
    {
      "epoch": 0.9312,
      "grad_norm": 0.0111083984375,
      "learning_rate": 0.0001,
      "loss": 0.0028,
      "step": 1455
    },
    {
      "epoch": 0.9344,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.0001,
      "loss": 0.0009,
      "step": 1460
    },
    {
      "epoch": 0.9376,
      "grad_norm": 0.00250244140625,
      "learning_rate": 0.0001,
      "loss": 0.0006,
      "step": 1465
    },
    {
      "epoch": 0.9408,
      "grad_norm": 0.00299072265625,
      "learning_rate": 0.0001,
      "loss": 0.0027,
      "step": 1470
    },
    {
      "epoch": 0.944,
      "grad_norm": 0.000110626220703125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1475
    },
    {
      "epoch": 0.9472,
      "grad_norm": 0.0003757476806640625,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1480
    },
    {
      "epoch": 0.9504,
      "grad_norm": 7.05718994140625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1485
    },
    {
      "epoch": 0.9536,
      "grad_norm": 0.0004425048828125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1490
    },
    {
      "epoch": 0.9568,
      "grad_norm": 0.0001220703125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1495
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.00020503997802734375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1500
    },
    {
      "epoch": 0.9632,
      "grad_norm": 0.014892578125,
      "learning_rate": 0.0001,
      "loss": 0.0014,
      "step": 1505
    },
    {
      "epoch": 0.9664,
      "grad_norm": 0.01470947265625,
      "learning_rate": 0.0001,
      "loss": 0.0012,
      "step": 1510
    },
    {
      "epoch": 0.9696,
      "grad_norm": 0.0004863739013671875,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 1515
    },
    {
      "epoch": 0.9728,
      "grad_norm": 0.0019683837890625,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 1520
    },
    {
      "epoch": 0.976,
      "grad_norm": 4.6253204345703125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1525
    },
    {
      "epoch": 0.9792,
      "grad_norm": 0.00469970703125,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 1530
    },
    {
      "epoch": 0.9824,
      "grad_norm": 0.00494384765625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 1535
    },
    {
      "epoch": 0.9856,
      "grad_norm": 0.00011920928955078125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1540
    },
    {
      "epoch": 0.9888,
      "grad_norm": 4.029273986816406e-05,
      "learning_rate": 0.0001,
      "loss": 0.0023,
      "step": 1545
    },
    {
      "epoch": 0.992,
      "grad_norm": 0.00013828277587890625,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1550
    },
    {
      "epoch": 0.9952,
      "grad_norm": 7.581710815429688e-05,
      "learning_rate": 0.0001,
      "loss": 0.0007,
      "step": 1555
    },
    {
      "epoch": 0.9984,
      "grad_norm": 0.000148773193359375,
      "learning_rate": 0.0001,
      "loss": 0.0023,
      "step": 1560
    },
    {
      "epoch": 1.0016,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.0001,
      "loss": 0.006,
      "step": 1565
    },
    {
      "epoch": 1.0048,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.0001,
      "loss": 0.0044,
      "step": 1570
    },
    {
      "epoch": 1.008,
      "grad_norm": 0.01556396484375,
      "learning_rate": 0.0001,
      "loss": 0.004,
      "step": 1575
    },
    {
      "epoch": 1.0112,
      "grad_norm": 0.01068115234375,
      "learning_rate": 0.0001,
      "loss": 0.0008,
      "step": 1580
    },
    {
      "epoch": 1.0144,
      "grad_norm": 0.0007781982421875,
      "learning_rate": 0.0001,
      "loss": 0.0006,
      "step": 1585
    },
    {
      "epoch": 1.0176,
      "grad_norm": 0.008056640625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 1590
    },
    {
      "epoch": 1.0208,
      "grad_norm": 0.000514984130859375,
      "learning_rate": 0.0001,
      "loss": 0.0007,
      "step": 1595
    },
    {
      "epoch": 1.024,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1600
    },
    {
      "epoch": 1.0272,
      "grad_norm": 0.001129150390625,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1605
    },
    {
      "epoch": 1.0304,
      "grad_norm": 0.11474609375,
      "learning_rate": 0.0001,
      "loss": 0.0019,
      "step": 1610
    },
    {
      "epoch": 1.0336,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.0001,
      "loss": 0.003,
      "step": 1615
    },
    {
      "epoch": 1.0368,
      "grad_norm": 0.01446533203125,
      "learning_rate": 0.0001,
      "loss": 0.0021,
      "step": 1620
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.0001,
      "loss": 0.0016,
      "step": 1625
    },
    {
      "epoch": 1.0432,
      "grad_norm": 0.00640869140625,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 1630
    },
    {
      "epoch": 1.0464,
      "grad_norm": 0.0018310546875,
      "learning_rate": 0.0001,
      "loss": 0.0021,
      "step": 1635
    },
    {
      "epoch": 1.0496,
      "grad_norm": 0.0004405975341796875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1640
    },
    {
      "epoch": 1.0528,
      "grad_norm": 0.0034332275390625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 1645
    },
    {
      "epoch": 1.056,
      "grad_norm": 0.000762939453125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1650
    },
    {
      "epoch": 1.0592,
      "grad_norm": 0.0002536773681640625,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1655
    },
    {
      "epoch": 1.0624,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.0001,
      "loss": 0.0032,
      "step": 1660
    },
    {
      "epoch": 1.0656,
      "grad_norm": 0.01055908203125,
      "learning_rate": 0.0001,
      "loss": 0.0013,
      "step": 1665
    },
    {
      "epoch": 1.0688,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.0001,
      "loss": 0.0017,
      "step": 1670
    },
    {
      "epoch": 1.072,
      "grad_norm": 0.00064849853515625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 1675
    },
    {
      "epoch": 1.0752,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 1680
    },
    {
      "epoch": 1.0784,
      "grad_norm": 0.000354766845703125,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 1685
    },
    {
      "epoch": 1.0816,
      "grad_norm": 0.00518798828125,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 1690
    },
    {
      "epoch": 1.0848,
      "grad_norm": 0.0026092529296875,
      "learning_rate": 0.0001,
      "loss": 0.0009,
      "step": 1695
    },
    {
      "epoch": 1.088,
      "grad_norm": 0.00125885009765625,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1700
    },
    {
      "epoch": 1.0912,
      "grad_norm": 0.00156402587890625,
      "learning_rate": 0.0001,
      "loss": 0.0006,
      "step": 1705
    },
    {
      "epoch": 1.0944,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 1710
    },
    {
      "epoch": 1.0976,
      "grad_norm": 0.0028533935546875,
      "learning_rate": 0.0001,
      "loss": 0.0013,
      "step": 1715
    },
    {
      "epoch": 1.1008,
      "grad_norm": 0.008544921875,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 1720
    },
    {
      "epoch": 1.104,
      "grad_norm": 0.001251220703125,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 1725
    },
    {
      "epoch": 1.1072,
      "grad_norm": 0.0032196044921875,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 1730
    },
    {
      "epoch": 1.1104,
      "grad_norm": 5.245208740234375e-05,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 1735
    },
    {
      "epoch": 1.1136,
      "grad_norm": 0.000293731689453125,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 1740
    },
    {
      "epoch": 1.1168,
      "grad_norm": 0.000751495361328125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1745
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.000194549560546875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1750
    },
    {
      "epoch": 1.1232,
      "grad_norm": 5.793571472167969e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1755
    },
    {
      "epoch": 1.1264,
      "grad_norm": 9.489059448242188e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1760
    },
    {
      "epoch": 1.1296,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.0001,
      "loss": 0.0013,
      "step": 1765
    },
    {
      "epoch": 1.1328,
      "grad_norm": 0.00244140625,
      "learning_rate": 0.0001,
      "loss": 0.0042,
      "step": 1770
    },
    {
      "epoch": 1.1360000000000001,
      "grad_norm": 0.0003643035888671875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1775
    },
    {
      "epoch": 1.1392,
      "grad_norm": 0.00762939453125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1780
    },
    {
      "epoch": 1.1424,
      "grad_norm": 1.5616416931152344e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1785
    },
    {
      "epoch": 1.1456,
      "grad_norm": 4.482269287109375e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1790
    },
    {
      "epoch": 1.1488,
      "grad_norm": 0.00013256072998046875,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 1795
    },
    {
      "epoch": 1.152,
      "grad_norm": 3.600120544433594e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1800
    },
    {
      "epoch": 1.1552,
      "grad_norm": 3.981590270996094e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1805
    },
    {
      "epoch": 1.1584,
      "grad_norm": 0.0009765625,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1810
    },
    {
      "epoch": 1.1616,
      "grad_norm": 0.0034637451171875,
      "learning_rate": 0.0001,
      "loss": 0.0016,
      "step": 1815
    },
    {
      "epoch": 1.1648,
      "grad_norm": 0.00775146484375,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 1820
    },
    {
      "epoch": 1.168,
      "grad_norm": 0.00029754638671875,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 1825
    },
    {
      "epoch": 1.1712,
      "grad_norm": 3.2901763916015625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0024,
      "step": 1830
    },
    {
      "epoch": 1.1743999999999999,
      "grad_norm": 0.0003795623779296875,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 1835
    },
    {
      "epoch": 1.1776,
      "grad_norm": 2.765655517578125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1840
    },
    {
      "epoch": 1.1808,
      "grad_norm": 0.1875,
      "learning_rate": 0.0001,
      "loss": 0.0021,
      "step": 1845
    },
    {
      "epoch": 1.184,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.0001,
      "loss": 0.0039,
      "step": 1850
    },
    {
      "epoch": 1.1872,
      "grad_norm": 0.0027008056640625,
      "learning_rate": 0.0001,
      "loss": 0.005,
      "step": 1855
    },
    {
      "epoch": 1.1904,
      "grad_norm": 0.0017852783203125,
      "learning_rate": 0.0001,
      "loss": 0.0011,
      "step": 1860
    },
    {
      "epoch": 1.1936,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0001,
      "loss": 0.0096,
      "step": 1865
    },
    {
      "epoch": 1.1968,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.0001,
      "loss": 0.0081,
      "step": 1870
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.0086669921875,
      "learning_rate": 0.0001,
      "loss": 0.0012,
      "step": 1875
    },
    {
      "epoch": 1.2032,
      "grad_norm": 0.000690460205078125,
      "learning_rate": 0.0001,
      "loss": 0.0009,
      "step": 1880
    },
    {
      "epoch": 1.2064,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.0001,
      "loss": 0.0027,
      "step": 1885
    },
    {
      "epoch": 1.2096,
      "grad_norm": 0.00012969970703125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1890
    },
    {
      "epoch": 1.2128,
      "grad_norm": 0.003631591796875,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 1895
    },
    {
      "epoch": 1.216,
      "grad_norm": 0.0004482269287109375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1900
    },
    {
      "epoch": 1.2192,
      "grad_norm": 9.584426879882812e-05,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 1905
    },
    {
      "epoch": 1.2224,
      "grad_norm": 0.00075531005859375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1910
    },
    {
      "epoch": 1.2256,
      "grad_norm": 0.00628662109375,
      "learning_rate": 0.0001,
      "loss": 0.0007,
      "step": 1915
    },
    {
      "epoch": 1.2288000000000001,
      "grad_norm": 0.002655029296875,
      "learning_rate": 0.0001,
      "loss": 0.0015,
      "step": 1920
    },
    {
      "epoch": 1.232,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.0001,
      "loss": 0.0012,
      "step": 1925
    },
    {
      "epoch": 1.2352,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 1930
    },
    {
      "epoch": 1.2384,
      "grad_norm": 0.00016689300537109375,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 1935
    },
    {
      "epoch": 1.2416,
      "grad_norm": 0.00014495849609375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1940
    },
    {
      "epoch": 1.2448,
      "grad_norm": 0.0002803802490234375,
      "learning_rate": 0.0001,
      "loss": 0.0089,
      "step": 1945
    },
    {
      "epoch": 1.248,
      "grad_norm": 0.0087890625,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 1950
    },
    {
      "epoch": 1.2511999999999999,
      "grad_norm": 8.106231689453125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0035,
      "step": 1955
    },
    {
      "epoch": 1.2544,
      "grad_norm": 0.0004405975341796875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 1960
    },
    {
      "epoch": 1.2576,
      "grad_norm": 0.015625,
      "learning_rate": 0.0001,
      "loss": 0.0015,
      "step": 1965
    },
    {
      "epoch": 1.2608,
      "grad_norm": 0.00543212890625,
      "learning_rate": 0.0001,
      "loss": 0.0007,
      "step": 1970
    },
    {
      "epoch": 1.264,
      "grad_norm": 0.00104522705078125,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 1975
    },
    {
      "epoch": 1.2671999999999999,
      "grad_norm": 0.015625,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 1980
    },
    {
      "epoch": 1.2704,
      "grad_norm": 0.00013256072998046875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1985
    },
    {
      "epoch": 1.2736,
      "grad_norm": 0.0091552734375,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 1990
    },
    {
      "epoch": 1.2768,
      "grad_norm": 0.000385284423828125,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 1995
    },
    {
      "epoch": 1.28,
      "grad_norm": 5.435943603515625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2000
    },
    {
      "epoch": 1.2832,
      "grad_norm": 8.392333984375e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2005
    },
    {
      "epoch": 1.2864,
      "grad_norm": 0.00024127960205078125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2010
    },
    {
      "epoch": 1.2896,
      "grad_norm": 0.00421142578125,
      "learning_rate": 0.0001,
      "loss": 0.0013,
      "step": 2015
    },
    {
      "epoch": 1.2928,
      "grad_norm": 0.00872802734375,
      "learning_rate": 0.0001,
      "loss": 0.0008,
      "step": 2020
    },
    {
      "epoch": 1.296,
      "grad_norm": 0.000392913818359375,
      "learning_rate": 0.0001,
      "loss": 0.0008,
      "step": 2025
    },
    {
      "epoch": 1.2992,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.0001,
      "loss": 0.0011,
      "step": 2030
    },
    {
      "epoch": 1.3024,
      "grad_norm": 0.00051116943359375,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 2035
    },
    {
      "epoch": 1.3056,
      "grad_norm": 0.0001697540283203125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2040
    },
    {
      "epoch": 1.3088,
      "grad_norm": 0.01336669921875,
      "learning_rate": 0.0001,
      "loss": 0.008,
      "step": 2045
    },
    {
      "epoch": 1.312,
      "grad_norm": 5.435943603515625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2050
    },
    {
      "epoch": 1.3152,
      "grad_norm": 0.000164031982421875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2055
    },
    {
      "epoch": 1.3184,
      "grad_norm": 0.0004558563232421875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2060
    },
    {
      "epoch": 1.3216,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.0001,
      "loss": 0.0012,
      "step": 2065
    },
    {
      "epoch": 1.3248,
      "grad_norm": 0.000415802001953125,
      "learning_rate": 0.0001,
      "loss": 0.0007,
      "step": 2070
    },
    {
      "epoch": 1.328,
      "grad_norm": 0.00104522705078125,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 2075
    },
    {
      "epoch": 1.3312,
      "grad_norm": 0.0003910064697265625,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 2080
    },
    {
      "epoch": 1.3344,
      "grad_norm": 0.00020313262939453125,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 2085
    },
    {
      "epoch": 1.3376000000000001,
      "grad_norm": 5.030632019042969e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2090
    },
    {
      "epoch": 1.3408,
      "grad_norm": 0.00090789794921875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2095
    },
    {
      "epoch": 1.3439999999999999,
      "grad_norm": 0.00037384033203125,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 2100
    },
    {
      "epoch": 1.3472,
      "grad_norm": 0.00014400482177734375,
      "learning_rate": 0.0001,
      "loss": 0.003,
      "step": 2105
    },
    {
      "epoch": 1.3504,
      "grad_norm": 0.00188446044921875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2110
    },
    {
      "epoch": 1.3536000000000001,
      "grad_norm": 0.0023956298828125,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 2115
    },
    {
      "epoch": 1.3568,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 2120
    },
    {
      "epoch": 1.3599999999999999,
      "grad_norm": 0.0103759765625,
      "learning_rate": 0.0001,
      "loss": 0.0009,
      "step": 2125
    },
    {
      "epoch": 1.3632,
      "grad_norm": 0.000926971435546875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2130
    },
    {
      "epoch": 1.3664,
      "grad_norm": 3.0159950256347656e-05,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 2135
    },
    {
      "epoch": 1.3696,
      "grad_norm": 0.00174713134765625,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 2140
    },
    {
      "epoch": 1.3728,
      "grad_norm": 3.1948089599609375e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2145
    },
    {
      "epoch": 1.376,
      "grad_norm": 0.00030517578125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2150
    },
    {
      "epoch": 1.3792,
      "grad_norm": 3.0279159545898438e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2155
    },
    {
      "epoch": 1.3824,
      "grad_norm": 5.030632019042969e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2160
    },
    {
      "epoch": 1.3856,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.0001,
      "loss": 0.0007,
      "step": 2165
    },
    {
      "epoch": 1.3888,
      "grad_norm": 0.00439453125,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 2170
    },
    {
      "epoch": 1.392,
      "grad_norm": 0.0004425048828125,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 2175
    },
    {
      "epoch": 1.3952,
      "grad_norm": 0.005523681640625,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2180
    },
    {
      "epoch": 1.3984,
      "grad_norm": 0.0007781982421875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2185
    },
    {
      "epoch": 1.4016,
      "grad_norm": 2.8252601623535156e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2190
    },
    {
      "epoch": 1.4048,
      "grad_norm": 0.000560760498046875,
      "learning_rate": 0.0001,
      "loss": 0.0043,
      "step": 2195
    },
    {
      "epoch": 1.408,
      "grad_norm": 0.0002574920654296875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2200
    },
    {
      "epoch": 1.4112,
      "grad_norm": 5.507469177246094e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2205
    },
    {
      "epoch": 1.4144,
      "grad_norm": 0.00087738037109375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2210
    },
    {
      "epoch": 1.4176,
      "grad_norm": 0.0057373046875,
      "learning_rate": 0.0001,
      "loss": 0.0009,
      "step": 2215
    },
    {
      "epoch": 1.4208,
      "grad_norm": 0.0093994140625,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 2220
    },
    {
      "epoch": 1.424,
      "grad_norm": 0.000213623046875,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 2225
    },
    {
      "epoch": 1.4272,
      "grad_norm": 0.00421142578125,
      "learning_rate": 0.0001,
      "loss": 0.0006,
      "step": 2230
    },
    {
      "epoch": 1.4304000000000001,
      "grad_norm": 0.0002956390380859375,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2235
    },
    {
      "epoch": 1.4336,
      "grad_norm": 0.00482177734375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2240
    },
    {
      "epoch": 1.4368,
      "grad_norm": 7.05718994140625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 2245
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2250
    },
    {
      "epoch": 1.4432,
      "grad_norm": 4.1484832763671875e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2255
    },
    {
      "epoch": 1.4464000000000001,
      "grad_norm": 0.000701904296875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2260
    },
    {
      "epoch": 1.4496,
      "grad_norm": 0.0123291015625,
      "learning_rate": 0.0001,
      "loss": 0.001,
      "step": 2265
    },
    {
      "epoch": 1.4527999999999999,
      "grad_norm": 0.007110595703125,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 2270
    },
    {
      "epoch": 1.456,
      "grad_norm": 0.00049591064453125,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 2275
    },
    {
      "epoch": 1.4592,
      "grad_norm": 0.00604248046875,
      "learning_rate": 0.0001,
      "loss": 0.0008,
      "step": 2280
    },
    {
      "epoch": 1.4624,
      "grad_norm": 9.965896606445312e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2285
    },
    {
      "epoch": 1.4656,
      "grad_norm": 0.00040435791015625,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2290
    },
    {
      "epoch": 1.4687999999999999,
      "grad_norm": 0.0001773834228515625,
      "learning_rate": 0.0001,
      "loss": 0.0077,
      "step": 2295
    },
    {
      "epoch": 1.472,
      "grad_norm": 2.6226043701171875e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2300
    },
    {
      "epoch": 1.4752,
      "grad_norm": 3.039836883544922e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2305
    },
    {
      "epoch": 1.4784,
      "grad_norm": 0.00025177001953125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2310
    },
    {
      "epoch": 1.4816,
      "grad_norm": 0.006805419921875,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 2315
    },
    {
      "epoch": 1.4848,
      "grad_norm": 0.002532958984375,
      "learning_rate": 0.0001,
      "loss": 0.0006,
      "step": 2320
    },
    {
      "epoch": 1.488,
      "grad_norm": 0.0169677734375,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 2325
    },
    {
      "epoch": 1.4912,
      "grad_norm": 8.726119995117188e-05,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 2330
    },
    {
      "epoch": 1.4944,
      "grad_norm": 1.7762184143066406e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2335
    },
    {
      "epoch": 1.4976,
      "grad_norm": 8.630752563476562e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2340
    },
    {
      "epoch": 1.5008,
      "grad_norm": 0.00075531005859375,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2345
    },
    {
      "epoch": 1.504,
      "grad_norm": 1.7523765563964844e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2350
    },
    {
      "epoch": 1.5072,
      "grad_norm": 1.990795135498047e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2355
    },
    {
      "epoch": 1.5104,
      "grad_norm": 9.870529174804688e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2360
    },
    {
      "epoch": 1.5135999999999998,
      "grad_norm": 0.00115966796875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2365
    },
    {
      "epoch": 1.5168,
      "grad_norm": 0.001068115234375,
      "learning_rate": 0.0001,
      "loss": 0.0008,
      "step": 2370
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.0001659393310546875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2375
    },
    {
      "epoch": 1.5232,
      "grad_norm": 0.00011730194091796875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2380
    },
    {
      "epoch": 1.5264,
      "grad_norm": 1.9550323486328125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 2385
    },
    {
      "epoch": 1.5295999999999998,
      "grad_norm": 9.107589721679688e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2390
    },
    {
      "epoch": 1.5328,
      "grad_norm": 7.82012939453125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 2395
    },
    {
      "epoch": 1.536,
      "grad_norm": 1.6927719116210938e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2400
    },
    {
      "epoch": 1.5392000000000001,
      "grad_norm": 1.6689300537109375e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2405
    },
    {
      "epoch": 1.5424,
      "grad_norm": 7.05718994140625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 2410
    },
    {
      "epoch": 1.5455999999999999,
      "grad_norm": 0.00075531005859375,
      "learning_rate": 0.0001,
      "loss": 0.001,
      "step": 2415
    },
    {
      "epoch": 1.5488,
      "grad_norm": 0.01153564453125,
      "learning_rate": 0.0001,
      "loss": 0.0011,
      "step": 2420
    },
    {
      "epoch": 1.552,
      "grad_norm": 0.000339508056640625,
      "learning_rate": 0.0001,
      "loss": 0.0008,
      "step": 2425
    },
    {
      "epoch": 1.5552000000000001,
      "grad_norm": 0.0022735595703125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2430
    },
    {
      "epoch": 1.5584,
      "grad_norm": 0.000431060791015625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 2435
    },
    {
      "epoch": 1.5615999999999999,
      "grad_norm": 0.001129150390625,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2440
    },
    {
      "epoch": 1.5648,
      "grad_norm": 0.00066375732421875,
      "learning_rate": 0.0001,
      "loss": 0.0008,
      "step": 2445
    },
    {
      "epoch": 1.568,
      "grad_norm": 0.00010156631469726562,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2450
    },
    {
      "epoch": 1.5712000000000002,
      "grad_norm": 0.00031280517578125,
      "learning_rate": 0.0001,
      "loss": 0.0022,
      "step": 2455
    },
    {
      "epoch": 1.5744,
      "grad_norm": 0.0007476806640625,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2460
    },
    {
      "epoch": 1.5776,
      "grad_norm": 0.00994873046875,
      "learning_rate": 0.0001,
      "loss": 0.0008,
      "step": 2465
    },
    {
      "epoch": 1.5808,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.0001,
      "loss": 0.0008,
      "step": 2470
    },
    {
      "epoch": 1.584,
      "grad_norm": 0.0098876953125,
      "learning_rate": 0.0001,
      "loss": 0.0008,
      "step": 2475
    },
    {
      "epoch": 1.5872000000000002,
      "grad_norm": 0.000102996826171875,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 2480
    },
    {
      "epoch": 1.5904,
      "grad_norm": 0.00010824203491210938,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 2485
    },
    {
      "epoch": 1.5936,
      "grad_norm": 0.00016307830810546875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2490
    },
    {
      "epoch": 1.5968,
      "grad_norm": 6.580352783203125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2495
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 2500
    },
    {
      "epoch": 1.6032,
      "grad_norm": 0.00018596649169921875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2505
    },
    {
      "epoch": 1.6064,
      "grad_norm": 7.62939453125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2510
    },
    {
      "epoch": 1.6096,
      "grad_norm": 0.0015869140625,
      "learning_rate": 0.0001,
      "loss": 0.0009,
      "step": 2515
    },
    {
      "epoch": 1.6128,
      "grad_norm": 0.00020122528076171875,
      "learning_rate": 0.0001,
      "loss": 0.0028,
      "step": 2520
    },
    {
      "epoch": 1.616,
      "grad_norm": 0.00058746337890625,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 2525
    },
    {
      "epoch": 1.6192,
      "grad_norm": 0.00017070770263671875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2530
    },
    {
      "epoch": 1.6223999999999998,
      "grad_norm": 5.340576171875e-05,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 2535
    },
    {
      "epoch": 1.6256,
      "grad_norm": 6.818771362304688e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2540
    },
    {
      "epoch": 1.6288,
      "grad_norm": 0.00014495849609375,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 2545
    },
    {
      "epoch": 1.6320000000000001,
      "grad_norm": 4.673004150390625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2550
    },
    {
      "epoch": 1.6352,
      "grad_norm": 6.580352783203125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2555
    },
    {
      "epoch": 1.6383999999999999,
      "grad_norm": 0.00011396408081054688,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2560
    },
    {
      "epoch": 1.6416,
      "grad_norm": 0.0036773681640625,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 2565
    },
    {
      "epoch": 1.6448,
      "grad_norm": 0.004425048828125,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 2570
    },
    {
      "epoch": 1.6480000000000001,
      "grad_norm": 8.678436279296875e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2575
    },
    {
      "epoch": 1.6512,
      "grad_norm": 0.00040435791015625,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2580
    },
    {
      "epoch": 1.6543999999999999,
      "grad_norm": 5.435943603515625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2585
    },
    {
      "epoch": 1.6576,
      "grad_norm": 0.0019378662109375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2590
    },
    {
      "epoch": 1.6608,
      "grad_norm": 2.5153160095214844e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2595
    },
    {
      "epoch": 1.6640000000000001,
      "grad_norm": 2.47955322265625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 2600
    },
    {
      "epoch": 1.6672,
      "grad_norm": 6.437301635742188e-05,
      "learning_rate": 0.0001,
      "loss": 0.0055,
      "step": 2605
    },
    {
      "epoch": 1.6703999999999999,
      "grad_norm": 4.410743713378906e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2610
    },
    {
      "epoch": 1.6736,
      "grad_norm": 0.01953125,
      "learning_rate": 0.0001,
      "loss": 0.0009,
      "step": 2615
    },
    {
      "epoch": 1.6768,
      "grad_norm": 0.000286102294921875,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 2620
    },
    {
      "epoch": 1.6800000000000002,
      "grad_norm": 4.9591064453125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2625
    },
    {
      "epoch": 1.6832,
      "grad_norm": 0.009765625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 2630
    },
    {
      "epoch": 1.6864,
      "grad_norm": 0.0001964569091796875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2635
    },
    {
      "epoch": 1.6896,
      "grad_norm": 0.001983642578125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2640
    },
    {
      "epoch": 1.6928,
      "grad_norm": 0.0001049041748046875,
      "learning_rate": 0.0001,
      "loss": 0.002,
      "step": 2645
    },
    {
      "epoch": 1.696,
      "grad_norm": 4.839897155761719e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2650
    },
    {
      "epoch": 1.6992,
      "grad_norm": 6.4849853515625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2655
    },
    {
      "epoch": 1.7024,
      "grad_norm": 0.00010347366333007812,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2660
    },
    {
      "epoch": 1.7056,
      "grad_norm": 0.00347900390625,
      "learning_rate": 0.0001,
      "loss": 0.0015,
      "step": 2665
    },
    {
      "epoch": 1.7088,
      "grad_norm": 0.00022220611572265625,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 2670
    },
    {
      "epoch": 1.712,
      "grad_norm": 0.000514984130859375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2675
    },
    {
      "epoch": 1.7151999999999998,
      "grad_norm": 0.000308990478515625,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2680
    },
    {
      "epoch": 1.7184,
      "grad_norm": 4.00543212890625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 2685
    },
    {
      "epoch": 1.7216,
      "grad_norm": 0.007568359375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2690
    },
    {
      "epoch": 1.7248,
      "grad_norm": 0.00115966796875,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 2695
    },
    {
      "epoch": 1.728,
      "grad_norm": 0.0076904296875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2700
    },
    {
      "epoch": 1.7311999999999999,
      "grad_norm": 0.00022602081298828125,
      "learning_rate": 0.0001,
      "loss": 0.0026,
      "step": 2705
    },
    {
      "epoch": 1.7344,
      "grad_norm": 9.632110595703125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2710
    },
    {
      "epoch": 1.7376,
      "grad_norm": 0.0118408203125,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 2715
    },
    {
      "epoch": 1.7408000000000001,
      "grad_norm": 0.00146484375,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 2720
    },
    {
      "epoch": 1.744,
      "grad_norm": 0.0023345947265625,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 2725
    },
    {
      "epoch": 1.7471999999999999,
      "grad_norm": 9.822845458984375e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2730
    },
    {
      "epoch": 1.7504,
      "grad_norm": 2.4318695068359375e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2735
    },
    {
      "epoch": 1.7536,
      "grad_norm": 9.393692016601562e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2740
    },
    {
      "epoch": 1.7568000000000001,
      "grad_norm": 1.823902130126953e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2745
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.0001773834228515625,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2750
    },
    {
      "epoch": 1.7631999999999999,
      "grad_norm": 1.6927719116210938e-05,
      "learning_rate": 0.0001,
      "loss": 0.0011,
      "step": 2755
    },
    {
      "epoch": 1.7664,
      "grad_norm": 0.0002307891845703125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2760
    },
    {
      "epoch": 1.7696,
      "grad_norm": 0.0021820068359375,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2765
    },
    {
      "epoch": 1.7728000000000002,
      "grad_norm": 0.01446533203125,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 2770
    },
    {
      "epoch": 1.776,
      "grad_norm": 0.0062255859375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2775
    },
    {
      "epoch": 1.7792,
      "grad_norm": 7.2479248046875e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2780
    },
    {
      "epoch": 1.7824,
      "grad_norm": 6.031990051269531e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2785
    },
    {
      "epoch": 1.7856,
      "grad_norm": 3.314018249511719e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2790
    },
    {
      "epoch": 1.7888,
      "grad_norm": 0.00011873245239257812,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2795
    },
    {
      "epoch": 1.792,
      "grad_norm": 0.00274658203125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2800
    },
    {
      "epoch": 1.7952,
      "grad_norm": 1.7762184143066406e-05,
      "learning_rate": 0.0001,
      "loss": 0.0026,
      "step": 2805
    },
    {
      "epoch": 1.7984,
      "grad_norm": 0.00110626220703125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2810
    },
    {
      "epoch": 1.8016,
      "grad_norm": 0.000553131103515625,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2815
    },
    {
      "epoch": 1.8048,
      "grad_norm": 0.00390625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 2820
    },
    {
      "epoch": 1.808,
      "grad_norm": 8.440017700195312e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2825
    },
    {
      "epoch": 1.8112,
      "grad_norm": 0.01214599609375,
      "learning_rate": 0.0001,
      "loss": 0.002,
      "step": 2830
    },
    {
      "epoch": 1.8144,
      "grad_norm": 1.3053417205810547e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2835
    },
    {
      "epoch": 1.8176,
      "grad_norm": 0.000514984130859375,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2840
    },
    {
      "epoch": 1.8208,
      "grad_norm": 1.990795135498047e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2845
    },
    {
      "epoch": 1.8239999999999998,
      "grad_norm": 2.968311309814453e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2850
    },
    {
      "epoch": 1.8272,
      "grad_norm": 2.2292137145996094e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2855
    },
    {
      "epoch": 1.8304,
      "grad_norm": 5.078315734863281e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2860
    },
    {
      "epoch": 1.8336000000000001,
      "grad_norm": 0.000812530517578125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2865
    },
    {
      "epoch": 1.8368,
      "grad_norm": 0.00077056884765625,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2870
    },
    {
      "epoch": 1.8399999999999999,
      "grad_norm": 0.007781982421875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2875
    },
    {
      "epoch": 1.8432,
      "grad_norm": 0.01385498046875,
      "learning_rate": 0.0001,
      "loss": 0.0007,
      "step": 2880
    },
    {
      "epoch": 1.8464,
      "grad_norm": 0.01507568359375,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 2885
    },
    {
      "epoch": 1.8496000000000001,
      "grad_norm": 8.96453857421875e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2890
    },
    {
      "epoch": 1.8528,
      "grad_norm": 0.003997802734375,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2895
    },
    {
      "epoch": 1.8559999999999999,
      "grad_norm": 0.000301361083984375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2900
    },
    {
      "epoch": 1.8592,
      "grad_norm": 2.4199485778808594e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2905
    },
    {
      "epoch": 1.8624,
      "grad_norm": 0.000530242919921875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2910
    },
    {
      "epoch": 1.8656000000000001,
      "grad_norm": 0.0048828125,
      "learning_rate": 0.0001,
      "loss": 0.0017,
      "step": 2915
    },
    {
      "epoch": 1.8688,
      "grad_norm": 0.00115203857421875,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 2920
    },
    {
      "epoch": 1.8719999999999999,
      "grad_norm": 0.001678466796875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2925
    },
    {
      "epoch": 1.8752,
      "grad_norm": 8.296966552734375e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2930
    },
    {
      "epoch": 1.8784,
      "grad_norm": 7.724761962890625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 2935
    },
    {
      "epoch": 1.8816000000000002,
      "grad_norm": 0.00075531005859375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2940
    },
    {
      "epoch": 1.8848,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.0001,
      "loss": 0.0082,
      "step": 2945
    },
    {
      "epoch": 1.888,
      "grad_norm": 9.34600830078125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2950
    },
    {
      "epoch": 1.8912,
      "grad_norm": 6.341934204101562e-05,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 2955
    },
    {
      "epoch": 1.8944,
      "grad_norm": 0.000598907470703125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2960
    },
    {
      "epoch": 1.8976,
      "grad_norm": 0.000759124755859375,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 2965
    },
    {
      "epoch": 1.9008,
      "grad_norm": 0.00115203857421875,
      "learning_rate": 0.0001,
      "loss": 0.0016,
      "step": 2970
    },
    {
      "epoch": 1.904,
      "grad_norm": 0.00017452239990234375,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2975
    },
    {
      "epoch": 1.9072,
      "grad_norm": 0.0005645751953125,
      "learning_rate": 0.0001,
      "loss": 0.0011,
      "step": 2980
    },
    {
      "epoch": 1.9104,
      "grad_norm": 0.0007781982421875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 2985
    },
    {
      "epoch": 1.9136,
      "grad_norm": 0.00010824203491210938,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 2990
    },
    {
      "epoch": 1.9167999999999998,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.0001,
      "loss": 0.0083,
      "step": 2995
    },
    {
      "epoch": 1.92,
      "grad_norm": 5.245208740234375e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3000
    },
    {
      "epoch": 1.9232,
      "grad_norm": 4.315376281738281e-05,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 3005
    },
    {
      "epoch": 1.9264000000000001,
      "grad_norm": 0.00014400482177734375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3010
    },
    {
      "epoch": 1.9296,
      "grad_norm": 0.01171875,
      "learning_rate": 0.0001,
      "loss": 0.0016,
      "step": 3015
    },
    {
      "epoch": 1.9327999999999999,
      "grad_norm": 0.000598907470703125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3020
    },
    {
      "epoch": 1.936,
      "grad_norm": 3.504753112792969e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3025
    },
    {
      "epoch": 1.9392,
      "grad_norm": 6.628036499023438e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3030
    },
    {
      "epoch": 1.9424000000000001,
      "grad_norm": 5.8650970458984375e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3035
    },
    {
      "epoch": 1.9456,
      "grad_norm": 6.866455078125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 3040
    },
    {
      "epoch": 1.9487999999999999,
      "grad_norm": 0.00048828125,
      "learning_rate": 0.0001,
      "loss": 0.0041,
      "step": 3045
    },
    {
      "epoch": 1.952,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.0001,
      "loss": 0.0007,
      "step": 3050
    },
    {
      "epoch": 1.9552,
      "grad_norm": 0.000919342041015625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 3055
    },
    {
      "epoch": 1.9584000000000001,
      "grad_norm": 0.0103759765625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 3060
    },
    {
      "epoch": 1.9616,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.0001,
      "loss": 0.0022,
      "step": 3065
    },
    {
      "epoch": 1.9647999999999999,
      "grad_norm": 0.00982666015625,
      "learning_rate": 0.0001,
      "loss": 0.0014,
      "step": 3070
    },
    {
      "epoch": 1.968,
      "grad_norm": 0.00089263916015625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 3075
    },
    {
      "epoch": 1.9712,
      "grad_norm": 0.0076904296875,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 3080
    },
    {
      "epoch": 1.9744000000000002,
      "grad_norm": 0.00604248046875,
      "learning_rate": 0.0001,
      "loss": 0.0006,
      "step": 3085
    },
    {
      "epoch": 1.9776,
      "grad_norm": 0.008544921875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3090
    },
    {
      "epoch": 1.9808,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 3095
    },
    {
      "epoch": 1.984,
      "grad_norm": 4.935264587402344e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3100
    },
    {
      "epoch": 1.9872,
      "grad_norm": 0.00018596649169921875,
      "learning_rate": 0.0001,
      "loss": 0.0035,
      "step": 3105
    },
    {
      "epoch": 1.9904,
      "grad_norm": 0.000698089599609375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3110
    },
    {
      "epoch": 1.9936,
      "grad_norm": 0.00113677978515625,
      "learning_rate": 0.0001,
      "loss": 0.0006,
      "step": 3115
    },
    {
      "epoch": 1.9968,
      "grad_norm": 7.200241088867188e-05,
      "learning_rate": 0.0001,
      "loss": 0.0008,
      "step": 3120
    },
    {
      "epoch": 2.0,
      "grad_norm": 9.5367431640625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3125
    },
    {
      "epoch": 2.0032,
      "grad_norm": 0.007568359375,
      "learning_rate": 0.0001,
      "loss": 0.0032,
      "step": 3130
    },
    {
      "epoch": 2.0064,
      "grad_norm": 0.0019073486328125,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 3135
    },
    {
      "epoch": 2.0096,
      "grad_norm": 0.0081787109375,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 3140
    },
    {
      "epoch": 2.0128,
      "grad_norm": 0.00014019012451171875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3145
    },
    {
      "epoch": 2.016,
      "grad_norm": 1.8358230590820312e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3150
    },
    {
      "epoch": 2.0192,
      "grad_norm": 0.005828857421875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3155
    },
    {
      "epoch": 2.0224,
      "grad_norm": 4.3392181396484375e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3160
    },
    {
      "epoch": 2.0256,
      "grad_norm": 0.0002460479736328125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3165
    },
    {
      "epoch": 2.0288,
      "grad_norm": 0.000545501708984375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3170
    },
    {
      "epoch": 2.032,
      "grad_norm": 2.849102020263672e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3175
    },
    {
      "epoch": 2.0352,
      "grad_norm": 0.007781982421875,
      "learning_rate": 0.0001,
      "loss": 0.0011,
      "step": 3180
    },
    {
      "epoch": 2.0384,
      "grad_norm": 0.0003719329833984375,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 3185
    },
    {
      "epoch": 2.0416,
      "grad_norm": 1.1682510375976562e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3190
    },
    {
      "epoch": 2.0448,
      "grad_norm": 0.000732421875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3195
    },
    {
      "epoch": 2.048,
      "grad_norm": 1.633167266845703e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3200
    },
    {
      "epoch": 2.0512,
      "grad_norm": 0.000431060791015625,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3205
    },
    {
      "epoch": 2.0544,
      "grad_norm": 3.123283386230469e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3210
    },
    {
      "epoch": 2.0576,
      "grad_norm": 2.491474151611328e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3215
    },
    {
      "epoch": 2.0608,
      "grad_norm": 1.8358230590820312e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3220
    },
    {
      "epoch": 2.064,
      "grad_norm": 2.7179718017578125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3225
    },
    {
      "epoch": 2.0672,
      "grad_norm": 0.0028839111328125,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 3230
    },
    {
      "epoch": 2.0704,
      "grad_norm": 0.00506591796875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3235
    },
    {
      "epoch": 2.0736,
      "grad_norm": 0.0004749298095703125,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 3240
    },
    {
      "epoch": 2.0768,
      "grad_norm": 0.001251220703125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3245
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.633167266845703e-05,
      "learning_rate": 0.0001,
      "loss": 0.0006,
      "step": 3250
    },
    {
      "epoch": 2.0832,
      "grad_norm": 5.817413330078125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3255
    },
    {
      "epoch": 2.0864,
      "grad_norm": 5.817413330078125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3260
    },
    {
      "epoch": 2.0896,
      "grad_norm": 0.007598876953125,
      "learning_rate": 0.0001,
      "loss": 0.0012,
      "step": 3265
    },
    {
      "epoch": 2.0928,
      "grad_norm": 7.200241088867188e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3270
    },
    {
      "epoch": 2.096,
      "grad_norm": 6.866455078125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3275
    },
    {
      "epoch": 2.0992,
      "grad_norm": 0.000553131103515625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 3280
    },
    {
      "epoch": 2.1024,
      "grad_norm": 0.015380859375,
      "learning_rate": 0.0001,
      "loss": 0.0009,
      "step": 3285
    },
    {
      "epoch": 2.1056,
      "grad_norm": 8.58306884765625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3290
    },
    {
      "epoch": 2.1088,
      "grad_norm": 0.01165771484375,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 3295
    },
    {
      "epoch": 2.112,
      "grad_norm": 1.2755393981933594e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3300
    },
    {
      "epoch": 2.1152,
      "grad_norm": 0.000213623046875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3305
    },
    {
      "epoch": 2.1184,
      "grad_norm": 2.0503997802734375e-05,
      "learning_rate": 0.0001,
      "loss": 0.0048,
      "step": 3310
    },
    {
      "epoch": 2.1216,
      "grad_norm": 0.00011348724365234375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3315
    },
    {
      "epoch": 2.1248,
      "grad_norm": 0.000301361083984375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3320
    },
    {
      "epoch": 2.128,
      "grad_norm": 0.00041961669921875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3325
    },
    {
      "epoch": 2.1312,
      "grad_norm": 0.0007781982421875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3330
    },
    {
      "epoch": 2.1344,
      "grad_norm": 0.00171661376953125,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 3335
    },
    {
      "epoch": 2.1376,
      "grad_norm": 8.7738037109375e-05,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 3340
    },
    {
      "epoch": 2.1408,
      "grad_norm": 0.000270843505859375,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3345
    },
    {
      "epoch": 2.144,
      "grad_norm": 7.867813110351562e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3350
    },
    {
      "epoch": 2.1471999999999998,
      "grad_norm": 0.0002269744873046875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3355
    },
    {
      "epoch": 2.1504,
      "grad_norm": 1.4781951904296875e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3360
    },
    {
      "epoch": 2.1536,
      "grad_norm": 0.0087890625,
      "learning_rate": 0.0001,
      "loss": 0.0013,
      "step": 3365
    },
    {
      "epoch": 2.1568,
      "grad_norm": 1.4066696166992188e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3370
    },
    {
      "epoch": 2.16,
      "grad_norm": 3.4332275390625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3375
    },
    {
      "epoch": 2.1632,
      "grad_norm": 0.0024566650390625,
      "learning_rate": 0.0001,
      "loss": 0.0007,
      "step": 3380
    },
    {
      "epoch": 2.1664,
      "grad_norm": 0.01202392578125,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 3385
    },
    {
      "epoch": 2.1696,
      "grad_norm": 8.058547973632812e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3390
    },
    {
      "epoch": 2.1728,
      "grad_norm": 0.00136566162109375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3395
    },
    {
      "epoch": 2.176,
      "grad_norm": 5.53131103515625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3400
    },
    {
      "epoch": 2.1792,
      "grad_norm": 0.00021266937255859375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3405
    },
    {
      "epoch": 2.1824,
      "grad_norm": 2.1576881408691406e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3410
    },
    {
      "epoch": 2.1856,
      "grad_norm": 2.288818359375e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3415
    },
    {
      "epoch": 2.1888,
      "grad_norm": 2.0623207092285156e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3420
    },
    {
      "epoch": 2.192,
      "grad_norm": 0.0003108978271484375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3425
    },
    {
      "epoch": 2.1952,
      "grad_norm": 0.001983642578125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3430
    },
    {
      "epoch": 2.1984,
      "grad_norm": 0.00531005859375,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 3435
    },
    {
      "epoch": 2.2016,
      "grad_norm": 0.0015716552734375,
      "learning_rate": 0.0001,
      "loss": 0.003,
      "step": 3440
    },
    {
      "epoch": 2.2048,
      "grad_norm": 0.0003223419189453125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3445
    },
    {
      "epoch": 2.208,
      "grad_norm": 0.0003376007080078125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3450
    },
    {
      "epoch": 2.2112,
      "grad_norm": 3.0040740966796875e-05,
      "learning_rate": 0.0001,
      "loss": 0.0035,
      "step": 3455
    },
    {
      "epoch": 2.2144,
      "grad_norm": 0.000858306884765625,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3460
    },
    {
      "epoch": 2.2176,
      "grad_norm": 0.00054168701171875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3465
    },
    {
      "epoch": 2.2208,
      "grad_norm": 1.9311904907226562e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3470
    },
    {
      "epoch": 2.224,
      "grad_norm": 0.0002803802490234375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3475
    },
    {
      "epoch": 2.2272,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 3480
    },
    {
      "epoch": 2.2304,
      "grad_norm": 4.76837158203125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3485
    },
    {
      "epoch": 2.2336,
      "grad_norm": 5.1975250244140625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3490
    },
    {
      "epoch": 2.2368,
      "grad_norm": 7.915496826171875e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3495
    },
    {
      "epoch": 2.24,
      "grad_norm": 7.212162017822266e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3500
    },
    {
      "epoch": 2.2432,
      "grad_norm": 7.05718994140625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3505
    },
    {
      "epoch": 2.2464,
      "grad_norm": 1.4960765838623047e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3510
    },
    {
      "epoch": 2.2496,
      "grad_norm": 0.017578125,
      "learning_rate": 0.0001,
      "loss": 0.0027,
      "step": 3515
    },
    {
      "epoch": 2.2528,
      "grad_norm": 1.1146068572998047e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3520
    },
    {
      "epoch": 2.2560000000000002,
      "grad_norm": 0.0012664794921875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3525
    },
    {
      "epoch": 2.2592,
      "grad_norm": 0.00010824203491210938,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3530
    },
    {
      "epoch": 2.2624,
      "grad_norm": 0.003509521484375,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3535
    },
    {
      "epoch": 2.2656,
      "grad_norm": 7.510185241699219e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3540
    },
    {
      "epoch": 2.2688,
      "grad_norm": 0.00157928466796875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3545
    },
    {
      "epoch": 2.2720000000000002,
      "grad_norm": 4.500150680541992e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3550
    },
    {
      "epoch": 2.2752,
      "grad_norm": 2.086162567138672e-05,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 3555
    },
    {
      "epoch": 2.2784,
      "grad_norm": 0.00011110305786132812,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3560
    },
    {
      "epoch": 2.2816,
      "grad_norm": 1.4722347259521484e-05,
      "learning_rate": 0.0001,
      "loss": 0.0007,
      "step": 3565
    },
    {
      "epoch": 2.2848,
      "grad_norm": 1.0788440704345703e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3570
    },
    {
      "epoch": 2.288,
      "grad_norm": 5.245208740234375e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3575
    },
    {
      "epoch": 2.2912,
      "grad_norm": 0.0019989013671875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3580
    },
    {
      "epoch": 2.2944,
      "grad_norm": 0.0069580078125,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 3585
    },
    {
      "epoch": 2.2976,
      "grad_norm": 6.22868537902832e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3590
    },
    {
      "epoch": 2.3008,
      "grad_norm": 6.580352783203125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3595
    },
    {
      "epoch": 2.304,
      "grad_norm": 5.185604095458984e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3600
    },
    {
      "epoch": 2.3072,
      "grad_norm": 2.288818359375e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3605
    },
    {
      "epoch": 2.3104,
      "grad_norm": 6.29425048828125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0067,
      "step": 3610
    },
    {
      "epoch": 2.3136,
      "grad_norm": 1.9669532775878906e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3615
    },
    {
      "epoch": 2.3168,
      "grad_norm": 1.2755393981933594e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3620
    },
    {
      "epoch": 2.32,
      "grad_norm": 2.9206275939941406e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3625
    },
    {
      "epoch": 2.3232,
      "grad_norm": 0.0078125,
      "learning_rate": 0.0001,
      "loss": 0.0006,
      "step": 3630
    },
    {
      "epoch": 2.3264,
      "grad_norm": 0.0005950927734375,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3635
    },
    {
      "epoch": 2.3296,
      "grad_norm": 0.0009613037109375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3640
    },
    {
      "epoch": 2.3327999999999998,
      "grad_norm": 3.409385681152344e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3645
    },
    {
      "epoch": 2.336,
      "grad_norm": 0.000598907470703125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3650
    },
    {
      "epoch": 2.3392,
      "grad_norm": 0.002532958984375,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 3655
    },
    {
      "epoch": 2.3424,
      "grad_norm": 1.3053417205810547e-05,
      "learning_rate": 0.0001,
      "loss": 0.0065,
      "step": 3660
    },
    {
      "epoch": 2.3456,
      "grad_norm": 3.4809112548828125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3665
    },
    {
      "epoch": 2.3487999999999998,
      "grad_norm": 9.655952453613281e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3670
    },
    {
      "epoch": 2.352,
      "grad_norm": 2.4199485778808594e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3675
    },
    {
      "epoch": 2.3552,
      "grad_norm": 0.00072479248046875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3680
    },
    {
      "epoch": 2.3584,
      "grad_norm": 0.000553131103515625,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3685
    },
    {
      "epoch": 2.3616,
      "grad_norm": 5.7220458984375e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3690
    },
    {
      "epoch": 2.3648,
      "grad_norm": 4.2438507080078125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3695
    },
    {
      "epoch": 2.368,
      "grad_norm": 0.0003528594970703125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3700
    },
    {
      "epoch": 2.3712,
      "grad_norm": 3.552436828613281e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3705
    },
    {
      "epoch": 2.3744,
      "grad_norm": 0.0120849609375,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 3710
    },
    {
      "epoch": 2.3776,
      "grad_norm": 2.2172927856445312e-05,
      "learning_rate": 0.0001,
      "loss": 0.004,
      "step": 3715
    },
    {
      "epoch": 2.3808,
      "grad_norm": 1.9669532775878906e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3720
    },
    {
      "epoch": 2.384,
      "grad_norm": 2.110004425048828e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3725
    },
    {
      "epoch": 2.3872,
      "grad_norm": 0.006866455078125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3730
    },
    {
      "epoch": 2.3904,
      "grad_norm": 0.00118255615234375,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 3735
    },
    {
      "epoch": 2.3936,
      "grad_norm": 1.2099742889404297e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3740
    },
    {
      "epoch": 2.3968,
      "grad_norm": 0.00213623046875,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 3745
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.0848045349121094e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3750
    },
    {
      "epoch": 2.4032,
      "grad_norm": 2.8371810913085938e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3755
    },
    {
      "epoch": 2.4064,
      "grad_norm": 1.4662742614746094e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3760
    },
    {
      "epoch": 2.4096,
      "grad_norm": 1.6927719116210938e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3765
    },
    {
      "epoch": 2.4128,
      "grad_norm": 1.2516975402832031e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3770
    },
    {
      "epoch": 2.416,
      "grad_norm": 1.8835067749023438e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3775
    },
    {
      "epoch": 2.4192,
      "grad_norm": 0.000213623046875,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 3780
    },
    {
      "epoch": 2.4224,
      "grad_norm": 8.153915405273438e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3785
    },
    {
      "epoch": 2.4256,
      "grad_norm": 2.288818359375e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3790
    },
    {
      "epoch": 2.4288,
      "grad_norm": 0.0001430511474609375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3795
    },
    {
      "epoch": 2.432,
      "grad_norm": 6.288290023803711e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3800
    },
    {
      "epoch": 2.4352,
      "grad_norm": 0.0002727508544921875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3805
    },
    {
      "epoch": 2.4384,
      "grad_norm": 1.4603137969970703e-05,
      "learning_rate": 0.0001,
      "loss": 0.0082,
      "step": 3810
    },
    {
      "epoch": 2.4416,
      "grad_norm": 2.0742416381835938e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3815
    },
    {
      "epoch": 2.4448,
      "grad_norm": 9.000301361083984e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3820
    },
    {
      "epoch": 2.448,
      "grad_norm": 1.4185905456542969e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3825
    },
    {
      "epoch": 2.4512,
      "grad_norm": 0.00021648406982421875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3830
    },
    {
      "epoch": 2.4544,
      "grad_norm": 4.172325134277344e-05,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 3835
    },
    {
      "epoch": 2.4576000000000002,
      "grad_norm": 6.377696990966797e-06,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3840
    },
    {
      "epoch": 2.4608,
      "grad_norm": 0.00145721435546875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3845
    },
    {
      "epoch": 2.464,
      "grad_norm": 9.59634780883789e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3850
    },
    {
      "epoch": 2.4672,
      "grad_norm": 0.0015869140625,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 3855
    },
    {
      "epoch": 2.4704,
      "grad_norm": 1.7404556274414062e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3860
    },
    {
      "epoch": 2.4736000000000002,
      "grad_norm": 0.00075531005859375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3865
    },
    {
      "epoch": 2.4768,
      "grad_norm": 7.420778274536133e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3870
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.4483928680419922e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3875
    },
    {
      "epoch": 2.4832,
      "grad_norm": 4.6253204345703125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3880
    },
    {
      "epoch": 2.4864,
      "grad_norm": 0.0013275146484375,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 3885
    },
    {
      "epoch": 2.4896,
      "grad_norm": 0.000568389892578125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3890
    },
    {
      "epoch": 2.4928,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0001,
      "loss": 0.0014,
      "step": 3895
    },
    {
      "epoch": 2.496,
      "grad_norm": 0.00555419921875,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 3900
    },
    {
      "epoch": 2.4992,
      "grad_norm": 0.00024318695068359375,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3905
    },
    {
      "epoch": 2.5023999999999997,
      "grad_norm": 3.7670135498046875e-05,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 3910
    },
    {
      "epoch": 2.5056000000000003,
      "grad_norm": 0.0002765655517578125,
      "learning_rate": 0.0001,
      "loss": 0.0036,
      "step": 3915
    },
    {
      "epoch": 2.5088,
      "grad_norm": 0.005462646484375,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3920
    },
    {
      "epoch": 2.512,
      "grad_norm": 3.886222839355469e-05,
      "learning_rate": 0.0001,
      "loss": 0.0007,
      "step": 3925
    },
    {
      "epoch": 2.5152,
      "grad_norm": 0.0020751953125,
      "learning_rate": 0.0001,
      "loss": 0.0008,
      "step": 3930
    },
    {
      "epoch": 2.5183999999999997,
      "grad_norm": 0.003082275390625,
      "learning_rate": 0.0001,
      "loss": 0.0011,
      "step": 3935
    },
    {
      "epoch": 2.5216,
      "grad_norm": 0.0002231597900390625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 3940
    },
    {
      "epoch": 2.5248,
      "grad_norm": 0.0017547607421875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3945
    },
    {
      "epoch": 2.528,
      "grad_norm": 8.487701416015625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 3950
    },
    {
      "epoch": 2.5312,
      "grad_norm": 0.000263214111328125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3955
    },
    {
      "epoch": 2.5343999999999998,
      "grad_norm": 0.0002307891845703125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 3960
    },
    {
      "epoch": 2.5376,
      "grad_norm": 0.000244140625,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3965
    },
    {
      "epoch": 2.5408,
      "grad_norm": 5.459785461425781e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3970
    },
    {
      "epoch": 2.544,
      "grad_norm": 3.075599670410156e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3975
    },
    {
      "epoch": 2.5472,
      "grad_norm": 0.0008087158203125,
      "learning_rate": 0.0001,
      "loss": 0.0024,
      "step": 3980
    },
    {
      "epoch": 2.5504,
      "grad_norm": 0.007781982421875,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 3985
    },
    {
      "epoch": 2.5536,
      "grad_norm": 0.0004863739013671875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 3990
    },
    {
      "epoch": 2.5568,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.0001,
      "loss": 0.0016,
      "step": 3995
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.010009765625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 4000
    },
    {
      "epoch": 2.5632,
      "grad_norm": 0.006134033203125,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 4005
    },
    {
      "epoch": 2.5664,
      "grad_norm": 0.0010833740234375,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4010
    },
    {
      "epoch": 2.5696,
      "grad_norm": 3.695487976074219e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4015
    },
    {
      "epoch": 2.5728,
      "grad_norm": 4.2438507080078125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4020
    },
    {
      "epoch": 2.576,
      "grad_norm": 2.4199485778808594e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4025
    },
    {
      "epoch": 2.5792,
      "grad_norm": 0.00144195556640625,
      "learning_rate": 0.0001,
      "loss": 0.0026,
      "step": 4030
    },
    {
      "epoch": 2.5824,
      "grad_norm": 0.0015716552734375,
      "learning_rate": 0.0001,
      "loss": 0.0009,
      "step": 4035
    },
    {
      "epoch": 2.5856,
      "grad_norm": 0.00049591064453125,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 4040
    },
    {
      "epoch": 2.5888,
      "grad_norm": 0.0022430419921875,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 4045
    },
    {
      "epoch": 2.592,
      "grad_norm": 0.0001544952392578125,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 4050
    },
    {
      "epoch": 2.5952,
      "grad_norm": 0.0076904296875,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 4055
    },
    {
      "epoch": 2.5984,
      "grad_norm": 0.000446319580078125,
      "learning_rate": 0.0001,
      "loss": 0.0007,
      "step": 4060
    },
    {
      "epoch": 2.6016,
      "grad_norm": 6.341934204101562e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4065
    },
    {
      "epoch": 2.6048,
      "grad_norm": 3.6716461181640625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4070
    },
    {
      "epoch": 2.608,
      "grad_norm": 0.000461578369140625,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4075
    },
    {
      "epoch": 2.6112,
      "grad_norm": 0.003204345703125,
      "learning_rate": 0.0001,
      "loss": 0.0011,
      "step": 4080
    },
    {
      "epoch": 2.6144,
      "grad_norm": 0.00286865234375,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 4085
    },
    {
      "epoch": 2.6176,
      "grad_norm": 6.67572021484375e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4090
    },
    {
      "epoch": 2.6208,
      "grad_norm": 0.00110626220703125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4095
    },
    {
      "epoch": 2.624,
      "grad_norm": 0.0003604888916015625,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4100
    },
    {
      "epoch": 2.6272,
      "grad_norm": 0.004302978515625,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4105
    },
    {
      "epoch": 2.6304,
      "grad_norm": 6.4849853515625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4110
    },
    {
      "epoch": 2.6336,
      "grad_norm": 0.000171661376953125,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 4115
    },
    {
      "epoch": 2.6368,
      "grad_norm": 5.781650543212891e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4120
    },
    {
      "epoch": 2.64,
      "grad_norm": 9.000301361083984e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4125
    },
    {
      "epoch": 2.6432,
      "grad_norm": 0.0030975341796875,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 4130
    },
    {
      "epoch": 2.6464,
      "grad_norm": 0.00131988525390625,
      "learning_rate": 0.0001,
      "loss": 0.0016,
      "step": 4135
    },
    {
      "epoch": 2.6496,
      "grad_norm": 0.000354766845703125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4140
    },
    {
      "epoch": 2.6528,
      "grad_norm": 7.724761962890625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4145
    },
    {
      "epoch": 2.656,
      "grad_norm": 4.231929779052734e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4150
    },
    {
      "epoch": 2.6592000000000002,
      "grad_norm": 1.704692840576172e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4155
    },
    {
      "epoch": 2.6624,
      "grad_norm": 6.616115570068359e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4160
    },
    {
      "epoch": 2.6656,
      "grad_norm": 6.884336471557617e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4165
    },
    {
      "epoch": 2.6688,
      "grad_norm": 5.453824996948242e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4170
    },
    {
      "epoch": 2.672,
      "grad_norm": 0.000179290771484375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4175
    },
    {
      "epoch": 2.6752000000000002,
      "grad_norm": 0.013916015625,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 4180
    },
    {
      "epoch": 2.6784,
      "grad_norm": 0.0031890869140625,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4185
    },
    {
      "epoch": 2.6816,
      "grad_norm": 1.7404556274414062e-05,
      "learning_rate": 0.0001,
      "loss": 0.0006,
      "step": 4190
    },
    {
      "epoch": 2.6848,
      "grad_norm": 4.291534423828125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4195
    },
    {
      "epoch": 2.6879999999999997,
      "grad_norm": 7.420778274536133e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4200
    },
    {
      "epoch": 2.6912000000000003,
      "grad_norm": 0.0001087188720703125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4205
    },
    {
      "epoch": 2.6944,
      "grad_norm": 3.647804260253906e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4210
    },
    {
      "epoch": 2.6976,
      "grad_norm": 0.01220703125,
      "learning_rate": 0.0001,
      "loss": 0.0019,
      "step": 4215
    },
    {
      "epoch": 2.7008,
      "grad_norm": 6.109476089477539e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4220
    },
    {
      "epoch": 2.7039999999999997,
      "grad_norm": 1.8477439880371094e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4225
    },
    {
      "epoch": 2.7072000000000003,
      "grad_norm": 0.0009765625,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4230
    },
    {
      "epoch": 2.7104,
      "grad_norm": 0.0002918243408203125,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 4235
    },
    {
      "epoch": 2.7136,
      "grad_norm": 7.62939453125e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4240
    },
    {
      "epoch": 2.7168,
      "grad_norm": 5.340576171875e-05,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 4245
    },
    {
      "epoch": 2.7199999999999998,
      "grad_norm": 5.930662155151367e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4250
    },
    {
      "epoch": 2.7232,
      "grad_norm": 0.0009613037109375,
      "learning_rate": 0.0001,
      "loss": 0.0008,
      "step": 4255
    },
    {
      "epoch": 2.7264,
      "grad_norm": 2.3603439331054688e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4260
    },
    {
      "epoch": 2.7296,
      "grad_norm": 2.110004425048828e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4265
    },
    {
      "epoch": 2.7328,
      "grad_norm": 1.055002212524414e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4270
    },
    {
      "epoch": 2.7359999999999998,
      "grad_norm": 0.00011110305786132812,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4275
    },
    {
      "epoch": 2.7392,
      "grad_norm": 0.0032196044921875,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 4280
    },
    {
      "epoch": 2.7424,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4285
    },
    {
      "epoch": 2.7456,
      "grad_norm": 5.626678466796875e-05,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 4290
    },
    {
      "epoch": 2.7488,
      "grad_norm": 0.000469207763671875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4295
    },
    {
      "epoch": 2.752,
      "grad_norm": 2.372264862060547e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4300
    },
    {
      "epoch": 2.7552,
      "grad_norm": 0.00115203857421875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4305
    },
    {
      "epoch": 2.7584,
      "grad_norm": 4.57763671875e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4310
    },
    {
      "epoch": 2.7616,
      "grad_norm": 0.0001354217529296875,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 4315
    },
    {
      "epoch": 2.7648,
      "grad_norm": 1.4185905456542969e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4320
    },
    {
      "epoch": 2.768,
      "grad_norm": 0.00012159347534179688,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4325
    },
    {
      "epoch": 2.7712,
      "grad_norm": 0.0087890625,
      "learning_rate": 0.0001,
      "loss": 0.0008,
      "step": 4330
    },
    {
      "epoch": 2.7744,
      "grad_norm": 0.00128936767578125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4335
    },
    {
      "epoch": 2.7776,
      "grad_norm": 7.0035457611083984e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4340
    },
    {
      "epoch": 2.7808,
      "grad_norm": 4.458427429199219e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4345
    },
    {
      "epoch": 2.784,
      "grad_norm": 1.4960765838623047e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4350
    },
    {
      "epoch": 2.7872,
      "grad_norm": 0.00145721435546875,
      "learning_rate": 0.0001,
      "loss": 0.0077,
      "step": 4355
    },
    {
      "epoch": 2.7904,
      "grad_norm": 1.2636184692382812e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4360
    },
    {
      "epoch": 2.7936,
      "grad_norm": 0.01092529296875,
      "learning_rate": 0.0001,
      "loss": 0.0019,
      "step": 4365
    },
    {
      "epoch": 2.7968,
      "grad_norm": 1.3828277587890625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4370
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.0002593994140625,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4375
    },
    {
      "epoch": 2.8032,
      "grad_norm": 8.96453857421875e-05,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 4380
    },
    {
      "epoch": 2.8064,
      "grad_norm": 0.00038909912109375,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4385
    },
    {
      "epoch": 2.8096,
      "grad_norm": 1.7404556274414062e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4390
    },
    {
      "epoch": 2.8128,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 4395
    },
    {
      "epoch": 2.816,
      "grad_norm": 1.9788742065429688e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4400
    },
    {
      "epoch": 2.8192,
      "grad_norm": 0.000514984130859375,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4405
    },
    {
      "epoch": 2.8224,
      "grad_norm": 3.981590270996094e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4410
    },
    {
      "epoch": 2.8256,
      "grad_norm": 1.3172626495361328e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4415
    },
    {
      "epoch": 2.8288,
      "grad_norm": 1.0192394256591797e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4420
    },
    {
      "epoch": 2.832,
      "grad_norm": 0.00017070770263671875,
      "learning_rate": 0.0001,
      "loss": 0.0003,
      "step": 4425
    },
    {
      "epoch": 2.8352,
      "grad_norm": 0.0004100799560546875,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 4430
    },
    {
      "epoch": 2.8384,
      "grad_norm": 0.011962890625,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 4435
    },
    {
      "epoch": 2.8416,
      "grad_norm": 0.000797271728515625,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 4440
    },
    {
      "epoch": 2.8448,
      "grad_norm": 4.00543212890625e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4445
    },
    {
      "epoch": 2.848,
      "grad_norm": 5.513429641723633e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4450
    },
    {
      "epoch": 2.8512,
      "grad_norm": 2.0265579223632812e-05,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 4455
    },
    {
      "epoch": 2.8544,
      "grad_norm": 0.00010013580322265625,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4460
    },
    {
      "epoch": 2.8576,
      "grad_norm": 0.00189971923828125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4465
    },
    {
      "epoch": 2.8608000000000002,
      "grad_norm": 1.9311904907226562e-05,
      "learning_rate": 0.0001,
      "loss": 0.0008,
      "step": 4470
    },
    {
      "epoch": 2.864,
      "grad_norm": 0.0150146484375,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4475
    },
    {
      "epoch": 2.8672,
      "grad_norm": 0.000732421875,
      "learning_rate": 0.0001,
      "loss": 0.0009,
      "step": 4480
    },
    {
      "epoch": 2.8704,
      "grad_norm": 0.00018596649169921875,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 4485
    },
    {
      "epoch": 2.8736,
      "grad_norm": 0.00049591064453125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4490
    },
    {
      "epoch": 2.8768000000000002,
      "grad_norm": 0.0032501220703125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4495
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.055002212524414e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4500
    },
    {
      "epoch": 2.8832,
      "grad_norm": 6.151199340820312e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4505
    },
    {
      "epoch": 2.8864,
      "grad_norm": 0.0003452301025390625,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4510
    },
    {
      "epoch": 2.8895999999999997,
      "grad_norm": 2.9325485229492188e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4515
    },
    {
      "epoch": 2.8928000000000003,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.0001,
      "loss": 0.0019,
      "step": 4520
    },
    {
      "epoch": 2.896,
      "grad_norm": 0.00994873046875,
      "learning_rate": 0.0001,
      "loss": 0.001,
      "step": 4525
    },
    {
      "epoch": 2.8992,
      "grad_norm": 0.0004177093505859375,
      "learning_rate": 0.0001,
      "loss": 0.0014,
      "step": 4530
    },
    {
      "epoch": 2.9024,
      "grad_norm": 0.000476837158203125,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 4535
    },
    {
      "epoch": 2.9055999999999997,
      "grad_norm": 0.0004215240478515625,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 4540
    },
    {
      "epoch": 2.9088000000000003,
      "grad_norm": 0.000247955322265625,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4545
    },
    {
      "epoch": 2.912,
      "grad_norm": 0.0040283203125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4550
    },
    {
      "epoch": 2.9152,
      "grad_norm": 0.00124359130859375,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4555
    },
    {
      "epoch": 2.9184,
      "grad_norm": 4.6253204345703125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4560
    },
    {
      "epoch": 2.9215999999999998,
      "grad_norm": 0.0004825592041015625,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4565
    },
    {
      "epoch": 2.9248,
      "grad_norm": 0.0002574920654296875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4570
    },
    {
      "epoch": 2.928,
      "grad_norm": 1.9431114196777344e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4575
    },
    {
      "epoch": 2.9312,
      "grad_norm": 0.003387451171875,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 4580
    },
    {
      "epoch": 2.9344,
      "grad_norm": 0.00160980224609375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4585
    },
    {
      "epoch": 2.9375999999999998,
      "grad_norm": 2.5987625122070312e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4590
    },
    {
      "epoch": 2.9408,
      "grad_norm": 7.009506225585938e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4595
    },
    {
      "epoch": 2.944,
      "grad_norm": 1.341104507446289e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4600
    },
    {
      "epoch": 2.9472,
      "grad_norm": 9.393692016601562e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4605
    },
    {
      "epoch": 2.9504,
      "grad_norm": 1.919269561767578e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4610
    },
    {
      "epoch": 2.9536,
      "grad_norm": 4.76837158203125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4615
    },
    {
      "epoch": 2.9568,
      "grad_norm": 9.417533874511719e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4620
    },
    {
      "epoch": 2.96,
      "grad_norm": 1.1980533599853516e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4625
    },
    {
      "epoch": 2.9632,
      "grad_norm": 0.00116729736328125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4630
    },
    {
      "epoch": 2.9664,
      "grad_norm": 0.0016632080078125,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4635
    },
    {
      "epoch": 2.9696,
      "grad_norm": 0.001373291015625,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4640
    },
    {
      "epoch": 2.9728,
      "grad_norm": 0.00014972686767578125,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 4645
    },
    {
      "epoch": 2.976,
      "grad_norm": 5.125999450683594e-06,
      "learning_rate": 0.0001,
      "loss": 0.0005,
      "step": 4650
    },
    {
      "epoch": 2.9792,
      "grad_norm": 6.67572021484375e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4655
    },
    {
      "epoch": 2.9824,
      "grad_norm": 1.1086463928222656e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4660
    },
    {
      "epoch": 2.9856,
      "grad_norm": 0.0002803802490234375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4665
    },
    {
      "epoch": 2.9888,
      "grad_norm": 3.218650817871094e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4670
    },
    {
      "epoch": 2.992,
      "grad_norm": 1.1086463928222656e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4675
    },
    {
      "epoch": 2.9952,
      "grad_norm": 1.6689300537109375e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4680
    },
    {
      "epoch": 2.9984,
      "grad_norm": 9.47713851928711e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4685
    },
    {
      "epoch": 3.0016,
      "grad_norm": 0.005615234375,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 4690
    },
    {
      "epoch": 3.0048,
      "grad_norm": 0.00191497802734375,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 4695
    },
    {
      "epoch": 3.008,
      "grad_norm": 0.0022125244140625,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4700
    },
    {
      "epoch": 3.0112,
      "grad_norm": 0.000110626220703125,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4705
    },
    {
      "epoch": 3.0144,
      "grad_norm": 2.1576881408691406e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4710
    },
    {
      "epoch": 3.0176,
      "grad_norm": 4.172325134277344e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4715
    },
    {
      "epoch": 3.0208,
      "grad_norm": 0.000225067138671875,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4720
    },
    {
      "epoch": 3.024,
      "grad_norm": 1.341104507446289e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4725
    },
    {
      "epoch": 3.0272,
      "grad_norm": 8.940696716308594e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4730
    },
    {
      "epoch": 3.0304,
      "grad_norm": 7.772445678710938e-05,
      "learning_rate": 0.0001,
      "loss": 0.0009,
      "step": 4735
    },
    {
      "epoch": 3.0336,
      "grad_norm": 5.316734313964844e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4740
    },
    {
      "epoch": 3.0368,
      "grad_norm": 2.2172927856445312e-05,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 4745
    },
    {
      "epoch": 3.04,
      "grad_norm": 0.00162506103515625,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4750
    },
    {
      "epoch": 3.0432,
      "grad_norm": 3.528594970703125e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4755
    },
    {
      "epoch": 3.0464,
      "grad_norm": 1.2099742889404297e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4760
    },
    {
      "epoch": 3.0496,
      "grad_norm": 0.003631591796875,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4765
    },
    {
      "epoch": 3.0528,
      "grad_norm": 2.7298927307128906e-05,
      "learning_rate": 0.0001,
      "loss": 0.0001,
      "step": 4770
    },
    {
      "epoch": 3.056,
      "grad_norm": 3.266334533691406e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4775
    },
    {
      "epoch": 3.0592,
      "grad_norm": 7.063150405883789e-06,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4780
    },
    {
      "epoch": 3.0624,
      "grad_norm": 1.2516975402832031e-05,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4785
    },
    {
      "epoch": 3.0656,
      "grad_norm": 0.0146484375,
      "learning_rate": 0.0001,
      "loss": 0.0009,
      "step": 4790
    },
    {
      "epoch": 3.0688,
      "grad_norm": 0.00020503997802734375,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 4795
    },
    {
      "epoch": 3.072,
      "grad_norm": 0.00091552734375,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 4800
    }
  ],
  "logging_steps": 5,
  "max_steps": 4800,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 90,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.883056636914381e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}