{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.691519635742743,
  "eval_steps": 500,
  "global_step": 10000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0002,
      "loss": 1.495,
      "step": 10
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.2119140625,
      "learning_rate": 0.0002,
      "loss": 1.3828,
      "step": 20
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.2373046875,
      "learning_rate": 0.0002,
      "loss": 1.2271,
      "step": 30
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.28515625,
      "learning_rate": 0.0002,
      "loss": 1.1604,
      "step": 40
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.2890625,
      "learning_rate": 0.0002,
      "loss": 1.099,
      "step": 50
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.1904296875,
      "learning_rate": 0.0002,
      "loss": 1.2541,
      "step": 60
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.2060546875,
      "learning_rate": 0.0002,
      "loss": 1.2232,
      "step": 70
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 1.229,
      "step": 80
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.2890625,
      "learning_rate": 0.0002,
      "loss": 1.1292,
      "step": 90
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.287109375,
      "learning_rate": 0.0002,
      "loss": 1.0562,
      "step": 100
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.208984375,
      "learning_rate": 0.0002,
      "loss": 1.1468,
      "step": 110
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.21875,
      "learning_rate": 0.0002,
      "loss": 1.1161,
      "step": 120
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.212890625,
      "learning_rate": 0.0002,
      "loss": 1.1258,
      "step": 130
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.279296875,
      "learning_rate": 0.0002,
      "loss": 1.0508,
      "step": 140
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.271484375,
      "learning_rate": 0.0002,
      "loss": 1.0176,
      "step": 150
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.2138671875,
      "learning_rate": 0.0002,
      "loss": 1.1983,
      "step": 160
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.2353515625,
      "learning_rate": 0.0002,
      "loss": 1.1958,
      "step": 170
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.236328125,
      "learning_rate": 0.0002,
      "loss": 1.125,
      "step": 180
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 1.0656,
      "step": 190
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.28125,
      "learning_rate": 0.0002,
      "loss": 0.9628,
      "step": 200
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.2021484375,
      "learning_rate": 0.0002,
      "loss": 1.0976,
      "step": 210
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.2373046875,
      "learning_rate": 0.0002,
      "loss": 1.1465,
      "step": 220
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.21875,
      "learning_rate": 0.0002,
      "loss": 1.0914,
      "step": 230
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.27734375,
      "learning_rate": 0.0002,
      "loss": 1.0155,
      "step": 240
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.291015625,
      "learning_rate": 0.0002,
      "loss": 0.9932,
      "step": 250
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.1982421875,
      "learning_rate": 0.0002,
      "loss": 1.1324,
      "step": 260
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.212890625,
      "learning_rate": 0.0002,
      "loss": 1.1367,
      "step": 270
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.248046875,
      "learning_rate": 0.0002,
      "loss": 1.0788,
      "step": 280
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.2470703125,
      "learning_rate": 0.0002,
      "loss": 1.051,
      "step": 290
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.26171875,
      "learning_rate": 0.0002,
      "loss": 1.0043,
      "step": 300
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.1982421875,
      "learning_rate": 0.0002,
      "loss": 1.0812,
      "step": 310
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.2294921875,
      "learning_rate": 0.0002,
      "loss": 1.0901,
      "step": 320
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.234375,
      "learning_rate": 0.0002,
      "loss": 1.0626,
      "step": 330
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.24609375,
      "learning_rate": 0.0002,
      "loss": 0.9682,
      "step": 340
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.326171875,
      "learning_rate": 0.0002,
      "loss": 0.9774,
      "step": 350
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.203125,
      "learning_rate": 0.0002,
      "loss": 1.1245,
      "step": 360
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.224609375,
      "learning_rate": 0.0002,
      "loss": 1.1601,
      "step": 370
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.2470703125,
      "learning_rate": 0.0002,
      "loss": 1.0697,
      "step": 380
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.2421875,
      "learning_rate": 0.0002,
      "loss": 0.993,
      "step": 390
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.287109375,
      "learning_rate": 0.0002,
      "loss": 0.9591,
      "step": 400
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.2216796875,
      "learning_rate": 0.0002,
      "loss": 1.1005,
      "step": 410
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.2294921875,
      "learning_rate": 0.0002,
      "loss": 1.0966,
      "step": 420
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.236328125,
      "learning_rate": 0.0002,
      "loss": 1.0549,
      "step": 430
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.265625,
      "learning_rate": 0.0002,
      "loss": 1.0008,
      "step": 440
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.3046875,
      "learning_rate": 0.0002,
      "loss": 0.9528,
      "step": 450
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.216796875,
      "learning_rate": 0.0002,
      "loss": 1.0955,
      "step": 460
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.236328125,
      "learning_rate": 0.0002,
      "loss": 1.0835,
      "step": 470
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.2333984375,
      "learning_rate": 0.0002,
      "loss": 1.0694,
      "step": 480
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.2890625,
      "learning_rate": 0.0002,
      "loss": 1.0206,
      "step": 490
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.3046875,
      "learning_rate": 0.0002,
      "loss": 0.9314,
      "step": 500
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.197265625,
      "learning_rate": 0.0002,
      "loss": 1.1097,
      "step": 510
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.2412109375,
      "learning_rate": 0.0002,
      "loss": 1.0696,
      "step": 520
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.2353515625,
      "learning_rate": 0.0002,
      "loss": 1.0366,
      "step": 530
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.259765625,
      "learning_rate": 0.0002,
      "loss": 1.0058,
      "step": 540
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.287109375,
      "learning_rate": 0.0002,
      "loss": 0.9348,
      "step": 550
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.22265625,
      "learning_rate": 0.0002,
      "loss": 1.0774,
      "step": 560
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.2294921875,
      "learning_rate": 0.0002,
      "loss": 1.0795,
      "step": 570
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.2373046875,
      "learning_rate": 0.0002,
      "loss": 1.0754,
      "step": 580
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0002,
      "loss": 0.997,
      "step": 590
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.3125,
      "learning_rate": 0.0002,
      "loss": 0.9705,
      "step": 600
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.2265625,
      "learning_rate": 0.0002,
      "loss": 1.1025,
      "step": 610
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.2265625,
      "learning_rate": 0.0002,
      "loss": 1.0675,
      "step": 620
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.224609375,
      "learning_rate": 0.0002,
      "loss": 1.0343,
      "step": 630
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 0.9763,
      "step": 640
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.28125,
      "learning_rate": 0.0002,
      "loss": 0.9035,
      "step": 650
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.21484375,
      "learning_rate": 0.0002,
      "loss": 1.0758,
      "step": 660
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.23828125,
      "learning_rate": 0.0002,
      "loss": 1.0746,
      "step": 670
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.2333984375,
      "learning_rate": 0.0002,
      "loss": 1.0232,
      "step": 680
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.9997,
      "step": 690
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.30078125,
      "learning_rate": 0.0002,
      "loss": 0.953,
      "step": 700
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.2109375,
      "learning_rate": 0.0002,
      "loss": 1.1189,
      "step": 710
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.2353515625,
      "learning_rate": 0.0002,
      "loss": 1.0821,
      "step": 720
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.2333984375,
      "learning_rate": 0.0002,
      "loss": 0.9964,
      "step": 730
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.2451171875,
      "learning_rate": 0.0002,
      "loss": 0.9904,
      "step": 740
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.294921875,
      "learning_rate": 0.0002,
      "loss": 0.9861,
      "step": 750
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.197265625,
      "learning_rate": 0.0002,
      "loss": 1.0426,
      "step": 760
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.23046875,
      "learning_rate": 0.0002,
      "loss": 1.1045,
      "step": 770
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.232421875,
      "learning_rate": 0.0002,
      "loss": 1.0563,
      "step": 780
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.2490234375,
      "learning_rate": 0.0002,
      "loss": 0.9486,
      "step": 790
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.337890625,
      "learning_rate": 0.0002,
      "loss": 0.8978,
      "step": 800
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.21875,
      "learning_rate": 0.0002,
      "loss": 1.038,
      "step": 810
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.228515625,
      "learning_rate": 0.0002,
      "loss": 1.0695,
      "step": 820
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.2275390625,
      "learning_rate": 0.0002,
      "loss": 0.9979,
      "step": 830
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.9699,
      "step": 840
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.26953125,
      "learning_rate": 0.0002,
      "loss": 0.928,
      "step": 850
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.205078125,
      "learning_rate": 0.0002,
      "loss": 1.0532,
      "step": 860
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.2158203125,
      "learning_rate": 0.0002,
      "loss": 1.0447,
      "step": 870
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.228515625,
      "learning_rate": 0.0002,
      "loss": 1.0093,
      "step": 880
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.9471,
      "step": 890
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.296875,
      "learning_rate": 0.0002,
      "loss": 0.9345,
      "step": 900
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.22265625,
      "learning_rate": 0.0002,
      "loss": 1.0507,
      "step": 910
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.2255859375,
      "learning_rate": 0.0002,
      "loss": 1.0093,
      "step": 920
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.23046875,
      "learning_rate": 0.0002,
      "loss": 1.0098,
      "step": 930
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.9844,
      "step": 940
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.291015625,
      "learning_rate": 0.0002,
      "loss": 0.9225,
      "step": 950
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.2021484375,
      "learning_rate": 0.0002,
      "loss": 1.0379,
      "step": 960
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.22265625,
      "learning_rate": 0.0002,
      "loss": 1.0356,
      "step": 970
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.2236328125,
      "learning_rate": 0.0002,
      "loss": 1.0603,
      "step": 980
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0002,
      "loss": 0.9772,
      "step": 990
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.28515625,
      "learning_rate": 0.0002,
      "loss": 0.9391,
      "step": 1000
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.201171875,
      "learning_rate": 0.0002,
      "loss": 1.0791,
      "step": 1010
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.2255859375,
      "learning_rate": 0.0002,
      "loss": 1.0449,
      "step": 1020
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.2333984375,
      "learning_rate": 0.0002,
      "loss": 1.0394,
      "step": 1030
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0002,
      "loss": 0.9459,
      "step": 1040
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.310546875,
      "learning_rate": 0.0002,
      "loss": 0.9259,
      "step": 1050
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.22265625,
      "learning_rate": 0.0002,
      "loss": 0.9938,
      "step": 1060
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.23828125,
      "learning_rate": 0.0002,
      "loss": 1.0291,
      "step": 1070
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.2451171875,
      "learning_rate": 0.0002,
      "loss": 1.0005,
      "step": 1080
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.23046875,
      "learning_rate": 0.0002,
      "loss": 0.9573,
      "step": 1090
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.28125,
      "learning_rate": 0.0002,
      "loss": 0.9081,
      "step": 1100
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.208984375,
      "learning_rate": 0.0002,
      "loss": 1.0571,
      "step": 1110
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.224609375,
      "learning_rate": 0.0002,
      "loss": 1.0457,
      "step": 1120
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.2412109375,
      "learning_rate": 0.0002,
      "loss": 1.025,
      "step": 1130
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0002,
      "loss": 0.9688,
      "step": 1140
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.265625,
      "learning_rate": 0.0002,
      "loss": 0.9028,
      "step": 1150
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.21484375,
      "learning_rate": 0.0002,
      "loss": 1.0761,
      "step": 1160
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.328125,
      "learning_rate": 0.0002,
      "loss": 1.042,
      "step": 1170
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.2470703125,
      "learning_rate": 0.0002,
      "loss": 0.9584,
      "step": 1180
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.2470703125,
      "learning_rate": 0.0002,
      "loss": 0.9269,
      "step": 1190
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.28515625,
      "learning_rate": 0.0002,
      "loss": 0.8703,
      "step": 1200
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.2109375,
      "learning_rate": 0.0002,
      "loss": 1.0499,
      "step": 1210
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.999,
      "step": 1220
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.232421875,
      "learning_rate": 0.0002,
      "loss": 0.9965,
      "step": 1230
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.267578125,
      "learning_rate": 0.0002,
      "loss": 0.937,
      "step": 1240
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.291015625,
      "learning_rate": 0.0002,
      "loss": 0.9117,
      "step": 1250
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.2275390625,
      "learning_rate": 0.0002,
      "loss": 1.0298,
      "step": 1260
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.2197265625,
      "learning_rate": 0.0002,
      "loss": 1.0433,
      "step": 1270
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.240234375,
      "learning_rate": 0.0002,
      "loss": 0.9567,
      "step": 1280
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.265625,
      "learning_rate": 0.0002,
      "loss": 0.9681,
      "step": 1290
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.28125,
      "learning_rate": 0.0002,
      "loss": 0.8989,
      "step": 1300
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.240234375,
      "learning_rate": 0.0002,
      "loss": 1.0711,
      "step": 1310
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.228515625,
      "learning_rate": 0.0002,
      "loss": 1.0023,
      "step": 1320
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.2373046875,
      "learning_rate": 0.0002,
      "loss": 0.9999,
      "step": 1330
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0002,
      "loss": 0.9522,
      "step": 1340
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.2734375,
      "learning_rate": 0.0002,
      "loss": 0.8813,
      "step": 1350
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.2275390625,
      "learning_rate": 0.0002,
      "loss": 1.0647,
      "step": 1360
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.24609375,
      "learning_rate": 0.0002,
      "loss": 1.0166,
      "step": 1370
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.224609375,
      "learning_rate": 0.0002,
      "loss": 0.9722,
      "step": 1380
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.9438,
      "step": 1390
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.2734375,
      "learning_rate": 0.0002,
      "loss": 0.9141,
      "step": 1400
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.220703125,
      "learning_rate": 0.0002,
      "loss": 1.0024,
      "step": 1410
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.244140625,
      "learning_rate": 0.0002,
      "loss": 0.9796,
      "step": 1420
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.236328125,
      "learning_rate": 0.0002,
      "loss": 1.025,
      "step": 1430
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.2451171875,
      "learning_rate": 0.0002,
      "loss": 0.9433,
      "step": 1440
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 0.8746,
      "step": 1450
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.2138671875,
      "learning_rate": 0.0002,
      "loss": 0.9818,
      "step": 1460
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.2314453125,
      "learning_rate": 0.0002,
      "loss": 1.0507,
      "step": 1470
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.2353515625,
      "learning_rate": 0.0002,
      "loss": 0.9784,
      "step": 1480
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.2421875,
      "learning_rate": 0.0002,
      "loss": 0.9124,
      "step": 1490
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.296875,
      "learning_rate": 0.0002,
      "loss": 0.895,
      "step": 1500
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.2158203125,
      "learning_rate": 0.0002,
      "loss": 1.0743,
      "step": 1510
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.2197265625,
      "learning_rate": 0.0002,
      "loss": 1.0102,
      "step": 1520
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.2197265625,
      "learning_rate": 0.0002,
      "loss": 0.9876,
      "step": 1530
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9359,
      "step": 1540
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0002,
      "loss": 0.8637,
      "step": 1550
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.2451171875,
      "learning_rate": 0.0002,
      "loss": 1.0772,
      "step": 1560
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.2373046875,
      "learning_rate": 0.0002,
      "loss": 1.0055,
      "step": 1570
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.2412109375,
      "learning_rate": 0.0002,
      "loss": 0.9339,
      "step": 1580
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.267578125,
      "learning_rate": 0.0002,
      "loss": 0.9126,
      "step": 1590
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.279296875,
      "learning_rate": 0.0002,
      "loss": 0.8885,
      "step": 1600
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.2080078125,
      "learning_rate": 0.0002,
      "loss": 1.0688,
      "step": 1610
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.21875,
      "learning_rate": 0.0002,
      "loss": 1.0258,
      "step": 1620
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.228515625,
      "learning_rate": 0.0002,
      "loss": 0.986,
      "step": 1630
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 0.9506,
      "step": 1640
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 0.9254,
      "step": 1650
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.21484375,
      "learning_rate": 0.0002,
      "loss": 1.0348,
      "step": 1660
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.23046875,
      "learning_rate": 0.0002,
      "loss": 1.0086,
      "step": 1670
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.234375,
      "learning_rate": 0.0002,
      "loss": 0.9324,
      "step": 1680
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.87,
      "step": 1690
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.267578125,
      "learning_rate": 0.0002,
      "loss": 0.8989,
      "step": 1700
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.2177734375,
      "learning_rate": 0.0002,
      "loss": 1.0533,
      "step": 1710
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.2060546875,
      "learning_rate": 0.0002,
      "loss": 1.0415,
      "step": 1720
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.2314453125,
      "learning_rate": 0.0002,
      "loss": 1.024,
      "step": 1730
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.263671875,
      "learning_rate": 0.0002,
      "loss": 0.9463,
      "step": 1740
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.291015625,
      "learning_rate": 0.0002,
      "loss": 0.8608,
      "step": 1750
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.2041015625,
      "learning_rate": 0.0002,
      "loss": 1.0173,
      "step": 1760
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.2265625,
      "learning_rate": 0.0002,
      "loss": 1.0341,
      "step": 1770
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.21875,
      "learning_rate": 0.0002,
      "loss": 1.012,
      "step": 1780
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.240234375,
      "learning_rate": 0.0002,
      "loss": 0.9792,
      "step": 1790
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.2578125,
      "learning_rate": 0.0002,
      "loss": 0.8659,
      "step": 1800
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.2275390625,
      "learning_rate": 0.0002,
      "loss": 0.9118,
      "step": 1810
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.2236328125,
      "learning_rate": 0.0002,
      "loss": 0.9863,
      "step": 1820
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.2431640625,
      "learning_rate": 0.0002,
      "loss": 1.0178,
      "step": 1830
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.234375,
      "learning_rate": 0.0002,
      "loss": 0.9463,
      "step": 1840
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.2578125,
      "learning_rate": 0.0002,
      "loss": 0.8579,
      "step": 1850
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.2265625,
      "learning_rate": 0.0002,
      "loss": 0.8987,
      "step": 1860
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.2138671875,
      "learning_rate": 0.0002,
      "loss": 1.0184,
      "step": 1870
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.23828125,
      "learning_rate": 0.0002,
      "loss": 1.0167,
      "step": 1880
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.2412109375,
      "learning_rate": 0.0002,
      "loss": 0.9397,
      "step": 1890
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.2734375,
      "learning_rate": 0.0002,
      "loss": 0.8526,
      "step": 1900
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.21875,
      "learning_rate": 0.0002,
      "loss": 0.9324,
      "step": 1910
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.2216796875,
      "learning_rate": 0.0002,
      "loss": 1.0039,
      "step": 1920
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.240234375,
      "learning_rate": 0.0002,
      "loss": 0.9614,
      "step": 1930
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.2470703125,
      "learning_rate": 0.0002,
      "loss": 0.9676,
      "step": 1940
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.267578125,
      "learning_rate": 0.0002,
      "loss": 0.8858,
      "step": 1950
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.205078125,
      "learning_rate": 0.0002,
      "loss": 0.9097,
      "step": 1960
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.2236328125,
      "learning_rate": 0.0002,
      "loss": 1.0206,
      "step": 1970
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.2578125,
      "learning_rate": 0.0002,
      "loss": 0.9845,
      "step": 1980
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.2314453125,
      "learning_rate": 0.0002,
      "loss": 0.9402,
      "step": 1990
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.283203125,
      "learning_rate": 0.0002,
      "loss": 0.8963,
      "step": 2000
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 0.8848,
      "step": 2010
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.234375,
      "learning_rate": 0.0002,
      "loss": 0.9937,
      "step": 2020
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.2490234375,
      "learning_rate": 0.0002,
      "loss": 0.9398,
      "step": 2030
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.26953125,
      "learning_rate": 0.0002,
      "loss": 0.9064,
      "step": 2040
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0002,
      "loss": 0.857,
      "step": 2050
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.216796875,
      "learning_rate": 0.0002,
      "loss": 0.8838,
      "step": 2060
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.23828125,
      "learning_rate": 0.0002,
      "loss": 1.0213,
      "step": 2070
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.2333984375,
      "learning_rate": 0.0002,
      "loss": 0.9989,
      "step": 2080
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9823,
      "step": 2090
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.2470703125,
      "learning_rate": 0.0002,
      "loss": 0.8393,
      "step": 2100
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.232421875,
      "learning_rate": 0.0002,
      "loss": 0.8814,
      "step": 2110
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.21875,
      "learning_rate": 0.0002,
      "loss": 1.0063,
      "step": 2120
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 1.0122,
      "step": 2130
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.2412109375,
      "learning_rate": 0.0002,
      "loss": 0.9377,
      "step": 2140
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.283203125,
      "learning_rate": 0.0002,
      "loss": 0.8951,
      "step": 2150
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.2431640625,
      "learning_rate": 0.0002,
      "loss": 0.88,
      "step": 2160
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.2431640625,
      "learning_rate": 0.0002,
      "loss": 1.0067,
      "step": 2170
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.248046875,
      "learning_rate": 0.0002,
      "loss": 0.9457,
      "step": 2180
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.2353515625,
      "learning_rate": 0.0002,
      "loss": 0.9101,
      "step": 2190
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.259765625,
      "learning_rate": 0.0002,
      "loss": 0.8667,
      "step": 2200
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.2470703125,
      "learning_rate": 0.0002,
      "loss": 0.8822,
      "step": 2210
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.23046875,
      "learning_rate": 0.0002,
      "loss": 1.0005,
      "step": 2220
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.2451171875,
      "learning_rate": 0.0002,
      "loss": 0.965,
      "step": 2230
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.267578125,
      "learning_rate": 0.0002,
      "loss": 0.9537,
      "step": 2240
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.26171875,
      "learning_rate": 0.0002,
      "loss": 0.8393,
      "step": 2250
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.236328125,
      "learning_rate": 0.0002,
      "loss": 0.9028,
      "step": 2260
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.2099609375,
      "learning_rate": 0.0002,
      "loss": 0.9456,
      "step": 2270
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.228515625,
      "learning_rate": 0.0002,
      "loss": 0.9766,
      "step": 2280
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.9445,
      "step": 2290
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0002,
      "loss": 0.8368,
      "step": 2300
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.2421875,
      "learning_rate": 0.0002,
      "loss": 0.8603,
      "step": 2310
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.2412109375,
      "learning_rate": 0.0002,
      "loss": 1.0516,
      "step": 2320
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.9612,
      "step": 2330
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.244140625,
      "learning_rate": 0.0002,
      "loss": 0.9333,
      "step": 2340
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.26171875,
      "learning_rate": 0.0002,
      "loss": 0.8959,
      "step": 2350
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.220703125,
      "learning_rate": 0.0002,
      "loss": 0.8876,
      "step": 2360
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.2314453125,
      "learning_rate": 0.0002,
      "loss": 0.9755,
      "step": 2370
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.232421875,
      "learning_rate": 0.0002,
      "loss": 0.984,
      "step": 2380
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.248046875,
      "learning_rate": 0.0002,
      "loss": 0.9038,
      "step": 2390
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.267578125,
      "learning_rate": 0.0002,
      "loss": 0.8772,
      "step": 2400
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.240234375,
      "learning_rate": 0.0002,
      "loss": 0.8802,
      "step": 2410
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.23828125,
      "learning_rate": 0.0002,
      "loss": 0.9951,
      "step": 2420
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.2451171875,
      "learning_rate": 0.0002,
      "loss": 0.9761,
      "step": 2430
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.259765625,
      "learning_rate": 0.0002,
      "loss": 0.9057,
      "step": 2440
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.2734375,
      "learning_rate": 0.0002,
      "loss": 0.8993,
      "step": 2450
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.2294921875,
      "learning_rate": 0.0002,
      "loss": 0.8513,
      "step": 2460
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.2421875,
      "learning_rate": 0.0002,
      "loss": 0.9897,
      "step": 2470
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.2333984375,
      "learning_rate": 0.0002,
      "loss": 0.985,
      "step": 2480
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.2431640625,
      "learning_rate": 0.0002,
      "loss": 0.9435,
      "step": 2490
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.3046875,
      "learning_rate": 0.0002,
      "loss": 0.866,
      "step": 2500
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.8878,
      "step": 2510
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.2109375,
      "learning_rate": 0.0002,
      "loss": 0.9561,
      "step": 2520
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.267578125,
      "learning_rate": 0.0002,
      "loss": 0.9653,
      "step": 2530
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.8876,
      "step": 2540
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.259765625,
      "learning_rate": 0.0002,
      "loss": 0.8824,
      "step": 2550
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 0.8765,
      "step": 2560
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.2421875,
      "learning_rate": 0.0002,
      "loss": 0.9614,
      "step": 2570
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0002,
      "loss": 0.9814,
      "step": 2580
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 0.8922,
      "step": 2590
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.3046875,
      "learning_rate": 0.0002,
      "loss": 0.8736,
      "step": 2600
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.2451171875,
      "learning_rate": 0.0002,
      "loss": 0.8832,
      "step": 2610
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.220703125,
      "learning_rate": 0.0002,
      "loss": 0.9943,
      "step": 2620
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.2265625,
      "learning_rate": 0.0002,
      "loss": 0.9602,
      "step": 2630
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.936,
      "step": 2640
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.27734375,
      "learning_rate": 0.0002,
      "loss": 0.8718,
      "step": 2650
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.234375,
      "learning_rate": 0.0002,
      "loss": 0.9041,
      "step": 2660
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.2236328125,
      "learning_rate": 0.0002,
      "loss": 0.9741,
      "step": 2670
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.2314453125,
      "learning_rate": 0.0002,
      "loss": 0.9613,
      "step": 2680
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.2373046875,
      "learning_rate": 0.0002,
      "loss": 0.9412,
      "step": 2690
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.2578125,
      "learning_rate": 0.0002,
      "loss": 0.907,
      "step": 2700
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.263671875,
      "learning_rate": 0.0002,
      "loss": 0.8255,
      "step": 2710
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.2392578125,
      "learning_rate": 0.0002,
      "loss": 0.9947,
      "step": 2720
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.259765625,
      "learning_rate": 0.0002,
      "loss": 0.9775,
      "step": 2730
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.2373046875,
      "learning_rate": 0.0002,
      "loss": 0.9293,
      "step": 2740
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.271484375,
      "learning_rate": 0.0002,
      "loss": 0.877,
      "step": 2750
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.212890625,
      "learning_rate": 0.0002,
      "loss": 0.9085,
      "step": 2760
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.21875,
      "learning_rate": 0.0002,
      "loss": 0.975,
      "step": 2770
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.2470703125,
      "learning_rate": 0.0002,
      "loss": 0.9957,
      "step": 2780
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.9567,
      "step": 2790
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.271484375,
      "learning_rate": 0.0002,
      "loss": 0.8587,
      "step": 2800
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 0.8616,
      "step": 2810
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 1.0032,
      "step": 2820
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.2470703125,
      "learning_rate": 0.0002,
      "loss": 0.9001,
      "step": 2830
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.2373046875,
      "learning_rate": 0.0002,
      "loss": 0.9542,
      "step": 2840
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.26953125,
      "learning_rate": 0.0002,
      "loss": 0.8459,
      "step": 2850
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.2392578125,
      "learning_rate": 0.0002,
      "loss": 0.8761,
      "step": 2860
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.22265625,
      "learning_rate": 0.0002,
      "loss": 1.0039,
      "step": 2870
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.2412109375,
      "learning_rate": 0.0002,
      "loss": 0.9821,
      "step": 2880
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.9451,
      "step": 2890
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0002,
      "loss": 0.887,
      "step": 2900
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.2333984375,
      "learning_rate": 0.0002,
      "loss": 0.9211,
      "step": 2910
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.232421875,
      "learning_rate": 0.0002,
      "loss": 0.9841,
      "step": 2920
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.234375,
      "learning_rate": 0.0002,
      "loss": 0.9731,
      "step": 2930
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.240234375,
      "learning_rate": 0.0002,
      "loss": 0.9368,
      "step": 2940
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.265625,
      "learning_rate": 0.0002,
      "loss": 0.8407,
      "step": 2950
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.234375,
      "learning_rate": 0.0002,
      "loss": 0.878,
      "step": 2960
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.2255859375,
      "learning_rate": 0.0002,
      "loss": 1.0246,
      "step": 2970
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.2431640625,
      "learning_rate": 0.0002,
      "loss": 0.9817,
      "step": 2980
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.2470703125,
      "learning_rate": 0.0002,
      "loss": 0.9605,
      "step": 2990
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0002,
      "loss": 0.8364,
      "step": 3000
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.2255859375,
      "learning_rate": 0.0002,
      "loss": 0.865,
      "step": 3010
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.22265625,
      "learning_rate": 0.0002,
      "loss": 0.9692,
      "step": 3020
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.240234375,
      "learning_rate": 0.0002,
      "loss": 0.9697,
      "step": 3030
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.2578125,
      "learning_rate": 0.0002,
      "loss": 0.9166,
      "step": 3040
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.263671875,
      "learning_rate": 0.0002,
      "loss": 0.8529,
      "step": 3050
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.2265625,
      "learning_rate": 0.0002,
      "loss": 0.9194,
      "step": 3060
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.2314453125,
      "learning_rate": 0.0002,
      "loss": 0.9443,
      "step": 3070
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.2421875,
      "learning_rate": 0.0002,
      "loss": 0.9519,
      "step": 3080
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.2470703125,
      "learning_rate": 0.0002,
      "loss": 0.9066,
      "step": 3090
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.26171875,
      "learning_rate": 0.0002,
      "loss": 0.8582,
      "step": 3100
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.2333984375,
      "learning_rate": 0.0002,
      "loss": 0.9266,
      "step": 3110
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 0.9802,
      "step": 3120
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.228515625,
      "learning_rate": 0.0002,
      "loss": 0.912,
      "step": 3130
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 0.9689,
      "step": 3140
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.27734375,
      "learning_rate": 0.0002,
      "loss": 0.8624,
      "step": 3150
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.2451171875,
      "learning_rate": 0.0002,
      "loss": 0.8598,
      "step": 3160
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.2294921875,
      "learning_rate": 0.0002,
      "loss": 0.986,
      "step": 3170
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.236328125,
      "learning_rate": 0.0002,
      "loss": 0.9528,
      "step": 3180
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.24609375,
      "learning_rate": 0.0002,
      "loss": 0.9345,
      "step": 3190
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.259765625,
      "learning_rate": 0.0002,
      "loss": 0.8386,
      "step": 3200
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.2275390625,
      "learning_rate": 0.0002,
      "loss": 0.8852,
      "step": 3210
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.212890625,
      "learning_rate": 0.0002,
      "loss": 0.9578,
      "step": 3220
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.23828125,
      "learning_rate": 0.0002,
      "loss": 0.9477,
      "step": 3230
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.2490234375,
      "learning_rate": 0.0002,
      "loss": 0.9011,
      "step": 3240
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.259765625,
      "learning_rate": 0.0002,
      "loss": 0.8294,
      "step": 3250
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.2412109375,
      "learning_rate": 0.0002,
      "loss": 0.8807,
      "step": 3260
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.248046875,
      "learning_rate": 0.0002,
      "loss": 0.972,
      "step": 3270
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.23828125,
      "learning_rate": 0.0002,
      "loss": 0.9898,
      "step": 3280
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.9307,
      "step": 3290
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.27734375,
      "learning_rate": 0.0002,
      "loss": 0.8186,
      "step": 3300
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8804,
      "step": 3310
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.21875,
      "learning_rate": 0.0002,
      "loss": 0.9899,
      "step": 3320
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.24609375,
      "learning_rate": 0.0002,
      "loss": 0.9439,
      "step": 3330
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.23828125,
      "learning_rate": 0.0002,
      "loss": 0.8727,
      "step": 3340
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.283203125,
      "learning_rate": 0.0002,
      "loss": 0.862,
      "step": 3350
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8661,
      "step": 3360
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.228515625,
      "learning_rate": 0.0002,
      "loss": 0.9943,
      "step": 3370
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.2451171875,
      "learning_rate": 0.0002,
      "loss": 0.9868,
      "step": 3380
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.244140625,
      "learning_rate": 0.0002,
      "loss": 0.9108,
      "step": 3390
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.8446,
      "step": 3400
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.89,
      "step": 3410
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.232421875,
      "learning_rate": 0.0002,
      "loss": 0.9935,
      "step": 3420
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.2421875,
      "learning_rate": 0.0002,
      "loss": 0.966,
      "step": 3430
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.244140625,
      "learning_rate": 0.0002,
      "loss": 0.9391,
      "step": 3440
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.265625,
      "learning_rate": 0.0002,
      "loss": 0.8886,
      "step": 3450
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.265625,
      "learning_rate": 0.0002,
      "loss": 0.8634,
      "step": 3460
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.21484375,
      "learning_rate": 0.0002,
      "loss": 0.9927,
      "step": 3470
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.2412109375,
      "learning_rate": 0.0002,
      "loss": 0.9513,
      "step": 3480
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.2373046875,
      "learning_rate": 0.0002,
      "loss": 0.9113,
      "step": 3490
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.26953125,
      "learning_rate": 0.0002,
      "loss": 0.8498,
      "step": 3500
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0002,
      "loss": 0.8547,
      "step": 3510
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.212890625,
      "learning_rate": 0.0002,
      "loss": 0.9485,
      "step": 3520
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.24609375,
      "learning_rate": 0.0002,
      "loss": 0.9402,
      "step": 3530
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.9405,
      "step": 3540
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 0.863,
      "step": 3550
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.2734375,
      "learning_rate": 0.0002,
      "loss": 0.7701,
      "step": 3560
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.244140625,
      "learning_rate": 0.0002,
      "loss": 0.8538,
      "step": 3570
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.2470703125,
      "learning_rate": 0.0002,
      "loss": 0.9473,
      "step": 3580
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.9183,
      "step": 3590
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.24609375,
      "learning_rate": 0.0002,
      "loss": 0.8304,
      "step": 3600
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.3046875,
      "learning_rate": 0.0002,
      "loss": 0.791,
      "step": 3610
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0002,
      "loss": 0.8672,
      "step": 3620
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.2470703125,
      "learning_rate": 0.0002,
      "loss": 0.9531,
      "step": 3630
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.2431640625,
      "learning_rate": 0.0002,
      "loss": 0.9388,
      "step": 3640
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.26953125,
      "learning_rate": 0.0002,
      "loss": 0.8646,
      "step": 3650
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.265625,
      "learning_rate": 0.0002,
      "loss": 0.7959,
      "step": 3660
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.2373046875,
      "learning_rate": 0.0002,
      "loss": 0.8864,
      "step": 3670
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.2373046875,
      "learning_rate": 0.0002,
      "loss": 0.952,
      "step": 3680
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9314,
      "step": 3690
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.2734375,
      "learning_rate": 0.0002,
      "loss": 0.8895,
      "step": 3700
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.28515625,
      "learning_rate": 0.0002,
      "loss": 0.7767,
      "step": 3710
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.23046875,
      "learning_rate": 0.0002,
      "loss": 0.9263,
      "step": 3720
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.234375,
      "learning_rate": 0.0002,
      "loss": 0.9469,
      "step": 3730
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.2451171875,
      "learning_rate": 0.0002,
      "loss": 0.8629,
      "step": 3740
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.2734375,
      "learning_rate": 0.0002,
      "loss": 0.8327,
      "step": 3750
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.28515625,
      "learning_rate": 0.0002,
      "loss": 0.7878,
      "step": 3760
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.2265625,
      "learning_rate": 0.0002,
      "loss": 0.9102,
      "step": 3770
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 0.8878,
      "step": 3780
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.24609375,
      "learning_rate": 0.0002,
      "loss": 0.868,
      "step": 3790
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.26953125,
      "learning_rate": 0.0002,
      "loss": 0.8576,
      "step": 3800
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0002,
      "loss": 0.7992,
      "step": 3810
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.2421875,
      "learning_rate": 0.0002,
      "loss": 0.8624,
      "step": 3820
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.2421875,
      "learning_rate": 0.0002,
      "loss": 0.9591,
      "step": 3830
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.248046875,
      "learning_rate": 0.0002,
      "loss": 0.9186,
      "step": 3840
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.27734375,
      "learning_rate": 0.0002,
      "loss": 0.8621,
      "step": 3850
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.294921875,
      "learning_rate": 0.0002,
      "loss": 0.8437,
      "step": 3860
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.259765625,
      "learning_rate": 0.0002,
      "loss": 0.8686,
      "step": 3870
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.2333984375,
      "learning_rate": 0.0002,
      "loss": 0.9382,
      "step": 3880
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.2490234375,
      "learning_rate": 0.0002,
      "loss": 0.9485,
      "step": 3890
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.26171875,
      "learning_rate": 0.0002,
      "loss": 0.8593,
      "step": 3900
    },
    {
      "epoch": 2.23,
      "grad_norm": 0.27734375,
      "learning_rate": 0.0002,
      "loss": 0.7782,
      "step": 3910
    },
    {
      "epoch": 2.23,
      "grad_norm": 0.234375,
      "learning_rate": 0.0002,
      "loss": 0.8612,
      "step": 3920
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9716,
      "step": 3930
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 0.9311,
      "step": 3940
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.271484375,
      "learning_rate": 0.0002,
      "loss": 0.863,
      "step": 3950
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.279296875,
      "learning_rate": 0.0002,
      "loss": 0.7972,
      "step": 3960
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.248046875,
      "learning_rate": 0.0002,
      "loss": 0.9061,
      "step": 3970
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.263671875,
      "learning_rate": 0.0002,
      "loss": 0.9467,
      "step": 3980
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.263671875,
      "learning_rate": 0.0002,
      "loss": 0.9015,
      "step": 3990
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.263671875,
      "learning_rate": 0.0002,
      "loss": 0.8434,
      "step": 4000
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.28515625,
      "learning_rate": 0.0002,
      "loss": 0.8208,
      "step": 4010
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.2431640625,
      "learning_rate": 0.0002,
      "loss": 0.89,
      "step": 4020
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9688,
      "step": 4030
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.2578125,
      "learning_rate": 0.0002,
      "loss": 0.9848,
      "step": 4040
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.26953125,
      "learning_rate": 0.0002,
      "loss": 0.8888,
      "step": 4050
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.298828125,
      "learning_rate": 0.0002,
      "loss": 0.8008,
      "step": 4060
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.232421875,
      "learning_rate": 0.0002,
      "loss": 0.8682,
      "step": 4070
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.2470703125,
      "learning_rate": 0.0002,
      "loss": 0.9475,
      "step": 4080
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.2490234375,
      "learning_rate": 0.0002,
      "loss": 0.9707,
      "step": 4090
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.263671875,
      "learning_rate": 0.0002,
      "loss": 0.856,
      "step": 4100
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.279296875,
      "learning_rate": 0.0002,
      "loss": 0.7628,
      "step": 4110
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.2314453125,
      "learning_rate": 0.0002,
      "loss": 0.8685,
      "step": 4120
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.2431640625,
      "learning_rate": 0.0002,
      "loss": 0.9101,
      "step": 4130
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.244140625,
      "learning_rate": 0.0002,
      "loss": 0.8846,
      "step": 4140
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.287109375,
      "learning_rate": 0.0002,
      "loss": 0.8627,
      "step": 4150
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.28515625,
      "learning_rate": 0.0002,
      "loss": 0.7946,
      "step": 4160
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.2353515625,
      "learning_rate": 0.0002,
      "loss": 0.8894,
      "step": 4170
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.23046875,
      "learning_rate": 0.0002,
      "loss": 0.9399,
      "step": 4180
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.8964,
      "step": 4190
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.259765625,
      "learning_rate": 0.0002,
      "loss": 0.8725,
      "step": 4200
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.30078125,
      "learning_rate": 0.0002,
      "loss": 0.8108,
      "step": 4210
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.2373046875,
      "learning_rate": 0.0002,
      "loss": 0.8888,
      "step": 4220
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.240234375,
      "learning_rate": 0.0002,
      "loss": 0.9212,
      "step": 4230
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.263671875,
      "learning_rate": 0.0002,
      "loss": 0.9241,
      "step": 4240
    },
    {
      "epoch": 2.42,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.9084,
      "step": 4250
    },
    {
      "epoch": 2.42,
      "grad_norm": 0.294921875,
      "learning_rate": 0.0002,
      "loss": 0.7934,
      "step": 4260
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0002,
      "loss": 0.9015,
      "step": 4270
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.2578125,
      "learning_rate": 0.0002,
      "loss": 0.944,
      "step": 4280
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.267578125,
      "learning_rate": 0.0002,
      "loss": 0.9105,
      "step": 4290
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.26171875,
      "learning_rate": 0.0002,
      "loss": 0.8353,
      "step": 4300
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.287109375,
      "learning_rate": 0.0002,
      "loss": 0.7583,
      "step": 4310
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.2373046875,
      "learning_rate": 0.0002,
      "loss": 0.9053,
      "step": 4320
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.244140625,
      "learning_rate": 0.0002,
      "loss": 0.961,
      "step": 4330
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.9519,
      "step": 4340
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0002,
      "loss": 0.9192,
      "step": 4350
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 0.7754,
      "step": 4360
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.2265625,
      "learning_rate": 0.0002,
      "loss": 0.8785,
      "step": 4370
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 0.8908,
      "step": 4380
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.248046875,
      "learning_rate": 0.0002,
      "loss": 0.9154,
      "step": 4390
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.2578125,
      "learning_rate": 0.0002,
      "loss": 0.9057,
      "step": 4400
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.287109375,
      "learning_rate": 0.0002,
      "loss": 0.8377,
      "step": 4410
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.2314453125,
      "learning_rate": 0.0002,
      "loss": 0.8707,
      "step": 4420
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.2431640625,
      "learning_rate": 0.0002,
      "loss": 0.9584,
      "step": 4430
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.267578125,
      "learning_rate": 0.0002,
      "loss": 0.9012,
      "step": 4440
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.8969,
      "step": 4450
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.326171875,
      "learning_rate": 0.0002,
      "loss": 0.8156,
      "step": 4460
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.2412109375,
      "learning_rate": 0.0002,
      "loss": 0.909,
      "step": 4470
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.2431640625,
      "learning_rate": 0.0002,
      "loss": 0.9195,
      "step": 4480
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0002,
      "loss": 0.9104,
      "step": 4490
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.271484375,
      "learning_rate": 0.0002,
      "loss": 0.8961,
      "step": 4500
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.271484375,
      "learning_rate": 0.0002,
      "loss": 0.8028,
      "step": 4510
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0002,
      "loss": 0.8662,
      "step": 4520
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 0.8866,
      "step": 4530
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0002,
      "loss": 0.9517,
      "step": 4540
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.259765625,
      "learning_rate": 0.0002,
      "loss": 0.8347,
      "step": 4550
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.30078125,
      "learning_rate": 0.0002,
      "loss": 0.789,
      "step": 4560
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.224609375,
      "learning_rate": 0.0002,
      "loss": 0.8661,
      "step": 4570
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 0.9314,
      "step": 4580
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.26171875,
      "learning_rate": 0.0002,
      "loss": 0.8867,
      "step": 4590
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.26171875,
      "learning_rate": 0.0002,
      "loss": 0.9347,
      "step": 4600
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.2890625,
      "learning_rate": 0.0002,
      "loss": 0.7914,
      "step": 4610
    },
    {
      "epoch": 2.63,
      "grad_norm": 0.240234375,
      "learning_rate": 0.0002,
      "loss": 0.8746,
      "step": 4620
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.234375,
      "learning_rate": 0.0002,
      "loss": 0.9408,
      "step": 4630
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.24609375,
      "learning_rate": 0.0002,
      "loss": 0.9056,
      "step": 4640
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0002,
      "loss": 0.8412,
      "step": 4650
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.28125,
      "learning_rate": 0.0002,
      "loss": 0.7981,
      "step": 4660
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 0.9191,
      "step": 4670
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 0.9577,
      "step": 4680
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0002,
      "loss": 0.8989,
      "step": 4690
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.259765625,
      "learning_rate": 0.0002,
      "loss": 0.8807,
      "step": 4700
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.28515625,
      "learning_rate": 0.0002,
      "loss": 0.7851,
      "step": 4710
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.2451171875,
      "learning_rate": 0.0002,
      "loss": 0.9028,
      "step": 4720
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.234375,
      "learning_rate": 0.0002,
      "loss": 0.9442,
      "step": 4730
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.267578125,
      "learning_rate": 0.0002,
      "loss": 0.9038,
      "step": 4740
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.26953125,
      "learning_rate": 0.0002,
      "loss": 0.8508,
      "step": 4750
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.2890625,
      "learning_rate": 0.0002,
      "loss": 0.8069,
      "step": 4760
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.244140625,
      "learning_rate": 0.0002,
      "loss": 0.8872,
      "step": 4770
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 1.0013,
      "step": 4780
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.2578125,
      "learning_rate": 0.0002,
      "loss": 0.8867,
      "step": 4790
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.265625,
      "learning_rate": 0.0002,
      "loss": 0.8711,
      "step": 4800
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.28515625,
      "learning_rate": 0.0002,
      "loss": 0.7886,
      "step": 4810
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.24609375,
      "learning_rate": 0.0002,
      "loss": 0.9002,
      "step": 4820
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.2490234375,
      "learning_rate": 0.0002,
      "loss": 0.9994,
      "step": 4830
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.2470703125,
      "learning_rate": 0.0002,
      "loss": 0.9225,
      "step": 4840
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.28515625,
      "learning_rate": 0.0002,
      "loss": 0.8509,
      "step": 4850
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.28125,
      "learning_rate": 0.0002,
      "loss": 0.7973,
      "step": 4860
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.236328125,
      "learning_rate": 0.0002,
      "loss": 0.8904,
      "step": 4870
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.2373046875,
      "learning_rate": 0.0002,
      "loss": 0.9189,
      "step": 4880
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0002,
      "loss": 0.8756,
      "step": 4890
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.2578125,
      "learning_rate": 0.0002,
      "loss": 0.8989,
      "step": 4900
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0002,
      "loss": 0.7792,
      "step": 4910
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.26171875,
      "learning_rate": 0.0002,
      "loss": 0.8501,
      "step": 4920
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.2451171875,
      "learning_rate": 0.0002,
      "loss": 0.9372,
      "step": 4930
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.267578125,
      "learning_rate": 0.0002,
      "loss": 0.9166,
      "step": 4940
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.2734375,
      "learning_rate": 0.0002,
      "loss": 0.8523,
      "step": 4950
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.28515625,
      "learning_rate": 0.0002,
      "loss": 0.777,
      "step": 4960
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.2333984375,
      "learning_rate": 0.0002,
      "loss": 0.8998,
      "step": 4970
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.2451171875,
      "learning_rate": 0.0002,
      "loss": 0.907,
      "step": 4980
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 0.9126,
      "step": 4990
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.271484375,
      "learning_rate": 0.0002,
      "loss": 0.8323,
      "step": 5000
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.302734375,
      "learning_rate": 0.0002,
      "loss": 0.7853,
      "step": 5010
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.2470703125,
      "learning_rate": 0.0002,
      "loss": 0.8995,
      "step": 5020
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0002,
      "loss": 0.943,
      "step": 5030
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.8728,
      "step": 5040
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 0.8486,
      "step": 5050
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.265625,
      "learning_rate": 0.0002,
      "loss": 0.7424,
      "step": 5060
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.2255859375,
      "learning_rate": 0.0002,
      "loss": 0.8708,
      "step": 5070
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.2470703125,
      "learning_rate": 0.0002,
      "loss": 0.9878,
      "step": 5080
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.271484375,
      "learning_rate": 0.0002,
      "loss": 0.9329,
      "step": 5090
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.287109375,
      "learning_rate": 0.0002,
      "loss": 0.8871,
      "step": 5100
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.310546875,
      "learning_rate": 0.0002,
      "loss": 0.7713,
      "step": 5110
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.2490234375,
      "learning_rate": 0.0002,
      "loss": 0.8863,
      "step": 5120
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.2451171875,
      "learning_rate": 0.0002,
      "loss": 0.98,
      "step": 5130
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.26171875,
      "learning_rate": 0.0002,
      "loss": 0.8952,
      "step": 5140
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0002,
      "loss": 0.8547,
      "step": 5150
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.27734375,
      "learning_rate": 0.0002,
      "loss": 0.7906,
      "step": 5160
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.2333984375,
      "learning_rate": 0.0002,
      "loss": 0.8786,
      "step": 5170
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9914,
      "step": 5180
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.26171875,
      "learning_rate": 0.0002,
      "loss": 0.9002,
      "step": 5190
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.271484375,
      "learning_rate": 0.0002,
      "loss": 0.8754,
      "step": 5200
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.283203125,
      "learning_rate": 0.0002,
      "loss": 0.7862,
      "step": 5210
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.2412109375,
      "learning_rate": 0.0002,
      "loss": 0.853,
      "step": 5220
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.2412109375,
      "learning_rate": 0.0002,
      "loss": 0.9762,
      "step": 5230
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.9244,
      "step": 5240
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.26953125,
      "learning_rate": 0.0002,
      "loss": 0.871,
      "step": 5250
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.302734375,
      "learning_rate": 0.0002,
      "loss": 0.8212,
      "step": 5260
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.298828125,
      "learning_rate": 0.0002,
      "loss": 0.8614,
      "step": 5270
    },
    {
      "epoch": 3.01,
      "grad_norm": 0.2333984375,
      "learning_rate": 0.0002,
      "loss": 0.9362,
      "step": 5280
    },
    {
      "epoch": 3.01,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.9234,
      "step": 5290
    },
    {
      "epoch": 3.02,
      "grad_norm": 0.263671875,
      "learning_rate": 0.0002,
      "loss": 0.7922,
      "step": 5300
    },
    {
      "epoch": 3.02,
      "grad_norm": 0.265625,
      "learning_rate": 0.0002,
      "loss": 0.8053,
      "step": 5310
    },
    {
      "epoch": 3.03,
      "grad_norm": 0.326171875,
      "learning_rate": 0.0002,
      "loss": 0.7281,
      "step": 5320
    },
    {
      "epoch": 3.03,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0002,
      "loss": 0.9209,
      "step": 5330
    },
    {
      "epoch": 3.04,
      "grad_norm": 0.27734375,
      "learning_rate": 0.0002,
      "loss": 0.891,
      "step": 5340
    },
    {
      "epoch": 3.04,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0002,
      "loss": 0.8865,
      "step": 5350
    },
    {
      "epoch": 3.05,
      "grad_norm": 0.287109375,
      "learning_rate": 0.0002,
      "loss": 0.8166,
      "step": 5360
    },
    {
      "epoch": 3.06,
      "grad_norm": 0.30859375,
      "learning_rate": 0.0002,
      "loss": 0.7621,
      "step": 5370
    },
    {
      "epoch": 3.06,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0002,
      "loss": 0.8545,
      "step": 5380
    },
    {
      "epoch": 3.07,
      "grad_norm": 0.26171875,
      "learning_rate": 0.0002,
      "loss": 0.8467,
      "step": 5390
    },
    {
      "epoch": 3.07,
      "grad_norm": 0.26953125,
      "learning_rate": 0.0002,
      "loss": 0.8838,
      "step": 5400
    },
    {
      "epoch": 3.08,
      "grad_norm": 0.27734375,
      "learning_rate": 0.0002,
      "loss": 0.7687,
      "step": 5410
    },
    {
      "epoch": 3.08,
      "grad_norm": 0.3203125,
      "learning_rate": 0.0002,
      "loss": 0.6956,
      "step": 5420
    },
    {
      "epoch": 3.09,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0002,
      "loss": 0.9383,
      "step": 5430
    },
    {
      "epoch": 3.1,
      "grad_norm": 0.2578125,
      "learning_rate": 0.0002,
      "loss": 0.9487,
      "step": 5440
    },
    {
      "epoch": 3.1,
      "grad_norm": 0.287109375,
      "learning_rate": 0.0002,
      "loss": 0.859,
      "step": 5450
    },
    {
      "epoch": 3.11,
      "grad_norm": 0.279296875,
      "learning_rate": 0.0002,
      "loss": 0.8361,
      "step": 5460
    },
    {
      "epoch": 3.11,
      "grad_norm": 0.31640625,
      "learning_rate": 0.0002,
      "loss": 0.7345,
      "step": 5470
    },
    {
      "epoch": 3.12,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 0.8926,
      "step": 5480
    },
    {
      "epoch": 3.12,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0002,
      "loss": 0.8956,
      "step": 5490
    },
    {
      "epoch": 3.13,
      "grad_norm": 0.265625,
      "learning_rate": 0.0002,
      "loss": 0.924,
      "step": 5500
    },
    {
      "epoch": 3.14,
      "grad_norm": 0.296875,
      "learning_rate": 0.0002,
      "loss": 0.8312,
      "step": 5510
    },
    {
      "epoch": 3.14,
      "grad_norm": 0.322265625,
      "learning_rate": 0.0002,
      "loss": 0.7387,
      "step": 5520
    },
    {
      "epoch": 3.15,
      "grad_norm": 0.279296875,
      "learning_rate": 0.0002,
      "loss": 0.9535,
      "step": 5530
    },
    {
      "epoch": 3.15,
      "grad_norm": 0.263671875,
      "learning_rate": 0.0002,
      "loss": 0.8703,
      "step": 5540
    },
    {
      "epoch": 3.16,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0002,
      "loss": 0.8652,
      "step": 5550
    },
    {
      "epoch": 3.16,
      "grad_norm": 0.28125,
      "learning_rate": 0.0002,
      "loss": 0.8261,
      "step": 5560
    },
    {
      "epoch": 3.17,
      "grad_norm": 0.330078125,
      "learning_rate": 0.0002,
      "loss": 0.7514,
      "step": 5570
    },
    {
      "epoch": 3.18,
      "grad_norm": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8952,
      "step": 5580
    },
    {
      "epoch": 3.18,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 0.9366,
      "step": 5590
    },
    {
      "epoch": 3.19,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 0.84,
      "step": 5600
    },
    {
      "epoch": 3.19,
      "grad_norm": 0.2890625,
      "learning_rate": 0.0002,
      "loss": 0.8177,
      "step": 5610
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.296875,
      "learning_rate": 0.0002,
      "loss": 0.7599,
      "step": 5620
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.267578125,
      "learning_rate": 0.0002,
      "loss": 0.8695,
      "step": 5630
    },
    {
      "epoch": 3.21,
      "grad_norm": 0.26953125,
      "learning_rate": 0.0002,
      "loss": 0.9257,
      "step": 5640
    },
    {
      "epoch": 3.22,
      "grad_norm": 0.279296875,
      "learning_rate": 0.0002,
      "loss": 0.8631,
      "step": 5650
    },
    {
      "epoch": 3.22,
      "grad_norm": 0.306640625,
      "learning_rate": 0.0002,
      "loss": 0.8063,
      "step": 5660
    },
    {
      "epoch": 3.23,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 0.7112,
      "step": 5670
    },
    {
      "epoch": 3.23,
      "grad_norm": 0.2578125,
      "learning_rate": 0.0002,
      "loss": 0.8807,
      "step": 5680
    },
    {
      "epoch": 3.24,
      "grad_norm": 0.26953125,
      "learning_rate": 0.0002,
      "loss": 0.8928,
      "step": 5690
    },
    {
      "epoch": 3.24,
      "grad_norm": 0.28125,
      "learning_rate": 0.0002,
      "loss": 0.8631,
      "step": 5700
    },
    {
      "epoch": 3.25,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 0.8192,
      "step": 5710
    },
    {
      "epoch": 3.26,
      "grad_norm": 0.337890625,
      "learning_rate": 0.0002,
      "loss": 0.7731,
      "step": 5720
    },
    {
      "epoch": 3.26,
      "grad_norm": 0.263671875,
      "learning_rate": 0.0002,
      "loss": 0.8969,
      "step": 5730
    },
    {
      "epoch": 3.27,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0002,
      "loss": 0.9397,
      "step": 5740
    },
    {
      "epoch": 3.27,
      "grad_norm": 0.263671875,
      "learning_rate": 0.0002,
      "loss": 0.8819,
      "step": 5750
    },
    {
      "epoch": 3.28,
      "grad_norm": 0.30859375,
      "learning_rate": 0.0002,
      "loss": 0.8199,
      "step": 5760
    },
    {
      "epoch": 3.28,
      "grad_norm": 0.2890625,
      "learning_rate": 0.0002,
      "loss": 0.7295,
      "step": 5770
    },
    {
      "epoch": 3.29,
      "grad_norm": 0.26171875,
      "learning_rate": 0.0002,
      "loss": 0.922,
      "step": 5780
    },
    {
      "epoch": 3.3,
      "grad_norm": 0.283203125,
      "learning_rate": 0.0002,
      "loss": 0.8984,
      "step": 5790
    },
    {
      "epoch": 3.3,
      "grad_norm": 0.265625,
      "learning_rate": 0.0002,
      "loss": 0.8534,
      "step": 5800
    },
    {
      "epoch": 3.31,
      "grad_norm": 0.291015625,
      "learning_rate": 0.0002,
      "loss": 0.7722,
      "step": 5810
    },
    {
      "epoch": 3.31,
      "grad_norm": 0.30859375,
      "learning_rate": 0.0002,
      "loss": 0.7367,
      "step": 5820
    },
    {
      "epoch": 3.32,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.8646,
      "step": 5830
    },
    {
      "epoch": 3.32,
      "grad_norm": 0.2734375,
      "learning_rate": 0.0002,
      "loss": 0.896,
      "step": 5840
    },
    {
      "epoch": 3.33,
      "grad_norm": 0.26953125,
      "learning_rate": 0.0002,
      "loss": 0.8458,
      "step": 5850
    },
    {
      "epoch": 3.34,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0002,
      "loss": 0.763,
      "step": 5860
    },
    {
      "epoch": 3.34,
      "grad_norm": 0.330078125,
      "learning_rate": 0.0002,
      "loss": 0.7361,
      "step": 5870
    },
    {
      "epoch": 3.35,
      "grad_norm": 0.259765625,
      "learning_rate": 0.0002,
      "loss": 0.9133,
      "step": 5880
    },
    {
      "epoch": 3.35,
      "grad_norm": 0.2734375,
      "learning_rate": 0.0002,
      "loss": 0.892,
      "step": 5890
    },
    {
      "epoch": 3.36,
      "grad_norm": 0.265625,
      "learning_rate": 0.0002,
      "loss": 0.856,
      "step": 5900
    },
    {
      "epoch": 3.36,
      "grad_norm": 0.30859375,
      "learning_rate": 0.0002,
      "loss": 0.8,
      "step": 5910
    },
    {
      "epoch": 3.37,
      "grad_norm": 0.314453125,
      "learning_rate": 0.0002,
      "loss": 0.7611,
      "step": 5920
    },
    {
      "epoch": 3.38,
      "grad_norm": 0.259765625,
      "learning_rate": 0.0002,
      "loss": 0.8691,
      "step": 5930
    },
    {
      "epoch": 3.38,
      "grad_norm": 0.287109375,
      "learning_rate": 0.0002,
      "loss": 0.8909,
      "step": 5940
    },
    {
      "epoch": 3.39,
      "grad_norm": 0.291015625,
      "learning_rate": 0.0002,
      "loss": 0.8433,
      "step": 5950
    },
    {
      "epoch": 3.39,
      "grad_norm": 0.3125,
      "learning_rate": 0.0002,
      "loss": 0.8023,
      "step": 5960
    },
    {
      "epoch": 3.4,
      "grad_norm": 0.318359375,
      "learning_rate": 0.0002,
      "loss": 0.7657,
      "step": 5970
    },
    {
      "epoch": 3.4,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 0.8957,
      "step": 5980
    },
    {
      "epoch": 3.41,
      "grad_norm": 0.263671875,
      "learning_rate": 0.0002,
      "loss": 0.874,
      "step": 5990
    },
    {
      "epoch": 3.41,
      "grad_norm": 0.2890625,
      "learning_rate": 0.0002,
      "loss": 0.8756,
      "step": 6000
    },
    {
      "epoch": 3.42,
      "grad_norm": 0.28515625,
      "learning_rate": 0.0002,
      "loss": 0.8275,
      "step": 6010
    },
    {
      "epoch": 3.43,
      "grad_norm": 0.333984375,
      "learning_rate": 0.0002,
      "loss": 0.7317,
      "step": 6020
    },
    {
      "epoch": 3.43,
      "grad_norm": 0.265625,
      "learning_rate": 0.0002,
      "loss": 0.927,
      "step": 6030
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.271484375,
      "learning_rate": 0.0002,
      "loss": 0.8982,
      "step": 6040
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.296875,
      "learning_rate": 0.0002,
      "loss": 0.8786,
      "step": 6050
    },
    {
      "epoch": 3.45,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 0.7958,
      "step": 6060
    },
    {
      "epoch": 3.45,
      "grad_norm": 0.31640625,
      "learning_rate": 0.0002,
      "loss": 0.7241,
      "step": 6070
    },
    {
      "epoch": 3.46,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0002,
      "loss": 0.8856,
      "step": 6080
    },
    {
      "epoch": 3.47,
      "grad_norm": 0.271484375,
      "learning_rate": 0.0002,
      "loss": 0.9041,
      "step": 6090
    },
    {
      "epoch": 3.47,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 0.859,
      "step": 6100
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.28515625,
      "learning_rate": 0.0002,
      "loss": 0.8297,
      "step": 6110
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.3359375,
      "learning_rate": 0.0002,
      "loss": 0.7489,
      "step": 6120
    },
    {
      "epoch": 3.49,
      "grad_norm": 0.263671875,
      "learning_rate": 0.0002,
      "loss": 0.9257,
      "step": 6130
    },
    {
      "epoch": 3.49,
      "grad_norm": 0.26171875,
      "learning_rate": 0.0002,
      "loss": 0.8566,
      "step": 6140
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.283203125,
      "learning_rate": 0.0002,
      "loss": 0.8396,
      "step": 6150
    },
    {
      "epoch": 3.51,
      "grad_norm": 0.296875,
      "learning_rate": 0.0002,
      "loss": 0.7752,
      "step": 6160
    },
    {
      "epoch": 3.51,
      "grad_norm": 0.318359375,
      "learning_rate": 0.0002,
      "loss": 0.7169,
      "step": 6170
    },
    {
      "epoch": 3.52,
      "grad_norm": 0.271484375,
      "learning_rate": 0.0002,
      "loss": 0.9082,
      "step": 6180
    },
    {
      "epoch": 3.52,
      "grad_norm": 0.26171875,
      "learning_rate": 0.0002,
      "loss": 0.8875,
      "step": 6190
    },
    {
      "epoch": 3.53,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0002,
      "loss": 0.8636,
      "step": 6200
    },
    {
      "epoch": 3.53,
      "grad_norm": 0.3046875,
      "learning_rate": 0.0002,
      "loss": 0.7987,
      "step": 6210
    },
    {
      "epoch": 3.54,
      "grad_norm": 0.3203125,
      "learning_rate": 0.0002,
      "loss": 0.7453,
      "step": 6220
    },
    {
      "epoch": 3.55,
      "grad_norm": 0.26171875,
      "learning_rate": 0.0002,
      "loss": 0.8934,
      "step": 6230
    },
    {
      "epoch": 3.55,
      "grad_norm": 0.26953125,
      "learning_rate": 0.0002,
      "loss": 0.8645,
      "step": 6240
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.283203125,
      "learning_rate": 0.0002,
      "loss": 0.8846,
      "step": 6250
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.306640625,
      "learning_rate": 0.0002,
      "loss": 0.7788,
      "step": 6260
    },
    {
      "epoch": 3.57,
      "grad_norm": 0.33203125,
      "learning_rate": 0.0002,
      "loss": 0.7235,
      "step": 6270
    },
    {
      "epoch": 3.57,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 0.9174,
      "step": 6280
    },
    {
      "epoch": 3.58,
      "grad_norm": 0.27734375,
      "learning_rate": 0.0002,
      "loss": 0.9251,
      "step": 6290
    },
    {
      "epoch": 3.59,
      "grad_norm": 0.310546875,
      "learning_rate": 0.0002,
      "loss": 0.8728,
      "step": 6300
    },
    {
      "epoch": 3.59,
      "grad_norm": 0.291015625,
      "learning_rate": 0.0002,
      "loss": 0.8223,
      "step": 6310
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.322265625,
      "learning_rate": 0.0002,
      "loss": 0.7791,
      "step": 6320
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.279296875,
      "learning_rate": 0.0002,
      "loss": 0.8652,
      "step": 6330
    },
    {
      "epoch": 3.61,
      "grad_norm": 0.26171875,
      "learning_rate": 0.0002,
      "loss": 0.8955,
      "step": 6340
    },
    {
      "epoch": 3.61,
      "grad_norm": 0.2890625,
      "learning_rate": 0.0002,
      "loss": 0.8949,
      "step": 6350
    },
    {
      "epoch": 3.62,
      "grad_norm": 0.30078125,
      "learning_rate": 0.0002,
      "loss": 0.805,
      "step": 6360
    },
    {
      "epoch": 3.63,
      "grad_norm": 0.345703125,
      "learning_rate": 0.0002,
      "loss": 0.7788,
      "step": 6370
    },
    {
      "epoch": 3.63,
      "grad_norm": 0.2578125,
      "learning_rate": 0.0002,
      "loss": 0.9124,
      "step": 6380
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.283203125,
      "learning_rate": 0.0002,
      "loss": 0.9254,
      "step": 6390
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.306640625,
      "learning_rate": 0.0002,
      "loss": 0.8955,
      "step": 6400
    },
    {
      "epoch": 3.65,
      "grad_norm": 0.287109375,
      "learning_rate": 0.0002,
      "loss": 0.8114,
      "step": 6410
    },
    {
      "epoch": 3.65,
      "grad_norm": 0.31640625,
      "learning_rate": 0.0002,
      "loss": 0.8126,
      "step": 6420
    },
    {
      "epoch": 3.66,
      "grad_norm": 0.291015625,
      "learning_rate": 0.0002,
      "loss": 0.9208,
      "step": 6430
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.2734375,
      "learning_rate": 0.0002,
      "loss": 0.8888,
      "step": 6440
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.283203125,
      "learning_rate": 0.0002,
      "loss": 0.8866,
      "step": 6450
    },
    {
      "epoch": 3.68,
      "grad_norm": 0.310546875,
      "learning_rate": 0.0002,
      "loss": 0.7911,
      "step": 6460
    },
    {
      "epoch": 3.68,
      "grad_norm": 0.33984375,
      "learning_rate": 0.0002,
      "loss": 0.7326,
      "step": 6470
    },
    {
      "epoch": 3.69,
      "grad_norm": 0.263671875,
      "learning_rate": 0.0002,
      "loss": 0.9256,
      "step": 6480
    },
    {
      "epoch": 3.69,
      "grad_norm": 0.267578125,
      "learning_rate": 0.0002,
      "loss": 0.8476,
      "step": 6490
    },
    {
      "epoch": 3.7,
      "grad_norm": 0.279296875,
      "learning_rate": 0.0002,
      "loss": 0.8977,
      "step": 6500
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.283203125,
      "learning_rate": 0.0002,
      "loss": 0.7827,
      "step": 6510
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.330078125,
      "learning_rate": 0.0002,
      "loss": 0.7202,
      "step": 6520
    },
    {
      "epoch": 3.72,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.9272,
      "step": 6530
    },
    {
      "epoch": 3.72,
      "grad_norm": 0.2734375,
      "learning_rate": 0.0002,
      "loss": 0.933,
      "step": 6540
    },
    {
      "epoch": 3.73,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 0.8725,
      "step": 6550
    },
    {
      "epoch": 3.73,
      "grad_norm": 0.30078125,
      "learning_rate": 0.0002,
      "loss": 0.8208,
      "step": 6560
    },
    {
      "epoch": 3.74,
      "grad_norm": 0.3203125,
      "learning_rate": 0.0002,
      "loss": 0.7662,
      "step": 6570
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.9297,
      "step": 6580
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.259765625,
      "learning_rate": 0.0002,
      "loss": 0.9671,
      "step": 6590
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.287109375,
      "learning_rate": 0.0002,
      "loss": 0.8773,
      "step": 6600
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.2890625,
      "learning_rate": 0.0002,
      "loss": 0.7993,
      "step": 6610
    },
    {
      "epoch": 3.77,
      "grad_norm": 0.3125,
      "learning_rate": 0.0002,
      "loss": 0.7291,
      "step": 6620
    },
    {
      "epoch": 3.77,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0002,
      "loss": 0.9011,
      "step": 6630
    },
    {
      "epoch": 3.78,
      "grad_norm": 0.306640625,
      "learning_rate": 0.0002,
      "loss": 0.8893,
      "step": 6640
    },
    {
      "epoch": 3.78,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0002,
      "loss": 0.8976,
      "step": 6650
    },
    {
      "epoch": 3.79,
      "grad_norm": 0.28515625,
      "learning_rate": 0.0002,
      "loss": 0.8049,
      "step": 6660
    },
    {
      "epoch": 3.8,
      "grad_norm": 0.322265625,
      "learning_rate": 0.0002,
      "loss": 0.7419,
      "step": 6670
    },
    {
      "epoch": 3.8,
      "grad_norm": 0.259765625,
      "learning_rate": 0.0002,
      "loss": 0.8809,
      "step": 6680
    },
    {
      "epoch": 3.81,
      "grad_norm": 0.259765625,
      "learning_rate": 0.0002,
      "loss": 0.8919,
      "step": 6690
    },
    {
      "epoch": 3.81,
      "grad_norm": 0.26953125,
      "learning_rate": 0.0002,
      "loss": 0.9121,
      "step": 6700
    },
    {
      "epoch": 3.82,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0002,
      "loss": 0.8353,
      "step": 6710
    },
    {
      "epoch": 3.82,
      "grad_norm": 0.322265625,
      "learning_rate": 0.0002,
      "loss": 0.7508,
      "step": 6720
    },
    {
      "epoch": 3.83,
      "grad_norm": 0.263671875,
      "learning_rate": 0.0002,
      "loss": 0.8558,
      "step": 6730
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.27734375,
      "learning_rate": 0.0002,
      "loss": 0.8868,
      "step": 6740
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.267578125,
      "learning_rate": 0.0002,
      "loss": 0.8733,
      "step": 6750
    },
    {
      "epoch": 3.85,
      "grad_norm": 0.271484375,
      "learning_rate": 0.0002,
      "loss": 0.8081,
      "step": 6760
    },
    {
      "epoch": 3.85,
      "grad_norm": 0.302734375,
      "learning_rate": 0.0002,
      "loss": 0.7546,
      "step": 6770
    },
    {
      "epoch": 3.86,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0002,
      "loss": 0.8928,
      "step": 6780
    },
    {
      "epoch": 3.86,
      "grad_norm": 0.263671875,
      "learning_rate": 0.0002,
      "loss": 0.8965,
      "step": 6790
    },
    {
      "epoch": 3.87,
      "grad_norm": 0.27734375,
      "learning_rate": 0.0002,
      "loss": 0.8731,
      "step": 6800
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 0.8397,
      "step": 6810
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.3203125,
      "learning_rate": 0.0002,
      "loss": 0.703,
      "step": 6820
    },
    {
      "epoch": 3.89,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.8832,
      "step": 6830
    },
    {
      "epoch": 3.89,
      "grad_norm": 0.26171875,
      "learning_rate": 0.0002,
      "loss": 0.9216,
      "step": 6840
    },
    {
      "epoch": 3.9,
      "grad_norm": 0.279296875,
      "learning_rate": 0.0002,
      "loss": 0.8275,
      "step": 6850
    },
    {
      "epoch": 3.9,
      "grad_norm": 0.28515625,
      "learning_rate": 0.0002,
      "loss": 0.8089,
      "step": 6860
    },
    {
      "epoch": 3.91,
      "grad_norm": 0.3046875,
      "learning_rate": 0.0002,
      "loss": 0.7456,
      "step": 6870
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.26171875,
      "learning_rate": 0.0002,
      "loss": 0.9036,
      "step": 6880
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.2734375,
      "learning_rate": 0.0002,
      "loss": 0.9218,
      "step": 6890
    },
    {
      "epoch": 3.93,
      "grad_norm": 0.2734375,
      "learning_rate": 0.0002,
      "loss": 0.8775,
      "step": 6900
    },
    {
      "epoch": 3.93,
      "grad_norm": 0.291015625,
      "learning_rate": 0.0002,
      "loss": 0.8538,
      "step": 6910
    },
    {
      "epoch": 3.94,
      "grad_norm": 0.30859375,
      "learning_rate": 0.0002,
      "loss": 0.7244,
      "step": 6920
    },
    {
      "epoch": 3.94,
      "grad_norm": 0.26171875,
      "learning_rate": 0.0002,
      "loss": 0.9341,
      "step": 6930
    },
    {
      "epoch": 3.95,
      "grad_norm": 0.28125,
      "learning_rate": 0.0002,
      "loss": 0.8836,
      "step": 6940
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.291015625,
      "learning_rate": 0.0002,
      "loss": 0.843,
      "step": 6950
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.2890625,
      "learning_rate": 0.0002,
      "loss": 0.7709,
      "step": 6960
    },
    {
      "epoch": 3.97,
      "grad_norm": 0.33203125,
      "learning_rate": 0.0002,
      "loss": 0.7457,
      "step": 6970
    },
    {
      "epoch": 3.97,
      "grad_norm": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9053,
      "step": 6980
    },
    {
      "epoch": 3.98,
      "grad_norm": 0.25390625,
      "learning_rate": 0.0002,
      "loss": 0.9145,
      "step": 6990
    },
    {
      "epoch": 3.98,
      "grad_norm": 0.283203125,
      "learning_rate": 0.0002,
      "loss": 0.88,
      "step": 7000
    },
    {
      "epoch": 3.99,
      "grad_norm": 0.291015625,
      "learning_rate": 0.0002,
      "loss": 0.7534,
      "step": 7010
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.3046875,
      "learning_rate": 0.0002,
      "loss": 0.7275,
      "step": 7020
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0002,
      "loss": 0.8967,
      "step": 7030
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.271484375,
      "learning_rate": 0.0002,
      "loss": 0.8987,
      "step": 7040
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.37109375,
      "learning_rate": 0.0002,
      "loss": 0.8529,
      "step": 7050
    },
    {
      "epoch": 4.02,
      "grad_norm": 0.2734375,
      "learning_rate": 0.0002,
      "loss": 0.7961,
      "step": 7060
    },
    {
      "epoch": 4.02,
      "grad_norm": 0.3046875,
      "learning_rate": 0.0002,
      "loss": 0.7495,
      "step": 7070
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.294921875,
      "learning_rate": 0.0002,
      "loss": 0.7246,
      "step": 7080
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.26171875,
      "learning_rate": 0.0002,
      "loss": 0.8711,
      "step": 7090
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.291015625,
      "learning_rate": 0.0002,
      "loss": 0.878,
      "step": 7100
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.388671875,
      "learning_rate": 0.0002,
      "loss": 0.7671,
      "step": 7110
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.302734375,
      "learning_rate": 0.0002,
      "loss": 0.7497,
      "step": 7120
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.33984375,
      "learning_rate": 0.0002,
      "loss": 0.7371,
      "step": 7130
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.2890625,
      "learning_rate": 0.0002,
      "loss": 0.8748,
      "step": 7140
    },
    {
      "epoch": 4.07,
      "grad_norm": 0.28125,
      "learning_rate": 0.0002,
      "loss": 0.8654,
      "step": 7150
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.30859375,
      "learning_rate": 0.0002,
      "loss": 0.8225,
      "step": 7160
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.314453125,
      "learning_rate": 0.0002,
      "loss": 0.7118,
      "step": 7170
    },
    {
      "epoch": 4.09,
      "grad_norm": 0.28125,
      "learning_rate": 0.0002,
      "loss": 0.7277,
      "step": 7180
    },
    {
      "epoch": 4.09,
      "grad_norm": 0.26171875,
      "learning_rate": 0.0002,
      "loss": 0.881,
      "step": 7190
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.2890625,
      "learning_rate": 0.0002,
      "loss": 0.882,
      "step": 7200
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.30078125,
      "learning_rate": 0.0002,
      "loss": 0.8092,
      "step": 7210
    },
    {
      "epoch": 4.11,
      "grad_norm": 0.322265625,
      "learning_rate": 0.0002,
      "loss": 0.7393,
      "step": 7220
    },
    {
      "epoch": 4.11,
      "grad_norm": 0.283203125,
      "learning_rate": 0.0002,
      "loss": 0.7119,
      "step": 7230
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.267578125,
      "learning_rate": 0.0002,
      "loss": 0.8961,
      "step": 7240
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.298828125,
      "learning_rate": 0.0002,
      "loss": 0.8183,
      "step": 7250
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.2890625,
      "learning_rate": 0.0002,
      "loss": 0.7742,
      "step": 7260
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.32421875,
      "learning_rate": 0.0002,
      "loss": 0.7203,
      "step": 7270
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.294921875,
      "learning_rate": 0.0002,
      "loss": 0.7242,
      "step": 7280
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0002,
      "loss": 0.8495,
      "step": 7290
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.279296875,
      "learning_rate": 0.0002,
      "loss": 0.8641,
      "step": 7300
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.3046875,
      "learning_rate": 0.0002,
      "loss": 0.8044,
      "step": 7310
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.3359375,
      "learning_rate": 0.0002,
      "loss": 0.747,
      "step": 7320
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.298828125,
      "learning_rate": 0.0002,
      "loss": 0.7301,
      "step": 7330
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.283203125,
      "learning_rate": 0.0002,
      "loss": 0.8736,
      "step": 7340
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.287109375,
      "learning_rate": 0.0002,
      "loss": 0.8345,
      "step": 7350
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.3125,
      "learning_rate": 0.0002,
      "loss": 0.8308,
      "step": 7360
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.33203125,
      "learning_rate": 0.0002,
      "loss": 0.7209,
      "step": 7370
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.287109375,
      "learning_rate": 0.0002,
      "loss": 0.7254,
      "step": 7380
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002,
      "loss": 0.8572,
      "step": 7390
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.27734375,
      "learning_rate": 0.0002,
      "loss": 0.8422,
      "step": 7400
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.361328125,
      "learning_rate": 0.0002,
      "loss": 0.8001,
      "step": 7410
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.326171875,
      "learning_rate": 0.0002,
      "loss": 0.7102,
      "step": 7420
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.28125,
      "learning_rate": 0.0002,
      "loss": 0.7417,
      "step": 7430
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.279296875,
      "learning_rate": 0.0002,
      "loss": 0.8495,
      "step": 7440
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.279296875,
      "learning_rate": 0.0002,
      "loss": 0.8218,
      "step": 7450
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.318359375,
      "learning_rate": 0.0002,
      "loss": 0.802,
      "step": 7460
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.3203125,
      "learning_rate": 0.0002,
      "loss": 0.7345,
      "step": 7470
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.298828125,
      "learning_rate": 0.0002,
      "loss": 0.7279,
      "step": 7480
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.298828125,
      "learning_rate": 0.0002,
      "loss": 0.9336,
      "step": 7490
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 0.8473,
      "step": 7500
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.30078125,
      "learning_rate": 0.0002,
      "loss": 0.8213,
      "step": 7510
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.310546875,
      "learning_rate": 0.0002,
      "loss": 0.7491,
      "step": 7520
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.298828125,
      "learning_rate": 0.0002,
      "loss": 0.77,
      "step": 7530
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.291015625,
      "learning_rate": 0.0002,
      "loss": 0.9229,
      "step": 7540
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.30859375,
      "learning_rate": 0.0002,
      "loss": 0.8667,
      "step": 7550
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.310546875,
      "learning_rate": 0.0002,
      "loss": 0.8125,
      "step": 7560
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.306640625,
      "learning_rate": 0.0002,
      "loss": 0.774,
      "step": 7570
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.306640625,
      "learning_rate": 0.0002,
      "loss": 0.761,
      "step": 7580
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.279296875,
      "learning_rate": 0.0002,
      "loss": 0.8943,
      "step": 7590
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.298828125,
      "learning_rate": 0.0002,
      "loss": 0.8648,
      "step": 7600
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.3125,
      "learning_rate": 0.0002,
      "loss": 0.7935,
      "step": 7610
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.3125,
      "learning_rate": 0.0002,
      "loss": 0.7123,
      "step": 7620
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0002,
      "loss": 0.7425,
      "step": 7630
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.296875,
      "learning_rate": 0.0002,
      "loss": 0.8549,
      "step": 7640
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.296875,
      "learning_rate": 0.0002,
      "loss": 0.876,
      "step": 7650
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.310546875,
      "learning_rate": 0.0002,
      "loss": 0.8254,
      "step": 7660
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.328125,
      "learning_rate": 0.0002,
      "loss": 0.7379,
      "step": 7670
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.27734375,
      "learning_rate": 0.0002,
      "loss": 0.7345,
      "step": 7680
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.283203125,
      "learning_rate": 0.0002,
      "loss": 0.9001,
      "step": 7690
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.287109375,
      "learning_rate": 0.0002,
      "loss": 0.8214,
      "step": 7700
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.30078125,
      "learning_rate": 0.0002,
      "loss": 0.7806,
      "step": 7710
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.322265625,
      "learning_rate": 0.0002,
      "loss": 0.7122,
      "step": 7720
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.2890625,
      "learning_rate": 0.0002,
      "loss": 0.7645,
      "step": 7730
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 0.9018,
      "step": 7740
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.298828125,
      "learning_rate": 0.0002,
      "loss": 0.8552,
      "step": 7750
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.30859375,
      "learning_rate": 0.0002,
      "loss": 0.8327,
      "step": 7760
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.3203125,
      "learning_rate": 0.0002,
      "loss": 0.7828,
      "step": 7770
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.306640625,
      "learning_rate": 0.0002,
      "loss": 0.716,
      "step": 7780
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0002,
      "loss": 0.911,
      "step": 7790
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 0.86,
      "step": 7800
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.30078125,
      "learning_rate": 0.0002,
      "loss": 0.8629,
      "step": 7810
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.310546875,
      "learning_rate": 0.0002,
      "loss": 0.7469,
      "step": 7820
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.287109375,
      "learning_rate": 0.0002,
      "loss": 0.7177,
      "step": 7830
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.2890625,
      "learning_rate": 0.0002,
      "loss": 0.9086,
      "step": 7840
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.34765625,
      "learning_rate": 0.0002,
      "loss": 0.8792,
      "step": 7850
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.314453125,
      "learning_rate": 0.0002,
      "loss": 0.8517,
      "step": 7860
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.353515625,
      "learning_rate": 0.0002,
      "loss": 0.7711,
      "step": 7870
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.294921875,
      "learning_rate": 0.0002,
      "loss": 0.7412,
      "step": 7880
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.2734375,
      "learning_rate": 0.0002,
      "loss": 0.8807,
      "step": 7890
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.341796875,
      "learning_rate": 0.0002,
      "loss": 0.8501,
      "step": 7900
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.30859375,
      "learning_rate": 0.0002,
      "loss": 0.7968,
      "step": 7910
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.314453125,
      "learning_rate": 0.0002,
      "loss": 0.7332,
      "step": 7920
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.298828125,
      "learning_rate": 0.0002,
      "loss": 0.7492,
      "step": 7930
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.28515625,
      "learning_rate": 0.0002,
      "loss": 0.87,
      "step": 7940
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.326171875,
      "learning_rate": 0.0002,
      "loss": 0.8625,
      "step": 7950
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.296875,
      "learning_rate": 0.0002,
      "loss": 0.8603,
      "step": 7960
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.33984375,
      "learning_rate": 0.0002,
      "loss": 0.8038,
      "step": 7970
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.32421875,
      "learning_rate": 0.0002,
      "loss": 0.7151,
      "step": 7980
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.287109375,
      "learning_rate": 0.0002,
      "loss": 0.8531,
      "step": 7990
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.296875,
      "learning_rate": 0.0002,
      "loss": 0.8473,
      "step": 8000
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.337890625,
      "learning_rate": 0.0002,
      "loss": 0.8129,
      "step": 8010
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.337890625,
      "learning_rate": 0.0002,
      "loss": 0.7445,
      "step": 8020
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.279296875,
      "learning_rate": 0.0002,
      "loss": 0.7548,
      "step": 8030
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 0.8742,
      "step": 8040
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.27734375,
      "learning_rate": 0.0002,
      "loss": 0.9106,
      "step": 8050
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.3203125,
      "learning_rate": 0.0002,
      "loss": 0.8259,
      "step": 8060
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.37890625,
      "learning_rate": 0.0002,
      "loss": 0.7902,
      "step": 8070
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.318359375,
      "learning_rate": 0.0002,
      "loss": 0.7934,
      "step": 8080
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.283203125,
      "learning_rate": 0.0002,
      "loss": 0.8658,
      "step": 8090
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 0.862,
      "step": 8100
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.310546875,
      "learning_rate": 0.0002,
      "loss": 0.8364,
      "step": 8110
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.302734375,
      "learning_rate": 0.0002,
      "loss": 0.7066,
      "step": 8120
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.2890625,
      "learning_rate": 0.0002,
      "loss": 0.7342,
      "step": 8130
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.2890625,
      "learning_rate": 0.0002,
      "loss": 0.9144,
      "step": 8140
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.283203125,
      "learning_rate": 0.0002,
      "loss": 0.8674,
      "step": 8150
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.287109375,
      "learning_rate": 0.0002,
      "loss": 0.8151,
      "step": 8160
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.33203125,
      "learning_rate": 0.0002,
      "loss": 0.7233,
      "step": 8170
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.318359375,
      "learning_rate": 0.0002,
      "loss": 0.7736,
      "step": 8180
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.267578125,
      "learning_rate": 0.0002,
      "loss": 0.89,
      "step": 8190
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.291015625,
      "learning_rate": 0.0002,
      "loss": 0.8526,
      "step": 8200
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.302734375,
      "learning_rate": 0.0002,
      "loss": 0.83,
      "step": 8210
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.337890625,
      "learning_rate": 0.0002,
      "loss": 0.7225,
      "step": 8220
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.283203125,
      "learning_rate": 0.0002,
      "loss": 0.7638,
      "step": 8230
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.271484375,
      "learning_rate": 0.0002,
      "loss": 0.8994,
      "step": 8240
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.287109375,
      "learning_rate": 0.0002,
      "loss": 0.8524,
      "step": 8250
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 0.8493,
      "step": 8260
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 0.7478,
      "step": 8270
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.30859375,
      "learning_rate": 0.0002,
      "loss": 0.7344,
      "step": 8280
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.283203125,
      "learning_rate": 0.0002,
      "loss": 0.8714,
      "step": 8290
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.279296875,
      "learning_rate": 0.0002,
      "loss": 0.8932,
      "step": 8300
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.30859375,
      "learning_rate": 0.0002,
      "loss": 0.8411,
      "step": 8310
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.318359375,
      "learning_rate": 0.0002,
      "loss": 0.7606,
      "step": 8320
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.283203125,
      "learning_rate": 0.0002,
      "loss": 0.7383,
      "step": 8330
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.287109375,
      "learning_rate": 0.0002,
      "loss": 0.8935,
      "step": 8340
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.27734375,
      "learning_rate": 0.0002,
      "loss": 0.885,
      "step": 8350
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.322265625,
      "learning_rate": 0.0002,
      "loss": 0.8561,
      "step": 8360
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.337890625,
      "learning_rate": 0.0002,
      "loss": 0.7718,
      "step": 8370
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.349609375,
      "learning_rate": 0.0002,
      "loss": 0.7581,
      "step": 8380
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.279296875,
      "learning_rate": 0.0002,
      "loss": 0.8594,
      "step": 8390
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.298828125,
      "learning_rate": 0.0002,
      "loss": 0.88,
      "step": 8400
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.322265625,
      "learning_rate": 0.0002,
      "loss": 0.8275,
      "step": 8410
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.31640625,
      "learning_rate": 0.0002,
      "loss": 0.7273,
      "step": 8420
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.294921875,
      "learning_rate": 0.0002,
      "loss": 0.7681,
      "step": 8430
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0002,
      "loss": 0.9001,
      "step": 8440
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.306640625,
      "learning_rate": 0.0002,
      "loss": 0.8591,
      "step": 8450
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 0.8366,
      "step": 8460
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.322265625,
      "learning_rate": 0.0002,
      "loss": 0.7378,
      "step": 8470
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 0.765,
      "step": 8480
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.298828125,
      "learning_rate": 0.0002,
      "loss": 0.8969,
      "step": 8490
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.283203125,
      "learning_rate": 0.0002,
      "loss": 0.8693,
      "step": 8500
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.32421875,
      "learning_rate": 0.0002,
      "loss": 0.8613,
      "step": 8510
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.33984375,
      "learning_rate": 0.0002,
      "loss": 0.779,
      "step": 8520
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 0.7373,
      "step": 8530
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.263671875,
      "learning_rate": 0.0002,
      "loss": 0.8844,
      "step": 8540
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.30078125,
      "learning_rate": 0.0002,
      "loss": 0.851,
      "step": 8550
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.3125,
      "learning_rate": 0.0002,
      "loss": 0.8388,
      "step": 8560
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.33984375,
      "learning_rate": 0.0002,
      "loss": 0.8049,
      "step": 8570
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 0.728,
      "step": 8580
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.3046875,
      "learning_rate": 0.0002,
      "loss": 0.85,
      "step": 8590
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.298828125,
      "learning_rate": 0.0002,
      "loss": 0.8518,
      "step": 8600
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.306640625,
      "learning_rate": 0.0002,
      "loss": 0.8337,
      "step": 8610
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.349609375,
      "learning_rate": 0.0002,
      "loss": 0.7444,
      "step": 8620
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.2890625,
      "learning_rate": 0.0002,
      "loss": 0.7731,
      "step": 8630
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0002,
      "loss": 0.8802,
      "step": 8640
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.279296875,
      "learning_rate": 0.0002,
      "loss": 0.9098,
      "step": 8650
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.30859375,
      "learning_rate": 0.0002,
      "loss": 0.834,
      "step": 8660
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.32421875,
      "learning_rate": 0.0002,
      "loss": 0.7543,
      "step": 8670
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.302734375,
      "learning_rate": 0.0002,
      "loss": 0.7567,
      "step": 8680
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.28515625,
      "learning_rate": 0.0002,
      "loss": 0.8749,
      "step": 8690
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.296875,
      "learning_rate": 0.0002,
      "loss": 0.902,
      "step": 8700
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.298828125,
      "learning_rate": 0.0002,
      "loss": 0.8387,
      "step": 8710
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.33203125,
      "learning_rate": 0.0002,
      "loss": 0.7634,
      "step": 8720
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.291015625,
      "learning_rate": 0.0002,
      "loss": 0.7561,
      "step": 8730
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0002,
      "loss": 0.8951,
      "step": 8740
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 0.8968,
      "step": 8750
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.302734375,
      "learning_rate": 0.0002,
      "loss": 0.8528,
      "step": 8760
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.330078125,
      "learning_rate": 0.0002,
      "loss": 0.7364,
      "step": 8770
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.298828125,
      "learning_rate": 0.0002,
      "loss": 0.7572,
      "step": 8780
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.2734375,
      "learning_rate": 0.0002,
      "loss": 0.8193,
      "step": 8790
    },
    {
      "epoch": 5.01,
      "grad_norm": 0.296875,
      "learning_rate": 0.0002,
      "loss": 0.8147,
      "step": 8800
    },
    {
      "epoch": 5.01,
      "grad_norm": 0.298828125,
      "learning_rate": 0.0002,
      "loss": 0.814,
      "step": 8810
    },
    {
      "epoch": 5.02,
      "grad_norm": 0.3359375,
      "learning_rate": 0.0002,
      "loss": 0.7789,
      "step": 8820
    },
    {
      "epoch": 5.03,
      "grad_norm": 0.361328125,
      "learning_rate": 0.0002,
      "loss": 0.7042,
      "step": 8830
    },
    {
      "epoch": 5.03,
      "grad_norm": 0.294921875,
      "learning_rate": 0.0002,
      "loss": 0.7396,
      "step": 8840
    },
    {
      "epoch": 5.04,
      "grad_norm": 0.380859375,
      "learning_rate": 0.0002,
      "loss": 0.8916,
      "step": 8850
    },
    {
      "epoch": 5.04,
      "grad_norm": 0.30859375,
      "learning_rate": 0.0002,
      "loss": 0.8322,
      "step": 8860
    },
    {
      "epoch": 5.05,
      "grad_norm": 0.3125,
      "learning_rate": 0.0002,
      "loss": 0.7591,
      "step": 8870
    },
    {
      "epoch": 5.05,
      "grad_norm": 0.33984375,
      "learning_rate": 0.0002,
      "loss": 0.6652,
      "step": 8880
    },
    {
      "epoch": 5.06,
      "grad_norm": 0.310546875,
      "learning_rate": 0.0002,
      "loss": 0.7564,
      "step": 8890
    },
    {
      "epoch": 5.07,
      "grad_norm": 0.30859375,
      "learning_rate": 0.0002,
      "loss": 0.8046,
      "step": 8900
    },
    {
      "epoch": 5.07,
      "grad_norm": 0.333984375,
      "learning_rate": 0.0002,
      "loss": 0.842,
      "step": 8910
    },
    {
      "epoch": 5.08,
      "grad_norm": 0.330078125,
      "learning_rate": 0.0002,
      "loss": 0.7599,
      "step": 8920
    },
    {
      "epoch": 5.08,
      "grad_norm": 0.353515625,
      "learning_rate": 0.0002,
      "loss": 0.6848,
      "step": 8930
    },
    {
      "epoch": 5.09,
      "grad_norm": 0.30078125,
      "learning_rate": 0.0002,
      "loss": 0.7208,
      "step": 8940
    },
    {
      "epoch": 5.09,
      "grad_norm": 0.294921875,
      "learning_rate": 0.0002,
      "loss": 0.8479,
      "step": 8950
    },
    {
      "epoch": 5.1,
      "grad_norm": 0.3125,
      "learning_rate": 0.0002,
      "loss": 0.8372,
      "step": 8960
    },
    {
      "epoch": 5.11,
      "grad_norm": 0.333984375,
      "learning_rate": 0.0002,
      "loss": 0.7274,
      "step": 8970
    },
    {
      "epoch": 5.11,
      "grad_norm": 0.330078125,
      "learning_rate": 0.0002,
      "loss": 0.6996,
      "step": 8980
    },
    {
      "epoch": 5.12,
      "grad_norm": 0.31640625,
      "learning_rate": 0.0002,
      "loss": 0.7768,
      "step": 8990
    },
    {
      "epoch": 5.12,
      "grad_norm": 0.30078125,
      "learning_rate": 0.0002,
      "loss": 0.837,
      "step": 9000
    },
    {
      "epoch": 5.13,
      "grad_norm": 0.330078125,
      "learning_rate": 0.0002,
      "loss": 0.8047,
      "step": 9010
    },
    {
      "epoch": 5.13,
      "grad_norm": 0.328125,
      "learning_rate": 0.0002,
      "loss": 0.774,
      "step": 9020
    },
    {
      "epoch": 5.14,
      "grad_norm": 0.341796875,
      "learning_rate": 0.0002,
      "loss": 0.6529,
      "step": 9030
    },
    {
      "epoch": 5.15,
      "grad_norm": 0.298828125,
      "learning_rate": 0.0002,
      "loss": 0.7248,
      "step": 9040
    },
    {
      "epoch": 5.15,
      "grad_norm": 0.296875,
      "learning_rate": 0.0002,
      "loss": 0.8272,
      "step": 9050
    },
    {
      "epoch": 5.16,
      "grad_norm": 0.328125,
      "learning_rate": 0.0002,
      "loss": 0.8244,
      "step": 9060
    },
    {
      "epoch": 5.16,
      "grad_norm": 0.31640625,
      "learning_rate": 0.0002,
      "loss": 0.7324,
      "step": 9070
    },
    {
      "epoch": 5.17,
      "grad_norm": 0.35546875,
      "learning_rate": 0.0002,
      "loss": 0.6828,
      "step": 9080
    },
    {
      "epoch": 5.17,
      "grad_norm": 0.296875,
      "learning_rate": 0.0002,
      "loss": 0.7614,
      "step": 9090
    },
    {
      "epoch": 5.18,
      "grad_norm": 0.302734375,
      "learning_rate": 0.0002,
      "loss": 0.8315,
      "step": 9100
    },
    {
      "epoch": 5.18,
      "grad_norm": 0.32421875,
      "learning_rate": 0.0002,
      "loss": 0.8298,
      "step": 9110
    },
    {
      "epoch": 5.19,
      "grad_norm": 0.33984375,
      "learning_rate": 0.0002,
      "loss": 0.7865,
      "step": 9120
    },
    {
      "epoch": 5.2,
      "grad_norm": 0.345703125,
      "learning_rate": 0.0002,
      "loss": 0.7191,
      "step": 9130
    },
    {
      "epoch": 5.2,
      "grad_norm": 0.296875,
      "learning_rate": 0.0002,
      "loss": 0.7779,
      "step": 9140
    },
    {
      "epoch": 5.21,
      "grad_norm": 0.302734375,
      "learning_rate": 0.0002,
      "loss": 0.8367,
      "step": 9150
    },
    {
      "epoch": 5.21,
      "grad_norm": 0.318359375,
      "learning_rate": 0.0002,
      "loss": 0.8405,
      "step": 9160
    },
    {
      "epoch": 5.22,
      "grad_norm": 0.32421875,
      "learning_rate": 0.0002,
      "loss": 0.7552,
      "step": 9170
    },
    {
      "epoch": 5.22,
      "grad_norm": 0.353515625,
      "learning_rate": 0.0002,
      "loss": 0.6737,
      "step": 9180
    },
    {
      "epoch": 5.23,
      "grad_norm": 0.306640625,
      "learning_rate": 0.0002,
      "loss": 0.7761,
      "step": 9190
    },
    {
      "epoch": 5.24,
      "grad_norm": 0.302734375,
      "learning_rate": 0.0002,
      "loss": 0.8427,
      "step": 9200
    },
    {
      "epoch": 5.24,
      "grad_norm": 0.31640625,
      "learning_rate": 0.0002,
      "loss": 0.8125,
      "step": 9210
    },
    {
      "epoch": 5.25,
      "grad_norm": 0.337890625,
      "learning_rate": 0.0002,
      "loss": 0.8025,
      "step": 9220
    },
    {
      "epoch": 5.25,
      "grad_norm": 0.3515625,
      "learning_rate": 0.0002,
      "loss": 0.7217,
      "step": 9230
    },
    {
      "epoch": 5.26,
      "grad_norm": 0.310546875,
      "learning_rate": 0.0002,
      "loss": 0.7732,
      "step": 9240
    },
    {
      "epoch": 5.26,
      "grad_norm": 0.3125,
      "learning_rate": 0.0002,
      "loss": 0.8663,
      "step": 9250
    },
    {
      "epoch": 5.27,
      "grad_norm": 0.3046875,
      "learning_rate": 0.0002,
      "loss": 0.8386,
      "step": 9260
    },
    {
      "epoch": 5.28,
      "grad_norm": 0.341796875,
      "learning_rate": 0.0002,
      "loss": 0.7779,
      "step": 9270
    },
    {
      "epoch": 5.28,
      "grad_norm": 0.357421875,
      "learning_rate": 0.0002,
      "loss": 0.689,
      "step": 9280
    },
    {
      "epoch": 5.29,
      "grad_norm": 0.310546875,
      "learning_rate": 0.0002,
      "loss": 0.7515,
      "step": 9290
    },
    {
      "epoch": 5.29,
      "grad_norm": 0.31640625,
      "learning_rate": 0.0002,
      "loss": 0.9001,
      "step": 9300
    },
    {
      "epoch": 5.3,
      "grad_norm": 0.31640625,
      "learning_rate": 0.0002,
      "loss": 0.7936,
      "step": 9310
    },
    {
      "epoch": 5.3,
      "grad_norm": 0.3359375,
      "learning_rate": 0.0002,
      "loss": 0.7395,
      "step": 9320
    },
    {
      "epoch": 5.31,
      "grad_norm": 0.3515625,
      "learning_rate": 0.0002,
      "loss": 0.6828,
      "step": 9330
    },
    {
      "epoch": 5.32,
      "grad_norm": 0.3125,
      "learning_rate": 0.0002,
      "loss": 0.766,
      "step": 9340
    },
    {
      "epoch": 5.32,
      "grad_norm": 0.3046875,
      "learning_rate": 0.0002,
      "loss": 0.8524,
      "step": 9350
    },
    {
      "epoch": 5.33,
      "grad_norm": 0.318359375,
      "learning_rate": 0.0002,
      "loss": 0.8122,
      "step": 9360
    },
    {
      "epoch": 5.33,
      "grad_norm": 0.34375,
      "learning_rate": 0.0002,
      "loss": 0.7972,
      "step": 9370
    },
    {
      "epoch": 5.34,
      "grad_norm": 0.3671875,
      "learning_rate": 0.0002,
      "loss": 0.7058,
      "step": 9380
    },
    {
      "epoch": 5.34,
      "grad_norm": 0.318359375,
      "learning_rate": 0.0002,
      "loss": 0.7869,
      "step": 9390
    },
    {
      "epoch": 5.35,
      "grad_norm": 0.29296875,
      "learning_rate": 0.0002,
      "loss": 0.8735,
      "step": 9400
    },
    {
      "epoch": 5.36,
      "grad_norm": 0.30859375,
      "learning_rate": 0.0002,
      "loss": 0.8089,
      "step": 9410
    },
    {
      "epoch": 5.36,
      "grad_norm": 0.33984375,
      "learning_rate": 0.0002,
      "loss": 0.7725,
      "step": 9420
    },
    {
      "epoch": 5.37,
      "grad_norm": 0.38671875,
      "learning_rate": 0.0002,
      "loss": 0.6743,
      "step": 9430
    },
    {
      "epoch": 5.37,
      "grad_norm": 0.310546875,
      "learning_rate": 0.0002,
      "loss": 0.7509,
      "step": 9440
    },
    {
      "epoch": 5.38,
      "grad_norm": 0.369140625,
      "learning_rate": 0.0002,
      "loss": 0.857,
      "step": 9450
    },
    {
      "epoch": 5.38,
      "grad_norm": 0.31640625,
      "learning_rate": 0.0002,
      "loss": 0.8284,
      "step": 9460
    },
    {
      "epoch": 5.39,
      "grad_norm": 0.345703125,
      "learning_rate": 0.0002,
      "loss": 0.7789,
      "step": 9470
    },
    {
      "epoch": 5.4,
      "grad_norm": 0.35546875,
      "learning_rate": 0.0002,
      "loss": 0.6707,
      "step": 9480
    },
    {
      "epoch": 5.4,
      "grad_norm": 0.34375,
      "learning_rate": 0.0002,
      "loss": 0.7618,
      "step": 9490
    },
    {
      "epoch": 5.41,
      "grad_norm": 0.330078125,
      "learning_rate": 0.0002,
      "loss": 0.8578,
      "step": 9500
    },
    {
      "epoch": 5.41,
      "grad_norm": 0.318359375,
      "learning_rate": 0.0002,
      "loss": 0.8171,
      "step": 9510
    },
    {
      "epoch": 5.42,
      "grad_norm": 0.361328125,
      "learning_rate": 0.0002,
      "loss": 0.7501,
      "step": 9520
    },
    {
      "epoch": 5.42,
      "grad_norm": 0.35546875,
      "learning_rate": 0.0002,
      "loss": 0.72,
      "step": 9530
    },
    {
      "epoch": 5.43,
      "grad_norm": 0.30078125,
      "learning_rate": 0.0002,
      "loss": 0.732,
      "step": 9540
    },
    {
      "epoch": 5.44,
      "grad_norm": 0.306640625,
      "learning_rate": 0.0002,
      "loss": 0.8673,
      "step": 9550
    },
    {
      "epoch": 5.44,
      "grad_norm": 0.31640625,
      "learning_rate": 0.0002,
      "loss": 0.8511,
      "step": 9560
    },
    {
      "epoch": 5.45,
      "grad_norm": 0.341796875,
      "learning_rate": 0.0002,
      "loss": 0.7353,
      "step": 9570
    },
    {
      "epoch": 5.45,
      "grad_norm": 0.36328125,
      "learning_rate": 0.0002,
      "loss": 0.7003,
      "step": 9580
    },
    {
      "epoch": 5.46,
      "grad_norm": 0.3203125,
      "learning_rate": 0.0002,
      "loss": 0.804,
      "step": 9590
    },
    {
      "epoch": 5.46,
      "grad_norm": 0.3125,
      "learning_rate": 0.0002,
      "loss": 0.8475,
      "step": 9600
    },
    {
      "epoch": 5.47,
      "grad_norm": 0.337890625,
      "learning_rate": 0.0002,
      "loss": 0.8454,
      "step": 9610
    },
    {
      "epoch": 5.48,
      "grad_norm": 0.3515625,
      "learning_rate": 0.0002,
      "loss": 0.7883,
      "step": 9620
    },
    {
      "epoch": 5.48,
      "grad_norm": 0.357421875,
      "learning_rate": 0.0002,
      "loss": 0.6765,
      "step": 9630
    },
    {
      "epoch": 5.49,
      "grad_norm": 0.310546875,
      "learning_rate": 0.0002,
      "loss": 0.7595,
      "step": 9640
    },
    {
      "epoch": 5.49,
      "grad_norm": 0.30078125,
      "learning_rate": 0.0002,
      "loss": 0.849,
      "step": 9650
    },
    {
      "epoch": 5.5,
      "grad_norm": 0.32421875,
      "learning_rate": 0.0002,
      "loss": 0.841,
      "step": 9660
    },
    {
      "epoch": 5.5,
      "grad_norm": 0.33984375,
      "learning_rate": 0.0002,
      "loss": 0.7854,
      "step": 9670
    },
    {
      "epoch": 5.51,
      "grad_norm": 0.337890625,
      "learning_rate": 0.0002,
      "loss": 0.7011,
      "step": 9680
    },
    {
      "epoch": 5.52,
      "grad_norm": 0.310546875,
      "learning_rate": 0.0002,
      "loss": 0.7428,
      "step": 9690
    },
    {
      "epoch": 5.52,
      "grad_norm": 0.287109375,
      "learning_rate": 0.0002,
      "loss": 0.8701,
      "step": 9700
    },
    {
      "epoch": 5.53,
      "grad_norm": 0.3125,
      "learning_rate": 0.0002,
      "loss": 0.7852,
      "step": 9710
    },
    {
      "epoch": 5.53,
      "grad_norm": 0.32421875,
      "learning_rate": 0.0002,
      "loss": 0.7718,
      "step": 9720
    },
    {
      "epoch": 5.54,
      "grad_norm": 0.3828125,
      "learning_rate": 0.0002,
      "loss": 0.7085,
      "step": 9730
    },
    {
      "epoch": 5.54,
      "grad_norm": 0.322265625,
      "learning_rate": 0.0002,
      "loss": 0.7796,
      "step": 9740
    },
    {
      "epoch": 5.55,
      "grad_norm": 0.296875,
      "learning_rate": 0.0002,
      "loss": 0.8341,
      "step": 9750
    },
    {
      "epoch": 5.55,
      "grad_norm": 0.330078125,
      "learning_rate": 0.0002,
      "loss": 0.8505,
      "step": 9760
    },
    {
      "epoch": 5.56,
      "grad_norm": 0.33984375,
      "learning_rate": 0.0002,
      "loss": 0.81,
      "step": 9770
    },
    {
      "epoch": 5.57,
      "grad_norm": 0.359375,
      "learning_rate": 0.0002,
      "loss": 0.6939,
      "step": 9780
    },
    {
      "epoch": 5.57,
      "grad_norm": 0.302734375,
      "learning_rate": 0.0002,
      "loss": 0.7882,
      "step": 9790
    },
    {
      "epoch": 5.58,
      "grad_norm": 0.330078125,
      "learning_rate": 0.0002,
      "loss": 0.8979,
      "step": 9800
    },
    {
      "epoch": 5.58,
      "grad_norm": 0.310546875,
      "learning_rate": 0.0002,
      "loss": 0.8045,
      "step": 9810
    },
    {
      "epoch": 5.59,
      "grad_norm": 0.32421875,
      "learning_rate": 0.0002,
      "loss": 0.7508,
      "step": 9820
    },
    {
      "epoch": 5.59,
      "grad_norm": 0.345703125,
      "learning_rate": 0.0002,
      "loss": 0.7107,
      "step": 9830
    },
    {
      "epoch": 5.6,
      "grad_norm": 0.333984375,
      "learning_rate": 0.0002,
      "loss": 0.7851,
      "step": 9840
    },
    {
      "epoch": 5.61,
      "grad_norm": 0.3203125,
      "learning_rate": 0.0002,
      "loss": 0.8831,
      "step": 9850
    },
    {
      "epoch": 5.61,
      "grad_norm": 0.341796875,
      "learning_rate": 0.0002,
      "loss": 0.7831,
      "step": 9860
    },
    {
      "epoch": 5.62,
      "grad_norm": 0.3359375,
      "learning_rate": 0.0002,
      "loss": 0.7963,
      "step": 9870
    },
    {
      "epoch": 5.62,
      "grad_norm": 0.357421875,
      "learning_rate": 0.0002,
      "loss": 0.6816,
      "step": 9880
    },
    {
      "epoch": 5.63,
      "grad_norm": 0.318359375,
      "learning_rate": 0.0002,
      "loss": 0.7638,
      "step": 9890
    },
    {
      "epoch": 5.63,
      "grad_norm": 0.294921875,
      "learning_rate": 0.0002,
      "loss": 0.8935,
      "step": 9900
    },
    {
      "epoch": 5.64,
      "grad_norm": 0.3203125,
      "learning_rate": 0.0002,
      "loss": 0.7828,
      "step": 9910
    },
    {
      "epoch": 5.65,
      "grad_norm": 0.353515625,
      "learning_rate": 0.0002,
      "loss": 0.7594,
      "step": 9920
    },
    {
      "epoch": 5.65,
      "grad_norm": 0.390625,
      "learning_rate": 0.0002,
      "loss": 0.7165,
      "step": 9930
    },
    {
      "epoch": 5.66,
      "grad_norm": 0.31640625,
      "learning_rate": 0.0002,
      "loss": 0.7579,
      "step": 9940
    },
    {
      "epoch": 5.66,
      "grad_norm": 0.3125,
      "learning_rate": 0.0002,
      "loss": 0.8643,
      "step": 9950
    },
    {
      "epoch": 5.67,
      "grad_norm": 0.33203125,
      "learning_rate": 0.0002,
      "loss": 0.8233,
      "step": 9960
    },
    {
      "epoch": 5.67,
      "grad_norm": 0.341796875,
      "learning_rate": 0.0002,
      "loss": 0.8198,
      "step": 9970
    },
    {
      "epoch": 5.68,
      "grad_norm": 0.34375,
      "learning_rate": 0.0002,
      "loss": 0.6659,
      "step": 9980
    },
    {
      "epoch": 5.69,
      "grad_norm": 0.302734375,
      "learning_rate": 0.0002,
      "loss": 0.7583,
      "step": 9990
    },
    {
      "epoch": 5.69,
      "grad_norm": 0.30859375,
      "learning_rate": 0.0002,
      "loss": 0.8282,
      "step": 10000
    }
  ],
  "logging_steps": 10,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 500,
  "total_flos": 1.3772789232384e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}